See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/283516284

Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and

Capillary Electrophoresis

Chapter · October 2015

DOI: 10.1016/B978-0-12-409547-2.11559-8

CITATIONS READS

4 20,378

5 authors, including:

Xavier Subirats Elisabet Fuguet

University of Barcelona University of Barcelona
52 PUBLICATIONS   726 CITATIONS    85 PUBLICATIONS   2,273 CITATIONS   

SEE PROFILE SEE PROFILE

Martí Rosés Elisabeth Bosch

University of Barcelona University of Barcelona
212 PUBLICATIONS   8,567 CITATIONS    188 PUBLICATIONS   8,022 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Modelization and characterization of HILIC View project

Physicochemical characterization and estimation of the activity of bioactive compounds View project

All content following this page was uploaded by Elisabeth Bosch on 12 October 2017.

The user has requested enhancement of the downloaded file.

This article was published in the Elsevier Reference Module in Chemistry, Molecular Sciences and
Chemical Engineering, and the attached copy is provided by Elsevier for the author’s benefit and
for the benefit of the author’s institution, for non-commercial research and educational use
including without limitation use in instruction at your institution, sending it to specific colleagues
who you know, and providing a copy to your institution’s administrator.

All other uses, reproduction and distribution, including without limitation commercial reprints,
selling or licensing copies or access, or posting on open internet sites, your personal or
institution’s website or repository, are prohibited. For exceptions, permission may be sought for
such use through Elsevier’s permissions site at:

http://www.elsevier.com/locate/permissionusematerial

Subirats X., Fuguet E., Rosés M., Bosch E. and Ràfols C. (2015) Methods for pKa Determination (I):
Potentiometry, Spectrophotometry, and Capillary Electrophoresis. In: Reedijk, J. (Ed.) Elsevier
Reference Module in Chemistry, Molecular Sciences and Chemical Engineering. Waltham, MA:
Elsevier. 29-Oct-15 doi: 10.1016/B978-0-12-409547-2.11559-8.

© 2015 Elsevier Inc. All rights reserved.

 Author's personal copy

Methods for pKa Determination (I): Potentiometry, Spectrophotometry,

and Capillary Electrophoresis
X Subirats, Universitat de Barcelona, Barcelona, Spain
E Fuguet, Universitat de Barcelona, Barcelona, Spain; Serra Hunter Programme, Generalitat de Catalunya, Barcelona, Spain
M Rosés, E Bosch, and C Ràfols, Universitat de Barcelona, Barcelona, Spain
ã 2015 Elsevier Inc. All rights reserved.

Theoretical and Operational Definitions of pH and pH Measurement 1

Acidic Dissociation Constant Definitions 2
Potentiometric Titrations for pKa Determination 3
Potentiometric Methods 3
pKa Calculation from Potentiometric Data 4
Spectrophotometric Measurements for pKa Determination 5
Single-Wavelength Spectrophotometric Approach 6
Multiwavelength Spectrophotometric Approach 7
pKa Determination by Capillary Electrophoresis 8
Capillary Zone Electrophoresis of Acid–Base Analytes 8
Buffer and Instrumental Considerations for pKa Determination by CZE 8
References 10

The acidity and basicity of an organic compound are significant physicochemical characteristics since the ionization state of a
molecule governs a number of relevant properties such as solubility, lipophilicity, and binding ability. The aim of this chapter is to
bring together the most significant methods for acidity constant determination. In all instances, proper pH measurements are
required, and, then, pH definition and practical procedure to get accurate pH values are carefully described. Since pure water is the
usual working solvent, the following discussion is referred to it. Due emphasis is given to the most used classical methods,
potentiometry and ultraviolet–visible spectroscopy, and to a separation technique, capillary electrophoresis, which nowadays is
strongly recommended for pKa evaluation and shows the advantage that it does not require pure samples for measurements.

Theoretical and Operational Definitions of pH and pH Measurement

The definition of pH, proposed by Sørensen, relates this quantity to the negative logarithm of the hydrogen ion activity, aH,
according to Ref. [1]

pH ¼  log aH + [1]

Despite activity and pH being dimensionless quantities, activity must be referred to a concentration scale through an activity
coefficient and so must be the pH quantity. This means that the same solution may have different pH values, which depend on the
scale that hydrogen ion concentration is measured. As long as the practical methods for the determination of acidity constants are
always referred to the molar scale, henceforward, the discussion will be also focused exclusively to the molar scale, and pH will be
considered as


pH ¼  log cH + gc, H + =c0 [2]

where c0 is an arbitrary constant, standing for the standard-state condition and numerically equivalent to 1 mol dm3, and gc, H + is
the single-ion activity coefficient of the hydrogen ion. Since pH is defined in terms of activity, it also depends on the standard state
of the activity, that is, the conditions for which the activity coefficient of hydrogen ion is considered to be equal to unity, and thus,
in this standard state, activity becomes numerically equal to concentration. In water, the standard state for aH + is infinite dilution of
hydrogen ion in pure water, for which gH + ! 1.2
The operational pH of a solution is obtained by comparison of the electromotive force of the test solution in an appropriate
potentiometric cell to the electromotive force of one or several standard reference solutions of known pH. Thus, the everyday pH
measurement of an unknown solution X is usually performed by the glass electrode combined with a reference electrode (silver–
silver chloride or calomel):


