Two-ComponentMonolayers

II. Surface Pressure--Area Relations for the OctadecanoI-Docosyl Sulphate System

I. S. C O S T I N 1 AND G. T. B A R N E S
Department of Chemistry, University of Queensland, St. Lucia, Queensland, Australia
Received August 12, 1974; accepted November 29, 1974

Surface pressure-area isotherms are reported for monolayers of octadecanol and docosyl
sulphate and their mixtures on substrates of 0.1 M KC1 at 298.3 K. Eleven different composi-
tions spanning the whole composition range were investigated. At all compositions and at all
surface pressures the monolayer areas are less than the added areas of the two pure components.
The area-composition plots show three distinct regions: alcohol-rich (x~t > 0.6), intermediate,
and sulphate-rich (XA < 0.18); and partial molecular areas of alcohol and of sulphate have
been determined for each region. The theory of Fowkes has been applied to the data and the
system shown to be nonideal in the terms of this theory. Excess free energies of mixing have
been calculated by the method of Goodrich. Values ranging from zero to -600 J mole-1 were
obtained. The contributions of the high area portions of the II-.~ isotherms were not included
in these calculations, but estimates of these contributions have been made from the evapora-
tion resistance data reported in Part III and values between --60 and --200 J mole-z were
found.

INTRODUCTION is p o s t u l a t e d if the m o n o l a y e r p r o p e r t y , or
I n considering two-component monolayers some simple function of it, is linearly d e p e n d e n t
we are essentially concerned with the varia- on monolayer composition. This may, or m a y
tions of m o n o l a y e r properties as functions of not, be reasonable, b u t the converse is almost
monolayer composition, and with the insights certainly justified : t h a t marked deviations from
this information m a y p r o v i d e a b o u t inter- a linear relationship, a n d in p a r t i c u l a r the
molecular forces in the monolayer. I n bulk occurrence of a m a x i m u m or minimum, indi-
liquid systems the concept of an ideal solu- cate a significant deviation from ideality.
tion has p l a y e d a k e y role, and similarly, in Deviations from ideality are frequently
two-dimensional solutions, a concept of ideal a t t r i b u t e d to differences in intermolecular
behavior lies behind the i n t e r p r e t a t i o n of m a n y forces. I n relating observed properties to the
monolayer properties. U n f o r t u n a t e l y it is much intermolecular forces in a m o n o l a y e r the term
more difficult to define i d e a l i t y for two-di- molecular interaction has often been loosely
mensional solutions t h a n it is for three-di- used to indicate t h a t the intermolecular forces
mensional solutions, and consequently there are stronger t h a n for an ideal mixture. Of
has been a t e n d e n c y for a notion of ideal be- course there is always interaction so we prefer
havior to develop for each monolayer p r o p e r t y the less misleading term excess interaction.
and for each theory. I t is b y no means c e r t a i n Q u a n t i t a t i v e l y , an excess interaction would be
t h a t these various notions of i d e a l i t y are con- expressed as an excess of some t h e r m o d y n a m i c
sistent with one another. Generally, " i d e a l i t y " function: entropy, enthalpy, or free energy;
1Present address: Science Department, Mt. Gravatt or in terms of an a c t i v i t y coefficient in the
Teachers' College, Mt. Gravatt, Queensland, Australia. surface.

106
Journal of CoUoid and Interface Science, Vol. 51, No, 1, April 1975 Copyright ~ 1975 by Academic Press, Inc.
All rights of reproduction in any form reserved.
 TWO-COMPONENT MONOLAYERS. II 107

60
1 I 1 I I I I

¥
z
E
50 --

1 \
0

!
40 -- I0

r~
30 --
03
(D
bJ
Iv
D_
20 --
hl
(D 3
It
n- 8
I0 -
09

0 I I
0"18 0.19 020 0"21 0'22 0"25 0"24 0"25 0'26 0"27
AREA PER MOLECULE, $/nm 2 molecule-!
FIG. 1. Surface pressure-area isotherms of octadecanol-docosyl sulphate monolayers of various
compositions. Substrate: 0.1 M KCI + Na2 EDTA; temperature: 298.3 K. Curve numbers refer to
monolayer composition (mole percent of octadecanol).

In this paper we will examine a number of tire range from pure octadecanol, through
monolayer properties which are derivable from various octadecanol-docosyl sulphate mixtures,
surface pressure-area studies and discuss their to pure docosyl sulphate. The surface pressure-
dependence on monolayer composition. The area isotherms are shown in Fig. 1. Each curve
system used consists of octadecanol-docosyl is the average of at least three separate deter-
sulphate monolayers on 0.1 M potassium minations. For each monolayer composition at
chloride solution. least one isotherm was run at a slower com-
pression rate than normal (speed 5, 6.9 mm
EXPERIMENTAL rain -t, instead of speed 7, 27.5 mm min-1),
The apparatus and techniques for determin- but no significant effect was ever observed.
ing 1I-ti isotherms are described in Part I (1). The eleven systems studied all formed con-
The substrate for all monolayers was 0.1 M densed monolayers under the experimental
KCI with added 0.0001 M Na2EDTA. The bulk conditions used. However the docosyl sulphate
substrate temperature was 298.3 -I- 0.4 K. The monolayer was more compressible than the
monolayer materials were : 1-octadecanol which octadecanol monolayer, and occupied, on
had been purified by formation of the urea average, larger areas of surface. In the mixed
clathrate and recrystallized from n-hexane (see monolayers, increasing the proportion of octao
Part I); and sodium docosyl sulphate which decanol resulted in a shift of the isotherm to
had been washed and twice recrystallized (see lower molecular areas at surface pressures less
Part I). The spreading solvent was 60% Merck than 13 mN m -1 (Fig. 2), and in an increase in
benzene and 40% Merck methanol. The con- the steepness of the isotherm at all surface
concentrations of monolayer material in the pressures (Fig. 1). However the compressi-
spreading solutions were between 1.8 and 3.7 bilities did not show any significant trend
mM. (Table I).
RESULTS With changing composition there is also a
In all, we have studied monolayers of eleven gradual change in the shape of the isotherm
different compositions distributed over the en- from the smoothly curved isotherm of the
Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
108 COSTIN AND BARNES

layers at high surface pressures (see Part I I I ,

Fig. l).

