International Journal of Pharmaceutics 388 (2010) 231–241

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsevier.com/locate/ijpharm

Pharmaceutical Nanotechnology

Enhanced bioavailability of lacidipine via microemulsion based transdermal gels:

Formulation optimization, ex vivo and in vivo characterization
Ramesh Gannu, Chinna Reddy Palem, Vamshi Vishnu Yamsani, Shravan Kumar Yamsani,
Madhusudan Rao Yamsani ∗
National Facilities in Engineering and Technology with Industrial Collaboration (NAFETIC) Centre, University College of Pharmaceutical Sciences,
Kakatiya University, Warangal 506 009, Andhra Pradesh, India

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of the present study was to develop and optimize the microemulsion based transdermal
Received 4 August 2009 therapeutic system for lacidipine (LCDP), a poorly water soluble and low bioavailable drug. The pseudo-
Received in revised form ternary phase diagrams were developed for various microemulsion formulations composed of isopropyl
20 December 2009
myristate, Tween 80 and Labrasol. The microemulsion was optimized using a three-factor, three-level
Accepted 24 December 2009
Box–Behnken design, the independent variables selected were isopropyl myristate, surfactant mixture
Available online 7 January 2010
(Tween 80 and Labrasol) and water; dependent variables (responses) were cumulative amount perme-
ated across rat abdominal skin in 24 h (Q24 ; Y1 ), flux (Y2 ), and lag time (Y3 ). Mathematical equations
Keywords:
Microemulsion
and response surface plots were used to relate the dependent and independent variables. The regression
Lacidipine equations were generated for responses Y1 , Y2 and Y3 . The statistical validity of the polynomials was estab-
Microemulgel lished, and optimized formulation factors were selected by feasibility and grid search. Validation of the
Box–Behnken optimization study with 10 confirmatory runs indicated high degree of prognostic ability of response sur-
Optimization face methodology. The gel of optimized formulation (ME-OPT) showed a flux of 43.7 ␮g cm−2 h−1 , which
Pharmacokinetics could meet the target flux (12.16 ␮g cm−2 h−1 ). The bioavailability studies in rabbits showed that about
Bioavailability 3.5 times statistically significant (p < 0.05) improvement in bioavailability, after transdermal adminis-
EVIV correlation
tration of microemulsion gel compared to oral suspension. The ex vivo–in vivo correlation was found to
have biphasic pattern and followed type A correlation. Microemulsion based transdermal therapeutic
system of LCDP was developed and optimized using Box–Behnken statistical design and could provide
an effective treatment in the management of hypertension.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction has been made to deliver LCDP transdermally across skin using
microemulsions.
Lacidipine (LCDP) is a calcium channel blocker used in the Microemulsion (ME) is defined as an O/W or W/O emulsion pro-
treatment of hypertension and atherosclerosis. It also possesses ducing a transparent product that has a droplet size from 10 to
antioxidant effect and is one of the most vascular selective of 100 nm and does not have the tendency to coalesce (Kreilgaard,
the dihydropyridines (Lee and Bryson, 1994; Mc Cormack and 2002; Lawrence and Rees, 2000). MEs are composed of oil phase,
Wagstaff, 2003). LCDP undergoes extensive first-pass hepatic surfactant, cosurfactant and aqueous phase at appropriate ratios
metabolism and has a mean absolute bioavailability of about 10% (Mohammed and Manoj, 2000). MEs have several specific physic-
(range 4–52%). LCDP is completely metabolized in the liver by ochemical properties such as transparency, optical isotropy, low
cytochrome P450 3A4 to pharmacologically inactive metabolites viscosity and thermodynamic stability (Lawrence and Rees, 2000;
(Mc Cormack and Wagstaff, 2003). In addition, its limited aque- Mohammed and Manoj, 2000; Baroli et al., 2000). It is promising
ous solubility contributes to its limited bioavailability. However for both transdermal and dermal delivery of drugs as an efficient
the low oral bioavailability restricts its use, therefore alternative route of drug administration (Kreilgaard, 2002; Rhee et al., 2001;
mode of delivery system is desirable, to deliver the drug at effec- Kreilgaard et al., 2000; Baboota et al., 2007; Kamal et al., 2007; Chen
tive concentrations to treat hypertension. There are no reports et al., 2007). Several mechanisms have been proposed to explain
on transdermal delivery of LCDP. In this investigation, an attempt the advantages of microemulsions for the transdermal and dermal
delivery of drugs (Chen et al., 2006; Zhao et al., 2006). First, the
thermodynamics towards the skin is increased due to large amount
∗ Corresponding author. Tel.: +91 870 2438844; fax: +91 870 2453508. of a drug incorporated in the formulation. Second, the increased
E-mail address: ymrao123@yahoo.com (M.R. Yamsani). thermodynamic activity of the drug may favor its partitioning into

0378-5173/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpharm.2009.12.050
232 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241

the skin. Third, the ingredients of microemulsion may reduce the and stirring to form a clear and transparent liquid. The resulting
diffusional barrier of the stratum corneum and increase the perme- microemulsions were tightly sealed and stored at ambient tem-
ation rate of drug via skin by acting as permeation enhancers. Also, perature, and their physical stability was measured by observing
the hydration effect of microemulsion on the stratum corneum periodically the occurrence of phase separation.
may influence the permeation ability of formulations (Mohammed
and Manoj, 2000). Although many drugs have been incorporated 2.4. Rheological measurements
in microemulsion for transdermal and dermal delivery, LCDP has
not been evaluated. In this study, O/W microemulsions contain- The rheological properties of MEs were measured using
ing LCDP have been developed after screening oils and surfactants. Brookfield Programmable DVIII+ Digital Rheometer (Brookfield
Pseudo-ternary phase diagrams were constructed to obtain the Engineering Laboratories Inc., MA, USA). The rheological measure-
components and their concentration ranges, and the microemul- ments were performed using a controlled stress rheometer with
sion formulations varied according to a Box–Behnken statistical the cone (24 mm) and plate geometry. The viscosity was deter-
design to find out the most suitable components ratio for the opti- mined by torque sweep from 10 to 110%. All the measurements
mized formulation. The formulation was also evaluated for in vivo were performed in triplicate at 25 ◦ C. The equilibrium time before
performance in rabbits. every measurement was 5 min and the sample volume used was
approximately 0.5 mL. Calculation of rheological properties were
2. Materials and methods performed using Rheocalc 32 software (Brookfield Engineering Lab-
oratories Inc., USA). The data was analyzed using “Power Law”
2.1. Materials (Bonacucina et al., 2004) as expressed by Eq. (1).

Lacidipine and Labrasol (Gatteffose, France) were gift samples  = KDn (1)
from Dr. Reddys Laboratories, Hyderabad, India. Liquid paraffin,
Tween 80, isopropyl myristate, polyethylene glycol 400 were pur- where  is shear stress; K is gel index (GI) or consistency index; D
chased from Merck, Mumbai, India. Oleic acid was purchased from is shear rate; and n is flow index. ‘Rheocalc 32’ software was used
SD Fine Chemicals, Mumbai, India. Double distilled water was used to automatically apply the model to generated data, and the value
for the preparation of microemulsions. All other chemicals and sol- of GI was recorded.
vents were of analytical reagent grade.
2.5. Preparation of rat abdominal skin
2.2. Solubility studies
The animal study was conducted in accordance with the
2.2.1. Screening of oils and surfactants for microemulsions approval of the Animal Ethical Committee, Kakatiya University,
To find out appropriate oil phase in microemulsions, the solu- India. Wistar rats weighing 150–200 g were sacrificed using anaes-
bility of LCDP in various oils such as oleic acid, isopropyl myristate thetic ether. The hair of test animals was carefully trimmed with
(IPM), liquid paraffin, eucalyptus oil and sesame oil and each oil electrical clippers and the full thickness skin was removed from the
containing 10% (v/v) of dimethyl formamide were measured. An abdominal region. The epidermis was prepared surgically by heat
excess amount of LCDP was added to 10 mL of oil and the mix- separation technique (Levang et al., 1999), which involved soaking
ture was shaken at 25 ◦ C for 72 h. The resulting suspension was the entire abdominal skin in water at 60 ◦ C for 45 s, followed by
then centrifuged (Mikro 220R, Tuttlingen, Germany) for 10 min at careful removal of the epidermis. The epidermis was washed with
10,000 rpm. The supernatant was filtered through a membrane fil- water and used for ex vivo permeability studies.
ter (0.45 ␮m) and the drug concentration in filtrate was determined
by high performance liquid chromatography (HPLC) analysis after 2.6. Ex vivo permeation studies
appropriate dilution. The oil phase that showed high solubility of
LCDP was used in the preparation of microemulsions containing Franz diffusion cell with a surface area of 3.56 cm2 was used
0.2% LCDP. for ex vivo permeation studies. The rat skin was mounted between
donor and receptor compartments of the diffusion cell with stratum
2.3. Construction of phase diagrams and formulation of LCDP corneum facing the donor compartment. Microemulsion (4 mL)
loaded microemulsions was placed in the donor compartment and 13 mL of phosphate
buffer saline (PBS, pH 7.4) containing 40% (v/v) of polyethylene gly-
To find out appropriate components in the formation of col (PEG) 400 was placed in receptor compartment. PEG 400 was
microemulsions, two safe and compatible nonionic surfactants, incorporated to maintain sink conditions and the contents of recep-
polysorbate 80 and a cosurfactant, Labrasol, were chosen. Pseudo- tor compartment were agitated at 400 rpm and was placed over
ternary phase diagrams were constructed using water titration a multi-magnetic stirrer (Cintex, Mumbai, India). The study was
method in order to obtain the concentration range of compo- conducted at 37 ◦ C and samples of 1 mL were collected at predeter-
nents for the existing range of microemulsions. The weight ratio mined time points and replenished with PBS (pH 7.4) containing
of surfactant to cosurfactant (SMix ) varied as 1:1, 1:2, 1:3, 2:1 and 40% (v/v) PEG 400. The cumulative amount of LCDP permeated
3:1. For each pseudo-ternary phase diagram at a specific surfac- was determined using HPLC (Gannu et al., 2009) and concentration
tant/cosurfactant weight ratio was mixed with oil at a ratio of 9:1, was corrected for sampling effects according to Eq. (2) (Hayton and
8:2, 7:3, 6:4, 5:5, 4:6, 3:7, 2:8 and 1:9 (w/w). Water was added Chen, 1982):
  C1 
drop by drop to each oily-surfactant mixture under magnetic stir-
ring at room temperature, until the mixture became clear at a
 VT n−1
Cn1 = Cn (2)
certain point. The concentrations of components were recorded in VT − VS Cn−1
order to complete the pseudo-ternary phase diagrams, and then
the contents of oil, surfactant, cosurfactant and water at appropri- where Cn1 is the corrected concentration of the nth sample, Cn is
ate weight ratios were selected based on these results. LCDP loaded 1
the measured concentration of LCDP in the nth sample, Cn−1 is the
microemulsions were prepared by dissolving LCDP in DMF:IPM corrected concentration in the (n − 1)th sample. Cn−1 is the mea-
(1:9) and surfactant mixture, adding the required quantity of water, sured concentration of the LCDP in the (n − 1)th sample, VT is the
 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241 233

total volume of the receiver fluid and VS is the volume of the sample 2.9. Droplet size determination
drawn.
The steady state flux (Jss ) was calculated from the slope of the The droplet size analysis of optimized microemulsion was deter-
steady state portion of the line in the plot of drug amount perme- mined using dynamic light scattering method employing a Zeta
ated Vs time (h). Permeability coefficient (Kp ) was calculated by Sizer (Malvern Instruments Ltd., Worcestershire WR 14 1XZ, UK).
dividing the flux with concentration of the drug in microemulsion.
The lag time was calculated from the intercept on the time axis in 2.10. Stability studies
the plot of cumulative amount permeated Vs time. The target flux
was calculated using the following Eq. (3). The enhancement ratio The stability study was conducted for formulation ME4 and ME
was calculated as the ratio of flux obtained with microemulsion and gel. Sufficient samples were placed in amber colour containers and
control (solution). were further placed at 40 ± 2 ◦ C/75 ± 5% R.H. (Skylab Instruments
& Engineering Pvt Ltd., Thane, India) for 6 months. Samples were
Css × CLt × BW
Target flux = (3) withdrawn at time intervals of 1, 2, 3 and 6 months. The ex vivo per-
A
meation study was conducted according to the procedure described
Css, the LCDP concentration at therapeutic level (8.6 ␮g L−1 ) and in Section 2.6 and drug content in the formulations was estimated
CLt the total body clearance, 83.9 mL h−1 (calculated from volume using HPLC (Gannu et al., 2009).
of distribustion, 2300 mL kg−1 and half life 19 h) (Clarke’s, 2007), BW
the standard human body weight of 60 kg, A represents the surface 2.11. Skin irritation studies
area of the diffusion cell (i.e. 3.56 cm2 ). The calculated target flux
value for LCDP was 12.16 ␮g cm−2 h−1 . The skin irritation study was performed using six rabbits. The
hair of rabbits on dorsal side was shaved with electrical shaver and
2.7. Experimental design ME gel (about 4 g) formulation was applied. The development of
erythema was monitored for 7 days.
Box–Behnken statistical design was used to statistically opti-
mize the formulation factors and evaluate main effects, interaction 2.12. In vivo bioavailability study in rabbits
effects and quadratic effects on the amount of LCDP permeated
in 24 h (Q24 ), flux and lag time. A 3-factor, 3-level Box–Behnken The animal study protocol was reviewed and approved by the
design was used to explore quadratic response surfaces and Institutional Animal Ethical Committee, University College of Phar-
constructing second-order polynomial models with Design Expert maceutical Sciences, Kakatiya University, India. White New Zealand
(Version 7.1, Stat-Ease Inc., Minneapolis, MN). The Box–Behnken rabbits weighing 3.30 ± 0.35 kg were selected for the study. The
design was specifically selected since it requires fewer runs than a bioavailability of LCDP (8 mg in 4 g of ME gel) from microemulsion
central composite design (CCD) in cases of three or four variables. based gel was compared with an oral suspension. The suspension
This cubic design is characterized by set of points lying at the was prepared by suspending 8 mg of LCDP in 5 mL of water con-
midpoint of each edge and center point of the multidimensional taining 0.5% (w/v) of sodium carboxy methyl cellulose. They were
cube (Box and Behnken, 1960). A design matrix comprising of allowed free access to food and water, until the night prior to dosing
17 experimental runs was constructed. The non-linear computer and were fasted for 10 h. Latin square cross over design was fol-
generated quadratic model is given as Y = b0 + b1 X1 + b2 X2 + lowed; the animals were divided into two groups each consisting
b3 X3 + b12 X1 X2 + b13 X1 X3 + b23 X2 X3 + b11 X12 + b22 X22 + b33 X32 of three rabbits. The rabbits to be used for application of ME gel were
where Y is the measured response associated with each factor shaved carefully with the help of electrical shaver before applica-
level combination; b0 is an intercept; b1 to b33 are regression tion of ME gel followed by cleaning with water. To one group, oral
coefficients computed from the observed experimental values of suspension (8 mg/5 mL) was administered through feeding tube
Y; and X1 , X2 and X3 are the coded levels of independent variables. followed by rinsing with 10 mL of water and ME gel to another
The terms X1 X2 and X 2 (i = 1, 2 or 3) represent the interaction and group in first phase. In second phase vice versa was followed and
quadratic terms, respectively (Govender et al., 2005; Chopra et al., was conducted after 15 days of wash out period. The ME gel was
2007). The dependent and independent variables selected were applied over a surface area of 4 cm2 and was covered with a water
shown in Table 2 along with their low, medium and high levels, impermeable back up membrane and was further fixed with the
which were selected based on the results from psuodo-ternary help of adhesive membrane. Blood samples (2.5 mL) from marginal
phase diagrams. The proportion of oil (X1 ), SMix (X2 ) and water ear vein were collected at preset intervals of 0.0, 0.5, 1, 2, 4, 8, 12,
(X3 ) used to prepare the 17 experimental trials and the respective 24, 36, 48 and 72 h; 0.0, 1.0, 2.0, 4.0, 6.0, 8.0, 12.0, 18.0, 24.0, 36.0,
observed responses are given in Table 2. 48.0 and 72.0 h respectively, after administration of oral suspen-
sion and application of ME gel. All blood samples were allowed to
2.8. Check point analysis and optimization model validation clot and centrifuged for 10 min at 4000 rpm. The serum was sepa-
rated and transferred into clean micro-centrifuge tubes and stored
Statistical validation of the polynomial equations generated by at −20 ◦ C until HPLC analysis. The amount of LCDP in the samples
Design Expert was established on the basis of ANOVA provision in was estimated using HPLC (Gannu et al., 2009).
the software. The models were evaluated in terms of statistically
significant coefficients and R2 values. Various feasibility and grid 2.13. Pharmacokinetic analysis
searches were conducted to find the optimum parameters. Various
3D response surface graphs were provided by the Design Expert Pharmacokinetic parameters of LCDP after administration of
software. Ten optimum checkpoint formulations were selected LCDP ME gel and oral suspension were estimated for each rab-
by intensive grid search over the whole experiment region to bit by using a computer program, KINETICA 2000 (Version 3.0,
validate the experimental model and polynomial equations. The Innaphase Corporation, Philadelphia, USA). Noncompartmental
optimized checkpoint formulation factors were evaluated for vari- analysis was used to calculate the pharmacokinetic parame-
ous response properties. The resultant experimental values of the ters, CMax , TMax and area under the curve (AUC). CMax (ng mL−1 )
responses were quantitatively compared with the predicted values and TMax (h) were the observed maximal drug concentra-
to calculate the percentage prediction error. tion and its time, respectively. The relative bioavailability F
234 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241

for transdermal drug delivery system was calculated using Eq. 2.15. Ex vivo–in vivo correlation
(4):
The cumulative amount of LCDP permeated across rat abdom-
[AUC]ME gel inal skin ex vivo from ME gel was compared against the extent of
Relative bioavailability = (4)
[AUC]Oral suspension absorption i.e., cumulative AUC values for a possible ex vivo–in vivo
correlation.

2.14. Statistical analysis 3. Results and discussion

Statistical comparisons were made using Student’s t-test using 3.1. Selection of components for microemulsions
Sigmastat software package (Jandel Corp., CA, USA). Results were
considered significant at 95% confidence interval (p < 0.05) and The solubility of LCDP in various oils was analyzed in order to
results were expressed as mean ± SD. select components for microemulsions. The solubility of LCDP was

Fig. 1. Pseudo-ternary phase diagrams of microemulsions composed of oil phase (IPM:DMF, 9:1), surfactant (Tween 80), cosurfactant (Labrasol) and water.
 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241 235

Table 1
Solubility of LCDP in oils.

Oil phase Solubility (mg)

Oleic acid 2.16 ± 0.115

Liquid paraffin 0.02 ± 0.001
Sesame oil 2.46 ± 0.130
IPM 3.81 ± 0.202
Eucalyptus oil 3.17 ± 0.168
Oleic acid + DMF 18.82 ± 0.999
Liquid paraffin + DMF 0.11 ± 0.005
Sesam oil + DMF 22.79 ± 1.209
IPM + DMF 42.83 ± 2.215
Eucalyptus oil + DMF 33.81 ± 1.794

found to be 42.83, 33.81, 22.79, 18.82 and 0.11 mg mL−1 respec-

tively in isopropyl myristate (IPM), eucalyptus oil, sesame oil, oleic
acid and liquid paraffin each containing 10% (v/v) of dimethyl
formamide (DMF). Plain oils showed low solubility (Table 1), to
solubilize LCDP large amounts of oil phase is needed, and there-
fore a solubilizer DMF was incorporated. IPM containing 10% (v/v)
DMF was selected as oil phase to solubilize LCDP that could fur-
ther maintain large concentration gradient towards skin. Previous
reports indicated that the superior dermal flux appeared mainly
due to the large solubilizing capacity of the microemulsions, which
led to larger concentration gradient towards the skin (Kreilgaard et
al., 2000; Sintov and Shapiro, 2004).

3.2. Construction of pseudo-ternary diagrams

Microemulsions were made up of IPM containing 10% (v/v)

of DMF as oil phase, a surfactant of Tween 80 and an aqueous
phase in the presence of cosurfactant (Labrasol). Fig. 1 presents
the pseudo-ternary phase diagrams with various weight ratios of
Tween 80/Labrasol. The transparent microemulsion region is pre-
sented in phase diagrams. The rest of the region on the phase
diagram represents the turbid and conventional emulsions based
on visual observation. From Fig. 1, it was found that the area region
of microemulsions isotropic region became enlarged as the SMix
decreased, reaching the maximum point at SMix of 1:3.

3.3. Rheological measurements Fig. 2. Ex vivo permeation profiles of LCDP from microemulsions, ME gel and control,
values represented are mean ± SD (n = 3).
CP-40 spindle was used for the viscometric characterization of
MEs. The decrease in viscosity of the MEs observed with an increas- the higher amount of LCDP permeated (7704.5 ␮g) with a flux of
ing shear rates, can be described well by an exponential function 64.7 ␮g cm−2 h−1 and a lag time (2.17 h). LCDP solution (control)
and hence the obtained data was analyzed using “Power Law” showed a cumulative amount of 910.3 ␮g permeated with a flux of
(Bonacucina et al., 2004) as expressed by Eq. (1). The GI value for 8.0 ␮g cm−2 h−1 and a lag time of 2.83 h. The enhancement ratios
different formulations is presented in Table 2. The gel index was of microemulsions were 3.7–8.1-fold higher than the control (LCDP
found to be ranging from 2.0 to 163.2. was dissolved in water containing 5% (v/v) of DMF and 40% (v/v) of
PEG 400). These results demonstrated that microemulsion had a
3.4. Ex vivo skin permeation experiments potent enhancement effect for transdermal delivery.
It was observed that the amount of LCDP permeated and flux was
The permeation profiles of LCDP microemulsions through rat decreased and prolonging lag time with increase in the amount of
skin are shown in Fig. 2. The permeation parameters including surfactant. The result might be due to a decreased thermodynamic
the cumulative amount at 24 h, flux and lag time (Table 2) for activity of drug in microemulsion at higher concentrations of sur-
all experimental formulations were calculated and the permeation factant. The drug permeation was found to decrease with increase
was found to follow zero-order model (R2 > 0.974). The cumulative in the G.I. At higher levels of oil the G.I. was found to be more and the
amount ranged from 3640.5 to 7704.5 ␮g, flux ranged from 29.7 to result is in accordance with earlier studies (Gallarate et al., 1990;
64.7 ␮g cm−2 h−1 and lag time ranged from 2.16 to 2.48 h indicat- Gasco et al., 1991) and might be attributed to the gel formation
ing that the permeation parameters of LCDP from microemulsions in the microemulsion that will increase its viscosity and further
were markedly influenced by the composition of microemulsions. decrease the permeation in the skin. The diffusion through the dou-
The oil phase, IPM selected in the present work also act as an ble layer microemulsion might be a rate-determining step, as the
effective penetration enhancer (Goldberg-Cettina et al., 1995). The viscosity plays an important role in controlling the release of the
surfactant and cosurfactant in the microemulsions may reduce the drug into the receptor compartment (Ho et al., 1996).
diffusional barrier of the stratum corneum by acting as perme- The high proportion of cosurfactant in SMix could lower the
ation enhancers (Peltola et al., 2003). Formulation ME4 showed interfacial tension of the surfactant in microemulsions, resulting
236 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241

Table 2
Variables and observed responses in Box–Behnken design for microemulsions.

Formulation Independent variables Dependent variables GI Kp × 10−3 (/cm)

2
X1 (g) X2 (g) X3 (mL) Y1 (␮g) Y2 (␮g/h/cm ) Y3 (h)

ME1 0 −1 1 6398.3 55.8 2.24 11.6 7.0

ME 2 0 0 0 6170.9 52.6 2.26 20.8 6.6
ME 3 1 −1 0 4561.1 38.6 2.16 80.4 4.8
ME 4 −1 0 1 7704.5 64.7 2.17 2.0 8.1
ME 5 0 −1 −1 5931.4 48.7 2.34 23.5 6.1
ME 6 0 0 0 6183.6 53.3 2.16 18.1 6.7
ME 7 1 1 0 3640.5 29.7 2.48 163.2 3.7
ME 8 1 0 1 5024.5 41.7 2.23 57.5 5.2
ME 9 −1 1 0 6683.4 55.9 2.30 3.1 7.0
ME 10 0 0 0 6255.3 53.3 2.18 16.4 6.7
ME 11 0 1 −1 5681.8 47.2 2.29 28.7 5.9
ME 12 −1 0 −1 7260.6 61.1 2.41 3.5 7.6
ME 13 1 0 −1 4175.8 35.2 2.39 92.5 4.4
ME 14 0 0 0 6323.5 53.4 2.26 17.5 6.7
ME 15 0 0 0 6270.4 51.4 2.27 19.6 6.4
ME 16 0 1 1 5381.4 44.9 2.16 49.4 5.6
ME 17 −1 −1 0 6914.6 59.2 2.37 4.8 7.4

Independent variables Levels used, actual (coded)

Low (−1) Medium (0) High (+1)

X1 = oil (g) 6 11 16
X2 = SMix (g) 10 13 16
X3 = water (mL) 68 76 84

in a more flexible and dynamic layer (Trotta et al., 1999; Wu et al., The ranges of other responses, Y1 and Y2 were 3640.5–7704.5 ␮g
2001). The drug in this energy-rich system can diffuse across the and 29.7–64.7 ␮g cm−2 h−1 , respectively.
flexible interfacial surfactant film between the phases; a thermo- The responses of these formulations ranged from a low drug
dynamic process that increases partitioning and diffusion into the penetration of 3640.5 ␮g (ME7, high level of oil and SMix and
stratum corneum. However at high levels of oil and surfactant mix- medium level of water) to a higher penetration of 7704.5 ␮g (ME4,
ture the drug permeation and flux were found to be decreased, this low level of oil, medium level of SMix and high level of water). For
was due to the affinity of LCDP towards oil phase and surfactant estimation of quantitative effects of the different combination of
system. factors and factor levels on Q24 , flux and lag time, the response
surface models were calculated with Design Expert software by
applying coded values of factor levels. The model described could
3.5. Formulation optimization by experimental design be represented as:

A three-factor, three-level Box–Behnken statistical experimen- Y1 (Q24 ) = 6240.74 − 1395.15X1 − 302.29X2 + 182.39X3
tal design was used to optimize the formulation variables as
− 172.35X1 X2 − 101.20X1 X3 − 191.83X2 X3
the response surface methodology requires 17 experiments. The
independent variables and the responses for all 17 experimen- − 298.86X12 − 491.98X22 + 99.47X32 (5)
tal runs are given in Table 2. The contour plots and 3D response
surface plots drawn using Design Expert software are shown in
Fig. 3. Based on the results of pseudo-ternary phase diagrams, Y2 (Flux) = 52.8 − 11.96X1 − 3.08X2 + 1.86X3 − 1.40X1 X2
appropriate ranges of the components were chosen. The oil phase
−0.73X1 X3 − 2.35X2 X3 − 2.71X12 − 4.24X22 + 0.59X32
concentration that could form microemulsion was found to be
6–16% and was selected as oil concentration to identify the (6)
optimum proportion of oil. Previous reports revealed that there
was a really tight relationship between the hydration effect of
the stratum corneum and the dermal permeation (Mohammed Y3 (lag time) = 2.22 − 0.034X1 + 0.02X2 − 0.044X3 − 0.098X1 X2
and Manoj, 2000), and the thermodynamic activity of drug in
microemulsions was a significant driving force for the release and − 0.05X1 X3 − 0.01 X2 X3 + 0.11X12 − 0.005X22
penetration of drug into skin (Mohammed and Manoj, 2000). Based
+ 0.04X32 (7)
on pseudo-ternary phase diagrams, the water content was deter-
mined between 68 and 84%. The surfactant mixture (surfactant,
cosurfactant, SMix 1:3), that could form clear microemulsion with
large area was selected as variable and was found to be 10–16%. 3.6. Fitting of data to the model
Design Expert software was used to optimize the formulation and
to develop the mathematical (quadratic) equations are shown in Formulation ME4 showed a significantly higher amount of drug
Eqs. (5)–(7). permeation (Y1 , Q24 ) and higher flux (Y2 ) among the formulations.
The responses, Q24 (Y1 ) and flux (Y2 ) were found to be signif- The responses observed for 17 formulations prepared were simul-
icantly higher (Y1 , 5381.4–7704.5 ␮g; Y2 , 44.9–64.7 ␮g cm−2 h−1 ) taneously fit to first order, second order and quadratic models using
only when the oil and SMix were used at 6 or 11% (v/v) and 10 or Design Expert 7.1.5. It was observed that the best fit model was
13% (w/v) concentration level respectively. The lag time (Y3 ) was quadratic model and the comparative values of R2 , standard devia-
found to be ranging from 2.16 to 2.48 h at low to high levels of SMix . tion and % coefficient of variation are given in Table 3 along with the
 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241 237

Fig. 3. Contour plot showing effect of (a) oil (X1 ) and SMix (X2 ), (b) oil (X1 ) and water (X3 ), (c) SMix (X2 ) and water (X3 ) on response Y2 (flux), corresponding response surface
plots (d–f).

regression equation generated for each response. A positive value independent variable X2 is having positive effect on the response
represents an effect that favors the optimization, while a negative lag time (Y3 ).
value indicates an inverse relationship between the factor and the The three-dimensional response surface plots (Fig. 3d–f) were
response. It is evident that the independent variable X3 (water) is drawn to estimate the effects of the independent variables on
having positive effect on the responses, Q24 (Y1 ) and flux (Y2 ). The response and to select the optimal formulation. The required
238 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241

Table 3
Summary of results of regression analysis for responses Y1 , Y2 and Y3 for fitting to quadratic model.

Quadratic model R2 Adjusted R2 Predicted R2 Adeq Precision SD % CV

Response (Y1 ) 0.9896 0.9761 0.8458 30.90 166.10 2.81

Response (Y2 ) 0.9950 0.9887 0.9502 45.04 0.99 2.00
Response (Y3 ) 0.7798 0.4966 −1.488 4.69 0.07 3.18

flux to reach therapeutic concentration calculated was found to interaction terms (X1 X2 , X1 X3 , X2 X3 , X12 , X22 , and X32 ) show how
be about 12.16 ␮g cm−2 h−1 . Hence, the penetration rate of opti- the Q24 changes when two variables are simultaneously changed.
mal formulations in the optimization process was set at above The negative coefficients for all 3 independent variables indicate an
12.16 ␮g cm−2 h−1 . Formulation, ME4 showed maximum flux of unfavorable effect on the Q24 , while the positive coefficients for the
64.7 ␮g cm−2 h−1 and could meet the target flux, calculated from interactions between 2 variables indicate a favorable effect on Q24 .
the pharmacokinetic parameters of LCDP indicating that the con- Among the 3 independent variables, the lowest coefficient value
centrations may be enough to elicit the pharmacological effect. is for X1 (−1395.15), indicating that this variable is insignificant in
prediction of Q24 .
3.7. Data analysis The value of R2 of Eq. (6) was found to be 0.9950, indicating
good fit (Table 3). The “Pred R-Squared” of 0.9502 is in reasonable
Formulations ME4, ME9, ME12 and ME17 had the higher Q24 agreement with the “Adj R-Squared” of 0.9887. “Adeq Precision” a
and flux. Table 4 shows the observed and predicted values with measure for the signal to noise ratio was found to be 45.04, indi-
residuals and percent error of responses for all the formulations. cating an adequate signal. The flux values of ME1, ME4, ME9, ME12
The Q24 and flux obtained at various levels of the 3 independent and ME17 were found to be more among the formulations. The flux
variables (X1 , X2 and X3 ) was subjected to multiple regression to values were found to be increased from medium to low levels of X1 ;
yield a second-order polynomial equation. The value of the corre- low to medium levels of variable X2 and low to high levels of X3 . The
lation coefficient (R2 ) of Eq. (5) was found to be 0.9896, indicating flux values measured for the different formulations showed wide
good fit (Table 3). The “Pred R-Squared” of 0.8458 is in reasonable variation (i.e., values ranged from a minimum of 29.7 ␮g cm−2 h−1
agreement with the “Adj R-Squared” of 0.9761. “Adeq Precision” in ME7 to a maximum of 64.7 ␮g cm−2 h−1 in ME4). The interac-
measures the signal to noise ratio. A ratio greater than 4 is desirable, tion terms (X1 X2 , X1 X3 , X2 X3 , X12 , X22 , and X32 ) show how the flux
the ratio of 30.90 (Table 3) indicates an adequate signal. The Q24 val- changes when 2 variables are simultaneously changed. The positive
ues measured for the different formulations showed wide variation coefficients X1 and X32 for the interactions between 2 variables indi-
(i.e., values ranged from a minimum of 3640.5 ␮g in ME7 to a max- cate a favorable effect on flux. Among the 3 independent variables,
imum of 7704.5 ␮g ME4). The results clearly indicate that the Q24 the lowest coefficient value is for X1 (−11.96), indicating that this
value is strongly affected by the variables selected for the study. The variable is insignificant in prediction of flux.
main effects of X1 , X2 , and X3 represent the average result of chang- The value of R2 of Eq. (7) was found to be 0.7798 (Table 3). The
ing one variable at a time from its low level to its high level. The signal to noise ratio was found to be 4.695, indicating an adequate

Table 4
Composition of checkpoint formulations, the predicted and experimental values of response variables and percentage prediction error.

Optimized formulation composition (X1 :X2 :X3 ) Response variable Experimental value Predicted value Percentage prediction error

Y1 7185.4 7244.5 −0.82

7.0:10.2:70.6 Y2 60.8 62.03 −2.02
Y3 2.41 2.38 1.24
Y1 7310.5 7265.8 0.61
6.7:12.7:73.7 Y2 61.8 61.4 0.65
Y3 2.37 2.36 −4.42
Y1 6987.6 7163.7 −2.52
8.0:13.4:83.0 Y2 59.4 60.4 −1.68
Y3 2.36 2.31 2.12
Y1 6445.5 6536.2 −1.41
10.0:12.4:72.5 Y2 53.8 55.1 −2.42
Y3 2.30 2.26 1.74
Y1 6910.1 6838.4 1.04
9.6:10.9:81.5 Y2 58.3 58.7 −0.69
Y3 2.28 2.26 2.59
Y1 6740.3 6658.9 1.21
9.5:12.8:69.5 Y2 54.8 55.9 −2.01
Y3 2.34 2.30 1.71
Y1 6487.6 6399.5 1.36
10.5:12.7:73.0 Y2 53.1 54.0 −1.69
Y3 2.28 2.25 3.43
Y1 5983.8 6141.4 −2.63
10.4:14.8:71.7 Y2 50.1 51.6 −2.99
Y3 2.30 2.26 2.59
Y1 6520.8 6452.1 2.59
9.1:15.3:74.3 Y2 53.8 53.3 1.66
Y3 2.25 2.24 −1.36
Y1 6115.3 6225.7 −1.81
12.5:10.9:83.0 Y2 52.4 53.5 −2.10
Y3 2.20 2.17 1.36
 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241 239

signal. The lag time values of ME3, ME6, ME10 and ME16 were
found to be less among the formulations, however the difference
was insignificant (p > 0.05). The lag time values were found to be
increased from low to high levels of X1 ; high to low levels of X2
and medium to low levels of variable X3. The results attributed to
that the deposition of drug from MEs within the layers of stratum
corneum, might increase the lag time. The solubilizing capacity and
affinity of drug in variable X2 could result in increased lag time. The
interaction terms (X1 X2 , X1 X3 , X2 X3 , X12 , X22 , and X32 ) show how
the lag time changes when 2 variables are simultaneously changed.
The positive coefficients (X2 , X2 X3 , X12 and X32 ) for the interactions
between 2 variables indicate a favorable effect on lag time. Among
the 3 independent variables, the lowest coefficient value is for X3
(−0.044), indicating that this variable is insignificant in prediction
of lag time.

3.8. Contour plots and response surface analysis

Two-dimensional contour plots and three-dimensional

response surface plots are shown in Fig. 3, which are very useful
to study the interaction effects of the factors on the responses.
These types of plots are useful in study of the effects of two factors
on the response at one time. In all the presented figures, the third
factor was kept at a constant level. All the relationships among
the three variables are non-linear, although exhibit a nearly linear
relationship of factor X2 with factors X1 and X3 , in the form of
almost straight lines up to the medium level of SMix (Fig. 3).
At higher concentrations of SMix these become curvilinear or
non-linear.
Factors X2 and X3 have curvilinear relationship at all levels of the
two variables on the response Y2 . Response surface plots show the
relationship between these factors even more clearly. The Q24 and
flux were found to be increased with increasing concentrations of
either SMix (up to medium level) or water at constant concentration
of oil phase.

3.9. Optimization

The optimum formulation was selected based on the criteria

of attaining the maximum value of Q24 , maximum flux and low
value of lag time by applying constraints on Y1 (5800 ≤ Y ≤ 8000),
Y2 (50 ≤ Y ≤ 65) and Y3 (2.00 ≤ Y ≤ 3.00). Upon trading of various
response variables and comprehensive evaluation of feasibility
and exhaustive grid search, the formulation composition with
oil concentration of 6.5%, SMix 11.4% and water 82.0% was found
to fulfill the maximum requisite of an optimum formulation
because of maximum Q24 (7347.2 ␮g), flux (62.9 ␮g cm−2 h−1 )
and low lag time (2.16 h) values. The optimized formulation was
formulated as microemulsion gel using hydroxy propyl methyl
cellulose (HPMC K 4 M) at 4% (w/v) in the microemulsion. The
microemulsion gel showed about 5820.8 ␮g of drug permeated
in 24 h, flux 46.1 ␮g cm−2 h−1 and with lag time of 2.22 h. The
flux of microemulsion gel was found to meet the target flux
(12.16 ␮g cm−2 h−1 ). Fig. 4. Linear correlation plots (a, b, c) between actual and predicted values.

3.10. Validation of response surface methodology describe the domain of applicability of RSM model. Linear correla-
tion plots between the actual and the predicted response variables
Ten checkpoint formulations were obtained from the RSM, the were shown in Fig. 4. The linear correlation plots drawn between
composition and predicted responses of which are listed in Table 4. the predicted and experimental values demonstrated high values of
To confirm the validity of the calculated optimal parameters and R2 (Q24 , 0.9421; flux, 0.9698; lag time, 0.9539) indicating goodness
predicted responses, the optimum formulations were prepared of fit.
according to the above values of the factors and subjected to ex vivo
permeation studies. From the results presented in Table 4, the pre- 3.11. Determination of droplet size
dicted error was below 5%, indicating that the observed responses
were very close to the predicted values. Percentage prediction error The parameters for physicochemical characters of the optimized
is helpful in establishing the validity of generated equations and to formulation (ME-OPT) were as follows: 30.5 nm for mean particle
240 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241

Fig. 6. Ex vivo–in vivo correlation of cumulative amount permeated ex vivo Vs AUC.

Fig. 5. Serum profiles of LCDP in Rabbits, after administration of oral suspension that could delay the permeation of LCDP from microemulsion gels
and ME gel, each containing 8 mg of LCDP, values represented are mean ± SD (n = 6). in contrast, suspension administered by oral route is an immediate
release dosage form. The overall mean value of AUC0−t by trans-
size with a poly dispersity index of 0.11 and 2.31 cP of gel index. All dermal route was 3.5 times higher than that of oral route, and the
microemulsion formulations were stable at ambient temperature difference was found to be statistically significant (p < 0.05) demon-
in the presence or absence of LCDP. No changes of particle size, strating improved bioavailability of LCDP from microemulsion gel.
phase separation and degradation of LCDP were observed during 6 This could be due to avoidance of first-pass hepatic metabolism by
months. transdermal route. The reported oral bioavailability of lacidipine
was 10% (Mc Cormack and Wagstaff, 2003), because of first-pass
3.12. Stability studies metabolism. In the present study the bioavailability of LCDP by
transdermal route was found to be 35.1%. Therefore for the effec-
The LCDP content in the stability samples after 6 months was tive management of chronic hypertension LCDP in the form of
found to be 98.6 and 98.8% in ME4 and ME gel respectively. The microemulsion gel could provide an effective treatment.
results reveal that LCDP was stable during the study. The ex vivo per-
meation profiles are shown in Fig. 2. Formulation ME4 and ME gel 3.15. Ex vivo–in vivo correlation
showed 7540.0 and 5715.8 ␮g of LCDP permeated in 24 h. The flux
values were found to be 46.10 and 43.7 ␮g cm−2 h−1 , respectively. Ex vivo–in vivo correlation between the cumulative % of drug
The results suggested that the formulations did not show signifi- permeated across rat abdominal skin and AUC showed a biphasic
cant difference (p > 0.05) in permeation profiles compared to that curve pattern (Fig. 6), which can be distinguished into two regions
of initial permeation profiles indicating that both the formulations for ME gel. Good linear correlation was observed with correla-
are stable. tion coefficients, R2 = 0.938 during lag phase and R2 = 0.993 during
absorption phase. Point to point correlation of ex vivo permeation
3.13. Skin irritation study of drug to in vivo performance was observed, indicating that it fol-
lows type A correlation (Emami, 2006). The slow permeation of
The skin irritation studies could not find any irritation, erythyma LCDP through skin in initial stages is explained as follows; in first
indicating that the ME gel is non-irritant. phase LCDP was released and permeated through skin and deposi-
tion of LCDP took place in skin layers and concentration build up
3.14. In vivo bioavailability studies was maintained. Permeation and concentration buildup at the skin
is the lag phase observed in the first region. Concentration built
The results from the bioavailability study (Table 5, Fig. 5) reveal up resulted in the flux establishment and AUC increased at a rapid
that LCDP is released and permeated well from microemulsion gel rate in the second phase. This indicates that initially drug perme-
by transdermal route, as compared to the oral suspension. The CMax ated into medium rapidly but it takes some time for permeation
of LCDP was found to be 97.4 and 110.0 ng mL−1 after administra- and absorption. Once the necessary flux is established, absorption
tion of oral suspension and ME gel respectively. The CMax , TMax and was rapid as large amount of drug is deposited in the layers of skin.
AUC profiles were compared. In four rabbits, CMax was higher for
transdermal route than oral route and in the remaining rabbits CMax 4. Conclusions
for oral route was higher than the transdermal route. The TMax val-
ues in all rabbits were higher for transdermal administration than The present study conclusively demonstrates the use of a
the oral administration and the difference was statistically signif- Box–Behnken statistical design is valid for predicting the Q24 , flux,
icant (p < 0.05). This difference was because of stratum corneum and lag time in optimization of microemulsion formulations. The

Table 5
Pharmacokinetic parameters of LCDP in rabbits after administration of oral suspension and ME Gel each containing 8 mg of LCDP, values represented are mean ± SD (n = 6).

Formulation CMax (ng mL−1 ) TMax (h) AUC0–t (ng h mL−1 ) AUC0–∞ (ng h mL−1 )

Oral suspension 97.4 ± 9.25 1.2 ± 0.41 782.4 ± 360.76 804.8 ± 372.82
ME gel 110.0 ± 28.55 13.0 ± 2.45 2272.0 ± 982.51 2702.2 ± 1226.55
 R. Gannu et al. / International Journal of Pharmaceutics 388 (2010) 231–241 241

derived polynomial equations and contour plots aid in predict- Gannu, R., Yamsani, V.V., Palem, C.R., Yamsani, S.K., Yamsani, M.R., 2009. Devel-
ing the values of selected independent variables for preparation of opment of high performance liquid chromatography method for Lacidipine
in rabbit serum: application to pharmacokinetic study. Anal. Chim. Acta 632,
optimum microemulsion formulations with desired properties. The 278–283.
developed microemulsion gel formulation was efficacious for the Gasco, M.R., Gallarate, M., Pattarino, F., 1991. In vitro permeation of azelaic acid from
delivery of lipophilc and poorly soluble drugs such as lacidipine. viscosized microemulsions. Int. J. Pharm. 69, 193–196.
Goldberg-Cettina, M., Liu, P., Nightingale, L., Kurihara-Bergstrom, T., 1995. Enhanced
The results demonstrated that the formulation was nonirritating transdermal delivery of estradiol in vitro using binary vehicles of isopropyl
and did not cause any erythyma upon transdermal administration. myristate and short-chain alkanols. Int. J. Pharm. 114, 237–245.
Results of bioavailability study showed improved permeation of Govender, S., Pillay, V., Chetty, D.J., Essack, S.Y., Dangor, C.M., Govender, T., 2005.
Optimisation and characterisation of bioadhesive controlled release tetracycline
the drug from the microemulsion gel compared to oral suspen- microspheres. Int. J. Pharm. 306, 24–40.
sion. Good ex vivo–in vivo correlation was obtained with correlation Hayton, W.L., Chen, T., 1982. Correction of perfusate concentration for sample
coefficients of 0.938 and 0.993 during lag and permeation phase removal. J. Pharm. Sci. 71, 820–821.
Ho, H.O., Hsiao, C.C., Sheu, M.T., 1996. Preparation of microemulsions using polyg-
respectively.
lycerol fatty acid esters as surfactant for the delivery of protein drugs. J. Pharm.
Sci. 85, 138–143.
Acknowledgements Kamal, M.A., Iimura, N., Nabekura, T., Kitagawa, S., 2007. Enhanced skin permeation
of diclofenac by ion-pair formation and further enhancement by microemulsion.
Chem. Pharm. Bull. 55, 368–371.
One of the authors (Ramesh Gannu) thanks AICTE, New Delhi, Kreilgaard, M., 2002. Influence of microemulsions on cutaneous drug delivery. Adv.
India for providing financial assistance in the form of National Doc- Drug Deliv. Rev. 54, 77–98.
toral Fellowship (NDF). The authors also acknowledge the liberal Kreilgaard, M., Pedersen, E.J., Jaroszewski, J.W., 2000. NMR characterisation and
transdermal drug delivery potential of microemulsion systems. J. Control
help of Dr. Reddys Laboratories, Hyderabad, India for providing Release 69, 421–433.
lacidipine and Labrasol (Gatteffose, France) in the form of gift sam- Lawrence, M.J., Rees, G.D., 2000. Microemulsion-basedmedia as novel drug delivery
ples, respectively. systems. Adv. Drug Deliv. Rev. 45, 89–121.
Lee, C.R., Bryson, H.M., 1994. Lacidipine: a review of its pharmacodynamic and
pharmacokinetic properties and therapeutic potential in the treatment of hyper-
References tension. Drugs 48, 274–296.
Levang, A.K., Zhao, K., Singh, J., 1999. Effect of ethanol/propylene glycol on the in
Baboota, S., Al-Azaki, A., Kohli, K., Ali, J., Dixit, N., Shakeel, F., 2007. Development vitro percutaneous absorption of aspirin, biophysical changes and macroscopic
and evaluation of a microemulsion formulation for transdermal delivery of barrier properties of the skin. Int. J. Pharm. 181, 255–263.
terbinafine PDA. J. Pharm. Sci. Technol. 61, 276–285. Mc Cormack, P.L., Wagstaff, A.J., 2003. Lacidipine a review of its use in the manage-
Baroli, B., Lopez-Quintela, M.A., Delgado-Charro, M.B., Fadda, A.M., Blanco-Mendez, ment of hypertension. Drugs 63, 2327–2356.
J., 2000. Microemulsions for topical delivery of 8-methoxsalen. J. Control Release Mohammed, C., Manoj, V., 2000. Aerosol-OT microemulsions as transdermal carriers
69, 209–218. of tatracaine hydrochloride. Drug Dev. Ind. Pharm. 26, 507–512.
Bonacucina, G., Martelli, S., Palmieri, G.F., 2004. Rheological, mucoadhesive and Peltola, S., Saarinen-Savolainen, P., Kiesvaara, J., Suhonen, T.M., Urtti, A.,
release properties of Carbopol gels in hydrophilic cosolvents. Int. J. Pharm. 282, 2003. Microemulsions for topical delivery of estradiol. Int. J. Pharm. 254,
115–130. 99–107.
Box, G.E.P., Behnken, D.W., 1960. Some new three level designs for the study of Rhee, Y.S., Choi, J.G., Park, E.S., Chi, S.C., 2001. Transdermal delivery of ketoprofen
quantitative variables. Technometrics 2, 455–475. using microemulsions. Int. J. Pharm. 228, 161–170.
Chen, H., Mou, D., Du, D., Chang, X., Zhu, D., Liu, J., Xu, H., Yang, X., 2007. Hydro- Sintov, A.C., Shapiro, L., 2004. New microemulsion vehicle facilitates percutaneous
gel thickened microemulsion for topical administration of drug molecule at an penetration in vitro and cutaneous drug bioavailability in vivo. J. Control. Release
extremely low concentration. Int. J. Pharm. 341, 78–84. 95, 173–183.
Chen, H.B., Chang, X.L., Du, D.R., Li, J., Xu, H.B., Yang, X.L., 2006. Microemulsion based Trotta, M., Gallarate, M., Pattarino, F., Carlotti, M.E., 1999. Investigation of the phase
hydrogel formulation of ibuprofen for topical delivery. Int. J. Pharm. 315, 52–58. behaviour of systems containing lecithin and 2-acyl lysolecithin derivatives. Int.
Chopra, S., Pati, G.V., Motwani, S.K., 2007. Release modulating hydrophilic matrix J. Pharm. 190, 83–89.
systems of losartan potassium: optimisation of formulation using statistical Wu, H., Ramachandran, C., Weiner, N.D., Roessler, B.J., 2001. Topical transport of
experimental design. Eur. J. Pharm. Biopharm. 66, 73–82. hydrophilic compounds using water-in-oil nano emulsions. Int. J. Pharm. 220,
Clarke, 2007. Clarke’s Analysis of Drugs and Poisons., www.medicinescomlete.com. 63–75.
Emami, J., 2006. In vitro–in vivo correlations: from theory to applications. J. Pharm. Zhao, X., Liu, J.P., Zhang, X., Li, Y., 2006. Enhancement of transdermal delivery of
Pharm. Sci. 9, 31–51. theophylline using microemulsion vehicle. Int. J. Pharm. 327, 58–64.
Gallarate, M., Gasco, M.R., Rua, G., 1990. In vitro release of azelaic acid from oil-in-
water microemulsions. Acta Pharm. Yugoslv. 40, 541–545.
 Progress in Retinal and Eye Research 21 (2002) 15–34

Microemulsions as ocular drug delivery systems: recent developments

and future challenges
Th.F. Vandamme*
!
Laboratoire de Chimie Therapeutique et Nutritionnelle, Faculte! de Pharmacie, Universite! Louis Pasteur, 74 Route du Rhin,
B.P. 24, 67401 Illkirch Cedex, France

Abstract

Eye drops are the most used dosage form by ocular route, in spite of low bioavailability and the pulsed release of the drug.
However, due to their intrinsic properties and specific structures, the microemulsions are a promising dosage form for the natural
defence of the eye. Indeed, because they are prepared by inexpensive processes through autoemulsification or supply of energy, and
can be easily sterilized, they are stable and have a high capacity of dissolving the drugs. The in vivo results and preliminary studies
on healthy volunteers have shown a delayed effect and an increase in the bioavailability of the drug. The proposed mechanism is
based on the adsorption of the nanodroplets representing the internal phase of the microemulsion, which constitutes a reservoir of
the drug on the cornea and should then limit their drainage. r 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Microemulsion; Ocular drug delivery systems; Bioavailability; Surfactant; Diffusion kinetics; Site-specific delivery

CONTENTS

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1. Formulation of microemulsions intended for ocular route . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.2. Structure of microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.3. Properties of microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4. Choice of various components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4.1. The surfactant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4.2. The co-surfactant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4.3. The organic phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4.4. The aqueous phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5. Preparation processes and examples of eye drop microemulsions . . . . . . . . . . . . . . . . . . . . 18
1.5.1. Autoemulsification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5.2. Process based on supply of energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2. Significance of microemulsions in ocular drug delivery devices . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1. Physico-chemical parameters of microemulsions intended for ophthalmic route . . . . . . . . . . . . 22
2.1.1. Improvement of solubilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1.2. Zeta potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

*Tel.: +33-90-24-41-06; fax: +33-90-24-43-17.

E-mail address: vandamme@pharma.u-strasbg.fr. (Th. Vandamme)

1350-9462/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 1 3 5 0 - 9 4 6 2 ( 0 1 ) 0 0 0 1 7 - 9
16 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

2.2.Results in vitro: use of diffusion cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.2.1. Study of diffusion kinetics through an artificial membrane . . . . . . . . . . . . . . . . . . . 25
2.2.2. Studies of diffusion kinetics of the drugs through an excised cornea of rabbit . . . . . . . . . 27
2.2.3. Comments on diffusion cells and artificial membranes . . . . . . . . . . . . . . . . . . . . . . 27
2.3. Results of in vivo trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.4. Results on man . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3. Conclusion and future challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

1. Formulation of microemulsions intended for ocular which water and oil should be separated by a regular or
route irregular interfacial layer.

1.1. Introduction 1.3. Properties of microemulsions

For several years, microemulsions have been investi- As described above, the main characteristic of the
gated as carriers used to find new drug delivery systems. microemulsions is their transparent appearance due to
For about a decade, their potential in ophthalmology the high level of dispersion of the internal phase, the size
has been studied by several research teams. The drug of which ranges from 100 to 1000 A ( (Aiache and
delivery system currently used in this field, namely eye Devissaguet, 1983 and Sharma and Shah, 1985). The
drops, has several disadvantages, such as a very low microemulsions are liquid and behave as a Newtonian
bioavailability (1–10%) (J.arvinen et al., 1995) of the liquid. They are not very viscous; the bicontinous
drugs which must be absorbed at this site and must be systems are less viscous than the nanodroplets systems.
inserted several times a day. Also, the absorption These dispersed systems are isotropic as opposed to the
through the corneal barrier is difficult due to the liquid crystals, which are anisotropic. In terms of the
lachrymal drainage. Therefore, microemulsions are an manufacture, their formation is spontaneous and does
interesting alternative. The industrial production and not require much energy, and are thermodynamically
sterilization are relatively simple and inexpensive; they stable (Langevin et al., 1985).
are thermodynamically stable and inherently provide the
capacity to make soluble lipophilic drugs, which 1.4. Choice of various components
depends also on the lipophilic phase used.
1.4.1. The surfactant
1.2. Structure of microemulsions The preferential adsorption of the surfactant enables
the modification of the physico-chemical properties of
In spite of their similarity, the terms ‘‘microemulsion’’ the interface due to its amphiphilic nature. While the
and ‘‘emulsion’’, characterize two very different systems surfactant concentration is 0.1% in weight in emulsions,
both by their physical and thermodynamic properties it accounts for at least 10% in microemulsions due to
and by their structure. In both cases, the systems consist the increase of the interface area between the aqueous
of an aqueous phase, a lipophilic phase, and a surfactant and oily phases (Langevin et al., 1985). This high
agent. A co-surfactant is also required for some concentration of surfactants can lead to ocular toxicity.
microemulsions. The addition of this compound results That is why, it might be better to decrease the quantity
in a homogeneously dispersed system, which can diffuse of surfactant and choose a non-spontaneous prepara-
the light, and, as opposed to emulsions, is thermo- tion process as described below.
dynamically stable (Bothorel, 1985). The ionic surfactants are generally too toxic to be
When referring to emulsions, the simplest representa- used for this application. Therefore, non-ionic surfac-
tion of a microemulsion is the droplet model. These tants are preferred (Attwood, 1994). These surfactants
systems actually exist when they have a low percentage are easily soluble in water due to the presence of ether
of oil or water in the internal phase (E10%). These functional groups. The most used surfactants in the
dispersions of oil or water nanodroplets in an external preparation of microemulsions are the poloxamers,
phase are stabilized by an interfacial film of surfactant polysorbates (Evitts et al., 1991), polyethylene glycol
and co-surfactant (Fig. 1a–d) (Lawrence, 1994). In the and tyloxapol (Fig. 1a–d).
case of larger quantities of oil or water, the structure of Amongst the polyethylene glycols (PEG), the PEG
nanodroplets that was described above does not apply. 200 is preferred due to its lowest viscosity. As for the
As a result, a bicontinuous system has been proposed in poloxamers, the polyoxyethylene glycol blocks are
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 17

origin of this compound before its use, as well as the

purification degree (variation of fatty acids content) as
its physical properties, composition and concentration
of phospholipids depend on them. Thus, for example,
the lecithin from egg yolk contains 21% of phosphati-
dylcholine, 22% of phosphatidylethanolamine and 19%
of phosphatidylinositol. The Miranol MHTs (aqueous
solution containing 3.4% of NaCl and 34.5% of
lauroamphodiacetate), another amphoteric surfactant,
is also well tolerated by the eye.

1.4.2. The co-surfactant

The co-surfactants have three functions: (i) they
provide very low interfacial tensions required for the
formation of microemulsions and their thermodynamic
stability, (ii) they can modify the curvature of the
interface based on the relative importance of their
apolar groups (Langevin et al., 1985; Pore, 1992), (iii)
finally, they act on the fluidity of the interfacial film. If
the film is too rigid, it prevents the formation of
microemulsion and results in a more viscous and
birefringent phase. The co-surfactant provides addi-
tional fluidity and is equivalent to a branched surfactant
(Fig. 2). The existence of unsaturated bounds on the
hydrocarbon chain of the surfactants equally increases
the fluidity of the film.
The co-surfactants that are used consist of small mole-
cules, which generally are alcohols with low molecular
weight and glycols with the length of the carbon chain
ranging from C2 and C10. Some alcohols, such as
pentanol and hexanol, are not used in pharmaceutical
applications due to their highly irritating nature. Amines
with short chains can also be used as co-surfactants.

1.4.3. The organic phase

The choice of the oily phase (which is dispersed for
ophthalmic drug delivery applications) is important
because it conditions both the existence of the micro-
emulsion and the solubilization of the drug. The same
type of oil does not generate both conditions; therefore

Fig. 1. Surfactants used in preparation of microemulsions (a) Polox-

amers; (b) Polysorbates; (c) Polyethylene glycol; (d) Tyloxapol.

hydrophilic whereas the polyoxypropylene blocks are (d) (e)

hydrophobic. The highest proportion of oxyethylene is
found in the poloxamer 188 (82% w/w). As a result, the
value of its hydrophilic/lipophilic balance is the highest
(HLB=29) and is preferred for the preparation of
lipophilic/hydrophilic (L/H) emulsion. Finally, polysor-
bate 80 is the surfactant, which has both a high (a) (b) (c)
hydrophilic/lipophilic balance (HLB=15) and the low-
est viscosity (Wade and Weller, 1994). Fig. 2. Influence of co-surfactants on the fluidity of the interfacial film.
Amphoteric surfactants, like lecithin, are interesting Linear surfactant (a), branched surfactant (b), co-surfactant (c), rigid
due to their low toxicity. It is imperative to specify the film (d), fluid film (e).
18 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

it is necessary to find the appropriate combination. Oils which can be determined by visual observation of
with excessively long hydrocarbon chains do not result various mixtures of surfactant, co-surfactant, oily phase,
in microemulsions. The shorter the chain, the deeper the and aqueous phase reported in a phase diagram.
penetration of the organic phase into the interfacial film Two techniques are presented in the literature, each of
and more important the existence range of the micro- them resulting in microemulsions: (i) ‘‘Exact’’ process by
emulsion. On the contrary, the capacity of solubilization autoemulsification; (ii) process based on supply of
by the organic phase increases with the length of the energy.
chain. Therefore, the choice is based on the solubility of
the drug. 1.5.1. Autoemulsification
Polar oils, such as triglycerides with medium or long Due to the spontaneous formation of the microemul-
chains (MCT, LCT) are also preferred instead of non- sions, they can be prepared in one step by mixing the
polar oils, based on the better solubility of drugs. In constituents with reduced roughness. The order of
addition, the organic phases should not be too polar the addition of the constituents is not considered a
because they prevent the formation of the microemul- critical factor for the preparation of microemulsions,
sion. The MCT, which are more water-soluble than the but it can influence the time required to obtain
LCT, are often preferred because they provide the equilibrium. This time will increase if the co-surfactant
dissolution of large concentrations of lipophilic drugs. is added to the organic phase, because its greater
The most often used external phases consist of solubility in this phase will prevent the diffusion in the
vegetable oils like soja oil, castor oil (Evitts et al., aqueous phase.
1991), triglycerides for which 95% of fatty acids are Table 1 and Fig. 3 show an example of the formula-
made of 8–10 carbon atoms, Myglyol 812s (triesters of tion of eye drop microemulsion and its preparation.
glycerol, capric and caprylic acids), isopropyl myristate, The dosage form, which was obtained, is translucent
fatty acids, such as oleic acid, and esters of saccharose, and the average diameter of the droplets (internal phase)
such as mono-, di- or tri-palmitates of saccharose. As is 15 nm.
these excipients are well tolerated by the eye, their Ruth et al. (1995) compared the efficacy of two
degree of purity must be high in order to prevent any alcohols, butanol and ethanol, in similar systems
contamination with potentially irritating substances. constituted of isopropylmyristate, egg lecithin and
water. The quantity of ethanol required for the
1.4.4. The aqueous phase preparation of microemulsions is seven times higher
The aqueous phase must contain several additives, than the quantity of butanol. The difference of efficacy
such as buffers, antibacterial and isotonic agents. They between co-surfactants is based on the length of the
can affect the area of existence of the microemulsions, carbon chain. To obtain the curvature required for the
and therefore they must be studied in the presence of formation of oily droplets in an external hydrophilic
other constituents of the microemulsions. Salinity affects phase (L/H emulsion), the alcohol must be inserted in
the phase diagrams when ionic surfactants are added and the interfacial area in order to reduce the rigidity of the
decreases the phase inversion temperature (PIT) of the lecithin film. The distribution of the alcohol between the
non-ionic surfactants. The preparation of microemul- interface and the continuous aqueous phase is based on
sions is very sensitive to temperature, if the PIT is close to its hydrophilic character. The ethanol has an interface/
the operating conditions. Adjustment of the initial pH at water distribution coefficient lower than the butanol.
7–8 is also important in order to minimize the hydrolysis Therefore, its higher solubility in water requires the use
of the phospholipids and the triglycerides to fatty acids, of higher quantities in order to obtain an interface with
which can decrease the pH of the microemulsion. similar mechanical properties to that obtained by using
The preservatives which are usually used in eye drop butanol.
formulations cannot be incorporated in microemulsions.
They must not interact with surfactants and result in
complexes, nor should they be absorbed in nanodroplets, Table 1
which would considerably decrease the bacterial activity Example of eye drops microemulsion (Gasco et al., 1988)
in bases. Thiomersal and chlorobutanol with concentra- Component Amount Function
tions of 0.01–0.2% can be used; their combination (% w/w)
increases their properties (Benita and Muchtar, 1992).
Timolol base 2.60 Drug
Egg lecithin 28.70 Surfactant
1.5. Preparation processes and examples of eye drop Isopropylmyristate (IPM) 11.70 Internal phase
microemulsions Octanoic acid 4.70 Ion pairing agent
1-butanol 14.90 Co-surfactant
a-Tocopherol 0.01 Antioxidizing agent
The preparation of microemulsions requires the
Isotonic phosphate buffer (pH=7.4) QSP 100 External phase
determination of the existence range of microemulsions,
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 19

IPM α-tocopherol Octanoic acid

Mixture

Solubilization
Lecithin

Solubilization
Timolol base

Phosphate buffer

Addition under

roughness

Solubilization
1-Butanol

Microemulsion

Fig. 3. Diagram for the preparation of eye drop microemulsion containing timolol base.

Siebenbrodt and Keipert (1993) (Table 2; Fig. 4) increase in the lipophilic character of the surfactant
studied the influence of non-ionic surfactants on the caused by the low interaction between the hydrophilic
stability of microemulsions based on the existence of segment and water, results in the release of an excess of
PIT. The changes in temperature can make the water which does not interact with poloxamers and
microemulsion go from a one-phase system (Winsor leads to Winsor II type systems (Fig. 5). On the
IV type) (Fig. 5) to multiphase systems (Eccleston, contrary, at a low temperature (o101C), the relative
1988). Thus, at a high temperature (>581C), the relative hydrophilic character increases and the excess of the
20 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Table 2 Winsor I Winsor II Winsor III Winsor IV

Example of eye drops microemulsion (Siebenbrodt and Keipert, 1993)

Component Amount Function

(% w/w) Oil
Oil
Chloramphenicol 0.5 Drug
Triacetine 20.0 Organic phase (internal phase)
Symperonics L64 15.0 Surfactant Water
Water
Propylene glycol 40.0 Co-surfactant
Water 24.5 Aqueous phase (external phase)
Fig. 5. Classification of Winsor.

research teams have chosen another preparation method

organic phase, the triacetin, for which the solubility in for microemulsions.
water decreases at low temperature, is rejected as phase
in excess, leading to a Winsor I type (Fig. 5). If a temp- 1.5.2. Process based on supply of energy
erature higher than 581C is not very probable under In this case, microemulsions are not obtained
normal conditions, a temperature lower than 101C can spontaneously. A decrease of the quantity of surfactants
be encountered during winter storage. That is why the results in the use of high-pressure homogenizers in order
authors study the stability of microemulsions at 51C. to obtain the desired size of droplets that constitute the
However, these two examples result in ocular irrita- internal phase as opposed to the former technique.
tions, which are induced by high concentrations of Benita and Levy (1993) have studied the efficacy of
surfactants. In order to overcome this problem, other various equipment for obtaining particles of different

Chloramphenicol Poloxamer Triacetine

Mixture

Water Propylene glycol

Mixture

Addition under

roughness

Microemulsion

Fig. 4. Diagram for the preparation of eye drop microemulsion containing chloramphenicol.
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 21

Table 3 the temperature depending on the sensitivity of the drug

Example of eye drops microemulsion (Muchtar et al., 1997) to heat. Cooling the preparation is required before its
Component Amount Function introduction in the high-pressure homogenizer, which
(% w/w) can raise the temperature. A blue opalescent micro-
Indomethacin 0.125 Drug emulsion is obtained which consists of 100 nm droplets.
MCT 10.0 Oily phase In the formulation presented in Table 3 and illustrated
Lipo.ıd E-80s 0.75 Surfactant in Fig. 6, Lipo.ıd E-80s is a mixture of phospholipids,
a-Tocopherol 0.02 Antioxydizing agent which contains 80% phosphatidylcholine, 8% phospha-
Miranol MHTs 0.5 Co-surfactant
tidylethanolamine, 3.6% non-polar lipids and 2% of
Glycerol 2.25 Osmotic agent
Distilled water QSP 100 Aqueous phase sphingomyelin (Zeevi et al., 1994). Miranol MHTs is an
aqueous solution containing 3.4% NaCl and 34.5%
lauroamphodiacetate which is an amphoteric surfactant.
sizes. Two steps are required: the first step produces a This surfactant has the advantage of easily penetrating
coarse emulsion (0.65 mm) by using a high-speed mixer the layer of phospholipids. The combination of these two
(Polytrons). The second step consists of using a high- surfactants is important for the stabilization of the
pressure homogenizer (Gaulins). The dispersion of the microemulsion, as the phospholipids alone are not
oily phase in the aqueous phase is also facilitated by sufficient. The concentrations of surfactants (0.75%
heating the phases before mixing themFthe choice of and 0.5%) are lower than those used in the examples 1

Figure
Lipoïd E-80 MCT Water

α -tocopherol
Glycerol Miranol

Indomethacin Mixture (70˚C)

pH 3.8
Mixture (70˚C)
Mixture (85˚C)

Dispersion

Cooling

Homogeneization

Microemulsion

Fig. 6. Diagram for the preparation of eye drop microemulsion containing indomethacin.
22 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Table 4 Keipert, 1993 determined the maximal percentage for

Example of a positively charged eye drops microemulsion (Elbaz et al., three drugs (indomethacin, chloramphenicol, sodium
1993)
diclofenac) when dissolved in each of the microemulsion
Component Amount Function constituents (Table 2 and Fig. 7).
(% w/w) In an aqueous solution, each constituent of the
Pilocarpine 1.0 Drug microemulsion increased the solubility of the three
MCT 8.0 Oily phase drugs, except for the triacetin of the sodium diclofenac,
Glycerol 2.25 Co-surfactant which is a salt. In addition, the microemulsion strongly
Poloxamers 188 2.0 Co-surfactant
improved the solubilization.
Lipo.ıd E-80s 1.0 Surfactant
Stearylamine 0.2 Positively charged lipid The choice of the components of the internal phase is
a-Tocopherol 0.1 Antioxydizing agent an important preliminary step in the establishment of
Distilled water QSP 100 Aqueous phase the formulation. In order to improve the solubilization
of the piroxicam and indomethacin, Klang et al. (1996)
and 2. The oily phase was chosen based on the and Muchtar et al. (1997) studied their solubility in
several internal phases (Figs. 8 and 9).
solubilization capacity of indomethacin. The quantities
However, the mixture which allows the optimal
of the internal phase and surfactants are in the range that
solubility of the drugs was not selected for the final
was recommended by Benita and Muchtar (1992): 1–
formulation for stability reasons, as in the case when
20% of oily phase in order to maintain the fluidity of the
oleic acid was used.
formulation, 0.2–1% phospholipids and 0.2–0.5% am-
photeric agents. The ratio between the quantities of
phospholipids and amphoteric surfactant must be 1 : 1 to
4 : 1Fa 2 : 1 ratio is preferred. 2.1.2. Zeta potential
Recently, while working on microemulsions obtained
The value of the pH is determined based on the
by the supply of energy, Benita and Levy (1993)
solubility of the drug in the oily phase. As the pKa of
theorized that the charge of the droplets which
indomethacin is 4.5, the adjustment of the pH to 3.8
constitute the internal phase could influence their
prevents its ionization and facilitates its localization in
absorption in the ocular route. As the corneal area is
the oily phase. The size of the internal phase of the micro-
negatively charged, the positively charged droplets
emulsion is 110 nm and has a
50 mV Zeta potential.
might bind to the sites. The residence time of the
The value of this parameter is important because the
charge of the internal phase provides information on the microemulsion on the cornea would increase and thus
influence the drug release. The charge is provided by a
efficacy of the induced electrical barrier.
positively charged surfactant, the stearylamine.
Table 4 presents the composition of a positively
This component has also been used in the past to
charged microemulsion. The size of the internal phase of
positively charge liposomes.
the microemulsion is 103 nm and the Zeta potential is
The toxicological studies are contradictory regarding
+8.6 mV. This charge depends on the surfactants in the
their ocular effects. Klang et al. (1994) studied the
interface. As the Poloxamers 188 is non-ionic and the
toxicity of this compound on rabbits based on a
phospholipids have an amphoteric character, the charge
mainly depends on the stearylamine. Elbaz et al., 1993, formula with no drug content, the composition of which
is presented in Table 4, and containing pilocarpine.
studied the influence of this amine on the Zeta potential,
In this study, the quantity of stearylamine has been set
that is, one that increases with the increase in the
at 0.3%.
quantity of stearylamine and reaches a maximum of
The instillation of this microemulsion eight times a
0.3%, which might indicate a saturation of the amine on
day for five days on four rabbits did not result in any
the interface. However, this concentration has no
irritation or inflammation. The cornea and its epithe-
significant effect on the size of the internal phase.
lium were not affected. In order to study a possible side
effect caused by systemic absorption, intravenous
injections were administered to mice; these did not
2. Significance of microemulsions in ocular drug delivery result in any signs of toxicity. Therefore, it was
devices concluded that the positively charged microemulsions
are appropriate for the ophthalmic route. However, no
2.1. Physico-chemical parameters of microemulsions article has been published to date, which demonstrates a
intended for ophthalmic route beneficial effect of this charge.
Increase in the lipophilic character of the drug by
2.1.1. Improvement of solubilization association with a counter-ion Gallarate et al. (1988,
The main advantage of the microemulsions is the 1993) studied the beneficial effect of a microemulsion
increase in the solubilization of drugs. Siebenbrodt and when two beta-blockers (timolol and levobunolol) are
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 23

Fig. 7. Histogram of the maximum percentage of ( ) indomethacin, (’) chloramphenicol and (&) sodium diclofenac in each component of
microemulsion.

Fig. 8. Solubility of piroxicam (mg/ml) in various solvents used in the preparation of microemulsions.

associated with a counter-ion in order to improve the lecithin. The formulation, which is prepared without
lipophilic character of the drugs. lecithin and presented in Table 1, is mixed until
The influence of the counter-ion (octanoic acid) on equilibrium is obtained. Then, the preparation is
the lipophilia of the drug (timolol) should be determined centrifuged in order to separate the continuous phase
based on the partition coefficient of the drug in the in which timolol was quantified. Five ratios of octanoic
external and internal phases of the microemulsion. As acid and isopropylmyristate are used for the internal
the microemulsion is a stable thermodynamic system, phase and the apparent partition coefficients are
it is impossible to determine the quantity of drug included in Table 5.
in the aqueous phase. Therefore, it is necessary The linear relationship between the two parameters
to remove a surfactantFin this case the egg yolk (R2 ¼ 0:980) demonstrates the influence of the quantity
24 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Fig. 9. Solubility of indomethacin (mg/ml) in various solvents used in the preparation of microemulsions.

Table 5 verify if its association with a counterion increases the

Apparent partition coefficient oil/buffer according to the percentage of
lipophilia of the drug. Gallarate et al. (1993) created five
octanoic acid in the oily phase (Gasco et al., 1988)
microemulsions based on various constituents, each of
Composition of the oily phase Apparent partition them containing levobunolol (Table 6). The relationship
octanoic acid: isopropylmyristate (% w/w) coefficient oil/buffer between the quantity of octanoic acid and the apparent
0 : 100 0.57 oil/water partition coefficient (without lecithin) is less
10.6 : 89.4 15.6 linear than in the previous case (R2 ¼ 0:747). This fact
15.9 : 84.1 23.6 can be attributed to the different compositions of the
21.1 : 78.9 25.3
preparations, in particular for microemulsion no. 5 in
28.4 : 71.6 35.5
Table 6, in which the pair of ions was distributed
differently in a water+polyethylene glycol (PEG) 200
mixture. Nevertheless, its influence on lipophilia is
of octanoic acid on the apparent oil/buffer partition obvious because the apparent partition coefficient
coefficient (Fig. 10). Therefore, the combination of ions between ethyloleate and water without octanoic acid at
in pairs demonstrates the increased lipophilia of the a pH of 6.75 is only 0.6.
drug. These experiments led to the conclusion that the
Levobunolol has an octanol/water partition coeffi- association in the pair of ions of the timolol and
cient higher than timolol. Therefore, it is important to levobunolol with the octanoic acid increases their

Fig. 10. Linear regression between the apparent oil/buffer partition coefficient and the quantity of counter-ions in the internal phase.
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 25

Table 6
Compositions of five eye drops microemulsions containing levobunolol and octanoic acid (Gallarate et al., 1993)

Component Function No. 1 No. 2 No. 3 No. 4 No. 5a

Levobunolol Drug 0.5 0.5 0.5 0.5 0.5

Lecithin from soybean Surfactant 5.7 9.2 8.4 8.7 6.2
Sodium octyl phosphate Surfactant 3.9 3.6 F F F
Sodium cholate Surfactant F F 3.5 6.6 6.7
Water External phase 70.8 65.7 69.2 71.8 62.0
Sodium hexanoate Co-surfactant 6.6 F 6.3 2.8 F
Monobutylglycerol Co-surfactant F 9.3 F F F
Capric/caprylic triglyceride Internal phase 3.2 3.0 3.0 3.2 F
Isopropylmyristate+octanoic acid (10% w/w) Internal phase+counter-ion 9.3 F F F F
Ethyloleate+octanoic acid (10% w/w) Internal phase+counter-ion F 8.7 9.1 9.4 8.1
PEG 200 Co-surfactant F F F F 19.5
PH 6.5 6.5 6.3 6.4 6.75
Apparent partition coefficient oil/water 2.5 2.2 2.0 2.0 1.1
a
Only this formulation is tolerated by the eye of rabbit.

lipophilia. In addition, these experiments led us to the 1% hydroxyethylcellulose and with the same viscosity as
following conclusions: the microemulsions (about 20 mPa.s at 251C), and from
aqueous solutions containing 15% Synperonic-L64s
* The pair of ions located in the internal phase provides (Figs. 11–13).
a delayed release by a reservoir effect. The concen- The studies of diffusion kinetics are conducted by
tration of the drug in the internal phase can be means of an artificial membrane. The single-layer
determined by modifying the concentration of the hydrophilic artificial membrane is placed between the
counter-ion and the partition coefficient of the drug; donor and the receiver compartments. The diffusion and
* the pair of ions located in the external phase should kinetic release decreased for each drug when they were
easily penetrate the lipophilic epithelium and thus formulated as microemulsions as opposed to the
increase the corneal bioavailability of the drug. formulations of aqueous solutions:
* 6.2 times less indomethacin was released from the
2.2. Results in vitro: use of diffusion cells
microemulsions after 2 h;
* 4.2 times less chloramphenicol was released from the
2.2.1. Study of diffusion kinetics through an artificial
microemulsions after 1 h;
membrane * 5.4 times less sodium diclofenac was released from
Siebenbrodt and Keipert (1993) compared the kinetic
the microemulsions after 6 h.
release of chloramphenicol, indomethacin and sodium
diclofenac which were obtained from aqueous solutions, The results obtained with aqueous solutions which
microemulsions (Table 2), aqueous solutions containing have a similar viscosity with microemulsions were

Fig. 11. Diffusion profiles of indomethacin (0.1%) incorporated in aqueous solutions (&) and microemulsion (’). The aqueous solution contained
only 0.1% of the drug due to the low solubility of indomethacin in water.
26 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Fig. 12. Diffusion profiles of chloramphenicol (0.5%) incorporated in aqueous solutions (&) and microemulsions (’).

Fig. 13. Diffusion profiles of sodium diclofenac (0.5%) incorporated in aqueous solutions (&), aqueous solutions containing 1.0% of
hydroxyethylcellulose (n), microemulsions (’).

reported only for sodium diclofenac (4.5 times less drug The microemulsion provides a delayed diffusion. This
released from the microemulsion after 6 h) (Fig. 13). effect, which was observed in vitro, is not the result of an
Nevertheless, the authors specified that, although the increase in viscosity. However, the composition and,
kinetic releases of the three drugs was significantly more precisely, the presence of the poloxamer can
slower when they were incorporated in aqueous solu- influence the kinetic release. Nevertheless, these conclu-
tions containing hydroxypropylcellulose compared to sions must be drawn with precaution due to the
aqueous solutions, the decrease in the kinetic release was limitation of the diffusion cell method.
insignificant compared to the decrease in the micro- Hasse and Keipert (1997) compared the release
emulsions. kinetics of the pilocarpine hydrocloride from an aqu-
As for the solutions containing Synperonic-L64s, eous solution and from microemulsions (Fig. 14). The
the authors confirmed that the kinetic release of the experimental conditions were similar to those developed
drugs was not significantly different compared to the by Siebenbrodt and Keipert (1993) namely the use of a
kinetic release in the microemulsions. The quantities hydrophilic artificial membrane. Two microemulsions
that were released after 6 h for the microemulsion and were prepared. Egg yolk lecithin was incorporated in the
the solution containing 15% Synperonic-L64s are the first, and PEG 1500 and 200 were used (Table 7) in the
following: second.
The diffusion kinetics of pilocarpine hydrochloride
* sodium diclofenac: 1.17 and 1.24 mg from the aqueous solution and microemulsion no. 2 did
* indomethacin: 0.22 and 0.32 mg not show any difference between the two formulations.
* chloramphenicol: 1.50 and 1.45 mg However, the diffusion kinetics of the pilocarpine
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 27

Fig. 14. Diffusion kinetics of pilocarpine hydrochloride from aqueous solutions (’), microemulsion 1 (m) and microemulsion 2 (n).

Table 7 two compositions did not have the same pH (3.8 for the
Compositions of two microemulsions prepared by Hasse and Keipert microemulsion and 6.8 for the aqueous eye drops), it is
(1997) difficult to reach conclusions. The pH of the Indocollyres
Component Function No. 1 No. 2 is high due to the large proportion of ionized
indomethacin at this pH (pKa=4.5) and its hydrosolu-
Pilocarpine hydrochloride Drug 2.0% 2.0%
Egg yolk lecithin Surfactant 1.7% F bility. Besides, the microemulsion was not formulated at
PEG 1500 Surfactant 13.1% 19.6% pH 3.8 for reasons of instability.
PEG 200 Surfactant 9.8% 4.9%
Isopropylmyristate Internal phase 6.6% 4.9%
2.2.3. Comments on diffusion cells and artificial
Water External phase QSP 100 QSP 100
membranes
Although the method of diffusion cells is often used, it
is not representative of the real situation in vivo. First,
hydrochloride from the microemulsion containing le- the artificial membranes are not always multilayered.
cithin was significantly different. For microemulsion no. The hydrophilic membranes (Nephrophans), which are
1, 25% less pilocarpine hydrochloride was diffused usually used in dialysis, are made of cellulose deriva-
through the artificial hydrophilic membrane after 6 h tives. Due to the nature of the material of these
than for the same concentration of drug from the membranes, the charge, which depends on the degree
microemulsion no. 2 and from aqueous solution. of esterification, must be determined. Second, the
Nevertheless, although the microemulsion containing volume in the donor compartment remains constant
lecithin seemed to prolong the diffusion of the drug in whereas the drainage physiologically eliminates part of
vitro, the results were less pronounced than these the drug present in the aqueous phase and can dilute the
observed with the sodium diclofenac. drug in tears. This phenomenon affects the concentra-
tion gradient and the diffusion of the drug through the
2.2.2. Studies of diffusion kinetics of the drugs through an cornea (Fig. 15).
excised cornea of rabbit The concentration gradient is different whether an
Muchtar et al. (1997) studied the diffusion of drugs aqueous solution or a microemulsion is studied. Trotta
through a rabbit cornea used as a membrane between et al. (1989) have evaluated the K permeability
the two compartments. The endothelium was placed at coefficients of several drugs formulated in an aqueous
the receiver. The experiments were conducted using a solution and in a microemulsion through a hydrophilic
microemulsion (Table 3) and an aqueous solution membrane according to the following formula:
(Indocollyres) containing indomethacin.
C2
C1 V
The coefficients of permeability were 6.9  10
5 cm/s K¼  ;
t2
t1 C0 A
for the indomethacin incorporated in a microemulsion
and 1.83  10
5 cm/s for the conventional aqueous eye where C0 is the initial concentration in the donor
drops containing the same drug. Nevertheless, as the compartment, C1 and C2 are the concentrations at time
28 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Donor compartment Receveir compartment Table 8

Permeability coefficients of drugs in aqueous solutions and in
1) diffusion cell microemulsions through an artificial hydrophilic membrane (Trotta
Ca Cb et al., 1989)

Drugs Aqueous solution Microemulsion

K K K0

Nitrofurazone 3.2 1.80 2.5

Phenylbutazone 1.4 0.70 1.5
Prednisone 1.9 0.56 1.8
2) in vivo Betamethasone 2.5 0.18 2.3
Menadione 3.6 0.05 3.8

Eliminated
by drainage the aqueous phase is eliminated by drainage. For
conventional eye drops, the fraction of the drug
eliminated by drainage is more important because only
Ca' Cb' the aqueous phase is present. Therefore, the concentra-
tion gradient is lower for a drug dissolved into an
Ca' - Cb' < Ca - Cb
aqueous solution, which leads to a reduced diffusion of
Fig. 15. Diffusion of a drug through an artificial hydrophilic drugs through the cornea. The permeability coefficients
membrane in vivo according to the concentration gradient. determined in vitro using a diffusion cell for conven-
tional aqueous eye drops and microemulsions are
amplified in vivo. That is why, the small difference
t1 and t2 in the receiver compartment, V is the volume of between the diffusion of a drug from an aqueous
the receiver compartment, A is the membrane area. solution and microemulsion no. 2 shown in Fig. 14 is
The permeability coefficients of the drugs formulated not representative of its performance in vivo. The same
as microemulsions are calculated based on the over- applies to eye drops in an aqueous phase containing 1%
all concentration of drugs in the external and hydroxyethylcellulose (Fig. 13). In this case, the diffu-
internal phase. Trotta et al. (1989) used the determina- sion of the pilocarpine hydrochloride is not representa-
tion of the apparent partition coefficient of the tive of the real diffusion of the drug due to the drainage
drugs incorporated in microemulsion to determine the decrease in vivo.
quantity of drug dissolved in the external aqueous
phase. These authors calculated the K 0 permeability 2.3. Results of in vivo trials
coefficients using the quantity of drug dissolved in
the aqueous phase of the microemulsion as initial Muchtar et al., 1992, have evaluated the activity of
concentration. delta-8-tetrahydrocannabinol that was incorporated in a
The permeability coefficients specified in Table 8 microemulsion dosage form on usual rabbits and on
can be explained by the mechanisms, which govern rabbits suffering from high blood pressure. The compo-
the release of drugs from each formulation. Because sition of the microemulsion is presented in Table 9.
only the drug dissolved in the external phase is The size of the droplets was 130741 nm and they had
able to penetrate the hydrophilic membrane, it is a
59 mV Zeta potential. These parameters were stable
obvious that the K permeability coefficients of the after the microemulsions were stored at 41C for 9
drugs dissolved in the aqueous solutions and K 0 months. The size of the droplets was 146747 nm and
coefficients of the drugs dissolved in the external the Zeta potential was
57.1 mV. Furthermore, the pH
aqueous phase of microemulsions are similar. As for remained stable and the microemulsion did not result in
the fraction of the drug dissolved in the internal oily any irritating reaction. The microemulsion dosage form
phase, it must be first transferred to the external phase provided a sustained reduction in intraocular pressure in
before being diffused in the receiver compartment. The the two groups of rabbits with a higher effect on the
K permeability coefficients are lower than the K 0 rabbits suffering from high blood pressure (Fig. 16).
coefficients due to the necessity of a transfer before the Although the number of trials on animals (six rabbits
diffusion. for each group) was low, the reproducibility of the
These experiments demonstrated that the two for- results was good. Nevertheless, the microemulsion
mulations would have a different behaviour in vivo. As a dosage form in this study was not compared with
result of the quick adsorption of the internal phase another dosage form; therefore, it did not provide
nanodroplets of the microemulsion on the corneal enough evidence regarding the benefits of this carrier
membrane, only the fraction of the drug dissolved in compared to others.
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 29

Fig. 16. Effect on intraocular pressure in healthy (&) rabbits and in rabbits suffering from high blood pressure (’) after instillation of 50 ml of
microemulsion (x: po0:05 and xx: po0:01) containing delta-8-tetrahydrocannabinol.

Table 9 Table 10
Composition of a microemulsion containing delta-8-tetrahydrocanna- Composition of a microemulsion dosage form containing pilocarpine
binol (d8THC) (Muchtar et al., 1992) (Naveh et al., 1994)

Components Function Amount (% w/w) Components Function Amount

8
(% w/w)
d -THC Drug 0.40
Soya oil Internal phase 14.00 Pilocarpine base Drug 1.70
Oleic acid Internal phase 6.00 Medium chain triglycerides (MCT) Internal phase 5.00
Egg phospholipids Surfactant 1.00 Lipo.ıd E-80s Surfactant 0.75
Pluronic F-68s Co-surfactant 2.00 Miranol MHT Surfactant 1.50
Glycerol Osmotic agent 2.25 a-Tocopherol Antioxydizing agent 0.02
a-Tocopherol Antioxydizing agent 0.02 Glycerol Co-surfactant 2.25
Distilled water External phase QSP 100 Distilled water External phase QSP 100

Naveh et al. (1994) conducted a randomized double- However, the intraocular pressure kept going down until
blind trial in on three groups, each consisting of twelve the 29th hour. Furthermore, the maximum reduction
healthy rabbits, and determined the effects after a single recorded for the intraocular pressure was only 18% for
administration of a microemulsion dosage form contain- regular eye drops and 28.5% for microemulsion eye
ing 1.7% pilocarpine base as opposed to an aqueous drops. The microemulsion dosage form provided both a
solution containing 2% pilocarpine hydrochloride delayed and delayed pharmacological action compared
(equivalent to 1.7% pilocarpine base). The composition to the pharmacological action of regular eye drops. This
of the microemulsion is reported in Table 10. The results observation led to the conclusion that the microemul-
in Fig. 17 are expressed in percentage of reduction of sion eye drops have a real advantage compared to
intraocular pressure. They were calculated based on a regular eye drops which must be administered four times
baseline, which provides the values of the intraocular a day due to the short duration of the pharmacological
pressure according to the time before the treatment. This action.
protocol obviated the mistakes due to fluctuations in The low-pressure ultrafiltration technique enabled the
intraocular pressure observed on the day of the determination of the ratio of the existing drug in the
experiment. The effects that were obtained using the internal phase. The experiment showed that 5.9% of the
two dosage forms were different. The eye drops modified pilocarpine was dissolved in the nanodroplets constitut-
the intraocular pressure earlier (2 h after instillation), ing the internal phase and 14.9% was located at the
but the duration of its pharmacological action was interface. Only 20.8% of the exiting drug in the internal
shorter (5 h). Eleven hours after the instillation, the phase was required to prolong the pharmacological
control values were the same as the values observed on action of the dosage form. Nevertheless, during the
the treated rabbits. The instillation of the microemulsion experiment, a decrease in the intraocular pressure has
demonstrated a pharmacological action only 5 h later. also been observed in the contralateral eye, which
30 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Fig. 17. Effect after a single administration of aqueous solution (&) containing 2% of pilocarpine hydrochloride and a microemulsion (’)
containing 1.7% of pilocarpine base (po0:05).

reveals a systemic effect. According to Naveh et al., Table 11

Composition of a microemulsion dosage form containing indometha-
1994, it appeared that the retention of 20.8% of the total
cin (Aviv et al., 1993)
pilocarpine content in the internal oil phase, and the oil–
water interface of the emulsion are sufficient to Components Function Amount
concomitantly enhance the ocular absorption of the (% w/w)
drug through the cornea, and also increasing the corneal Indomethacin Drug 0.20
concentration of pilocarpine. From these authors, this Medium chain triglycerides (MCT) Internal phase 17.00
combined effect would account for both the prolonged Lipo.ıd E-80s Surfactant 3.00
Tween 80s Surfactant 1.00
hypotensive effect and might lead to systemic absorption
a-Tocopherol Antioxydizing agent 0.02
and contralateral effect. Methyl parahydroxybenzoate Preservative agent 0.10
In order to verify the extension of the residence time Propyl parahydroxybenzoate Preservative agent 0.02
of the microemulsion in the conjunctival sac, Beilin et al., Glycerol Co-surfactant 2.25
1995, added a fluorescent marker to the formulations. EDTA Chelating agent 0.10
Distilled water External phase QSP 100
After the administration of the eye drops after 1, 2, 3, 5,
10 and 15 min, 200 ml of saline solution was used to rinse
the eye. Afterwards the intensity of the fluorescence was
Table 12
determined. The intensity was higher for the micro- Concentrations (mM) of indomethacin in the anterior chamber after
emulsion dosage form. One minute after the adminis- instillation of a solution and a microemulsion (Aviv et al., 1993)
tration, 39.9710.2% of the fluorescence was measured
Time (h) Solution 1% Microemulsion 0.2%
for the microemulsion as compared to only 6.871.8%
a
for regular eye drops. Furthermore, the areas under the 0.5 2.271.7 (7) 1.170.1 (7)
curves of the pupillary diameters according to time were 1 1.070.5 (6) 0.970.3 (7)
3 0.670.3 (6) 0.370.1 (4)
7.570.7 mm/h for the microemulsion dosage form and
a
5.770.6 mm/h for regular eye drops, the difference Mean7standard deviation (number of eyes).
between the two values being significant (po0:05). The
microemulsion eye drops provided a much more delayed
pharmacological action due to a delayed residence time indomethacin was twice bigger than for the regular
in the conjunctival sac following a more significant aqueous solution than for the microemulsion. The area
contact time with the cornea. under the curve (AUC) was 2.2 times bigger for the
Aviv et al., 1993, determined the concentrations of microemulsion, which means an increase of the bioa-
indomethacin in the anterior chamber of rabbits after vailability for the colloidal dosage form.
the administration of 50 ml of aqueous solution contain- After comparing the diffusion profiles of three
ing 1% of the drug or microemulsion eye drops compositions of pilocarpine in vitro (Table 7, Fig. 14),
containing 0.2% of the same drug. The composition of Hasse and Keipert (1997) studied their pharmacological
the tested microemulsion is specified in Tables 11 and effect in vivo by using six rabbits for each group
12. After half an hour, the determined concentration of (Fig. 18). The obtained results were different from those
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 31

Fig. 18. Miotic activity determined after instillation of 50 ml of aqueous solution (’), microemulsion no. 1 (m) and microemulsion no. 2 (n) each one
containing 2.0% of pilocarpine hydrochloride.

Table 13
Concentrations of timolol and volumes of instilled dosage forms (Gasco et al., 1989)

Dosage forms Concentration of timolol (% w/v) Total instilled volume (ml)

Microemulsion 2.6 150

Aqueous solution containing some timolol 1.4 150
Aqueous solution containing some timolol associated to octanoic acid 0.33 300

observed in vitro (Fig. 14). The two microemulsions a small quantity of the drug in this form was dissolved in
provided a delayed release compared to the release of aqueous solution (0.33%) compared to the regular
the drug incorporated in the aqueous solution. timolol dissolved in aqueous solution (1.4%). However,
The calculation of the AUC provided a comparison the solubilization in the internal phase of the micro-
between the bioavailability of various dosage forms: the emulsion was higher (2.6%).
areas under the curve are 56% and 68% bigger for the These results led to the conclusion that it is preferable
microemulsions no. 1 and no. 2 than for regular aqueous to instil the microemulsion dosage form because a low
solution. volume of a more concentrated solution can be instilled.
After determining the percentage of counter-ion The experiment was conducted on three groups of eight
which will replace the isopropylmyristate in the oily rabbits and the quantity of timolol was measured out in
phase in order to obtain a better apparent oil/buffer the aqueous humor at regular intervals. The results
partition coefficient for timolol (28.4% octanoic acid/ specified in Fig. 19 were calculated so that they match
71.6% isopropylmyristate), Gasco et al. (1989) com- the quantities of instilled drug.
pared the pharmacological action of the microemulsion The influence of the pair of ions is obvious since the
(Table 1) to the effects observed after the administration areas under the curves of the microemulsion and
of an aqueous solution containing timolol or timolol aqueous solution containing the pair of ions were 3.5
combined with octanoic acid (timolol/octanoate molar and 4.2 times bigger than for the aqueous solution
ratio 1 : 10) in vivo on rabbits. However, the drug containing timolol.
concentrations and instilled volumes for the three Although the bioavailability of the microemulsion
dosage forms was different (Table 13). The volumes does not reach the maximum, its pharmacokinetic
were not instilled all at once, which would have resulted profile is different from the aqueous solutions. The
in the loss of a large part of the drug by drainage. Fifty absorption of the microemulsion takes longer and
microliters were placed in the conjunctival sac of each timolol is still determined 2 h after the instillation.
eye at every 4 min. The concentrations of timolol in each Thus, the existence of the pair of ions in the external
dosage were different based on the solubility of the phase increased the absorption of timolol whereas the
timolol in the phase in which it was dissolved. As the nanodroplets of the internal phase, which remained on
pair of ions is more lipophilic than the drug alone, only the cornea, act as a reservoir of the drug.
32 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Fig. 19. Determination of timolol concentrations in aqueous humor according to time after numerous instillations of microemulsion containing
timolol and octanoic acid (&), aqueous solution containing timolol and octanoic acid (m) and aqueous solution containing timolol (n).

Fig. 20. Pupillary diameters after instillation of 50 ml of microemulsion containing pilocarpine base (’) and microemulsion without drug (&).

2.4. Results on man Table 14

Composition of eye drops microemulsion dosage form containing
pilocarpine (Aviv et al., 1993)
No experimental study has been conducted with
microemulsions prepared by autoemulsification. How- Components Function Amount
ever, several trials were conducted with microemulsions (% w/w)
prepared by supply of energy. Anselem et al. (1993), Pilocarpine base Drug 1.70
and Melamed et al. (1994) prepared microemulsions Medium chain triglycerides (MCT) Internal phase 4.25
containing adaprolol maleate. The size of the droplets of Lipo.ıd E-75s Surfactant 0.75
Tyloxapol Surfactant 1.00
the internal phase was 100730 nm. The physico- a-Tocopherol Antioxidizing agent 0.02
chemical parameters at room temperature and at 451C EDTA Chelating agent 0.10
have not changed for more than 6 months. According Chlorbutanol Preservative agent 0.20
to these authors, no ocular irritation was noticed in Thiomersal Preservative agent 0.01
the group of forty healthy volunteers as opposed to Glycerol Co-surfactant 2.25
Distilled water External phase QSP 100
regular eye droplets. The depressor effect was delayed;
the intra-ocular pressure was still high 6 and 12 h
after the instillation of the microemulsion and contained
0.2 or 0.4% of the drug as compared with the control Also Aviv et al. (1993) conducted human studies with
group which received the microemulsion without any a microemulsion the composition of which is specified in
drug. Table 14.
 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34 33

Fig. 21. Variations of intraocular pressure (treated eye (&) and contralateral (’)) after instillation of microemulsion containing pilocarpine base.

A single instillation of microemulsion or correspond- requires several daily administrations. The eye is not an
ing placebo, namely microemulsion without any drug, easy-accessible organ, mainly because of the lachrymal
was administered to twenty healthy volunteers. The drainage and corneal barrier.
determined parameters were the pupillary diameter and That is why, microemulsions are an interesting
variation of intra-ocular pressure (Fig. 20). The effect of alternative. In point of view of production and
the microemulsion which contains pilocarpine is obvious sterilization, microemulsions are simple and inexpen-
as compared to the placebo and was noticed within 1 h sive; they are stable and provide by their structure the
from instillation. The return to the initial values was solubilization of lipophilic drugs, which depends on the
noticed within 12 h (Fig. 21). This experimental trial also choice of the oily phase. The choice of surfactants
showed that a pressure variation follows a systemic mainly depends on the lack of toxicity. That is why, the
absorption in the contralateral eye. most frequently used surfactants are the non-ionic
Another experimental trial (Garty and Lusky, 1994), poloxamers or amphoteric lecithin. Moreover, in order
which used pilocarpine in ocular drug deliveries, was to avoid the risk of ocular intolerance, it is preferable to
conducted in order to compare the activity of the micro- reduce the quantity of surfactants under 5%. In this
emulsion instilled twice a day with a generic dosage form case, a two-step process based on supply of energy might
instilled four times a day to forty hypertensive patients. be required, namely dispersion and homogenization, in
All treatment was stopped two to three weeks before the order to obtain the desired size of particles (100–
trial. For seven days, the patients randomly received 150 nm). The use of higher quantities of surfactants
treatment with microemulsions or regular eye drops. No and co-surfactants (40–50%) leads to a spontaneous
local side effect has been reported. The intraocular formation of the microemulsion and a reduction of the
pressure decreased with 25% in both groups during all size (50 nm) of the droplets constituting the internal
this period. No significant difference has been noticed phase.
between the two treatments. These results proved that It is difficult to reach a conclusion from the results in
the microemulsion extended the action of the drug and vitro of the diffusion cell experiments. However, the in
two daily administrations have the same result as four vivo studies on rabbits showed a delayed effect of the
instillations of regular eye drops. drugs incorporated in microemulsions, which improves
their bioavailability. The proposed mechanism of the
delayed action is based on the adsorption of the
3. Conclusion and future challenges nanodroplets on the cornea which are not eliminated
by the lachrymal drainage and act as reservoirs of the
Although microemulsions are known since 1928, their drugs. The first studies conducted on healthy volunteers
potential as drug carriers for the ocular route has been are promising and confirmed the results obtained on
investigated by research teams only within the last rabbits.
decade. Some pilocarpine-based microemulsions were able to
In ophthalmology, the regular dosage form, namely delay the activity of the drug; two daily instillations of
eye drops, has several disadvantages: the low bioavail- microemulsion are equivalent to four instillations of
ability of the drugs that are absorbed systemically conventional drops.
34 Th.F. Vandamme / Progress in Retinal and Eye Research 21 (2002) 15–34

Nevertheless, complementary studies are necessary. In Garty, N., Lusky, M., 1994. Pilocarpine in submicron emulsion
particular, the quantity of drug in nanodroplets formulation for treatment of ocular hypertension: a phase II
generates a reservoir effect. The choice of the oily phase clinical trial. Invest. Ophthalmol. Vis. Sci. 35 (4), 2175.
Gasco, M.R., Gallarate, M., Trotta, M., Bauchiero, L., Gremmo, E.,
is also important and must take into consideration the Chiappero, O., 1989. Microemulsions as topical delivery vehicles:
existence range of microemulsions. The formation of ocular administration of timolol. J. Pharm. Biochem. Anal. 7 (4),
pairs of ions also improved this characteristic. On one 433–439.
hand, the future challenges will consist in the prepara- Hasse, A., Keipert, S., 1997. Development and characterization of
tion of microemulsions able to deliver the drug microemulsions for ocular application. Eur. J. Pharm. Biopharm.
43, 179–183.
following a zero kinetic order. On the other hand, J.arvinen, K., J.arvinen, T., Urtti, A., 1995. Ocular absorption following
comparisons between microemulsions containing nega- topical delivery. Adv. Drug Deliv. Rev. 16, 3–19.
tively and positively charged nanodroplets would have Klang, S.H., Baszkin, A., Benita, S., 1996. The stability of piroxicam
to be developed in order to determine if the affinity incorporated in a positively charged submicron emulsion for ocular
administration. Int. J. Pharm. 132, 33–44.
between cornea and microemulsion can be improved
Klang, S.H., Frucht-Pery, J., Hoffman, A., Benita, S., 1994.
and if an increase of the corneal residence time is Physicochemical characterization and acute toxicity evaluation of
possible. From a methodological point of view, the a positively charged submicron emulsion vehicle. J. Pharm.
challenges in this field will have to focus on the in vitro Pharmacol. 46, 986–993.
models with extrapolation of results to in vivo results. Langevin, D., Meunier, J., Cazabat, A.-M., 1985. Les microemulsions.
La recherche 167 (16), 720–728.
Lawrence, J., 1994. Surfactant systems: microemulsions and vesicles as
vehicles for drug delivery. Eur. J. Drug Metab. Pharmacokinet. 3,
References 257–269.
Melamed, S., Kurtz, S., Greenbaum, A., Haves, J.F., Neumann, R.,
Aiache, J.M., Devissaguet, J.Ph., 1983. Les microemulsions. In: Garty, N., 1994. Adaprolol maleate in submicron emulsion, a novel
Galenica 5 Agents de surface et e! mulsions. Lavoisier, pp. 313–320. soft b-blocking agent, is safe and effective in human studies. Invest.
Anselem, S., Beilin, M., Garty, N., 1993. Submicron emulsion as Ophthalmol. Vis. Sci. 35 (4), 1387.
ocular delivery system for adaprolol maleate, a soft b-blocker. Muchtar, S., Abdulrazik, M., Benita, S., 1997. Ex-vivo permeation
Pharm. Res. 10 (suppl.), S205. study of indomethacin from a submicron emulsion through albino
Attwood, D., 1994. Microemulsions. In: Collo.ıdal Drug Delivery rabbit cornea. J. Controlled Release 44 (1), 55–64.
Systems. Marcel Dekker, New York, pp. 66, 31–71. Muchtar, S., Almong, S., Torraca, M.T., Saettone, M.F., Benita, S.,
Aviv, H., Friedman, D., Bar-Ilan, A., Vered, M., 1993. Submicron 1992. A submicron emulsion as ocular vehicle for delta-8-
emulsions as ocular drug delivery vehicles. PCT WO 94/05298. tetrahydrocannabinol: effect on intraocular pressure in rabbits.
Beilin, M., Bar-Ilan, A., Amselem, S., 1995. Ocular retention time of Ophtalmic Res. 24, 142–149.
submicron emulsion (SME) and the miotic response to pilocarpine Naveh, N., Muchtar, S., Benita, S., 1994. Pilocarpine incorporated
delivered in SME. Invest. Ophthalmol. Vis. Sci. 36 (4), S166. into a submicron emulsion vehicle causes an unexpectedly
Benita, S., Levy, M.Y., 1993. Submicron emulsions as colloidal drug prolonged ocular hypotensive effect in rabbits. J. Ocul. Pharmacol.
carriers for intravenous administration: comprehensive physico- 10 (3), 509–520.
chemical characterization. J. Pharm. Sci. 82 (11), 1069. Pore, J. (1992) Les microemulsions. In: Emulsions, micro!emulsions,
Benita, S., Muchtar, S., 1992. Ophthalmic compositions. Patent NoEP e! mulsions multiples. Techniques des corps gras, pp. 91–143.
0521 799 A1. Ruth, H., Attwood, D., Ktistis, G., Taylor, C.J., 1995. Phase studies
Bothorel, P., 1985. Les microemulsions. L’actualit!e chimique 1, 23–27. and particle size analysis of oil-in-water phospholipid microemul-
Eccleston, G.M., 1988. Microemulsions. In: Encyclopedia of Pharma- sions. Int. J. Pharm. 116, 253–261.
ceutical Technology, Vol 9, pp. 375–421. Sharma, M.K., Shah, D.O. (1985) Introduction to macro and
Elbaz, E., Zeevi, A., Klang, S., Benita, S., 1993. Positively charged microemulsions, theory and applications. ACS Symposium Series,
submicron emulsions, a new type of colloidal drug carrier. Int. J. pp. 1–18.
Pharm. 96, R1–R6. Siebenbrodt, I., Keipert, S., 1993. Poloxamer-systems as potential
Evitts, D.P., Olejnik, O., Musson, D.G., Bilgood, A.M., 1991. ophtalmic microemulsions. Eur. J. Pharm. Biopharm. 39 (1),
Aqueous ophtalmic microemulsions of tepoxalin. European patent 25–30.
application 0 480 690 A1. Trotta, M., Gasco, M.R., Morel, S., 1989. Release of drugs from oil-
Gallarate, M., Gasco, M.R., Trotta, M., 1988. Influence of octanoic water microemulsions. J. Controlled Release 10, 237–243.
acid on membrane permeability of timolol from solutions and from Wade, A., Weller, P.J., (Eds.), 1994. Handbook of Pharmaceutical
microemulsions. Acta Pharm. Technol. 34 (2), 102–105. Excipients, 2d Edition. The pharmaceutical press.
Gallarate, M., Gasco, M.R., Trotta, M., Chetoni, P., Saettone, M.F., Zeevi, A., Klang, S., Benita, S., 1994. The design and characterization
1993. Preparation and evaluation in vitro of solutions and o/w of a positively charged submicron emulsion containing a sunscreen
microemulsions containing levobunolol as ion-pair. Int. J. Pharm. agent. Int. J. Pharm. 108, 57–68.
100, 219–225.
 3. Microemulsions

This chapter is devoted to another important property of surfactants,

that of stabilization of water-oil films and formation of microemulsions. These
are a special kind of colloidal dispersion that have attracted a great deal of
attention because of their ability to solubilise otherwise insoluble materials.
Industrial applications of microemulsions have escalated in the last 40 years
following an increased understanding of formation, stability and the role of
surfactant molecular architecture. This chapter reviews main theoretical
features relevant to the present work and some common techniques used to
characterize microemulsion phases.

3.1 MICROEMULSIONS: DEFINITION AND HISTORY

One of the best definitions of microemulsions is from Danielsson and

Lindman [1] “a microemulsion is a system of water, oil and an amphiphile
which is a single optically isotropic and thermodynamically stable liquid
solution”. In some respects, microemulsions can be considered as small-scale
versions of emulsions, i.e., droplet type dispersions either of oil-in-water (o/w)
or of water-in-oil (w/o), with a size range in the order of 5−50 nm in drop
radius. Such a description, however, lacks precision since there are significant
differences between microemulsions and ordinary emulsions (or
macroemulsions). In particular, in emulsions the average drop size grows
continuously with time so that phase separation ultimately occurs under
gravitational force, i.e., they are thermodynamically unstable and their
formation requires input of work. The drops of the dispersed phase are

59
generally large (> 0.1 µm) so that they often take on a milky, rather than a
translucent appearance. For microemulsions, once the conditions are right,
spontaneous formation occurs. As for simple aqueous systems, microemulsion
formation is dependent on surfactant type and structure. If the surfactant is
ionic and contains a single hydrocarbon chain (e.g., sodium dodecylsulphate,
SDS) microemulsions are only formed if a co-surfactant (e.g., a medium size
aliphatic alcohol) and/or electrolyte (e.g., 0.2 M NaCl) are also present. With
double chain ionics (e.g., Aerosol-OT) and some non-ionic surfactants a co-
surfactant is not necessary. This results from one of the most fundamental
properties of microemulsions, that is, an ultra-low interfacial tension between
the oil and water phases, γo/w. The main role of the surfactant is to reduce γo/w
sufficiently − i.e., lowering the energy required to increase the surface area −
so that spontaneous dispersion of water or oil droplets occurs and the system is
thermodynamically stable. As described in Section 3.2.1 ultra-low tensions are
crucial for the formation of microemulsions and depend on system composition.

Microemulsions were not really recognized until the work of Hoar and
Schulman in 1943, who reported a spontaneous emulsion of water and oil on
addition of a strong surface-active agent [2]. The term “microemulsion” was
first used even later by Schulman et al. [3] in 1959 to describe a multiphase
system consisting of water, oil, surfactant and alcohol, which forms a
transparent solution. There has been much debate about the word
“microemulsion” to describe such systems [4]. Although not systematically used
today, some prefer the names “micellar emulsion” [5] or “swollen micelles” [6].
Microemulsions were probably discovered well before the studies of Schulmann:
Australian housewives have used since the beginning of last century
water/eucalyptus oil/soap flake/white spirit mixtures to wash wool, and the first
commercial microemulsions were probably the liquid waxes discovered by
Rodawald in 1928. Interest in microemulsions really stepped up in the late
1970's and early 1980's when it was recognized that such systems could
improve oil recovery and when oil prices reached levels where tertiary recovery
methods became profit earning [7]. Nowadays this is no longer the case, but

60
other microemulsion applications were discovered, e.g., catalysis, preparation
of submicron particles, solar energy conversion, liquid−liquid extraction
(mineral, proteins, etc.). Together with classical applications in detergency and
lubrication, the field remains sufficiently important to continue to attract a
number of scientists. From the fundamental research point of view, a great deal
of progress has been made in the last 20 years in understanding microemulsion
properties. In particular, interfacial film stability and microemulsion structures
can now be characterized in detail owing to the development of new and
powerful techniques such as small-angle neutron scattering (SANS, as
described in Chapter 4). The following sections deal with fundamental
microemulsion properties, i.e., formation and stability, surfactant films,
classification and phase behaviour.

3.2 THEORY OF FORMATION AND STABILITY

3.2.1 Interfacial tension in microemulsions

A simple picture for describing microemulsion formation is to consider a

subdivision of the dispersed phase into very small droplets. Then the
configurational entropy change, ∆Sconf, can be approximately expressed as [8]:

∆Sconf = −nk B [ln φ + {(1 − φ ) / φ }ln (1 − φ )] (3.2.1)

where n is the number of droplets of dispersed phase, kB is the Boltzmann

constant and φ is the dispersed phase volume fraction. The associated free
energy change can be expressed as a sum of the free energy for creating new
area of interface, ∆Aγ12, and configurational entropy in the form [9]:

∆G form = ∆Aγ12 − T∆Sconf (3.2.2)

61
where ∆A is the change in interfacial area A (equal to 4πr2 per droplet of radius
r) and γ12 is the interfacial tension between phases 1 and 2 (e.g., oil and water)
at temperature T (Kelvin). Substituting Eq. 3.2.1 into 3.2.2 gives an expression
for obtaining the maximum interfacial tension between phases 1 and 2. On
dispersion, the droplet number increases and ∆Sconf is positive. If the surfactant
can reduce the interfacial tension to a sufficiently low value, the energy term in
Eq. 3.2.2 (∆Aγ12) will be relatively small and positive, thus allowing a negative
(and hence favourable) free energy change, that is, spontaneous
microemulsification.

In surfactant-free oil−water systems, γo/w is of the order of 50 mN m-1,

and during microemulsion formation the increase in interfacial area, ∆A, is very
large, typically a factor of 104 to 105. Therefore in the absence of surfactant,
the second term in Eq. 3.2.2 is of the order of 1000 kBT, and in order to fulfill
the condition ∆Aγ12 ≤ T∆Sconf, the interfacial tension should be very low
(approximately 0.01 mN m-1). Some surfactants (double chain ionics [10, 11]
and some non-ionics [12]) can produce extremely low interfacial tensions −
typically 10-2 to 10-4 mN m-1 − but in most cases, such low values cannot be
achieved by a single surfactant since the CMC is reached before a low value of
γo/w is attained. An effective way to further decrease γo/w is to include a second
surface-active species (either a surfactant or medium-chain alcohol), that is a
co-surfactant. This can be understood in terms of the Gibbs equation extended
to multicomponent systems [13]. It relates the interfacial tension to the
surfactant film composition and the chemical potential, µ, of each component in
the system, i.e.,

dγ o / w = −∑ (Γi dµi ) ≈ −∑ (Γi RTd ln Ci ) (3.2.3)

i i

where Ci is the molar concentration of component i in the mixture, and Γi the

surface excess (mol m-2). Assuming that surfactants and co-surfactants, with
concentration Cs and Cco respectively, are the only adsorbed components (i.e.,
Γwater = Γoil = 0), Eq. 3.2.3 becomes:

62
 dγ o / w = −Γs RTd ln Cs − Γco RTd ln Cco (3.2.4)

Integration of Eq. 3.2.4 gives:

Cs C co
γ o / w = γ oo / w − ∫ Γs RTd ln C s − ∫ Γco RTd ln C co (3.2.5)
0 0

Eq. 3.2.5 shows that γ oo / w is lowered by two terms, both from the surfactant

and co-surfactant (of surface excesses Γs and Γco respectively) so their effects
are additive. It should be mentioned, however, that the two molecules should
be adsorbed simultaneously and should not interact with each other (otherwise
they lower their respective activities), i.e., are of completely different chemical
nature, so that mixed micellisation does not occur.

Figure 3.1 shows typical low interfacial tensions found in microemulsions,

in this case spanning ~ 1 to 10-3 mN m-1. The effect of salt concentration is
consistent with changes in the phase behaviour, which are discussed in more
detail in Section 3.3 and Figure 3.2 below.

3.2.2 Kinetic instability

Internal contents of the microemulsion droplets are known to exchange,

typically on the millisecond time scale [16, 17]. They diffuse and undergo
collisions. If collisions are sufficiently violent, then the surfactant film may
rupture thereby facilitating droplet exchange, that is the droplets are kinetically
unstable. However, if one disperses emulsions sufficiently small droplets (< 500
Å), the tendency to coalesce will be counteracted by an energy barrier. Then
the system will remain dispersed and transparent for a long period of time
(months) [18]. Such an emulsion is said to be kinetically stable [19]. The
mechanism of droplet coalescence has been reported for AOT w/o
microemulsions [16]; the droplet exchange process was characterized by a
second order rate constant kex, which is believed to be activation controlled

63
(hence the activation energy, Ea, barrier to fusion) and not purely diffusion
controlled. Other studies [20] have shown that the dynamic aspects of
microemulsions are affected by the flexibility of the interfacial film, that is film
rigidity (see Section 3.3.2), through a significant contribution to the energy
barrier. Under the same experimental conditions, different microemulsion
systems can have different kex values [16]: for AOT w/o system at room
temperature, kex is in the range 106−109 dm3 mol-1 s-1, and for non-ionics CiEj,
108−109 dm3 mol-1 s-1 [16, 17, 20]. In any case, an equilibrium droplet shape
and size is always maintained and this can be studied by different techniques
[20].

Figure 3.1 Oil-water interfacial tension between n-heptane and aqueous NaCl
solutions as a function of salt concentration in the presence of AOT surfactant.
The values were determined by spinning drop tensiometry. The AOT surfactant
concentration is 0.050 mol dm-3, temperature 25°C.

0
log [ γo/w / (mN m-1)]

-1

-2

-3

0.02 0.04 0.06 0.08 0.10

[NaCl] / (mol dm-3)

64
3.3 PHYSICOCHEMICAL PROPERTIES

This section gives an overview of the main parameters characterizing

microemulsions. References will be made to related behaviour for planar
interfaces presented in Chapter 2.

3.3.1 Predicting microemulsion type

A well-known classification of microemulsions is that of Winsor [21] who

identified four general types of phase equilibria:

• Type I: the surfactant is preferentially soluble in water and oil-in-water

(o/w) microemulsions form (Winsor I). The surfactant-rich water phase
coexists with the oil phase where surfactant is only present as
monomers at small concentration.
• Type II: the surfactant is mainly in the oil phase and water-in-oil (w/o)
microemulsions form. The surfactant-rich oil phase coexists with the
surfactant-poor aqueous phase (Winsor II).
• Type III: a three-phase system where a surfactant-rich middle-phase
coexists with both excess water and oil surfactant-poor phases (Winsor
III or middle-phase microemulsion).
• Type IV: a single-phase (isotropic) micellar solution, that forms upon
addition of a sufficient quantity of amphiphile (surfactant plus alcohol).

Depending on surfactant type and sample environment, types I, II, III or

IV form preferentially, the dominant type being related to the molecular
arrangement at the interface (see below). As illustrated in Figure 3.2, phase
transitions are brought about by increasing either electrolyte concentration (in
the case of ionic surfactants) or temperature (for non-ionics). Table 3.1
summarizes the qualitative changes in phase behaviour of anionic surfactants
when formulation variables are modified [22].

65
 Various investigators have focused on interactions in an adsorbed
interfacial film to explain the direction and extent of interfacial curvature. The
first concept was that of Bancroft [23] and Clowes [24] who considered the
adsorbed film in emulsion systems to be duplex in nature, with an inner and an
outer interfacial tension acting independently [25]. The interface would then
curve such that the inner surface was one of higher tension. Bancroft’s rule was
stated as “that phase will be external in which the emulsifier is most soluble”;
i.e., oil-soluble emulsifiers will form w/o emulsions and water-soluble
emulsifiers o/w emulsions. This qualitative concept was largely extended and
several parameters have been proposed to quantify the nature of the surfactant
film. They are briefly presented in this section. Further details concerning the
these microemulsion types and their location in the phase diagram will be given
in Section 3.3.3.

The R-ratio
The R-ratio was first proposed by Winsor [21] to account for the
influence of amphiphiles and solvents on interfacial curvature. The primary
concept is to relate the energies of interaction between the amphiphile layer
and the oil and water regions. Therefore, this R-ratio compares the tendency
for an amphiphile to disperse into oil, to its tendency to dissolve in water. If
one phase is favoured, the interfacial region tends to take on a definite
curvature. A brief description of the concept is given below, and a full account
can be found elsewhere [26].
In micellar or microemulsion solutions, three distinct (single or
multicomponent) regions can be recognized: an aqueous region, W, an oil or
organic region, O, and an amphiphilic region, C. As shown in Figure 3.3, it is
useful to consider the interfacial zone as having a definite composition,
separating essentially bulk-phase water from bulk-phase oil. In this simple
picture, the interfacial zone has a finite thickness, and will contain, in addition
to surfactant molecules, some oil and water.

66
Figure 3.2 Winsor classification and phase sequence of microemulsions
encountered as temperature or salinity is scanned for non-ionic and ionic
surfactant respectively. Most of the surfactant resides in the shaded area. In
the three-phase system the middle-phase microemulsion (M) is in equilibrium
with both excess oil (O) and water (W).

2 phases 3 phases 2 phases

excess O
oil
W/O
M
O /W
excess
W water

Winsor I III II

Increasing temperature
(non-ionic surfactants)

Increasing salinity
(ionic surfactants)

Table 3.1 Qualitative effect of several variables on the observed phase

behaviour of anionic surfactants. After Bellocq et al. [22]

Scanned variables (increase) Ternary diagram transition

Salinity I ⌫III ⌫II

Oil: Alkane carbon number II ⌫III ⌫I
Alcohol: low M.W.a I ⌫III ⌫II
Alcohol: high M.W.b I ⌫III ⌫II
Surfactant: lipophilic chain length I ⌫III ⌫II
Temperature II ⌫III ⌫I

67
Figure 3.3 Interaction energies in the interfacial region of an oil-surfactant-
water system.
Ao o

ALL

ALc o

ALcw
AHco

AH cw
AHH

Aw w

OIL PHA S E: O A c o = A Lc o + A Hco

A c w = A Lcw + A Hc w
SURFA CTA NT
:C
LAY ER

WA TER PHA S E: W

Cohesive interaction energies therefore exist within the C layer, and these
determine interfacial film stability. They are depicted schematically in Figure
3.3: the cohesive energy between molecules x and y is defined as Axy, and is
positive whenever interaction between molecules is attractive. Axy is depicted as
the cohesive energy per unit area between surfactant, oil and water molecules
residing in the anisotropic interfacial C layer. For surfactant−oil and
surfactant−water interactions Axy can be considered to be composed of two
additive contributions:

A xy = A Lxy + A Hxy (3.3.1)

68
where ALxy quantifies interaction between nonpolar portions of the two
molecules (typically London dispersion forces) and AHxy represents polar
interactions, especially hydrogen bonding or Colombic interactions. Thus, for
surfactant−oil and surfactant−water interactions, cohesive energies to be
considered are:

• A co = A Lco + A Hco (3.3.2)

• A cw = A Lcw + A Hcw (3.3.3)

AHco and ALcw are generally very small values and can be ignored.

Other cohesive energies are those arising from the following interactions:
• water−water, Aww
• oil−oil, Aoo
• hydrophobic−hydrophobic parts (L) of surfactant molecules, ALL
• hydrophilic−hydrophilic parts (H) of surfactant molecules, AHH

The cohesive energy Aco evidently promotes miscibility of the surfactant

molecules with the oil region, and Acw with water. On the other hand, Aoo and
ALL oppose miscibility with oil, while Aww and AHH oppose miscibility with water.
Therefore, interfacial stability is ensured if the difference in solvent interactions
in C with oil and water bulk phases is sufficiently small. Too large a difference,
i.e., too strong affinity of C for one phase or the other, would drive to a phase
separation.

Winsor expressed qualitatively this variation in dispersing tendency by:

A co
R= (3.3.4)
A cw

To account for the structure of the oil, and the interactions between surfactant
molecules, an extended version of the original R-ratio was proposed [26]:

69
 R=
(A co − A oo − A )
LL
(3.3.5)
(A cw − A ww − A )
HH

As mentioned before, in many cases, AHco and ALcw are negligible, so Aco and
Acw can be approximated respectively to ALco and AHcw.

In brief, Winsor's primary concept is that this R-ratio of cohesive

energies, stemming from interaction of the interfacial layer with oil, divided by
energies resulting from interactions with water, determines the preferred
interfacial curvature. Thus, if R > 1, the interface tends to increase its area of
contact with oil while decreasing its area of contact with water. Thus oil tends
to become the continuous phase and the corresponding characteristic system is
type II (Winsor II). Similarly, a balanced interfacial layer is represented by R =
1.

Packing parameter and microemulsion structures

Changes in film curvature and microemulsion type can be addressed
quantitatively in terms of geometric requirements. This concept was introduced
by Israelachivili et al. [27] and is widely used to relate surfactant molecular
structure to interfacial topology. As described in Section 2.3.3, the preferred
curvature is governed by relative areas of the head group, ao, and the tail
group, v/lc (see Figure 2.6 for the possible aggregate structures). In terms of
microemulsion type,

• if ao > v/lc, then an oil-in-water microemulsion forms,

• if ao < v/lc, then a water-in-oil microemulsion forms,
• if ao ≈ v/lc, then a middle-phase microemulsion is the preferred
structure.

Hydrophilic−Lipophilic Balance (HLB)

Another concept relating molecular structure to interfacial packing and
film curvature is HLB, the hydrophilic−lipophilic balance. It is generally

70
expressed as an empirical equation based on the relative proportions of
hydrophobic and hydrophilic groups within the molecule. The concept was first
introduced by Griffin [28] who characterized a number of surfactants, and
derived an empirical equation for non-ionic alkyl polyglycol ethers (CiEj) based
on the surfactant chemical composition [29]:

HLB = (E j wt % + OH wt % )/ 5 (3.3.6)

where Ej wt% and OH wt% are the weight percent of ethylene oxide and
hydroxide groups respectively.

Davies et al. [30] proposed a more general empirical equation that

associates a constant to the different hydrophilic and hydrophobic groups:

HLB = [(n H × H ) − (n L × L )] + 7 (3.3.7)

where H and L are constants assigned to hydrophilic and hydrophobic groups

respectively, and nH and nL the number of these groups per surfactant
molecule.

For bicontinuous structures, i.e., zero curvature, it was shown that HLB ≈
10 [31]. Then w/o microemulsions form when HLB < 10, and o/w
microemulsion when HLB > 10. HLB and packing parameter describe the same
basic concept, though the latter is more suitable for microemulsions. The
influence of surfactant geometry and system conditions on HLB numbers and
packing parameter is illustrated in Figure 3.4.

71
Figure 3.4 Effect of molecular geometry and system conditions on the
packing parameter and HLB number (after Israelachvili [31]).

Larger, more hydrated head groups Smaller, less hydrated head groups
Lower ionic strength, larger d/lc Higher ionic strength, smaller d/lc
Lower pH (cationics), higher pH (anionics) Higher pH (cationics), lower pH (anionics)
Lower temperature (non-ionics) Higher temperature (non-ionics)

LARGER ao SMALLER ao

ao
d
O/W lc v W/O

v/aolc 1/3 1/2 1 2 3

HLB 40 20 10 2 1

Structure of
Surfactant Bilayer Inverted
Micelle
aggregate vesicle micelle

SMALLER v, LARGER lc LARGER v, SMALLER lc

Single, saturated chains Branched, unsaturated chains

Shorter chains, larger d/lc Double-chains, higher temperature
Less oil-penetration Greater oil-penetration
Higher MW oil Co-surfactant addition

Phase Inversion Temperature (PIT)

Non-ionic surfactants form water−oil microemulsions (and emulsions)
with a high temperature sensitivity. In particular, there is a specific phase
inversion temperature (PIT) and the film curvature changes from positive to
negative. This critical point was defined by Shinoda et al. [32]:

• if T < PIT, an oil-in-water microemulsion forms (Winsor I),

• if T > PIT, a water-in-oil microemulsion forms (Winsor II),
• at T = PIT, a middle-phase microemulsion exists (Winsor III) with a
spontaneous curvature equal to zero, and a HLB number (Eq. 3.3.6)
approximately equal to 10.

72
The HLB number and PIT are therefore connected; hence the term HLB
temperature is sometimes employed [33].

3.3.2 Surfactant film properties

an alternative, more physically realistic, approach is to consider

mechanical properties of a surfactant film at an oil−water interface. This film
can be characterized by three phenomenological constants: tension, bending
rigidity, and spontaneous curvature. Their relative importance depends on the
constraints felt by the film. It is important to understand how these parameters
relate to interfacial stability since surfactant films determine the static and
dynamic properties of microemulsions (and emulsions). These include phase
behaviour and stability, structure, and solubilisation capacity.

Ultra-low interfacial tension

Interfacial (or surface) tensions, γ, were defined in Chapter 2 for planar
surfaces, and the same principle applies for curved liquid−liquid interfaces, i.e.,
it corresponds to the work required to increase interfacial area by unit amount.
As mentioned in Section 3.2.1, microemulsion formation is accompanied by
ultra-low interfacial oil−water tensions, γo/w, typically 10-2 to 10-4 mN m-1. They
are affected by the presence of a co-surfactant, as well as electrolyte and/or
temperature, pressure, and oil chain length. Several studies have been reported
on the effect of such variables on γo/w. In particular, Aveyard and coworkers
performed several systematic interfacial tension measurements on both ionics
[34, 35] and non-ionics [36], varying oil chain length, temperature, and
electrolyte content. For example, in the system water−AOT−n-heptane, at
constant surfactant concentration (above its CMC), a plot of γo/w as a function
of electrolyte (NaCl) concentration shows a deep minimum that corresponds to
the Winsor phase inversion; i.e., upon addition of NaCl, γo/w decreases to a
minimum critical value (Winsor III structure), then increases to a limiting value
close to 0.2−0.3 mN m-1 (Winsor II region). At constant electrolyte
concentration, varying temperature [34], oil chain length and co-surfactant

73
content [35] have a similar effect. With non-ionics, a similar tension curve and
phase inversion are observed, but on increasing temperature [36]. In addition,
when increasing surfactant chain length, the interfacial tension curves shift to
higher temperatures and the minimum in γo/w decreases [37]. Ultra-low
interfacial tensions cannot be measured with standard techniques such as Du
Nouy Ring, Wilhelmy plate, or drop volume (DVT). Appropriate techniques for
this low tension range are spinning drop tensiometry (SDT) and surface light
scattering [38].

Spontaneous curvature
Spontaneous (or natural or preferred) curvature Co is defined as the
curvature formed by a surfactant film when a system consists of equal amounts
of water and oil. Then, there is no constraint on the film, which is free to adopt
the lowest free energy state. Whenever one phase is predominant, there is a
deviation from Co. In principle, every point on a surface possesses two principal
radii of curvature, R1 and R2 and their associated principal curvatures are C1 =
1/R1 and C2 = 1/R2. Mean and Gaussian curvatures are used to define the
bending of surfaces. They are defined as follows [39]:

Mean curvature: C = ½ (1/R1 + 1/R2)

Gaussian curvature: κ = 1/R1 × 1/R2

C1 and C2 are determined as follows: every point on the surface of the

surfactant film has two principal radii of curvature, R1 and R2 as shown in
Figure 3.5. If a circle is placed tangentially to a point p on the surface and if the
circle radius is chosen so that its second derivative at the contact point equals
that of the surface in the direction of the tangent (of normal vector, n), then
the radius of the circle is a radius of curvature of the surface. The curvature of
the surface is described by two such circles chosen in orthogonal (principal)
directions as shown in Figure 3.5(a).

74
 For a sphere, R1 and R2 are equal and positive (Figure 3.5(b)). For a
cylinder R2 is indefinite (Figure 3.5(c)) and for a plane, both R1 and R2 are
indefinite. In the special case of a saddle, R1 = −R2, i.e., at every point the
surface is both concave and convex (Figure 3.5(d)). Both a plane and saddle
have the property of zero mean curvature.

The curvature Co depends both on the composition of the phases it

separates and on surfactant type. One argument applied to the apolar side of
the interface is that oil can penetrate to some extent between the surfactant
hydrocarbon tails. The more extensive the penetration, the more curvature is
imposed toward the polar side. This results in a decrease of Co since, by
convention, positive curvature is toward oil (and negative toward water). The
longer the oil chains, the less they penetrate the surfactant film and the smaller
the effect on Co. Recently, Eastoe et al. have studied the extent of solvent
penetration in microemulsions stabilized by di-chained surfactants, using SANS
and selective deuteration. Results suggested that oil penetration is a subtle
effect, which depends on the chemical structures of both surfactant and oil. In
particular, unequal surfactant chain length [40-43] or presence of C = C bonds
[44] result in a more disordered surfactant/oil interface, thereby providing a
region of enhanced oil mixing. For symmetric di-chained surfactants (e.g.,
DDAB and AOT), however, no evidence for oil mixing was found [42]. The
effect of alkane structure, and molecular volume on the oil penetration was also
investigated with n-heptane, and cyclohexane. The results indicate that
heptane is essentially absent from the layers, but the more compact
cyclohexane has a greater penetrating effect [43].