Glass electrodejsolution S or XjKCl c  3:5moldm3 reference electrode [3]

where S and X stand for the standard and the unknown test solutions, respectively. The overall electromotive force is given by the
equation:

Reference Module in Chemistry, Molecular Sciences and Chemical Engineering http://dx.doi.org/10.1016/B978-0-12-409547-2.11559-8 1

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
2 Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis

Table 1 pHOS values for operational standard reference solutions in water

T ( C)

Operational standard reference solution 15 20 25 30 37 40

0.1 mol kg1 potassium tetraoxalate – 1.475 1.479 1.483 1.490 1.493
0.05 mol kg1 sodium hydrogen diglycolate 1.642 1.644 1.646 1.648 1.649 1.650
0.05 mol kg1 sodium hydrogen diglycolate 3.476 3.484 3.492 3.502 3.519 3.527
Saturated (at 25  C) potassium hydrogen tartrate – – 3.556 3.549 3.544 3.542
0.05 mol kg1 potassium hydrogen phthalate 3.998 4.000 4.005 4.011 4.022 4.027
0.1 mol dm3 acetic acid + 0.1 mol dm3 sodium acetate 4.647 4.645 4.644 4.643 4.647 4.650
0.1 mol dm3 acetic acid + 0.01 mol dm3 sodium acetate 4.714 4.712 4.713 4.715 4.722 4.726
0.02 mol kg1 piperazine phosphate 6.364 6.310 6.259 6.209 6.143 6.116
0.025 mol kg1 disodium hydrogen phosphate + 0.025 mol kg1 potassium 6.891 6.873 6.857 6.843 6.828 6.823
dihydrogen phosphate
0.03043 mol kg1 disodium hydrogen phosphate + 0.008695 mol kg1 potassium 7.441 7.423 7.406 7.390 7.369 –
dihydrogen phosphate
0.04 mol kg1 disodium hydrogen phosphate+ 0.01 mol kg1 potassium dihydrogen 7.466 7.445 7.428 7.414 7.404 –
phosphate
0.05 mol kg1 Tris hydrochloride + 0.01667 mol kg1 Tris 7.933 7.788 7.648 7.513 7.332 7.527
0.05 mol kg1 sodium tetraborate 9.288 9.233 9.182 9.134 9.074 9.051
0.01 mol kg1 sodium tetraborate 9.275 9.225 9.179 9.138 9.066 9.066
0.025 mol kg1 sodium hydrogen carbonate + 0.025 mol kg1 sodium carbonate 10.098 10.045 9.995 9.948 9.889 9.866
Saturated (at 20  C) calcium hydroxide 12.780 12.602 12.431 12.267 12.049 11.959

Source: Inczédy, J.; Lengyel, T.; Ure, A. M.; Gelencsér, A.; Hulanicki, A. IUPAC Compendium of Analytical Nomenclature, Definitive Rules 1997, 3rd ed.; Blackwell: Oxford, 1998,
http://www.iupac.org/publications/analytical_compendium/.

E ¼ E0 + Ej  gpH [4]

where ideally g ¼ 0.059 V at 25  C, E0 is a constant that involves the standard potential of both glass and reference electrodes, and Ej
is the liquid junction potential of the salt bridge in the cell. To calibrate a potentiometric system, three procedures are common:
one-point, two-point, and multipoint calibration. In most practical applications, the two-point calibration using two standard
buffer solutions with pH values pHS1 and pHS2 (usually pH 4 and 7 or sometimes 9) is used. This procedure determines the
experimental g value and assumes the liquid junction potentials of the two standards and the test solutions are equal.
If the respective potentials of the two buffers are ES1 and ES2, and EX is the potential of the unknown sample, the pH value of
the unknown (pHX) is

EX  ES1
pHX ¼ pHS1  [5]
g
with

ES1  ES2
g¼ [6]
pHS2  pHS1
The IUPAC-recommended buffers for the standardization of potentiometric systems are gathered in Table 1.