DISCUSSION
"7 The discussion has been divided into a num-
o
ber of sections, in each of which a particular
-6
E approach to the concepts of ideality and non-
% ideality in two-component monolayers is ex-
c
amined with respect to the present data for
the octadecanol-docosyl sulphate system.
w"
.J

0
W t_ COLLAPSE PRESSURE AND MONOLAYER
.J
0 STABILITY
r A relationship between the collapse pressure
0_
and the molecular interactions in a mixed
w monolayer has been suggested (2, 3), an en-
hanced collapse pressure being taken to indicate
stronger interactions than in an ideal mixture.
However there are difficulties in the determina-
v ,v
tion of meaningful collapse data (4, 5), and
o o.z o.4 o.6 os I.O moreover, for a thermodynamic analysis of
COMPOSITION, xA
the results the collapse point must represent
FIG. 2. Area per molecule as a function of composi- an equilibrium between the monolayer and the
tion for octadecanol-docosyl sulphate monolayers at collapsed material. In the present work we have
various surface pressures, using data from Fig. 1. Point considered several criteria of collapse.
symbols indicate different surface pressures as in list Experimentally, collapse is recognized as a
of symbols (1).
sharp drop in the surface pressure. In the
present work this occurred at 56 -q- 1 m N m -1
docosyl sulphate to the two-region isotherm of with monolayers of all compositions. However
the octadecanol. These two regions can be the surface pressure drop was often accom-
distinguished in the isotherms of monolayers panied by slight overflow of the water surface
with alcohol mole fractions greater than 0.5.
The transition point (-/iT, IIT) between the two
TABLE I
regions is sharp in the isotherm of the pure COMPRESSIBILITIES OF O C T A D E C A N O L - D o c o s Y L
alcohol and becomes less pronounced as the SULPHATE MONOLAYERS
alcohol content decreases.
xA I I / m N m -1 A/nm~ g/l*N m -~
With some of the mixed monolayers, molecu- molecule-1
lar areas were observed which were smaller
than those of a pure octadecanol monolayer. 0.0 10 0.254 4.7
There are several considerations which suggest 40 0.228 3.1
0.1 10 0.237 5.3
that these smaller areas are real and have not 40 0.208 3.9
arisen from the experimental difficulties which 0.4 10 0.213 5.9
were discussed earlier (Part I). Of these con- 40 0.191 1.6
siderations, the most important are the 0.7 10 0.201 5.2
reproducibility and self-consistency of the area 40 0.191 0.6
1.0 10 0.201 5.2
data as shown in Fig. 2, and the enhanced 40 0.194 0.7
evaporation resistances of these mixed mono-
J o u r n a l o f Colloid a n d I n t e r f a c e S c i e n c e , Vol. 51, No. 1, April 1975
 TWO-COMPONENT MONOLAYERS. II 109

and it seems that we were not observing 50 I I I I

monolayer collapse, but a critical surface pres- ,i
sure for overflow. This would always occur at
the same surface pressure as the water was
always adjusted to the same height and the
geometry of the air/water/teflon triple inter-
~o
face was therefore reproduced in each
experiment. ~-" 2o

When a monolayer is compressed to a sur-

face pressure which is higher than its equilib- 40
rium spreading pressure and lower than the
collapse pressure, and is held at constant area,
it usually exhibits a slow downward drift in ~ 3o
surface pressure (Fig. 3). The observation of
such drifts provides a measure of the stability 2o I I I I
0 0'2 0"4 0'6 0"8 I'0
of the monolayer. The necessary data were
COMPOSITION, xA
available from the records of surface pressure
made during the measurement of evaporation FIG. 4. Highest surface pressure at which the mono-
resistance (1). The surface pressure was noted layer meets a given stability criterion, plotted as a
at the completion of each step in the procedure: function of composition. Curve 1 : drift of 1 mN m-I or
less in time interval Atl~; curve 2: drift of 1 mM m-1 or
immediately after compression, t = 0 sec; at
less in time interval Ate; curve 3: drift in surface
t = 300 sec; and at t = 500 sec (Fig. 3). pressure corresponding to a relative area decrease of
Using this data, we examined three different 10-~ or less in time interval At,.
arbitrary stability limits (5). For each limit
we determined the highest surface pressure ing to a relative area decrease of 10-3 or less in
at which the monolayer satisfied the stability the time interval At2.
criterion. The criteria were
These results are shown in Fig. 4.
(a) a drift of 1 m N m - : or less in the time Relating monolayer stability to the strength
interval Atl2 (Fig. 3); of molecular interaction is a procedure of
(b) a drift of 1 m N m -1 or less in the time dubious validity in view of the variety of
interval At,~ (Fig. 3) ; mechanisms which m a y contribute to in-
(c) a drift in surface pressure correspond- stability (5). Nevertheless, an enhanced
collapse pressure has sometimes been taken to
indicate excess interaction in mixed mono-
P'?I2 1.1
layers (6, 7). In our results (Fig. 4) there does
I~
Q:
?~tl I' I A t2 not appear to be any correlation between
if)
m
monolayer stability as measured by any one
w
of the above criteria and the excess free energy
o,.
w of mixing (Figs. 10 and 13).
o
a
MOLECULAR AREAS