75
Figure 3.5 Principal curvatures of different surfaces. (a) Intersection of the
surfactant film surface with planes containing the normal vector (n) to the
surface at the point p. (b) convex curvature, (c) cylindrical curvature, (d)
saddle-shaped curvature. After Hyde et al. [39].
n

(a) R1 R2

R1 > 0 R1 = ∝ R1 < 0
(b) (c) (d)
R2 > 0 R2 > 0 R2 > 0

76
Surfactant type, and nature of the polar head group, also influences Co through
different interactions with the polar (aqueous) phase:

• For ionic surfactants electrolyte content and temperature affect the

spontaneous curvature in opposite ways. An increase in salt concentration
screens electrostatic head group repulsions − i.e., decreases head group
area − so the film curves more easily toward water, leading to a decrease in
Co. Raising temperature has two effects: (1) an increase in electrostatic
repulsions between head groups due to higher counterion dissociation, so
Co increases; (2) more gauche conformations are induced in the surfactant
chains, which become more coiled, resulting in a decrease in Co. Therefore
the combined effects of temperature on the apolar chains and on
electrostatic interactions are competitive. The electrostatic term is believed
to be slightly dominant, so Co increases weakly with increasing
temperature.

• For non-ionic surfactants, unsurprisingly, electrolytes have very little effect

on Co, whereas temperature is a critical parameter due to the strong
dependence of their solubility (in water or oil) on temperature. For
surfactants of the CiEj type as temperature increases water becomes a less
good solvent for the hydrophilic units and penetrates less into the surfactant
layer. In addition, on the other side of the film, oil can penetrate further into
the hydrocarbon chains, so that increasing temperature for this type of
surfactant causes a strong decrease in Co. This phenomenon explains the
strong temperature effects on the phase equilibria of such surfactants as
shown in Figure 2.8 (see Chapter 2).

Thus, by changing external parameters such as temperature, nature of

the oil or electrolyte concentration, the spontaneous curvature can be tuned to
the appropriate value, and so drive transitions between Winsor systems. Other
factors affect Co in a similar fashion; they include varying the polar head group,

77
type and valency of counterions, length and number of apolar chains, adding a
co-surfactant, or mixing surfactants.
Film bending rigidity
Film rigidity is an important parameter associated with interfacial
curvature. The concept of film bending energy was first introduced by Helfrich
[45] and is now considered as an essential model for understanding
microemulsion properties. It can be described by two elastic moduli [46] that
measure the energy required to deform the interfacial film from a preferred
mean curvature:
• the mean bending elasticity (or rigidity), K, associated with the mean
curvature, that represents the energy required to bend unit area of
surface by unit amount. K is positive, i.e., spontaneous curvature is
favoured;
• the factor K is associated with Gaussian curvature, and hence accounts
for film topology. K is negative for spherical structures or positive for
bicontinuous cubic phases.
Theoretically, it is expected that bending moduli should depend on surfactant
chain length [47], area per surfactant molecule in the film [48] and electrostatic
head group interactions [49].

The film rigidity theory is based on the interfacial free energy associated
with film curvature. The free energy, F, of a surfactant layer at a liquid
interface may be given by the sum of an interfacial energy term, Fi, a bending
energy term, Fb, and an entropic term, Fent. For a droplet type structure this is
written as [50]:

⎡K ⎤
F = Fi + Fb + Fent = γA + ∫ ⎢ (C1 + C 2 − 2C o ) + KC1C 2 ⎥ dA + nk B Tf (φ )
2

⎣2 ⎦
(3.3.8)

where γ is the interfacial tension, A is the total surface area of the film, K is the

mean elastic bending modulus, K is the Gaussian bending modulus, C1 and C2

78
are the two principal curvatures, Co the spontaneous curvature, n is the number
of droplets, kB is the Boltzmann constant, and f(φ)is a function accounting for
the entropy of mixing of the microemulsion droplets, where φ is the droplet
core volume fraction. For dilute systems where φ < 0.1, it was shown that
f (φ ) = [ln(φ ) − 1] [50]. Microemulsion formation is associated with ultra-low

interfacial tension, γ so the γA term is small compared to Fb and Fent and can be
ignored as an approximation.
As mentioned previously, the curvatures C1, C2 and Co can be expressed
in terms of radii as 1/R1, 1/R2, and 1/Ro respectively. For spherical droplets, R1
= R2 = R, and the interfacial area is A = n4πR 2 . Note that R and Ro are core
radii rather than droplet radii [50]. Solving Eq. 3.3.8 and dividing by area A, the
total free energy, F, for spherical droplets (of radius R) is expressed as:

2
F ⎛1 1 ⎞ K ⎡k T ⎤
= 2K⎜⎜ − ⎟⎟ + 2 + ⎢ B 2 f (φ )⎥ (3.3.9)
A ⎝ R Ro ⎠ R ⎣ 4πR ⎦

For systems where the solubilisation boundary is reached (WI or WII region), a
microemulsion is in equilibrium with an excess phase of the solubilisate and the
droplets have achieved their maximum size, i.e., the maximum core radius,

max . Under this condition the minimization of the total free energy leads to a
R av

relation between the spontaneous radius, Ro, and the elastic constants K and
K [51]:

R av 2K + K k B T
max
= + f (φ ) (3.3.10)
Ro 2K 8πK

A number of techniques have been used to determine K and K

separately, in particular, ellipsometry, X-ray reflectivity, and small-angle X-ray
scattering (SAXS) techniques [52-54]. De Gennes and Taupin [55] have
developed a model for bicontinuous microemulsions. For Co = 0 the layer is

79
supposed to be flat in the absence of thermal fluctuations. They introduced the
term ξK, the persistence length of the surfactant layer that relates to K via:

ξ K = a exp(2πK / k B T) (3.3.11)

where a is a molecular length and ξK is the correlation length for the normals to
the layer, i.e., the distance over which this layer remains flat in the presence of
thermal fluctuations. ξK is extremely sensitive to the magnitude of K. When K
>> kBT, ξK is macroscopic, i.e., the surfactant layer is flat over large distances
and ordered structures such as lamellar phases may form. If K is reduced to ~
kBT then ξK is microscopic, ordered structures are unstable and disordered
phases such as microemulsions may form. Experiments reveal that K is typically
between 100 kBT for condensed insoluble monolayers [56] and about 10 kBT for
lipid bilayers [57-59] but can decrease below kBT in microemulsion systems
[60]. The role of K is also important, however, there are few measurements of
this quantity in the literature [e.g., 53, 61]. Its importance in determining the
structure of surfactant−oil−water mixtures is still far from clear.

An alternative, more accessible, method to quantify film rigidities is to

calculate the composite parameter (2K + K ) using tensiometry and SANS

techniques. This parameter can be derived for droplet microemulsion at the

solubilisation boundary, WI or WII system, by combining the radius of the
droplet with interfacial tensions or droplet polydispersity. Two expressions can
be derived from Eq. 3.3.9 and 3.3.10.

1. Using the interfacial tension γo/w (measured by SLS or SDT) and the
maximum mean core radius Rmax
av
(measured by SANS see Chapter 4):
γo/w at the interface between microemulsion and excess phases at the
solubilisation boundary can be expressed in terms of these elastic moduli and
R av
max [52]. Any new area created must be covered by a monolayer of

surfactant, and so this energy may be calculated in the case of WI or WII

systems since the surfactant monolayer is taken from around the curved

80
microemulsion droplets [56]. To do this it is necessary to unbend the surfactant
film, introducing a contribution from K, of 2K /( R av
max ) . The resulting change in
2

the number of microemulsion droplets introduces an entropic contribution and a

contribution due to the change in topology involving K , of K /(R av
max ) . So the
2

interfacial tension between the microemulsion and excess phase is given by:

2K + K k BT
γo/w = + f (φ ) (3.3.11)
av
(R max ) 2
4π(R av
max )
2

which gives for the bending moduli:

k BT
f (φ )
2K + K = γ o / w (R av
max ) −
2
(3.3.12)
4π
2. Using the Schultz polydispersity width p = σ/Rmax
av
obtained from SANS
analysis (see Chapter 4):
Droplet polydispersity relates to the bending moduli through thermal
fluctuations of the microemulsion droplets. Safran [62] and Milner [63]
described the thermal fluctuations by an expansion of the droplet deformation
in terms of spherical harmonics. The principal contribution to these fluctuations
was found to arise from the deformation mode l = 0 only [50]; and l = 0
deformations are fluctuations in droplet size, i.e., changes of the mean droplet
radius and hence the droplet polydispersity. In the case of the two phase
equilibria at maximum solubilisation (WI or WII), this polydispersity, p, may be
expressed as a function of K and K :

u o2 k BT
p =
2
= (3.3.13)
4π 8π(2K + K ) + 2k B Tf (φ )

where uo is the fluctuation amplitude for the l = 0 mode. This polydispersity is

max [64], and Eq. 3.3.13

given by the SANS Schultz polydispersity parameter σ/R av

can be written:

81
 k BT k T
2K + K = − B f (φ ) (3.3.14)
8π(σ/R max )
av 2
4π

Therefore Eq. 3.3.12 and 3.3.14 give two accessible expressions for the
sum (2K + K ) using data from SANS and tensiometry. This approach has been

shown to work well with non-ionic films in WI systems [50, 65], and also
cationic [64] and zwitterionic [66] layers in WII microemulsions. Figure 3.6
shows results for these latter two classes of system, as a function of surfactant
alkyl carbon number n-C. The good agreement between equation 3.3.12 and
3.3.14 suggests they can be used with confidence. These values are in line with
current statistical mechanical theories [48], which suggest that K should vary as
n-C2.5 to n-C3, whereas there is only a small effect on K .

82
Figure 3.6 Film rigidities (2K + K ) as a function of total alkyl carbon number
n-C from Winsor II microemulsions. The lines are guides to the eye.

4.0

(2K + K ) ~ n - C2.35
3.5

3.0

phospholipids
(2K + K) / kBT

2.5
zwitterionic

2.0

1.5

1.0
dialkylammonium
cationics
0.5

0.0
15 20 25 30 35 40

n-C
n-C

k BT k BT
SANS (2K + K ) = − f(φ )
8π(σ/R max ) av 2
4π

SANS & k BT
(2K + K ) = γ o/w (Rmax )2 − f(φ )
av

tensiometry 4π

83
3.3.3 Phase behaviour

Solubilisation and interfacial properties of microemulsions depend upon

pressure, temperature and also on the nature and concentration of the
components. The determination of phase stability diagrams (or phase maps),
and location of the different structures formed within these
water(salt)−oil−surfactant−alcohol systems in terms of variables are, therefore,
very important. Several types of phase diagram can be identified depending on
the number of variables involved. In using an adequate mode of
representation, it is possible to describe not only the limits of existence of the
single and multiphase regions, but also to characterize equilibria between
phases (tie-lines, tie-triangles, critical points, etc.). Below is a brief description
of ternary and binary phase maps, as well as the phase rule that dictates their
construction.

Phase rule
The phase rule enables the identification of the number of variables (or
degrees of freedom) depending on the system composition and conditions. It
is generally written as [67]:
F=C−P+2 (3.3.15)

where F is the number of possible independent changes of state or degrees of

freedom, C the number of independent chemical constituents, and P the
number of phases present in the system. A system is called invariant,
monovariant, bivariant, and so on, according to whether F is zero, 1, 2, and so
on. For example, in the simplest case of a system composed of three
components and two phases, F is univariant at a fixed temperature and
pressure. This means that the mole or weight fraction of one component in one
of the phases can be specified but all other compositions in both phases are
fixed. In general, microemulsions contain at least three components: oil (O),
water (W) and amphiphile (S), and as mentioned previously a co-surfactant
(alcohol) and/or an electrolyte are usually added to tune the system stability.
These can be considered as simple O−W−S systems: whenever a co-surfactant

84
is used, the ratio oil:alcohol is kept constant and it is assumed that the alcohol
does not interact with any other component so that the mixture can be treated
(to a first approximation) as a three-component system. At constant pressure,
the composition−temperature phase behaviour can be presented in terms of a
phase prism, as illustrated in Figure 3.7. However, the construction of such a
phase map is rather complex and time consuming so it is often convenient to
simplify the system by studying specific phase-cuts. The number of variables
can be reduced either by keeping one term constant and/or by combining two
or more variables. Then, ternary and binary phase diagrams are produced.

Ternary phase diagrams

At constant temperature and pressure, the ternary phase diagram of a
simple three-component microemulsion is divided into two or four regions as
shown in Figure 3.8. In each case, every composition point within the single-
phase region above the demixing line corresponds to a microemulsion.
Composition points below this line correspond to multiphase regions comprising
in general microemulsions in equilibrium with either an aqueous or an organic
phase or both, i.e., Winsor type systems (see Section 3.3.1).

Any system whose overall composition lies within the two-phase region
(e.g., point o in Figures 3.8(a) and 3.8(c)) will exist as two phases whose
compositions are represented by the ends of the “tie-line”, i.e., a segment
formed by phases m and n. Therefore, every point on a particular tie-line has
identical coexisting phases (m and n) but of different relative volumes. When
the two conjugate phases have the same composition (m = n), this
corresponds to the plait (or critical) point, p.

If three phases coexist (Figure 3.8(b)), i.e., corresponding to WIII, the

system at constant temperature and pressure is, according to the phase rule,
invariant. Then, there is a region of the ternary diagram that consists of three-
phase systems having invariant compositions and whose boundaries are tie-
lines in the adjacent two-phase regions that surround it. This region of three-
phase invariant compositions is therefore triangular in form and called “tie-

85
triangle” [26]. Any overall composition, such as point q (Figure 3.8(b)) lying
within the tie-triangle will divide into three phases having compositions
corresponding to the vertices A, B, and C of the triangle. The compositions A, B,
and C are invariant in the sense that varying the position q, the overall
composition, throughout the triangle will result in variations in the amounts of
the phases A, B, and C but not in their composition.

86
Figure 3.7 The phase prism, describing the phase behaviour of a ternary
system at constant pressure.

Temp.
W O

Figure 3.8 Ternary diagram representations of two- and three-phase regions

formed by simple water−oil−surfactant systems at constant temperature and
pressure. (a) Winsor I type, (b) Winsor II type, (c) Winsor III type systems.

surfactant

1φ

surfactant 2φ 2φ surfactant
B

3φ q
A C

1φ water oil 1φ
(b)

2φ m 2φ
p p m
o o

n n
water oil water oil
(a) (c)

Salinity

Winsor type I III II

87
Binary phase diagrams
As mentioned previously, ternary diagrams can be further simplified by
fixing some parameters and/or combining two variables together (e.g., water
and electrolyte into brine, or water and oil into water−to−oil ratio), i.e.,
reducing the degrees of freedom. Then, determining the phase diagram of such
systems reduces to a study of a planar section through the phase prism.
Examples of such pseudo-binary diagrams are given in Figures 3.9 to 3.11 for
non-ionic and anionic surfactants.
Figure 3.9 shows the schematic phase diagram for a non-ionic
surfactant−water−oil ternary system. Since temperature is a crucial variable in
the case of non-ionics, the pseudo-binary diagram is represented by the planar
section defined by φw = φo, where φw and φo are the volume fractions of water
and oil respectively. Then, at constant pressure, defining the system in a single-
phase region requires the identification of two independent variables (F = 2),
i.e., temperature and surfactant concentration. The section shown in Figure
3.9(b) can be used to determine TL and TU, the lower and upper temperatures,
respectively, of the phase equilibrium W+M+O (with M, the microemulsion
phase), and the minimum amount of surfactant necessary to solubilise equal
amounts of water and oil, denoted Cs* [68]. The lower Cs* the more efficient
the surfactant. Figure 3.10 illustrates the determination of a second possible
section for a non-ionic surfactant−water−oil ternary system: pressure and
surfactant concentration are kept constant, leaving the two variables,
temperature and water−to−oil ratio (φw-o). This diagram shows the various
surfactant phases obtained as a function of temperature and water−to−oil ratio
[68]. The third example (Figure 3.11) concerns an anionic surfactant, Aerosol-
OT. In order to obtain F = 2 when defining the ternary W−O−S system in a
single-phase region at constant pressure, the surfactant concentration
parameter is fixed. Then, the two variables are temperature and w, the
water−to−surfactant molar ratio defined as w = [water] / [surfactant]. w
represents the number of water molecules solubilised per surfactant molecule,
so that this phase diagram characterizes the surfactant efficiency, as a
microemulsifier.

88
Figure 3.9 Binary phase behaviour in ternary microemulsion systems formed
with non-ionic surfactants. (a) Illustration of the section through the phase
prism at equal water and oil content. (b) Schematic phase diagram plotted as
temperature versus surfactant concentration Cs. TL and TU are the lower and
upper temperatures, respectively, of the phase equilibrium W+M+O. T* is the
temperature at which the three-phase triangle is an isosceles, i.e., when the
middle-phase microemulsion contains equal amounts of water and oil. This
condition is also termed 'balanced'. Cs* is the surfactant concentration in the
middle-phase microemulsion at balanced conditions. ‘Lam’ denotes a lamellar
liquid crystalline phase. After Olsson and Wennerström [68].

Temp.
(a) W O

TU
W+M
Temperature

T* W+M+O M Lam+M Lam

TL O+M

Cs*
(b) Surfactant conc.

89
Figure 3.10 Binary phase behaviour in ternary microemulsion systems formed
with non-ionic surfactants. (a) Illustration of a section at constant surfactant
concentration through the phase prism. (b) Schematic phase diagram, plotted
as temperature versus volume fraction of oil, φo, at constant surfactant
concentration. Also shown are various microstructures found in different
regions of the microemulsion phase, M. At higher temperatures the liquid phase
is in equilibrium with excess water (M+W), and at lower temperatures with
excess oil (M+O). At intermediate temperatures a lamellar phase is stable at
higher water contents and higher oil contents, respectively. After Olsson and
Wennerström [68].

Temp.
(a) W O

normal monolayer
bilayer

40

35 M+W
Temperature / °C

M+O
30 Lam Lam
M
M+W
25
M+O

20

0 0.2 0.4 0.6 0.8 1

(b) Oil Volume Fraction

reverse
bilayer

90
Figure 3.11 Pseudo-binary phase diagram in ternary microemulsion systems
formed with the anionic surfactant Aerosol-OT (AOT) in various straight-chain
alkane solvents. The water–to–surfactant molar ratio, w, is plotted versus
temperature at constant surfactant concentration and pressure. Alkane carbon
numbers are indicated; ringed numbers correspond to the lower temperature
(solubilisation) boundary, TL, and un-ringed numbers to the upper temperature
(haze) boundary, TU. The single phase microemulsion region is located between
TL and TU. Below TL the system consists of a microemulsion phase in
equilibrium with excess water (WII type), and above TU the single
microemulsion phase separates into a surfactant-rich phase and an oil phase.
After Fletcher et al. [16].

5 5
6 6
100 7
7
8 8
9

10

w
50 11

12

0
0 15 30 45
T / °C

91
REFERENCES

1. Danielsson, I.; Lindman, B. Colloids Surf. A 1981, 3, 391.

2. Sjöblom, J.; Lindberg, R.; Friberg, S. E. Adv. Colloid Interface Sci. 1996,
125.
3. Schulman, J. H.; Stoeckenius, W.; Prince, M. J. Phys. Chem. 1959, 63,
1677.
4. Shinoda, K.; Friberg, S. Adv. Colloid Interface Sci. 1975, 4, 281.
5. Adamson, A. W. J. Colloid Interface Sci. 1969, 29, 261.3.
6. Friberg, S. E.; Mandell, L.; Larsson, M. J. Colloid Interface Sci. 1969, 29,
155.
7. Shah, D. O., Ed. ‘Surface Phenomena in Enhanced Recovery’ Plenum Press,
1981, New York.
8. Overbeek, J. Th. G. Faraday Discuss. Chem. Soc. 1978, 65, 7.
9. Tadros, Th. F.; Vincent, B. in ‘Encyclopaedia of Emulsion Technology’
Becher, P. Ed., Vol. 1, Marcel Dekker, 1980, New York.
10. Kunieda, H.; Shinoda, K. J. Colloid Interface Sci. 1980, 75, 601.
11. Chen, S. J.; Evans, F. D.; Ninham, B. W. J. Phys. Chem. 1984, 88, 1631.
12. Kahlweit, M.; Strey, R.; Busse, G. J. Phys. Chem. 1990, 94, 3881.
13. Hunter, R. J. ‘Introduction to Modern Colloid Science’ Oxford University
Press, 1994, Oxford.
14. Lekkerkerker, H. N. W.; Kegel, W. K.; Overbeek, J. Th. G. Ber. Bunsenges
Phys. Chem. 1996, 100, 206.
15. Ruckenstein, E.; Chi, J. C. J. Chem. Soc. Faraday Trans. 1975, 71, 1690.
16. Fletcher, P. D. I.; Howe, A. M.; Robinson, B. H. J. Chem. Soc. Faraday
Trans. 1 1987, 83, 985.
17. Fletcher, P. D. I.; Clarke, S.; Ye, X. Langmuir 1990, 6, 1301.
18. Biais, J.; Bothorel, P.; Clin, B.; Lalanne, P. J. Colloid Interface Sci. 1981, 80,
136.
19. Friberg, S.; Mandell, L.; Larson, M. J. Colloid Interface Sci. 1969, 29, 155.
20. Fletcher, P. D. I.; Horsup, D. I. J. Chem. Soc. Faraday Trans. 1 1992, 88,
855.

92
21. Winsor, P. A. Trans. Faraday Soc. 1948, 44, 376.
22. Bellocq, A. M.; Biais, J.; Bothorel, P.; Clin, B.; Fourche, G.; Lalanne, P.;
Lemaire, B.; Lemanceau, B.; Roux, D. Adv. Colloid Interface Sci. 1984, 20,
167.
23. Bancroft, W. D. J. Phys. Chem. 1913, 17, 501.
24. Clowes, G. H. A. J. Phys. Chem. 1916, 20, 407.
25. Adamson, A. W. ‘Physical Chemistry of Surfaces’ Interscience, 1960, p 393.
26. Bourrel, M.; Schechter, R. S. ‘Microemulsions and Related Systems’ Marcel
Dekker, 1988, New York.
27. Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. J. Chem. Soc. Faraday
Trans. 2 1976, 72, 1525.
28. Griffin, W. C. J. Cosmetics Chemists 1949, 1, 311.
29. Griffin, W. C. J. Cosmetics Chemists 1954, 5, 249.
30. Davies, J. T. Proc. 2nd Int. Congr. Surface Act. Vol. 1 Butterworths, 1959,
London.
31. Israelachvili, J. N. Colloids Surf. A 1994, 91, 1.
32. Shinoda, K.; Saito, H. J. Colloid Interface Sci. 1969, 34, 238.
33. Shinoda, K.; Kunieda, H. in ‘Encyclopaedia of Emulsion Technology’ Becher,
P. Ed., Vol. 1, Marcel Dekker, 1983, New York.
34. Aveyard, R.; Binks, B. P.; Clarke, S.; Mead, J. J. Chem. Soc. Faraday Trans.
1 1986, 82, 125.
35. Aveyard, R.; Binks, B. P.; Mead, J. J. Chem. Soc. Faraday Trans. 1 1986,
82, 1755.
36. Aveyard, R.; Binks, B. P.; Fletcher, P. D. I. Langmuir 1989, 5, 1210.
37. Sottmann, T.; Strey, R. Ber. Bunsenges Phys. Chem. 1996, 100, 237.
38. Langevin, D., Ed. ‘Light Scattering by Liquid Surfaces and Complementary
Techniques’ Marcel Dekker, 1992, New York.
39. Hyde, S.; Andersson, K.; Larsson, K.; Blum, Z.; Landh, S.; Ninham, B. W.
‘The Language of Shape’ Elsevier, 1997, Amsterdam.
40. Eastoe, J.; Dong, J.; Hetherington, K. J.; Steytler, D. C.; Heenan, R. K. J.
Chem. Soc. Faraday Trans. 1996, 92, 65.

93
41. Eastoe, J.; Hetherington, K. J.; Sharpe, D.; Dong, J.; Heenan, R. K.;
Steytler, D. C. Langmuir 1996, 12, 3876.
42. Eastoe, J.; Hetherington, K. J.; Sharpe, D.; Dong, J.; Heenan, R. K.;
Steytler, D. C. Colloids Surf. A 1997, 128, 209.
43. Eastoe, J.; Hetherington, K. J.; Sharpe, D.; Steytler, D. C.; Egelhaaf, S.;
Heenan, R. K. Langmuir 1997, 13, 2490.
44. Bumajdad, A.; Eastoe, J.; Heenan, R. K.; Lu, J. R.; Steytler, D. C.; Egelhaaf,
S. J. Chem. Soc. Faraday Trans. 1998, 94, 2143.
45. Helfrich, W. Z. Naturforsch. 1973, 28c, 693.
46. Kellay, H.; Binks, B. P.; Hendrikx, Y.; Lee, L. T.; Meunier, J. Adv. Colloid
Interface Sci. 1994, 9, 85.
47. Safran, S. A.; Tlusty, T. Ber. Bunsenges. Phys. Chem. 1996, 100, 252.
48. Szleifer, I.; Kramer, D.; Ben-Shaul, A.; Gelbart, W. M.; Safran, S. J. Chem.
Phys. 1990, 92, 6800.
49. Winterhalter, M.; Helfrich, W. J. Phys. Chem. 1992, 96, 327.
50. Gradzielski, M.; Langevin, D.; Farago, B. Phys. Rev. E 1996, 53, 3900.
51. Safran, S. A. in ‘Structure and Dynamics of Strongly Interacting Colloids and
Supramolecular Aggregates in Solution’ Vol. 369 of NATO Advanced Study
Institute, Series C: Mathematical and Physical Sciences, Chen, S. H.; Huang,
J. S.; Tartaglia, P. Ed.; Kluwer, Dortrecht, 1992.
52. Meunier, J.; Lee, L. T. Langmuir 1991, 46, 1855.
53. Kegel, W. K.; Bodnar, I.; Lekkerkerker, H. N. W. J. Phys. Chem. 1995, 99,
3272.
54. Sicoli, F.; Langevin, D.; Lee, L. T. J. Chem. Phys. 1993, 99, 4759.
55. De Gennes, P. G.; Taupin, C. J. Phys. Chem. 1982, 86, 2294.
56. Daillant, J.; Bosio, L.; Benattar, J. J.; Meunier, J. Europhys. Lett. 1989, 8,
453.
57. Shneider, M. B.; Jenkins, J. T.; Webb, W. W. Biophys. J. 1984, 45, 891.
58. Engelhardt, H; Duwe, H. P.; Sackmann, E. J. Phys. Lett. 1985, 46, 395.
59. Bivas, I.; Hanusse, P.; Botherel, P.; Lalanne, J.; Aguerre-Chariol, O. J.
Physique 1987, 48, 855.
60. Di Meglio, J. M.; Dvolaitzky, M. Taupin, C. J. Phys. Chem. 1985, 89, 871.

94
61. Farago, B.; Huang, J. S.; Richter, D.; Safran, S. A.; Milner, S. T. Progr.
Colloid Polym. Sci. 1990, 81, 60.
62. Safran, S. A. J. Chem. Phys. 1983, 78, 2073.
63. Milner, S. T.; Safran, S. A. Phys. Rev. A 1987, 36, 4371.
64. Eastoe, J.; Sharpe, D.; Heenan, R. K.; Egelhaaf, S. J. Phys. Chem. B 1997,
101, 944.
65. Gradzielski, M.; Langevin, D. J. Mol. Struct. 1996, 383, 145.
66. Eastoe, J.; Sharpe, D. Langmuir 1997, 13, 3289.
67. Rock, P. A. ‘Chemical Thermodynamics’ MacMillan, 1969, London.
68. Olsson, U.; Wennerström, H. Adv. Colloid Interface Sci. 1994, 49, 113.

95
 INTERNATIONALE PHARMACEUTICA SCIENCIA
| April-June 2011 | Vol. 1 | Issue 2 |
Available online http://www.ipharmsciencia.com
ISSN 2231-5896
©2011 IPS
REVIEW ARTICLE

Microemulsions- Potential Carrier for Improved Drug Delivery

ABSTRACT Sajal Kumar Jha1*, Sanjay Dey1,
Roopa Karki2
The novel carriers have been exploited through almost all the routes of
administration. Association of drugs with carriers is normally noncovalent, based
1 Department of
on collective strength of weak binding forces. Many newer carriers are evolving with
Pharmaceutics, Bengal College
the advent of technology and the demand of targeted delivery like microemulsions.
of Pharmaceutical Sciences &
Microemulsions are clear, stable, isotropic mixtures of oil, water and surfactant,
Research, Durgapur, India.
frequently in combination with a cosurfactant. These systems are currently of
2. Department of Pharmaceutics, Malla
interest to the pharmaceutical scientist because of their considerable potential to act
Reddy College of Pharmacy,
as drug delivery vehicles by incorporating a wide range of drug molecules. They
Hyderabad, India.
offer the advantage of spontaneous formation, ease of manufacturing and scale-up,
thermodynamic stability, and improved drug solubilization and bioavailability.
Date of Submission: 01-06-2011
Preparing a pharmaceutically acceptable dosage form demands a clear
Date of Acceptance: 12-06-2011
understanding of the micro-emulsion structure, phase behavior, factors leading to
Conflict of interest: Nil
its thermodynamic stability, factors influencing drug release from the formulation,
Source of support: None
requirements of ideal microemulsion excipients, and the potential uses and
limitations of the microemulsion system.
Key Words: Microemulsions, Surfactants, Cosurfactants, Thermodynamic
stability, Phase behavior, Novel carriers.

Introduction locations, which lead to improved transportation of

In search of safe and effective therapy, the drugs toward the diseased locations. Further, it
development of new drugs has been the common improves the chances of the availability of the drug at
practice historically. However, it involved a long the specific receptor site and enhances drug receptor
gestation period in terms of time, efforts, and huge interaction through mediation of specialized
cost. Later on, it was realized that the issues pertaining composition and design of the carrier systems. All
to efficacy and safety are largely influenced by the these factors tend to potentiate the degree of
distribution of the drug within the biological system, as pharmacodynamic response, the safety and patient
there is appreciable deviation from the desired site of compliance being the immediate benefits [1].
action, i.e., the target site. This objective, hitherto un- Microemulsions are dispersions of nanometer-sized
accomplished gave way to an alternate approach of droplets of an immiscible liquid within another liquid.
drug delivery, wherein the carrier systems were used to Droplet formation is facilitated by the addition of
deliver the molecules to specific receptor sites without surfactants and often also cosurfactants. Through the
afflicting the normal tissues and organs of the body. years much has been written on the formation and
The fundamentals lie in hosting the drug in carefully stability of these oil-dispersed-in-water (o/w) or water-
designed carriers to bring favorable change(s) in its dispersed-in-oil (w/o) systems. Nevertheless, the
surrounding microenvironment, and consequently, its cosmetic formulator still seeks to understand and
delivery. It is the modification(s) in physicochemical create the most favorable cosmetically eloquent and
characteristics of the molecules and in the barrier functional products possible. Aesthetically appealing
properties of the biological membranes at various products can be formulated as trans- parent o/w or
Address for correspondence w/o dispersions called microemulsions. The possible
Mr. Sajal Kumar Jha application for these systems range from products with
Email: sajal.kumar.jha@gmail.com
25 Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2
 Sajal Kumar Jha, et al: Microemulsions- Potential Carrier for Improved Drug Delivery

an extended shelf life to delivery systems for active entropic contribution is very large dispersion entropy
ingredients [2]. arising from the
These versatile systems are currently of great mixing of one phase in the other in the form of large
technological and scientific interest to the researchers number of small droplets. However there are also
because of their potential to incorporate a wide range expected to be favorable entropic contributions arising
of drug molecules (hydrophilic and hydrophobic) due from other dynamic processes such as surfactant
to the presence of both lipophilic and hydrophilic diffusion in the interfacial layer and monomer-micelle
domains. These adaptable delivery systems provide surfactant exchange. Thus a negative free energy of
protection against oxidation, enzymatic hydrolysis and formation is achieved when large reductions in surface
improve the solubilization of lipophilic drugs and tension are accompanied by significant favorable
hence enhance their bioavailability. In addition to oral entropic change. In such cases, microemulsion is
and intravenous delivery, they are amenable for spontaneous and the resulting dispersion is
sustained and targeted delivery through ophthalmic, thermodynamically stable[4-5].
dental, pulmonary, vaginal and topical routes. Types of microemulsion systems:
Microemulsions are experiencing a very active According to Winsor, there are four types of
development as reflected by the numerous publications microemulsion phases exists in equilibria, these phases
and patents being granted on these systems [3]. are referred as Winsor phases [6-7]. They are,
Theories of Microemulsion Formation 1. Winsor I: With two phases, the lower (o/w)
Historically, three approaches have been used to microemulsion phases in equilibrium with the upper
explain microemulsion formation and stability. They excess oil.
are as follows- 2. Winsor II: With two phases, the upper
• Interfacial or mixed film theories. microemulsion phase (w/o)
• Solubilization theories. microemulsion phases in equilibrium with lower
• Thermodynamic treatments. excess water.
The free energy of microemulsion formation can be 3. Winsor III: With three phases, middle
considered to depend on the extent to which surfactant microemulsion phase (o/w plus w/o,
lowers the surface tension of the oil water interface and called bicontinous) in equilibrium with upper excess
change in entropy of the system such that, oil and lower excess water.
Gf = γ a - T S 4. Winsor IV: In single phase, with oil, water and
Where, Gf = free energy of formation surfactant homogenously mixed.
A = change in interfacial area of
Advantages of Microemulsion over Other
microemulsion
Dosage Forms
S = change in entropy of the system
T = temperature • Increase the rate of absorption

γ = surface tension of oil water interphase • Eliminates variability in absorption

• Helps solublize lipophilic drug

It should be noted that when a microemulsion is
formed the change in A is very large due to the large • Provides a aqueous dosage form for water

number of very small droplets formed. In order for a insoluble drugs

microemulsion to be formed (transient) negative value • Increases bioavailability

of was required, it is recognized that while value of is
• Various routes like tropical, oral and
positive at all times, it is very small and it is offset by
intravenous can be used to deliver the product
the entropic component. The dominant favorable

Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2 26

 Sajal Kumar Jha, et al: Microemulsions- Potential Carrier for Improved Drug Delivery

• Rapid and efficient penetration of the drug The role of surfactant in the formulation of
moiety microemulsion is to lower the interfacial tension which
will ultimately facilitates dispersion process during the
• Helpful in taste masking
preparation of microemulsion and provide a flexible
• Provides protection from hydrolysis and around the droplets. The surfactant should have
oxidation as drug in oil phase in o/w appropriate lipophilic character to provide the correct
microemulsion is not exposed to attack by curvature at the interfacial region. Generally, low HLB
water and air. surfactants are suitable for w/o microemulsion,
whereas high HLB (>12) are suitable for o/w
• Liquid dosage form increases patient
microemulsion. Following are the different surfactants
compliance.
are mainly used for microemulsion-
• Less amount of energy requirement. Polysorbate (Tween 80 and Tween 20), Lauromacrogol
300, Lecithins, Decyl polyglucoside (Labrafil M 1944
Components of Microemulsion System LS), Polyglyceryl-6-dioleate (Plurol Oleique), Dioctyl
A large number of oils and surfactant are available but sodium sulfosuccinate (Aersol OT), PEG-8
their use in the microemulsion formulation is caprylic/capril glyceride (Labrasol).
restricted due to their toxicity, irritation potential and Cosurfactants
unclear mechanism of action. Oils and surfactant Cosurfactants are mainly used in microemulsion
which will be used for the formulation of formulation for following reasons:
microemulsion should be biocompatible, non-toxic, • They allow the interfacial film sufficient
clinically acceptable, and use emulsifiers in an flexible to take up different curvatures
appropriate concentration range that will result in mild required to form microemulsion over a wide
and non-aggressive microemulsion. The emphasis is, range of composition.
excipients should be generally regarded as safe
(GRAS)[3]. • Short to medium chain length alcohols

Oil Phase (C3-C8) reduce the interfacial tension and

The oil component influences curvature by its ability to increase the fluidity of the interface.
penetrate and swell the tail group region of the • Surfactant having HLB greater than
surfactant monolayer. As compare to long chain 20 often require the presence of cosurfactant
alkanes, short chain oil penetrate the tail group region to reduce their effective HLB to a value within
to a greater extent and resulting in increased negative the range required for microemulsion
curvature (and reduced effective HLB). Following are formulation.
the different oil are mainly used for the formulation of
Following are the different cosurfactant mainly used
microemulsion:
for microemulsion:
• Saturated fatty acid-lauric acid, myristic acid, sorbitan monoleate, sorbitan monosterate, propylene
capric acid glycol, propylene glycol monocaprylate (Capryol 90),

• Unsaturated fatty acid-oleic acid, linoleic acid, 2-(2-ethoxyethoxy)ethanol (Transcutol) and ethanol.
linolenic acid Preparation of Microemulsion
Following are the different methods are used for the
• Fatty acid ester-ethyl or methyl esters of
preparation of microemulsion[3]:
lauric, myristic and oleic acid.
-Phase titration method
The main criterion for the selection of oil is that -Phase inversion method
the drug should have high solubility in it. This will
Phase titration method
minimize the volume of the formulation to deliver the
Microemulsions are prepared by the spontaneous
therapeutic dose of the drug in an encapsulated form.
emulsification method (phase titration method) and
Surfactants
27 Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2
 Sajal Kumar Jha, et al: Microemulsions- Potential Carrier for Improved Drug Delivery

can be portrayed with the help of phase diagram. As be achieved by changing the temperature of the
quaternary phase diagram (four component system) is system, forcing a transition from an o/w
time consuming and difficult to interpret, pseudo microemulsion at low temperature to a w/o
ternary phase diagram is constructed to find out the microemulsion at higher temperatures (transitional
different zones including microemulsion zone, in phase inversion). During cooling, the system crosses a
which each corner of the diagram represents 100% of point zero spontaneous curvature and minimal surface
the particular components. Pseudo-ternary phase tension, promoting the formation of finely dispersed
diagrams of oil, water, and co-surfactant/surfactants oil droplets. Apart from temperature, salt
mixtures are constructed at fixed concentration or pH value may also be considered.
cosurfactant/surfactant weight ratios. Phase diagrams A transition in the radius of curvature can be obtained
are obtained by mixing of the ingredients, which shall by changing the water volume fraction. Initially water
be pre-weighed into glass vials and titrated with water droplets are formed in a continuous oil phase by
and stirred well at room temperature. Formation of successively adding water into oil. Increasing the water
monophasic/ biphasic system is confirmed by visual volume fraction changes the spontaneous curvature of
inspection. In case turbidity appears followed by a the surfactant from initially stabilizing a w/o
phase separation, the samples shall be considered as microemulsion to an o/w microemulsion at the
biphasic. In case monophasic, clear and transparent inversion locus.
mixtures are visualized after stirring; the samples shall Evaluation of Microemulsion
be marked as points in the phase diagram. The area Evaluation of microemulsion is very challenging due to
covered by these points is considered as the their small droplet size with fluctuating boundaries
microemulsion region of existence. and complex structure. Following are different
parameters are evaluated for the microemulsion:
Phase behavior study can be used to differentiate
microemulsions from liquid crystals and coarse
emulsions. This is carried out by visual observations,
phase contrast microscope and freeze fracture
transmission electron microscope.
Viscosity measurement can indicate the presence
of rod-like or worm-like reverse micelles. Conductivity
measurement provide means of determining whether a
microemulsion is oil-continuous or water-continuous,

Fig. A ternary phase diagram portraying various as well as providing means of monitoring percolation

structures a) o/w microemulsion; b) w/o or phase inversion phenomena. Dielectric

microemulsion; c) bicontinuous microemulsion; d) measurements are powerful means probing both the

and e) various dispersions. structural and dynamic features of microemulsion

systems.

Phase inversion method Isotropic nature of microemulsion and their optical

Phase inversion of microemulsion is carried out upon clarity can be evaluated by spectroscopic techniques,

addition of excess of the dispersed phase or in particularly by comparison to conventional

response to temperature. During phase inversion microemulsions. Pulse field gradient NMR has been

drastic physical changes occur including changes in used to measure self-diffusion coefficient of the various

particle size that can ultimately affect drug release both components and yield information on the mobility and

in vitro and in vivo. For non-ionic surfactants, this can microenvironment. Scatter methods such as dynamic

Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2 28

 Sajal Kumar Jha, et al: Microemulsions- Potential Carrier for Improved Drug Delivery

and static light scattering (SANS) and small angle X- they exhibit a higher physical stability in plasma than
ray scattering (SAXS) as well as cryo transmission liposome’s or other vehicles and the internal oil phase
electron microscopy have been used to determining is more resistant against drug leaching. Several
particle size of microemulsion. sparingly soluble drugs have been formulated into o/w
microemulsion for parenteral delivery. An alternative
Application of microemulsion in delivery of approach was taken by Von Corsewant and Thoren[9]
drug in which C3-C4 alcohols were replaced with
During the last two decades, microemulsions have parenterally acceptable cosurfactants, polyethylene
been promisingly used as drug delivery system for its glycol (400) / polyethylene glycol (660) 12-
advantages include their thermodynamic stability, hydroxystearate / ethanol, while maintaining a flexible
optical clarity and ease of penetration. The role of surfactant film and spontaneous curvature near zero to
microemulsion as drug delivery system shall be obtain and almost balanced middle phase
discussed here in. microemulsion.
Oral delivery Topical delivery
The development of effective oral delivery systems has Topical administration of drugs can have advantages
always been challenging to researchers because drug over other methods for several reasons, one of which is
efficacy can be restricted by instability or poor the avoidance of hepatic first-pass metabolism of the
solubility in the gastrointestinal fluid. Microemulsions drug and related toxicity effects. Another is the direct
have the potential to enhance the solubilization of delivery and targetability of the drug to affected areas
poorly soluble drugs (particularly BCS class II or class of the skin or eyes. Now a day, there have been a
IV) and overcome the dissolution related number of studies in the area of drug penetration into
bioavailability problems. Due to the presence of polar, the skin. They are able to incorporate both hydrophilic
nonpolar and interfacial domains, hydrophilic drugs (5-flurouracil, apomorphine hydrochloride,
including macromolecules can be encapsulated with diphenhydramine hydrochloride, tetracaine
varying solubility. These systems have been protecting hydrochloride, methotrexate) and lipophilic drugs
the incorporated drugs against oxidation, enzymatic (estradiol, finasteride, ketoprofen, meloxicam,
degradation and enhance membrane permeability. felodipine, triptolide) and enhance their permeation.
Presently, Sandimmune Neoral(R) (Cyclosporine A), Since formation of microemulsion formation requires
Fortovase(R) (Saquinavir), Norvir(R) (Ritonavir) etc. high surfactant concentration, the skin irritation
are the commercially available microemulsion aspect must be considered especially when they are
formulations. Microemulsion formulation can be intended to be applied for a longer period[10].
potentially useful to improve the oral bioavialbility of Ophthalmic delivery
poorly water soluble drugs by enhancing their In conventional ophthalmic dosage forms, water
solubility in gastrointestinal fluid [8]. soluble drugs are delivered in aqueous solution while
Parenteral delivery water insoluble drugs are formulated as suspension or
The formulation of parenteral dosage form of lipophilic ointments. Low corneal bioavailability and lack of
and hydrophilic drugs has proven to be difficult. O/w efficiency in the posterior segment of ocular tissue are
microemulsions are beneficial in the parenteral some of the serious problem of these systems. Recent
delivery of sparingly soluble drugs where the research has been focused on the development of new
administration of suspension is not required. They and more effective delivery systems. Microemulsions
provide a means of obtaining relatively high have emerged as a promising dosage form for ocular
concentration of these drugs which usually requires use.
frequent administration. Other advantages are that

29 Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2

 Sajal Kumar Jha, et al: Microemulsions- Potential Carrier for Improved Drug Delivery

Chloramphenicol, an antibiotic used in the treatment Current and Future Developments

of trachoma and keratitis, in the common eye drops During the last two decades lot of research work has
hydrolyzes easily. Lv et al. investigated the been carried out on microemulsion system for
microemulsion composed of Span 20, Tween 20, providing novel solutions to overcome the problems of
isopropylmyristate and water as potential drug poor aqueous solubility of highly lipophilic drug
delivery systems for eye drops. Chlroamphenicol was compounds and provide reproducible bioavailability.
entrapped in the o/w microemulsion free of alcohol. Industrial point of view, it can be easily scaled up with
The authors revealed that microemulsion formulation considering relative cost of commercial production.
content much lower glycol (main hydrolysis product) Microemulsion can also be used for cosmetic purpose
than that in the commercial eye drops at the end of the and drug targeting. Now a day, researcher work is
accelerated experiments. Thus, a remarkable increase focused on the production of safe, efficient and more
in the chloramphenicol stability was observed in the compatible microemulsion constituents which will
microemulsion formulations[11]. further enhance the utility of this novel delivery
Fialho et al. studied microemulsion based system.
dexamethasone eye drops which showed better
tolerability and higher bioavailability. The formulation References:
showed greater penetration in the eye which allowed 1) Morganti P, Ruocco E, Wolf R, Ruocco V, (2001).

the possibility of decreasing dosing frequency and "Percutaneous absorption and delivery systems." Clin
Dermatol. 19: 489-501.
thereby improve patient compliance [12].
2) Marsh A , Clark B. J., Kevin D, Altria, A review of the
Nasal delivery
background, operating parameters and applications
Recently, microemulsions are being studied as a
of microemulsion liquid chromatography (MELC)
delivery system to enhance uptake of drug through
Journal of Separation Sciences., 28, 2023-2032
nasal mucosa. In addition with mucoadhesive polymer (2005).
helps in prolonging residence time on the mucosa. 3) Talegaonkar S, Azeem A, Ahmad F J, Khar R K,
Lianly et al. investigated the effect of diazepam on the Pathan S A, Khan Z I, Microemulsions: a novel
emergency treatment of status epilepticus. They found approach to enhanced drug delivery Recent Pat Drug

that the nasal absorption of diazepam fairly rapid at 2 Deliv Formul., 2, 238-257 (2008).
4) Schulman J.H, Stoeckenius, W., Prince, L.M, J. Phys.
mg kg-1 dose with maximum drug plasma
Chem., 63, 1677-1680. (1959). Mechanismof
concentration reached within 2-3 min[13].
formation and structure o f microemulsions by
Drug targeting
electron Microscopy .
Drug targeting to the different tissues has evolved as
5) Prince L.M, J. Colloid Interface Sci., 23,165-173
the most desirable goal of drug delivery. By altering (1967). A theory of aqueous emulsions I. Negative
pharmacokinetics and biodistribution of drugs and interfacial tension at the oil/water interface
restricting their action to the targeted tissue increased 6) Aboofazeli R, LawrenceM.J, Int. J. Pharm., 93,161-175
drug efficacy with concomitant reduction of their toxic (1993). Investigations into the formation and

effects can be achieved. Shiokawa et al. reported a characterization of phospholipid microemulsions. I.

novel microemulsion formulation for tumor targeting Pseudo-ternary phase diagrams of systems containing
water-lecithin-alcohol-isopropyl myristate
of lipophilic antitumor antibiotic aclainomycin A
7) Hasse A, Keipert, S, Eur. J. Pharm. Biopharm., 430,
(ACM). They reported that a folate-linked
179-183 (1997). Development and characterization of
microemulsion is feasible for tumour targeted ACM
microemulsions for ocular application
delivery. They also reported that folate modification 8) Hsiu-O Ho, Chih-Chuan Hsiao, Ming-Thau Sheu
with a sufficiently long PEG chain on emulsions is an J.Pharm Sci., 85 ,138-143 (1996). Preparation of
effective way of targeting emulsion to tumour cells[14].

Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2 30

 Sajal Kumar Jha, et al: Microemulsions- Potential Carrier for Improved Drug Delivery

microemulsions using polyglycerol fatty acid esters as

surfactant for the delivery of protein drugs.
9) Corswant C, Thoren P, Engstrom S, J. Pharm. Sci., 87,
200-208(1998). Triglyceride-based microemulsion
for intravenous administration of sparingly soluble
substances.
10) Dreher F, Walde P, Walther P, Wehrli E , J. Control.
Rel., 45,131-140(1997). Interaction of a lecithin
microemulsion gel with human stratum corneum and
its effect on transdermal transport.
11) Lv FF, Li N, Zheng LQ, Tung CH, Eur J Pharm
Biopharm., 62,288-294 (2006). Studies on the
stability of the chloramphenicol in the microemulsion
free of alcohols.
12) Fialho SL, da Silva-Cunha A. Cl i n Ex p e r Op h t h a
lmo l., 2004; 32(6):626-632. New vehicle based on a
microemulsion for topical ocular administration of
dexamethasone.
13) Syamasri Gupta, S.P. Moulik Journal of
Pharmaceutical Sciences 97, 22-45(2008).
Biocompatible microemulsions and their prospective
uses in drug delivery.
14) Shiokawa T, Hattori Y, Kawano K. Clin Cancer Res.,
11, 2018 (2005). Effect of Polyethylene Glycol Linker
Chain Length of Folate-Linked Microemulsions
Loading Aclacinomycin A on Targeting Ability and
Antitumor Effect In vitro and In vivo.

31 Internationale Pharmaceutica Sciencia Apr-Jun 2011 Vol 1 Issue 2

 Journal of Controlled Release 161 (2012) 62–72

Contents lists available at SciVerse ScienceDirect

Journal of Controlled Release

journal homepage: www.elsevier.com/locate/jconrel

Review

Applications of NMR in the characterization of pharmaceutical microemulsions

Rania M. Hathout a,⁎, Timothy J. Woodman b
a
Department of Pharmaceutics and Industrial Pharmacy, Faculty of Pharmacy, Ain Shams University, Cairo, Egypt
b
Department of Pharmacy and Pharmacology, Bath University, UK

a r t i c l e i n f o a b s t r a c t

Article history: Microemulsions have successfully proven themselves as useful vehicles for drugs through the different routes
Received 15 January 2012 of administration because they can confer on drugs greater water solubility and bioavailability. The ability to
Accepted 20 April 2012 understand the structural aspects of these important drug delivery systems is essential to the progress of this
Available online 1 May 2012
science. The use of NMR techniques in pharmaceutical and drug delivery science is increasing especially in
the characterization field. This review demonstrates the major and novel NMR methods and techniques
Keywords:
Microemulsion
used in understanding and characterizing the different microemulsion components, types and structures.
NMR © 2012 Elsevier B.V. All rights reserved.
Pharmaceutical
Diffusion coefficient
Relaxation

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2. NMR applications in characterizing pharmaceutical microemulsion systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1. Chemical shifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1.1. Determination of the locus of molecules in the microemulsion interiors . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1.2. Determination of the state of drug molecules within the microemulsion droplets . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.2. Self-diffusion NMR (SD-NMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.2.1. Determination of the microemulsion type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.2.2. Determination of aggregates in microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.2.3. Determination of the drug locus in microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.2.4. Optimization of fractional composition of microemulsion transdermal vehicles . . . . . . . . . . . . . . . . . . . . . . . . 70
2.3. Relaxation times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.3.1. Determination of the drug locus in the microemulsion using the longitudinal (spin–lattice, T1) relaxation times . . . . . . . . 71
2.3.2. Characterization of microemulsions using the translational (spin–spin, T2) relaxation times . . . . . . . . . . . . . . . . . 71
3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

1. Introduction investigation are determined not only by the chemical structures

and sizes of the individual components, but also by the physical envi-
The need to characterize and quantify the mobility of particular ronment in which they are located and the surrounding molecules
components in multi-component dispersions is becoming a very im- that exert obvious effects [2].
portant issue and has had a profound impact on colloid science [1]. Microemulsions are colloidal dispersions that may be defined as
The physical properties of the multi-component dispersions under stable, low viscosity mixtures of water (or buffer), oil, and surfac-
tant(s) [3–6]. They are self-aggregated systems in which oil and
water are homogeneously mixed due to the presence of amphiphiles.
⁎ Corresponding author at: Department of Pharmaceutics and Industrial Pharmacy,
Faculty of Pharmacy, Ain Shams University, Monazamet ElWehda ElAfriqia St., Cairo,
They are isotropic in nature with typical structural units in the size
Egypt. Tel.: + 20 100 5252919, + 20 02 22912685; fax: + 20 02 24011507. range of 10–100 nm. This small size accounts for their transparent ap-
E-mail address: r_hathout@yahoo.com (R.M. Hathout). pearance. They differ from conventional emulsions not only by their

0168-3659/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconrel.2012.04.032
 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72 63

much smaller structural size and transparency but in particular by swollen micelles and normal micelles is very difficult [5]. However,
their thermodynamic stability, as they allow for a long-lived stabiliza- in between these extremes, the microemulsion components typically
tion of mixed oil/ water systems, which otherwise can never be form non-spherical aggregates, which may become more continuous
achieved. Oral, ocular, pulmonary, nasal, vaginal, topical, transdermal in the phase with the highest volume fraction. These aggregates usu-
and intravenous routes are the main administration routes to which ally change into bicontinuous structures by titration with the phase of
the microemulsion technique can be applied [7–22]. Microemulsions the lowest volume fraction, and through these structures, they rapid-
are becoming increasingly complex through the inclusion of more ly invert to ‘reversed’ aggregates (Fig. 1). This behavior is distinct from
sophisticated surfactants, polymers, biomolecules and inorganic that displayed in emulsions, where sudden inversions of the systems
nanoparticles [23]. Therefore, several different internal structures from oil-in-water to water-in-oil and vice versa usually occur [24].
can be observed within the same microemulsion region. The micro- They are dynamic systems in which the interface is continuously
emulsion structure is greatly influenced by the physicochemical and spontaneously fluctuating [25]. Many microemulsions, com-
properties and respective ratios of its components. Microemulsion monly those based on ionic surfactants, also contain a cosurfactant,
structures may vary from emulsion like normal or reverse swollen the role of which is to decrease the surface tension between water
micelles, to aggregates of w/o or o/w types, to bicontinuous struc- and oil to very small values approaching zero; sufficient for micro-
tures, depending on the ratio of the constituents (Fig. 1). In very di- emulsion formation. Usually short-chain and medium-chain alco-
lute systems, with only a small percentage of oil or water phase, hols, like ethanol, glycerol, propylene glycol, butanol, pentanol,
microemulsion structures may approach regular or reverse ‘swollen hexanol or Transcutol [26,27] are used. As the size of micro-
micelle’ droplet-like shapes where the distinction between these emulsion aggregates is smaller than the wavelength of visible

Fig. 1. Microemulsion structures (oil is represented in grey color, water is represented in white).
Taken with permission from Ref. [24].
64 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72

light, and since the structures can be altered by very slight changes is also capable of discriminating between different sites within an
in composition and temperature, the direct examination of micro- individual molecule, ensuring that a plethora of information is avail-
emulsion structures (e.g. by light microscopy) is very difficult able. There are also many other recognized advantages of NMR; (1)
[28]. The efficient use of these colloidal dispersions in scientific and in- there is no need for any particular sample preparation, (2) the sam-
dustrial applications is crucially dependent on the understanding of ples can be large enough to exclude effects connected to the wall of
their microstructure [29]. In this context, the detailed understanding the sample holder, (3) impurities usually have negligible effects, (4)
of such complex systems requires increasingly more refined, compre- maintaining a homogeneous, stable, and accurate temperature in
hensive and sensitive characterization. To this end, sophisticated tech- the whole sample volume is usually simple, (5) the sample can have
nologies and indirect measurement techniques are often employed to any viscosity and consistency and can be turbid, and finally,
obtain information about their internal structure and spatial orientation (6) using specialized equipment, the sample can also be subjected
[30]. These techniques, which include scattering methods [3,31–33], to shear or to an electric field [41,42]. In this review we provide an
conductivity [34], electron microscopy [35], differential scanning cal- overview of the wide use of NMR in characterizing microemulsion
orimetry (DSC) [36], small angle X-ray scattering [29,37] and fluores- systems.
cence quenching [38,39], have been used to clarify the type and the
structure of microemulsions. Despite their utility, all these methods 2. NMR applications in characterizing pharmaceutical microemulsion
have limitations. systems
More recently, the use of NMR techniques, which were classically
used to characterize micellar systems, has attracted attention for the 2.1. Chemical shifts
characterization of colloids of pharmaceutical interest e.g. the micro-
emulsions [40]. To gain additional information about the behavior of 2.1.1. Determination of the locus of molecules in the microemulsion
these complicated systems, we report here different results of NMR interiors
studies. These studies show that in certain circumstances a great Proton chemical shifts originate from differences in their sur-
deal of information may be obtained using NMR techniques. NMR rounding electronic environments, from both interactions within
can be readily applied to several of the most common nuclei found the molecule and from external molecular contacts. Therefore, they
in most components of a microemulsion, i.e. 1H, 13C and others, and are prone to local interactions and may serve as a sensitive probe to

Fig. 2. Chemical structures of the different components for the microemulsion system in Ref. [44]. The most sensitive groups to chloramphenicol addition are highlighted.
 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72 65

Table 1
1
H-NMR chemical shifts of different functional groups of drug-free and drug-loaded
testosterone (1% w/v) oleic acid/Tween 20®/Transcutol®/water microemulsions (15,
30.8, 30.8 and 23, %w/w respectively) [Ref. [46]—supplementary material].

Functional δ (ppm) Δδ
group 0 1 (δ1 − δ0)
Drug-free ME (δ ) ME with 1% (w/v) drug δ )

CH3 0.855 0.888 0.033

CH2 1.162 1.191 0.029
CH2CH2CO 1.543 1.571 0.028
CH2CH 1.989 2.02 0.031
CH2CO 2.208 2.23 0.022
OCH2CH2O 3.646 3.668 0.022
CH2OCO 4.171 4.190 0.019
OH 4.653 4.652 0.001
CH 5.287 5.318 0.031

Fig. 3. Chemical structures of the different components of the microemulsion system in

Ref. [45]. The 1H-NMR studied groups of the surfactant are pointed out.
Interestingly, in a previous study (although not pharmaceutical),
Shi et al., 2003 [45] reported on the solubilization of styrene in sodi-
changes taking place in their vicinity [40]. In this context, 1H-NMR um dodecyl sulfonate (DS)/n-butanol/water O/W microemulsion
spectroscopy has been used to locate the possible site of drug mole- consisting of 75%, (w/w) water and a ratio of DS:n-butanol = 7:3
cules in the microemulsion system using oleic acid (oil), Tween80 (w/w) using 1H NMR. Fig. 3 demonstrates the chemical structures of
(surfactant), ethanol (cosurfactant) and phosphate buffer (pH7.4) the studied microemulsion components and points out the surfactant
(aqueous phase). Chemical shifts of the surfactant in the presence of groups that were focused on in this work.
the drug (isoniazid, 0.5 and 1 wt.%) were determined employing Fig. 4 shows the dependence of chemical shifts (δ) or their
D2O as internal locking agent. The changes in the chemical shifts of changes (δ − δ0) for the α-CH2, (CH2)10, and ω-CH3 groups of DS mol-
different functional groups were very small. This limited the possibil- ecules on the mole fraction c (c = mole styrene / mole styrene and DS)
ity of the presence of the drug near the surfactant molecules, al- of styrene in the microemulsion. As can be seen from Fig. 4, the chem-
though at higher concentrations of the drug some changes in the ical shifts of the α-CH2, (CH2)10, and ω-CH3 groups were proportional
chemical shifts of oxyethylene groups were seen and it was conclud- to the mole fraction for c b 0.312. When c was higher than 0.312, the
ed that most of the drug was present in the dispersion medium with chemical shifts tended to a constant value. It was then deduced that
fairly small amounts near the interface [43]. the styrene molecules increase their partitioning into the palisade
In 2006, Lv et al. [44] determined that after the addition of chlor- layer until c surpasses 0.312. Since the 1H nuclei of α-CH2, (CH2)10,
amphenicol to a microemulsion consisting of isopropyl myristate as and ω-CH3 groups of DS molecules are greatly affected by the circula-
the oily phase, Span 20 and Tween 20 as the surfactant/cosurfactant tion effect of the benzene rings of styrene, their chemical shifts moved
mixture and deuterated water as the aqueous phase, that the chemi- to higher field with increasing styrene. The solubilization of styrene
cal shifts of the protons of methylene groups near the ether bond has reached saturation in the palisade layer at c = 0.312, and further
(α2-CH2) were the most sensitive to the addition of the drug styrene went into the inner cores of the microemulsion droplets [45].
(~0.2 ppm) followed by the chemical shifts of the protons of the oxy- Adopting the same ideas, Hathout et al. [46] confirmed the pres-
ethylene groups (CH2CH2O)n (~0.15 ppm). The changes in the chem- ence of testosterone inside the oily domains of bicontinuous micro-
ical shifts of other functional groups, such as ω-CH3, (CH2)n, β1-CH2 emulsions consisting of oleic acid/Tween 20®(Polyoxyethylene 20
and α1-CH2 were all very small (b0.02 ppm). It was then concluded sorbitan monolaurate)/Transcutol®(Diethylene glycol monoethyl
that the chloramphenicol molecules were solubilized neither in the ether)/water, with slight amounts near the interfacial film formed
palisade layer nor in the oily core of the o/w microemulsions drops, of the surfactant and the cosurfactant. The proton change in the
but in the hydrophilic shells of the microemulsion drops, which are chemical shifts of oleic acid distinct functional groups (CH3, CH2CH
composed of many oxyethylene groups. Fig. 2 shows the structure and CH) after drug incorporation (1%) in a formulation consisting of
of the different components of the studied microemulsion system. 15.4, 30.8, 30.8 and 23, % w/w of the components respectively were

Table 2
13
C-NMR chemical shifts of different functional groups of drug-free and drug-loaded
testosterone (1% w/v) oleic acid/Tween 20®/Transcutol®/water microemulsions
(15.4, 30.8, 30.8 and 23, %w/w respectively) [Ref. [46]—supplementary material].

Functional δ (ppm) Δδ
group (δ1 − δ0)
Drug-free ME (δ0) ME with 1% drug (δ1)

CH3 14.117 14.806 0.689

CH2 22.896 22.926 0.073
29.477 29.513 0.036
29.631 29.660 0.029
30.004 30.034 0.030
CH2CHCO 25.131 25.183 0.052
CH2C=C 27.359 27.396 0.037
CH3CH2CH2 32.188 32.225 0.037
CH2CO 34.086 34.130 0.044
C=C 129.725 129.871 0.146
COOH 175.782 175.723 − 0.059
C–O 70.067 70.126 0.029
72.375 72.441 0.066
Fig. 4. Changes of proton chemical shifts of DS due to styrene presence.
C–OH 60.863 60.914 0.051
Taken with permission from Ref. [45].
66 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72

Fig. 5. Components structures and 1H-NMR assignment of peaks for the microemulsion oleic acid/Tween 20®/Transcutol/water (15.4, 30.8, 30.8, and 23, %w/w respectively).

the highest. These results were confirmed using 13C NMR, where C=C NMR peaks and this means that differences in linewidth can provide
and −CH3 carbons showed the highest difference in chemical shifts. information on the state of drug molecules within the microemulsion
The results are collated in Tables 1 and 2. Figs. 5 and 6 depict the droplets. Itoh et al. [47] incorporated a lipid lowering drug viz. N-[2-
structures of the different microemulsion components together with (3,5-di-tert-butyl-4-hydroxyphenethyl)-4,6-difluorophenyl]-N′‐[4-(N-
the 1H NMR and 13C NMR assignment of peaks for the unloaded oleic benzylpiperidyl)] urea (N-4472) in a microemulsion system consisting
acid/Tween 20®/Transcutol®/water investigated microemulsion. of Gelucire® 44/14 (mixture of 30% glycerol ester and 70% PEG-ester
These results explained the indifferent physical behavior of the studied with fatty acids) and polyoxyethylene (60) hydrogenated castor oil
microemulsion system regarding pH, conductivity measurements and (HCO-60)® and sodium dodecyl sulfate (SDS). In a complementary
rheological properties after drug incorporation. study the authors studied the complex formed between N-4472 and
L-ascorbic acid (VC). The chemical structures of N-4472 and VC are pro-
2.1.2. Determination of the state of drug molecules within the microemulsion vided in Fig. 7. They used the line-widths obtained from 1H-NMR peaks
droplets at 6.5, 7 and 7.5 ppm, corresponding to aromatic rings I, II and III respec-
The linewidth of a given NMR signal is often related to the mobil- tively (Fig. 7), to propose two modes of molecular states : a molecular
ity of the nuclei. Typically slow molecular tumbling results in broad dispersed state and a self‐associated one [48]: a molecularly dispersed

13
Fig. 6. Components structures and C-NMR assignment of peaks for the microemulsion oleic acid/Tween 20®/Transcutol/water (15.4, 30.8, 30.8, and 23, %w/w respectively).
 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72 67

case, for a bicontinuous structure, with both solvents forming do-

mains with no confined structures that extend over macroscopic dis-
tances, both solvents have diffusion coefficients of the same order of
magnitude as that of the neat solvent [29]. Accordingly, a good corre-
lation has been demonstrated between the characteristics of the var-
ious microemulsion structures and the molecular mobility of the oil,
water and surfactant(s) phases, respectively [49–53]. Table 3 summa-
rizes the cases obtained from self-diffusion results of the micro-
emulsion components.
Translational motion in colloidal dispersions, either through diffu-
sion or fluid flow is most effectively studied by means of pulsed gra-
Fig. 7. Chemical structures of (a) N-4472 and (b) VC. The interacting sites are
highlighted.
dient spin‐echo (PGSE) NMR [54]. Self-diffusion coefficients are
usually measured using a gradient spin-echo pulse sequence, modu-
lating the amplitude of fixed-length gradient pulses, and diffusion of
state was recorded in the microemulsion vehicle resulting in narrow a molecule is directly related to the decay of the peak heights in the
NMR signals and a self-associated state was observed in an aqueous recorded 1H-NMR spectra, according to the following equation [55]:
state resulting in slow tumbling of the molecules and consequently     
broad NMR resonances. Ig 2 2 2 d
ln ¼ − γ d G Δ− D
Io 3
2.2. Self-diffusion NMR (SD-NMR)
Where Ig and Io are intensities of NMR signal in presence and ab-
2.2.1. Determination of the microemulsion type sence of field gradient pulses, respectively, γ is the gyromagnetic con-
Self-diffusion is a term used to describe the random movements of stant for the hydrogen atom, d is the duration of the z gradient pulse,
a molecule in an isotropic solution without thermal gradients. G is the gradient strength, △ the time interval between the gradient
According to the Stoke-Einstein equation (D = kT / (6πηr), where pulses and finally D is the self-diffusion coefficient.
D is the self-diffusion coefficient, η the viscosity of the medium and SD-NMR, sometimes referred to as “Diffusion-ordered nuclear
r the (macro-) molecular radius), the self-diffusion of a (macro-) mol- magnetic resonance (NMR) spectroscopy (DOSY)” is a powerful tech-
ecule is dependent on the molecular size and shape and on the nique that enables the discrimination of molecular species by the use
molecule-solvent interactions. Stilbs and co-workers [49–53], have of their apparent translational diffusion coefficients. It is considered
shown that these rules can also be applied to microemulsion systems. as a useful tool for the study of complex mixtures [54]. A DOSY exper-
The self-diffusion of constituents in a microemulsion, relative to those iment performed on a mixture results in a two dimensional spectrum
of the neat components, can be related to the degree of freedom and which displays conventional NMR spectral information in the direct
molecular interaction of the phases, and can vary by various magni- dimension and the diffusion coefficients in the indirect dimension
tudes depending on the degree of structural restriction. [56]. The DOSY spectra of microemulsions consisting of isopropyl
Molecular self-diffusion is sensitive to friction, obstruction, and myristate (IPM) as the oil, AOT and Tween85 as the surfactants and
solvation, but in particular, any confinement into closed domains dra- water were obtained by Liu et al., 2009 [57]. The DOSY spectra are
matically reduces the long range molecular displacement that is stud- presented in Fig. 8 with chemical shifts on the horizontal axis and dif-
ied through these investigations. For molecules in closed domains, fusion on the vertical axis. The diffusion axis was logarithmically
e.g. droplets, diffusion is determined by droplet diffusion, i.e., by treated and self-diffusion coefficient values were expressed in
droplet size, droplet–droplet and droplet–solvent interaction. Thus, m 2 s − 1. Peaks were picked from the 2D DOSY spectra and diffusion
for an oil-in-water droplet structure with water as the continuous coefficients were determined from the peak summits.
medium, the diffusion coefficient of an oil molecule (dispersed The diffusion coefficient of each pure compound was then calcu-
phase) is typically 2–3 orders of magnitude lower than that of a lated as the mean value of each observed coefficient. Diffusion coeffi-
neat solvent molecule, while diffusion of a water molecule is only cients obtained were those found for the different compounds
slightly retarded in comparison with its neat solvent coefficient. On characteristic peaks viz. methylene signal of IPM at 2.4 ppm, the pro-
the other hand, for a structure of water droplets in oil, a reverse situ- ton signal near sulfonic acid of AOT at 3.8 ppm and the proton signal
ation can be predicted. In both cases, surfactant diffusion is obviously of water at 4.7 ppm. The diffusion coefficient of each component in
slow and roughly approaches that of the dispersed phase. In another the microemulsion was examined and recorded. To evaluate the
self-diffusion data in terms of microstructure, the relative diffusion
coefficients, D/Do, of each component in the nano droplets of the
Table 3
microemulsion are calculated by relating its diffusion coefficient in
Self-diffusion results usually recorded for the different microemulsion structures.
the microemulsion vicinity to the diffusion coefficient in the pure so-
Case Microemulsion type Self-diffusion results lution. Accordingly, relative diffusion coefficients of the water and
1 Micellar (o/w) droplet Water diffusion will be rapid, in the same order of the oil were obtained by dividing water (Dw) and oil (Do) diffusion
structure magnitude as that of neat water. Oil and coefficients in the microemulsion by the diffusion coefficient of
surfactant diffusion will be slow and within the water (Dowater) and oil in the neat phase (Dooil). As indicated in Fig. 9,
same order of magnitude.
at b20 wt.% aqueous phase, the microemulsion was found to contain
2 Inverted micellar (w/o) Oil diffusion will be rapid, in the same order of
droplet structure magnitude as that of neat oil. Water and discrete water droplets dispersed in the oil phase, since the relative
surfactant diffusion will be slow and within the diffusion coefficients of water and IPM differ by one order of magni-
same order of magnitude. tude. This suggested a W/O droplet microemulsion. In the micro-
3 Bicontinuous structures Both water and oil diffusion will be rapid, only
emulsions containing 20–40 wt.% aqueous phase, water showed
slightly slower than those of the neat liquids.
Surfactant diffusion will be low due to the similar diffusion coefficient for the oil, the diffusion of water and
constitution of the interfacial film, but slightly IPM was 2.0–4.8 × 10 − 12 m 2 s − 1, 1.1–1.2 × 10 − 12 m 2 s − 1, respective-
higher than cases 1 and 2 as the interface is ly. Therefore a bicontinuous microstructure was suggested for these
fluctuating and less restricted than in the micellar formulations, as the relative diffusion coefficients of water and IPM
shapes
were of the same order of magnitude.
68 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72

Fig. 8. DOSY spectra of microemulsions with fixed surfactant concentration at 20.0 wt.% (the weight ratio of AOT to Tween 85 was 2): (a) ME1: 10.0 wt.% water; (b) ME2: 20.0 wt.%
water; (c) ME3:30.0 wt.% water; (d) ME4: 40.0 wt.% water. The DOSY spectra are presented with chemical shifts on the horizontal axis and diffusion coefficients (log scale) on the ver-
tical axis.
Taken with permission from Ref. [57].

It is now well documented that if the D/Do values of water and oil capric triglycerides as the oily phase [59], experiments were carried
differ by more than one order of magnitude (10 − 10 vs 10 − 9 m 2 s − 1 out by varying the gradient strength and keeping all other timing pa-
for example), discrete particles of the slowly diffusing component rameters constant. Δ and d were used as 100 ms and 8 ms, respective-
are suggested, whereas if the D/Do values of water and oil are of the ly. G varied from 1.7 to 32.3 G cm − 1 in 32 steps. The very large
same order of magnitude, a bicontinuous structure is implied [58]. variations that were observed in the relative diffusion coefficients of
Fast diffusion (>10 − 9 m 2 s − 1) is characteristic of free molecules in water and oil on varying their weight fractions indicated micro struc-
solution while a small diffusion coefficient (b10 − 12 m 2 s − 1) suggests tural transitions from O/W structure at low oil to water ratio to W/O
the presence of macromolecules or bound (or restricted immobilized) at higher ratios passing through bicontinuous structures.
molecules. In another study, self‐diffusion measurements were conducted on
In a similar study comprising the microemulsion system con- a microemulsion consisting of 22% (w/w) polyoxyethylene (4) lauryl
sisting of water as the aqueous phase, sucrose laurate/ethoxylated ether (Brij-30), 22% (w/w) ethanol, 45% (w/w) eucalyptus oil and 10%
mono-di-glycerides as mixed non-ionic surfactants and capylic– (w/w) distilled water [60]. The obtained D/Do values for water and oil
 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72 69

O/W microstructure occurred. In the fourth region, a discrete O/W

microstructure was found [68].
In another study [69], two pseudo-ternary microemulsion systems
comprising isopropyl myristate, soybean lecithin, water, ethanol and
either decyl glucoside or capryl–caprylyl glucoside as surfactant
were investigated for their potential to form microemulsion tem-
plates to produce nanoparticles as drug delivery vehicles for proteins
and peptides. All of the microemulsion and nanoparticle ingredients
used were pharmaceutically acceptable and biocompatible. Phase di-
agrams were established and characterized using polarizing light mi-
croscopy, viscosity, conductivity, electron microscopy, differential
Fig. 9. Relative diffusion coefficients of water (o) and oil (Δ) in the microemulsion sys- scanning calorimetry and self-diffusion NMR. Self-diffusion coeffi-
tem consisting of isopropyl myristate, Tween85, AOT and water.
cients indicated that the surfactant mixtures in the samples diffused
Taken with permission from Ref. [57].
almost at the same rate as in their neat state. Accordingly the forma-
tion of solution-type microemulsions (which is merely a molecular
were 0.21 and 0.32, respectively. Hence, a bicontinuous structure dispersion of all the components) was concluded. Despite the absence
with both solvents forming domains that extend over macroscopic di- of “droplets” or swollen micelles in this microemulsion structures, it
mensions, was suggested [61]. In addition, the polar domain was also has been shown that nanoparticles can be prepared by interfacial po-
defined by the cosurfactant ethanol, which exhibited a self-diffusion lymerization of such systems [62].
coefficient value occurring in the same range (0.37). Recently, Hathout et al. confirmed the presence of bicontinuous
In the same context, different structural types of microemulsions were microstructures in all the investigated microemulsions across the di-
characterised and used for the preparation of poly(alkylcyanoacrylate) lution line starting at 20% (w/w) oil and 80% (w/w) surfactant–
nanoparticles by interfacial polymerization [62,63]. A phase diagram cosurfactant mixture for the microemulsion system oleic acid/
comprising ethyl oleate, polyoxyethylene 20 sorbitan mono-oleate/ Tween 20®/Transcutol®/water, as D/Do values for water and oil
sorbitan mono laurate ,butanol and water was prepared. Solution were always at the same order of magnitude [46].
type, bicontinuous and water in oil droplets were characterised. The In summary, self-diffusion coefficients are often used to distin-
entrapment of the model protein ovalbumin within these particles guish between W/O, bicontinuous, and O/W microemulsions. In a
was up to 95%, depending on the amount of monomer used for poly- typical O/W microemulsion, the sequence will be Do ≪ Dw. In a typi-
merization and the type of microemulsion used as the polymerization cal W/O microemulsion, the order will be Do ≫ Dw while in the
template. Surprisingly, the particles formed with similar characteris- bicontinuous phase, both Do and Dw are similar (in the order of
tics from templates having different microstructure. It was concluded 10 − 9 m 2 s − 1) [58].
then that the dynamics within the template (stirring, viscosity) or in-
deed interfacial phenomena relating to the solid–liquid interface ap-
pear to be more important for the determination of nanoparticle 2.2.2. Determination of aggregates in microemulsions
morphology and characteristics than the microstructure of the tem- NMR techniques can provide information about structure com-
plate system. plexity and aggregate size and shape in microemulsions [50,61,70].
Boonme et al. [64] used a variety of techniques to determine the In a recent study, the analysis of 2H and 13C relaxation data demon-
type and structure of microemulsion colloidal samples formed in iso- strated that the addition of certain alcohols and oils to micellar solu-
propyl palmitate (IPP)/water/Brij 97:1-butanol (2:1) system. At a tions decreases the degree of aggregation [71]. Furthermore, it was
constant concentration of the surfactant (Brij 97)/cosurfactant (1-bu- suggested that aggregates which are predominantly water or oil con-
tanol) mixture, the microemulsions of various ratios of water and oil tinuous will typically display self-diffusion coefficients like those
were produced and characterized. Several techniques including con- recorded for the O/W and W/O droplets, respectively, but with slower
ductivity, viscosity, self-diffusion NMR, DSC, and Cryo-field emission diffusion of the `continuous' phase and faster diffusion of the `dis-
scanning electron microscopy (Cryo-FESEM) were combined to in- persed' phase, approaching the diffusion coefficients recorded for
vestigate the types of microemulsions and transitions between the bicontinuous structures. Simple, non-structured solutions will
them. The self-diffusion NMR measurements determined precisely display rapid values and basically provide different self-diffusion co-
the transition from water/oil to oil/water microemulsion type to efficients of all constituents. Mixtures of different structures may co-
occur at a point corresponding to a water concentration of 35% exist (e.g., droplets and aggregates). This occurs due to the fluctuation
(wt/wt). of surfactant between the interfacial film and the aqueous–oil phase
Spernath et al. [65] also used a microemulsion system comprising or because the oil or water molecules may be partly integrated into
R(+) limonene as the oily phase and water as the aqueous phase [66] the surfactant film. In these situations, self-diffusion coefficients will
for the solubilization of a very important nutraceutical; lycopene. Ly-
copene is an essential carotenoid that gives the characteristic red Table 4
color to tomatoes. It is a lipophilic compound that is usually insoluble Self-diffusion coefficients for oleic acid, Tween20, water and testosterone in different
in water and in most food grade oils. Clinical studies have indicated water composition microemulsion formulations (MEs) composed of oleic acid,
Tween20, Transcutol and water.
the important role of lycopene in reducing risk factors for chronic dis-
eases such as cancer, coronary heart disease, and aging [67]. The au- Water percentage Da (H2O) Da Da Da
thors [65] used PGSE-NMR to correlate the changes in solubilization in the MEs (m2s− 1)b (oleic acid) (Tween 20) (Testosterone)
investigated (%) (m2s− 1)b (m2s− 1)b (m2s− 1)b
capacity to microstructure transitions as the water concentration pro-
gressively increased, and to study the effect of lycopene on these mi- 5 1.18 0.24 0.63 0.24
10 1.37 0.16 0.6 0.22
crostructure transition points. Four different regions were indicated
15 1.54 0.16 0.61 0.23
in the solubilization capacity indicative of the microstructure transi- 18 1.75 0.15 0.63 0.19
tion along the dilution line. The first region indicated the formation 20 1.73 0.12 0.673 0.24
of W/O microstructure. The second region indicated the transition 23 1.81 0.11 0.68 0.23
from W/O microstructure to a bicontinuous microemulsion. In the a
D represents the self-diffusion coefficient value.
third region, a transition from a bicontinuous microemulsion to an b
All D values are multiplied by a factor of 1010.
70 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72

results]. The self-diffusion coefficients of this drug in the micro-

emulsion formulations along the dilution line starting at 20% w/
w oleic acid and 80% w/w surfactant/cosurfactant mixture, having dif-
ferent water compositions (from 5% to 23%, w/w) were closely related
to the self-diffusion coefficients of the oil (oleic acid) (Table 4). Fig. 10
shows a schematic diagram demonstrating the locus of testosterone in
the oily domains of the bicontinuous structures of the microemulsions
under investigation.

2.2.4. Optimization of fractional composition of microemulsion transder-

mal vehicles
Self-diffusion coefficients of venpocetine from different micro-
emulsion formulations were determined in order to investigate the in-
fluence of their different composition (with equal drug concentration)
on the drug transdermal delivery. A strong correlation (0.9034) was
obtained between the self-diffusion coefficients obtained and the trans-
dermal permeation rates across human skin. It was then suggested that
the fractional composition of a given microemulsion vehicle can be op-
timized using the values of the self-diffusion coefficients with no need
Fig. 10. A schematic diagram demonstrating the locus of testosterone in the bicontinuous for the transdermal permeation experiments which can sometimes be
structures of the microemulsion system oleic acid/Tween 20®/Transcutol®/water. more difficult to conduct [74].
Nevertheless, a strong correlation was observed between the trans-
dermal permeation through porcine skin of finasteride hormone in an
be a weighted average of the diffusion within the different structures, eucalyptus oil containing microemulsion system and its self-diffusion
and interpretation of the studies become more complex [25,30,72]. coefficient obtained by NMR [60]. An increased solubility in the apolar
phase of the microemulsion, which can be deduced from a lower relative
2.2.3. Determination of the drug locus in microemulsions diffusion coefficient for this drug in the NMR experiments, may be the
In pharmaceutical microemulsions, information about drug solu- reason for the facilitated transport of finasteride. Similar attempts to
bility and disposition within the phases is also reflected in the self- correlate the drug release with NMR derived self-diffusion coefficients
diffusion coefficient of the drug relative to the other respective have also been previously published for Labrasol/Plurolisostearique/
ingredients. isostearylicisostearate/water microemulsions by Kreilgaard et al. [30].
Fanun [73] studied a microemulsion system containing diclofenac
sodium which was formed with water as the aqueous phase, sucrose 2.3. Relaxation times
laurate + ethoxylated mono-di-glyceride as a mixed non-ionic surfac-
tant mixture and isopropyl myristate + ethanol as the oily phase. The return to equilibrium of net magnetization is called relaxation,
Viscosity, conductivity and NMR measurements confirmed that a pro- and is comprised of two mechanisms, spin–lattice relaxation (T1) and
gressive transformation of the water in-oil to bicontinuous and inver- spin–spin relaxation (T2). T1 characterizes the rate at which longitu-
sion to oil-in-water microemulsions occurred upon dilution with dinal magnetization returns to thermal equilibrium, and takes place
water. The relative diffusion coefficients of diclofenac sodium (D/Do) via a variety of interactions of the nuclei with their surrounding envi-
was very low (below 0.1) for water contents below 20 wt.% and ronment, known as the lattice, which is essentially the rest of the uni-
above 70 wt.%. For water content equal 50 wt.% the diclofenac relative verse. Longitudinal magnetization is dissipated to the lattice as
diffusion coefficient reached its maximum value of 0.18. These values rotational and vibrational energy, ultimately increasing the lattice
indicated that the locus of diclofenac sodium is strongly bound to the temperature.
interface for water contents below 20 wt.% and above 70 wt.%. How- Spin–spin relaxation (T2) is the return of transverse magnetization
ever, for water contents between 30 and 60 wt.%, more space was to equilibrium, and involves only other nuclei. T2 relaxation is a com-
available to the diclofenac sodium to move, indicating less binding plex phenomenon, but is essentially the process of decoherence of the
to water. transverse nuclear spin magnetization. Two main mechanisms play a
In another study by Hua et al. [74] the self-diffusion of the drug part. Firstly, random fluctuations of the local magnetic field lead to
venpocetine was very slow compared to its pure values indicating small instantaneous changes in the precession frequency of individu-
its association with the inner phase of the investigated system (the al spins. Secondly, spin–spin coupling leads to small differences in
oily phase in this case) due to its low solubility in water. precession rates for the coupled nuclei, again leading to decoherence.
Work by Hathout et al. showed that testosterone associates with Typically, but not always, T2 times are shorter than T1. Relaxation
the lipophilic phase of microemulsion formulations containing oleic times are usually indicative of molecular motion and in this context
acid, Tween20 and Transcutol (ratio 1:1) and water [unpublished can provide information on microemulsion structure [42].

Table 5
Summary of NMR applications in the characterization of microemulsions.

Characterisation mode NMR application used Refs.

Tracing the microemulsion formation Translational (spin–spin, T2) relaxational times [1]
Determination of the microemulsion type Self-diffusion NMR [46,49–53,56,59,60,62–65,69]
Determination of the locus of molecules in the microemulsion interiors. Chemical shifts [43–46,73,74,30]
Self-diffusion NMR
longitudinal (spin–lattice, T1) relaxation times
Determination of the state of drug molecules within the microemulsion droplets Chemical shifts [48]
Determination of aggregation in microemulsions Self-diffusion NMR [50,61,70,71]
Optimization of fractional composition of microemulsion Self-diffusion NMR [30,60,74]
 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72 71

2.3.1. Determination of the drug locus in the microemulsion using the systems. These transitions can be investigated very elegantly using
longitudinal (spin–lattice, T1) relaxation times the stated NMR techniques such as the SD-NMR. It is our contention
Relaxation times are usually related via correlation times to the that such a level of information is essential for efficient optimization.
mobility of a molecule and may thus be used to assess the site of
incorporation of the drug in the microemulsion phases. Drug Acknowledgment
disposition within microemulsion structures can be examined by
measurement of NMR longitudinal (T1) relaxation of the drug in the The authors would like to thank Professor Richard Guy (Department
vehicle compared to that of the drug in its neat phase [24]. T1 values of Pharmacy and Pharmacology—University of Bath—UK) for his help,
basically gives information about the re-orientational processes of a support and useful discussions.
molecule after exposing to an external magnetic field, which reflects
the environment of the substance [30]. References
In a study on the transdermal delivery of lidocaine, the relaxation
[1] M. Garcia-Fuentes, D. Torres, M. Martin-Pastor, M.J. Alonso, Application of NMR
times of lidocaine in microemulsions consisting of water as the aque-
spectroscopy to the characterization of PEG-stabilized lipid nanoparticles, Lang-
ous phase, isostearylic isostearate as the oily phase, labrasol as the muir 20 (2004) 8839–8845.
surfactant and plurol isostearique as the cosurfactant, were closely [2] P. Griffiths, P. Stilbs, NMR self-diffusion studies of polymeric surfactants, Curr.
similar to the value obtained in neat isostearylic isostearate (0.77 s). Opin. Colloid Interface Sci. 7 (2002) 249–252.
[3] J.F. Billman, E.W. Kaler, Structure and phase-behavior in 4 component nonionic
This suggested that lidocaine was primarily located in the oil phase microemulsions, Langmuir 7 (1991) 1609–1617.
and that a large portion of the drug was free, i.e. not immobilized in [4] M.J. Lawrence, Microemulsions as drug delivery vehicles, Curr. Opin. Colloid In-
the surfactant interface [24]. This result is logically accepted as it terface Sci. 1 (1996) 826–832.
[5] M.J. Lawrence, G.D. Rees, Microemulsion-based media as novel drug delivery sys-
goes in accordance with the better solubility of lidocaine in the lipo- tems, Adv. Drug Deliv. Rev. 45 (2000) 89–121.
philic components than in water [49]. [6] K.J. Lissant, Basic theory, in: K.J. Lissant (Ed.), Emulsions and Emulsion Technolo-
gy, Marcel Dekker Inc, New York, 1974, pp. 4–10.
[7] H.Y. Karasulu, Microemulsions as novel drug carriers: the formation, stability, ap-
2.3.2. Characterization of microemulsions using the translational plications and toxicity, Expert Opin. Drug Deliv. 5 (2008) 119–135.
(spin–spin, T2) relaxation times [8] A.S.B. Goebel, R.H.H. Neubert, J. Wohlrab, Dermal targeting of tacrolimus using
Sizes and shapes are related to frictional factors for re-orientational colloidal carrier systems, Int. J. Pharm. 404 (2011) 159–168.
[9] S.K. Singh, P.R. Verma, B. Razdan, Development and characterization of a lovastatin-
and translational rates. Translational motion can contribute to nuclear
loaded self-microemulsifying drug delivery system, Pharm. Dev. Technol. 15 (2010)
relaxation rates through intermolecular dipole–dipole interaction 469–483.
[54]. In a study adopted for calculating the translational relaxation [10] S. Gupta, S.P. Moulik, Biocompatible microemulsions and their prospective uses in
drug delivery, J. Pharm. Sci. 97 (2008) 22–45.
time for lecithin in deuterated water, the aliphatic chains (δ = 0.9–
[11] S. Heuschkel, A. Goebel, R.H.H. Neubert, Microemulsions—modern colloidal carri-
2.8 ppm) of lecithin showed T2 values that were much smaller than er for dermal and transdermal drug delivery, J. Pharm. Sci. 97 (2008) 603–631.
their T1 value. This indicated that these groups were tightly packed, [12] B.K. Kang, S.K. Chon, S.H. Kim, S.Y. Jeong, M.S. Kim, S.H. Cho, H.B. Lee, G. Khang,
most probably forming a micelle-type structure [1]. This finding can Controlled release of paclitaxel from microemulsion containing PLGA and evalu-
ation of anti-tumor activity in vitro and in vivo, Int. J. Pharm. 286 (2004)
be adopted for tracing microemulsion formation as well, in systems 147–156.
containing oil, water and surfactant(s). [13] G. Kantarci, I. Ozgungey, H.Y. Karasulu, T. Guneri, G. Basdemir, In vitro permeation
Table 5 summarises the use of NMR applications in the micro- of diclofenac sodium from novel microemulsion formulations through rabbit skin,
Drug Dev. Res. 65 (2005) 17–25.
emulsion domain. [14] A. Kogan, N. Garti, Microemulsions as transdermal drug delivery vehicles, Adv.
Colloid Interface Sci. 123–126 (2006) 369–385.
3. Conclusion [15] K.M. Park, C.K. Kim, Preparation and evaluation of flurbiprofen-loaded micro-
emulsion for parenteral delivery, Int. J. Pharm. 181 (1999) 173–179.
[16] S. Peltola, P. Saarinen-Savolainen, J. Kiesvaara, T.M. Suhonen, A. Urtti, Micro-
Microemulsions, comprised as they are of multiple components, emulsions for topical delivery of estradiol, Int. J. Pharm. 254 (2003) 99–107.
exist in a wide variety of states, and the simple relation between com- [17] N. Subramanian, S.K. Ghosal, S.P. Moulik, Enhanced in vitro percutaneous absorp-
tion and in vivo anti-inflammatory effect of a selective cyclooxygenase inhibitor
position and performance, whilst useful, misses much of the inherent usingmicroemulsion, Drug Dev. Ind. Pharm. 31 (2005) 405–416.
complexity of the structures that are found. This article has intro- [18] T.K. Vyas, A.K. Babbar, R.K. Sharma, S. Singh, A. Misra, Preliminary brain-targeting
duced the range of NMR techniques that can be applied to such struc- studies on intranasal mucoadhesive microemulsions of sumatriptan, AAPS
PharmSciTech 7 (2006).
tures, and has shown that by judicious application of NMR techniques
[19] H. Zhang, F.Q. Feng, J. Li, X. Zhan, H.W. Wei, H.X. Li, H.Y. Wang, X.D. Zheng, Formu-
much useful information can be obtained, allowing for optimization lation of food-grade microemulsions with glycerol monolaurate: effects of short-
based on the properties of the microemulsion and other parameters, chain alcohols, polyols, salts and nonionic surfactants, Eur. Food Res. Technol. 226
such as the location of drug components in a given formulation. (2008) 613–619.
[20] A.H. Elshafeey, A.O. Kamel, M.M. Fathallah, Utility of nanosizedmicroemulsion for
Whilst chemical shifts provide information about the drug locus and transdermal delivery of tolterodine tartrate: ex-vivo permeation and in-vivo
the possible formation of aggregates, self-diffusion techniques pro- pharmacokinetic studies, Pharm. Res. 26 (2009) 2446–2453.
vide a more complete picture of the microemulsion formulation in- [21] A.O. Hassan, A.H. Elshafeey, Nanosized particulate systems for dermal and trans-
dermal delivery, J. Biomed. Nanotechnol. 6 (2010) 621–633.
cluding its type, drug locus inside it, aggregation and gives a [22] A.H. Elshafeey, E.R. Bendas, O.H. Mohamed, Intranasal microemulsion of sildenafil
measure that can be used in optimizing the formulation. The use of citrate: in vitro evaluation and in vivo pharmacokinetic study in rabbits, AAPS
relaxation times (longitudinal and translational) can be used as a pre- PharmSciTech 10 (2009) 361–367.
[23] M. Gradzielski, Recent developments in the characterisation of microemulsions,
cise alternative for self-diffusion techniques in characterizing the Curr. Opin. Colloid Interface Sci. 13 (2008) 263–269.
microemulsion systems and determining of drug locus. [24] M. Kreilgaard, Influence of microemulsions on cutaneous drug delivery, Adv.
It should be noted that the combination of various experimental Drug Deliv. Rev. 54 (2002) S77–S98.
[25] A.C. Lam, R.S. Schechter, The theory of diffusion in microemulsion, J. Colloid Inter-
and novel techniques such as Cryo-field emission scanning electron face Sci. 120 (1987) 56–63.
microscopy and scattering methods, together with appropriate theo- [26] R.M. Hathout, S. Mansour, N. Mortada, A.S. Geneidi, R.H. Guy, Uptake of micro-
retical models is required in order to deduce a more refined structural emulsion components into the stratum corneum and their molecular effects on
skin barrier function, Mol. Pharm. 7 (2010) 1266–1273.
picture of the particularly complex microemulsion systems. Tech-
[27] R.M. Hathout, S. Mansour, A.S. Geneidi, N.D. Mortada, Visualization, dermato-
niques other than the NMR yield more global information regarding pharmacokinetic analysis and monitoring the conformational effects of a micro-
the mesomorphic structure, whilst NMR yields information on the emulsion formulation in the skin stratum corneum, J. Colloid Interface Sci. 354
molecular level of the structure. Accordingly, the described NMR (2011) 124–130.
[28] V.B. Junyaprasert, P. Boonme, S. Songkro, K. Krauel, T. Rades, Transdermal deliv-
methods are complementary to these techniques that are not sensi- ery of hydrophobic and hydrophilic local anesthetics from o/w and w/o Brij 97-
tive to the diverse transitions fluently occurring in the microemulsion based microemulsions, J. Pharm. Pharm. Sci. 10 (2007) 288–298.
72 R.M. Hathout, T.J. Woodman / Journal of Controlled Release 161 (2012) 62–72

[29] O. Regev, S. Ezrahi, A. Aserin, N. Garti, E. Wachtel, E.W. Kaler, A. Khan, Y. Talmon, [52] P. Stilbs, B. Lindman, NMR measurements on microemulsions, Prog. Colloid
A study of the microstructure of a four-component nonionic microemulsion by Polym. Sci. 69 (1984) 39–47.
cryo-TEM, NMR, SAXS, and SANS, Langmuir 12 (1996) 668–674. [53] P. Stilbs, Fourier transform pulsed-gradient spin-echo studies of molecular diffu-
[30] M. Kreilgaard, E.J. Pedersen, J.W. Jaroszewski, NMR characterisation and transder- sion, Prog. Nucl. Magn. Reson. Spectrosc. 19 (1987) 1–45.
mal drug delivery potential of microemulsion systems, J. Control. Release 69 [54] K.F. Morris, C.S. Johnson, Resolution of discrete and continuous molecular-size
(2000) 421–433. distributions by means of diffusion-ordered 2D NMR-spectroscopy, J. Am.
[31] S.H. Chen, S.L. Chang, R. Strey, P. Thiyagarajan, Small-angle neutron-scattering in- Chem. Soc. 115 (1993) 4291–4299.
vestigation of structural inversion in a 3-component ionic microemulsion, J. Phys. [55] I. Lonnqvist, B. Hσkansson, B. Balinov, O. Soderman, NMR self-diffusion studies of
Condens. Matter 3 (1991) F91–F107. the water and the oil components in a W/O/W emulsion, J. Colloid Interface Sci.
[32] S. Magazu, G. Maisano, F. Mallamace, N. Micali, Growth of fractal aggregates in 192 (1997) 66–73.
water solutions of macromolecules by light-scattering, Phys. Rev. A 39 (1989) [56] M.J. Thrippleton, N.M. Loening, J. Keeler, A fast method for the measurement of
4195–4200. diffusion coefficients: one-dimensional DOSY, Magn. Reson. Chem. 41 (2003)
[33] G. Maisano, F. Mallamace, N. Micali, Dynamical properties of a potassium oleate 441–447.
microemulsion determined by photon-correlation spectroscopy, Phys. Rev. A 39 [57] H.Z. Liu, Y.J. Wang, Y.Y. Lang, H.M. Yao, Y. Dong, S.M. Li, Bicontinuous cyclosporin
(1989) 4103–4108. A loaded water-AOT/Tween 85-isopropylmyristate microemulsion: structural
[34] S. Bisal, P.K. Bhattacharya, S.P. Moulik, Conductivity study of microemulsions— characterization and dermal pharmacokinetics in vivo, J. Pharm. Sci. 98 (2009)
dependence of structural behavior of water oil systems on surfactant, 1167–1176.
cosurfactant, oil, and temperature, J. Phys. Chem. 94 (1990) 350–355. [58] D. Libster, A. Aserin, N. Garti, A novel dispersion method comprising a nucleating
[35] J. Biais, M. Mercier, P. Bothorel, B. Clin, P. Lalanne, B. Lemanceau, Microemulsions agent solubilized in a microemulsion, in polymeric matrix: I. Dispersion method
and electron microscopy, J. Microsc. 121 (1981) 169–178. and polymer characterization, J. Colloid Interface Sci. 299 (2006) 172–181.
[36] N. Garti, A. Aserin, S. Ezrahi, I. Tiunova, G. Berkovic, Water behavior in nonionic [59] M. Fanun, Structure probing of water/mixed nonionic surfactants/caprylic–capric
surfactant systems I: subzero temperature behavior of water in nonionic micro- triglyceride system using conductivity and NMR, J. Mol. Liq. 133 (2007) 22–27.
emulsions studied by DSC, J. Colloid Interface Sci. 178 (1996) 60–68. [60] B. Biruss, H. Kahlig, C. Valenta, Evaluation of an eucalyptus oil containing topical
[37] M. Kahlweit, R. Strey, D. Haase, H. Kunieda, T. Schmeling, B. Faulhaber, M. drug delivery system for selected steroid hormones, Int. J. Pharm. 328 (2007)
Borkovec, H.F. Eicke, G. Busse, F. Eggers, T. Funck, H. Richmann, L. Magid, O. 142–151.
S÷derman, P. Stilbs, J. Winkler, A. Dittrich, W. Jahn, How to study micro- [61] B. Lindman, U. Olsson, O. Soderman, Characterization of microemulsions by NMR,
emulsions, J. Colloid Interface Sci. 118 (1987) 436–453. in: P. Kumar, K.L. Mittal (Eds.), Handbook of Microemulsion Science and Technol-
[38] J. Lang, N. Lalem, R. Zana, Quaternary water in oil microemulsions. 1. Effect of al- ogy, Marcel Dekker, New York, 1999, pp. 309–356.
cohol chain-length and concentration on droplet size and exchange of material [62] K. Krauel, N.M. Davies, S. Hook, T. Rades, Using different structure types of micro-
between droplets, J. Phys. Chem. 95 (1991) 9533–9541. emulsions for the preparation of poly(alkylcyanoacrylate) nanoparticles by inter-
[39] J. Lang, N. Lalem, R. Zana, Quaternary water-in-oil microemulsions. 2. Effect of facial polymerization, J. Control. Release 106 (2005) 76–87.
carboxylic-acid chain-length on droplet size and exchange of material between [63] K. Krauel, A. Graf, S.M. Hook, N.M. Davies, T. Rades, Preparation of
droplets, J. Phys. Chem. 96 (1992) 4667–4671. poly(alkylcyanoacrylate) nanoparticles by polymerization of water-free micro-
[40] D. Waysbort, S. Ezrahi, A. Aserin, R. Givati, N. Garti, H-1 NMR study of a U-type emulsions, J. Microencapsul. 23 (2006) 499–512.
nonionic microemulsion, J. Colloid Interface Sci. 188 (1997) 282–295. [64] P. Boonme, Characterization of microemulsion structures in the pseudoternary
[41] F. Istvan, NMR spectroscopy of micelles and related systems, J. Mol. Liq. 117 phase diagram of isopropyl palmitate/water/Brij97: 1-butanol, AAPS PharmSciTech
(2005) 117–137. 7 (2006) E45.
[42] Y.E. Shapiro, Structure and dynamics of hydrogels and organogels: an NMR spec- [65] A. Spernath, A. Yaghmur, A. Aserin, R.E. Hoffman, N. Garti, Food-grade micro-
troscopy approach, Prog. Polym. Sci. 36 (2011) 1184–1253. emulsions based on nonionic emulsifiers: media to enhance lycopene solubiliza-
[43] S.K. Mehta, G. Kaur, K.K. Bhasin, Incorporation of antitubercular drug isoniazid in tion, J. Agric. Food Chem. 50 (2002) 6917–6922.
pharmaceutically accepted microemulsion: effect on microstructure and physical [66] N. Garti, A. Yaghmur, A. Aserin, A. Spernath, R. Elfakess, S. Ezrahi, Solubilization of
parameters, Pharm. Res. 25 (2008) 227–236. active molecules in microemulsions for improved environmental protection, Col-
[44] F.F. Lv, N. Li, L.Q. Zheng, C.H. Tung, Studies on the stability of the chloramphenicol loids Surf. A 230 (2003) 183–190.
in the microemulsion free of alcohols, Eur. J. Pharm. Biopharm. 62 (2006) [67] P.M. Bramley, Is lycopene beneficial to human health? Phytochemistry 54 (2000)
288–294. 233–236.
[45] Y.C. Shi, Y.S. Wu, G.Z. Li, W.Z. Zhu, Solubilization of styrene and acrylamide mono- [68] A. Spernath, A. Aserin, Microemulsions as carriers for drugs and nutraceuticals,
mers in microemulsions, J. Dispersion Sci. Technol. 24 (2003) 229–236. Adv. Colloid Interface Sci. 128–130 (2006) 47–64.
[46] R.M. Hathout, T.J. Woodman, S. Mansour, N.D. Mortada, A.S. Geneidi, R.H. Guy, [69] A. Graf, E. Ablinger, S. Peters, A. Zimmer, S. Hook, T. Rades, Microemulsions con-
Microemulsion formulations for the transdermal delivery of testosterone, Eur. taining lecithin and sugar-based surfactants: nanoparticle templates for delivery
J. Pharm. Sci. 40 (2010) 188–196. of proteins and peptides, Int. J. Pharm. 350 (2008) 351–360.
[47] K. Itoh, S. Matsui, Y. Tozuka, T. Oguchi, K. Yamamoto, Improvement of physico- [70] B. Lindman, O. Soderman, H. Wennerstrom, Surfactant solutions—new methods
chemical properties of N-4472 Part II: characterization of N-4472 microemulsion of investigations, Marcel Dekker, New York, 1987.
and the enhanced oral absorption, Int. J. Pharm. 246 (2002) 75–83. [71] M. Monduzzi, A. Ceglie, B. Lindman, O. Soderman, A 2H and 13C NMR multifield
[48] K. Itoh, Y. Tozuka, T. Oguchi, K. Yamamoto, Improvement of physicochemical relaxation study of aqueous aggregate systems of sodium dodecyl sulfate, J. Col-
properties of N-4472 part I formulation design by using self-micro emulsifying loid Interface Sci. 136 (1990) 113–123.
system, Int. J. Pharm. 238 (2002) 153–160. [72] A.C. Lam, R.S. Schechter, A study of diffusion and electrical-conduction in micro-
[49] J. Carlfors, I. Blute, V. Schmidt, Lidocaine in microemulsion—a dermal delivery emulsions, J. Colloid Interface Sci. 120 (1987) 42–55.
system, J. Dispersion Sci. Technol. 12 (1991) 467–482. [73] M. Fanun, Conductivity, viscosity, NMR and diclofenac solubilization capacity
[50] B. Lindman, P. Stilbs, M.E. Moseley, Fourier-transform NMR self-diffusion and studies of mixed nonionic surfactants microemulsions, J. Mol. Liq. 135 (2007) 5–13.
micro-emulsion structure, J. Colloid Interface Sci. 83 (1981) 569–582. [74] L. Hua, P. Weisan, L. Jiayu, Z. Ying, Preparation, evaluation, and NMR characteriza-
[51] P. Stilbs, M.E. Moseley, B. Lindman, Fourier-transform NMR self-diffusion mea- tion of vinpocetine microemulsion for transdermal delivery, Drug Dev. Ind.
surements on micro-emulsions, J. Magn. Reson. 40 (1980) 401–404. Pharm. 30 (2004) 657–666.
 Institut für Lebensmitteltechnologie

Solubility and Stability of Natural Food

Colorants in Microemulsions

Inaugural – Dissertation

Zur

Erlangung des Grades

Doktor-Ingenieur
(Dr.-Ing)

der

Hohen Landwirtschaftlichen Fakultät

der

Rheinschen Friedrich-Wilhelms-Universität

zu Bonn

vorgelegt am 18.03.2002

von
Mohamed Awad Saad Abd El-Galeel

aus
El-Beheira, Ägypten
Referent: Prof. Dr. Benno Kunz

Koreferent: Prof. Dr. Rudolf C. Galensa

Tag der mündlichen Prüfung: 10.06.2002

Gedruckt bei: ----------------------------------------------------

Acknowledgements

I wish to express my sincerest gratitude to my supervisor Prof. Dr. Benno Kunz, head of the
Department of Food Technology, University of Bonn for his help in planning this work,
valuable guidance, support and continuously encouragment.

I would like to thank Prof. Dr. Rudolf C. Galensa, dean of Agriculture Faculty, Bonn
University for accepting the evaluation of my thesis.

Sincere appreciations are expressed to my colleague Alaa Arafat M. Hayallah in

Pharmaceutical Institute, University of Bonn, for his help in correction the writing this
thesis.

Many thanks to my parents especially my mother for her great and unending encouraging
and first of all her upbringing for me.

My thanks go also to my ture wife Azza Abd El-Karim Moustafa for maitaining a
comfortable atmosphere and for her support during the long hard work of the thesis. Special
thanks also to my family in Egypt.

My thanks go to my colleague Dr. Birgit Ditgens, Sandra Maaß, Bernd Stefer and Jenny
Weißbrodt in Dapartment of Food Technology, University of Bonn for their help in
translation and correction the summary in German. I would like to thank everybody of the
staff of the Department of Food Technology, University of Bonn and everybody else who is
not named for their assistance and help.

Finally, I would like to thank the Egyptian Government and the Food Technology
Dapartment, Faculty of Agriculture in Kafr El-Sheikh, Tanta University, Egypt for the
financial support during my stay in Germany.
 Löslichkeit und Stabilität natürlicher Lebensmittelfarbstoffe in
Mikroemulsionen

Mohamed Awad Saad Abd El-Galeel

Es wurden verschiedene Mikroemulsionssysteme hergestellt unter Verwendung natürlicher

Öle wie reines Pfefferminzöl und Mischungen mit anderen Speiseölen (Sojaöl, Erdnussöl,
Rapsöl), verschiedenen Emulgatoren , wie Lecithin, Monoolein oder Tween20 und einer
wässrigen Phase (Wasser oder eine 20%ige Lösung von NaCl, Saccharose oder
Zitronensäure) ohne oder mit Ethanol als Co-Emulgator. Es wurde die Löslichkeit von
Curcumin in diesen Mikroemulsionssystemen untersucht. Darüber hinaus wurde die
Stabilität des in den Mikroemulsionen gelösten Curcumins gegen ultraviolettes Licht (UV)
sowie gewöhnliches elektrisches Licht gemessen. Die Abnahme des Curcumin-Farbtons
wurde mithilfe eines Farbmessgerätes (L*a*b*-System) bestimmt und als Größe zur
Bewertung der Stabilität herangezogen.
Die Ergebnisse zeigen, dass unter Verwendung geeigneter Komponenten Mikroemulsionen
zur Verwendung in Lebensmitteln hergestellt werden können. Diese Mikroemulsionen
erhöhen die Löslichkeit von Curcumin erheblich. Die Löslichkeit von Curcumin in diesen
Mikroemulsionen ist besser als seine Löslichkeit in Ethanol, Speiseölen und Pfefferminzöl.
Tween20/Pfefferminzöl-Mikroemulsionen zeigen die höchste Löslichkeit für Curcumin
sowie die größte Wasseraufnahmekapazität. Alle in den Stabilitätstests vermessenen
Mikroemulsionen lieferten gute Ergebnisse für die Stabilität von Curcumin gegen UV- und
normales elektrisches Licht für einen langen Zeitraum, ausgenommen die Mikroemulsionen,
deren wässrige Phase aus NaCl-Lösung bestand.
Die in dieser Arbeit hergestellten Mikroemulsionen scheinen vielversprechend für die
Erhöhung der Löslichkeit und die Stabilisierung von Curcumin und anderer natürlicher
Lebensmittelfarbstoffe mit vergleichbaren Eigenschaften (unlöslich in Wasser, schlecht
löslich in Pflanzenölen und lichtempfindlich) zu sein.
 Solubility and Stability of Natural Food Colorants in
Microemulsions

Mohamed Awad Saad Abd El-Galeel

Several microemulsion systems were prepared by using different natural oils such as
peppermint oil alone or mixed with a common edible oils (soybean, peanut or rapeseed oil),
different surfactants such as lecithin, monoolein or Tween20 and an aqueous solution (water
or 20% solution of NaCl, sucrose or citric acid ) without or with ethanol as cosurfactant. The
solubility of curcumin in these microemulsion systems was investigated. The stability of
curcumin solubilized in these microemulsions against UV light and normal electric light was
also studied. The stability of curcumin was determined as the change in curcumin color
shade by using Measuring Color Instrument.
The results obtained indicated that microemulsions can be prepared by using a suitable
components for food applications. These microemulsions greatly enhanced the solubility of
curcumin. The solubility of curcumin in these microemulsions is more higher than that in
ethanol, edible oils and peppermint oil. Tween20/peppermint oil microemulsions exhibited
the highst solubility for curcuminn and optimum water solubilization capacity. All
microemulsions on the basis of the stability tests offered good results for curcumin stability
against UV and normal electric light for long period, except microemulsions prepared by
NaCl solution as aqueous phase.
These prepared microemulsions are suggested to be promising for solubility and stability of
curcumin and other natural food colorants which have similar characteristics (insoluble in
water, poorly soluble in vegetable oils and sensitive to light).
Table of Contents

1 Introduction------------------------------------------------------------- 1

2 Theoretical Background---------------------------------------------- 5
2.1 Characteristic of microemulsions------------------------------------------------ 5
2.2 Structure and formation of microemulsions------------------------------------ 6
2.2.1 Role of surfactant and cosurfatant---------------------------------------------- 8
2.2.2 Oils used in the microemulsion preparation------------------------------------ 10
2.3 Applications of microemulsions-------------------------------------------------- 12
2.4 Food microemulsions------------------------------------------------------------- 14
2.4.1 Earlier work on food microemulsions applications---------------------------- 17
2.4.2 Surfactants in food microemulsions--------------------------------------------- 18
2.5 Natural colorants------------------------------------------------------------------ 21
2.5.1 Chemistry of natural colorants--------------------------------------------------- 22
2.5.2 Stability of natural colorants----------------------------------------------------- 26
2.5.3 Curcumin--------------------------------------------------------------------------- 27

3 Objectives---------------------------------------------------------------- 29

4 Materials and Methods----------------------------------------------- 30

4.1 Materials---------------------------------------------------------------------------- 30
4.1.1 Chemicals used in the microemulsion preparation----------------------------- 30
4.1.1.1 Surfactants------------------------------------------------------------------------- 30
4.1.1.1.1 Lecithin----------------------------------------------------------------------------- 31
4.1.1.1.2 Tween20---------------------------------------------------------------------------- 31
4.1.1.1.3 Monoolein-------------------------------------------------------------------------- 31
4.1.1.2 Oils---------------------------------------------------------------------------------- 31
4.1.1.2.1 Peppermint oil--------------------------------------------------------------------- 32
4.1.1.2.2 Edible oils-------------------------------------------------------------------------- 32
4.1.1.3 Cosurfactants---------------------------------------------------------------------- 32
4.1.1.4 Water------------------------------------------------------------------------------- 33
4.1.2 Other chemicals-------------------------------------------------------------------- 33
4.1.2.1 Curcumin--------------------------------------------------------------------------- 33
4.1.2.2 Sodium chlorid, Sucrose and Citric acid---------------------------------------- 34
4.1.3 Equipments and Instruments----------------------------------------------------- 34
4.1.3.1 Lamps------------------------------------------------------------------------------- 34
4.1.3.1.1 Normal electric light-------------------------------------------------------------- 34
4.1.3.1.2 Ultraviolet (UV) lamp------------------------------------------------------------ 34
4.1.3.2 Color Measuring Instrument----------------------------------------------------- 34
4.1.3.3 pH-Meter--------------------------------------------------------------------------- 34
4.1.3.5 Vortex mixer----------------------------------------------------------------------- 35
4.2 Methods---------------------------------------------------------------------------- 35
4.2.1 Preparation of microemulsions--------------------------------------------------- 35
4.2.1.1 Lecithin microemulsions---------------------------------------------------------- 35
4.2.1.2 Monoolein microemulsions------------------------------------------------------- 36
 II

4.2.1.3 Tween20 microemulsions.-------------------------------------------------------- 37

4.2.2 Solubility of curcumin------------------------------------------------------------- 38
4.2.2.1 Solubility in solvents-------------------------------------------------------------- 38
4.2.2.2 solubility in mixtures before aqueous phase addition ------------------------- 38
4.2.2.3 Solubility in microemulsions----------------------------------------------------- 38
4.2.3 Determination of the pH-value -------------------------------------------------- 39
4.2.4 Light stability of curcumin-------------------------------------------------------- 39
4.2.4.1 Deremination of the color shade------------------------------------------------- 41

5 Results------------------------------------------------------------------- 43
5.1 Preparatiom of microemulsions-------------------------------------------------- 43
5.1.1 Lecithin/peppermint oil microemulsions --------------------------------------- 43
5.1.2 Lecithin/peppermint oil/soybean oil microemulsions-------------------------- 44
5.1.3 Lecithin/peppermint oil/peanut oil microemulsions---------------------------- 45
5.1.4 Lecithin/peppermint oil/rapeseed oil microemulsions------------------------- 45
5.1.5 Monoolein microemulsions------------------------------------------------------- 46
5.1.6 Tween20 microemulsions.-------------------------------------------------------- 47
5.2 Solubility of curcumin------------------------------------------------------------- 47
5.2.1 Curcumin solubility in solvents-------------------------------------------------- 47
5.2.2 Curcumin solubility in microemulsions------------------------------------------ 49
5.2.2.1 Curcumin solubility in lecithin/peppermint oil microemulsions--------------- 49
5.2.2.2 Solubility in lecithin/peppermint oil/soybean oil microemulsions------------- 50
5.2.2.3 Solubility in lecithin/peppermint oil/peanut oil microemulsions-------------- 51
5.2.2.4 Solubility in lecithin/peppermint oil/rapeseed oil microemulsions------------ 52
5.2.2.5 Solubility in monoolein microemulsions---------------------------------------- 53
5.2.2.6 Solubility in peppermint oil/Tween20 (1:1) microemulsions------------------ 53
5.2.2.7 Solubility in peppermint oil/Tween20 (1:4) microemulsions------------------ 54
5.3 Determunation of pH-value------------------------------------------------------ 55
5.4 Light stability of curcumin in microemulsions---------------------------------- 57
5.4.1 Light stability in lecithin/peppermint oil microemulsions --------------------- 57
5.4.1.1 Effect of UV light----------------------------------------------------------------- 57
5.4.1.2 Effect of darkness----------------------------------------------------------------- 58
5.4.1.3 Effect of normal electric light---------------------------------------------------- 59
5.4.2 Stability in lecithin/peppermint oil/soybean oil microemulsions-------------- 60
5.4.2.1 Effect of Uvlight------------------------------------------------------------------- 60
5.4.2.2 Effect of darkness----------------------------------------------------------------- 61
5.4.2.3 Effect of normal electric light---------------------------------------------------- 62
5.4.3 Stability in lecithin/peppermint oil/peanut oil microemulsions---------------- 63
5.4.3.1 Effect of Uvlight------------------------------------------------------------------- 63
5.4.3.2 Effect of darkness----------------------------------------------------------------- 64
5.4.3.3 Effect of normal electric light---------------------------------------------------- 65
5.4.4 Stability in lecithin/peppermint oil/rapeseed oil microemulsions------------- 66
5.4.4.1 Effect of Uvlight------------------------------------------------------------------- 66
5.4.4.2 Effect of darkness----------------------------------------------------------------- 67
5.4.4.3 Effect of normal electric light---------------------------------------------------- 68
5.4.5 Stability in monoolein microemulsions------------------------------------------ 70
5.4.6 Stability in peppermintoil/Tween20 (1:1) microemulsions-------------------- 71
 III

5.4.6.1 Effect of Uvlight------------------------------------------------------------------- 71

5.4.6.2 Effect of darkness----------------------------------------------------------------- 72
5.4.6.3 Effect of normal electric light---------------------------------------------------- 73
5.4.7 Stability in peppermint oil/Tween20 (1:4) microemulsions------------------- 73
5.4.7.1 Effect of UV light----------------------------------------------------------------- 73
5.4.7.2 Effect of darkness----------------------------------------------------------------- 74
5.4.7.3 Effect of normal electric light---------------------------------------------------- 74

6 Discussion---------------------------------------------------------------- 76
6.1 Preparation of microemulsions--------------------------------------------------- 76
6.1.1 Lecithin/peppermint oil microemulsions---------------------------------------- 77
6.1.2 Lecithin/peppermint oil/edible oil microemulsions----------------------------- 79
6.1.3 Monoolein microemulsions------------------------------------------------------- 82
6.1.4 Tween20 microemulsions-------------------------------------------------------- 83
6.2 Solubility of curcumin ------------------------------------------------------------ 84
6.2.1 Curcumin solubility in solvents-------------------------------------------------- 85
6.2.2 Curcumin solubility in microemulsions------------------------------------------ 85
6.2.2.1 Curcumin solubility in lecithin/peppermint oil microemulsions--------------- 85
6.2.2.2 Solubility in lecithin/peppermint and soybean oil microemulsions----------- 86
6.2.2.3 Solubility in lecithin/peppermint and peanut oil microemulsions------------- 87
6.2.2.4 Solubility in lecithin7peppermint and rapeseed oil microemulsions---------- 87
6.2.2.5 Curcumin solubility in monoolein microemulsions----------------------------- 88
6.2.2.6 Curcumin solubility in Tween20 microemulsions------------------------------ 89
6.3 Light stability of curcumin in microemulsions---------------------------------- 90
6.3.1 Curcumin stability in lecithin/peppermint oil microemulsions---------------- 90
6.3.2 Curcumin stability in lecithin/peppermint oil/edible oil microemulsions ---- 91
6.3.3 Curcumin stability in monoolein microemulsions------------------------------ 91
6.3.4 Curcumin stability in Tween20 microemulsions-------------------------------- 92
6.4 Conclusion------------------------------------------------------------------------- 93

7 Summary---------------------------------------------------------------- 96

8 References--------------------------------------------------------------- 98
1 Introduction

Microemulsions are generally defined as isotropic, transparent, thermodynamically stable

mixtures of at least three components: a water, an oil and a surfactant; usually in
combination with a cosurfactant, typically a short chain alcohol [Aboofazeli and Lawrence
1994]. Microemulsions are isotropic systems of infinite stability, where, the surfactant and
the cosurfactant are principally located at the surface separating the two immiscible liquids
to stabilize their mutual dispersion [Bourrel and Schechter 1988]. An interesting
characteristic of microemulsions is that when even a small amount of a mixture of surfactant
and cosurfactant is added to biphasic water-oil system, a thermodynamically stable,
transparent and isotropic mixture spontaneusly forms [Ho et al., 1996].
When the components being used are safe for human consumption, microemulsions become
important in such fields as foods, cosmetics and pharmaceuticals [Kunieda and Shinoda
1982]. Microstructural studies of microemulsions have been given great attention because of
their physicochemical properties and various applications of commercial importance [Zaks
and Klibanov, 1985]. Important in microemulsions is fundamental because of their presence
in nature and applications in the food and pharmaceutical industries. A microemulsion is a
stable and transparent solution of several components (usually oil, water and surfactant) with
characteristic wavelength less than 100 nm to 200 nm [J. Research National Instit. Standards
Technol. 1994].
The use of a microemulsion as reaction medium avoids the problem of insolubility
frequently occourred with triglycerides and other lipophilic substrates. In addition, it opens
new synthetic possibilities [Holmberg and Österberg 1987]. Recently, microemulsions have
been identified as potential delivery systems for lipophilic agents due to their transparent or
translucent appearance, stability for long time, high solubilization capacity and ease of
formation [Malmsten 1996].
Since the first microemulsion system was described by Schulman and Hoar [1943] an
extensive number of papers have been published in this field. Most of the systems published
are, however, unacceptable for pharmaceutical use. A suitable oil phase for pharmaceutical
uses would be vegetable oils [Von Corswant et al., 1997]. Microemulsions prepared by
using vegetable oils or fatty acid esters could be used, for example, in cosmetics and food
products, where there is a demand towards environmentally more acceptable formulations
 2

[Abillon et al., 1986]. Oils from natural sources and their derivatives, triglycerides, are
considered to be harmless to the environment [Busch 1992], such systems with natural oils
are but infrequently described in the literature [Alander and Warnheim 1989]. Few studies
using triglycerides as the lipophilic phase in a microemulsion have been studied [Aboofazeli
et al., 1995]. The preparation of microemulsions with mineral oils, synthetic surfactants, and
-if necessary- alkanols as cosurfactants has been intensively studied in the literature.
Because, however, these components are harmful the proplem is how to prepare
microemulsions with nontoxic oils, surfactants and cosurfactants for possible applications in
pharmaceutical industry [Kahlweit et al., 1997]. The naturally occurring phospholipid
(lecithin) are biocompatible and legislatively acceptable for food applications [Svensson et
al., 1996].
There are very few acceptable examples of ingestible microemulsions for food uses, even
though much has been accomplished in recent years in the general field of microemulsions.
microemulsions are suitable as the delivery system for water soluble and oil soluble
nutrients, and flavors in foods. An ingestible, cosurfactant free system, with no off-taste or
change in performance is suitable for this purpose [El-Nokaly et al., 1991].
The advantages of a microemulsions over emulsions, or other solutions are either improved
stability or solubility characteristics. Microemulsions also have the potential ability to
solubilize both lipophilic and hydrophilic agents, which allows for a variety of flavoring and
coloring materials having vastly different physical properties to be dissolved in the system
[Friberg and Burasczenska 1978]. An additional important feature of phospholipid-based
microemulsions in their ability to solubilize larger guest molecules [Peng and Luisi 1990].
From the above mentioned, the problem in the preparation of microemulsions generally is to
find suitable components for food products.
Natural colorants are organic colorants that are obtained from natural edible sources using
recognized food preparation methods, for example curcumin (from turmeric). Natural colors
have always formed part of man`s normal diet and have, therefore, been safely consumed for
countless generations. The desirability of retaining the natural color of food is self-evident
but almost the demands of industry are such that additional color is important. Contrary to
many reports, natural sources can offer a comprehensive range of attractive colors for use in
the food industry. Food quality is first judged on the basis of its color.The color of a food
therefore influence not only the perception of flavor, but also that of attraction and quality.
One of the advantages of using natural colors is that they are generally more widely
accepted in food-stuffs than synthetic colors It is only in the last 100 years or so that
 3

synthetic colors have been added to food. For centuries prior to this, natural products in the
form of spices, berries and herbs were used to develop the color and flavor of food. During
this century, the use of synthetic color has gradually increased at the expense of these
products of natural origin, due mainly to their availability and lower relative cost. In the last
20 years following the delisting of several synthetic colors there has been an increase in the
use of colors derived from natural sources [Henry 1996].
The problem when using natural colors has been their lack of stability [Lauro 1998]. It
should remembered that natural colors are a diverse group of colorants with widely differing
solubility and stability preporties [Henry 1996]. Few plants have attracted the importance of
scientists and been the subject of scientific studies. One from these plants has been
investigated is Curcuma longa Linn [Cooper et al., 1994]. Curcuma extracts have been
shown to give a number of functions. It has been reported that administration of pure or
commercial grade curcumin in the diet decreases the incidence of tumors in mice and also
reduces tomour size. Histopathological test of the tumors showed that dietary curcumin
inhibits the number of papillomas and squamous cell carcinomas of the forestomach as well
as the number of adenomas and adenocarcinomas of the duodenum and colon [Azuine and
Bhide 1994]. Turmeric has been used as a spice for many thousands of years. Curcumin is
the principal color present in the rhizome of the turmeric plant. It is produced by
crystallization from oleoresin, which obtained by solvent extraction of the ground turmeric,
and has a purity level of around 95% [Henry 1996].
It is important to note that the clinical trials of curcumin in human cancer patients are in
progress [Ravindranath and Chandrasekhare 1980]. In addition, chemopreventive properties
in skin and forestomach carcinogenesis and various pharmaceutical applications have been
reported [Rao et al., 1995]. Curcumin is an antioxidant that inhibits lipid peroxidation in rat
liver microsomes, and a scavenger of reactive oxygen species that reduces the formation of
inflammatory compouds such as prostaglandins and leukotrienes [Reddy and Lokesh 1992].
Curcumin has been extensively used to color and flavor in foods. Phenolic pigment is
primarily caused the yellow color by curcumin [Cooper et al., 1994].
Curcumin is oil soluble but some blends of curcumin may be both oil and water insoluble.
Curcumin needs the addition of gums, stabilizers or emulsifiers in order to render its
miscibility in water. It is important that these ingredients are compatible with the food
system to which the color is being required [Henry 1996]. Specification for curcumin
[European Commission Document III/5218/94-rev.4, April 1995] states that the dye content
must be not less than 90% when measured spectrophotometrically at 426 nm in ethanol, and
 4

has ratio of curcumin to essentail oil 99:1. Curcumin is not an ideal product for direct use in
the food industry since it is insoluble in water and has poor solubility in most oils used in
food products. Thus it important for curcumin to be converted into convenient application
form. In many countries, this is achieved by dissolving the curcumin in a mixture of food-
grade solvent and accepted emulsifier. In this form, the product contains 4 to 10% curcumin
and easily dissolve in water. Polysorbate 80 is the favoured emulsifier/diluent for such
products since it is an ideal carrier for curcumin. Curcumin is sensitive to light and this
factor is the one that generally limits its use in foods [Henry 1996].
The problems for curcumin usually are its solubility and stability against light, because it is
insoluble in aqueous solutions and poorly soluble in fats or oils that used in foods and it is
sensitive to light. Microemulsions have infinite stability and large solubility capacity for
lipophilic and hydrophilic substances as well as ability to solubilize larger guest molecules.
Microemulsions also were used successfully to solubilize and stabilize of some
pharmaceutical components such as cyclosporin A by Gao et al., [1998] and to stabilize of
some nutrients such as ascorbic acid against oxidation by Gallarate et al., [1999]. For these
reasons, microemulsions were chosen to investigate the curcumin solubility in order to use it
with food products and study the possibility of these microemulsions to protect the
curcumin against light during the storage.
 5

2 Theoretical Background

2.1 Characteristic of microemulsions

The most characteristic difference between an emulsion and a microemulsion is their
appearance. An emulsion is turbid while the microemulsion almost is transparent. The
reason for this difference in appearance is the size of the droplets. For an emulsion the
droplets are similar or greater than the wave length of light and light is reflected off their
droplets. The emulsion, hence, appears turbid because the light cannot penetrate through it.
On the other hand the size of microemulsion droplets is smaller than the wave length of
light, and the interaction with light is limited to scattering. The light beam passes through
with but little loss; consequently the microemulsion appears transparent. The
microemulsions are thermodynamically stable with few exceptions, but the emulsions are
not thermodynamically stable because the interfacial energy is positive and dominant in total
free energy, where its droplet is of a size that the bending energy is negligible and the
surface free energy is large and positive; a few mN/m. The surface free energy of the
microemulsion has two components stretching (positive contribution) and bending (negative
contribution). The two cancel each other and the total surface free energy is extremely small
about 10-3 mN/m (figure 1). Moreover, The microemulsions form spontaneously or need
gentle shaken to miscible the components for short time (few minutes). The most
characteristic differences between an emulsion and a microemulsion are presented in Table 1
[Friberg and Kayali 1991]. These microemulsdions con take different forms which include
among others oil water and oil/water bicontinuous microemulsions. Phase change can occur
between these different forms due to changes in either individual component concentrations
or other thermodynamic conditions [Rosen 1989].

Table 2.1. Characteristics of Emulsions and Microemulsions [Friberg and Kayali 1991]

Emulsion Microemulsion
Appearance Turbid Transparent
Droplet size, µm radius 0.15 – 100 0.0015 - 0.15
Formation Mechanical or Chemical Energy Spontaneous
Thermodynamic Stability No Yes (No)
 6

Figure 2.1. The curvature in an emulsion droplet (A) is extremely small and the bending component of surface
energy is not significant. A change in curvature does not lead to a change in the free energy. In the
microemulsion droplet (B), on the other hand, a change in curvature leads to a pronounced change in free
energy; e.g., the bending component of the surface free energy is pronounced [Friberg and Kayali 1990].

The microemulsions possess special characteristics of relatively large interfacial area, ultra
low interfacial tension and large solubility capacity as compared to many other colloidal
systems [El-Nokaly et al., 1991].
The surfactant to cosurfactant ratio greatly affected the physicochemical characteristics of
the resultant microemulsion systems obtained using polyoxyethylated castor oil (Cremophor
EL) as a surfactant, Transcutol as a cosurfactant and carylic/capric triglyceride (Captex 355)
as an oil. The stable microemulsion with its high solubility of cyclosporin A (poorly water-
soluble drug), small droplet size and fast dispersion rate was obtained from a mixture
composed of 10:5:4 ratio of Cremophor EL:Transcutol:Captex 355. The enhanced
bioavailability of cyclosporin A loaded in this microemulsion system might be attributed to
the small droplet size of microemulsion systems [Gao et al., 1998].

