Site-specific acid-base properties of tenoxicam

Article abstract of DOI:10.1002/hlca.200790174

The Hammett approach, as a new deductive tool, was introduced to characterize the otherwise inaccessible minor protonation pathway of tenoxicam (1), the non-steroidal anti-inflammatory drug. A total of eight compounds, constituting a systematic series of side chain-substituted analogues of tenoxicam and piroxicam (2), were synthesized and studied in terms of acid-base properties and Hammett constants to identify the ideal replacement of the unprotonated pyridin-2-yl group, a key moiety in both molecules. Hammett constants of the phenyl substituents have been found to be in a linear correlation with the experimental log K values of the enolate sites, the basic moiety of the extended conjugated system in this family of piroxicam derivatives. Then, a similar correlation was observed for the analogous tenoxicam derivatives. After identifying the 2-aza Hammett constant of the pyridin-2-yl group and the corresponding log K value, the site-specific acid-base properties of tenoxicam could be quantitated. This novel method is assessed to be a fine-tuning tool to find the ideal substituent by using analogue-based deductive method to obtain site-specific constants of the minor protonation/deprotonation pathway in drugs and biomolecules. The tenoxicam microconstant values indicate that the enolate moiety is of extremely low basicity (reflected by the log k^O = 3.70 and log k^O _N = 1.09 values), which can, however, be interpreted in terms of the peculiar ring system and the overwhelming electron-withdrawing effects of the adjacent heteroatoms. A diagram depicting the pH-dependent distribution of 1 microspecies is also presented.

Full text of DOI:10.1002/hlca.200790174

Helvetica Chimica Acta – Vol. 90 (2007)
1681
Site-Specific Acid – Base Properties of Tenoxicam
by Krist o´ f K o´ czi a´ n, Gergely Vçlgyi, J o´ zsef Kçkçsi, and B e´ la Nosz a´ l*
Department of Pharmaceutical Chemistry, Semmelweis University, Research Group for Drugs of Abuse
and Doping Agents, Hungarian Academy of Sciences, H o˝ gyes E. u. 9, H-1092 Budapest
(
phone/fax: þ 3612170891; e-mail: nosbel@gytk.sote.hu)
The Hammett approach, as a new deductive tool, was introduced to characterize the otherwise
inaccessible minor protonation pathway of tenoxicam (1), the non-steroidal anti-inflammatory drug. A
total of eight compounds, constituting a systematic series of side chain-substituted analogues of
tenoxicam and piroxicam (2), were synthesized and studied in terms of acidÀbase properties and
Hammett constants to identify the ideal replacement of the unprotonated pyridin-2-yl group, a key
moiety in both molecules. Hammett constants of the phenyl substituents have been found to be in a linear
correlation with the experimental log K values of the enolate sites, the basic moiety of the extended
conjugated system in this family of piroxicam derivatives. Then, a similar correlation was observed for the
analogous tenoxicam derivatives. After identifying the 2-aza Hammett constant of the pyridin-2-yl group
and the corresponding log K value, the site-specific acid – base properties of tenoxicam could be
quantitated. This novel method is assessed to be a fine-tuning tool to find the ideal substituent by using
analogue-based deductive method to obtain site-specific constants of the minor protonation/deproto-
nation pathway in drugs and biomolecules. The tenoxicam microconstant values indicate that the enolate
O
O
moiety is of extremely low basicity (reflected by the log k ¼ 3.70 and log k ¼ 1.09 values), which can,
N
however, be interpreted in terms of the peculiar ring system and the overwhelming electron-withdrawing
effects of the adjacent heteroatoms. A diagram depicting the pH-dependent distribution of 1
microspecies is also presented.
Introduction. – The protonation state of the non-steroidal anti-inflammatory drugs
(
NSAIDs) highly influences their pharmacokinetic processes and the concomitant
bioavailability. It is also related to undesired side effects such as gastrointestinal or
cardiovascular toxicity [1 – 3].