, , J , I The area of a two component monolayer can

300 500 be compared with that of the unmixed com-
TIME, t/s
ponents at the same surface pressure to give
FIG. 3. S u r f a c e p r e s s u r e d r i f t s w i t h t i m e ( s c h e m a t i c ) . the excess area or excess area of mixing, A,~iE:
C u r v e 1 : t y p i c a l b e h a v i o r of s u l p h a t e - r i c h m o n o l a y e r s ;
curve 2: typical behavior of alcohol-rich monolayers. ~AE = A.,s - (XAA;~+ x~i~). [13
Journal of Colloid and InterJace Science, Vol. 51, No. 1, April 1975
 110 COSTIN AND BARNES

ponent monolayer can be treated as a two-

dimensional binary solution: that is, that the
-- 0 water which supports the monolayer can be
ignored in the thermodynamic treatment. Such
an approach has several advantages: much of
the thermodynamic theory of three-dimen-
% sional binary systems can be adapted to the
-0'01
two-dimensional case; and a knowledge of the
amount of water in the surface is not required.
On the other hand, the water may well be the
most significant component, and the theories
of Fowkes (10, 11) and Lucassen-Reynders
'~ -0'02
(12, 13) (discussed later) are based on this
premise.
For the octadecanol-docosylsulphate system
the plots of area per molecule against composi-
-0"03 i i i i I tion (Fig. 2) show that the monolayer areas
0'2 0"4 0"6 0'8 I'0 for mixed films are less than the areas of the
COMPOSITION, xA corresponding unmixed films at all composi-
tions and at all surface pressures. In other
FIG. 5. Excess area per molecule as a function of
composition at various surface pressures. Point symbols words, the excess areas are always negative, as
indicate different surface pressures as in list of symbols shown in Fig. 5. According to the treatment
(1). outlined above, these observations mean that
the components are miscible and that the sur-
If the surface solution is ideal, or if the two face solution is nonideal.
components are immiscible, the excess area will However Phillips and Joos (14) have re-
be zero and the plot of HAs against xx will be cently pointed out that a plot of ~i against XA
linear (8). Deviation from these conditions which is not linear overall, but does contain a
indicates miscibility and nonideality (9). Im- linear segment, can be an indication that the
plicit in these definitions of excess area and of components are only partially miscible. Thus
ideality is the assumption that a two-corn- in the present case, the linear region between

~ 60 I I I I I I I I I l

~ 40
~ 30

o I 0.112 Ii I
o. 4 0.16 • 2
PARTIAL MOLECULAR AREA, ]A /rim2 molecule'l
FIC. 6. Surface pressure as a function of partial molecular area of octadecanol in octadecanol-docosyl
sulphate monolayers. BB, alcohol-rich; ¢, intermediate; O, sulphate-rich composition regions.
Journal of Colloid and Interface Science, Vol. 51, No. I, April 1975
 TWO-COMPONENT MONOLAYERS. II 111

60 I I I I i I I I I
z
~ 5o

~40
W
n,"

~ 30
u.I

a. 20
w
¢j

u. Io
n--

o.i 8 0.20 0.22 0.24 (>26

PARTIAL MOLECULAR AREA, A s / r i m 2 m o l e c u l e "1

FIG. 7. Surface pressure as a function of partial molecular area of docosylsulphate in octadecanol-

docosyl sulphate monolayers.., alcohol-rich; 0, intermediate; e, sulphate-rich compositionregions.

x~, = 0.18 and XA = 0.6 could represent a two- times been adduced when sharp changes in
phase monolayer, one phase being a saturated the area-composition plot occured at simple
solution of the sulphate in the alcohol (XA molecular ratios (3, 19, 20). In addition it
= 0.6) and the other being a saturated solution has been suggested that molecules such as
of the alcohol in the sulphate (XA = 0.18). cholesterol have the ability to condense ex-
It is difficult to test the miscibility of a two- panded monolayers (21). Our present results
component monolayer (8, 15, 16), but there can be satisfactorily explained without in-
has been a suggestion that evaporation re- voking either of these concepts.
sistance data could be used for this purpose For a nonideal surface solution, defined in
(17). This suggestion has not yet been properly terms of Eq. [-1], the molecular area can be
explored, but if it is valid then our evaporation expressed as a function of the partial molecular
resistance data (Part III) indicate complete areas (AA, -ds) of the two components:
miscibility rather than the interpretation dis-
~{AS = XAYA + XSXS. [-2]
cussed above. For the present we will therefore
assume complete miscibility. These partial molecular areas can be deter-
Examination of the area data shows that mined graphically by a standard procedure
there are three regions of composition exhibit- (22). In applying this method to the present
ing different patterns of behavior and with data (Fig. 2) it was assumed that the plots of
well-defined transitions between them at _JAS(=~) against xx are linear within each
xA = 0.18 and xA = 0.6. We will refer to these of the three composition regions. Extrapola-
regions as the sulphate-rich, intermediate, and tion of these straight lines to XA = 1 and to
alcohol-rich composition ranges. xA = 0 gives respectively the partial molecular
Area-composition plots similar to those in areas of the alcohol and of the sulphate for
Fig. 2 have been reported by several workers that particular composition region and surface
and much of this work has been reviewed by pressure. In Figs. 6 and 7 we show these
Dervichian (3) and by Cadenhead (18). In this results presented in the same manner as
earlier work the emphasis has been on molecules surface pressure-area isotherms.
of greater structural complexity than in the In the sulphate-rich monolayers the partial
present system and this has undoubtedly in- molecular areas of docosyl sulphate are
creased the difficulties of interpretation. The identical with the areas per molecule of a
formation of molecular complexes has some- pure docosyl sulphate monolayer. The octa-
Journal ol Colloid and Interface Science, V o l . 51, N o . 1, A p r i l 1 9 7 5
112 COSTIN AND BARNES

