Kinetics and mechanism of the pyridinolysis of S-4-nitrophenyl 4-substituted thiobenzoates in aqueous ethanol

Article abstract of DOI:10.1021/jo049260f

The pyridinolysis of S-4-nitrophenyl 4-X-substituted thiobenzoates (X = H, Cl, and NO₂; 1, 2, and 3, respectively) is studied kinetically in 44 wt % ethanol-water, at 25.0 °C and an ionic strength of 0.2 M (KCl). The reactions are measured spectrophotometrically (420-425 nm) by following the appearance of 4-nitrobenzenethiolate anion. Pseudo-first-order rate coefficients (k_obsd) are obtained throughout, under excess of amine over the substrate. Plots of k_obsd vs [free amine] at constant pH are linear with the slope (k_N) independent of pH. The Bronsted-type plot (log k_N vs pK_a of the conjugate acids of the pyridines) for the reactions of thiolbenzoate 1 is curved with a slope at high pK _a, β₁ = 0.20, and slope at low pK_a, β₂ = 0.94. The pK_a value for the center of the Bronsted curvature is pK_a⁰ = 9.7. The pyridinolysis of thiolbenzoates 2 and 3 show linear Bronsted-type plots of slopes 0.94 and 1.0, respectively. These results and other evidence indicate that these reactions occur with the formation of a zwitterionic tetrahedral intermediate (T^±). For the pyridinolysis of thiolbenzoate 1, breakdown of T^± to products (k₂ step) is rate-limiting for weakly basic pyridines and T^± formation (k₁ step) is rate-determining for very basic pyridines. The k₂ step is rate-limiting for the reactions of thiolbenzoates 2 and 3. The smallest pK _a⁰ value for the reaction of 1 is due to a the weakest electron withdrawal of H (relative to Cl and NO₂) in the acyl group, which results in the smallest k_-1/k₂ ratio. The pK _a⁰ values for the title reactions are smaller than those for the reactions of secondary alicyclic amines with thiolbenzoates 1-3. This is attributed to a lower leaving ability from the T^± of pyridines than isobasic alicyclic amines. The lower pK_a⁰ value found for the pyridinolysis of 2,4-dinitrophenyl benzoate (pK _a⁰ = 9.5), compared with that for the pyridinolysis of 1, is explained by the greater nucleofugality from T^± of 2,4-dinitrophenoxide than 4-nitrobenzenethiolate, which renders the k _-1/k₂ ratio smaller for the reactions of the benzoate relative to thiolbenzoate 1. The title reactions are also compared with the aminolysis of similar thiolbenzoates in other solvents to assess the solvent effect.

Full text of DOI:10.1021/jo049260f

Kin etics a n d Mech a n ism of th e P yr id in olysis of S-4-Nitr op h en yl
-Su bstitu ted Th ioben zoa tes in Aqu eou s Eth a n ol
4
Enrique A. Castro,* Melissa Vivanco, Raul Aguayo, and J os e´ G. Santos
Facultad de Qu ´ı mica, Pontificia Universidad Cat o´ lica de Chile, Casilla 306, Santiago 22, Chile
ecastro@puc.cl
Received April 30, 2004
The pyridinolysis of S-4-nitrophenyl 4-X-substituted thiobenzoates (X ) H, Cl, and NO
2
; 1, 2, and
3
, respectively) is studied kinetically in 44 wt % ethanol-water, at 25.0 °C and an ionic strength
of 0.2 M (KCl). The reactions are measured spectrophotometrically (420-425 nm) by following the
appearance of 4-nitrobenzenethiolate anion. Pseudo-first-order rate coefficients (kobsd) are obtained
throughout, under excess of amine over the substrate. Plots of kobsd vs [free amine] at constant pH
are linear with the slope (k
conjugate acids of the pyridines) for the reactions of thiolbenzoate 1 is curved with a slope at high
pK , â ) 0.20, and slope at low pK , â ) 0.94. The pK value for the center of the Brønsted curvature
is pK ) 9.7. The pyridinolysis of thiolbenzoates 2 and 3 show linear Brønsted-type plots of slopes
N N a
) independent of pH. The Brønsted-type plot (log k vs pK of the
a
1
0
a
2
a
a
0
.94 and 1.0, respectively. These results and other evidence indicate that these reactions occur
(
with the formation of a zwitterionic tetrahedral intermediate (T ). For the pyridinolysis of
(
thiolbenzoate 1, breakdown of T to products (k
2
step) is rate-limiting for weakly basic pyridines
step) is rate-determining for very basic pyridines. The k step is rate-limiting
for the reactions of thiolbenzoates 2 and 3. The smallest pK value for the reaction of 1 is due to
(
and T formation (k
1
2
0
a
the weakest electron withdrawal of H (relative to Cl and NO
2
) in the acyl group, which results in
0
the smallest k-1/k
reactions of secondary alicyclic amines with thiolbenzoates 1-3. This is attributed to a lower leaving
2
ratio. The pK values for the title reactions are smaller than those for the
a
(
0
a
ability from the T of pyridines than isobasic alicyclic amines. The lower pK value found for the
0
pyridinolysis of 2,4-dinitrophenyl benzoate (pK ) 9.5), compared with that for the pyridinolysis of
a
(
1
, is explained by the greater nucleofugality from T of 2,4-dinitrophenoxide than 4-nitroben-
2
zenethiolate, which renders the k-1/k ratio smaller for the reactions of the benzoate relative to
thiolbenzoate 1. The title reactions are also compared with the aminolysis of similar thiolbenzoates
in other solvents to assess the solvent effect.