Oxicams, an important group of NSAIDs, act by blocking the formation of
prostaglandins through the inhibition of COX isoenzymes [4]. Concerning molecular
properties, oxicams are certainly examples of spectroscopic and structural peculiarities
due to the various heteroatom sites suitable for coordination, protonation, H-bonding
interactions, and tautomeric and conformational changes. To understand their
mechanism of action at the molecular level, and, also, to design more effective
derivatives, it is necessary to know their site-specific acidÀbase behavior in solution.
The site-specific protonation constants (microconstants and the less frequently
applicable group constants) are physico-chemical parameters of crucial importance to
describe the interactions of ampholyte drugs and biomolecules in the body [5][6].
Macroconstants, the most frequently used parameters in acidÀbase equilibria, describe
only the overall acidÀbase properties of the chemical entity of two or more ionizing
groups. In principle, however, they cannot be assigned to individual functional groups.
ꢀ
2007 Verlag Helvetica Chimica Acta AG, Zürich

1
682
Helvetica Chimica Acta – Vol. 90 (2007)
Thus, the protonation microconstants are essential parameters to characterize
protonation equilibria at the molecular and submolecular levels [7][8]. Their
significance is even greater in biological systems where the minor protonation isomer
can well be the reactive one in highly specific (patho)biochemical processes [6][9].
Two fundamental approaches exist to determine microconstants: combined
spectroscopic/pH-metric direct methods and deductive methods by using synthetic
model compounds. The former is widespread, and it frequently includes UV/VIS or
NMR techniques (see refs. cit. in [6]; for some recent articles, see [10 – 13]). However,
it requires some basic criteria in the molecule to be met. It is not applicable if
acidÀbase properties of the functional groups are very different, or the protonation
state of one of the proton-binding sites cannot be monitored selectively by spectro-
scopic methods, or the molecule is too complicated for spectral resolution. In these
cases, the deductive method can be effective. It uses an auxiliary ligand which is usually
a close derivative of the parent compound, but it contains a reduced number of
functional groups. Macroconstant of the auxiliary molecule can then be handled as the
appropriate microconstant of the main compound. Common such replacements for
ÀCOOH are ÀCOOMe [14][15] and ÀCONH [16].
2
Earlier, we resolved the piroxicam (2) protonation at the microscopic level [15].
Here, we report the microspeciation of tenoxicam (¼4-hydroxy-2-methyl-N-(pyridin-2-
yl)-2H-thieno[2,3-e][1,2]thiazine-3-carboxamide 1,1 dioxide; 1), which differs from 2 in
a fused five-membered thiophene heterocycle, replacing the fused benzene ring of 2.
Compound 1 has some advantages in therapy. It has a longer half-life and one-a-day
advantage over 2. Thus, it has achieved a significant role in the treatment of rheumatic
diseases over the last decades [17][18].
Compound 1 has a bifunctional amphoteric structure: the enolic acid and the
pyridin-2-yl moieties represent the acidic and basic sites, respectively. Accordingly, four
individual microspecies exist in protic solvents (e.g., in simple aqueous solutions and in
tissue fluids) namely anionic, zwitterionic, uncharged, and cationic ones (Scheme 1).

Helvetica Chimica Acta – Vol. 90 (2007)
Scheme 1. Microspeciation Scheme of 1
1683
The fused thienothiazine system and the conjugated enol-carboxamide moieties change
significantly upon protonation/deprotonation, as shown by the strongly pH-dependent
UV and NMR spectra.
The protonation properties of 1 have been studied at various levels [19][20].
Bernhard and Zimmermann [19] determined the macroscopic protonation properties
of some oxicams by UV-titrations. They reported log K ¼ 5.34 ꢀ 0.06 and log K ¼
1
2
1
.07 ꢀ 0.11 for 1. Tsai et al. [20] made estimations about the site-specific protonation
constants by using the following considerations. They assumed that the average of the
protonation macroconstants of isoxicam (3) and meloxicam (4) can approximate the
microconstant of the anionic-uncharged protonation equilibrium of 2. They also
assumed that replacement of the pyridin-2-yl group in 2 with a 5-methylisoxazol-3-yl
ring in 3 (log K ¼ 3.93) and 5-methylthiazol-2-yl ring in 4 (log K ¼ 4.08) means the
elimination of the basic pyridin-2-yl moiety without significant influence on the acidity
of the enolic OH group. They introduced the average of the macroconstants of 3 and 4
(
log K ꢁ 4) to calculate the microconstants of 2. Applying the microconstants, they
have also calculated the interactivity parameter that quantitates how protonation at
one site reduces the basicity of the other site. Finally, the interactivity parameter of 2
was combined with the macroconstants of 1, which allowed the approximation of the
microconstants of 1 [20].