decanol, on the other hand, has extremely low alcohol-rich region appear to be slightly lower
partial molecular areas. Thus it would appear than the areas per molecule of most one-
that, throughout this composition range, the component monolayers and of octadecanol
spacing of the docosyl sulphate ions is deter- monolayers in particular. It seems likely that
mined primarily by the same factors as in these lower areas reflect the greater strength
pure docosyl sulphate monolayers. The low of the head group-head group attraction when
partial molecular areas of the alcohol indicate an ion interacts with a dipole, compared with
that this sulphate spacing is such that the the interaction between two dipoles, which is
alcohol molecules can be accomodated between the usual situation in a one-component, highly
the sulphate ions with only a v e r v small in- coherent, monolayer. Other data showing a
crease in the monolayer area. similar effect are to be seen in the areas of
The apparent linearity of the ./I-XA plot monolayers of carboxylic acids when the pH of
in the alcohol-rich region (Fig. 2) means that the substrate is raised and partial ionization
the partial molecular areas of octadecanol are of the monolayer ensues (23).
the same as the areas per molecule of pure The II-ti isotherms for the intermediate
octadecanol monolayers. In this region the region (Figs. 6 and 7) have been calculated
partial molecular areas of the docosyl sulphate on the assumption that the area-composition
give an isotherm which is very similar in form plots (Fig. 2) are linear within this composi-
to the H-z{ isotherm of a condensed mono- tion range: i.e., that the partial molecular
layer of a nonionized long-chain compound areas are constant. This interpretation of the
such as an alcohol or a carboxylic acid at low data is unequivocal for low surface pressures,
pH (Fig. 7). Indeed, this isotherm represents but at high surface pressures (/> 25 naN m -1)
the II-~ isotherm of a docosyl sulphate some nonlinearity of the area-composition
monolayer with the effect of the mutual repul- curves is evident. Thus the isotherms presented
sion of the ionized sulphate groups removed. for this region in Figs. 6 and 7 represent the
Obviously this is only possible because of the average partial molecular areas.
separation of the sulphate ions in the matrix The isotherm for the alcohol shows two
of alcohol molecules. If the mole fraction of distinct regions, the region at higher surface
sulphate is increased to the extent that some pressure having larger partial molecular areas
sulphate ions must be near neighbors in the than that at lower surface pressures. This ap-
monolayer the partial molecular area of sul- pears to be a consequence of the large changes
phate will begin to rise. It is possible to make in the partial molecular area of the sulphate
a rough calculation of the composition at which with surface pressure: at low surface pressure
this transition will occur. If we assume that ,4s is large and the alcohol can be partly ac-
the monolayer is a two-dimensional hexagonal comodated in gaps between the sulphate
array of vertically oriented cylinders, the molecules, while at high II this is not possible
maximum ratio of sulphate to alcohol molecules to the same extent.
without any sulphate ions being nearest Fowkes (11) has also studied an alcohol-
neighbors is 1:2 corresponding to xA = 0.67. sulphate system, but there are significant
Reference to Fig. 2 shows that the transition differences between his system and ours:
between the alcohol-rich region and the inter-
Fowkes used hexadecyl sulphate on a sub-
mediate region occurs at about XA = 0.63 if
strate containing sodium chloride (c. 1.2 M)
the transition is assumed to be sharp. A more
gradual transition could well begin at a higher and consequently the molecular areas of the
mole fraction of alcohol. In either case the pure sulphate monolayer and of the sulphate-
agreement is reasonable. rich mixed monolayers were very much larger
It is interesting that the magnitudes of the than in our system. In Fowkes' system the
partial molecular areas of the sulphate in the sulphate-rich monolayers gave partial molecu-
Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
 TWO-COMPONENT~_MONOLAYERS. II 113

lar areas e of sulphate which were almost 0,6

I I I I
I
identical with those of pure sulphate, but the
partial molecular areas of the alcohol were "~ 0.5
almost zero. It seems, therefore, that the E

much greater spacing of the sulphate ions in

Fowkes' system would allow addition of alcohol ~o.4 '~
without any significant change in monolayer
area up to composition of xA ~< 0.5. An area- ~ 0'3
composition plot prepared from Fowkes' data LO
is shown in Fig. 8. .J

The differences between these two systems

are also demonstrated when they are analyzed U.I
by means of Fowkes' theory of two-dimensional 0"1

solutions. This will be discussed later. W

n~
"Condensation" has been defined by Caden- 0
0'2 0'4 0"6 0'8 I'0
head and Phillips (9) as a process leading to
COMPOSITION, XA
negative values of the excess area of mixing
(see Eq. [1]). This definition is clear, but its FIG. 8. Area per molecule as a function of composi-
application is customarily restricted to the tion for hexadecanol-hexadecylsulphate monolayers.
effect of adding small amounts of a substance Data of Fowkes (11). Point symbols indicate different
which forms condensed monolayers to a sub- surface pressures as in list of symbols (1).
stance which, in the pure state, forms ex-
it is inappropriate to use the term "condensa-
panded monolayers. Furthermore, there are
tion" in such cases where the component being
connotations which are not necessarily ap-
"condensed" is already in a condensed state.
plicable to all cases of condensation. Often
We suggest that the term "contraction"
the existence of a strong excess interaction is
should be employed to describe any system for
implied, but condensation can also arise from
which the excess area of mixing is negative and
an athermal, space-filling, mechanism (3, 24,
that a new and more restricted definition
25), and additional evidence must be used to
should be formulated for the term "con-
distinguish between these possibilities. The
densation."
implication that in condensation the molecules
We have already seen that it is possible to
of an expanded component are drawn closer
explain the contraction of docosyl sulphate by
together is only correct in the rare case where
octadecanol in terms of a space-filling mech-
the partial molecular area of the other com-
anism, and that the apparent contraction of
ponent is negative (26).
octadecanol by docosyl sulphate results in an
The present system differs from the systems
approximately constant area per hydrocarbon
studied by Cadenhead and Phillips (9) and
chain (for xA > 0.6) because the electrostatic
from most of the systems hitherto examined
repulsion between sulphate ions is negligible.
(3, 26) in that both components form con-
Whether there is, in addition to these mecha-
densed monolayers. It is interesting, therefore,
nisms, any contribution from an excess inter-
to find that the present system fulfils the formal
action is a point which will be discussed when
requirement for condensation of docosyl sul-
the free energy of mixing data are considered.
phate monolayers by octadecanol and, indeed,
of octadecanol by docosyl sulphate. Clearly, FOWKES' THEORY
I n t h e t h e o r e t i c a l t r e a t m e n t of his d a t a , F o w k e s '
i n c l u d e s w a t e r a s a c o m p o n e n t so t h a t t h e t e r m partial
In the theory developed by Fowkes (10, 11,
molecular area as u s e d b y h i m h a s a d i f f e r e n t m e a n i n g to 27) a distinction is made between monolayers
t h a t i n t e n d e d here. which are "surfactant-continuous" and those
Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
114 COSTIN AND BARNES