In tr od u ction
thiobenzoates (1, 2, and 3) in aqueous ethanol proceed
by stepwise mechanisms. This was concluded on the
3
We have been interested lately in the mechanisms of
the aminolysis of S-aryl thiobenzoates (aryl thiolben-
zoates). To our knowledge, there have been only a few
reports in the literature on the mechanisms of these
reactions.1 There are contrasting results concerning
some of these reports. The reactions of aryl thiolbenzoates
with substituted anilines in methanol were claimed to
be concerted, namely, a single-step mechanism, without
basis of the curved Brønsted-type plots obtained for the
two former aminolyses and the linear plot found for the
latter reactions. For the aminolysis of thiolbenzoates 1
3
and 2 the Brønsted slopes at low pK
a
are â
2
) 0.86 and
are â
.27 and 0.10, respectively. These results were inter-
-3
0
0
.84, respectively, whereas those at high pK
a
1
)
3
preted as a change in the rate-determining step, from a
(
zwitterionic tetrahedral intermediate (T ) breakdown to
2
a
the formation of a zwitterionic tetrahedral intermediate.
(
3
T formation, as the secondary amine basicity increases.
2
b
On the other hand, the benzylaminolysis and pyridino-
lysis2c of the same thiolbenzoates in acetonitrile were
The Brønsted slope value of 0.81 exhibited by the
reactions of thiolbenzoate 3 is consistent with rate-
found to be stepwise, through the formation and break-
down of the zwitterionic intermediate.
We have recently found that the reactions of secondary
(
3
limiting breakdown to products of the intermediate T .
The reactions of 2,4-dinitrophenyl 4-X-substituted ben-
zoates (X ) Cl, CN, NO ) with a series of pyridines in
aqueous ethanol show linear Brønsted plots of slopes
.94, 0.96, and 0.90, respectively, in agreement with
stepwise mechanisms where the breakdown of the inter-
alicyclic amines with S-4-nitrophenyl 4-X-substituted
2
0
(
(
1) Castro, E. A. Chem. Rev. 1999, 99, 3505.
2) (a) Lee, I.; Shim, C. S.; Lee, H. W. J . Chem. Res., Synop. 1992,
9
0. (b) Lee, I.; Koh, H. J . New J . Chem. 1996, 20, 131. (c) Koh, H. J .;
Han, K. L.; Lee, I. J . Org. Chem. 1999, 64, 4783.
3) Castro, E. A.; Bessolo, J .; Aguayo, R.; Santos, J . G. J . Org. Chem.
(
4a-c
mediate T to products is the rate-determining step.
For the pyridinolysis of 2,4-dinitrophenyl benzoate (X )
H) a curved Brønsted plot was found, with slopes â 0.9
(
2
003, 68, 8157.
2
1
0.1021/jo049260f CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/08/2004
J . Org. Chem. 2004, 69, 5399-5404
5399

Castro et al.