Microconstants of the minor protonation pathway obtained by the above method
for piroxicam (2), however, differ by 0.6 log k units from the more recent
corresponding values obtained by the well-established O-methyl deductive method
[
15], while macroconstants and microconstants of the major protonation pathway are in
agreement within 0.04 log units.
Our study aimed, therefore, at the possible most accurate and reliable character-
ization of 1 at the microscopic level. This goal necessitated a new approach, since the
O-methyl-derivative formation for 1 has been proven infeasible, as we have recently

1
684
Helvetica Chimica Acta – Vol. 90 (2007)
shown [21] due to fast hydrolytic decomposition of O-methyl-tenoxicam in aqueous
media. For this purpose, we synthesized auxiliary compounds with no ionizing group
other than the enolic OH, but with pyridin-2-yl-mimicking, electron-withdrawing
character in the side chain. The set of new compounds provided a possibility to assign
Hammett constants to certain ring replacements after determining calibration lines
between acidity and ring-substituent constants. By means of this novel method, the
microspeciation of 1 could be carried out, based on experimental data of model
molecules in this family, and the related observations in their pK values and Hammett
a
constants.
Results and Discussion. – Preliminary experiments demonstrated that characteristic
differences occur in positions of UV spectra, and also, in the protonation constants of
tenoxicam analogues, depending on the quality and position of the substituent in the
phenyl moiety. Substituent effects play a significant role in a variety of physical and
chemical properties. For example, they influence the reactivity of the substrates, the
rates of nucleophilic substitutions [22], positions of UV absorption bands [23], NMR
chemical shifts [24], and the conformations of molecules [15]. The nature and position
of substituents also affect the protonation constants (log K values) [25].
Thus the choice of the right substituent in deductive modelling is of crucial
importance. Since the introduction of Hammett constants (s) [26], these parameters
have been eminent means for quantifying substituent effects. The Hammett constant,
by definition, is the difference between the logarithmic protonation constants of the
unsubstituted and substituted benzoic acids, taking into account the type and position
of the substituent [27 – 29]. The slope of the correlation line between log K values and
Hammett constants for benzoic acid is 1.00. For other molecules, the slope may differ
from unity, depending on the sensitivity of the basic site, and various steric and
inductive effects in the molecule.
Electron-withdrawing and steric effects of an adjacent substituent are the main
factors that influence the acidÀbase properties of such a molecule. For oxicams, the
conjugated enolic carboxamide structure and the relatively long distance between the
enolic and the pyridin-2-yl moieties, the electronic effects highly influence the extent of
acidity. Thus, they can be characterized in terms of Hammett constants. As a
consequence, acidity – Hammett constant correlation relationship can be established.
O
Since the most reliable value for log k , one of the microconstants in the minor
protonation pathway of 2 is 4.60 [15], an ideally substituted phenyl moiety would also
set log K ¼ 4.60. We assumed that such a molecule does not exactly exist, but close
derivatives can be prepared.
Our synthetic efforts to produce piroxicam derivatives 6a – 6d (Scheme 2) resulted
in compounds with log K values in the range of 4.40 – 4.81, and the corresponding
2
Hammett constants between 0.00 and 0.53 (Table 1). A close correlation (R ¼ 0.990)
has been found between the log K values and the Hammett constants (Fig. 1),
reinforcing the common origin of these two parameters.
By substitution of the value 4.60 into the linear-regression equation, an ideal
Hammett constant of 0.29 could be marked. This represents the substituent on the
phenyl group that would exert exactly the same influence on the enolic site as the
pyridin-2-yl moiety. The results are in good correlation with the work of Katritzky et al.