Acidic Dissociation Constant Definitions

The acid–base equilibriums of any acidic or basic solute in aqueous solution can be expressed as follows:

HA z + H2 O⇄A z1 + H3 O + [7]

where z is the charge of the acidic form of the solute, HA. In this instance, the thermodynamic acidity constant relates activities and
concentrations of the different species present in the solvent according to
 
aH + aAz1 aH + A z1 gAz1
Ka ¼ ¼ [8]
aHAz ½A z gHAz

or
 
A z1 gAz1
pKa ¼ pH  log [9]
½HA z gHAz

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis 3

where gAz1 and gHAz stand for the activity coefficients of the species indicated in the subscripts, which are usually estimated through
the well-known Debye–Hückel equation:

z2 AI1=2
 log ss g ¼ [10]
1 + a0 BI1=2
where

1:8246  106
A¼
3=2 [11]
s eT

being T the absolute temperature and se the static dielectric constant. The Bates–Guggenheim convention (a0 ¼ 4.56 Å) is
commonly adopted, and then, at 25  C, a0B ¼ 1.5 and A ¼ 0.5.
Equation [8] defines the thermodynamic acidity constant, that is, the one in which all the involved species are expressed by their
activity. Nevertheless, it is very common to use mixed constants, Ka0 , in which the proton contribution is given by the activity, but
other species contribute by their concentration. This is because current determination methods use the glass electrode to measure
the pH, and then, the measured quantity is the proton activity, and spectrometric, potentiometric, electrophoretic, and other kind
of techniques provide the ratio between the concentrations of the basic and acidic species. In addition, it is also useful to define the
concentration equilibrium constant, Ka00 , to relate the concentrations of all the species in equilibrium. Thus, the three types of
equilibrium constants are related as follows:
aH + aAz1 0 g z1 00 gH + gA z1
Ka ¼ ¼ Ka A ¼ Ka [12]
aHAz gHAz gHAz

Potentiometric Titrations for pKa Determination

Potentiometric Methods
Potentiometric methods are based on the measurement of the activity of hydrogen ions in an acid–base system of known ratio
between basic and acidic forms of the compound.3 Usually, the pH is measured through a well-calibrated glass combination
electrode, over a concentration range of 0.005–0.05 M and in 2–12 aqueous pH range. To calibrate the potentiometric system, the
already described two-point procedure is commonly used.
The pKa is conveniently determined by a titration of the basic or acidic compound with a standardized strong acid or base. The
pH is measured at various stages of neutralization along the titration, and the concentration ratio of basic and acidic forms is
calculated from the amounts of the titrated compound and the added titrant. The plot of pH against the volume of the added titrant
is called potentiometric titration curve. To illustrate it, the titration curve for methylparaben in aqueous solution is given in Figure 1
(a), which shows the resistance to change the pH in its first part. This is because the titrant additions neutralize the acid converting
it in its conjugate base and a buffered region is obtained. When both species show the equal activity, the pH of the solution equates
to the pKa of the substance (eqn [9]). Then, from the buffered region of the curve, and in particular in the half-neutralization point,
a reliable pKa value can be obtained. The curve inflection point, called equivalence point, points out complete neutralization of the
titrated compound, and it allows the determination of its concentration.
Titration curves for two diprotic acids are given in Figure 1. Thus, Figure 1(b) shows two separated buffered regions, and, from
them, values for pKa1 and pKa2 can be easily obtained. On the contrary, in Figure 1(c), the two buffered regions are overlapped
because the ionization of the second group starts before the end of the neutralization of the first acidic group. Then, all species

00
Figure 1 Titration curves in aqueous medium and I ¼ 0.15 M and 25  C of (a) methylparaben, pKa ¼ 8:15 at I ¼ 0.15 M and 25  C (ct ¼ 0.062 M); (b)
00 00 00 00
piperazine dihydrochloride, pKa ¼ 5:66 and pKa ¼ 9:80 at 25  C (ct ¼ 5  103 M); (c) epicatechin, pKa ¼ 8:73 and pKa ¼ 9:41 at 25  C
(ct ¼ 2  103 M).

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
4 Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis

involved in the two ionization steps must be taken into account simultaneously for pKa value calculations. In many cases, before
starting titration, the analyte is converted in its fully protonated species by adding standard HCl solution, and, then, the excess of
HCl added and the protonated analyte are titrated. Similarly, when the compound is poorly soluble in acidic medium, a back
titration of the compound can be carried out starting at pH 11 and using HCl as titrant until precipitation. The potentiometric
titrations can be easily automatized, and two or even more pKa values can be obtained for polyprotic compounds. The accuracy of
results depends on the calibration system and also on the titration speed. For accurate pKa values (0.02 pKa units), a sufficient
equilibration time after each titrant addition has to be allowed.3b Common potentiometers are equipped with ordinary combi-
nation glass electrodes, and their standardization through ordinary aqueous buffers is recommended.

pKa Calculation from Potentiometric Data

The usual method to calculate pKa values is based on the mass balances of the species in equilibrium and the electroneutrality of the
solution.3a Each one of the consecutive acid–base equilibriums of a fully protonated species, HnAz, can be expressed as
aHni Azi aH +
Hni + 1 A zi + 1 ⇄Hni A zi + H + Kai ¼ [13]
aHni + 1 Azi + 1

where n is the total number of ionizable groups, z the maximum positive charge (in the fully protonated species), and Kai the
thermodynamic dissociation constant. If such compound is titrated with a standardized strong base (BOH), the mass balance and
electroneutrality of the solution require that