and â
1
0.4 at low and high pK
a
, respectively, consistent
TABLE 1. Exp er im en ta l Con d ition s a n d kobsd Va lu es for
a
th e P yr id in olysis of S-4-Nitr op h en yl Th ioben zoa te (1)
with a stepwise process and a change in the rate-limiting
step.4d
pyridine
substituent
10 [N]tot
(M)c
2
103kobsd
-1
no. of
runs
b
pH
F
N
(s
)
With the aim to analyze the discrepancies concerning
some of the reactions above and to clarify the mecha-
nisms of the aminolysis of aryl thiolbenzoates, in the
present work we study kinetically the pyridinolysis of
thiolbenzoates 1-3 in aqueous ethanol. By a comparison
between these reactions the effect of the nonleaving group
of the substrate on the kinetics and mechanism can be
analyzed. Comparison of the reactions of the present
work with those of secondary alicyclic amines with the
same thiolbenzoates in the same solvent will allow
evaluation of the effect of the amine nature. By compar-
ing the present reactions with the aminolysis of thiol-
benzoates in other solvents and the pyridinolysis of 2,4-
dinitrophenyl benzoates in aqueous ethanol, the influence
3,4-diamino
9.15 0.333 0.23-2.09
2.5-18.7
2.7-26.9
4.0-37.0
9
9
9
8
8
8
7
6
7
7
8
10
10
9
9
7
5
9
9
.45 0.500 0.23-2.33
.75 0.667 0.23-2.33
4
4
-dimethylamino 8.84 0.333 0.207-2.07 1.45-11.2
9.14 0.500 0.207-1.86 2.38-18.6
9
.44 0.667 0.207-1.66 2.71-22.0
-amino
8.68 0.333 0.70-2.09
8
9
1.42-5.41
.98 0.500 1.16-2.33
3.43-8.62
.28 0.667 0.23-2.09
1.66-11.9
d
4-amino-3-bromo 7.55 0.817 0.49-2.23
0.075-0.368
0.085-0.390
0.037-0.203
0.038-0.220
0.020-0.112
0.029-0.122
0.0147-0.0556
0.0163-0.0551
3
7.85
d
d
0.899 0.35-1.73
3
4
,4-dimethyl
-methyl
7.55 0.987 1.98-19.8
7
d
.85 0.993 1.98-19.8
d
7.55 0.994 1.98-19.8
2
d
7.85 0.997 1.98-19.8
4
3-methyl
7.55
7
d
d
0.998 3.96-19.8
.85 0.999 3.96-19.8
of the solvent and the leaving group, respectively, can
be assessed.
a
In 44 wt % ethanol-water, at 25.0 °C, ionic strength 0.2 M
(
KCl). b Fraction of free amine. c Concentration of total amine (free
amine plus its conjugate acid). Under the presence of phosphate
buffer 0.01 M.
d
TABLE 2. Exp er im en ta l Con d ition s a n d kobsd Va lu es for
th e P yr id in olysis of S-4-Nitr op h en yl 4-Ch lor oth io-
ben zoa te (2)a
2
103kobsd
-1
pyridine
substituent
10 [N]tot
(M)c
no. of
runs
b
pH
F
N
(s )
Exp er im en ta l Section
3
,4-diamino
9.15 0.333 0.099-0.99 2.94-18.9
.45 0.500 0.099-0.99 4.19-26.7
9.75 0.667 0.099-0.99 6.92-38.0
4-dimethylamino 8.84 0.333 0.099-0.99 1.04-7.90
10
10
10
8
9
Ma ter ia ls. The series of pyridines were purified either by
4
d,5
distillation or recrystallization.
Thiolbenzoates 1-3 were
6
synthesized as described. Their melting points were in ac-
9
.14 0.500 0.099-0.99 1.77-11.8
8.98 0.500 0.099-0.99 1.11-7.70
.28 0.667 0.099-0.99 1.23-11.4
9
7
1
13
cordance with literature values, and their H and C NMR
spectra and elemental analyses agreed with their structures.
Kin etic Mea su r em en ts. The reactions were studied spec-
trophotometrically at 420-425 nm by means of a diode array
instrument. At these wavelengths an increase of absorbance
was observed due to the appearance of 4-nitrobenzenethiolate
anion. The experimental conditions of the reactions were 44
wt % ethanol-water solutions, at 25.0 ( 0.1 °C, and an ionic
strength of 0.2 M (maintained with KCl). Two or three pH
values were employed for the reactions of each pyridine. For
the reactions of the substrates with the three more basic
pyridines, the pH was kept constant by the amine and its
conjugate acid, whereas for the reactions with the other
pyridines, phosphate was used as an external buffer. At least
a 10-fold excess of total amine over the substrate was
4
-amino
10
10
7
9
8
10
7
8
7
7
9
d
4
-amino-3-bromo 7.55 0.817 0.248-2.48 0.167-0.618
d
d
7.85
0.899 0.173-1.73 0.289-0.678
3,4-dimethyl
7.25 0.974 2.97-14.9
0.0455-0.181
0.0284-0.186
0.0520-0.169
0.0575-0.196
0.0785-0.230
0.0311-0.198
0.0150-0.0975
0.0262-0.150
0.0053-0.151
d
7.55 0.987 1.49-14.9
d
7
.85 0.993 2.97-14.9
d
4
3
-methyl
-methyl
7.25 0.988 5.94-26.7
d
7
7
.55 0.994 8.91-29.7
d
.85 0.997 2.97-29.7
d
7.25 0.995 2.97-29.7
9
9
10
d
7
7
.55 0.998 4.95-49.5
d
.85 0.999 4.95-49.5
a
In 44 wt % ethanol-water, at 25.0 °C, ionic strength 0.2 M
(
KCl). b Fraction of free amine. c Concentration of total amine (free
d
-5
amine plus its conjugate acid). Under the presence of phosphate
buffer 0.01 M.
employed. The initial substrate concentration was (2-5) × 10
M in all runs.