Helvetica Chimica Acta – Vol. 90 (2007)
1685
Scheme 2
Fig. 1. Correlation between protonation constants and Hammett constants of 6a – 6d
Table 1. Protonation Constants of 6a – 6d and 7a – 7d as a Function of Substituents and Hammett
Constants
Substituent
Hammett constant
log K of 6a – 6da)
log K of 7a – 7da)
H
p-Cl
m-CF3
0.00
0.23
0.43
0.53
4.81
4.67 )
4.50 )
4.40 )
3.89
3.77
3.59
3.53
b
b
b
p-CF
3
a) Standard deviation is within 0.01 – 0.03. b) Using Yasuda – Shedlovsky extrapolation.
[
30], who assessed a 2-aza s value of 0.268 of pyridin-2-yl group, based on IR intensity
o
measurements.
An analogous set of tenoxicam derivatives, 7a – 7d, were prepared and charac-
terized in our laboratory (details in Exper. Part). The correlation between the
protonation constant of 7a – 7d and the Hammett constants of the substituents of the
2
phenyl group (Table 1) gave also a linear fit with R > 0.98 (Fig. 2). Due to the high
structural similarity between 2 and 1, especially in the region where inductive, electron-
withdrawing effects take place, a moiety with a Hammett value of 0.29 can well be
O
assumed to mark the log k constant of 1 in an analogous linear regression diagram.

1
686
Helvetica Chimica Acta – Vol. 90 (2007)
Fig. 2. Correlation between protonation constants and Hammett constants of 7a – 7d
Applying the Hammett constant of 0.29, a protonation constant of 3.70 was pointed out
O
for the ideal phenyl analogue thus for the log k of the parent compound. All the
microconstants could be calculated, using the macroconstants values (log K ¼ 5.34 ꢀ
1
0
.06, log K ¼ 1.07) [19] and the following relationships between macro- and micro-
2
constants:
N
O
K ¼ k þ k
(1)
(2)
1
N
O
N
O
N
O
K K ¼ k k ¼ k k
1
2
In Table 2, all the protonation microconstants of 1 are listed.
Table 2. Summary of the Microconstants of 2 and 1
Microconstant
2
1
Tsai et al. [20]
Tak a´ cs-Nov a´ k et al. [15]
Tsai et al. [20]
This work
5.32
3.70 )
1.09
2.71
log kN
5.44
ca. 4 )
5.40
4.60
1.92
5.34 )
a
O
b
a
c
log k
log k
3.17 )
O
a
1.88
3.32
1.07 )
N
log kN
2.72 )
d
3.24 )
a
O
a
b
)
Calculated assuming same interactivity parameters of 2 and 1. ) Average of log K values of 3 and 4.
c) Determined by the novel deductive method (estimated standard deviation is 0.03). d) Determined
through O-methylpiroxicam.
The microconstants show that the two oxicam derivatives have nearly identical
O
O
NPir
O
Ten
O
NTen
interactivity parameters (log kPir À log k ¼ 2.68, log k À log k
¼ 2.61), an ob-
vious consequence of the structural homology.
Using these microconstant values, the pH-dependent distribution of the four
microspecies could be calculated for tenoxicam (Fig. 3). The zwitterionic species

Helvetica Chimica Acta – Vol. 90 (2007)
1687
Fig. 3. Logarithmic distribution curves for microspecies of 1
dominates over the uncharged form by a factor of 42. In fact, the major protonation
route goes through the zwitterionic form in 97.7%. It can also be seen that the whole
protonation takes place in the acidic region.
Conclusions. – This study of oxicams demonstrates that extended Hammett s
parameters can be used to account for substituent effects of pyridin-2-yl heterocycles in
determination of protonation microconstants. In this novel deductive method,
Hammett constants, harnessing molecular analogies, have been used to elucidate
protonation microconstants in 1.
The close correlation between log K values and Hammett constants for both 1 and 2
derivatives is a promising finding to determine hardly accessible macro- and micro-
constants for compounds in structurally related molecular families. The precise
microconstants provide the possibility to design more reliable COX – enzyme binding
operational model.
This work has been supported by OTKA (National Scientific Research Foundation, Hungary)
T 43579 grant.