which are "water-continuous." For a mono- values for z from which inferences about the
layer which is wholly or partly "water-con- monolayer structure mav be drawn.
tinuous" the surface pressure is treated as an The value of z is determined by plotting the
osmotic pressure and is related to the mole measured e{ against [Nw/z(NA + Ns)], the
fraction of water in the surface layer (xw 8) by values of which are calculated from the mea-
the relationship : sured surface pressures by Eqs. [3] and [4]
and by assuming ideality (Eq. [6]). It is also
1I = -- ( k T / X w ) In xwSfw s, [3]
necessary to assume that AA and Ks are inde-
where fw ~ is the relevant activity coefficient. pendent of surface pressure and to assign a con-
In calculating Xw~ all species relevant to the stant value of Aw. In essence, Fowkes attrib-
development of an osmotic-type pressure in utes any difference between the molecular
the "water-continuous" region of the surface areas of the mixed monolayers and of the cor-
must be considered. Consequently, allowances responding unmixed monolayers to a change in
must be made for the dissociation or associa- the number of water molecules in the surface
tion of monolayer molecules and for mono- rather than to changes in the partial molecular
layer species which are not in the "water- areas of the monolayer molecules as in the
continuous" layer. The factor, z, is introduced approach discussed earlier. It is therefore
for this purpose. With a two-component mono- possible for a system to exhibit large negative
layer, z could have different values for each excess areas of mixing (Eq. [i-l) and yet be
component, but for ease of computation an ideal in the sense intended by Fowkes.
average z is used. Thus: An analysis of our data for the octadecanol-
docosyl sulphate system by the Fowkes method
xw ~ = ~Vw/E~Vw + z ( X , + x s ) ] . [4]
is shown in Fig. 9 and the resultant values of z
For a two-component monolayer with water are given in Table II. In these calculations Aw
counted as a third component, Eq. [2] is was taken as 0.097 nm 2, the value used by
written: Fowkes (11). It is useful to compare these
results with those obtained by Fowkes (11)
A NA for the hexadecanol-hexadecyl sulphate
NA + Ns N. + Ns system.
With both systems the plots of Eq. [-5] were
Ns Nw curved, but the curves appeared to approach
+ ,is + - - Aw
A:A + Ns NA + Ns limiting straight lines at low ./i. It was the
slopes of the limiting lines which were used to
Ni A's calculate z values. This failure to find the
- - - -AA ~ A s
NA + Ns IVA + Ns linear relationship expected from Eq. [5] was
more marked in our system than in that of
Nw Fowkes. There are several possible explana-
+ z2w. [s]
z(-¥A + Ns) tions : the values of the partial molecular areas
may not be independent of surface pressure,
In this theory, an ideal surface solution is or the assumption of ideality (Eq. [6]) may
defined as one for which the activity coefficient
be incorrect.
of water in the surface, fw s is unity:
For the hexadecanol-hexadecyl sulphate
fw s-- 1. [6] system, Fowkes found values of z which cor-
Fowkes has used these equations ([3] to [-6]) responded to the fraction of sulphate in the
to show that experimental data are consistent monolayer provided xs/> 0.67. Thus for these
with the assumption that the surface solu- sulphate-rich monolayers, Z/xs was unity.
tion is ideal (Eq. [6]) and to determine Fowkes interpreted this as evidence that the
Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
 TWO-COMPONENT MONOLAYERS. II 115

0'25 I I I I 0 clude, therefore, that the docosyl sulphate does

not behave ideally in the sense of the Fowkes
0'24 - 0 0
theory.
oo Lucassen-Reynders (13) has recently pub-
0
@ lished a treatment of two-component mono-
ME0"23
layers which is based on the same osmotic-type
O
approach as the Fowkes' theory (Eq. [3]).
e~ 0.22 @ e However, she does not distinguish between
"water-continuous" and "surfactant-continu-
0

~ 0,2I
ous" layers, preferring a more rigorous ap-
g, 3 proach in which all deviations from ideality
IE
o are assigned to the activity coefficients. We
re 0.20 have not yet analyzed our data according to
LO
O..
this theory.
E 0'19
.w
~V 60 v

E X C E S S F R E E E N E R G I E S OF M I X I N G

o.18 I I I I A method for determining the excess free

o 0.5 l.O ~.5 z.o z.5 energy of mixing in a two-component mono-
COMPOSITION, N w / z ( NA+Ns)
layer has been devised by Goodrich (28). The
FIG. 9. Analysis of data of Fig. 1 by the method of free energy here is defined by
Fowkes (11). Curve numbers refer to monolayer com-
position (mole percent of octadecanol). G= U - TS + P V - - T A , [7-]
and the excess free energy is given by an
sulphate, which is "water-continuous," be-
expression analogous to Eq. [-1-] :
haves ideally, while the alcohol, which is
"surfactant continuous," is in a "dehydrated AGE = GAs 1 (xAGX + xsG~). [83
state" and "rests on top of the surface layer
Goodrich showed that the excess free energy
of water molecules." Values of Z/Xs less than
one indicate partial dehydration of the sul- TABLE II
phate and its withdrawal from the "surface ANALYSIS OF OCTADECANOL-DocosYL SULPHATE
solution 'state.' " A n alternative interpretation MONOLAYERS ACCORDING TO THE
of these results in terms of Eq. [2] has been THEORY OF FOWKES (10)