2.2 Structure and formation of microemulsions

Bansal et al., [1980] reported that the conditions necessary for microemulsion formation are:
- large adsorption of surfactant or surfactant/cosurfactant mixture at the interface between
the oil and water which achieved by choosing a surfactant mixture with proper hydrophilic-
lipophilic-balances (HLB).
- high fluidity of the interface. The interfacial fluidity can be developed by using a proper
cosurfactant or an optimum temperature.
 7

- optimum curvature. The importance of oil penetration in the surfactant/cosurfactant film

and the appropriate surfactant/cosurfactant structures.
The microemulsions can take different types which include, among others, oil, water and
oil/water bicontinuous microemulsions. Phase change can occur between these different
forms due to changes in nature or concentrations of individual component or other
thermodynamic conditions [Rosen 1989].
A microemulsion that contains a relatively low content of oil confined within small isolated
droplets dispersed in water is known as oil-in-water (O/W) microemulsion, while the reverse
type (small amount of water dispersed in large amount of oil) is a water-in-oil (W/O)
microemulsion. Upon continuously increasing the water-to-oil ratio in a W/O
microemulsion, phase inversion occurs. An intermediate transparent, isotropic bicontinuous
structure may form during such an inversion, involving both oil and water-continuous
domains separated by interfacial surfactant film [Singh, et al., 1994]. Under appropriate
conditions the microemulsion system is miscible with both oil and aqueous phase. However,
the microemulsion system partitions into three phases, a surfactant-rich phase, a surfactant-
rich aqueous phase and a surfactant-rich oil phase. The surfactant-rich phase is called a
middle phase microemulsion [Abe, et al., 1986]. It is in the middle phase microemulsion
where a surfactant shows the greatest solubilizing power for both water and oil; here it also
gives ultra small values of interfacial tensions between oil and water which are less than 10-2
mNm-1 under proper conditions [Kunieda and Shinoda 1982]. The nature and structure of
the surfactant, cosurfactant and oil are important factors in the prepration of microemulsions
[Ho, et al., 1996]. The ability of phospholipids to form microemulsions with alkanes has
been studied by several authors [Shinoda and Kaneko 1988, Shinoda et al., 1991 and 1993,
Schurtenberger et al., 1993, and Kahlweit et al., 1995]. Shinoda et al., [1993] have shown
that it is possible to form microemulsions with equal amounts of hexadecane and aqueous
phase with only 2.5% soybean phospholipid using 1-propanol as the cosurfactant, and
Kahlweit et al., [1995] have systematically studied the influence of chain length of both the
phospholipid and the hydrocarbon on the microemulsion phase behavior.
Aboofazeli et al., [1994] studied partial phase diagrams of systems containing water, egg
lecithin, propanol and different polar oils such as Miglyol 812 and soybean oil, and found
that the influence of the oil and the ratio of egg lecithin to propanol on the microemulsion
area is significant. Due to the salting-out effect, the addition of NaCl decreases the required
amount of pentanediol (as cosurfactant), but has only a small effect on the lecithin required
(as surfactant) [Kahlweit et al., 1995].
 8

2.2.1 Role of surfactant and cosurfactant:

Surfactant and cosurfactant are mainly located at the surface separating the two immiscible
liquids (usually oil and water) to stabilize their mutual dispersion [Bourrel and Schechter
1988]. A hydrophilic surfactant adsorbs strongly to an interface toward the air or toward an
oil because of its dual structure with a hydrocarbon tail with limited interaction with water
(the hydrophobic part) and a polar group with strong interaction with water (the hydrophilic
part). This adsorption acts a reduction of the interfacial free energy. An lipophilic surfactant
does not adsorb toward the oil/air interface, but does so toward an oil/water interface. For
surfactant concentrations above a certain limit in water (the critical micellization
cocentration,c.m.c.) the added surfactant forms micelles, figure 2(A), and the adsorption to
the interface does not increase with surfactant concentration. In an oil, the oil soluble
surfactants and water form inverse micelles, figure (2B), in a step-wise process. These two
structures are especially significant in systems in which an ionic surfactant and a long chain
alcohol are combined, because this system illustrates, with high clarity, the fundamental
difference between the stabilizing system for a microemulsion and an emulsion. Once this
difference is distinguished the difficulties with microemulsions in food products easy to
comprehend [Friberg and Kayali 1991]. A cosurfactant is almost a medium chain fatty
alcohol, acid or amines [Lang et al., 1984].
The role of the cosurfactant together with the surfactant is to reduce the interfacial tension
down to a very small even transient negative value at which the interface would expand to
form ulta small dispersed droplets, and consequently adsorb more surfactant and
surfactant/cosurfactant until their bulk condition is depleted enough to give interfacial
tension positive again. This process known as „spontaneous emulsification“ forms the
microemulsion. Thus, based on ability of the cosurfactant to affect the solvent properties of
oil and/or water and to penetrate the surfactant interfacial monolayer, it can:
- Reduce further the interfacial tension. Increase the fluidity of interfaces.
- Destroy liquid crystalline and/or gel structures which prevent the microemulsion
formation.
- Adjust HLB value and spontaneous curvature of the interface by changing surfactant
partitioning characteristics.
- Decrease the sensitivity to structure fluctuations and brings formulation to its optimum
state [Kunieda et al., 1988].
 9

Figure 2.2.(A) normal micelle, in this case the surfactant hydrocarbon chains (blck) point toward the inner part
surrounded by the polar parts (unfilled circles).(B) inverse micelle in this case the hydrocarbon chains point
outward while the polar groups are concentrated in the center [Friberg and Kayali 1991].

The nature and concentration of the surfactant become of too importance to obtain an
optimum solubilization in a given W/O microemulsion [El-Nokaly et al., 1991]. Poly-
oxyethylene sorbitan trioleate (Tween 85) is a nonionic surfactant which has some important
properties for microemulsion preparation and protein solubilization [Komives et al., 1994].
The cosurfactant can act by either interchelating between surfactant molecules at the
interface between oil and water and/or by decreasing the aqueous phase hydrophilicity. In
the preparation of a balanced lecithin microemulsion, the cosurfactant has an additional role
in that it can also act to decrease the tendency of lecithin to form a highly rigid film [Binks
et al., 1989], thus allowing the interfacial film to take up the different curvatures required to
prepare balanced microemulsions [De Gennes and Taupin, 1982]. Because alkanols are, in
general, toxic, can substituted by alkanediols. The inexpensive alkanols can use in industry
as cosolvents and their toxicity can be toleranted, whereas in pharmacy one may have to use
nontoxic alkanediols [Kahlweit et al., 1995].
 10

2.2.2 Oils used in the microemulsion preparation

The preparation of microemulsions with mineral oils, synthetic surfactants, either ionic or
nonionic, and -if necessary- alkanols as cosolvents, has become a well-established practice
[Kahlweit 1995]. Because, however, mineral oils and synthetic amphiphiles, as well as
alkanols, in general, harmful, the problem is how to prepare microemulsions with suitable
components for possible applications in pharmaceutical industry [Kahlweit et al., 1997].
Unfortunately, most work to date studying microemulsions has utilized oils, surfactants and
cosurfactants unacceptable for pharmaceutical purposes. In order to make these systems
pharmaceutically acceptable, it is necessary to prepare such systems by using nontoxic and
safe components [Aboofazeli and Lawrence 1994].
Oils from natural sources and their derivatives, e.g. triglycerides and fatty acid methyl
esters, are easily degraded by microorganisms and are considered to be harmless to the
environment [Busch 1992]. The formation of bicontinuous microemulsions with mineral oils
has been intensively investigated in model experiments [Kahlweit et al., 1990] and for
application in industrial products [Schwuger 1995], such systems with natural oils and esters
are but infrequently published in the literature [Alander and Warnheim 1989].
An acceptable lipophilic phase for pharmaceutical uses would be a vegetable oils [Von
Corswant et al., 1997]. The extension of a microemulsion region is generally dependent on
the oil nature. This is due to differences in oil penetration into the surfactant layer
[Monduzzi et al., 1997]. Relatively few studies using triglycerides as the lipophilic phase in
a microemulsion have been studied [Aboofazeli et al., 1995]. Joubran et al., [1994] studied
microemulsions of soybean oil, polyoxyethylene(40) sorbitan-hexaoleate, and water-ethanol.
They found that the extension of the water-in-oil (W/O) microemulsion regions were
significantly dependent on temperature. Moreover, water-ethanol ratios also affected the
phase behavior. In these systems large amounts of the surfactant had to be used to form a
microemulsion with equal amounts of triglyceride and aqueous phase.
Triglycerides form bicontinuous microemulsions only at relatively high temperatures with a
high amount of hydrophobic surfactants owing to their high molecular weight. Esters of
fatty acids, however, offer suitable phase behavior forming three phases at appropriate
temperature and moderate surfactant content [Mönig et al., 1996].
 11

Alander and Warnheim [1989a] observed that microemulsion phase diagrams for high-
molecular-weight triglycerides exhibit smaller homogeneous regions than low-molecular-
weight esters and hydrocarbons. This effect has been attributed to better penetration of the
interfacial film by small oil molecules which aids in obtaining optimal curvature of W/O or
O/W droplets [Walde et al., 1990].
Triglycerides, in particular large triglycerides such as peanut oil, are significantly more
difficult to solubilize into microemulsions than hydrocarbons or alkyl esters [Alander and
Warnheim 1989a].
El-Nokaly et al., [1991] explained the reasons which make difficults concerning to form a
triglyceride micoemulsions:
- Triglycerides are semi-polar comparing with hydrocarbons.
- A surfactant of higher hydrophile-lipophile-balances (HLB) is thus needed to favor the
water-in-oil system, with lower solubility in bulk and increased adsorption at the interface.
- In case of triglyceride, the ratio of surfactant/water is high.
-The surfactant efficiency is decreased if it lost to the bulk and is unavailable to the
interface.
Edible triglycerides such as soybean, rapeseed, or sunflower oils contain long alkyl chains
mainly C16, C18,C20 and C22. The oil may be very bulky to penetrate the interfacial layer to
assist the formation of the optimum curvature, figure 2.3. Reports of oil being solubilized in
the aggregates palisade layer may be due to the shortness of the alkyl chains in the
triglycerides used [Kunieda et al., 1988].
To increase the triglyceride microemulsion regions, a different strategy was applied by using
a suitable hydrotrope to destabilize the liquid crystalline phase of the triglyceride, surfactant
and water, which leads preferentially to the formation of the microemulsion [Joubran et al.,
1993]. The preparation of triglyceride microemulsions can be achieved by incorporating
sucrose and short chain alcohols such as ethanol. The alcohol acts synergistically with
sucrose to destabilize the liquid crystalline mesophase [Joubran et al., 1994]. Sucrose
enhanced the formation of the oil-in-water microemulsion phase while destroying the water-
in-oil microemulsion phase. Triglycerides containing unsaturated or short-chain fatty acids
have improved solubility in oil-in-water microemulsions compared to triglycerides with
saturated or long-chain fatty acids [Parris et al., 1994].
 12

Figure 2.3. Optimum Curvature, Ro=Radius of Spontaneous Curvatures [Kunieda et al., 1988].

2.3 Applications of microemulsions

As a consequence of their unusual thermodynamic properties, microemulsions are of
considerable industrial importance in tertiary oil recovery [Langvin 1984], extraction of
biomolecules from fermentation broths [Göklen and Hatton 1985] and as liquid membrane
carrier agents [Tonder and Xenakis 1982]. W/O microemulsions with little water content
exhibit the best cleaning results for oil soil [Dörfler et al., 1995]. So far microemulsions
have not been exploited for pharmaceutical purposes [Aboofazeli and Lawrence 1994].
In recent years, microemulsions have been identified as potential drug delivery systems for
lipophilic drugs due to their transparent or translucent appearance, long term stability, high
solubilization capacity, and ease of preparation [Malmsten 1996]. Water-in-oil (W/O)
microemulsions are a particularly attractive system for biotechnological applications
[Stamatis et al., 1995]. Microemulsions of edible oils in a matrix of water and different
hydrotropes have been used as carriers for flavors or essential oils [Wolf and Hauakotta
1989].
Since alcohols of medium chain length (generally used as cosurfactant) tend to posses
unacceptable toxicity/irritation profiles (see Table 2.2, their use in foods is very limited.
Unlike, the ethanol ingestion is of known consequence. Thus in small amounts (5%), use of
ethanol is completely acceptable for food products [Osborne et al., 1991].
 13

Table 2.2 Safety considerations of Medium Chain Alcohols [Osborne et al., 1991]

Alcohol Oral LD50 in Rat Eye Irritation Skin Irritation

(mg/kg)
2-Butanol 6480 Moderate Threshold Conc. 7.8%
(least irritation of 4 isomers)
Hexanol 720 Severe Mild
Octanol 1790 Moderate Threshold Conc. 12%
Decanol 472 Sereve Sereve

Oral LD50 for mouse, oral LD50 for humans 0.5-5 mg/kg

Microemulsions have the potential ability to solubilize both lipophilic and hydrophilic
species, which allows for a variety of flavoring and coloring agents having vastly different
physical properties to be dissolved/solubilized within a system. A major disadvantage of
microemulsion systems for ingestion is the traditional need for a medium chain alcohol such
as pentanol to function as a cosurfactant [Friberg and Burasczenska 1978].
Microstructural studies of microemulsions have been given considerable attention because
of their interesting physicochemical properties and various applications of commercial
importance [Zaks and Klibanov 1985]. Interest in microemulsions is substantial because of
their ubiquitous presence in nature and applications in the food and pharmaceutical
industries. Further studies are under way with the goal of aiding development of improved
properties and stability of microemulsions used in many processed food products [J.
Research National Instit. Standards Technol. 1994].
Microemulsions containing vegetable oils or fatty acid esters could be used, for example, in
cosmetics and food products, where there is a demand for environmentally more acceptable
formulations [Abillon et al., 1986]. Phospholipid-based (W/O) microemulsions have ability
to solubilize larger guest molecules such as enzymes [Peng and Luisi 1990]. It has been
shown earlier that lipases in microemulsions can be successfully used to catalyze
esterification reactions [Kolisis et al., 1990].When ionic surfactants are used, the resulting
microemulsions are often called reverse micelles [Hatton 1989].
 14

Many enzymatic reactions require biphasic media, polar media for the solubilization of
enzymes, and organic media solubilization substrates. Reverse micelles provide a larger
polar/apolar interfacial area, hence improving the interaction between enzyme and substrate
[Hayes and Gulari 1990]. Use of reverse micelles have many advantages including easier
media and enzyme preparation, less mass-transfer limitations (due in part to its large degree
of interfacial area), and simpler control and monitoring of water content [Hayes and Gulari
1991]. The use of a microemulsion as reaction medium eliminates the problem of
insolubility frequently encountered with triglycerides and other lipophilic substrates. In
addition, it opens novel synthetic possibilities. For instance, lipase catalyzed
interesterification can be used to produce triglycerides, which is interest for the production
of synthetic cocoa butter [Holmberg and Österberg 1987].
From an industrial point of view, the microemulsions could be attractive systems for making
stable products with low fat content [Friberg et al., 1990]. Monoglycerides are nonionic
surfactants widely used as emulsifiers in the food and pharmaceutical areas. Normally they
are produced by alcoholysis of corresponding triglyceride with two equivalents of glycerol.
The reaction requires high temperatures (210-240 0 C) and the use of a transesterfication
catalyst. After work-up the effective yield of triglyceride to monoglyceride conversion is 40-
50%. Monoglycerides have been obtained in 80% yield by enzyme catalyzed hydrolysis of
the corresponding triglyceride. The reaction was carried out in an oil-rich microemulsion
formulated without cosurfactant [Holmberg and Österberg 1988]. The ability of W/O
microemulsion to isolate and selectively extract proteins is well-known [Pires et al., 1996].
The use of nonionic surfactant microemulsions for protein extraction has attracted attention
recently [Vasudevan et al., 1995].

2.4 Food microemulsions

The food emulsions have always generated great interest. However, little has been done with
microemulsions applicable to the complex world of foods. Microemulsions have been the
subject of much fundamental research that focuses on noningestible systems. For example,
applications to non-food uses such as tertiary oil recovery, fuel, cosmetics, and household
have received considerable attention. Larsson`s monoglyceride/water/oil system was the
first practical system to be published [El-Nokaly et al., 1991].
Since the first microemulsion system was described by Schulman and Hoar [1943] an
extensive number of papers have been published in this area. Most of the systems described
are, however, not suitable for pharmaceutical use [Von Corswant et al., 1997].
 15

The yield from this search (Food micremulsion) was a mere four references none older than
1987. The inclusion of microemulsions is a matter of looking to the future [Becher 1991]. It
should point out that a recent review of applications of micremulsions contained but a single
reference to foods [Gillberg 1984]. In fact, a number of recent books on food emulsions,
edited by; respectively, Friberg 1976, Dickinson 1987, and Dickinson and Stainsby, 1988,
contain no instance of food microemulsions (except for one minor reference in Dickinson
1987). This lack of interest may possibly be ascribed to a number of related factors. First,
the high levels of emulsifying agents normally encountered in microemulsions quite simply
serves as an economic barrier. Second, this same high level of emulsifier might well raise
legal problems in securing approval from the FDA. Third, of course, there is the simple fact
that it is apparently quite difficult to make microemulsions of the fats and oils used in foods
[Becher 1991]. From these considerations, food microemulsions, it follows automatically
that a surfactant/cosurfactant combination, which is optimal for a microemulsion, is of little
use in order to stabilize an emulsion. This is aserious disadvantage when double emulsions,
W/O/W, are formulated. For a system of this kind a W/O microemulsion emulsified into
water would in princible be a very attractive option, because the W/O part which is the
difficult part to stabilize would now be thermodynamically stable. However, stabilization
using surfactant combination has fundamental difficulties. The surfactant combination for
the microemulsion will rapidly exchange with the one for the emulsion, which leads to
destabilization for both the emulsion and the microemulsion. This dilemma has been
resolved in an elegant manner by Larsson et al., [1980]. They used a surfactant to stabilize
the W/O microemulsion but avoided the problem of emulsion part by using a polymer as its
stabilizing agent. The polymer, being water soluble, is virtually insoluble in the oil part of
the microemulsion and its dimensions prevents its inclusion into W/O droplets. It will,
hence, not interfere with the W/O microemulsion stabilization system.
As pointed out earlier liquid triglycerides do not lend themselves to microemulsion
formulation with the traditional technique. Addition of liquid triglycerides to the inverse
micellar solution results in a phase change to a lamellar liquid crystal. Hence, a different
strategy must be employed in order to prepare a microemulsion with triglycerides. One
solution is to attack the problem from the opposite side. Realizing that the microemulsions
are obtained by destabilizing a liquid crystal, figure 2.4, it appears reasonable to approach
the problem with a liquid crystal as starting point instead. This means forming a liquid
crystal containing triglycerides and destabilizing it by addition of a suitable compound. This
compound should be considered as a potent cosurfactant destabilizing the liquid crystalline
 16

phase. The common cosurfactants are, however, not useful because of their toxicity. The
destabilization was instead obtained by the use of hydrotropes [Friberg and Rydhag 1971].
These are compounds, the action of which is the destabilization of liquid crystals as was
early demonstrated [Lawrence and Pearson 1964]. A large number of them are allowed into
food products. The solution in the system of water/1-monocaprylin/sodium xylene sulfonate
is the largest one and it was used to dissolve a triglyceride, trioctanoin. The amount of
triglyceride dissolved was very indeed with a maximum of 13.5% by weight of triglyceride.
Other approaches have resulted in similar results. So, for example, does the system
water/monocaprylin/tricaprylin which was prepared by Ekwall [1975], shows very little
water solubilization into the oil. To reach 15% by weight of water, 50% of the stabilizers
were needed [Friberg and Kayali 1991].

Figure 2.4 In a combination of water, a surfactant and a long chain alcohol, the areas for solutions of normal
and inverse micelles, are separated by a lamellar liquid crystal [Friberg and Rydhag 1971].
 17

There are very few available examples of ingestible water-in-oil or oil-in-water

microemulsion systems for food products, even though much has been accomplished in
recent years in the general field of microemulsions. A water-in-oil (W/O) microemulsion is
suitable as the delivery system for water soluble nutrients and flavors in foods. An
ingestible, cosurfactant free system, with no off-taste or change in performance (soybean oil,
water and commercially available food surfactants such as polyglycerol oleate, polyglycerol
linoleate monoglyceride and polyoxyethylene sorbitol oleate) was chosen for this purpose.
The surfactants are evaluated based on their structures and performance in solubilizing water
in high triglyceride concentration range [El-Nokaly et al., 1991].
Microemulsions prepared by using triglycerides and other food components, poly-
oxyethylen(40) sorbitol hexaoleate, anonionic surfactant, were studied by Joubran et al.,
[1993], and soybean oil was used as a typical triglyceride.

2.4.1 Earlier work on food microemulsions applications:

Reports on milk fortified with vitamin A solubilized in a microemulsion have been
published [Duxbury 1988]. Dissolving essential oils in water with and without an alcohol
cosurfactant for aromatization of beverage or pharmaceutical formulations, have been
reported [Wolf and Havekotte 1989]. A new system for hydrolyzing milk fat was developed
by using a microemulsion system containing reversed micelles [Chen and Pai 1991].
Controlled hydrolysis of milk fat by lipase is applied in the dairy industry to produce
lipolyzed milk fat with butter-like or cheese-like flavor [Kilara 1985]. The microemulsion
single phase region of a peppermint oil/Tween20/water system which was studied by
Treptow [1971]. Friberg and Rydhag [1971] solubilized up to 15 wt% tricaprylin in an
isotropic aqueous solution of monocaprylin and a hydrotrope such as sodium xylene
sulfonate. Treptow [1971] dissolved less than 10% soybean oil in water with a 30/70
surfactant mixture of Tween 20 and G1045 respectively. Ascorbic acid was added to the
emulsified systems, such as microemulsions, and its stability against oxidation was studied
at 45 0 C in aerobic conditions and compared with that in aqueous solutions at different pH
values. All emulsified systems provided protection to ascorbic acid, as its degradation rate,
was slower in emulsified systems than in aqueous solutions [Gallarate et al., 1999].
The rate of oxidation of linolic acid and ethyl linoleate in O/W microemulsions was studied
by Carlotti et al., [1995]. The results showed that the structure of the interface in the
microemulsions was particularly important to protect the systems from auto-oxidation.
 18

An alcohol or acid cosurfactant was needed to form the microemulsion, with the exception
of the work done by Gulik and Larsson [1984] and Troptow [1971]. Such cosurfactants are
usually not acceptable for taste, safety, or performance reasons in ingestible oil
formulations. The sparcity of work available on water solubilization in triglyceride for food
applications is obviously due to difficulties inherent in the structure of oil, in finding
appropriate cosurfactants, and the need to use minimum amounts of food approved
surfactants not to adversely affect the oil properties [El-Nokaly et al., 1991].

2.4.2 Surfactants in food microemulsions

When the surfactant being used is safe for human consumption, microemulsion become
important in such fields as foods, cosmetics and pharmaceuticals [Kunieda and Shinoda
1982]. Surfactants in food are usually called emulsifiers whether their intended use is
emulsification or not. An Acceptable Daily Intake (ADI) value has been allocated to most
food emulsifiers by health authorities in many countries (FDA, FAO, EEC) [El-Nokaly et
al., 1991]. Chemically, most food surfactants are esters of fatty acids with naturally
occurring alcohols and acids. The primary raw materials for food surfactants production are
fats and oils which can be utilized directly or after having been hydrogenated, fractionated,
or split to fatty acids and glycerol [Lauridsen 1986], they have to impart no taste or smell on
foods. They are mostly nonionic surfactants with few exceptions such as succinic, citric, and
diacetyl tartaric acids esters of monoglycerides and soaps. Amphoteric lecithin is the only
food approved surfactant containing a positive charge. Surfactants listed in Table 2.3, were
tested for W/O microemulsion formation at 12:1 surfactant/water by weight. The oleate and
linoleate groups were preferred for testing because of their unsaturation and its known effect
in increasing fluidity of interfacial film. The sorbitan and sorbitol esters of oleate and
linoleate did not give microemulsions at surfactant/water ratio below 12:1. Branching should
be another factor in increasing the interfacial fluidity. The special kink in the linoleate
molecule which makes its alignment difficult, thus leading to an increase in the interfacial
fluidity. The presence of unsaturation was found to be important. The monooleate and
linoleate could act as cosurfactant since they are different in hydrocarbon moiety size
comparing with dilinoleate [El-Nokaly et al., 1991]. For legislative reasons synthetic
surfactants are in any case of limited use in food synthesis applications. A group of
surfactants which possess no such limitations are the naturally occurring phospholipid
(lecithin) surfactants which are biocompatible and legislatively acceptable [Svensson et al.,
1996].
 19

Table 2.3 Chemical Classfication of Food Emulsifiers and Legal Status (US FDA 21 CFR* )
General Class Example ADI Values** mg/kg
body wt/day
Partial Glycerides Mono- and diglycerides Not limited
Lactic acid esters of monoglycerides Not limited
Acetic acid esters of monoglycerides Not limited
Citric acid esters of monoglycerides Not limited
Diacetyl tartaric acid esters of 50
monoglycerides
Succinic acid esters of 3% by wt. of shortening;
monoglycerides 5% by wt. of flour
Fatty Acid Esters of Propylene glycol esters of fatty acids 25
Polyols
Polyglycerol esters of fatty acids 25
Sucrose esters of fatty acids 10
Sorbitan esters 25
Ethoxylated Emulsifiers Ethoxylated partial glycerides --

Polysorbates 25
Phosphatides, Lecithin Not limited
Phosphorylated
Partial Glycerides Fractionned phosphatides --
Phosphorylated monoglyceride --
Miscellaneous Sodium strearoyl-lactylate 20
Calcium stearoyl-lactylate 20
Salts of fatty acids (Na, K) Not limited
Sodium dodecyl sulfate

*
CFR Code of Federal Regulations. ** Acceptable Daily Intake (ADI) for man
 20

Lecithin is obtained for commercial purposes from oil seeds (in particular soja) and egg
yolk. It is a phosphatid and thus a fat-similar substance. In the lecithin molecule phosphoric
acid is connected on the one hand to a primary hydroxy group of glycerin bound and on the
other hand with choline figure 2.5. Soybean lecithin contains only to a third lecithin. It
contains further 15% inisitphosphatids and approximately 25% kephalin with other
connections. The emulsifying agent effect of the lecithins is based on the fact that
hydrophilically and the fatty acid remainders behave the zwitterion structure formed by the
phosphate and cholin hydrophob. Lecithin is used e.g. during the margarine production, in
order to decrease of the surface tension a fine distribution of the water phase [Baltes 1995].
O

O R1
O
H O
R2 O
O P O
CH3
O
_
N CH3
+
H3C

Figure 2.5 : Lecithin (general structure)[Baltes 1995]. R1 , R2 . typically normal aliphatic remainders
with 15 or 17 carbon atoms and up to 4 cis -double bounds.

Phosphatidylcholine (PC) is a naturally occuring surfactant which is excellent in terms of

biodegradability and human safety. In fact, many PCs have been utilized as fat emulsifiers
[Ishii et al., 1990]. However, few studies have been made on preparation and properties of
microemulsions with phosphatidylcholine [Kaneko and Shinoda 1988]. None of studies have
investigated the potential of non-alcohol cosurfactants to produce microemulsions with
lecithin. The most commonly used cosurfactants are simple low molecular weight alcohols
[Aboofazeli et al., 1994]. The ability of phospholipids to form microemulsions with alkanes
has been studied by several authors [Kaneko and Shinoda 1988, Shinoda et al., 1991]. Leser
et al., [1993] have studied the phase behavior and microstructure of soybean
phosphatidylcholine, medium chain triglyceride, and short chain alcohols.
 21

Lecithin has a fairly high critical packing parameter (CPP) of around 0.8 [Cornell et al.,
1986]. If lecithin containing systems are to form microemulsions over a reasonably wide
range of oil and water concentrations, the effective CPP of lecithin needs to be reduced. One
way this can be achieved is the addition of a short chain cosurfactant. A short chain
cosurfactant can act to reduce the effective CPP in two ways, firstly by its incorporation into
the interfacial film and/or secondly by dissolving in the aqueous phase thereby making it
less hydrophilic. In addition, in the case of a long hydrophobic chain surfactant like lecithin,
a short chain cosurfactant can also act to increase the fluidity of the interfacial surfactant
layer, thereby reducing the tendency of lecithin to form highly rigid films [Binks et al.,
1989] and allowing the interfacial film sufficient flexibility to take up the different
curvatures required to form balanced microemulsions [De Gennes and Taupin 1982].
Soybean lecithin is both strongly hydrophobic, due to long hydrocarbon chains, and strongly
lipophobic due to zwitterionic polar head groups. At room temperature soybean lecithin and
water form a lamellar liquid crystalline phase between 0.35 and 0.07 mass fraction of water
[Bergenstahl and Fontell 1983]. In water-oil systems lecithin is slightly too lipophilic to
form microemulsions spontaneously [Backlund et al., 1994]. This means that lecithin need
a cosurfactant such as a short-chain alcohol to form microemulsions [Shinoda et al., 1991]
It is important to note that the differences between grades of lecithin are considerably less
than the effect of changing the alcohol as cosurfactant [Aboofazli and Lawrence 1994]. The
major factor influecing the phase behavior of microemulsions prepared by lecithin as
surfactant is the nature and mixing ratio of the cosurfactant used [Aboofazeli et al., 1995].

2.5 Natural colorants

Recently the safety of food colorants has been a very controversial topic but most of the
criticism has been directed towards the synthetic colorants. The natural colorants have been
relatively free of criticism may be due to the blief that most are derived from food sources
that have been consumed for many years [Francis 1996]. According to scientific
investigations humans take up the largest part of their information visually. Also, food is
assess first on the basis their outside appearance. A part from form and general presentation
the color contributes substatially to the quality assessment by the consumer. For different
product groups a different expectation attitude developed concerning the chroma. Generally
products with a more intensive coloring are preferred. Processing or storage of a product can
lead by oxidative or enzymatic influences to the impairment of the natural colors. In order to
arrange the product optically responding, in the context of the legal regulations coloring
 22

materials can be added to food. Other food, which is colorless due to their production
process, e.g. lemonade beverages, is colored for the rise of benefit value. The synthetic
colorants used in the food-stuffs industry possess an intensive coloring strength with at the
same time good solubility, are more stable and easier to process than natural food colorants.
The application type of natural food colorants are however so far limited, since they have
some disadvantages in comparing with the artificial colorants [Täufel et al., 1993]. They are
however relatively safe dealing with toxicological check and incompatibility reactions
[Baltes 1995].
Natural colors are organic colorants that are derived from edible sources using reconized
food preparation methods. One of the advantages of using natural colors is that they are
generally more widely permitted in food-stuffs than synthetic colors. Contrary to many
reports, natural sources can provide a comprehensive range of attractive colors for use in the
modern food industry. However, although the color of fruits and vegetables can vary during
the season and processing can cause color loss, food manfacturers need to ensure uniformity
of product appearance [Henry 1996]. Also colors may be added to foods to give color to
certain foods such as sugar confectionery, ice lollies and soft trinks, which would otherwise
be virtually colorless [Food Advisry Committee, FdAC/REP/4, London 1987].

2.5.1 Chemistry of natural colorants

The coloring of chemical connections is based on the absorption and emission of light. The
visible light (approx. 400-800 Nm) is electromagnetic radition, by the wavelength ? and the
frequency ? to describe leaves itself (2.1) [Skoog and Leary 1996]:

? .? = c (2.1)

c is the speed of light (in vacuum around 2,99 . 1010 cm . s-1 ).

E=h.? (2.2)

A light quantum has the energy (2.2) whereby h the Planck`quantum of action is (6,626 .
10-34 Js) [Skoog and Leary 1996]. If light quantum of suitable frequency meets a molecule in
the initial state, it can become lively, i.e. suitable electron is raised to a higher energy level.
 23

In order to arrive again into the energy-poorer and thus more stable initial state, the
molecule is anxious to deliver the taken up energy again. The put on condition must remain
received a certain time, in order to cause coloredness. The material appears in the
wavelength the corresponding complementary color. If a connection absorbs e.g. in the
violet one (around 420 Nm), then it appears for the human eye yellow [Breitmaier and Jung
1983] figure. 2.6 shows a spectrum of the visible light.

Figure 2.6: Spectrum of the visible light [Breitmaier and Jung 1983].

Regarding the orbit of molecule involved four different electron transitions can be
differentiated [Breitmaier and Jung 1983; and Skoog and Leary 1996] (Figure 2.7)

n? ? * -Transition
If electron of a non-binding a few of electrons becomes lively on an anti-binding ? * -
orbital of a double bond, one speaks of a n ? ? * -transition. These transitions are in all
connections with hytroatomen and neighouring double bond possible, e.g. carbonyl or azo
connections. The n ? ? *-transition probability is only small. Therefore the appropriate
absorption bands are likewise small.
 24

? ? ? * -Transition.
Transition is raised an electron to a ? * -orbital, speaks one of a ? ? ? * -transition This type
from transitions is in all insatiated connections possible, so in alkenes, alkines, aromatics
and further in transition probability is very high. Therefore the appropriate absorption bands
are intensive.

n ? ? *-Transition
If n-electrons to anti-binding ? *-orbitals are a neighbouring linkage, one speaks lively of
n ? ? *-transitions. Generally, they can be caused by radiation of the region between 150
and 250 Nm. They are in satisfied connections with heteroatomen possible, e.g. with
alcohols, amines or with alkyl halides.

? ? ? *-Transition
If ? -electrons become lively to transition in the appropriate anti-bending ? *-orbital, one
speaks of ? ? ? *-transitions. They are possible in all organic compounds. The energy
which can be spent is very large and corresponds to radiation frequencies within the vacuum
ultraviolet area.

Figure 2.7: Electronic molecule energy levels [Skoog and leary 1996].
 25

Substituents, which shift the absorption to longer-wave light, are called Chromophore
(chromophore groups). They possess bonding electrons with relatively low excitation
energies. When the absorption capacity is the larger, the electron system is the more
expanded. One differentiates two types from chromophors (Figure 2.8).

C O N N NO2 N O

? -Chromophore (double- triple bonds, aromatics)

C C C C

Figure 2.8: Examples for chromophor groups

Usually, it requires several chromophor groups in the molecule, in order to shift the light
absorption into the visible spectral region [Täufel et al., 1993]. The conjugation of electrons
leads to a further delocalization, which degrades the ? *-energy levels. Consequently the
absorption maximal shift to longer wavelengths [Skoog and Leary 1996]. The increase of
? max one defines as red or bathochrom shift. Fewer significantly is the shift absorption
maximum to shorter wavelengths. The so-called blue shift is probably due to an increase of
solvating the free pair of electrons, whereby the energy of the n-orbital is lowered [Skoog
and Leary 1996]. Auxochrome (auxochrome groups) are substituents with n-electrons,
however without linkages, which extend the conjugation system of the chromophors and
thus for an increase of the chroma contribute. Examples for this are hydroxyl and the amino
group (also substitutes) [Schröter et al., 1995].
 26

2.5.2 Stability of natural colorants

Natural colorants are found in many plants and animals. They fulfil different functions e.g.
attracting from insects to dusting the blooms. Natural colorings were used since ever ago for
coloring (e.g. in the classic antiquity magenta for very valuable textiles). Due to their
molecular structure natural colorings are sensitive for the light and oxygen. The conjugation
of chromphors (Figure 2.9), frequently occurring in natural colorants, lead with appropriate
suggestion by light to the delocalization of ? -electrons. From it results a theoretical
„turning“ of the double bonds, due to whose different mesomer phases can be trained.

OH
+
HO O
X-
OR
OH

Figure 2.9: Natural colorings (ß-Carotin, a Carotinoid; Pelargonidin, an Anthocyanidin if R = H, an

Anthocyan if R = sugar remainder, X = halide, usually Cl)

The developing partial charges favors a depositing of oxygen radicals. Thereby the
delocalization of the electrons is limited. The consequence is a reduced light absorption
readiness and thus a weaker color impression. The reaction mechanism wide on the basis the
molecular structure represented by curcumin (Figure 2.10).
 27

H
O O

H3CO OCH3

HO OH

H -
O O
+
H3CO OCH3

HO OH

H -
O O
+
H3CO OCH3
O
O
HO OH

Figure 2.10: Suggestion of natural colorings by UV-light and attack by acidity radikal for the
example of curcumin.

2.5.3 Curcumin
The yellow coloring curcumin originates from the roots of the turmeric plants resident in
South-east Asia (bot.: Curcuma longa). It is produced by extraction of the degreased root.
Chemical formula of curcumin is C21 H20 O6, molecular weight is 368.4 and color shade is
lemon yellow at pH3, orange at pH10 [Food Advisory Commettee, FdAC/REO/4, HMSO,
London 1987]. It is discussed that curcumin has an anti-carcinogenic effect [Falbe and
Regitz 1990]. Curcumin has been shown to inhibit lipoxygenase and cycloxygenase [Huang
et al., 1991] and thereby inhibits the superoxide generation and prevents tumour promotion
[Brian and Cizela 1989]. Curcumin has been successfully used in several experimental
carcinogenesis models. Recently, curcumin has also been shown to inhibit protein kinase C,
and thereby inhibits oncogene activation [Reddy and Aggarwal 1994].
 28

It has been suggested that the hydroxy groups on the benzene rings, double bonds in the
alkene portion of the molecule and/or the central B-diketone moiety could be responsible for
the high biological activity of curcumin [Ruby et al., 1995].
Curcumin also increased the level of glutathione which would prevent thiol depletion
occurring during apoptosis [Jaruga et al., 1998]. Curcumin is an interesting molecule
because of the variety of biological effects. It possesses in addition to its potent anticancer
activity. However, its exact mechanism of action is not very clear. It suggested that its
anticancer activity could be mainly due to its ability to induce apoptosis in tumor cells [Khar
et al., 1999].

OH O
MeO OMe

HO OMe

Figure 2.11. Curcumin structure [Baltes 1995].

It has been shown that curcumin has a poor light stability. The stability of curcumin was
strongly improved by lowering the pH or by adding glutathione, N-acetyl-L-cysteine,
ascorbic acid and rat liver microsomes [Cooper et al., 1994]. The oral LD50 for curcumin in
mice is higher than 2g [Srimal and Dhawan 1973]. Monkeys fed 500 mg/kg/day for 9
months of turmeric showed no adverse effects [Bhavanishanker et al., 1986].
The solubility is very different as a function of solvents: Curcumin is solubly in vegetable
oils, however insolubly in water, in ethanol solve itself 3 g.L-1 . The light fastness is small,
particularly in solve form in products with high water activity . The coloring agent is heat-
proof until 150 0 C. It possesses a good acid resistance. Above a pH value of 9 a color
changes from yellow to orange. The molecular structure is represented in figure 2.11. The
coloring agent is certified only under marking for certain food. Curcumin uses among other
things for the coloring of curry, mustard, sweet goods, eis-crem, fruit and vegetable prepared
and fine pastry. The admissible in each case are maximum quantites thereby to consider
(guideline 94/36/EG) [Falbe and Regitz 1990]
 29

3 Objectives

In recent years, microemulsions have been given considerable attention due to their special
important physicochemical properties such as their transparent or translucent appearance,
long term stability, high solubilization capacity for lipophilic and hydrophilic substances,
and ease of preparation. Because of that microemulsions are very important in various
applications of commercial industries.
Huge number of papers have been published in microemulsion area. Unfortunately, most of
work to date studying microemulsions used oils, surfactants and cosurfactants unsuitable for
food applications and there are very few available examples of microemulsion systems for
this purpose. In order to make these systems , it is necessary to formulate such systems
using nontoxic and safe ingredients. The problem is how to prepare microemulsions with
harmless oils, surfactants, and cosurfactants for possible applications in food products.
The present investigation aimed to prepare microemulsions by using inexpensive and
suitable components for food applications. The microemulsions can be used as medium for
water and oil soluble nutrients because they contain both aqueous phase and oil phase.
Curcumin is very important since it can be used as anticarcinogenic, natural food colorant
and antioxidant. Curcumin is insoluble in water and sensitive towards. For these reasons this
study carried out also to investigate the ability of these microemulsions for solubility of
curcumin and to study the stability of curcumin solubilized in them against light during
storage, making stress on the following items :
- The preparation of various microemulsion systems.
- The determination of curcumin solubility in these systems.
- Study the stability of curcumin solubilized in these microemulsions against light by
determination the effect of UV light and normal electric light on its color shade during
storage.
 30

4 Materials and Methods

4.1 Materials

4.1.1 Chemicals used in microemulsion preparation

The chemicals generally used in the preparation of microemulsions are four: surfactant,
cosurfactant, oil and water. The microemulsions are formed mainly by using at least three
components an oil, a water and a surfactant. Surfactants are necessary for the solubility of
water in oil, forming therefore a microemulsion. Some surfactants are more soluble in oil
than in water or vice versa, in which case the addition of a cosurfactant can assist the
obtaition of a balanced microemulsion.

4.1.1.1 Surfactants
It is well known that oils are insoluble in water. Surfactant or mixture of surfactant and
cosurfactant are usually located at the surface separating two immiscible liquids for the
stability of the mutual dispersion [Bourrel and Schechter, 1988]. A surfactant phase with a
bicontinuous structure consists of interwined water and oil lamellae separated by a
surfactant layer with low or zero curvature, hence it has the highest mutual solubility
between oil and water [Sonesson et al., 1991]. Surfactants are nonionic, ionic or cationic.
All surfactants used in this study were nonionic, because as reported by Costantinides the
nonionic surfactants are known to be less affected by pH and ionic stregth changes
[Costantinides and Yiv, 1995] and some data have observed that nonionic surfactants offer
the advantage of a high chemical stability [Hofland et al., 1994]. Microemulsions stabilized
by nonionic surfactants may need no cosurfactants [Aboofazeli et al., 1994].
Three types of surfactants were used: lecithin, monoolein and Tween20. These surfactants
were selected because they are safe for food uses and with the exception of Tween20, the
acceptable daily intake (ADI) is not limited. Moreover, they are inexpensive and available
compared to other surfactants.
 31

4.1.1.1.1 Lecithin
Phosphatidylcholine is a naturally occurring emulsifier excellent in terms of
biodegradability and human safety [Ishii et al., 1990]. In this study 3-sn-phosphatidylcholine
(lecithin) from soybean containing 40% phosphatidylcholine was used as surfactant to
prepare the microemulsions. It was purchased from Fluka and was used as received.
Aboofazeli and Lawrence [1994] have investigated the influence of purity of lecithin on
producing balanced microemulsions and they noted that the differences between grades of
lecithin are considerably littile. Therefore, it is not unreasonable to use lecithin containing
40% phosphatidylcholine to prepare food microemulsions where it is commercially
available and inexpensive.

4.1.1.1.2 Tween 20
Tween 20 (Polyoxyethylen-sorbitan-monolaurat) was obtianed from Fluka. It was selected
because it can be used in food products with acceptable daily intake (ADI) 10 mg/k. It also
has a high ability for solubility of aqueous phas in peppermint oil to prepare transparent
microemulsion without a cosurfactant. In addition it is able to dissolve curcumin which is
used as natural coloring material, easily available and economical.

4.1.1.1.3 Monoolein
Monoglycerides are nonionic surfactants widely used as emulsifiers in the food and
pharmaceutical industries [Holmberg and Österberg 1988]. DL-? -Monoolein (Glycerin-1-
monooleate) used in this work, consists of 38-40% diglyceride, 16-18% triglyceride and 4-
5% free glycerol in addition to the essentail component (mono-olein). It was purchased from
Fluka. Its Selection was due to the suitability of its uses in food, its low cost and could be
used both as surfactant and as oil.

4.1.1.2 Oils
The oils used in this study were vegetable oils such as peppermint, soybean, peanut and
rapeseed oil. As reported by Winkler [1992], oils obtained from natural sources and their
derivatives, e.g. triglycerides, are easily degraded by microorganisms and are considerd to
be harmless to the environment. Such microemulsion systems containing these oils are but
infrequently published in the literature [Paul and Moulik 1991]. Triglycerides are preferred
 32

as they have been used in intraveous fat emulsions for more than 20 years and are therefore
well characterized from the toxicological aspect [[Davis et al., 1983].

4.1.1.2.1 Peppermint oil

This oil is extracted from Menthe piperta L., plant and it was purchased from Fluka. This is
a special oil due to its contents of 30-40% menthol, 15-32% menthone and 3-10% menthyl
acetate. It was selected for this work because due to its effectivness in microemulsion
preparation by using lecithin or Tween20 as surfactants without or with cosurfactants. It
could used peppermint oil alone or mixed with a common edible oil as lipophilic phase. In
addition, it quickens the solubility of lecithin in edible oils where the lecithin solubility is
very difficult.

4.1.1.2.2 Edible oils

Edible oils which used in this work are soybean, peanut and rapeseed oil which were
purchased from Fluka. Soybean oil contains 41% saturated and 84% unsaturated fatty acids;
peanut oil 17% saturated and 83% unsaturated fatty acids; and rapeseed oil 7% saturated and
92% unsaturated fatty acids. These oils are a suitable oil phase in microemulsions for food
applications. Microemulsions containing these oils are important and interesting where there
is a demand for environmentally more acceptable formulations.

4.1.1.3 Cosurfactants
The most commonly used cosurfactants are simple low molecular weight alcohols. To data,
however, very little work has investigated the use of other cosurfactants. The presence of the
cosurfactant in the interfacial region causes the destruction of the liquid crystalline phase
(that prevents the formation of microemulsion) thereby allowing the production of balanced
microemulsions [Aboofazeli et al., 1994]. Alkanols as cosurfactants dissolve mainly in the
interfacial film and thereby, change its properties. Hitherto, two models have been proposed.
The first model is based on the “wedge model“ suggested first by Langmuir [1916] who
assumed that the interfacial layer possesses a natural curvature. On the basis of this model,
Mitchell and Ninham [1981] assumed alkanol molecules to intrude the interfacial layer
thereby, changing its curvature. In the second model, proposed by De Gennes and Toupin
[1982]. It is suggested that the alkanol molecules adsorb preferentially at the region of
strong curvature of the interfacial layer, thereby reducing its rigidity [Kahlweit et al., 1995].
 33

The role of cosurfactant together with the surfactant is to lower the interfacial tension down
to a very small even transiet negative value at which the interface would expand to make
fine dispersed droplets, and subsequently adsorb more surfactant and surfactant/cosurfactant
until their bulk condition is depleted enough to make interfacial tension positive again. This
process known as “spontaneous emulsification“ forms the microemulsion [Kunieda et al.,
1988]. In the production of a balanced lecithin microemulsion, the cosurfactant has an
additional role in that it can also act to lower the tendency of lecithin to form a highly rigid
film [Binks et al., 1989].

Ethanol
Absolute ethanol was obtained from Merck. It was used in this work as cosurfactant to
prepare microemulsions in particular by using lecithin or monoolein as surfactant. In
addition, using it assists the solubility of lecithin in edible oils. It is known alcohols, in
general, are unsuitable for food uses, but using a few amount (about 5%) of ethanol in food
products is acceptable [Osborne et al., 1991].

4.1.1.4 Water
Water was used after double destillation because the hard water may to be difficult in the
preparation of microemulsions. Water was used as aqueous phase to form most
microemulsion systems.

4.1.2 Other Chemicals

4.1.2.1 Curcumin
Curcumin was used in this study as natural coloring material for the determination of its
solubility in microemulsions and stability of curcumin solubilized in these microemulsions
against light. Curcumin was supplied by Dr. Marcus GmbH.

4.1.2.2 Sodium chloride, Sucrose and Citric acid

These chemicals were obtained from Merck and were used as electrolyte to determine the
effect of them on the preparation of microemulsions and also their effect on the solubility
and stability of curcumin dissolved in microemulsions.
All the chemicals were used as received.
 34

4.1.3 Equipments and Instruments

4.1.3.1 Lamps

4.1.3.1.1 Normal electric lamp

Type: elite, q 528, 100 W, 230 V. This lamp was used to give a similar light at all times for
the samples stored under light and to know the effect of light on the curcumin dissolved in
microemulsions durnig storage.

4.1.3.1.2 Ultraviolet (UV) lamp

Luminous element: OSRAM Dulux L; 24 W / 41. This lamp was used to give Ultraviolet
light for the samples stored in order to determine the stability of curcumin solubilized in
microemulsions.

4.1.3.2 Color Measuring Instrument

Minolta Chroma-Meter CR-310.
Type: Remmissions-color measuring instrument after three scales-processes for object
colors.
Measuring receiver: 6 silizium-photodiodes (two flash processes) filtered for the primary
color sensitivy values of green, red and blau light.
Light saurce: Xenon-flash tube (PXA).
Measuring field ? : 50 mm.
This instrument was used in this study to measure the color shade of curcumin solubilized in
microemulsions before and after storage under various light fields.

4.1.3.3 PH-Meter
Type: Knick, pH-Meter 765 Calimatic
It was used to measure pH-value of prepared microemulsions to know the effect of pH-value
on the stability of curcumin.
 35

4.1.3.5 Vortex mixer

Type : IKAMAG REO
It was used to shake the mixtures during the prepration of microemulsions. This mixer work
by using a electric magnetic Vortex. Another type of mixer (JANKE & KUNKEL, IKA-
Labortechnik, VF2) was used to shake the test tube during the solubility of curcumin.

4.2. Methods

4.2.1 Preparation of Microemulsions

Microemulsions were prepared in three systems defined on the basis of surfactant type:
?? Lecithin-microemulsion
?? Monoolein-microemulsion
?? Tween20-microemulsion.

4.2.1.1 Lecithin-microemulsion
Lecithin-microemulsion was prepared by using either peppermint oil or mixture of
peppermint and an edible oil as lipophilic phase and water as aqueous phase without or with
ethanol as cosurfactant. In case of lecithin-microemulsion containing peppermint oil and
water, a three ratio of peppermint oil and lecithin were used (1:1, 2:1 or 3:1). A mass of 10g
of lecithin was weighed in a 250 ml beaker and the appropriate amount of peppermint oil
(10, 20 or 30g) was added to it. In case of using ethanol, various percentages (2, 4, 6, 8 or
10%) were separately added to lecithin and oil mixture. The beaker was then closed and
allowed to stand overnight to dissolve the lecithin in oil. To reduce or shorten the time of
lecithin dissolution in oil, the Vortex mixer can be used to shake the mixture. The use of
ethanol also assists in mixture time reduction. After dissolving lecithin in oil, the mixture
was titrated with water by using a 10 ml syringe equipped with an injection needle (to
control the amount of added water) until onset of turbidity. The mixture was shaken after
each addition of water for a short time (about 1 min) by hand or by using a Vortex mixer.
The amount of water which is used to prepare the microemulsion was determined by
weighing before and after the titration. The experiment was carried out at room temperature
(20 ? 2 °C). Since only the single phase microemulsion region was of important no attempt
was made to obtain the other phases.
 36

Lecithin-microemulsion containing peppermint oil mixed with an edible oil was prepared at
three ratios of peppermint oil, edible oil (soybean, peanut or rapeseed oil) and lecithin
(1:1:2, 1:1:1 or 1:2:1). In this case, the appropriate amount of lecithin (10g) was weighed in
a 250 ml beaker and a fixed amount of peppermint oil (5g in case of ratio 1:1:2 or 10g in
case of ratio 1:1:1 and 1:2:1) was added. An amount of edible oil (5g in case of ratio 1:1:2
or 10g in case of ratio 1:1:1 and 1:2:1) was then added to the mixture of peppermint oil and
lecithin. In case of using ethanol an appropriate percentage (2, 4, 6, 8, 10, 12 or 14% by
weight) was added in this stage. Since lecithin does not easily dissolve in edible oils, and
the prepration of microemulsions by using these compounds with lecithin was difficult, the
addition of peppermint oil with the proper ratio was important. Moreover, using peppermint
oil favours the dissolution of lecithin, likewise is the use of ethanol. In addition, using either
peppermint oil or ethanol increase the fluidity of the mixture and reduces the production of
foam during the addition of water to lecithin. After incorporating peppermint oil, edible oil
and lecithin in the proper ratio (1:1:2, 1:1:1 or 1:2:1) without or with ethanol, the beaker was
closed and allowed overnight or shaken by using a Vortex mixer for about 2 hours to
dissolve the lecithin in oil. The mixture was then titrated with water by using a 10 ml
syringe equipped with an injection needle until the onset of trubidity. The mixture was
shaken for a short time (about 2 min) by the hand or by using a Vortex mixer after each
addition of water. The amount of water required to prepare a transparent microemulsion was
determined by weighing before and after titration.

4.2.1.2 Monoolein-microemulsions
Monoolein-microemulsion systems could not be produced by using monoolein alone or
mixed with soybean oil, but the addition of ethanol as cosurfactant is necessary to solubilize
the water for the preparation of a transparent and stable solution. Using Tween20 also as
another surfactant was applicable for water solubility in monoolein alone without soybean
oil. In case of using monoolein alone as surfactant and as oil, an amount of 10g was weighed
in a 250 ml beaker and different percentages of ethanol (4, 8, 12, 16% by weight) were
singly added to the monoolein. The mixture was then titrated with water by using a syringe
equipped with an injection needle and the titration was continued as long as the mixture was
transparent. The mixture was shaken for a sufficiently long time after each titration by a
Vortex mixer to solubilize the water. The amount of water was determined by weighing the
beaker before and after titration. In case of using soybean oil, a fixed amount (5g) was added
 37

to a fixed amount of monoolein (5g) and the different pecentages of ethanol (4, 8, 12 or
16%) were then added. The mixture of monoolein, soybean oil and ethanol was titrated with
water by using a syringe until trubidity appearance.
The mixture was shaken after each titration of water by hand or by using a Vortex mixer for
a sufficient time (about 1 min.). When using Tween20 with monoolein, a desired amount of
monoolein (5g) was taken in a 250 ml beaker and a same amount of Tween20 was added.
The mixture was titrated with water by using a syringe until the appearance of trubidity. The
mixture was shaken after each addition of water for a short time and the maximum amount
of water required to form a transparent and stable mixture was determined by weighing the
mixture before and after the titration with water. The use of ethanol with the mixture of
monoolein and Tween20 lead to the increase of water solubility, in this case, different
percentages of ethanol (4, 8, 12 or 16%) were added to the mixture containing 5g of
monoolein and 5g of Tween20 and the titration was done with water.

4.2.1.4 Tween20-microemulsions
Tween20-microemulsions were prepared by using peppermint oil as lipophilic phase and
water or an aqueous solution of 20% of either NaCl, sucrose or citric acid. The ratios of
peppermint oil and Tween20 used in the preparation of micremulsions were 1:1, 1:2, 1:3,
1:4, or 1:5. The aqueous solution of NaCl, sucrose or citrc acid was prepared by dissolving
20g of the solid substance in 80g water.

A known weight of peppermint oil (10g) was weighed in a 250 ml beaker and different
amounts of Tween20 (10, 20, 30, 40, or 50g) were separately added to make the proper ratio
of peppermint oil and Tween20. The mixture of peppermint oil and Tween 20 was then
titrated with water or with an aqueous solution which was prepared previously and the
titration was continued until production of transparent microemulsion with a maximum
amount of water or aqueous solution and before the appearance of trubidity. It is to be
emphasized that there is a first trubidity followed by a phase of clearness and then a next
trubidity. The mixture was shaken after each addition of aqueous phase for a short time by
using a Vortex mixer. The amount of aqueous solution required to prepare microemulsion
was calculated by the difference between the weight of the mixture before and after titration.
 38

4.2.2 Solubility of Curcumin

Solubility of curcumin in this reseach means the maximum amount of curcumin dissolved in
the dissolution medium without sediment.

4.2.2.1 Solubility in solvents

Curcumin (1 mg) was added to a test tube with 10g of solvent ( water, ethanol, Tween20,
peppermint oil, soybean oil,....etc.). Curcumin weight was gradually increased under
vigorous shaking effect of a Vortex shaker after each addition for a long time to dissolve the
curcumin. The addition of curcumin was continued until the appearance of undissolved
curcumin in the bottom of the test tube. The amount of dissolved curcumin was quantified
by subtracting the weight of dissolution medium before the addition of curcumin from its
weight after dissolving the curcumin.

4.2.2.2 Solubility in mixtures before aqueous phase addition

In this case, a stock solution of oil, peppermint oil alone or mixed with soybean, peanut, or
rapeseed oil and a surfactant such as lecithin or Tween20 at the appropriate weight ratio was
prepared. This stock solution was used either alone or mixed with a known weight of
ethanol. Also, a stock solution of aqueous phase, either water or an aqueous solution of 20%
NaCl, sucrose or citric acid and Tween20 was prepared. In all cases, a certain amount of the
prepared stock solution (10g) was weighed into a test tube and then an a little amount of
curcumin (1 mg) was added as described previously.

4.2.2.3 Solubility in microemulsions

First the microemulsion systems were prepared in a 250 ml beaker by using a fixed ratio of
oil, surfactant, cosurfactant (if used) and various contents of aqueous phase. A certain
weight of microemulsion system (10g) was taken in a test tube and a little weight of
curcumin (1mg) was then added and the dissolution was carried out as previous. The
dissolved amount of curcumin was estimated by difference between the weight of
microemulsion before and after the dissolving of curcumin.
Note : All procedures for prepration of microemulsions and solubilization of curcumin were
carried out at room temperature (20°C + 2°C) also were perfomed with three repetitions and
the average was calculated.
 39

4.2.3 Determination of the pH-value

After the dissolution of curcumin in the microemulsion, 5 g was taken from each sample in a
test tube to measure the pH-value by using pH-Meter.

4.2.4 Light stability of curcumin

Stability of curcumin solubilized in microemulsion against light was determined in this work
according to the change of color shade (h°) during storage under UV lamp, normal electric
lamp or in the darkness.
The description of color by the use of the wavelength alone is unreliable because this
method gives insuficient informations. Part of the spectrum cannot be completely absorbed
and the rest is scattered or reflexed [Breitmaier and Jung 1983]. Several systems were
developed for the precise determination of colors.
The L* a* b* system describs the color by using of three dimensional coordinate systems.
The color shade (h°) can be determined from a* b*-values according to the following
equetion:

h° = arctan (b* / a*) (4.1)

The L*-value indicates the lightness and it forms a third virtical axis to a* and b* axis.
 40

Figure (4.1) shows the a* and b* of color slab for the L* a* b*-system.
 41

4.2.4.1 Determination of the color shade

The color shade of curcumin was measured by using the Color Measuring Instrument (L* a*
b*-system) according to the above mentioned equation (4.1). The preparation of
microemulsion systems was carried out as previous method which was described in section
preparation of microemulsions, with exception, the aqueous phase content here was fixed.
Such content was 3% in case of lecithin- and monoolein-microemulsion, but in case of
Tween20-microemulsions, the aqueous phase content was 11% or 54.5% in ratio 1:1 or 1:4
of peppermint oil and Tween20 respectively. When using the ethanol with the
microemulsions based lecithin, the content was 5%, and with the monoolein-microemulsion
containing monoolein either alone, mixed with soybean oil or mixed with Tween20, the
ethanol content was 10%.
The solubilization of curcumin was perfomed as described method in section solubility of
curcumin in microemulsions with exception the curcumin content here was fixed (0.2% in
case of monoolein-microemulsions and 0.4% in other microemulsions
After that, each sample was divided into three parts in a small petri dishes each one
containing about 6 g and then the color shade (h°) was measured by using Color Measuring
Instrument before storage of the samples (Zero time).
The first part of petri dishs was stored in the dark place and was tightly covered with
aluminum seals for 60 days.
The second part was placed under normal electric lamp (100 W) for 60 days
The third was kept under UV lamp (OSRAM Dulux L; 24 W/41) for 30 days.
The places for the second and third parts were equipped with vintilator to remove the warm
air by the lamps to avoid generated temperatures increase during storage.
The samples were withdrawn to measure the color shade of curcumin solubilized in
microemulsions at gradually increased intervals. The measurments of curcumin color shade
were carried out by using Color Measuring Instrument which gives three values L*, a* and
b* and from a* and b* values the color shade can be calculated acorrding to the above
equation (4.1). The petri dishes were covered with their covers during storage and during
measuring of hue moreover, they were also covered with additional black cover during the
measurment processing.
 42

The resultant value indicates the color of sample such as the range from 105 to 95 means
slight yellow, from 95 to 80 is yellow, from 80 to 70 indicate orange, from 70 to 60 is slight
red, and from 60 to 45 is deep red as shows in color slab (Figure 4.1). The values of color
shade for the samples in this study ranged from 105 to 45.
Concerning the control was selected two mediums for this reason , the first was prepared by
using the ethanol diluted with water at 4:1 then the curcumin was solubilized in this medium
either at ratio 0.4% and it was used with lecithin-microemulsions or at ratio 0.2% and it was
used with monoolein-microemulsions. The second medium of control was prepared by using
Tween 20 diluted with water at ratio 1:1 and the curcumin was solubilized at ratio 0.4%and
it was used with Tween20-microemulsions.
 43

5 Results

Microemulsions were prepared in three systems defined on basis the type of the surfactant.
The first system is lecithin microemulsions which prepared as the following content
formulations: lecithin/pepperment oil/water or lecithin/peppermint oil/an edible oil/water
without or with ethanol in either cases.
The second is monoolein microemulsions which prepared as the following content
formulations: monoolein/water/ethanol, monoolein/soybean oil/water/ethanol or monoolein/
Tween20/water without or with ethanol.
The third is Tween20 microemulsions which prepared by the following content
formulations: Tween20/pepperment oil/water or Tween20/peppermint oil/an aqueous
solution 20% of NaCl, sucrose or citric acid.
Solubility of curcumin was determined in solvents which used in the prepration of
microemulsions and also in the prepared microemulsions. Stability of curcumin dissolved in
microemulsion systems against light was determined as the change in its color shade (h°)
during storage under UV-light, normal electric light or in darkness at room temperature for
30 day in case of UV-light or 60 days in other cases.

5.1 Preparation of microemulsions

5.1.1 Lecithin/peppermint oil microemulsions

Figure (5.1) shows that microemulsion can be prepared by using peppermint oil, lecithin
(with various ratio; 1:1, 2:1 and 3:1) and water without or with ethanol. In case of the system
free ethanol, maximum amount of water required to prepare microemulsion was 9, 15 and
11 in three various ratios of peppermint oil/lecithin respectively. In case of the system
prepared with ethanol, maximum amount of water increased when using ratio 1:1 of
peppermint oil and lecithin where was 18% by using 4% of ethanol after that the maximum
water decreased with increasing the ethanol. On the other hand, the maximum water
decreased in case of ratios 2:1 and 3:1 of peppermint oil and lecithin with increasing the
ethanol. The highest amount of water was obtained by using ratio 1:1 of peppermint oil :
lecithin (4% of ethanol), while the lowest amount of water was obtained by using ratio 3:1
of peppermint oil : lecithin
 44

20

15
Pe:Le 1:1
Water%

10 Pe:Le 2:1
Pe:Le 3:1
5

0
0 2 4 6 8 10
Ethanol%

Figure 5.1 Optimum amount of water to prepare microemulsion by using peppermint oil (Pe), lecithin (Le)
without or with ethanol

5.1.2 Lecithin/peppermint/soybean oil microemulsions

It should be noted from figure 5.2 that microemulsion was prepared by using water,
peppermint oil mixed with soybean oil and lecithin with or without ethanol as cosurfactant.
The water content was relatively small in the system free ethanol. On the other hand, when
using ethanol the water content increased at first with increasing the ethanol content in three
ratios of peppermint oil/soybean oil/lecithin, after that decreased, where it reached 23%, in
case of ratio 1:1:2 and using 12% of ethanol, and 17% in case of ratio 1:1:1 and using 6% of
ethanol and 12% in ratio 1:2:1 and using 8% of ethanol after that decreased. It should be
also shown that the highest water content was by using ratio 1:1:2 and the lowest content
was by using ratio 1:2:1 of peppermint oil/soybean oil/lecithin either with or without

25

20
Pe:So:Le(1:1:2)
Water %

15
Pe:So:Le(1:1:1)
10
Pe:So:Le(1:2:1)
5

0
0 2 4 6 8 10 12 14
Ethanol %

ethanol.

Figure 5.2 Optimum amount of water to prepare microemulsion by using peppermint oil (Pe), soybean oil (So)
and lecithin (Le) without or with ethanol.
 45

5.1.3 Lecithin/peppermint oil/peanut oil microemulsions

From the results given in figure 5.3 shows that the water content required to prepare
microemulsion without ethanol was small in case of ratios 1:1:1 and 1:2:1 of peppermint
oil/peanut oil/lecithin, and it was somewhat higher in case of ratio 1:1:2. On the other hand,
by using ethanol the water content increased at the first with increasing the ethanol where
the maximum contents of water were 20%, 14% and 13% in case of ratio 1:1:2, 1:1:1 and
1:2:1, by using 12%, 10%and 6% respectively, after that they decreased with increasing the
ethanol.

25

20
Pe:Pn:Le(1:1:2)
Water %

15
Pe:Pn:Le(1:1:1)
10 Pe:Pn:Le(1:2:1)

0
0 2 4 6 8 10 12 14
Ethanol %
Figure 5.3 optimum amount of water to prepare microemulsion by using peppermint oil (Pe), peanut oil (Pn)
and lecithin (Le) without or with ethanol.

5.1.4 Lecithin/peppermint oil/rapeseed oil microemulsions

The microemulsion could be prepared by using peppermint oil combined with rapeseed oil
as lipophilic phase, water and lecithin with or without ethanol as cosurfactant. The water
contents by using three ratios of peppermint oil, rapeseed oil and lecithin (1:1:2, 1:1:1 and
1:2:1) with or without ethanol were illustrated in figure 5.4. It should be observed that the
water content in system free ethanol was smaller in case of ratio 1:2:1 than that in other
ratios. By using ethanol, the water content increased at the first, after that decreased with
increasing the ethanol in all ratios of peppermint oil/rapeseed oil/lecithin. It should be noted
also that the maximum content of water was obtained by using ratio 1:1:2 and by using 12%
of ethanol. While the maximum contents in ratios 1:1:1 and 1:2:1 were obtained by using
10% and 14% of ethanol respectively.
 46

20

15 Pe:Rp:Le(1:1:2)
Water %

Pe:Rp:Le(1:1:1)
10 Pe:Rp:Le(1:2:1)

0
0 4 6 8,06 10 12 14 16
Ethanol %

Figure 5.4 Optimum amount of water to prepare microemulsion by using peppermint oil
(Pe), rapeseed oil (Rp) and lecithin (Le) without or with ethanol.

5.1.5 Monoolein microemulsions

The data given in figure 5.5 indicated that the transparent microemulsion can be produced
by using water, monoolein and ethanol. It can be also produced when mixing the monoolein
with soybean oil or Tween20 at ratio 1:1 and ethanol. It should be noted that in case of
mixing monoolein with Tween20, the microemulsion can prepared without ethanol but in
other cases the addition of ethanol was necessary. The optimum amount of water required to
prepare microemulsion was markedly higher in case of using Tween20 than that in case of
using monoolein alone or mixed with soybean oil, however in case of using soybean oil it
was the lowest one.

30

25

20
Mo and Tw
Water %

15 Mo and Eth
Mo and So
10

0
0 4 6 8 10 12 14 16
Ethanol %
Figure 5.5 Optimum amount of water to prepare microemulsion by using ethanol (Eth) and
monoolein (Mo) alone, mixed with soybean oil (So) or Tween20 (Tw) at ratio 1:1.
 47

From this figure shows also that the maximum amount of water increased with increasing
the ethanol in all systems but this increment was remarkable in case of using Tween20,
likewise at the first the increment was higher than that after addition 12% of ethanol.

5.1.6 Tween20 microemulsios

The microemulsions could be prepared by using peppermint oil, Tween 20 as surfactant and
an aqueous solution without cosurfactant as shown in figure 5.6. The aqueous solutions
which used were water, sucrose 20%, NaCl 20% and citric acid 20 %. It should be noted
also that the microemulsion prepared by using various ratios of peppermint oil and Tween
20. The aqueous solution content increased strongly with increasing of Tween20 and the
highest content of aqueous solution was obtained in case of sucrose solution, while the
lowest content was obtained when using of NaCl 20% as aqueous solution and in case of
water or citric acid the aqueous solution was approximatly similar. The differences of
aqueous solution content between all cases of using aqueous solution were small.

80
70
aquious solution %

60
with water
50 with NaCl
40 with Suc
with C.a
30
20
10
0
01:00 01:01 01:01 01:02 01:03 01:04 01:04 01:05
Pe:Tw 20

Figure 5.6 Optimum amount of aqueous solution to prepare microemulsion by using

peppermint oil (Pe) and Tween20 (Tw) at various ratios. NaCl=sodium chlorid, Suc=sucrose
and C.a=citric solution 20%.
 48

5.2 Solubility of Curcumin

5.2.1 Curcumin solubility in solvents

The solubility of curcumin in each solvent used to prepare the microemulsion systems
(water, peppermint oil, soybean oil, peanut oil, rapeseed oil, ethanol, monoolein and
Tween20) was presented in table 5.1. From data in this table it can be showed that curcumin
could not be dissolved in the water, and aqueous solutions of 20% NaCl, sucrose and citric
acid. In case of soybean, peanut or rapeseed oil it dissolved slightly where the solubility was
0.02% in these oils. The solubility of curcumin was more in case of peppermint oil, ethanol
and monoolein than that in previous components but the maximized solubilization was
achieved in case of Tween 20, where it reached 6 %.

Table 5.1 Solubility of Curcumin in solvents used to produce the microemulsions

Component water So Pn Rp NaCl Suc C.a Eth Pe Mo Tw20

curcumin % 0.00 0.02 0.02 0.02 0.00 0.00 0.00 0.45 0.55 0.25 6.00

So=soybean oil, Pn=peanut oil, Rp=raps oil, Suc=sucrose solution 20%, C.a=citric acid solution 20%, Eth=
ethanol, Pe=peppermint oil, Mo=monoolein and Tw20=Tween 20.
 49

5.2.4 Solubility of curcumin in microemulsions

5.2.2.1 Curcumin solubility in lecithin/peppermint oil microemulsions

From the results in figures 5.7 and 5.8 shows that the solubility of curcumin in
lecithin/peppermint oil/water (without or with ethanol) microemulsions ranged between 1.2
and 0.6. The solubility in case of ratio 2:1of peppermint oil and lecithin was somewhat
higher than those in other ratios and when using ethanol the solubility was slightly higher
than that without ethanol. From the same figures can be noted that the solubility of curcumin
decreased gradually with increasing the water content.

1,2
1,1
Curcumin %

1 Pe:Le(1:1)
0,9 Pe:Le(2:1)
0,8 Pe:Le(3:1)

0,7
0,6
0,5
0 2 4 6 8
Water %
Figure 5.7: Effect of water content on the solubility of curcumin in peppermint oil (Pe)/lecithin (Le)
microemulsions without ethanol .

1,3
1,2
1,1
Curcumin %

Pe:Le(1:1)
1
Pe:Le(2:1)
0,9
Pe:Le(3:1)
0,8
0,7
0,6
0,5
0 2 4 6 8
Water %
Figure 5.8: Effect of water content on the solubility of curcumin in peppermint oil (Pe)/lecithin (Le)
microemulsions with 5% ethanol.
 50

5.2.2.2 Solubility in lecithin/peppermint oil/soybean oil microemulsions

The results in figures 5.9 and 5.10 indicated that the solubility of curcumin in
microemulsions containing water, peppermint oil, soybean oil and lecithin without or with
ethanol ranged between 0.6 and 0.3%. In case of microemulsions prepared without ethanol
the solubility firstly increased and no more increase with increasing the water content until
2% after that decreased in case of ratios 1:1:2 and 1:1:1 of peppermint, soybean and lecithin
but in case of ratio 1:2:1 the solubility remained the same with increasing the water content.
On the other hand, when using of ethanol the solubility of curcumin increased with addition
of 2% water and no change with increasing the water content in case of ratio 1:1:2 of
peppermint oil, soybean oil and lecithin, while in case of ratios 1:1:1 and 1:2:1 the solubility
was stable with addition of 2% water and then decreased with increasing the water content.
From figure 5.9 can be observed that the solubility of curcumin was higher in case of ratio
1:1:2 than those in case of ratios 1:1:1 and 1:2:1.

0,65
0,6
0,55
Curcumin %

0,5 Pe:So:Le(1:1:2)
0,45 Pe:So:Le(1:1:1)
0,4 Pe:So:Le(1:2:1)
0,35
0,3
0,25
0,2
0 1 2 3
Water %
Figure 5.9 Effect of water content on the solubility of curcumin in peppermint oil (Pe)/soybean oil (So)/lecithin
(Le) microemulsions without ethanol.

0,65
0,6
0,55
Curcumin %

0,5 Pe:So:Le(1:1:2)
0,45 Pe:So:Le(1:1:1)
0,4 Pe:So:Le(1:2:1)
0,35
0,3
0,25
0,2
0 2 4 6
Water %
Figure 5.10 Effect of water content on the solubility of curcumin in peppermint oil (Pe)/soybean oil
(So)/lecithin microemulsions with 5% ethanol.

5.2.2.3 Solubility in lecithin/peppermint oil/peanut oil microemulsions

 51

The results illustrated in figures 5.11 and 5.12 showed that the solubility of curcumin in
microemulsions containing water, peppermint oil, peanut oil and lecithin without or with
ethanol ranged between 0.6 and 0.4%. In case of microemulsions prepared without ethanol
the solubility increased with addition of 1% water and no more increase with increasing the
water content to 2% after that decreased especially in case of ratios 1:1:1 and 1:2:1 of
peppermint oil, peanut oil and lecithin, while in case of ratio 1:1:2 the solubility decreased
with addition 1% of water and stabilized with increasing the water content.

0,65
0,6
Curcumin %

0,55 Pe:Pn:Le(1:1:2)
0,5 Pe:Pn:Le(1:1:1)
0,45 Pe:Pn:Le(1:2:1)

0,4
0,35
0,3
0 1 2 3
Water %
Figure 5.11 Effect of water content on the solubility of curcumin in peppermint oil (Pe)/peanut oil (Pn)/lecithin
(Le) microemulsions without ethanol.

0,65
0,6
Curcumin %

0,55 Pe:Pn:Le(1:1:2)
0,5 Pe:Pn:Le(1:1:1)
0,45 Pe:Pn:Le(1:2:1)
0,4
0,35
0,3
0 2 4 6
Water %
Figure 5.12 Effect of water content on the solubility of curcumin in peppermint oil (Pe)/peanut oil (Pn)/lecithin
(Le) with 5% ethanol.

On the other hand, when using ethanol, the solubility of curcumin in case of ratio 1:1:2
increased when addition of 2% water and then decreased with increasing the water. In case
of ratios 1:1:1 and 1:2:1 as shows in figure 5.12.
 52

5.2.2.4 Solubility in lecithin/peppermint oil/rapeseed oil microemulsions

As shows in figures 5.13 and 5.14 the solubility of curcumin in these systems ranged
between 0.6 and 0.3. From figure 5.13 apparent that, in case of no using ethanol the
solubility in case of ratioes 1:1:1 and 1:2:1 of peppermint oil, rapeseed oil and lecithin was
stable after addition 1% water then decreased gradually with increasing the water.
Reversible results were found in case of ratio 1:1:2 where the solubility somewhat increased
with increasing the water. On the other hand, when using the ethanol and ratio 1:1:1 and
1:2:1(peppepmint oil/rapeseed oil/lecithin) the solubility decreased after addition the water
until reached the water content to 4% after that it stabilized with increasing the water to 6%
as to be seen in figure 5.14. In case of ratio 1:1:2 the solubility decreased after addition of
2% water and the reduction continued with increasing the water content as shows in the
same figure .

0,55
0,5
0,45
Curcumin %

Pe:Rp:Le(1:1:2)
0,4 Pe:Rp:Le(1:1:1)
0,35 Pe:Rp:Le(1:2:1)
0,3
0,25
0,2
0 1 2 3
Water %
Figure 5.13 Effect of water content on the solubility of curcumin in peppermint oil (Pe)/rapeseed oil
(Rp)/lecithin (Le) microemulsions without ethanol.

0,65
0,6
Curcumin %

0,55 Pe:Rp:Le(1:1:2)
0,5 Pe:Rp:Le(1:1:1)

0,45 Pe:Rp:Le(1:2:1)

0,4
0,35
0,3
0 2 4 6
Water %
Figure 5.14: Effect of water content on the solubility of curcumin in peppermint oil (Pe)/rapeseed oil
(Rp)/lecithin with 5% ethanol.
 53

5.2.2.5 Solubility in monoolein microemulsions

Figure 5.15 showed that the solubility of curcumin in microemulsions containing monoolein
and 10% ethanol or monoolein mixed with soybean oil at ratio 1:1 and 10% ethanol was
relatively small where it ranged between 0.35 and 0.2, and in case of mixing soybean oil
with the monoolein the solubility was lower than that in case of monoolein without soybean
oil.

3
2,8
2,6
2,4
2,2
2
Mo and Eth
Curcumin %

1,8
1,6 Mo,So and
Eth
1,4 Mo and Tw
1,2
1
0,8
0,6
0,4
0,2
0
0 1 2 3 4
Water %
Figure 5.15 Effect of water content on the solubility of curcumin in monoolein microemulsions: monoolein
(Mo)/ethanol (Eth), monoolein (Mo)/soybean oil (So)/ethanol (Eth)and monoolein (Mo)/Tween20 (Tw).

On the other hand, in case of system prepared by using monoolein and Tween20 at ratio 1:1
was consideribly great in comparing to that in case of monoolein either alone or mixed with
soybean oil where ranged between 2.8% and 2.3%. The same figure shoued that the
solubility decreased gradually with increasing the water content in all systems.

5.2.4.6 Solubility in peppermint oil/Tween20 (1:1) microemulsions

The results presented in figure 5.16 indicated that the solubility of curcumin in
microemulsion containing peppermint oil, Tween 20 at ratio 1:1 and water, sodium chlorid
20%, sucrose 20% or citric acid solution 20% was high in comparing with previous systems
of microemulsions, where it ranged between 3.5 and 2.5%.
 54

The same figure indicated that the solubility was highest in case of citric acid solution while
it was lowest in case of water. Apparent also from this figure that the solubility decreased
slightly after addition of 2% aqueous solution and with increasing the aqueous solution it
decreased gradually.

3,6
3,4
3,2 Water
Curcumin %

3 NaCl
2,8 Suc
C.a
2,6
2,4
2,2
2
0 2 10 12 14
Aqueous solution %
Figure 5.16: Effect of aqueous solution content on the solubility of curcumin in peppermint oil/Tween20 (1:1)
microemulsions. NaCl=sodium chlorid, Suc=sucrose and C.a=citric acid solution 20%.

5.2.2.7 Solubility in peppermint oil/Tween20 (1:4) microemulsions

Microemulsions in this system were prepared by using peppermint oil/Tween20 at ratio 1:4
and aqueous phase (water or 20% solution of NaCl, sucrose or citric acid). The results given
in figure 5.17 showed that the solubility of curcumin in this system was markedly high in
comparing with that in other prepared microemulsions, where the solubility in this case was
about 6% before addition the aqueous solution and after addition of 5% aqueous solution
decreased to about 4%. It could be noted also that the solubility of curcumin decreased
gradually with increasing the aqueous solution. The differences between the solublity of
curcumin in all cases of aqueous solution were small.
 55

5
Water
Curcumin %

4 NaCl 20%
Suc 20%
3 C.a 20%

0
0 5 25 45
Aqueous solution %
Figure 5.17 Effect of aqueous solution content on the solubility of curcumin in peppermint oil/Tween20 (at
ratio 1:4) microemulsions. NaCl=sodium chlorid, Suc=sucrose and C.a=citric acid solution 20%.

5.3 Determination of pH value

The pH-value in microemulsions was measured and the data presented in table 5.2. From
this table it should be noted that the pH-value in MEs which prepared by using peppermint
oil (or mixed of peppermint oil with an edible oil), lecithin and water ranged between 5.00
and 5.30 while in case of microemulsions containing monoolein (or mixed of monoolein
with soybean oil), ethanol and water it was 6.56 (or 6.85). In case of microemulsions
containing peppermint oil, Tween20 at ratio 1:1 and water or sucrose solution the pH-value
was somewhat high where it was 7.8 or 7.4 respectively but in case of microemulsions
containing peppermint oil, Tween20 at ratio 1:4 and water or sucrose solution was 6.80.
From the same table apparent that the pH-value in case of using NaCl solution as aqueous
phase, peppermint oil and Tween20 either at ratio 1:1 or 1:4 was 6.50 or 6.20 while in case
of using the citric acid solution as aqueous phase to prepare microemulsions with
peppermint oil and Tween20 either at ratio 1:1 or 1:4 the pH-value was small where it was
3.70 or 2.10.
 56

Table 5.2 the pH-values of microemulsions

System of microemulsions pH-value System of microemulsions pH-value
Pe,Le 1:1and water 5.30 Pe,Rp,Le 1:2:1 and water 5.00
Pe,Le 2:1 and water 5.26 Mo,Eth and water 6.56
Pe,Le 3:1 and water 5.20 Mo,So 1:1,Eth and water 6.85
Pe,So,Le 1:1:2 and water 5.30 Pe,Tw20 1:1 and water 7.80
Pe,So,Le 1:1:1 and water 5.27 Pe,Tw20 1:1 and NaCl 6.50
Pe,So,Le 1:2:1 and water 5.17 Pe,Tw20 1:1 and Suc 7.40
Pe,Pn,Le 1:1:2 and water 5.30 Pe,Tw20 1:1 and C.a 3.70
Pe,Pn,Le 1:1:1 and water 5.25 Pe,Tw20 1:4 and water 6.80
Pe,Pn,Le 1:2:1 and water 5.00 Pe,Tw20 1:4 and NaCl 6.20
Pe,Rp,Le 1:1:2 and water 5.30 Pe,Tw20 1:4 and Suc 6.80
Pe,Rp,Le 1:1:1 and water 5.20 Pe,Tw20 1:4 and C.a 2.10
Pe=peppermint oil, Le=lecithin, So=soybean oil, Pn=peanut oil, Mo=monoolein,
Tw20=Tween20, Eth=ethanol, NaCl=sodium chlorid, Suc=sucrose andC.a=citric acid
 57

5.4 Light stability of curcumin in microemulsions

Stability of curcumin solubilized in microemulsions against light was determined according
to the change in color shade (h°) during storage the samples under UV light for 30 days, in
the darkness or under normal electric light for 60 days. The color shade was measured by
using Measuring Color Instrument.

5.4.1 Light stability in lecithin/peppermint oil microemulsions

Lecithin microemulsions containing peppermint oil as lipophilic phase were prepared at
three ratios 1:1, 2:1 or 3:1 0f peppermint oil and lecithin, and 3% water without or with 5%
ethanol. An aqueous solution 80% of ethanol was used as control. Curcumin was solubilized
in microemulsion samples and in control at concentration 0.4%.The samples and control
containing curcumin were stored under UV light, in darkness or under normal electric light
and the stability of curcumin was determined as the change in its color shade during the
storage.
5.4.1.1 Effect of UV light
Figures 5.18 and 5.19 shows that the changes in color shade of samples either without or
with ethanol were very slight but in control the changes were relatively great where the color
shade decreased gradually from 90 to 45. This Indicated that the color of curcumin
solubilized in the samples was stable (yellow) while it in control changed from yellow to
red. Apparent from the same figures that the changes in color of curcumin solubilized in
microemulsions containing peppermint oil and lecithin at all ratios were similar.

100

90
Color shade

80 Pe:Le (1:1)
Pe:Le(2:1)
70
Pe:Le(3:1)
60 Control
50

40
0 2 24 48 96 144 240 288 360 480 600 720
Time (Hours)

Figure 5.18 Changes in color shade of lecithin (Le)/peppermint oil (Pe) microemulsions
containing curcumin during storage under UV light.
 58

100
90
Color shade

80 Pe:Le(1:1)
70 Pe:Le(2:1)

60 Pe:Le(3:1)
Control
50
40
0 2 24 48 96 144 240 288 360 480 600 720
Time(Hours)

Figure 5.19 Changes in color shade of lecithin (Le)/peppermint oil (Pe)/5% ethanol
microemulsions containing curcumin during storage under UV light.

5.4.1.2 Effect of darkness

As showed in figures 5.20 and 5.21 the changes in hue of samples either without or with
ethanol were unremarkable, but in case of control, these changes were small where the color
shade decraesed from 90 to 82. These results shows that the color of curcumin solubilized in
control or in microemulsions stored in darkness was somewhat stable. It should be also
noted that the changes in color shade were similar in all ratios of peppermint oil and lecithin
which used to prepare the microemulsion.

105
100
95 Pe:Le(1:1)
Color shade

90 Pe:Le(2:1)
85 Pe:Le(3:1)
80 Control
75
70
65
60
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.20) Changes in color shade of lecithin (Le)/peppermint oil (Pe) microemulsions
containing curcumin during storage in darkness.
 59

110
105
Pe:Le(1:1)
100
Color shade

Pe:Le(2:1)
95
Pe:Le(3:1)
90
Control
85
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.21) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/5% ethanol
microemulsions containing curcumin during storage in darkness.

5.4.1.3 Effect of normal electric light

The changes in color shade of samples in this case were very slight where it changed from
100 to 85 as in figures 5.22 and 5.23. These figures showed also that the changes in color
shade in case of control were higher than those in case of the samples. These results
indicated that the color of curcumin solubilized in microemulsions stored under normal
electric light changed from slight yellow to yellow and in case of control from yellow to
orange.

110
105
100 Pe:Le(1:1)
Color shade

95 Pe:Le(2:1)
90 Pe:Le(3:1)

85 Control

80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time(days)

Figure (5.22) Changes in color shade of lecithin (Le)/peppermint oil (Pe) microemulsions
containing curcumin during storage under normal electric light.
 60

110
105
100 Pe:Le(1:1)
Color shade

95 Pe:Le(2:1)
90 Pe:Le(3:1)
85 Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.23) Changes in color shade lecithin (Le)/peppermint oil (Pe)/5% ethanol
microemulsions and containing curcumin during storage under normal electric light.

5.4.2 Stability in lecithin/peppermint oil/soybean oil microemulsions

5.4.2.1 Effect of UV light

Alteration in color shade of samples and control stored under UV light were given in figures
5.24 and 5.25. The color shade of control changed greatly while it decreased from 90 at the
begining of storage period, to 46 after 30 days, this means the color changed from yellow to
red. The change in case of microemulsions was very slight in all ratios of peppermint oil,
soybean oil and lecithin..

110
100
Pe:So:Le(1:1:2)
90
Color shade

Pe:So:Le(1:1:1)
80
Pe:So:Le(1:2:1)
70 Control
60
50
40
0 2 24 48 96 144 240 288 360 480 600 720
Time(hours)

Figure (5.24) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/soybean oil (So)
microemulsions containing curcumin during storage under UV light
 61

110
100 Pe:So:Le(1:1:2)
Color shade

90
Pe:So:Le(1:1:1)
80
70 Pe:So:Le(1:2:1)
60 Control
50
40
0 2 24 48 96 144 240 288 360 480 600 720
Time (hours)

Figure (5.25) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/soybean oil
(So)/5% ethanol microemulsions containing curcumin during storage under UV-light.

5.4.2.2 Effect of darkness

It should be noted from figures 5.26 and 5.27 that the color shade of samples and control
stored in darkness for 60 days were somewhat stable. However, the color was remained
slight yellow in case of samples at all ratios of peppermint oil, soybean oil and lecithin either
without or with ethanol and yellow in case of control.

105

100
Pe:So:Le(1:1:2)
Color shade

95
Pe:So:Le(1:1:1)
90 Pe:So:Le(1:2:1)
85 Control

80

75
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.26) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/soybean oil(So)
microemulsions containing curcumin during storage in darkness
 62

110
105
100 Pe:So:Le(1:1:2)
Hue (h°)

95 Pe:So:Le(1:1:1)
90 Pe:So:Le(1:2:1)
85 Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.27) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/soybean oil (So)
microemulsions containing curcumin with5% ethanol during storage in darkness.

5.4.2.3 Effect of normal electric light

The results in figures 5.28 and 5.29 showed that the changes in color shade of samples
without ethanol were lower than those with ethanol where the differences in the color shade
between the zero time of storage and after 60 days were about 8 and 12 respectively. On the
other hand, the changes in color shade of control were higher than those in samples where
the color shade of control decreased from 90 at the zero time to 72 after 60 days of storage.
These results indicated that the changes in the color of samples stored under normal electric
light were slight where it changed from slight yellow to yellow but in case of control it
changed from yellow to orange. In addition, the differences between the various ratios of
peppermint oil, soybean oil and lecithin which used to prepare microemulsion were slight.

110
105
100 Pe:So:Le(1:1:2)
Color shade

95 Pe:So:Le(1:1:1)
90
Pe:So:Le(1:2:1)
85
Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.28) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/soybean oil (So)
microemulsions containing curcumin during storage under normal electric light.
 63

110
105
100 Pe:So:Le(1:1:2)
Color shade

95 Pe:So:Le(1:1:1)
90 Pe:So:Le(1:2:1)
85
Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.29) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/ soybean oil
(So)/5% ethanol microemulsions containing curcumin during storage under normal electric
light.

5.4.3 Stability in lecithin/peppermint oil/peanut oil microemulsions

5.4.3.1 Effect of UV light

The presented results in figures 5.30 and 5.31 showed that the changes in color shade of
samples at ratio 1:2:1 of peppermint oil, peanut oil and lecithin were unremarkable either
without or with ethanol, while these changes in case of ratios 1:1:2 and 1:1:1 were slight
where decreased from 95 to 85. On the other hand, reversible results were found in case of
control, where the color shade decreased from 90 to 46 after 30 days. Generally, it should be
concluded that the color of samples was somewhat stable where it remained yellow in all
cases, but the color of control changed from yellow at the begining of storage period to red
after 30 days.

110
100 Pe:Pn:Le(1:1:2)
Color shade

90 Pe:Pn:Le(1:1:1)
80
Pe:Pn:Le(1:2:1)
70
Control
60
50
40
0 2 24 48 96 144 240 288 360 480 600 720
Time (hours)

Figure (5.30) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/peanut oil (Pn)
microemulsions containing curcumin during storage under UV light.
 64

110

100

90 Pe:Pn:Le(1:1:2)
Color shade

80 Pe:Pn:Le(1:1:1)
Pe:Pn:Le(1:2:1)
70
Control
60

50

40
0 2 24 48 96 144 240 288 360 480 600 720
Time (hours)

Figure (5.31) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/peanut oil (Pn)/5%
ethanol microemulsions containing curcumin during storage under UV light.

5.4.3.2 Effect of the darkness

The results in figures 5.32 and 5.33 indicated that the color shade of samples stored in the
darkness was stable and the color remained yellow after 60 days of storage at all ratios of
peppermint oil, peanut oil and lecithin either without or with ethanol. It can be observed
from the same figures that the changes in color shade of control were small where it
decreased from 90 to 82 after 60 days of storage, in addition the color in this case remained
yellow.

110
105
Pe:Pn:Le(1:1:2
100
)
Color shade

95 Pe:Pn:Le(1:1:1
90 )
Pe:Pn:Le(1:2:1
85 )
Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.32) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/peanut oil (Pn)
microemulsions containing curcumin during storage in darkness.
 65

110
105
100 Pe:Pn:Le(1:1:2)
Color shade

95 Pe:Pn:Le(1:1:1)
90 Pe:Pn:Le(1:2:1)
85
Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.33) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/peanut oil (Pn)/5%
ethanol microemulsions containing curcumin during storage in darkness.

5.4.3.3 Effect of normal electric light

In this case the color shade of samples either without or with ethanol at all ratios of
peppermint oil, peanut oil and lecithin was somewhat stable and it of control decreased from
90 at zero time to 72 after 60 days of storage as showed in figures 5.34 and 5.35. From these
results it can be said that the color of curcumin solubilized in microemulsion prepared by
using peppermint oil, peanut oil, lecithin and water either without or with ethanol during the
storage under normal electric light for 60 days was stable where it remained yellow while
the color of control which containing curcumin solubilized in ethanol 80% and stored at the
same condations changed from yellow to orange.

110
105
100 Pe:Pn:Le(1:1:2)
Color shade

95 Pe:Pn:Le(1:1:1)

90 Pe:Pn:Le(1:2:1)

85 Control

80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time(days)

Figure (5.34) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/peanut oil (Pn)
microemulsions containing curcumin during storage under normal electric light.
 66

110
105
100 Pe:Pn:Le(1:1:2)
color shade

95 Pe:Pn:Le(1:1:1)
90 Pe:Pn:Le(1:2:1)
85
Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.35) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/peanut oil (Pn)/5%
ethanol microemulsions containing curcumin during storage under normal electric light.

5.4.4 S tability in lecithin/peppermint/rapeseed oil microemulsions

In this case microemulsions were prepared by using pepprmint oil mixed with rapeseed oil
as lipophilic phase and lecithin at ratio 1:1:2, 1:1:1 or 1:2:1. The water was added at 3%
without or with 5% ethanol. Curcumin was solubilized at 0.4% in the microemulsion
samples and in control which prepared by using 80% ethanol in water.

5.4.4.1 Effect of UV light

The changes in color shade of curcumin solubilized in microemulsions containing
peppermint oil, rapeseed oil and lecithin at ratios 1:1:2, 1:1:1, 1:2:1 and water without or
with ethanol as well as the changes in color shade of control during storage under UV light
for 30 days presented in figures 5.36 and 5.37. From these figures, it should be noted that
the changes in color shade of control were higher than those of samples. The color in case
of control changed gradually from yellow to red while in case of samples it was
approximatly stable in case of ratio 1:2:1 or changed slightly from yellow to orange in other
ratios. Apparent from the same figures that the changes in color shade were somewhat
higher in samples which with ethanol than those without it.
 67

100

90 Pe:Rp:Le(1:1:2)
Color shade

80 Pe:Rp:Le(1:1:1)
Pe:Rp:Le(1:2:1)
70 Control
60

50

40
0 2 24 48 96 144 240 288 360 480 600 720
Time(hours)

Figure (5.36) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/rapeseed oil (Rp)
microemulsions containing curcumin during storage under UV light.

110

100

90 Pe:Rp:Le(1:1:2)
Color shade

Pe:Rp:Le(1:1:1)
80
Pe:Rp:Le(1:2:1)
70 Control
60

50

40
0 2 24 48 96 144 240 288 360 480 600 720
Time (hours)

Figure (5.37) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/rapeseed oil
(Rp)/5% ethanol microemulsions containing curcumin during storage under UV light.

5.4.4.2 Effect of the darkness

The color shade was stable in case of microemulsion samples and it decreased slightly in
case of control where the color remained in range slight yellow (color shade 98) and yellow
(color shade 90) of samples and control respectively after 60 days of storage in the darkness.
The changes in color shade were similar at all ratios of peppermint oil, rapeseed oil and
lecithin either without or with ethanol. These results were illustrated in figures 5.38 and
5.39.
 68

105

100
Pe:Rp:Le(1:1:2)
Color shade

95
Pe:Rp:Le(1:1:1)
90 Pe:Rp:Le(1:2:1)
85 Control

80

75
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.38) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/rapeseed oil (Rp)
microemulsions containing curcumin during storage in darkness.

100

95 Pe:Rp:Le (1:1:2)
Color shade

Pe:Rp:Le (1:1:1)
90 Pe:Rp:Le (1:2:1)
Control
85

80
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64
Time (days)

Figure (5.39) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/rapeseed oil
(Rp)/5% ethanol microemulsions containing curcumin during storage in darkness.

5.4.4.3 Effect of normal electric light

It should be noted in figures 5.40 and 5.41 that the color shade of microemulsion samples
was stable at all ratios either without or with ethanol after 60 days of storage under normal
electric light while in case of control it decreased from 90 at the begining of storage to 70
after 60 days. It is clear from these results that the color of microemulsion samples was
stable comparing with that in case of control where it changed from yellow to orange.
 69

110
105
100 Pe:Rp:Le(1:1:2)
Color shade

95 Pe:Rp:Le(1:1:1)
90 Pe:Rp:Le(1:2:1)
85 Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.40) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/rapeseed oil (Rp)
microemulsions containing curcumin during storage under normal electric light.

110
105
100 Pe:Rp:Le(1:1:2)
Color shade

95 Pe:Rp:Le(1:1:1)
90 Pe:Rp:Le(1:2:1)
85 Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.41) Changes in color shade of lecithin (Le)/peppermint oil (Pe)/rapeseed oil
(Rp)/5% ethanol microemulsions containing curcumin during storage under normal electric
light.
 70

5.4.5 Stability in monoolein microemulsions

The monoolein microemulsions were prepared by using 3% water, 10% ethanol and
monoolein either alone, mixed with soybean oil or with Tween20 at ratio 1:1. The control
used in this case was 80% of ethanol in the water. Curcumin was solubilized in these
microemulsions and in control at concentration 0.2%. The stability of curcumin was
determined as the change of color shade occurred during the storage under UV light, in
darkness or under normal electric light. The results were illustrated in figures 5.42, 5.43 and
5.44. It can be observed that the color shade of all microemulsion samples during storage
either under UV light, in darkness or under electric light was stable and the color was lemon
yellow at all time of storage. In case of control the color shade changed during storage under
UV and normal electric light but it was approximatly stable during storage in the darkness.
The change of control in case of storage under UV light was higher than that under normal
electric light where the color changed from yellow to red and from yellow to orange
respectively.

110
100
Mo,Eth and Wa
Color shade

90
Mo,So,Eth and Wa
80
70 Mo,Tw20,Eth and Wa
60 Control
50
40
0 2 24 48 96 144 240 288 360 480 600 720
Time (Hours)

Figure (5.42) Changes in color shade of monoolein microemulsions containing curcumin

during storage under UV light. Mo= monoolein, Eth=ethanol, Wa= water, Tw20=Tween20
and So=soybean oil.
 71

110
105
100 Mo,Eth and Wa
Color shade

95 Mo,So,Eth and Wa
90 Mo,Tw20,Eth and Wa
85
Control
80
75
70
0 1 2 4 68 10 15 20 25 30 40 50 60
Time (days)

Figure (5.43) Changes in color shade of monoolein microemulsions containing curcumin

during storage in darkness. Mo=monoolein, Eth=ethanol, Wa=water, Tw20=Tween20 and
So=soybean oil.

110
105
Mo,Eth and Wa
100
Color shade

95 Mo,So,Eth and Wa
90 Mo,Tw20,Eth and Wa
85
Control
80
75
70
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.44) Changes in color shade of monoolein microemulsions containing curcumin

during storage under normal electric light. Mo=monoolein, Eth=ethanol, Wa=water and
So=soybean oil.

5.4.6. Stability in peppermint oil/Tween20 (1:1) microemulsions

The microemulsions in this case were prepared by using peppermint oil, Tween20 at ratio
1:1 and water or an aqueous solution of 20% NaCl, sucrose or citric acid. The curcumin was
dissolved in these microemulsion at 0.4%(w/w). The samples were stored under UV light for
30 days, in darkness or under normal elecric light for 60 days. The control (with Tween20
microemulsions) was solution of Tween20 and water at ratio 1:1.

5.4.6.1 Effect of UV light

It could be noted in figure 5.45 that the color shade of microemulsion samples containing
water, sucrose or citric acid as aqueous phase was somewhat stable and the color was yellow
to end of storage period but in case of microemulsion sample prepared by NaCl solution as
aqueous phase the change in color shade was great where it decreased gradually from 85 at
first to 65 at the end of storage period (720 hours), the color of this sample ranged between
 72

yellow at zero time and orange brown at the end of storage. As for control the color shade
decreased gradually from 77 at the begining of storage to 52 at the end, thus the color
changed from orange at first to red at the end of storage.

5.4.6.2 Effect of darkness

In this case the color shade of microemulsion samples containing water, sucrose or citric
acid was stable and the color was lemon yellow in case of citric acid and yellow in other
three cases as shows in figure 5.46. Likewise, in case of control the color shade was
somewhat stable. On the other hand, the color shade of sample containing NaCl solution
changed strongly where decreased from 95 at first to 45 after 60 days of storage, the color in
this case changed from yellow to red at the end of storage.

110
Pe:Tw and Wa
100
Pe, Tw and NaCl
color shade

90
Pe, Tw and Suc
80 Pe, Tw and C.a
70 Control

60
50
0 2 24 48 96 144 240 288 360 480 600 720
Time (hours)

Figure (5.45) Changes in color shade of peppermint oil (Pe)/Tween20 (Tw)

microemulsions containing curcumin during storage under UV light. Pe : Tw 1:1,
Wa=water, NaCl= sodium chlorid, Suc=sucrose, C.a=citric acid

110
100 Pe,Tw and Wa
90 Pe,Tw and NaCl
Color shade

80 Pe,Tw and Suc

70 Pe,Tw and C.a
60 Control
50
40
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.46) Changes in color shade of peppermint oil (Pe)/Tween20 (Tw) microemulsions
containing curcumin during storage in darkness. Pe : Tw =1:1, Wa=water, NaCl=sodium
chlorid, Suc= sucrose and C.a=citric acid
 73

5.4.6.3 Effect of normal electric light

The results illustrated in figures 5.47 showed that the color shade of microemulsion samples
containing water, sucrose or citric acid as aqueous phase was approximatly stable, while in
case of microemulsion prepared by using NaCl solution the color shade decreased strongly
from 95 at the first to reached 45 after 60 days of storage. In case of control the color shade
decreased gradually from 90 at the first to 60 at the end of storage.

110
100 Pe,Tw and Wa
Pe,Tw and NaCl
Color shade

90
80 Pe,Tw and Suc
70 Pe,Tw and C.a
60 Control
50
40
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.47) Changes in color shade of peppermint oil (Pe)/Tween20 (Tw) microemulsions
containing curcumin during storage under normal electric light. Pe : Tw =1:1, Wa=water,
NaCl=sodium chlorid, Suc= sucrose and C.a=citric acid

5.4.7 Stability in peppermint oil/Tween20 (1:4) microemulsions

5.4.7.1 Effect of UV light

It could be observed from figure 5.48 that the color shade of microemulsion samples was
somewhat stable, where the color remained yellow in all cases during storage, but in case of
control the color shade decreased gradually from 85 at zero time to 50 after 30 days of
storage, thus the color in this case changed from yellow at the begining to red at the end of
storage period.
 74

100
90 Pe,Tw and Wa
Color shade

Pe,Tw and NaCl

80
Pe,Tw and Suc
70
Pe,Tw and C.a
60
Control
50

40
0 2 24 48 96 144 240 288 360 480 600 720
Time (hours)

Figure (5.48) Changes in color shade of peppermint oil (Pe)/Tween20 (Tw) 1:4
microemulsions containing curcumin during storage under UV light. Wa=water,
NaCl=sodium chlorid, Suc=sucrose and C.a=citric acid

5.4.7.2 Effect of darkness

In this case the color shade of microemulsion samples containing water, sucrose or citric
acid was stable, where the color remained yellow to end of storage period and similar result
was obtained in case of control as shows in figure 5.49. On the other hand, the color shade
of microemulsion sample containing NaCl as aqueous phase decreased gradually from 90 at
the begining to 65 after 60 days of storage, thus the color in this case changed from yellow
to orange.

110
Pe,Tw and Wa
100
Pe,Tw and NaCl
Color shade

90 Pe,Tw and Suc

80 Pe,Tw and C.a

Control
70

60
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.49) Changes in color shade of peppermint oil (Pe)/Tween20 (Tw) 1:4
microemulsions containing curcumin during storage in darkness. Wa=water, NaCl=sodium
chlorid, Suc=sucrose and C.a=citric acid

5.4.7.3 Effect of normal electric light

From the results presented in figure 5.50, it should be noted that the color shade of
microemulsion samples containing water, sucrose or citric acid as aqueous phase was stable
during storage under normal electric light for 60 days and the color remained yellow. On the
other hand the color shade in case of sample containing NaCl solution as aqueous phase and
 75

also in case of the control the color shade decreased gradually but the change was higher in
case of sample containing NaCl solution than that in case of control where the color changed
from yellow to red and from yellow to orange respectively.

110
100 Pe,Tw and Wa
Color shade

90 Pe,Tw and NaCl

Pe,Tw and Suc
80
Pe,Tw and C.a
70
Control
60
50
0 1 2 4 6 8 10 15 20 25 30 40 50 60
Time (days)

Figure (5.50) Changes in color shade of peppermint oil (Pe)/Tween20 (Tw) 1:4
microemulsions containing curcumin during storage under normal electric light. Wa=water,
NaCl=sodium chlorid, Suc=sucrose and C.a=citric acid
 76

6 Discussion
This work investigates the possibility of preparation microemulsions from suitable
components for food uses. The microemulsions possess special characteristics of high
solubility capacity, long term stability and ease of preparation. They also have the potential
ability to solubilize both lipophilic and hydrophilic substances. Curcumin is a natural
colorants and has problems for its uses direct in foods due to its insolubility in water and
poor solubility in vegetable oils as well as it is sensitive to light. This work also investigates
solubility and stability of curcumin when dissolved in these microemulsions. It is expected
that the microemulsions due to their previous characteristics will offer good results to
solubilize and stabilize curcumin against light during storage.

6.1 Preparation of microemulsions

Microemulsions containing vegetable oils could be used in food products, where there is a
demand for environmentally more safety formulations [Abillon et al., 1986]. When the
components used in the preparation of microemulsions (surfactant, oil and cosurfactant) are
safe for human consumption, they become important in such fields as foods, cosmetics and
pharmaceuticals [Kunieda and shinoda 1982]. Since the first microemulsion system was
described by Schulman and Hoar [1943] an extensive number of papers have been published
in this area, unfortunately, most of the systems described are not suitable for human
consumption either in foods or in pharmaceuticals [Von Corswant et al., 1997]. Aboofazeli
et al., [1995] stated that most of the microemulsion systems investigated are unsuitable for
human health, mainly because of the ingredients used. By far the majority of work to date
has involved the use of ionic surfactants, alcohol cosurfactants and oils such as hexane and
benzene. All of which are unsuitable for the purposes of pharmaceutical and food
formulations. Few studies using triglycerides as the lipophilic phase in microemulsions
have been published. Although oils from natural sources and their derivatives, triglycerides
are considerd to be harmless to the human and the environment [Busch 1992], such
microemulsion systems with natural oils are but infrequently described in the literature
[Alander and Warnheim 1989a]. There are very few available examples of acceptable
microemulsion systems for food applications. The microemulsion system without
cosurfactant, with available food surfactants and oils, and no off-taste or change in
 77

performance is suitable for this purpose [El-Nokaly et al., 1991]. The preparation of
microemulsions with mineral oils, synthetic surfactants and alkanols has become a well
established practice and these components are, in general, harmful. The problem arises how
to prepare microemulsions with nontoxic substances [Kahlweit et al., 1997]. So, the problem
is then to find suitable components with physiological compatibility in order to produce
microemulsions for possible applicactions in food products.
Various microemulsion systems were prepared by using a number of natural oils as
lipophilic phase such as peppermint oil alone or mixed with one of a common edible oil
such as soybean, peanut or rapeseed oil. A suitable food surfactant such as lecithin,
monoolein or Tween20 was used as emulsifier. The aqueous phase was either water or an
aqueous solution of 20% NaCl, sucrose or citric acid in the absence or presence of ethanol as
cosurfactant. The microemulsions were defined on the basis of the surfactant type: lecithin-
microemulsions, monoolein-microemulsions or Tween20-microemulsions.

6.1 1 Lecithin/peppermint oil microemulsions

Microemulsions could be produced by using peppermint oil alone as lipophilic phase,
lecithin as surfactant and water without ethanol as in figure 5.1. Microemulsions could also
be prepared by using these components with ethanol. It was observed that the use of high
quantity of lecithin is considerably difficult to solubilize in peppermint oil, due to the
formation of a high viscosity mixture. It has also observed that the solubility of a greater
lecithin mass in peppermint oil needs a longer time. This problem is due to the waxy
semisolid structure of soybean lecithin which used in this study. On the other hand, very
small water content could be solubilized in the mixture of peppermint oil and lecithin to
form microemulsion, when using a small amount of lecithin. This may be due to the
insufficient surfactant content, in particular, according to Bergenstahl and Fontell [1983]
soybean lecithin is both strongly hydrophobic due to long hydrocarbon chains, and strongly
lipophobic due to zwitterionic polar head groups.
From the same figure (5.1) it could be noted that microemulsions were produced by using
peppermint oil, lecithin and water without cosurfactant. This result is very important
because the formation of microemulsions by using lecithin as surfactant, natural oils
(suitable for food products) and water is one of the objectives of this work. This result may
be due to the fact that, microemulsions stabilized by a nonionic surfactant can be produced
without cosurfactant as reported by Aboofazeli et al., [1994]. Therefore the common
 78

cosurfactant (short chain alcohols) are not useful because of their toxicity [Friberg and
Rydhang, 1971]. Cosurfactants are not easy to find in foods and their addition is not a
theoretical requirement [El-Nokaly et al., 1991]. Moreover, most work to date studying
microemulsions has utilised oils, surfactants and cosurfactants unsuitable for pharmaciutical
purposes [Aboofazeli and Lawrence 1994]. On the other hand this result is different from
that is obtained by Shinoda et al., [1991] where they found that lecithin will not form
microemulsions without the aid of a short-chain alcohol as cosurfactant. Aboofazeli et al.,
[1994] stated that none of studies have investigated the potential of non-alcohol
cosurfactants to produce microemulsions with lecithin. Binks et al., [1989] said that if
lecithin containing systems are to form microemulsions over a reasonably wide range of oil
and water concentrations, the effective a fairly high critical packing parameter (CPP) of
lecithin needs to be reduced and a short chain cosurfactant can act to reduce the effective
CPP by its incorporation into the interfacial film, in addition a short chain cosurfactant can
also act to increase the fluidity of the interfacial surfactant layer, thereby reducing the
tendency of lecithin to form highly rigid film.
This difference between the findings obtained in this study and those in above mentioned
references may be due to the peppermint oil which used in this work has special structure
which different from the structure of oils used in those studies where it contains menthol,
menthyl acetate and menthone, the menthol may then act as cosurfactant.
The location and extension of a microemulsion region is generally dependent on the oil
structure, this is due to differences in oil penetration into the surfactant [Monduzzi et al.,
1997]. Alander and Warnheim, [1989b] found that the molecular weight of the oil affected
the stability of microemulsions.
El-Nokaly found that the nature and concentration of the surfactant are important to obtain a
maximum solubility of water in microemulsion [El-Nokaly et al., 1991]. The same result
was obtained, where the maximum amount of water solubilized in the system when the
prepration of microemulsions without ethanol, was higher in case of ratio 2:1 of peppermint
oil and lecithin than those in case of ratio 3:1. But the decrease of the water amount in case
of ratio 1:1 compared to in case of ratio 2:1 may be due to the increasing of menthol in the
latter ratio (owing to the increasing of peppermint oil) which may act as cosurfactant and
this increases the efficiency of lecithin to solubilize a more amount of water as shows in the
same figure.
 79

The maximum amount of water decreased in general, when using the ethanol with
peppermint oil and lecithin. The maximum amount of water increased only when using a
high amount of lecithin and a certain amount of ethanol (4%). This may be due to the
ethanol efficiency as cosurfactant depends on the ratio of lecithin to ethanol in the mixture.
Aboofazeli et al., [1995] reported that the major factor influencing on the lecithin based
microemulsions is the nature and mixing ratio of the cosurfactant used.

6.1.2 Lecithin/peppermint oil/edible oil microemulsions

The possibility to form microemulsions by using edible oils such as soybean, peanut or
rapeseed oil was investigated in the initial experiments. Unfortunately, it was seen that using
these oils to prepare microemulsion was vrey difficult because the solubility of lecithin in
these oils needs a high amount of ethanol and this is unsuitable for food uses. In addition,
water is insoluble in such mixture containing edible oils and lecithin either without or with
ethanol thereby, microemulsion could not be obtained. This result is in agreement with
that reported by Alander and Warnheim [1989] where they found that triglycerides, in
particular long alkyl chain triglycerides such as peanut oil, are considerably more difficult to
solubilize into microemulsions than hydrocarbons or alkyl esters.
Kunieda et al., [1988] maintained that the edible oils such as soybean, rapeseed or sunflower
oil contain long alkyl chains thereby the oil may be too bulky to penetrate the interfacial
film to assist the formation of the optimum curvatures. El-Nokaly et al., [1991] explained
this problem on the basis that the triglycerides are semi-polar compared to hydrocarbons and
the emulsifier efficiency is decreased if it is lost to the bulk and is unavailable to the
interface. Friberg and Kayali [1991], stated that as pointed out earlier, liquid triglycerides do
not lend themselves to microemulsion formation with the traditional technique hence, a
different strategy must be employed in order to prepare a microemulsion with these
compounds. The vegetable oils such as soybean and peanut oils are difficult to solubilize in
microemulsions because they contain long bulky alkyl chains (mainly C16,C18 and C20)
and preferentially form liquid crystalline mesophases which prevent the formation of
microemulsion [Joubran et al., 1994].
The peppermint oil has a strong flavor and using of it in high amount is unsuitable for taste
as well as it is relatively expensive comparing with the edible oils. Attempet was carried out
to make a mixture of peppermint oil and one suitable edible oil as a lipohhilic phase to avoid
or decrease the disadvantages of peppermint oil. The results obtained from these
 80

experiments were recorded in figures 5.2, 5.3 and 5.4. The results indicate that the use of
high amount of lecithin gives a high viscosity mixture and needs a very long time for its
solubility. This problem may be due to the waxy semisolid structure of lecithin (soybean
lecithin) and/or both strongly hydrophobic and lipophobic characters. The hydrophobic
nature due to long hydrocarbon chains on the other hand, lipophobic nature due to
zwitterionic polar head groups [Bergenstahl and Fontell 1983]. The using a small amount of
lecithin is difficult for the miscibility of water in the mixture containing peppermint oil
and edible oil to prepare microemulsions, especially when using a high amount of
triglycerides used in this case. Triglyceride-microemulsions needs to a large ratio of
surfactant to water as usually form microemulsions only with a high content of hydrophobic
surfactants according to the study by Mönig et al., [1996].
Due to the edible oils need a high amount of surfactant to prepare microemulsions [El-
Nokaly et al., 1991 and Möing et al., 1996] and these oils are difficult to sloubilize in
microemulsions [Alander and Warnheim 1989a and Joubran et al., 1994], the optimum
miscible amount of water to prepare a microemulsion was somewhat lower with the case in
which peppermint / soybean oil system used as a lipophilic phase than the case in which
peppermint oil alone used (see figure 5.1).
The water required to obtain the microemulsion containing peppermint, soybean oil and
lecithin decreased also with increasing the soybean oil and decreasing the lecithin as shows
in figure 5.2. This attributed to as reported by, Kuneida et al., [1988] the edible triglycerides
contain long alkyl chains, therefore the oil may be too bulky to penetrate the interfacial film
and/or according to El-Nokaly et al., [1991] in case of triglyceride microemulsions, the
emulsifier efficiency is decreased due to its inclusion in the bulk and becomes unable to the
interface. Consequently, the mixture solubility capacity for water decreases with increasing
the soybean oil content.
The effect of ethanol addition into the mixture containing peppermint oil, soybean oil and
lecithin on the maximum amount of water required to obtain the microemulsions was
investigated and the results were given also in figure 5.2. From these results shows that the
required water increased with increasing the ethanol but this increment reached to a certain
limit after that decreased with increasing the ethanol. When using the ethanol, it should be
observed that the maximum amount of water increased with increasing the lecithin content
in the mixture. The results obtained by using a mixture from peppermint mixed with
soybean oil, lecithin and water without or with ethanol were illustrated in figure 5.2. It is
clear from these results that the free cosurfactant microemulsions could be prepared. This
 81

result is important, where the cosurfactants are ,in general, harmful [Kahlweit et al., 1997]
and usually not acceptable for taste, safety, or performance [El-Nokaly et al., 1991].
Moreover, in these microemulsion systems the peppermint oil content was relatively small
comparing with that in the system containing peppermint oil and lecithin without edible oil.
Thus, the strong flavor of peppermint oil could be reduced. This result different from these
obtained by several outhers [Shinoda et al., 1991, Aboofazeli et al., 1994, Ho, et al., 1996]
where they found that the use of cosurfactant (short chain alcohol) is necessary to produce
lecithin based microemulsions. This can be explain on the basis that the nature of oil affect
the efficiency of lecithin and/or the peppermint oil used (in this study) act as cosurfactant
because it contain an alcohol (menthol).
It should be observed also that, the optimun amount of water increased by addition a small
amount of ethanol after that the optimum amount of water decreased with increasing the
ethanol. The results in figure 5.3 demonstrates that the microemulsions could be prepared by
using water, a mixture of peppermint oil with peanut oil and lecithin without addition of
cosurfactant. The microemulsions were prepared by using the same components (peppermint
oil, peanut oil, lecithin and water) with ethanol as cosurfactant. The effect of ethanol on the
optimum amount of water required to prepare the microemulsions was illustrated also in
figure (5.3). Concerning the comparison of results obtained by using peanut oil instead of
soybean oil which used in the previous experiment, it could be concluded that, the results in
either cases were approximatly similar with exception of few differences. These differences
included that the maximum amount of water when using soybean oil was slightly higher
than that when using peanut oil especially with increasing the lecithin content (ratio 1:1:2)
either without or with ethanol. The optimum amount of water when using soybean oil in
case of ratio 1:1:1 needs amount of ethanol lower than that in peanut oil under the same
conditions. This could be due to the differences in the chemical structure between soybean
oil and peanut oil, where the first oil contains unsaturated fatty acids relatively higher than
the latter oil. This explain is agreement with that repored by Parris et al., [1994], where they
fount that the nature of oil affect the oil miscibility in microemulsions. They found also that
triglycerides containing unsaturated or short chain fatty acids had improved miscibility of oil
in microemulsions compared to triglycerides with saturated or long chain fatty acids.
Microemulsions free cosurfactant were also prepared by using peppermint mixed with
rapeseed oil as lipophilic phase, lecithin and water and the results were illustrated in figure
5.4. The effect of ethanol in this microemulsion system was illustrated in the same figure
(5.4). It was observed that the effect of ethanol on the water miscibility in this system was to
 82

some extent similar to the above case. The results obtained in this case (using rapeseed oil)
compared to the results in case of using soybean oil (fig. 5.2) or peanut oil (fig. 5.3)
indicated that the results in the three cases were almost similar with some exceptions. These
exceptions included that the optimum amount of water when using rapeseed oil without
ethanol was higher especially in case of ratio 1:1:1of peppermint oil, rapeseed oil and
lecithin than that when using soybean oil or peanut oil at the same conditions. On the other
hand, on addition of ethanol the maximum amount of water was somewhat higher in case of
using rapeseed oil than those in case of soybean or peanut oil. This is thought to be due to
the difference in the chemical structure of the used edible oils. Aboofazeli et al., [1994]
studied partial phase diagrams of systems containing water, egg lecithin, propanol and
different polar oils such as Miglyol 812 and soybean oil, and they found that the influence
of the oil and the ratio of egg lecithin to propanol on the microemulsion area was
remarkable.

6.1.3 Monoolein microemulsions

The monooleate and linoleate could act as cosurfactant since they are different in
hydrocarbon moiety size. The oleate and linoleate groups are suitable as surfactant because
of their unsaturation and its known effect in increasing the interfacial fluidity [El-Nokaly et
al., 1991]. The pure mono-or di-glycerides are very expensive, therefore its use in food
industry is not economic. Monoolein used in this study contains mixture of mono- and di-
glycerides as well as triglycerides, therefore it can be used as surfactant and oil to prepare
the microemulsions. The addition of any amount of water into either monoolein alone or
mixed with soybean oil could not produced a microemulsion. But when using either
additional surfactant such as Tween20 or ethanol (as cosurfactant) the microemulsion could
be formed as shows in figure 5.5. This means, using of monoolein either alone or mixed
with soybean oil is insufficient to prepare microemulsions, this is attributed to the surfactant
content (mono-and di-glycerides) present in monoolein are small and the triglyceride
microemulsions need a high amount of surfactant [El-Nokaly et al., 1991 and Möing et al.,
1996]. Friberg and Kayali [1991] reported that triglycerides do not lend directly to
microemulsion formulation with the traditional technique hence, a different strategy must be
employed to prepare a microemulsion with these compounds. In this exepriment triglyceride
microemulsions were prepared by using two surfactants (monoolein and Tween20) or with
relatively high content of ethanol as cosurfactant. Troptow [1971] used two surfactants
 83

(Tween20 and G1045) to prepare soybean oil microemulsion and found that less than 10%
of oil could be dissolved in water with a 30/70 of surfactant mixture respectively.
It should be also noted that from this figure (5.5) the maximum amount of water to obtain
microemulsion was higher in case of using monoolein alone than that in case of mixture of it
and soybean oil. The addition of Tween20 as another emulsifier with monoolein allowed to
increase the amount of water required to form the microemulsion. This indicates that the use
of a high amount of surfactant has a great effect on water miscibility in monoolein
microemulsions. The nature and concentration of the surfactant become of utmost
importance to obtain a maximum solubility in a given water/oil microemulsion [El-Nokaly
et al., 1991]. Apparent also from the results in the same figure (5.5) that the water solubility
increased gradually with increasing of ethonol in all cases. This may be ascribed to the
effect of cosurfactant on surfactant efficiency, where the role of cosurfactant together with
the surfactant is to lower the interfacial tension down to a very small even transient negative
value at which the interface would expand to form fine dispersed droplets, and subsequently
adsorb more surfactant and surfactant / cosurfactant until their bulk condition is
depleted enough to make interfacial tension positive again, this process known as
spontaneous emulsification forms the miceoemulsion [Kunieda et al., 1988], and it is
assumed that the alkanol molecules adsorb preferntially at the region of strong curvature of
the interfacial film, thereby reducing its rigidity [Kahlweit et al., 1995]

6.1.4 Tween20 microemulsions

Polyoxyethylene sorbitan fatty acid esters have some interesting characteristics for
microemulsion formation and macro-molecules such as protein solubilization [Komives et
al., 1994]. Troptow [1971] prepared microemulsion by using Tween20, peppermint oil and
water without cosurfactant. Ayala et al., [1992] described the ability of a polyoxyethylene
sorbitan fatty acid esters (Tweens)-isopropanol microemulsion in hexane to solubilize pure
proteins and they found that this type of the surfactants forms stable reverse micelles which
can solubilize large amounts of protein and water in hexane. Accorrding to these data,
preparation of microemulsions by using Tween20 was tried. In this work Tween20
microemulsion was used to solubilize and stabilize curcumin.
 84

Microemulsions were prepared by using peppermint oil and Tween20 at various ratios with
water or 20% aqueous solution of NaCl, sucrose or citric acid. The results were illustrated
in figure 5.6, showed that when using a high amount of peppermint oil (more than in ratio
1:1), the microemulsion could not formed, since water was insoluble in such mixture. This
may be due to inefficiency of the emulsifier content. The optimum amount of aqueous phase
much increased and linearly with increasing the Tween20 in the mixture as seen in figure
5.5. It should be also noted that the maximum amount of aqueous phase solubilized to
produce Tween20-microemulsion was slightly lower only in case of NaCl relative to other
cases which were approximatly similar. This is possibly due to the salting out effect of
NaCl, since the elektrolyte lowers the solubility of nonionic surfactants in water [Brandt et
al., 1997]. The much and linear increase of maximum amount of aqueous phase with
increasing the Tween20 content is may be due to the high hydrofilicity of Tween20. This
result in agreement with that obtained by Osborne et al., [1991], where they found that the
surfactant concentration has a significant influence on the water solubility capacity of the
microemulsion system containing dioctyl sodium sulfosuccinate-sorbitan monolaurate as
surfactant.
The results obtained from these attempets showed that the use of this system (peppermint
oil/Tween20/water) exhibit great ability to incorporate a high amount of water within the
microemulsion system without cosurfactant.

6.2 Solubility of curcumin

Curcumin is insoluble in aqueous solutions and has poor solubility in other solvents [Henry
1996]. Microemulsions have the potential ability to solubilize both lipophilic and
hydrophilic species, which allows for a variety of flavoring and coloring agents having
vastly different physical properties to be dissolved within a system [Fribeg and
Burasczenska 1978]. In recent years, microemulsions have been identified as potential
delivery systems for lipophilic substrates due to their transparent appearance, long term
stability, high solubility capacity and ease of preparation [Malmsten 1996]. This study
describes attempt to formulate microemulsions containing suitable components for food
applications and use them to devolop solubility and stability of curcumin.
Curcumin solubility was determined on the basis of the maximum amount of curcumin
solubilized in the dissolution medium without siedment. The solubility was meassured as
 85

percent of curcumin in the dissolution medium by weight. Dissolution mediums were either
each solvent used to prepare microemulsions, each mixture contained the oil, surfactant and
cosurfactant (if used) or each microemulsion system prepared in this study.

6.2.1 Curcumin solubility in solvents

Curcumin is not an ideal product for direct use with food products since it is insoluble in
water and has poor solubility in other solvents. Thus it is usual for curcumin to be changed
into a convenient application form. This is achieved by dissolving the curcumin in amixture
of food-grade solvent and accepted emulsifier [Henry 1996].
The results were recorded in table 5.1 It could be noted that the curcumin is insoluble in
water and other aqueous solution of 20% NaCl, sucrose or citric acid, while the solubility
was very slight in case of edible oils. On the other hand, the curcumin solubility in case of
ethanol, pepperment oil or monoolein was higher than that in edible oils. The curcumin
solubility in Tween20 was considerably high comparing with other solvents. This may be
due to the curcumin is soluble in the emulsifiers owing to its nature, where they contain both
a lipophilic and hydrophilic group.

6.2.2 Solubility of curcumin in microemulsions

The solubility of curcumin in microemulsions containing oil, surfactant and oil in absence or
presence of ethanol was determined. The effect of water content in microemulsion system
on the curcumin solubility was invistgated and the results recorded in figures from 5.7 to
5.17.

6.2.2.1 Curcumin solubility in lecithin/peppermint oil microemulsions

The curcumin solubility in system prepared by using peppermint oil and lecithin decreased
slightly after addition the water as shows in figure 5.7. This may be due to the curcumin is
insoluble in water and/or the water decreases the solubility capacity of oil. From a
formulation viewpoint, the increased oil content in microemulsions may provide a greater
opportunity for the solubility of poorly water soluble agents [Gao et al., 1998]. Comparing
the results in figure 5.7 with those in figure 5.8 could be noted that the solubility increased
slightly after addition of the ethanol. This is thought to be due to the ethanol act to reduce
the tendency of lecithin to form a highly rigid film [Binks et al., 1989], thus lowering the
interfacial fluidity thereby increases the solubility capacity of the mixture. A similar result
 86

was obtained by Gao et al., [1998],where they found that the solubility of cyclosporin A
increased slightly with increasing the cosurfactant when dissolving cyclosporin A in
microemulsions based on triglyceride, polyoxyethylated castor oil as surfactant and
Transcutol as cosurfactant and they suggested that this slight increase may be due to the
excess of cosurfactant exists in the water phase and increased the solubility.
Microemulsions containing peppermint oil/lecithin/water offered a good results for
curcumin solubility where the solubility increased to 2 folds or 50 folds comparing with that
in peppermint oil or edible oil respectively. The effect of ethanol addition as cosurfactant on
the curcumin solubilty was slight. The use of ratio 2:1 of peppermint oil and lecithin
exihibted better result for curcumin solubility compared to that with 1:1 or 3:1. The
increasing of water content in the peppermintoil/lecithin microemulsion led to slight
decreasing the solubility of curcumin.

6.2.2.2 Solubility in lecithin/peppermint and soybean oil-microemulsions

It could be observed that the solubility increased with increasing the lecithin in mixture
contained peppermint and soybean oil as lipophilic phase as seen in figure 5.9. This is can
be explained due to the effect of lecithin as emulsifier on the solubility of curcumin. Peng
and Luisi [1990] stated that an important feature of phospholipid microemulsions in their
ability to solubilize larger guest molecules. The solubility increased also with adding a few
amount of water but after adding more than 2% the solubility almost decreased. The increase
of solubility at the first may be due to the solubility capacity of microemulsions containing
all components (required to their producing) is higher than that in case of absence one or
more of essential component [Gao et al., 1998]. But the decreasing after addition more than
2% of water may be ascribed to the excess of water decreases the lecithin efficiency, in
addition the curcumin is insoluble in water. On the other hand, the solubility increased
slightly when addition the ethanol into the system contained lecithin, peppermint mixed with
soybean oil and water (fogure 5.10) compared to the same system without ethanol. This is
possible due to the effect of ethanol on the mixture interfacial fluidity. From the same figure
(5.10) apparent also the solubility decreased when using ethanol with a high amount of
water content. This may be due to the excess of water incorporates with ethanol and
decreases its efficiency.
 87

6.2.2.3 Solubility in lecithin/peppermint and peanut oil-microemulsions

It could be noted that the solubility increased after addition a few amount of water into the
mixture contained a lower content of lecithin and decreased with increasing the water as
shows in figure 5.11. The increment (at the first) may be due to the microemulsions possess
high solubilization capacity compared to many other solutions [Friberg and Burasczenska,
1978; El-Nokaly et al., 1991 and Von Corswant et al., 1997] owing to extremely small of
their size droples [Goa et al., 1998]. The decreasing of solubility with increasing the water
content than 3% may be ascribed to the curcumin insolubility in water.
The curcumin solubility in the mixture containing a high content of lecithin decreased
slightly with addition a few amount of water and no more increase with increasing the water.
This may be due to the water decreases the lecithin efficiency on the curcumin solubility
because a part of lecithin dissolve in the water, but the no change in curcumin solubility
occurred with increasing the water can be explain on the basis, the water although decreases
the lecithin efficiency, leads to lowering the viscosity thereby increases the mixture
solubility.
The solubility of curcumin decreased slightly when using ethanol with increasing the water
in mixtures containing a smaller content of lecithin as to be seen in figure 5.12. This
decrease may be due to the curcumin is insoluble in water and/or the excess of water
incorporate with a part of ethanol. In contrast, in case of using a higher content of lecithin
and afew amount of water, the solubility of curcumin increased when using ethanol. This is
thought to be the ethanol increases the mixture fluidity consequently, the solubility capacity
increased. With increasing the water, the solubility decreased slightly in the same mixture,
this may be due to a part of ethanol dissolve in the excess water.

6.2.2.4 Solubility in lecithin/peppermint and rapeseed oil-microemulsions

The solubility of curcumin decreased slightly with increasing the water in case of the system
containing a smaller lecithin content, this may be attributed to the combination of water with
lecithin, so decreases its efficiency for solubility of curcumin. In case of the system
containing a higher lecithin and a lower rapeseed oil, the solubility increased slightly after
addition the water and this increase continued with increasing the water as to be seen in
figure 5.13. This is thought attributed to decreasing the mixture viscosity thereat increased
curcumin solubility.
 88

The curcumin solubility increased slightly when using the ethanol as cosurfactant with
peppermint and rapeseed oil/lecithin microemulsion as to be seen in figure 5.14. This may
be ascribed to that the ethanol increases the mixture fiuidity. Kahlweit et al., [1995]
stated that alkanols adsorb preferentially at the region of strong curvature of interfacial
layer, thereby lowering its rigidity. As shows in the same figure (5.14) the solubility
decreased slightly with increasing the water, this may be due to a part of ethanol dissolved in
the excess water, thereat decreased the available amount of ethanol.
The curcumin solublity in edible oil microemulsion was more better than in edible oil itself,
where the solubility increased from 0.02% in edible oil to 0.5% in case of edible oil
microemulsion. The curcumin solubility was however, lower when using the edible oil with
peppermint oil to prepare microemulsions than that when using peppermint oil without
edible oil. The curcumin solubility decreased with increasing the edible oil content in
peppermint-edible oil microemulsion, but in case of use a higher content of peanut oil the
solubility of curcumin increased slightly. The increase of water content in microemulsions
containing peppermint oil mixed with edible oil has slightly effet on the curcumin solubility.
Likewise, the effect of ethanol addition as cosurfactant on the curcumin solubility was
unremarkable.

6.2.2.5 Curcumin solublity in monoolein microemulsions

The solubility of curcumin in monoolein microemulsions containing ethanol, and water was
higher than that in monoolein singly. While the solubility of curcumin in monoolein
microemulsion containing soybean oil, ethanol and water was similar with that in monoolein
alone as shows in figure 5.15. This indicates that microemulsions possess high solubility
capacity [Malmsten 1996] due to their ultra low interfacial tension and extremely small size
droplets. As reported by Gao et al., [1998] the solubility of cyclosporin A ( poorly soluble in
the water ) in a system containing all components for producing microemulsions increased
markedly compared with those of system without surfactant or cosurfactant. It should be
also noted that the solubility of curcumin increased strongly when using Tween20 as
additional surfactant with monoolein to prepare microemulsions comparing with that in
other monoolein microemulsions as shows in the same figure (5.15). This is thought to be
due to the effect of Tween20 as emulsifier on the mixture solubility capacity.
Ayala et al., [1992] described the ability of a polyoxyethylene sorbitan triolate (Tween85) -
isopropanol microemulsion in hexane to solubilize pure proteins and they found that the
 89

nonionic surfactant Tween85 forms stable reverse micelles in hexane which can solubilize
large amounts of protein and water in hexane. It could be noted from the same figure (5.15)
that the solubility in monoolein microemulsions contained soybean oil was the lowest one,
this indicates that the curcumin is poor soluble in a long chain triglycerides. It could be
shown that the solubility of curcumin decreased slightly with increasing the water content
due to insolubility curcumin in water. The previous results showed that, the solubility of
curcumin in lecithin microemulsion was higher than that in monoolein microemulsions (with
exception of monoolein microemulsion containing Tween20). The use of Tween20 with
monoolein to prepare microemulsion enhanced the solubilization capacity water and
curcumin in microemulsions.