Pseudo-first-order rate coefficients (kobsd) were found in all
cases. For the reactions of the three thiolbenzoates with the
three more basic pyridines and that of thiolbenzoates 2 and 3
with 4-amino-3-bromopyridine, the values of kobsd were ob-
tained by means of the kinetic software of the spectrophotom-
eter, after ca. 3 half-lives. After these reaction times a slow
decrease of absorbance at 420-425 nm was observed, due to
the formation of bis(4-nitrophenyl) disulfide (see below). For
the reactions of the three thiolbenzoates with the three less
basic pyridines and that of thiolbenzoate 1 with 4-amino-3-
bromopyridine, the values of kobsd were determined by the
initial rate method.8
The experimental conditions of the reactions and the kobsd
values are shown in Tables S1-S3 in Supporting Information
and in Tables1-3.
Deter m in a tion of p K Va lu es. The pK values of 3,4-
a a
diamino and 4-amino-3-bromo pyridines were determined by
a potentiometric method, in 44 wt % ethanol-water, at 25.0
(
0.1 °C, and an ionic strength of 0.2 M (maintained with KCl).
The values obtained are 9.45 ( 0.09 and 6.90 ( 0.08,
(
4) (a) Castro, E. A.; Valdivia, J . L. J . Org. Chem. 1986, 51, 1668.
b) Castro, E. A.; Becerra, M. C. Bol. Soc. Chil. Quim. 1988, 33, 67. (c)
Castro, E. A.; Steinfort, G. B. J . Chem. Soc., Perkin Trans. 2 1983,
53. (d) Castro, E. A.; Santander, C. L. J . Org. Chem. 1985, 50, 3595.
5) Bond, P. M.; Castro, E. A.; Moodie, R. B. J . Chem. Soc., Perkin
Trans. 2 1976, 68.
6) Bunton, C. A.; Foroudian, H. J .; Kumar, A. J . Chem. Soc., Perkin
Trans. 2 1995, 33.
7) Cilento, G. J . Am. Chem. Soc. 1953, 75, 3748. Kanaoka, Y.;
Tanizawa, K.; Sato, E.; Yonemitsu, O.; Ban, Y. Chem. Pharm. Bull.
respectively.
(
P r od u ct Stu d ies. One of the products of the pyridinolysis
of the thiolbenzoates is 4-nitrobenzenethiolate anion. The other
final product is the corresponding benzoate anion. These
anions and not their conjugate acids were identified as
4
(
(
(
(8) Satterthwait, A. C.; J encks, W. P. J . Am. Chem. Soc. 1974, 96,
7018. Ba-Saif, S.; Luthra, A. K.; Williams, A. J . Am. Chem. Soc. 1987,
109, 6362.
1
967, 15, 593.
5
400 J . Org. Chem., Vol. 69, No. 16, 2004

Pyridinolysis of Thiobenzoates
TABLE 3. Exp er im en ta l Con d ition s a n d kobsd Va lu es for
described above. Oxidation of benzenethiolate and 4-nitroben-
zenethiolate anions has been previously observed.
1
1
th e P yr id in olysis of S-4-Nitr op h en yl 4-Nitr oth ioben zoa te
a
(
3)
2
103kobsd
-1
pyridine
substituent
10 [N]tot
no. of
runs
Resu lts a n d Discu ssion
b
(M)c
pH
F
N
(s )
3
,4-diamino
9.15 0.333 0.099-0.99 8.8-92.3
10
9
9
10
10
10
10
8
10
8
9
8
6
The rate law found for all of the reactions subjected to
this work is given by eqs 1 and 2, where NPS , S, and N
represent 4-nitrobenzenethiolate anion, the thiolbenzoate
,and the free pyridine, respectively. The rate constants
9
9
.45 0.500 0.099-0.99 20.5-136
.75 0.667 0.099-0.89 19.9-176
-
4
4
4
3
-dimethylamino 9.14 0.500 0.099-0.99 13.3-80.6
.44 0.667 0.099-0.99 10.6-90.0
8.98 0.500 0.099-0.99 4.20-54.2
9
-amino
k
0
and k
N
are those for solvolysis and pyridinolysis of the
9
.28 0.667 0.099-0.99 15.5-75.3
substrates, respectively.
