Full text of DOI:10.1093/chromsci/41.6.323

Journal of Chromatographic Science, Vol. 41, July 2003
Determination of pKa Values by Liquid Chromatography
Markus Manderscheid and Thomas Eichinger*
Aventis Pharma Deutschland GmbH, Drug Innovation & Approval, Frankfurt am Main, Germany
This method is applicable if the drug is acceptably soluble and
stays in solution during the titration procedure. If there are dif-
Abstract
ferences between the protonized and the nonprotonized form,
the procedure can only be used in a limited way. Furthermore, a
relatively high amount of test material is needed.
In this paper, we investigate the potential of a high-performance
liquid chromatography technique to determine pKa values of drug
candidates that show poor solubility in water. The determination of
pKa values by this method is in principle not new, but it exhibits
simplicity, requires lower quantities of drugs and solvents, and
minimal analysis time. The method is an alternative to existing
methodology, in which this determination is not readily feasible.
To compensate for the disadvantages of titration, and consid-
ering the very small amounts of available test materials, an alter-
native method needs to be developed.
In principle, the use of liquid chromatography (LC) to deter-
mine pKa values is not new (6–8). However, these methods do
not fulfill the requirements of a high-throughput lab, in which
short retention times, employment of microgram quantities of
test materials, the use of simple and reasonably priced equip-
ment, and a variety of appropriate chromatographic columns are
required.
To evaluate our method, three well-known substances were
chosen to demonstrate its appropriateness: the organic acids (a)
piretanide, (b) furosemide, and (c) glyburide (Figure 2).
The LC method is based on the different retention behavior of
the protonized and the nonprotonized form of the test material.
The retention time is determined in relationship to the pH-value
of the mobile phase by reversed-phase HPLC. The pKa value is
the point of inflection in the resulting sigmoidal curve, which
can be easily achieved.
Introduction
The pKa value is a main item in the biophysical characteriza-
tion of a drug and may be helpful in predicting the behavior of a
drug under in vivo conditions. Because a correlation exists
between the pKa value and the solubility of the drug in different
media, it is possible to make predictions in the behavior of
absorption in the organism and, as a result of this, the closely
linked bioavailability.
Current properties of new drugs include poor solubility char-
acteristics, so the need to generate a high-performance liquid
chromatography (HPLC)-based method to predict the pKa values
of such drugs was recognized. The described method is attractive
because of its simplicity and ability to use a variety of isocratic
HPLC systems, resulting in the requirement for minimal
amounts of drug substance. In addition, because of the high
quality and acceptable throughput and further information
obtained for the test material, the method is attractive for the
pharmaceutical industry.
The test materials, with the exception of glyburide, are readily
soluble in water and show a moderate acidic behavior (Table I).
To measure pKa values of acids and bases in the conventional
way, titration is applied to get the pKa value at the half-neutral-
ization point, which is ideally the point of inflection in the titra-
tion curve (Figure 1) (1–5). For acids, the following relationship
exists:
OH– + XH → X– + H2O
The following exists for bases:
X + H+ → HX+
Eq. 1
Titrant (mL)
Eq. 2
Figure 1. Determination of the point of inflection in the titration curve of an
acid (schematic).
* Author to whom correspondence should be addressed: email Thomas.Eichinger@aventis.com.
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
323

Journal of Chromatographic Science, Vol. 41, July 2003
chemicals and solvents were of ACS-reagent grade. Furosemide
(lot no. B030), piretanide (lot no. E041), and glyburide (lot no.
B008) were obtained from Aventis Pharma Deutschland GmbH
(Frankfurt am Main, Germany).
The HPLC mobile phases were prepared from deionized water,
acetonitrile, sodium chloride (1 g/L), phosphoric acid (85%, 2
mL/L), and sodium hydroxide (10N, to adjust the pH values,
starting at pH 7.0).
The relation of organic solvent to water was chosen in a way
that the substance peak did not overlap with the peak of injection
and the run would result in acceptable retention times in the
acidic pH range, in which the substance showed strong interac-
tion with the column because of hampered dissociation of the
corresponding acid.
The disadvantage of the use of an organic modifier and
the subsequent “incorrect” pKa value can be overcome by a
simple experiment.
Experimental
Instrumentation
Measurements were carried out with an isocratic HPLC system
(HP 1090, Agilent, Waldronn, Germany) with a standard UV
detector (Gynkotek SP-4, Germering, Germany). The detection
wavelengths for piretanide, furosemide, and glyburide were 226,
232, and 227 nm, corresponding to their respective UV maxima.
A Superspher 100 RP18 endcapped column (75-mm × 4-mm ×
5-µm particle size) was used for the analysis of furosemide, a
Lichrospher 60 RP select B column for piretanide, and a
Lichrospher 100 RP8 column for glyburide (all Merck,
Darmstadt, Germany).
An SP 4270 integrator (Spectra Physics, Egelsbach, Germany)
was used to obtain the chromatogram and data calculations.