6.2.2.6 Curcumin solubility in Tween20 microemulsions

The curcumin solubility was determined in Tween20 microemulsions and the results were
illustrated in figures 5.16 and 5.17. From these results apparent that the solubility in
Tween20 microemulsions was considerably high compared to that in lecithin- or monoolein-
microemulsions. This indicates that Tween20 has a high ability to solubilize the curcumin,
this is due to its structure where it contain polyoxyethylene which may have high ability for
solubility of macro-molecules [Komives et al., 1994].
The curcumin solubility in the system containing a high content of Tween was higher than
that in the system containing a smaller content, indicated that the Tween content has a great
influence on the curcumin solubility. Apparent also from the results in figure 5.16 that the
solubility in case of microemulsion containing citric acid as aqueous phase was the highest
one. The slight increment of solubility when using citric acid is thought attributed to its
effect on the interficial tension of system. The effect of water addition on the curcumin
solubility is dependent on the content of Tween, where the solubility of curcumin decreased
slightly with increasing the aqueous phase in the system containing a smaller content of
Tween (figure 5.16). The solublity of curcumin decreased considerably with increasing the
aqueous phase when producing microemulsion with higher Tween20 content as shows in
figure 5.17. This may be due to the decrease of oil content in such system. As reported by
Gao et al., [1998] from a formulation viewpoint, the increased oil content in microemulsions
may offer a greater opportunity for the solubilization of poorly water-soluble agents.
 90

6.3 Light stability of curcumin in microemulsions

Safety of food colorants has been a very controversial topic in recent years but most of the
criticism has been usually towards the synthetic colorants. The natural colorants have been
relatively few of criticism may be due to the belief that most are derived from well known
food sources that have been consumed for many years [Francis 1996]. Curcumin is an
interesting molecule because of the variety of biological effects, it possesses in addition to
its potent anticancer activity [Khar et al., 1999]. The advantages of using natural colorants
are that they are generally more widely permitted in food-stuffs than synthetic colorants and,
contrary to many reports, natural sources can provide a comprehensive range of attractive
colors for use in the modern food industry [Henry 1996]. The problem with natural colorants
has been their lak of stability for the light and oxygen, which limits its use [Lauro 1998].
Curcumin changes in the light and its sensibility to light is a factor that usually limits its use
in foods [Henry 1996].
Microemulsions possess long term stability [Friberg and Burasczenska 1978, Bourrel and
Schechter 1988, and Malmsten 1996]. So, the attmpt for curcumin stability against light by
using microemulsions was studied. Microemulsions containing curcumin were stored under
UV light for 30 days, in darkness or under normal electric light for 60 days to investigate the
stability of curcumin solubilized in microemulsions against light. The stability was
measured as the change in color shade during the storage.

6.3.1 Curcumin stability in lecithin/peppermint oil microemulsions

The curcumin color shade was measured in the samples and control before and during
storage. The results were illustrated in figures from 5.18 to 5.23. From these results, it
should be noted that the color shade was stable in all samples stored either under UV light,
in darkness or under normal electric light, but it decreased markedly in case of control
praticularly, during storage under the light. The illustrated results indicated that the
microemulsions increased the stability of curcumin against light. This attributed to the
infinite stability of microemulsions [Bourrel and Schechter 1988], microemulsions are
thermodynamically stable because the surface free energy of them has two components
stretching (posative contribution) and bending (negative contribution) the two cancel
each other and the total surface free energy is extremely small about 10-3 mN/m, and the
interaction with light is limited [Friberg and Kayali 1991]. It should be also noted that the
 91

concentration of lecithin and oil in microemulsion system has no effect on the stability of
curcumin, further the effect of alcohol is not obviously. Ho et al., [1996] studied the stability
of insulin in microemulsions prepared by using polyglycerol fatty acid esters as surfactant
combined with short chain alcohols and they found that the stability of insulin in these
systems was quite acceptable, and the influence of alcohols on the stability of insulin in
microemulsions seems to be somewhat different, but not so obviously.

6.3.2 Curcumin stability in lecithin/peppermint/edible oil microemulsions

The change in the color shade of curcumin solubilized in microemulsions containing
soybean oil was illustrated in figures from 5.24 to 5.29. The results for lecithin/peppermint
and peanut oil microemulsions were illustrated in figures from 5.30 to 5.35. The resultes for
lecithin/peppermint and rapeseed oil microemulsions were illustrated in figures 5.36 to 5.41.
From these results, it could be observed that no changes in the curcumin color shade in case
of microemulsion samples comparing with those in control where the changes in control
were considerable during starage under UV or normal electric light. It was clear that
microemulsions exihibt a greater opportunity for the stability of curcumin against light.
Apparent also from these results that the nature and concentration of oil and lecithin in
microemulsion systems have no great effect on the stability of curcumin. Likeweise, the
effect of ethanol as cosurfactant in the microemulsions on the curcumin stability is not
obvious. Although the curcumin is sensitive to light as reported by [Henry 1996], the
microemulsions offered very good results for its stability against light for long period.
Friberg and Burasczenska [1978] reported that the advantages of a microemulsions over
emulsions, suspentions, or solutions are improved stability and solubility characteristics.
Since microemulsions are thermodynamicaly stable, separation problems that frequently
occurr with emulsions are avaided. Malmsten [1996] stated that, in recent years,
microemulsions have been identified as potential delivery systems for lipophilic agents due
to their transparent or translucent appearance, stability for long time, large solubility
capacity and ease of preparation.

6.3.3 Curcumin stability in monoolein microemulsions

The results for stability of curcumin solubilized in monoolein microemulsions were recorded
in figures 5. 42, 5.43 and 5.44. The results shows that the color shade was stable in case of
samples of monoolein microemulsions containing curcumin under UV light, in darkness or
 92

under normal electric light. The change in color shade of curcumin solubilized in control
was considerable during storage under light, and it was slight during storage in darkness.
This indicates that the monoolein microemulsions give a good stability for curcumin against
light for long period. The sensibility of curcumin to light is the one of the factors that
usually to be limited its application in foods [Henry 1996]. Monoglycerides are nonionic
surfactants widely used in the food and pharmaceutical fields [Holmberg and Österberg
1988], further some data have shown that nonionic surfactants exhibit the advantage of a
high chemical stability [Hofland et al., 1994].

6.3.4 Curcumin stability in Tween20 microemulsions

The results for the stability of curcumin in Tween20 microemulsions were recorded in
figures from 5.45 to 5.50. It could be showed that, all Tween20 microemulsion samples
exhibited a quite acceptable results for stability of curcumin against light for long period
comparing with that in control, exept the samples in which prepared by using NaCl as
aqueous phase. Gallarate et al., [1999] found that the rate of ascorbic acid oxidation, as its
degradation rate, was slower in microemulsions than that in aqueous solutions. In case of
Tween20-microemulsions prepared with NaCl as aqueous phase the color shade changed
markedly from yellow to red. This result is thought due to the effect of sodium cation in
NaCl solution on the curcumin color shade. Henry [1996] reported that, for curcumin, in
general, cations will tend to induce a more orange brown color shade.
The change in color shade of samples containing NaCl was higher during storage in
darkness or under normal electric light than that under UV light. This may be due to the UV
light has effect on the ionization process of NaCl and/or it has effect on the efficiency of
sodium cation on curcumin color shade. It should be noted that the curcumin color shade
tend to be lemon yellow with low pH value and orange with high pH value. This result in
agreement with that reported in Food Advisory Committee, FdAC/REP/4, HMSO, London,
[1987].
It could be also observed that the influence of pH value, concentration of oil and surfactant,
nature of oil and surfactant, presence of alcohol as cosurfactant in microemulsions on the
stability of curcumin during storage was no remarkable.
Carlotte et al., [1995] studied the rate of oxidation of linolic acid and ethyl linoleate in O/W
microemulsions and the results showed that the structure of the interface in the
microemulsions was especially significant to protect the systems from auto-oxidation, while
 93

the nature of the emulsifier does not seems to greatly affect the oxidation of unsaturared
molecules. Ho et al., [1996] found that the microemulsions maintained similar viscosity
during storage at room temperature for long period (several months), indicating that the
system was thermodynamically stable. In contrast, the viscosities of macroemulsions were
variable. Further, several microemulsion systems were demonstrated to be promising for
oral delivery of insulin based on the results of stability expriments and acid-protection
efficiency.
They found also that the influence of alcohols as cosurfactant on the stability of insulin in
microemulsions seems to be not so obviously.

6.4 Conclusion
Several microemulsion systems were prepared by using different natural oils such as
peppermint oil alone or mixed with a common edible oils (soybean, peanut or rapeseed oil),
different surfactants such as lecithin, monoolein or Tween20 and an aqueous solution
(water, 20% solution of NaCl, sucrose or citric acid ) without or with ethanol as
cosurfactant. The solubility of curcumin in these microemulsion systems was investigated.
The stability of curcumin solubilized in these microemulsions against light was also studied
during storage under UV-light for 3o days, in darkness and under normal electric light for 60
days. The stability of curcumin was determined as the change in curcumin color shade by
using Measuring Color Instrument.
The obtianed results can be concluded as follows :
1- Microemulsions could be prepared by using peppermint oil, lecithin and water without
cosurfactant
2- The optimum amount of water required to prepare microemulsions in this case, was
somewhat remarkable.
3- Microemulsions were also prepared by using these components with ethanol and the
optimum amount of water in this case decreased slightly with increasing the ethanol.
4- Microemulsions could not be prepared by using edible oils as lipophilic phase and
lecithin either without or with ethanol, but the use of peppermint oil with edible oil was
necessary.
5- The optimum amount of water to obtain microemulsions by the use of edible oil and
peppermint oil as lipophilic phase was lower than that in case of use peppermint oil
singly.
 94

6- The optimum amount of water almost decreased with increasing the edible oil content in
the mixture.
7- When addition the ethanol as cosurfactant into the system containing peppermint oil,
edible oil and lecithin, it was observed that the the optimum amount of water increased
at the first with increasing the ethanol to reached a certain amount after that decreased
with increasing the ethanol.
8- The efficiency of ethanol as cosurfactant on the water solubility in the microemulsion
system depended on the lecithin content, where its efficiency increase with increasing
the lecithin content.
9- No wide differences were noted between the edible oils used, where the nature of edible
oil does not seems to greatly affect the solubility of water in the microemulsion system.
10- The preparation of microemulsions by using monoolein alone as surfactant could not be
achieved.
11- The addition of ethanol or other emulsifier such as Tween20 with monoolein was
important to solubilize the water in the system containing monoolein either alone or
mixed with an edible oil.
12- Using Tween20 with monoolein has greatly affect on the solubilization of water in the
mixture to prepare microemulsion, where the maximum amount of water miscible with
the system was considerably remarkable compared to that when using ethanol with
monoolein either alone or mixed with soybean oil.
13- The use of soybean oil with monoolein decreases the water solubilized in the system.
14- The miscibilty of water increased linearly with increasing the ethanol either in case of
monoolein alone, mixed with soybean oil or mixed with Tween20.
15- Microemulsions were prepared by using peppermint oil, Tween20 and water or an
aqueous solution 20% of NaCl, sucrose or citric acid without cosurfactant.
16- The solubility of aqueous phase increased with increasing the Tween20 content.
17- NaCl seems to have the lowest solubility compared among all the tested aqueous
solution, while the other aqueous solutions given similar results.
18- Generally, the solubilization capacity for aqueous phase in case of using Tween20 as
surfactant was higher than that in case of using lecithin or monoolein.
19- An attempt to emulsify edible oils with Tween20 failed, where the mixture of Tween20
and edible oil was trubid and the turbidity increased after addition the aqueous phase.
20- The results obtained for solubility of curcumin in microemulsions showed that the
solubility of curcumin in microemulsionss was greatly higher than that in edible oil.
 95

21- Solubility of curcumin in microemulsions containing peppermint oil as lipophilic phase

was markedly higher than that in peppermint oil. Curcumin solubility in lecithin
microemulsions containing edible oil increased slightly after addition a few amount of
water, incontrast, in case of lecithin microemulsions containing peppermint oil.
Solubility of curcumin in lecithin microemulsions produced by using peppermint oil was
higher than that in lecithin microemulsions produced by using peppermint and edible oil.
22- Solubility of curcumin in Tween20-microemulsions was considerably higher than that in
lecithin or monoolein-microemulsions. The solubility of curcumin in Tween20-
microemulsionss increased with increasing the Tween content, but decreased with
increasing the aqueous phase.
23- The light stability of curcumin in microemulsions was studied as the change in the color
shade during storage and the results indicated that, with exception the microemulsions
prepared by using NaCl as aqueous phase, all microemulsion foemulations exhibited a
very good results for the stability of curcumin against either UV or normal electric light
for long period. In case of microemulsions formed by using NaCl as aqueous phase the
change in color shade during storage was remarkable either under light or in darkness.
24- The influence of ethanol which added as cosurfactant on the stability seems to be not
obviously, further pH value has no affect on the stability. The nature and concentration
of oil or surfactant used to formulate the microemulsions have also no effect on the
stability of curcumin.
Finally, Microemulsions could be prepared by using a suitable components for food
applications. These microemulsions could be used to improve the solubility of curcumin and
other natural food colorants which are insoluble in water and poorly soluble in vegetable
oils. These microemulsions were demonstrated to be promising for stability of curcumin and
other natural food colorants which are sensitive to light.
 96

7 Summary
Several microemulsion formulations were prepared by using various vegetable oils such as
peppermint oil alone or mixed with a common edible oils (soybean, peanut or rapeseed oil),
different surfactants such as lecithin, monoolein or Tween20 and an aqueous solution (water
or aqueous solution of 20% NaCl, sucrose or citric acid ) without or with ethanol as
cosurfactant. The solubility of curcumin in these microemulsion systems was investigated.
The stability of curcumin dissolved in these microemulsions against light was also studied
during storage under UV-light for 3o days, in darkness and under normal electric light for 60
days.
The obtianed results indicated that microemulsions could be prepared by using peppermint
oil, lecithin and water without cosurfactant. Microemulsions were also prepared by using
these components with ethanol as cosurfactant. Microemulsions could not be prepared by
using edible oils as lipophilic phase and lecithin either without or with ethanol, but the use
of peppermint oil with edible oil was necessary. The optimum amount of water required to
form microemulsions by the use of edible oil and peppermint oil was lower than that in case
of use peppermint oil singly.
The preparation of microemulsions by using monoolein alone as surfactant could not be
achieved. The addition of ethanol or other emulsifier such as Tween20 with monoolein was
important to solubilize the water in the system containing monoolein either alone or mixed
with an edible oil. The use of Tween20 with monoolein has greatly affect on the mixture
solubilization capacity for water. The miscibilty of water increased linearly with increasing
the ethanol either in case of monoolein alone, mixed with soybean oil or mixed with
Tween20.
Microemulsions were prepared by using peppermint oil, Tween20 and water or an aqueous
solution of 20% NaCl, sucrose or citric acid without cosurfactant. An attempt to emulsify
edible oils with Tween20 failed. The solubility of aqueous phase increased with increasing
the Tween20 content. Generally, the solubilization capacity for aqueous phase in case of
using Tween20 as surfactant was higher than that in case of using lecithin or monoolein.
The results obtained for solubility of curcumin showed that the solubility of curcumin in
microemulsionss was greatly higher than that in edible oil. Solubility of curcumin in
microemulsions containing peppermint oil alone as lipophilic phase was markedly higher
than that in peppermint oil. The effect of ethanol addition into lecithin microemulsions on
 97

the solubility of curcumin was small. Solubility of curcumin in Tween20-microemulsions

was considerably higher than that in lecithin or monoolein-microemulsions.
The light stability of curcumin in microemulsions was studied as the change in the color
shade during storage. The results indicated that microemulsions exhibited very good results
for the stability of curcumin against either UV or normal electric light for long period except
the microemulsion prepared by using NaCl as aqueous phase, where the change in color
shade of this microemulsion was great.
The influence of ethanol which added as cosurfactant on the curcumin stability seems to be
not obviously, further pH value has no effect on the stability. The nature and concentration
of oil or surfactant used to formulate the microemulsions have also no effect on the stability
of curcumin.
 98

8 References
1. Abe, M.; Schechter, D.; Schechter, R. S.; Wade, W. H.; Weerasooriya, U.; Yiv, S., J.
Colloid Interface Sci., 114 (1986), 342.
2. Abillon, O.; Chatenary, D.; Guest, D.; Langevin, D.; Meunier, J., Low interfacial
tension in miroemulsion systems. pp. 1159-1166 In: K. L. Mittal and P. Bothorel (eds.),
Surfactants in solution, Vol. 6 Plenum, New York. (1986).
3. Aboofazeli, R.; Lawrence, M.J. Investigations into the formation and characterization of
phospholipid microemulsions. II. Pseudo-ternary phase diagrams of systems containing
water-lecithin-isopropyl myristate and alcohol: influence of purity of lecithin. Int. J.
Pharm., 106 (1994) 51-61.
4. Aboofazeli, R.; Lawrence, C.B.; Wicks, S.R.; Lawrence, M.J. Investigations into the
formation and characterization of phospholipid microemulsions. III. Pseudo-ternary
phase diagrams of systems containing water-lecithin-isopropylmyristate and either an
alkanoic acid, amine, alkanediol, polyethylene glycol alkyl ether or alcohol as
cosurfactant. International Journal of Pharmaceutics, 111 (1994) 63-72.
5. Aboofazeli, R.; Patel, N.; Thomas, M.; Lawrence M. J., Investigations into the
formation and characterization of phospholipid microemulsions. IV. Pseudo-ternary
phase diagrams of systems containing water-lecithin-alcohol and oil; the influence of
oil. Int. J. Pharm. 125 (1995), 107-116.
6. Alander, J.; Warnheim, T. Model microemulsions containing vegetable oil part 1: Non-
ionic surfactant systems. JAOCS., 66 (1989a), 1656-1660.
7. Alander, J.; Warnheim, T. Model microemulsions containing vegetable oil part 2: Ionic
surfactant systems. JAOCS., 66 (1989b), 1661-1665.
8. Ayala, G. A.; Kamat, S.; Beckman, E. J.; Russell, A. J. Protein extraction and activity in
reverse micelles of a nonionic detergent. Biotechnol. Bioeng., 39 (1992), 806-814.
9. Azuine, M. A.; Bhide, S. V., J. Ethnopharmacol., 44 (1994), 211.
10. Backlund, S.; Rantala, M.; Molander, O., Characterization of lecithin-based
microemulsions used as media for a chlesterol oxidase-catalyzed reaction. Colloid
Polym. Sci., 272 (1994), 1098-1103.
11. Baltes, W., Lebensmittelchemie. 4 Auflage. Berlin, Heidelberg: Springer, (1995).
12. Bansal, V. K.; Shah, D. O.; O´Connel, J. P., J. Colloid Interface Sci., 75 (1980), 462-
475.
 99

13. Becher, P., Food Emulsions., In: „Microemulsions and Emulsions in Food“, El-Nokaly,
M. and Cornell, D., Eds. ACS Symposium Series No.448, American Chemical Society,
Washington, DC, (1991), p. 1.
14. Bergenstahl, B.; Fontell, K., Prog. Colloid Polym. Sci., 68 (1983), 48-52.
15. Bhavanishanker, T. N.; Murthy, K. N.; Murthy, K. S., J. Food Sci. & Technol., 23
(1986), 287.
16. Binks, B.P.; Meunier, J.; Langevin, D. Characterisitic sizes, film rigidity and interfacial
tensions in microemulsion systems. Prog. Colloid Polym. Sci., 79 (1989) 208-213.
17. Bourrel, M.; Schechter, R.S. Microemulsions and Related Systems. Formulations
Solvency, and Physical Poroperties. Surfactant Science Series 30, Marcel Dekker, New
York (1988), 27.
18. Brandt, M.; Wehling, A.; Schumpe, A. Microemulsions with Alkaylpolyglucocides.
Chemie Ingenieur Technik., 69 (1997), 500-504.
19. Breitmaier, E.; Jung, G. Organische Chemie II. 1 Aufl. Stuttgart: Thieme, (1983).
20. Brian, J. C.; Cizela, W. Arachidonic acid potentiates superoxide anion radical
production by murine peritoneal macrophages stimulated with tumour promoters;
Carcinogenesis, 10 (1989), 1769-1775.
21. Busch, C., Fett Wiss. Technol. 94 ( 1992), 380.
22. Carlotti, M.E.; Trotta, M.; Gasco, M. R. Influence of some components of emulsions
and microemulsions on the oxidation of linoleic acid and ethyl linoleate. S.T.P. Pharma
Sci., 5 (1995) 379-383.
23. Chen,J. P.; Pai, H., Hydrolysis of milk fat with lipase in reversed micelles. J. Food Sci.,
56 (1991), 234-237.
24. Constatinides, P. P.; Yiv, S. H. Particle size determination of phase inverted water-in-oil
microemulsions under different dilution and storage conditions. Int. J. Pharm. 115
(1995), 225-234.
25. Cooper, T. H.; Clark, G. ; Guzinki, J., in : Chi-Tang Ho (Ed.), Am. Chem. Soc.,
Washington, DC, 23 (1994) 231-236.
26. Cornell, B.A.; Middlehurst, J.; Separovic, F. Small unilamellar phospholipid vesicles
and the theory of membrane formation. Faraday Disc. Chem. Soc., 81 (1986) 163-167.
27. Davis, S. S.; Hadgraft, J.; Palin, K. J., in: Encyclopedia of emulsion technology; Becher,
P., Ed.; Marcel Dekker: New York/basel, 2 (1983), 159-238.
28. De Gennes, P.G.; Taupin, C. Microemulsions and flexibility of oil/water interfaces. J.
Phys. Chem., (1982) 2294-2304.
 100

29. Dickinson, E., Food Emulsions and Foams. Royal Society of Chemistry., London
(1987).
30. Dickinson, E.; Stainsby, G. Advances in food emulsions and foams, Elsevier Applied
Sci., London and New York, (1988).
31. Dörfler, H.-D.; Große, A.; Krüßmann, H. Microemulsions and their application in
model washing tests. Tenside Surf. Det., 32 (1995), 484-491.
32. Duxbury, D. D., Food Processing., May (1988), 62-64.
33. Ekwall, P., Advances in Liquid Crystals. Brown, G. H., Ed.; Academic: New York, 1
(1975), p. 1.
34. El-Nokaly, M.; George Hiler, Sr.; Joseph MeGrady, Solubilization of water and water-
soluble components in triglycerides. In: „Microemulsions and Emulsions in foods“, El-
Nokaly, M.; Cornell, D., Eds., ACS Symposium Series No. 448, American Chemical
Society, Washington, DC, (1991), p.26-43.
35. European Commission Document III/5218/94-rev.4, April 1995
36. Falbe, J.; Regitz, M., Hg. RÖMPP Chemie Lexikon. 9., erweiterte Auflage. Band 2.
Stuttgart, New York: Thieme Verlag, (1990).
37. Food Advisry Committee, FdAC/REO/4, HMSO, London (1987).
38. Food Advisry Committee, FdAC/REP/4, London (1987).
39. Francis, F. J., Safety of food colorants. pp. 112-130. In: Natural Food Colorants, second
edition. Edited by Hendry, G. A. F. and Houghton, J. D., (1996).
40. Friberg, S. E., Food Emulsions, Ed., Marcel Dekker: New York, (1976).
41. Friberg, S. E.; Burasczenska, I., Progr. Colloid & Polymer Sci., 63 (1978), 1.
42. Friberg, S. E.; Kayali, I. Surfactant association structures, microemulsions, and
emulsions in foods. In: „Microemulsions and Emulsions in food“, Magda, A., and
Donald C., Eds., ACS Symposium Series No. 448, American Chemical Society,
Washington, DC, (1991), p.7-24.
43. Friberg, S. E.; Rydhag, L., J. Am. Oil Chem. Soc. 48 (1971), 113-115.
44. Friberg, S. E., In: Food Emulsions. Larsson, K.; Friberg, S. E., Eds.; Marcel Dekker:
New York, (1990), p. 7.
45. Gallarate, M.; Carlotti, M. E.; Trotta, M.; Bovo, S. On the stability of ascorbic acid in
emulsified systems for topical and cosmetic use. International Journal of
Pharmaceutics., 188 (1999) 233-241.
 101

46. Gao, Z.-G.; Choi, H.-G.; Shin, H.-J.; Park, K.-M.; Lim, S.-j.; Hwang, K.-J.; Kim, C.-K.
Physicochemical characterization and evaluation of a microemulsion system for oral
delivery of cyclosporin A. International Journal of Pharamaceutics, 161 (1998), 75-86.
47. Gillberg, G., in „Emulsion and Emulsion Technology“, Lissant, K. J. Ed., Marcel
Dekker Inc., New York and Basel, Vol. III (1984) p.1.
48. Göklen, K. E.; Hatton, T. A. Protein extractions using reverse micelles. Biotechnol.
Porg., 1 (1985), 69-74.
49. Gulik-Krzywicki, T.; Larsson, K., Chemistry and Physics of Lipids, 35 (1984), 127-132.
50. Hatton , T. A. Reversed micellar extraction of proteins. In: J. F. Scamehorn and J. H.
Harwell (eds.), Surfactant-based separation process. Marcel Dekker, New York, Basel.
(1989), 55-90.
51. Hayes, D. G.; Gulari, E. Esterification reactions of lipase in reverse micelles.
Biotechnol. Bioeng., 35 (1990), 793-801.
52. Hayes, D. G.; Gulari, E. 1-Monoglyceride production from lipase-catalyzed
esterification of glycerol and fatty acid in reverse micelles. Biotechnol. Bioeng., 38
(1991), 507-517.
53. Henry, B. S. Natural food colours. pp. 40-79 In: Natural Food Colorants, second edition.
Edited by Hendry, G. A. F. and Houghton, J. D., (1996).
54. Ho, H-O.; Hsiao, C-C.; Sheu, M-T., Preparation of microemulsions using polyglycerol
fatty acid esters as surfactant for dielivery of protein drugs. Journal of Pharmaceutical
Sciniece, 85 (1996), 138-143.
55. Hofland, H. E. J.; van der Geest, R.; Bodde, H. E.; Junginer, H. E.; Bouwstra, J. A.,
Pharm. Res., 11 (1994), 659-664.
56. Holmberg, K.; Österberg, E., Progress Colloid Polym. Sci. 74 (1987), 150.
57. Holmberg, K.; Österberg, E. Enzymatic preparation of monoglycerides in
microemulsion. JAOCS., 65 (1988), 1544-1548.
58. Huang, T. S.; Lee, S. C.; Linn, J. K. Suppression of C-Jun/AP-1 activation by an
inhibitor of tumor promotion in mouse fibriblast cells. Proc. Natl. acad. Sci. USA, 88
(1991), 5292-5296.
59. Ishii, F.; Sasaki, I.; Ogata, H., Pharm. Pharmacol. 42 (1990), 513.
60. Jaruga, E.; Salvioli, S.; Dobrucki, Chrul, S.; Bandorowicz-Pikula, J.; Sickora, E.;
Franceschi, C.; Cossarizza, A.; Bartosz, G., FEBS LETT. 433 (1998), 287-293.
61. J. Research National Instit. Standers Technol. „ Structure of triglyceride
microemulsions for processed food“, 99 (1994), 210.
 102

62. Joubran, R. F.; Cornell, D. G.; Parris, N. Microemulsione of triglyceride and nonionic
surfactant –effect of temperature and aqueous phase composition. Colloids and Surfaces
A: Physicochem. Eng. Aspects 80 (1993), 153-160.
63. Joubran, R. F.; Parris, N.; Lu, D.; Trevino, S. Synergetic effect of sucrose and ethanol
on formation of triglyceride microemulsions. J. Dispersion Sci. Technol., 15 (1994),
687.
64. Kahlweit, M., J. Phys. Chem., 99 (1995), 1281.
65. Kahlweit, M.; Strey, R.: Busse, G. J. Phys. Chem., 94 (1990), 3881.
66. Kahlweit, M.; Busse, G.; Faulhaber, B. Preparing microemulsions with alkyl mono-
glucosides and the role of n-alkanols. Langmuir, 11 (1995), 3382-3387.
67. Kahlweit, M.; Busse, G.; Faulhaber, B. Preparing nontoxic microemulsions. 2.
Langmuir 13 (1997), 5249-5251.
68. Khar, A.; Mubarak, A. A.; Pardhasaradhi, B. V. V.; Begum, Z.; Anjum, R. Antitumor
activity of curcumin is mediated through the induction of apoptosis in AK-5 tumor cells.
FEBS Letters, 445 (1999), 165-168.
69. Kilara, A. Enzyme-modified lipid ingredients. Process Biochem., 19 (1985), 35.
70. Kolisis, F. N.; Valis, T. P.; Xenakis, A. Lipase catalyzed esterification of fatty acid in
nonionic microemulsions. Ann. NY Acad. Sci., 613 (1990), 674-678.
71. Komives, C.; Osborne, D.; Russell, A. J. Degradation of pesticides in a continuous-flow
two phase microemulsion reactor. Biotechnol. Prog., 10 (1994), 340-343.
72. Kunieda, H.; Shinoda, K., J. Dispersion Sci. Technol., 3 (1982), 233-244.
73. Kunieda, H.; Asaoka, H.: Shinoda, K., J. Phys. Chem. 92 (1988), 185.
74. Lang, J.; Rueff, R.; Dinh-Cao, M.; Zana, R., J. Colloid Interface Sci., 101 (1984), 184.
75. Langmuir, I., Met. Chem. Eng., 15 (1916), 468.
76. Langvin, D., In „Reverse Micelles, Proceedings of the European Science Foundation
Workshop“,eds. Luisi, P. L.; Straub, B. E. Plenum Press, New York., (1984), 287.
77. Larsson, K., J. Disp. Sci. Technol., 1 (1980), 267.
78. Lauridsen, J. B., Food Surfactants, Their Structure and Polymophism; Technical Paper
No. TP 908-1e, (1986), Grindsted Products A/S, Braband, Denmark.
79. Lauro, G. Interview with a color expert. Food Technol. 52 (1998), 76.
80. Lawrence, A.; Pearson, J., Proc. Int. Cong. Subs. 2, 709 (1964).
81. Leser, M. E.; Markoci, K.; Luisi, P. L. Reverse micelles in protein separation: The use
of silica for the back-transfer process. Biotechnol. Bioeng., 41 (1993), 489-492.
 103

82. Malmsten, M., In: Microemulsions: Fundamentals and applied aspects; Kumar, P.;
Mittal, K. L., Eds.; Marcel Dekker, Inc.: New York, (1996).
83. Mitchell, J. D.; Ninham, B. W., J. Chem. Soc. Faraday Trans. II 77 (1981), 601.
84. Mönig, K.; Haegel, F.-H.; Schwuger, M.J. Microemulsions with Plant Oils. Tenside
Surf. Det. 33 (1996) 228-231.
85. Monduzzi, M.; Caboi, F.; Larche, F.; Olsson, U., DDAB Microemulsions-Dependence
on the Oil Chain Length. Langmuir, 13 (1997), 2184-2190.
86. Osborne, D. W.; Pesheck, C. V.; Chipman, R. J. Dioctyl sodium sulfosuccinate-sorbitan
monolaurate microemulsions. In: „Microemulsions and Emulsions in foods“, El-
Nokaly, M.; Cornell, D., Eds., ACS Symposium Series No. 448, American Chemical
Society, Washington, DC, (1991), p. 63-79.
87. Parris, N.; Joubran, R. F.; Lu, D. P. Triglyceride microemulsions: Effect of nonionic
surfactants and the nature of the oil. J. Agrc. Food Chem., 42 (1994), 1295-1299.
88. Paul, B. K.; Moulik, S. P., Indian J. Biochem. Biophys., 28 (1991), 174-183.
89. Peng, Q.; Luisi, P. L. The behavior of proteases in lecithin reserve micelles. Eur. J.
Biochem. 188 (1990), 471-480.
90. Pires, M. J.; Aires-Barros, M. R.; Cabral, J. M. S., Biotechnol. Prog., 12 (1996), 290.
91. Pulla Reddy, A. C.; Lokesh, B.R., Mol. Cell. Biochem., 111 (1992), 117-124.
92. Rao, C. V. ; Rivenson, A.: Simi, B.; Reddy B. S. Chemoprevention of colon
carcinogenesis by dietary curcumin, a naturally occorring plant phenolic compound.
Cancer Research 55 (1995), 259-266.
93. Ravindranath, V.; Chandrasekhare, N., Toxicology, 16 (1980), 259-266.
94. Reddy, S.; Aggarwal, B. B. Curcumin is a noncompetitive and selective inhibitor of
phosphorylase kinase. FEBS Lett. 341 (1994), 19-22.
95. Rosen, M. J. Surfactants and interfacial phenomena. 2 nd edition. Wiley Interscience,
New York. (1989).
96. Ruby, A: J.; Kuttan, G.; Dinesh Babu, K. V.; Rajasekharan, K. N.; Kuttan, R. Antitumor
and atioxidant activity of natural curcuminoids. Cancer Lett. 94 (1995), 79-83.
97. Schröter, W.; Lautenschläger, K.-H.; Bibrack, H., Taschenbuch der Chemie. 17.,
durchgesehene Aufl. Frankfurt a. M.: Harri Deutsch, (1995).
98. Schulman, J. H.; Hoar, T. P. Trasparent water in oil dispersions: oleopathic
hydromicelle. Nature 152 (1943), 102-103.
 104

99. Schurtenberger, P.; Peng, Q.; Leser, M. E.; Luisi, P.-L. Structure and phase behavior
of lecithin based microemulsions: A study of the chain length dependence. Journal of
colloid and interface science, 156 (1993), 43-51.
100. Schwuger, M. J.; Stickdorn, K.; Schomäcker, R., Chem. Rev., 95 (1995), 849.
101. Shinoda, K.; Kaneko, T., J. Disperation Sci. Technol., 9 (1988), 555-559.
102. Shinoda, K.; Araki, M.; Sadaghiani, A.; Khan, A.: Lindman, B. Lecithin based
microemulsions; phase behavior and microstructure. J. Phys. Chem. 95 (1991) 989-993.
103. Shinoda, K.; Shibata, Y.; Lindman, B., Interfacial tensions for lecithin micremulsions
including the effect of surfactant and polymer addition. Langmuir, 9 (1993), 1254-1257.
104. Singh, C. P.; Skagerlind, P.; Holmberg, K.; Shah, D. O., A Compaison between lipase-
catalyzed esterfication of oleic acid with glycerol in monolayer and microemulsion
systems. J. Am. Oil Chem. Soc., 71 (1994), 1405-1409.
105. Skoog, D. A.; Leary, J. J., Insrrumentelle Analytik 1. Auflage. Berlin, Heidelberg:
Springer-Verlage, (1996).
106. Sonesson, C.; Holmberg, K., J. Colloid Interface Sci., 141 (1991), 239.
107. Srimal, R. C.; Dhawan, B. N., J. Pharmacy and Pharmacol., 25 (1973), 447.
108. Stamatis, H.; Xenakis, A.; Dimitriadis, E.; Kolisis, F. N. Biotechnol. Bioeng., 45
(1995)
109. Svensson, M.; Rees, G. D.; Robinson, B. H.; Stephenson, G. R. Phospholipid-
stabilised water-in-oil microemulsions: A study of ester synthesis by Humicola
lanuginosa lipase. Colloids and Surface B: Biointerfaces, 8 (1996) 101-111.
110. Töufel, A.; Ternes, W.; Tunger, L.; Zobel, M., (Hrsg.). Lebensmittel-Lexikon. 3.
Auflage Band 1. Hamburg: Behr´s Verlag, (1993).
111. Tonder, C.; Xenakis, A., Transport of solubilized pyrene by using O/W
microemulsions. Colloid Polym. Sci., 260 (1982), 232-233.
112. Treptow, R. S.. Research and Development Report, June 1, (1971), The procter and
Gamble Company, Cincinnati, OH.
113. Vasudevan, M.; Tahan, K.; Wiencek, M. Surfactant structure effects in protein
separations using nonionic microemulsions. Biotech. Bioeng., 46 (1995), 99-108.
114. Von Corswant, C.; Engström, S.; Söderman, O., Microemulsions based on soybean
phosphatidylcholine and triglycerides. Phase behavior and microstructure. Langmuir 13
(1997), 5061-5070.
115. Walde, P.; Giuliani, A. M.; Boicelli, C. A.; Luisi, P. L., Chem. Phys. Lipids, 53 (1990),
265-282.
 105

116. Winkler, H.-G., Chem. Lab. Biotech., 43 (1992), 378.

117. Wolf, P. A.; Havekotta, M. J., US Patent 4 835 002, (1989).
118. Zaks, A.; Klibanov, A. M. Enzyme catalyzed process in organic solvents. Proc. Natl.
Acad. Sci. USA, 82 (1985), 3192- 3196.
 106

Curriculum Vitae
Personal data
Familiy Name Abd El-Galeel
First Name Mohamed Awad Saad
Date of Birth 23 Jaunuary 1963
Place of Birth El-Etehad, El-Delengat, El-Beheira, Egypt
Nationality Egyptian
Sex Male
Marital Status Married

Education

1969-1974 Primary School

1975-1977 Middle School

1978-1981 Secondary School

1982-1986 Faculty of Agriculture, Department of Food Technology, Tanta University,

Kafr El-Sheikh, Egypt
Degree: B. Sc. In Agriculture Science

1987-1993 Position: Demonstrator in Dept. Of Food Technol., Fac. Of Agric., Tanta

Uni., Kafr El-Sheikh, Egept
Degree: Magister in Agric. Sci. (Food Technol.)

1994-1997 Position: Assistant Lecturer in Fac. Of Agric., Tanta Uni., Kafr El-Sheikh
Egypt

Since 04/1997 Ph. D. Scholarship of Egyptian Goverment, Institute of Food Technology,

University of Bonn, Germany
 Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

Terpene microemulsions for transdermal curcumin delivery: Effects of terpenes

and cosurfactants
Chi-Hsien Liu ∗ , Fu-Yen Chang, De-Kai Hung
Graduate Institute of Biochemical and Biomedical Engineering, Chang Gung University, 259 Wen-Hwa 1st Road, Kwei-Shan, Taoyuan 333, Taiwan

a r t i c l e i n f o a b s t r a c t

Article history: Microemulsion systems composed of terpenes, polysorbate 80, cosurfactants, and water were investi-
Received 27 May 2010 gated as transdermal delivery vehicles for curcumin. Pseudoternary phase diagrams of three terpenes
Received in revised form 15 July 2010 (limonene, 1,8-cineole, and ␣-terpineol) at a constant surfactant/cosurfactant ratio (1:1) were con-
Accepted 10 August 2010
structed to illustrate their phase behaviors. Limonene combined with cosurfactants like ethanol,
Available online 17 August 2010
isopropanol, and propylene glycol were employed as microemulsion ingredients to study their potential
for transdermal curcumin delivery. The transdermal delivery efficacy and skin retention of curcumin
Keywords:
were evaluated using neonate pig skin mounted on a Franz diffusion cell. The curcumin permeation rates
Transdermal
Curcumin
in the limonene microemulsion studied were 30- and 44-fold higher than those of 1,8-cineole and ␣-
Microemulsion terpineol microemulsions, respectively. Significant effects on the skin permeation rates were observed
Limonene from microemulsions containing different limonene/water contents. Histological examination of treated
skin was performed to investigate the change of skin morphologies. Characteristics such as droplet size,
conductivity, interfacial tension, and viscosity were analyzed to understand the physicochemical prop-
erties of the transdermal microemulsions. In conclusion, microemulsions loaded with curcumin were
successfully optimized for transdermal delivery after screening various terpenes, cosurfactants, and
limonene/water ratios. These results indicate that the limonene microemulsion system is a promising
tool for the percutaneous delivery of curcumin.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction of the patch in an area under the eye indicating mildness to the
skin of the developed curcuminoid patch. Additionally, the use of
Curcuminoids and curcumin exhibit multiple biological activ- curcumin for the chemoprevention and treatment of various skin
ities including anti-inflammatory, anticancer, antioxidant, wound diseases like scleroderma, psoriasis, and skin cancer was reviewed
healing, and antimicrobial effects [1]. Curcumin is used to treat vari- [10]. New microemulsion vehicles to efficiently deliver curcumin
ous disorders like tumors [2] and proinflammatory chronic diseases through the skin barrier may provide new therapeutic approaches
[3]. Nevertheless, widespread clinical application of curcumin in for many skin diseases.
cancer and other diseases has been limited due to its insufficient The stratum corneum is the top layer of the skin, and it is dif-
aqueous solubility and bioavailability. The poor absorption, rapid ficult for most drugs to be delivered into and through it. Many
metabolism, and rapid systemic elimination contribute to the low strategies have been employed to enhance dermal and transdermal
bioavailability of curcumin [4]. Many ways to improve the aqueous delivery. These include the use of chemical penetration enhancers,
solubility and oral bioavailability of curcumin were developed by preparation of supersaturated drug delivery systems, electrically
making curcumin nanoparticles [5], micelles [6], and phosphatidyl- driving molecules into or through the tissues by iontophoresis,
choline complexes [7]. Nevertheless, there is a significant first-pass and physically disrupting the skin structure by electroporation or
effect of the liver that metabolizes drugs delivered by the oral sonophoresis. Microemulsions provide another promising alterna-
route. Transdermal delivery represents an attractive alternative to tive for the dermal and transdermal delivery of both hydrophilic
oral delivery for local and systemic therapeutic uses [8]. Recently, and lipophilic drugs. Microemulsions are clear, stable, isotropic
Boriwanwattanarak et al. developed a hydrogel patch containing mixtures of oil, water, and surfactants [11]. Microemulsions offer a
curcumin for application for cosmetic purposes [9]. No skin irrita- powerful carrier system for drug delivery because of their high sol-
tion was observed in volunteers after a single 30 min application ubilization capacity, transparency, thermodynamic stability, ease
of preparation, and high diffusion rate [12–14]. Microemulsions
were applied to deliver different lipophilic drugs including flu-
∗ Corresponding author. Tel.: +886 3 211 8800x3146; fax: +886 3 211 8668. conazole [15], penciclovir [16], quercetin [17], hydrocortisone [18],
E-mail address: chl@mail.cgu.edu.tw (C.-H. Liu). progesterone [19], and indomethacin [20] through the transdermal

0927-7765/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.colsurfb.2010.08.018
64 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70

route. Those in vitro and in vivo studies demonstrated that drugs 2.3. Preparation and characterization of microemulsions
incorporated into microemulsions can efficiently penetrate the
skin. Microemulsions were prepared by mixing the oil with the
Essential oils such as terpenes are applied as additives in foods, surfactant or surfactant/cosurfactant mixture before adding the
cosmetics, and pharmaceutical products. Terpenes with repeating required amount of water under magnetic stirring. Curcumin
isoprene units are currently receiving much attention as penetra- (0.4 wt%) was then added to the prepared microemulsion. No
tion enhancers due to their low toxicities compared to synthetic phase change was noted after addition of the drug or after 30 min
penetration enhancers [21]. For example, limonene was reported equilibration in the water bath. microemulsion The transder-
to enhance the delivery of ketoprofen and aceclofenac through a rat mal delivery experiments were performed at 32 ◦ C to simulate
skin model [22,23]. Eucalyptus oil containing the essential oil 1,8- human skin temperature. The physicochemical characteristics of
cineole was used as a vehicle for the transdermal delivery of steroid the microemulsions are not significantly affected when the tem-
hormones [24] and hydrocortisone [18]. Additionally, ␣-terpineol perature was changed from 25 to 32 ◦ C. Viscosity determinations
was used as a transdermal enhancer for zidovudine and buspirone employed a Brookfield viscometer (DV II+, Brookfield Engineer-
[25,26]. However, there have been few studies investigating ter- ing Laboratories, Stoughton, MA, USA). The pH and conductivity
pene microemulsions for dermal curcumin delivery. The objective of the microemulsions were measured with a pH meter (WTW,
of this study was to investigate the constituents of oils, surfactants, Inolab 720, Burladingen, Germany) and conductivity meter (Con
and a cosurfactant on the formation of stable microemulsions using 500, Eutech Instruments, Singapore). Interfacial tension measure-
pseudoternary phase diagrams. The optimal combination of ter- ments were carried out at room temperature using a thin platinum
pene and the cosurfactant was found by screening their effects on plate attached to a transducer amplifier (Kyowa CBVP-A3, Saitama,
formulation stability and curcumin transdermal delivery. Transder- Japan). During the measurement, the plate is dipped into the test
mal delivery efficacy and skin retention of curcumin were evaluated microemulsions. The tensiometer measures the pulling force of
using neonate pig skin mounted on a Franz diffusion cell. Finally, the liquid on the plate and calculates the interfacial tension by
the effects of limonene/water ratios on the physicochemical char- dividing by the known plate size. The average particle size and poly-
acteristics of the microemulsions were studied to interpret the dispersity index were characterized using a Zetasizer Nano ZS 90
relation between transdermal delivery efficiency and the formu- (Malvern, Worcestershine, UK) at a fixed angle of 90◦ and a temper-
lations. ature of 25 ◦ C. Polydispersity index is an indication of the particle
size distribution. The smaller the polydispersity index is, the more
consistent the size distribution of dispersion is [39]. The refrac-
2. Materials and methods
tive index of the tested samples was in the range of 1.40–1.43.
Microemulsion samples were measured without dilution with
2.1. Materials
water due to the immiscible property of the microemulsion and
water.
1,8-Cineole, ␣-terpineol, limonene, tetrahydrofuran, and
ethanol were obtained from Sigma Chemical (St. Louis, MO, USA).
2.4. In vitro drug release
Propylene glycol and isopropanol were from Echo Chemicals
(Taipei, Taiwan). Curcumin was procured from Masterasia (Taipei,
Full-thickness skin from the porcine ear is a generally
Taiwan). Polysorbate 80 (commercially known as Tween 80) was
accepted model of permeation for human dermatological research.
purchased from IL-Shin Emulsifier (Seoul, Korea). All reagents
Researchers use pig ear skin for in vitro transdermal diffusion
were used without further purification. The water used in this
experiments [27,28]. In this study, the skin of the outside of the
study was freshly purified by the Milli-Q Gradient A10 system
ears of corpses of new-born piglets was used to study the per-
(Millipore, Molsheim, France).
meation of curcumin in a Franz diffusion assembly. The skin was
peeled from the underlying cartilage after cutting along the tips
2.2. Construction of the pseudoternary phase diagrams of the ears. The ear skin was mounted between the donor and
receptor compartments with the stratum corneum side facing the
Three terpenes (including 1,8-cineole, ␣-terpineol, and donor compartment. The donor medium consisted of 0.5 ml of
limonene), polysorbate 80, and ethanol were selected as the vehicle containing curcumin. The receptor medium (5.5 ml) was
oily phase, surfactant, and cosurfactant, respectively. Surfac- ethanol: pH 7.4 phosphate-buffered saline (PBS, 1:1) to maintain
tant/cosurfactant mixtures were tested at a ratio of 1:1 (w/w). sink conditions. The available diffusion area between the cells
The pseudoternary phase diagrams were constructed using the was 0.785 cm2 . The stirring rate and temperature were respec-
water titration method at room temperature [18]. For each phase tively kept at 600 rpm and 32 ◦ C. At appropriate intervals, the
diagram, mixtures of oil and surfactant or surfactant/cosurfactant entire receptor medium was withdrawn and immediately replaced
mixtures were prepared at weight ratios of 0.5:9.5, 1:9, 1.5:8.5, with equal volumes of fresh buffer. The permeated amount of
2:8, 2.5:7.5, 3:7, 3.5:6.5, 4:6, 5:5, 6:4, 7:3, 8:2, and 9:1. These curcumin was determined by high-performance liquid chromatog-
mixtures were titrated drop-wise with water under gentle mag- raphy (HPLC). The curcumin retained in the skin was measured
netic stirring. After being equilibrated for 30 min at the room after the permeation experiment (24 h). The skin was washed
temperature, the systems were visually characterized. In the three times using a cotton cloth containing ethanol. A skin sam-
study of Garti et al., the emulsion samples were vigorously stirred ple with the 0.785 cm2 permeated area was cut, weighed, and
and allowed to equilibrate for at least 24 h before they were then homogenized in a 50% ethanol solution. The resulting solution
characterized [38]. However, the 30 min equilibration time was was centrifuged, and the surpernatant was analyzed by HPLC. The
enough to reach the phase equilibrium in this study. Besides, permeated skin (0.785 cm2 ) was equivalent to 0.042 ± 0.008 mg
the addition of curcumin at 0.4% concentration only had a minor (n = 10).
effect on the phase behavior of microemulsion. Single-phase,
transparent mixtures with a droplet size <50 nm were designated 2.5. Curcumin analysis by chromatography
microemulsions. Coarse emulsions with an opaque appearance
in the phase diagram were illustrated as the non-microemulsion Quantification of curcumin was achieved using an HPLC system
region. (Jasco, Tokyo, Japan) consisting of a pump, a UV detector, and a
 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70 65

Table 1
Effects of essential oils and cosurfactants on curcumin dermal deliverya .

Oil Cosurfactant Curcumin permeated at 24 h (␮g cm−2 ) Curcumin in skin (␮g)b Flux (␮g cm−2 h−1 ) Kp (×10−4 cm h−1 )

Limonene – 0.71 ± 0.31c

6.94 ± 3.55c
0.02 ± 0.01c
0.05 ± 0.01c
Limonene Isopropanol 0.73 ± 0.07c 1.60 ± 0.43 0.03 ± 0.01c 0.08 ± 0.01c
Limonene Propylene glycol 1.58 ± 0.94c 2.66 ± 0.74 0.07 ± 0.05c 0.16 ± 0.12c
Limonene Ethanol 30.00 ± 4.77c 6.19 ± 3.27c 1.25 ± 0.25c 3.12 ± 0.63c
1,8-Cineole Ethanol 0.97 ± 0.13c 0.96 ± 0.05 0.04 ± 0.01c 0.10 ± 0.02c
␣-Terpineol Ethanol 0.38 ± 0.19c 1.56 ± 0.62 0.28 ± 0.10c 0.07 ± 0.03c
– Ethanol 0.01 ± 0.01 1.73 ± 0.88 0.01 ± 0.01 0.01 ± 0.01

Each value represents the mean ± SD (n = 3).

a
The composition of microemulsion is 20% oil, 35% polysorbate 80, 35% cosurfactants, and 10% water. Curcumin is added at the concentration of 0.4%.
b
Permeated skin with 0.785 cm2 area and 0.042 ± 0.008 mg weight in the Franz cell was homogenized and extracted to analyze the curcumin retained in skin.
c
p < 0.05 compared with the curcumin ethanol solution.

Microsorb-C18 column (Varian, Lake Forest, CA, USA). The mobile 3. Results and discussion
phase consisted of a 1% (w/v) citric acid solution, adjusted to pH 3.0
using a 45% (w/v) potassium hydroxide solution, and tetrahydrofu- 3.1. Effects of terpenes and cosurfactants in microemulsions on
ran in the ratio of 50:50 (v:v). The flow rate of the mobile phase was the percutaneous delivery of curcumin
1.0 ml/min [29]. The column effluent was monitored at 430 nm, and
the chromatographic data analysis was performed with the Borwin Effects of cosurfactants combined with three terpenes (1,8-
Program (vers. 1.5, Jasco). cineole, terpineol, and limonene) on curcumin delivery were
investigated in order to find the stable formulation with the best
2.6. Histological examination of skin specimens transdermal efficacy. The skin permeation studies employed full-
thickness piglet skin from the outer side ear. The cosurfactants
Porcine skin samples were treated with PBS and the optimized tested in this study included short-chain alcohols like ethanol, iso-
microemulsion (formulation B) respectively for 24 h on a Franz propanol, and propylene glycol which are relatively tolerated by
diffusion cell. Thereafter, skin samples were fixed in PBS solution the skin. Table 1 shows that the flux of curcumin through the
containing 10% formalin. Skin was cut vertically, dehydrated using skin was ranked in the following order ethanol > propylene gly-
ethanol, embedded in paraffin, and stained with hematoxylin and col > isopropanol > cosurfactant-free when limonene was used as
eosin (H&E) staining. These samples were then observed under light the oil. The amount of curcumin retained in the skin had a simi-
microscope (Olympus BX51, Tokyo, Japan) using 40× magnifica- lar order as the transdermal flux of curcumin. Increasing the chain
tion. length as the cosurfactant changed from ethanol to isopropanol
decreased the flux of curcumin. However, increasing the number
of hydroxyl groups as the cosurfactant changed from isopropanol
2.7. Data analysis
to propylene glycol increased the curcumin flux, and it remained
higher than that obtained in the cosurfactant-free system. The cur-
Cumulative amounts of curcumin permeated with time were
cumin flux of the limonene microemulsion with ethanol as the
used to calculate the transdermal drug flux, which was obtained
cosurfactant was 19- and 41-fold higher than those of propylene
from the slope of the regression line fitted to the linear portion of
glycol and isopropanol, respectively. It is important to note that
the profile. Extrapolation of this line intercepts the x-axis at a time
the addition of curcumin did not affect the isotropic structure of
equal to the lag time. Student’s t-test was used for the statistical
microemulsions. Although propylene glycol and isopropanol were
analysis. A 0.05 level of significance (p < 0.05) was considered sig-
successfully applied as cosurfactants in several microemulsion sys-
nificant. The skin flux can be experimentally determined from the
tems for various drugs [30,31], ethanol was found to the best
following equation [22]:
cosurfactant for curcumin delivery in a limonene microemulsion.
(dQ /dt) Similar results were obtained when comparing cosurfactants for
J=
A hydrocortisone transdermal flux with eucalyptus oil microemul-
sions [18]. For the tested cosurfactants, the amount of curcumin
where J is the steady-state flux (␮g cm−2 h−1 ), A is the diffusion
remaining in the skin was ranked in the same order when limonene
area of skin tissue (cm2 ) through which drug permeation takes
was used as the oil. Therefore ethanol was chosen as the cosurfac-
place, and dQ/dt is the amount of drug passing through the skin
tant for evaluation of different terpene microemulsions. Terpenes
per unit time at a steady-state (␮g h−1 ). The cumulative amount
are biocompatible oils that can enhance the efficacy of transdermal
of curcumin permeating through the porcine skin was plotted as a
delivery. Three terpenes of limonene, 1,8-cineole, and ␣-terpineol
function of time. The permeation rate of curcumin through piglet
were used to compose the oil phase of the microemulsions, and
ear skin was calculated from the slope of the linear portion of the
their effects on curcumin skin delivery were also evaluated. Among
plot. The intercept of the extrapolated linear region with the x-axis
the tested cyclic monoterpenes, the limonene microemulsion had
gave the lag time. The following equation was used to calculate the
the best transdermal delivery efficacy compared to those of 1,8-
permeability coefficient (Kp ):
cineole and ␣-terpineol. Cineole and ␣-terpineol microemulsions
J exhibited higher transdermal curcumin delivery compared to the
Kp = control (terpene-free) ethanol solution. The curcumin retained
Co
in the skin increased in the order limonene > ␣-terpineol > 1,8-
where Kp has the units of cm h−1 , and Co represents the curcumin cineole. The curcumin flux for limonene microemulsions was 30-
concentration (4000 ␮g ml−1 ). and 44-fold higher than those of 1,8-cineole and ␣-terpineol,
Statistical analysis of difference between various parameters respectively. The transdermal delivery of curcumin with the three
was performed using unpaired Student’s t-test. A 0.05 level of prob- terpenes significantly differed as shown in Table 1. Addition-
ability was taken as the level of significance. ally all terpenes could form a clear microemulsion in the tested
66 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70

terpenes including 1,8-cineole, menthol, and camphor [23]. Ter-

penes are reported to enhance drug diffusivity and partitioning
into the stratum corneum by disturbing the skin lipid bilayers
[11]. However, their enhancing function also depends on the drug
lipophilicity and vehicle composition. Previously, non-polar group
containing limonene provided better permeation effects for ace-
clofenac and ketoprofen than polar group-containing terpenes like
␣-terpineol and eucalyptus oil [22,23]. The microemulsion com-
posed of limonene and ethanol had the highest dermal delivery
potentials among the tested terpenes and cosurfactants herein.
Their increased curcumin flux may have been caused by a decrease
in curcumin’s affinity to the oil phase in the microemulsion. The
enhanced partitioning into the stratum corneum by limonene
might also account for the increased flux as evidenced by the
increased curcumin retained in the skin. The phase behavior of
microemulsions composed of three terpenes was investigated
using pseudoternary phase diagrams in the next section.

3.2. Phase diagram construction and effects of limonene

concentration on curcumin delivery
Fig. 1. The pseudoternary phase diagrams of microemulsion composed of oils,
polysorbate 80/ethanol, and water. ME represents the microemulsion region with
droplet size < 50 nm. NME stands for the non-microemulsion region with an opaque Phase diagram was a suitable approach to determine the
appearance. composition of the aqueous phase, oil phase, surfactant, and cosur-
factant from which the transparent, one-phase microemulsions
were formed. The composition of microemulsions was fixed at
oils and cosurfactants. The composition was chosen based on the
a constant polysorbate 80/ethanol mass ratio of 1:1. Limonene,
phase diagram (Fig. 1) and our preliminary experiments. The clear
eucalyptus oil, and ␣-terpineol were employed as the oil phase
microemulsions could be formed with the various oils and cosur-
to construct the microemulsion diagrams with polysorbate 80 and
factants tested in Table 1 under the constitution of 20% oil, 35%
ethanol as the surfactant and cosurfactant, respectively. Polysor-
polysorbate 80, 35% cosurfactants, and 10% water. Interestingly,
bate 80 was chosen as the surfactant because it is completely
Fang et al. found that 5% ␣-terpineol in an ethanol solution was
miscible with the three terpenes. Ethanol is a good cosurfac-
the best enhancer for curcumin permeation for 1 h pretreatment of
tant for many microemulsion systems. A stable microemulsion
rat skin [32]. However, their curcumin flux was 2.5-times less than
was formed when the content of polysorbate 80 and ethanol
our results, which might have been due to the different vehicles
was more than 60% as indicated by the pseudoternary phase
and animal skins used. Among the tested terpenes and cosurfac-
diagrams (Fig. 1). Pseudoternary phase diagrams are frequently
tants, the combination of limonene and ethanol had the highest
used to obtain the concentration ranges of the oil, surfactant,
transdermal delivery efficiency. The curcumin permeation of the
and cosurfactant for clear microemulsion formation. Construct-
combination of limonene and ethanol was 5-fold higher than that
ing phase diagrams makes it easy to determine the concentration
of the pure limonene microemulsion. Various terpenes were used
range of components where a transparent microemulsion exists.
as skin permeation enhancers for aceclofenac [22]. Limonene is
Fig. 1 exhibits the phase diagrams composed of water, surfac-
a better enhancer than 1,8-cineole for transdermal delivery. The
tant (polysorbate 80), cosurfactant (ethanol), and three terpenes.
addition of eucalyptus oil even hindered the delivery of Aceclofenac
Limonene had a smaller microemulsion area compared to those of
compared to the enhancer free control. Only limonene resulted
␣-terpineol and 1,8-cineole. Only 20% of the total area of the phase
in powerfully enhanced activity for ketoprofen among the tested
diagram was a transparent region when limonene was used as the
oil. The microemulsion zone occupied about 40 and 41%, respec-
tively, of the total area when ␣-terpineol and 1,8-cineole were used.
The fact that the limonene microemulsion needed more surfac-
tant/cosurfactant to form a clear microemulsion than eucalyptus oil
and ␣-terpineol might have been due to the high hydrophobic prop-
erty of limonene. All terpenes could form a stable microemulsion
region when the concentration of polysorbate 80/ethanol exceeded
64 wt%. The droplet in the microemulsion zone was around the
20 nm range (data not shown). A milky and opaque colloid solu-
tion (non-microemulsion, NME) appeared for the tested terpenes
when the concentration of polysorbate 80/ethanol was <38 wt%.
Terpene emulsions within this area were not stable, and phase
separation appeared when mixing was stopped. The area of the
microemulsion zone in the phase diagram is influenced by the
physicochemical properties of the oil phase, aqueous phase, sur-
factant, and cosurfactant. In addition, some essential conditions
are necessary for microemulsion formation including the exis-
tence of a very low surface tension at the oil–water interface, the
presence of a highly fluid interfacial film of surfactant, and the
penetration and association of oil molecules with the interfacial
Fig. 2. Effects of limonene and polysorbate 80 concentration on curcumin perme-
ated and retained in the excised pig ear skin with fixed water (10%) and polysorbate surfactant film [11]. In order to study the effective threshold for
80/ethanol ratio (1:1). The error bar stands for S.D. (n = 3). limonene on curcumin delivery, effects of several microemulsions
 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70 67

Table 2
Compositions and delivery characteristics of the curcumin microemulsionsa .

Limonene Water Curcumin in skin (␮g)b Flux (␮g cm−2 h−1 ) Kp (×10−4 cm h−1 )

A 30 0 9.29 ± 0.56c
0.50 ± 0.04c
1.24 ± 0.09c
B 25 5 29.88 ± 1.64c 1.10 ± 0.09c 2.74 ± 0.22c
C 20 10 25.83 ± 3.62c 0.68 ± 0.19c 1.69 ± 0.48c
D 15 15 23.01 ± 1.07c 0.61 ± 0.06c 1.52 ± 0.14c
E 10 20 22.60 ± 0.82c 0.21 ± 0.05 0.53 ± 0.10
F 5 25 14.66 ± 1.46 0.12 ± 0.08 0.30 ± 0.19

Each value represents the mean ± SD (n = 3).

a
Concentration of polysorbate 80 and ethanol is the same, 35%.
b
Permeated skin with 0.785 cm2 area and 0.042 ± 0.008 mg weight in the Franz cell was homogenized and extracted to analyze the curcumin retained in skin.
c
p < 0.05 compared with formulation F.

with various limonene concentrations were evaluated sequen- may account for the increased curcumin retention in the skin
tially. [22]. Importantly, transdermal curcumin permeation of the test
In the stable microemulsion region of the phase diagram, effects microemulsions was significantly affected by the limonene/water
of the limonene concentration on curcumin delivery were inves- ratios. The microemulsion (formulation B) containing 5% water
tigated by fixing the concentration of water at 10% and the ratio and 25% limonene had the best delivery efficiency, followed by
of polysorbate 80/ethanol at 1:1. Curcumin permeating into and the microemulsion containing 10% water and 20% limonene. For
retained in the excised pig ear skin was significantly affected transdermal delivery to take place, the drug encapsulated in the
when curcumin increased to 20 wt%. The permeation of curcumin vehicles has to be released from the microemulsion, partitioned
was very low when the concentration of limonene was ≤10 wt%. into the skin, and then diffused into the receptor solution. The
Interestingly, curcumin permeation in the microemulsion contain- skin plays a role as a reservoir for curcumin, which then sequen-
ing 20% limonene was 30-fold higher compared to that of 10% tially diffuses into deeper tissues (Table 2). It is important to note
limonene. Curcumin retention in the skin with the microemulsion that the microemulsion containing 5% water and 25% limonene
containing 20% limonene was 7-fold higher compared to that with achieved a 2-fold higher curcumin flux than that of the water-free
10% limonene after 24 h of permeation (Fig. 2). There was no signifi- microemulsion with only 30% limonene. The small amount of water
cant difference in the permeation flux when limonene was <10 wt%, in the microemulsion may have hydrated proteins in the stratum
which indicates that the effective threshold for limonene should be corneum and caused a disordering of the lipid bilayers in cor-
within the range of 10–20 wt%. Although water and the polysorbate neous cells [34]. Therefore, the cumulative curcumin permeation
80/ethanol ratio were respectively fixed at 10% and 1, the con- of microemulsion containing 5% water was improved compared to
centration of polysorbate 80 and ethanol was not fixed and varied the water-free microemulsion. However, the rate of transdermal
from 35 to 45% in these experiments. Therefore, the concentration curcumin delivery was reduced when the water content increased.
change of polysorbate 80 and ethanol on curcumin delivery needed In addition, the lag time is a permeation parameter that mainly
to be further studied in order to determine the effective threshold of depends on the diffusivity of the drug through the skin [18].
limonene under the microemulsion conditions. Recently, oil/water There were no significant differences in lag times among the six
ratios were found to be important for colloidal stability and trans- microemulsions indicating that the procedure of curcumin release
dermal drug delivery in several microemulsion systems [18,20,33]. from the microemulsions was not the limiting step (Fig. 3). The
The limonene/water ratios on physicochemical properties and cur- detailed curcumin retention in the skin and permeation results for
cumin delivery in microemulsions were studied as the following the six microemulsions are summarized in Table 2. The safety of the
section. transdermal delivery vehicle is an important factor in formulation
development. Histological examination of treated (formulation B)
and control (PBS) skin was performed using light microscopy to
3.3. Effects of limonene/water ratios on transdermal curcumin
delivery

When the concentrations of polysorbate 80 and ethanol were

>35 wt%, limonene formed a stable microemulsion with the aid
of ethanol as a cosurfactant in the phase diagram. In order to
investigate the delivery effects of the limonene/water fraction, the
detailed characterization and delivery of six potential microemul-
sion vehicles (composition referred to as formulation A–F in Table 2
and the phase diagram Fig. 1) were evaluated. The test formula-
tions were selected so that all formulations contained the same
concentrations of polysorbate 80 and ethanol (35% each) with
the remaining 30% being limonene and water. This selection thus
allowed us to investigate the effects of limonene and water con-
tents on the microemulsion characteristics and curcumin skin
delivery from the microemulsions. Furthermore, 0.4 wt% curcumin
was simultaneously added to the microemulsions to evaluate their
potential for dermal delivery. Curcumin retention in the skin was
slightly affected by the limonene/water ratios in the microemul-
sions (Fig. 3).
The microemulsion containing 5% water and 25% limonene
Fig. 3. Permeation profiles of curcumin through the excised pig ear skin from
showed the highest curcumin retention in the skin (Table 2). microemulsions A–F with various limonene/water ratios (mean ± S.D., n = 3). For-
Lipid extraction and osmotic expansion by the microemulsion mulation details are presented in Table 2.
68 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70

Fig. 4. HE staining skin micrographs after 24 h PBS treatment (A) and after 24 h Microemulsion (formulation B) treatment (B) with 40× magnification, scale bar: 100 ␮m.

investigate skin changes after H&E staining. Micrographs of control The polydispersion indexes of all formulations were larger than
and treated samples demonstrated similar skin layers as shown in 0.33 indicating these microemulsions had broad size distributions.
Fig. 4(A) and (B). A normal morphology of the stratum corneum The effects of terpene and water ratios on the physicochemical
was observed in both skins treated with microemulsion and PBS. properties of microemulsions were also studied. The conductiv-
No apparent epidermis changes were observed in the two samples. ity and viscosity of the microemulsions are used to describe the
Photograph of the skin sample with 24 h microemulsion treatment colloidal microstructure of the selected formulations. In addition,
showed a thin keratin layer, a normal dermis, and subcutaneous drug encapsulation was reported to affect the microstructure of
tissues. Keratin fragmentation and disruption of the lipid bilayers the delivery vehicles when added to microemulsions [35]. In this
was observed in the 24 h microemulsion-treated skin. Ablation of study, the conductivity, viscosity, pH, and surface tension of both
stratum corneum may have contributed to the enhancing effect of unloaded and curcumin-loaded microemulsions were character-
microemulsion on curcumin permeation. Photograph of the control ized. These results might provide some clues to explain why the
skin sample showed a reduplicated stratum corneum with five to limonene/water content affected the efficacy of curcumin der-
six distinct layers. No other changes were observed in the histol- mal delivery. The conductivity and viscosity of the unloaded and
ogy of microemulsion-treated skin. Some area of the skins (both curcumin-loaded microemulsions are presented in Fig. 6. When
microemulsion treated and control) had loose voids and empty the conductivity of microemulsions with increasing water content
spaces in the dermal region. followed a bell-shaped curve, this phenomenon is known as the
These observations support the in vitro curcumin permeation percolation transition indicating that the system changes from one
data. Transdermal delivery by microemulsions was previously of isolated oil droplets to an interconnected bicontinuous struc-
demonstrated to be safe for various drugs [6,18,32]. For example, ture [33,35]. In this study, the conductivity of the microemulsions
a triptolide microemulsion has no irritation in mouse skins but a increased from 15 to 35 ␮S/cm when the water content increased
triptolide aqueous solution was irritating [30]. This result suggests from 0 to 25%.
microemulsion can reduce the irritant effect of some materials. No percolation transition was observed in the limonene-
However, more in vivo tests are needed in order to evaluate containing microemulsions, which may have been due to the
whether the developed microemulsion is safe for the transdermal limited water content (<25 wt%). These increased conductivity lev-
delivery of curcumin. The average size and polydispersity index of els can be explained by the dilution of the oil/water microemulsion
the formulations with various limonene/water ratios were charac- with the added water which decreased the concentration of the
terized in Fig. 5. Formulation A, a water-free microemulsion, had dispersed oil droplets. One of the most special and useful prop-
the largest size than the other water-containing formulations. The erties of microemulsions is the typically low interfacial tension
average sizes of the tested microemulsions were less than 9.1 nm at liquid–liquid interfaces. The incorporation of curcumin slightly
which satisfied the size criterion of microemulsions. decreased the conductivity of the microemulsions. Furthermore,
the interfacial tension of the six formulations slightly increased

Fig. 5. The average droplet sizes and particle size distributions with various Fig. 6. Conductivity and pH for microemulsions A–F with various limonene/water
limonene/water ratios (mean ± S.D., n = 3). ratios (n = 3). Formulation details are presented in Table 2.
 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70 69

In addition, the decrease in limonene (to <15%) in the microemul-

sion also contributed to the decreased curcumin flux as shown in
Fig. 2.
Various mechanisms were reported to enhance transdermal
drug delivery from terpene microemulsions [11,18,32]. The first
mechanism is that limonene in the microemulsion can enter the
skin as a monomer to increase curcumin’s solubility in the skin.
Limonene in the skin increases the partitioning of the drug into the
skin and creates a high drug concentration within the upper layers
of the skin. This results in a higher concentration gradient which
is the driving force for transdermal drug delivery. This possibility
was supported by our results that a high flux was accompanied
by high curcumin retention in the skin. The second possibil-
ity is that limonene is a well-known skin penetration enhancer.
Limonene is known to produce greater skin penetration enhance-
ment when applied in combination with ethanol as a cosurfactant.
However, the limonene concentration should be >20% to signifi-
Fig. 7. Viscosity and surface tension for microemulsions A–F with various
limonene/water ratios (n = 3).
cantly enhance curcumin delivery (Fig. 2). The third mechanism
depends on the possibility of direct drug transfer from microemul-
sion droplets to the stratum corneum. As the drug molecules are
from 26 to 28 mN/m with an increase in the water content (Fig. 7). distributed in the microstructure of the microemulsion system with
Therefore, the interfacial tension was not the major factor deter- its very small droplet size, a very large surface area for drug transfer
mining the transdermal flux because the interfacial tension of the to the skin is created. This explains the superiority of microemul-
tested microemulsions was very low compared to the interfacial sion formulations over the curcumin ethanol solution. Accordingly,
tension of water (72.0 mN/m). The fact that interfacial tension the effect of the formulation on the skin may play a major role in
slightly increases with increasing water content might be due to the the improved skin permeation. The composition of the microemul-
surfactant partition from oil/water and air/liquid interfaces to the sions was found to be important in determining the efficiency of
water phase [36]. The change in the microstructure of microemul- microemulsions as skin drug delivery systems.
sion caused by the reorganization of surfactant might contribute to
the increase in viscosity after the increasing water content [37].
4. Conclusions
Similar observations in microemulsions viscosity and interfacial
tension elevated by water addition were reported by Podlogar et
Microemulsions have received great attention in recent years
al. [33] and Djordjevic et al. [35]. In addition, the pH value was
for various therapeutic delivery applications. Pseudoternary phase
found to have only slightly decreased from 7.1 for both unloaded
diagrams were used to illustrate the phase behaviors of ter-
and curcumin-loaded microemulsions. The addition of curcumin
pene microemulsions. The curcumin permeation rates in the
to the microemulsions only slightly affected the final pH value.
limonene microemulsion studied significantly increased compared
The carbon dioxide in water will be converted into carbonic acid,
to that of eucalyptus oil and ␣-terpineol microemulsions. The skin
which dissociates to liberate hydrogen ions and reduce the pH
permeation rates greatly increased when the limonene/water con-
value when the water content was higher than 5%. The viscosity
tents were appropriately adjusted. Microemulsions composed of
increased with an increase in water and a decrease in limonene
limonene, polysorbate 80, ethanol, and water showed enhanced
in the microemulsion. The viscosity range of the microemulsions
transdermal curcumin delivery by reducing the diffusional barrier
prepared was 10–40 ␮s/cm with an increase in the water content
of the stratum corneum and the viscosity. Microemulsions loaded
from 0 to 25%. Previously, the viscosity of drug-loaded microemul-
with curcumin were successfully developed for dermal delivery
sions was reported to influence drug partitioning into the skin.
after screening various terpenes and cosurfactants and optimiz-
The higher the viscosity was, the lower the drug delivery effi-
ing the limonene/water ratio. This limonene microemulsion system
cacy was [35]. Due to their amphiphilic nature surfactants could
was demonstrated to be a promising tool for the percutaneous
form various structures in oil and water phases depending on the
delivery of curcumin.
respective concentrations. In this study a high concentration (35%)
of polysorbate 80 was constitutively added in the formulations
(A–F). The surfactant micelles start to elongate in one or in two Acknowledgements
dimensions, leading to their transformation to bicontinuous and
lamellar structures under high surfactant conditions [33]. It is well The authors are indebted to Professor Jia-You Fang at Chang
known that change in water/oil ratio affects the physicochemi- Gung University for his suggestions on this work. The project
cal property and microstructure of microemulsions [11–13]. The was supported by grants from the National Science Council (NSC
total amount of oil and water was 30% in the tested microemul- 98-2221-E-182-013) and Chang Gung Memorial Hospital (CMRPD
sions, while the dominant ingredients were 35% ethanol and 35% 170081), Taiwan.
surfactant. Since the water and oil was less than the surfactant
or ethanol, microemulsion structure could not be distinguished References
among the O/W, W/O or bicontinuous types. Besides, our conduc-
tivity results demonstrated no percolation transition happened in [1] R.K. Maheshwari, A.K. Singh, J. Gaddipati, R.C. Srimal, Life Sci. 78 (2006)
2081–2087.
the tested water/oil ratios. This result indicated that the structure of
[2] G. Kuttan, K.B. Kumar, C. Guruvayoorappan, R. Kuttan, Adv. Exp. Med. Biol. 595
the microemulsions was not changed from the isolated droplets to (2007) 173–184.
an interconnected bicontinuous structure. Actually no attempt was [3] V.P. Menon, A.R. Sudheer, Adv. Exp. Med. Biol. 595 (2007) 105–125.
made to distinguish among the microemulsion types in this study. [4] P. Anand, A.B. Kunnumakkara, R.A. Newman, B.B. Aggarwal, Mol. Pharm. 4
(2007) 807–818.
The limonene microemulsion with increased water had low perme- [5] S. Bisht, G. Feldmann, S. Soni, R. Ravi, C. Karikar, A. Maitra, A. Maitra, J.
ation flux which may have been caused by the increased viscosity. Nanobiotechnol. 5 (2007).
70 C.-H. Liu et al. / Colloids and Surfaces B: Biointerfaces 82 (2011) 63–70

[6] K. Maiti, K. Mukherjee, A. Gantait, B.P. Saha, P.K. Mukherjee, Int. J. Pharm. 330 [23] Y.S. Rhee, J.G. Choi, E.S. Park, S.C. Chi, Int. J. Pharm. 228 (2001) 161–170.
(2007) 155–163. [24] B. Biruss, H. Kahlig, C. Valenta, Int. J. Pharm. 328 (2007) 142–151.
[7] T.H. Marczylo, R.D. Verschoyle, D.N. Cooke, P. Morazzoni, W.P. Steward, A.J. [25] M. Al-Khalili, V.M. Meidan, B.B. Michniak, AAPS Pharm. Sci. 5 (2003).
Gescher, Cancer Chemother. Pharmacol. 60 (2007) 171–177. [26] S.T.K. Narishetty, R. Panchagnula, J. Control Release 95 (2004) 367–379.
[8] M.R. Prausnitz, R. Langer, Nat. Biotechnol. 26 (2008) 1261–1268. [27] C. Balaguer-Fernandez, A. Femenia-Font, S.D. Rio-Sancho, V. Merino, A. Lopez-
[9] P. Boriwanwattanarak, K. Ingkaninan, N. Khorana, J. Viyoch, Int. J. Cosmet. Sci. Castellano, J. Pharm. Sci. 97 (2008) 2102–2109.
30 (2008) 205–218. [28] C.M. Heard, S. Johnson, G. Moss, C.P. Thomas, Int. J. Pharm. 317 (2006) 26–31.
[10] R.L. Thangapazham, A. Sharma, R.K. Maheshwari, Adv. Exp. Med. Biol. 595 [29] Y. Pak, R. Patek, M. Mayersohn, J. Chromatogr. B: Anal. Technol. Biomed. Life
(2007) 343–357. Sci. 796 (2003) 339–346.
[11] A. Kogan, N. Garti, Adv. Colloid Interface Sci. 123–126 (2006) 369–385. [30] H. Chen, X. Chang, T. Weng, X. Zhao, Z. Gao, Y. Yang, H. Xu, X. Yang, J. Control
[12] Y. Gao, Y. Wang, Y. Ma, A. Yu, F. Cai, W. Shao, G. Zhai, Colloids Surf. B: Biointer- Release 98 (2004) 427–436.
faces 71 (2009) 306–314. [31] S. Peltola, P. Saarinen-Savolainen, J. Kiesvaara, T.M. Suhonen, A. Urtti, Int. J.
[13] T.P. Formariz, L.A. Chiavacci, V.H.V. Sarmento, C.V. Santilli, E.S. Tabosa do Egito, Pharm. 254 (2003) 99–107.
A.G. Oliveira, Colloids Surf. B: Biointerfaces 60 (2007) 28–35. [32] J.Y. Fang, C.F. Hung, H.C. Chiu, J.J. Wang, T.F. Chan, J. Pharm. Pharmacol. 55 (2003)
[14] S.K. Mehta, G. Kaur, K.K. Bhasin, Colloids Surf. B: Biointerfaces 60 (2007) 95–104. 593–601.
[15] Y.G. Bachhav, V.B. Patravale, Int. J. Pharm. 365 (2009) 175–179. [33] F. Podlogar, M. Gasperlin, M. Tomsi, A. Jamnik, M.B. Roga, Int. J. Pharm. 276
[16] W. Zhu, A. Yu, W. Wang, R. Dong, J. Wu, G. Zhai, Int. J. Pharm. 360 (2008) (2004) 115–128.
184–190. [34] A. Azeem, F.J. Ahmad, R.K. Khar, S. Talegaonkar, AAPS Pharm. Sci. Technol. 10
[17] F.T.M.C. Vicentini, T.R.M. Simi, J.O. Del Ciampo, N.O. Wolga, D.L. Pitol, M.M. (2009) 1093–1103.
Iyomasa, M.V.L.B. Bentley, M.J.V. Fonseca, Eur. J. Pharm. Biopharm. 69 (2008) [35] L. Djordjevic, M. Primorac, M. Stupar, D. Krajisnik, Int. J. Pharm. 271 (2004)
948–957. 11–19.
[18] G.M. El Maghraby, Int. J. Pharm. 355 (2008) 285–292. [36] S.K. Mehta, G. Kaur, K.K. Bhasin, Pharm. Res. 25 (2008) 227–236.
[19] B. Biruss, C. Valenta, Int. J. Pharm. 349 (2008) 269–273. [37] S. Sajjadi, Colloids Surf. A: Physicochem. Eng. Aspects 299 (2008) 73–78.
[20] G.M. El Maghraby, Colloids Surf. B: Biointerfaces 75 (2010) 595–600. [38] N. Garti, A. Yaghmur, M.E. Leser, V. Clement, H.J. Watzke, J. Agric. Food Chem.
[21] M.K. Das, A. Bhattacharya, S.K. Ghosal, Drug Deliv. 13 (2006) 425–431. 49 (2001) 2552–2562.
[22] J. Lee, Y. Lee, J. Kim, M. Yoon, W.C. Young, Arch. Pharm. Res. 28 (2005) [39] C.H. Liu, S.Y. Yu, Colloids Surf. B: Biointerfaces 79 (2010) 509–515.
1097–1102.
 international
journal of
ELSEVIER International Journal of Pharmaceutics 120 (1995) 127-136
pharmaceutics

Invited Review

Using microemulsion formulations for oral drug delivery of

therapeutic peptides
J.M. Sarciaux, L. Acar, P.A. Sado *
Laboratoire de Pharmacie Gal~nique, Biopharmacie et Pharmacie Clinique, Facult~ des Sciences Biologiques et Pharmaceutiques, 2
Avenue du Professeur Leon Bernard. Unit,ersit~ de Rennes 1, Rennes, France
Received 7 July 1994; accepted 28 October 1994

Abstract

Peptide drugs are increasingly becoming a very important class of therapeutic agents with the rapid advances in
the field of biotechnology engineering. However, these drugs are generally not suitable for oral administration. In
this review, the main physico-chemical and biopharmaceutical characteristics of peptides are summarized. The
obstacles to peptide drug absorption and the different possibilities for solving these difficulties are listed. Results
using this formulation approach for oral drug delivery of peptides are apparently promising with some specific
peptides such as cyclosporin. Various mechanisms are only beginning to be understood and further investigations
need to be performed in this area to explain the results obtained with some peptides.

Keywords: Peptide; Microemulsion; Oral drug delivery

I. Introduction (Samanen, 1985). Aliments that might be treated

more effectively include cancer (Triptorelin®),
M a n a g e m e n t of illness through medication is mental disorders, hypertension (renin inhibitors),
about to enter an era of rapid growth in the area autoimmune diseases and certain cardiovascular
of pharmaceuticals that are of peptide and pro- and metabolic diseases (Lee, 1986).
tein origin. Decades of peptide research have One typical example is the discovery of cy-
created a wide variety of biomedical peptide hor- closporin, an established immunosuppressant in
mones, synthetic peptides, enzyme substrates and the m a n a g e m e n t of organ transplant patients.
inhibitors (Lee, 1986). The growing interest can This peptide is a neutral, lipophilic, cyclic unde-
be ascribed to the increased understanding of capeptide (Kahan, 1989).
their role in physiology and therapy as well as the Although they are highly potent and specific in
established capability of producing a large quan- their physiological functions, most of them are
tity by sophisticated biotechnological processes difficult to administer orally. They are usually not
therapeutically active by oral administration. In
fact, they have an extremely short biological half-
life when administered parenterally and repeated
* Corresponding author. injections are often needed. These compounds

0378-5173/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved

SSDI 0378-5173(94)00386-6
128 J.M. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136

require parallel development of viable delivery the principle causes of poor membrane perme-
systems to improve their systemic bioavailability ation of peptides. Possible solutions and exam-
(Verhoef et al., 1990). ples of applications are given. As the focus of this
In reality, the commercial success of peptide review is on the use of microemulsions for pep-
drugs for medication will depend on the develop- tides, this new drug delivery system will be pre-
ment of non-parenteral routes of administration sented. Therapeutic peptide drugs using this new
such as nasal, oral, and rectal or on the successful strategy formulation will be given.
development of other novel approaches such as
implantable delivery systems, self-regulating de-
livery systems to overcome the drawbacks associ- 2. Physico.chemical and biopharmaceutical char-
ated with parenteral administration. Frequent in- acteristics of peptides
jections are poorly accepted by patients except
those suffering from life-threatening diseases. Proteins are the most abundant components of
The oral dosage form is for several reasons the cells and are constructed from the same 20 amino
most popular way of administration. One reason acids. They are macromolecules with molecular
is patient compliance which is important in long- weights ranging from approx. 5000 to several mil-
term therapy with potent drugs as peptides. In an lions.
attempt to improve patient acceptance, new Each protein molecule is a polymer with a-
strategies are used to improve the oral bioavail- amino acids linked together in sequential manner
ability of peptides. Bioavailability is the fraction by peptide bonds. The resulting polymers are
or percentage of a dose that reaches the systemic called peptides. The term polypeptide refers to
circulation intact, when not directly injected into the peptides which contain about eight or more
the circulation (Devissaguet, 1982). amino acids whereas oligopeptides are those pep-
Bioavailability is clinically important because tides which have less than eight amino acids.
pharmacologic and toxic effects are proportional Polypeptides that contain from about 50 to as
to both dose and bioavailability. When bioavail- many as 2500 amino acids are called proteins
ability is very low as in the case of most peptides, (Banga and Chien, 1988). The polypeptide chain
inter- and intrasubject variability in bioavailability of a protein is folded into a specific three-dimen-
is magnified and incomplete. Therefore, the ma- sional structure which is referred to as the con-
jor reason for trying to maximize oral bioavail- formation of the protein.
ability is to maximize our ability to control drug Based on the conformation, proteins can be
concentrations and the effect of analysis. The classified into two major classes: fibrous and
biological potency characteristic of these globular. The second class, globular proteins, in-
molecules requires methods that are exquisitely cludes peptide pharmaceuticals whereas the first
sensitive (Samanen, 1985). class forms the basic structural elements in the
Cost may be another driving force. If bioavail- connective tissue (Banga and Chien, 1988; Sama-
ability averages 10%, then 90% of a dose is nen, 1991). Unless their N- and C-termini are
wasted. For drugs that are expensive to produce, blocked through cyclization, amide formation, or
wasting 90% of the material may not be accept- esterification, peptides and proteins are usually
able. It must be considered as an important fac- charged and hydrophilic molecules (Lee, 1988).
tor for peptide drugs (Aungst, 1993). In reality, Some exceptions exist such as cyclosporin with a
maximizing bioavailability contributes to increas- partition coefficient of 991 (in n-octanol, pH 7.4)
ing cost-effectiveness. (Kahan, 1989). Peptides can adopt different
To improve oral absorption and bioavailability structures: linear, cyclic, cyclic disulfide and bi-
of these compounds, many different strategies cyclic (Wieland and Bodanszky, 1991).
can be used. This review presents first briefly the Peptides that occur naturally as cyclic struc-
structural aspects of peptides and their major tures typically require the cyclic structure for
physico-chemical properties. Then, it examines activity. Both somatostatin and vasopressin are
 J.M. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136 129

cyclic by virtue of the disulfide bridges between luminal pancreatic enzymes, and for tripeptides
two cysteine residues in the peptides. Analogues and larger peptides to be metabolized by intesti-
of either peptide that lack cyclic structures are nal cytosolic enzymes (Verhoef et al., 1990).
inactive (Samanen, 1985). Another important barrier is the physical bar-
The oral route presents a severe obstacle for rier. A specific peptide to be absorbed must
peptide drugs due to the enzymatic and physical bypass the unstirred water layer, mucous layer,
barriers of the intestinal tract. The enzymatic apical and basal cell membranes and cell con-
barrier is by far the most important barrier. The tents, tight junctions, basement membrane and
oral bioavailability of most peptides is less than the wall of lymph and blood capillaries (Van
10% (Lee, 1988). The exceptions are cyclosporin Hoogdalem et al., 1989).
and thyrotropin-releasing hormone which are re- Peptide drugs can pass from the lumen into
sistant to proteolysis (Lee, 1988). the bloodstream via two routes: transcellularly, in
The enzymatic barrier, composed of exo- and which molecules are transported into and through
endopeptidases, is well designed for the digestion epithelial cells and transferred into the systemic
of peptides to a mixture of amino acids and thus circulation. This can occur either as a result of
it is not surprising that the oral intact absorption passive, active transport or membrane invagina-
of peptides is difficult (Verhoef et al., 1990). tion or intercellularly, in which molecules pass
Hydrolysis of peptides occurs at several sites: directly into the circulation via the junctions be-
luminally, at the brush border and intracellularly tween the epithelial cells (Wilson et al., 1989).
(Lee, 1988). Transport via the transcellular route requires that
One way to evaluate whether a compound is luminal mucus and stagnant water layers must be
degraded by luminal enzymes is via in vitro stud- passed and subsequently at least the apical mem-
ies with simulated gastric or intestinal fluids or brane, the cytoplasm and the basolateral mem-
using washing from the lumen (Aungst, 1993). brane.
Intestinal peptidases and their substrate specifici- In contrast, the paracellular route comprises
ties were recently reviewed by Bai and Amidon transport of drugs via the tight junctions and
(1992). subsequently the interstitial space. It is therefore
Different features of the enzymatic barrier in principle, a hydrophilic route and transport is
must be considered in order to understand the mainly limited by the size a n d / o r charge of the
poor bioavailability of peptides. First, unlike most tight junctions. In addition, this non-selective
small drug molecules that are metabolized in the route comprises less physical and metabolic barri-
liver primarily, peptidcs usually are susceptible to ers than the transcellular route and offers there-
degradation, in the blood and in the kidney (Lee fore interesting possibilities for peptide drug
et al., 1991). transport (Smith et al., 1992).
Secondly, a given peptide is susceptible to It has been suggested that transport via the
degradation at more than one linkage within the tight junctions is dynamically regulated in re-
backbone and each locus of hydrolysis is medi- sponse to physiological, pathological and experi-
ated by a certain peptidase. Substance P, for mental conditions (Cereijido et al., 1988). The
example, an undecapeptide is susceptible to different pathways for a peptide to pass the in-
degradation by at least five enzymes (Bunnett et testinal membrane have been reviewed by differ-
al., 1985). Modification of one linkage still leaves ent authors (Wilson et al., 1989; Friedman and
the rest of the peptide drug vulnerable to prote- Amidon, 1990; Jani et al., 1992).
olysis. However, peptides, even if pancreatic pro- Absorption is a dynamic process and many
teolysis (trypsin, a-chymotrypsin) is avoided, are factors affect it such as dissolution, ionization,
subject to the potential attack by brush border partitioning and blood flow. Poor membrane per-
peptidases and cytosolic peptidases of intestinal meation as one cause of quantified low oral
epithelial cells (Verhoef et al., 1990). There is a bioavailability can be evaluated in many ways
general tendency for proteins to be degraded by such as by measuring rates of intestinal mem-
130 J.M. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136

brane permeation in vitro and in situ (Aungst,

1993). The distinction between absorption and
PEPTIDE DRUGS
bioavailability should be noted. Achievement of
good absorption may not ultimately translate into
good bioavailability if the molecules are subject
to rapid metabolism or clearance as is the case Dosage form and Physiological factors [ Physico--chcmical factors
for the majority of peptides (Smith et al., 1992). formulation
I I
Artusson and Karlsson (1990) recently demon- Food, imestinal nlotdity Property ot pepud¢ :
strated a proportionality between in vitro perme- flow rate, transit time pKa, molecular size,
ability coefficients using a cell line derived from a Metabolism: lipophilicity,stability,
human colon carcinoma (Caco-2), and values for I,arn e n ,membrane surface solubility, flow rate,
mtracellularly aqueous boundary layer
the percentage of oral absorption in humans for
20 drugs and peptides. In general, it is thought Pallcreaticenzymes
pH intestinal, blood flow
that for simple drug molecules, a strong correla- IYIIICOIIS
tion exists between the octanol-water partition Absorptionmechanism:
coefficient and absorption (Ho et al., 1983). How- tra~scellularly,paracellularly

ever, this relationship does not seem to apply to

the transport of some peptides across Caco-2 cell
monolayers (Conradi et al., 1992). It has been /
shown that within a series of neutral peptides, /
flux through Caco-2 intestinal membranes in vitro
was correlated with the hydrogen bonding ten- Absorption

dency of the compound rather than with octanol-

water partitioning (Conradi et al., 1992). Burton
et al. (1993), have shown an apparent participa- Lymphaticpathway ( Blood pathway
tion of an apically polarized active resistance to without first pass metabolism First pass metabolism
peptide transport in Caco-2 ceils. Such active
polarized systems may serve as an additional bar-
rier to the non-specific uptake of peptides from
Systemic circulation
intestinal lumen.
With in situ methods, an estimation of absorp- Fig. 1. Mainfactorsinvolvedin peptide drug absorption in the
tion can be made by measuring drug disappear- intestine.
ance from the intestinal lumen and appearance in
blood (Swintosky and Pogonowska-Wala, 1982;
Schurgers et al., 1986). However, the results might Fig. 1 summarizes schematically the different
be different from the in vivo data because of elements to be considered in oral peptide
factors such as GI motility, bile flow and food bioavailability. What are the different strategies
intake (Aungst, 1993). to improve the oral bioavailability of peptides?
In a closed intestinal loop, the dose is exposed
to a limited portion of the intestine and so the
specific formulation is more concentrated than it 3. Different strategies to improve oral absorption
might be in vivo. It is important to measure drug of peptides
appearance in blood because if only in situ mea-
surements of luminal disappearance are evalu- One of the first strategies is chemical modifi-
ated, false interpretation of the data can be made cation to obtain more stabilized and more
if the drug binds in the membrane. Nevertheless, lipophilic drugs. Different examples exist in the
in situ methods can help in understanding the literature. A-72517, a dipeptide renin inhibitor, is
mechanisms of absorption. a structural relative of A-64662 (a first generation
 J.M. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136 131

renin inhibitor) that is intravenously efficacious reproducible delivery of a therapeutic peptide

and has been shown to lack oral bioavailability drug dose in the GI tract with minimal toxicity.
(Kleinert et al., 1992). Another strategy is based on the concept that
A-72517 is the more lipophilic compound with the co-administration of a peptide and an in-
a log P of 4.6 (in octanol-water, pH 7.4) as hibitor of its metabolism can reduce presystemic
compared with A-64662. This compound has oral metabolism resulting in increased fractional ab-
bioavailability of 8, 24, 32 and 53% in the mon- sorption. An example of the successful use of a
key, rat, ferret and dog, respectively (Kleinert et metabolism inhibitor is the increase in insulin
al., 1992). The P3-site residue of A-72517 imparts absorption in rats and dogs when co-administered
proteolytic stability and the sulfonamide moiety with a chymotrypsin (Fujii et al., 1985; Schilling
improves potency 10-fold as compared with A- and Mitra, 1991).
64662 (Buhlmayer et al., 1988). A-72517 is under The release of a specific peptide from a carrier
clinical evaluation in human subjects. system to a region of the GI tract where digestive
Improvement of lipophilicity is reported to be enzyme activity is low, such as colon and where
the mechanism by which esterification of the uptake into the lymphatics is maximized, would
dipeptide angiotensin-converting enzyme in- be most feasible for the peroral absorption of
hibitor enalaprilat to the monoester enalapril in- these drugs. However, the presence of bacteria
creases oral bioavailability (Wyvratt and Patchett, and fecal matter in the colon and their effects on
1985). the stability of peptides are not well documented
Another approach is dealing with the co-ad- (Kraeling and Ritschel, 1992). Soft gelatin cap-
ministration of absorption enhancers such as sules coated with azo-cross-linked co-polymers of
EDTA, salicylates, surfactants, bile salts, fatty styrene and hydroxyethyl methacrylate have been
acids, and unsaturated cyclic ureas. Swenson and used with insulin and vasopressin (Saffran et al.,
Curatolo (1992), have reviewed extensively the 1986).
different types of enhancers. Most of them are Further investigations need to be performed in
able to improve oral absorption but do so by this area. Van Den Mooter et al. (1993) have
creating disorders in the phospholipid domain in shown that the degradation of azo-polymers by
the membrane. Some can cause major tissue intestinal microflora can be influenced by the
damage (Swenson and Curatolo, 1992). type of polymers (their hydrophilicity).
Enhancers may act by transcellular or paracel- Another important approach to improve oral
lular pathways or both (Lee, 1990). Although they bioavailability of peptides is the formulation ap-
have been demonstrated to enhance drug uptake proach. This strategy has been examined exten-
(Nakada et al., 1988), their applications are not sively to render insulin more stable to proteolysis
without problems considering safety and mucosal in the GI tract. The formulations that have been
damage. For example, the co-carcinogenic and tested, for different peptide drugs, include lipo-
co-mutagenic effects of bile salts discourage the somes (Woodley, 1986; Chiang and Weiner, 1987).
development of bile salts containing pharmaceuti- Contradiction in results is a problem with lipo-
cal formulations (Breuer and Goebell, 1985). somes. This is due to the different techniques
However, some absorption enhancers are appar- used by different team researchers.
ently safe without causing major tissue damage Many references also exist on the use of emul-
(Hastewell et al., 1994). sions, nanocapsules or nanospheres (Couvreur
Today, the design and search for safe and and Puisieux, 1993), and microspheres to increase
effective penetration enhancers are becoming an the absorption of peptides following oral adminis-
active area of research in peptide drug delivery. tration. The combination of different approaches
The trend appears to be towards enhancers that may be interesting as a protease inhibitor com-
are based on natural body constituents as glyc- bined with microspheres (Morishita et al., 1992).
erides (Swenson and Curatolo, 1992). In fact, the However, none of the above formulation ap-
enhancer of choice will be one which enables the proaches improves oral insulin absorption suffi-
132 J.M. Sarciaux et al. /International Journal of Pharmaceutics 120 (1995) 127-136

ciently to be of therapeutic benefit. Another Table 1

strategy is to use microemulsions for oral drug Microemulsion NF 37, especially for soft or hard gelatin
capsules (Gattefosse S.A.)
delivery of peptides.
Raw materials Source Type % w/w
Labrafac lipo ® Gattefosse lipophilic phase 47
4. Characteristics and different uses of mi- Labrasol ® Gattefosse 'surfactant' 31.3
Plurol oleique ® Gattefosse 'co-surfactant' 13.26
croemulsions
Water hydrophilic phase 8.44

Having remained unknown for a considerable

time to the pharmaceutical field, microemulsions
can now be certainly considered as a very inter-
esting drug delivery approach for peptides. Mi- tensio-active properties instead of some known
croemulsions were first introduced by the British surfactants (Table 1).
chemist J.H. Schulman in 1943 (Hoar and Schul- Microemulsions can be used in different areas
man, 1943). A microemulsion is considered to be of applications: optimisation of blood substitutes,
a quaternary system. assisted recovery of petrol, reactivity in specific
Microemulsion systems contain a media (Rico, 1986) (Berthod, 1983) and also for
surfactant/co-surfactant blend which when added oral formulation of peptide drugs (Ritschel, 1991).
to a two-phase hydrophilic/lipophilic mixture, Among the physical measurements that are
form a stable, optically clear, isotropic, colloidal useful for identifying microemulsion systems are
system apparently thermodynamically stable with light scattering, sedimentation, centrifugation,
ultralow interracial tension (Borkovec, 1992). The rheology, conductivity and nuclear magnetic reso-
microstructure of these systems consists of nance (NMR), X-ray and electron microscopy.
micro-droplets of a size less than 100 nm Each measurement tells us something about the
(Bhargava et al., 1987). system (Rico, 1986).
A typical diagram of microemulsion is pre- The interest to potential users of microemul-
sented in Fig. 2. Polyglycolized fatty acids can be sions as novel oral drug delivery systems is moti-
incorporated in a microemulsion formulation with vated by different reasons: technologically, mi-
croemulsions are interesting because of their
spontaneous formation at room temperature,
considerable solubilizing properties for active
Water
I00~ principles, possible sterilization by filtration, and
high physical stability (Bhargava et al., 1987).
They are of physiological interest because they
can modify the active bioavailability of a specific
drug (Ritschel, 1991). They are thus expected to
offer several benefits for oral dosage forms such
as improved clinical potency and potentially de-
creased toxicity (Bhargava et al., 1987; Vergnault
and Metziger, 1992).
Through the appropriate selection of bioac-

Cosurfactant 100 80 60 40
\,oo
20 Oil
ceptable compounds, oral dosage forms can be
designed as in the use of soft gel capsules. An
example of a microemulsion designed especially
Fig. 2. A phase diagram for an emulsion containing water, oil, for hard and soft gelatin capsules is presented in
surfactant, and cosurfactant. The numbers along the edges
indicate the percentages of water, oil, and cosurfactant in the
Table 1. Based on different references (Tarr and
emulsion. The concentration of surfactant is held constant Yalkowsky, 1989), microemulsion formulation has
(Bhargava et al., 1987). been used with a specific immunosuppressant
 J.M. Sarciaux et aL / International Journal of Pharmaceutics 120 (1995) 127-136 133

peptide, cyclosporin. Due to its very low water inter- and intra-variabilities with ME formulation
solubility (Yee, 1991), the oral dosage form of yield a more consistent and predictable concen-
cyclosporin is an olive-oil based solution (Sandoz, tration-time profile (Kovarik et al., 1994).
1993). This was subsequently modified by using A study was conducted this time in 18 renal
maize oil in a soft gelatin capsule. The mean stable transplant patients at 6-26 months post-
absolute bioavailability of cyclosporin from these transplant to evaluate the safety and steady-state
formulations (oral solution and soft gelatin cap- pharmacokinetics of this new formulation of Cy-
sules) is approx. 30% but varies markedly as has closporin. Both formulations with mean doses of
been demonstrated in several transplant patient 139 + 27 mg (Sandimmune) vs 120 + 19 mg
populations (Ptachinski et al., 1986). (Neoral Sandimmune) are well tolerated (Mueller
Recently, a new oral formulation (Sandim- et al., 1993).
mune Neoral) was developed which incorporates All these results are very encouraging and
the drug in a microemulsion pre-concentrate con- explain the new interest in microemulsion formu-
taining a surfactant, lipophilic and hydrophilic lation.
solvents and ethanol. Other peptides such as SKF-106760, a fibrino-
A preliminary study showed that a specific gen receptor antagonist, have been formulated in
faster releasing microemulsion (ME) formulation microemulsions at pharmacologically relevant lev-
increased the extent of absorption on average by els. Intraduodenal bioavailabilities of SKF-106760
49% compared with the soft gelatin capsule in 12 from a solution (S) and the microemulsion (ME)
healthy male volunteers (Drewe et al., 1992). were investigated in the conscious rat. ME ad-
To investigate the influence of a fat-rich meal, ministration of SKF-106760 resulted in significant
on the pharmacokinetics of cyclosporin from this absorption enhancement compared with the solu-
new ME, a randomized, four-way crossover study tion (21.9+5.7 vs 0 . 5 + 0 . 3 ( n = 3 ) ) (Con-
was conducted in 24 healthy male volunteers. stantinides et al., 1993). Furthermore, there were
Simple oral doses of 300 mg Sandimmune and no signs of irritation upon macroscopic observa-
180 mg Sandimmune Neoral (the ME formula- tion of the intestinal tract of each rat at 4-6 h
tion) were each administered after starting a high post-dosing.
fat meal. Over a 48 h period, after each adminis- Based on different observations and results,
tration, whole blood cyclosporin concentrations Ritschel (1991) proposed a schematic diagram on
were determined by a specific monoclonal ra- the hypothesized mechanism of peptide absorp-
dioimmunoassay method. tion from microemulsions given orally. To sum-
The results demonstrated that the influence of marize, after by-pass of the aqueous boundary
food is less pronounced on the rate and extent of layer, mucins and epithelial cells, peptide drugs
cyclosporin absorption from Sandimmune Neoral can reach the systemic circulation via two ways:
and does not perturb the subsequent blood con- lymphatic vessels via the thoracic duct (where
centration profile (Mueller et al., 1994). In addi- first-pass metabolism is avoided) or via mesen-
tion, this microemulsion formulation can have a teric veins where a first-pass effect may occur
beneficial influence on compliance. (Ritschel, 1991).
The same doses as in the study before were According to different results, systemic pep-
given in a randomized, sequential bioreplication tide uptake from ME in the GI tract is dependent
study in the same number of patients to investi- on particle size, type of lipid phase of ME, di-
gate the inter- and intra-individual variability of gestibility of lipid used, type of emulsifying agents,
cyclosporin pharmacokinetics form Sandimmune pH and shedding of enterocyte cells. Only the
Neoral compared with Sandimmune (Kovarik et most lipophilic compounds (i.e, those with a log
al., 1994). octanol-water partition coefficient of more than
The results demonstrated that altering the oral 5) with good solubility in triglycerides are prefer-
formulation has a considerable and significant entially absorbed via the lymphatics (Myers and
impact on pharmacokinetic variability. Reduced Stella, 1992). This is largely due because the
134 J.M. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136

blood is much more efficient in clearing absorbed Borkovec, M., Phenomenological theories of globular mi-
compounds than is lymph. croemulsions. Adv. Colloid. Interface. Sci., 37 (1992) 195-
217
Therefore, targeting a specific lipophilic pep-
Breuer, N. and Goebell, H., The role of bile acids in colonic
tide drug via the lymphatic pathway may be inter- carcinogenesis. Klin. Wochenschr., 63 (1985) 97-105.
esting. Buhlmayer, P., Caselli, A., Fuhrer, W., Goschkler, R., Rasetti,
V., Rueger, H., Stanton, J.L., Criscione, L. and Wood,
J.M., Synthesis and biological activity of some transition-
state inhibitors of human renin. J. Med. Chem., 31 (1988)
5. Conclusion 1839.
Bunnett, N.W., Orloff, M.S. and Turner, A.J., Catabolism of
substance P in the stomach wall of the rat. Life. Sci., 37
The specific aim of this review article was to
(1985) 599-606.
summarize the difficulties in designing efficient Burton, P.S., Conradi, R.A., Hilgers, A.R. and Ho, N.F.H.,
oral dosage forms for peptide drugs and to pre- Evidence for a polarized effiux system for peptides in the
sent different strategies. This article emphasizes apical membrane of Caco-2 cells. Biochem. Biophys. Res.
the new oral formulation strategy, i.e., the use of Commun., 190 (1993) 760-766.
microemulsion formulations. This formulation Cereijido, M., Gonzlez-Mariscal, L., Avila, G., and Contreras,
R.G., Tight junctions. CRC Crit. Rev. Anat. Sci., 1 (1988)
approach can help to improve the oral bioavail- 171-192.
ability of peptides as is the case for cyclosporin. Chiang, C.M. and Weiner, N., Gastrointestinal uptake of
However, examination of different aspects such liposomes. In vivo studies. Int. J. Pharm., 40 (1987) 143-150.
as the toxicity, safety, and stability of ME needs Conradi, R.A., Hilgers, A.R., Ho, N.F.H. and Burton, P.S.,
The influence of peptide structure on transport across
to be undertaken on a specific peptide. The rela-
Caco-2 cells. Peptide bond modification which results in
tively high surfactant concentration of this improved permeability. Pharm. Res., 9 (1992) 435-439.
micro-emulsified system may result in localized Constantinides, P.P., Scalart, J.P., Lancaster, C., Marcello, J.,
cytotoxic effects. Nevertheless, the results ob- Marks, G., Ellens, H. and Smith, P., Water in oil mi-
tained with cyclosporin using this new oral drug croemulsions containing medium-chain glycerides: formu-
lation and absorption enhancement evaluation in the rat.
delivery system, are very encouraging for the fu-
Proc. Int. Symp. Controlled Release Bioact. Mater., 20 (1993)
ture of oral efficient peptide drugs. 184-185.
Couvreur, P. and Puisieux, F., Nano and microparticles for
the delivery of polypeptides and proteins. Adv. Drug. Del.
Rev., 10 (1993) 141-162.
References Devissaguet, J.P., Evaluation de la biodisponibilit6 des m~di-
caments. In Aiache, J.M., Devissaguet, J.P., Guyot-Her-
Artusson, P. and Karlsson, J., Correlation between oral drug mann, A.M. (Eds), Galenica 2, Biopharmacie, Technique
absorption in humans and apparent drug permeability et Documentation, Paris, 1982, pp. 101-146.
coefficients in human intestinal epithelial (Caco-2) cells. Drewe, J., Meier, R., Vonderscher, J., Kiss, D., Posanski, U.,
Biochim. Biophys. Res. Commun., 175 (1990) 880-885. Kissel, T. and Gyr, K., Enhancement of the oral absorp-
Aungst, B.J., Novel formulation strategies for improving oral tion of cyclosporin in man. Br. J. Clin. Pharmacol., 34
bioavailability of drugs with poor membrane permeation (1992) 60-64.
or presystemic metabolism. J. Pharm. Sci., 82 (1993) 979- Friedman, D.I. and Amidon, G.L., Characterization of the
987. intestinal transport parameters for small peptide drugs. J.
Bai, J.P.F. and Amidon, G.L., Degradation of insulin by Controlled Release, (1990) 141-146.
trypsin and alpha-chymotrypsin. Pharm. Res., 9 (1992) Fujii, S., Yokohama, T., Ikegaya, K., Sato, F. and Yohoo, N.,
969-978. Promoting effect of the new chymotrypsin inhibitor FK-448
Banga, A.K. and Chien, Y.W., Systemic delivery of therapeu- on the intestinal absorption of insulin in rats and dogs. J.
tic peptides and proteins. Int. J. Pharm., 48 (1988) 15-50. Pharm. Pharmacol., 37 (1985) 545-549.
Berthod, A., Contribution ~ l'6tude physico-chimique des Hastewell, J., Lynch, S., Fox, R., Williamson, I., Skelton-
milieux disperses. Cas d'une 6mulsion de perfluorotributy- Stroud, P. and Mackay, M., Enhancement of human calci-
lamine utilisable comme substitut du sang. Th~se de doc- tonin absorption across the rat colon in vivo. Int. J. Pharm.,
torat d'6tat es Sciences, Lyon I, 1983, pp. 13-63. 101 (1994) 115-120.
Bhargava, H.N., Narurkar, A. and Lieb, L.M., Using mi- Ho, N.F.H., Park, J.Y., Ni, P.F. and Higuchi, W.I., Advancing
croemulsions for drug delivery. Pharm. Technol., 11 (1987) quantitative and mechanistic approaches in interfacing
46-54. gastrointestinal drug absorption studies in animals and
 ZM. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136 135

humans. In Crouthamel, W. and Sarapu, A.C. (Eds), Ani- human calcitonin in rats. Z Pharmacobio. Dyn., 11 (1988)
mal Models for Oral Drug Delivery in Man: In Situ and In 395 -401.
ViL,o Approaches, American Pharmaceutical Association, Ptachinski, R.J., Venkataramanan, R. and Burckart, G.J.,
Washington, DC, 1983, pp. 27-105. Clinical pharmacokinetics of cyclosporin. Clin. Pharma-
Hoar, T.P. and Schulman, J.H., Transparent water in oil cokinet., 11 11986) 107-132.
dispersions: the oleopathic hydro-micelle. Nature, 152 Rico, I., Les micro~mulsions: d~finition et applications pra-
(1943) 102-103. tiques. Les Entretiens de Carla, VII (1986) 77-92.
Jani, P.U., McCarthy, D.E. and Florence, A.T., Nanosphere Ritschel, W.A., Microemulsions for improved peptide absorp-
and microsphere uptake via Peyer's patches: observation tion from the gastrointestinal tract. Methods Find. Exp.
of the rate of uptake in the rat after a single oral dose. Int. Clin. Pharmacol., 13 (1991) 205-220.
J. Pharm., 86 (1992) 239-246. Saffran, M., Kumar, G.S., Savariar, C., Burnham, J.C.,
Kahan, B.D., Cyclosporine. N. Engl. J. Med., 321 11989) 1725- Williams, F. and Neckers, D.C., A new approach to the
1737. oral delivery of insulin and other peptide drugs. Science,
Kleinert, H.D., Rosenberg, S.H., Baker, W.R., Stein, H.H., 223 (1986) 1081-1084.
Klinghofer, V., Barlow, J., Spina, K., Polakowski, J., Ko- Samanen, J., Biomedical polypeptides. A wellspring of phar-
var, P., Cohen, J. and Denissen, J., Discovery of a maceuticals. In Gebelein, C.G. and Carraher, C.E. (Eds),
peptide-based renin inhibitor with oral bioavailability and BioactiL,e Polymeric Systems - An OL'ert~iew, Plenum, New
efficacy. Science, 257 (1992) 1940-1943. York, 1985, pp. 279-344.
Kraeling, M.E.K. and Ritschel, W.A., Development of a Samanen, J., Physical biochemistry of peptide drugs: struc-
colonic release capsule dosage form and the absorption of ture, properties and stabilities of peptides compared with
insulin. Methods Find. Exp. Clin. Pharmacol., 14 (1992) proteins. In Lee, V.H.L. (Ed), Peptide and Protein Drug
199-209. DeliL~ery, Dekker, New York, 1991, pp. 137-166.
Kovarik, J.M., Mueller, E.A., Van Bree, J.B., Tetzloff, W. and Sandoz, Sandimmun ' ~ ciclosporine. Dossier d'information
Kutz, K., Reduced inter and intraindividual variability in technique. IV~me ~dition, Rueil-Malmaison, France
cyclosporine pharmacokinetics from a microemulsion for- (1993).
mulation. J. Pharm. Sci., 83 (1994) 444-446. Schilling, R.J. and Mitra, A.K., Degradation of insulin by
Lee, V.H.L., Enzymatic barriers to peptide and protein ab- trypsin and alpha-chymotrypsin. Pharm. Res., 8 (1991)
sorption. Crit. Retd. Ther. Drug Carrier Syst., 5 (1988) 69-97. 721-727.
Lee, V.H.L., Peptide and protein drug delivery: opportunities Schurgers, N., Bijdendijk, J., Tukker, J.J. and Crommelin,
and challenges. Pharm. Int., (1986) 208-212. D.J.A., Comparison of four experimental techniques for
Lee, V.H.L., Protease inhibitors and penetration enhancers as studying drug absorption kinetics in the anesthetized rat in
approaches to modify peptide absorption. J. Controlled situ. J. Pharm. Sci., 75 (1986) 117-119.
Release, 13 (1990) 213-223. Smith, P.L., Wall, D.A., Gochoco, C.H. and Wilson, G.,
Lee, V.H.L., Dodda-Kashi, S., Grass, G.M. and Rubas, W., Routes of delivery: case studies oral absorption of pep-
Oral route of peptide and protein drug delivery. In Lee, tides and proteins. AdL,. Drug Del. ReLy., 8 11992) 253-290.
V.H.L. (Ed.), Protein and Peptide Drug Deliuery, Dekker, Swenson, E.S. and Curatolo, W.J., Means to enhance penetra-
New York, 1991, Ch. 16, pp. 691-738. tion intestinal permeability enhancement for proteins,
Morishita, M., Morishita, I., Takayama, K., Machida, Y., and peptides and other polar drugs: mechanisms and potential
Nagai, T., Novel oral microspberes of insulin with protease toxicity. AdL~. Drug Del. ReL,., 8 (1992) 39-92.
inhibitor protecting from enzymatic degradation, Int. J. Swintosky, J.V. and Pogonowska-Wala, E., The in-situ rat gut
Pharm., 78 11992) 1-7. technique. Pharm. Int., (1982) 163-167.
Mueller, E.A., Kovarik, J.M., Van Bree, J.B., Grevel, J., Tart, B.D. and Yalkowsky, S.H., Enhanced intestinal absorp-
Lucker, P.W. and Kutz, K., Influence of a fat-rich meal on tion of cyclosporine in rats through the reduction of emul-
the pharmacokinetics of a new oral formulation of cy- sion droplet size. Pharm Res., 6 (1989) 40-43.
closporine in a crossover comparison with the market Van Den Mooter, G., Samyn, C. and Kinget, R., Azopolymers
formulation. Pharm. Res., 11 (1994) 151-155. for colon-specific drug delivery. Influence of the type of
MueUer, E., Kovarik, J., Van Bree, J., Lison, A. and Kutz, K., azo polymer on the degradation by intestinal microflora.
Safety and steady-state pharmacokinetics of a new oral Int. J. Pharm., 97 (1993) 133-139.
formulation of cyclosporin A in renal transplant patients. Van Hoogdalem, E.J., De Boer, A.G. and Breimer, D.D.,
6th Congress European Society For Organ Transplantation, Intestinal drug absorption enhancement: an overview.
Rhodos, Greece, 1993, p. 131. Pharmac Ther., 44 (1989) 407-443.
Myers, R.A. and Stella, V.J., Factors affecting the lymphatic Vergnault, G. and Metziger, P., Oral anhydrous non-ionic
transport of penclonedine (NSC-338720), lipophilic cyto- microemulsions administered in softgel capsules. AAPS,
toxic drug: comparison to DDT and hexachlorobenzene. San Antonio, TX, 1992.
Int. J. Pharm., 80 (1992) 51-62. Verhoef, J.C., Bodd6, H.E., De Boer, A.G., Bouwstra, J.A.,
Nakada, Y., Awata, N., Nakamichi, C. and Sugimoto, I., The Junginger, H.E., Merkus, F.W.H.M. and Breimer, D.D.,
effects of additives on the oral mucosal absorption of Transport of peptide and protein drugs across biological
136 J.M. Sarciaux et al. / International Journal of Pharmaceutics 120 (1995) 127-136

membranes. Eur. J. Drug Metab. Pharmacokinet., 15(2) Woodley, J.F., Liposomes for oral administration of drugs.
(1990) 83-93. Crit. Rev. Drug Carrier Syst., 2 (1986) 1-18.
Wieland, T. and Bodanszky, M., The World of Peptides, a Brief Wyvratt, M.J. and Patchett, A.A., Recent development in the
History of Peptide Chemistry, Springer, Berlin, 1991, Ch. 1, design of Angiotensin-converting enzyme inhibitors. Med.
pp. 1-21. Res. Rev., 5 (1985) 483-531.
Wilson, C.G., Washington, C. and Washington, N., Overview Yee, G.C., Dosage forms of cyclosporine. Pharmacotherapy,
of epithelial barriers and drug transport. In Wilson, C.G. 11 (1991) 149S-152S.
and Washington, N. (Eds), Physiological Pharmaceutics.
Biological Barriers to Drug Absorption, Wiley, New York,
1989, pp. 11-20.
View pictures
View other presentations
Microemulsions as drug delivery systems

Olivier Midler
Marketing Director
Gattefossé group

The concept of microemulsion was first introduced by Hoar and Schulman in 1943 (1) ; they
prepared the first microemulsions by dispersing oil in an aqueous surfactant solution and adding
an alcohol as a co-surfactant, leading to a transparent, stable formulation.
The existence of this theoretical structure was later confirmed by use of various technologies, and
we can today adopt the definition given by Attwood (2) : “a microemulsion is a system of water,
oil, and amphiphilic compounds (surfactant and co-surfactant) which is a transparent, single
optically isotropic, and thermodynamically stable liquid”.
The main difference between emulsions and microemulsions lies in the size — and shape — of
the particles dispersed in the continuous phase : these are at least an order of magnitude smaller
in the case of microemulsions ( 10 – 200 nm) than those of conventional emulsions (1 – 20 µm).
Also, whereas emulsions consist of roughly spherical droplets of one phase dispersed into the
other, microemulsions constantly evolve between various structures ranging from droplet-like
swollen micelles to bicontinuous structures, making the usual “oil in water” and “water in oil”
distinction sometimes irrelevant (3).
Microemulsions are formed when and only when (i) the interfacial tension at the oil/water
interface is brought to a very low level and (ii) the interfacial layer is kept highly flexible and
fluid. These two conditions are usually met by a careful and precise choice of the components
and of their respective proportions, and by the use of a “co-surfactant” which brings flexibility to
the oil/water interface.
These conditions lead to a thermodynamically optimised structure, which is stable — as opposed
to conventional emulsions — and does not require high input of energy (i.e. through agitation) to
be formed.
Because the size of the particles is much smaller than the wavelength of visible light,
microemulsions are transparent and their structure cannot be observed through an optical
microscope.
Microemulsions have been used in a wide range of industries, starting with enhanced oil recovery
in the 70’s, expanding to a wide range of chemicals and entering the pharmaceutical and cosmetic
formulation area a decade ago.
Because of their unique solubilization properties, microemulsions have attracted increasing
attention as potential drug delivery systems, either as vehicles for topical applications or as
bioavailability enhancers for poorly water soluble active pharmaceutical ingredients (API).

In topical formulations, microemulsions have been proved to increase the cutaneous absorption of
both lipophilic and hydrophilic API’s when compared to conventional vehicles (emulsions, pure
oils, aqueous solutions, etc.). In an extensive review of this type of applications, Kreilgaard et. al.
(3) attribute this performance to a generally higher solubility of the API’s in microemulsions,
generating an increased concentration gradient towards the skin. The role of penetration
enhancers played by the amphiphilic components of the microemulsion and the internal mobility
of the drug within the vehicle also contribute to the overall performance of microemulsions in
dermal or transdermal drug delivery.
Many topical formulations currently marketed are based on microemulsions:
View pictures
View other presentations
The dog shampoo "Allermyl" marketed by Virbac is, to our knowledge, the first application of
microemulsions to a therapeutic cleansing product ; the use of microemulsion allowed the
addition of a higher content of essential fatty acids to the formula.
Solvium is a topical Ibuprofen gel marketed by Chefaro (Akzo). In this case, microemulsion has
been used to formulate a poorly soluble active at a dose of 5% into a perfectly transparent gel.
But the most promising potential in microemulsions comes from the advantage they bring to oral
route formulations.
Microemulsions enhance the bioavailability of poorly soluble drugs by maintaining them in
molecular dispersion in the GI tract and extending the absorption window available in the GI
lumen.
This lead to a faster absorption allowing a more rapid onset of drug action. But microemulsions
also overcome food effect and reduce subject to subject variability by levelling the differences in
digestive capabilities.
Many examples of microemulsion based formulations are now on the market ; Among them, the
performances of microemulsions are well demonstrated in the reformulation of Cyclosporin A by
Novartis into a microemulsion based formulation marketed under the trade mark Neoral® : this
has increased the bioavailability nearly by a factor 2. In addition, Neoral shows a much faster
onset of action than the earlier version Sandimmune®, a reduced inter/intra-subject variability
and a much lower impact of food intake on cyclosporin pharmacokinetics.
Further promising development have been brought by Self Micro Emulsifying Drug Delivery
Systems (SMEDDS .), patented by Gattefossé in the 90’s (5): these are “latent” microemulsions
in the form of a stable, water-free combination of surfactants, co-surfactants and lipophilic phase,
which creates a microemulsion when diluted in water or body fluids. Such systems combine the
advantages of microemulsions with a water-free formulation protecting sensitive API from the
chemical degradation they would undergo in an aqueous medium.
The performance of SMEDDS has been demonstrated by Gattefossé on Simvastatin, a well
known anti-hyperlipidemic drug marketed by Merck (Zocor), which is subject to extensive
hepatic first pass metabolism. The use of a SMEDDS formulation gives a spectacular (four fold)
increase in bioavailability in dogs (6).
Microemulsions are powerful formulation tools for poorly soluble API’s, both for the oral and
topical administration routes. The availability of efficient, non toxic surfactants and co-

surfactant now makes them a very attractive and feasible option to overcome the bioavailability
problems frequently encountered in the development of modern drugs.
1. Hoar T.P. and Schulman J. H. “Transparent water in oil dispersions: the oleopathic
hydromicelle”. Nature 152: 102 – 103, 1943.
2. Attwood D. Microemulsions in Colloidal drug delivery systems (J. Kreuter ed.), Marcel
Dekker, New York 1994.
3. Kreilgaard M. “Influence of microemulsions on cutaneous drug delivery”. in Bulletin
Technique Gattefossé N°95 : 79 – 100, 2002
4. Meinzer A., Mueller E. and Vonderscher J. in Bulletin technique Gattefossé N°88 : 21 – 26
5. Farah N. and Denis J. “Orally administrable composition capable of forming a
microemulsion”, French Patent FR-B-2 710 535
6. Benameur H., Jannin V., and Roulot D. "Composition pharmaceutique à usage oral
comprenant un principe actif susceptible de subir un important effet de premier passage
intestinal", French patent application N° FR-A-2 827 770.
 International Journal of Pharmaceutics 436 (2012) 536–544

Contents lists available at SciVerse ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsevier.com/locate/ijpharm

Pharmaceutical Nanotechnology

A propofol microemulsion with low free propofol in the aqueous phase:

Formulation, physicochemical characterization, stability and pharmacokinetics
WeiHui Cai a , WanDing Deng a , HuiHui Yang a , XiaoPing Chen b , Fang Jin a,∗
a
Shanghai Engineering Research Center of Respiratory Medicine, Shanghai Institute of Pharmaceutical Industry, Shanghai 200437, PR China
b
Beijing Shiqiao Biopharmaceutical Co. Ltd., Beijing 101301, PR China

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of this study was to develop a propofol microemulsion with a low concentration of free
Received 11 April 2012 propofol in the aqueous phase. Propofol microemulsions were prepared based on single-factor exper-
Received in revised form 19 June 2012 iments and orthogonal design. The optimal microemulsion was evaluated for pH, osmolarity, particle
Accepted 4 July 2012
size, zeta potential, morphology, free propofol in the aqueous phase, stability, and pharmacokinetics in
Available online 17 July 2012
beagle dogs, and comparisons made with the commercial emulsion, Diprivan® . The pH and osmolarity
of the microemulsion were similar to those of Diprivan® . The average particle size was 22.6 ± 0.2 nm,
Keywords:
and TEM imaging indicated that the microemulsion particles were spherical in appearance. The concen-
Propofol
Microemulsion
tration of free propofol in the microemulsion was 21.3% lower than that of Diprivan® . Storage stability
Emulsion tests suggested that the microemulsion was stable long-term under room temperature conditions. The
Free propofol pharmacokinetic profile for the microemulsion showed rapid distribution and elimination compared to
Diprivan® . We conclude that the prepared microemulsion may be clinically useful as a potential carrier
for propofol delivery.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction soybean oil (100 mg/mL), glycerol (22.5 mg/mL), and purified egg
lecithin (12 mg/mL), and this was launched in the United Kingdom
Propofol (2,6-diisopropylphenol) is a short-acting hypnotic and New Zealand in 1986, and in the United States in November
agent that is commonly used for both induction and maintenance 1989 (Thompson and Goodale, 2000). Later, ethylenediaminete-
of anesthesia. It offers various advantages over other agents, such traacetic acid and sodium metabisulfite were added as antibacterial
as rapid onset, short duration of action, and excellent quality of agents (Han et al., 2001; Langevin, 1999; Thompson and Goodale,
recovery (Gupta et al., 2004; Langley and Heel, 1988). Propofol is 2000).
a highly lipophilic liquid, with a partition coefficient (log P) of 3.8 In spite of the success of the current formulation, several
and a water solubility of 154 ␮g/mL (Babu and Godiwala, 2004). drawbacks have been reported for this lipid-based emulsion,
This high lipophilicity allows propofol to easily permeate the blood including emulsion instability (Park et al., 2003), pain upon injec-
brain barrier, resulting in a rapid anesthetic effect. Furthermore, tion (Sundarathiti et al., 2007), hyperlipidemia (Wolf et al., 2001)
propofol has a low tendency to accumulate in the body due to a and rapid growth of microorganisms (Wachowski et al., 1999). In
high clearance rate and short elimination half-life (Langley and order to overcome the limitations of lipid-based emulsions, vari-
Heel, 1988). Despite such attractive properties as an anesthetic, the ous approaches have been exploited by researchers, including use
development of an injectable formulation of propofol has been a of microemulsions (Ryoo et al., 2005) and prodrugs (Pergolizzi et al.,
particular challenge due to its poor water solubility. 2011), and formation of complexes with cyclodextrins (Trapani
Initial clinical trials with propofol were conducted in Europe et al., 2004). Fospropofol (2,6-diisopropylphenoxymethyl phos-
in 1977 using a 1% preparation formulated in Cremophor EL, a phate, disodium salt), a novel prodrug of propofol, has recently been
polyethoxylated castor oil (Baker and Naguib, 2005). Due to the approved in the United States, and has been associated with less
high incidence of anaphylaxis observed with the Cremophor EL pain at the injection site than propofol. However, the pharmacoki-
formulation, an oil-in-water emulsion (Diprivan® , Astra-Zeneca, netic characteristics of propofol liberated from fospropofol differ
London, UK) was developed, containing propofol (10 mg/mL), from that of propofol emulsion, with lower peak concentrations
and more prolonged elimination half-life (Pergolizzi et al., 2011).
For this reason, fospropofol has a slow onset of action compared to
∗ Corresponding author. Tel.: +86 21 55514600 143; fax: +86 21 55514600 425. propofol, such that more time may be required to achieve adequate
E-mail address: jfsipi@163.com (F. Jin). sedation. A cyclodextrin-based formulation of propofol (sulfobutyl

0378-5173/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ijpharm.2012.07.008
 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544 537

ether-␤-cyclodextrin; Captisol® ) has been developed, but adverse Ltd. (Changsha, Hunan, China). Calcium disodium edetate (EDTA)
hemodynamic effects associated with this formulation have been was provided by Sinopharm Chemical Reagent Co. Ltd. (Shanghai,
observed (Egan et al., 2003). China). Methanol (HPLC-grade) was purchased from Dikma Tech-
Microemulsions have gained great attention as drug delivery nologies Inc. (Lake Forest, CA, USA). Diprivan® was supplied by
systems in pharmaceutical research. These offers several potential Astra-Zeneca Ltd. (London, UK). Distilled water, made in house, was
advantages over other techniques, such as thermodynamic stabil- used for preparation of all solutions and dilutions. Dialysis mem-
ity, ability to solubilize hydrophobic drugs and ease of manufacture brane was purchased from Greenbird (Shanghai, China). All other
and scale-up (Gupta and Moulik, 2008). A microemulsion-based reagents were of analytic grade.
propofol, AquafolTM (Daewon Pharmaceutical Co. Ltd., Seoul,
Republic of Korea), has been developed in an attempt to eliminate 2.2. Formulation optimization
the adverse events of propofol emulsion. This was initially formu-
lated with 8% polyethylene glycol 660 hydroxystearate (Solutol® Surfactants and oil known to have parenteral acceptability were
HS 15) and 5% tetrahydrofurfuryl alcohol polyethylene glycol ether. screened as an injectable microemulsion-based drug delivery sys-
Although the clinical characteristics at induction and recovery did tem for propofol. Solutol® HS 15, egg lecithin and glycocholic acid
not show significant differences between AquafolTM (microemul- sodium salt were chosen as emulsifier and co-emulsifiers, respec-
sion) and Diprivan® (emulsion), a significantly higher incidence tively. MCT, glycerol and EDTA were used as the oil phase, osmotic
and greater severity of pain on injection were observed with pressure regulator and antioxidant, respectively.
microemulsion propofol (Sim et al., 2009). The concentration of Propofol microemulsion was prepared using the following
free propofol in the aqueous phase is known to be associated with method. Propofol, Solutol® HS 15 and egg lecithin were dissolved
the intensity of pain on injection (Ohmizo et al., 2005), and indeed in MCT (the oil phase) at 60–70 ◦ C. Glycocholic acid, glycerol and
the measured concentration of free propofol in the aqueous phase EDTA were dispersed in water as the water phase, and the pH value
was found to be seven times greater for microemulsion propofol of this phase was adjusted to 7.0–8.0, using sodium hydroxide solu-
than for lipid emulsion propofol (Sim et al., 2009). Subsequently, tion (0.1 mol/L). A microemulsion solution was then prepared by
AquafolTM was reformulated with 10% purified poloxamer 188 and mixing the two phases, using constant stirring (500 rpm) for about
0.7% Solutol® HS 15. The reformulated AquafolTM was approved in 30 min at 60–70 ◦ C. Finally, the microemulsion was sterilized by
Korea and found to be as effective and safe as Diprivan® , but the autoclaving at 121 ◦ C for 15 min.
incidence of pain on injection with the microemulsion remained The influence of concentrations of Solutol® HS 15, MCT, propo-
significantly higher than that with emulsion propofol, due to a fol, egg lecithin and glycocholic acid (w/v) on particle size and
higher concentration of free propofol in the aqueous phase (Jung hemolysis were studied (the compositions of the microemulsions
et al., 2010; Lee et al., 2008). This highlighted the urgent need to are shown in Table 1). An L9 (34 ) orthogonal design was established
develop a microemulsion with a lower free concentration of propo- to optimize the formulations, as shown in Table 2, with concentra-
fol. tion of propofol setting as 1%, using hemolysis as an index.
In one of our previous studies, we examined the concentra- The particle sizes of the emulsions were analyzed with a
tions of free propofol in the aqueous phase for formulations made dynamic light scattering (DLS) technique, using a Malvern laser
using lipid emulsions, Cremophor EL micelles, microemulsions and particle size analyzer (Nano-ZS 90, Malvern, UK). The hemolysis
mixed micelles, and compared the in vitro hemolytic activities of was determined in vitro. Briefly, erythrocytes were isolated from
these formulations, using vehicles without drug as controls. Unex- rabbit fresh whole blood, separated by centrifugation at 3000 rpm
pectedly, it was found that the degree of hemolysis increased with for 5 min, and washed three times with ten volumes of normal
the concentration of free propofol (Cai and Jin, 2008), suggest- saline solution. Then, 2 mL of this erythrocyte solution was resus-
ing that hemolysis was induced by free propofol in the aqueous pended in 100 mL normal saline (2%). Immediately following this,
phase. This raised the possibility that in vitro hemolytic activity 0.3 mL volumes of microemulsion were added to tubes, along with
might be exploited as a useful index to allow optimization of intra- 2.5 mL volumes of the erythrocyte dispersion, and normal saline
venous propofol formulations. We hypothesized that formulations was added to the tubes to obtain a final volume of 5 mL. Positive and
with low in vitro hemolytic activity were likely to have a low free negative controls were prepared by the addition of water (2.5 mL)
propofol concentration in the aqueous phase. and normal saline (2.5 mL) to 2.5 mL samples of the erythrocyte
The aims of this study were to develop new propofol microemul- dispersion. Following incubation at 37 ◦ C for 3 h, the samples were
sions with a low concentration of free propofol in the aqueous centrifuged at 3000 rpm for 5 min, and 0.4 mL of the resulting super-
phase. The formulations were screened for the degree of hemol- natant was diluted to 5 mL with an ethanol/HCl mixture (39 parts
ysis (%) and particle size. The optimal propofol microemulsion was of 99% ethanol (v/v) and 1 part of 37% hydrochloric acid (w/v)). The
evaluated for pH, osmolarity, particle size, zeta potential, morphol- amount of released hemoglobin was measured spectrophotometri-
ogy, and free propofol in the aqueous phase. In addition, a stability cally at a wavelength of 398 nm (UV 240 Spectrometer, Shimadzu,
study was carried out under both room temperature and acceler- Japan), as previously described (Jumaa and Muller, 2000). The
ated conditions, and the pharmacokinetic properties of propofol hemolysis values were calculated using the following equation:
microemulsions and emulsions were compared in beagle dogs.
As − A0
Hemolysis (%) = × 100 (1)
A100 − A0
2. Materials and methods
where As was the absorbance of the sample, A0 was the absorbance
2.1. Materials of the negative control and A100 was the absorbance of the positive
control.
Propofol and glycocholic acid were kindly provided by Beijing
Shiqiao Biopharmaceutical Co. Ltd. (Beijing, China). Polyethylene 2.3. pH and osmolarity
glycol 660 hydroxystearate (Solutol® HS 15) was purchased from
BASF (Berlin, Germany). Egg lecithin and medium chain triglyc- The pH and osmolarity of the optimized microemulsions and
eride (MCT) were supplied by Lipoid (Ludwigshafen, Germany). propofol emulsion (1% Diprivan® ) were measured using a pH meter
Glycerol was purchased from Hunan Erkang Pharmaceutical Co. (PHS-25, Shanghai Precision & Scientific Instrument Co. Ltd., China)
538 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544

Table 1
Composition (%, w/v) of the propofol microemulsions (n = 3).