X
i¼n   Vo Mt
ct ¼ Hni A zi ¼ [14]
i¼0
Vo + V

X
i¼n  
½B +  + ½H +   ½OH  + ðz  iÞ Hni A zi ¼ 0 [15]
i¼0

with

VMBOH
cBOH ¼ ½B +  ¼ [16]
Vo + V
where ct and cBOH correspond to the analytical concentration of the analyte and basic titrant, respectively, in any particular point of
the titration. Vo is the initial volume, V is the volume of titrant added, and Mt and MBOH are the molarity of the analyte and
standardized base, respectively.
In the case of monoprotic compounds or when the pKa values of polyprotic substances are conveniently separated (more than
three units), a simple expression for each pKa can be obtained:
 
Hni + 1 A zi + 1 g zi + 1
pKai ¼ pH + log   + log Hni + 1 A [17]
Hni A zi gHni Azi
   
where Hni + 1 A zi + 1 and Hni A zi can be determined from the analytical concentrations of compound and titrant added.
According to eqns [14]–[16], eqn [17] can be rewritten as follows:

ct  cBOH  ½H +  + ½OH  gH A zi + 1

pKai ¼ pH + log + log ni + 1 [18]
cBOH + ½H +   ½OH  gHni Azi

If the pKai value is between 5 and 9, both [H+] and [OH] become usually negligible, but if pKai is lower than 5, [H+] must be
included in the calculations, and an iterative process must be used. Similarly, when pKai is higher than 9, [OH] must be kept in the
calculations. When the consecutive pKa values differ in less than three pKa units, the resolution of the equilibrium system becomes
more difficult because all the involved species must be taken into account. Fortunately, different computer programs are available
to compute reliable pKa values.
Another method to calculate pKa values is based on the formation function method, that is, the Bjerrum function ( n), which
stands for the average number of bound protons per unit of analyte concentration,4 and it is expressed by

Total bound proton cH +  ½H + 

n ¼ ¼ [19]
Total analyte ct
where cH + and ct are the total concentrations of the proton and analyte, respectively. The difference between the total and free
hydrogen concentration is equal to the concentration of the bound hydrogen ions, and it can be obtained by measuring the pH of
the solution and computing the total hydrogen ion concentration. Thus, cH + can be obtained by the following expression:

cH + ¼ ct + cHCl  cBOH + ½OH  [20]

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis 5

00
Figure 2 Bjerrum function plot of phenylethylamine in aqueous medium: pKa ¼ 9:42 at I ¼ 0.15 M and T ¼ 25  C (ct ¼ 0.025 M). Inflection point
indicates the pKa value.

where cHCl is the concentration of strong acid (typically HCl) added before the titration to ensure the analyte is completely in its
protonated species.
Therefore, an experimental n value can be calculated at each data point of the potentiometric titration. However, for a
compound in which the fully protonated species is HnAz, cH + should be expressed as

X
i¼n  
cH + ¼ ½ H +  + i Hi A zn + i [21]
i¼1

The substitution of eqn [21] and eqn [14] in eqn [19] yields

X
i¼n  
i Hi A zn + i
i¼1
n ¼ [22]
Xi¼n  
zi
Hni A
i¼0

By combining eqn [13] and eqn [22], the following expression is obtained:

X
j¼n
0
pKaj  ipH
X
i¼n
i10 j¼ni + 1
i¼1
n ¼ [23]
X
j¼n
0
pKaj  ipH
X
i¼n
1+ 10 j¼ni + 1
i¼1

Equation [23] shows that n is a nonlinear function of the solution acidity. Then, on the basis of the experimentally determined n
at each point of the titration (eqns [19] and [20]), the acidity constants included in eqn [23] can be determined by a nonlinear
curve-fitting analysis. Figure 2 shows an example of Bjerrum function obtained in aqueous solution for phenylethylamine.
It should be noticed that the inflection point corresponds to the pKa value.

Spectrophotometric Measurements for pKa Determination

As explained, potentiometry is useful for aqueous pKa determination if the compound is soluble enough in water (at least 105 M)
in an adequate range of pH (i.e., 2–12). pKa determinations in much diluted solutions (105–106 M) can be performed by
spectrophotometry if the compound shows a chromophore near to the ionizable group.3a,c The method is based on the different
spectra of molecular and ionized species. Intermediate spectra between these ones will be obtained in solutions with different ratios
of both species, as shown in Figure 3. This ratio can be easily calculated because it depends solely on the pH at which the solution is
measured, and it is determined by measuring the absorbance of solutions at pH values where the two forms coexist. Traditionally,
the absorbance was measured at one specific wavelength. Nowadays, many spectrophotometric methods are based on
multiwavelength data.