To titrate the pH values of the acetonitrile buffers to 0.01 pH
units accuracy, a MetrOhm 605-pH-Meter (Herisau, Switzer-
land) was used.
Three elution phases differing in their amounts of organic sol-
vent were chosen. Extrapolation to the 100% water value was per-
formed by linear regression analysis. The generated results for the
three model compounds were close to the theoretical values and
substantiate the appropriateness of the applied methodology.
Chemicals and reagents
Deionized water was produced using a Milli-Q185 Plus water
system (Millipore, Milford, MA) for all aqueous solutions. All
A
B
–0.13
–0.20
–0.27
– 0.50
–0.47
– 1.00
–0.95
– 1.50
–1.85
–1.60
–2.15
– 2.00
– 2.50
– 3.00
– 3.50
–2.75
–3.07
–2.77
–3.25
3.00
4.00
5.00
6.00
7.00
pH
C
6.22
5.85
5.42
4.87
7.00
6.00
5.00
4.00
3.00
2.00
1.00
3.95
3.30
2.47
1.83
1.45
1.31
1.23
1.17
1.13
Figure 2. Chemical structures of the tested drugs: (A) piretanide, (B)
furosemide, and (C) glyburide.
3.00
4.00
5.00
6.00
7.00
pH
Table I. Solubility Data of the Tested Drugs*
Regression Analysis
Substance
Solvent
Solubility (mg/100 mL)
5
4
3
2
1
0
Furosemide
Water
Buffer pH 7.5
8.2
2400
3.80
3.9
15
4.1
20
4.0
Piretanide
Glyburide
Water
Buffer pH 7.0
5.3
2800
0
5
10
25
30
Water
Buffer pH 7.5
0.16
2.4
Acetonitrile (%)
Figure 3. Example for the analysis of the obtained data: furosemide with 20%
acetonitrile as organic modifier.
* The values were taken from the List of Pharmaceutical Substances, 12th ed. Eschborn,
Germany, 2000.
324

Journal of Chromatographic Science, Vol. 41, July 2003
Procedure (sample preparation)
sample solution until two retention time values were equal. In
this way, we established the relationship between the retention
time of each substance and the pH value (Figure 3).
The used pH values ranged from 2.0 to 7.0; measuring started
at pH 7.0 in steps of 0.3 units.
One milligram or less of each test material was dissolved in
acetonitrile to obtain a solution having a concentration of 0.1
µg/10 µL. The column was equilibrated by rinsing with the
mobile phase for 5 min, at 3.0 mL/min flow.
We have used a flow rate of 1.0 mL/min and injected 10 µL of
Table III. Comparison of pKa Values*
Results and Discussion
Furosemide
Piretanide
Glyburide
In the conventional way, pKa values of drugs are detected is by
titration (2–4). However, when the drug is only poorly soluble or
does not stay in the solvents during the titration procedure, this
procedure can be used only in a limited way. In this study, we
used the LiChroCART 25-4 column cartridge system, which is
available for the resins Superspher 100 RP18 endcapped,
Literature
3.9
3.8
3.9
4.0
5.3
5.5
Found after
regression
* The literature values are taken from reference 9.
Table II. Summary of the Collected Data for Furosemide, Piretanide, and Glyburide*
Concentration of organic modifier (acetonitrile)
Furosemide (DtR/DpH)
Piretanide (DtR/DpH)
Glyburide (DtR/DpH)
pH value
15%
20%
25%
15%
20%
25%
30%
35%
0.00
40%
0.00
7.0
6.7
6.6
6.5
6.4
6.3
6.1
6.0
5.8
5.7
5.5
5.4
5.3
5.2
5.1
5.0
4.9
4.8
4.7
4.6
4.5
4.4
4.3
4.2
4.1
4.0
3.9
3.8
3.7
3.6
3.5
3.4
3.3
3.2
3.0
0.00
0.00
–0.98
0.00
0.00
0.00
–0.07
–0.17
0.00
–2.23
–1.10
–0.14
–1.44
–0.24
–5.32
–0.03
-0.20
–0.23
–0.53
–0.87
–0.13
–0.20
–0.27
–0.47
–3.20
–3.47
–1.30
–1.17
–0.87
0.79
–0.28
–0.70
–0.02
–0.13
–3.16
–8.22
–0.56
–1.52
–8.32
–2.67
–2.25
–6.98
–6.40
–0.21
–0.44
–1.75
–1.60
–1.47
–2.73
3.90
–0.47
–1.93
–3.50
–5.30
–0.95
–19.74
–4.14
–4.90
–0.87
–0.30
–0.73
–0.53
–1.60
–2.77
–0.82
–1.10
–32.63
–39.27
–37.40
0.00
–9.30
–0.23
–5.00
–13.87
–6.87
–1.25
–7.87
–3.25
–5.40
–14.30
–13.60
–12.00
–11.55
–3.07
–2.75
–2.15
–1.85
–1.15
–1.00
–7.70
–8.15
–7.60
–8.45
–6.50
–0.90
–0.75
* Obtained with different concentrations of organic modifier in dependence of the pH value.