Factor Level Propofol Solutol® HS 15 MCT Egg lecithin Glycocholic acid Glycerol EDTA
®
Solutol HS 15 1 1 5 5 0 0 2 0.005
2 1 7 5 0 0 2 0.005
3 1 9 5 0 0 2 0.005
4 1 11 5 0 0 2 0.005
5 1 13 5 0 0 2 0.005
MCT 1 1 9 0 0 0 2 0.005
2 1 9 1 0 0 2 0.005
3 1 9 3 0 0 2 0.005
4 1 9 5 0 0 2 0.005
5 1 9 7 0 0 2 0.005
Propofol 1 0 9 5 0 0 2 0.005
2 1 9 5 0 0 2 0.005
3 2 9 5 0 0 2 0.005
4 3 9 5 0 0 2 0.005
5 4 9 5 0 0 2 0.005
Egg lecithin 1 1 9 5 0 0 2 0.005
2 1 9 5 1 0 2 0.005
3 1 9 5 1.5 0 2 0.005
4 1 9 5 2 0 2 0.005
5 1 9 5 2.5 0 2 0.005
Glycocholic acid 1 1 9 5 2 0 2 0.005
2 1 9 5 2 0.25 2 0.005
3 1 9 5 2 0.5 2 0.005
4 1 9 5 2 1 2 0.005
5 1 9 5 2 1.5 2 0.005

and osmometer (Model 3250, Advanced Instruments Inc., USA), the ends, bound, and placed in 10 mL microemulsion or lipid emul-
respectively. The measurements were carried out in triplicate. sion for 24 h at 25 ± 1 ◦ C. After removal from the microemulsion or
emulsion, the tubes were dabbed, and their contents taken up by a
2.4. Particle size and zeta potential syringe in order to transfer them (within 10 s) into glass vials for use
by an auto-injector. Concentrations of free propofol in the aqueous
The particle sizes and zeta potentials of propofol microemul- phase were measured by high-performance liquid chromatogra-
sion and emulsion (1% Diprivan® ) were determined using a DLS phy (Shimadzu LC-10 AD; Shimadzu Corporation, Japan) using a
technique with a Malvern laser particle size analyzer (Nano-ZS 90, C18 column (Kromasil C18, 150 mm × 4.6 mm, 5 ␮m; Eka Chemi-
Malvern, UK), after equilibration for 2 min. The cell temperature cals, Bohus, Sweden), using a methanol–water–acetic acid mixture
used was 25 ◦ C; the scattering angle was 90◦ . Particle size data were (85:14:1) as the mobile phase. The flow rate was 1.0 mL/min, the
reported as the intensity-weighted distribution, and results were injection volume was 20 ␮L, and components of the column efflu-
the mean of three independent test runs. ent were monitored using an ultraviolet detector at a wavelength
of 265 nm.
2.5. Morphology
2.7. Storage stability
Morphology was observed with transmission electron
microscopy (TEM) using a JEM-1230 electron microscope (JEOL The physicochemical stability of the propofol microemulsion
Ltd., Japan). Negative staining of samples was performed as fol- was characterized both in long-term room and accelerated con-
lows: one drop of sample solution was placed onto a copper grid ditions of (shown in Table 3). The stability parameters, including
coated with carbon; the sample droplet was tapped with filter appearance, pH, particle size, and concentrations of drug and
paper to remove surface water and air-dried for 5 min, and this related substances, were determined as a function of storage time.
was followed by application of 2% phosphotungstic acid to deposit The pH value and particle size were determined as described
nanoparticles on the grid. The samples were air-dried before above. The concentrations of drug and of related substances were
observations were made. measured by high-performance liquid chromatography with the
chromatographic conditions described above. Test solution and ref-
2.6. Concentration of free propofol erence solution containing 0.1 mg/mL propfol were prepared in
methanol for assay. The propofol assay value was calculated as
The propofol microemulsion and emulsion (1% Diprivan® ) were follows:
dialyzed using a dialysis membrane, with a cutoff molecular weight Ai × Cs
of approximately 3500–4000 Da (Yamakage et al., 2005). Dialysis Assay (%) = × 100 (2)
As
tubes, filled with a solution of 2.5% (w/v) glycerol, were bent at
where Ai and As were the peak areas of test solution and reference
solution, respectively, and Cs was the concentration of the reference
Table 2
Factors and levels of orthogonal design (%, w/v). solution.

Factor A B C D
Table 3
Level 1 5 1 1.5 0.25 Storage conditions for the microemulsions (RH, relative humidity).
Level 2 9 3 2.0 0.5
Level 3 13 5 2.5 1.0 Group Storage conditions Time points

®
A: concentration of Solutol HS 15; B: concentration of MCT; C: concentration of Long-term 25 ± 2 ◦ C/60% ± 5% RH 0, 3, 6, 9, 12 months
egg lecithin; D: concentration of glycocholic acid. Accelerated 40 ± 2 ◦ C/75% ± 5% RH 0, 1, 2, 3, 6 months
 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544 539

Test solution (1 mg/mL propofol), reference solution Table 4

Results of orthogonal design.
(0.01 mg/mL propofol), European Pharmacopoeia chemical
reference substance solution of propofol impurity J (5 ␮g/mL), A B C D Hemolysis (%)
and impurity E (5 ␮g/mL) were prepared in methanol for the 1 5 1 1.5 0.25 61.2
measurement of related substances. The concentrations of total 2 5 3 2.0 0.5 10.3
impurities, impurity J, and impurity E were determined. 3 5 5 2.5 1.0 14.7
4 9 5 1.5 0.5 32.8
5 9 1 2.0 1.0 71.1
2.8. Pharmacokinetics study in beagle dogs 6 9 3 2.5 0.25 4.8
7 13 3 1.5 1.0 53.2
8 13 5 2.0 0.25 14.3
An in vivo pharmacokinetic study was performed in 6 bea- 9 13 1 2.5 0.5 31.4
gle dogs (3 male, 3 female, 9–10 kg), supplied by the Shanghai
Slac Laboratory Animal Company (China). The dogs used for this K1 86.2 163.7 147.2 80.3
study were housed individually under standard conditions (12 h K2 108.7 68.3 95.7 74.5
day/night cycle, temperature 21–23 ◦ C, humidity 45–55%), and K3 98.9 61.8 50.9 139
k1 28.7 54.6 49.1 26.8
fasted overnight before the experiment, with free access to water.
k2 36.2 22.8 31.9 24.8
Animals were randomly assigned to control and test groups, with k3 33.0 20.6 17.0 46.3
3 dogs in each group. A two-period cross-over design was used k 7.5 31.8 32.1 21.5
to compare propofol microemulsion and propofol emulsion (1% A: concentration of Solutol® HS 15; B: concentration of MCT; C: concentration of
Diprivan® ), with a 1-week washout period. All animals received egg lecithin; D: concentration of glycocholic acid.
the same dose of propofol (0.4 mg/kg body weight) intravenously.
Blood samples (1.5 mL) were collected from a limb vein at 1, 3, 5,
increasing glycocholic acid levels (Fig. 1E). One possibility is that
10, 20, 40, 60, 90, 120, 180 and 240 min after injection. A plasma
more sodium glycocholate micelles formed when the concentra-
sample (0.2 mL) was transferred into a centrifuge tube, and stored
tion of glycocholic acid exceeded 0.5%, and bile salt micelles have
at −20 ◦ C for later analysis.
been shown to have strong hemolytic activity (Narain et al., 1998).
Saturated ammonium sulfate, methanol and acetonitrile (0.2 mL
The results of the orthogonal design study are shown in Table 4.
of each) were added to the plasma (0.2 mL). After vortexing for
The concentrations of Solutol® HS 15, MCT, egg lecithin and gly-
1 min and centrifugation at 15,000 rpm for 15 min, 100 ␮L of super-
cocholic acid (w/v) were chosen as the most influential factors
natant was analyzed using an HPLC/FID method (HP1100 series
(labeled as A, B, C and D). Taking the hemolysis as an index, the
HPLC system, Agilent Technologies, UK). A ZORBAX-C18 column
four factors were investigated at three different levels. k1, k2 and
(4.6 ␮m × 250 mm, Agilent Technologies, UK) was chosen to sepa-
k3 were the average hemolysis of Level 1, Level 2 and Level 3 for
rate the samples. Wavelengths of 276 nm (excitation) and 310 nm
each factors. The level which got the lowest hemolysis was chosen
(emission) were used for fluorescence detection. The mobile phase
as the optimal level of each factor. For all the experiments in the
was methanol–water solution (75/25, v/v), with a flow rate of
orthogonal design, the hemolysis were within a range of 4.8–71.1%.
1.0 mL/min.
The ranking of the four factors in this experiment was C > B > D > A,
The pharmacokinetic parameters including distribution half-life
and the individual levels within each factor were ranked as: A,
(t1/2˛ ), elimination half-life (t1/2ˇ ), apparent volume of distri-
2 > 3 > 1; B, 1 > 2 > 3; C, 1 > 2 > 3; D, 3 > 1 > 2. The optimal formula-
bution (Vd ), clearance of drug (CL) and total area under the
tion was found to be A1 B3 C3 D2 , which was shown to induce no
concentration–time curve (AUC(0–240 min) ) were assessed by using
in vitro hemolysis.
compartmental analysis method with a Drug and Statistics (DAS)
In previous studies, pseudo-ternary phase diagrams commonly
v.2 software (Mathematical Pharmacology Professional Commit-
have been used for screening microemulsion formulations (Fu et al.,
tee of China, Shanghai, China) 3.0. All the data are presented as
2006; Nornoo et al., 2008). Using this approach, the physical states
mean ± standard deviation (S.D.). The statistical significance of dif-
were usually represented on a pseudo-ternary phase diagram with
ferences was determined by one-way ANOVA using SPSS software
one axis representing water, the second representing oil, and the
version 13.0 (SPSS Inc., Chicago, IL, USA). A value of p < 0.05 was
third representing the weight ratio of the surfactant to the co-
considered to be significant.
surfactant. The optimal formulation was selected on the basis of its
ability to incorporate large amounts of water into the surfactant:oil
3. Results and discussion mixture and load a large quantity of drug into the microemulsion.
In the present study, the goal was to obtain a microemulsion with
3.1. Formulation optimization the least hemolytic activity and lowest concentration of free propo-
fol. For this reason, orthogonal design, instead of a pseudo-ternary
Prior to orthogonal design, five factors were pre-screened by phase diagram method, was used for formulation screening.
varying only one factor at a time. The influences of Solutol® HS Microemulsions are spontaneously forming colloidal disper-
15, MCT, propofol, egg lecithin and glycocholic acid on particle size sions of either oil-in-water (o/w) or water-in-oil (w/o) stabilized
and hemolysis are shown in Fig. 1. The mean particle size of each by an interfacial film of surfactant(s) and cosurfactant(s) (Gupta
sample was less than 100 nm, and decreased with increasing con- and Moulik, 2008). The appearance and structural model of the
centrations of surfactants (Solutol® HS 15 and egg lecithin), and optimal microemulsion is shown in Fig. 2. It may be seen that the
with decreasing concentrations of oils (MCT and propofol). The microemulsion was a clear liquid with a yellow tinge. With regard
degree of hemolysis increased with the concentration of propofol to structure, propofol was dissolved in MCT, which served as the oil
in the formulation, such that hemolysis reached 100% when the for- core of the microemulsion; Solutol® HS 15, egg lecithin and sodium
mulation contained 2% propofol. Addition of Solutol® HS 15, MCT glycocholate together formed a corona that surrounded the oil core.
and egg lecithin were all found to decrease hemolysis. Interest- Solutol® HS 15 is a low molecular weight surfactant, recom-
ingly, at the lower concentrations of glycocholic acid used, particle mended as a non-ionic solubilizing agent that may be added to
size and hemolysis changed slightly with increasing glycocholic injection solutions.
acid levels. When concentrations of glycocholic acid were larger LD50 intravenously in mice, rats and beagle dogs are 3.160,
than 0.5%, particle size and hemolysis increased apparently with 1.212 and 3.160 g/kg respectively (Mittwollen, 2002). Its main
540 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544

Mean Particle Size (nm)

A Hemolysis (%) B
Mean Particle Size (nm)
Hemolysis (%)
100 90 100 100
Mean Particle Size (nm)

80 80

Mean Particle Size (nm)

80 80

Hemolysis (%)

Hemolysis (%)
60 70 60 60

40 60 40 40

20 50 20 20

0 40 0 0
5 7 9 11 13 0 1 3 5 7
Solutol HS 15 (%, w/v) MCT (%, w/v)

Mean Particle Size (nm)

C
Mean Particle Size (nm)
D Hemolysis (%)
Hemolysis (%)
120 60 70
100
100

Mean Particle Size (nm)

Mean Particle Size (nm)

50 60
80

Hemolysis (%)
Hemolysis (%)
80
40 50
60
60
40 30 40
40

20 20 30
20

0 0 10 20
0 1 2 3 4 0 1 1.5 2 2.5
Propofol (%, w/v) Egg Lecithin (%, w/v)

Mean Particle Size (nm)

E Hemolysis (%)
100 100
Mean Particle Size (nm)

80 80
Hemolysis (%)

60 60

40 40

20 20

0 0
0 0.25 0.5 1 1.5
Glycocholic Acid (%, w/v)

Fig. 1. Effect on particle size and hemolysis of varying the concentration of (A) Solutol® HS 15, (B) MCT, (C) propofol, (D) egg lecithin and (E) glycocholic acid (n = 3).

Fig. 2. Physical appearance and structural model of the propofol microemulsion.

 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544 541

Table 5 A microemulsion is a thermodynamically stable system, with

Average particle size, polydispersity index and zeta potential of propofol microemul-
droplet diameter usually within the range of 10–140 nm (Karasulu,
sion and Diprivan® (mean ± S.D., n = 3).
2008). The optimal microemulsion exhibited bimodal particle size
Formulation Average particle size Polydispersity index Zeta potential distributions with peak 1 at about 7 nm and peak 2 at about 45 nm.
(nm) (mV)
The average particle size was 22.6 ± 0.2 nm. The D(50), D(90) and
Microemulsion 22.6 ± 0.2 0.464 ± 0.010 −13.5 ± 1.6 D(99) values (particle diameter at 50, 90 and 99% of the cumu-
Diprivan® 233.8 ± 2.6 0.109 ± 0.180 −38.5 ± 1.3 lative particle size distribution) were 37.4 ± 1.4, 67.9 ± 4.1, and
96.5 ± 8.4 nm, respectively.
Propofol emulsion, and other lipid emulsions for intravenous
component is a polyglycol ester of 12-hydroxystearic acid, and delivery, are commonly manufactured with oil droplets averaging
its relatively bulky lipophilic portion allows it to better solubilize 150–300 nm in size (Baker and Naguib, 2005). The average parti-
drugs (Cho et al., 2010). Egg lecithin is widely employed as an emul- cle size of the Diprivan® emulsion was 233.8 ± 2.6 nm, with D(50),
sifier in parenteral emulsions and liposomes, due to its good safety D(90) and D(99) values of 243 ± 2.5, 379 ± 8.7, and 509 ± 10.1 nm,
(Guan et al., 2011). Sodium glycocholate is a bile salt that may respectively. No emulsion droplet larger than 1 ␮m was found.
be used to solubilize phosphatidylcholine, forming a clear mixed The visible appearance of the microemulsion or emulsion results
micellar solution that solubilizes substances with poor water solu- from the properties of the smaller particles. Particles less than one-
bility (Hammad and Muller, 1998). It is used as an ingredient with quarter the average wavelength of visible light (0.560 ␮m) allow
soybean lecithin to solubilize fat soluble vitamins in CernevitTM - light to pass through, and hence have an almost clear appearance
12 (multivitamins for infusion). MCT is a medium-chain (6–12 (Baker and Naguib, 2005). The propofol microemulsion was a clear
carbons) fatty acid ester of glycerol. In parenteral nutrition, it liquid with a particle size less than 100 nm. However, the propofol
has been found that replacement of long-chain triglyceride (LCT) emulsion appeared as a white milky solution due to the presence
fat solutions with MCT/LCT-enriched lipid emulsions results in of much larger droplets. The milky, opaque appearance of propo-
lower serum triglyceride levels during lipid infusion, and faster fol emulsion can hide visible macroparticles that would be noticed
elimination of triglycerides after completion of administration. It easily in the transparent microemulsion, an important advantage
is therefore possible that the use of MCT/LCT instead of an LCT lipid for a solution that is to be injected. Finkelstein et al. (1990) found
emulsion as the solvent for propofol may decrease the incidence of particles (presumably glass) ranging in length from approximately
development of hypertriglyceridemia (Theilen et al., 2002). In our 2.5–17 mm in Diprivan® . These types of visible matter need to
study, the content of MCT was 50 mg/mL, a 50% decrease compared be checked for after manufacture of the injections; however, the
to that of LCT in Diprivan® . Considering the lower content and opaque appearance of the injectable emulsion makes it difficult to
rapid elimination of MCT, it is possible that use of our propofol observe such macroparticles, posing a potential danger to patients.
microemulsion might decrease the incidence of hyperlipidemia. The transparency of our propofol microemulsion, similar to that of
some other commonly used injected solutions, allows for inspec-
3.2. pH and osmolarity tion for visible matter.

Pain on intravenous injection is likely to be caused by formula-

tions of drug with extremely un-physiological osmolalities or pH 3.4. Morphology
values (Klement and Arndt, 1991); for this reason, it is important to
determine the pH and osmotic pressure of the parenteral formula- Propofol microemulsion had a spherical appearance, according
tion to identify any need for adjustment of these. The pH of propofol to our observations using TEM (an image is shown in Fig. 4). The
microemulsion was found to be 6.72 ± 0.02, which was a little lower diameters of the microemulsion particles observed by TEM were
than that of Diprivan® (7.5 ± 0.04). The osmotic pressures of propo- in good agreement with the particle sizes determined as described
fol microemulsion and Diprivan® were 289.3 ± 1.5 mOsmol/L and above.
292.1 ± 1.8 mOsmol/L, respectively; both these formulations there-
fore were slightly hypotonic to 0.9% (w/v) saline solution.

3.3. Particle size and zeta potential

The average particle size, polydispersity index and zeta poten-

tial of propofol microemulsion and emulsion are shown in Table 5,
and the particle size distribution spectrum for the microemulsion
is presented in Fig. 3.

Fig. 3. Particle size distribution spectrum for the propofol microemulsion. Fig. 4. TEM image of propofol microemulsion.
542 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544

lower than that in an LCT emulsion, whereas AquafolTM has a sig-

nificantly higher level of free propofol, possibly due to a lack of lipid
in this formulation.
In our previous study, it was observed that the hemolysis
increased with the concentration of free propofol (Cai and Jin,
2008). In the present study, the optimal microemulsion, which
induced no in vitro hemolysis, was found to have a significantly
lower free propofol concentration than Diprivan® . We conclude
that our propofol microemulsion may induce less pain on injection
than Diprivan® .

Fig. 5. Free propofol concentration in the aqueous phase of the microemulsion and 3.6. Storage stability
Diprivan® (mean ± S.D., n = 3).

To determine the stability of the propofol microemulsion over

an extended period of time, accelerated and long-term room tem-
3.5. Concentration of free propofol perature stability tests were performed, as described in the relevant
guidelines (Tables 6 and 7). Significant indications of instability
Diprivan® was found to have a free propofol concentration of were observed in the accelerated stability test, especially changes
20.2 ␮g/mL in the aqueous phase, i.e. 0.2% of the total propofol in the pH and mean particle size; the propofol assay also showed a
content, a value similar to that reported previously (Müller and decrease of over 2%, and total related substances increased by more
Harnisch, 2000). In contrast, our propofol microemulsion showed than 0.15%. The propofol microemulsion was more stable under
a significant reduction in the free propofol concentration of 21.3%, long-term storage conditions at room temperature. Although the
compared with the commercial preparation (p < 0.01) (Fig. 5). mean particle size increased (somewhat surprisingly) by 10 nm
The concentration of free propofol in the aqueous phase has during long-term storage, possibly due to degradation of excipi-
been demonstrated to be associated with the intensity of pain on ents, there were minimal changes in propofol assay levels, pH and
injection. Scott et al. (1988) speculated that free propofol produces related substances over the course of the 12-month storage time,
injection pain by affecting the plasma kallikrein–kinin system. implying that propofol was chemically stable in the microemulsion.
However, Sim et al. (2009) have reported that bradykinin gener-
ation via activation of the plasma kallikrein–kinin system is not
3.7. Pharmacokinetics study in beagle dogs
related to propofol-induced pain: neither propofol itself nor the
lipid solvent were found to activate the plasma kallikrein–kinin
In our experiments, the relationship between the peak area (Y)
system in their tests.
and propofol concentration (X, ranging from 0.01 to 7 ␮g/mL) was
Several agents, including lidocaine and sufentanil, are com-
described by the following equation:
monly used with propofol in clinical practice. These agents have
been reported to successfully decrease propofol-induced pain, but Y = 446.585X − 4.551 (R2 = 0.9999) (3)
without reducing the free propofol concentration (Yamakage et al.,
2005): it was found that the free propofol levels in microemulsion The %RSD values for intraday and interday precision were less
and lipid emulsion mixed with lidocaine were not different from than 3.2% and 15.7%, respectively. The absolute recovery of propo-
those mixed with saline (as controls). These favorable actions of fol in rat plasma was 77.4–82.4%, and the relative recovery was
lidocaine may be attributed to its inhibition of pain transmission 96.3–114.2%. The limit of detection was found to be 0.01 ␮g/mL.
via the free nerve endings of vessels. The mean plasma concentration–time profiles for propofol
Propofol-LCT/MCT lipid emulsion has been reported to reduce microemulsion and Diprivan® are shown in Fig. 6. The initial plasma
the incidence and intensity of injection pain, compared with concentration of propofol for the microemulsion was slightly
a propofol-LCT formulation (Sundarathiti et al., 2007). In addi- higher, but overall, the plasma concentration profile for propofol
tion, it has been found that the concentration of free propofol was not significantly different from that of Diprivan® . In addition,
in a propofol-LCT/MCT emulsion (about 14 ␮g/mL) was signifi- there were slightly rapid distribution and elimination phases for
cantly lower (by 26%) than that of a propofol-LCT emulsion (about the microemulsion than Diprivan® .
19 ␮g/mL) (Müller and Harnisch, 2000).
Wang et al. (2007) hypothesized that the more stable propo-
2.4
Concentration of propofol (µg/ml)

fol microemulsion might generate less free propofol, due to a

2.2 Microemulsion
reduction in leakage or cracking, and thereby reduce pain at
2
the injection site. However, the approved Korean microemul- Diprivan®
1.8
sion, AquafolTM , was found to have a free propofol concentration 1.6
of 63.3 ± 1.2 ␮g/mL, which was significantly higher than that of 1.4
propofol-LCT emulsion (12.4 ± 0.7 ␮g/mL). This correlated with the 1.2
incidences of pain on injection, which were observed to be 81.9% 1
with AquafolTM and 29.2% with LCT emulsion (Jung et al., 2010). 0.8
The types of lipid used in the formulations may contribute to the 0.6
0.4
observed differences in free propofol concentrations. It has been
0.2
reported that the partitioning coefficients for propofol differ sub-
0
stantially, depending on the nature of the oil used. The partitioning
1 3 5 10 20 40 60 90 120 180 240
coefficient was found to be higher in an MCT/water system com-
pared to a soybean oil/water system (Müller and Harnisch, 2000), Time (min)
and utilizing MCT in a propofol microemulsion could distinctly
Fig. 6. Profiles for mean plasma concentration of propofol vs. time for the opti-
reduce the free propofol content. These factors may explain why mized microemulsion ( ) and Diprivan® ( ), after intravenous administration
the free propofol concentration in an LCT/MCT lipid emulsion is (mean ± S.D., n = 6).
 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544 543

Table 6
Results of the accelerated stability test for propofol microemulsion (mean ± S.D., n = 3).

Test item Initial 1 month 2 months 3 months 6 months

Appearance Slightly yellow clear Slightly yellow clear Slightly yellow clear Slightly yellow clear Slightly yellow clear
liquid liquid liquid liquid liquid
pH 6.72 ± 0.02 6.67 ± 0.01 6.57 ± 0.06 6.29 ± 0.01 6.15 ± 0.01
Mean particle size (nm) 22.6 ± 0.2 25.2 ± 0.2 26.0 ± 0.2 28.5 ± 0.4 31.0 ± 0.1
Assay (%) 103.63 ± 0.29 102.49 ± 0.22 102.68 ± 0.11 102.41 ± 0.21 101.59 ± 0.26
Total related 0.078 ± 0.016 0.137 ± 0.014 0.146 ± 0.033 0.148 ± 0.014 0.249 ± 0.037
substances (%)
Impurity J (%) 0.002 ± 0.000 0.003 ± 0.000 0.004 ± 0.000 0.004 ± 0.000 0.015 ± 0.003
Impurity E (%) 0.052 ± 0.001 0.091 ± 0.002 0.083 ± 0.001 0.097 ± 0.004 0.122 ± 0.003

Table 7
Results of the long-term stability test for propofol microemulsion (mean ± S.D., n = 3).

Test item Initial 3 months 6 months 9 months 12 months

Appearance Slightly yellow clear Slightly yellow clear Slightly yellow clear Slightly yellow clear Slightly yellow clear
liquid liquid liquid liquid liquid
pH 6.72 ± 0.02 6.65 ± 0.01 6.57 ± 0.03 6.53 ± 0.04 6.67 ± 0.05
Mean particle size (nm) 22.6 ± 0.2 21.8 ± 0.1 22.1 ± 0.1 26.5 ± 0.2 32.1 ± 0.6
Assay (%) 103.63 ± 0.29 103.46 ± 0.30 102.86 ± 0.14 102.69 ± 0.40 102.26 ± 0.09
Total related 0.078 ± 0.016 0.145 ± 0.031 0.166 ± 0.011 0.120 ± 0.011 0.118 ± 0.026
substances (%)
Impurity J (%) 0.002 ± 0.000 0.010 ± 0.000 0.010 ± 0.000 0.002 ± 0.000 0.002 ± 0.000
Impurity E (%) 0.052 ± 0.001 0.058 ± 0.000 0.067 ± 0.001 0.082 ± 0.001 0.079 ± 0.001

The main pharmacokinetic parameters were calculated based receiving microemulsion propofol, which may be explained by
on two-compartment modeling, and are summarized in Table 8. a smaller elimination rate constant from the hypothetical effect
The initial plasma drug concentration was 1.468 ± 0.729 ␮g/mL for compartment (ke0 ).
the microemulsion, which rapidly declined to 0.282 ± 0.102 ␮g/mL In contrast, Morey et al. (2006) found that propofol microemul-
within 10 min. This sharp fall in the serum concentration may sion and emulsion had similar pharmacokinetics, and that the
be attributed to a high distribution rate. The distribution half- microemulsion formulation did not modify thromboelastographic
life (t1/2˛ ) was 1.158 ± 0.416 min, which was approximately 60% parameters in swines. Besides possible species–species variation,
shorter than that for Diprivan® (2.898 ± 1.322 min) (p < 0.05). The these differing findings may be due to variations in the types and
data shown in Table 8 indicated that the microemulsion propo- concentrations of the excipients associated with the microemul-
fol was rapidly distributed, penetrated quickly into tissues, and sion. For example, the surfactants used in AquafolTM were 10%
was rapidly eliminated. Furthermore, the microemulsion also had a purified poloxamer 188 and 0.7% Solutol HS 15, whereas the surfac-
shorter elimination half-life (t1/2ˇ ) than that of Diprivan® (p < 0.05). tants used in Morey’s research were 5% purified poloxamer 188 and
Propofol in the microemulsion may require less time to be released 0.21% sodium laurate. Alterations in the composition, including the
from the oil phase than that in the emulsion. Some other param- surfactant used, may induce marked changes in the physical char-
eters, such as apparent volume of distribution (Vd ), clearance (CL) acteristics and pharmacokinetic properties of the microemulsion.
and total area under the concentration–time curve (AUC(0–240 min) ), In the present study, different excipients were used to pre-
showed no significant differences between the microemulsion and pare the microemulsion, and different pharmacokinetic properties
emulsion formulations (p > 0.05). were observed, to those reported above. The rapid distribution and
Other investigators have also examined the pharmacokinetic release of propofol from the microemulsion may translate to more
properties of propofol microemulsions. Jung and coworkers (Jung rapid onset and recovery than the commercial emulsion. However,
et al., 2010) studied the pharmacokinetic and pharmacodynamic further pharmacokinetic and pharmacodynamic studies with an
characteristics of a propofol microemulsion (AquafolTM ) and infusion of propofol microemulsion would be needed to confirm
emulsion (Diprivan® ) in patients. The microemulsion propofol this hypothesis.
showed smaller V1 (central volume of distribution) and more
extensive peripheral distribution than that of the emulsion, which 4. Conclusions
induced more prolonged recovery of microemulsion propofol. Fur-
thermore, a slightly slower onset of sleep was observed in patients In this study, a propofol microemulsion was successfully pre-
pared using Solutol® HS 15, egg lecithin and glycocholic sodium
salt (as emulsifier and co-emulsifiers), with MCT as the oil phase.
Table 8
The microemulsion was a clear liquid with a particle size less than
Pharmacokinetic parameters for propofol after intravenous administration of the
microemulsion and Diprivan® in beagle dogs (mean ± S.D., n = 6). 100 nm, according to measurement of particle size and TEM imag-
ing. Its transparency enabled easier screening for visible matter
Parameter Microemulsion Diprivan®
than in the commercial emulsion. In addition, the content of oil
t1/2˛ (min) 1.158 ± 0.416*
2.898 ± 1.322 was low in the microemulsion, which may decrease the incidence
t1/2ˇ (min) 30.564 ± 19.334* 56.600 ± 17.315 of hyperlipidemia. The microemulsion also showed a significantly
Vd (L/kg) 0.202 ± 0.120 0.447 ± 0.457
CL (L/min/kg) 0.023 ± 0.007 0.022 ± 0.007
lower free propofol concentration than the commercial emulsion,
AUC(0–240 min) (mg min/L) 16.849 ± 4.560 17.579 ± 6.154 which we hypothesize would induce less pain on injection. Storage
stability tests indicated that the microemulsion was stable under
t1/2˛ , distribution half-life; t1/2ˇ , elimination half-life; Vd , apparent volume of distri-
bution; CL, clearance of drug; AUC(0–240 min) , total area under the concentration–time room conditions for 12 months. Furthermore, our microemulsion
curve. demonstrated superior pharmacokinetics by increasing distribu-
*
p < 0.05 compared with Diprivan® . tion and elimination rates, which may translate to more rapid onset
544 W. Cai et al. / International Journal of Pharmaceutics 436 (2012) 536–544

and recovery than the commercial emulsion. Further studies are Klement, W., Arndt, J.O., 1991. Pain on i.v. injection of some anaesthetic agents
needed to focus on the tissue distribution, safety and efficacy of the is evoked by the unphysiological osmolality or pH of their formulations. Br. J.
Anaesth. 66, 189–195.
propofol microemulsion, to evaluate its potential for clinical use. Langevin, P.B., 1999. Propofol containing sulfite-potential for injury. Chest 116,
1140–1141.
Acknowledgements Langley, M.S., Heel, R.C., 1988. Propofol. A review of its pharmacodynamic and
pharmacokinetic properties and use as an intravenous anaesthetic. Drugs 35,
334–372.
Financial support from Science and Technology Commission of Lee, E.H., Lee, S.H., Park, D.Y., Ki, K.H., Lee, E.K., Lee, D.H., Noh, G.J., 2008. Physicochem-
Shanghai Municipality (Project No. 0952NM04500) is gratefully ical properties, pharmacokinetics, and pharmacodynamics of a reformulated
microemulsion propofol in rats. Anesthesiology 109, 436–447.
acknowledged. Mittwollen, J.P., 2002. Solutol HS 15 Information for Customers. BASF Aktienge-
sellschaft, Berlin, Germany, p. 16.
References Morey, T.E., Modell, J.H., Shekhawat, D., Shah, D.O., Klatt, B., Thomas, G.P., Kero, F.A.,
Booth, M.M., Dennis, D.M., 2006. Anesthetic properties of a propofol microemul-
sion in dogs. Anesth. Analg. 103, 882–887.
Babu, M.K., Godiwala, T.N., 2004. Toward the development of an injectable dosage
Müller, R., Harnisch, S., 2000. Physicochemical characterization of propofol-loaded
form of propofol: preparation and evaluation of propofol-sulfobutyl ether 7-
emulsions and interaction with plasma proteins. Eur. Hosp. Pharm. 6, 24–31.
beta-cyclodextrin complex. Pharm. Dev. Technol. 9, 265–275.
Narain, P.K., DeMaria, E.J., Heuman, D.M., 1998. Lecithin protects against plasma
Baker, M.T., Naguib, M., 2005. Propofol: the challenges of formulation. Anesthesiol-
membrane disruption by bile salts. J. Surg. Res. 78, 131–136.
ogy 103, 860–876.
Nornoo, A.O., Osborne, D.W., Chow, D.S., 2008. Cremophor-free intravenous
Cai, W.H., Jin, F., 2008. Relationship between free propofol concentrations of differ-
microemulsions for paclitaxel I: formulation, cytotoxicity and hemolysis. Int.
ent delivery systems for injection and in vitro hemolytic activities. Chinese J.
J. Pharm. 349, 108–116.
Pharm. 39, 22–27.
Ohmizo, H., Obara, S., Iwama, H., 2005. Mechanism of injection pain with long and
Cho, J., Cho, J.C., Lee, P., Lee, M., Oh, E., 2010. Formulation and evaluation of an alterna-
long-medium chain triglyceride emulsive propofol. Can. J. Anaesth. 52, 595–599.
tive triglyceride-free propofol microemulsion. Arch. Pharm. Res. 33, 1375–1387.
Park, J.W., Park, E.S., Chi, S.C., Kil, H.Y., Lee, K.H., 2003. The effect of lidocaine on the
Egan, T.D., Kern, S.E., Johnson, K.B., Pace, N.L., 2003. The pharmacokinetics and phar-
globule size distribution of propofol emulsions. Anesth. Analg. 97, 769–771.
macodynamics of propofol in a modified cyclodextrin formulation (Captisol)
Pergolizzi Jr., J.V., Gan, T.J., Plavin, S., Labhsetwar, S., Taylor, R., 2011. Perspectives
versus propofol in a lipid formulation (Diprivan): an electroencephalographic
on the role of fospropofol in the monitored anesthesia care setting. Anesthesiol.
and hemodynamic study in a porcine model. Anesth. Analg. 97, 72–79.
Res. Pract. 2011, 458920.
Finkelstein, A., Lokhandwala, B.S., Pandey, N.S., 1990. Particulate contamination of
Ryoo, H.K., Park, C.W., Chi, S.C., Park, E.S., 2005. Development of propofol-
an intact glass ampule. Anesthesiology 73, 362–363.
loaded microemulsion systems for parenteral delivery. Arch. Pharm. Res. 28,
Fu, X., Feng, F., Huang, B., 2006. Physicochemical characterization and evaluation of
1400–1404.
a microemulsion system for antimicrobial activity of glycerol monolaurate. Int.
Scott, R.P., Saunders, D.A., Norman, J., 1988. Propofol: clinical strategies for prevent-
J. Pharm. 321, 171–175.
ing the pain of injection. Anaesthesia 43, 492–494.
Guan, T., Miao, Y., Xu, L., Yang, S., Wang, J., He, H., Tang, X., Cai, C., Xu, H., 2011.
Sim, J.Y., Lee, S.H., Park, D.Y., Jung, J.A., Ki, K.H., Lee, D.H., Noh, G.J., 2009. Pain on
Injectable nimodipine-loaded nanoliposomes: preparation, lyophilization and
injection with microemulsion propofol. Br. J. Clin. Pharmacol. 67, 316–325.
characteristics. Int. J. Pharm. 410, 180–187.
Sundarathiti, P., Boonthom, N., Chalacheewa, T., Jommaroeng, P., Rungsithiwan, W.,
Gupta, S., Moulik, S.P., 2008. Biocompatible microemulsions and their prospective
2007. A comparison of propofol-LCT with propofol-LCT/MCT on pain of injection.
uses in drug delivery. J. Pharm. Sci. 97, 22–45.
J. Med. Assoc. Thailand 90, 2683–2688.
Gupta, A., Stierer, T., Zuckerman, R., Sakima, N., Parker, S.D., Fleisher, L.A., 2004.
Theilen, H.J., Adam, S., Albrecht, M.D., Ragaller, M., 2002. Propofol in a medium- and
Comparison of recovery profile after ambulatory anesthesia with propofol,
long-chain triglyceride emulsion: pharmacological characteristics and potential
isoflurane, sevoflurane and desflurane: a systematic review. Anesth. Analg. 98,
beneficial effects. Anesth. Analg. 95, 923–929.
632–641.
Thompson, K.A., Goodale, D.B., 2000. The recent development of propofol (DIPRI-
Hammad, M.A., Muller, B.W., 1998. Increasing drug solubility by means of bile
VAN). Intens. Care Med. 26 (Suppl. 4), S400L 404.
salt-phosphatidylcholine-based mixed micelles. Eur. J. Pharm. Biopharm. 46,
Trapani, A., Laquintana, V., Lopedota, A., Franco, M., Latrofa, A., Talani, G., Sanna,
361–367.
E., Trapani, G., Liso, G., 2004. Evaluation of new propofol aqueous solutions for
Han, J., Davis, S.S., Washington, C., 2001. Physical properties and stability of two
intravenous anesthesia. Int. J. Pharm. 278, 91–98.
emulsion formulations of propofol. Int. J. Pharm. 215, 207–220.
Wachowski, I., Jolly, D.T., Hrazdil, J., Galbraith, J.C., Greacen, M., Clanachan, A.S.,
Jumaa, M., Muller, B.W., 2000. Lipid emulsions as a novel system to reduce the
1999. The growth of microorganisms in propofol and mixtures of propofol and
hemolytic activity of lytic agents: mechanism of the protective effect. Eur. J.
lidocaine. Anesth. Analg. 88, 209–212.
Pharm. Sci. 9, 285–290.
Wang, H., Cork, R., Rao, A., 2007. Development of a new generation of propofol. Curr.
Jung, J.A., Choi, B.M., Cho, S.H., Choe, S.M., Ghim, J.L., Lee, H.M., Roh, Y.J., Noh, G.J.,
Opin. Anaesthesiol. 20, 311–315.
2010. Effectiveness, safety, and pharmacokinetic and pharmacodynamic char-
Wolf, A., Weir, P., Segar, P., Stone, J., Shield, J., 2001. Impaired fatty acid oxidation in
acteristics of microemulsion propofol in patients undergoing elective surgery
propofol infusion syndrome. Lancet 357, 606–607.
under total intravenous anaesthesia. Br. J. Anaesth. 104, 563–576.
Yamakage, M., Iwasaki, S., Satoh, J., Namiki, A., 2005. Changes in concentra-
Karasulu, H.Y., 2008. Microemulsions as novel drug carriers: the formation, stability,
tions of free propofol by modification of the solution. Anesth. Analg. 101,
applications and toxicity. Expert Opin. Drug Deliv. 5, 119–135.
385–388.
 Applications and Use of Microemulsions
Kai Lun LEE, CID:00664757
Department of Chemical Engineering and Chemical Technology,
Imperial College London, November 2010

Abstract
During the past five decades since the discovery of microemulsions by Jack H. Shulman, there have been
huge progresses made in applying microemulsion systems in a plethora of research and industrial
processes. While it is beyond the scope of this paper to give a complete review of all significant
developments and applications, it will attempt to highlight several recent developments in applications
that might interest readers for whom this paper serves as an introduction to microemulsions. In that note,
the relevance of this paper and the truncated scientific background on microemulsions are first discussed.

INTRODUCTION other food and pharmaceutical applications,

and in the petrochemical industry.
Microemulsions, or µ-emulsions, are isotropic
mixtures of oil, water and surfactant; usually RELEVANCE
with a co-surfactant and the oil being a
mixture of different hydrocarbons and The significance and potential that
olefins1. In contrast to ordinary emulsions researchers attach to microemulsions are in
which are kinetically stable but are no small part due to their unique properties
thermodynamically unstable and will phase that are low interfacial tension, high
separate2, microemulsions are thermodynamic stability, high interfacial area,
thermodynamically stable and therefore do and the ability to dissolve immiscible liquids.
not require high inputs of energy or shear
conditions for their formation. They are also These have allowed researchers to discover,
clear, as compared to emulsions which are over the past few decades, many practical
cloudy. Despite being named microemulsions, uses for microemulsions that are employed in
the droplet sizes of the dispersed phase in many industrial processes and consumer
microemulsions are generally in the products, either as new processes and
magnitude of ~10nm. Apart from being more products or as more efficient processes and
commercially viable due to its lower energy better or cheaper products. Examples
requirements, microemulsions are also described in this paper include drugs,
attracting the interest of researchers due to cosmetics, electronics, and petrol. In short,
their potential as drug delivery vehicles, in microemulsions have and can continue to
enable us to make better consumer products.
1
There is particularly strong interest in the
Chandra, A(1992). Microemulsions: An Overview
potential of microemulsions as novel drug
Retrieved from
delivery systems, as they are believed to hold
http://www.pharmainfo.net/reviews/microemulsi
ons-overview the key to making better orally delivered
drugs that can be absorbed better, and only at
2
K, Shinoda. B,Lindman (1987). Organised selected lengths along the gastrointestinal
surfactant systems: microemulsions, Langmuir 3 tract. With pharmaceutical revenue reaching
135–149
497 billion USD in 2007, the monetary
incentive alone is considerable. Microemulsions are usually characterised by
ternary phase diagrams, which three edges
BACKGROUND are the components of a microemulsion,
namely, oil, water and surfactant. Any co-
The concept of microemulsions was surfactant used are usually grouped together
introduced by Professor, Jack H. Shulman at with the surfactant at a fixed ratio and treated
Columbia University in 19593. The definition as a pseudo-component.
of a microemulsion have varied with time and
location, but the more commonly accepted The below shows a hypothetical ternary phase
view is that of a “system of water, oil and diagram of a microemulsion; illustrating that 2
amphiphile which is a single optically isotropic phase systems forms at very high surfactant
and thermodynamically stable liquid concentrations.
4
solution” and that in a microemulsion the
surfactant is located at a certain boundary
between the oil and aqueous phases, giving
the microemulsion a definite microstructure.

The abovementioned surfactant molecules, in

most cases, comprise of a polar head which
make up a small fraction of the molecular
volume and a non-polar tail. These two
regions allow them to interact with the polar
aqueous phase and the non-polar oil phase.
Surfactant molecules associate into different
Fig 1. Tenary phase diagram of microemulsion (courtesy of Advanced
forms, including spherical micelle, rod micelle Drug Delivery Reviews)
laminar phase and hexagonal phase, to
minimise the gibbs free energy of the system Co-surfactants are usually used in conjunction
and to “optimise solvation requirements”.5 with surfactants due to most single chain
The formation and stability of microemulsions surfactants being incapable of reducing the
can be described by interface science, interface tension of oil and water to form a
chemical solubility theories, or by microemulsion. The most common co-
thermodynamic explanations; the last one of surfactants are medium chain alcohols, which
which shows that a net release of free energy reduce the tension and increase the fluidity of
is obtained when favourable entropic the oil-water interface, thereby increasing the
contributions from the mixing of small entropy of the system6. These medium chain
droplets in the continuous phase and the alcohols also increase the motility of the
diffusion of surfactant in the inter-facial layer surfactants’ non-polar tail region, allowing
are larger than the unfavourable contribution greater penetration by oil molecules and
of the reduction of surface tension5, resulting therefore stabilising the system and
in a thermodynamically stable dispersion. facilitating the formation of a microemulsion.

3 In spite of their ease of formation, it is

J.H. Schulman, W. Stoeckenius, L.M. Prince,
important to accurately characterise
Mechanism of formation and structure of micro
emulsions by electron microscopy, J. Phys. Chem. microemulsions, especially for industrial
63 (1959) 1677–1680 processes. Characterisation of microemulsions
4
I. Danielsson, B. Lindman, The definition of a includes both macroscopic measurements and
microemulsion, Colloids and Surfaces 3 (1981)
6
391–392 M.J. Lawrence, Surfactant systems:
5
M.J. Lawrence, G.D. Rees, Microemulsions-based microemulsions and vesicles as vehicles for drug
media as novel drug delivery systems, Advanced delivery, Eur. J. Drug Metab. Pharmacokinet. 3
Drug Delivery Reviews 45 (2000) 89–121 (1994) 257–269.
microenvironment methods. The former when a colon-release microemulsion was
include viscosity measurements which designed and used9.
indicates the presence (or absence) of certain Microemulsions employed in the form of
surfactant forms, conductivity measurements oral-delivered drugs can come with a self-
which can determine the continuous and emulsifying ability. These are called Self-
dispersed phases, as well as dielectric microemulsifying drug delivery systems
measurements which gives insight to the (SMEDDSs). Such SMEDDS particularly interest
structure and dynamics of the particular researchers because of their ability to deliver
microemulsion7. On the other hand, hydrophobic drugs. With almost 40% of new
microenvironment studies can involve pulsed drug compounds hydrophobic10, the potential
field NMR and scattering methods such as commercial viability of SMEDDS is great.
light scattering, neutron scattering and X-ray
scattering. In spite of the promise of increased
dissolution and bioavailability of APIs in oral
RECENT DEVELOPMENTS forms, w/o microemulsions and SMEDDSs
have not been commercially or industrially
Water in Oil Drug Delivery Microemulsions exploited. This is in part due to the lack of
knowledge as to how drug molecules are
The rationale behind exploring water in oil distributed between the oil and aqueous
(w/o) microemulsions is to protect water phase, metabolism of the oil phase, and most
soluble drug molecules, in particular proteins importantly, how the absorption in human
and peptides from metabolism and to bodies (as compared to in animals) of oral
overcome physical barriers. These are dosage drugs are affected if the oil is non-
particularly attractive as such drug molecules digestible11. This depends on the combination
are heat-sensitive and these emulsions do not of the drug and the oil phase (usually a lipid),
require high temperatures to form. and can increase, inhibit or not affect
absorption. The difficulty in understanding
More interestingly, w/o microemulsions with these processes arises from the in vivo
encapsulated Active Pharmaceutical behaviour of drugs due to several physical and
Ingredients (APIs) are found to convert into oil physiological factors beyond the scope of this
in water (o/w) microemulsions upon the report11.
addition of particular aqueous fluids8, which
results in the release of the API. This allows That being said, further research and insight
w/o microemulsions to be designed to on the relationship among drug structures,
selectively release APIs at required locations microemulsion compositions and absorption
along the gastrointestinal tract. Typical can resolve current drug development issues
microemulsions used for this purpose and exploit the advantages of microemulsions
incorporate fatty-acid esters as the oil phase in drug delivery systems.
and isopropanol as the co-surfactant. Studies
carried out using these microemulsions
successfully showed that the absorption of
vasopressin in the intestines of rats was
9
tripled when compared to using aqueous W. A. Ritschel. Microemulsions for improved
solutions; and also resulted in much higher peptide absorption from the gastrointestinal tract.
bioavailability of insulin in the colon of dogs Meth. Find. Exp. Clinic. Pharmacol. 13:205-220
(1993)
10
M. J. Patel, S. S. Patel, N. M. Patel A Self
Microemulsifying Drug Delivery System
7
Y. Feldman, N. Kozlovich, I. Nir, N. Garti, Dielectric International Journal of Pharmaceutical Sciences
spectroscopy of microemulsions, Colloids Surfaces: Review and Research Volume 4 Issue 3 (Sept 2010)
11
Physicochem. Eng. Aspects 128 (1997) 47–61. P. P. Constantinides Lipid Microemulsions for
8
A. J. Owen, S. H. Yiv, A. Sarkahian, Convertible Improving Drug Dissolution and Oral Absorption:
microemulsion formation 29 October 1992 Pharmaceutical Research Vol 12 (1995)
Water in Carbon Dioxide Microemulsions Other surfactants with similar attributes such
as having tail groups with weak van der
Supercritical Carbon Dioxide is a “useful Waal’s forces and low solubility in water, an
replacement for organic solvents to minimise ionic head group with a tendency to leave
waste and volatile organic carbon emissions” CO2, and most importantly, carbon fluoride
as it is “non-flammable, non-toxic and the branching which bends the interface of water
least expensive solvent after water.”12 The can possibly act as suitable surfactants for
apparent advantages of carbon dioxide over water in CO2 microemulsions.
organic solvents have been tempered by
difficulties in dispersing water in carbon
dioxide to form a stable water-in-CO2 Synthesis of Nanosized Bismuth Oxyhalide
microemulsion. Recent efforts have, in Particles and other nanoparticles using
particular, been geared towards making CO2 microemulsions
accessible to proteins and similar hydrophilic
chemicals through water in CO2 Bismuth Oxyhalides are crucial to a number of
microemulsions. applications including BiOCl in cosmetics and
in cracking of butane and BiOI as
Theoretically, a dispersion of a hydrophilic or “components for color filters in transparent
lipophilic phase in CO2 could be stabilised by nanocomposite materials”14. Like other
surfactants with “CO2-philic” tails and nanosized particles, their shape, size and size
hydrophilic or lipophilic heads respectively. distribution are crucial to their applications in
The difficulty however, was finding such a “catalysis, electronics, miniaturization, and
surfactant that could disperse water in CO2. ceramics14”.
Attempts to locate such a surfactant involved
experimenting with 150 different substances Reverse microemulsions are often used in the
in the decade preceding13 the 1996 discovery preparation of nanoparticles to avoid broad
that commercially available ammonium particle size distributions that often result
carboxylate perflouropolyether (PFPE) with an from precipitation. A reverse microemulsion
average molecular weight of 740 are suitable differs from an emulsion in that the tail
surfactants for water in CO2 microemulsions groups of the surfactants orientate outwards
at pressures below 300bar13. Johnston and into the continuous phase and their head
team (1996) noted that the use of PFPE groups towards the dispersed phase.
created a thermodynamically stable aqueous
dispersed phase in continuous CO2 that more A relatively new method of using
importantly, did not alter the three microemulsions to synthesis the nanoparticles
dimensional conformation of proteins which is through a multi-microemulsion route. Such
are important to their functions; having a route involves sets of microemulsions with
demonstrated the results through the use of the same water, oil and surfactant used and
Fourier Transform Infrared Spectroscopy, similar ratios of the 3 components, but with
ultraviolet absorbance, fluorescence and each of the reactants for the nanoparticles
electron paramagnetic resonance dissolved in the aqueous phase of different
experiments. microemulsions. Intermicellar exchange of
reactants present in the reverse micelles
subsequently occurs and the nanoparticles
form. (See illustration on next page.) Studies
12 show that “the multi-microemulsion route
K.P.Johnston et al Water-in-Carbon Dioxide
Microemulsions: An
Environment for Hydrophiles Including Proteins
Science, New Series, Vol. 271, No. 5249 (Feb. 2,
14
1996), pp 624-626 J. Henle et al. Nanosized BiOX (X ) Cl, Br, I)
13
K. A. Consani and R. D. Smith, J. Supercritical Particles Synthesized in Reverse
Fluids 3, 51 (1990). Microemulsions Chem. Mater. 2007, pp366-373
yields finer particles and narrow size size using microemulsions also found an
distributions.15” application in the study of reaction kinetics in
catalytic processes. Some reaction kinetics
and mechanism are dependent on the catalyst
and/or its size and well-defined catalysts in
the nanometer range deposited on supports
are very often indispensible to study them.

It has been found that water-in-oil

microemulsions are particularly advantageous
over other methods of preparing nanoparticle
catalysts on supports, such as electron beam
lithography, colloidal lithography and spin-
coating, as they “can be formed at
atmospheric pressure and at room
temperature and that large sample volumes
relatively easily can be obtained.18” This
process involves adding support to the
microemulsion suspension and subsequently
destabilising the microemulsion by the
addition of a destabilizer such as
Fig 2. Multi-microemulsions route of synthesising nanoparticles.
tetrahydrofurane which removes the
Courtesy of Henle(2007)14 surfactant.

By varying the concentrations of the salt Enhanced Oil Recovery using Microemulsions
solutions in the aqueous phase and the water
to surfactant ratio, Henle and team (2007) The use of microemulsions is of high interest
managed to characterize and formulate the in many aspects of crude oil exploitation, but
relationship between those conditions and none more so that in enhanced oil recovery.
the particle size in the range of 3 to 22nm; In cases where the pressure exerted by
allowing them to manipulate characteristics of gushing sea water on the oil phase is not able
the nanoparticles such as “the band gap, to overcome capillary forces sufficiently,
absorption edge, and color of BiOI microemulsions are the key to extracting
nanoparticles”. more than just a minor portion of crude oil.
Properly balanced microemulsions are able to
Further on, processing and preparations of do so by drastically reducing the interfacial
other nano-sized substances also employed tension to the magnitude of 0.001 mN m-1 19.
the use of microemulsions such as for This is also known as chemical flooding.
nanocrystalline hydroxyapatite16 and for
single crystal BaMoO4 nanofibers17. Many difficulties are encountered in creating
a suitable microemulsion with temperature
The ability to narrow the size distribution, gradients required large, many ionic
create finer nanoparticles, and control their surfactants precipitating when contacted with
brine, and most non-ionic surfactants
15
Wang, J. A et al. Today 2001, 68, pp21.
16
L. Hong et al Processing of nanocrystalline
hydroxyapatite particles via reverse
18
microemulsions, Journal of Materials Science, v H.H.Ingelsten et al Deposition of Platinum
43, n 1, p 384-389, January 2008 Nanoparticles, Synthesized inWater-in-Oil
17
Z. Li et al Synthesis of single crystal BaMoO4 Microemulsions, Langmuir 2002, 18, pp1811-1818
19
nanofibers in CTAB reverse microemulsions L. Lake, Enhanced Oil Recovery, Prentice-Hall Inc,
Materials Letters, v 59, n 1, p 64-68, January 2005 1989.
unsuitable20. Other issues include adsorption is in part due to their ease of formation and
of microemulsion components on rocks, and stability.
varying salinities and temperatures of the sea.
Over the past three decades, however, there This paper has highlighted several recent
has been significant progress made on the developments in the use or studies on the use
recovery of residual oil in particular chemical of microemulsions both for industrial and
flooding with microemulsions and many research purposes. It is regretted that it is
patents issued, mostly concerned with “the beyond the scope and length of this paper to
chemistry of surfactant-based processes highlight all outstanding recent developments
including the use of chemical solutions to involving the use of microemulsions and is
decrease the surface tension between oil and therefore limited only to the discussions on
the flooding medium, screening of surfactants the application of microemulsions in novel
for oil recovery efficiency, chemical surfactant drug delivery systems, use of CO2 as a solvent,
designs and formulation to mobilize residual in enhanced oil recovery and in the synthesis
oil and other factors in the chemistry of of nanoparticles. In the first development
surfactant-based chemical flooding discussed, microemulsions enable the
21
processes. ” selective release Active Pharmaceutical
Ingredients along specified lengths in the
Recent reviews have focused on the gastrointestinal tract by inverting from a
application of suitable microemulsion systems water-in-oil emulsion to an oil-in-water
in specifically conditions of the North Sea emulsion. In the second, the discovery of a
leading to many patents filed by various suitable surfactant allows CO2 to form a
companies involved in oil recovery there; microemulsion in water thereby possibly
while researchers are also excited by the eliminating the need to use organic solvents
possible use of bacteria and biosurfactants which may be toxic, flammable and expensive.
that are able to create microemulsions. These In the third, microemulsions enable a much
substitutes are, in addition to being able to larger fraction of highly valuable residual oil
multiply within capillaries, capable of trapped under sea beds to be recovered by
physically modifying solid surfaces and reducing interfacial tensions. Research on
lowering viscosity of the oil phase22, and are biosurfactants, which additionally may
viewed as having great potential of lowering multiply themselves and lower viscosity of the
recovery costs. oil phase, is also ongoing. In the last
development discussed, nanoparticles with
wide ranges of applications in industry,
CONCLUSION research and consumer products are able to
be made finer, with a narrower size
Microemulsions, a relatively recent discovery distribution and better controlled properties
in 1959, have found applications in a wide by using microemulsions in their production.
variety of chemical and industrial processes.
Their wide use in both research and industry The field of microemulsion applications is far
from being fully established and researchers
face many difficulties yet are intrigued by
their immense potential in many areas due to
20
J. Sjiiblom, R. Lindbergh, S.Friberg their extremely low interfacial tension,
Microemulsions - phase equilibria characterization, thermodynamic stability, large interfacial area
structures, applications and chemical reactions and ability to dissolve immiscible liquids. With
Advances In Colloid and Interface Science 1996
continued research and studies,
pp125-287
21
B.Y.Jamaloei Insight to the chemistry of microemulsions could possibly play a much
surfactant-based enhanced oil recovery January bigger role in our lives by affecting the ways
2009 industries work and research is conducted.
22
B.K.Paul, S.P.Moulik Uses and applications of
microemulsions Current Science 80 April 2001
 Food Chemistry 132 (2012) 799–807

Contents lists available at SciVerse ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Nanoemulsion- and emulsion-based delivery systems for curcumin:

Encapsulation and release properties
Kashif Ahmed, Yan Li, David Julian McClements, Hang Xiao ⇑
Biopolymers and Colloids Research Laboratory, Department of Food Science, University of Massachusetts, Amherst, MA 01003, United States

a r t i c l e i n f o a b s t r a c t

Article history: Curcumin has been reported to have many biological activities, but its application as a functional ingredient
Received 20 June 2011 is currently limited because of its poor water-solubility and bioaccessibility. This study investigated the
Received in revised form 28 September 2011 impact of different lipid-based formulations on curcumin encapsulation and bioaccessibility. Oil-in-water
Accepted 8 November 2011
nanoemulsions (r < 100 nm), or conventional emulsions (r > 100 nm), were prepared with different lipids:
Available online 15 November 2011
long, medium, and short chain triacylglycerols (LCT, MCT and SCT, respectively). An in vitro model simulat-
ing small intestine digestion conditions characterised rate and extent of lipid phase digestion. A centrifu-
Keywords:
gation method determined fraction of curcumin released into mixed micelles after digestion
Curcumin
Bioavailability
(bioaccessibility). Initial digestion rate decreased in the order SCT > MCT > LCT, while final digestion extent
Bioaccessibility decreased in the order MCT > SCT > LCT. The bioaccessibility of curcumin decreased in the order
Digestion MCT > LCT  SCT and appeared to be slightly higher in conventional emulsions than in nanoemulsions.
pH-stat Ó 2011 Elsevier Ltd. All rights reserved.
Emulsions
Nanoemulsions
Delivery systems

1. Introduction wavelength of light so that they only scatter light weakly. The bio-
availability of encapsulated lipophilic components within the gas-
Emulsion-based delivery systems are being used increasingly for trointestinal tract may be increased by using nanoemulsions
encapsulating lipophilic bioactive components in the pharmaceuti- because of their relatively small droplet size (Patel & Sawant,
cal industry (Porter, Trevaskis, & Charman, 2007). Oil-in-water 2007). One of the objectives of the present study was therefore to
(O/W) emulsions, or nanoemulsions, can be prepared by solubilising determine the influence of droplet size on the digestion and release
the lipophilic bioactive components within the oil phase, and then characteristics of emulsion-based delivery systems.
homogenising this phase with an aqueous phase containing a The digestion of lipids within the gastrointestinal tract is a com-
water-soluble emulsifier. The size of the droplets produced depends plex process and its impact on the release and uptake of any encap-
on the composition of the system and the homogenisation method sulated lipophilic components is not fully understood. In vitro
used. O/W conventional emulsions and nanoemulsions are both digestion models have recently gained much attention as a tool for
thermodynamically unstable systems that consist of emulsifier- understanding the basic physicochemical processes that occur dur-
coated lipid droplets dispersed within an aqueous medium (Solans, ing lipid digestion and the release of encapsulated components
Izquierdo, Nolla, Azemar, & Garcia-Celma, 2005). However, the two (McClements & Li, 2010). In particular, in vitro lipolysis models
systems can be distinguished from each other based on their particle (‘‘pH-stat methods’’) have been developed as a powerful means of
sizes: the mean droplet radius is <100 nm for nanoemulsions, but quantifying the lipid digestion process (Pouton, 2006, Zangenberg,
>100 nm for conventional emulsions. The relatively small size of Mullertz, Kristensen, & Hovgaard, 2001a, 2001b) and the release of
the droplets in nanoemulsions means that they often have different lipophilic substances into the various colloid phases formed during
physicochemical and biological properties than conventional emul- lipid digestion, e.g., mixed micelles (Porter et al., 2007). The fraction
sions. Nanoemulsions have been reported to have better stability to of a lipophilic component released into the mixed micelle phase
particle aggregation and gravitation separation due to their small after lipid digestion can be taken as a marker of its bioaccessibility.
droplet sizes (Solans et al., 2005; Sonneville-Aubrun, Simonnet, & The bioactivity of a component, which is a measure of its specific bio-
L’Alloret, 2004). Nanoemulsions that are optically transparent can logical affect, is determined by the fraction available for absorption.
be prepared by making the droplet size much smaller than the Ultimately, a successful in vitro digestion model should be able to
predict in vitro–in vivo correlations because this is a prerequisite
⇑ Corresponding author. Tel.: +1 413 545 2281; fax: +1 413 545 1262. for rational development of effective delivery systems (Fricker
E-mail address: hangxiao@foodsci.umass.edu (H. Xiao). et al., 2010).

0308-8146/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2011.11.039
800 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807

In this study, we used curcumin as a model lipophilic compo- 225A, Fisher Scientific), at ambient temperature, for 10 min to re-
nent that could be incorporated into functional food products. move non-dissolved (crystalline) curcumin, and then the superna-
Curcumin is a natural polyphenolic phytochemical extracted tant was collected. Preliminary experiments indicated that the
from the powdered rhizomes of turmeric (Curcuma longa) spice. curcumin remained dissolved in all the oil phases when stored at
Curcumin has been reported to have a number of potentially 25–60 °C for 24 h or more. The supernatant was then diluted to
beneficial biological activities, e.g., anti-tumour, anti-oxidant, an appropriate concentration to be analysed by a UV–VIS spectro-
anti-microbial and anti-inflammatory properties (Duvoix et al., photometer (Ultraspec 3000 Pro, GE Healthcare). A cuvette con-
2005). However, curcumin has an extremely low water solubility taining pure oil was used as a reference cell to zero the
(11 ng/ml), which makes it difficult to incorporate into many spectrophotometer. Wavelengths used, to obtain kmax, were depen-
food products (Tonnesen, Masson, & Loftsson, 2002), as well as dent upon oil type and ranged from 415 to 420 nm. A calibration
a low bioavailability, which means that its beneficial attributes curve was prepared for each oil phase by measuring the absor-
may not be realised even if it is ingested (Maiti, Mukherjee, Gan- bance as a function of curcumin concentrations. Results are re-
tait, Saha, & Mukherjee, 2007). Many attempts have therefore ported as the average of three measurements on freshly prepared
been made to improve the water solubility and bioavailability samples.
of curcumin. Studies have shown curcumin can be encapsulated
within complexes formed using either polysaccharides or phos- 2.3. Nanoemulsion and conventional emulsion preparation and
pholipids (Liu, Lou, Zhao, & Fan, 2006; Tonnesen et al., 2002). characterisation
Recent studies have shown that emulsion-based delivery systems
can also be used to encapsulate curcumin (Bisht et al., 2007; Oil-in-water conventional emulsions, or nanoemulsions,
Mukerjee & Vishwanatha, 2009; Wang et al., 2008; Yu & Huang, containing curcumin were prepared. Concentrations are expressed
2010). These studies indicate that emulsion-based systems can as weight percentage (wt.%), which is equal to 100 multiplied by
greatly increase the bioavailability of curcumin compared to the mass of the component divided by the total mass of the sample.
crystalline curcumin dispersed within water. An aqueous emulsifier solution containing 1 wt.% protein was pre-
In the present study we use an in vitro digestion model to sys- pared by dispersing b-lactoglobulin into PBS and stirring for at least
tematically examine some of the major factors that impact the bio- two hours to ensure complete hydration. The oil phase was prepared
accessibility of curcumin encapsulated within emulsion-based by adding 0.15 wt.% curcumin into the heated oil (60 °C), magneti-
delivery systems. In particular, we examined the influence of oil cally stirring for 10 min, and then sonicating for 20 min. Coarse
droplet size (emulsions versus nanoemulsions) and lipid phase emulsions were prepared by homogenising 10 wt.% oil phase, with
composition (triacylglycerol molecular weight) on lipid digestion 90 wt.% aqueous emulsifier solution, using a M133/1281 high-speed
and curcumin bioaccessibility. The knowledge gained from this blender (Biospec Products Inc., ESGC, Switzerland) for 2 min at room
study should facilitate the rational design of delivery systems for temperature. For some experiments these coarse emulsions were
enhancing the bioavailability of curcumin and other health- used as conventional emulsions, for other experiments they were
promoting lipophilic components. further homogenised to produce nanoemulsions. Nanoemulsions
were prepared by passing the coarse emulsions through a high pres-
sure homogeniser five times at 9000 psi (620 bars) (Microfluidics
2. Materials and methods
M-110Y Microfluidizer™, MFIC Corporation, Newton, MA, USA)
with a F20 Y 75 lm interaction chamber. After preparation, all
2.1. Materials
emulsions were stored at 4 °C.
The particle size distribution (PSD) of the nanoemulsions and
Curcumin (1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadi-
conventional emulsions was measured using dynamic light scat-
ene-3,5-dione) was used as a model lipophilic nutraceutical compo-
tering (Zetasizer Nano-ZS, Malvern Instruments, Worcestershire,
nent. Curcumin (95.2% pure, from Curcuma longa) was purchased
U.K.), with each individual measurement being the average of 12
from Bepharm Ltd. (Shanghai, China) and used without further
runs. Results were reported as the average diameter (Z-average)
purification. Powdered b-lactoglobulin (b-Lg) was provided by
and width (polydispersity index-PDI). Emulsions were diluted
Davisco Foods International Inc. (BioPURE, Le Sueur, MN). Lipid
prior to analysis using buffer solution (5 mM phosphate, pH 7.0)
phases with predominantly long, medium, and short chain triacyl-
to avoid the effects of multiple scattering.
glycerols (LCT, MCT, and SCT, respectively) were used. Corn oil (LCT)
was purchased from a local supermarket, MiglyolÒ 812 (MCT) was
2.4. In vitro digestion
purchased from SASOL (Houston, TX), and Tributyrin (98%) (SCT)
was purchased from Sigma–Aldrich (St. Louis, MO). All were used
The dynamic in vitro digestion model used was a modification of
without further purification. Lipase (from porcine pancreas pancre-
those described previously (Zangenberg et al., 2001a, 2001b; Mun,
atin) and bile extract (porcine) were also obtained from Sigma–
Decker, Park, Weiss, & McClements, 2006). Emulsion and buffer
Aldrich. Double deionized water was used in all experiments. A
solution (30.0 ml in total) were placed in a clean beaker, in a water
5 mM pH 7 phosphate buffer solution (PBS) was produced by dis-
bath, at 37.0 °C for 10 min and adjusted to pH 7 with NaOH solu-
persing 0.011% of anhydrous NaH2PO4 (Sigma–Aldrich, St. Louis,
tion (range from 0.05 to 1 M). Then, 4.0 ml of bile extract solution,
MO) and 0.058% of anhydrous Na2HPO4 (Fisher) in water.
containing 187.5 mg of bile extract (pH 7.0, PBS) and 1.0 ml of
CaCl2 solution, containing 110 mg of CaCl2 (pH 7.0, PBS), was added
2.2. Curcumin solubility into the 30 ml emulsion under stirring. The resultant mixture was
then adjusted to pH 7. Finally, 2.5 ml of freshly prepared pancrea-
The maximum solubility of curcumin in each lipid phase was tin suspension, containing 60 mg of lipase (pH 7.0, PBS), was added
determined using a spectrophometry method. Excess powdered to the mixture. The bile solution and lipase solution concentrations
curcumin was added to an oil phase heated to 60 °C, magnetically were 5.0 and 1.6 mg/ml, respectively. At this point, an automatic
stirred for 10 min, and then sonicated for 20 min. A temperature of titration experiment was started. The pH-stat (Metrohm, USA
60 °C was used based on previous studies of curcumin dissolution Inc.) was used to monitor and control the pH (at pH 7) of the diges-
in oils (Wang et al., 2008). The resulting crystalline slurry was cen- tion solution. The volume of added NaOH solution (500 mM) was
trifuged at 1750 rpm (1170g) using a bench top centrifuge (Model assumed to be equal to the amount of free fatty acids generated
 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807 801

by lipolysis of the initial triacylglycerols. Digestion experiments emulsion-based delivery system is mainly determined by the max-
were performed for 2 h. imum amount that can be incorporated into the lipid phase. Curcu-
The amount of free fatty acids released was calculated using the min has a relatively high melting temperature (183 °C) and is
following equations: therefore crystalline at ambient temperatures. Consequently, it
  must be dissolved within the lipid phase prior to making an oil-
ðmoil Þ ð1000Þ
V Max ¼ 2 ð1Þ in-water emulsion, since it is very difficult to homogenise lipid
ðMW oil Þ ðC NaOH Þ phases containing crystalline material, due to their tendency to
block homogeniser channels and promote droplet coalescence.
V Exp The amount of a highly lipophilic material that can be dissolved
%FFA Released ¼  100% ð2Þ
V Max in an oil phase depends on the molecular characteristics of the
Here, moil is the total mass of oil present in the reaction vessel (g), oil (e.g., molecular weight, polarity, and interactions). For this rea-
MWoil is the molecular weight of the oil (g per mol), CNaOH is the son, we initially measured the maximum amount of curcumin that
concentration of sodium hydroxide in the titration burette (mol could be solubilised within different lipid phases at 60 °C:
per 1000 cm3), and VMax is the volume of NaOH titrated into the LCT = 0.30 ± 0.10 wt.%; MCT = 0.79 ± 0.2 wt.%; SCT = 2.98 ± 0.18
reaction vessel to neutralise the FFA released assuming that all the wt.%; LCT:SCT = 1.94 ± 0.44 wt.%.
triacylglycerols are converted to two free fatty acids. Finally, VExp The maximum amount of curcumin that could be solubilised in-
is the actual volume of NaOH titrated into the reaction vessel to neu- creased as the average molecular weight of the carrier oil molecules
tralise the FFA released during the experiment. A relatively high decreased. This phenomenon can be attributed to differential inter-
calcium level (20 mM) was used in the simulated small intestine action and excluded volume effects (Huyskens & Haulaitpirson,
conditions to ensure that all of the lipid phases were completely 1985). Shorter chain triglycerides have more polar groups (oxygen)
digested within the experimental timeframe (Li & McClements, per unit mass than longer chain triglycerides, and hence there
2010). should be more dipole–dipole interactions between polar groups
on the carrier lipid and curcumin molecules for SCT, which may fa-
vour solubilisation. There is also an excluded volume effect associ-
2.5. Bioaccessibility determination
ated with incorporation of curcumin molecules into a lipid phase.
When a curcumin molecule is introduced into a lipid phase there
After in vitro digestion, emulsions were centrifuged (12,500 rpm,
will be a region around it from which the centre of the lipid mole-
29,700g), Sorvall RC 6C Plus, DuPont), at 25 °C for 30 min. The
cules are excluded. The concentration of lipid molecules in this ex-
emulsions separated into an opaque sediment phase at the bottom,
cluded zone is therefore effectively zero, whereas their
a clear micelle phase in the middle, and sometimes an oily or
concentration in the bulk lipid phase is equal to the number of lipid
creamed phase at the top. Aliquots (5 ml) of the micelle phase were
molecules per unit volume. The presence of an exclusion zone is
collected using a syringe, vortexed with 5 ml of chloroform, and then
thermodynamically unfavourable due to an osmotic effect associ-
centrifuged at 1750 rpm (1170g), at room temperature, for 10 min
ated with this concentration gradient. The thickness of the depletion
(Fisher Scientific 225A, Fisher). The bottom chloroform layer was
zone increases with increasing molecular weight of the lipid mole-
collected, while the top layer was vortexed with another 5 ml of
cules, hence this effect may be greater for LCT than SCT. The solubil-
chloroform, and centrifuged (1750 rpm, 1170g), at room tempera-
ity of the curcumin in the LCT:SCT mix was close to the average of its
ture, for 10 min. The second bottom chloroform layer was added
solubility in pure LCT and SCT. In other words, there was no signifi-
to the previously set aside chloroform layer, mixed, and analysed
cant synergistic or antagonistic effect on curcumin solubility by
by UV–VIS spectrophotometer (Ultrospec 3000 Pro, GE). A cuvette
mixing the two oils. Based on these results, we used 0.15 wt.% curcu-
containing pure chloroform was used as a reference cell to zero
min in the lipid phase to prepare all of the conventional emulsions
the spectrophotometer. The concentration of curcumin extracted
and nanoemulsions, so that they could be compared based on
from a sample was then determined from a previously prepared
similar, fully dissolved curcumin concentrations.
calibration curve of absorbance versus curcumin concentration in
chloroform.
3.2. Influence of oil type on nanoemulsion characteristics
2.6. Structural characterisation of micelle phase
A series of oil-in-water nanoemulsions was produced by micro-
fluidization that contained 10 wt.% lipid phase and 0.15 wt.% curcu-
The size of particles present in the micelle phase after digestion
min. The mean particle size (Z-average diameter) and polydispersity
was characterised by dynamic light scattering. After in vitro diges-
index (PDI) of these nanoemulsions are summarised in Table 1, and
tion, emulsions were centrifuged (12,500 rpm, 29,700g Sorvall RC
selected particle size distributions are shown in Fig. 1. The particle
6C Plus, DuPont) at 25 °C, for 30 min. Aliquots were taken from
size distributions of the MCT and LCT:SCT nanoemulsions were sim-
the micelle phase and were either filtered (0.45 lm), or left unfil-
ilar to that of the LCT nanoemulsions.
tered. The particle size distribution of the micelle phase was then
Nanoemulsions containing relatively small particles (d <
measured using dynamic light scattering (Zetasizer Nano-ZS,
200 nm) with narrow particle size distributions (PDI < 0.2) could
Malvern Instruments, Worcestershire, UK), with each individual
be formed when LCT or MCT were used as the lipid phase. On the
measurement being the average of 12 runs. The samples were di-
other hand, emulsions containing relatively large particles
luted prior to analysis using buffer solution (5 mM phosphate, pH
7.0) to avoid the effects of multiple scattering.
Table 1
Particle size characteristics of 10% oil-in-water nanoemulsions and emulsions
3. Results and discussion
containing 0.15% curcumin prepared using different lipid phases.

3.1. Curcumin solubility in oil phase Oil type Z-average (nm) PDI (nm)
LCT 181 ± 9 0.19 ± 0.02
Curcumin has a relatively low water-solubility (<0.005 wt.%) MCT 174 ± 2 0.16 ± 0
SCT 1981 ± 1061 0.75 ± 0.31
and a high oil–water partition coefficient (log P = 3.1), and so the
LCT:SCT 182 ± 0 0.13 ± 0.01
maximum amount of curcumin that can be incorporated into an
802 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807

(d > 1900 nm) with broad particle size distributions (PDI > 0.6) were on the kinetics of lipid digestion (Fig. 2). In general, the FFA release
formed when SCT was used as the lipid phase. After 24 h storage, the profiles followed a similar trend: an initial rapid increase in FFA
LCT and MCT nanoemulsions maintained a homogeneous optically during the first 2000 s, followed by a more gradual increase at
opaque yellowish appearance, whereas the SCT emulsions rapidly longer digestion times (Fig. 2a). The initial rate of FFA release ap-
separated into a yellow creamy or oily layer, on top of a transparent peared to be faster for the emulsions containing either SCT or
serum layer. These results showed that the SCT emulsion was highly LCT:SCT than for the emulsions containing either MCT or LCT
unstable to droplet growth and creaming. Recent research has (Fig. 2b). There are a number of possible explanations for this phe-
attributed this effect to Ostwald ripening because of the relatively nomenon. First, SCTs have some water solubility and therefore
high water solubility of the low molecular weight triacylglycerol some digestion may occur in the aqueous phase. Second, SCT mol-
(Li, Le Maux, Xiao, & McClements, 2009). Ostwald ripening is the ecules may protrude more into the aqueous phase than MCT or LCT
process where larger droplets grow at the expense of smaller ones molecules because of their short hydrocarbon chains, thereby facil-
due to molecular diffusion of oil through the aqueous phase. The itating the ability of lipase molecules to hydrolyse the ester bonds.
driving force for droplet growth is the increase in oil solubility in Third, the relatively small digestion products of SCTs may be able
the aqueous phase for droplets with high curvatures (small sizes). to rapidly leave the droplet surfaces.
The sample containing 50:50 LCT:SCT contained relatively small The final amount of FFAs produced also depended strongly on
lipid droplets (d < 200 nm) with a narrow particle size distribution lipid type: 68% for LCT; 90% for SCT and LCT:SCT; 113% for
(PDI < 0.2), which were stable to particle growth and gravitational
separation. This effect can be attributed to the fact that Ostwald
ripening can be inhibited in emulsions prepared from oils with rel-
a 120
atively high water solubility (such as SCT), by mixing them with
oils with relatively low water solubility (such as LCT) (Kabalnov
& Shchukin, 1992; Li et al., 2009). The physicochemical origin of 100
this effect is due to a compositional ripening process (which arises
due to an imbalance in molecular compositions of different drop-
lets) that counteracts the Ostwald ripening process (which arises 80
due to an imbalance in droplet curvatures).
% FFA Released

60
3.3. Influence of oil type & concentration on in vitro lipid digestion
SCT*
In this section, we quantified the influence of carrier oil type 40
and concentration on the rate and extent of lipid digestion using MCT*

an in vitro lipid digestion model. A series of oil-in-water nanoemul-

sions containing 0.15 wt.% curcumin in the lipid phase was pro- 20
LCT:SCT*

duced with different types and amounts of lipid. The kinetics of

lipid digestion was then monitored by measuring the amount of LCT

free fatty acids released over time after the nanoemulsions were 0
introduced into the simulated small intestinal fluid. 0 2000 4000 6000 8000
Lipid type: Initially, we compared the rate and extent of lipid Digestion Time (s)
digestion for emulsions prepared using different carrier oil phases,
at a constant carrier oil concentration (1 wt.% in the reaction
vessel). The nature of the carrier oil type clearly had a major impact
b 80

SCT*

14
MCT
60
LCT
SCT LCT:SCT*
12
% FFA Released

LCT
10
40
Volume (%)

6 20

0
2 0 200 400 600 800
Digestion Time (s)
0
10 100 1000 10000 Fig. 2. Effect of carrier lipid type on rate and extent of lipid digestion, measured
Particle Diameter (nm) using an in vitro digestion model that simulates small intestine conditions. A final
lipid concentration of 1 wt.% in the simulated small intestinal fluid was used.
Fig. 1. Particle size distributions for 10 wt.% oil-in-water emulsions, containing Results with ‘‘’’ were statistically different from the LCT samples. The reaction
10 wt.% lipid phase (LCT or SCT) and 0.15% curcumin. The PSD of MCT and LCT:SCT vessel contained 5 mg/ml bile, 1.6 mg/ml lipase, and 20 mmoles calcium. The
were similar to those for LCT. overall profile is shown in Fig. 2a and the initial states in Fig. 2b.
 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807 803

MCT. The relatively low amount of LCT digested may be due to the digestion was a measure of curcumin bioaccessibility. The trans-
fact that long chain fatty acids tend to accumulate at droplet parent micelle phase was collected after ultracentrifugation of
surfaces unless there is a physicochemical mechanism to remove the digested emulsion and extraction via chloroform.
them, such as bile salt solubilisation or calcium precipitation The bioaccessibility of curcumin clearly depended on the type
(Sek, Porter, Kaukonen, & Charman, 2002). The accumulation of and amount of carrier lipid present in the nanoemulsion-based
FFA at the droplet surface has previously been shown to inhibit li- delivery systems, ranging from as low as 1% to as high as 58% (Ta-
pase digestion of emulsified triglycerides. It is possible that there ble 2). The curcumin had a very low bioaccessibility (1%) when it
was insufficient bile salts or calcium present in the reaction vessel was encapsulated in an emulsion containing only SCT as the lipid
to remove all of the long chain FFA produced by digestion. Short or phase. This effect can be attributed to the fact that short chain fatty
medium chain FFAs are known to be more readily dispersible in acids, such as those produced by digestion of SCT, do not form
aqueous phases than long chain FFAs, and therefore do not tend mixed micelles that are capable of solubilising highly lipophilic
to accumulate at the droplet surfaces and inhibit lipid digestion components (Fatouros & Mullertz, 2008). Presumably, the curcu-
(Sek et al., 2002). This would account for the relatively high min precipitated from solution when there were no mixed micelles
amount of FFA digestion that occurred when the lipid phase was available to solubilise and transport them, thereby decreasing the
either SCT or MCT. Nevertheless, it does not account for the fact amount incorporated into the micelle phase. These results suggest
that the amount of FFA released by digestion was appreciably high- that utilisation of SCT in lipid-based delivery systems should be
er than 100% for the MCT sample. The calculation of the percentage avoided because any encapsulated components may not be
of FFA released from the lipid droplets during digestion assumes: bioavailable.
(i) that only two FFA molecules are released per TAG molecule; On the other hand, emulsions containing MCT or LCT within the
(ii) no other components release protons during the digestion pro- carrier lipid were able to substantially increase the bioaccessibility
cess. It is possible that some of the medium chain monoacylgylce- of curcumin, which can be attributed to their ability to form mixed
rols (MAGs) produced during digestion were broken down to FFAs micelles capable of solubilising highly lipophilic components. For
and glycerol, thereby meaning that more than two FFAs were pro- these systems, the bioaccessibility of curcumin tended to increase
duced per TAG. Alternatively, other components in the reaction as the total lipid concentration increased, due to the increase in
vessel may have produced protons during the digestion process, mixed micelles available to incorporate the curcumin. The only
such as phospholipids or proteins, although this effect should have exception to this trend was the emulsion containing LCT, where
been similar for all the lipid phases studied. Other researchers have the bioaccessibility of curcumin was similar at 1.5 and 2.0 wt.% li-
also reported that lipid digestion rates were substantially higher on pid (Table 2). This may have been because an appreciable amount
MCT, as compared to LCT emulsions (Deckelbaum et al., 1990). The of the LCT was not digested at the highest lipid concentration, and
samples containing SCT and LCT:SCT were almost fully digested hence some of the curcumin may not have been released from the
within the digestion period studied. Interestingly, the amount of droplets. In summary, we postulate that there are two competing
LCT:SCT digestion (90%) was not the average of the LCT digestion physiochemical mechanisms that determine the dependence of
(68%) and SCT digestion (90%), suggesting that the short chain curcumin bioaccessibility on lipid concentration. First, bioaccessi-
triglycerides might facilitate the digestion of the LCT. bility may increase with increasing lipid content, because more
Lipid concentration: The amount of lipid present in the small mixed micelles are formed to solubilise the released curcumin.
intestine may vary depending on the type and amount of food in- Second, the bioaccessibility may decrease with increasing lipid
gested. We therefore investigated the influence of lipid concentra- content (for LCT systems), because a greater fraction of lipid phase
tion on the rate and extent of lipid digestion. The full digestion remains non-digested, which means some of the curcumin is not
profiles for the emulsions containing LCT, MCT and SCT are shown released from the droplets into the surrounding micelle phase.
in Fig. 3. The amount of FFA released after 120 min of digestion is Overall, our results suggest that MCT is the most effective lipid
summarised for all of the samples in Table 2. phase for increasing the bioaccessibility of curcumin, whereas SCT
For the emulsions containing LCT, MCT and LCT:SCT there was a is the most ineffective lipid phase.
significant decrease in the amount of FFA produced as the total li-
pid concentration was increased from 1 to 2 wt.% (Fig. 3a and b and 3.5. Influence of oil droplet size on in vitro digestion and
Table 2). This effect may be attributed to a number of physico- bioaccessibility
chemical mechanisms: (i) as the lipid concentration increased,
the amount of lipase per unit surface area of oil droplets decreased; It has been suggested that the bioavailability of lipophilic com-
(ii) the amount of bile salts present may have been insufficient to ponents is greater in nanoemulsions than conventional emulsions
solubilise all of the digestion products (FFA and MAG) produced (Acosta, 2009). We therefore compared the in vitro digestion and
at higher lipid concentrations; (iii) at the higher lipid concentra- curcumin bioaccessibility of two delivery systems with the same
tions, the amount of calcium present may have been insufficient overall composition, but very different droplet sizes: nanoemul-
to precipitate all the FFAs produced by digestion and remove them sions (d < 200 nm) and conventional emulsions (d > 10 lm). It
from the droplet surfaces. For the emulsions containing SCT, there should be noted that the nanoemulsions prepared using SCT were
was no significant effect of total lipid concentration on the rate or highly unstable to droplet growth, so that the droplet size rapidly
extent of lipid digestion (Fig. 3c). This may have been because of increased above this 1 lm limit.
the much greater water dispersibility of the shorter chain fatty The rate and extent of FFAs released from emulsions containing
acids produced by digestion. Short chain FFAs can readily move 1%, 1.5% or 2% of LCT, MCT, SCT or LCT:SCT was measured for the
into the aqueous phase even in the absence of bile salts and nanoemulsions and conventional emulsions. Representative data
calcium. for the systems containing 1% MCT as the lipid phase is shown in
Fig. 4. The total amount of FFAs produced after 120 min of lipid
3.4. Influence of oil type & concentration on in vitro bioaccessibility digestion for all the nanoemulsions and conventional emulsions
studied are reported in Tables 2 and 3, respectively. We found no
In this section, we quantified the influence of carrier oil type major differences between the digestion profiles of the
and concentration on the bioaccessibility of curcumin measured nanoemulsions and conventional emulsions (Fig. 4). Previous stud-
using an in vitro digestion model. It was assumed that the fraction ies have found that the rate of lipid digestion increased when the
of the original curcumin that ended up in the micelle phase after droplet size decreased, which was attributed to the increase in the
804 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807

a b
140

100
120

80 100

% FFA Released
% FFA Released

80
60

60
40 1% 1%

40
1.5%* 1.5%

20
2%* 20 2%*

0 0
0 2000 4000 6000 8000 0 2000 4000 6000 8000
Digestion Time (s) Digestion Time (s)

c 120

100

80
% FFA Released

60

40
1%

1.5%
20
2%

0
0 2000 4000 6000 8000

Digestion Time (s)

Fig. 3. Effect of lipid concentration (1, 1.5 or 2 wt.%) on the rate and extent of lipid digestion, measured using an in vitro digestion method’’ (a) LCT; (b) MCT; (c) SCT. Results
with ‘‘’’ were statistically different from the 1% lipid samples. The reaction vessel contained 5 mg/ml bile, 1.6 mg/ml lipase, and 20 mmoles calcium.

Table 2
Influence of lipid type and concentration on the bioaccessibility of curcumin and the total amount of FFA released after 120 min of digestion of oil-in-water nanoemulsions. The
asterix () show samples that are significantly different from the samples containing 1% lipid. The initial mean particle diameters (Z-average) are also shown for each sample.

Lipid concentration (%) LCT (Z-av. = 181 ± 9 nm) MCT (Z-av. = 174 ± 2 nm) SCT (Z-av. = 1981 ± 1061 nm) LCT:SCT (Z-av. = 182 ± 0 nm)
Bioavailability (%) FFA (%) Bioavailability (%) FFA (%) Bioavailability (%) FFA (%) Bioavailability (%) FFA (%)
1 20 ± 10 92 ± 9 8±1 135 ± 9 1±0 93 ± 3 3±1 99 ± 8
1.5 40 ± 6⁄ 80 ± 9 20 ± 7⁄ 120 ± 6 1±1 97 ± 15 9 ± 3⁄ 91 ± 2
2 41 ± 4⁄ 68 ± 6 58 ± 6⁄ 113 ± 6 1±1 92 ± 8 20 ± 3⁄ 89 ± 3

surface area of the lipid phase exposed to the aqueous phase (Li & amount of lipase present (i.e., so not all of the surfaces can be satu-
McClements, 2010; Lundin & Golding, 2009). The reason that we rated with lipase), then increasing the surface area further may not
did not see a similar effect maybe because different experimental have a major impact on the digestion rate.
conditions were used: lipid concentrations; emulsifier type; and In general, the trends in the bioaccessibility of the curcumin were
simulated small intestinal fluid composition. For example, if the similar for the conventional emulsions (Table 3) as for the
lipid concentration is already relatively high compared to the nanoemulsions (Table 2): at the highest lipid concentration,
 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807 805

140

120

100
% FFA Released

80

60

40
Nanoemulsion
20
Emulsion
Fig. 5. Creaming stability of conventional emulsions (left) and nanoemulsions
0 (right) after 4 h storage at ambient temperature (before digestion).
0 1000 2000 3000 4000 5000 6000 7000 8000
Digestion Time (s)
conventional emulsion, to encapsulate a highly lipophilic compo-
Fig. 4. Comparison of in vitro digestion profiles of a nanoemulsion and a
conventional emulsion containing 1 wt.% MCT.
nent such as curcumin. Indeed, there may even be an advantage
in using larger droplet sizes in lipid-based delivery systems con-
taining LCT as the lipid phase in terms of maximum achievable bio-
accessibility. Nevertheless, there may be other advantages to using
curcumin bioaccessibility decreased in the following order MCT > nanoemulsions rather than conventional emulsions to encapsulate
LCT  SCT; curcumin bioaccessibility increased with increasing li- and deliver lipophilic functional components. During storage at
pid concentration, for emulsions containing MCT or LCT. Neverthe- ambient temperatures, conventional emulsions were observed to
less, there were some appreciable differences between the cream within a few hours, while nanoemulsions were much more
nanoemulsions and conventional emulsions. In particular, the cur- stable (Fig. 5). After 4 h storage, the conventional emulsion con-
cumin bioaccessibility appeared to be appreciably higher when sisted of a yellowish cream layer on top of a slightly turbid serum
LCT was used as the lipid phase in the conventional emulsions (Table layer, whereas the nanoemulsion consisted of a homogeneous opti-
3) than in the nanoemulsions (Table 2). This effect may have been cally opaque system.
because a greater fraction of the LCT appeared to have been digested Nanoemulsions containing LCT or MCT as the lipid phase were
in the conventional emulsions (Table 3) than in the nanoemulsions stored in a refrigerator (4 °C), for up to 10 days, without any notice-
(Table 2), leading to a greater release of curcumin into the aqueous able phase separation, i.e., droplet creaming or curcumin sedimen-
phase. The reason that a greater amount of lipid phase was digested tation. On the other hand, nanoemulsions containing SCT showed
in a conventional emulsion than in an equivalent nanoemulsion is phase separation within a few hours, which was due to the Ost-
currently unknown. It is possible that the interfacial layer of globular wald ripening instability mechanism discussed earlier. The stabil-
proteins (b-lactoglobulin) that initially coated the lipid droplets had ity of the SCT nanoemulsions could be greatly improved by
some impact on lipid digestion or release (Lee & McClements, 2010). incorporating 50% LCT into the lipid phase as described in an ear-
The protein layer may have inhibited the ability of lipase to come lier section.
into close contact with the lipids, or it may have inhibited the release
of the digested lipids. Globular proteins tend to form a two-dimen-
sional network of cross-linked molecules at oil–water interfaces. As 3.6. Influence of lipid type & concentration on micelle phase structure
digestion proceeds, this protein coating may have collapsed and
formed a thick crinkled layer that trapped any remaining lipid, or li- Finally, we examined the influence of lipid type and concentra-
pid digestion products, inside (Lee & McClements, 2010). This effect tion on the structure of the micelle phase formed at the end of the
would be more important for LCT than the other lipids because they lipid digestion process. The particle size distribution of the micelle
are least dispersible in water. phase was measured using dynamic light scattering, and the mean
Previous studies have also found that the bioaccessibility of lipo- particle size (Z-average diameter) and polydispersity index (PDI)
philic substances in lipid-based delivery systems is more closely are reported in Table 4. The mean particle diameters were rela-
correlated to the solubilisation capacity of the lipid digestion prod- tively large (d > 100 nm) and the distributions were relatively
ucts than to the initial droplet size (Porter, Kaukonen, Boyd, Ed- broad (PDI > 0.2). Particles ranging in diameter from 10 to over
wards, & Charman, 2004; Sek, Boyd, Charman, & Porter, 2006). 1000 nm were observed in the particle size distributions, which
Our results suggest that there is no advantage, from a bioaccessibil- suggested that the micelle phase contained a mixture of micelles
ity point of view, in using a nanoemulsion, rather than a and vesicles. The relatively large mean diameters can be attributed

Table 3
Influence of lipid type and concentration on the bioaccessibility of curcumin and the total amount of FFA released after 120 min of digestion of oil-in-water coarse emulsions. The
asterix () show samples that are significantly different from the samples containing 1% lipid. The initial mean particle diameters (d43) are also shown for each sample.

Lipid concentration (%) LCT (d43 = 20 ± 11 lm) MCT (d43 = 22 ± 6 lm) SCT (d43 = 14 ± 2 lm) LCT:SCT (d43 = 19 ± 5 lm)
Bioavailability (%) FFA (%) Bioavailability (%) FFA (%) Bioavailability (%) FFA (%) Bioavailability (%) FFA (%)
1 38 ± 22 98 ± 16 3±0 139 ± 9 0±0 110 ± 6 5±3 104 ± 4
1.5 56 ± 9 88 ± 8 17 ± 6 132 ± 13 0±0 110 ± 3 20 ± 6⁄ 103 ± 5
2 50 ± 10 73 ± 10 59 ± 5⁄ 123 ± 8 0±0 107 ± 4 45 ± 7⁄ 103 ± 4
806 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807

Table 4
Particle size characteristics of the structures in the micelle phase collected from digested nanoemulsions prepared with different lipid types and concentrations.

Oil type 1% Lipid 1.5% Lipid 2% Lipid

Mean diameter (nm) PDI Mean diameter (nm) PDI Mean diameter (nm) PDI
LCT 99 ± 14 0.26 ± 0.03 113 ± 5 0.37 ± 0.18 161 ± 10 0.42 ± 0.09
MCT 161 ± 39 0.51 ± 0.08 121 ± 13 0.20 ± 0.01 68 ± 1 0.30 ± 0.04
LCT:SCT – – 131 ± 19 0.26 ± 0.04 120 ± 13 0.29 ± 0
SCT – – – – – –

to the fact that larger particles scatter light much more effectively References
than small ones, and therefore tend to dominate the signal in dy-
namic light scattering experiments. Acosta, E. (2009). Bioavailability of nanoparticles in nutrient and nutraceutical
delivery. Current Opinion in Colloid & Interface Science, 14(1), 3–15.
As well as forming mixed micelles, bile salts, phospholipids, and Bisht, S., Feldmann, G., Soni, S., Ravi, R., Karikar, C., & Maitra, A. (2007). Polymeric
lipid digestion products may assemble into various other types of nanoparticle-encapsulated curcumin (‘‘nanocurcumin’’): A novel strategy for
association colloids during the lipid digestion process, including human cancer therapy. Journal of Nanobiotechnology, 5, 3.
Deckelbaum, R. J., Hamilton, J. A., Moser, A., Bengtssonolivecrona, G., Butbul, E.,
vesicles and lamellar structures. The concentration and structure Carpentier, Y. A., et al. (1990). Medium-chain versus long-chain triacylglycerol
of these association colloids is likely to change throughout the emulsion hydrolysis by lipoprotein-lipase and hepatic lipase – implications for
digestion process and to depend on the molecular characteristics the mechanisms of lipase action. Biochemistry, 29(5), 1136–1142.
Duvoix, A., Blasius, R., Delhalle, S., Schnekenburger, M., Morceau, F., Henry, E., et al.
of the carrier oil, which may affect their ability to incorporate lipo-
(2005). Chemopreventive and therapeutic effects of curcumin. Cancer Letters,
philic bioactive ingredients such as curcumin. In future studies, it 223(2), 181–190.
therefore would be useful to characterise the concentration and Fatouros, D. G., & Mullertz, A. (2008). In vitro lipid digestion models in design of
structure of the various association colloids formed throughout drug delivery systems for enhancing oral bioavailability. Expert Opinion on Drug
Metabolism & Toxicology, 4(1), 65–76.
the lipid digestion process in more detail, e.g., using electron Fricker, G., Kromp, T., Wendel, A., Blume, A., Zirkel, J., Rebmann, H., et al. (2010).
microscopy or X-ray diffraction methods. Phospholipids and lipid-based formulations in oral drug delivery. Pharm Res.
Huyskens, P. L., & Haulaitpirson, M. C. (1985). A new expression for the
combinatorial entropy of mixing in liquid-mixtures. Journal of Molecular
4. Conclusions Liquids, 31(3), 135–151.
Kabalnov, A. S., & Shchukin, E. D. (1992). Ostwald ripening theory – applications to
fluorocarbon emulsion stability. Advances in Colloid and Interface Science, 38,
This study has shown that physically stable curcumin-loaded
69–97.
nanoemulsions can be prepared using a variety of different lipid Lee, S. J., & McClements, D. J. (2010). Fabrication of protein-stabilized nanoemulsions
phases: LCT; MCT; LCT:SCT. On the other hand, stable nanoemul- using a combined homogenization and amphiphilic solvent dissolution/
evaporation approach. Food Hydrocolloids, 24(6–7), 560–569.
sions could not be prepared using pure SCT because of Ostwald rip-
Li, Y., Le Maux, S., Xiao, H., & McClements, D. J. (2009). Emulsion-based delivery
ening effects associated with the relatively high solubility of this systems for tributyrin, a potential colon cancer preventative agent. Journal of
lipid in water. The maximum amount of curcumin that could be Agricultural and Food Chemistry, 57(19), 9243–9249.
loaded into the lipid phase was shown to decrease as the molecular Li, Y., & McClements, D. J. (2010). New mathematical model for interpreting pH-stat
digestion profiles: Impact of lipid droplet characteristics on in vitro
weight of the lipid phase increased: SCT > MCT > LCT. The rate and digestibility. Journal of Agricultural and Food Chemistry, 58(13), 8085–8092.
extent of lipid digestion also depended on lipid type. The initial Liu, A. C., Lou, H. X., Zhao, L. X., & Fan, P. H. (2006). Validated LC/MS/MS assay for
rate decreased in the following order SCT > MCT > LCT, while the curcumin and tetrahydrocurcumin in rat plasma and application to
pharmacokinetic study of phospholipid complex of curcumin. Journal of
total amount of FFA produced at the end of digestion decreased Pharmaceutical and Biomedical Analysis, 40(3), 720–727.
in the following order MCT > SCT > LCT. The bioaccessibility of cur- Lundin, L., & Golding, M. (2009). Structure design for healthy food. Australian Journal
cumin, assumed to be equal to the fraction of curcumin incorpo- of Dairy Technology, 64(1), 68–74.
Maiti, K., Mukherjee, K., Gantait, A., Saha, B. P., & Mukherjee, P. K. (2007). Curcumin-
rated into the micelle phase after digestion, decreased in the phospholipid complex: Preparation, therapeutic evaluation and pharmacokinetic
following order MCT > LCT  SCT. The bioaccessibility of curcumin study in rats. International Journal of Pharmaceutics, 330(1–2), 155–163.
appeared to be slightly higher in conventional emulsions than in McClements, D. J., & Li, Y. (2010). Structured emulsion-based delivery systems:
Controlling the digestion and release of lipophilic food components. Advances in
nanoemulsions, but nanoemulsions had much better physical sta-
Colloid and Interface Science, 159(2), 213–228.
bility. The bioaccessibility results are interpreted in terms of the Mukerjee, A., & Vishwanatha, J. K. (2009). Formulation, characterization and
influence of lipid type and concentration on: (i) the formation of evaluation of curcumin-loaded PLGA nanospheres for cancer therapy.
Anticancer Research, 29(10), 3867–3875.
mixed micelles available to solubilise the digestion products
Mun, S., Decker, E. A., Park, Y., Weiss, J., & McClements, D. J. (2006). Influence of
formed – an increase in the number of mixed micelles leads to interfacial composition on in vitro digestibility of emulsified lipids: Potential
an increase in bioaccessibility; (ii) the amount of undigested lipid mechanism for chitosan’s ability to inhibit fat digestion. Food Biophysics, 1(1),
remaining after the digestion period – an increase in the amount 21–29.
Patel, D., & Sawant, K. K. (2007). Oral bioavailability enhancement of acyclovir by
of undigested lipid leads to a decrease in bioaccessibility. These re- self-microemulsifying drug delivery systems (SMEDDS). Drug Development and
sults have important implications for the design and fabrication of Industrial Pharmacy, 33(12), 1318–1326.
delivery systems to encapsulate and release highly lipophilic func- Porter, C. J. H., Kaukonen, A. M., Boyd, B. J., Edwards, G. A., & Charman, W. N. (2004).
Susceptibility to lipase-mediated digestion reduces the oral bioavailability of
tional ingredients. danazol after administration as a medium-chain lipid-based microemulsion
formulation. Pharmaceutical Research, 21(8), 1405–1412.
Acknowledgement Porter, C. J. H., Trevaskis, N. L., & Charman, W. N. (2007). Lipids and lipid-based
formulations: Optimizing the oral delivery of lipophilic drugs. Nature Reviews
Drug Discovery, 6(3), 231–248.
This work was supported in part by an EPA-NSF-NIFA (AFRI) Pouton, C. W. (2006). Formulation of poorly water-soluble drugs for oral
joint Grant (2010-05266) program, NIH Grant CA139174, a special administration: Physicochemical and physiological issues and the lipid
formulation classification system. European Journal of Pharmaceutical Sciences,
call Grant from Massachusetts Center for Agriculture, and a CVIP
29(3–4), 278–287.
Grant from the University of Massachusetts Amherst. The authors Sek, L., Boyd, B. J., Charman, W. N., & Porter, C. J. H. (2006). Examination of the
have declared no conflict of interest. impact of a range of pluronic surfactants on the in-vitro solubilisation
 K. Ahmed et al. / Food Chemistry 132 (2012) 799–807 807

behaviour and oral bioavailability of lipiclic formulations of atovaquone. Journal Wang, X. Y., Jiang, Y., Wang, Y. W., Huang, M. T., Ho, C. T., & Huang, Q. R. (2008).
of Pharmacy and Pharmacology, 58(6), 809–820. Enhancing anti-inflammation activity of curcumin through O/W
Sek, L., Porter, C. J. H., Kaukonen, A. M., & Charman, W. N. (2002). Evaluation of the nanoemulsions. Food Chemistry, 108(2), 419–424.
in-vitro digestion profiles of long and medium chain glycerides and the phase Yu, H. L., & Huang, Q. R. (2010). Enhanced in vitro anti-cancer activity of curcumin
behaviour of their lipolytic products. Journal of Pharmacy and Pharmacology, encapsulated in hydrophobically modified starch. Food Chemistry, 119(2),
54(1), 29–41. 669–674.
Solans, C., Izquierdo, P., Nolla, J., Azemar, N., & Garcia-Celma, M. J. (2005). Nano- Zangenberg, N. H., Mullertz, A., Kristensen, H. G., & Hovgaard, L. (2001a). A dynamic
emulsions. Current Opinion in Colloid & Interface Science, 10(3–4), 102–110. in vitro lipolysis model I. Controlling the rate of lipolysis by continuous addition
Sonneville-Aubrun, O., Simonnet, J. T., & L’Alloret, F. (2004). Nanoemulsions: A new of calcium. European Journal of Pharmaceutical Sciences, 14(2), 115–122.
vehicle for skincare products. Advances in Colloid and Interface Science, 108–09, Zangenberg, N. H., Mullertz, A., Kristensen, H. G., & Hovgaard, L. (2001b). A dynamic
145–149. in vitro lipolysis model II: Evaluation of the model. European Journal of
Tonnesen, H. H., Masson, M., & Loftsson, T. (2002). Studies of curcumin and Pharmaceutical Sciences, 14(3), 237–244.
curcuminoids. XXVII. Cyclodextrin complexation: Solubility, chemical and
photochemical stability. International Journal of Pharmaceutics, 244(1–2),
127–135.
 Current Opinion in Colloid & Interface Science 9 (2004) 264 – 278
www.elsevier.com/locate/cocis

y
Microemulsion dynamics and reactions in microemulsions
M.A. López-Quintela a,*, C. Tojo b, M.C. Blanco a, L. Garcı́a Rio a, J.R. Leis a
a
Department of Physical Chemistry, University of Santiago de Compostela, E-15782 Santiago de Compostela, Spain
b
Faculty of Sciences, University of Vigo, E-36200 Vigo, Spain

Available online 11 August 2004

Abstract

Microemulsions are very versatile reaction media which nowadays find many applications, ranging from nanoparticle templating to
preparative organic chemistry. On one hand, for the synthesis of nanomaterials microemulsions represent a well-established technique that
can be used to fine control the particle size of many inorganic and organic materials, as well as latexes. On the other hand, the
thermodynamical stable and microheterogeneous nature of microemulsions, used as reaction media, induces drastic changes in the reagent
concentrations, and this can be specifically used for tuning the reaction rates. In particular, amphiphilic organic molecules can accumulate and
orient at the oil – water interface inducing regiospecificity in organic reactions. In this review, we will show the recent tendencies of the use of
microemulsions for the preparation of nanoparticles, and also as particularly interesting organic reaction media.
D 2004 Elsevier Ltd. All rights reserved.

Keywords: Microemulsions; Inverse micelles; Nanomaterials; Latexes; Core – shell nanoparticles; Microheterogeneous reaction media; Pseudo-phase model;
Water – oil interface properties; Catalytic properties

1. Introduction microemulsion. Understanding how the chemical reaction

proceeds depends on the ratio between the characteristic
Microemulsions have been used as chemical reactors chemical reaction time, sr, and sex. When the reaction is
because of their special interfacial properties allowing an very slow in comparison with the microemulsion dynamics,
intimate contact, at nanoscale level, of hydrophilic and sr/sexH1, the reaction ‘‘sees’’ the microemulsion as a static
hydrophobic domains. The dynamic character of these object, and a pseudo-phase model can be applied. In
nano-reactors is one of the most important features, which contrast, for chemical reactions with sr/sex< or c1, the
has to be taken into account for a comprehensive under- dynamics of the microemulsion has to be taken into account
standing of chemical reactions carried out in these media. to explain the chemical reaction.
However, this important fact is usually overlooked. Scheme
1 shows a picture that is useful to understand the role that
the microemulsion dynamics can have on the chemical 2. Microemulsion dynamics
reaction.
By microemulsion dynamics, we mean the fact that the If one assumes that microemulsion domains are formed
domains are not static ones, but are in continuous move- by spherical droplets, the characteristic droplet’s collision
ment and collision with each other. In each collision, time in microemulsions can be easily calculated assuming
material interchange can takes place. The whole process that the droplets diffuse through a continuous medium
of motion – collision exchange can be characterised through with viscosity, g. Then, the collision rate constant is given
a parameter sex, which is characteristic of each kind of by kD=(8/3)kBT/gc109 M1 s1 for a typical low vis-
cosity solvent. Because usually the droplet’s volume
fraction Uc0.1, i.e., [droplet]c103 M for a typical
y
droplet’s size c10 nm, the encounter rate constant kenc106
Madrid, March 11th, 2004: En memoria de todas las vı́ctimas s1. Thus, the average collision time (encounter time) is
inocentes de éste y todos los demás actos de violencia (In memory of all the
innocent victims of this and all other violent actions). senc1 As.
* Corresponding author. Tel.: +34-981-595998; fax: +34-981-595012. It is well known that not all droplets’ collisions are
E-mail address: qfarturo@usc.es (M.A. López-Quintela). effective for material exchange. This can be taken into

1359-0294/$ - see front matter D 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cocis.2004.05.029
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 265

account introducing an encounter rate factor, c, which for a 3. Formation of particles in microemulsions [3]
particular material (reactant), depends on the film flexibil-
ity [1], i.e., kex=cken. For rigid films like AOT micro- 3.1. Ionic surfactants
emulsions, cc10 3 —that is, only 1 in each 1000
collisions is effective for the reactants’ exchange [2]. 3.1.1. AOT microemulsions
However, for flexible films, this value can reach up to Sodium bis(2-ethylhexyl) sulfosuccinate is an anionic
cc101 [1]. Then the microemulsion exchange character- surfactant commonly known as AOT. When a small
istic time sex is in the range c10 As<sex<1 ms. amount of AOT is dissolved in organic solvents, thermo-
According to Scheme 1, reactions with a half-life dynamically stable reverse micelles are formed. These
reaction time, sr, much larger than 0.01 –1 ms are as- micelles consist of a hydrophilic core compartmentalised
sumed to occur in a ‘‘static’’ pseudo-phase system, in by the hydrophilic head group of the AOT, and with the
which the exchange of material does not play any role in hydrophobic alkyl tails extending into the nonpolar con-
the kinetics (region III, Scheme 1). However, for reactions tinuous phase solvent.
with a half-life reaction time (sr) far below 0.01 –1 ms,
the droplet’s exchange is the control factor of the kinetics 3.1.1.1. Growth versus stabilising mechanisms. The size
(I). This occurs only for reactions that are near diffusion- of the micelle core is described by the molar ratio of water
controlled. The interplay of droplet’s exchange and chem- to surfactant molecules in solution, W=[water]/[surtactant].
ical reaction has to be taken into account for reactions For materials such as CdS, ZnS and AgCl, it has been
with sr near sex (II). Therefore, to understand chemical observed that the particle size is controlled by the size of
reactions in microemulsions, there is a need to know the micelle, i.e., reverse micelles act as a template.
which model is most suitable for a particular reaction. However, time-resolved studies on the formation of Cu
However, there are not many available data for this nanoparticles by Cason et al. [4..] have shown that the
classification, and any study in this direction will be of final particle size in AOT/alkane micelles is independent
valuable importance. Because film flexibility is mainly of W, although the particle growth rate is a function of W
determined by the surfactant, it is convenient to classify and the bulk solvent type. This leads us to propose an
reactions located in regions I and II based on the alternate view: that the sizes are largely controlled by
surfactants being used. This will restrict the sex values, solvent stabilisation of the particles, and the surfactant
corresponding to each surfactant, to a small range that, in acting as a stabilising ligand [5]. This argument has also
turn, depends on other variables (such as cosurfactant, oil, been proposed by Kitchens et al. [6], in their study of the
droplet’s size) and the specific reactant to be exchanged. solvent effects on the growth rate and final particle size of
A more direct comparison of the different experimental copper metallic nanoparticles. Their experimental results
results can then be achieved. support the assumption that the surfactant has a double
To simplify this revision, we will first refer to reactions influence: on the particle growth and the stabilisation
leading to the formation of nanoparticles in Section 3. process. Initially, the surfactant provides an initiation site,
Because of the possible influence of the microemulsion the micelle core, for the reduction of the metal followed by
dynamics, we will make a classification of reactions particle growth through intermicellar exchange. At the
according to the surfactants being used. In Section 4, we latter end of the particle growth, the surfactant acts as a
concentrate on other types of reactions, which can be stabilising ligand with a weak interaction between the
explained by a pseudo-phase model, and therefore will metal particle and the surfactant head group. The authors
be classified according to the type of the reaction. concluded that growth rate and particle size are inversely

Scheme 1. Reaction time (sr) versus microemulsion dynamics (sex).