Experimental Part
General. Materials. Tenoxicam (1) was a generous gift from Mih a´ ly Kata, Department of
Pharmaceutical Technology, University of Szeged. Saccharin sodium was obtained from Hungaropharma
(
H-Budapest). All other reagents and solvents for the synthetic work were purchased from Sigma (D-
Steinheim) and were used as received. Of the chemicals used for pH electrode calibration and log K
determination, KCl (anal. grade), KOH, and HCl were from Sigma, and MeOH (HPLC grade) from
Merck (D-Darmstadt). KOH and HCl were standardized by titration against primary standards KH-
phthalate and (amino)tris(hydroxymethyl)methane, resp. (Sigma). Bidistilled H O was used throughout
2
this study. M.p.: in open cap. tubes on Stuart Scientific SMP melting-point apparatus. UV/VIS Spectra:
3
À1
À1 1
Jasco V-550 spectrophotometer at r.t.; lmax in nm, e in dm mol cm . H-NMR Spectra: Varian Unity
Inova 600 MHz spectrometer in (D )DMSO at r.t.; d in ppm rel. TMS, J in Hz.
6
Ethyl 4-Hydroxy-2-methyl-2H-1,2-benzothiazine-3-carboxylate 1,1-Dioxide (5). It was prepared
from the readily accessible saccharin sodium in three steps as described in [31 – 33]. Saccharin sodium

1
688
Helvetica Chimica Acta – Vol. 90 (2007)
(
61.5 g, 0.30 mol) was dissolved in 240 ml of DMF and reacted with ethyl 2-chloroacetate (equiv.
quantity) at 1008 for 3 h according to [31]. The soln. was poured onto ice-water (4 ꢂ excess in volume),
that resulted in 57.7 g (72%) of the ethyl (3-oxo-1,2-benzoisothiazol-2(3H)-yl)acetate 1,1-dioxide) as a
white precipitate. The product (20 g, 0.074 mol) was dissolved in 3 equiv. instead of 2 [32] of freshly
prepared EtONa soln. (5.12 g of Na in 80 ml abs. EtOH). The mixture was kept at 558 for 1 h so that a
Gabriel – Colman rearrangement could occur. HCl (9 ml, of 9%) was added during continuous stirring
and not allowing the temp. to rise above 58. The whole mixture was kept in a refrigerator overnight. The
white precipitates were filtered off to yield 13.6 g (68%) of ethyl 4-hydroxy-2H-benzo[1,2]thiazine-3-
carboxylate 1,1-dioxide. According to [33], 10 g (0.037 mol) of the former product was combined to 8 ml
of MeI, 40 ml of H O, 140 ml of EtOH, and 40 ml of 1n NaOH. After standing at r.t. for 18 h, the yellow
2
precipitation was filtered, washed with H O, and dried to yielding 7.5 g (72%) of 5. M.p. 161 – 1638. UV/
2
1
VIS (MeOH): 322 (14000). H-NMR (600 MHz, (D )DMSO): 1.33 (t, J ¼ 7.1, MeCH ); 2.82 (s, MeN);
6
2
4
.36 (q, J ¼ 7.1, MeCH ); 7.84 (t, J ¼ 7.4, HÀC(6)); 7.86 (t, J ¼ 7.4, HÀC(7)); 7.88 (d, J ¼ 7.1, H ÀC(5)); 8.04
2
(d, J ¼ 7.1, H ÀC(8)).
Piroxicam Analogues Suitable for Deductive Characterization (6a – 6d). Compound 5 (0.28 g,
0
.001 mol) was suspended in DMF (3 ml), excesses (3ꢂ) of substituted anilines were added, and the
mixtures were heated under N for 12 h at 140 – 1508 (oil bath). The mixtures were cooled and then
2
poured into 15 ml of HCl soln. containing mol-equivalent quantity of acid. The precipitous mixtures were
placed into refrigerator and kept there for 16 h, then filtered, and the precipitates were washed 3 times
with H O and cold i-PrOH.