325

Journal of Chromatographic Science, Vol. 41, July 2003
Lichrospher 60 RP select B, and Lichrospher 100 RP8 (Merck). As
mobile phases, water–acetonitrile mixtures with different pH
values were applied with a flow rate of 1.0 mL/min and detection
at appropriate wavelengths.
drug and solvent, and low time requirement. As a conclusion, one
could suggest it as a convenient alternative to the conventional
methods.
We calculated the retention time divided by the pH value and
plotted the resulting ratio in a coordinate system with the pH
value on the abscissa. The resulting curve had a U-shaped form,
in which one can estimate the pKa value at the minimum of the
curve. We estimated the pKa value for the tested drugs by regres-
sion analysis of the pKa values found at the different acetonitrile
concentrations (Table II), which corresponded to values reported
in literature and thus appeared readily acceptable (Table III)
(9–11). Using the three selected drugs having similar chemical
structures, we could show that one can easily measure the pKa
values despite different solubilities. Other advantages of this
method are the short retention time, which leads to a short run
time that results in a high-quality output (2–3 compounds/day)
and a minimum in drug and solvent requirements with resulting
low-running costs.
By regression analysis, the influence of the modifier was con-
sidered. Further positive side effects are that one can get
additional valuable information of the drug concerning its chro-
matographic behavior (i.e., the peak shape for the investigated
stationary and mobile phases).
By connecting the pKa values, a hockey-stick-shaped curve was
obtained (12). If one uses the data that resulted from measure-
ments at higher amounts of the organic modifier to perform the
regression analysis, one will get slightly lower pKa values that will
not lay on the straight line resulting from the data obtained from
measurements at smaller amounts of organic solvent (12).
Therefore, the investigator should use the smallest amounts of
organic solvent and the shortest columns possible. Reduction in
the organic solvent concentration will lead to a prolonged retention
time, and the time needed to perform a single run will increase.
References
1. G. Rücker, M. Neugebauer, and G.G. Willems: Instrumentelle
Pharmazeutische Analytik, 2nd ed. VWG-Verlag, Stuttgart, Germany,
1992, pp. 335–36.
2. S. Beck and C.D. Bevan. The Determination of pKa Values. Hoechst
Pharmaceutical Research Laboratories, Internal report, pp. 1–12,
1982.
3. A. Fini, P.D. Maria, A. Guarnieri, and L. Varoli. Acidity constants of
sparingly water-soluble drugs from potentiometric determinations
in aqueous dimethyl sulfoxide. J. Pharm. Sci. 76: 48–52 (1987).
4. V. Martinez, M.I. Maguregui, R.M. Jimenez, and R.M. Alonso.
Determination of the pKa values of beta-blockers by automated
potentiometric titrations. J. Pharm. Biomed. Anal. 23(2-3): 459–68
(2000).
5. R.G. Franz. Comparisons of pKa and log P values of some car-
boxylic and phosphonic acids: synthesis and measurement. AAPS
Pharmsci. 3 (2): 1–13 (2001).
6. E. Bosch, P. Bou, H. Allemann, and M. Roses. Retention of ionizable
compounds on HPLC. pH scale in methanol-water and the pK and
pH values of buffers. Anal. Chem. 68: 3651–57 (1996).
7. B. Uno, N. Okumura, and S. Kawai. Reversed-phase high-perfor-
mance liquid-chromatographie behavior of phthalic acid and
terephthalic acid in the pH region around the 2. pKa value. Bull.
Chem. Soc. Jpn. 67: 3356–59 (1994).
8. H.Y. Ando and T. Heimbach. pKa determinations by using a HPLC
equipped with a DAD as a flow injection apparatus. J. Pharm.
Biomed. Anal. 16: 31–37 (1997).
9. N. Sistovaris, Y. Hamachi, and T. Kurikio. Multifunctional sub-
stances–determination of pKa values by various methods.
Fresenius’ J. Anal. Chem. 340: 345–49 (1991).
10. W. Martindale and J.E.F. Reynolds. Martindale, The Complete Drug
Reference, 30th ed. Pharmaceutical Press, London, U.K., 1993, p.
xxii.
11. Mizukami. Merck Index. S. Budavari, Ed. Merck Inc., Rahway, NJ,
Monographs, 4372, 1989.
12. A. Albert and E.P. Serjeant. Ionization Constants of Acids and Bases.
John Wiley & Sons, New York, NY, 1962, pp. 66–67.
Conclusion
This paper investigated the potential of an analytical method to
determine the pKa values of drug candidates that show poor solu-
bility. The advantages of this method are its simplicity, low need for
Manuscript accepted May 16, 2003.
326