266 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

related, where a decrease in growth rate corresponds to a phobic chain and introduces a discontinuity in the interfacial
larger particle size at a specific W value. Increased film of the water pool, thus promoting the intermicellar
interaction between the solvent and the surfactant tails exchange and leading to smaller though more numerous
results in a more stable micelle system and enhanced particles. Indeed, a higher intermicellar exchange implies a
ability to stabilise larger particles while reducing the higher consumption of the reactants during the nucleation
intermicellar exchange. In contrast, the micelle core tem- stage, which means that less are left for the growth stage.
plating effect, observed in CdS, ZnS and AgCl, is Similar results were found by Bagwe and Khilar [15], who
accounted for because these nanoparticles are not pure used NP-5 as a cosurfactant to synthesise Ag particles in
metals and it is possible that a weaker interaction between AOT/n-heptane/water.
the particle surface and the ionic headgroup of the surfac-
tant exists, thus inhibiting the steric stabilization of par- 3.1.1.3. Supercritical CO2 microemulsions. Increasing at-
ticles larger than the micelle core. In each of these cases, tention has been paid to the synthesis of nanoparticles in
the obtained particle sizes rarely exceeded the size of the water-in-supercritical CO2 microemulsions [16 –18]. One
micelle core. The surfactant-stabilized nanoparticle droplets problem of using conventional water-in-oil microemulsions
do not only act as microreactors, but also inhibit the for nanoparticle synthesis is the separation and removal of
aggregation of particles, because the surfactants can be solvent from products. Supercritical carbon dioxide used as
adsorbed onto the particle surface when the particle size a solvent offers several advantages such as fast reaction
approaches the water pool size (for capping, see, e.g. Ref. speed, rapid separation and easy removal of solvent from
[3]). Manna et al. [7] reported the possibility that the nanoparticles. This method has been used to produce Ag
strong interactions between Au and SH groups of AOT and Cu [16] and CdS and ZnS [18] nanoparticles. Hydro-
help to form a ‘‘three-dimensional self-assembled mono- genation of olefins catalysed by Pd nanoparticles in a
layer’’ onto the particle surface, which controls the growth water-in-CO2 microemulsion has also been reported by
and stabilisation of the nanocrystals. Ohde et al. [17].
Different stabilising agents have been used to inhibit the The selectivity coefficients for the counterion exchange
nanoparticle growth in AOT microemulsions. As an example, in the water – AOT – heptane microemulsion interface were
trioctyl phosphine oxide (TOPO) has been used to stabilize determined by using a pseudo-phase ion exchange formal-
iron nanoparticles [8], and dodecanthiol [9] and bis(2-ethyl- ism [19]. Theoretical results have been successfully com-
hexyl)amine (BEA) [10] were likewise used to stabilize pared to quenching of the RuL34 luminescence emission
cadmium sulfide nanoparticles. The stabilization mechanism measurements.
consists in the rapid formation of a chemically bonded layer Finally, the catalytic activity of metallic particles syn-
of oriented molecules on the nanoparticle surface. Formation thesised in AOT microemulsions has been a field of high
of a thin layer of SiO2 on different nanoclusters offers a new activity [8,17,20 –22]—see Section 4.3.
challenge in synthesis. By changing the thickness of the shell
and the particle radius, the overlap of the wave functions and 3.1.1.4. Preparation of bimetallic particles. Chen et al.
band gap can be changed, which represents a major interest in [23. – 25] have reported the synthesis of Au – Ag [23.], Au –
the semiconductor field. Therefore, silica-coated nanopar- Pd [24] and Pd – Pt [25] bimetallic nanoparticles in water/
ticles have attracted the attention of many researchers in AOT/isooctane microemulsions. The following mechanism
recent years. As an example, the size of Ag particles and the is proposed: for Au – Ag [23.] and Au – Pd [24], the reduc-
thickness of the coating can be controlled by manipulating the tion rate is so large that almost all of the ions are reduced
relative rates of the hydrolysis and condensation reactions of before the formation of nuclei. Then, the atoms start to
tetraethoxysilane [11] within the microemulsion. Silica nano- aggregate to form the nuclei. Since the nucleation rate of Au
particles have also been used to cover a ZnFe2O4 magnetic is much faster than that of Ag [23.] or Pd [24], the nuclei of
core [12..] and rhodium nanoparticles [13]. the bimetallic system should be mainly formed from Au
atoms, and the composition of the nuclei might have a
3.1.1.2. Cosurfactant. Marchand et al. [14.] have studied higher Au concentration than that of the feeding solution,
the synthesis of MoSx particles in AOT/water/n-heptane i.e., Au might act as the seed for the formation of the
microemulsions. The addition of NP-5, as a nonionic bimetallic particle. All nuclei might be formed almost at the
cosurfactant, at a small concentration compared to that of same time. After that, Ag or Pd atoms codeposite onto the
AOT, leads to a substantial decrease of the mean micellar nuclei and grow to their final sizes. The faster growth rate of
size, resulting in a significant decrease of the nanoparticle Au than Ag or Pd leads to the enrichment of Ag or Pd in the
size. This can be attributed to a higher fluidity of the outer layer of the bimetallic nanoparticle. For the case of
interfacial film and a higher mean curvature of the droplets. Pd – Pt [25], the formation rate of Pd nanoparticles is faster
Therefore, the cosurfactant increases the fluidity of the than Pt nanoparticles, but the difference is less than Au –Ag
interface and thus the kinetics of the intermicellar exchange, and Au – Pd particles. Consequently, the nucleus might
which in turns ensures a more homogeneous repartition of contain both Pd and Pt codeposited at a similar deposition
reactants among droplets. NP-5 has a cycloalkane hydro- rate, so that a homogeneous alloy structure is obtained. It is
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 267

interesting to note that the same Pd – Pt synthesis using a of the reduction. Atoms formed at a latter period will collide
different reduction agent, i.e., different chemical reaction, with the nuclei already formed, and larger particles are
leads to a different final nanoparticle size. obtained. As hydrazine concentration increases, the en-
hanced reduction rate favours the generation of considerably
3.1.2. CTAB microemulsions higher number of nuclei, and leads to smaller Ni nano-
Cetyl trimethylammonium bromide (C16H33 – (CH3)3 – particles. When the concentration ratio of hydrazine to NiCl
N+Br, CTAB) provides a very flexible film, which gives is large enough, the reduction is much faster than the
rise to a high exchange dynamic of the micelles. In recent nucleation. The nucleation rate is not further raised and
years, an increasing number of works used CTAB as the number of nuclei is held constant with the increase of
surfactant because of CTAB reverse micellar systems, hydrazine concentration. Then, the size of the nanoparticles
which show remarkably more solubilization capacity of remains constant. The effect of NiCl concentration on final
high concentration aqueous salt solution than AOT-based size was also investigated [29.]. In contrast to previous
systems. results [26,28], Ni size increases as the NiCl concentration is
Using water-CTAB-n-octane microemulsions, with 1- increased. However, experimental conditions are hardly
butanol as cosurfactant, Porta et al. [26] obtained smaller comparable because a hydrazine excess is used.
Au nanoparticles as the reactant concentration increases. ZrO2 – Y2O3 nanoparticles has been obtained in a CTAB/
They discussed the possibility that alcohol behaves as hexanol/water microemulsion by Fang and Yang [30]. In their
capping agent at high concentrations [27]. As expected, report, these authors noted a peculiar behaviour: nanoparticle
higher pentanol contents favours exchange dynamics and size distribution is narrowed down in two cases, increasing
leads to larger particles, even larger that the original droplet water content at fixed surfactant concentration or decreasing
diameter. Husein et al. [28] obtained similar results for the the surfactant content at fixed W. Both cases led to larger
synthesis of silver chloride in water – dioctyldimethylammo- droplets. The greater number of metallic ions existing in a
nium chloride –n-decanol-isooctane. They explained these large droplet would cause more nuclei to form in a reverse
results assuming that, at high alcohol content, the particle micelle. When a water pool, containing free metallic atoms,
size increases due to decreasing of the interactions between fuses with such a reverse micelle, these nuclei will grow up at
the nanoparticles and the stabilising surfactant layer. In the same speed. As the droplet sizes increases (because of the
addition, these authors also found that the particle size increase of the water content), the surfactant film becomes
decreases as the concentration of AgNO3 increases. This thinner, thus accelerating the exchange process. The high
result is explained, assuming that smaller particles are exchange rate could lead to a uniform nucleation and growth
formed when there is a larger number of nuclei. At a fixed process. For water pools with smaller sizes, nucleation only
surfactant concentration and a fixed W, increasing the occurs in a little number of micelles at the very beginning of
amount of reactant reduces the Ag+ ion occupancy number. the precipitation reaction, because most of them do not
More droplets carrying a AgCl concentration higher than the contain enough metal ions to form a critical nucleus. As a
critical nucleation concentration are formed, and the rate of result, due to diffusion, new nuclei will form as a function of
nucleation becomes less dependent on the intermicellar time. Particles already existing and newly emerging will grow
exchange of solubilizate. The larger number of nuclei at a different rate, causing a broad size distribution.
provides more seeds for particle growth and results in This kind of microemulsions have also been used to
particles with smaller diameter. produce other materials: spinel ZnAl2O4 [31], perovskite
Chen and Wu [29.] studied the influence of both reac- LaMnO3 [31], bioceramic hydroxyapatite [32], cerium oxide
tants, metal salt and reducing agent, on the final Ni nano- [33], and coated materials like, SiO2 coated Pt, Pd and Pt/
particle size in a water/CTAB/n-hexanol microemulsion. At Ag [34]. Finally, the catalytic activity of nanoparticles
a constant NiCl concentration, the size of Ni nanoparticles synthesised in CTAB microemulsions was also studied [34].
decreases with the increase of hydrazine concentration and
then approaches to a constant value. This fact can be 3.1.3. PFPE microemulsions
explained from the influence of reduction rate on the The anionic surfactant perfluoropolyethercarboxylic acid
nucleation. Collisions between several atoms must occur was converted to its ammonium salt by reaction with excess
for a nucleation—with this probability at a much lower rate ammonium hydroxide. PFPE – NH4 has an average formula
than the probability for collisions between one atom and a of [CF3O(CF2CF(CF3)O)f3CF2COO][NH4]+.
nucleus already formed. Once the nuclei are formed, the This surfactant was recently used in water-in-carbon
growth would be superior to nucleation. In addition, if all of dioxide microemulsions [35 –38] to produce Ag nanopar-
the nuclei were formed almost at the same time and grew at ticles [35] and Ti [36].
the same rate, nanoparticles would be monodisperse. Thus,
the number of nuclei formed at the very beginning deter- 3.1.4. Sodium lauryl sulfate
mines the number and size distribution of the obtained Sodium lauryl sulfate (SLS) was used by Wang et al. [39]
particles. At low hydrazine concentration, the reduction rate to synthesise PbS nanoparticles by a sonochemical method.
is slow, and only few nuclei are formed at the initial period A probable mechanism for the formation of nanocrystalline
268 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

PbS particles in a toluene-in-water microemulsion with the Finally, the catalytic activity of metallic particles syn-
inducement of ultrasound irradiation is proposed. thesised in Triton X microemulsions has also been reported
[44,46,47].
3.2. Nonionic surfactants
3.2.2. NP-5, NP-9 and NP-12 microemulsions
Nonionic surfactants are very common in the literature, Poly(oxyethylene)5 nonylphenol ether (NP-5), poly(oxy-
and are mainly used to prevent possible counterion ethylene)9 nonylphenol ether (NP-9) and poly(oxyethy-
interactions. lene)12 nonylphenol ether (NP-12) have been recently
used to prepare different particles, such as hydroxyapatite
3.2.1. Triton X microemulsions [50], metallic bismuth [51], and also applied in catalytic
Triton X-100 [Polyoxyethylene(9)4-(1,1,3,3-tetramethyl- activity studies [52,53].
butyl)phenyl ether] has been used to prepare different kind Rh nanoparticles have been synthesised in NP-5/cyclo-
of nanoparticles: CeO2[40], Ce1x – ZrxO2 [40], Ce – Tb hexane microemulsion [54..]. At a fixed RhCl3 and
mixed oxides [41], Al2O3 [42], Y2O3:Eu3+ [43], TiO2 surfactant concentration, and fixed W value, ultrafine Rh
[44], silver halides [45]. Zhang and Chan [46] studied particles were obtained using different reducing agents
the formation of Pt– Ru bimetallic nanoparticles using a (H2, NaBH4 and N2H4). Different final particle sizes were
water-in-oil reverse microemulsion of water/Triton X-100/ obtained, depending on the reactants’ nature.
propanol/cyclohexane. The composition in the Pt – Ru Pt –Ru bimetallic nanoparticles have been prepared by
nanoalloy is found to be the same as that in the original using these nonionic surfactants [52]. Silica-coated Rh
precursor solution. They noted that the bimetallic particle nanoparticles was also reported [54..].
size is relationally largest at higher precursor concentra-
tion, but it showed a plateau at very low and very high 3.2.3. Brij microemulsions
metal salt concentrations. This size dependence on con- Brij30, a nonionic surfactant with a short hydrocarbon
centration of precursor was similar to that reported by chain (polyoxyethylene(4) lauryl ether: H 3 (CH 2 ) 11
Chen and Wu [29.] for Ni nanoparticles prepared with a (CH2CH2O)4OH), was used to study the immobilization
cationic surfactant. They concluded that the size of nano- of ZnS nanoparticles synthesised in microemulsions to
particles appears to be limited by nucleation at low silica [54..], and to prepare silica coated iron oxide
concentration and limited by collisions with hydrazine [48].
droplets at high precursor concentrations. It is interesting In a very interesting paper, Grasset et al. [12..]
to note that they proposed the existence of highly improb- compared various surfactants for the synthesis of silica-
able multiple collisions. Zhang and Chan [47] also studied coated zinc ferrite nanoparticles. The coating is achieved
the synthesis of Pt – Co nanoparticles using the same by using a ferrofluid-in-oil microemulsion to which tet-
microemulsion. They noted that larger particles would be raethoxysilane (TEOS) is added. They show that the most
formed if contact between two Pt– Co containing micro- uniform and spherical coatings are achieved by using
emulsions occurred before their individual contacts with Brij30 or a mixture of AOT and Brij30 (50/50 wt.%) as
the microemulsion carrying the reducing agent. the surfactant phase, being the mixture of surfactants the
Silica-coated iron oxide nanoparticles were studied by optimal one. In this way, relatively monodisperse particles
Santra et al. [48], using different precipitating agents in the range 40– 80 nm with a ferrite core of 4 –6 nm are
(NH4OH or NaOH) and surfactants (Triton X, Brij 97 and obtained. Results also show that poorer coatings are
Igepal). They found that the ultrasmall synthesized particles obtained using pure AOT or SDS (with added propanol)
(<5 nm) aggregate in different morphologies. Results are surfactants. Although the authors attribute this different
explained, assuming that the extent of surfactant adsorption behaviour of surfactants to differences in the flexibility
onto the particle surface varies depending on the experi- and interfacial tension of the surfactant film, a definite
mental conditions. The more ordered structures observed explanation is still missing.
with Brij 97 are attributed to its longer hydrophobic chain
(compared to Triton X and Igepal), promoting a more 3.2.4. Igepal microemulsions
ordered particle aggregation, due to stronger hydropho- Pentaoxyethylene-glycol-nonyl-phenyl ether, commonly
bic – hydrophobic interactions between the oleyl groups known by Igepal-CO520, is a nonionic surfactant used by
attached to adjacent nanoparticles. The authors do not Bae et al. [55] to study the influence of [water]/[TEOS]
exclude the influence of the ultrasounds used during the molar ratio on the final size of Pd and Pd/SiO2. They
reaction on the adsorption process. Different results were concluded that the particle size and the thickness of the
obtained by Tartaj and Serna [49]. These authors found that coating can be controlled by manipulating the relative
the nature of the surfactant (Igepal CO-720 or Triton X- rates of the hydrolysis and condensation reaction of
100) did not significantly affect the microstructure of the TEOS.
prepared iron oxide nanoparticles, using cyclohexane as oil Interesting contribution on the influence of W on the
and n-hexanol as cosurfactant. nanoparticle size was given by Nanni and Dei [56].
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 269

Special attention has been paid to the study of magnetic Kaler model—and its extension [68]—provide a sufficiently
properties of different nanoparticles synthesised in this kind reliable general description. This model has been used to
of microemulsions: silica-coated CoFe2O4 and MnFe2O4 predict the evolution of nanoparticle size of hexyl methac-
spinel ferrite nanoparticles [57], silica-coated iron oxide rylate in a microemulsion of water, n-hexyl methacrylate,
particles [49] and iron oxide-doped alumina nanoparticles dodecyltrimethylammonium bromide (DTAB) and didode-
[58]. cyldimethylammonium bromide (DDAB) [69..].
As examples of recent works performed in this area, we
3.2.5. Other surfactants note that:

3.2.5.1. PEG. Polyethylene glycol (PEG) has been used to Tauer et al. [70] have investigated the polymerization
obtain Cs-doped alumina nanoparticles [59] and barium- kinetics by calorimetry and DLS. The results have been
stabilised alumina nanoparticles [60]. explained by the classical nucleation theory.
Cosurfactants effects on the microemulsion polymeriza-
3.2.5.2. Span – Tween 80. Span –Tween 80, a commercial tion of styrene have been studied by Puig et al. [71].
mixture of sorbitol monooleate and polysorbate 80, was Although the kinetics of polymerization slows down in
used to prepare TiO2 nanoparticles in microemulsions [61]. the presence of alcohol, no particle trend was noticed on
The main factor affecting nanoparticle sizes and the physical the particle size.
properties of the nanoparticles was the W ratio. Xu et al. [72] have prepared polystyrene microlatex using
a polymerizable surfactant and CTAB. In this way,
3.2.5.3. Polyoxyethylene 4 lauryl ether. Polyoxyethylene monodisperse latex particles with diameters ranging from
4 lauryl ether was used to prepare Pd, Pt and Pt/Pd nano- 50 to 80 nm could be obtained. The dependence of the
particles, which showed a high catalytic activity [62]. particle concentration, the latex particle size and the
copolymer molar mass on the polymerization time is
3.2.5.4. Polyoxyethylene 15 cetyl ether. Polyoxyethylene discussed in conjunction with the effect of the monomer
15 cetyl ether was used by Tago et al. [63] as a surfactant to concentration.
obtain SiO2-coated CeO2 nanoparticles. They observed that Larpent et al. [73] showed that polymerisation of a
the type of particle-forming agents affects the efficiency of reactive monomer in microemulsions followed by post-
silica coating on the CeO2 nanoparticles. functionalisation allows the synthesis of ligand-function-
alized nanoparticles in the 15 – 25 nm diameter range.
3.2.5.5. Epikuron 170. Epikuron 170 is a lecithin (min. Nanosize polymers of about 60 nm of PMMA have been
67% phosphatidylcholine) used as a surfactant to prepare obtained by Zhang et al. [74] at relatively low surfactant
nimesulide, a molecule of pharmaceutical interest [64]. The concentrations and with a relatively high polymer content
size seems to be independent of either the W ratio or the (c30 wt.%). Jiang et al. [75] have also prepared PMMA
concentration of the active compound. Debuigne et al. [64] in the 20– 40 nm range.
proposed that the constancy of the size suggests that the size
is controlled by thermodynamic stabilization of the nano-
particles with the surfactant molecules. 4. Microemulsions as reaction media

3.2.5.6. S-1670. S-1670 is a nonionic surfactant of food In recent years, microemulsions have been evaluated as
grade sucrose fatty acid ester which was used to prepare reaction media for a variety of chemical reactions. In
CdS and PbS nanoparticles by Khiew et al. [65]. This preparative organic chemistry, microemulsions have been
surfactant is a commercial food grade additive, and used to overcome reactant solubility problems due to the
provides a suitable microenvironment for the preparation ability of microemulsions to solubilize both polar and non-
of materials with narrow size distribution and high polar substances and to compartmentalize and concentrate
monodispersity. reactants. Recent reviews from Holmber et al. [76,77] show
that microemulsions can be regarded as an alternative to
3.3. Polymerization reactions biphasic systems with added phase transfer agent. By
combining microemulsion and phase transfer approaches,
The use of microemulsions to prepare latexes with very high reactions rates have been obtained. Organic
particle sizes below approximately 100 nm is a very impor- molecules with polar and nonpolar regions will accumulate
tant topic with many potential applications in drug delivery, at the oil – water interface of microemulsions. They will
microencapsulation, etc. For this reason, a considerable orient at the interface in such a way that the polar part of the
amount of activity has been conducted in this area (see, molecule extends into the water domain and the non-polar
e.g., Refs. [66,67]). part extends into the hydrocarbon domain. This tendency for
Although there is currently no general scheme for the orientation at the interface can exploited to induce regio-
kinetics of polymerization in microemulsions, the Morgan– specificity in an organic reaction. In fact, bromination of
270 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

phenols and anisols gives a higher para/ortho ratio than in which the internal structure may be oil-in-water drop-
conventional bromination. The strong preference for para lets, bicontinuous, or water-in-oil droplets, the whole
substitution can be of preparative interest. Photocyclodad- solution is divided into oil, surfactant film, and water
dition of 9-substituted anthracenes is also controlled by regions (pseudo-phases) with the surfactant lining the
using water-in-oil microemulsions as reaction media [78]. boundary between the oil and water regions. Each region
Photoirradiation of most of the substituted anthracenes is treated as a separate phase or pseudo-phase, and the
incorporated in the microemulsions almost exclusively partitioning of components between the regions depends
yielded the head-to-head photocyclomers. on their free energies of transfer between the pseudo-
phases. The rate of transfer or reactants is assumed to be
4.1. Chemical reactivity in microemulsions much faster than the observed rate of the reaction (region
III, Scheme 1). In other words, distribution of the reagents
Experimental evidences and models about the effect of is always described by equilibrium constant throughout the
such systems on chemical reactivity should parallel the use course of the reaction.
of microemulsions as reaction media. In particular, the
microheterogeneous nature of microemulsions induces se- 4.1.1. Solvolytic reactions
vere changes of reagent concentrations. Their local concen- Solvolysis reaction rates are usually written as unim-
trations can increase or decrease in relation to their bulk olecular rate constants and they are the simplest example to
concentrations, thus allowing the tuning of reaction rates. be studied in microemulsions. Solvolysis of substituted
Also, the properties of local reaction media are quite benzoyl chlorides was studied in AOT/isooctane/water
different from those of the bulk solutions as a consequence microemulsions [79..]. In order to apply the pseudo-phase
of the intense local electric fields. These fields affect all the model, it was assumed that benzoyl chlorides are only
relevant parameters that modulate the reaction rates. In present at the interface and the oil pseudo-phase. This is
order to carry out a quantitative interpretation of the quite reasonable, when one takes into account their low
influence of the microemulsion on the reactivity, it is solubility in water. This distribution (Scheme 2) yields to
necessary to know the concentrations of the reagents in the conclusion that the only pseudo-phase, where benzoyl
the various pseudo-phases/microenvironments of the system chlorides and water are present, is the interface of the
and the corresponding rate constants. A kinetic model based microemulsion and hence the reaction should take place at
on the formalism of the pseudo-phase was devised to the interface.
explain the reactivity in water in oil microemulsions. In Kinetic results allow several conclusions: (i) Reaction
the pseudo-phase model for homogeneous microemulsions, rate changes with W as a consequence of the changes of

Scheme 2.
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 271

Scheme 3.

the physical properties of the interface. (ii) Rate constant takes place in the water pool and does not involve the
for high water content, Wi50, is much lower than in bulk anionic surfactant, which consistent with repulsive interac-
water due to the insufficient hydration of the interface. (iii) tions between anionic 6NBIC and the AOT headgroups.
Influence of W on the rate constant is determined by the
reaction mechanism. For benzoyl chlorides following a 4.1.2. Neutral molecule-anion reactions
dissociative mechanism (path a in Scheme 3 ), rate con- The microheterogeneous distribution of microemulsions
stant increases with W, whereas for associative mecha- allows the compartmentalisation of reagents, which is
nisms, rate constants increase when W decreases (path c in clearly shown in the case of reactions between a neutral
Scheme 3). molecule (distributed along the different microenviron-
When comparing rates of solvolysis of benzoyl fluoride, ments according to its hydrophobicity) and an anionic
chloride and bromide in AOT based microemulsions [80.], nucleophile. Depending on the reagent hydrophobicity
the ratio kBr/kF decreases 40 times when W decreases from and the charge of the head groups, it is possible to observe
50 to 2. Such decrease of the ratio kBr/kF agrees with a more large catalytic or inhibitory effects. Recent results obtained
efficient associative solvolytic mechanism on decreasing the for basic hydrolysis of 4-nitrophenylacetate in AOT-based
water content of the microemulsion. microemulsions [83] show that reaction rate in micro-
Recent studies on solvolysis of benzoyl chlorides and 4- emulsions is considerably lower than that observed in bulk
nitrophenylchloroformiate in CO2-induced microemulsions water because OH is exclusively present in the water
[81.] of (EO)27(PO)61(EO)27 (P104; EO=ethylene oxide, droplet. 4-Nitrophenylacetate is quite hydrophobic and this
PO = propylene oxide)/p-xylene/CO2/H2O show that the means that only a small fraction is present in the water
observed rate constant of both substrates increase signifi- droplet and hence a remarkable inhibition of hydrolysis is
cantly with W, and that W has a larger influence on the observed. Bimolecular rate constants show that the hydro-
hydrolysis of benzoyl chloride. The different influence of W lysis rate increases on decreasing the water content. This
on the two reactions can be explained in terms of the behaviour has been attributed to desolvation of OH ions
different reaction mechanisms. Bunton et al. studied the when the water content of the microemulsion decreases,
decarboxylation reaction. . .(CTPABr) and AOT microemul- which means an increase in OH reactivity.
sions in CCI4 analysing the relationship between rate Hao [84] studied the basic hydrolysis of aspirin and 2,4-
constants and water properties [82]. With CTPABr, decar- dinitrochlorobenzene in CTABr/butanol/octane/water and
boxylation is much faster than in water, and its addition sodium dodecylsulfonate/butanol/styrene/water microemul-
slows the reaction. The anionic substrate in the interior of sions. Experimental results show that hydrolysis rate is
CTPABr microemulsions interacts with the CTPA+ head- greatly affected by the microstructure and the structural
group, which assists charge delocalisation in the transition transitions of microemulsions. Hydrolysis rates are higher
state, and reactions at Wi0 are faster than those in water by in water-in-oil microemulsions and decrease with the addi-
a factor of approximately 106. With AOT, decarboxylation tion of water. The rates increase in bicontinuous micro-
of 6NBIC has a rate similar to that in water and there is no emulsions and decrease in oil-in-water microemulsions. The
catalysis, with 0«!<W<10—indicating that decarboxylation transition points of the hydrolysis rates occurred at the two
272 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

microemulsion structural transitions points from w/o to the micelles of LTABr. The comparison of the bimolecular
bicontinuous and from bicontinuous to o/w. Nevertheless, rate constants at the interface of the microemulsion with
these results should be considered with caution because those in the aqueous medium show three types of behav-
when water content is increased, it is possible that the iours. These behaviours are well differentiated according to
distribution of reagents is being also affected—which could the hydrophobicity of the amines, which reflects its local-
in turn modify the reaction rates. isation in different zones of the interface of the micro-
emulsion. These zones of the interface will have different
4.1.3. Cation-neutral molecule reactions polarities and will consequently give rise to different
The acid hydrolysis of phthalomohydroxamic acid has effects on the rate of the reactions.
been studied in AOT/isooctane/water microemulsions [85]. Nitroso group transfer reactions were also used to study
Reaction rate is higher in microemulsions than in bulk kinetic behaviour in two different types of four component
water due to compartmentalisation of reagents. It was microemulsions: (i) mixed quaternary microemulsions
observed that reaction rate slows down when increasing (AOT/SDS/isooctane/water) and (ii) quaternary microemul-
AOT concentration at constant W. This behaviour is due to sions with cosurfactant (TTABr/hexanol/isooctane/water).
the competition of H+ and Na+ for binding to the interface For mixed microemulsions (AOT/SDS/isooctane/water),
of the microemulsion droplets. It was also observed that both surfactants will be located at the interface isooctane/
reaction rate becomes faster when moving from W=4 to water [89.], and it is possible to assume that the total
W=12. This behaviour is considered as a consequence of surfactant concentration is the sum of both. The effect of
changes in the properties of interfacial water when chang- addition of SDS to AOT/isooctane/water microemulsions is
ing the microemulsion composition. The acid hydrolysis of similar to the addition of more AOT. Nevertheless, when
some tailor-made Schiff bases having flexible spacers using this type of microemulsions as reaction media, we
between aldimine groups and alkoxy groups at ortho or should consider the increment of the interfacial volume
para position in the benzene ring has been investigated in related with inclusion of the second surfactant. In the case
anionic (SDS) and cationic (CTAB) microemulsions [86]. of quaternary microemulsions with cosurfactant, it should
The change in reactivity, due to change in the spacer be considered that this cosurfactant would be distributed
length and position of the alkoxy group in the Schiff between the interface and the oil phase. The inclusion of
bases, has been explained based on the location sites of part of the cosurfactant in the interface implies an increment
the reaction centre at different polarity pockets of the of its volume and the corresponding local dilution of
reaction media. reagents. Moreover, the solvation ability of alcohol (cosur-
factant) at the interface implies a lower presence of water
4.1.4. Neutral molecule-neutral molecule reactions and the corresponding lower polarity of the interface [90..].
Pseudo-phase model was also used in studies of many Addition of cosurfactant also significantly changes the
bimolecular reactions. In this case, distribution of both continuous pseudo-phase. Reagents not present in the iso-
reagents among the different pseudo-phases should be octane in AOT/isooctane/water are now present in the oil
taken into account, and reaction should take place in those phase formed by isooctane and part of the cosurfactant
where both reagents are present. Local concentrations of (hexanol).
reagents in each pseudo-phase should be defined in order For nucleophilic aromatic substitution reactions [91], a
to analyse kinetic results. Successful explanation was mechanistic change was reported. The rate-limiting step is
achieved for nitroso group transfer reactions [87] to amines now the formation of the intermediate. For W=10 and at
of very different hydrophobicity. In all cases, the results high amine concentrations, the base catalysed reaction in the
assume that the reaction occurs at the interface between the benzene pseudo-phase predominates over the interface re-
water droplet and isooctane and that the dependence of the action. Thus, the observed rate constant decreases with AOT
observed rate constants on the reaction conditions is concentration because of the reactant distributions. Howev-
largely governed by the relative affinities of the amines er, the reaction rate is accelerated at least 3 orders of
for the different phases present in the medium. The magnitude in benzene/BHDC/water microemulsions with
hydrophobicity of the amines determines its distribution respect to the pure solvent, suggesting that the reaction
between the three pseudo-phases and whether the rate occurs at the interface. The catalytic effects of cationic
constants increase, decrease, or remain essentially constant microemulsions have been considered as a consequence of
as the droplet size increases. A comparative study of the interaction between the zwitterionic intermediate and the
nitroso group transfer from N-methyl-N-nitroso-p-toluene- ammonium head of BHDC.
sulfonamide to secondary amines with different hydropho- Another example of a catalytic effect derived from
bic character has been carried out in aqueous micelles, compartmentalisation of the reagents is the hydrolysis of
vesicles and AOT-based microemulsions [88]. By compar- acetylsalicylic acid in AOT/supercritical ethane/water micro-
ing the rate constants in the three colloidal systems, it can emulsions in the presence of imidazole catalyst [92]. An
be suggested that the interface of the microemulsion is less increase of the rate constant by 55 times was observed in
hydrophobic than the vesicular one and more than that of AOT/supercritical ethane microemulsions compared to the
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 273

reaction in aqueous buffer. The reaction had a strong microemulsions prepared with CTAB in n-dodecane/CHCl3
dependence on the water content of micelles, which decel- and isooctane/n-hexanol has been obtained [95.] for 2,4,6-
erated to a large extent when the water content was trimethylbenzenediazonium (1-ArN2+) and 2,6-dimethyl-4-
increased. The acceleration of the reaction in microemul- hexadecylbenzenediazonium (16-ArN2+) tetrafluoroborates.
sions compared to aqueous buffer and the water content Quantitative analysis of the reaction products of 1-ArN2+and
dependence have been confirmed to be due to compartmen- 16-ArN2+with water and bromide ion, the corresponding
talization of the water soluble reactant and catalyst in the phenol and bromo derivatives (Scheme 4), yielded the local
aqueous core of the microemulsion. concentrations of Br-in the water pool and micellar interface
of CTAB microemulsions. Experimental data indicate that
4.1.5. Modification of the pseudo-phase model the degree of counterion dissociation from CTAB micro-
The pseudo-phase model have been modified to take into emulsions in n-dodecane/CHCl3 reaches a value of i0.2
account the specificity of the reaction medium in explaining above W=15. This value is very similar to that found in
the kinetic behaviour of the reaction between [Ru aqueous micelles and may reflect an intrinsic property
(NH3)5pz]2+and S2O82in AOT/water/oil microemulsions related to specific, noncoulombic interactions between the
[93..]. New approaches incorporated to the model are: (i) bromide ion and the tetralkylammonium headgroup.
distribution of substrates between the interface and water The chemical trapping methodology has been applied to
droplet is now described in terms of an equation similar to study the interfacial composition of AOT/isooctane/water
Langmuir equation; (ii) the association of the substrate to microemulsions [96]. The obtained results demonstrate that
the interface is the result of two electrostatic contributions, the interfacial regions of AOT-based microemulsions are
one dependent on the surface potential and the other which densely populated by the sulfosuccinate head-groups of
is specific for each reaction. AOT and the interfacial concentrations of water are signif-
icantly lower than the molar concentration of bulk water.
4.2. Alteration of physical properties of the microemulsion Interfacial water concentration increases with W but the
maximum value obtained for W=44 is about 32 M, signif-
The water microdroplets formed in the central region of icantly lower than the value of 55.5 M for bulk water. This
the microemulsions provide a useful hydrophilic reaction difference is in keeping with studies of solvolysis reactions
field. The reaction field effects of the water pool on various at the interface of AOT-based microemulsiones, where the
photoreactions and thermal reactions and physical transi- reaction constant for high W values is significantly lower
tions have attracted considerable interest. Knowledge of the than the corresponding value in bulk water.
water pool properties—local viscosity, local polarity, local Chemical trapping methodology have been applied by
acidity, and their heterogeneous structure—is required to Romsted and Zhang [97] to the determination of distribution
fully understand the unique reaction systems in this hydro- constants of tert-butylhydroquinone, TBHQ, in a fluid,
philic nanospace. In this way, considerable interest has been opaque, model food emulsion composed of the nonionic
paid in recent years to studies focusing on the character- emulsifier C12E6, octane and water. The distribution con-
isation of composition and physical properties of the differ- stants for partitioning of TBHQ between the oil and surfac-
ent pseudo-phases. Interfacial concentrations in aqueous tant, and between the aqueous and surfactant film were
cationic micelles and microemulsions can be measured by obtained by fitting the changes in the first order rate
a phenyl cation trapping method developed by Chaudhuri constants with emulsifier volume fraction for the reaction
and co-workers [94]. Chemical trapping of bromide ions in of 4-hexadecyl-2,6-dimethylbenzenediazomium ion with
TBHQ, by following the formation of the product hexa-
decyl-2,6-dimethylbenzene by HPLC.
The relationship between the pH of the aqueous solution
adjusted before solubilization and the water pool local pH
estimated with pH-sensitive probes has been studied in AOT/
heptane/water microemulsions. Hasegawa [98] has estimated
the local pH from the excitation spectra of the pH-sensitive
fluorescence probe, pyranine, solubilized into the water core
of the microemulsion as a function of the pH of the aqueous
solutions to be solubilized. This probe shows two distinct
excitation bands corresponding to the neutral and basic
forms. When NaOH/HCl was used for pH adjustment, even
if an acidic or alkaline aqueous solution was solubilized into
the microemulsion, the solubilized probes surprisingly
reported an almost constant intensity ratio over a wide pH
range. This result suggests that the water pools of AOT-based
Scheme 4. microemulsions have buffer-like action. By using a highly
274 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

water soluble model compound, disodium ethanedisulfonate

Et(SO3Na)2, Hasegawa show that this system also had a
buffer-like feature. Aqueous solutions with various pH
values were added to aqueous solutions of Et(SO3Na)2,
where the pH of the mixed solutions was kept at pHi5,
independent of the pH of the used aqueous solutions, in the
range pH=3 –10. The results led to the conclusion that a
considerable number of the AOT sulfonate groups localising
on the water/oil interface are responsible for the buffer-like Scheme 5.
action. The buffer mechanism can be explained by the
suppressed ionic dissociation of the AOT sulfonate groups In spite of the clear advantages in using microemulsions
in the reverse micellar aggregation state. as reaction media for reactions catalysed by metallic cations
and other catalysts, there is a lack of kinetic studies explain-
4.3. Catalytic process in microemulsions ing the role of the microemulsion. Recently, Fanti et al.
[103..] have studied the hydrolytic reactivity of ligands
Use of microemulsions as reaction media is compatible featuring a 6-alkylaminomethylpyridine or an N-alkylethy-
with the simultaneous use of other catalysts. Oil-in-water lenediamine, as chelating subunits, in the presence of Cu(II)
microemulsions based on a nonionic surfactant have been in AOT/isooctane/water microemulsions. The substrates of
used as reaction media to oxidize aqueous azo dyes [99]. choice were the p-nitrophenyl esters of picolinic acid
Different combinations of manganese porphyrins and lipo- (PNPP), of acetic acid (PNPA), and of diphenylphosphoric
philic acids were employed as oxidation catalysts. Phase acid (DPPNPP). In the presence of Cu(II) complexes of
transfer catalysis can also be used in a microemulsion hydroxy-functionalized ligands, such as 1a or 1b (Scheme
system in which case a further acceleration of the reaction 5), the cleavage of PNPP is a million-fold faster than in the
may be obtained [100]. In biphasic systems, the role of the absence of Cu(II) and any ligand.
phase transfer agent is to transfer the nucleophilic anion The effect of Cu(II) alone is much more important than
from the aqueous to the organic phase. Once in the organic that observed in water solution; the rate accelerations at the
phase, the nucleophile becomes highly reactive because (i) same Cu(II) concentrations being 2 orders of magnitude
the degree of solvation is low, and (ii) the large cations used larger. The difference can be ascribed to the favourable
as phase transfer agents do not form strong ion pairs with partition of the metal ion and the substrate in the water pool
the nucleophile in the organic phase; thus, the anion behaves of the reverse micelle. The main characteristics of the
as a ‘‘naked’’ ion. In the microemulsion approach, there is reaction mechanism are conserved, and as in normal micel-
no transfer of reagent from one environment to another. lar aggregates, the main source of the kinetic effects
Addition of phase transfer agents enhance catalytic efficien- observed is the high concentration of the reactants in the
cy of microemulsions in the reaction of lipophilic epoxide aggregate core. However, in the case of reversed metal-
with sodium sulfite; nevertheless, the mechanism of the lomicelles, the high concentration of organic species in the
catalysis is still unclear. Alkenes with carbon numbers water pool is determined by the presence of Cu(II) and by
higher than 8 are hydroformylated in the presence of the possibility of coordination with the metal ion rather than
cobalt-based catalysts. An alternative for solving the prob- by their hydrophilic or lipophilic nature.
lem of miscibility between oil (higher alkene) and water In an attempt to clarify the reasons for a different
(aqueous catalyst solution) is the addition of surfactants to behaviour of catalysis by metallic cations in micelles and
form microemulsions. Haumann et al. [101] have studied microemulsions, a kinetic and thermodynamic study on the
the application of water-soluble catalysts based on cobalt for complexation reaction of metallic cations by bidentate
the combined isomerization and hydroformylation of 7- ligands was recently reported. The complexation constants
tetradecene or less reactive higher alkenes by using Rh of Ni2+ and Co2+ with pyridine-2-azo-p-dimethylaniline
catalysts [102]. The catalyst is highly active and converts (PADA) were investigated in AOT/isooctane/water micro-
the internal alkene into the corresponding branched alde- emulsions [104]. A complexation reaction is always the first
hyde with high regioselectivity. In order to obtain linear step in any process catalysed by Lewis acids. In all cases,
aldehydes from an internal alkene feedstock, cobalt-based the values of the complexation constants are greater than
catalysts can be used. The cobalt catalyst allows isomeriza- those obtained in bulk water. By applying the pseudo-phase
tion of the double bond first, followed by hydroformylation. formalism, the microscopic complexation constants were
More than 50% of the internal alkene is isomerised into a 1- obtained showing that their values decrease as W increases
alkene before hydroformylation although there is no detect- and are always lower than the values obtained in bulk water.
able concentration of 1-alkene. In order to increase the The interaction of interfacial water with the surfactant
linear to branched ratio of the obtained aldehydes, either headgroup in the microemulsion causes an increase in the
the reaction temperature had to be lowered or the excess of electronic density on the oxygen atoms of water and a
ligand had to be increased. consequent increase in the interaction H2O. . .M2+, which
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 275

in turn leads to a greater stabilisation of the ion with respect material for in situ hydrogenation in supercritical CO2. The
to that in bulk water. As the water content decreases, this hydrogenation of 4-methoxycinnamic acid to 4-methoxy-
interaction becomes stronger but the quantity of water hydrocinnamic acid, hydrogenation of trans-stilbene to
molecules available for hydration decreases. This situation 1,2-diphenylethane and hydrogenation of maleic acid to
results in destabilisation of the ion and an increase in its succinic acid were performed in supercritical CO2 micro-
complexation capacity. Formation of Ni2+-PADA and Co2+- emulsions catalysed by Pd nanoparticles (size range of
PADA complexes have been kinetically studied in AOT- about 5– 10 nm). Further studies [107.] have shown that
based microemulsions showing that the complexation rate rhodium nanoparticles dispersed in CO2 microemulsions
constants decrease as both AOT concentration and W are also effective catalysts for rapid hydrogenation of
increase, being in all cases greater than in bulk water. arenes in supercritical CO2.
Analysis of the kinetic data shows that the complexation
mechanism in the microemulsion is compatible with that in
bulk water; the catalytic effects are derived from the Acknowledgements
increase in the local reagent concentrations and the modifi-
cation of water properties when W is varied. Financial support from Ministerio de Ciencia y Tecno-
In recent years, the widespread synthesis of nanosized logı́a (Projects BQU2002-01184 and MAT2002-00824) and
solid particles in w/o microemulsions has opened the possi- Xunta de Galicia (Projects PGIDT03-PXIC20905PN and
bility of new colloidal systems with potential catalytic PGDIT03-PXIC20907PN)) is gratefully acknowledged.
properties. Although their catalytic potential has been recog-
nised, only a few research groups have so far studied the
kinetics of reactions catalysed by such particles. Spiro and de References and recommended reading
Jesus [105 .. ] have shown that the oxidation of p-
Me2NC6H4NH2 by Co(NH3)5Cl2+ can proceed in a buffered
[1] López-Quintela MA, Rivas J, Blanco MC, Tojo C. Synthesis of
water/AOT/n-heptane microemulsion, and that it is strongly
nanoparticles in microemulsions. In: Marzán Liz LM, Kamat PV,
catalysed by nanoparticles of palladium. However, the editors. Nanoscale Materials. Chapter 6. Boston: Kluwer Academic
reaction did not reach completion because of a side Publishing/Plenum; 2003. p. 135 – 55.
reaction between the semiquinonediimine formed and the [2] Fletcher PDI, Howe AM, Robinson BH. The kinetics of solubili-
surfactant. Kinetic results allow the authors to conclude sate exchange between water droplets of a water-in-oil microemul-
sion. J Chem Soc, Faraday Trans 1 1987;83:985 – 1006.
that the rate-determining step of the catalysis is probably
[3] For a recent review in this journal, see. López-Quintela MA. Syn-
diffusion of Co(NH 3 ) 5 Cl 2+ ions through a layer of thesis of nanomaterials in microemulsions: formation mechanisms
adsorbed p-Me2NC6H4NH2 to reach the metal surface. and growth control. Curr Opin Colloid Interface Sci 2003;8:137 – 44.
Electrons are then transferred via the metal from adsorbed [4] Cason J, Miller ME, Thompson JB, Roberts CB. Solvent Effects on
p-Me2NC6H4NH2 molecules to Co(NH3)5Cl2+ ions. Spiro .. Copper Nanoaprticle Growth Behaviour in AOT Reverse Micelle
Systems. J Phys Chem B 2001;105:2297 – 302.
and de Jesus [106] have also studied the oxidation of
One of the few experimental examples trying to correlate the intermicellar
N,N,N,N-tetramethyl-p-phenylenediamine (TMPPD) by exchange rate and subsequent particle growth rate with the different syn-
Co(NH3)5Cl2+ ions catalysed by palladium nanoparticles thesis parameters (water content, bulk solvent, addition of cosolvents and
in an aqueous buffer/AOT/n-heptane microemulsions. The cosurfactants.
activation energy of the catalytic reaction decreased from [5] Shah PS, Holmes JD, Johnston KP, Korgel BA. Size-selective
dispersion of dodencanthiol-coated nanocrystals in ligand and su-
97 kJ mol1 at 15 jC to 39 kJ mol1 at 40 jC. These
percritical ethane by density tuning. J Phys Chem B 2002;10:
values are all greater than those recorded for the 2545 – 51.
corresponding p-Me2NC6H4NH2 reaction. Electrochemical [6] Kitchens C, McLeod MC, Roberts B. Solvent effects on the growth
studies showed that the Co(NH3)5Cl2+ reduction current at .. and steric stabilization of copper metallic nanoparticles in AOT
a rotating Pd electrode decreased in the presence of reverse micelle systems. J Phys Chem B 2003;107:11331 – 8.
This study shows that the bulk organic phase influences both the particle
micromolar amounts of added TMPPD. This finding,
growth rate and the particle size of Cu nanoparticles synthesized in a AOT
together with the marked temperature variation of the microemulsion. A model is proposed which accurately predict the final Cu
activation energy, indicated that adsorption of TMPPD and Ag nanoparticle sizes obtained in various alkanes.
on the Pd particles affected the rate-determining step of [7] Manna A, Imae T, Yogo T, Aoi K, Okazaki M. Synthesis of gold
the catalysis. Slow diffusion of Co(NH3)5Cl2+ through nanoparticles in a Winsor II type microemulsion and their charac-
terization. J Colloid Interface Sci 2002;256:297 – 303.
adsorbed TMPPD species is likely to be the slow step in
[8] Guo L, Huang Q, Li X, Shihe Y. Iron nanoparticles: synthesis and
the catalysis followed by electron transfer via the metal applications in surface enhanced Raman scattering and electroca-
from the adsorbed diamine to the cobalt ion at the surface. talysis. Phys Chem Chem Phys 2001;3:1661 – 5.
Using water/AOT/supercritical CO2 microemulsions, [9] Khomane RB, Manna A, Mandale AB, Kulkarni BD. Synthesis and
Ohde and coworkers [17] showed that hydrogen gas can characterization of dodecanethiol-capped cadmium sulfide nanopar-
cause reduction of a number of metal ions including Pd2+
dissolved in the water core of the microemulsion. After . of special interest
reduction, the hydrogen gas can also serve as a starting .. of outstanding interest
276 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

ticles in a Winsor II microemulsion of diethyl ether/AOT/water. Chiavarone L, Spagnolo V, Lugarà M. Synthesis and characteriza-
Langmuir 2002;18:8237 – 40. tion of CdS nanoclusters in a quaternary microemulsion: the role of
[10] Caponetti E, Pedone L, Martino DC, Pantò V, Turco Liveri V. Syn- the cosurfactant. J Phys Chem B 2000;104:8391 – 7.
thesis, size control, and passivation of CdS nanoparticles in water/ [28] Husein M, Rodil E, Vera J. Formation of silver chloride nanopar-
AOT/n-heptane microemulsions. Mater Sci Eng C 2003;23:531 – 9. ticles in microemulsions by direct precipitation with the surfactant
[11] Li T, Moon J, Morrone AA, Mecholsky JJ, Talham DR, Adair JH. counterion. Langmuir 2003;19:8467 – 74.
Preparation of Ag/SiO2 nanosize composites by a reverse micelle [29] Chen D-H, Wu S-H. Synthesis of nickel nanoparticles in water in oil
and sol – gel technique. Langmuir 1999;15:4328 – 34. . microemulsions. Chem Mater 2000;12:1354 – 60.
[12] Grasset F, Labhsetwar N, Li D, Park CC, Saito N, Haneda H, Cador The influence of both reactants on final nanoparticle size and number of
.. O, Roisnel T, Mornet S, Duguet E, Portier J, Etourneau J. Synthesis particles is discussed. Some hypothesis about the particle formation mech-
and magnetic characterization of zinc ferrite nanoparticles with dif- anism are proposed.
ferent environments: powder, colloidal solution, and zinc ferrite – [30] Fang X, Yang C. An experimental study on the relationship between
silica core – shell nanoparticles. Langmuir 2002;18:8209 – 16. the physical properties of CTAB/hexanol/water reverse micelles and
This is a rigorous and interesting study about the preparation of uniform ZrO2Y2O3 nanoparticles prepared. J Colloid Interface Sci 1999;
silica coatings onto zinc ferrite nanoparticles using ferrofluids/surfactant/ 212:242 – 51.
heptane microemulsions. Experimental conditions for the whole process are [31] Giannakas AE, Vaimakis TC, Ladavos AK, Trikalitis PN, Pomo-
clearly described, and reasons for the different steps used in the synthesis nis PJ. Variation of surface properties and textural features of
are given. spinel ZnAl2O4 and perovskite LaMnO3 nanoparticles prepared
[13] Tago T, Shibata Y, Hatsuta T, Miyajima K, Kishida M, Tashiro S, via CTAB – butanol – octane – nitrate salt microemulsions in the re-
Wakabayashi K. Synthesis of silica-coated rhodium nanoparticles in verse and bicontinuous state. J Colloid Interface Sci 2003;259:
reversed micellar solution. J Mater Sci 2002;37:977 – 82. 244 – 53.
[14] Marchand KE, Tarret M, Lechaire JP, Normand L, Kasztelan S, Cseri [32] Koumoudilis GC, Katsoulidis AP, Ladavos AK, Pomonis PJ,
. T. Investigation of AOT-based microemulsions for the controlled Trapalis AT, Vaimakis TC. Preparation of hydroxyapatite via micro-
synthesis of MoSx nanoparticles: an electron microscopy study. Col- emulsion route. J Colloid Interface Sci 2003;259:254 – 60.
loid Surf, A Physicochem Eng Asp 2003;214:239 – 48. [33] Wu Z, Zhang J, Benfield RE, Ding Y, Grandjean D, Zhang Z, Ju
An interesting study about the influence of cosurfactant on the kinetics of X. Structure and chemical transformation in cerium oxide nano-
the intermicellar exchange. particles coated by surfactant cetyltrimethylammonium bromide
[15] Bagwe RP, Khilar KC. Effects of the intermicellar exchange rate and (CTAB): an X-ray absorption spectroscopy study. J Phys Chem
cations on the size of silver chloride nanoparticles formed in reverse B 2002;106:4569 – 77.
micelles of AOT. Langmuir 1997;13:6432 – 8. [34] Yu KMK, Yeung CMY, Thompsett D, Tsang SC. Aerogel-coated
[16] Ohde H, Hunt F, Wai CM. Synthesis of silver and copper nano- metal nanoparticle colloid as novel entities for the synthesis of
particles in a water-in-supercritical-carbon dioxide microemulsion. defined supported metal catalysis. J Phys Chem B 2003;107:
Chem Mater 2001;13:4130 – 5. 4515 – 26.
[17] Ohde H, Wai CM, Kim H, Kim J, Ohde M. Hydrogenation of [35] Sun Y, Atorngitjawat P, Meziani MJ. Preparation of silver nanopar-
Olefins in Supercritical CO2 catalyzed by palladium nanoparticles ticles via rapid expansion of water in carbon dioxide microemulsion
in a water-in-CO2 microemulsion. JACS 2002;124:4540 – 1. into reductant solution. Langmuir 2001;17:5707 – 15710.
[18] Ohde H, Ohde M, Bailey F, Kim H, Wai CM. Water-in-oil micro- [36] Stallings WE, Lamb HH. Synthesis of nanostructured titania pow-
emulsions as nanoreactors for synthesizing CdS and ZnS nanopar- ders via hydrolysis of titanium isopropoxide in supercritical carbon
ticles in supercritical CO2. Nano Lett 2002;2:721 – 4. dioxide, Langmuir 2003;19:2989 – 2994.
[19] Goncalves SAP, De Pauli SH, Tedesco AC, Quina FH, Okano LT, [37] Zhang R, Liu J, He J, Han B, Wu W, Jiang T, Liu Z, Du J. Organic
Bonilha JBS. Counterion exchange selectivity coefficients at water- reactions and nanoparticle preparation in CO2-induced water/P104/
in-oil microemulsion interface. J Colloid Interface Sci 2003;267: p-xylene microemulsions. Chem Eur J 2003;9:2167 – 72.
494 – 9. [38] Nagashima K, Lee Jr CT, Xu B, Johnston KP, DeSimone JM,
[20] Yadav OP, Palmqvist A, Cruise N, Holmberg K. Synthesis of plat- Johnson Jr CS. NMR studies of water transport and proton ex-
inum nanoparticles in microemulsions and their catalytic activity for change in water-in-dioxide microemulsions. J Phys Chem B 2003;
the oxidation of carbon dioxide. Colloids Surf, A Physicochem Eng 107:1962 – 8.
Asp 2003;221:131 – 4. [39] Wang H, Zhang J-R, Zhu J-J. Sonochemical preparation of lead
[21] Liao M-H, Chen D-H. Characteristics of magnetic nanoaprticles- sulfide nanocrystals in an oil-in-water microemulsion. J Cryst
bound YADH in water/AOT/isooctane microemulsions. J Mol Catal, Growth 2002;246:161 – 8.
B Enxym 2002;18:81 – 7. [40] Rodriguez JA, Hanson JC, Kim J-Y, Liu G. Properties of CeO2
[22] Ingelsten HH, Béziat J, Bergkvist K, Palmqvist A, Skoglundh M, and Ce1-x ZrxO2 nanoparticles: X-ray absorption near-edge spec-
Qiuhong H, Falk LKL, Holmberg K. Deposition of platinum nano- troscopy, density functional and time-resolved X-ray diffraction
particles, synthesized in water-in-oil microemulsions, on alumina studies. J Phys Chem B 2003;107:3535 – 43.
supports. Langmuir 2002;18:1811 – 8. [41] Hungrı́a AB, Martı́nez-Aries A, Fernández-Garcı́a M, Iglesias-Juez
[23] Chen D-H, Chen C-J. Formation and characterization of Au – Ag A, Guerrero-Ruiz A, Calvino CC, Conesa JC, Soria J. Structural,
. bimetallic nanoparticles in water-in-oil microemulsions. J Mater morphological, and oxygen handling properties of nanosized
Chem 2002;12:1557 – 62. cerium – terbium mixed oxides prepared by microemulsions. Chem
An interesting report on the synthesis of bimetallic nanoparticles. Mater 2003;15:4309 – 16.
[24] Wu M-L, Chen D-H, Huang T-C. Synthesis of Au/Pd bimetallic [42] Pang Y-X, Bao X. Aluminium oxide nanoparticles prepared by wa-
nanoparticles in reverse micelles. Langmuir 2001;17:3877 – 83. ter-in-oil microemulsions. J Mater Chem 2002;12:3699 – 704.
[25] Wu M-L, Chen D-H, Huang T-C. Preparation of Pd/Pt bimetallic [43] Pang Q, Shi J, Liu Y, Xing D, Gong M, Xu N. A novel approach for
nanoparticles in water/AOT/isooctane microemulsions. J Colloid preparation of Y2O3:Eu3+ nanoparticles by microemulsions—micro-
Interface Sci 2001;243:102 – 8. wave heating. Mater Sci Eng B 2003;103:57 – 61.
[26] Porta F, Prati L, Rossi M, Scari G. Synthesis of Au(0) nanoparticles [44] Andersson M, Österlund L, Ljungström S, Palmqvist A. Preparation
from W/O microemulsions. Colloids Surf, A Physicochem Eng Asp of nanosize anatase and rutile TiO2 by hydrothermal treatment of
2002;211:43 – 8. microemulsions and their activity for photocatalytic wet oxidation of
[27] Curri ML, Agostiano A, Manna L, Della Monica M, Catalano M, phenol. J Phys Chem B 2002;106:10674 – 9.
 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278 277

[45] Xu S, Li Y. Different morphology at different reactant molar ratios: [66] Candau F. In: Paleos CM, editor. Polymerization in Organized
synthesis of silver halide low-dimensional in microemulsions. J Ma- Media. Paris: Gordon & Breach, 1992. Chap. 4.
ter Chem 2003;13:163 – 5. [67] Antonietti M, Basten R, Lohmann S. Polymerization in microemul-
[46] Zhang X, Chan K-Y. Water-in-oil microemulsions synthesis of plat- sions—a new approach to ultrafine, highly functionalized polymer
inum – ruthenium nanoparticles, their characterization and electroca- dispersions. Macromol Chem Phys 1995;196:441 – 66.
talytic properties. Chem Mater 2003;15:451 – 9. [68] De Vries R, Co CC, Kaler EW. Macromolecules 2001;34:3233 – 44.
[47] Zhang X, Chan K-Y. Microemulsion synthesis and electrocatalytic [69] Hermanson KD, Kaler EW. Kinetics and mechanism of the multiple
properties of platinum – cobalt nanoparticles. J Mater Chem 2002; .. addition microemulsion polymerization of hexyl methacrylate. Mac-
12:1203 – 6. romolecules 2003;36:1836 – 42.
[48] Santra S, Tapec R, Theodoropoulou N, Dobson J, Hebard A, Tan W. The Morgan – Kaler model, and its extended version, are used to examine
Synthesis and characterization of silica-coated iron oxide nanopar- the polymerization mechanisms and to predict the evolution of particle
ticles in microemulsions: the effects of nonionic surfactants. Lang- size.
muir 2001;17:2900 – 6. [70] Tauer K, Ramı́rez AG, López RG. Effect of the surfactant concen-
[49] Tartaj P, Serna CJ. Microemulsion-assisted synthesis of tunable tration on the kinetics of oil in water microemulsion polymerization:
superparamagnetic composites. Chem Mater 2002;14:4396 – 402. a case study with butyl acrylate. C R Chim 2003;6:1245 – 66.
[50] Bose S, Saha SK. Synthesis and characterization of hydroxyap- [71] Puig JE, Mendizábal E, Delgado S, Arellano J, López-Serrano F.
atite nanopowders by emulsion technique. Chem Mater 2003;15: Cosurfactant effects on the microemulsion characterization of sty-
4464 – 9. rene. C R Chim 2003;6:1267 – 73.
[51] Fang J, Stokes KL, Wiemann JA, Zhou WL, Dai J, Chen F, O’Connor [72] Xu XJ, Siow KS, Wong MK, Gan LM. Microemulsion polymeriza-
CJ. Microemulsion-processed bismuth nanoparticles. Mater Sci Eng tion of styrene using a polymerizable nonionic surfactant and a
B 2001;83:254 – 7. cationic surfactant. Colloid Polym Sci 2001;279:879 – 86.
[52] Liu Z, Lee JY, Han M, Chen W, Gan LM. Synthesis and charac- [73] Larpent C, Amigoni-Gerbier S, De Sousa Delgado A. Synthesis of
terization of PtRu/C catalysts from microemulsions and emulsions. metal-complexing nanoparticles by post-functionalisation of reactive
J Mater Chem 2002;12:2453 – 8. nanolatexes produced by microemulsion polymerisation. C R Chim
[53] Althues H, Kaskel S. Sulfated zirconia nanoparticles synthesized in 2003;6:1275 – 83.
reverse microemulsions: preparation and catalytic properties. Lang- [74] Zhang Y, Guo T, Hao G, Song M, Zhang B. Nanosize polymer
muir 2002;18:7428 – 35. latices made by microemulsion copolymerization: preparation and
[54] Hanaoka T, Hayashi H, Tago T, Kishida M, Wakabayashi K. In situ characterization. J Appl Polym Sci 2003;90:3625 – 30.
.. immobilization of ultrafine particles synthesized in a water/oil [75] Jiang W, Yang W, Zeng X, Fu S. Structure and properties of
microemulsion. J Colloid Interface Sci 2001;235:235 – 40. poly(methyl methacrylate) particles prepared by a modified micro-
This is an interesting example of the influence of the chemical reaction rate emulsion polymerization. J Polym Sci 2004;42:733 – 41.
on the final particle size. [76] Holmber K. Organic reactions in microemulsions. Curr Opin Colloid
[55] Bae D-S, Han K-S, Adair JH. Synthesis and microstructure of Pd/ Interface Sci 2003;8:187 – 96.
SiO2 nanosized particles by reverse micelle and sol/gel processing. [77] Häger M, Currie F, Holmber K. Organic reactions in microemulsions.
J Mater Chem 2002;12:3117 – 20. Top Curr Chem 2003;227:53 – 74.
[56] Nanni A, Dei L. Ca(OH)2 nanoparticles from w/o microemul- [78] Wu D-Y, Zhang L-P, Wu L-Z, Wang B-J, Tung Ch-H. Water-in-oil
sions. Langmuir 2003;19:933 – 8. microemulsions as microreactors to control the regioselectivity in the
[57] Vestal C, Zhang ZJ. Synthesis and magnetic characterization of Mn photocycloaddition of 9-substituted anthracenes. Tetrahedron Lett
and Co spinel ferrite/silica nanoparticles with tunable magnetic core. 2002;43:1281 – 3.
Nano Lett 2003;3:1739 – 43. [79] Garcı́a-Rı́o L, Leis JR, Moreira JA. Reactivity in water/oil micro-
[58] Tartaj P, Tartaj J. Microstructural evolution of iron-oxide-doped alu- .. emulsions. Influence of sodium bis/2-ethylhexyl)sulfosuccinate/iso-
mina nanoparticles synthesized from microemulsions. Chem Mater octane/water microemulsions on the solvolysis mechanism of
2002;14:536 – 41. substituted benzoyl chlorides. J Am Chem Soc 2000;122:10325 – 34.
[59] You Z, Balint I, Aika K. Synthesis of thermally stable Cs-doped Changes in the physical properties of microemulsions on changing W in-
alumina nanoparticles by microemulsion method. Chem Lett 2002; duce changes in the solvolysis reaction rates. As a consequence, the balance
1090-1. between associative and dissociative mechanisms is shifted to a more as-
[60] Balint I, You Z, Aika K. Morphology and oxide phase control in the sociative one on decreasing the water content.
microemulsion mediated synthesis of barium stabilized alumina [80] Garcı́a-Rı́o L, Leis JR, Mejuto JC. Solvolysis of benzoyl halides in
nanoparticles. Phys Chem Chem Phys 2002;4:2501 – 3. . AOT/isooctane/water microemulsions. Influence of the leaving
[61] Zhand R, Gao L. Preparation of nanosized titania by hydroly- group. Langmuir 2003;19:3190 – 7.
sis of alkoxide titanium in micelles. Mater Res Bull 2002;37: The ionizing power of the interface of the microemulsions changes with W,
1659 – 66. and as a consequence also changes the solvolysis rate of benzoyl halides.
[62] Yashima M, Falk LKL, Palmqvist AEC, Holmberg K. Structure The influence of the ionizing power on the solvolysis rate is for
and catalytic properties of nanosized alumina supported platinum Br>Cl>F.
and palladium particles synthesized by reaction in microemulsion. [81] Zhang R, Liu J, He J, Han B, Wu W, Jiang T, Liu Z, Du J. Organic
J Colloid Interface Sci 2003;268:348 – 56. . reactions and nanoparticle preparation in CO2-induced water/P104/
[63] Tago T, Tashiro S, Hashimoto Y, Wakabayashi K, Kishida M. p-xylene microemulsions. Chem Eur J 2003;9:2167 – 72.
Synthesis and optical properties of SiO2-coated CeO2 nanoparticles. The different influence of W on the hydrolysis of benzoyl chloride and p-
J Nanopart Res 2003;5:55 – 60. nitrophenyl chloroformate can be ascribed to the different reaction mecha-
[64] Debuigne F, Cuisenaire J, Jeunieau L, Masereel B, Nagy JB. Syn- nisms and the expected changes in the nucleophilicity and polarity of water
thesis of nimesulide nanoparticles in the microemulsion epikuron/ in the microemulsions.
isopropyl myristate/water/n-butanol (or isopropanol). J Colloid In- [82] Bunton CA, Garreffa A, Germani R, Onori G, Santucci A, Savelli G.
terface Sci 2001;243:90 – 101. Relation between the IR spectrum of water and decarboxylation
[65] Khiew PS, Radiman S, Huang NM, Ahmad MdS. Studies on the kinetics in microemulsions. Prog Colloid Polym Sci 2001;118:
growth and characterization of CdS and PbS nanoparticles using 103 – 6.
sugar-ester nonionic water-in-oil microemulsion. J Cryst Growth [83] Garcı́a-Rı́o L, Mejuto JC, Pérez M. Modification of reactivity by
2003;254:235 – 43. changing microemulsion composition. Basic hydrolysis of nitro-
278 M.A. López-Quintela et al. / Current Opinion in Colloid & Interface Science 9 (2004) 264–278

phenyl acetate in AOT/isooctane/water systems. Unpublished hexadecyltrimethylammonium bromide in chloroform/n-dodecane

results. and isooctane/n-hexanol by chemical trapping. Langmuir 2001;17:
[84] Hao J. Effect of the structures of microemulsions on chemical reac- 1060 – 8.
tions. Colloid Polym Sci 2000;278:150 – 4. The chemical trapping methodology was used to investigate the interface of
[85] Ghosh KK, Tiwary LK. Microemulsions as reaction media for a cationic microemulsions. Authors evaluate the degree of counterion disso-
hydrolysis reaction. J Dispers Sci Technol 2001;22:343 – 8. ciation and their results are close to those obtained in direct micelles.
[86] Mishra BK, Patel N, Dash PK, Behera GB. Reactions in micro- [96] Srilakshmi GV, Chaudhuri A. ‘‘Snap-Shooting’’ the interface of AOT
emulsion media: Schiff bases with targeting/anchoring module as reverse micelles: use of chemical trapping. Chem Eur J 2000;6:
kinetic sensor to map the polarity pocket of a microemulsion droplet. 2847 – 53.
Int J Chem Kinet 2001;33:458 – 64. [97] Romsted LS, Zhang J. Kinetic method for determining antioxi-
[87] Garcı́a-Rı́o L, Hervés P, Mejuto JC, Pérez-Juste J, Rodriguez- dant distributions in model food emulsions: distribution constants
Dafonte P. Pseudophase approach to reactivity in microemulsions: of t-butylhydroquinone in mixtures of octane, water, and a non-
quantitative explanation of the kinetics of the nitroso group transfer ionic emulsifier. J Agric Food Chem 2002;50:3328 – 36.
reactions between N-methyl-N-nitroso-p-toluenesulfonamide and [98] Hasegawa M. Buffer-like action in water pool or aerosol OT reverse
secondary alkylamines in water/AOT/isooctane microemulsions. micelles. Langmuir 2001;17:1426 – 31.
Ind Eng Chem Res 2003;42:5450 – 6. [99] Häger M, Holmberg K, da Rocha Gonsalves AM, Sera AC. Oxida-
[88] Garcı́a-Rı́o L, Hervés P, Mejuto JC, Pérez-Juste J, Rodrı́guez- tion of azo dyes in oil-in-water microemulsions catalyzed by metal-
Dafonte P. Comparative study of nitroso group transfer in colloidal loporphyrins in presence of lipophilic acids. Colloid Surf A 2001;
aggregates: micelles, vesicles and microemulsions. J New Chem 183 – 185:247 – 57.
2003;27:372 – 80. [100] Häger M, Holmberg K. A substitution reaction in an oil-in-water
[89] Garcı́a-Rı́o L, Hervés P, Mejuro J, Pérez-Juste J, Rodriguez-Dafonte microemulsion catalyzed by a phase transfer catalyst. Tetrahedron
. P. Pseudophase approach to the transfer of the nitroso group in water/ Lett 2000;41:1245 – 8.
AOT/SDS/isooctane quaternary microemulsions. Langmuir 2000; [101] Haumann M, Koch H, Schomäcker R. Hydroformylation in micro-
16:9716 – 21. emulsios: conversion of an internal long chain alkene into a linear
The pseudophase model accounts for addition of a new surfactant to a three aldehyde using a water soluble cobalt catalyst. Catal Today 2003;
component microemulsion. The fourth component (another surfactant) is 79 – 80:43 – 9.
considered to be located at the interface and its effect will be an increase [102] Haumann M, Yildiz H, Koch H, Schomäcker R. Hydroformylation
in the volume of the phase. of 7-tetradecene using Rh-TPPTS in a microemulsion. Appl Catal A
[90] Garcı́a-Rı́o L, Leis JR. Reactivity in quaternary water in oil micro- 2002;236:173 – 8.
. emulsions: 2. Different distribution of the reagents changing from [103] Fanti M, Mancin F, Tecilla P, Tonellato U. Ester cleavage catalysis in
three- to four-component microemulsions. J Phys Chem B 2000;104: .. reversed micelles by Cu(II) complexes of hydroxy-functionalized
6618 – 25. ligands. Langmuir 2000;16:10115 – 22.
The influence of the cosurfactant, a medium chain alcohol, is not only to This study mix the catalytic influence of microemulsions and copper
modify the properties of the interface but also the properties of the contin- cations on the ester hydrolysis. The kinetic response of the microemul-
uous medium, because the cosurfactant will be distributed between the sions is quite different from that of direct aqueous micelles, mainly be-
continuous medium and the interface. cause the apparent insensitivity of the reaction rate to variables such as
[91] Correa NM, Durantini EN, Silber JJ. Influence of anionic and cat- the pH of added water, the water content and the hydrophilic character of
ionic reverse micelles on nucleophilic aromatic substitution reaction the ligand.
between 1-fluoro-2,4-dinitrobenzene and piperidine. J Org Chem [104] Fernández E, Garcı́a-Rı́o L, Knörl A, Leis JR. Metal – Ligand
2000;65:6427 – 33. complexation in water in oil microemulsions: I. Thermodynamic
[92] Shervani Z, Ikushima Y. Compartmentalization of reactants in su- approach. Langmuir 2003;19:6611 – 9.
percritical fluid micelles. J New Chem 2002;26:1257 – 60. [105] Spiro M, de Jesus DM. Nanoparticle catalysis in microemulsions:
[93] López-Cornejo P, Pérez P, Garcı́a F, Vega R, Sánchez F. Use of .. oxidation of N, N-dimethyl-p-phenylenediamine by cobalt (III) pen-
.. the pseudophase model in the interpretation of reactivity under taamine chloride catalyzed by colloidal palladium in water/AOT/n-
restricted geometry conditions. An application to the study of the heptane microemulsions. Langmuir 2000;16:2464 – 8.
[Ru(NH3)5pz]2+ + S2O82-electron-transfer reaction in different This is an interesting study on mixing the catalytic effect of microemul-
microheterogeneous systems. J Am Chem Soc 2002;124:5154 – 64. sions and nanoparticle on oxidation reactions.
The pseudophase model was modified in order to take into account the [106] Spiro M, de Jesus DM. Catalysis by palladium nanoparticles in
specificity of the reaction medium. These modifications describe the anti- microemulsions. Langmuir 2000;16:4896 – 900.
cooperative character of the binding of the substrates to the surfaces and the [107 ] Ohde M, Ohde H, Wai CH. Catalytic hydrogenation of arenes with
changes of the electric potential at the surfaces. . rhodium nanoparticles in a water in supercritical CO2 microemul-
[94] Chaudhuri JA, Loughlin JA, Romsted LS, Yao J. Arenediazonium sion. Chem Commun 2002:2388 – 9.
salts: new probes of the interfacial compositions of association col- A beautiful example of how nanoparticle can catalyze an hydrogenation
loids: 1. Basic approach, methods, and illustrative applications. J Am reaction in microemulsions. Moreover, the continuous medium of the
Chem Soc 1993;115:8351 – 61. microemulsions is supercritical CO2 yielding a more environmental
[95] Cuccovia IM, Dias LG, Maximiano FA, Chaimovich H. Analysis of friendly process.
. the bromide ion distribution in the water pool of reverse micelles of
 Available online at www.sciencedirect.com

International Journal of Pharmaceutics 345 (2007) 9–25

Review

Stable drug encapsulation in micelles and microemulsions

Ajit S. Narang a,1 , David Delmarre b,1 , Danchen Gao c,∗,1
a
Biopharmaceutics R&D, Bristol-Myers Squibb, PO Box 191, Mail Stop 85A-167A, New Brunswick, NJ 08903, USA
b Capsugel Pharmaceutical R&D Center, Parc d’Innovation, Rue Tobias Stimmer, B.P. 30442, 67412 Illkirch Graffenstaden Cedex, France
c Anchen Pharmaceuticals, Inc., 5 Goodyear, Irvine, CA 92618, USA

Received 8 June 2007; received in revised form 26 August 2007; accepted 30 August 2007
Available online 8 September 2007

Abstract
Oral absorption of hydrophobic drugs can be significantly improved using lipid-based non-particulate drug delivery systems, which avoid the
dissolution step. Micellar and microemulsion systems, being the most dispersed of all, appear the most promising. While these systems show
high drug entrapment and release under sink conditions, the improvement in oral drug bioavailability is often unpredictable. The formulation and
drug-related biopharmaceutical aspects of these systems that govern oral absorption have been widely studied. Among these, preventing drug
precipitation upon aqueous dilution could play a predominant role in many cases. Predictive ability and quick methods for assessment of such
problems could be very useful to the formulators in selecting lead formulations. This review will attempt to summarize the research work that
could be useful in developing these tools.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Bioavailability; Hydrophobic drugs; Micelles; Microemulsions; Precipitation; SEDDS; SMEDDS; Self-emulsifying; Solubilization; Self-
microemulsifying

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.1. Solutions, emulsions, microemulsions, and micelles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2. Components of micelles and microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3. Characterization of microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.4. Drug entrapment and structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.5. Microemulsions for protein and peptide delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2. Drug loading capacity in micelles and microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1. Solubilization capacity in reverse micelles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2. Dilutability as monophasic systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3. Solubilization capacity in diluted microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3. Drug precipitation and solute crystallization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1. In vivo drug precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2. Prediction of in vivo drug precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3. Avoiding in vivo drug precipitation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4. Mechanism of solute crystallization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.5. Preventing drug crystallization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.6. Combined use of solubilization approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

∗ Corresponding author. Tel.: +1 949 639 8143.

E-mail address: danchen.gao@anchen.com (D. Gao).
1 Formerly at Morton Grove Pharmaceuticals, Inc., 50 Lakeview Pkwy #127,

Vernon Hills, IL 60061, USA.

0378-5173/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpharm.2007.08.057
10 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

4. Other factors influencing bioavailability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

4.1. Lymphatic transport and lipolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.2. Inhibition of drug efflux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.3. Dispersion size of emulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

1. Introduction and may be limited in their daily and per-dose uptake levels. For-
mulators aim to develop systems with maximum drug loading
Oral liquid dosage forms are often required of new molecules, capacity while using minimum possible amounts of surfactants
especially at the discovery and pre-clinical stages of drug devel- and cosolvents. These limitations lead formulators to a limited
opment, and of existing molecules as a part of product life-cycle range of compositions.
management. When permitted by the aqueous solubility and In addition, micelles and microemulsions can be metastable
stability of the drug substance, a simple solution in water is pre- with respect to drug solubility and show drug precipitation upon
ferred, e.g., Prozac® oral solution. More often, however, drug dilution or crystallization over a period of storage. In vivo drug
solubility (in relation to its required concentration) and stability precipitation upon dilution in stomach can lead to failure in
are the limiting factors. Hydrophobic drugs may be formulated bioavailability enhancement and compromise the competitive
as emulsions and suspensions, e.g., Megace ES® suspension and advantage of this dosage form. In vitro drug crystallization in
Diprivan® emulsion. Drugs that show rapid degradation in aque- a micellar solution or microemulsion could be very slow and
ous media can be formulated as either powder for suspension, dependent on temperature and handling of the formulation. The
e.g., Augmentin® , Amoxil® , and Zegerid® ; powder for solu- ready-to-use formulations are expected to have a shelf life of at
tion, e.g., Zerit® ; oily solution, e.g., Aquasol E® (Vitamin E) least 2 years, while concentrates (SEDDS and SMEDDS) are
soft gelatin capsules; or oily suspension, e.g., Accutane® soft expected to be physically and chemically stable after reconsti-
gelatin capsules. Hydrolysis-sensitive hydrophobic drugs may tution for the duration of the therapy or until administration.
also be formulated as oily concentrates called self-emulsifying Examples of commercialized SMEDDS formulations include
drug delivery systems (SEDDS) that form an emulsion upon cyclosporine (Neoral® ), ritonavir (Norvir® ), and saquinavir
addition of water or an aqueous solution with mild agitation, (Fortovase® ) (Cooney et al., 1998, Porter and Charman, 2001).
e.g., Sandimmune® oral solution. Very few SEDDS and SMEDDS formulations have been
Emulsions and suspensions allow the drug to be administered commercialized because of limitations in the usage level of
as a dispersed oil solution or as suspended particles, respec- excipients, e.g., surfactants and cosolvents, and the unpre-
tively. These dosage forms, however, have particulate nature dictable improvement of oral bioavailability due to possibility of
and show phase separation upon storage due to their thermo- drug precipitation upon aqueous dilution in vivo. Predictive abil-
dynamic instability. In contrast, micelles and microemulsions ity and quick methods for assessment of such problems could
do not show the physical instability in terms of agglomeration be very useful to the formulators in selecting lead formulations.
or separation of the dispersed phase. These systems also have This review will attempt to summarize the research work that
lower dispersed phase size (≤200 nm) than emulsions, giving could be useful in developing these tools.
them transparency. Also, these dosage forms allow the drug to
be formulated as both ready-to-use aqueous solutions and as 1.1. Solutions, emulsions, microemulsions, and micelles
non-aqueous concentrates. The concentrate may be a solution,
reverse micellar solution, or a microemulsion, which is diluted Simple aqueous drug solutions involve hydrogen-bonding
with water immediately before administration, or administered and dipole interactions of drug molecules with the surround-
as it is and gets diluted with gastric fluids in vivo. In cases where ing water. Hydrophobic drugs have low solubility because
they form transparent microemulsions upon dilution, the con- of lower capacity for these interactions. In such cases, the
centrates are known as the self-microemulsifying drug delivery solute–solvent interactions can be qualitatively as well as quan-
systems (SMEDDS). SEDDS, SMEDDS, and micellar systems titatively changed to improve the drug solubility. For example,
offer further advantage over conventional emulsions in the sig- pH can be adjusted with buffers to increase ionization of a
nificantly reduced energy requirement for their preparation, such weakly acidic or a weakly basic drug, resulting in higher ion-
that simple mixing is enough for their formation. SEDDS and dipole solute–solvent interactions. Cosolvent addition reduces
SMEDDS may also be administered as concentrates, e.g., in the dielectric constant of water and facilitates hydrophobic
a soft gelatin capsule, and expected to form solubilized drug interactions of drug molecules with the solvent system. Sol-
containing micelles or microemulsions in vivo upon dilution in ubility may also be increased by drug complexation with
stomach. a hydrophilic compound, e.g., hydroxypropyl-␤-cyclodextrin
The use of SEDDS, SMEDDS, and micellar systems is (HPBCD). Hydrophobic and/or specific ionic interactions lead
limited by their drug loading capacity and the usage level of to drug entrapment in HPBCD, which, in turn, is soluble in
excipients. Surfactants and cosolvents can be toxic at high doses water. In addition, incorporation of amphiphilic surfactants in
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 11

aqueous solutions can solubilize hydrophobic drugs by different concentration on the stability of an emulsion or a microemulsion
mechanisms. is more profound in the case of ionic surfactant than nonionic
Surfactants have both hydrophilic and lipophilic properties surfactants. Additionally, for pharmaceutical applications, ionic
and are characterized by their hydrophile–lipophile balance surfactants are not preferred due to toxicological concerns.
(HLB) values. Surfactants with an HLB value >10 are predomi- Microemulsions often include a cosurfactant. A cosurfactant
nantly hydrophilic and favor the formation of o/w emulsions, is an amphiphilic molecule that substantially accumulates with
while surfactants with HLB values <10 are hydrophobic the surfactant at the interfacial layer. Usually a very low HLB
and form w/o emulsions. High HLB surfactants are used to cosurfactant is used with a high HLB surfactant to modify the
form aqueous solutions or dispersions of hydrophobic drug overall HLB of the system. Unlike surfactant, the cosurfactant
molecules. may not be capable of forming self-associated structures like
Surfactants in solution below their critical micellization micelles on its own. Several kinds of molecules including non-
concentration (CMC) improve drug solubility by providing ionic surfactants and alcohols can function as cosurfactants in a
regions for hydrophobic drug interactions in solution. Above the given system. The quantity of a cosurfactant in a system is usu-
CMC, surfactants self-aggregate in defined orientation to form ally less than that of the surfactant and it often serves to modify
micelles with a hydrophobic core and a hydrophilic surface. The the overall HLB value of the system.
hydrophobic core enhances the entrapment of drug, thus increas- Cosolvents are often included in microemulsion formulations
ing its solubility. In the presence of a significant amount of oil, to increase drug solubility by cosolvency and to stabilize the
surfactants concentrate on the oil/water interface forming emul- dispersed phase. In addition to making the environment more
sions, wherein the drug is solubilized in the internal oil phase. hydrophobic by reducing the dielectric constant of water, cosol-
When the oil content is low, minute oil-entrapped surfactant vents increase the amount of molecularly dispersed surfactant
globules are produced, which are known as swollen-micelles or in the aqueous phase. Availability of free surfactant aids in drug
microemulsions. Drug may be solubilized in the oily core and/or solubilization by creating pockets of hydrophobic regions within
on the interface of these structures. The predominant location the aqueous phase. Examples of surfactants, cosurfactants, and
of drug solubilization depends on its hydrophobicity and inter- cosolvents that have been used in commercial lipid-based prod-
actions with the surfactant and/or cosurfactant. Microemulsions ucts are listed in Table 1. In addition, various surfactants and
differ from micelles in the presence of oil and from emulsions cosurfactants have been listed with their HLB values, chemical
in the amount of the dispersed phase. Furthermore, microemul- classification, and commercial names in US patent application
sions often require a cosolvent and/or cosurfactant to facilitate PCT/US00/32255. Also, Strickley has summarized the solubi-
their formation. lizing excipients used in commercial formulations (Strickley,
Both microemulsions and micelles are useful for preparing 2004).
aqueous solutions of hydrophobic drugs. Several recent reviews Structurally, the dispersed phase of microemulsions con-
have summarized physical and biopharmaceutical aspects of sists of microstructures of oil-entrapped pockets stabilized by
these systems (Constantinides, 1995; Flanagan and Singh, surfactant/cosurfactant accumulation on the oil/water bound-
2006; Gursoy and Benita, 2004; Pouton, 2000; Pouton, 1997; ary, similar to conventional emulsions. In addition, surfactant
Lawrence and Rees, 2000; Humberstone and Charman, 1997). molecules self-associate to form micelles in the bulk phase.
The physical nature of these systems, mechanism of drug entrap- These structures coexist in equilibrium, with their relative abun-
ment, as well as the physicochemical interactions of constituents dance determined by the proportions of different components.
determine their drug solubilization capacity and physical stabil- In addition, the size and shape of oil molecules relative to the
ity during storage and upon dilution. hydrophobic region of the surfactant determine the extent of oil
entrapment in the surfactant layer.
1.2. Components of micelles and microemulsions Microemulsion formation is a function of composition of the
system. The composition ranges with respect to the number of
Pharmaceutical microemulsions are typically composed of phases that exist in a system are graphically demonstrated as a
oil and surfactant in water, and often also include a cosurfactant phase diagram. A ternary phase diagram, with three corners of
and/or a cosolvent. SMEDDS contain the non-aqueous com- a triangle representing three components of a system, describes
ponents of microemulsions and readily disperse upon dilution phase regions. A pseudo-ternary phase diagram is used for sys-
in aqueous phase with mild agitation to form microemulsions. tems of more than three components, when the ratio of at least
SMEDDS are often preferred over microemulsion formulations two of the components is kept constant and represented by one
for hydrolytically sensitive drugs and their low volume enables of the axis of the triangle. A hypothetical phase diagram of a
packing into soft gelatin capsules for oral administration. three component system is presented in Fig. 1, representing
The surfactant used in microemulsion formation could be oil, water, and emulsifier as the three phases of the triangle
ionic or nonionic, which determines the stabilizing interactions (Prince, 1975). At different concentrations of each component,
of the hydrophilic end of the surfactant with the aqueous phase. macroemulsions or emulsions, micelles, or inverted micellar
Thus, while a nonionic surfactant is stabilized by dipole and structures are formed. The L1 and L2 phases in these diagrams
hydrogen bond interactions with the hydration layer of water correspond to the normal and inverted micelles, and correspond-
on its hydrophilic surface, an ionic surfactant is additionally ing types of microemulsions, respectively. The microemulsions
stabilized by the electrical double layer. Thus, the effect of salt transition into each other with variation in composition through
12 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

Table 1
Examples of surfactants, cosurfactants, and cosolvents used in commercial lipid-based formulations
Excipient name (commercial name) Examples of commercial products in which it has been used

Surfactants/cosurfactants
Polysorbate 20 (Tween 20) Targretin soft gelatin capsule
Polysorbate 80 (Tween 80) Gengraf hard gelatin capsule
Sorbitan monooleate (Span 80) Gengraf hard gelatin capsule
Polyoxyl-35-castor oil (Cremophor EL) Gengraf hard gelatin capsule, Ritonavir soft gelatin capsule
Polyoxyl-40-hydrogenated castor oil (Cremophor RH40) Neoral soft gelatin capsule, Ritonavir oral solution
Polyoxyethylated glycerides (Labrafil M 2125Cs) Sandimmune soft gelatin capsules
Polyoxyethylated oleic glycerides (Labrafil M 1944Cs) Sandimmune oral solution
d-␣-Tocopheryl polyethylene glycol 1000 succinate (TPGS) Agenerase soft gelatin capsule, Agenerase oral solution
Cosolvents
Ethanol Neoral soft gelatin capsule, Neoral oral solution, Gengraf hard gelatin capsule,
Sandimmune soft gelatin capsule, Sandimmune oral solution
Glycerin Neoral soft gelatin capsule, Sandimmune soft gelatin capsule
Propylene glycol Neoral soft gelatin capsule, Neoral oral solution, Lamprene soft gelatin capsule,
Agenerase soft gelatin capsule, Agenerase oral solution, Gengraf hard gelatin
capsule
Polyethylene glycol Targretin soft gelatin capsule, Gengraf hard gelatin capsule, Agenerase soft gelatin
capsule, Agenerase oral solution
Lipid ingredients
Corn oil mono-, di-, tri-glycerides Neoral soft gelatin capsule, Neoral oral solution
dl-␣-Tocopherol Neoral oral solution, Fortovase soft gelatin capsule
Fractionated triglyceride of coconut oil (medium-chain triglyceride) Rocaltrol soft gelatin capsule, Hectorol soft gelatin capsule
Fractionated triglyceride of palm seed oil (medium chain triglyceride) Rocaltrol oral solution
Mixture of mono- and di-glycerides of caprylic/capric acid Avodart soft gelatin capsule
Medium chain mono- and di-glycerides Fortovase soft gelatin capsule
Corn oil Sandimmune soft gelatin capsule, Depakene capsule
Olive oil Sandimmune oral solution
Oleic acid Ritonavir soft gelatin capsule, Norvir soft gelatin capsule
Sesame oil Marinol soft gelatin capsule
Hydrogenated soybean oil Accutane soft gelatin capsule, Vesanoid soft gelatin capsule
Hydrogenated vegetable oils Accutane soft gelatin capsule, Vesanoid soft gelatin capsule
Soybean oil Accutane soft gelatin capsule
Peanut oil Prometrium soft gelatin capsule
Beeswax Vesanoid soft gelatin capsule

Fig. 1. (A) A hypothetical ternary phase diagram representing three components of the system (water, emulsifier (E), and oil) as three axis of an equilateral triangle.
Different compositions of the formulation result in the formation of different phase structures: normal micellar solution, inverted micellar solution, macroemulsions or
emulsions, o/w microemulsions, w/o microemulsions, and various transition phases represented by cylinders and lamellae structures. The conventionally designated
L1 phase consists of micelles and o/w microemulsions while the L2 phase consists of inverted micelles and w/o microemulsions (Prince, 1975). (B) Schematic
representation of the dispersed phase structure of micelles, reverse micelles, o/w microemulsions, and w/o microemulsions.
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 13

intermediate liquid crystalline phases, which are viscoelastic upon dilution with water to form o/w microemulsions, which can
gels composed of hexagonal array of water cylinders adjacent be done by dropwise addition, static serial dilution, or dynamic
to the w/o phase and a lamellar phase of swollen bimolecular injection (Li et al., 1998). Modified in vitro tests can be used
leaflets adjacent to the o/w phase (Prince, 1975). These phases for more accurate assessment of tendency for drug precipita-
are characterized by the presence of birefringence, as opposed to tion (Gao et al., 2004; Gao et al., 2003). Solubilization capacity
microemulsion regions which are optically isotropic. Incorpo- of the drug is measured by saturation solubility evaluation in
ration of cosurfactant and/or cosolvent increases the one-phase different components and component mixtures (Aramaki et al.,
region. Construction of phase diagrams enables determination 2001).
of aqueous dilutability and range of compositions that form a Drugs can be incorporated in microemulsions by the phase
monophasic region. inversion temperature (PIT) method (Brime et al., 2002) and
in SMEDDS by dissolving the drug in the hydrophilic or the
1.3. Characterization of microemulsions hydrophobic component(s). The PIT method involves mixing
drug solution with microemulsions and applying heat to form
Characterization of reverse micelles, SMEDDS, and transparent drug loaded systems. In addition, drug release rate
microemulsions involves the physical and chemical tests related studies may be carried out, when desired, in Franz diffusion
to oral liquid dosage forms, e.g., assay, uniformity of content, cell across the donor and acceptor compartments separated by
stability of the active (impurities), appearance, pH, viscosity, a semipermeable membrane (Peltola et al., 2003; Spiclin et al.,
density, conductivity, surface tension, size and zeta poten- 2003) or using US Pharmacopeial methods for dissolution test-
tial of the dispersed phase, etc. with respect to the effect ing (Porter and Charman, 2001).
of the composition on physical parameters (Podlogar et al.,
2004). Additionally, differential scanning calorimetry (DSC) 1.4. Drug entrapment and structure
provides information on the interactions of different compo-
nents and polarization microscopy using crossed polarizers is Location of the solubilized drug in microemulsion systems
employed to confirm isotropicity of the formulation (Neubert depends on the hydrophobicity and structure of the solute.
et al., 2005). Size of the dispersed phase in o/w microemul- Enhanced drug solubility in microemulsion and micellar sys-
sions has been measured by photon correlation spectroscopy tems usually arises from the solubilization at the interface. The
(PCS) and total-intensity light scattering (TILS) techniques interface-associated solute, in turn, may affect the size and
(Malcolmson et al., 2002). The use of scattering techniques, e.g., shape of the microemulsion droplets. For example, incorporation
static light scattering (SLS), dynamic light scattering (DLS), of hydrophobic amino acids in di-2-ethylhexyl sulfosuccinate
and small-angle neutron scattering (SANS), for dispersed (AOT) reverse micelles (Leodidis and Hatton, 1990a; Leodidis
phase size measurement requires correction for non-ideality and Hatton, 1990b; Leodidis and Hatton, 1991a; Leodidis and
of the hard sphere model arising from interparticle interac- Hatton, 1991b) and w/o microemulsions (Yano et al., 2000)
tions in concentrated microemulsions (Shukla et al., 2002; leads to their association at the interface, and they may act as
Shukla et al., 2003). Structural features of microemulsions cosurfactants. Upon comparing the solubilization of glycine,
have been studied using self-diffusion nuclear magnetic reso- l-histidine, and l-phenylalanine in AOT stabilized water-in-
nance (SD NMR) (Spernath et al., 2003; Johannessen et al., isooctane microemulsions, Yano et al. observed that hydrophilic
2004) and small-angle X-ray scattering (SAXS) (Garti et al., amino acid glycine was solubilized primarily in the dispersed
2006). aqueous phase while hydrophobic amino acids, l-histidine and
During the development of these systems, pseudo-ternary l-phenylalanine, migrated to the AOT interface layer (Yano et
phase diagrams are constructed by titrating a reverse micelle al., 2000). Furedi-Milhofer et al. obtained similar results with the
mix with one of the components and observing visually for solubilization of aspartame in water/isooctane/AOT microemul-
transparency and through crosspolarizers for optical isotropy sions (Furedi-Milhofer et al., 2003). Aspartame was solubilized
(Moreno et al., 2003). Maintenance of monophasic character- at the interface and resulted in a sharp reduction of surface ten-
istics and drug solubility is tested upon dilution with water. sion depending on aspartame concentration, indicating its role
Phase stability of formed microemulsions is evaluated by accel- as a cosurfactant.
erated tests such as centrifugation or freeze thaw cycles (Brime The maximum amount of solubilized hydrophobic drug is
et al., 2002). Partitioning behavior of drug in the dispersed dependent on the curvature of the interface. Surfactant layer
phase of these systems has been studied by electrokinetic chro- on the interface has a positive curvature towards the dispersed
matography (EKC) for both micelles (Ishihama et al., 1994) phase, which is determined both by the relative volume of
and microemulsions (Huie, 2006), and by gel permeation chro- dispersed phase and the spontaneous curvature of surfactant
matography (GPC) in micelles (Scherlund et al., 2000). The log molecules. Entrapment of drug molecules in the interface is
of capacity factor obtained by EKC of hydrophobic compounds facilitated, leading to higher drug loading capacity, if the spon-
in microemulsions correlated well with their octanol water par- taneous curvature is lower than the actual curvature. Higher
tition coefficients (log P) (Mrestani et al., 1998). In addition, this spontaneous curvature, on the other hand, leads to lower drug
dosage form is tested to evaluate the tendency for drug precipita- loading capacity at the interface.
tion or crystallization by physical observation upon undisturbed Partitioning of the drug into the interface was quantified
storage at room temperature and refrigerated conditions, and by the interfacial partition coefficient by Leodidis and Hatton
14 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

(Leodidis and Hatton, 1990a). Using phase equilibrium analyses The beneficial effects of microemulsions in these applications
on the solubilization of amino acids in AOT reverse micelles, the were attributed to the prevention of degradation in the gastro-
authors showed that interfacial partition coefficient of the solute intestinal environment and the permeability enhancing effect of
depended weakly on surfactant concentration and did not depend the lipid components.
on solute concentration and aggregate geometry. It depended Microemulsion systems have also been claimed to improve
strongly on the factors that affect surface pressure or bending storage stability of proteins. For example, Owen and Yiv (US
moment of the surface film, e.g., solvent type and external elec- Patent #5,633,226) disclose improved chemical stability of
trolyte type and concentration. Also, Testard and Zemb showed a horse radish peroxidase after storage in w/o microemulsions as
general linear relationship between induced curvature variation compared to aqueous solution. In addition, w/o microemulsion-
and solute content of the interfacial film for a hydrophobic solute based media have been utilized for immobilization of water
using nonionic surfactant based o/w microemulsions (Testard soluble enzymes, such as lipase, in the internal, dispersed
and Zemb, 1999). aqueous phase for biocatalytic conversion of water-insoluble
These studies indicate that hydrophobic solute is solubilized substrates in the outer non-aqueous layer (Schuleit and Luisi,
at the interface of reverse micellar and microemulsion systems 2001; Madamwar and Thakar, 2004). In a similar application
and its solubility is affected by system variables that affect the of enhancing enzyme mediated catalysis of non-aqueous sub-
curvature of the interfacial film. Moreover, the presence of the strates, water soluble protein myoglobin was cross-linked to
solute itself affects the system, depending on the nature of the poly(l-lysine), which was in turn covalently attached to oxi-
solute and the surfactant. The phenomenon of drug solubiliza- dized cathode, in an o/w microemulsion environment such that
tion at the interface affects not only drug loading capacity but the protein was present in the water-rich external environment,
also drug precipitation upon dilution. For example, for a drug while the reactant, styrene, was present in the internal oil-rich
whose solubilization capacity at the interface has been increased environment. Catalysis of epoxidation of styrene by myoglobin
with the use of a cosurfactant, dilution with aqueous phase lead- in this system was higher than aqueous solution, which increased
ing to cosurfactant migration away from the interface can lead further in the presence of bicontinuous microemulsion system
to dramatic reduction in drug loading capacity, causing precipi- (Vaze et al., 2004).
tation. In all these applications hydrophilic peptides or proteins
were dissolved in the aqueous phase at or below their solubil-
1.5. Microemulsions for protein and peptide delivery ity levels. This review, however, will focus on solubilization
of hydrophobic molecules in SMEDDS and diluted o/w
Improvement in the oral bioavailability of hydrophobic cyclic microemulsions while preventing physical instability of drug
peptides, like cyclosporine A, using SEDDS and SMEDDS is separation by crystallization on storage or precipitation upon
discussed in Section 3.1 and Section 4.3. SMEDDS systems aqueous dilution, with particular relevance to oral administra-
have also shown promise in improving the oral bioavailability tion.
of hydrophilic linear peptides and proteins. For example, Cilek et
al. tested the oral absorption of recombinant human insulin dis- 2. Drug loading capacity in micelles and microemulsions
solved in the aqueous phase of w/o microemulsions composed of
Labrafil® , lecithin, ethanol, and water in streptozotocin-induced Pharmaceutical micellar and microemulsion systems are
diabetic male Wistar rats. The authors demonstrated significant usually formulated as oil + surfactant ± cosurfactant/cosolvent
improvement in oral pharmacological availability compared mixtures that exist as reverse micelles or w/o type microemul-
with insulin solution, although it was ∼0.1% compared with sions. These systems are diluted with water in vivo or before
sub-cutaneous administration (Cilek et al., 2005). On the other administration. Solubilization or drug loading capacity in these
hand, Kraeling and Ritschel found that the oral pharmacological systems refers to the drug concentration achievable in reverse
availability of insulin microemulsions as compared to intra- micelles and the ability of these systems to undergo aqueous
venous insulin in beagle dogs was 2.1%, which further increased dilution as monophasic systems.
to 6.4% with the encapsulation of gelled microemulsions in Drug precipitation from a self-emulsifying drug delivery sys-
hard gelatin capsules along with the protease inhibitor apro- tem is a consequence of concentration exceeding the equilibrium
tinin and coating of the capsules for colonic release (Kraeling solubilization capacity. Consequently, systems formulated to
and Ritschel, 1992). Improved oral delivery of insulin from have drug solubilization capacity much higher than the required
microemulsion system was also demonstrated by others (Cho concentration would be expected to show the least propensity
and Flynn, 1989). for precipitation in vivo. Drug loading or solubilization capac-
Improved oral bioavailability from the w/o microemulsion ity in the system also determines the minimum volume per unit
system was also shown for the linear water-soluble nonapeptide dose that can be formulated. Thus, an understanding of fac-
leuprolide acetate (Zheng and Fulu, 2006) and dipeptide N- tors influencing drug loading capacity while maintaining the
acetylglucosaminyl-N-acetylmuramic acid (Lyons et al., 2000). capability of the system to undergo monophasic dilution with
Also, intra-gastric administration of w/o microemulsion of epi- water and minimizing the tendency for drug precipitation or
dermal growth factor was more effective in healing acute gastric crystallization in diluted systems is essential to the design of
ulcers in rats as compared to both intra-peritoneal and intra- stable and appropriately low-volume systems for drug delivery
gastric aqueous solution administration (Celebi et al., 2002). applications.
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 15

2.1. Solubilization capacity in reverse micelles to be caused by the change in the locus of drug solubilization
associated with microstructural transitions during aqueous dilu-
Micellar and microemulsion systems are often able to solubi- tion (Spernath et al., 2003). In addition, migration of water
lize higher amount of drug than its individual components. For miscible cosurfactant away from the interface upon aqueous
example, Spernath et al. reported that the solubility of lycopene, dilution could lead to reduced drug solubilization capacity at the
a hydrophobic carotenoid obtained from tomatoes, in the reverse interface. Evaluation of drug solubilization capacity at different
micelles of (R)-(+)-limonene (limonene) and polysorbate 60 dilution levels allows the formulator to define the appropriate
(Tween 60® ) (4:6) was 2500 ppm, about three times higher than dilution range for a given formulation with minimum likelihood
in either individual component (700 ppm in (R)-(+)-limonene of drug precipitation.
and 800 ppm in Tween 60® ) (Spernath et al., 2002). Higher
solubilization capacity in reverse micellar systems was also 2.2. Dilutability as monophasic systems
noted for phytosterol, whose solubility was 150,000 ppm in the
reverse micelles of limonene and Tween 60® (4:6), about six An approach to improve the dilutability of drug containing
times higher than in either individual component (25,000 ppm surfactant/oil reverse micelles with aqueous phase is to expand
in each) (Spernath et al., 2003). This higher capacity for sol- the monophasic/isotropic region through a wide range of com-
ubilization was attributable to the interfacial locus of drug positions. When the expanded isotropic region covers aqueous
solubilization, which has higher solubilization capacity than the dilutability through a range of compositions with different water
core. Higher solubilization capacity at the interface is a function content, called ‘dilution line’, the systems so formed have been
of drug–surfactant interactions leading to drug association at called dilutable U-type microemulsions. An example of the role
the interface. These interactions depend on the hydrophobicity, of surfactant in determining the monophasic region and dilution
functional groups, and shape of both the drug and the surfac- line are represented in Fig. 2 (Spernath et al., 2006). The dilu-
tant/cosurfactant. The shape influences sub-molecular proximity tion line N73 in Fig. 2A represents 7:3 composition of the ethyl
or fit of interacting molecules to maximize interactions. Thus, laurate/acetic acid (1:3) and phosphatidyl choline (PC)/Tween
different excipients and different grades of similar excipients 80® /propylene glycol (PG) (1:3:10) axis in reverse micelles (in
can show markedly different solubilization capacity for a given the absence of water). Upon progressive addition of water, the
drug. system progresses to the third axis of the phase diagram along
The solubilization capacity progressively decreases upon the dilution line N73 through the monophasic region (Fig. 2A).
aqueous dilution, as the micellar system passes through swollen Therefore, both the composition of the formulation and the area
w/o reverse micelles, to bicontinuous phase, to o/w microemul- of the monophasic region are important to ensuring successful
sion system. This reduction in solubilization capacity is thought aqueous dilution without ‘breaking’ the microemulsions.

Fig. 2. Phase diagram of a 6-component system and factors influencing monophasic region. (A) demonstrates 1-phase and 2-phase regions of a ethyl laurate/acetic
acid (1:3) system stabilized with mixed surfactants PC/HECO40/PG (1:3:10) and an aqueous dilution line N73 from the non-aqueous reverse micelles to the water
axis. AT represents the percentage of monophasic region. (B) and (C) represent the influence of using Tween 60® (B) versus triglycerol monooleaste (C) on the
monophasic region. (D) represents the variation in the percentage of isotropic or monophasic region with the use of different chain length acid surfactants (Spernath
et al., 2006).
16 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

The role of HLB of the surfactant in determining the area of depending on the type of acid used. As shown in Fig. 2D, pro-
monophasic region is illustrated in an extreme case in Fig. 2B pionic acid was the most efficient in increasing the area of the
and C. The isotropic or single phase region of 5-component isotropic region in systems stabilized with PC, polyoxyethylene-
system composed of limonene, water, ethanol, propylene gly- 40-hydrogenated castor oil (HECO40 or Cremophor RH40® ),
col, and Tween 60® (Fig. 2B) reduced significantly when the and PG in 1:3:10 weight ratio. The area of isotropic region
hydrophilic surfactant, Tween 60® (HLB 14.9), was replaced progressively decreased with increasing carbon chain length of
with a hydrophobic surfactant, triglycerol monooleate (HLB 6.2) organic acid (Spernath et al., 2006). This behavior is similar to
(Fig. 2C) (Spernath et al., 2006). Aqueous dilution of reverse that observed with alcohols and is postulated to proceed through
micelles of the latter system would invariably result in ‘breaking’ similar mechanisms (Garti et al., 2001; Hou and Shah, 1987).
of the microemulsion system into two phases.
Certain formulation approaches can lead to increase in the 2.3. Solubilization capacity in diluted microemulsions
monophasic region. Addition of polyols, e.g., glycerin and
propylene glycol; short-chain alcohols, e.g., ethanol; and organic Drug solubilization capacity in microemulsions vis-à-vis
acids, e.g., propionic acid, increase the monophasic region of corresponding micelles and the oil used for solubilization
o/w microemulsions (Garti et al., 2001). These additives act was evaluated by Malcolmson et al. (1998). The authors
as cosolvents, by promoting solubility of the drug in the bulk used 2% o/w microemulsions and micelles of nonionic sur-
phase, and/or cosurfactants, by affecting interfacial structure and factant polyoxyethylene-10-oleyl ether (Brij 96) to solubilize
promoting drug solubility at the interface. the hydrophobic drug testosterone propionate (log P 4.78) and
Aqueous dilutability of w/o reverse micellar or microemul- studied the role of the type of oil on drug solubility in microemul-
sion systems proceeds through a series of structural changes sions. As shown in Table 2, drug solubility was higher in
from w/o to bicontinuous to o/w system, which concurrently microemulsions than corresponding micelles and the oil, which
involves changes in drug solubilization capacity. Factors affect- was attributed to drug solubilization in the interfacial surfactant
ing water solubilization capacity of w/o microemulsions before monolayer.
their breakdown into bicontinuous structures were reported by The type of oil significantly influenced drug solubility in
Hou and Shah (Hou and Shah, 1987). Addition of water to a microemulsions. This was due to oil penetration in the surfac-
w/o microemulsion system could result in water incorporation tant monolayer, causing a dilution of the polyoxyethylene region
in the dispersed phase. The growth of microemulsion droplets of the surfactant that lies close to the hydrophobic region and
without coalescence during this process is limited by either the contributes to drug solubility. Variations in the oil molecular vol-
radius of curvature of the interface or the attractive interac- ume, polarity, size, and shape led to variations in its penetration
tions among droplets (Hou and Shah, 1987). For the systems of the surfactant monolayer and influence on drug solubilization.
where solubilization capacity for water is limited by the curva- The authors concluded that the ability of an o/w microemulsion
ture of the interfacial layer, reduction in spontaneous curvature to increase drug solubility over the equivalent micelle depends
by modification of the interface or the continuous phase can on both the solubility of drug in the dispersed phase, influence
result in increased solubilization. For systems where solubiliza- of oil on the nature of microemulsion droplet, and the site of
tion capacity is limited by the critical droplet radius, reduction in drug solubilization within the surfactant aggregate. The use of
attractive forces among droplets would increase the solubiliza- large molecular volume polar oils, e.g., caprylic acid triglyc-
tion capacity of water (Hou and Shah, 1987). These principles erides (Miglyol 812® ), was recommended to maximize drug
provide useful insights to the analogous scenario of solubi- solubilization in microemulsions.
lization of hydrophobic solute in the dispersed phase of o/w The role of surfactant type and percent aqueous phase
microemulsions. Thus, incorporating components that increase composition on the solubilization capacity in diluted o/w
the spontaneous curvature and/or increase solute–interface inter- microemulsions was reported by Spernath et al. (2002). Solubi-
actions can be useful in increasing drug solubilization while lization of lycopene in microemulsions stabilized by different
maintaining monophasic characteristics of the system. surfactants in 25% limonene/ethanol/Tween 60® (1:1:3 and
By partitioning into the interface, short-chain alcohols and 1:1:8) and 75% water containing o/w microemulsions was
acids alter the molecular structure of the interface and decrease a function of the HLB of surfactants (Fig. 3A). Maximum
the spontaneous curvature, thus leading to higher solubilization lycopene solubilization was observed using Tween 60® (HLB
capacity for the dispersed phase. In reverse micelles, when the 14.9), which reduced dramatically when more hydrophilic sur-
system is rich in oil and poor in surfactants, the surfactant mix- factants, e.g., Tween 40® and Tween 20® (HLB 16.7) were used
ture has a tendency to partition mainly into the oil phase and its (Spernath et al., 2002). This indicated a suitable range of HLB
level at the interface is below the concentration that is needed to of surfactant or system to maximize drug solubilization. This
form a large area of w/o microemulsions. Ethanol, however, has range could be drug specific, but is usually 10–16.
a tendency to penetrate the interface at low surfactant content to Solubilization capacity of lycopene was also dependent on the
form mixed films (Spernath et al., 2006). Thus, ethanol enlarges aqueous phase dilution of a 1:1:3 mixture of limonene, ethanol
the isotropic region by increasing the flexibility of the surfactant and Tween 60® (Fig. 3B). Four different regions were identified
film. in terms of lycopene solubilization capacity along the aqueous
Use of organic acids as a cosurfactant also leads to signifi- dilution line. The solubilization capacity decreases dramatically
cant increase in the isotropic region of microemulsion formation upon increasing aqueous phase content of the system from 0
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 17

Table 2
Solubility of testosterone propionate in micelles, various oils, and corresponding microemulsions at two different surfactant (Brij 96) concentrations
Oil type Solubility in oil (%w/w) Drug contribution from oil content to Solubility in micelles/microemulsions (%w/v) at surfactant level of
the solubility in microemulsions
15% 20%

Micelles – 0.000 0.365 0.430

Tributyrin 8.78 0.176 0.553 0.641
Miglyol 812 6.20 0.124 1.150 1.300
Soybean oil 3.42 0.068 0.531 0.656
Ethyl butyrate 18.64 0.373 0.471 0.486
Ethyl caprylate 12.17 0.243 0.489 0.599
Ethyl oleate 5.79 0.116 0.497 0.641
Heptane 0.92 0.018 0.354 0.486
1-Heptene 4.28 0.086 0.402 0.424
Hexadecane 1.70 0.034 0.431 0.520
1-Hexadecene 1.74 0.035 0.389 0.573

Abbreviations: DMTG: dimethoxytetraethylene glycol. Note: Table modified from Malcomson et al. to report only mean values. Solubility in water 0.009% (w/w).

to 20% (region I), remains almost unchanged from 20 to 50% 3. Drug precipitation and solute crystallization
(region II), increases again from 50 to 67% (region III), and then
reduces upon further dilution (region IV). Drug precipitation upon oral administration and in vivo dilu-
Solubilization capacity of lycopene was related to the struc- tion of a SEDDS or SMEDDS formulation is a rapid process that
tural transitions taking place during aqueous dilution of the involves solute exclusion from the solution whose solubilization
reverse micelle system. Structural transitions in the system were capacity for the drug has suddenly reduced. In addition to the
studied by self-diffusion nuclear magnetic resonance (SD NMR) drug and formulation variables, this process is affected by con-
to calculate diffusion coefficients of water and limonene in ditions in the gastrointestinal tract and the fate of lipids upon
systems with and without lycopene, as a function of aqueous coming in contact with gastrointestinal fluids. Approaches to
dilution. The decrease in drug solubilization capacity in region minimize and models to mimic in vivo drug precipitation could
I was related to increasing interactions between the surfactant be helpful in improving bioavailability from these systems.
and water molecules, with a gradual swelling of reverse micelles, In contrast, in vitro drug crystallization from diluted micelles
leaving less surfactant available for interaction with the solute. and microemulsions involves formation of solute crystals over
Region II was associated with gradual transformation of the prolonged undisturbed storage. This process is usually slow,
system into a biocontinuous phase structure, while the interfa- temperature dependent, and influenced by such factors gov-
cial area remains almost unchanged. Over region III, the system erning crystallization as saturation solubility of the drug in the
changed from a bicontinuous to an o/w microstructure, which system. A system with lower drug solubility will show higher
was strengthened in region IV (Spernath et al., 2002). These propensity for crystallization, and vice versa. A comparison of
results indicate that the amount of aqueous phase dilution influ- tendency of several formulations to crystallize over time can
ences solute solubilization capacity upon dilution of the reverse be observed upon undisturbed storage of samples under refrig-
micelles to o/w microemulsions, which is related to the structural erated conditions, which accelerates solute crystallization, or
state of the system. Assuming fasted state gastric fluid volume by using modified in vitro tests (Gao et al., 2004; Gao et al.,
of ∼50 mL, SMEDDS that show highest solubilization capacity 2003). Therefore, modeling in vitro drug crystallization can help
at this dilution would, therefore, be expected to have the least develop ready-to-use oral and parenteral microemulsion dosage
tendency for drug precipitation in vivo. forms of drugs.

Fig. 3. Solubilization capacity in microemulsions as a function of surfactant type and aqueous dilution. (A) represents the solubilization capacity of lycopene in
microemulsions of composition (1, solid bars) (R)-(+)-limonene/ethanol/Tween 60® (1:1:3) and 75% aqueous phase and (2, hatched bars) limonene/ethanol/Tween
60® (1:1:8) and 75% aqueous phase. (B) represents lycopene solubilization as a function of aqueous weight percent in the microemulsions in relation to the structural
transition regions of the microemulsion (Spernath et al., 2002).
18 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

3.1. In vivo drug precipitation model (Sek et al., 2002) incorporates the use of temperature,
enzymes, and pH control to simulate in vivo conditions, fol-
Lipid solutions often achieve higher oral absorption than lowed by ultracentrifugation, and separation of the formulation
corresponding solid dosage forms of hydrophobic drugs (Shen into three phases: an aqueous phase containing bile salts, fatty
and Zhong, 2006), particularly class II (low solubility, high acids, and monoglycerides along with dissolved drug (which
permeability) compounds as per the biopharmaceutics classifi- is considered available for absorption), a lipid phase contain-
cation system (Lindenberg et al., 2004). However, improvement ing undigested diglycerides and triglycerides, and a sediment
of bioavailability upon presenting a hydrophobic drug in the containing undissolved fatty acids (Dahan and Hoffman, 2006).
solution or emulsion form can be compromised if the drug Fig. 4A represents the distribution of vitamin D3 molecules
precipitates from the dosage form in vivo. In several cases, across the aqueous and sediment phase using long-chain triglyc-
avoidance of drug precipitation could be the predominant factor erides (LCT) and medium chain triglycerides (MCT) in the
governing improvement of oral bioavailability from lipid vehi- formulation. Upon 5-fold reduction of the amount of lipid in the
cles than the size of the dispersed phase. The SEDDS, SMEDDS, formulation, drug precipitation was evident with increasing per-
and micellar systems have different levels of drug dispersion. centage of drug in the sediment (Fig. 4B). This experiment shows
The dispersion size, upon in vivo dilution and bile-surfactants that in vitro simulation studies could be extrapolated to evaluate
induced emulsification, of SMEDDS is expected to be smaller the in vivo drug precipitation tendency of the formulation.
than that of SEDDS, which, in turn, would be smaller than that
of a lipid-solution of drug. The influence of dispersion size on 3.3. Avoiding in vivo drug precipitation
bioavailability has been observed for several molecules, e.g.,
Vitamin E (Julianto et al., 2000), cyclosporine (Trull et al., 1995), Increasing the solubilization capacity of the formulation sig-
and halofantrine (Khoo et al., 1998); while it is limited for some nificantly over the desired drug concentration could help avoid
others, e.g., atovaquone (Sek et al., 2006), danazol (Porter et al., in vivo drug precipitation. Formulations that can be diluted with
2004), and ontazolast (Hauss et al., 1998) (Table 3). water in vitro without drug precipitation are likely to be more
For example, the self-emulsifying formulations had equiva- stable under in vivo conditions than those that are not dilutable.
lent bioavailability to corresponding lipid-solution formulations These aspects are discussed in Section 2.
for atovaquone (log P 5.31) (Sek et al., 2006) and danazol (log P Another approach in this direction is to promote the forma-
4.53) (Porter et al., 2004) in dogs, and for ontazolast (log P 4.00) tion of supersaturated drug solution in vivo by incorporation of
(Hauss et al., 1998) in rats. The bioavailability of all these formu- hydrophilic polymeric ingredients in the formulation that act
lations was higher than the corresponding aqueous suspensions. as precipitation inhibitors. The supersaturated drug solutions
These studies suggest that the role of dispersion size in improv- will eventually precipitate due to the thermodynamic instability
ing oral bioavailability could be limited depending on the drug, of the system, but if the precipitation is delayed long enough
the animal species, or other overriding factors. in vivo to cover the drug absorption time, bioavailability from
Presentation of a hydrophobic drug in a dissolved form these systems can be improved. Several common pharmaceutical
improves oral absorption as compared to a corresponding solid excipients act as precipitation inhibitors, e.g., methyl cellulose
or suspension dosage form by avoiding the dissolution step. In (MC), hydroxypropyle methylcellulse (HPMC), HPMC phtha-
all cases, lack of in vivo precipitation plays a predominant role in late (HPMCP), sodium carboxymethyl cellulose (Na CMC), and
improving oral bioavailability of hydrophobic compounds. The polyvinylpyrrolidone (PVP) (Hasegawa et al., 1988; Raghavan
assessment and minimization of the tendency for precipitation et al., 2001a; Raghavan et al., 2000; Raghavan et al., 2001b;
of drugs, both in vivo and in vitro, upon aqueous dilution of Simonelli et al., 1970). For example, Gao et al. demonstrated
dosage forms is important to their utilization in improving the the improved oral bioavailability of an experimental hydropho-
oral bioavailability of hydrophobic drugs. bic drug, PNU-91325, with the use of 20 mg/g HPMC in
the formulation using both cosolvent and SEDDS formulation
3.2. Prediction of in vivo drug precipitation approaches. The bioavailability improvement with the incorpo-
ration of HPMC in a PEG 400 cosolvent-based formulation
Development of a lipid formulation of a hydrophobic com- was >4-fold, while it was ∼2-fold for supersaturable SEDDS
pound presents overabundance of choices of vehicles (de Smidt formulation using Cremophor EL® compared with a micelle
et al., 2004) and the development strategies are mostly empirical formulation using Tween 80® (Gao et al., 2004). In application
(Dahan and Hoffman, 2006). Formulation choices can be com- to SMEDDS formulation, inclusion of HPMC was demonstrated
pared with respect to their tendency towards drug precipitation to increase the bioavailability of paclitaxel more than 9-fold in
in vivo by such empirical tests as dilutability in water in vitro rats (Gao et al., 2003).
and the rate of drug crystallization.
The tendency for in vivo drug precipitation in a formulation 3.4. Mechanism of solute crystallization
is often also evident in absorption simulation experiments. For
example, Dahan and Hoffman used an in vitro lipolysis model to The efficiency of a system to solubilize drug is commonly
perform in vitro in vivo correlation (IVIVC) between lipolysis of interpreted in terms of the amount of drug dissolved over a short
solubilized lipophilic solute, vitamin D3 , and oral bioavailabil- period of time with reasonable degree of agitation. Whether
ity (Dahan and Hoffman, 2006). The dynamic in vitro lipolysis nucleation and crystallization would subsequently occur in such
Table 3
Relative bioavailability of lipid-based formulations of hydrophobic drugs
Drug name (log P value) Species tested Test product Reference product Increase in AUC

Formulation AUC (Mean ± S.D.) Formulation AUC (Mean ± S.D.)

Vitamin E (log P 9.96) Humans Tween 80, Span 80, and AUC0−∞ = 210.7 ± 63.0 h ␮g/mL Natopherol® soft AUC0−∞ = 94.6 ± 80.0 h ␮g/mL ∼2-fold
Vitamin E dissolved in palm gelatin capsules
oil in the proportion 4:2:4 to (solution in soybean oil)

A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

form SEDDS
Cyclosporine (log P 4.29) Humans SMEDDS, Neoral® soft SEDDS, Sandimmune® ∼6.5-fold
gelatin capsules soft gelatin capsules
Halofantrine (log P 9.20) Dogs SEDDS, MCT AUC0−∞ = 5313 ± 1956 h ng/mL SMEDDS, MCT AUC0−∞ = 5426 ± 2481 h ng/mL None
SMEDDS, LCT AUC0−∞ = 6973 ± 2388 h ng/mL ∼1.3 fold
Atovaquone (log P 5.31) Dogs Solution in lipids + ethanol AUC0−73h = 31.8 ± 9.3 h ␮g/mL Aqueous suspension AUC0−73h = 9.4 ± 1.0 h ␮g/mL ∼3.4-fold
SMEDDS, AUC0−73h = 31.8 ± 8.4 h ␮g/mL ∼3.4-fold
lipids + Cremophor
EL® + ethanol
SMEDDS, lipids + Pluronic AUC0−73h = 33.7 ± 13.0 h ␮g/mL ∼3.4-fold
121® + ethanol
Danazol (log P 4.53) Dogs SMEDDS, LCT AUC0−10h = 270.5 ± 38.5 h ng/mL Micronized powder AUC0−10h = 35.3 ± 5.2 h ng/mL ∼7-fold
SMEDDS, MCT AUC0−10h = 47.7 ± 29.5 h ng/Ml ∼1.3-fold
Lipid solution, LCT AUC0−10h = 340.2 ± 64.4 h ng/mL ∼9-fold
Ontazolast (log P 4.00) Rats SEDDS, 1:1 mix of Gelucire AUC0−8h = 752 ± 236 h ng/mL Aqueous suspension, AUC0−8h = 65 ± 15 h ng/mL ∼11-fold
44/14® and Peceol® Tween 80® + HPMC
SEDDS, 8:2 mix of Gelucire AUC0−8h = 877 ± 104 h ng/mL ∼13-fold
44/14® and Peceol®
SEDDS, Peceol® AUC0−8h = 528 ± 68 h ng/mL ∼8-fold
Emulsion, soybean AUC0−8h = 1003 ± 270 h ng/mL ∼15-fold
oil + Tween 80®
Atorvastatin (log P 6.26) Dogs SMEDDS, Labrafil® , AUC0−24h = 2613.0 ± 367.6 h ng/mL Lipitor® Tablets 10 mg AUC0−24h = 1738.0 ± 207.9 h ng/mL ∼1.5-fold
Cremophor RH40® ,
propylene glycol
SMEDDS, Estol® , AUC0−24h = 2568.3 ± 408.0 h ng/mL Lipitor® Tablets 10 mg AUC0−24h = 1738.0 ± 207.9 h. ng/mL ∼1.5-fold
Cremophor RH40® ,
propylene glycol
SMEDDS, Labrafac® , AUC0−24h = 2520.81 ± 308.4 h ng/mL Lipitor® Tablets 10 mg AUC0−24h = 1738.0 ± 207.9 h ng/mL ∼1.5-fold
Cremophor RH40® ,
propylene glycol

Abbreviations: LCT, long-chain triglycerides; MCT, medium chain triglycerides.

19
20 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

Fig. 4. Distribution of Vitamin D3 molecules across the aqueous phase and the sediment of the dynamic in vitro lipolysis medium using high (A) or 5-times lower
(B) lipid load of its long-chain triglyceride (LCT) or medium-chain triglyceride (MCT) solution. Modified from Dahan and Hoffman (Dahan and Hoffman, 2006).

a system depends on relative levels of drug solubilized vis-à-vis Assuming the saturation concentration of drug in the system
its saturation concentration in the system. Above saturation con- with dilution follow the double lines as marked, reduction in
centration, the rate of nucleation would depend on actual solute drug concentration with dilution in the formulation would lead to
concentration in the system and other factors, e.g., seed crystals, tendency for precipitation along either of lines 1, 2, or 3 depend-
leading to either immediate or delayed drug precipitation. ing upon the starting drug concentration in the system. Based on
Principles governing solute precipitation with progressively the amount by which drug concentration in the system exceeds
increasing concentration in solution were elaborated by LaMer the saturation concentration and the length of dilution line along
and Dinegar in the study of formation of monodisperse col- which it exceeds, dilution along line 1 would be expected to
loids (LaMer and Dinegar, 1950). In the classical LaMer lead to faster drug precipitation than line 2, while a system
diagram, solute concentration progressively increases in solu- diluted along line 3 would be expected to maintain the drug
tion beyond saturation concentration until it reaches a threshold in the solubilized state throughout.
for nucleation (the concentration that would lead to immediate, Formulation modifications tend to influence the saturation
heterogeneous nucleation and solute precipitation). Thereafter, drug concentration in the SMEDDS as well as upon dilution.
crystal growth occurs on the formed nuclei leading to reduc- Thus, in addition to formulation approaches to minimize and
tion of solution concentration until the saturation concentration inhibit drug precipitation, starting drug concentration plays a
is reached (Fig. 5). Nucleation can occur heterogeneously on crucial role in determining the window of permissible drug con-
impurity centers or homogeneously through spontaneous nucle- centrations upon dilution that do not lead to precipitation.
ation. The former leads to fewer, larger crystals than the latter
(Beattie, 1989). 3.5. Preventing drug crystallization
This principle could be extrapolated to the hypothetical sce-
nario of drug concentration in micellar and microemulsion High solubilization capacity of reverse micelles, however,
systems as illustrated in Fig. 6. This figure represents drug con- is of limited use in improving oral bioavailability if aqueous
centration (y-axis) in a reverse micelle upon progressive dilution phase dilution were to cause migration of the solubilized drug
with water (x-axis) to form an o/w microemulsion. Saturation molecule from interface to the outer aqueous phase, followed by
drug concentration in the system upon dilution is non-linear
(Garti et al., 2006; Spernath et al., 2002; Spernath et al., 2003).

Fig. 5. LaMer diagram representing the time dependence of concentration Fig. 6. A hypothetical set of scenarios for SEDDS, SMEDDS, and micellar
required for monodispersity. This figure illustrates the supersaturation region systems depicting different possibilities for drug supersaturation upon aqueous
of drug solubility between the saturation and the concentration that would lead dilution. With the defined saturation drug concentrations at each composition of
to immediate, heterogeneous nucleation in the case of monodisperse colloids the system over the dilution curve, different starting drug concentrations would
(LaMer and Dinegar, 1950). lead to different outcomes in drug precipitation upon dilution.
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 21

drug precipitation, and uncontrolled absorption (Spernath et al., lization at a given temperature. In the o/w microemulsions
2006). It is important, therefore, to develop systems that main- solubilizing a hydrophobic solute, the primary location of drug
tain high drug solubilization upon aqueous dilution of reverse in the system would influence the preferred site of nucleation. In
micelles. cases where drug resides at the interface along with surfactant
The problem of drug crystallizing out of solution upon aque- (and sometimes also cosurfactant) molecules, molecular pack-
ous dilution of systems that form micelles, emulsions, and ing and structure of the amphiphilic surfactant and drug at the
microemulsions has been widely discussed in several patent interface would play a role in facilitating or inhibiting nucle-
documents, which also discuss ways to address this issue. ation. For example, resemblance of molecular structure of the
Drug crystallization of aqueous oil/surfactant solutions of the emulsifier to that of the crystallizing solute, which affects prox-
hydrophobic drug fenofibrate (log P 5.58) was assessed by sim- imity and packing of solute molecules, could increase nucleation
ple physical observation of appearance of crystals immediately and the rate of crystallization (Davey et al., 1996). Therefore,
upon addition of water (US 2004/0005339 A1). The authors choice of a surfactant with reference to its molecular structure
proposed the use of a water-miscible solubilizer that allows resemblance to that of the hydrophobic solute could influence
complete drug dissolution and prevents or minimizes drug crys- the rate of drug crystallization from a microemulsion.
tallization in the formulation upon coming in contact with an
aqueous environment. Liang et al. (US 7,022,337 B2) extended 3.6. Combined use of solubilization approaches
the observation for possible crystallization up to 24 h. The use of
solubilizers such as N-alkyl derivatives of 2-pyrrolidone, ethy- A combination of pH control with the use of micelliza-
lene glycol monoether, C8–12 fatty acid esters of polyethylene tion, cosolvency, or complexation is the first choice approach to
glycol helped maintain drug in solution upon dilution with water. increase the solubility of hydrophobic drugs. Theoretical treat-
Another approach that has been proposed to prevent the pre- ment of the increase in solubility observed with a combination of
cipitation of drug upon aqueous dilution is to balance the HLB pH and other approaches has involved segregation of the contri-
value of surfactants used in the formulation. Preferentially water- bution of the ionized and the unionized species to solubilization
soluble surfactants have an HLB value of greater than 10, while (Li et al., 1999a). The increase in solubility achieved with a
surfactants that have higher solubility in oil have a value of less combination of cosolvent (ethanol) or micellization (polysor-
than 10. Chacra-Vernet et al. describe in US patent application bate 20) with pH modulation was demonstrated by Li et al. using
2004/0052824 A1 that the risk of recrystallization of drug is flavopiridol as a model compound, which is weakly basic with
the greatest when using hydrophilic SEDDS, i.e., which con- an apparent pKa of 5.68 and intrinsic solubility of 0.025 mg/mL
tain a hydrophilic surfactant and co-surfactant with having HLB (Li et al., 1999b). Flavopiridol solubility increased linearly with
values greater than 12. Although these formulations do help to the increase in surfactant content of solution, with a slope that
solubilize hydrophobic drugs, they may not lead to the desired increased with the reduction in pH. In contrast, increasing the
improvement in bioavailability. To prevent crystallization of the proportion of cosolvent led to logarithmic increase in flavopiri-
drug upon aqueous dilution, these authors proposed the use of dol solubility at all pH conditions, with the greatest increase at
small quantities of lipophilic phase with very low HLB values, acidic pH. These approaches may be incorporated in microemul-
and the essential presence of a cosurfactant which is also a good sion formulation to increase the saturation concentration and
solvent for the drug. solubilization capacity of the system.
The tendency for solute crystallization is amply demonstrated Aqueous solubility of a nonelectrolyte is also influenced by
in studies that have deliberately sought to achieve new crys- both the type and concentration of the electrolyte present in
tal forms of molecules by using microemulsions. For example, solution. The reduction in solubility of a hydrophobic drug in
Furedi-Milhofer et al. prepared new polymorphs of aspartame the presence of a salt or electrolyte is a function of salt con-
by crystallization from microemulsions (Furedi-Milhofer et al., centration, as described by the Setschenow equation (Ni et al.,
1999). The authors produced water/isooctane microemulsions 2000). This “salting-out” effect of electrolytes is also depen-
of the artificial sweetener aspartame using diisooctyl sulfosucci- dent on the molar volume, aqueous solubility, and the log P of
nate as a surfactant. Amount of surfactant and temperature were the solute (Shukla et al., 2003). Presence of electrolytes and
the primary factors determining the amount of aspartame which salts also affects the critical micellar concentration (CMC) of
could be solubilized. Aspartame was primarily located at the surfactants and the structure of micelles and microemulsions.
water/oil interface and acted as a cosurfactant. Crystallization These considerations should be taken into account with the use
of aspartame was achieved by slow cooling of the microemul- of ionized pharmaceutical excipients in these formulations.
sion to 5 ◦ C. For drugs solubilized in the w/o microemulsions,
nucleation could occur in either the dispersed water droplets 4. Other factors influencing bioavailability
or at the interface. The type of crystals formed depends on the
location of the drug in the system. Crystallization at the inter- In addition to drug precipitation in the gastrointestinal
face leads to the formation of long crystals, while crystallization tract, drug bioavailability from self-emulsifying formulations
initiated in the dispersed phase results in short crystals. is influenced by biopharmaceutical properties of the lipid,
For pharmaceutical applications, preventing the crystalliza- e.g., lipolysis; and the drug, e.g., lymphatic transport, enteric
tion is the desired goal. The tendency for crystallization is metabolism, and efflux. Lipid-based formulations can influence
reflected in the crystallization temperature or time to crystal- the bioavailability of hydrophobic drugs through several mech-
22 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

anisms, e.g., stimulation of pancreatic and biliary secretions, increase the bioavailability for drugs which are known substrates
prolongation of gastrointestinal residence time, stimulation of of P-gp efflux pumps.
lymphatic transport, increased intestinal wall permeability, and
reduced metabolism and efflux pump activity. 4.3. Dispersion size of emulsions

4.1. Lymphatic transport and lipolysis Presenting the drug in the dissolved form using lipid-based
formulations provides significant improvement of oral absorp-
Lipid digestion in the formulation increases the disper- tion as compared to an oral solid or suspension dosage form. This
sion of the drug, which promotes its absorption. Lipolysis advantage can be further improved in several cases by reducing
rate of medium chain triglycerides (MCT) is higher than the dispersion size of the dosage form. The reduction in disper-
long-chain triglycerides (LCT), which has been shown to influ- sion size of cyclosporine A (log P 4.29) SEDDS formulation,
ence the bioavailability of hydrophobic drugs from lipid-based Sandimmune® , to its SMEDDS formulation, Neoral® , improved
dosage forms. Bioavailability from a lipid-based formula- its bioavailability by ∼6.5-fold (Trull et al., 1995) (Table 1).
tion can be reduced by the use of lipolysis inhibiting Similarly, Julianto et al. (2000) observed that the self-
surfactants, e.g., polyoxyethylene-10-oleoyl ether (Brij 96® ), emulsifying formulation of Vitamin E (log P 9.96) had ∼3-fold
polyoxyle-35-castor oil (Cremophor EL® ), Cremophor RH40® , higher extent of absorption than its solution in soybean oil
and polysorbate 80 (Crillet 4® ) (US patents 5,645,856 and (Natopherol® soft gelatin capsules). The SEDDS formulation
6,096,338) in cases where lipolysis is important to drug absorp- consisted of Tween 80® , sorbitan monooleate (Span 80® ), and
tion. Rate of lipolysis of various lipids and formulations can Vitamin E dissolved in palm oil in the proportion 4:2:4. These
be compared in vitro. The effect of lipids on lymphatic drug results indicated that, in addition to bile mediated emulsification
transport, however, can overwhelm the difference in their rate and absorption mechanism, formulation-induced in vivo emulsi-
of lipolysis. fication was useful in enhancing drug absorption. Similar results
Dahan and Hoffman evaluated the impact of using short (C2 , were shown by Yap and Yuen for tocotrienols, which belong to
triacetin), medium (C8–10 , glyceryl tricaprylate/caprate (Cap- the Vitamin E family (Yap and Yuen, 2004). Thus, given other
tex 355® )), and long-chain (C18 , peanut oil) triglycerides (SCT, things being equal, SMEDDS formulation is expected to have
MCT, and LCT, respectively) on hydrophobic drug absorp- higher bioavailability than the SEDDS formulation because of
tion as a function of lymphatic transport of the drug molecule lower dispersed phase size.
and lipolysis of the formulation (Dahan and Hoffman, 2006).
They selected progesterone (log P 4.0) and vitamin D3 (log P 5. Conclusions
9.1) as hydrophobic drugs, of which only the latter has signif-
icant lymphatic transport. Bioavailability of progesterone from Lipid-based systems are a promising choice for the deliv-
the formulations followed the trend MCT > LCT > SCT which ery of hydrophobic molecules. These systems could be lipid
strongly correlated with in vitro lipolysis data of these formu- solution, emulsions, microemulsions, SEDDS, SMEDDS, or
lations, while that of vitamin D3 was LCT > MCT > SCT and micellar systems. These systems avoid the dissolution step upon
did not correlate with the lipolysis data (MCT > LCT > SCT). oral administration and differ from one another with respect to
These results were explained as a stimulation of lipid turnover the size of the dispersed phase and the content of surfactant
in enterocytes by LCT, which led to increased lymphatic trans- and other ingredients. They help improve the bioavailability of
port pathway capacity (Dahan and Hoffman, 2006). Increased hydrophobic drugs through several mechanisms, e.g., facilita-
lymphatic transport can also reduce hepatic metabolism of drugs tion of in vivo dispersion through the added surfactant, lipolysis
that have significant first pass effect. Thus, to maximize bioavail- of constituent lipids, increased lymphatic transport, etc. Micel-
ability of a hydrophobic drug from the lipidic formulation, the lar and microemulsion systems, being the most dispersed of all,
choice of excipients should also take into consideration biophar- appear the most promising.
maceutical properties of the drug. The use of lipid-based delivery systems has become increas-
ingly popular for pre-clinical studies since most of the new
4.2. Inhibition of drug efflux molecular entities are highly hydrophobic. Several studies have
reviewed the formation of these systems, the role of composi-
Absorbed drug molecules entering the enterocyte are tion on phase diagram, and drug release and bioavailability from
exposed to metabolizing enzymes, e.g., cytochrome P-450 3A4 these systems. While improved drug entrapment and release is
(CYP3A4), or can be secreted back into the gastrointestinal observed in almost all cases, improvement in bioavailability is
lumen by P-glycoprotein (P-gp) efflux pumps on the entero- often unpredictable. Several studies have focused on formulation
cyte membrane. The impact of formulation ingredients on the and drug-related biopharmaceutical aspects that are important in
biopharmaceutical properties of drugs is also illustrated by the governing oral bioavailability. These factors include precipita-
inhibition of drug efflux pumps by certain formulation ingre- tion of drug in vivo, digestability of lipids in the formulation,
dients. For example, common pharmaceutical excipients like overall HLB of surfactant mix in the system, intestinal efflux
polyethylene glycol, Tween 80® , and Cremophor EL® , have pumps and metabolizing enzymes, contribution of lymphatic
been shown to inhibit P-gp activity (Hugger et al., 2002). Their transport of drug to its absorption, etc. The design of SEDDS,
inclusion in the formulation, therefore, can be expected to SMEDDS, and micellar systems presents a plethora of choices
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 23

that appear equivalent on surface and are usually selected empir- Constantinides, P.P., 1995. Lipid microemulsions for improving drug dissolution
ically. Incorporation of these formulation and biopharmaceutical and oral absorption: physical and biopharmaceutical aspects. Pharm. Res.
considerations into the design of these systems will help improve 12, 1561–1572.
Cooney, G.F., Jeevanandam, V., Choudhury, S., Feutren, G., Mueller, E.A.,
their in vivo performance. Eisen, H.J., 1998. Comparative bioavailability of Neoral and Sandimmune
Among factors that influence the bioavailability of drugs from in cardiac transplant recipients over 1 year. Transplant. Proc. 30, 1892–1894.
these systems, lack of drug precipitation upon aqueous dilution Dahan, A., Hoffman, A., 2006. Use of a dynamic in vitro lipolysis model to
plays the predominant role in many cases. While several factors rationalize oral formulation development for poor water soluble drugs: cor-
need to be incorporated into the design of SEDDS, SMEDDS, relation with in vivo data and the relationship to intra-enterocyte processes
in rats. Pharm. Res. 23, 2165–2174.
and micellar drug delivery systems, as discussed in Section 5 Davey, R., Hilton, A., Garside, J., de la Fuente, M., Edmondson, M., Rains-
above, due attention needs to be given to the propensity of these ford, P., 1996. Crystallisation of oil-in-water emulsions. Amphiphile directed
systems for precipitation in vivo upon oral administration. While nucleation in aqueous emulsions of m-chloronitrobenzene. J. Chem. Soc.
this aspect has been recognized by several studies and empirical Faraday Trans. 92, 1927–1933.
rationale for minimizing the tendency of drug for precipitation de Smidt, P.C., Campanero, M.A., Troconiz, I.F., 2004. Intestinal absorption
of penclomedine from lipid vehicles in the conscious rat: contribution of
from the system have been developed, there remains a need to emulsification versus digestibility. Int. J. Pharm. 270, 109–118.
have predictive ability and objective parameters for assessing Flanagan, J., Singh, H., 2006. Microemulsions: a potential delivery system for
this risk. bioactives in food. Crit. Rev. Food Sci. Nutr. 46, 221–237.
Some key features of these systems can be useful in address- Furedi-Milhofer, H., Garti, N., Kamyshny, A., 1999. Crystallization from
ing these needs. For example, solubilization capacity of the microemulsions—a novel method for the preparation of new crystal forms
of aspartame. J. Cryst. Growth 198/199, 1365–1370.
system can be increased much above the required drug concen- Furedi-Milhofer, H., Kamishny, A., Yano, J., Aserin, A., Garti, N., 2003. Crys-
tration, so that it remains below the saturation and nucleation tallization of organic compounds in reverse micelles. III. Solubilization of
concentration of the drug in the system and upon dilution. aspartame. Langmuir 19, 5984–5990.
The aspects that affect solubilization capacity and saturation Gao, P., Guyton, M.E., Huang, T., Bauer, J.M., Stefanski, K.J., Lu, Q., 2004.
concentration as both undiluted reverse micelles and diluted Enhanced oral bioavailability of a poorly water soluble drug PNU-91325 by
supersaturatable formulations. Drug Dev. Ind. Pharm. 30, 221–229.
microemulsions, as well as dilutability as a single phase system, Gao, P., Rush, B.D., Pfund, W.P., Huang, T., Bauer, J.M., Morozowich, W.,
have been reviewed. Some in vitro models can be extrapolated Kuo, M.S., Hageman, M.J., 2003. Development of a supersaturable SEDDS
to predict the relative tendency of formulations for in vivo drug (S-SEDDS) formulation of paclitaxel with improved oral bioavailability. J.
precipitation. The use of some polymeric hydrophilic excipients Pharm. Sci. 92, 2386–2398.
in the formulation can help prevent or delay drug precipitation Garti, N., Avrahami, M., Aserin, A., 2006. Improved solubilization of cele-
coxib in U-type nonionic microemulsions and their structural transitions
by the formation of a supersaturated state upon aqueous dilution. with progressive aqueous dilution. J. Colloid Interface Sci. 299, 352–
These studies provide the background and basis on which 365.
models to predict, and approaches to prevent, in vivo drug pre- Garti, N., Yaghmur, A., Leser, M.E., Clement, V., Watzke, H.J., 2001. Improved
cipitation may be developed. These efforts will help improve the oil solubilization in oil/water food grade microemulsions in the presence of
outcome of formulation efforts towards improving the bioavail- polyols and ethanol. J. Agric. Food Chem. 49, 2552–2562.
Gursoy, R.N., Benita, S., 2004. Self-emulsifying drug delivery systems (SEDDS)
ability of hydrophobic drugs. for improved oral delivery of lipophilic drugs. Biomed. Pharmacother. 58,
173–182.
Hasegawa, A., Taguchi, M., Suzuki, R., Miyata, T., Nakagawa, H., Sugimoto,
Acknowledgements
I., 1988. Supersaturation mechanism of drugs from solid dispersions with
enteric coating agents. Chem. Pharm. Bull. (Tokyo) 36, 4941–4950.
We thank Dr. Patrick McGrath for critical reading of this Hauss, D.J., Fogal, S.E., Ficorilli, J.V., Price, C.A., Roy, T., Jayaraj, A.A., Keirns,
manuscript and very helpful comments. J.J., 1998. Lipid-based delivery systems for improving the bioavailability and
lymphatic transport of a poorly water-soluble LTB4 inhibitor. J. Pharm. Sci.
87, 164–169.
References Hou, M.J., Shah, D.O., 1987. Effects of molecular structure of the interface and
continuous phase on solubilization of water in water/oil microemulsions.
Aramaki, K., Hayashi, T., Katsuragi, T., Ishitobi, M., Kunieda, H., 2001. Effect Langmuir 3, 1086–1096.
of adding an amphiphilic solubilization improver, sucrose distearate, on the Hugger, E.D., Novak, B.L., Burton, P.S., Audus, K.L., Borchardt, R.T., 2002. A
solubilization capacity of nonionic microemulsions. J. Colloid Interface Sci. comparison of commonly used polyethoxylated pharmaceutical excipients
236, 14–19. on their ability to inhibit P-glycoprotein activity in vitro. J. Pharm. Sci. 91,
Beattie, J., 1989. Monodisperse colloids of transition metal and lanthanide com- 1991–2002.
pounds. Pure Appl. Chem. 61, 937–941. Huie, C.W., 2006. Recent applications of microemulsion electrokinetic chro-
Brime, B., Moreno, M.A., Frutos, G., Ballesteros, M.P., Frutos, P., 2002. matography. Electrophoresis 27, 60–75.
Amphotericin B in oil-water lecithin-based microemulsions: formulation Humberstone, A.J., Charman, W.N., 1997. Lipid-based vehicles for the oral
and toxicity evaluation. J. Pharm. Sci. 91, 1178–1185. delivery of poorly water soluble drugs. Adv. Drug Deliv. Rev. 25, 103–
Celebi, N., Turkyilmaz, A., Gonul, B., Ozogul, C., 2002. Effects of epidermal 128.
growth factor microemulsion formulation on the healing of stress-induced Ishihama, Y., Oda, Y., Uchikawa, K., Asakawa, N., 1994. Correlation of octanol-
gastric ulcers in rats. J. Control Release 83, 197–210. water partition coefficients with capacity factors measured by micellar
Cho, Y.W., Flynn, M., 1989. Oral delivery of insulin. Lancet 2, 1518–1519. electrokinetic chromatography. Chem. Pharm. Bull. (Tokyo) 42, 1525–
Cilek, A., Celebi, N., Tirnaksiz, F., Tay, A., 2005. A lecithin-based microemul- 1527.
sion of rh-insulin with aprotinin for oral administration: investigation of Johannessen, E., Walderhaug, H., Balinov, B., 2004. Aqueous microemulsions
hypoglycemic effects in non-diabetic and STZ-induced diabetic rats. Int. J. of a fluorinated surfactant and oil studied by PFG-NMR: transformation
Pharm. 298, 176–185. from threadlike to spherical micelles. Langmuir 20, 336–341.
24 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25

Julianto, T., Yuen, K.H., Noor, A.M., 2000. Improved bioavailability of vita- microemulsions using different experimental methods. Int. J. Pharm. 276,
min E with a self emulsifying formulation. Int. J. Pharm. 200, 53– 115–128.
57. Porter, C.J., Charman, W.N., 2001. In vitro assessment of oral lipid based for-
Khoo, S.M., Humberstone, A.J., Porter, C.J., Edwards, G.A., Charman, mulations. Adv. Drug Deliv. Rev. 50 (Suppl. 1), S127–S147.
W.N., 1998. Formulation design and bioavailability assessment of lipidic Porter, C.J., Kaukonen, A.M., Boyd, B.J., Edwards, G.A., Charman, W.N., 2004.
self-emulsifying formulations of halofantrine. Int. J. Pharm. 167, 155– Susceptibility to lipase-mediated digestion reduces the oral bioavailability of
164. danazol after administration as a medium-chain lipid-based microemulsion
Kraeling, M.E., Ritschel, W.A., 1992. Development of a colonic release cap- formulation. Pharm. Res. 21, 1405–1412.
sule dosage form and the absorption of insulin. Methods Find. Exp. Clin. Pouton, C.W., 1997. Formulation of self-emulsifying drug delivery systems.
Pharmacol. 14, 199–209. Adv. Drug Deliv. Rev. 25, 47–58.
Lamer, V., Dinegar, R., 1950. Theory, production and mechanism of formation Pouton, C.W., 2000. Lipid formulations for oral administration of drugs: non-
of monodispersed hydrosols. J. Am. Chem. Soc. 72, 4847–4854. emulsifying, self-emulsifying and ‘self-microemulsifying’ drug delivery
Lawrence, M.J., Rees, G.D., 2000. Microemulsion-based media as novel drug systems. Eur. J. Pharm. Sci. 11 (Suppl. 2), S93–S98.
delivery systems. Adv. Drug Deliv. Rev. 45, 89–101. Prince, L.M., 1975. Microemulsions versus micelles. J. Colloid Interface Sci.
Leodidis, E.B., Hatton, T.A., 1990a. Amino acids in AOT reversed micelles. 1. 52, 182–188.
Determination of interfacial partition coefficients using the phase-transfer Raghavan, S.L., Kiepfer, B., Davis, A.F., Kazarian, S.G., Hadgraft, J., 2001a.
method. J. Phys. Chem. 94, 6400–6411. Membrane transport of hydrocortisone acetate from supersaturated solu-
Leodidis, E.B., Hatton, T.A., 1990b. Amino acids in AOT reversed micelles. 2. tions; the role of polymers. Int. J. Pharm. 221, 95–105.
The hydrophobic effect and hydrogen bonding as driving forces for interfa- Raghavan, S.L., Trividic, A., Davis, A.F., Hadgraft, J., 2000. Effect of cellulose
cial solubilization. J. Phys. Chem. 94, 6411–6420. polymers on supersaturation and in vitro membrane transport of hydrocor-
Leodidis, E.B., Hatton, T.A., 1991a. Amino acids in reversed micelles. 3. Depen- tisone acetate. Int. J. Pharm. 193, 231–237.
dence of the interfacial partition coefficient on excess phase salinity and Raghavan, S.L., Trividic, A., Davis, A.F., Hadgraft, J., 2001b. Crystallization
interfacial curvature. J. Phys. Chem. 95, 5943–5956. of hydrocortisone acetate: influence of polymers. Int. J. Pharm. 212, 213–
Leodidis, E.B., Hatton, T.A., 1991b. Amino acids in reversed micelles. 4. Amino 221.
acids as cosurfactants. J. Phys. Chem. 95, 5957–5965. Scherlund, M., Malmsten, M., Holmqvist, P., Brodin, A., 2000. Thermosetting
Li, P., Tabibi, S.E., Yalkowsky, S.H., 1999a. Solubilization of flavopiridol by pH microemulsions and mixed micellar solutions as drug delivery systems for
control combined with cosolvents, surfactants, or complexants. J. Pharm. periodontal anesthesia. Int. J. Pharm. 194, 103–116.
Sci. 88, 945–947. Schuleit, M., Luisi, P.L., 2001. Enzyme immobilization in silica-hardened
Li, P., Tabibi, S.E., Yalkowsky, S.H., 1999b. Solubilization of ionized and organogels. Biotechnol. Bioeng. 72, 249–253.
un-ionized flavopiridol by ethanol and polysorbate 20. J. Pharm. Sci. 88, Sek, L., Boyd, B.J., Charman, W.N., Porter, C.J., 2006. Examination of the
507–509. impact of a range of Pluronic surfactants on the in-vitro solubilisation
Li, P., Vishnuvajjala, R., Tabibi, S.E., Yalkowsky, S.H., 1998. Evaluation of in behaviour and oral bioavailability of lipidic formulations of atovaquone.
vitro precipitation methods. J. Pharm. Sci. 87, 196–199. J. Pharm. Pharmacol. 58, 809–820.
Lindenberg, M., Kopp, S., Dressman, J.B., 2004. Classification of orally admin- Sek, L., Porter, C.J., Kaukonen, A.M., Charman, W.N., 2002. Evaluation of the
istered drugs on the World Health Organization Model list of essential in-vitro digestion profiles of long and medium chain glycerides and the phase
medicines according to the biopharmaceutics classification system. Eur. J. behaviour of their lipolytic products. J. Pharm. Pharmacol. 54, 29–41.
Pharm. Biopharm. 58, 265–278. Shen, H., Zhong, M., 2006. Preparation and evaluation of self-microemulsifying
Lyons, K.C., Charman, W.N., Miller, R., Porter, C.J., 2000. Factors limiting drug delivery systems (SMEDDS) containing atorvastatin. J. Pharm. Phar-
the oral bioavailability of N-acetylglucosaminyl-N-acetylmuramyl dipeptide macol. 58, 1183–1191.
(GMDP) and enhancement of absorption in rats by delivery in a water-in-oil Shukla, A., Janich, M., Jahn, K., Krause, A., Kiselev, M.A., Neubert, R.H., 2002.
microemulsion. Int. J. Pharm. 199, 17–28. Investigation of pharmaceutical oil/water microemulsions by small-angle
Madamwar, D., Thakar, A., 2004. Entrapment of enzyme in water-restricted scattering. Pharm. Res. 19, 881–886.
microenvironment for enzyme-mediated catalysis under microemulsion- Shukla, A., Janich, M., Jahn, K., Neubert, R.H., 2003. Microemulsions for
based organogels. Appl. Biochem. Biotechnol. 118, 361–369. dermal drug delivery studied by dynamic light scattering: effect of inter-
Malcolmson, C., Barlow, D.J., Lawrence, M.J., 2002. Light-scattering studies of particle interactions in oil-in-water microemulsions. J. Pharm. Sci. 92, 730–
testosterone enanthate containing soybean oil/C18:1E10/water oil-in-water 738.
microemulsions. J. Pharm. Sci. 91, 2317–2331. Simonelli, A.P., Mehta, S.C., Higuchi, W.I., 1970. Inhibition of sulfathiazole
Malcolmson, C., Satra, C., Kantaria, S., Sidhu, A., Lawrence, M.J., 1998. Effect crystal growth by polyvinylpyrrolidone. J. Pharm. Sci. 59, 633–638.
of oil on the level of solubilization of testosterone propionate into nonionic Spernath, A., Aserin, A., Garti, N., 2006. Fully dilutable microemulsions embed-
oil-in-water microemulsions. J. Pharm. Sci. 87, 109–116. ded with phospholipids and stabilized by short-chain organic acids and
Moreno, M.A., Ballesteros, M.P., Frutos, P., 2003. Lecithin-based oil-in-water polyols. J. Colloid Interface Sci. 299, 900–909.
microemulsions for parenteral use: pseudoternary phase diagrams, charac- Spernath, A., Yaghmur, A., Aserin, A., Hoffman, R.E., Garti, N., 2002. Food-
terization and toxicity studies. J. Pharm. Sci. 92, 1428–1437. grade microemulsions based on nonionic emulsifiers: media to enhance
Mrestani, Y., Neubert, R.H., Krause, A., 1998. Partition behaviour of drugs in lycopene solubilization. J. Agric. Food Chem. 50, 6917–6922.
microemulsions measured by electrokinetic chromatography. Pharm. Res. Spernath, A., Yaghmur, A., Aserin, A., Hoffman, R.E., Garti, N., 2003.
15, 799–801. Self-diffusion nuclear magnetic resonance, microstructure transitions, and
Neubert, R.H., Schmalfuss, U., Wolf, R., Wohlrab, W.A., 2005. Microemulsions solubilization capacity of phytosterols and cholesterol in Winsor IV food-
as colloidal vehicle systems for dermal drug delivery. Part V. Microemul- grade microemulsions. J. Agric. Food Chem. 51, 2359–2364.
sions without and with glycolipid as penetration enhancer. J. Pharm. Sci. 94, Spiclin, P., Homar, M., Zupancic-Valant, A., Gasperlin, M., 2003. Sodium ascor-
821–827. byl phosphate in topical microemulsions. Int. J. Pharm. 256, 65–73.
Ni, N., El-Sayed, M.M., Sanghvi, T., Yalkowsky, S.H., 2000. Estimation of the Strickley, R.G., 2004. Solubilizing excipients in oral and injectable formulations.
effect of NaCl on the solubility of organic compounds in aqueous solutions. Pharm. Res. 21, 201–230.
J. Pharm. Sci. 89, 1620–1625. Testard, F., Zemb, T., 1999. Excess of solubilization and curvature in nonionic
Peltola, S., Saarinen-Savolainen, P., Kiesvaara, J., Suhonen, T.M., Urtti, A., microemulsions. J. Colloid Interface Sci. 219, 11–19.
2003. Microemulsions for topical delivery of estradiol. Int. J. Pharm. 254, Trull, A.K., Tan, K.K., Tan, L., Alexander, G.J., Jamieson, N.V., 1995.
99–107. Absorption of cyclosporin from conventional and new microemulsion oral
Podlogar, F., Gasperlin, M., Tomsic, M., Jamnik, A., Rogac, M.B., 2004. Struc- formulations in liver transplant recipients with external biliary diversion. Br.
tural characterisation of water-Tween 40/Imwitor 308-isopropyl myristate J. Clin. Pharmacol. 39, 627–631.
 A.S. Narang et al. / International Journal of Pharmaceutics 345 (2007) 9–25 25

Vaze, A., Parizo, M., Rusling, J.F., 2004. Enhanced rates of electrolytic styrene Yap, S.P., Yuen, K.H., 2004. Influence of lipolysis and droplet size on tocotrienol
epoxidation catalyzed by cross-linked myoglobin-poly(l-lysine) films in absorption from self-emulsifying formulations. Int. J. Pharm. 281, 67–
bicontinuous microemulsions. Langmuir 20, 10943–10948. 78.
Yano, J., Furedi-Milhofer, H., Wachtel, E., Garti, N., 2000. Crystallization of Zheng, J.Y., Fulu, M.Y., 2006. Decrease of genital organ weights and plasma
organic compounds in reversed micelles. I. Solubilization of amino acids in testosterone levels in rats following oral administration of leuprolide
water-isooctane-AOT microemulsions. Langmuir 16, 9996–10004. microemulsion. Int. J. Pharm. 307, 209–215.
 Advanced Drug Delivery Reviews 64 (2012) 175–193

Contents lists available at SciVerse ScienceDirect

Advanced Drug Delivery Reviews

journal homepage: www.elsevier.com/locate/addr

Microemulsion-based media as novel drug delivery systems☆

M. Jayne Lawrence a,*, Gareth D. Rees b,*
a
Department of Pharmacy, King’s College London, Franklin Wilkins Building, 150 Stamford Street, London SE1 9NN, UK
b
SmithKline Beecham R&D, St. George’s Avenue, Weybridge, Surrey KT13 0DE, UK

a r t i c l e i n f o a b s t r a c t

Available online 13 September 2012 Microemulsions are clear, stable, isotropic mixtures of oil, water and surfactant, frequently in combination
with a cosurfactant. These systems are currently of interest to the pharmaceutical scientist because of their
considerable potential to act as drug delivery vehicles by incorporating a wide range of drug molecules. In
order to appreciate the potential of microemulsions as delivery vehicles, this review gives an overview of
the formation and phase behaviour and characterization of microemulsions. The use of microemulsions
and closely related microemulsion-based systems as drug delivery vehicles is reviewed, with particular em-
phasis being placed on recent developments and future directions.
© 2012 Published by Elsevier B.V.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
2. Overview of microemulsion formation and phase behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
2.1. Theories of microemulsion formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
2.2. Phase behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
2.3. The role of surfactant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
3. Microemulsion characterisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
4. Microemulsion-based systems in drug delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
4.1. Pharmaceutically acceptable excipients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
4.2. Low viscosity microemulsion systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
4.3. High viscosity systems and microemulsion gels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
4.4. SMEDDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
5. Recent developments and future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
5.1. Component selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
5.2. Phase behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
5.3. Block copolymer micelles and microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.4. Comparative drug delivery studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.5. Oral delivery — SEDDs and SMEDDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.6. Parenteral, pulmonary and ocular delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
5.7. Topical delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
5.8. New developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
5.9. Alternative surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

☆ PII of original article: S0169-409X(00)00103-4. The article was originally published 1. Introduction
in Advanced Drug Delivery Reviews 45 (2000) 89–121.
* Corresponding authors. Tel.: +44-0207-848-4808; fax: +44-0207-848-4800
(M.J. Lawrence). Tel.: +44-1932-822179; fax: +44-1932-822120 (G.D. Rees).
The microemulsion concept was introduced as early as the 1940s
E-mail addresses: jayne.lawrence@kcl.ac.uk (M.J. Lawrence), gareth.d.rees@sb.com by Hoar and Schulman who generated a clear single-phase solution
(G.D. Rees). by titrating a milky emulsion with hexanol [1]. Schulman and

0169-409X/$ – see front matter © 2012 Published by Elsevier B.V.

http://dx.doi.org/10.1016/j.addr.2012.09.018
176 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

coworkers (1959) subsequently coined the term microemulsion [2], apolar liquids such as alkanes. In this case the orientation of the sur-
and it has since been defined and indeed redefined on many occa- factant molecules are reversed compared to those adopted in aqueous
sions. For the purposes of this review, however, the microemulsion solution. This reorientation serves to optimise the solvation require-
definition provided by Danielsson and Lindman in 1981 will be used ments of the surfactant and minimises the free energy of the system
as the point of reference [3]. Microemulsions are thus defined as ‘a overall. When surfactants are incorporated into immiscible mixtures
system of water, oil and amphiphile which is a single optically isotro- of oil and water, the surfactant molecules can locate at the oil/water
pic and thermodynamically stable liquid solution.’ interface which is thermodynamically very favourable. A number of
In practice, the key difference between emulsions and micro- phases can result which may be structured on the microscopic or
emulsions are that the former, whilst they may exhibit excellent macroscopic scale, one example of a phase structured on the micro-
kinetic stability, are fundamentally thermodynamically unstable and scopic scale is an optically isotropic microemulsion phase. The sche-
will eventually phase separate [4]. Another important difference con- matic given in Fig. 1 gives an indication of a few of the wide variety
cerns their appearance; emulsions are cloudy while microemulsions of possible self-association structures that surfactants can form in
are clear or translucent. In addition, there are distinct differences in the presence of water, oil or combinations of all three. Although out-
their method of preparation, since emulsions require a large input side the scope of this review many of the structures shown in Fig. 1, as
of energy while microemulsions do not. The latter point has obvious well as some of those not shown, have potential for use as drug deliv-
implications when considering the relative cost of commercial pro- ery systems.
duction of the two types of system. Fig. 2 shows schematic representations of the three types of
It is also useful to note that under the definition given, self- microemulsions which are most likely to be formed depending on com-
microemulsifying drug delivery systems (SMEEDS) are not micro- position. It can be seen while the three structures shown are quite dif-
emulsions, although they may be considered to be a closely related ferent, in each there is an interfacial surfactant monolayer separating
system. A SMEDD typically comprises a mixture of surfactant, oil the oil and water domains. Note that while the oil-in-water (o/w) and
and drug (known as the concentrate) which when introduced into water-in-oil (w/o) droplets are represented in Fig. 2 as spheres, they
the body is rapidly dispersed to form droplets of approximately the may be asymmetric in shape, frequently adopting the shape of a prolate
same size range as those observed in microemulsion systems. Once ellipsoid. The presence of o/w microemulsion droplets is likely to be a
dispersed such systems would be expected to behave in vivo much feature in microemulsions where the volume fraction of oil is low. Con-
the same way as oil-in-water (o/w) microemulsions. versely, w/o droplets are likely when the volume fraction of water is
The above broad definition does not require a microemulsion to low, and in systems where the amounts of water and oil are similar, a
contain any microstructure, in other words it includes systems that bicontinuous microemulsion may result. In the latter case, both oil
are co-solvents, that is, systems in which the constituent components and water exist as a continuous phase in the presence of a continuously
are molecularly dispersed. Most researchers in the field agree how- fluctuating surfactant-stabilised interface with a net curvature of zero.
ever that for a microemulsion to be formed it is important that the The relationship between micelles and o/w microemulsion drop-
system contains some definite microstructure, in other words there lets as well as between reverse micelles and w/o microemulsion
is a definite boundary between the oil and water phases at which droplets has been debated on a number of occasions. Clearly there
the surfactant is located. In order to gain an understanding of the is a transition through the series (reverse) micelle, swollen micelle
reasons for microemulsion formation, it is first useful to consider and microemulsion droplet but by definition micelles and reverse
the properties of amphiphiles, such as surfactants, in solution. micelles are not microemulsions. Distinguishing between swollen
Conventional surfactant molecules comprise a polar head group micelles and microemulsion droplets is largely a semantic exercise,
region and an apolar tail region, the latter having the larger molecular but it is recognised that as the ratio of dispersed phase to surfactant
volume particularly in the case of ionic surfactants. On dispersal in increases, the physicochemical properties approach those of the pure
water, surfactants self-associate into a variety of equilibrium phases, solvent. It should be noted that while the definition used in the pres-
the nature of which stems directly from the interplay of the various ent study does not differentiate between a swollen micelle and a
inter and intermolecular forces as well as entropy considerations. microemulsion, other researchers in the field do however make this
Surfactants also self-associate in non-aqueous solvents, particularly distinction.