2
4
-Hydroxy-2-methyl-N-phenyl-2H-1,2-benzothiazine-3-carboxamide 1,1-Dioxide (6a). White crys-
1
talline solid. Yield: 57%. M.p. 216 – 2178. UV/VIS (MeOH): 334 (17000), 240 (14000). H-NMR
(
600 MHz, (D )DMSO): 2.87 (s, MeN); 7.19 (m, HÀC(4’)); 7.39 (t, J ¼ 7.1, H ÀC(3’), HÀC(5’)); 7.68 –
6
7.76 (m, HÀC(6), HÀC(7)); 7.85 – 7.96 (m, HÀC(5), HÀC(2’), HÀC(6’)); 8.04 (d, J ¼ 7.4, HÀC(8));
1
0.30 (s, OH).
N-(4’-Chlorphenyl)-4-Hydroxy-2-methyl-2H-1,2-benzothiazine-3-carboxamide 1,1-Dioxide (6b).
1
White crystalline solid. Yield: 64%. M.p. 230 – 2328. UV (MeOH): 334 (23000), 249 (19000). H-NMR
(
600 MHz, (D )DMSO): 2.86 (s, MeN); 7.45 (d, J ¼ 7.8, HÀC(3’), HÀC(5’)); 7.79 (d, J ¼ 7.8, HÀC(2’),
6
HÀC(6’)); 7.86 – 7.95 (m, HÀC(5), HÀC(6), HÀC(7)); 8.04 (d, J ¼ 7.4, HÀC(8)); 10.42 (s, OH).
4-Hydroxy-2-methyl-N-[3-(trifluoromethyl)phenyl]-2H-1,2-benzothiazine-3-carboxamide 1,1-Diox-
ide (6c). White crystalline solid. Yield: 60%. M.p. 199 – 2018. UV (MeOH): 334 (24000), 249 (18000).
1
H-NMR (600 MHz, (D )DMSO): 2.88 (s, MeN); 7.54 (d, J ¼ 8.0, HÀC(4’)); 7.64 (t, J ¼ 8.0, HÀC(5’));
6
7
.88 – 7.95 (m, HÀC(5), HÀC(6), HÀC(7)); 8.06 (d, J ¼ 7.6, HÀC(8)); 8.12 (d, J ¼ 8.0, HÀC(6’)); 8.19 (s,
HÀC(2’)); 10.57 (s, OH).
4-Hydroxy-2-methyl-N-[4-(trifluoromethyl)phenyl]-2H-1,2-benzothiazine-3-carboxamide 1,1-Diox-
ide (6d). White crystalline solid. Yield: 56%. M.p. 237 – 2398. UV (MeOH): 334 (20000), 253 (16000).
1
H-NMR (600 MHz, (D )DMSO): 2.88 (s, MeN); 7.77 (d, J ¼ 8.6, HÀC(3’), HÀC(5’)); 7.88 – 7.95 (m,
6
HÀC(5), HÀC(6), HÀC(7)); 8.02 (d, J ¼ 8.6, HÀC(2’), HÀC(6’)); 10.60 (s, OH).
Synthesis of Tenoxicam Analogues 7a – 7d. The substituted benzene derivatives of tenoxicam were
prepared as described in [21]. Detailed description of the procedures and products for phenyl, 3-
(
trifluoromethyl)phenyl, and 4-(trifluoromethyl)phenyl derivatives are provided therein.
N-(4-Chlorophenyl)-4-hydroxy-2-methyl-2H-thieno[2,3-e][1,2]thiazine-3-carboxamide 1,1-Dioxide
(
(
7c). Yellow crystals. Yield: 60%. M.p. 192 – 1948. UV (MeOH): 386 (19000), 296 (14000), 270
1
15000). H-NMR (600 MHz, (D )DMSO): 2.99 (s, MeN); 7.23 (d, J ¼ 7.8, HÀC(3’), HÀC(5’)); 7.41 (d,
6
J ¼ 7.8, HÀC(2’), HÀC(6’)); 7.39 (d, J ¼ 5.2, HÀC(7)); 7.86 (d, J ¼ 5.2, HÀC(8)); 10.55 (s, OH).