d
-amino-3-bromo 7.55 0.817 0.25-2.23
0.51-1.61
d
7
.85 0.899 0.17-1.73
0.42-1.33
d
d[NPS-]
,4-dimethyl
7.25 0.974 0.99-9.90
7
7
7.25 0.988 0.99-9.90
7
7
7.25 0.995 0.99-9.90
7
7
0.058-0.47
d
.55 0.987 0.99-9.90
0.126-0.537
0.077-0.485
0.043-0.242
0.035-0.252
0.048-0.260
0.0174-0.0902
0.0295-0.0975
0.0427-0.115
)
kobsd[S]
(1)
(2)
d
.85 0.993 0.99-9.90
dt
kobsd ) k + k [N]
d
4
3
-methyl
d
.55 0.994 0.99-9.90
8
9
8
8
0
N
d
.85 0.997 0.99-9.90
d
-methyl
d
.55 0.998 0.99-8.91
The value of k
0
was much lower than that of k
N
[N] in
d
.85 0.999 0.99-9.90
7
eq 2, except for the slow reactions of the three less basic
pyridines, where the pyridinolysis term in eq 2 was also
a
In 44 wt % ethanol-water, at 25.0 °C, ionic strength 0.2 M
b
c
(KCl). Fraction of free amine. Concentration of total amine (free
small. The values of k
reactions carried out under the presence of external
phosphate) buffer. The values of k for all reactions were
obtained as the slopes of linear plots of kobsd against [N].
The k values were found to be pH-independent; these
0
varied with pH only for the
d
amine plus its conjugate acid). Under the presence of phosphate
buffer 0.01 M.
(
N
products since the experimental pH values were larger than
those of the pK of 4-nitrobenzenethiol and the corresponding
a
N
a
are shown in Table 4, together with the pK values of
the conjugate acids of the pyridines.
As expected, the values of k in Table 4 increase as
N
the pyridine basicity increases. As found for the reactions
of the same thiolbenzoates with secondary alicyclic
benzoic acid, respectively. The identification was realized by
comparison of the UV-vis spectra after completion of the
reactions (before oxidation of 4-nitrobenzenethiolate took place,
see below) with those of an equimolar mixture of authentic
samples of 4-nitrobenzenethiolate and the corresponding ben-
zoate anion, under the same reaction conditions.
amines, the k
N
values also increase as the electron-
An intermediate in the reactions under scrutiny presumably
is the amidic cation, 1-(4-X-benzoyl)-substituted pyridinium,
as judged by a small increase and later decrease of absorbance
at 300-302 nm, observed in the reactions of the three more
basic pyridines. For the reactions of the other pyridines, a good
isosbestic point at ca. 300 nm was noted, indicating that in
these cases the amidic cation is highly unstable toward
hydrolysis. This is reasonable since the less basic pyridines
are better leaving groups in the hydrolysis reactions. Even the
amidium cations derived from very basic pyridines are rapidly
hydrolyzed. 1-Benzoyl-4-(dimethylamino)pyridinium could not
be detected in the reaction of this pyridine with 4-nitrophenyl
attracting effect of the 4-substituent on the acyl group
of the substrate increases.3
With the values of pK
a
for the pyridinium ions and
those of k for the reactions subjected to the present
N
work, the Brønsted-type plots were obtained. These are
shown in Figure 1 for the pyridinolysis of substrates 1
and 3. Figure 2 exhibits the Brønsted-type plot for the
reactions of thiolbenzoate 2.
The curved line shown in the Brønsted-type plot for
the reactions of thiolbenzoate 1 in Figure 1 was calcu-
lated by a semiempirical equation (eq 3) based on the
9
benzoate in water. Nevertheless, the acetylpyridinium deriva-
tives are more stable toward hydrolysis, as indicated by the
fact that 1-acetyl-4-(dimethylamino)pyridinium has been ob-
served spectrophotometrically in the reaction of this pyridine
with acetic anhydride and 2,4-dinitrophenyl thiolacetate in
hypothesis of a zwitterionic tetrahedral intermediate on
5,12,13
the reaction pathway.
Curved Brønsted plots have
been attributed to a change in the rate-limiting step, from
decomposition of the intermediate to its formation as the
1
0
aqueous solution. On the other hand, these types of cations
are stable in nonaqueous solvents, since 1-(4-nitrobenzoyl)-
pyridinium was isolated in the pyridinolysis of S-4-bromophen-
5
,12,13
pK
a
of the conjugate acid of the amine increases.
The
curved Brønsted line for the pyridinolysis of thiolbenzoate
2
c
1 was calculated by nonlinear least-squares fitting,
yl 4-nitrotiobenzoate in acetonitrile.
0
For the title reactions a slow decrease of absorbance at 420
nm was observed after a long reaction time. This can be
attributed to oxidation of 4-nitrobenzenethiolate anion to form
bis(4-nitrophenyl) disulfide. The identification of this product
was carried out by comparison of the UV-vis spectra of some
of the reactions at very long reaction times with an authentic
sample of bis(4-nitrophenyl) disulfide under the same experi-
mental conditions. To avoid the interference of this oxidation
reaction with the kinetic measurements for the formation of
yielding the following parameters: log k ) 0.54 ( 0.05,
N
0
a
pK ) 9.7 ( 0.1, â
1
) 0.20 ( 0.1, and â ) 0.94 ( 0.05.