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
6 Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis

Figure 3 Variation of UV–visible spectrum in aqueous solution of bromocresol green with pH.

Single-Wavelength Spectrophotometric Approach

At a specific wavelength (named analytical wavelength) and assuming that Beer’s law is fulfilled for the protonated and deproto-
nated species of the acid, the observed absorbance of a solution that contains both species will be the sum of the absorbance of both
absorbing species.
Considering each one of the consecutive equilibriums in eqn [13],

HA⇄A + H + [24]

where the charges are omitted for simplicity, the pKa0 is expressed by

0 ½A 
pKa ¼ pH  log [25]
½HA

At pH near to pKa, the total concentration, ct, and the absorbance at the measured wavelength, l, can be expressed as follows:

ct ¼ ½HA  + ½A  [26]

A ¼ elHA b½HA + elA b½A  [27]

where elHA elA

and are the molar absorptivities at the measurement wavelength of the [HA] and [A], respectively, and b is the path
length. Combining eqns [26] and [27], the following expressions are obtained:

A  elA bct
½HA ¼ [28]
elHA b  elA b

elHA bct  A
½A  ¼ [29]
elHA b  elA b

Substituting eqns [28] and [29] in eqn [25],

0 elHAz bct  A
pKa ¼ pH  log [30]
A  elA bct

The terms elHAz bct and elAbct correspond to the absorbance of the protonated and deprotonated species that can be easily obtained
by measuring the absorbance of the solution in acidic (pH < pKa  2) and basic (pH > pKa + 2) media. Then, eqn [30] can be
rewritten as

0 Aa  A
pKa ¼ pH  log [31]
A  Ab
where Aa and Ab are the absorbance in acidic and basic media, respectively.

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis 7

Therefore, a good estimation of pKa0 value can be easily obtained by measuring potentiometrically the pH and spectro-
photometrically the absorbance at the analytical wavelength of three solutions at the same ionic strength: acidic (fully protonated
species) and basic (fully deprotonated species) solutions and one more solution with intermediate pH. The analytical wavelength,
l, is that at which one species absorbs strongly and the other one has no absorption at all or, at least, the one at which the difference
between the absorbance of the neutral and the absorbance of the ionic species is higher.
For accurate determination, not only one but six to eight solutions of the sample in identical concentration but with different
pH values near to pKa must be prepared and measured.3b A good alternative is the spectroscopic titration based on the titration of
the sample solution with a standardized strong acid or base. At various stages of neutralization, the pH is measured and absorbance
at l is acquired. This method can be easily automatized. Thus, eqn [31] can be rewritten as

0
elHA + elA 10pHpKa
A ¼ bct 0 [32]
1 + 10pHpKa

and when A is plotted versus pH, a sigmoidal curve is obtained. The inflection point corresponds to pKa0 value, which can be also
computed by the first derivative approach or by fitting the experimental data to eqn [32]. In the same way, a more complex
expression can be deduced for polyprotic compounds:

X
i
0
ipH  pK aj
X
n
elHn Az + elHni Azi 10 j¼1

i¼1
A ¼ bct [33]
X
i
0
ipH  pK aj
X
n
j¼1
1+ 10
i¼1

where HnAz stands for the fully protonated species, elH Azi indicates the molar absorptivity of the corresponding species, and Kaj0 is
ni
the mixed acidity constant. For compounds with two ionizable groups sufficiently separated, two inflection points are obtained
when absorbance is plotted versus pH, and the two pKa0 values can be easily determined. When a substance has two ionizing groups
with close pKa values (inside a range of three pKa unities), the two pKa0 values and also the molar absorptivity of the intermediate
species must be simultaneously determined.3b The errors in the pKa0 determination can be due to a decomposition of the compound
or a failure of one or both species to follow Beer’s law. The first can be detected by comparing the measurements obtained in
different conditions of the stock solution (pH, time, and temperature).
Provided that the other precautions commonly employed are observed, and that the spectral characteristics of the compound
are favorable, results of high accuracy can be expected if a constant temperature is maintained for both the pH and the absorbance
measurements.