indicated earlier. In particular, unit value for Series a xA xs z $/x8

z/xs would correspond to a partial molecular
area of alcohol of zero, and replotting of 2 0.00 1.00 0.46 0.46
i 0.00 1.00 0.43 0.43
Fowkes' data shows that this indeed the case
2 O.10 0.90 0.36 0.40
for low surface pressures (Fig. 8). 1 0.11 0.89 0.41 0.46
Table II shows that the analysis of the octa- 2 0.18 0.82 0.31 0.38
decanol-docosyl sulphate system yields quite 1 0.20 0.80 0.31 0.39
different results. The values of z/xs are always 2 0.25 0.75 0.21 0.28
less than 0.5, even for the pure sulphate mono- 2 0.33 0.67 0.16 0.24
1 0.39 0.61 0.15 0.25
layer. According to the Fowkes theory, this 2 0.50 0.50 0.09 0.18
would mean that only a fraction of the sulphate 1 0.58 0.42 0.10 0.2
molecules are in the "surface solution 'state.'" 2 0.70 0.30 0.06 0.2
This interpretation appears very reasonable 1 0.80 0.20 0.06 0.3
when one considers that the docosyl sulphate 2 0.87 0.13 0.06 0.5
has a much longer hydrocarbon chain than the a Series 1 refers to a n earlier set of results n o t g e n e r a l l y
hexadecyl sulphate used by Fowkes. We con- discussed here.

Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
116 COSTIN AND BARNES

TABLE III
ExcEss FREE ENERGIES OF MIXING FOR OCTADECANOL--DOCOSYLSULPHATEMONOLAYERSON
0.1 M KC1 AT 298.3 K

H / m N m -I 5 10 15 20 25 30 35 40 45 50
xA AGE/J mole-~

0.10 30 65 100 135 180 225 270 320 370 425

0.18 50 105 160 220 290 360 435 505 585 665
0.25 50 105 165 230 300 375 455 525 610 685
0.33 55 115 175 240 310 390 465 540 610 680
0.40 60 120 180 245 320 395 465 535 605 670
0.50 55 115 175 240 310 375 440 500 555 610
0.60 60 115 175 245 310 375 435 495 550 600
0.70 50 95 145 200 250 300 345 385 425 465
0.87 20 50 75 100 130 150 175 200 220 240

could be calculated from the II-~ isotherms The integrals in Eq. [-91 were evaluated
of the mixed monolayer and of the monolayers from the II-x{ isotherms of the relevant mono-
of the two pure components layers by the method of counting squares. By
this means, values of AG E were obtained at
AG E =
L
[£ n
~AsdlI XA
f] 2{).dH
several surface pressures for each composition
of the mixed monolayer. These values are
tabulated in Table III and plotted in Fig. 10.
The estimated error in a single value of 2~GE
1 is -4-211 J mole-1, where I I / m N m -1 is the
upper limit of the integration. However the
where lI* is a surface pressure below which the self-consistency of the data for any given sur-
monolayers can be assumed to mix ideally. face pressure suggests that the actual error is
He suggested that the excess free energy of much less than this estimate.
mixing of the monolayer vapors would be It is possible to draw two types of curve
negligible in comparison with that of the through the points for any one surface pres-
condensed monolayers so that in evaluating sure. Figures 10 and l l illustrate the alter-
the intergrals it would be possible to ignore native curves for several surface pressures,
the expanded and the condensed-expanded Curves of the first type (Fig. 10) correspond
transition regions [-areas greater than ~{L in in shape to the curves found by Goodrich (28)
Fig. 1 of Part I (1)] where the surface pressure for a similar mixed monolayer system: they
is too low for detection by the normal film are smooth curves with minima. In our
balance: thus II* ~ 0. system the minimum is very shallow at low
It should be noted that the concept of surface pressures, and as the surface pressure
ideality used in the Goodrich theory is consis- is increased the minimum becomes deeper and
tent with that implied by Eq. [-1]. seems to shift to lower alcohol concentrations.
In applying the Goodrich approach to the The minima are too fiat and the results not
present system, in which both monolayers are sufficiently accurate for the positions of the
condensed, we initially considered that a minima to be fixed with any certainty. In fact,
negligible error would be introduced if II* was the second type of curve (Fig. 1 l) with a nearly
taken as the surface pressure at which the horizontal region at intermediate compositions,
monolayer becomes condensed. With our fits the results equally well.
apparatus this surface pressure was effectively For the hexadecanol-hexadecyl sulphate
zero. system Goodrich (28) found a minimum in the
Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
 TWO COMFONENT MONOLAYERS. II 117

0'1 the mixed monolayer has the same shape as

that of the corresponding ideal mixture, but
is shifted to lower areas. This in turn means
that the excess area is independent of surface
"i"D
pressure-a conclusion which could have also
-0'l
E
O been reached by considering the relationship

=" -0'2

<3

>_~ -0-3 Reference to Fig. 5, which shows values of

(9
re"
U.I
A~g determined directly from the II-~
Z
UJ - 0 " 4 isotherms, shows that they are indeed con-
bJ
stant for any given composition with xn > 0,5.
Ld
For the other monolayers giving curved plots
.0.5 ¸
in Fig. 12, Eq. [10] indicates that A~ E be-
comes more negative with increasing pressure
-0"6 and again reference to Fig. 5 confirms this.
It is interesting to note that all of the curves
in Fig. 12 pass through the origin smoothly
-0.7
0 0,2 0'4 0'6 0-8 1.0 without any marked changes in slope. This
COMPOSITION, xA might appear to support the assumption made
in evaluating AGE, that the high area parts of
FIG. i0. Excess free energy of mixing, as measured
by integration under the II-~ isotherms of Fig. 1,
the II-~{ isotherms could be ignored. However
plotted against composition. Point symbols indicate several workers have argued that this assump-
different surface pressures as in list of symbols (1). tion is not valid (29, 30) and that there can be
a significant contribution to the excess free
excess free energy curve at mole fraction 0.5 energy arising from the high area, low surface
and surface pressure 32 mN m -1 of approxi- pressure regions of the monolayers. This
mately --1250 J mole-1. The corresponding assertion has now been given convincing ex-
minimum for the octadecanol-docosyl sulphate
system (Fig. 10) is found at an alcohol mole 0 I I I I
fraction of about 0.4 and has a depth of
- 4 3 0 4- 60 J mole-L However, these values L-
co -0'l
are not comparable because the two systems
differ in several important ways : in the concen- %
tration and nature of the salt in the substrate, ,q - 0 ' 2

in the chain lengths of the monolayer mole- >:

(9
cules, and in the different alcohol and sulphate n-
UJ - 0 ' 3
chain lengths in our system compared with Z
UJ
equal chain lengths in the system used by b.I
Goodrich. bJ
n- - 0 ' 4
Lu
The relationship between AGE and surface
pressure is shown graphically in Fig. 12. For -o.5 I I
monolayers with xA > 0.5 these plots are 0 02 0.4 0'6 0'8 1.0
COMPOSITfON, XA
linear, while those monolayers richer in sul-
phate give plots which are slightly curved. The FIG. 11. Selected data from Fig. 10 showing alter-
linear plots indicate that the II-~ isotherm of native curves.
Journal of Colloid and Interface Science, V o l . 51, N o . 1, A p r i l 1975
 118 COSTIN AND BARNES

700

600

"T 5O0
-6
E

<I. 4 0 0

>:
re 3 0 0
W
Z
ti.J

h, 200
,'it
U.

100

0
0 I0 20 50 40 50
SURFACE PRESSURE, 7 T / m N m -I

FIG. 12. Excess free energy data from Fig. 10 as a function of surface pressure. Point symbols indicate
different monolayer compositions as in list of symbols (1).

perimental support by the work of Pagano and area portions of the I I - ~ isotherms by adding
Gershfeld (31). In the system they examined, values of AGgE [-obtained by extrapolation of
the contribution from the high area region was AG+E - AG,,e~ E plots (Part I I I , Fig. 5)-] to
-

twice as great as that from the condensed the measured excess free energies (Part I I I ,
region. Eq. [9]). These curves may be compared to
We have not made any measurements on those in Fig. 14 where we have plotted values
our system at high areas. However, we show of AG E estimated from evaporation resistance
in Part I I I that it may be possible to use data using the equation:
evaporation resistance data to estimate the
AG +E = -- KAG E, [-11-]
contribution that the high area regions would
make to the values of AG E. These values are and the value of K determined in Part I I I .
shown in Fig. 13. Unfortunately it has not At the higher surface pressures, the curves in
been possible to make reasonable estimates for Figs. 13 and 14 are in reasonable agreement,
some of the sulphate-rich monolayers because but there are marked differences at low sur-
of limitations in the accuracy of the evapora- face pressures. These differences arise mainly
tion resistance data. I t should also be em- from the gross uncertainties in AG %E values
phasized that the estimates which are given in for monolayers with low evaporation resist-
Fig. 13 are dependent on the validity of the ances (see Part I I I ) , but there are some in-
relationship between AG E and AG +E. These dications that this may not be the sole explana-
points are discussed more fully in Part I I [ . tion. The maxima and minima in the curves for
Also shown in Fig. 13 are some values of 10 m N m -1 (Fig. 14) and 5 m N m -1 (not shown)
AG E which have been corrected for the high are supported by data at more than one corn-
Journal of Colloid and Interface Science, Vol. 51, No. 1, April 1975
 TWO-COMPONENT MONOLAYERS. II 119

position, suggesting that they should not be 0"2

I I I I
entirely attributed to random errors. Thus
there is still some doubt about the validity of
Eq. Ell~ at low surface pressures.
The positive values of the excess free energy
of mixing shown in Fig. 13 suggest the possi- ~J
bility of a phase separation in the monolayers
at low surface pressures. To examine this -0"1
question we need to calculate the free energy
of mixing, AG. This is given by the relation
Z -0'2
b.I
AG = AG E "k- AG id,

where AGid can be calculated from

[-12~
"' "'
-o.s "
# X/ '

AGid = RT(xA in XA + Xs in Xs). [-13-]

-0,4 I I
For an equimolar composition we thus have 0 0'2 0'4 0'6 0-8 t'0
COMPOSITION, xA
AxGE = 0.2 kJ mole -1 (from Fig. 13),
FIG. 14. Excess free energy of mixing as a function
and of composition. Values obtained from evaporation
AG id = -- 1.7 kJ mole -1, resistance data using Eq. [15-1, with K = 6.3. Curve
numbers refer to surface pressure (raN m-l).

0.3 1 I I I
whence
AG = -- 1.5 kJ mole -1.
0.2 -- , ~ / ~ -
The two-dimensional solution state is there-
fore more stable, thermodynamically, than
o,,- / the corresponding unmixed state in this case
o / where the most positive value of AG E was
/ ~
used. Other cases would show more negative
% values of AG. This observation supports our
original assumption that the octadecanol and
-0.1
; k the docosyl sulphate were completely miscible

.-o" in the monolayer state.