2
0
0
a
The parameters k and pK are those corresponding to
N
the center of the Brønsted curvature and â
1
2
and â are
5
,12
the slopes at high and low pK values, respectively.
a
(11) (a) Hupe, D. J .; J encks, W. P. J . Am. Chem. Soc. 1977, 99, 451.
(b) Castro, E. A.; Ureta, C. J . Org. Chem. 1989, 54, 2153. (c) Antonello,
S.; Daasbjerg, K.; J ensen, H.; Taddei, F.; Maran, F. J . Am. Chem. Soc.
4
-nitrobenzenethiolate, two different methods (for fast and
2
003, 125, 14905. (d) Shah, S. T. A.; Khan, K. M.; Fecker, M.; Voelter,
W. Tetrahedron Lett. 2003, 44, 6789.
12) Castro, E. A.; Moodie, R. B. J . Chem. Soc., Chem. Commun.
1973, 828.
(13) (a) Gresser, M. J .; J encks, W. P. J . Am. Chem. Soc. 1977, 99,
slow reactions) were employed to obtain the values of kobsd, as
(
(
9) Um, I.-H.; Baek, M.-H.; Han, H.-J . Bull. Korean Chem. Soc. 2003,
4, 1245.
10) (a) Castro, C.; Castro, E. A. J . Org. Chem. 1981, 46, 2939. (b)
Castro, E. A.; Ureta, C. J . Chem. Soc., Perkin Trans. 2 1991, 63.
2
(
6963. (b) Gresser, M. J .; J encks, W. P. J . Am. Chem. Soc. 1977, 99,
6970.
J . Org. Chem, Vol. 69, No. 16, 2004 5401

Castro et al.
TABLE 4. Va lu es of p Ka for Con ju ga te Acid s of P yr id in es a n d kN for th e P yr id in olysis of S-4-Nitr op h en yl Th ioben zoa te
(
1), S-4-Nitr op h en yl 4-Ch lor oth ioben zoa te (2), a n d S-4-Nitr op h en yl 4-Nitr oth ioben zoa te (3)a
kN (s- M-1
1
)
pyridine
substituent
pKa
1
2
3
3
4
4
4
3
4
3
,4-diamino
-dimethylamino
-amino
-amino-3-bromo
,4-dimethyl
-methyl
9.45
9.14
8.98
6.90
5.68
5.35
4.92
2.3 ( 0.1
5.0 ( 0.2
26 ( 2
14 ( 1
10 ( 1
1.8 ( 0.1
2.4 ( 0.1
0.86 ( 0.07
1.7 ( 0.1
-2
-4
-4
-4
-2
-3
-4
-4
-2
-3
-3
-4
(2.3 ( 0.1) × 10
(9.4 ( 0.5) × 10
(5.4 ( 0.2) × 10
2.4 ( 0.1) × 10
(2.6 ( 0.1) × 10
(1.1 ( 0.1) × 10
(6.6 ( 0.3) × 10
(3.0 ( 0.2) × 10
(7.2 ( 0.5) × 10
(4.7 ( 0.3) × 10
(2.3 ( 0.1) × 10
(8.4 ( 0.8) × 10
-methyl
a
Both the pKa and kN values were determined in 44 wt % ethanol-water, at 25.0 °C, ionic strength 0.2 M (KCl).
F IGURE 1. Brønsted-type plots obtained in the pyridinolysis
of 1 (O) and 3 (b) in 44 wt % ethanol-water, at 25.0 °C, and
an ionic strength of 0.2 M (KCl).
F IGURE 2. Brønsted-type plot obtained in the pyridinolysis
of 2 in 44 wt % ethanol-water, at 25.0 °C, and an ionic
strength of 0.2 M (KCl).
kN
1 + a
2
0
a
log
) â (pK - pK ) - log
2
a
0
N
( k
)
[
]
0
a
The amidium cation 4 in Scheme 1 undergoes hydroly-
sis to yield the substituted benzoate anion and the
corresponding pyridine, according to the analysis of
products. On the other hand, 4-nitrobenzenethiolate (5)
is very slowly oxidized under the reaction conditions to
yield the corresponding disulfide (see Experimental Sec-
tion).
log a ) (â - â )(pK - pK )
(3)
2
1
a
According to the value of the slopes of the linear
Brønsted-type plots for the pyridinolysis of thiolbenzoates
and 3 (â ) 0.94 ( 0.05 and 1.0 ( 0.1, respectively), the
most likely mechanism for these reaction is that de-
2
scribed in Scheme 1, where breakdown of the tetrahedral
(
intermediate (T ) to products (k
2
step) is the rate-
In these cases, the first step
a
Table 5 shows the values of pK for the center of the
determining step.1
is at equilibrium.