Multiwavelength Spectrophotometric Approach

Measurements by single-wavelength spectrophotometric method are nontrivial if the acid–base equilibriums comprise more than
two ionization steps or if the reacting components are not stable within two pH units of the pKa values. Furthermore, calculations
require the knowledge of the molar absorptivities of the protonated and deprotonated species, and it may be difficult to apply for
polyprotic substances with close pKa values. Alternatively, computer programs from multiwavelength spectrophotometric data
have been proposed and successfully used. Most of these methods involve a least-squares approach whereby the differences
between the theoretical and experimental absorbance are minimized by means of the Gauss–Newton–Marquardt algorithm.5 Then,
the pKa and the molar absorptivity of individual reacting species are treated as adjustable parameters. Computer programs for
multiwavelength spectrophotometric data also allow scrutinizing ionizable compounds with overlapping pKa values and/or similar
absorption spectra. Chemometric approaches based on target factor analysis (TFA) are also applied to determine pKa from
spectroscopic titrations. Specifically, the absorbance data matrix is decomposed into a linear combination of principal components
using Principal Component Analysis (PCA) method. Based on a suggested reaction model, TFA treatment can be employed to
transform the mathematical solution with the components being identified into one with physical meaning. However, although
the molar absorptivities are usually not required for analysis, explicit equations for the equilibrium expressions are necessary to
rotate the eigenvectors to give the correct concentration profile.5b,6 The merits of the multiwavelength-TFA approach over the
traditional ones are that it enables the simultaneous determination of the pKa values, the number of independent light-absorbing
species present in the chemical system, and their absorption spectra from a single titration, without prior knowledge of the optical
properties of the components. This could be useful for structure identification purposes and for assignment of pKas to functional
groups. The TFA method is able to resolve the spectral data of the multiprotic ionization system even if the absorption spectra of
individual species are very similar.7 However, it requires some spectral dissimilarity for each ionization species, thus giving rise to a
pH-dependent spectrum caused by the ionization of the molecule.6a

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
8 Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis

pKa Determination by Capillary Electrophoresis

Capillary Zone Electrophoresis of Acid–Base Analytes
The determination of pKa by capillary zone electrophoresis (CZE or simply CE) is based on the measurement of both the effective
mobility of an acid–base compound and the pH in a series of buffers at constant ionic strength. The pKa is obtained by fitting the
measured effective mobility as a function of pH to an appropriate function, which depends on the number of ionizable groups.8
The advantages of CE methods include that there is the consumption of low sample amounts, no purity of sample compounds is
required, and a variety of detection techniques can be successfully used.8a–e,9
Effective mobility (meff) of a compound in a given pH buffer can be easily determined from the applied voltage (V), the effective
capillary length to the detector (LD), the total capillary length (LT), the migration time of the analyte – or time that takes the charged
analyte to reach the detection point (tm) – and the electroosmotic flow time (tEOF) measured by the migration time of a neutral
marker compound, according to eqn [34]:
 
LD LT 1 1
meff ¼  [34]
V tm tEOF

In normal-mode CZE, cations move to the negative electrode faster than the EOF, and thus, they have positive electrophoretic
mobilities. Anions are attracted by the positive pole but pulled to the negative pole by the EOF (which is usually higher), and thus,
they have negative electrophoretic mobilities.
For an acid–base compound, partially charged, the observed effective mobility is proportional to the ionization degree of the
compound. Equation [35] is a general expression that relates meff of a compound to its pKa and the pH of the buffer
solution:8a,e,g,h,9e

X
i
0
ipH  pKaj
X
n
j¼1
mHn Az + mHni Azi 10
i¼1
meff ¼ [35]
X
i
0

X
n ipH  pKaj
j¼1
1+ 10
i¼1

where mHn Az indicates the mobility of the fully protonated species, mHni Azi the mobilities of the successive deprotonated species,
and Kaj0 the mixed acidity constants, related to the thermodynamic ones by the activity coefficients (g) (eqn [12]). Since the
mobilities and the pKa0 depend on the ionic strength of the solution, all buffers are prepared at the same ionic strength, and ionic
activity coefficient correction is done for each pKai value. Thus, the different mixed acid–base pKaj0 values of the studied compound,
together with the mobilities of its pure forms, can be calculated by fitting the experimental mobilities to the pH of the buffers
through eqn [35]. A scheme of the procedure for a diprotic acid is presented in Figure 4.10 The two pKaj0 values agree with the
inflection points of the mobility versus pH plot.