Spectrophotometric log K Determination. The pH-metric titrations were performed at 25.0 ꢀ 0.18 and
at constant ionic strength (0.15 mm KCl) on a GLpKa automatic titrator (Sirius Analytical Instruments
TM
Ltd., Forest Row, UK). The four-parameter technique (Four Plus method) was used for electrode
calibration in both aqueous medium and MeOH/H O mixtures [34][35]. The Refinement Pro software
2
(
Sirius) was used to record and evaluate the titration curves. The log K values of samples were measured
TM
spectrophotometrically using the D-PAS
UV spectrometer attachment for GLpKa [36]. In the
spectrophotometric method, multi-wavelength UV absorbance of the sample soln. is monitored

Helvetica Chimica Acta – Vol. 90 (2007)
1689
throughout the titration using the fiber-optic probe in-situ. This method allows significantly smaller
À5
sample concentrations for log K measurements (10 m or below).
UV-pH Titration in Aqueous Medium. For all compounds, a stock sample soln. of ca. 1.0-mm
concentration was prepared in MeOH. Aliquots (100 ml or 500 ml) of the stock solns. were then
transferred into a sample vial containing 15 ml of 0.15m KCl soln. to achieve the required sample
concentration. Sample concentrations of 6 – 30 mm were employed. In each experiment, the pH of the
sample soln. of 15-ml volumes was adjusted, using 0.5m KOH, to ca. 8 and then titrated with 0.5m HCl to a
suitably low value (ca. 2). Spectral data were recorded in the region of 200 – 700 nm after each pH
measurement. About 20 pH readings and absorption spectra were collected from each titration. The
titrations were carried out at constant ionic strength (I ¼ 0.15m KCl) and temp. (T¼ 25.0 ꢀ 0.58) under
N . A minimum of three parallel measurements were carried out, and the log K values were calculated
2
TM
with RefinementPro software.
UV-pH Titration in MeOH/H O Mixtures and Extrapolation of Sample log K’ Values Measured in
2
MeOH/H O Mixtures to Aqueous log K Values. The cosolvent dissociation constants of H O-insoluble
2
2
compounds 6b, 6c, and 6d were determined in various MeOH/H O mixtures between 16 – 44 wt-%. The
2
same titration protocol was performed as described above. Each sample was measured at least in four
different MeOH/H O mixtures. The measured log K’ values in MeOH/H O solns. were extrapolated to
2
2
aq. conditions according to the Yasuda – Shedlovsky procedure [37][38], which is the most widely used
method among cosolvent techniques [39][40].
REFERENCES
[
[
[
[
[
[
1] A. L. Blobaum, L. J. Marnett, J. Med. Chem. 2007, 50, 1425.
2] M. R. Weir, R. S. Sperling, A. Reicin, B. J. Gertz, Am. Heart J. 2003, 146, 591.
3] F. Reymond, P. A. Carrupt, B. Testa, H. H. Girault, Chem.–Eur. J. 1999, 5, 39.
4] A. D. Rodrigues, DrugMetab. Dispos. 2005, 33, 1567.
5] B. Nosz a´ l, J. Phys. Chem. 1986, 90, 4104.
6] B. Nosz a´ l, in ꢁBiocoordination Chemistry: Coordination Equilibria in Biologically Active Systemsꢂ,
Ed. K. Burger, Ellis Horwood, Chichester, 1990, p. 18.
[
7] J. Bjerrum, Z. Phys. Chem. 1923, 106, 209.
[
8] K. Tak a´ cs-Nov a´ k, B. Nosz a´ l, I. Hermecz, G. Kereszturi, B. Pod a´ nyi, G. Sz a´ sz, J. Pharm. Sci. 1990, 79,
1023.
[
9] B. Nosz a´ l, V. Schellerkrattiger, R. B. Martin, J. Am. Chem. Soc. 1982, 104, 1078.
[
[
[
[
[
[
10] Z. Mandic, V. Gabelica, J. Pharmaceut. Biomed. 2006, 41, 866.
11] P. I. Nagy, G. Vçlgyi, K. Tak a´ cs-Nov a´ k, Mol. Phys. 2005, 103, 1589.
12] M. H. Langlois, M. Montagut, J. P. Dubost, J. Grellet, M. C. Saux, J. Pharm. Biomed. 2005, 37, 389.
13] Z. Szak a´ cs, M. Kraszni, B. Nosz a´ l, Anal. Bioanal. Chem. 2004, 378, 1428.