,3-5,10,12,13
0
a
Brønsted curvature (pK ) for the reactions of secondary
alicyclic (SA) amines3 and pyridines (this work) with
Scheme 1 can also describe the mechanism of the
pyridinolysis of thiolbenzoate 1. In this case the k step
is rate-determining for all of the pyridines (pK < 9.7).
For (hypothetical) pyridines of pK values similar to those
ca. 9.7), k-1 is similar to k and there is no
thiolbenzoates 1-3.
As seen in Table 5, the increase of the pK value with
0
a
2
a
increasing electron withdrawal of the substituent on the
acyl group of the thiolbenzoate, shown by the SA ami-
nolysis, is also present in the pyridinolysis. This trend
is also in agreement with that found by Gresser and
J encks in the reactions of quinuclidines with diaryl
carbonates.1 This effect can be explained by the follow-
ing way: as the substituent in the nonleaving group
becomes more electron-withdrawing the central carbon
a
0
a
of pK (pK
a
2
1
,5,13
clear rate-determining step.
The experimental points of log k
N
a
vs pK for the
3b
pyridinolysis of thiolbenzoate 2 can also be accom-
modated by the semiempirical equation based on the
hypothesis of the tetrahedral intermediate. In this case,
0
N
0
a
(
the best fitting parameters are log k ) 0.94 ( 0.07, pK
of the zwitterionic tetrahedral intermediate (T ) becomes
more positive and favors the push provided by the sulfur
)
9.9 ( 0.1, â
1
) 0.20 ( 0.1, and â
2
) 0.97 ( 0.07.
5
402 J . Org. Chem., Vol. 69, No. 16, 2004

Pyridinolysis of Thiobenzoates
SCHEME 1
TABLE 5. Va lu es of p Ka for th e Cen ter of th e Br øn sted
trinitrophenyl thiolacetates show curved biphasic Brøn-
0
0
Cu r va tu r e (p K ) for th e Rea ction s of Secon d a r y Alicyclic
a
sted-type plots; the pK values are larger for the reac-
a
a
(
SA) Am in es a n d P yr id in es w ith Th iolben zoa tes 1-3
10b
tions of SA amines. According to eq 4, this indicates a
thiolbenzoate
SA aminesb
pyridinesc
greater nucleofugality for SA amines. The SA aminolyses
of S-(2,4-dinitrophenyl) and S-(2,4,6-trinitrophenyl) ethyl
thiocarbonates in water are concerted,16 whereas the
pyridinolyses of the same substrates in the same solvent
1
2
3
10.0
10.4
>11
9.7
g9.9
>10
(
17
a
are stepwise, through the intermediate T . Similarly,
the SA aminolysis of 2,4-dinitrophenyl and 2,4,6-tri-
nitrophenyl methyl carbonates,18 both in water, are
concerted, in contrast to the pyridinolysis of the same
In 44 wt % ethanol-water, at 25.0 °C, ionic strength 0.2 M
b
c
(
KCl). Data taken from ref 3. Data from this work
(
atom in the leaving group of T to expel the amine. The
1
9,20
(
compounds in water, which are stepwise.
The higher
amino moiety in T cannot exert a push to expel the
(
instability of the intermediate T formed with the SA
amines, compared with that formed with isobasic pyr-
idines, has been attributed to the greater nucleofugality
leaving group since it does not possess an electron pair.
As a result, electron withdrawal from the nonleaving
(
group favors expulsion of the amine from T relative to
(
of SA amines from T relative to that of isobasic
that of the leaving group. According to the hypothesis of
the tetrahedral intermediate, an equation (eq 4) can be
derived that relates the ratio of rate constants for amine
pyridines.1
0b,17-20
The pyridinolyses of 2,4-dinitrophenyl 4-X-substituted
(
0
benzoates (X ) H, Cl, CN, and NO ) in 44 wt % ethanol-
2
and leaving group expulsion from T (k-1/k
2
) to the pKa
are little sensitive
eq 4 predicts that as the
4d
value.14 Since the values of â
water show a curved Brønsted-type plot for X ) H and
2
and â
1
linear plots for the other benzoates.4 The center of the
a-c
to the substrate nature,1
,3-5,13,15
Brønsted plot curvature for X ) H is located at pK ) p
a
k
2
-1/k ratio becomes larger (as amine expulsion is more
0
a
0
K ) 9.5 and the linear plots exhibit slopes of ca. 0.9,
which indicates that pKa is larger than 10 for these
reactions. This trend (pK increases as X becomes more
favored), the pK value also increases.
a
0
0
a
k-1
0
a
log
) (â - â )(pK - pK )
(4)
electron-withdrawing) is similar to that found for the
2
1
a
(
)
k2
3
reactions of thiolbenzoates with SA amines and pyr-
idines (this work) in the same solvent.