Fig. 1. Schematic representation of the most commonly encountered self-association structures in water, oil or a combination thereof.
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 177

Fig. 2. Schematic representation of the three most commonly encountered microemulsion microstructures: (a) oil-in-water, (b) bicontinuous, and (c) water-in-oil microemulsion.

Depending upon the nature of the oil, in particular its size relative entropy arising from the mixing of one phase in the other in the form
to the hydrophobic chain of the surfactant, the oil may penetrate to of large numbers of small droplets. However, there are also expected
varying extents into the surfactant tails of the interfacial monolayer. to be favourable entropic contributions arising from other dynamic
This is shown schematically in Fig. 3 for a bicontinuous microemulsion; processes such as surfactant diffusion in the interfacial layer and
a similar effect has been proposed to occur in both o/w and w/o monomer-micelle surfactant exchange. Thus a negative free energy of
microemulsions. formation is achieved when large reductions in surface tension are
accompanied by significant favourable entropic change. In such cases,
2. Overview of microemulsion formation and phase behaviour microemulsification is spontaneous and the resulting dispersion is
thermodynamically stable.
2.1. Theories of microemulsion formation Qualitatively we know that several factors determine whether a
w/o or o/w system is formed. Intuitively, we would summise that
Historically, three approaches have been used to explain micro- the most likely microemulsion would be that in which the phase
emulsion formation and stability. These are: (i) interfacial or mixed with the smaller volume fraction forms the droplets, and indeed
film theories [2,5]: (ii) solubilisation theories [6–8]: and (iii) thermo- this is very often although by no means exclusively the case. By
dynamic treatments [9–11]. An in depth discussion of these theories their very nature, o/w microemulsion droplets generally have a larger
are beyond the scope of this review but has been addressed by others effective interaction volume than w/o droplets. In the case of ionic
[12]. However, an admittedly simplified thermodynamic rationalisa- surfactants this is attributable to the presence of an electrical double
tion is presented below. The free energy of microemulsion formation layer at the surface of the o/w droplet which introduces a strong
can be considered to depend on the extent to which surfactant lowers repulsive term. For o/w microemulsions stabilised by a non-surfactant,
the surface tension of the oil–water interface and the change in entro- although there is hydration shell associated with the polar headgroups,
py of the system such that, the predominant repulsive factor can be attributed to steric interactions.
Additionally, it is pertinent to note that it is easier to arrange surfactant
ΔGf ¼ γ ΔA − T ΔS at an interface with high curvature, i.e., small droplets, if the surfactant
tails extend outwards into a continuous oil phase. This is also entropi-
where ΔGf is the free energy of formation, γ is the surface tension
cally more favourable as the hydrocarbon tails have more directional
of the oil–water interface, ΔA is the change in interfacial area on
freedom. As a result, interfacial tension tends to be lower for a w/o
microemulsification, ΔS is the change in entropy of the system which
microemulsion than for an o/w microemulsion, thereby making their
is effectively the dispersion entropy, and T is the temperature. It should
preparation a more facile process. It should also be remembered howev-
be noted that when a microemulsion is formed the change in ΔA is very
er, that while microemulsions are thermodynamically stable there may
large due to the large number of very small droplets formed. Originally
be kinetic barriers to their formation. As a consequence, the order of
workers proposed that in order for a microemulsion to be formed a
component addition may impact on the ease of preparation, and in
(transient) negative value of γ was required, it is now recognised that
some cases mechanical agitation or the input of heat will assist more
while value of γ is positive at all times, it is very small (of the order of
rapid microemulsification.
fractions of mN/m), and is offset by the entropic component. The dom-
inant favourable entropic contribution is the very large dispersion
2.2. Phase behaviour

The relationship between the phase behaviour of a mixture and its

composition can be captured with the aid of a phase diagram. Compo-
sitional variables can also be studied as a function of temperature and
pressure, although with the exception of microemulsions prepared
using supercritical or near critical solvents [13,14], or with liquefied
chlorofluorocarbon [15] and HFA propellants [16], the large majority
of systems are studied under conditions of ambient pressure. The
phase behaviour of simple microemulsion systems comprising oil,
water and surfactant can be studied with the aid of ternary phase
diagram in which each corner of the diagram represents 100% of
Fig. 3. Penetration of oil molecules between the hydrophobic chains of the interfacial
surfactant monolayer in a bicontinuous microemulsion. Note the greater extent of oil
that particular component. More commonly, however, and almost
penetration when the film curves towards water (i.e. in a region of reverse curvature). always in the case of microemulsions in pharmaceutical applications,
Modified from [235]. the microemulsion will contain additional components such as a
178 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

cosurfactant and/or drug. The cosurfactant is also amphiphilic with an

affinity for both the oil and aqueous phases and partitions to an
appreciable extent into the surfactant interfacial monolayer present
at the oil–water interface. The cosurfactant need not necessarily be
capable of forming association structures in its own right. A wide
variety of molecules can function as cosurfactants including non-
ionic surfactants [17,18], alcohols [19,20], alkanoic acids, alkanediols
and alkyl amines [21]. Surprisingly few studies have examined the
effect of drug on phase behaviour, this is despite the fact that a
large number of drug molecules are themselves surface active [22]
and as such would be expected to influence phase behaviour. Fig. 4
shows the effect of the presence of drug (either fenoprofen or
fenoprofen sodium) on the phase behaviour of reverse phospholipid
aggregates.
In the case where four or more components are investigated,
Fig. 5. A hypothetical pseudo-ternary phase diagram of an oil/surfactant/water system with
pseudo-ternary phase diagrams are used where a corner will typically emphasis on microemulsion and emulsion phases. Within the phase diagram, existence
represent a binary mixture of two components such as surfactant/ fields are shown where conventional micelles, reverse micelles or water-in-oil (w/o)
cosurfactant, water/drug or oil/drug. The number of different phases microemulsions and oil-in-water microemulsions are formed along with the bicontinuous
present for a particular mixture can be visually assessed. Microstruc- microemulsions. At very high surfactant concentrations two phase systems are observed.

tural features can also be investigated with the aid of a wide variety of
techniques, which are discussed later in this review. A highly sche-
matic (pseudo) ternary phase diagram illustrating these features is systems. Clearly, however, time constraints impose a physical limit
presented in Fig. 5. It should be noted that not every combination of on the length of time systems can be left to equilibrate and conse-
components produce microemulsions over the whole range of possi- quently the elimination of metastable states can be difficult to ensure
ble compositions, in some instances the extent of microemulsion for- in practice, although centrifugation can be useful to speed up any sep-
mation may be very limited. aration. References to rapid screening procedures have appeared in
Constructing phase diagrams is time consuming, particularly when the literature [23].
the aim is to accurately delineate a phase boundary, as the time taken Outside the microemulsion region, particularly for compositions
for the system to equilibrate can be greatly increased as the phase close to the oil–water binary axis, there is insufficient surfactant to
boundary is approached. Heat and sonication are often used, particu- facilitate the formation of a single microemulsion phase. In this case
larly with systems containing nonionic surfactants, to speed up the multiple phases may exist, the complexity of which increases with
process. The procedure most often employed is to prepare a series of the number of components in the mixture. Within this region, and
(pseudo) binary compositions and titrate with the third component, indeed other multiphase regions of the ternary phase diagram, micro-
evaluating the mixture after each addition. Care must be taken to emulsions can exist in equilibrium with excess water or oil phases.
ensure not only that the temperature is precisely and accurately These multiphase systems can be conveniently described using the
controlled, but also that observations are not made on metastable Winsor classification [24]. In the Winsor classification, the one

Fig. 4. Schematic representation of the effect of drug on the association of phospholipid in an isopropylmyristate (IPM) system. Top, from left to right: phospholipid molecules in
IPM in the absence of water and drug form ellipsoidal reverse micelles, increasing the amount of water present causes these aggregates to transform first into rod-like micelles and
then finally lamellar liquid crystals. Bottom, from left to right: drug solubilisation either in its free acid or sodium salt form causes a change in shape of the colloidal aggregates —
with fenoprofen acid rod like micelles transform into more spherical ones, with fenoprofen salt rod-like micelles transform into extremely long rods. Modified from [222].
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 179

phase microemulsions that are generally explored as drug delivery and non-ionic, can be very effective at increasing the extent of the
systems are known as Winsor IV systems. microemulsion region. Examples of non-ionics include polyoxyethylene
Transitions between the various phases mapped out in these surfactants such as Brij 35 (C12E23) or a sugar esters such as sorbitan
phase diagrams can be driven by the further addition of one of the monooleate (Span 80). Phospholipids are a notable example of Zwit-
components, addition of a new component such as drug or electro- terionic surfactants and exhibit excellent biocompatibility. Lecithin
lyte, or by changing the temperature. Transitions from w/o to o/w preparations from a variety of sources including soybean and egg
microemulsions may occur via a number of different structural states are available commercially and contain diacylphosphatidylcholine
including bicontinuous, lamellar and also multiphase systems. Micro- as its major constituent [19,21,25,26]. Quaternary ammonium alkyl
emulsions stabilised by non-ionic surfactants, especially those based salts form one of the best known classes of cationic surfactants,
on polyoxyethylene, are very susceptible to temperature because a with hexadecyltrimethylammonium bromide (CTAB) [27–29], and
decrease in surfactant solubility occurs with increasing temperature, the twin-tailed surfactant didodcecylammonium bromide (DDAB)
and as a result systems stabilised by non-ionic surfactants or mixtures amongst the most well known [30–32]. The most widely studied an-
thereof often have characteristic phase inversion temperatures (PITs), ionic surfactant is probably sodium bis-2-ethylhexylsulphosuccinate
with the PIT of the microemulsion varying with a range of experimen- (AOT) which is twin-tailed and is a particularly effective stabiliser of
tal factors including the amount and nature of the oil present and the w/o microemulsions [18,30,33–37].
nature of the surfactant(s) present. This latter case is well illustrated Attempts have been made to rationalise surfactant behaviour in
in Fig. 6 which shows the effect on PIT of increasing the amount of the terms of the hydrophile–lipophile balance (HLB) [38], as well as the
lipophilic sugar surfactant present in a mixed C12E8:sucrose dilaurate critical packing parameter (CPP) [39,40]. Both approaches are fairly
system. It should be noted that the presence of electrolyte or in some empirical but can be a useful guide to surfactant selection. The HLB
cases drug, especially if lipophilic in nature, can act to lower the PIT, takes into account the relative contribution of hydrophilic and hydro-
illustrating the importance of determining microemulsion phase phobic fragments of the surfactant molecule. It is generally accepted
behaviour in the presence of drug. Although it is considered that the that low HLB (3–6) surfactants are favoured for the formation of
polyoxyethylene surfactants are the most sensitive, other non-ionic w/o microemulsions whereas surfactants with high HLBs (8–18) are
surfactants such as the alkylamine-N-oxides and the sugar surfactants preferred for the formation of o/w microemulsion systems. Ionic sur-
are also sensitive to changes in temperature. In contrast microemulsions factants such as sodium dodecyl sulphate which have HLBs greater
stabilised by ionic surfactants have little or no sensitivity to temperature. than 20, often require the presence of a cosurfactant to reduce their
The presence of a PIT can cause problems for the exploitation of effective HLB to a value within the range required for microemulsion
microemulsions stabilised by non-ionic surfactants as drug delivery formation.
systems. This is a particular problem where formulations are intended In contrast, the CPP relates the ability of surfactants to form partic-
for intravenous administration as autoclaving, the preferred means ular aggregates to the geometry of the molecule itself. The CPP can be
of sterilisation, is likely to destabilise the microemulsion. However, calculated using the following equation:
sterilisation by filtration remains an option for low viscosity droplet-
containing microemulsions. In order to avoid any complications due CPP ¼ v=a:l
to the presence of a PIT, the intended temperature of use should be
30 K below the PIT. Although it is possible to envisage circumstances where v is the partial molar volume of the hydrophobic portion of
whereby the presence of a phase transition in use can be used to de- the surfactant, a is the optimal head group area and l is the length
rive a benefit, for example to release drug at a pre-determined site in of the surfactant tail. The latter parameter is often expressed as lc,
the body. that is the critical length of the hydrophobic chain, generally assumed
to be 70–80% of its fully extended length. The CPP is a measure of the
2.3. The role of surfactant preferred geometry adopted by the surfactant, and as a consequence
is predictive of the type of aggregate that is likely to form. The effect
The single-phase microemulsion systems of interest in this review of changing CPP is illustrated in Fig. 7 but put simply, cone-shaped
are classified as Winsor IV. The surfactants used to stabilise such surfactants will pack at curved interfaces whereas surfactants whose
systems may be: (i) non-ionic, (ii) zwitterionic, (iii) cationic, or geometry can be represented by truncated cones or rectangular
(iv) anionic surfactants. Combinations of these, particularly ionic blocks prefer to form worm-like micelles or lamellar structures [41].
Of course, changes in microemulsion composition will modify the
microenvironment of the surfactant, which will lead to changes in
the apparent CPP of the surfactant. For example in a microemulsion
system, penetration of small oil molecules between the hydrocarbon
tails would be expected to increase the effective surfactant hydro-
phobe volume, whereas large molecular volume oils would not be
expected to exert much effect on the CPP [42]. Similarly, increases
in ionic strength would be expected to result in a decrease in the
effective head group area of ionic surfactants as the double layer
shrinks and screening of the head groups allows closer approach.
The presence of hydrophilic molecules such as glycerol and sorbitol
in the aqueous phase will also influence optimal head group area by
altering the solubility of the head group in the aqueous phase.
Because of these effects, water-soluble hydrophilic materials have
Fig. 6. Effect of temperature and weight fraction of surfactant mixture on the phase been used as to aid microemulsion formation as can be seen in
behaviour of a water/C12E8/sucrose dilaurate/heptane system. The weight fraction Fig. 8. However, whilst such materials may be referred to as cosurfac-
of the lipophilic sucrose dilaurate in the surfactant mixture are 0.5 (closed circles), tants, this description is misleading, as they are not amphiphilic in
0.7 (open triangle) and 0.8 (open circle). The water/heptane ratio is 50/50 (w/w). I, II their own right. As with traditional cosurfactants their presence can
and III indicate one-, two-, and three phase regions, with the one phase region being the
microemulsion region. Note that as the proportion of the lipophilic sucrose dilaurate
lead to destruction of the microemulsion upon dilution. Finally, the
increases, the I phase microemulsion region is formed at higher temperatures. Redrawn effect of temperature on these parameters is especially pertinent for
from [131]. non-ionics such as polyoxyethylene alkyl ethers as the polyoxyethylene
180 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

Fig. 7. Effect of molecular moieties and solution conditions on the CPP of a surfactant and the resulting range of possible surfactant aggregates in water or aqueous solution.
Redrawn from [236].

(PEO) group is dehydrated with increasing temperature. This has the to form, a point made in a number of pertinent microemulsions
effect of altering substantially the CPP and in extremis is manifested reviews [43–48]. Medium chain length alcohols which are commonly
by phase separation or phase inversion. added as cosurfactants, have the effect of further reducing the interfa-
In most cases, single-chain surfactants alone are unable to reduce cial tension, whilst increasing the fluidity of the interface thereby
the oil/water interfacial tension sufficiently to enable a microemulsion increasing the entropy of the system [44,45,48]. Medium chain length
alcohols also increase the mobility of the hydrocarbon tail and also
allow greater penetration of the oil into this region. Furthermore,
any alcohol present may also influence the solubility properties of
the aqueous and oily phases due to its partitioning between these
phases. It has also been suggested that some oils, for example the
ethyl esters of fatty acids, also act as ‘cosurfactants’ by penetrating
the hydrophobic chain region of the surfactant monolayer [49].
All of the aforementioned mechanisms are considered to facilitate
microemulsion formation. In the case of microemulsions stabilised
by ionic surfactants, the addition of alkanols also serves to reduce
repulsive interactions between the charged head groups.
A number of double chain surfactants such as AOT and DDAB
are able to form microemulsions without the aid of cosurfactants
[18,30–37,50,51]. These surfactants are characterised by having
small head groups in comparison to their hydrocarbon tails. Phospha-
tidylcholine or lecithin is also a twin-tailed surfactant, but in this case
it is generally necessary to include a cosurfactant in order to disrupt
the lamellar structures which characterise its biological behaviour.
Thus medium chain alcohols have been successfully used as cosurfac-
tants for the formation of lecithin-based microemulsions [19,26].
Interestingly w/o microemulsions have been prepared using short
diacyl chain lecithins and small molecular volume oils where it is pos-
sible that the small molecular volume oils penetrate the hydrophobic
chain region thereby facilitating microemulsion formation [52].

3. Microemulsion characterisation

In contrast to their ease of preparation, it is a far from trivial mat-

Fig. 8. Partial phase diagrams of the system isopropyl myristate (O), polysorbate 40
and sorbitol (S) and water (W) showing areas of existence of o/w microemulsions at
ter to characterise the microstructure of a microemulsion, yet such
37 °C for polysorbate/sorbitol mass ratios of a, 1/1; b, 1/1.5; c, 1/2 d, 1/2.5; e, 1/3; knowledge is essential for their successful commercial exploitation.
and f, 1/3.5. Redrawn from [86]. For example, it has been shown that the rate of release of sodium
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 181

salicylate from a lecithin-based microemulsions, is dependent upon associated with microemulsions include their thermodynamic stability,
their microstructure [53]. optical clarity and ease of preparation. The existence of microdomains
Microemulsions have been evaluated using a wide range of differ- of different polarity within the same single-phase solution enables
ent techniques over the years, but a complementarity of methods is both water-soluble and oil-soluble materials to be solubilised, and at
generally required in order to fully characterise these systems. At the same time if this is so desired. Furthermore it is also possible to in-
the macroscopic level viscosity, conductivity and dielectric methods corporate amphiphilic drugs into the microemulsion, sometimes even
provide useful information [29,35,36,54–56]. Viscosity measurements leading to an increase in the extent of existence of the microemulsion
for example can indicate the presence of rod-like or worm-like region. It should be noted that solubilisate partitions between the
reverse micelles [55,56], and conductivity measurements provide a microemulsion droplet and continuous phase and that while there
means of determining whether a microemulsion is oil-continuous may be a preferred site of solubilisation within the microemulsion
or water-continuous, as well as providing a means of monitoring per- droplet, solubilisate may be located at one of a number of sites. For
colation or phase inversion phenomena [37,55,57]. Dielectric mea- example the likely preferred sites of incorporation of a lipophilic,
surements are a powerful means of probing both the structural and water-insoluble drug into an o/w microemulsion are the disperse oil
dynamic features of microemulsion systems [37,58,59]. phase and/or hydrophobic tail region of the surfactant molecule, while
The isotropic nature of microemulsions and their optical clarity a water-soluble material would be most likely to be incorporated into
makes their study by spectroscopic techniques straightforward, the disperse aqueous phase of a water-in-oil droplet. In some instances
particularly in comparison to conventional macroemulsions. Pulsed the viscosity of the microemulsion may be tailored for a given applica-
field gradient NMR for example has been used extensively to measure tion through formulation changes, or in some instances through the
self-diffusion coefficients of the various components and yields incorporation of specific gelling agents such as Carbopol [90] or gelatin
information on the mobility and microenvironment [25,28,32,38,56, [91] as shown in Fig. 9.
60–68]. Scattering methods have also been invaluable in elucidating Table 1 illustrates how a number of self-associating surfactant sys-
microemulsion structure and methods employed include dynamic tems including microemulsions could be used in drug delivery appli-
and static light scattering [16,28,52,55,62,69–74], small-angle neu- cations, and we will go on to discuss specific cases in the following
tron scattering (SANS) [35,60,62,75–79] and small-angle X-ray scat- sections of this review. All of the systems shown in Table 1 have the
tering (SAXS) [30,60–62,76,79,80]. Indeed the value of scattering potential to protect labile compounds, but it is still the case that
methods is exemplified by the work of Tabony who identified the there are few examples of microemulsion-based drug delivery sys-
presence of a cubic phase in the bicontinuous microemulsion region tems used in commercial drug formulations.
[77,78]. These techniques have also been extremely useful in the The attraction of o/w microemulsion systems lies in their ability to
development of microemulsion models such as the cubic random incorporate hydrophobic drugs into the apolar oil phase thereby en-
cell (CRC) [76,81] and disordered open connected (DOC) models hancing their solubility [43–48,88,89]. However it is worth noting
[30,79]. Neutron scattering methods using contrast variation have that most drugs are not especially soluble in hydrocarbon oils, rather
been used to probe the nature of the oil penetration into the interfa- the polarity of the majority of poorly water-soluble drugs favour
cial surfactant monolayer of the microemulsion [82]. Freeze-fracture their solubilisation in small/medium molecular volume oils such as
electron microscopy has also been used to study microemulsion tributyrin or Miglyol 812. In fact formulating a drug in a hydrocarbon
structure, however extremely rapid cooling of the sample is required oil-in-water microemulsion may offer no advantage in terms of
in order to maintain structure and minimise the possibility of artifacts solubilisation over the corresponding micelle [92]. The dispersal of
[75,83–85]. the drug as a solution in nanometre-sized droplets enhances the
A potentially serious limitation with some of these methods of rate of dissolution into a contacting aqueous phase, and in vivo gener-
analysis lies in the requirement to dilute the microemulsion systems ally results in an increase in drug bioavailability. It is also noteworthy
in order to eliminate particle–particle interactions. This is because that the use of o/w microemulsions in drug delivery is more straight-
dilution can drive a phase transition or a molecular reorganisation, forward than is the case with w/o microemulsions. This is because the
and is therefore a particular problem for techniques such as viscometry, droplet structure of o/w microemulsions is often retained on dilution
NMR self-diffusion measurements and those relying on scattering. It is by a biological aqueous phase, thereby permitting oral as well as par-
therefore often necessary to work with systems containing a relatively enteral administration. However, the process of dilution will result in
high dispersed phase concentration and to account for interparticulate the gradual desorption of surfactant located at the droplet interface.
interactions by use of a model. On the one hand this provides an oppor- This process is thermodynamically driven by the requirement of
tunity to extract useful information regarding particle–particle interac- surfactant to maintain an aqueous phase concentration equivalent
tions, but on the downside it to makes the structural characterisation of to its critical micelle concentration (CMC) under the prevailing
concentrated systems extremely problematic.
In spite of the abovementioned complications, much of the work
reported in the pharmaceutical literature has been conducted using
concentrated microemulsion systems. For the most part where particle
sizes obtained using photon correlation spectroscopy, the measure-
ments quoted remain uncorrected, not least because such corrections
are far from trivial. Only a few studies have attempted to correct for
these interactions [86,87], yet without such corrections, these data
should only be used to establish the presence of microemulsion struc-
ture. Whilst neglecting correction factors, some workers have attempted
to correlate apparent droplet sizes obtained in concentrated systems
with oral bioavailability. Unsurprisingly, this approach has met with
very limited success.

4. Microemulsion-based systems in drug delivery

Fig. 9. Influence of the addition of gelatin to a water-in-dodecane microemulsion
Microemulsions have generated considerable interest over the stabilised by AOT on viscosity and conductivity. [Water]/[Surfactant] molar ratio of
years as potential drug delivery systems [43–48,88,89]. Advantages 45 and AOT molarity of 0.15. Redrawn from [36].
182 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

Table 1 and tetraethylene glycol monododecyl ether (C12E4). Based on the re-
Equilibrium phase structures encountered in oil–water–surfactant systems. From [46]a sults of these studies a number of workers have developed guidelines
Micelles/swollen micelles/microemulsions to aid in the formulation of microemulsions. However care must be
Reverse micelles/reverse swollen micelles/reverse microemulsions taken when extrapolating ‘guidelines’ developed for such systems
Worm (polymer)-like micelles to pharmaceutically acceptable systems. For example it is widely
Cubic/reverse cubic phases
reported that it is not possible to solubilise an oil which is larger
Hexagonal/reverse hexagonal
Oil and/or water swollen lamellar phases than the hydrophobic chain length of the surfactant, yet recent stud-
a ies have shown that using C18:1E10, it is possible to solubilise long
In addition to the above, one phase structures, a range of two and three phase
systems are observed, in which one of the above exists in equilibrium with an excess chain triglyceride to a greater extent than its medium or short chain
of oil and/or an excess of water. counterparts [42,92].
Nevertheless, quite a number of studies have deliberately chosen
to employ naturally occurring excipients such as lecithins and glycer-
conditions of temperature, pH and ionic strength. Because non-ionic ides, and there is even advice in the literature on the formulation of
surfactants typically have lower CMCs than their ionic counterparts, supposedly non-toxic microemulsions [102]. The regulatory status
o/w microemulsion dosage forms based on non-ionic surfactants of the different excipients will also depend on their intended use.
and designed for oral or parenteral use are likely to offer superior in Thus while a reasonable range of surfactants would be deemed ac-
vivo stability. ceptable for use in topical formulations, the number considered safe
In contrast, the use of w/o microemulsions for oral or parenteral for oral and especially parenteral use would be very restricted in
drug delivery is complicated by the fact that they are destabilised to comparison. This observation may go part way to explaining why a
a much greater extent when diluted by an aqueous phase. This is large number of researchers investigating microemulsions for drug
due to the increase in the volume fraction of the aqueous phase delivery have concentrated on developing topically applied delivery
which increases the ratio of water to surfactant (ωo) leading to drop- systems. Many of the surfactants and oils that are regarded as accept-
let growth and eventually percolation. If the dilution continues, phase able are food grade materials or have a history of use in the pharma-
separation or inversion may occur and this will result in load dump- ceutical arena [103], for example as parenteral emulsion dosage
ing. However, there are advantages to be gained from formulating forms [104]. There are a number of comprehensive studies in the lit-
drugs in w/o microemulsion systems. Peptide drugs, for example, erature concerning the preparation and characterisation of lecithin-
generally have little or no activity when delivered orally and are based microemulsions [15,19–21,25,26,52,56,59,65–69,87,89,97,99,
highly susceptible to proteolysis in the gastrointestinal tract [93]. 105–125].
Parenteral drug administration, especially for chronic conditions is Non-ionic surfactants can be useful alternatives to naturally occur-
not well accepted by patients and can lead to issues with compliance. ring surfactants, and polyoxyethylene sorbitan n-acyl esters (Tweens),
Consequently, the oral delivery of labile drugs is the focus of growing for example, have been reported to have minimal toxicity. Although
attention, particularly as many of the new therapeutic agents in there are some restrictions, the use of polyoxyethylene sorbitan
development are hydrophilic drugs such as peptides or oligonucleo- monooleate (Tween 80) and polyoxyethylene sorbitan monolaurate
tides. Hydrophilic drugs of this kind can be successfully incorporated (Tween 20) appear acceptable for oral or parenteral use [126]. Impor-
into the dispersed aqueous phase of w/o microemulsion droplets tantly in some cases non-ionic surfactants such as the polyoxyethylene
where they are afforded some protection from enzymatic degradation n-alkyl ethers (CnEm) are able to form microemulsions without the
when administered orally [88]. In addition, the presence of surfactant need for cosurfactant [127,128]. This is helpful as it reduces the
and in some cases cosurfactant, for example medium chain diglycer- complexity of the phase behaviour, and eliminates the requirement
ides in many cases serves to increase membrane permeability thereby for inclusion of medium chain alcohols, since these cosurfactants have
increasing drug uptake [88,89,94–98]. Drug delivery forms based on a poor toxicity profile. Furthermore, the insensitivity of non-ionic
w/o microemulsions can also be employed where dilution by an microemulsions to pH and electrolyte concentration relative to their
aqueous phase is less likely to occur, such as after intramuscular in- ionic counterparts represents an added benefit. There are consequently
jection [99]. Similarly, microemulsions and microemulsion gels have a large number of formulation studies involving non-ionic surfactants
found application as topical agents where the surfactants and in as microemulsion excipients [18,38,86,92,127–133]. However, the bio-
some cases the oil phase itself act as penetration enhancers to facili- degradability of many non-ionic surfactants raises issues with regard
tate transdermal drug delivery [85,90,91,100,101]. to long-term toxicity, especially in chronic use.
Clearly then, in order to aid successful formulation, it is essential It is evident from the literature that there is a move toward the use
to study the phase behaviour of potential combinations of water, sur- of alternative ‘safe’ surfactants other than lecithin and the non-ionic
factant and oil. However, the studies can only ever act as a broad pre- surfactants named above. Thus there are a number of studies that
dictor of the likely fate of a microemulsion dosage form after delivery. have been conducted using n-alkyl amine N-oxides either in isolation
For oral dosage forms in particular, it is not possible to realistically or in combination with lecithin [42,114]. Interestingly these surfac-
model the in vivo interactions of microemulsion components with tants are much more rapidly biodegradable than the polyoxyethylene
the complex variety of food materials and digestive fluids present in n-alkyl ethers. Furthermore there has also been a recent report detail-
the gastrointestinal tract. ing the potential of a biodegradable version of the amine N-oxide
surfactant, dimethyldodecylamine N-oxide, for use in microemulsion
4.1. Pharmaceutically acceptable excipients formation [134]. Sugar surfactants such as alkyl glucosides have also
received a good deal of attention [64,135–140]. Interestingly while
Perhaps the most significant problem associated with formulating sugar surfactants are widely considered as biodegradable they still
microemulsions is the difficulty associated with excipient acceptabil- exhibit a level of haemolytic activity on a par with that exhibited by
ity. The vast majority of phase behaviour studies have been carried the polyoxyethylene n-alkyl ethers. The use of sucrose fatty acid
out using surfactants and oils, which do not have regulatory approval esters as surfactants in the stabilisation of microemulsion phases
for use in pharmaceutical products. The majority of the work reported has also been investigated [75,85,141–144]. Surfactants based on
in the scientific literature concerns microemulsion systems based on polyglycerol fatty acid esters have been suggested as ‘orally safe’ in
either hydrocarbon oils such as heptane or dodecane, or cyclic oils a study designed to investigate the potential of microemulsions for
such as cyclohexane. The most commonly used surfactants contain the delivery of protein drugs [145]. Monoglyceride has also been
hydrophobes of 12 carbons such as sodium dodecyl sulphate (SDS) used as surfactant in its own right to stabilise a triglyceride-in-
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 183

water microemulsion system [83], and by Constantinides and co- vitamin K and steroids such as hydrocortisone, prednisolone and
workers [96,146] in a self-microemulsifying drug delivery system betamethasone [34,106]. The solubilisation of pilocarpine and chlor-
(SMEDDS). amphenicol was looked at in both AOT and DDAB w/o microemulsions
In many cases, a requirement for cosurfactant causes difficulty in [31], whilst the trandermal delivery of glyceryl trinitrate has been
the formulation of acceptable microemulsions because the majority tested in vivo from a variety of related AOT systems [149]. As has
of studies have chosen to employ medium chain length alcohols as been indicated previously, the advantage of twin tailed surfactants
the cosurfactant of choice. Unfortunately, there are significant toxicity like AOT and DDAB is that w/o microemulsions form easily without
and irritancy issues with these materials, which preclude their use the aid of cosurfactant. However, when AOT microemulsions are for-
in pharmaceutical formulations. In addition, the aqueous solubility mulated in combination with non-ionic surfactants, the extent of the
of cosurfactants in o/w microemulsion systems is often higher microemulsion region is often markedly increased, and in addition
than that of the principal surfactant. Consequently, when the o/w the temperature range over which the microemulsion is stable is gen-
microemulsion is diluted, the cosurfactant partitions more strongly erally improved [18,33,43]. Fig. 10 gives an example of the increase in
to the aqueous phase. This has the effect of depleting the cosurfactant microemulsion area of existence achievable in a mixed AOT-non-ionic
concentration at the oil/water interface, thereby destabilising the surfactant system over that seen when either surfactant is used alone.
microemulsion droplet. Alternatives to medium chain alcohols have Total parenteral nutrition emulsions have provided a starting point
been evaluated such as short chain amines [147] and alkanoic acids from which suitable oils might be selected for formulation into phar-
[21], however these cosurfactants behave in much the same way as maceutically acceptable microemulsions. Medium chain triglycerides
the alcohols and toxicity remains an issue. A number of small, rela- such as Miglyol 812 have been used quite frequently [42,128,148],
tively polar molecules are also thought to act as cosurfactants. Etha- but fatty acid esters are also a popular choice. Of these, isopropyl
nol, for example, has been used as a ‘cosurfactant’ in both o/w and myristate (IPM) is the most popular [19,21,34,38,66,67,69,71,74,86,
w/o microemulsion systems. Polyhydric alcohols such as sorbitol 90,106,115,117–120,123,124,141,150–154], but isopropyl palmitate
and sucrose have also been used as additives to facilitate micro- (IPP) has also found application [150,155], especially in micro-
emulsification but the resulting o/w formulations were unstable on emulsion gels [100,151,156,157]. Ethyl or methyl esters of lauric,
dilution with water as the aqueous solubility of these materials is myristic and oleic acid have also been employed as pharmaceutical
high [86,148]. excipients for microemulsion formulation [63,92,99,108,151]. Ryan
The utility of ionic surfactants is also relatively limited in pharma- and Kaler have employed alkyl ethylene glycol ethers such as ethylene
ceutical dosage forms, however the use of ionic surfactants in formu- glycol diethyl ether and ethylene glycol dibutyl ether as the oil phase.
lation studies has been reported on a number of occasions [18,33]. The combination of these rather hydrophilic oil phases with alkyl
Coformulation of the anionic surfactant AOT with a number of differ- glucoside surfactants enhances the surfactant solubility in the oil
ent drugs as a microemulsion dosage form has been more widely phase and microemulsification can be achieved without the need for
reported. Examples, which have appeared in the literature, include cosurfactant [137,139]. However, with one or two exceptions it is

Fig. 10. Partial ternary phase diagram for Tween 85 (dot) and AOT (solid) alone, or as a surfactant mixture in combination with isopropylmyristate (oil) and H2O at 298 K.
Compositions to the right of the phase boundary are optically clear, single phase, water-in-oil microemulsions. The weight ratios of Tween 85:AOT employed were 2:1 (dash),
1:1 (dash-dot-dot) and 1:2 (long dash). [S. Kantaria, G.D. Rees, M.J. Lawrence, unpublished observations.]
184 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

clear that the majority of oils intended for pharmaceutical use are
large and semi-polar, and therefore rather different to the alkane oils
most commonly reported in the scientific literature.

4.2. Low viscosity microemulsion systems

At relatively low dispersed phase volume fractions, the micro-

emulsion generally contains nanometre-sized droplets of oil or
water. If the droplets are non-interacting the resulting microemulsion
will be low viscosity and may therefore be suitable for oral, parenter-
al, pulmonary or even ocular delivery. The potential of these systems
for use topically has not been overlooked, however it is recognised
that more viscous microemulsion systems are preferred for this appli-
cation. Suitable systems would include those containing extended
structures such as cylindrical or polymeric worm-like or micelles
[52,59] or which have been thickened through the addition of a spe-
cific gelling agent [90,91]. Cubic or hexagonal mesophases iden-
tified in the bicontinuous microemulsion region may also be of
sufficiently high viscosity to have application as topical delivery
systems [38,85,142].
Oil-in-water microemulsions have been used to solubilise steroidal
drugs such as prednisolone, hydrocortisone and betamethasone, testos-
terone and its esters and progesterone [34,92,106,127]. Interestingly, it
has been noted that hydrophobic drugs need to have a significant solu-
bility in the dispersed oil phase for the o/w microemulsion system to
offer a marked benefit over the micellar system alone [127]. However,
the oils in which the drug is most soluble do not necessarily form
microemulsions with the highest drug solubilisation capacity [92].
The utility of w/o microemulsions for the sequestration of labile
peptide drugs was demonstrated using an analogue of luteinizing Fig. 11. (a) Release of diclofenac diethylamine from vehicle with 5% diclofenac diethylamine
hormone-releasing hormone (LHRH) to control testosterone levels and various concentrations of phospholipids at 20 °C. (A,B) Microemulsions with (A) 5%
in rats [99]. The LHRH was solubilised in a lecithin-stabilised system phospholipids and (B) 6% phospholipids; (C–E) anisotropic gel with (C) 9.3%, (D) 13.8%
with ethyl oleate as the oil phase and hexanoic acid as cosurfactant, and (E) 20% phospholipids, respectively. (b) Effective diffusion coefficient of diclofenac
diethylamine from vehicle with 5% diclofenac diethylamine vs content of phospholipids.
and was administered as an intramuscular injection.
Data represent mean and standard deviation of three duplicate determinations. Redrawn
The peroral delivery of insulin, cyclosporine A and vasopressin from from [112].
microemulsion formulations has been reviewed by Ritschel [158]. Inter-
estingly, some drugs such as diclofenac diethylamine are amphiphiles
in their own right and indeed this drug self assembles under aqueous successfully conducted using indomethacin, diclofenac and chloram-
conditions to yield a liquid crystal phases. Coformulation of diclofenac phenicol [160].
diethylamine with phospholipid, not unexpectedly, gives rise to some Antifungals including clotrimazole, cyclopiroxolamine and econazole
complex phase behaviour, and microemulsion phases have been identi- nitrate have been incorporated into IPM and IPP microemulsions
fied from the ternary phase diagram [159]. Diclofenac release and designed for topical use and stabilised by Tween 80 [150]. Azelaic acid,
permeation through excised human skin was subsequently reported a drug used in the treatment of pigmentary disorders, has also been
[112] with the researchers clearly demonstrating the importance of investigated as a topical microemulsion formulation. In this case the
the nature of the phase structure of the amount of drug released azelaic acid was solubilised in a viscosised o/w microemulsion and tested
(Fig. 11). Cholesteryl ester prodrugs of ibuprofen and flufenamic acid in a hairless mouse model where flux enhancements were observed
have also been incorporated into phospholipid microemulsions. In this [161]. The local anaesthetic pentacaine has been incorporated into a
case the dispersed oil phase was modified by the addition of cholesteryl w/o microemulsion and subjected to in vivo testing on rabbits [162].
oleate [109]. Interestingly, microemulsions have not only been used directly as
An interesting non-aqueous microemulsion formulation based drug delivery vehicles, but have also been used indirectly as a means
on lecithin/glycerol/soybean oil has been reported. The formulation of producing solid drug-loaded nanoparticles. In this case, solid lipid
was designed for parenteral delivery and contained the octyl ester nanoparticles were synthesised by spraying an o/w microemulsion
of ibuprofen. Modification of the microemulsion formulation with a into cold water at a temperature below the melting point of the
poloxamer alters the absorption process with the result that the oil. The microemulsion itself is typically stabilised by lecithin and
drug is targeted to reticuloendothelial system-rich organs [113]. uses a mixture of capric and palmitic acid as the oil phase. The micro-
A novel pressurised aerosol system has been devised for the pulmo- emulsion in addition may contain taurodeoxycholate and alkyl
nary delivery of salbutamol using lecithin-stabilised microemulsions phosphates as cosurfactants. The resulting drug loaded particles
formulated in trichlorotrifluoroethane [15]. The ocular use of micro- could be isolated, washed and freeze dried [163–169]. The proper-
emulsion systems has also been indicated and the β-blocker levobunolol ties of w/o microemulsions were similarly exploited by utilising
has been incorporated into o/w microemulsions with this purpose droplets as supermolecular templates for the synthesis of drug-loaded
in mind [107]. Pilocarpine has also been incorporated into a micro- polybutylcyanoacrylate nanoparticles [170,171].
emulsion designed for ocular use, which was stabilised by sucrose
fatty acid ester surfactants [141]. Poloxamer-stabilised microemulsions 4.3. High viscosity systems and microemulsion gels
containing triacetin, castor oil, water and propylene glycol have also
been formulated as potential ophthalmics. Formulations were prepared A bicontinuous microemulsion stabilised by a mixture of non-
over a range of compositions, and in vitro drug permeation studies were ionic surfactants was designed for the topical delivery of the local
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 185

anaesthetic lidocaine [38]. Comparative studies have shown that an to demonstrate greatly improved pharmacodynamics using systems
o/w microemulsion gel containing the analgesic anti-inflammatory based on medium chain triglyceride (MCT) and ethoxylated glyceryl
flufenamic acid and stabilised by PEG-7 glyceryl coccoate (Cetiol trioleate (Tagat TO) [178]. More recently, self-emulsifying w/o micro-
HE) outperforms the corresponding macroemulsion, hydrogel and emulsions based on MCTs such as Captex 355 and Captex 8000 have
cream. The formulation was tested in vitro and in vivo by topical been reported. The systems contained a mixture of mono and diglycer-
administration on rats [172]. ides (Capmul MCM) in combination with Tween 80 as surfactant. The
The interaction of cylindrical or worm-like micelles in micro- bioavailabilities of calcein, a water-soluble marker, and an RGD peptide
emulsion formulations tend to gives rise to high viscosity systems. were shown to be significantly increased using a microemulsion con-
The proposed structure of a worm-like (or spaghetti-like) micelle centrate and preformulated w/o microemulsions compared to the con-
is shown in Fig. 12. Lecithin in particular is known to form such sys- trol aqueous formulation [96,97,146]. The bioavailability of a poorly
tems in microemulsions at low water content and the utility of the water soluble 5α-reductase inhibitor has similarly been shown to be
resulting organogels as novel matrices for the transdermal transport improved in Beagle dogs [179]. It is also notable that the presence of
of drugs has been investigated [100]. Scopolamine, broxaterol and liquid crystalline phases in the pseudo binary oil/surfactant mixtures
propranolol have been incorporated into lecithin organogels based are claimed to be a feature of the most efficient SEDDS [180].
on either cyclohexane, isooctane or IPM. Ten-fold enhancements of The most notable example of a SMEDDS relates to the oral delivery
permeation rates through excised human skin were reported com- of cyclosporin A, in particular the commercial Neoral ® formulation.
pared to solvent only controls [100,173]. Aromatic tetra-amidines Cyclosporin A is a cyclic undecapeptide used as an immunosuppres-
with antitumour activity have also been incorporated into lecithin sant in transplantation surgery, and in contrast to most peptide
organogels for transdermal drug delivery [151]. The in vivo efficacy drugs is hydrophobic. The original Sandimmune ® formulation was
was determined in nude mice carrying xenografted tumour cells based on a solution of cyclosporin A in vegetable oil. Although the
and was considered to be encouraging. Methyl nicotinate has been coadministration of triglyceride with the cyclosporine A improved
formulated in a lecithin IPM organogel and tested in vivo with its bioavailability, there was considerable pharmacodynamic inter-
human subjects [174]. The utility of these lecithin organogels has and intra-patient variability as can be seen in Table 2 and Fig. 13.
been supported by a human skin irritation study, which showed a This variation can be ascribed to the proposed mechanism of uptake
very low irritancy potential for the soybean lecithin/IPP/water system in vivo which is considered to be related to the lipolysis of the triglyc-
[156,175]. eride yielding lower partial glycerides which then act as emulsifiers
and enhance drug uptake [181]. The Neoral ® formulation uses an
4.4. SMEDDs isotropic concentrated blend of surfactant based on medium chain
length partial glycerides, a medium chain length triglyceride oil and
Self-emulsifying drug delivery systems (SEDDS) and SMEDDS drug. Exposure of this concentrate to water results in formation of
can be described as isotropic solutions of oil and surfactant, which initially a w/o microemulsion which on further mixing with water
form o/w (micro)emulsions on mild agitation in the presence of undergoes phase inversion to yield an o/w microemulsion. The deliv-
water [176,177]. The utility of SEDDS has been investigated by ery of cyclosporine A via microemulsion formulations has been con-
Charman and coworkers who, although unable to show enhanced bio- sidered in some detail [158,182,183], and the superiority of the
availability of an investigational lipophilic drug WIN 54954, were able Neoral® microemulsion preconcentrate over the original Sandimmune®

Fig. 12. Microemulsion-based lecithin gels. (a) schematic representation of the formation of lecithin gels upon addition of water to small lecithin reverse micelles in apolar solvents.
(b) localisation of solubilised ‘guest’ molecules within the lecithin gels. Lipophilic drug (stripped bar); hydrophilic drug (black circle); and amphiphilic drug (shaded head with
attached tail). Redrawn from [209].
186 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

Table 2 where surfactant selection is altered. Alternatively, and as has just

Mean (CV%) pharmacokinetic parameters following twice-daily dosing with Sandimmune® been discussed, changes in bioavailability may occur where there
(SIM) or Neoral® by 11 stable renal transplant patients. From [184]a
are significant differences in the biodegradability of formulation
Parameter SIM fasting SIM non-fasting Neoral fasting Neoral non-fasting components which might lead to the production or otherwise of addi-
Tmax (h) 2.1 (33.3) 2.6 (76.9) 1.5 (33.3) 1.2 (33.3) tional surfactant species.
Cmax (μg/l) 663 (34.5) 528 (40.5) 997 (20.0) 892 (35.8) An interesting SMEDD variant has been employed for the dermal
Cmin (μg/l) 78 (30.8) 92 (29.3) 94 (22.3) 100 (23.0) delivery of β-blockers in an anhydrous surfactant/IPM vehicle [155].
AUC (μg-h/l) 2645 (25.7) 2432 (24.3) 3454 (17.6) 3028 (19.7)
After topical administration with an occlusive patch, water partitions
PTF% 261 (23.4) 212 (36.8) 317 (18.0) 309 (31.1)
into the SMEDD forming a microemulsion gel phase. As the extent
a
All concentrations measured in whole blood at steady state. of hydration increases, the drug solubility decreases resulting in a
PTF%
AUC was measured over a dosing interval. PTF%=percentage peak–trough fluctuation.
supersaturated system which the authors claim results in enhanced
drug activity and improved pharmacodynamics [192]. In a related
formulation has been demonstrated on several occasions [184–186]. report, organogels were prepared from lecithin and a fatty acid
Furthermore, the available data has been reviewed by Noble and ester oil containing indomethacin [193]. This pseudo binary system
Markham who confirm the conclusion that Neoral ® offers more was effectively anhydrous but in common with the previously men-
predictable and more extensive drug absorption than the standard tioned study, water uptake would be expected to result in the forma-
Sandimmune ® formulation [187]. tion of the standard lecithin microemulsion gel which has been
Interestingly, it has been noted that the bioavailability of vaso- previously discussed. As expected, enhanced permeation rates were
pressin and insulin from w/o microemulsions is higher when the oil reported for the indomethacin gel in a hairless mouse model.
phase is based on straight-chain rather than branched-chain fatty
acid esters [182]. It is likely that the observed differences in bioavail- 5. Recent developments and future directions
ability are related, at least in part, to the large reduction in lipolytic
activity exhibited by lipases toward branched chain fatty acid sub- For the purposes of this review, recent developments will for the
strates [188]. After administration, the microemulsion formulated most part constitute an evaluation of the literature in the area of
with straight chain fatty acid esters will undergo rapid enzymatic hy- microemulsions and microemulsion-based systems for drug delivery
drolysis being degraded in the gastrointestinal tract. The breakdown for the period beginning 1997 to date.
products are surface active and will stabilise any (micro)emulsion
that may form, as well as acting as membrane permeation enhancers 5.1. Component selection
[189]. As a consequence of the important role played by metabolic
processes in vivo, formulators should be aware that certain hydro- It is notable that an increasing proportion of new studies recognise
philic surfactants such as Brij 96/Brij 97 (C18:1E10), Tween 80 and the benefit associated with employing as far as possible, pharmaceu-
polyoxyethylene 40 hydrogenated castor oil (Cremophor RH40) tically acceptable surfactants, cosurfactants and oils [194]. Naturally
have been shown to inhibit lipolysis in vitro. Clearly if this behaviour occurring surfactants and oils remain an attractive option, and the
is mirrored in vivo one of the principal mechanisms facilitating drug phase behaviour and microstructure of microemulsions based on
uptake would be compromised. It is therefore encouraging to observe soybean phosphatidylcholine and triglycerides has recently been
that co-incubation of these hydrophilic surfactants with lipophilic reported, although propanol was used as the cosurfactant [65]. The
surfactants such as oleic acid or medium chain length partial glycer- addition of cosurfactants such as alkanol phosphocholines have
ides reverses the inhibition [181]. been shown to increase the extent of the microemulsion region in
It is also notable that in the case of w/o microemulsion systems, lecithin/triglyceride systems [120]. However the regulatory status of
there is no obvious correlation between droplet size and oral bioavail- these particular alkanol phosphocholines, which were synthesised
ability. This contrasts with the known relationship between o/w ‘in house’ is unclear, and once again short chain alcohols including
emulsion droplet size and bioavailability [190,191]. Ignoring the ethanol, propanol and butanol were used as cosolvents/cosurfactants.
problems of sizing such concentrated systems, this behaviour is The same group has also reported on the phase behaviour of systems
expected since the w/o microemulsion phase inverts or phase sepa- containing lecithin and 2-acyl lysolecithin derivatives coformulated
rates after administration, thereby releasing any drug previously with medium chain triglycerides or IPM as the oil phase and ethanol
compartmentalised in the aqueous droplet core. Where differences as cosurfactant/cosolvent. These modified phospholipids also increase
in bioavailability would be predicted for w/o systems are occasions the extent of the microemulsion region most probably by increasing
the fluidity of the surfactant film [124]. Significantly, most of the
work investigating the formation of lecithin-based microemulsions
has reported an improvement in the extent of the w/o region. In
contrast, relatively little success has been achieved in the water-rich
part of the phase diagram. However a recent study describes the for-
mation of o/w lecithin-based microemulsions in combination with
amine N-oxide surfactants [195].

5.2. Phase behaviour

The phase behaviour of systems stabilised by non-ionics other

than polyethylene glycol ethers continues to be a prominent area of
research activity. A number of phase behaviour studies involving
dermatologically acceptable sugar surfactants such as alkyl glucosides
have appeared. As has been indicated previously, the low solubility of
Fig. 13. Representative cylcosporine blood concentration profiles from a renal trans- sugar esters in hydrophobic alkane oils generally necessitates the use
plant patient given the currently marketed formulation Sandimmune® (SIM) or the of a cosurfactant such as a medium chain alcohol in order to facilitate
new formulation (Neoral) without food (a.m.) or with food (p.m.). Redrawn from microemulsification. However, unlike their polyoxyethylene surfac-
[184]. tant counterparts, surfactants based on sugar esters are much less
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 187

sensitive to temperature [74]. Until recently it was not possible to and castor oil, water and propylene glycol over a range of composi-
prepare microemulsions stabilised by alkyl glucosides in the absence tions [160].
of a cosurfactant. However it has recently been reported, that by Block copolymer microemulsions are likely to become the focus of
employing a polar oil phase based on alkyl ethylene glycol ethers, in more intense research over the next few years, particularly where
which the alkyl glucosides are more soluble, it is possible to prepare drug delivery applications are concerned. Mechanistically, the dy-
glycoside stabilised microemulsions in the absence of cosurfactant namics associated with such systems have much in common with
[137–139]. An interesting alternative approach in microemulsions conventional microemulsion stabilised by non-polymeric surfactants.
stabilised by octyl monoglucoside has been to employ geraniol, a However, the time-scales associated with molecular events such as
non-toxic perfume alcohol (C10H17OH) as a cosurfactant/cosolvent unimer insertion or exchange can be several orders of magnitude
[140]. As is the case with addition of alcohols or alkyl glycerol ethers slower in the case of block copolymers [207]. The aforementioned dif-
to alkyl glucosides, the addition of geraniol allows phase inversion ferences would be expected to have considerable impact on micro-
of the system to occur. Microemulsions formulations stabilised by emulsion stability and the control of drug release in vivo.
sucrose monolaurate and sucrose dilaurate containing ethyl and cetyl
2-(hexylethyl)-2-hexanoate as the oil phase and diethyleneglycol 5.4. Comparative drug delivery studies
monoethyl ether as cosurfactant have also been reported. These sys-
tems were characterised using FFEM, SANS and viscosity [75]. The for- A number of comparative reports have appeared in the literature,
mation of microemulsion and liquid crystal phases in biocompatible some of which have evaluated the utility of microemulsion for-
sucrose alkanoate systems has also been claimed [61]. Very recently, mulations against alternative delivery systems including liposomes,
a study has appeared examining the water solubilisation capacity micelles, emulsions and creams [208,209]. Regrettably, the picture is
of microemulsions stabilised by a variety of sucrose esters including often complicated as the same components are not always employed
mono, di and polyesters of lauric, palmitic, stearic and oleic acid. Medi- in the production of the various systems under test. Nevertheless, the
um chain triglyceride was employed as the oil phase, however the study antitumour drug camptothecin, for example, has been tested in vitro
employed medium chain alkanols as cosurfactants [144]. for anti-proliferative activity toward cultured human leukaemic K
The influence of drug incorporation on the phase behaviour and 562 cells as a micellar solution, a microemulsions and incorporated
structure of microemulsion system based on water, propanol, lecithin into liposomes. Activities in all cases were moderately enhanced over
and medium chain triglyceride has also been investigated [125]. The the control. Isostearyl isostearate was employed as the oil phase and
preparation of non-toxic microemulsions has been described for the microemulsion stabilised by a mixture of Labrosol® and Plurol
mixtures of IPM or orange oil with lecithin. Notably, alkane diols isostearate® [208]. The peroral delivery of parasympatholytic trospium
such as pentane-1,2-diol and octane-1,2-diol were employed as pos- chloride has been compared using two w/o microemulsion formula-
sible non-toxic replacements for propanol and pentanol, respectively tions as well as a cyclodextrin formulation and an aqueous solution
[119]. Interestingly in their study involving the use of neural net- control. The trospium carries a quaternary nitrogen and is positively
works to predict microemulsion phase behaviour Richardson and co- charged which may play a role in its poor peroral bioavailability (3 to
workers predicted that 1,2 hexanediol would be a suitable surfactant 11%). Uptake of trospium may also be inhibited by formation of insolu-
for the production of balanced lecithin-based microemulsions [196]. ble complexes with acidic residues of mucopolysaccharides. This would
In terms of drug solubilisation capacities, microemulsions should be consistent with the observation that ion pair formation increases
fare better than micelles because of the extra locus for solubilisation its lipophilicity and improves bioavailability. Disappointingly, in this
provided by the oil phase. However, a recent phase behaviour study particular study, bioavailabilities were either equivalent or lower than
has shown that whilst polar oils offered better drug solubility, the aqueous solution control [210]. Bicontinuous microemulsions based
solubility of the drug in the microemulsion itself was also determined on sucrose monolaurate or sucrose dilaurate have been used for the
by the extent to which low molecular volume oils penetrate the PEO topical delivery of niflumic acid, a potent anti-inflammatory. The 1%
region and destroy drug solubilisation sites [92]. microemulsion was as effective as the 3% commercially available
niflumic acid ointment [85]. An in vitro comparative study of the percu-
taneous absorption of propranolol from an emulsion, an o/w micro-
5.3. Block copolymer micelles and microemulsions emulsion and a micellar solution stabilised by Tween 80 and Span 80
has also been reported. Although the o/w microemulsion was superior
It is noticeable from the literature that there has been a significant to the emulsion system, both of which contained IPM, the micellar
increase in the number of papers describing the use of block copoly- delivery system containing propranolol was the most effective [152].
mer systems for drug delivery [197–204]. Given that both block copol-
ymers and conventional surfactants are amphiphilic, it is unsurprising 5.5. Oral delivery — SEDDs and SMEDDs
that there is great similarity between the self-association and adsorp-
tion properties of these two classes of compounds. However, an The formulation of w/o microemulsions for use as SEDDS or
advantage of amphiphilic block copolymers over conventional surfac- SMEDDS has been investigated using blends of low and high HLB
tants is the relative ease with which the physicochemical properties surfactants, which were commercially available and pharmaceutically
can be ‘tailored’ to suit a given application. Specific examples might in- acceptable, typically sorbitan esters and Tween 80. The oil phase
clude the glass transition temperature (Tg) of the hydrophobic core comprised long or medium chain length glycerides [73]. The micro-
and the CMC [197,200]. In addition, amphiphilic block copolymers emulsions were characterised by a variety of techniques including
may be less haemolytic than conventional surfactants [201]. As is conductivity, viscosity and photon correlation spectroscopy (PCS).
the case with conventional surfactants, the addition of oil or water re- In a related study, an optimised o/w microemulsion formulation for
spectively to block copolymer micelles or reverse micelles can lead to the delivery of cyclosporin A prepared using Cremophor EL ® as
the formation of o/w or w/o microemulsion systems, respectively. The surfactant, Transcutol ® as a cosurfactant and Captex 355 ® as the oil
physicochemical characterisation of block copolymer microemulsions phase has been reported. Bioavailability enhancements of 3.3 and
has been reported, and some of their potential applications examined 1.25 were observed relative to the Sandimmune® and Sandimmune
[205,206]. However, their formulation and potential application in Neoral ® formulations [211].
drug delivery has received little attention to date with the exception The formulation of SEDDS has also been recently reviewed and a
of a study by Siebenbrodt and Keipert which examined the for- number of pertinent observations made [98]. Concerns have been
mulation of Poloxamer-stabilised microemulsions containing triacetin raised for example regarding administration of dosage forms containing
188 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

high concentrations of surfactant, for example, and opportunities to malarial halofantrine has also been reported. The SEDDS and SMEDDS
evaluate chronic toxicology are limited in comparison to tablet formula- were isotropic mixtures of medium or long chain triglyceride, mono-
tions [95]. As has been indicated already, the presence of water or an- glyceride (Capmul MCM), drug and ethanol. Six- to eight-fold improve-
other polar cosolvent in a SEDDS formulation may mean that the ments in bioavailability were observed relative to tablet formulations
concentrate is itself a microemulsion. In highlighting factors that predis- [216].
pose efficient microemulsions the authors note that self-emulsification
requires least energy close to the PIT, which is also where the 5.6. Parenteral, pulmonary and ocular delivery
capacity for water solubilisation is enhanced. Mechanisms of self-
emulsification discussed include the dynamic formation of liquid crys- The preparation and evaluation of flurbiprofen-loaded o/w micro-
talline units at the oil/water interface, at least for simple SEDDS for emulsions is one of the few recent reports of delivery systems
example based on Tween 85, a medium chain triglyceride and drug designed for parenteral use. The systems of interest were prepared
[98]. For more complex SEDDS, a mechanism of ‘diffusion and stranding’ using ethyl oleate as the oil phase and Tween 20 as surfactant. The
producing fine emulsions or microemulsions may also operate. The drug solubility was eight times higher than in buffer, but there was
formulation of a SMEDDS containing flurbiprofen and based on phos- no significant difference in the pharmacokinetics after administration
pholipid, ethyl oleate and ethanol has also been reported. After paren- in rats [217]. Pharmaceutically acceptable microemulsions designed
teral administration, significant increases in flurbiprofen half-life and for i.v. administration have recently not only been formulated and
an altered biodistribution pattern were observed [122] as is clearly illus- characterised, but also tested in vivo for hemodynamic response
trated in Fig. 14. In another highly pertinent study, the influence of [68]. The microemulsions comprised Miglyol 810N (MCT), soybean
phase transformation on indomethacin release from microemulsions phosphatidylcholine (Epicuron 200), PEG 400, poly(ethylene glycol)
is examined [123]. The systems of interest were prepared from water (660)-12-hydroxystearate and ethanol. PFG–NMR indicated that the
(42%), IPM, lecithin, lysolecithin and ethanol. The compositions were microemulsions formed over a range of compositions were bicontinuous,
selected such that emulsions, microemulsions, or liquid crystal were even at high oil concentrations. After administration, the bicontinuous
formed on dilution. With the exception of phase transformations to microemulsions form o/w emulsions on dilution. In vitro studies showed
liquid crystals, the release rate of indomethacin was considered by the the resulting droplets were small, with mean radii typically in the
authors to be too rapid for controlled drug delivery. However, the range 60–200 nm. Solubilisation studies using felodipine (a calcium
release rate of drugs in general from these systems will also depend antagonist) and an antioxidant H 290/58 were conducted, however in
on the oil/water partition coefficient and factors such as droplet size vivo studies were performed using the drug-free systems. The i.v. admin-
and specific drug-excipient interactions. Compositional variables have istration of the microemulsion formulations was performed by infusion
been shown to affect the in vitro release of drug from microemulsion into conscious rats over a 5-min period. Doses up to 0.5 ml/kg had no
formulations in a number of other cases [53,71]. significant effect on acid–base balance, blood gases, plasma electrolytes,
A pharmacokinetic study with the Sandimmune cyclosporin A arterial blood pressure or heart rate [68].
Neoral ® microemulsion concentrate exhibited the expected improve- The formulation of a water-in-HFA propellant microemulsion
ments in bioavailabilty and inter/intra-patient variability and was stabilised by fluorocarbon non-ionic surfactant and intended for
shown to facilitate the effective management of psoriaris [212]. pulmonary delivery has been described [16]. To date these workers
Very recently, pharmacokinetic trials on microemulsion formulations have only presented the phase behaviour of the systems together
for the oral administration of Neoral and a macrolide immunosup- with some light scattering results to prove the formation of a micro-
pressant SDZ RAD have been evaluated in non-human primates emulsion. As yet no data has been reported on the incorporation of
and were shown to have promising efficacy and tolerability [213]. drug into these systems.
Valsopdar, a P-glycoprotein modulator and an analogue of cyclospor- The development and characterisation of o/w microemulsions
ine, has also been tested as an oral SEDDS formulation both free and designed for ocular use has recently been reported [118]. The micro-
in gelatin soft-capsules in a clinical trial against an i.v. infusion of emulsions containing pilocarpine were formulated using lecithin,
the drug. Absolute bioavailability was 60% and the rate and extent propylene glycol and PEG 200 as cosurfactants, and IPM as the oil
of drug absorption comparable to that of the IV infusion [214,215]. phase. The formulations were low viscosity fluids with a refractive
The formulation and performance of SEDDS containing the anti- index lending themselves to ophthamological application. The test
microemulsions were non-irritant in rabbit eyes or hen egg mem-
brane. A prolonged pharmacological effect was observed in vivo com-
pared to the drug administered as a simple aqueous solution. This
may have been related to increased bioavailability or enhanced reten-
tion or both. However, prolonged release was not observed in vitro
using a cellulose membrane as permeability barrier.

5.7. Topical delivery

A number of recent reports detail microemulsion formulations

designed for topical or transdermal application [218,219]. Both o/w
and a w/o microemulsions have been evaluated in a hairless mouse
model for the delivery of prostaglandin E1. The microemulsions
were based on oleic acid or Gelucire 44/14 as the oil phase and
were stabilised by a mixture of Labrasol (C8 and C10 polyglycolysed
glycerides) and Plurol Oleique CC 497 as surfactant [219]. Although
enhanced delivery rates were observed in the case of the o/w micro-
emulsion, the authors concluded that the penetration rates were
Fig. 14. The plasma concentration–time profiles of flurbiprofen after intravenous inadequate for practical use from either system. The use of lecithin/
administration of flurbiprofen solution, commercial product (Lipfen®, flurbiprofen
axetil-loaded emulsion) and flurbiprofen-loaded microemulsion (EO: lecithin: DSPE–PEG:
IPP/water microemulsion for the transdermal transport of indometh-
flurbiprofen=8:3:1:2) equivalent to 2.5 mg/kg as flurbiprofen to rats (n=5). Redrawn acin and diclofenac has also been reported. Fourier transform infra
from [122]. red (FTIR) spectroscopy and differential scanning calorimetry (DSC)
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 189

showed the IPP organogel had disrupted the lipid organisation in albumin from a sorbitan monooleate w/o organogel injected intramuscu-
human stratum corneum after a 1 day incubation [157]. larly to mice has been observed [100].
The transdermal delivery of the hydrophilic drug diphenhydra-
mine hydrochloride from a w/o microemulsion into excised human 5.8. New developments
has also been investigated. The formulation was based on combina-
tions of Tween 80 and Span 20 with IPM. However two additional Environmentally responsive drug delivery systems are an interest-
formulations were tested containing cholesterol and oleic acid, re- ing development and phase changes that occur after administration
spectively. Cholesterol increased drug penetration whereas oleic acid triggered by changes in temperature, pH or ionic strength can be par-
had no measurable effect, but the authors clearly demonstrated that ticularly useful. One example of such behaviour involves the phase
penetration characteristics can be modulated by compositional selec- transformation of a reverse micellar solution of lecithin in IPM to a
tion [153]. Compositional effects have also been investigated for the lamellar liquid crystal. In this case the transition was triggered by
skin permeation of felodipine, a calcium antagonist, from o/w micro- contact of the reverse micellar solution with a biological aqueous
emulsions stabilised by a surfactant mixture containing Tween 20 phase, resulting in the controlled release of the anti-inflammatory
and taurodeoxycholate. IPM was used as the oil phase with benzyl fenoprofen [222]. Very recently, similar behaviour has been exploited
alcohol as cosurfactant [90]. Characterisation of lecithin/alkanol/ by the use of thermosetting microemulsions as delivery systems for
dodecane/water systems containing lidocaine and designed for topical periodontal anaesthesia. In this case, a block copolymer liquid micro-
use has shown the presence of L1 (micelles) and L2 (reverse micelles) emulsion containing lidocaine and prilocaine was designed to form a
isotropic droplet phases and a bicontinuous phase is the mesophase gel after in vivo administration to the periodontal pocket [223].
between L1 and L2 [121]. A combination of drug delivery strategies Thermoresponsive polymeric block copolymer micelles based on poly
has been employed in a recent report describing the formulation (N-isopropylacrylamide) and poly(butylmethacrylate) and containing
of an inclusion complex of the anti-inflammatory piroxicam with adriamycin have also been reported [202]. The abovementioned fluid–
β-cyclodextrin in an o/w microemulsion. Again designed for topical gel transitions might be usefully exploited in an injectable fluid dosage
use, the microemulsion system contained IPM and was stabilised by form, which undergoes subcutaneous gelation after administration.
the cationic surfactant hexadecyltrimethylammonium bromide [154]. An interesting phase behaviour study has been conducted using
In a related study this time employing an anionic surfactant, the trans- mixtures of water, glyceryl monoleate and Pluronic F-127, the latter
dermal permeation of glyceryl trinitrate through mouse skin was found being a triblock PEP–PPO–PEO copolymer. A phase resembling a
to be enhanced by around a factor of around 10 by formulating the cubic bicontinuous microemulsion was identified which could be dis-
drug in AOT micelles and reverse micelles and w/o microemulsions. persed in water, and after microfluidisation yielded nanometre-sized
Irritation studies showed that the normal micelles caused moderate ‘cubosomes’. The authors suggested these ‘cubosomes’ might be expected
irritation but that AOT reverse micellar solutions caused little or no to have interesting properties as novel drug delivery vehicles [224].
erythema [149]. Another important development, which in the longer term should
An interesting application of gelatin microemulsion-based organogels simplify the process of excipient selection, is the use of artificial neu-
(MBGs) which exploits the presence of surfactant-stabilised conducting ral networks to predict microemulsion phase behaviour [196].
aqueous channels has been their use in the iontophoretic transdermal
delivery of a model hydrophilic drug. One of the structures proposed 5.9. Alternative surfactants
for these MBGs is given in Fig. 15. The MBGs were prepared using a vari-
ety of pharmaceutically acceptable surfactants and oils including Tween Another area of growing interest is the use of fluorinated surfactants
80 and IPM [91]. Novel sorbitan monostearate organogels have also for the stabilisation of microemulsion systems [225]. Fluorosurfactants
been prepared from vegetable oils and IPM. Prepared at elevated tem- are more surface-active than their hydrocarbon counterparts, the
peratures and then cooled, the surfactant self-assembles into inverse CMCs are typically two orders of magnitude lower, and importantly
vesicles and then rod-shaped tubules. The organogels are opaque and fluorosurfactants are less haemolytic [225]. The biomedical uses of
thermoreversible, and may have been suggested as novel delivery vehi- fluorinated materials has been recently reviewed [226,227], and the
cles for drugs and antigens [220]. The inclusion of a water component low toxicity of a number of fluorinated solvents noted. Specific applica-
allows the formulation of w/o organogels which in common with gelatin tions include blood substitutes such as Oxygent®, which is a low viscos-
MBGs, may have percolative electroconductive channels and can be used ity fluid emulsion comprising 60% fluorocarbon, perfluordecyl bromide
to solubilise hydrophilic drugs and vaccines as well as hydrophobic mate- as stabiliser, egg yolk lecithin as emulsifier and buffer [227]. Micro-
rials in the continuous oil phase [221]. Prolonged release of bovine serum emulsion formulations have also been described for use as a blood
substitute, which employed a fluorocarbon oil and was stabilised by
Montanox 80, a hydrogenated surfactant, which the authors claimed
was a biocompatible [228,229]. Pulmonary aerosols based on fluorocar-
bons have also been indicated. The toxicological and regulatory status of
fluorinated surfactants has not advanced to the point that they may be
regarded as safe excipients in pharmaceutical formulations. Neverthe-
less, it is encouraging to observe the increasing number of reports
describing the formation of microemulsion systems stabilised by fluori-
nated surfactants, or as surfactant mixtures with conventional
hydrocarbon-based surfactants. In this context, it is also interesting to
note that lipase and lipoxygenase have been successfully incorporated
into a water-in-supercritical CO2 microemulsions stabilised by a
fluorosurfactant with retention of catalytic activity [230]. The incorpo-
ration of peptides and proteins into microemulsion systems stabilised
by fluorosurfactants without loss of biological activity has therefore
been clearly demonstrated. Fluorinated liposomes have been prepared
using fluorocarbon–hydrocarbon diblocks [231,232], and the circula-
Fig. 15. Proposed microemulsion-based gel (MBG) structure based in small angle neu- tion time in mice of fluorinated phospholipid vesicles was prolonged
tron scattering [237]. three-fold in comparison to equivalent non-fluorinated vesicles [233].
190 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

The ability to use fluorinated amphiphiles in the formulation of emul- [17] H. Sagitani, S. Friberg, Microemulsion systems with a non-ionic cosurfactant,
J. Disper. Sci. Technol. 1 (1980) 151–164.
sions, gels, microemulsions and vesicles indicate they have considerable [18] D.W. Osborne, C.A. Middleton, R.L. Rogers, Alcohol-free microemulsions, J. Dis-
potential as drug delivery systems, although the toxicological and regu- persion Sci. Technol. 9 (1988) 415–423.
latory hurdles will need to be overcome. [19] D. Attwood, C. Mallon, C.J. Taylor, Phase studies of oil-in-water phospholipid
microemulsions, Int. J. Pharm. 84 (1992) R5–R8.
[20] R. Aboofazeli, N. Patel, M. Thomas, M.J. Lawrence, Investigations into the forma-
tion and characterisation of phospholipid microemulsions. IV. Pseudo-ternary
6. Conclusion phase diagrams of systems containing water–lecithin–alcohol and oil; the influ-
ence of oil, Int. J. Pharm. 125 (1995) 107–116.
To date microemulsions have been shown to be able to protect la- [21] R. Aboofazeli, C.B. Lawrence, S.R. Wicks, M.J. Lawrence, Investigations into the for-
mation and characterisation of phospholipid microemulsions. III. Pseudo-ternary
bile drug, control drug release, increase drug solubility, increase bio- phase diagrams of systems containing water–lecithin–isopropyl myristate and ei-
availability and reduce patient variability. Furthermore, it has ther an alkanoic acid, amine, alkanediol, polyethylene glycol alkyl ether or alcohol
proven possible to formulate preparations suitable for most routes as cosurfactant, Int. J. Pharm. 111 (1994) 63–72.
[22] D. Attwood, A.T. Florence, Surfactant Systems: Their Chemistry, Pharmacy and
of administration. There is still however a considerable amount of Biology, Chapman and Hall, London, 1983.
fundamental work characterising the physico-chemical behaviour of [23] H.L. Rosano, J.L. Cavello, D.L. Chang, J.H. Whittham, Microemulsions: a commen-
microemulsions that needs to be performed before they can live up tary on their preparation, J. Soc. Cosmet. Chem. 39 (1988) 201–209.
[24] P.A. Winsor, Hydrotropy, solubilisation and related emulsification processes,
to their potential as multipurpose drug delivery vehicles.
J. Chem. Soc. Faraday Trans. 44 (1) (1948) 376–398.
It is of interest to note that the first study investigating the use of [25] K. Shinoda, M. Araki, A. Sadaghiani, A. Khan, B. Lindman, Lecithin-based
microemulsions as potential drug delivery vehicles was reported in microemulsions: phase behaviour and microstructure, J. Phys. Chem. 95 (1991)
1974 by Attwood et al. [234]. However the field lay virtually dormant 989–993.
[26] R. Aboofazeli, M.J. Lawrence, Investigations into the formation and characteriza-
until a review on the subject was published by Bhargava et al. in 1987 tion of phospholipid microemulsions: I Pseudo-ternary phase diagrams of sys-
[43]. Since then there has been a very gradual increase in the number tems containing water–lecithin–alcohol–isopropyl myristate, Int. J. Pharm. 93
of research papers published on the topic until 1994 since when (1993) 161–175.
[27] G.D. Rees, B.H. Robinson, Esterification reactions catalysed by Chromobacterium
about 20 papers detailing the pharmaceutical use of microemulsions viscosum lipase in CTAB-based microemulsion systems, Biotechnol. Bioeng. 45
have been published each year. This small number of papers contrasts (1995) 344–355.
very sharply with liposomes where the number of publications deal- [28] M. Giustini, G. Palazzo, G. Colafemmina, M. Della Monica, M. Giomini, A. Ceglie,
Microstructure and dynamics of the water-in-oil CTAB/n-pentanol/n-hexane/
ing specifically with their use as drug delivery vehicles runs into the water microemulsion: spectroscopic and conductivity study, J. Phys. Chem.
order of 300 per year. Interestingly liposomes were first proposed as 100 (1996) 3190–3198.
a delivery system in 1972, just a few years prior to Attwood and co- [29] S.K. Mehta, X.X. Kawaljit, Isentropic compressibility and transport properties of
CTAB-alkanol-hydrocarbon-water microemulsion systems, Colloids Surfaces A:
workers paper. This lack of research in the field does not mean that Physicochem. Eng. Aspects 136 (1998) 35–41.
that microemulsions offer any less potential as delivery systems [30] I.S. Barnes, S.T. Hyde, B.W. Ninham, P.-J. Derian, M. Drifford, G.G. Warr, T.N.
than liposomes, indeed it is pertinent to note that it took considerably Zemb, The disordered open connected model of microemulsions, Progr. Colloid
Polym. Sci. 76 (1988) 90–95.
less time for a microemulsion product (i.e. Neroal ®) to get onto the
[31] T. Skodvin, J. Sjöblem, J.O. Saeten, B. Gestblom, Solubilisation of drugs in
market than the first liposomal drug delivery system. microemulsions as studied by dielectric spectroscopy, J. Colloid Interface Sci.
155 (1993) 392–401.
[32] M. Olla, M. Monduzzi, L. Ambrosone, Microemulsions and emulsions in
References DDAB/water/oil systems, Colloids Surfaces A: Physicochem. Eng. Aspects 160
(1999) 23–36.
[1] T.P. Hoar, J.H. Schulman, Transparent water-in-oil dispersions: the oleopathic [33] K.A. Johnston, D.O. Shah, Effect of oil chain length and electrolytes on water
hydro-micelle, Nature 152 (1943) 102–103. solubilisation in alcohol-free pharmaceutical microemulsions, J. Colloid Interface
[2] J.H. Schulman, W. Stoeckenius, L.M. Prince, Mechanism of formation and struc- Sci. 107 (1985) 269–271.
ture of micro emulsions by electron microscopy, J. Phys. Chem. 63 (1959) [34] M. Trotta, M.R. Gasco, F. Pattarino, Diffusion of steroid hormones from o/w
1677–1680. microemulsions: influence of the cosurfactant, Acta Pharm. Technol. 36 (1990)
[3] I. Danielsson, B. Lindman, The definition of a microemulsion, Colloids and 226–231.
Surfaces 3 (1981) 391–392. [35] J. Bergenholtz, A.A. Romagnoli, N.J. Wagner, Viscosity, microstructure and
[4] K. Shinoda, B. Lindman, Organised surfactant systems: microemulsions, Langmuir interparticle potential of AOT/H2O/n-decane inverse microemulsions, Langmuir
3 (1987) 135–149. 11 (1995) 1559–1570.
[5] L.M. Prince, A theory of aqueous emulsion. I. Negative interfacial tension at the [36] E. Acosta, D.H. Kurlat, M. Bisceglia, B. Ginzberg, L. Baikauskas, S.D. Romano, In-
oil/water interface, J. Colloid Interface Sci. 23 (1967) 165–173. duced electric birefringence and viscosity studies in microemulsions, Colloids
[6] K. Shinoda, H. Kunieda, Conditions to produce so-called microemulsions. Factors Surfaces A: Physicochem. Eng. Aspects 106 (1996) 11–21.
to increase the mutual solubility of oil and water by solubilizer, J. Colloid Inter- [37] M. D’Angelo, D. Fioretto, G. Onori, L. Palmieri, A. Santucci, Dynamics of water-
face Sci. 42 (1973) 381–387. containing sodium bis(2-ethylhexyl)sulfosuccinate (AOT) reverse micelles: a
[7] K. Shinoda, S. Friberg, Microemulsions. Colloidal aspects. Adv. Colloid Interface high-frequency dielectric study, Phys. Rev. E 54 (1996) 993–996.
Sci. 4 (1975) 281–300. [38] J. Carlfors, I. Blute, V. Schmidt, Lidocaine in microemulsion — a dermal delivery
[8] S. Friberg, I. Burasczenska, Microemulsions in the water–potassium oleate– system, J. Disp. Sci. Technol. 12 (1991) 467–482.
benzene system, Progr. Colloid Polymer Sci. 63 (1978) 1–9. [39] J.N. Israelachvilli, D.J. Mitchell, B.W. Ninham, Theory of self assembly of hydro-
[9] E. Ruckenstein, J.C. Chi, Stability of microemulsions, J. Chem. Soc. Faraday Trans. carbon amphiphiles into micelles and bilayers, J. Chem. Soc. Faraday Trans. II
71 (1975) 1690–1707. 72 (1976) 1525–1567.
[10] J.Th.G. Overbeek, Microemulsions, a field at the border between lyophobic and [40] D.J. Mitchell, B.W. Ninham, Micelles, vesicles and microemulsions, J. Chem. Soc.
lyophilic colloids, Faraday Disc. Chem. Soc. 65 (1978) 7–19. Faraday. Trans. II 77 (1981) 601–629.
[11] E. Ruckenstein, R. Krishnan, Effect of electrolytes and mixtures of surfactants on [41] B. Jonsson, B. Lindman, K. Holmberg, B. Kronberg, Behaviour of concentrated
the oil–water interfacial tension and their role in formation of microemulsions, surfactant systems, in: Surfactants and Polymers in Aqueous Solution, John
J. Colloid Interface Sci. 76 (1980) 201–211. Wiley & Sons, Chichester, 1999, pp. 61–89.
[12] T.F. Tadros, Microemulsions — an overview, in: K.L. Mittal, B. Lindman (Eds.), [42] W. Warisnoicharoen, A.B. Lansley, M.J. Lawrence, Nonionic oil-in-water
Proc. 4th Int. Symp. Surfactants in Solution, Plenum, New York, 1984, microemulsions: the effect of oil type on phase behaviour, Int. J. Pharm. 198
pp. 1501–1532. (2000) 7–27.
[13] J. Eastoe, B.M.H. Cazelles, D.C. Steytler, J.D. Holmes, A.R. Pitt, T.J. Wear, R.K. [43] H.N. Bhargava, A. Narurkar, L.M. Lieb, Using microemulsions for drug delivery,
Heenan, Water-in-CO2 microemulsions studied by small-angle neutron scatter- Pharm. Tech. 11 (1987) 46–52.
ing, Langmuir 13 (1997) 6980–6984. [44] D. Attwood, Microemulsions, in: J. Kreuter (Ed.), Colloidal Drug Delivery Systems,
[14] Y. Ikushima, N. Saito, K. Hatakeda, M. Arai, Water/AOT/ethane microemulsion Dekker, New York, 1994, pp. 31–71.
under supercritical conditions as a reaction medium, J. Supercrit. Fluids 13 [45] J. Eccleston, Microemulsions, in: J. Swarbrick, J.C. Boylan (Eds.), Encyclopedia
(1998) 217–224. of Pharmaceutical Technology, Vol. 9, Marcel Dekker, New York, 1994,
[15] R.M. Evans, S.J. Farr, The development of novel, pressurised aerosols formulated pp. 375–421.
as solutions, J. Biopharm. Sci. 3 (1992) 33–40. [46] M.J. Lawrence, Surfactant systems: microemulsions and vesicles as vehicles for
[16] N. Patel, M. Marlow, M.J. Lawrence, Microemulsions: a novel pMD1 formulation, drug delivery, Eur. J. Drug Metab. Pharmacokinet. 3 (1994) 257–269.
in: Drug Delivery to the Lungs IX, LondonThe Aerosol Society, Bristol, 1998, [47] M.J. Lawrence, Microemulsions as drug delivery vehicles, Curr. Opin. Colloid In-
pp. 160–163. terface Sci. 1 (1996) 826–832.
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 191

[48] S. Tenjarla, Microemulsions: an overview and pharmaceutical applications, Crit. [79] T.N. Zemb, I.S. Barnes, P.-J. Derian, B.W. Ninham, Scattering as a critical test
Rev. Therapeutic Drug Carrier Systems 16 (1999) 461–521. of microemulsion structural models, Progr. Colloid Polym. Sci. 81 (1990)
[49] W. Warisnoicharoen, A.B. Lansley, M.J. Lawrence, Light scattering investigations 20–29.
on dilute non-ionic oil-in-water microemulsions, AAPS Pharm. Sci. 2 (2000) [80] M. Hirai, R.K. Hirai, H. Iwase, S. Arai, S. Mitsuya, T. Takeda, H. Seto, M. Nagao, Dy-
429–448. namics of w/o AOT microemulsions studied by neutron spin echo, J. Phys. Chem.
[50] F.C. Larche, P. Delord, Structures and stability of isotropic phases in the AOT– Solids 60 (1999) 1359–1361.
decane–water system, Fluid Phase Equilib. 20 (1985) 257–264. [81] E.W. Kaler, S. Prager, A model of dynamic scattering by microemulsions, J. Col-
[51] J. Eastoe, G. Fragneto, B.H. Robinson, T.F. Towey, R.K. Heenan, F.J. Lena, Variation of loid Interface Sci. 86 (1982) 359–369.
surfactant counterion and its effect on the structure and properties of Aerosol– [82] J. Eastoe, K.J. Hetherington, D. Sharpe, J. Dong, R.K. Heenan, D. Steytler, Mixing of
OT-based water-in-oil microemulsions, J. Chem. Soc. Faraday Trans. 88 (1992) alkanes with surfactant monolayers in microemulsions, Langmuir 12 (1996)
461–467. 3876–3880.
[52] P. Schurtenberger, Q. Peng, M.E. Leser, P.L. Luisi, Structure and phase behaviour [83] T. Gulik-Krzywicki, K. Larsson, An electron microscopy study of the L2-phase
of lecithin-based microemulsions: a study of chain length dependence, J. Colloid (microemulsion) in a ternary system: triglyceride/monoglyceride/water, Chem.
Interface Sci. 156 (1993) 43–51. Phys. Lipids 35 (1984) 127–132.
[53] P. Khoshnevis, S.A. Mortazavi, M.J. Lawrence, R. Aboofazeli, In-vitro release of [84] P.K. Vinson, J.G. Sheehan, W.G. Miller, L.E. Scriven, H.T. Davis, Viewing
sodium salicylate from water-in-oil phospholipid microemulsions, J. Pharm. microemulsions with freeze-fracture transmission electron microscopy, J. Phys.
Pharmacol. 49 (S4) (1997) 47. Chem. 95 (1991) 2546–2550.
[54] S.K. Mehta, R.K. Dewan, K. Bala, Percolation phenomenon and the study of con- [85] M.A. Bolzinger, T.C. Carduner, M.C. Poelman, Bicontinuous sucrose ester
ductivity, viscosity and ultrasonic velocity in microemulsions, Phys. Rev. E 50 microemulsion: a new vehicle for topical delivery of niflamic acid, Int. J. Pharm. 176
(1994) 4759–4762. (1998) 39–45.
[55] Z.J. Yu, R.D. Neuman, Reversed micellar solution-to-bicontinuous microemulsion [86] D. Attwood, C. Mallon, G. Ktistis, C.J. Taylor, A study on factors influencing the
transition in sodium bis(2-ethylhexyl) phosphate/n-heptane/water system, Lang- droplet size in nonionic oil-in-water microemulsions, Int. J. Pharm. 88 (1992)
muir 11 (1995) 1081–1086. 417–422.
[56] R. Angelico, G. Palazzo, G. Colafemmina, P.A. Cirkel, M. Giustini, A. Ceglie, Water [87] R. Aboofazeli, D.J. Barlow, M.J. Lawrence, Particle size analysis of concentrated
diffusion and head group mobility in polymer-like reverse micelles: evidence of phospholipid microemulsions. I. Total intensity light scattering, AAPS Pharm.
a sphere-to-rod-to-sphere transition, J. Phys. Chem. B 102 (1998) 2883–2889. Sci. 2 (2000) 449–470.
[57] S.K. Mehta, X.X. Kavaljit, K. Bala, Phase behaviour, structural effects, volumetric [88] J.M. Sarciaux, L. Acar, P.A. Sado, Using microemulsion formulations for drug de-
and transport properties in nonaqueous microemulsions, Phys. Rev. E 59 (1999) livery of therapeutic peptides, Int. J. Pharm. 120 (1995) 127–136.
4317–4325. [89] P.P. Constantinides, Lipid microemulsions for improving drug dissolution and
[58] Y. Feldman, N. Kozlovich, I. Nir, N. Garti, Dielectric spectroscopy of microemulsions, oral absorption: physical and biopharmaceutical aspects, Pharm. Res. 12
Colloids Surfaces A: Physicochem. Eng. Aspects 128 (1997) 47–61. (1995) 1561–1572.
[59] P.A. Cirkel, J.P.M. van der Ploeg, G.J.M. Koper, Branching and percolation in lecithin [90] M. Trotta, S. Morel, M.R. Gasco, Effect of oil phase composition on the skin perme-
wormlike micelles studied by dielectric spectroscopy, Phys. Rev. E 57 (1998) ation of felodipine from (o/w) microemulsions, Pharmazie 52 (1997) 50–53.
6875–6883. [91] S. Kantaria, G.D. Rees, M.J. Lawrence, Gelatin-stabilised microemulsion-based
[60] O. Regev, S. Ezrahi, A. Aserin, N. Garti, E. Wachtel, E.W. Kaler, A. Khan, Y. Talmon, organogels: rheology and application in iontophoretic transdermal drug deliv-
A study of the microstructure of a four-component nonionic microemulsion by ery, J. Control. Rel. 60 (1999) 355–365.
cryo-TEM, NMR, SAXS and SANS, Langmuir 12 (1996) 668–674. [92] C. Malcolmson, C. Satra, S. Kantaria, A. Sidhu, M.J. Lawrence, Effect of oil on the
[61] N. Nakamura, Y. Yamaguchi, B. Hakansson, U. Olsson, T. Tagawa, H. Kunieda, For- level of solubilization of testosterone propionate into nonionic oil-in-water
mation of microemulsion and liquid crystal in biocompatible sucrose alkanoate microemulsions, J. Pharm. Sci. 87 (1998) 109–116.
systems, J. Disp. Sci. Technol. 20 (1999) 535–557. [93] V.H.L. Lee, Enzymatic barriers to peptide and protein absorption, Crit. Rev. Ther.
[62] M. Kahlweit, R. Strey, D. Haase, H. Kunieda, T. Schmeling, B. Faulhaber, M. Borkovec, Drug Carrier Syst. 5 (1988) 69–97.
H.-F. Eicke, G. Busse, F. Eggers, T.H. Funck, H. Richmann, L. Magid, O. Soderman, P. [94] E.C. Swenson, W.J. Curatolo, Intestinal permeability enhancement for proteins,
Stilbs, J. Winkler, A. Dittrich, W. Jahn, How to study microemulsions, J. Colloid Inter- peptides and other polar drugs: mechanisms and potential toxicity, Adv. Drug
face Sci. 118 (1987) 436–453. Deliv. Rev. 8 (1992) 39–92.
[63] J. Alander, T. Warnheim, Model microemulsions containing vegetable oils Part 2: [95] E.S. Swenson, W.B. Milisen, W. Curatolo, Intestinal permeability enhancement:
Ionic surfactant systems, JAOCS 66 (1989) 1661–1665. efficacy, acute local toxicity and reversibility, Pharm. Res. 11 (1994) 1132–1142.
[64] W.O. Parker Jr., C. Genova, G. Carignano, Study of micellar solutions and [96] P.P. Constantinides, J.-P. Scalart, S. Lancaster, J. Marcello, G. Marks, H. Ellens, P.L.
microemulsions of an alkyl oligoglucoside via NMR spectroscopy, Colloids Surfaces Smith, Formulation and intestinal absorption enhancement evaluation of
A: Physicochem. Eng. Aspects 72 (1993) 275–284. water-in-oil microemulsions incorporating medium-chain glycerides, Pharm.
[65] C. von Corswant, S. Engström, O. Söderman, Microemulsions based on soybean Res. 11 (1994) 1385–1390.
phosphatidylcholine and triglycerides. Phase behaviour and microstructure, [97] P.P. Constantinides, C.M. Lancaster, J. Marcello, D.C. Chiossone, D. Orner, I.
Langmuir 13 (1997) 5061–5070. Hidalgo, P.L. Smith, A.B. Sarkahian, S.H. Yiv, A.J. Owen, Enhanced intestinal ab-
[66] C. von Corswant, O. Soderman, Effect of adding isopropyl myristate to sorption of an RGD peptide from water-in-oil microemulsions of different com-
microemulsions based on soybean phosphatidylcholine and triglycerides, Lang- position and particle size, J. Control. Rel. 34 (1995) 109–116.
muir 14 (1998) 3506–3511. [98] C.W. Pouton, Formulation of self-emulsifying drug delivery systems, Adv. Drug.
[67] C. von Corswant, C. Olsson, O. Soderman, Microemulsions based on soybean Del. Rev. 25 (1997) 47–58.
phosphatidylcholine and isopropyl myristate — effect of addition of hydrophilic [99] M.R. Gasco, F. Pattarino, F. Lattanzi, Long-acting delivery systems for peptides:
surfactants, Langmuir 14 (1998) 6864–6870. reduced plasma testosterone levels in male rats after a single injection, Int. J.
[68] C. von Corswant, P. Thoren, S. Engstrom, Triglyceride-based microemulsion from Pharm. 62 (1990) 119–123.
intravenous administration of sparingly soluble substances, J. Pharm. Sci. 87 [100] H. Willimann, P. Walde, P.L. Luisi, A. Gazzaniga, F. Stroppolo, Lecithin organogels
(1998) 200–208. as matrix for transdermal transport of drugs, J. Pharm. Sci. 81 (1992) 871–874.
[69] H. Saint Ruth, D. Attwood, G. Ktistis, C.J. Taylor, Phase studies and particle size [101] S. Murdan, B. van den Bergh, G. Gregoriadis, A.T. Florence, Water-in-sorbitan
analysis of oil-in-water phospholipid microemulsions, Int. J. Pharm. 116 monostearate organogels (water-in-oil gels), J. Pharm. Sci. 88 (1999) 615–619.
(1995) 253–261. [102] M. Kahlweit, G. Busse, B. Faulhaber, Preparing microemulsions with alkyl
[70] A. Shioi, M. Harada, M. Tanabe, Static light scattering from oil-rich microemulsions monoglucosides and the role of n-alkanols, Langmuir 11 (1995) 3382–3387.
containing polydispersed cylindrical aggregates in sodium bis(2-ethylhexyl) phos- [103] R. Duro, C. Souto, J.L. Gomez-Amoza, R. Martinez-Pacheco, A. Concheiro, Interfa-
phate system, J. Phys. Chem. 99 (1995) 4750–4756. cial adsorption of polymers and surfactants: implications for the properties of
[71] G. Ktistis, Effect of polysorbate 80 and sorbitol concentration on in vitro release disperse systems of pharmaceutical interest, Drug Devel. Ind. Pharm. 25
of indomethacin from microemulsions, J. Disp. Sci. Technol. 18 (1997) 49–61. (1999) 817–829.
[72] C. Eugster, G. Rivara, G. Forni, S. Vai, Marigenol®-concentrates comprising Taxol [104] A.G. Floyd, Top ten considerations in the development of parenteral emulsions,
and/or Taxan esters as active substances, Panminerva Medica 38 (1996) 234–242. Pharm. Sci. Technol. Today 2 (1999) 134–143.
[73] P.P. Constantinides, J.P. Scalart, Formulation and physical characterisation of [105] Y.W. Cho, M. Flynn, Oral delivery of insulin, Lancet 2 (1989) 1518–1519.
water-in-oil microemulsions containing long versus medium chain length glycer- [106] B. Fubini, M.R. Gasco, M. Gallarate, Microcalorimetric study of microemulsions
ides, Int. J. Pharm. 158 (1997) 57–68. as potential drug delivery systems. II. Evaluation of enthalpy in the presence
[74] D. Hantzschel, S. Enders, H. Kahl, K. Quitzsch, Phase behaviour of quaternary sys- of drugs, Int. J. Pharm. 50 (1989) 213–217.
tems containing carbohydrate surfactants–water–oil–cosurfactant, Phys. Chem. [107] M. Gallarate, M.R. Gasco, M. Trotta, P. Chetoni, M.F. Saettone, Preparation and
Chem. Phys. 1 (1999) 5703–5710. evaluation in vitro of solutions and o/w microemulsions containing levobunolol
[75] M.A. Bolzinger, M.A. Thevenin, J.L. Grossiord, M.C. Poelman, Characterisation of a as ion pair, Int. J. Pharm. 100 (1993) 219–225.
sucrose ester microemulsion by freeze fracture electron micrograph and small [108] F. Pattarino, E. Marengo, M.R. Gasco, R. Carpignano, Experimental design and
angle neutron scattering experiments, Langmuir 15 (1999) 2307–2315. partial least squares in the study of complex mixtures: microemulsions as
[76] L. Auvray, J.P. Cotton, R. Ober, C. Taupin, Evidence for zero mean curvature drug carriers, Int. J. Pharm. 91 (1993) 157–165.
microemulsions, J. Phys. Chem. 88 (1984) 4586–4589. [109] J.L. Murtha, H.Y. Ando, Synthesis of the cholesteryl prodrugs cholesteryl ibuprofen and
[77] J. Tabony, Formation of cubic structures in microemulsions containing equal vol- cholesteryl flufenamate and their formulation into phospholipid microemulsions,
umes of water and oil, Nature 319 (1986) 400. J. Pharm. Sci. 83 (1994) 1222–1228.
[78] J. Tabony, Occurrence of liquid-crystalline mesophases in microemulsion disper- [110] P. Schurtenberger, C. Cavaco, Polymer-like lecithin reverse micelles. 1. A light
sions, Nature 320 (1986) 338–340. scattering study, Langmuir 10 (1994) 100–108.
192 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193

[111] P.P. Constantinides, S.H. Yiv, Particle size determination of phase-inverted [143] N. Garti, V. Clement, M. Leser, A. Aserin, M. Fanun, Sucrose ester microemulsions,
water-in-oil microemulsions under different dilution and storage conditions, J. Molec. Liq. 80 (1999) 253–296.
Int. J. Pharm. 115 (1995) 225–234. [144] N. Garti, A. Aserin, M. Fanun, Non-ionic sucrose ester microemulsions for food
[112] K. Kriwet, C.C. Müller-Goymann, Diclofenac release from phospholipid drug systems applications. Part 1. Water solubilisation, Colloids Surfaces A: Physicochem.
and permeation through excised human stratum corneum, Int. J. Pharm. 125 Eng. Aspects 164 (2000) 27–38.
(1995) 231–242. [145] H.-O. Ho, C.C. Hsiao, M.-T. Sheu, Preparation of microemulsions using polyglycerol
[113] M.-J. Li, M.-H. Lee, C.-K. Shim, Inverse targeting of drugs to reticuloendothelial fatty acid esters as surfactant for the delivery of protein drugs, J. Pharm. Sci. 85
system-rich organs by lipid microemulsions emulsified by poloxamer 338, Int. (1996) 138–143.
J. Pharm. 113 (1995) 175–187. [146] P.P. Constantinides, G. Welzel, H. Ellens, P.L. Smith, S. Sturgis, S.H. Yiv, A.B. Owen,
[114] C. Satra, M. Thomas, M.J. Lawrence, The solubility of testosterone propionate in Water-in-oil microemulsions containing medium-chain fatty acids/salts: formulation
oil-in-water microemulsions, J. Pharm. Pharmacol. 47 (1995) 1126. and intestinal absorption enhancement evaluation, Pharm. Res. 13 (1996) 210–215.
[115] M. Trotta, E. Ugazio, M.R. Gasco, Pseudo-ternary phase diagrams of lecithin-based [147] K.R. Wormuth, E.W. Kaler, Amines as microemulsion cosurfactants, J. Phys.
microemulsions: influence of monoalkylphosphates, J. Pharm. Pharmacol. 47 Chem. 91 (1987) 611–617.
(1995) 451–454. [148] R. Joubran, N. Parris, D. Lu, S. Trevino, Synergetic effect of sucrose and ethanol on
[116] M.E. Leser, W.C. van Evert, W.G.M. Agterof, Phase behaviour of lecithin–water– formation of triglyceride microemulsions, J. Disp. Sci. Technol. 15 (1994)
alcohol–triacylglycerol mixtures, Colloids Surfaces A: Physicochem. Eng. Aspects 687–704.
116 (1996) 293–308. [149] M. Varshney, T. Khanna, M. Changez, Effects of AOT micellar systems on the trans-
[117] M. Trotta, R. Cavalli, E. Ugazio, M.R. Gasco, Phase behaviour of microemulsion dermal permeation of glyceryl trinitrate, Colloids and Surface B: Biointerfaces 13
systems containing lecithin and lysolecithin as surfactants, Int. J. Pharm. 143 (1999) 1–11.
(1996) 67–73. [150] M.J. Garcia-Celma, N. Azemar, M.A. Pes, C. Solans, Solubilisation of antifungal
[118] A. Hasse, S. Keipert, Development and characterisation of microemulsions for drugs in water/POE(20) sorbitan monooleate/oil systems, Int. J. Pharm. 105
ocular application, Eur. J. Pharm. Biopharm. 43 (1997) 179–183. (1994) 77–81.
[119] M. Kahlweit, G. Busse, B. Faulhaber, Preparing nontoxic microemulsions. 2, [151] C. Nastruzzi, R. Gambari, Antitumor activity of (trans)dermally delivered aro-
Langmuir 13 (1997) 5249–5251. matic tetra-amidines, J. Control. Rel. 29 (1994) 53–62.
[120] M. Trotta, F. Pattarino, G. Grosa, Formation of lecithin-based microemulsions [152] G. Ktistis, I. Niopas, A study on the in-vitro percutaneous absorption of propran-
containing n-alkanol phosphocholines, Int. J. Pharm. 174 (1998) 253–259. olol from disperse systems, J. Pharm. Pharmacol. 50 (1998) 413–418.
[121] S.Y. Choi, S.G. Oh, S.Y. Bae, S.K. Moon, Effect of short-chain alcohols as cosurfactants [153] U. Schmalfuß, R. Neubert, W. Wohlrab, Modification of drug penetration into
on pseudo-ternary phase diagrams containing lecithin, Korean J. Chem. Eng. 16 human skin using microemulsions, J. Control. Rel. 46 (1997) 279–285.
(1999) 377–381. [154] M.E.A. Dalmora, A.G. Oliveira, Inclusion complex of piroxicam with beta-cyclodextrin
[122] K.M. Park, M.K. Lee, K.J. Hwang, C.K. Kim, Phospholipid-based microemulsions of and incorporation in hexadecyltrimethylammonium bromide based microemulsion,
flurbiprofen by the spontaneous emulsification process, Int. J. Pharm. 183 Int. J. Pharm. 184 (1999) 157–164.
(1999) 145–154. [155] J. Kemken, A. Ziegler, B.W. Müller, Investigations into the pharmacodynamic ef-
[123] M. Trotta, Influence of phase transformation on indomethacin release from fects of dermally administered microemulsions containing β-blockers, J. Pharm.
microemulsions, J. Control. Rel. 60 (1999) 399–405. Pharmacol. 43 (1991) 679–684.
[124] M. Trotta, M. Gallarate, F. Pattarino, M.E. Carlotti, Investigation of the phase be- [156] F. Dreher, P. Walde, P.L. Luisi, P. Elsner, Human skin irritation of a soybean leci-
haviour of systems containing lecithin and 2-acyl lysolecithin derivatives, Int. thin microemulsion gel and of liposomes, Proc. Int. Symp. Control. Rel. Bioact.
J. Pharm. 190 (1999) 83–89. Mater. 22 (1995) 640–641.
[125] C. von Corsant, P.E.G. Thoren, Solubilization of sparingly soluble active compounds in [157] F. Dreher, P. Walde, P. Walther, E. Wehrli, Interaction of a lecithin
lecithin-based microemulsions: influence on phase behavior and microstructure, microemulsion gel with human stratum corneum and its effect on transdermal
Langmuir 15 (1999) 3710–3717. transport, J. Control. Rel. 45 (1997) 131–140.
[126] A.H. Kibbe (Ed.), Handbook of Pharmaceutical Excipients, 3rd Edition, Phar- [158] W.A. Ritschel, Microemulsions for improved peptide absorption from the
maceutical Press, London, 2000, pp. 407–423. gastrointestinal tract, Meth. Find Exp. Clin. Pharmacol. 13 (1991) 205–220.
[127] C. Malcolmson, M.J. Lawrence, A comparison of the incorporation of model ste- [159] K. Kriwet, C.C. Müller-Goyman, Mutual interactions between diclofenac
roids into non-ionic micellar and microemulsion systems, J. Pharm. Pharmacol. diethylamine and phospholipids — investigations on the microstructure of the
45 (1993) 141–143. arisen systems, Pharmazie 49 (1994) 187–191.
[128] C. Malcolmson, M.J. Lawrence, Three-component non-ionic oil-in-water [160] I. Siebenbrodt, S. Keipert, Poloxamer-systems as potential ophthalmics II.
microemulsions using polyethylene ether surfactants, Colloids and surfaces B: Microemulsions, Eur. J. Pharm. Biopharm. 39 (1993) 25–30.
Biointerfaces 4 (1995) 97–109. [161] F. Pattarino, M.E. Carlotti, M.R. Gasco, Topical delivery systems for azelaic acid: effect
[129] J. Alander, T. Warnheim, Model microemulsions containing vegetable oils Part 1: of the suspended drug in microemulsion, Pharmazie 49 (1994) 72–73.
Non-ionic surfactant systems, JAOCS 66 (1989) 1656–1660. [162] M. Zabka, M. Benkova, Microemulsions containing local anaesthetics. Part 6: In-
[130] N.J. Kale, L.V. Allen Jr., Studies on microemulsions using Brij 96 as surfactant and fluence of microemulsion vehicle on in vivo effect of pentacaine, Pharmazie 50
glycerin, ethylene glycol and propylene glycol as cosurfactants, Int. J. Pharm. 57 (1995) 703–704.
(1989) 87–93. [163] M.R. Gasco, R. Cavalli, M.E. Carlotti, Timolol in lipospheres, Pharmazie 47 (1992)
[131] H. Kunieda, N. Ushio, A. Nakano, M. Miura, Three-phase behaviour in a mixed 119–121.
sucrose alkanoate and polyethyleneglycol alkyl ether system, J. Colloid Interface [164] S. Morel, M.R. Gasco, R. Cavalli, Incorporation in lipospheres of [D-Trp-6] LHRH,
Sci. 159 (1993) 37–44. Int. J. Pharm. 105 (1994) R1–R3.
[132] W. Warisnoicharoen, A.H. Lansley, M.J. Lawrence, The cytotoxicity of nonionic [165] R. Cavalli, D. Aquilano, M.E. Carlotti, M.R. Gasco, Study by X-ray powder diffrac-
oil-in-water microemulsions to a human cell line, in: Drug Delivery To the tion and differential scanning calorimetry of two model drugs, phenothiazine
Lungs, Vol. VI, The Aerosol Society, 1995, pp. 40–43. and nifedipine, incorporated into lipid nanoparticles, Eur. J. Pharm. Biopharm.
[133] K.-H. Oh, J.R. Baran, W.H. Wade, V. Weerasooriya, Temperature insensitive 41 (1995a) 329–333.
microemulsion phase behaviour with nonionic surfactants, J. Disper. Sci. [166] R. Cavalli, S. Morel, M.R. Gasco, P. Chetoni, M.F. Saettone, Preparation and eval-
Technol. 16 (1995) 165–188. uation in vitro of colloidal lipospheres containing pilocarpine as ion pair, Int.
[134] S. Tolle, T. Zuberi, M.J. Lawrence, Physicochemical and solubilisation properties J. Pharm. 117 (1995b) 243–246.
of N,N-dimethyl-N-(3-dodecyloxypropyl)-amineoxide: a biodegradable nonion- [167] R. Cavalli, O. Caputo, M.E. Carlotti, M. Trotta, C. Scarnecchia, M.R. Gasco,
ic surfactant, J. Pharm. Sci. 89 (2000) 798–806. Sterilisation and freeze-drying of drug-free and drug-loaded solid lipid
[135] N.B. Desai, Esters of sucrose and glucose as cosmetic materials, Cosmetics and nanoparticles, Int. J. Pharm. 148 (1997) 47–54.
Toiletries 105 (2) (1990) 99–107. [168] R. Cavalli, O. Caputo, E. Marengo, F. Pattarino, M.R. Gasco, The effect of the com-
[136] M. Kahlweit, G. Busse, B. Faulhaber, Preparing nontoxic microemulsions with ponents of microemulsions on both size and crystalline structure of solid lipid
alkyl monoglucosides and the role of alkanediols as cosolvents, Langmuir 12 nanoparticles (SLN) containing a series of model molecules, Pharmazie 53
(1996) 861–862. (1998) 392–396.
[137] L.D. Ryan, E.W. Kaler, Role of oxygenated oils in n-alkyl β-D-monoglucoside [169] S. Morel, E. Ugazio, R. Cavalli, M.R. Gasco, Thymopentin in solid lipid
microemulsions phase behaviour, Langmuir 13 (1997) 5222–5228. nanoparticles, Int. J. Pharm. 132 (1996) 259–261.
[138] L.D. Ryan, K.-V. Schubert, E.W. Kaler, Phase behaviour of microemulsions made [170] R. Carpignano, M.R. Gasco, S. Morel, Optimisation of doxorubicine incorporation
with n-alkyl monoglucosides and n-alkyl polyglycol ethers, Langmuir 13 and of the yield of polybutylcyanoacrylate nanoparticles, Pharm. Acta Helv. 66
(1997) 1510–1518. (1991) 28–32.
[139] L.D. Ryan, E.W. Kaler, Effect of alkyl sulphates on the phase behaviour and [171] M.R. Gasco, S. Morel, M. Trotta, I. Viano, Doxorubicine englobed in
microstructure of alkyl polyglucoside microemulsions, J. Phys. Chem. B 102 polybutylcyanoacrylate nanocapsules: behaviour in vitro and in vivo, Pharm.
(1998) 7549–7556. Acta Helv. 66 (1991) 47–49.
[140] C. Stubenrauch, B. Paeplow, G.H. Findenegg, Microemulsions supported by octyl [172] T.H. El-Faham, S.M. El-Shanawany, M.G. Abdel-Mohesen, In vitro and in vivo
monoglucoside and geraniol. 1. The role of alcohol in the interfacial layer, Lang- evaluation of a transparent oil–water gel formulation of flufenamic acid, Eur.
muir 13 (1997) 3652–3658. J. Pharm. Biopharm. 38 (1992) 180–185.
[141] S. Keipert, G. Schulz, Microemulsions with sucrose fatty ester surfactants. Part 1. [173] S. Bhatnagar, S.P. Vyas, Organogel-based system for transdermal delivery of
In-vitro characterisation, Pharmazie 49 (1994) 195–197. propanolol, J. Microencapsulation 11 (1994) 431–438.
[142] M.A. Thevenin, J.L. Grossiord, M.C. Poelman, Sucrose esters/cosurfactant [174] F.P. Bonina, L. Montenegro, N. Scrofani, E. Esposito, R. Cortesi, E. Menegatti, C.
microemulsion systems for transdermal delivery: assessment of bicontinuous Nastruzzi, Effects of phospholipid based formulations on in vitro and in vivo per-
structures, Int. J. Pharm. 137 (1996) 177–186. cutaneous absorption of methyl nicotinate, J. Control. Rel. 34 (1995) 53–63.
 M.J. Lawrence, G.D. Rees / Advanced Drug Delivery Reviews 64 (2012) 175–193 193

[175] F. Dreher, P. Walde, P.L. Luisi, P. Elsner, Human skin irritation studies of a lecithin [206] J.P. Spatz, S. Mossmer, M. Moller, Mineralisation of gold nanoparticles in a block
microemulsion gel and of lecithin liposomes, Skin Pharmacol. 9 (1996) 124–129. copolymer microemulsion, Chemistry – A Eur. J. 2 (1996) 1552–1555.
[176] R.W. Greiner, D.F. Evans, Spontaneous formation of a water-continuous emul- [207] B. Michels, G. Waton, R. Zana, Dynamics of micelles of poly(ethylene oxide)
sion from water-in-oil microemulsion, Langmuir 6 (1990) 1793–1796. poly(propylene oxide) poly(ethylene oxide) block copolymers in aqueous solu-
[177] N.H. Shah, M.T. Carvajal, C.I. Patel, M.H. Infeld, A.W. Malick, Self-emulsifying tions, Langmuir 13 (1997) 3111–3118.
drug delivery systems with polyglycolyzed glycerides for improving in vitro dis- [208] R. Cortesi, E. Esposito, A. Maietti, E. Menegatti, C. Nastruzzi, Formulation study
solution and oral absorption of lipophilic drugs, Int. J. Pharm. 106 (1994) 15–23. for the antitumor drug camptothecin: liposomes, micellar solutions and a
[178] S.A. Charman, W.N. Charman, M.C. Rogge, T.D. Wilson, F.J. Dutko, C.W. Pouton, microemulsion, Int. J. Pharm. 159 (1997) 95–103.
Self-emulsifying drug delivery systems: formulation and biopharmaceutical evalu- [209] R. Cortesi, C. Nastruzzi, Liposomes, micelles and microemulsions as new delivery
ation of an investigational lipophilic compound, Pharm. Res. 9 (1992) 87–93. systems for cytotoxic alkaloids, Pharm. Sci. Technol. Today 2 (1999) 288–298.
[179] B. Matuszewska, L. Hettrick, J.V. Bondi, D.E. Storey, Comparative bioavailability [210] P. Langguth, A. Kubis, G. Krumbiegel, W. Lang, H.P. Merkle, W. Wachter, L.H.
of L-683,453, a 5α-reductase inhibitor, from a self-emulsifying drug delivery Spahn, R. Weyhenmeyer, Intestinal absorption of the quaternary trospium
system in Beagle dogs, Int. J. Pharm. 136 (1996) 147–154. chloride: permeability-lowering factors and bioavailabilities for oral dosage
[180] D.Q.M. Craig, S.A. Barker, D. Banning, S.W. Booth, An investigation into the forms, Eur. J. Pharm. Biopharm. 43 (1997) 265–272.
mechanisms of self-emulsification using particle size analysis and low frequency [211] Z.-G. Gao, H.-G. Choi, H.-J. Shin, K.-M. Park, S.-J. Lim, K.-J. Hwang, C.-K. Kim,
dielectric spectroscopy, Int. J. Pharm. 114 (1995) 103–110. Physicochemical characterisation and evaluation of a microemulsion system
[181] K. Hutchison, Digestible emulsions and microemulsions for optimum oral deliv- for oral delivery of cyclosporin A, Int. J. Pharm. 161 (1998) 75–86.
ery of hydrophobic drugs, B.T. Gattefossé 87 (1994) 67–74. [212] P. Erkko, H. Granlund, M. Nuutinen, S. Reitamo, Comparison of cyclosporin A
[182] W.A. Ritschel, Microemulsions for improved peptide absorption from the gas- pharmacokinetics of a new microemulsion formulation and standard oral prep-
trointestinal tract, Meth. Find. Exp. Clin. Pharmacol. 13 (1993) 205–220. aration in patients with psoriasis, Brit. J. Derm. 136 (1997) 82–88.
[183] W.A. Ritschel, Microemulsion technology in the reformulation of cyclosporine: [213] B. Hausen, T. Ikonen, N. Briffa, G.J. Berry, U. Christians, R.C. Robbins, L. Hook, N.
the reason behind the pharmacokinetic properties of Neoral, Clin. Transplant. Serkova, L.Z. Benet, W. Schuler, R.E. Morris, Combined immunosuppression with
10 (1996) 364–373. cyclosporine (Neoral) and SDZ RAD in non-human primate lung transplantation:
[184] D.W. Holt, E.A. Mueller, J.M. Kovarik, J.B. van Bree, K. Kutz, The pharmacokinetics systematic pharmacokinetic-based trials to improve efficacy and tolerability,
of Sandimmune Neoral: a new oral formulation of cyclosporine, Transplant. Transplantation 69 (2000) 76–86.
Proc. 26 (1994) 2935–2939. [214] J.M. Kovarik, E.A. Muller, F. Richard, W. Tetzloff, Optimizing the absorption of
[185] J.M. Kovarik, E.A. Mueller, J.B. van Bree, W. Tetzloff, K. Kutz, Reduced inter and valspodar, a P-glycoprotein modulator. Part II: Quantifying its pharmacokinetic
intraindividual variability in cyclosporine pharmacokinetics from a variability and refining the bioavailability estimate, J. Clin. Pharmacol. 37
microemulsion formulation, J. Pharm. Sci. 83 (1994) 444–446. (1997) 1009–1014.
[186] E.A. Mueller, J.M. Kovarik, J.B. van Bree, W. Tetzloff, J. Grevel, K. Kutz, Improved [215] E.A. Muller, J.M. Kovarik, Y. Uresin, S.S. Preisig-Fluckiger, S. Hensel, P.W. Lucker, B. Holt,
dose linearity of cyclosporine pharmacokinetics from a microemulsion formula- Optimizing the absorption of valspodar, a P-glycoprotein modulator. Part I: Selecting
tion, Pharm. Res. 11 (1994) 301–304. an oral formulation and exploring its clinical pharmacokinetics/dynamics, J. Clin.
[187] S. Noble, A. Markham, Cyclosporin: a review of the pharmacokinetic properties, Pharmacol. 37 (1997) 1001–1008.
clinical efficacy and tolerability of a microemulsion-based formulation [216] S.-M. Khoo, A.J. Humberstone, C.J.H. Porter, G.A. Edwards, W.N. Charman, For-
(Neoral®), Drugs 50 (1995) 924–941. mulation design and bioavailability assessment of lipidic self-emulsifying for-
[188] G.D. Rees, M.D.G. Nascimento, T.R.J. Jenta, B.H. Robinson, Reverse enzyme synthesis mulations of halofantrine, Int. J. Pharm. 167 (1998) 155–164.
in microemulsion-based organogels, Biochim. Biophys. Acta 1073 (1991) 493–501. [217] K.M. Park, C.K. Kim, Preparation and evaluation of flurbiprofen-loaded
[189] P.-Y. Yeh, P.L. Smith, H. Ellens, Effect of medium chain glycerides on the physiological microemulsions for parenteral delivery, Int. J. Pharm. 181 (1999) 173–179.
properties of rabbit intestinal epithelium in vitro, Pharm. Res. 11 (1994) 1148–1154. [218] M.B. Delgado-Charro, G. Iglesias-Vilas, J. Blanco-Méndez, M.A. López-Quintela,
[190] R.A. Myers, V.J. Stella, Systemic bioavailability of penclomedine (NSC-338720) M.A.J.-P. Marty, R.H. Guy, Delivery of a hydrophilic solute through the skin
from oil-in-water emulsions administered intraduodenally in rats, Int. J. Pharm. from novel microemulsion systems, Eur. J. Pharm. Biopharm. 43 (1997) 37–42.
78 (1992) 217–226. [219] H.O. Ho, M.C. Huang, L.C. Chen, A. Hsia, K.T. Chen, H.S. Chiang, B.W. Spur, P.Y.K.
[191] T.T. Kararli, T.E. Needham, M. Griffin, G. Schoenhard, L.J. Ferro, L. Alcorn, Oral de- Wong, M.T. Sheu, The percutaneous delivery of prostaglandin E1 and its alkyl es-
livery of a renin inhibitor compound using emulsion formulations, Pharm. Res. 9 ters by microemulsions, Chin. Pharm. J. 50 (1998) 257–266.
(1992) 888–893. [220] S. Murdan, G. Gregoriadis, A.T. Florence, Novel sorbitan monostearate
[192] J. Kemken, A. Ziegler, B.W. Müller, Influence of supersaturation on the pharma- organogels, J. Pharm. Sci. 88 (1999) 608–614.
codynamic effect of bupranol after dermal administration using microemulsions [221] S. Murdan, G. Gregoriadis, A.T. Florence, Non-ionic surfactant based organogels
as a vehicle, Pharm. Res. 9 (1992) 554–558. incorporating niosomes, S.T.P. Pharma Sci. 6 (1996) 44–48.
[193] M. Fujii, K. Shiozawa, T. Henmi, S. Yamanouchi, H. Suzuki, N. Yamashita, M. [222] C.C. Muller-Goymann, H.-J. Hamann, Sustained release from reverse micellar so-
Matsumoto, Skin permeation of indomethacin from gel formed by fatty-acid lutions by phase transformations into lamellar liquid crystals, J. Control. Rel. 23
ester and phospholipid, Int. J. Pharm. 137 (1996) 117–124. (1993) 165–174.
[194] M.J. Garcia-Celma, Solubilisation of drugs in microemulsions, Indust. Appl. [223] M. Scherlund, M. Malmsten, P. Holmqvist, A. Brodin, Thermosetting microemulsions
Microemulsions 66 (1997) 123–145. as delivery systems for periodontal anesthesia, Int. J. Pharm. 194 (2000) 103–116.
[195] C. Satra, M. Thomas, M.J. Lawrence, Formulating oil-in-water microemulsions for [224] J. Gustafsson, H. Ljusberg-Wahren, M. Almgren, K. Larsson, Cubic lipid–water
pulmonary drug delivery, in: Drug Delivery To the Lungs, Vol. VI, The Aerosol phase dispersed into submicron particles, Langmuir 12 (1996) 4611–4613.
Society, 1995, pp. 40–43. [225] P. LoNostro, S.M. Choi, C.Y. Ku, S.H. Chen, Fluorinated microemulsions: a study of
[196] C.J. Richardson, A. Mbanefo, R. Aboofazeli, M.J. Lawrence, D.J. Barlow, Prediction phase behaviour and structure, J. Phys. Chem. B 103 (1999) 5347–5352.
of phase behaviour in microemulsion systems using artificial neural networks, [226] M.P. Krafft, J.G. Riess, Highly fluorinated amphiphiles and colloidal systems, and
J. Colloid Interface Sci. 187 (1997) 296–303. their applications in the biomedical field. A contribution, Biochimie 80 (1998)
[197] K. Kataoka, G.S. Kwon, M. Yokoyama, T. Okano, Y. Sakurai, Block copolymer mi- 489–514.
celles as vehicles for drug delivery, J. Control. Rel. 24 (1993) 119–132. [227] J.G. Riess, M.P. Krafft, Advanced fluorocarbon-based systems for oxygen and drug
[198] X.C. Zhang, J.K. Jackson, H.M. Burt, Development of amphiphilic diblock copoly- delivery, and diagnosis, Artificial Cells Blood Substitutes and Immobilisation Bio-
mers as micellar carriers of taxol, Int. J. Pharm. 132 (1996) 195–206. technology 25 (1997) 43–52.
[199] X.C. Zhang, H.M. Burt, G. Mangold, D. Dexter, D. von Hoff, L. Mayer, W.L. Hunter, [228] C. Cecutti, I. Rico, A. Lattes, A. Novelli, A. Rico, G. Marion, A. Graciaa, J. Lachaise,
Anti-tumour efficacy and biodistribution of intravenous polymeric micellar pac- New formulation of blood substitutes: optimisation of novel fluorinated
litaxel, Anti-cancer Drugs 8 (1997) 696–701. microemulsions, Eur. J. Med. Chem. 24 (1989) 485–492.
[200] T. Inoue, G. Chen, K. Nakamae, A.S. Hoffman, An AB block copolymer of [229] C. Cecutti, A. Novelli, I. Rico, A. Lattes, A new formulation for blood substitutes,
oligo(methy methacrylate) and poly(acrylic acid) for micellar delivery of hydro- J. Dispersion Sci. Technol. 11 (1990) 115–123.
phobic drugs, J. Control. Rel. 51 (1998) 221–229. [230] J.D. Holmes, D.C. Steytler, G.D. Rees, B.H. Robinson, Bioconversions in a
[201] B.G. Yu, T. Okano, K. Kataoka, G. Kwon, Polymeric micelles for drug delivery: water-in-CO2 microemulsions, Langmuir 14 (1998) 6371–6376.
solubilisation and haemolytic activity of amphotericin B, J. Control. Rel. 53 [231] J.G. Riess, Fluorinated vesicles, J. Drug Target. 2 (1994) 455–468.
(1998) 131–136. [232] J.G. Riess, Introducing a new element — fluorine — into the liposomal mem-
[202] J.E. Chung, M. Yokoyama, M. Yamato, T. Aoyagi, Y. Sakurai, T. Okano, Thermo- brane, J. Liposome Res. 5 (1995) 413–430.
responsive drug delivery from polymeric micelles constructed using block co- [233] C. Santaella, F. Frezard, P. Vierling, J.G. Riess, Extended in vivo blood circulation
polymers of poly(N-isopropylacrylamide) and poly(butylmethacrylate), time of fluorinated vesicles, FEBS Lett. 336 (1993) 481–484.
J. Control. Rel. 62 (1999) 115–127. [234] D. Attwood, L.R.J. Currie, P.H. Elworthy, Studies of solubilized micellar solutions.
[203] Y. Matsumura, M. Yokoyama, K.Z. Kataoka, T. Okano, Y. Sakurai, T. Kawaguchi, I. Phase studies and particle size analysis of solutions with noionic surfactants,
T. Kakizoe, Reduction of the side effects of an anti-tumour agent, KRN5500, by J. Colloid Interface Sci. 46 (1974) 249–256.
incorporation of the drug into polymeric micelles, Jpn. J. Cancer Res. 90 (1999) [235] D. Fennell Evans, H. Wennerstrom, Micro- and macro-emulsions, in: The Colloidal
122–128. Domain — Where Physics, Chemistry, Biology and Technology Meet, 1st Edition,
[204] S.A. Hagan, A.G.A. Coombes, M.C. Garnett, S.E. Dunn, M.C. Davis, L. Illum, S.S. vol. 1, VCH Publishers, 1994, pp. 451–469.
Davis, S.E. Harding, S. Purkiss, P.R. Gellert, Polylactide-poly(ethylene glycol) [236] J. Israelachvili, The science and application of emulsions — an overview, Colloids
copolymers as drug delivery systems.1. Characterization of water dispersible Surf. A: Physicochem. Eng. Aspects 91 (1994) 1–8.
micelle-forming systems, Langmuir 12 (1996) 2153–2161. [237] P.J. Atkinson, B.H. Robinson, A.M. Howe, R.K. Heenan, Structure and stability of
[205] P. Schmidt-Winkel, C.J. Glinka, G.D. Stucky, Microemulsion templates for microemulsion-based organogels, J. Chem. Soc. Faraday Trans. 87 (1991)
mesoporous silica, Langmuir 16 (2000) 356–361. 3389–3397.
 Current Opinion in Colloid & Interface Science 17 (2012) 274–280

Contents lists available at SciVerse ScienceDirect

Current Opinion in Colloid & Interface Science

journal homepage: www.elsevier.com/locate/cocis

Progress in microemulsion characterization

Durga P. Acharya ⁎, Patrick G. Hartley 1
CSIRO Division of Materials Science and Engineering, Bayview Avenue, Clayton, VIC 3168, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Microemulsions are technologically important complex fluids. In many applications, they are required to
Received 3 July 2012 accommodate functional additives such as drugs, polymers, and nanoparticles, which increase the complexity
Accepted 5 July 2012 of the systems further. Multiple complementary techniques are required in order to understand and manip-
Available online 14 July 2012
ulate microemulsion behavior. Recent advances in analytical techniques as well as theoretical models have
allowed observation of the nanoscale morphology of different microemulsion forms of increasing complexity
Keywords:
Microemulsions
at ever greater resolution. At the same time, there is increasing interest in the study of dynamic processes
Small-angle scattering within microemulsions using scattering techniques as well as techniques such as fluorescence correlation
NMR spectroscopy and ultrafast IR spectroscopy. This paper reviews these and other recent developments in the
Characterization characterization of microemulsion systems.
Surfactants © 2012 Elsevier Ltd. All rights reserved.
Solubilization

1. Introduction characterization techniques, which are the focus of this review article,
are then employed to reveal the detailed physicochemical properties
The term microemulsion first entered the scientific vernacular in of this microemulsion region, and most importantly their variations
1959, through the work of Jack H. Schulman and others at Columbia as a function of composition and environment. More recently, these
University [1]. They described these systems as ‘optically isotropic, studies have extended to studying compositions of increasing com-
transparent oil and water dispersions of droplets whose diameters plexity and to elucidate the substructure of the polar and non‐polar
were ‘somewhat greater than swollen micelles’. Since then the defini- compartments within the system. Dynamic structural rearrangements
tion has been refined several times to include key aspects such as the within microemulsions due to various stimuli are also a focus of cur-
importance of amphiphiles, the observation that microemulsions are rent research. Recent developments across the spectrum of techniques
thermodynamically stable (as distinct from conventional emulsions applied to microemulsion characterization are reviewed in this article,
and ‘mini- or ‘nano-’ emulsions), that they can exist as bicontinuous which is intended as an update to an excellent earlier review in this
systems, and more recently that their aqueous phase can be replaced journal [5].
with other polar components such as room temperature ionic liquids.
Nonetheless, the basic properties which Schulman identified have led 2. Electron microscopy
to the development of microemulsion systems in important applica-
tions, such as the delivery of food, consumer products and pharma- Electron microscopy (EM) techniques were instrumental in the
ceuticals, as well as in petrochemical recovery [2–4]. description of microemulsions by Schulman in 1959 [1]. They contin-
Microemulsion science has relied heavily on characterization ue to develop, and offer the potential for direct visualization of the
techniques, both from the perspective of fundamental elucidation of microstructure of microemulsions and other lyotropic mesophases
phase and nanostructural behavior, as well as through identifying at b 5 nm resolution. Transmission (TEM) and scanning (SEM) tech-
the unique properties which make them ‘fit for function’ in the appli- niques have been used to study internal and surface mesophase
cations mentioned above. nanostructure [6,7]. The general approach to TEM analysis is to pre-
The classical starting point for microemulsion characterization is pare a sample as a thin film, such that an electron beam can pass
the ternary phase diagram shown in Fig. 1. Represented simply, this through it. The image is collected with contrast resulting from differ-
involves generating oil, water and surfactant mixtures, and identify- ential absorption and scattering of electrons from different regions of
ing compositions which form a single phase. More complex the sample. As with all EM based techniques, imaging artefacts due to
sample preparation and dehydration are a major consideration. This
has led to the development of cryogenic preparation (cryo-TEM)
⁎ Corresponding author. Tel.: +61 3 9545 2518.
E-mail addresses: Durga.Acharya@csiro.au (D.P. Acharya), Patrick.Hartley@csiro.au
and freeze-fracture (FFEM) techniques [8].
(P.G. Hartley). The cryo-TEM technique combines cryogenic preparation techniques
1
Tel.: +61 3 9545 2595. with conventional transmission electron microscopy. Here, a thin film

1359-0294/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cocis.2012.07.002
 D.P. Acharya, P.G. Hartley / Current Opinion in Colloid & Interface Science 17 (2012) 274–280 275

it difficult to distinguish frozen water from microemulsion droplets.