Buffer and Instrumental Considerations for pKa Determination by CZE

The running pH buffers, also known as background electrolytes, play a crucial role in the pKa determination by CE. Errors in pH
measurement, unsuitable buffered pH range, and interactions of the buffers with the compound or, even the capillary, may result in
a significant error in the determined pKa value. The pH must be adequately measured in well-buffered solutions at constant ionic
strength and temperature to avoid errors in pKa determination. pKa and CE performance are very dependent on temperature.11
Numerous factors influence temperature inside the capillary, such as the applied voltage, the buffer conductivity, the capillary
dimensions (diameter and total length), and the thermoregulation system (controlled by an air- or liquid-cooling system).8a
Applied voltage should be high enough to obtain an acceptable running time but low enough to remain in the linear region of
Ohm’s law and prevent excessive Joule heating.12 The concentration of the buffer is also critical since the running time is directly
proportional to its square root. Thus, dilute buffers decrease running time and also Joule heating, but they present low buffer
capacity.8c For common capillaries, about 50 cm long and 50 mm in diameter, 0.05 M buffers and 15–20 kV applied voltages may
be adequate compromises.8e,g For shorter capillaries, the voltage should be reduced.8h
The buffer pH should be measured just before the mobility measurements. Running buffers with pH higher than 10 may be
rapidly altered by CO2 absorption from the air and/or electrolysis caused by the voltage applied.8b,e,g Therefore, they should be
prepared immediately before use and frequently replaced in the CE instrument. At least, measurements performed in 5–6 different
pH buffers should be taken for each pKa value to be determined. Most buffers should have pH values close to the compound pKa
(ideally in the pKa  1 range), but it is also convenient that at least one buffer presents a pH value much higher and much lower
than the pKa value for a proper estimation of the limiting mobilities of the pure ionic forms of the compound, which will lead to a
higher precision in pKa calculation. Buffers with pH pKa  1 for a monoprotic neutral acid or pH
pKa +1 for a monoprotic
neutral base should be discarded because the effective mobility of the neutral forms is not calculated by the models since it is zero.
For frequent routine pKa determination, it is convenient to dispose of a set of pH buffers differing in 0.5–1.0 pH units.

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis 9

Figure 4 A schematic workflow of pKa determination by CE-UV and CE-MS. Adapted with permission from Wan, H.; Thompson, R. A.; Drug Discov.
Today Technol. 2005, 2, 171–178. Copyright 2005 Elsevier.

Table 2 Recommended buffers according to Refs. [8e] and [8h]

Acid–base constituents pKa pH range Buffer preparation

H3PO4–H2PO4 2.12 2.0–3.0 NaH2PO4 + HCl

HCOOH–HCOO 3.75 3.0–4.5 NaHCOO + HCl
CH3COOH–CH3COO 4.76 3.7–5.8 NaCH3COO + HCl
BisTrisH+–BisTris 6.48 5.5–7.5 BisTrisHCl + NaOH
TrisH+–Tris 8.08 7.0–9.0 TrisHCl + NaOH
CHES–CHES 9.50 8.4–10.1 NaCHES + HCl
CAPS–CAPS 10.40 9.4–11.6 NaCAPS + HCl

BisTris, bis(2-hydroxyethyl)-amino-tris(hydroxymethyl)-methane; BisTrisHCl, hydrochloride of BisTris; Tris,

tris(hydroxymethyl)aminomethane; TrisHCl, hydrochloride of Tris; CHES, 2-(cyclohexylamino)ethanesulfonic acid;
NaCHES, sodium 2-(cyclohexylamino)ethanesulfonate; CAPS, 3-(cyclohexylamino)-1-propanesulfonic acid; NaCAPS,
sodium 3-(cyclohexylamino)-1-propanesulfonate.

The nature of the buffer should be carefully chosen to avoid buffer interactions, which may lead to wrong pKa determination.
Thus, some buffers must be avoided because they produce errors in pKa determination of many types of compounds.8e,13 However,
phosphoric/hydrogen phosphate and formate buffers are adequate buffers for pKa determination of quite strong acids, such as
amino acids.8h Taking into account the published constraints, Table 2 lists a set of recommended buffers for the whole pH range,
which includes the buffer components to be used in buffer preparation. Notice that to keep the ionic strength of the buffer constant,
the recommended procedure8e is to prepare a solution of the monocharged ionic component of the buffer (usually a sodium or
chloride salt) at a fixed concentration, higher than the desired ionic strength (e.g., 0.1–0.2 M for a final 0.05 M ionic strength), add
a small volume of a more concentrated strong acid or base (e.g., 0.5 M HCl or 0.5 M NaOH) until the desired pH is achieved, and
then dilute to the desired ionic strength.
Buffers recommended in Table 2 were proposed for UV detection, which is the preferred detection method for CZE because it is
simple and sensitive enough for most compounds. They may also be used for amperometric and contactless conductivity detection,
which have been also developed for the detection of non-UV-absorbing compounds in CZE.8a,h However, the proposed buffers
may not be volatile enough for mass spectrometry detection, which is an interesting alternative detection method because of its

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
 Author's personal copy
10 Methods for pKa Determination (I): Potentiometry, Spectrophotometry, and Capillary Electrophoresis

sensitivity, universality, and capability for simultaneous detection of several compounds, which increases the throughput of
pKa determination. The big challenge of using this detection method is to find volatile and stable enough buffers to cover the
whole pH range.8a