CONCLUSION
"' -0.3 -
\\ \ The first question that must be considered
in any study of a two-component monolayer
system concerns the miscibility of the com-
ponents in the monolayer (8, 15). For the
octadecanol-docosyl sulphate system there is
-0.5
evidence of several kinds which shows that the
f I I I
0 0"2 0"4 0'6 0'8 I'O system is completely miscible. Evidence based
COMPOSITION, xA on evaporation resistance data is not conclu-
sive as the validity of this test has not yet been
FIG. 13. Excess free energy of mixing as a function
property explored. A more convincing agru-
of composition. ©, AGUE;other curves: data of Fig. 10
corrected for AG,E. Point symbols indicate different ment comes from the large negative values of
surface pressures as in list of symbols (1). the free energies of mixing.
Journal of Colloid and Interface Science, V o l . 51, N o . 1, A p r i l 1975
120 COSTIN AND BARNES

The second question concerns the ideality as indicators of ideality or nonideality is dis-
of the two-dimensional solution, and here we cussed in the following paper (Part III).
encounter the problem of defining ideality
discussed in the introduction. REFERENCES
It was suggested there that a linear depen-
dence of some monolayer property on composi- 1. COSTIN, I. S., AND BARNES, G. T., J. Colloid
Interface Sci. 51, 94 (1975).
tion might possibly be an indication of ideality,
2. CRISP, D. J., Res., London 2, Suppl. Surface Chem.
but we can see clearly from the examination of 17, 23 (1949).
monolayer stability using several different 3. DERVlCHIAN, D. G., in "Surface Phenomena in
criteria that some monolayer properties have Chemistry and Biology" (J. F. Danielli, K. G. A.
no apparent relation to the ideality of the Pankhurst, and A. C. Riddiford, eds.) p. 70.
system. Pergamon, New York, 1958.
Probably the concept of ideality most 4. BROOKS, J. H,, AND ALEXANDER, A. E., in "Re-
tardation of Evaporation by Monolayers :
generally accepted is that based on the excess
Transport Process" (V. K. La Mer, ed.), p. 245.
free energy of mixing, with proper allowance Academic Press, New York, 1962.
for the high area parts of the surface pressure- 5. GAINES, G. L., JR., "Insoluble Monolayers at
area isotherms. Using this criterion, we con- Liquid-Gas Interfaces," p. 143-151. Wiley, New
York, 1966.
clude that the octadecanol-docosyl sulphate
6. SCHULMAN,J. H., STENHAGEN,E., Proc. Roy. Soc.,
system is nonideal, exhibiting marked nega-
London, Set. B 126, 356 (1938).
tive deviations from ideality at high surface 7. GODDARD,E. D., AND SCHULMAN,J. H., J. Colloid
pressures and small positive deviations at low Sci. 8, 329 (1953).
surface pressures (Fig. 13). The nonideality 8. GAINES, G. L., JR., See Ref. (5, p. 282, 288).
of the system is also evident in the area- 9. CADENHEAD,D. A., AND PHILLIPS, M. C., Advan.
composition plots (Fig. 2), the monolayer Chem. Set. 84, 131 (1968).
being more contracted than an ideal one at 10. FOWKES, F. M., J. Phys. Chem. 66, 385 (1962)"
all compositions and surface pressures. 11. FOWKES, F. M., d. Phys. Chem. 67, 1982 (1963).
12. LUCASSEN-REYNDERS,E. H., J. Colloid Interface
The latter statement implies a concept of
Sci. 41, 156 (1972).
ideality based on Eq. [-1] and compatible with 13. LUCASSEN-REYNDERS,E. H., J. Colloid Interface
the excess free energy criterion. However, Sci. 42, 554 (1973).
Lucassen-Reynders (13) disputes both of 14. PHILLIPS, M. C. ANDJoos, P. Kolloid Z. Z. Polym.
these definitions of ideality because the im- 238, 499 (1970).
portant contributions of the water are ignored, 15. PAGANO, R. E., AND GERSHFELD, N. L., J. Phys.
and proposes an alternative definition based Chem. 76, 1238, 1244 (1972).

on the osmotic approach to monolayer proper- 16. PAGANO,R. E., AND GERSHFELD,N. L., J. Colloid
Interface Sci. 44, 382 (1973).
ties. However, this approach also encounters
17. GAINES,G. L. JR., See Ref. (5, p. 290).
difficulties in that the partial molecular areas
18. CADENHEAD,D. A., in "Recent Progress in Sur-
of the water in the surface and of the monolayer face Science," (J. F. Danielli, A. C. Riddiford,
components cannot be determined and and M. Rosenberg, eds.), Voh 3 p. 169, 1970.
arbitrary assumptions must be made about 19. SHAH,D. 0., J. Colloid Interface Sci. 32, 577 (1970).
their values and about their dependences on 20. SHAH, D. 0., J. Colloid Interface Sci. 37, 744
composition and surface pressure. We propose (1971).
21. LEATHES,J. B., Lancet 208, 853 (1925).
to examine these questions in more detail in a
22. LEwis, G. N., RANDALL, M., PITZER, K. S., AND
later publication. BREWER, L., "Thermodynamics," 2nd ¢d., p.
The use of evaporation resistance data and 207. McGraw-Hill, New York, 1961.
the derived excess free energies of activation 23. SPINK, J. H., d. Colloid Sci. 18, 512 (1963).
Journal of Colloid and Interface Science, Vol. 51, No. 1. April 1975
 TWO-COMPONENT MONOLAYERS. II 121

24. SHAH, D. O., AND SCHULMAN,J. H., Y. Lipid Res. 28. GOODRXCH,F. C., Proc. Internat. Congr. Surface
8, 215, 227 (1967). Activity, 2nd, I, 85 (1957).
29. SHAH, D. O., Arm CAVVS,R. W., J. Colloid Inter-
25. SHAH, D. O., AND SC~ULUAN,J. H., Advan. Chem.
face Sci. 27, 319 (1968).
Ser. 84, 189 (1968). 30. GERSHFXLD, N. L., Tech. Surface Colloid Chem.
26. BARNES, G. T., Specialist Periodical Rept: Colloid Phys. 1, 1 (1972).
Sci. 2, in press. 31. PAGANO,R. E., Am) GERSHVELD,N. L., Y. Colloid
27. Fow~zt:s, F. M., J. Phys. Chem. 65, 355 (1961). Interface Sci. 41, 311 (1972).

Journal of Colloid and Inter.faceScience, Vol. 51, No. 1, April 1975