14] L. Ebert, Z. Phys. Chem. 1926, 121, 385.
15] K. Tak a´ cs-Nov a´ k, J. Kçkçsi, B. Pod a´ nyi, B. Nosz a´ l, R. S. Tsai, G. Lisa, P. A. Carrupt, B. Testa, Helv.
Chim. Acta 1995, 78, 553.
[
[
[
[
[
16] B. Nosz a´ l, P. S a´ ndor, Anal. Chem. 1989, 61, 2631.
17] P. A. Todd, S. P. Clissold, Drugs 1991, 41, 625.
18] Z. Unlu, K. Ay, J. Musculoskel. Pain 2006, 14, 37.
19] E. Bernhard, F. Zimmermann, Arzneim.-Forsch. 1984, 34 – 1, 647.
20] R. S. Tsai, P. A. Carrupt, N. Eltayar, Y. Giroud, P. Andrade, B. Testa, F. Bree, J. P. Tillement, Helv.
Chim. Acta 1993, 76, 842.
[
[
21] K. K o´ czi a´ n, J. Kçkçsi, K. Maz a´ k, B. Nosz a´ l, Helv. Chim. Acta 2005, 88, 2355.
22] T. H. Lowry, K. S. Richardson, ꢁMechanism and Theory in Organic Chemistryꢂ, 3rd edn., Harper
Collins, New York, 1987.
[
23] J. Ohshita, H. Kai, A. Takata, T. Iida, A. Kunai, N. Ohta, K. Komaguchi, M. Shiotani, A. Adachi, K.
Sakamaki, K. Okita, Organometallics 2001, 20, 4800.
[
24] C. R. Kaiser, E. A. Basso, R. Rittner, Magn. Reson. Chem. 2001, 39, 643.

1
690
Helvetica Chimica Acta – Vol. 90 (2007)
[
[
[
[
[
[
25] W. H. Beck, M. Liler, D. G. Morris, J. Chem. Soc., Perkin Trans. 2 1977, 1876.
26] L. P. Hammett, J. Am. Chem. Soc. 1937, 59, 96.
27] L. P. Hammett, ꢁPhysical Organic Chemistryꢂ, 1st edn., McGraw Hill, New York, 1940, p. 121.
28] L. P. Hammett, ꢂPhysical Organic Chemistryꢂ, 2nd edn., McGraw Hill, New York, 1972, p. 347.
29] K. Sekigawa, Tetrahedron 1972, 28, 505.
30] A. R. Katritzky, C. R. Palmer, F. J. Swinbourne, T. T. Tidwell, R. D. Topsom, J. Am. Chem. Soc. 1969,
91, 636.
[
[
[
[
[
[
[
[
[
[
31] M. S. Park, E. S. Chang, M. S. Lee, S. K. Kwon, Bull. Korean Chem. Soc. 2002, 23, 1836.
32] H. Zinnes, R. A. Comes, F. R. Zuleski, A. N. Caro, J. Shavel, J. Org. Chem. 1965, 30, 2241.
33] J. G. Lombardino, E. H. Wiseman, W. M. McLamore, J. Med. Chem. 1971, 14, 1171.
34] A. Avdeef, J. Pharm. Sci. 1993, 82, 183.
35] A. Avdeef, J. E. A. Comer, S. J. Thomson, Anal. Chem. 1993, 65, 42.
36] K. Y. Tam, K. Tak a´ cs-Nov a´ k, Anal. Chim. Acta 2001, 434, 157.
37] M. Yasuda, Bull. Chem. Soc. Jpn. 1959, 32, 429.
38] T. Shedlovsky, in ꢁElectrolytesꢂ, Ed. B. Pesce, Pergamon, New York, 1962, p. 146.
39] K. Tak a´ cs-Nov a´ k, K. J. Box, A. Avdeef, Int. J. Pharm. 1997, 151, 235.
40] A. Avdeef, K. J. Box, J. E. A. Comer, M. Gilges, M. Hadley, C. Hibbert, W. Patterson, K. Y. Tam, J.
Pharm. Biomed. 1999, 20, 631.
Received June 5, 2007