References
1. Sørensen, S. P. L.; Linderstrøm-Lang, K. Compt. Rend. Trav. Lab. Carlsberg 1924, 15(6), 1.
2. (a) Inczédy, J.; Lengyel, T.; Ure, A. M.; Gelencsér, A.; Hulanicki, A. IUPAC Compendium of Analytical Nomenclature, Definitive Rules 1997, 3rd ed; Oxford: Blackwell, 1998; http://
www.iupac.org/publications/analytical_compendium/; (b) Bates, R. G. Determination of pH: Theory and Practice, 2nd ed.; New York: Wiley, 1973; (c) Bates, R. G. Crit. Rev. Anal.
Chem. 1981, 10(3), 247–278.
3. (a) Serjeant, E. P. Potentiometry and Potentiometric Titrations; Wiley: New York, 1984; (b) Albert, A.; Serjeant, E. P. The Determination of Ionization Constants: A Laboratory
Manual; Chapman and Hall: London, 1971; (c) Cookson, R. F. Chem. Rev. 1974, 74(1), 5–28.
4. (a) Šůcha, L.; Kotrlý, S. Solution Equilibria in Analytical Chemistry; Van Nostrand Reinhold: London, 1972; (b) Hartley, F. R.; Burgess, C.; Alcock, R. Solution Equilibria; Ellis
Horwood Ltd.: Chichester, 1980.
5. (a) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43(10), 1739–1753; (b) Tam, K. Y.; Takács-Novák, K. Pharm. Res. 1999, 16(3), 374–381; (c) Tam, K. Y.; Hadley, M.;
Patterson, W. Talanta 1999, 49(3), 539–546.
6. (a) Allen, R. I.; Box, K. J.; Comer, J. E.; Peake, C.; Tam, K. Y. J. Pharm. Biomed. Anal. 1998, 17(4–5), 699–712; (b) Tam, K. Y. Anal. Lett. 2000, 33(1), 145–161.
7. Mitchell, R. C.; Salter, C. J.; Tam, K. Y. J. Pharm. Biomed. Anal. 1999, 20(1–2), 289–295.
8. (a) Henchoz, Y.; Bard, B.; Guillarme, D.; Carrupt, P.-A.; Veuthey, J.-L.; Martel, S. Anal. Bioanal. Chem. 2009, 394(3), 707–729; (b) Poole, S. K.; Patel, S.; Dehring, K.;
Workman, H.; Poole, C. F. J. Chromatogr. A 2004, 1037(1–2), 445–454; (c) Cleveland, J. A., Jr.; Benko, M. H.; Gluck, S. J.; Walbroehl, Y. M. J. Chromatogr. A 1993, 652(2),
301–308; (d) Babić, S.; Horvat, A. J.M; Mutavdžić Pavlović, D.; Kaštelan-Macan, M. Trends Anal Chem. 2007, 26(11), 1043–1061; (e) Fuguet, E.; Reta, M.; Gibert, C.; Rosés, M.;
Bosch, E.; Ràfols, C. Electrophoresis 2008, 29(13), 2841–2851; (f ) Šlampová, A.; Křivánková, L.; Gebauer, P.; Boček, P. J. Chromatogr. A 2009, 1216(17), 3637–3641;
(g) Herrero-Martı́nez, J. M.; Sanmartin, M.; Rosés, M.; Bosch, E.; Ràfols, C. Electrophoresis 2005, 26(10), 1886–1895; (h) Henchoz, Y.; Schappler, J.; Geiser, L.; Prat, J.;
Carrupt, P.-A.; Veuthey, J.-L. Anal. Bioanal. Chem. 2007, 389(6), 1869–1878; (i) Janos, P. J. Chromatogr. A 2004, 1037(1–2), 15–28.
9. (a) Pang, H.-m.; Kenseth, J.; Coldiron, S. DDT 2004, 9(24), 1072–1080; (b) Gong, X.; Figus, M.; Plewa, J.; Levorse, D. A.; Zhou, L.; Welch, C. J. Chromatographia 2008,
68(3/4), 219–225; (c) Avdeef, A.; Testa, B. Cell. Mol. Life Sci. 2002, 59(10), 1681–1689; (d) Fuguet, E.; Ràfols, C.; Bosch, E.; Rosés, M. J. Chromatogr. A 2009, 1216(17),
3646–3651; (e) Ishihama, Y.; Nakamura, M.; Miwa, T.; Kajima, T.; Asakawa, N. J. J. Pharm. Sci. 2002, 91(4), 933–942; (f ) Fuguet, E.; Ràfols, C.; Bosch, E.; Rosés, M. Chem.
Biodiversity 2009, 6(11), 1822–1827.
10. Wan, H.; Thompson, R. A. Drug Discov. Today Technol. 2005, 2(2), 171–178.
11. Reijenga, J. C.; Gagliardi, L. G.; Kenndler, E. J. Chromatogr. A 2007, 1155(2), 142–145.
12. Baker, D. R. Capillary Electrophoresis; Wiley: New York, 1995; p. 244.
13. Fuguet, E.; Ràfols, C.; Bosch, E.; Rosés, M. Buffered Background Electrolytes for Capillary Electrophoresis. Reference Module in Chemistry, Molecular Sciences and Chemical
Engineering, Elsevier: Amsterdam, 2015.

Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, (2015)
View publication stats