Nevertheless, the technique can be used differentiate bicontinuous
from droplet type microemulsions. Etching of the samples, which
involves increasing temperature to around − 90 °C for a few minutes
to allow surface ice to sublime, and application of an ultra-thin con-
ductive surface coating (e.g. platinum) to prevent charging helps to
minimize interference from frozen water and improve the image
quality, resulting in overall increase in the usefulness of the technique
in characterizing droplet and bicontinuous microemulsions [14].

3. Scattering techniques

Scattering techniques involving X-rays, neutrons and light have

been used to obtain quantitative information on size, shape and mor-
phology of microemulsions. The basic principle of these techniques
involves applying an incident beam of radiation to the sample, and
recording the intensity and angle of the scattered beam. The scattering
arises from the interaction of the radiation with regions of different
refractive index (light scattering), electron density (X-ray scattering)
or nuclear composition (neutron scattering). In the small-angle
x-ray scattering (SAXS) technique, the recorded scattering profile at
low angles is fitted to suitable models to extract information about
the shape, size and nanostructure of scattering elements such as
those found in microemulsions. To reach the scale of interest for
Fig. 1. Schematic ternary phase diagram of a mixture of water, oil and surfactant.
microemulsions (>10 nm) scattering angles smaller than 1° must be
Oil-in-water (O/W) microemulsion (L1), water-in-oil (W/O) microemulsion (L2) and
bicontinuous microemulsion regions are highlighted. 2ϕ and 3ϕ are two-phase and
used.
three phase regions, respectively. Adapted from ref. [4]. In small-angle neutron scattering (SANS), neutrons from a reactor
source are scattered by the atomic nuclei of the sample. Different
nuclei or even different isotopes of the same element have different
of sample is rapidly vitrified without ice or sample crystallisation, and abilities to scatter neutrons, expressed as their characteristic scatter-
structures captured in the film are observed directly in a frozen- ing length density (SLD). There is no systematic variation of SLD with
hydrated state, i.e. very close to their native non-dehydrated state. A density or atomic number and it can even be negative. Hence,
recent review article covers experimental technique and applications through the use of isotopic substitution (e.g. chemical substitution
of cryo-TEM in the studying different colloidal systems [6]. of hydrogen atoms with deuterium), variable contrasts in neutron
Cryo-TEM has been commonly used for the characterization of var- scattering length density can be created within samples to allow a
ious mesophases and their colloidal dispersions [6]. Some noticeable specific phase of interest to be studied without significant impact on
advances have been made in the use of electron microscopy in charac- the physico-chemical properties. An advantage of the use of neutrons
terizing microemulsions including the bicontinuous type in recent is that neutrons are non-destructive, relative to X-rays, and hence
years. Cryo-TEM has been used to characterize microemulsions and radiation damage rarely occurs in a neutron based experiment. On
other mesophases including sponge phases [9–11]. Bicontinuous the other hand, the relatively slow data acquisition times and neces-
microemulsion phases are seen to have characteristic zig-zag channel- sity for a powerful neutron source (e.g. nuclear reactor) for neutron
like complex structures whereas in Water-in-oil/microemulsion sys- scattering experiments limits their applicability in the study of
tems, small droplets are seen on a continuous background. A sponge dynamic phenomena.
(L3) phase that forms at very low loading of oil exhibits a bicontinuous Dynamic light scattering (DLS), also known as photon correlation
network structure [10]. spectroscopy, can be used to analyse microemulsion droplet size via
In freeze fracture TEM (FFTEM), samples are cleaved under vacu- determination of hydrodynamic radius which can be extracted from
um. The fracture plane typically follows the interior (hydrophobic) measurements of the diffusion constants of diluted dispersed phase
domains of membrane like structures providing a “face on” view of (droplets) undergoing Brownian motion. The technique employs
samples. It generally produces a clear picture of the network structure relatively simple equipment and involves very short experimental
of bicontinuous microemulsions and/or droplets within discontinuous times [15,16].
microemulsion [7,12,13] and therefore can be considered as comple- Application of SAXS in determining shape and size of microemulsion
mentary to cryo-TEM for the characterization of microemulsions. droplets relies on the difference in the ability of oil and water phases
Scanning electron microscopy (SEM), another closely related to scatter x-rays. This property has been commonly used to estimate
EM technique, allows the direct mapping of surface features of the radius of a confined phase in O/W or W/O microemulsions [17,18].
microemulsions, but suffers from many of the sample preparation Meanwhile Time resolved SAXS (TR-SAXS) has been used to follow
problems discussed earlier with respect to non-cryogenic TEM tech- the change in shape and size of microemulsion droplets to study the
niques, and hence has been less widely applied. Field emission SEM kinetics of phase separation [19].
(FESEM) provides narrower probing beams at low as well as high Recent advances in computational hardware and software have
electron energy, resulting in both improved spatial resolution and led to considerable research efforts in developing and refining the
minimized sample charging and damage. Cryo-FESEM has also been techniques which are used to deconvolute scattering data. Broadly speak-
used to analyse the surface morphological structure of bicontinuous ing, these can be classified as model-dependent or model-independent
and discontinuous microemulsions [7,14]. Although the sample prep- methods. Model-dependent methods consist of fitting experimental
aration in this technique is less invasive than in FF-TEM, since it only data to scattering functions derived from well defined radial density
involves the freezing and fracturing of the sample with a knife on the profiles for the scattering object. They are particularly useful when the
stage of the microscope, formation of ice crystals during transfer of internal structure of the particle is known from other experimental
the samples often cannot be completely avoided which can make techniques or can be predicted. Model-independent approaches such
276 D.P. Acharya, P.G. Hartley / Current Opinion in Colloid & Interface Science 17 (2012) 274–280

as generalized indirect Fourier transformation (GIFT) involve Fourier of water on its dynamic state in Aerosol OT-hydrocarbon based re-
transformation of the scattering data to obtain a pair distance distribution verse micelles/microemulsions [30]. Here, the SAXS experiment pro-
function (PDDF) which is related to form factor and contains information vided microemulsion drop shape, size and polydispersity while NIR
about shape and size of the particle. The resulting PDDF can be further revealed the hydrogen bonding states of water molecules at different
deconvoluted to obtain the radial distribution profile of the scattering levels of hydration.
particle [20]. Although there is no assumption of the shape of the particle Time resolved-SAXS combined with time resolved turbidity mea-
to obtain its form factor, the derivation of the structure factor, which is a surement has also been employed to simultaneously monitor the pro-
contribution to the scattering of a particle arising from interparticle inter- file of growth of oil droplets and shrinking of microemulsion droplets
actions, involves an assumption of an appropriate model for interparticle as a result of phase separation. Using high flux x-ray beams available
interactions. Hence, the approach is truly model independent only at low at synchrotron facilities, it is possible to capture scattering patterns at
droplet density where interparticle interaction is negligible. Recently, very short exposure time, typically fractions of a second. This has been
Nagao et al. [21] proposed a “relative form factor” method to obtain a exploited to study the morphological evolution of microemulsion
droplet density dependence of the form factor without any assumption systems in close to real time [31]. Structural transitions within a poly-
of the profile of structure factor. The method, when applied to reverse merizing microemulsion containing methacrylate-terminated siloxane
type droplet microemulsions, allowed the form factor and structure macromonomer as an oil phase were reported recently [31]. The time
factor to be extracted without any assumption of the structure factor evolution of scattering data during the polymerization reaction induced
even in the dense droplet region, although the structure factor obtained by x-rays were analyzed by both Teubner–Stery and core-shell models.
from the method could not be explained by any existing models. The TS model showed that the periodicity of the oil and water domains
There are various models available for the model dependent anal- decreased and correlation length increased during the polymerization,
ysis of small angle scattering data from microemulsions. The most indicating shrinkage of the systems in agreement with the results
commonly used model for the analysis of scattering data from discon- obtained from the analysis of data by the core-shell model.
tinuous type microemulsion systems is the core-shell model [22]. The Increasingly complex microemulsion mixtures are the subject of
basic form of the model considers a radial density profile varying in a recent SAXS studies. The formation of a transient polymer bridged
step-like fashion resulting in a sharp jump at the boundary of concen- droplet network was interpreted from the structure factor extracted
tric domains of different densities (for example, at the hydrophobic from the SAXS data in a study of the addition of an ABA triblock copol-
core and hydrophilic shell of O/W microemulsions). A recent paper ymer to the AOT-water-decane inverse microemulsion system [18].
[23] discussed the limitations of such a model and proposed an ana- SANS has been used to characterize both droplet [29,32–34] and
lytical and continuous density distribution function with a widely bicontinuous microemulsions [26,27,29,35] with a focus on studying
variable shape. It was used to derive an analytical scattering form factor the evolution of droplet size, and packing at and around the droplet
to describe the scattering from particles with the radial density profile percolation threshold. Reviews on the applications of SANS to -scCO2,
of homogeneous spheres, shells, or core-shell particles. Analytical ionic liquid microemulsions, and microemulsions stabilized by surfac-
expressions have also been derived for the scattering function from tant and copolymer mixtures are available [36].
network junctions, and compared with measured small-angle scatter- Neutron spin echo (NSE) spectroscopy, a technique closely related
ing data obtained from a nonionic microemulsion system consisting of to SANS, is an excellent tool for measuring the dynamic properties of
cylindrical elements and Y-junctions formed by interconnected cylin- microemulsions, such as diffusion of droplets as well as membrane
ders [24]. Recently, scattering functions have been derived for core- dynamics and shape fluctuations both in terms of spatial and time
shell structured hard spheres with single as well as multiple (layered) resolution. In this inelastic neutron scattering technique, a polarized
shells with polydispersity in both core radii and overall diameters. beam of neutrons with a velocity distribution is passed through a
These models were also tested on a well characterized discontinuous sample placed between two magnetic regions with opposite field
O/W type microemulsion at various concentrations of oils and surfac- directions. A small change in the velocity of neutrons caused by ex-
tant in the system [25]. change of energy during the scattering process is measured. Analysis
Scattering data from bicontinuous microemulsions is commonly of the inelastic broadening of SANS intensity gives the dynamic struc-
analysed by the Teubner–Strey (TS) model which gives two charac- ture factor which is related to detailed real-space dynamics within
teristic lengths — repeat distance and correlation length — to describe the system [37]. A recent review article covers the basic principles,
the structure [26–28]. Freiberger et al. applied the TS model as well as recent developments, and application of NSE to soft matter [38].
the model independent GIFT approach to fit scattering data from SANS and NSE spectroscopy have been used to study the shape and
nonionic microemulsion systems undergoing phase transitions from structure of microemulsion droplets and their translational diffusion
O/W to W/O type microemulsion via a bicontinuous phase [29]. It when a reversed droplet microemulsion system is cooled from room
was found that the GIFT approach could be used to fit the scattering temperature down to 220 K [39]. The translational and rotational
data from both the discontinuous and bicontinuous phases. The pic- dynamics of confined water molecules within water droplets have
ture of the system based on the form factors and structure factors been studied over more than three decades in time from pico- to
extracted from the GIFT approach is interesting. It suggests a gradual nanoseconds by quasielastic neutron scattering which showed slow
increase in the size of particles, but no significant change in the parti- (interfacially bound) and fast (core associated) motions [40]. Such
cle structure/shape when the system moved from the discontinuous studies have been extended to reverse microemulsions containing
to the bicontinuous. This is an example of the fact that a good fit to glycerol as the dispersed phase in investigations of the dynamics of
a model does not guarantee that the model considered provides a cor- confined supercooled glycerol [41]. Nagao and Seto [42] analyzed
rect description of the system. However, the variation in the extracted NSE data from concentrated W/O microemulsion systems by studied
parameters of structure factors was consistent with the expected the variation in shape fluctuation and bending elasticity as a function
structural changes when a normal discontinuous microemulsion trans- of the concentration of water (dispersed phase). Wellart et al. [43]
forms to first to bicontinuous then reversed type. As expected, the TS used NSE spectroscopy, in conjunction with dynamic light scattering
model on the same system showed good fits to the scattering data for technique, to study the influence of temperature variations on the
bicontinuous systems only [29]. bending elastic constant of bicontinuous microemulsions formed
Combining SAXS with other techniques allows the correlation of by sugar surfactants which show low temperature dependence of
phase structural information with different microemulsion proper- phase behavior. Combination of NSE and DLS techniques, allowed
ties. For instance, SAXS in combination with near infrared (NIR) spec- local thermally excited undulations of the interface to be expressed
troscopy has been used to characterize the effect of the confinement in terms of the bending elasticity to be separated from long-range
 D.P. Acharya, P.G. Hartley / Current Opinion in Colloid & Interface Science 17 (2012) 274–280 277

collective motions, found the undulation of interface independent of relaxation of water molecules, with two-dimensional proton–proton
temperature over wide range [43]. nuclear Overhauser effect spectroscopy and molecular modelling, it
Meanwhile, the Dynamic Light scattering (DLS) technique has has been shown that the location of the fluorophore complex depended
been used to characterize discontinuous microemulsions often in on the size of the aqueous core of the microemulsion, with the probe
combination with other characterization techniques such as cryo-SEM molecule being located deep inside the surfactant layer for small drop-
[44], SAXS or SANS [18,45,46], turbidimetry [47], nuclear magnetic let and nearer the interface for larger droplets [48].
resonance spectroscopy [48–50], and absorption and emission spec-
troscopy [51–53]. A recent paper describes a new approach for DLS 5. Spectroscopic techniques
data analysis and applies it to a nonionic microemulsion system to ex-
tract information such as free surfactant concentration, micellar molar Several spectroscopic techniques have been used to study differ-
composition, surfactant interfacial area and aggregation numbers [16]. ent aspects of microemulsion structures and properties. The absorp-
tion and steady-state emission spectroscopy of probe molecules
4. Nuclear magnetic resonance (NMR) solubilized in a microemulsion system can probe the polarity of the
microemulsion at their solvation location [58]. Time-resolved emis-
NMR has been used in the characterization of microemulsions to sion spectroscopy also provides dynamics and rotation relaxation of
study diffusion properties of components and relaxation behavior. A solvent in microemulsions [58–60].
review article covering the theoretical basis of NMR spectroscopy in Chemiluminescence techniques have also been employed to study
microemulsions is available [54]. Fourier transform pulsed gradient transitions between polar and non-polar environments in microemulsion
spin echo (FT-PGSE) technique allows one to measure self-diffusion systems. For instance, Tetrakis(dimethylamino)ethylene chemilumines-
coefficients of oil, water and surfactant molecules in the same experi- cence (TDE CL) is based on the transport of quenching species from the
ment. In this technique NMR peak attenuation is used to measure the continuous nonpolar phase to the interior polar region of aggregates. It
displacement (translational diffusion) of spins due to specific nuclei in involves the measurement of emission intensity of TDE CL as a function
the lipid/surfactant and water over a defined period in a magnetic field of time at various concentrations of surfactant and then fitting the data
gradient. Attenuation is measured as a function of field gradient pulse to estimate the decay constant which, when plotted against surfactant
duration, and the results can be fitted to yield the diffusion coefficient. concentration, shows a change in trend at the critical micellar concentra-
The technique permits one to differentiate between discontinuous tion [61]. An advantage of CL analysis over earlier techniques employing
and bicontinuous microemulsions and also to determine whether a fluorescence probes, is the absence of potentially interfering external
discontinuous microemulsion is W/O type or O/W type, since the light sources and therefore the possibility of measuring emissions against
oil- or water-continuous phase in discontinuous microemulsions a completely dark background. The sensitivity of the emission intensity of
has a self-diffusion coefficients close to that of the bulk liquid. In TDE CL to variation in solvent viscosity can also be exploited for measur-
bicontinuous microemulsions, free diffusion of solvent phases is ing the viscosity for reverse microemulsion droplets in solvent (oil).
restricted in two directions due to the presence of the surfactant Fluorescence correlation spectroscopy (FCS), which may be con-
stabilised interface, resulting in reduced diffusion coefficients. Hence, sidered as a miniaturization of DLS, measures the tiny spontaneous
simultaneously high diffusion coefficients of both component solvents fluctuations in fluorescence intensity of molecules in a very small
provide evidence of bicontinuous structures, and a sharp change in the volume (spot), and quantifies it by temporarily autocorrelating the
diffusion properties of one of the solvents corresponds to the transition recorded intensity signal to obtain the variation in local concentration
from bicontinuous to discontinuous microemulsions. The method has of fluorescent species and hence their diffusion coefficient [62]. The
been used, in combination with conductivity and SANS to characterize technique is an excellent tool for measuring molecular diffusion
gelled polymerizable microemulsions and changes in structure as a and size under extremely dilute conditions. In very dilute conditions,
function of temperature [35]. Multinuclear NMR (1H, 31P and 19F) tech- droplets of dispersed particles of reverse microemulsions become
niques have been used to characterize microemulsions consisting of invisible in light scattering experiment due to the so-called optical
water, a fluorinated drug solubilized in oil, and various phospholipid matching point condition where the dielectric constant of the dis-
stabilizers by measuring the diffusion coefficients of the individual com- persed phase becomes nearly same as that of the bulk. Another diffi-
ponents [55]. In another application of NMR diffusimetry, the internal culty with the light scattering method applied to microemulsion
structure of water-in-diesel microemulsions formulated with nonionic system arises from the variation of refractive index of droplets with
surfactants and with varying amounts of water has been investigated increase or decrease in droplet size and consequently the water
by determining the diffusion coefficients of diesel hydrocarbons, water content as well as surfactant aggregation number of droplets. In a
and the surfactants. It was been shown that as the amount of water polydispersed system, this introduces a refractive index mismatch
increased, the water domains first swell and then multiplied [56]. The among the droplets of different sizes which produces an unusual var-
technique is not limited to normal microemulsions but can also be iation in the droplet size distribution in light scattering measurement
applied to characterize microemulsions containing ionic liquids [50]. as the system approaches the optical matching point. FCS has been
The NMR relaxation technique for characterizing microemulsions shown to overcome these difficulties of light scattering and extract
involves measuring the molecular relaxations of component mole- accurate size and polydispersity of reversed microemulsion droplets
cules. Using models, it can provide information about aggregate using a fluorescent marker [63] and a model that incorporates the
shape and size, and it is sensitive in picking up subtle changes in size polydispersity of droplets. A methodology for the correction of
droplet shape and size without any interference from droplet interac- refractive index mismatch and fitting analysis of correlation curves
tions at high volume fractions. The technique has been used obtain in- to extract has also been shown. Although FCS is able to bypass diffi-
formation on water content and the structural organization of reverse culties of light scattering techniques in dilute solutions, its main lim-
microemulsions formed by AOT surfactants with Mg 2 + and different itation is that the method is not suitable for bigger particles especially
alkali metal ions, M + [57]. Free water and water molecules coordinat- when the droplet size becomes comparable to or bigger than the size
ed with the Mg 2+ ion gave two separate NMR signals, indicating that of the spot the FCS is looking at. Hence the method is complementary
the environment in the Mg(AOT)2 microemulsion is different from to DLS rather than a substitution.
microemulsions of MAOT. The technique has also been used to obtain Fourier transform Infrared (FTIR) spectroscopy has been used to
information on the location and dynamics of a redox fluorophore, distinguish between the local environments of water molecules con-
[Ru(bpy)3] 2+ which is used to study the microstructure inside fined in the core of reverse microemulsions because of its high sensi-
reverse microemulsions. Combining longitudinal or spin–lattice (T1) tivity to interactions between water molecules in the water pool, and
278 D.P. Acharya, P.G. Hartley / Current Opinion in Colloid & Interface Science 17 (2012) 274–280

between water and surfactant at the interface [64,65]. The technique immiscible homopolymers stabilized by a copolymer lying at the inter-
has been applied to water-in-scCO2 microemulsions stabilized by a mix- face) reveal rich rheological behavior, with Newtonian behavior at low
ture of fluorinated surfactants including fluorinated AOT-analogue shear rates and shear thinning under moderate shear [78]. At higher
surfactant by studying the stretching and bending bands of water and shear rate, the system showed shear induced bulk phase separation
the S = O stretching band of the surfactant over wide range of water/ with the two homopolymer-rich phases being ejected, with the sample
CO2 ratios. The number of water molecules in CO2 as dissolved, associat- resembling a blend of immiscible polymers under very high shear [78].
ed with the core, or at the water-surfactant interface as a function of It was found that the shear-induced macroscopic phase separation is a
water content in the system was determined [64]. Performing measure- characteristic feature of polymeric microemulsions irrespective of the
ments in the Near Infrared (NIR) region further addressed problems viscosity contrast between homopolymer phases.
relating to competing absorption in the mid-IR region [66].
Ultrafast IR spectroscopic techniques have also been employed to 7. Conductivity
study the hydrogen bonding network and dynamics of water molecules
in reverse micelles/microemulsions [67] or at the interface of AOT Electrical conductivity remains a simple and inexpensive technique
reverse micelles [68]. The technique measures orientational relaxation for microemulsion characterization. It primarily reveals whether an
of water molecules by exploring the OD stretch of dilute HOD in water aqueous or oil phase or both phases are continuous. The conductivity
and provides a quantitative determination of the dynamics of water as measurement technique has been used to determine the type of
a function of size and nature of confining structure. microemulsion, and to estimate phase boundaries resulting from
Dielectric spectroscopy is another spectroscopic technique which changes in composition or temperature [9,35,79]. In combination with
can provide information about the morphology of microemulsions other characterization technique, conductivity has also been used in
and dynamics of different polar groups and aggregates by measuring microemulsions containing ionic liquids [44,50,52] in which the ionic
the variation of conductivity and dielectric constant over a wide range liquid replaces the high conductivity aqueous phase in conventional
of frequencies ranging from 10 −5 to 10 9 Hz, or correspondingly microemulsions.
dielectric relaxation processes over a wide range of time from 10 4
to 10 −11 s, due to polarization at the interface between oil and 8. Ionic liquid microemulsions
water phases. The theoretical framework and applications of dielec-
tric spectroscopy in soft material particles have been covered in Nonaqueous microemulsions consisting of room temperature ionic
review articles [69,70]. Dielectric spectra of the microemulsion sys- liquids (RTILs) as the polar phase have drawn considerable interest in
tem can be described by a relaxation function consisting of two par- recent years. A recent review article provides further information on
tially overlapping but distinguishable relaxation components at high ionic liquid based microemulsions [80]. RTIL microemulsions exhibit
(above 500 MHz) and low frequency regions which are attributed many similarities with aqueous base microemulsions, such as droplet
to correlated movement of surfactant at the water surfactant inter- and bicontinuous structures. Many characterization techniques used
face and combination of slow interfacial relaxation of water and inter- for aqueous microemulsions are also applicable to RTIL microemulsions.
facial polarization. An additional dielectric relaxation contribution Micellar aggregates and microemulsions containing alkane, non-
reported recently has been attributed to additional dielectric polari- ionic polyoxyethylene alkyl ether surfactant and ethylammonium
zation mechanics arising from a correlated motion of the anionic ammonium nitrate (EAN) or propylammonium nitrate (PAN) as
group of the ionized surfactant molecule occurring at the water–sur- polar phase have been characterized by SANS and SAXS [81–83].
factant interface [71]. Dielectric spectroscopy proves to be a powerful SAXS analysis along a line of dilution from oil-rich to RTIL-rich
technique in the study of the structural changes such as organization conditions at a fixed concentration of nonionic surfactant within a
and rearrangements of droplets forming clusters when microemulsions single phase microemulsion domain in the ternary phase diagram
undergo percolation in reverse microemulsions [71,72] as well as in has been used to propose a microstructural change with increase
ionic liquid microemulsions [73]. in IL content [84].
The effect of adding water to an IL-in-oil microemulsion has been
6. Rheology investigated by using DLS, FTIR, 1H NMR, 2D-rotational Overhauser
effect spectroscopy, ITC and conductivity measurement technique.
The rheological properties of microemulsions depend on the type, Added water does not penetrate the surfactant layer but remains
shape and number density of aggregates present, as well as the inter- hydrogen bonded to the peripheral second and third oxyethylene
actions between these aggregates. Hence, microstructural changes units of surfactant Triton X-100 of the microemulsion droplets in
such as sphere-rod or discontinuous to bicontinuous transitions are the vicinity of polar head groups, and results in a transition of large
reflected in microemulsion rheology. Bicontinuous microemulsions elliptical droplets into small-sized spherical structures and subse-
exhibit a Newtonian behavior (constant viscosity) at low to medium quent increase in the number of droplets [85]. Steady state and time
shear rates but shear thinning is observed at high shear rates, proba- resolved emission spectroscopy using coumarin as probes has been
bly due to fragmentation of the bicontinuous structure. Discontinuous used to study the effect of added water on solvent and rotational
microemulsions on the other hand show Newtonian behavior over relaxation dynamics in water-in-RTIL microemulsion [50,86]. With
a wider range of shear rates [74]. However, differentiation of the types increase in water content of the microemulsions, the probe molecules
of microemulsions or identification of the structure of the microemulsion are gradually shifted to the aqueous core and experience free water-
cannot be done purely on the basis of rheological data, because this mac- like environment. On the other hand, addition of an ionic liquid to aque-
roscopic property is not sensitive enough to detect subtle microstructural ous microemulsion system also tunes the curvature of microemulsion.
changes such as the transformation of microemulsion to a wormlike Here, the ionic liquid behaves like an electrolyte but more effectively,
micellar phase induced by temperature [75]. Hence, rheometry has promoting ultralow interfacial tension, induces a Winsor I → Winsor
most often been used in combination with other techniques in the char- III → Winsor II transition [87].
acterization of microemulsions. For example, it has been used in conjunc-
tion with SANS and a phase study to measure the percolation threshold 9. Summary
and nanostructure of microemulsions containing a telechelic polymer
and silica nanoparticles [76,77]. Microemulsions are complex liquids with various structures and
Rheological studies of polymeric bicontinuous microemulsions physico-chemical behaviors. Technological application of these systems
(which contain continuously interpenetrating domains of two often requires additional components such as drugs, polymers, and
 D.P. Acharya, P.G. Hartley / Current Opinion in Colloid & Interface Science 17 (2012) 274–280 279

nanoparticles which increases the complexity of the systems further. [15] Silva EJ, Zaniquelli MED, Loh W. Light-scattering investigation on microemulsion
formation in mixtures of diesel oil (or hydrocarbons) plus ethanol plus additives.
Multiple complementary techniques are required in microemulsion Energy Fuel 2007;21:222-6.
characterization. Recent advances in electron microscopy have allowed [16] Lemyre JL, Lamarre S, Beaupre A, Ritcey AM. A new approach for the characterization
observation of the nanoscale morphology of different microemulsion of reverse micellar systems by dynamic light scattering. Langmuir 2010;26:10524-31.
[17] Deen GR, Pedersen JS. Phase behavior and microstructure of C12E5 nonionic
forms. However, there is ample room to improve, especially in improved microemulsions with chlorinated oils. Langmuir 2008;24:3111-7.
sample preparation and in reducing artefacts such as those arising from [18] Blochowicz T, Gogelein C, Spehr T, Muller M, Stuhn B. Polymer-induced transient
ice formation. Scattering techniques, which are complementary to imag- networks in water-in-oil microemulsions studied by small-angle x-ray and
dynamic light scattering. Phys Rev E 2007;76.
ing techniques in the sense that they provide nanostructural informa- [19] Deen GR, Oliveira CLP, Pedersen JS. Phase behavior and kinetics of phase separa-
tion in ambient conditions, have been extensively used to characterize tion of a nonionic microemulsion of C(12)E(5)/water/1-chlorotetradecane upon
various type of microemulsions based on model dependent and model a temperature quench. J Phys Chem B 2009;113:7138-46.
[20] Glatter O. Fourier transformation and deconvolution. In: Lindner P, Zemb Th, editors.
independent data analysis. New models have been proposed to refine
Neutrons, X-rays and light: scattering methods applied to soft condensed matter.
droplet structure and accommodate more complex network structures Amsterdam: Elsevier; 2002. p. 103-24. [Chapter 5].
within bicontinuous systems. However, model choice requires care, [21] Nagao M, Seto H, Yamada NL. Interaction between droplets in a ternary
since a good fit to scattering data does not guarantee that the extracted microemulsion evaluated by the relative form factor method. Phys Rev E 2007;75.
[22] Zemb T. Scattering by microemulsions. In: Lindner P, Zemb Th, editors. Neutrons,
information is valid. Analysis and model fitting of scattering data con- X-rays and light: scattering methods applied to soft condensed matter. Amsterdam:
tinues to be a developing research field as a result. Elsevier; 2002. p. 317-50. [Chapter 13].
Interests in the study of dynamic processes within microemulsions [23] Foster T. Universal analytical scattering form factor for shell-, core-shell, or homo-
geneous particles with continuously variable density profile shape. J Phys Chem B
are increasingly important. Although scattering techniques are excel- 2011;115:10207-17.
lent in extracting structural information, their main weakness is in the [24] Foster T, Safran SA, Sottmann T, Strey R. Scattering form factors for self-assembled
area of capturing information about dynamic processes at very short network junctions. J Chem Phys 2007;127.
[25] Nayeri M, Zackrisson M, Bergenholtz J. Scattering functions of core-shell-struct
time scales. Neutron spin echo spectroscopy has emerged as a closely ured hard spheres with Schulz-distributed radii. J Phys Chem B 2009;113:
related but complementary technique to SANS because of its ability to 8296-302.
catch information on the dynamics of microemulsion droplets such as [26] Dave H, Gao F, Schultz M, Co CC. Phase behavior and SANS investigations of edible
sugar-limonene microemulsions. Colloids Surf A Physicochem Eng Asp 2007;296:
interfacial fluctuation and translational movement. 45-50.
Ultrafast IR spectroscopy, chemiluminescence and fluorescence [27] Wellert S, Karg M, Imhof H, Steppin A, Altmann HJ, Dolle M, et al. Structure of
correlation spectroscopy (FCS) are emerging techniques in the area biodiesel based bicontinuous microemulsions for environmentally compatible
decontamination: a small angle neutron scattering and freeze fracture electron
of microemulsion characterization. The ability of FCS to see the dy-
microscopy study. J Colloid Interface Sci 2008;325:250-8.
namics of molecules at very small sample volume and low concen- [28] Acosta EJ, Szekeres E, Harwell JH, Grady BP, Sabatini DA. Morphology of ionic
tration has made it attractive for the characterization of dilute microemulsions: comparison of SANS studies and the net-average curvature
solutions. On the other hand techniques like NMR diffusimetry and (NAC) model. Soft Matter 2009;5:551-61.
[29] Freiberger N, Moitzi C, de Campo L, Glatter O. An attempt to detect bicontinuity
relaxation techniques, DLS, and conductimetry are commonly used from SANS data. J Colloid Interface Sci 2007;312:59-67.
to characterize microemulsions due to the availability of equipment [30] Balakrishnan S, Javid N, Weingartner H, Winter R. Small-angle x-ray scattering
and ease of data analysis. They continue to be widely applied to and near-infrared vibrational spectroscopy of water confined in aerosol-OT re-
verse micelles. Chemphyschem 2008;9:2794-801.
classical and novel microemulsions containing ionic liquids and [31] Peng SH, Guo QP, Hughes TC, Hartley PG. In situ synchrotron SAXS study of
polymers. polymerizable microemulsions. Macromolecules 2011;44:3007-15.
[32] Foster T, Sottmann T, Schweins R, Strey R. Small-angle neutron scattering from
giant water-in-oil microemulsion droplets. I. Ternary system. J Chem Phys
2008;128.
References [33] Foster T, Sottmann T, Schweins R, Strey R. Small-angle-neutron-scattering from
giant water-in-oil microemulsion droplets. II. Polymer-decorated droplets in a
[1] Schulman JH, Stoeckenius W, Prince LM. Mechanism of formation and structure of quaternary system. J Chem Phys 2008;128.
micro emulsions by electron microscopy. J Phys Chem 1959;63:1677-80. [34] Silva BFB, Marques EF, Olsson U, Linse P. Size, shape, and charge of salt-free
[2] Gupta S, Moulik SP. Biocompatible microemulsions and their prospective uses in catanionic microemulsion droplets: a small-angle neutron scattering and model-
drug delivery. J Pharm Sci 2008;97:22-45. ing study. J Phys Chem B 2009;113:10230-9.
[3] Gupta S. Biocompatible microemulsion systems for drug encapsulation and delivery. [35] Stubenrauch C, Tessendorf R, Salvati A, Topgaard D, Sottmann T, Strey R, et al.
Curr Sci 2011;101:174-88. Gelled polymerizable microemulsions. 2. Microstructure. Langmuir 2008;24:
[4] Paul BK, Moulik SP. Uses and applications of microemulsions. Curr Sci 2001;80: 8473-82.
990–1001. [36] Melnichenko YB, Wignall GD. Small-angle neutron scattering in materials science:
[5] Gradzielski M. Recent developments in the characterisation of microemulsions. recent practical applications. J Appl Phys 2007;102.
Curr Opin Colloid Interface Sci 2008;13:263-9. [37] Watson MC, Brown FLH. Interpreting membrane scattering experiments at the
[6] Kuntsche J, Horst JC, Bunjes H. Cryogenic transmission electron microscopy mesoscale: the contribution of dissipation within the bilayer. Biophys J 2010;98:
(cryo-TEM) for studying the morphology of colloidal drug delivery systems. Int J L9–L11.
Pharm 2011;417:120-37. [38] Farago B. Recent developments and applications of NSE in soft matter. Curr Opin
[7] Krauel K, Girvan L, Hook S, Rades T. Characterisation of colloidal drug delivery systems Colloid Interface Sci 2009;14:391-5.
from the naked eye to Cryo-FESEM. Micron 2007;38:796-803. [39] Spehr T, Frick B, Grillo I, Falus P, Muller M, Stuhn B. Structure and dynamics of
[8] Belkoura L, Stubenrauch C, Strey R. Freeze fracture direct imaging: a new freeze reverse micelles containing supercooled water investigated by neutron scatter-
fracture method for specimen preparation in cryo-transmission electron micros- ing. Phys Rev E 2009;79.
copy. Langmuir 2004;20:4391-9. [40] Spehr TL, Frick B, Zamponi M, Stuhn B. Dynamics of water confined to reverse AOT
[9] Lutz R, Aserin A, Wachtel EJ, Ben-Shoshan E, Danino D, Garti N. A study of the micelles. Soft Matter 2011;7:5745-55.
emulsified microemulsion by SAXS, Cryo-TEM, SD-NMR, and electrical conductivity. [41] Blochowicz T, Gouirand E, Fricke A, Spehr T, Stuhn B, Frick B. Accelerated dynamics
J Dispers Sci Technol 2007;28:1149-57. of supercooled glycerol in soft confinement. Chem Phys Lett 2009;475:171-4.
[10] Wolf L, Hoffmann H, Talmon Y, Teshigawara T, Watanabe K. Cryo-TEM imaging of [42] Nagao M, Seto H. Concentration dependence of shape and structure fluctuations
a novel microemulsion system of silicone oil with an anionic/nonionic surfactant of droplet microemulsions investigated by neutron spin echo spectroscopy. Phys
mixture. Soft Matter 2010;6:5367-74. Rev E 2008;78.
[11] Margulis-Goshen K, Kesselman E, Danino D, Magdassi S. Formation of celecoxib [43] Wellert S, Karg M, Holderer O, Richardt A, Hellweg T. Temperature dependence of the
nanoparticles from volatile microemulsions. Int J Pharm 2010;393:230-7. surfactant film bending elasticity in a bicontinuous sugar surfactant based
[12] Zhao XY, Xu J, Zheng LQ, Li XW. Preparation of temperature-sensitive microemulsion- microemulsion: a quasielastic scattering study. Phys Chem Chem Phys 2011;13:3092-9.
based gels formed from a triblock copolymer. Colloids Surf A Physicochem Eng Asp [44] Rabe C, Koetz J. CTAB-based microemulsions with ionic liquids. Colloids Surf A
2007;307:100-7. Physicochem Eng Asp 2010;354:261-7.
[13] Sharma SC, Tsuchiya K, Sakai K, Sakai H, Abe M, Komura S, et al. Formation [45] Causse J, Oberdisse J, Jestin J, Lagerge S. Small-angle neutron scattering study of
and characterization of microemulsions containing polymeric silicone. Langmuir solubilization of tributyl phosphate in aqueous solutions of L64 pluronic triblock
2008;24:7658-62. copolymers. Langmuir 2010;26:15745-53.
[14] Lutter S, Koetz J, Tiersch B, Kosmella S. Polymer-modified bicontinuous [46] Debbabi K, Guittard F, Eastoe J, Rogers S, Geribaldi S. Reverse water-in-
microemulsions used as a template for the formation of nanorods. J Dispers Sci fluorocarbon microemulsions stabilized by new polyhydroxylated nonionic fluo-
Technol 2009;30:745-52. rinated surfactants. Langmuir 2009;25:8919-26.
280 D.P. Acharya, P.G. Hartley / Current Opinion in Colloid & Interface Science 17 (2012) 274–280

[47] Deen GR, Pedersen JS. Nucleation of an oil phase in a nonionic microemulsion- [67] Pieniazek PA, Lin YS, Chowdhary J, Ladanyi BM, Skinner JL. Vibrational spectroscopy
containing chlorinated oil upon systematic temperature quench. J Phys Chem B and dynamics of water confined inside reverse micelles. J Phys Chem B 2009;113:
2010;114:7769-76. 15017-28.
[48] Binks DA, Spencer N, Wilkie J, Britton MM. Magnetic resonance studies of a redox probe [68] Fayer MD, Levinger NE. Analysis of water in confined geometries and at interfaces.
in a reverse sodium bis(2-ethylhexyl)sulfosuccinate/octane/water microemulsion. In: Yeung ES, Zare RN, editors. Annual Reviews, vol. 3. Annual Review of Analytical
J Phys Chem B 2010;114:12558-64. Chemistry; 2010. p. 89–107. [Annual Review of Analytical Chemistry, vol. 3].
[49] Sedgwick MA, Trujillo AM, Hendricks N, Levinger NE, Crans DC. Coexisting aggre- [69] Cametti C. Dielectric properties of soft-particles in aqueous solutions. Soft Matter
gates in mixed aerosol OT and cholesterol microemulsions. Langmuir 2011;27: 2011;7:5494-506.
948-54. [70] Bonincontro A, Cametti C. Interfacial characterization of mesoscopic particle sus-
[50] Pramanik R, Sarkar S, Ghatak C, Rao VG, Sarkar N. Ionic liquid containing pensions by means of radiowave dielectric spectroscopy: a minireview. Colloids
microemulsions: probe by conductance, dynamic light scattering, diffusion- Surf A Physicochem Eng Asp 2004;246:115-20.
ordered spectroscopy nmr measurements, and study of solvent relaxation dynamics. [71] Cametti C. Dielectric and conductometric properties of highly heterogeneous col-
J Phys Chem B 2011;115:2322-30. loidal systems. Riv Nuovo Cim 2009;32:185-260.
[51] Sagisaka M, Hino M, Oasa J, Yamamoto M, Yoda S, Takebayashi Y, et al. Character- [72] Wipf R, Jaksch S, Stuhn B. Dynamics in water-AOT-n-decane microemulsions with
ization of water/supercritical CO(2) microemulsion by UV-visible spectroscopy poly(ethylene glycol) probed by dielectric spectroscopy. Colloid Polym Sci 2010;288:
and dynamic light scattering. J Oleo Sci 2009;58:75-83. 589-601.
[52] Pramanik R, Sarkar S, Ghatak C, Rao VG, Setua P, Sarkar N. Microemulsions with [73] Lian YW, Zhao KS. Dielectric analysis of micelles and microemulsions formed in a
surfactant TX100, cyclohexane, and an ionic liquid investigated by conductance, hydrophilic ionic liquid. I. Interaction and percolation. J Phys Chem B 2011;115:
DLS, FTIR measurements, and study of solvent and rotational relaxation within 11368-74.
this microemulsion. J Phys Chem B 2010;114:7579-86. [74] Sharma G, Wilson K, van der Walle CF, Sattar N, Petrie JR, Kumar M.
[53] Sarkar S, Pramanik R, Ghatak C, Rao VG, Sarkar N. Characterization of 1-ethyl-3- Microemulsions for oral delivery of insulin: design, development and evaluation
methylimidazolium bis(trifluoromethylsulfonyl)imide ([Emim][Tf(2)N])/TX-100/ in streptozotocin induced diabetic rats. Eur J Pharm Biopharm 2010;76:159-69.
cyclohexane ternary microemulsion: investigation of photoinduced electron transfer [75] da Silva GC, de Morais WA, Neto AAD, Dantas TNC, Fonseca JLC. The relationship
in this RTIL containing microemulsion. J Chem Phys 2011;134. between rheology and dynamic light scattering for a xylene/water/ButOH/C12E9
[54] Furo I. NMR spectroscopy of micelles and related systems. J Mol Liq 2005;117: microemulsion. Colloids Surf A Physicochem Eng Asp 2012;397:42-7.
117-37. [76] Puech N, Mora S, Testard V, Porte G, Ligoure C, Grillo I, et al. Structure and rheo-
[55] Hoeller S, Kahlig H, Valenta C. Multinuclear NMR characterisation and dermal deliv- logical properties of model microemulsion networks filled with nanoparticles.
ery of fluorinated drugs in soybean-microemulsion systems. J Pharm Sci 2009;98: Eur Phys J E 2008;26:13-24.
2686-95. [77] Puech N, Mora S, Phou T, Porte G, Jestin J, Oberdisse J. Microemulsion nanocomposites:
[56] Lif A, Nyden M, Holmberg K. Water-in-diesel microemulsions studied by NMR phase diagram, rheology and structure using a combined small angle neutron scatter-
diffusometry. J Dispers Sci Technol 2009;30:881-91. ing and reverse Monte Carlo approach. Soft Matter 2010;6:5605-14.
[57] Stahla ML, Baruah B, James DM, Johnson MD, Levinger NE, Crans DC. H-1 NMR [78] Zhou N, Bates FS, Lodge TP, Burghardt WR. Shear flow behavior of a dynamically
studies of aerosol-OT reverse micelles with alkali and magnesium counterions: symmetric polymeric bicontinuous microemulsion. J Rheol 2007;51:1027-46.
preparation and analysis of MAOTs. Langmuir 2008;24:6027-35. [79] Li XC, He GH, Zheng WJ, Xiao GK. Study on conductivity property and microstruc-
[58] Saha R, Rakshit S, Mitra RK, Pal SK. Microstructure, morphology, and ultrafast ture of TritonX-100/alkanol/n-heptane/water microemulsion. Colloids Surf A
dynamics of a novel edible microemulsion. Langmuir 2012;28:8309-17. Physicochem Eng Asp 2010;360:150-8.
[59] Pramanik R, Sarkar S, Ghatak C, Rao VG, Sarkar N. Solvent and rotational relaxa- [80] Kunz W, Zemb T, Harrar A. Using ionic liquids to formulate microemulsions: cur-
tion study in ionic liquid containing reverse micellar system: a picosecond fluo- rent state of affairs. Curr Opin Colloid Interface Sci 2012;17:205-11.
rescence spectroscopy study. Chem Phys Lett 2011;512:217-22. [81] Atkin R, Warr GG. Phase behavior and microstructure of microemulsions with a
[60] Seth D, Chakraborty A, Setua P, Sarkar N. Dynamics of solvent and rotational room-temperature ionic liquid as the polar phase. J Phys Chem B 2007;111:9309-16.
relaxation of coumarin-153 in room-temperature ionic liquid 1-butyl-3-methyl [82] Atkin R, Bobillier SMC, Warr GG. Propylammonium nitrate as a solvent for amphi-
imidazolium tetrafluoroborate confined in poly(oxyethylene glycol) ethers phile self-assembly into micelles, lyotropic liquid crystals, and microemulsions.
containing micelles. J Phys Chem B 2007;111:4781-7. J Phys Chem B 2010;114:1350-60.
[61] Huang CC, Hohn KL. Tetrakis(dimethylamino)ethylene chemiluminescence (TDE [83] Araos MU, Warr GG. Structure of nonionic surfactant micelles in the ionic liquid
CL) characterization of the CMC and the viscosity of reversed microemulsions. ethylammonium nitrate. Langmuir 2008;24:9354-60.
J Phys Chem B 2010;114:2685-94. [84] Harrar A, Zech O, Hartl R, Bauduin P, Zemb T, Kunz W. [emim][etSO(4)] as the Polar
[62] Haustein E, Schwille P. Fluorescence correlation spectroscopy: novel variations of phase in low-temperature-stable microemulsions. Langmuir 2011;27:1635-42.
an established technique. Annu Rev Biophys Biomol Struct 2007;36:151-69. [85] Gao Y, Hilfert L, Voigt A, Sundmachert K. Decrease of droplet size of the reverse
[63] Pal N, Verma SD, Singh MK, Sen S. Fluorescence correlation spectroscopy: an efficient microemulsion 1-butyl-3-methylimidazolium tetrafluoroborate/triton X-100/
tool for measuring size, size-distribution and polydispersity of microemulsion drop- cyclohexane by addition of water. J Phys Chem B 2008;112:3711-9.
lets in solution. Anal Chem 2011;83:7736-44. [86] Seth D, Chakraborty A, Setua P, Sarkar N. Interaction of ionic liquid with water with
[64] Takebayashi Y, Mashimo Y, Koike D, Yoda S, Furuya T, Sagisaka M, et al. Fourier trans- variation of water content in 1-butyl-3-methyl-imidazolium hexafluorophosphate
form infrared spectroscopic study of water-in-supercritical CO2 microemulsion as a ([bmim][PF6])/TX-100/water ternary microemulsions monitored by solvent and
function of water content. J Phys Chem B 2008;112:8943-9. rotational relaxation of coumarin 153 and coumarin 490. J Chem Phys 2007;126.
[65] Sagisaka M, Oike DK, Mashimo Y, Yoda S, Takebayashi Y, Furuya T, et al. [87] Liu LP, Bauduin P, Zemb T, Eastoe J, Hao JC. Ionic liquid tunes microemulsion
Water/supercritical CO(2) microemulsions with mixed surfactant systems. Lang- curvature. Langmuir 2009;25:2055-9.
muir 2008;24:10116-22.
[66] Takebayashi Y, Sagisaka M, Sue K, Yoda S, Hakuta Y, Furuya T. Near-infrared spec-
troscopic study of a water-in-supercritical CO2 microemulsion as a function of the
water content. J Phys Chem B 2011;115:6111-8.
 Indian Journal of Pharmaceutical Education and Research Association of Pharmaceutical Teachers of India

Microemulsions as Promising Delivery Systems: A Review

1 2 3 4 5 6
Vinod Singh* , Bushettii S.S , Appala Raju S , Rizwan Ahmad , Mamta Singh and Anupam Bisht
SBS PG Institute of Biomedical Sciences & Research , Dehradun, Uttrankhand.(India)1,5,6
HKE's College of Pharmacy, Gulbarga, Karnataka (India)2,3.
Oman medical college, Sohar, (Oman)4

ABSTRACT Submitted: 8-11-2010 Revised: 14-1-2011 Accepted: 25-4-2011

Microemulsions are clear, stable, isotropic liquid mixtures of oil, water and surfactant, frequently in combination with a cosurfactant. The
aqueous phase may contain salt(s) and/or other ingredients, and the "oil" may actually be a complex mixture of different hydrocarbons and
olefins. In contrast to ordinary emulsions, microemulsions form upon simple mixing of the components and do not require the high shear
conditions generally used in the formation of ordinary emulsions.
Microemulsion are having unique properties, namely, ultralow interfacial tension, large interfacial area, thermodynamic stability and the ability
to solubilize otherwise immiscible liquids. Microemulsion are having wide applications and uses such as in pharmaceuticals, cosmetics, cutting
oils, biotechnology, food , agrochemicals, environmental detoxification, analytical applications, microporus media synthesis etc.

INTRODUCTION Microemulsions containing the oil and aqueous phase,

surfactant and cosurfactant (cos), are optically transparent
It is well established that large amounts of two immiscible
liquids (e.g. water and oil) can be brought into a single phase mixtures with a very small droplet size (G140 nm).3,4
(macroscopically homogeneous but microscopically Microemulsions have been increasing in popularity and
heterogeneous) by addition of an appropriate surfactant or a garnering more attention in recent years, because they may
surfactant mixture. This unique class of optically clear, enhance the transdermal absorption of drug molecules by
thermodynamically stable and usually low viscous solutions, increasing drug solubilities and modifying their partition
called 'microemulsions', have been the subject of extensive coefficients.5 A hydrogel base is used very often in topical
research over the last two decades primarily because of their formulations.6,7 . The hydrogel formulation was prepared and
scientific and technological importance. studied as a vehicle for its permeation potential.

Microemulsions are clear, stable, isotropic liquid mixtures of Advantages :

oil, water and surfactant, frequently in combination with a a. Thermodynamically stability and require minimum
cosurfactant. Microemulsions are bicontinuous systems that energy for formation.
are essentially composed of bulk phases of water and oil
b. compatibility in manufacturing
separated by a surfactant/cosurfactant rich interfacial region.
These systems have advantages over conventional emulsions c. Enhanced drug solubilization and improved
in that they are thermodynamically stable liquid systems and bioavailability.
are spontaneously formed.1. Microemulsions are currently the d. Micro emulsion are having wide applications in
subject of many investigations because of their wide range of colloidal drug delivery systems for the purpose of
potential and actual utilizations. The high capacity of drug targeting and controlled release.
microemulsions for drugs makes them attractive formulations
Method of Preparation
for pharmaceuticals. These systems also offer several benefits
for oral administration, including increased absorption, 1. Phase Titration Method
improved clinical potency and decreased toxicity.2 Microemulsions are prepared by the spontaneous
emulsification method (phase titration method) and can be
*Address for Correspondence:
depicted with the help of phase diagrams. Construction of
Vinod Singh, Department of Pharmaceutics, SBSPG Institute of Biomedical Sciences and phase diagram is a useful approach to study the complex
Research, Balawala, Dehradun, Uttrakhand, India. series of interactions that can occur when different
E-mail: vinod.panipat@gmail.com components

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 392
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

are mixed. Microemulsions are formed along with various alkanols are called cosurfactants, they lower the interfacial
association structures (including emulsion, micelles, tension between oil and water
lamellar, hexagonal, cubic, and various gels and oily sufficiently low for almost spontaneous formation. The
dispersion) depending on the chemical composition and miscibility of oil, water and amphiphile (surfactant plus
concentration of each component. The understanding of their cosurfactant) depends on the overall composition which is
phase equilibria and demarcation of the phase boundaries are system specific.
essential aspects of the study.
When English chemist J.H. Schulman introduced the term
As quaternary phase diagram (four component system) is "microemulsion" in 1943 he described the transition from a
time consuming and difficult to interpret, pseudo ternary stable oil-rich mixture to a stable water-rich mixture.
phase diagram is often constructed to find the different zones Microemulsions contain a polar component, water, and a non
including microemulsion zone, in which each corner of the polar component, oil, which makes them capable of
diagram represents 100% of the particular component . The solubilizing a wide spectrum of substances. They measure in
region can be separated into w/o or o/w microemulsion by size from 3 to 300 nanometers in droplet diameter, are
simply considering the composition that is whether it is oil transparent and thermodynamically stable. Due to these
rich or water rich. Observations should be made carefully so special properties microemulsions offer a high potential for
that the metastable systems are not included. numerous practical applications. Consequently,
2. Phase Inversion Method microemulsions may be used for enhanced oil recovery,
cosmetic formulations, edible coatings for food, and for drug
Phase inversion of microemulsions occurs upon addition of
delivery systems as both transdermal or oral administrative
excess of the dispersed phase or in response to temperature.
vehicles for the controlled release of dosages.
During phase inversion drastic physical changes occur
Microemulsions also have industrial applications, one of
including changes in particle size that can affect drug release
them being the synthesis of polymers. Microemulsion
both in vivo and in vitro. These methods make use of
polymerization is a complex heterogeneous process where
changing the spontaneous curvature of the surfactant. For
transport of monomers, free radicals and other species (such
non-ionic surfactants, this can be achieved by changing the
as chain transfer agent, co-surfactant and inhibitors) between
temperature of the system, forcing a transition from an o/w
the aqueous and organic phases, takes place.8 Compared with
microemulsion at low temperatures to a w/o microemulsion at other heterogeneous polymerization processes (suspension or
higher temperatures (transitional phase inversion). During emulsion) microemulsion polymerization is a more
cooling, the system crosses a point of zero spontaneous complicated system. Polymerization rate is controlled by
curvature and minimal surface tension, promoting the monomer partitioning between the phases, particle
formation of finely dispersed oil droplets. This method is nucleation, and adsorption and desorption of radicals. Particle
referred to as phase inversion temperature (PIT) method. stability is affected by the amount and type of surfactant and
Instead of the temperature, other parameters such as salt pH of dispersing medium.
concentration or pH value may be considered as well instead Several authors have reported preparation of microemulsions
of the temperature alone. Additionally, a transition in the using alcohols of short or medium length chains (e.g.,
spontaneous radius of curvature can be obtained by changing butanol, heptanol or pentanol) as co-surfactants.9 These
the water volume fraction. By successively adding water into substances limit the potential application of microemulsion
oil, initially due to their toxic and irritant properties. A selection of
water droplets are formed in a continuous oil phase. components for microemulsions suitable for pharmaceutical
Increasing the water volume fraction changes the use involves a consideration of their toxicity and, if the
spontaneous curvature of the surfactant from initially systems are to be used topically, their irritation and sensitivity
stabilizing a w/o microemulsion to an o/w microemulsion at properties.10 The ionic surfactants are generally too toxic to be
the inversion locus. used for preparation of lipid emulsions; therefore, non ionic
Microemulsions can be prepared by controlled addition of surfactants, such as the poloxamers, polysorbates,
lower alkanols (butanol, pentanol and hexanol) to milky polyethylene glycol are preferred. Polysorbate 80 is widely
emulsions to produce transparent solutions comprising applied to pharmaceutical preparations, including ophthalmic
dispersions of either water-in-oil (w/o) or oil-in-water (o/w) preparations, due to its history of usefulness and safety, and it
in nanometer or colloidal dispersions (~ 100 nm). The lower is listed in the United States Pharmacopoeia- National

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 393
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

Formulary, the European Pharmacopoeia and the Japanese Fig.1: Microemulsion Forming System
Pharmacopoeia.11.. With the recent improvements in aseptic
processing and the availability of new well-tolerated
emulsifieres (polysorbate 80), emulsion technology is
currently under evaluation for topical cyclosporine A
delivery. Ding 12 developed a castor oil in water
microemulsion. This microemulsion is stabilized by
polysorbate 80 where the active substance cyclosporine A
remains stable over 9 months and causes only mild discomfort
and slight hyperemia on the rabbit eyes applied 8 times per
day during 7 days. This encouraging result allowed the
formulations to undergo clinical trials of phase II and III in dry
eye disease. The II phase trial performed on 162 patients
demonstrated good tolerance of the emulsion.13
Ternary and quaternary phase diagrams:
The knowledge on the phase manifestations of the pseudo- Fig. 2: Schematic representation of microemulsion
ternary (water/amphiphile/oil) or explicitly quaternary
microstructures
(water/surfactant/cosurfactant/oil) mixtures has been
systematized. At low surfactant concentration, there is a
sequence of equilibria between phases, commonly referred to
as Winsor phases14, they are
Winsor I: with two phases, the lower (oil/water, o/w)
microemulsion phase in equilibrium with the upper excess
oil;
Winsor II: with two phases, the upper microemulsion phase should be stressed that the o/w microemulsion droplets
(water/oil, w/o) in equilibrium with excess water; generally have a larger effective interaction volume than w/o
droplets. Also, whereas emulsions consist of roughly
Winsor III: with three phases, middle microemulsion phase
spherical droplets of one phase dispersed into the other,
(o/w plus w/o, called bicontinuous) in equilibriumwith upper
microemulsions constantly evolve into various structures
excess oil and lower excess water;
ranging from ''droplet-like'' swollen micelles to bicontinuous
Winsor IV: in single phase, with oil, water and surfactant structures, frequently making the usual o/w and w/o
homogeneously mixed. distinctions irrelevant. Because the size of the particles is
Inter-conversion among the above-mentioned phases can be much smaller than the wavelength of visible light,
achieved by adjusting proportions of the constituents. microemulsions are transparent and their structure cannot be
Simultaneous presenceof two microemulsion phases, one in observed through an optical microscope.
contact with water and the other in contact with oil is also The extents of formation of w/o, o/w and bicontinuous
possible. 15. This may be considered as an extension of microemulsions can be understood from the phase
Winsor's classification forming the fifth category. A equilibrium studies. Such studies may often become complex
microemulsion forming systems is shown in figure 1. with the appearance of tiny or extended additional zones of
Another important consideration in the formation of viscous gel and liquid crystalline phases 16 , the establishment
microemulsions is related to the packing parameter, which is of their boundary demarcations is time consuming and
important for structures with high curvatures (Fig. 2). laborious.
Surfactants must have the proper molecular volume Most commonly used methods and techniques to acquire
dimensions and proportions to effectively pack into a micellar information on the particle dimension and shape, their
structure. Oil phases with high molecular volume fractions diffusion coefficient and polydispersity, aggregation and
(such as triglycerides) will pack less efficiently and will have dynamics of coalescence, state of the water pool,
difficulties in entering between the surfactant tails. This is thermodynamics of formation, etc. of the compartmentalized
also reflected in the isotropic regions of a phase diagram. It systems of microemulsion are related to conductance,

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 394
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

viscosity, ultrasound, static and dynamic light scattering, polymerization kinetics, the most characteristic feature of
neutron and low angle X-ray scattering, nuclear magnetic which is the compartmentalization, where the radicals
resonance, growing inside the particles are separated from each other,
dielectric relaxation, time resolved fluorescence quenching, thus suppressing termination to a high extent and, as a
transmission electron microscopy, calorimetry, etc. consequence, providing high rates of polymerization.

Differences between Emulsion & Microemulsion Microemulsions can be sterilized by filtration and their
production is relatively simple and inexpensive. Because of
In contrast to ordinary emulsions, microemulsions form upon
these properties, they have attracted a great interest as drug
simple mixing of the components and do not require the high
delivery vehicles17,18. Microemulsions can be applied as liquid
shear conditions generally used in the formation of ordinary
emulsions. The two basic types of microemulsions are direct membrane carriers to transport lipophilic substances through
(oil dispersed in water, o/w) and reversed (water dispersed in an aqueous medium or to carry hydrophilic substances across
oil, w/o). lipoidal medium. They are proposed for oral, topical, dermal,
transdermal, parentenal and pulmonary administration of
The main difference between emulsions and microemulsions
drugs19. Although microemulsions have been known for a
is in the size and shape of the droplets that are dispersed in the
long period, their potential as vehicles for topical ocular drug
continuous phase, reflecting the differences in the
delivery has been investigated only within the last decade20.
thermodynamic stability of the two systems (Table 1).
The main problem in a microemulsion application is a high
Emulsions are kinetically stable but thermodynamically
unstable, and after storage or aging, droplets will coalesce and concentration and a narrow range of physiologically
the two phases separate. In contrast, microemulsions are acceptable surfactants and co-surfactants. 21,22. On the other
thermodynamically stable and will not separate into the hand, the large surfactant concentration determines their
corresponding phases. It should be stressed that the term stability.
''mini-emulsions'' was coined by some authors to describe Factors Affecting the Microemulsion:
emulsion droplets of submicron size with improved
The formation of microemulsion will depend on the following
stabilities, other scientists may call those emulsions
factors
''nanoemulsions.'' While nanoemulsions do not have a long
shelf life, they frequently are freshly prepared and used. It a. Packing ratio: The HLB of surfactant determines the type
should also be stressed that in some studies, the authors of microemulsion through its influence on molecular packing
neglect to test stability and consider mini- or nanoemulsions and film curvature. The analysis of film curvature for
to be true microemulsions. The kinetics of microemulsion surfactant association's leadings to the formation of
polymerization has much in common with emulsion microemulsion.

Table 1: Difference between miniemulsion, macroemulsion and microemulsion.

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 395
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

b. Property of surfactant, oil phase and temperature: The intermediate temperature, microemulsion coexists with
type of microemulsion depends on the nature of surfactant. excess water and oil phases and forms bicontinuous structure.
Surfactant contains hydrophilic head group and lipophilic tail c. The chain length, type and nature of cosurfactant: Alcohols
group. The areas of these group, which are a measure of the
are widely used as a cosurfactant in microemulsions. Addition
differential tendency of water to swell head group and oil to
of shorter chain cosurfactant gives positive curvature effect as
swell the tail area are important for specific formulation when
alcohol swells the head region more than tail region so, it
estimating the surfactant HLB in a particular system. When a
becomes more hydrophilic and o/w type is favoured, while
high concentration of the surfactant is used or when the
longer chain cosurfactant favours w/o type w/o type by
surfactant is in presence of salt, degree of dissociation of polar
alcohol swelling more in chain region than head region.
groups becomes lesser and resulting system may be w/o type.
Diluting with water may increase dissociation and leads to an Evaluation / Characterization of microemulsion
o/w system. Ionic surfactants are strongly influenced by The microemulsions are evaluated by the following
temperature. It mainly causes increased surfactant counter- techniques. They are
ion dissociation. The oil component also influences curvature
(I) Phase behavior studies: visual observations, phase
by its ability to penetrate and hence swell the tail group region
contrast microscopy and freeze fracture transmission electron
of the surfactant monolayer. Short chains oils penetrate the
microscopy can be used to differentiate microemulsions and
lipophilic group region to a great extent and results in
coarse emulsions. Clear isotropic one-phase systems are
increased negative curvature. Temperature is extremely
identified as microemulsions whereas opaque systems
important in determining the effective head group size of
showing bifringence when viewed by cross polarized light
nonionic surfactants. At low temperature, they are
hydrophilic and form normal o/w system. At higher microscopy may be taken as liquid crystalline system.
temperature, they are lipophilic and form w/o systems. At an (II) Scattering Techniques: Scattering techniques such as

Some Research Work on Microemulsions

Drug Name Route Purpose/Result
Acyclovir 23 oral Improved bioavailability
Flurbiprofen 24 Parenteral Increased the solubility
Apormorphine HCL 25 Transdermal Increased the permeability
Ketoprofen 26 Transdermal Enhancement of permeability
Prilocainne-HCL 27 Transdermal Increased the solubility
Estradiol 28 Transdermal Improvement in solubilization
Aceclofenac 29 Dermatological Increased the solubility
Piroxicam 30 Oral Increased the solubility
Diclofenac 31 Transdermal Permeability enhancement
Dexamethasone 32 Topical Ocular Enhanced the Bioavailability
Carbamazepine 33 Intranasal Enhanced bioavailability
Chloramphenicol 34 Ocular Increased the solubility
Ibuprofen 35 Parenteral Increased the solubility
Sumatriptan 36 Intranasal Enhanced the Bioavailability
Ibuprofen 37 Topical Increasing the solubility
Clonixic acid 38 transdermal For better absorption
Itraconazole 39 Parenteral For better absorption
Timolol 40 Ophthalmic For better absorption
Terbinafine 41 Transdermal Permeability enhancement
Fenofibrate 42 Self-Micro emulsifying Increasing the solubility
Progesterone 43 Dermal Increased the chemical Stability
Clopidogrel 44 oral Solubility enhancement

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 396
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

small angle neutron scattering, small angle X-ray scattering respect to its individual ingredients.
and light scattering have found applications in studies of (IX)Drug release studies
microemulsion structure, particularly in case of dilute
monodisperese spheres, when polydisperse and/or (A) In-vitro drug release
concentrated systems such as those frequently seen in The diffusion study can be carried out on a modified Franz
microemulsions. diffusion cell, within volume of 20mL. The receptor
(III) Transmittance test compartment was filled with of buffer .The donor
compartment was fixed with cellophane membrane ,
Stability of the optimized microemulsion formulation with containing the microemulsion formulation and the plain drug
respect to dilution was checked by measuring transmittance at solution, separately. At predetermined time intervals samples
a specific wavelength with a UV spectrophotometer . were withdrawn from the receptor compartment and analyzed
(IV) Globule size and zeta potential measurements for drug content, using a UV spectrophotometer at specific
The globule size and zeta potential of themicroemulsion can wavelength.
be determined by dynamic light scattering, using a Zetasizer (B) Ex-vivo drug release
HSA 3000. Ex -vivo drug release into buffer was studied using intestinal
(V) Viscosity measurements membrane within a Franz diffusion cell. Microemulsion
formulation and plain drug solution were placed in the donor
Rheological behavior of the formulation can be observed by
compartment of two separate diffusion cells and the
using a Brookfield LVDV ΙΙΙ+ cone and plate (CP) viscometer
temperature of each cell was maintained at 37 ±2°C. The
(Mfg: Brookfield, USA) using rheocal software at a
amount of drug released from the microemulsion formulation
temperature. Change in the rheological characteristics help
can be estimated spectrophotometrically at specific
in determining the microemulsion region and its separation
wavelength, by withdrawing samples from the receptor
from other region. Bicontinuous microemulsion are dynamic
compartment at predetermined time intervals.
structures with continuous fluctuations occurring between the
bicontinuous structure, swollen reverse micelle, and swollen Applications of microemulsions
micelles There has been a revolution in the last two decades in the
(VI) Electrical conductivity utilization of microemulsion systems in a variety of
pharmaceutical, chemical, industrial processes etc.
The water phase was added drop wise to a mixture of oil,
surfactant and co-surfactant and the electrical conductivity of Microemulsion in pharmaceutical
formulated samples can be measured using a conductometer Liquid crystalline, miceller and emulsion forming systems
(CM 180 conductivity meter, Elico, India) at ambient are widely used in pharmaceutical preparations. The easy
temperature and at a constant frequency of 1 Hz. formation, remarkable environment independent stability,
(VII) Drug stability excellent solubilization capacity, etc. favour microemulsions
to be a better proposition over other compartmentalized
The optimized microemulsion was kept under cold condition
systems. The dispersed phase, lipophilic or hydrophilic(o/w
(4-8 o C), room temperature and at elevated temperature (50 ±
or w/o type) can act as a potential reservoir of lipophilic or
2 o C). After every 2 months the microemulsion can be
hydrophilic drugs that can be partitioned between the
analyzed for phase separation, % transmittance, globule size
dispersed and the continuous phases. Coming in contact with
and % assay.
a semipermeable membrane, such as skin or mucous
(VIII) Drug solubility membrane, the drug can be transported through the barrier. 45.
Drug was added in excess to the optimized microemulsion Both lipophilic and hydrophilic drugs can be administered
formulation as well as each individual ingredient of the together in the same preparation.
formulation. After continuous stirring for 24 h at room Low viscous formulations using microemulsions with
temperature, samples were withdrawn and centrifuged at suitable protein compatible surfactants can be used as
6000 rpm for 10 min. The amount of soluble drug in the injection solutions, for they are miscible with blood in any
optimized formulation as well as each individual ingredient of ratio. In contrast to emulsions, microemulsions cause
the formulation was calculated by subtracting the drug minimum immuno reactions or fat embolism. Proteins are not
present in the sediment from the total amount of drug added. denatured in microemulsions although they are unstable at
The solubility of drug in microemulsion was compared with high or low temperatures. The total dose of the drug can be

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 397
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

reduced when applied through the microemulsion route and polyorganosiloxane (a nonionic surfactant) and an acid and/or
thus side effects can be minimized. a metal salt.
Toxicity, bio-incompatibility of surfactants and Cosmetic microemulsions (transparent and translucent) of
cosurfactants, requirement of high concentrations for silicone oils was produced by emulsion polymerization
formulations and other relevant factors such as maintenance tehnique. Ultrafine emulsions prepared by condensation
of thermodynamic stability in the temperature range between method have some advantages in cosmetic and medical
0 0 C and 40 0 C, salinity, constant pressure during storage, low products, as they have excellent stability and safety and their
solubilizing capacity for high molecular weight drug (and droplet size can be readily controlled. Ultrafine emulsions can
oil), etc. limit the uses of microemulsions in the be regarded as thermodynamically unstable microemulsions,
pharmaceutical and medicinal fields. as they are o/w emulsions with droplet size similar to
An application of o/w microemulsion in the pharmaceutical microemulsion. Tokuoka et al.48 studied the solubilization of
industry is the use of strongly hydrophobic fluorocarbons (as several systems consisting of water, surfactant and synthetic
oils) to produce short-time blood plasma substitutes to perfumes (viz. d-limonene, a-ionone, benzyl accetate, linalol,
maintain the supply of oxygen in the living systems. The eugenol and a-hexylcinnamaldehyde), clarifying (a) the
components to be used must have low allergic potential, good influence of fragrance structure on the phase regions in a
physiological compatibility and high biocompatibility. The water/nonionic surfactant systems, (b) the distribution co-
biocompatibility requirements of the amphiphiles are efficient between micelles and the bulk phase, and (c) the
fulfilled by lecithins, non-ionic surfactants (Brijs, Arlacel partition between dissolved and solubilized perfume
186, Spans and Tweens). components on their volatility. In this, the phase equilibria in
water, lecithin, soybean oil and vanillin have been studied.
Garcia-Celma 45 has reviewed microemulsions as drug
delivery systems for different types of drugs, viz. Microemulsions in analytical applications
antineoplastics/ antitumour agents (doxorubicin, idarubicin, Microemulsions are widely used in the field of analytical
tetrabenzamidine derivative), peptide drugs (cyclosporine, techniques such as chromatography, laser-excited
insulin, vassopressin), sympatholytics (bupranolol, timolol, photoionization spectroscopy, etc. In microemulsion
levobunolol, propanolol), local anesthetics (lidocaine, electrokinetic chromatography (MEEKC), characterization
benzocaine, tetracaine, heptacaine), steroids (testosterone,
of solute hydrophobicity was carried out 49 , which provides a
testosterone propionate, testosterone enanthate,
quick and reproducible method to obtain hydrophobic
progesterone, medroxyprogestorane acetate),
parameters for solvents. Microemulsions are able to enhance
anxiolytics(benzodiazepines), antiinfective drugs
analytical spectroscopic techniques by functioning as
(cloitrimazole, ciclopirox olamine, econazole nitrate,
solubilized media, spectral shift reagents, intensity
tetracycline hydrochloride), vitamins (menadione, ascorbic
amplification agents, etc. The utilization of microemulsion
acid), anti-inflammatory drugs(butibufen, indomethacin),
media in analytical spectroscopy and the analytical
and dermological products( tyrocine, azelaic acid, octyl
sensitivities of the three systems o/w, w/o and bicontinuous
dimethyl PABA, 2- ethyl hexyl p-methoxy cinnamate).
microemulsion have been assessed. A series of studies have
Enzyme doped silica nanoparticles (ceramic drug carrier) in been reported on the determination of aluminium, zinc,
the aqueous core of reverse micelles and microencapsulation copper, cadmium, manganese ions using both microemulsion
of diospyrin, a plant-derived bisnapthoquinol of potential and mixed microemulsion systems. These studies are mostly
chemotherupic activity have been very recently reported. 46 published in the journals published by the Chinese Chemical
Microemulsions in cosmetics Society.50
It is believed that microemulsion formulation will result in a Microemulsions in biotechnology
faster uptake into the skin. Cost, safety, appropriate selection Recently, interest on microemulsions is being focused for
of ingredients are key factors in the formulation of various applications in biotechnology, viz, enzymatic
microemulsions. Skin care microemulsions contain 47, sodium reactions, immobilization of proteins and bioseparation.
alkyl sulfate, tetraethylene glycol monododecyl ether, Microemulsions are advantageous over other multiphase
lecithin, dodecyl oligoglucoside, alkyl dimethyl amine oxide, equilibrium systems because of simultaneous solubilization
propanol,hexadecane, isopropyl myristate have been used as of polar and nonpolar reactants in the same solution, shifting
surfactants, cosurfactants and oils respectively. of the equilibrium position of the reaction and the separation
Hair care microemulsions contain an amino-functional of products by physical means. However, bio-incompatibility

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 398
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

of the amphiphiles used poses a serious limitation in the The crystalline colloidal array self assemble into either face
advancement of this field. The prospects of biotechnological centered or body centered cubic form. Just as atomic crystals
applications have also been reviewed 51,52. Enzyme reactions diffract X-rays that fulfill the Bragg condition, CCAs diffract
(catalysis) in microemulsion media have widely been studied. ultraviolet, visible and near-infrared light, depending on the
The use of microemulsion for enzyme catalysis is not lattice spacing. Colloidal particles of inorganic materials,
arbitrary for enzymes under in vivo condition function in the such as silica or organic polymers, such as poly (N-
cell as well as at the interface of hydrophobic and hydrophilic isopropylacrylamide) have been synthesized having
domains of cell and tissue containing lipids and other natural periodicity of the order of ~200 nm. Asher et al. 58 and Holtz et
amphiphiles. al.59 have developed a novel sensing material from a
Enzymatic reactions in microemulsions: polymerized crystalline colloidal array (PCCA) which is a
The potential advantages of employing enzymes in media of mesoscopically periodic crystalline colloidal array of
low water content, i.e. w/o microemulsions are: (i) increased spherical polystyrene colloids within a thin, intelligent
solubility of nonpolar reactants; (ii)possibility of shifting polymer hydrogel film. They have fabricated a sensor,
thermodynamic equilibria in favour of condensation; (iii) utilizing a crown ether as the recognization agent that can
improvement of thermal stability of the enzymes, enabling detect Pb2+ in the 0.1  M– 20 mM (~20 ppb – ~ 400 ppm)
reactions to be carried out at higher temperature. Catalysis by concentration range. The sensors for glucose and galactose
a large number of enzymes in microemulsion media has been utilising glucose oxidase or b-D-galactosidase as the
studied for a variety of reactions, such as synthesis of esters, recognization entities have been developed. Besides sensing
peptides and sugar acetals; transesterifications; various glucose, this sensor can estimate dissolved oxygen
hydrolysis reactions; glycerolysis; oxidation and reduction concentration in the presence of constant glucose
and steroid transformation. The conformation and activity of concentration. Development of thermally tunable photonic
an enzyme depend on  ([water]/[surfactant]); the enzyme is crystal of poly (N-isopropylacrylamide) (PNIPAM), a novel
thus sensitive to amount of surrounding water. CCA photoionic crystals with variable sphere sizes and
variable array periodicity and sensors that change volume in
Gomez-Puyon has carried the work on behaviour of enzymes
response to nonionic molecular recognition processes such as
in microemulsions53
antibody/antigen interactions have been attempted.
Immobilization of protein in microemulsion: In the field of
protein immobilization, microemulsion medium has been CONCLUSION
found to be a good proposition. Immobilization of a variety of Microemulsions are having a vast and significant potential in
proteins on suitable solid surfaces using microemulsion drug delivery as well as in the industrial process. Researchers
media has been successfully carried out54. are working in this field for drug release, coatings, dyes,
Microemulsions for bioseparations: The possibility of agrochemicals and in enzyme reaction. In the future
microemulsions to extract biopolymers (proteins and prospects, microemulsions will be used in synthesis of
enzymes) from an aqueous phase has been explored. nanoparticles and as a industrial chemical sensors. The role of
Microemulsions are gentle solvents for extraction of proteins microemulsion in providing novel solutions to overcome the
without altering their enzymatic or functional properties problems of poor aqueous solubility of highly lipophilic drug
although the process can readily be scaled by conventional compounds and provide high, more consistent and
liquid–liquid extraction techniques. The pH, ionic strength, reproducible bioavailability. Furthermore, these formulations
type of salt, concentration of solvent and temperature can be easily manufactured in term of the relative cost of
influence the partition of a protein.55,56 commercial production. Topical products are now employing
the microemulsion technology are likely to emerge.
Microemulsion as a chemical sensor materials.
Microemulsions can also be used to achieve drug targeting
Microemulsions as novel crystalline colloidal arrays (CCA) however challenges remain, primarily because of the layers of
are new findings which acts as novel chemical sensors. A barriers that these systems need to overcome to reach to the
intelligent photoionic crystalline colloidal array self target. Recent research work is focused on the production of
assemblies have been developed, 57 which can have use in safe, efficient and more compatible microemulsion
medicine, environmental chemistry, process control and constituents which will further enhance the utility of these
remote sensing. These are mesoscopically periodic fluid novel vehicles. Microemulsion in today's world can be
materials, that diffract light satisfying the Bragg condition.
accepted as full of potential in a novel drug delivery systems.

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 399
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

REFERENCES 18. Pattarino F., Marengo E., Gasco M. R., Carpignano R.

1. Cortesi R and Nastruzzi C. Liposomes, micelles and Int. J. Pharm 1993; 91, 157 .
microemulsions as new delivery systems for cytotoxic 19 Tenjarla S: Therapeutic Drug Carrier Systems 1999; 16:
alkaloids. PSTT 1999; 2: 288-98 461
2 Baße A and Keipert S. Development and 20. Vandamme T. Prog. Retin. Eye Res 2002; 21: 15
characterization of microemulsions for
21. Corswant Ch., Thoren P., Engstrom S. J. Pharm.
ocularapplication. Europ. J. Pharm. Biopharm 1997;
Sci1998; 87: 200
43: 179-83
22. Aboofazeli R., Patel N., Thomas M., Lawrence M.J. Int.
3. Paul BK, Moulik SP. Microemulsions: an overview. J
J. Pharm 1995; 125: 107
Disper Sci Technol 1997;18: 301 -306.
23. Ghosh PK, Majithiya RK, Umrethia ML, and Murthy
4. Prince LM. Microemulsions. In: Lissant KJ (ed).
RSR. Design and Development of Microemulsion Drug
Emulsions and Emulsion Technology. New York :
Delivery System of Acyclovir for Improvement of Oral
Marcel Dekker; 1974. 125-178.
Bioavailability. AAPS PharmSciTech 2006; 7 (3)
5. Delgado-Charro MB, Iglesias-Vilas G, Blanco-Mendez Article 77
J, Lopez- Quintela MA, Marty JP, Guy RH. Delivery of a
24. Park KM and Kim CK. Preparation and evaluation of
hydrophilic solute through the skin from novel
flurbiprofen-loaded microemulsions for parental
microemulsion systems. Eur J Pharm Biopharm 1997;
delivery. Int.J.Pharm 1999; 181: 173-179.
43: 37-42.
6. Celebi N, Kıslal O, Tarımcı N. The effect of β- 25. Peira E, Scolari P and Gasco MR. Transdermal
cyclodextrin and penetration additives on the release of permeation of apomorphine through hairless mouse skin
naproxen from ointment bases. Pharmazie 1993; 48: from microemulsion. Int. J. Pharm 2001; 226: 47-51.
914-917. 26. Rhee YS, Choi JG, Park ES and Chi SC. Transdermal
7. Gupta P, Vermani K, Garg S. Hydrogels from controlled delivery of ketoprofen using microemulsions.
release to pH-responsive drug delivery. Drug Discov Int.J.Pharm2001; 228: 161-170.
Today 2002; 7: 569-579. 27. Kreilgard M, Peedersen EJ and Jaroszewski JW. NMR
8. Turner SR, Siano DB and J. Bock,A. Microemulsion characterization and transdermal drug delivery potential
Process for Producing Acrylamide-Alkyl Acrylamide of microemulsion system. J. Controlled Rel 2000; 69:
Copolymers, U. S. Patent No. 4,521,580, June 1985. 421-433.
9. Gasco M R, Gallarate M, Trotta M, Bauchiero L, 28. Peltola S, Saarinen SP, Kiesavaara J and Urttia STM.
Gremmo E, Chiappero O. J. Pharm. Biomed. Anal 1989; Microemulsions for topical delivery of estradiol. Int. J.
7: 433 Pharm 2003; 254: 99-107
10. Siebenbrodt I., Keipert S.: Eur. J. Pharm. Biopharm 29. Yang JH, Kim YI and Kim KM. Preparation and
1993; 39: 25 evaluation of aceclofenac microemulsions for
11. Yamaguchi M, Yasueda S, Isowaki A, Yamamoto M. et transdermal delivery system. Arch.Pharm. Res 2002;
al. Int. J. Pharm 2005; 301: 121 25: 534-540.

12. Ding S.: U.S. Patent 5, 474, 979, (1995). 30. Andrade SM and Costa SM. Fluorescence quenching of
acridine orange in microemulsions induced by the non-
13. Lallemand F.: Eur. J. Pharm. Biopharm 2003; 56: 307
steroidal anti inflammatory drug piroxicam.
14. Winsor P A. Solvent Properties of Amphiphilic Photochem. Photobiol. Sci 2003; 2: 605-610.
Compounds, Butterworth, London, 1954.
31. Kweon, J.H., Chi, S.C. and Park, E.S. 2004.
15. Kunieda H, Asaoka H and Shinoda K. J. Phys. Chem Transdermal delivery of diclofenac us ing
1988; 92: 185 microemulsions. Arch. Pharm. Res.27: 351-356.
16. Mukherjee K, Mukherjee D C and Moulik S P. J. Colloid 32. Fialho SL and Cunha DS. New vehicle based on a
Interface Sci 1997; 187: 327. microemulsion for topical ocular administration of
17. Aboofazeli R., Lawrence M.J.: Int. J. Pharm 1993; 93: dexamethasone. Clin. Experiment Ophthalmol 2004.
161 32: 626-632.

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 400
 Vinod Singh et al Microemulsions as a promising delivery systems: A Review

33. Surjyanarayan Mandal, Snigdha Das Mandal. Design 44. Vandana B Patel, Hirenkumar D Kukadiya, Rajshree
and development of carbamazepine mucoadhesive Mashru, Naazneen Surti and Surjyanarayan Mandal.
microemulsion for intranasal delivery: an ex-vivo study. Development of Microemulsion Formulation for the
International Journal of Pharmaceutical Sciences Solubility Enhancement of Clopidogrel. Iranian
Review and Research 2010; 3(1), 56-60. Journal of Pharmaceutical Research 2010; 9 (4): in press
34. Lv FF, Zheng LQ and Tung CH. Phase behavior of the 45. Solans C and Kunieda H (eds), Industrial Applications
microemulsions and stability of the chloramphenicol in of Microemulsions, Marcel Dekker Inc., New York,
microemulsion based ocular drug delivery system. Int.
1997; Tadros, Th. F., p. 199; Dungan, S. R., pp.
J. Pharm2005; 14: 237-246.
147–174; Gasco, M. R., pp. 97–122; Garcia-Celma, M.
35. Zhao X, Chen D, Gao P, Ding P and Li K. Synthesis of J., pp. 123–145; Holmberg, K., pp. 69–95.
ibuprofen eugenol ester and its microemulsion
46. JainT K, Roy I, De T K and Maitra A N. J. Am. Chem.
formulation for parental delivery. Chem. Pharm. Bull
Soc 1998; 120: 11092
2005; 53:1246-1250.
36. Vyas TK, Babbar AK, Sharma RK, Singh S and Misra 47. Shinoda K, Shibata Y and Lindman B. Langmuir 1993;
A. 2006. Preliminary brain targeting studies on 9:1254.
intranasal mucoadhesive microemulsions of 48. Tokuokas Y, Uchiyama H, Abe M and Christian S D.
sumatriptan. AAPS Pharm. Sci. Tech 2006; 20: E8. Langmuir 1995; 11: 725.
37. Chen H, Chang X, Du D, Li J, Xu H and Yang X. 49. Ishihama Y, Oda Y, Uchikawa K and Asakawa N. Anal.
Microemulsion based hydrogel formulation of Chem 1995; 67: 1588
ibuprofen for topical delivery. Int. J. Pharm 2006; 315:
52-58. 50. Guo R and Zhu X J. Chin. Univer 1987; 8: 508.

38 Jung-Mi Lee, Kyung-Mi Park, Soo-Jeong Lim, Mi- 51. Levashov A V, Khmelnitsky Y L, Klyachko N L,
Kyung Lee, Chong-Kook Kim Microemulsion Chernyak V Y and Martinek K J. Colloid Interface Sci
formulation of clonixic acid: solubility enhancement 1992; 88: 444.
and pain reduction. Journal of Pharmacy and 52. Larsson K, Lipids: Molecular Organisation, Physical
Pharmacology 2002; 54(1): 43–49 Functions and Technical Applications, Oily Press,
39. Rhee YS, Park CW, Nam TY, Shin YS, Chi SC and Park Dundee, Scotland, 1994; ch. 9.
ES. Formulation of parental microemulsion containing 53. Gomez-Puyon A. (ed.) Biomol. Org. Solvents, CRC
itraconazole. Arch Pharm. Res 2007; 30: 114-123.
Press, Boca Raton, 1992.
40. Li C C, Abrahamson M, Kapoor Y and Chauhan A.
54. Holmberg K, Bergstrom K, Brink C, Osterberg E,
Timolol transport from microemulsions trapped in
Tiberg F and Harris J. Adhesion Sci. Technol 1993;7:
HEMA gels. J. Colloid Interface Sci 3007; 315: 297-
503
306.
55. Kelley B D, Wang D I C and Hatton T A. Biotechnol.
41. Baboota S, AL-Azaki A, Kohli K, Ali J, Dixit N and
Shakeel F. Development and evaluation of a Bioeng 1993, 42: 1199 -1209.
microemulsion formulation for transdermal delivery of 56. Adachi M, Harada M, Shioi A and Sato Y. J. Phys.
terbinafine. PDA J. Pharm. Sci. Technol 2007; 61: 276- Chem 1991; 95: 7925.
285.
57. Holtz J H and Asher S A. Nature 1997; 389: 829
42. Patel AR and Vavia PR. Preparation and in vivo
58. Asher S A, Weissman J M, Sunkara H B, Pan G, Holtz
evaluation of SMEDDS containing fenofibrate. AAPS
J, Liu L and Kesavamoorthy R. In: Polymers for
2007; 9: E344.
Advanced Optical Applications (eds) Jenekhe S A and
43. Biruss B and Valenta C. The advantage of polymer Wynne K J. Washington DC 1997.
addition to a non-ionic oil in water microemulsion for
the develop delivery of progesterone. Int.J.Pharm 2008; 59. Holtz J H, Holtz J S W, Munro C H and Asher S A. Anal.
349: 269-273. Chem 1998; 70: 780.

********

Ind J Pharm Edu Res, Oct-Dec, 2011/ Vol 45/ Issue 4 401
 Indian Journal of Pharmaceutical Education and Research Association of Pharmaceutical Teachers of India

Microemulsions For Topical Use– A Review

Nirmala Grampurohit, Padmini Ravikumar* and Rashmi Mallya
Dr. Bhanuben Nanavati College of Pharmacy, Gate No.1, Mithibai College Campus, V.M. Road,Vile Parle (West), Mumbai - 400 056.

ABSTRACT Submitted: 20/7/2009 Revised: 19/9/2009 Accepted: 1/4/2010

Microemulsions have emerged as novel vehicles for drug delivery which allow sustained or controlled release for percutaneous, peroral, topical,
transdermal, ocular and parenteral administration of medicaments. They offer the advantage of spontaneous formation, ease of manufacturing
and scale-up, thermodynamic stability, improved drug solubilization of hydrophobic drugs and bioavailability. While microemulsions are used in
several fields, this article focuses on the reported investigations for topical applications which exhibit minimal systemic absorption.

KEYWORDS : Microemulsion, topical delivery , co-surfactant, antioxidant.

INTRODUCTION hydrophobic drugs and bioavailability. Also,

microemulsions that have inverse micellar structure may be
Topical preparations pertain to medicaments applied to the
less comedogenic than either creams or solutions3.
surface of a part of the body and is a term used to describe
formulations that have effects only in a specific area of the 4
Hoar and Schulman introduced the word microemulsion,
body and are formulated in such a manner that the systemic which they defined as a transparent solution obtained by
absorption of the medicament is minimal. The methods titrating a normal coarse emulsion with medium-chain
involved in conventional topical drug delivery basically alcohols. Microemulsions are thermodynamically stable
involve either assisting or manipulating the barrier function of isotropic systems in which two immiscible liquids (water and
the skin (topical antibiotics, antibacterials, emollients, oil) are mixed to form a single phase by means of an
sunscreen agents) or breaching the horny layer at the appropriate surfactant or its mixture. The short to medium
molecular scale so as to direct drugs to the viable epidermal chain alcohols are generally considered as co-surfactants in
and dermal tissues without using oral, systemic or other the microemulsion system. The presence of surfactant and co-
therapies. surfactant in the system makes the interfacial tension very
There has been a revolution in the last two decades in the low. Therefore microemulsions form spontaneously, with an
5
utilization of microemulsion systems in a variety of chemical average droplet diameter of 10 to 140 nm .
and industrial processes. Microemulsions have shown a wide Microemulsions have the ability to deliver larger amounts of
range of applications starting with enhanced oil recovery in water and topically applied agents into the skin than water
the 70's, expanding to a wide range of chemicals and entering alone or other traditional vehicles such as lotions or creams
the pharmaceutical and cosmetic formulation area a decade because they act as a better reservoir for a poorly soluble drug
1
ago . through their capacity for enhanced solubilization .
6

Microemulsions have advantages over both colloidal systems

The main difference between macroemulsions and
under investigation and conventional emulsions, suspensions
microemulsions lies in the size and shape of the particles
and micellar solutions and may provide alternative drug
2
dispersed in the continuous phase: these are at least an order of
carriers . They are promising delivery systems which allow
magnitude smaller in the case of microemulsions (10-200 nm)
sustained or controlled drug release for percutaneous, peroral,
than those of conventional emulsions (1-20 m).
topical, transdermal, ocular and parenteral administration of
Macroemulsions consist of roughly spherical droplets of one
medicaments. They offer the advantage of spontaneous
phase dispersed into the other whereas microemulsions
formation, ease of manufacturing and scale-up,
constantly evolve between various structures ranging from
thermodynamic stability, improved drug solubilization of
droplet-like swollen micelles to bicontinuous structures,
Address for Correspondence:
making the usual “oil in water” and “water in oil” distinction
Padmini Ravikumar, Dr. Bhanuben Nanavati College of Pharmacy, Gate No.1, Mithibai
College Campus, V.M. Road,Vile Parle (West), Mumbai - 400 056. sometimes irrelevant.
E- mail: pdmnravikumar@yahoo.com
In topical formulations, microemulsions have been proved to

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 100
 Padmini et al.: Microemulsions For Topical Use– A Review.

13,14
increase the cutaneous absorption of both lipophilic and surfactants from the toxicity point of view . The
hydrophilic medicaments when compared to conventional biocompatibility requirements of the amphiphiles are
vehicles (emulsions, pure oils, aqueous solutions). In an fulfilled by lecithins and non-ionic surfactants (Brijs, Arlacel
7 1
extensive review of this type of applications, Kreilgaard et al 186, Spans, Tweens and AOT) .
attribute this performance to a generally higher solubility of
The following examples are commonly used formulations
the medicaments in microemulsions, generating an increased
components of microemulsions
concentration gradient towards the skin. The role of
penetration enhancers played by the amphiphilic components - Oil: Ethyl oleate, Mineral oil, Isopropyl myristate,
of the microemulsion and the internal mobility of the drug Decanol, Oleic acid, Vegetable oils (Coconut oil, Safflower
within the vehicle also contribute to the overall performance oil, Soyabean oil, Olive oil), Medium chain length
of microemulsions in dermal or transdermal drug delivery. triglyceride (Mygliol 812).

While microemulsions are used in several fields, in this - Surfactant: Polysorbate (Tween 80 and Tween 20),
review an attempt has been made to emphasize on the Lauromacrogol 300, Lecithins, Decyl polyglucoside
reported studies for topical applications which exhibit (Labrafil M 1944 LS), Polyglyceryl-6-dioleate (Plurol
minimal systemic absorption. Oleique), Dioctyl sodium sulfosuccinate (Aerosol OT), PEG-
8 caprylic/capril glyceride (Labrasol).
Formulation Considerations
The challenges in formulating topical microemulsions are: - Co-surfactant: Sorbitan monooleate, Sorbitan

1. Determining systems that are non-toxic, non-irritating, monostearate, Propylene glycol, Propylene glycol
non-comedogenic and non-sensitizing. monocaprylate (Capryol 90), 2-(2-ethoxyethoxy)ethanol
(Transcutol P) and Ethanol.
2. Formulating cosmetically elegant microemulsions.
Applications Of Microemulsions
The microemulsion formulation must have low allergic
15,16
potential, good physiological compatibility and high Microemulsions are promising delivery systems that
biocompatibility. allow sustained or controlled drug release for percutaneous,
peroral, topical, transdermal, ocular and parenteral
The components involved in the general formulation of
administration. Enhanced absorption of drugs, modulation of
microemulsions include (a) an oil phase (b) an aqueous phase
the kinetics of the drug release and decreased toxicity are
containing hydrophilic active ingredients [preservatives and
several advantages in the delivery process. The following is a
buffers may be included] (c) a primary surfactant [anionic,
compilation of reported literature for topical microemulsions.
non-ionic or amphoteric] (d) secondary surfactant or
cosurfactants. Antifungal

Generally non-ionic surfactants are chosen because of their Superficial mycoses usually respond to topical therapy. In the
good cutaneous tolerance, lower irritation potential and settling of eczema, topical antifungal agents such as
toxicity. Microemulsions can be formulated using single- ketoconazole are used to reduce the fungal infection caused
chain surfactants or double chain surfactants. Single chain by Pityrosporum ovale or Malassezia furfur.
surfactants do not lower the oil water interfacial tension
Antifungal agents e.g miconazole, ketoconazole, and
sufficiently and hence cosurfactants are required. Double
itraconazole being lipophilic in nature have been formulated
chained surfactants like sulfosuccinates can form
as microemulsions to impart to them the advantages like
microemulsions in the absence of cosurfactants but are too
8
ease of preparation due to spontaneous formation,
toxic for general pharmaceutical applications . The
thermodynamic stability, transparent and elegant appearance,
cosurfactants even though being indispensable in the
increased drug loading, enhanced penetration through the
formulation of microemulsions, have exhibited toxicity e.g.
biological membranes, increased bioavailability compared to
medium chain length alcohols9. Hence judicious choice of
conventional dosage forms.17, 18.
surfactants and cosurfactants is of great importance. The use
of polyoxyethylenealcohol ethers has been reported as Microemulsions of poorly water soluble antifungal drugs
cosurfacants 1 0 , 11 , 1 2 . Microemulsions prepared from miconazole, ketoconazole, and itraconazole were designed
phospholipids such as lecithins are preferred over synthetic and developed by Puranajoti et al19 using either mineral oil or

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 101
 Padmini et al.: Microemulsions For Topical Use– A Review.

olive oil as an oil phase. Various combinations of surfactant and permeation coefficient as compared to marketed cream
®
and cosurfactant were used, including Labrafil M 1944 CS and ointment formulation. In vivo antiviral studies performed
and Plurol® Oleique (1:1); Labrafil® M 1944 CS and Plurol® in female mice against induced herpes simplex virus I
® ®
Oleique (1:2); or Labrafil M 1944 CS, Capmul MCM C-8, infection indicated that a single application of
®
Pluro Oleique, and dehydrated ethyl alcohol (3:3:1:1). microemulsion formulation containing 2.5% Transcutol, 24
hours post-injection resulted in complete suppression of
Microemulsions of poorly water-soluble antifungal agents 24
development of herpetic skin lesions .
were successfully developed with in vitro release rates
comparable to that of the gel formulation. The results of the Anti acne
work done on miconazole nitrate formulated as positively Novel drug delivery strategies like microemulsions can play a
charged microemulsions indicate optimized drug targeting pivotal role in improving the topical delivery of antiacne
without a concomitant increase in systemic absorption. L- agents by enhancing their dermal localization with a
alanine benzyl ester, an ester of a natural amino acid, is an 25
concomitant reduction in their side effects . Microemulsions
20
appropriate ionic charge-inducing agent . of azelaic acid, a bioactive molecule used in many skin
Microemulsion based gels for vaginal delivery of disorders, prepared using the monosodium salt (AZA-Na)
clotrimazole and fluconazole were developed and compared have been evaluated as delivery vehicles. Dialysis membrane
with the marketed clotrimazole gel (Candid-V® gel) by in experiments showed decreasing permeability to AZA-Na,
vitro methods. These microemulsion based gels showed and this was related to its partition at the microemulsion
significantly higher in vitro bioadhesion, antifungal activity interface. The results suggested that microemulsions
as compared to that of Candid-V® gel. Fluconazole containing AZA-Na could be used to optimize drug targeting
26
microemulsion based gel did not exhibit vaginal irritation21,22. in acne treatment .

Antiviral To increase the solubility of azelaic acid in the dispersed oil

phase of microemulsion containing polysorbate 20, butanol,
A study was done to investigate and evaluate microemulsion
decanol:dodecanol (2/1) and water, the pH of aqueous phase
and microemulsion-based hydrogel as a topical delivery
was lowered and propylene glycol was added. An increased
system for penciclovir in comparison with a commercial
partitioning into the lipophilic phase was noted as propylene
cream. The results of permeation test in vivo in mice showed
that as compared with the commercial cream, microemulsion- glycol concentration was increased. Microemulsion thus
based hydrogel and microemulsion could significantly provided a vehicle in which azelaic acid was dissolved rather
increase the permeation of penciclovir into both epidermis than suspended as in a cream. Moreover, reservoir effect
and dermis. Stability tests showed that microemulsion-based achieved by partitioning into the oil could prolong its release
hydrogel stored at 4 °C for 3 months had no significant change over several hours. It showed a 10 fold increase in the amount
in physicochemical properties. Skin irritation test in rabbits of drug released (upto 27-30% of initial amount) from the
demonstrated that single application or multiple applications microemulsion when compared to a cream clinically used in
27
of microemulsion-based hydrogel did not cause any erythema treatment of skin disorders .
or edema. Thus, it can be concluded that microemulsion- Antioxidants
based hydrogel could be a promising vehicle for topical
23 Antioxidants have been used in dermatological and cosmetic
delivery of penciclovir .
products because of their property of scavenging and
Acyclovir containing microemulsion-based formulations for destroying aggressive oxidizing agents and free radicals that
topical delivery were developed using isopropyl are involved in various skin conditions.
myristate/Captex 355/Labrafac as an oil phase, Tween 20 as
surfactant, Span 20 as cosurfactant, and In animals, topical application of alpha-tocopherol has shown
water:dimethylsulfoxide (1:3) as an aqueous phase. to exert photoprotective effects by reducing the number of
Transcutol, eucalyptus oil, and peppermint oil were used as sunburn cells, UV B induced damage and inhibiting
permeation enhancers. In vitro permeation studies through photocarcinogenesis. An o/w or w/o microemulsion of
mice skin were performed using Franz diffusion cells. The vitamin E delivered the vitamin predominantly to the
optimum formulation containing 2.5% Transcutol as the epidermis avoiding accumulation in organs other than the
penetration enhancer showed 1.7-fold enhancement in flux skin. The cream or lotion preparations of the same amount of

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 102
 Padmini et al.: Microemulsions For Topical Use– A Review.

vitamin results in excessive accumulation in the organs28. agent for w/o microemulsions and 0.5% (w/w) xanthan gum
for the o/w microemulsions. The presence of thickening agent
Newer studies show that combined applications of various
and the location of sodium ascorbyl phosphate in the
antioxidants can increase their potency as compared with a
29
microemulsion influenced the in vitro drug release profiles.
single antioxidant alone. Branka Rozman et al have
When incorporated in the internal aqueous phase, sustained
developed a temperature-sensitive microemulsion gel as an
release profiles were observed. This study confirmed
effective and safe delivery system suitable for simultaneous
microemulsions as suitable carrier systems for topical
topical application of a hydrophilic vitamin C and a lipophilic
application of sodium ascorbyl phosphate32.
vitamin E. By changing water content of liquid o/w
33
microemulsion, a gel like microemulsion with temperature- Spiclin et al studied the stability of o/w and w/o type of
sensitive rheological properties was formed. The microemulsions for topical use containing ascorbyl palmitate
temperature-driven changes in its microstructure were and sodium ascorbyl phosphate which are derivatives of
confirmed by rotational rheometry, viscosity measurements ascorbic acid that differ in stability and hydrophilic and
and droplet size determination. The release studies have lipophilic properties. The stability of the less stable derivative
shown that the vitamin release at skin temperature from gel- ascorbyl palmitate was tested under different conditions to
like microemulsion were comparable to those from o/w evaluate the influence of initial concentration, location in
microemulsion and were much faster and more complete than microemulsion, dissolved oxygen and storage conditions.
from o/w microemulsion conventionally thickened with High concentrations of ascorbyl palmitate reduced the extent
polymer (carbomer). of its degradation. In contrast, sodium ascorbyl phosphate
was stable in both types of microemulsion and was shown to
Non-thickened (o/w, w/o and gel-like) and thickened (with
be convenient as an active ingredient in topical preparations.
colloidal silica) microemulsions were studied as carriers for
In the case of ascorbyl palmitate, long-term stability in
vitamin C and E using reconstructed human epidermis
selected microemulsions was not adequate.
(RHE). The amounts of these vitamins accumulated in and
permeated across the RHE were determined, together with Investigation on the amphiphilic antioxidant ascorbyl
factors affecting skin deposition and permeation. Notable palmitate and its effectiveness against free radical formation
34
differences were observed between formulations. The was proven by Polona Jurkovic et al . When applied on the
absorption of vitamins C and E in RHE layers was in general skin, ascorbyl palmitate decreased the level of formation of
enhanced by microemulsions and the vitamins incorporated free radicals. Its effectiveness depended significantly on the
in the outer phase of the microemulsion exhibited greater carrier system, the type of microemulsion and its
absorption than that when vitamins were in the inner phase. concentration while the time of application had no influence
Addition of thickener enhanced the deposition of vitamins E on its effectiveness. Oil in water microemulsions delivered
and C in the RHE .
30 ascorbyl palmitate to the skin significantly better than water
in oil microemulsions. In both types of microemulsions, the
Various delivery systems of alpha-tocopherol (1%) were effectiveness increased at higher concentrations of ascorbyl
formulated, which included simple solution, gels, emulsions, palmitate..
and microemulsions. The hydroalcoholic gel delivered
significantly higher amounts of alpha-tocopherol into the In order to develop alternative formulations for topical
receptor than the other gels used. A microemulsion containing administration of retinoic acid, Michele Trotta et al35
isopropyl myristate emerged as the best delivery system for evaluated microemulsions as delivery vehicles. Oil in water
31
alpha-tocopherol amongst all the systems studied . and water in oil microemulsion formulations were prepared
using water, isopropyl myristate, lecithin, caprylyl–capryl
Microemulsions of w/o and o/w type for topical application glucoside and ethanol or 1,2 hexanediol. The results
containing sodium ascorbyl phosphate (hydrophilic suggested that o/w microemulsions containing a counter ion
derivative of ascorbic acid) were formulated and compared can be used to optimise drug targeting without a concomitant
with topical application of ascorbyl palmitate which is a increase in systemic absorption.
lipophilic derivative of vitamin C. To obtain liquid
microemulsions appropriate for topical application, their Miscellaneous skin conditions:
viscosity was increased by adding thickening agents. The following are examples depicting the use of
Colloidal silica 4% (w/w) was chosen as a suitable thickening microemulsions in varied skin conditions:

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 103
 Padmini et al.: Microemulsions For Topical Use– A Review.

An o/w microemulsion formulated using lecithin and an alkyl temozolomide acid hexyl ester through silicon membranes
glucoside as mild, non-irritant surfactants was proposed as a and rat skin resulting in less drug retention within the skin,
cosmetic vehicle for arbutin and kojic acid which are while oleic acid microemulsion systems demonstrated higher
naturally occurring whitening agents. The photostability to solubilising ability and a higher concentration of
UVB irradiation of both whitening agents was determined in temozolomide acid hexyl ester retained within the skin40.
aqueous solutions and in microemulsions, and also in the
The abilities of an o/w microemulsion of ethyl oleate with
presence of the perfumed compositions. The stability of
Tween 80 as emulsifier and n-pentanol as a co-emulsifier
arbutin and kojic acid was higher in microemulsions than in
36
were investigated to inactivate suspensions of vegetative cells
aqueous solutions .
of Salmonella spp. Escherichia coli Pseudomonas
Subramanian et al 37studied the topical delivery of celecoxib aeruginosa, Staphylococcus aureus and Listeria
using microemulsion as the vehicle for the treatment of UV B monocytogenes and was found to be effective against all five
induced skin cancer. Various oil to cosurfactant ratios were microorganisms. The abilities of these microemulsions to
studied to identify the formulation variables for reduce preformed biofilms of the five bacteria were also
microemulsion formation. The effect of these variables on investigated and was found to be effective41.
skin permeation of celecoxib was evaluated. Topical anti-
A microemulsion gel-based system of babchi oil (Psoralea
inflammatory effect of celecoxib was assessed and it showed
corylifolia) was studied for the treatment of psoriasis which
higher permeation rate and significant anti-inflammatory could provide improved permeation of the drug through the
activity. The studied microemulsion formulations have a skin and increased patient compliance. The chief constituent
prospect for use as a potential vehicle for treatment of UV B of babchi oil is psoralen, a photoactive furocoumarin which
induced skin cancer. reduces cell proliferation. Moreover, babchi oil, in addition to
38
Baroli developed and evaluated alternative microemulsion providing psoralen also acts as an oily phase for
formulations for topical administration of 8-Methoxsalen microemulsion system. The presence of surfactant and
and related furocumarins for the treatment of cosurfactant increases the permeation. Eight marketed
hyperproliferative skin diseases in association with long- samples of babchi oil were used for the preparation of
wavelength UVA light using water, isopropyl myristate microemulsions which were subjected to different
(IPM) and Tween 80: Span 80: 1,2-Octanediol (3:1:1.2 w/w ) thermodynamic stability tests and characterized for droplet
and results suggest that the studied microemulsion system is size, viscosity and refractive index. In vitro skin permeation
suitable. of babchi oil through rat abdominal skin was determined by
the Franz diffusion cell. The in vitro skin permeation profile
A combination of inhibitors of cyclo-oxygenase-2 and 5-
of a formulation consisting of 1.67% v/v of babchi oil, 8.33%
lipoxygenase applied via a microemulsion delivery system
v/v of oleic acid, 55% v/v of Tween 80: Transcutol-P (1:1) and
was proven to be effective in topically inhibiting skin
35% v/v of distilled water was significant when compared
carcinogenesis. The results clearly showed that topical
with other microemulsion formulations. This formulation
treatment with microemulsions containing celecoxib alone or
was converted into microemulsion gel by adding 1%
celecoxib plus zileuton significantly inhibited skin
Carbopol-940 and was tested for its in vivo antiinflammatory
carcinogenesis and that a combination of both agents had the effects determined by footpad edema. The results suggested
39
best results . that microemulsion gel is a potential vehicle for improved
Temozolomide acid hexyl ester used in the treatment of skin topical delivery of psoralen and that microemulsion gels are
cancer has poor solubility and instability. Microemulsion potential vehicles for improved topical delivery of babchi
42
systems were formulated with either oleic acid or isopropyl oil .
myristate as the oil phase and tocopheryl polyethylene glycol Microemulsions containing Aerosol OT, Tween 85, isopropyl
1000 succinate as a surfactant. Topical formulations of oleic myristate and water were observed to possess a potentially
acid or isopropyl myristate demonstrated beneficial improved skin bioavailability of cyclosporine A for topical
solubilising ability and provided a stable environment for the delivery against autoimmune skin disorders. In animal studies
drug. In permeation studies, the isopropyl myristate the amount of drug deposition into the skin and subcutaneous
microemulsion systems with inclusion of isopropyl alcohol fat was respectively almost 30 and 15-fold higher than the
(IPA) as a co-surfactant significantly increased permeation of concentrations compared with oral administration. The

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 104
 Padmini et al.: Microemulsions For Topical Use– A Review.

systemic distribution in blood, liver and kidney was much a longer time when compared with a conventional preparation
lower following topical administration as compared to oral that could allow for a decreased number of applications of eye
administration. The study indicated that because of high drops per day.
local concentrations and minimal distribution to other organs
Microemulsions composed of Span 20, Tween 20, isopropyl
via the circulation, topical microemulsion is a suitable
43
myristate and water were investigated as potential drug
vehicle for cyclosporin A . delivery systems for eye drops containing chloramphenicol
46
Ocular by Lv F.F. et al . Chloramphenicol in the conventional eye
Eye drops account for 90% of the available ophthalmic drops hydrolyzes easily to glycol whereas in the
formulations due to their simplicity and convenience. microemulsion it was trapped into the oil-in-water system.
However, rapid precorneal loss caused by drainage and high The results revealed that the content of the glycols in the
tear fluid turnover is amongst the major problems associated microemulsion formulation was much lower than that in the
with topical ophthalmic drug delivery. Only 5% of the applied commercial eye drops at the end of the accelerated stability
drug in eye drops penetrates the cornea and reaches the experiments. It implied that the stability of the
intraocular tissues with the rest of the dose undergoing chloramphenicol in the microemulsion formulations was
transconjunctival absorption or drainage via the nasolacrimal increased remarkably. The microemulsion was found to be
duct before transnasal absorption. This results in loss of drug almost free of toxicity and irritation.
into the systemic circulation and provides undesirable Spermicidal
systemic side effects Accordingly, microemulsions provided 47
O.D'Cruz described a formulation of novel gel-
a promising alternative with improved ocular retention,
microemulsions (GM) as nontoxic, dual-function
increased corneal drug absorption and reduced systemic side
intravaginal spermicides, which could be used as delivery
effects whilst maintaining the simplicity and convenience of
vehicles for lipophilic drug substances targeting sexually
the dosage form as eye drops.
transmitted pathogens. These GMs comprising oil-in-water
44
Judy Chan et al evaluated microemulsion based phase microemulsion and polymeric hydrogels were designed to
transition systems for ocular delivery of pilocarpine solubilize lipophilic antiviral/antimicrobial agents and
hydrochloride (a model hydrophilic drug). They used two exhibited rapid spermicidal activity in human semen and was
non-ionic surfactants sorbitan mono laurate and compared against nonoxynol-9-based detergent spermicide
polyoxyethylene sorbitan mono-oleate with ethyl oleate (oil (Gynol II). Spermicidal GM has shown unprecedented
component) and water. These systems undergo phase change potential as dual function microbicidal contraceptives to
from microemulsions to liquid crystalline and to coarse improve vaginal bioavailability of poorly soluble
emulsion with a change in viscosity depending on water antimicrobial agents without causing significant vaginal
content. Incorporation of pilocarpine hydrochloride did not damage.
affect the phase behaviour. Thus, phase transition
Cosmetics
microemulsion is promising for ocular drug delivery as it
provides the fluidity with its viscosity being increased after There is growing recognition of the potential benefits of
application and increasing ocular retention while retaining microemulsions in the field of cosmetics in addition to drug
the therapeutic efficiency. delivery. They are now being widely investigated for
45
preparing personal care products with superior features such
Fialho and da Silva-Cunha aimed to develop and
as having improved product efficiency, stability, appearance
characterize an oil-in-water microemulsion containing
and minimal irritation. They are well suited for the
dexamethasone, and evaluate its ocular irritation and the
preparation of various cosmetic products such as
pharmacokinetics. The developed system showed an
moisturizing and soothing agents, sunscreens,
acceptable physicochemical behavior and presented good
antiperspirants, body cleansing agents, hair conditioners and
stability for three months. The ocular irritation test used
after shave formulations. Microemulsions are also suitable in
suggested that the microemulsion did not provide significant
perfumery so as to minimize the quantity of organic solvents48.
alteration to eyelids, conjunctiva, cornea or iris. This
formulation showed greater penetration of dexamethasone in CONCLUSION
the anterior segment of the eye and also release of the drug for
Microemulsions are commercially feasible, simple and

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 105
 Padmini et al.: Microemulsions For Topical Use– A Review.

convenient novel vehicles for delivery of medicaments which formation and characterization of phospholipid
can enhance drug absorption with reduced systemic side microemulsions. I. Pseudo-ternary phase diagrams of
effects. They can be used to optimise drug targeting without a systems containing water-lecithin-alcohol-isopropyl
concomitant increase in systemic absorption. Appropriate myristate. Int. J. Pharm. 1993;93(1-3):161–175.
excipient selection and safety evaluation especially of the 14. Attwood D, Mallon C, and Taylor CJ . Phase studies and
cosurfactants is crucial in the formulation of microemulsions. particle size analysis of oil-in-water phospholipid
They can be potential drug delivery systems for the delivery microemulsions. Intl. J of Pharm.1995;116(2) : 253-261.
of more than one medicament simultaneously.
15. Kumar P, Mittal KL. Handbook of microemulsion
REFERENCES science and technology. New York :Marcel Dekker
1. Bidyut KP, Satya PM. Uses and applications of Inc;1999.457–497; 549–603 ;679–712; 755–77.
microemulsions. Current Science 2001;80(8): 990-1001. 16. Solans C, Kunieda H. Industrial applications of
2. Jain NK. Progress in controlled and novel drug delivery microemulsions. New York:Marcel Dekker Inc; 1997.
st
systems. 1 ed. New Delhi: CBS publishers and 199, 147–174, 97–122, 123–145, 69–95.
distributors; 2004.324. 17. Tenjarla SN. Microemulsions: An overview and
3. Osborne D, Anton HA, Topical Drug Delivery pharmaceutical applications. Critical Reviews TM in
Formulations. Informa Health Care, 1990;357 Therapeutic Drug Carrier Systems.1999;16:461–521.

4. Hoar TP, Schulman JH. Transparent water in oil 18. Lieberman HA, Rieger MM, Banker GS. Pharmaceutical
dispersions: the oleopathic hydromicelle. Nature Dosage Forms: Disperse Systems. 2nd ed. Vol 1.New
1943;152: 102 – 103. York :Marcel Dekker Inc; 1996. 211–281, 315–370.

5. Attwood D, Kreuter J. Colloidal Drug Delivery Systems. 19. Puranajoti PR, Patil T, Sheth PD, Bommareddy GP,
New York:Marcel Dekker;1994.31–71. Egbaria DK. Design and Development of Topical
Microemulsion for Poorly Water-Soluble Antifungal
6. Derle DV, Sagar BSH, Microemulsion as a vehicle for
Agents The Journal of Applied Research 2002;2(1).
transdermal permeation of nimesulide. Ind. J. Pharm.
Sci. 2006; 68(5): 622-625. 20. Peira E, Carlotti ME, Trotta C, Cavalli R, Trotta M.
Positively charged microemulsions for topical
7. Kreilgaard M. Influence of microemulsions on cutaneous
application. Int J Pharm. 2008;346(1-2):119-23.
drug delivery. Bulletin Technique Gattefossé 2002;95 :79
– 100. 21. Yogeshwar B, Vandana P. Microemulsion-Based Vaginal
Gel of Clotrimazole: Formulation, In vitro Evaluation,
8. Jain NK. Progress in controlled and novel drug delivery
st
and Stability Studies. AAPS PharmSciTech.
systems. 1 ed. New Delhi: CBS publishers and
2009;10(2):476-481.
distributors; 2004.319-320.
22. Yogeshwar B, Vandana P. Microemulsion based vaginal
9. Mittal KL, Pramod P. Handbook of microemulsions
gel of fluconazole: formulation, in vitro and in vivo
science and technology. New Delhi:CRC Press; 1999.
evaluation. Intl J Pharm. 2009;365(1-2):175-179.
767.
23. Weiwei Zhu, Chenyu Guo, Aihua Yu, Yan Gao, Fengliang
10. Jayakrishnan A, Kalaiarasi K, Shah DO. Microemulsion:
Cao and GuangXi Zhai. Microemulsion-based hydrogel
Evolving technology for cosmetic applications. J.
formulation of penciclovir for topical delivery.
Soc.Cosmet.Chem. 1983;34:334.
International Journal of Pharmaceutics 2009;378(1-
11. Johnson, KA, and Shah DO. Effect of oil chain length and 2):152-158.
electrolytes on water solubilization in alcohol-free
24. Shishu, Sunita Rajan, Kamalpreet Development of novel
pharmaceutical microemulsions. J of Colloid and
microemulsion based topical formulations of Acyclovir
Interface Sci. 1985;107(1): 269-271.
for the treatment of cutaneous herpetic infections.AAPS
12. Shinoda, K, Shibata Y, Lindman B. Interfacial Tensions PharmSciTech. 2009;10(2):559-565.
for Lecithin Microemulsions Including the Effect of
25. Date AA, Naik B, Nagarsenker MS. Novel Drug Delivery
Surfactant and Polymer Addition.
Systems: Potential in Improving Topical Delivery of
Langmuir 1993; 9:1254.
Antiacne Agents. Skin Pharmacology and Physiology
13. Aboofazeli R, Lawrence MJ. Investigations into the 2006; 19:2-16.

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 106
 Padmini et al.: Microemulsions For Topical Use– A Review.

26. Peira E. Carlotti ME, Cavalli R, Trotta M. Journal of 38. Baroli B, López-Quintela MA, Delgado-Charro MB,
drug delivery science and technology: Azelaic acid Fadda AM, Blanco-Méndez Microemulsions for topical
sodium salt in the formulation of microemulsions for delivery of 8-methoxsalen. J of Controlled Release.
topical applications. 2006;16(5): 375-379. 2000;3;69(1):209-218.
27. Gallarate MR, Gasco MR, Rua G. In vitro release of 39. Fegn L, Wang Z. Topical chemoprevention of skin cancer
azelaic acid form oil in water microemulsions. Acta in mice, using combined inhibitors of 5-lipoxygenase and
Pharm Jugosla .1990;40:533. cyclo-oxygenase-2. The Journal of Laryngology &
28. Martini MC, Bobin MF, Flandin H, Caillaud F, Cotte J. Otology 2009;123(8):880–884.
Role of microemulsions in the percutaneous absorption 40. Suppasansatorn, Panassay, Nimmannit, Ubonthip,
of alpha-tocopherol. J Pharm Belg. 1984;39(6):348-54. Conway et al.Microemulsions as topical delivery
29. Branka R, Alenka Z, Francoise F, Mirjana G. vehicles for the anti-melanoma prodrug, temozolomide
Temperature-Sensitive Microemulsion Gel: An Effective hexyl ester (TMZA-HE). J of Pharm. and Pharmacol.
Topical Delivery System for Simultaneous Delivery of 2007; 59(6): 787-794.
Vitamins C and E. AAPS PharmSciTech. 2009;10(1):54- 41. Teixeira PC, Leite GM, Domingues RJ, Silva J, Gibbs PA,
61. Ferreira JP. Detection of biofilm formation among the
30. Branka R, Mirjana G, Estelle T, Fabrice P and Francoise clinical isolates of staphylococci: An evaluation of three
F. Simultaneous absorption of vitamins C and E from different screening methods. International Journal of
topical microemulsions using reconstructed human Food Microbiology 2007;118(1):15-19.
epidermis as a skin model. European J of Pharm and 42. Ali J, Akhtar N, Sultana Y, Baboota S, Ahuja A.
Biopharmaceutics 2009;72(1):69-75. Antipsoriatic microemulsion gel formulations for topical
31. Rangarajan M, Zatz J. Effect of formulation on the topical drug delivery of babchi oil (Psoralea corylifolia).
delivery of alpha-tocopherol. J of Cosmet Sci. Methods Find Exp Clin Pharmacol. 2008;30(4):277-85.
2003;54(2):161-74. 43. Hongzhuo L. , Yongjun W. , Yiyong L. , Huimin Y. , Yang
32. Spiclin P, Homar M, Zupancic VA Gasperlin M. Sodium D. , Sanming Li. Bicontinuous Cyclosporin-a loaded
ascorbyl phosphate in topical microemulsions. Intl J Wa t e r - A O T / Tw e e n 8 5 - i s o p r o p y l m y r i s t a t e
Pharm. 2003;256(1-2):65-73. microemulsion: Structural characterization and dermal
33. Spiclin P, Gasperlin M, Kmetec V. Stability of sodium pharmacokinetics in vivo. J Pharm Sci. 2009;98:1167-
ascorbyl phosphate in topical microemulsions. Intl. J 1176.
Pharm. 2001;222(2):271-9. 44. Chan J, Gamal MM, El Maghraby, Jennifer PC, Raid GA.
34. Polona Jurkovic , Marjeta Sentjurc, Mirjana Gas perlin, Phase transition water-in-oil microemulsions as ocular
Julijana Kristl, Slavko Pecar. Skin protection against drug delivery systems: In vitro and in vivo evaluation.
ultraviolet induced free radicals with ascorbyl palmitate Intl. J of Pharm. 2007;328(1):65-71.
in microemulsions Eur J Pharm Biopharm 45. Fialho SL, da Silva-Cunha A. New vehicle based on a
2003;56(1):59-66. microemulsion for topical ocular administration of
35. Trotta M, Ugazio E, Peira E, Puritano C. Influence of ion dexamethasone. Clin Exper Ophthalmol.
pairing on topical delivery of retinoic acid from 2004;32(6):626.
microemulsions. J of Controlled Release 2003;86(2- 46. Lv FF, Li N, Zheng LQ, Tung CH. Studies on the stability
3):315-321. of the chloramphenicol in the microemulsion free of
36. Peira E, Carlotti ME, Cavalli R, Trotta M. Azelaic acid alcohols. Eur J Pharm Biopharm. 2006;62(3):288-94.
sodium salt in the formulation of microemulsions for 47. D'Cruz O. Gel-microemulsions as vaginal spermicides
topical applications. J of Drug Delivery Sci. and Tech. and intravaginal drug delivery vehicles. Contraception
2006;16(5) : 375-379. 2001;64(2):113-123.
37. Subramanian N, Ghosal SK, Moulik SP. Topical delivery 48. Azeem A, Rizwan M, Ahmad FJ et al. Emerging role of
of celecoxib using microemulsion. Acta Pol Pharm. microemulsions in cosmetics. Recent Pat Drug Deliv
2004;61(5):335-41. Formul. 2008;2(3):275-89.

Ind J Pharm Edu Res, Jan-Mar, 2011/ Vol 45/ Issue 1 107