0
a
Table 5 shows that the value of pK for the reactions
0
a
The fact that the pK value obtained in the pyridi-
of a given thiolbenzoate with SA amines is larger than
that for the reactions of the same substrate with pyr-
idines. This can also be accounted for through eq 4, as
follows. It is known that SA amines are better nucle-
nolysis of 2,4-dinitrophenyl benzoate in 44 wt % ethanol-
4d
water is smaller than that found in the reactions of
thiolbenzoate 1 with the same amines in the same solvent
(
this work) can be attributed to the superior leaving
(
ofuges from a T intermediate than isobasic pyridines
(
ability from T of 2,4-dinitrophenoxide compared with
-nitrobenzenethiolate. This stems from two causes: (i)
(
see below). This means that the value of k-1 is larger
for an SA amine compared with an isobasic pyridine. On
the other hand, the value of k is not affected by the
amine basicity or nature.1 Therefore, the ratio k-1/k
for a given SA amine is larger than that for an isobasic
pyridine. Since the values of â and â are little sensitive
eq 4 shows that the pK value
4
21
the lower pK
compared to that of 4-nitrobenzenethiol (pK
water), and (ii) the fact that phenoxide anions are a
a
of 2,4-dinitrophenol (pK
a
4.1 in water)
4.6 in
2
a
3a
2
11b
better leaving group than isobasic benzenethiolate anions
2
1
to the amine nature,1
,13,15
0
a
(
16) Castro, E. A.; Iba n˜ ez, F.; Salas, M.; Santos, J . G. J . Org. Chem.
991, 56, 4819. Castro, E. A.; Salas, M.; Santos, J . G. J . Org. Chem.
1994, 59, 30.
should be larger for the reactions with SA amines, as
observed.
1
(
17) Castro, E. A.; Pizarro, M. I.; Santos, J . G. J . Org. Chem. 1996,
The better nucleofugality of SA amines compared to
6
1, 5982.
(
that of isobasic pyridines from the intermediate T has
(18) Castro, E. A.; Aliaga, M.; Campodonico, P.; Santos, J . G. J . Org.
been demonstrated in many studies. The reactions of SA
amines and pyridines with 2,4-dinitrophenyl and 2,4,6-
Chem. 2002, 67, 8911. Castro, E. A.; Cubillos, M.; Santos, J . G. J . Org.
Chem. 2001, 66, 6000.
(19) Castro, E. A.; Gil, F. J . J . Am. Chem. Soc. 1977, 99, 7611.
20) Castro, E. A.; Iba n˜ ez, F.; Lagos, S.; Schick, M.; Santos, J . G. J .
(
(
14) Castro, E. A.; Araneda, C. A.; Santos, J . G. J . Org. Chem. 1997,
2, 126.
15) J encks, W. P.; Gilchrist, M. J . Am. Chem. Soc. 1968, 90, 2622.
Org. Chem. 1992, 57, 2691.
(21) Albert, A.; Serjeant, E. P. The Determination of Ionization
Constants; Chapman and Hall: London, 1971; p 87.
6
(
J . Org. Chem, Vol. 69, No. 16, 2004 5403

Castro et al.
from a tetrahedral intermediate.22 Therefore, the value
of k for 2,4-dinitrophenoxide should be larger than that
for 4-nitrobenzenethiolate. This means a lower k-1/k
tion of T conferred by both solvents should be similar.
On the other hand, anilines are only little better leaving
groups than isobasic pyridines24 and should confer to T
only a slightly greater destabilization relative to isobasic
pyridines. Furthermore, the reactions of SA amines with
(
13b
2
(
2
ratio for a given pyridine in the reactions with the
benzoate substrate and, according to eq 4, a lower value
0
a
3
of pK , as observed.
the above substrates in aqueous ethanol are stepwise,
The pyridinolysis of aryl 4-nitrothiolbenzoates in ac-
etonitrile exhibit biphasic Brønsted-type plots that have
been explained by stepwise mechanisms.2 It is known
that the zwitterionic tetrahedral intermediate is less
stable in less polar solvents.1 The fact that the pyridi-
nolysis of thiolbenzoate 3 in aqueous ethanol is stepwise
despite the fact that SA amines are better nucleofuges
than isobasic anilines.24
c
For the reactions of S-aryl 4-X-thiobenzoates (X ) H,
2
Cl, NO ) with benzylamines in acetonitrile linear Brøn-
3b
2b
sted-type plots were found. These plots were drawn
using the pK values in water but not in acetonitrile. The
pK value in water for the most basic amine employed
a
(this work) and the same reactions in acetonitrile are
a
stepwise indicates that the change of solvent from
aqueous ethanol to acetonitrile does not destabilize the
intermediate sufficiently as to change the mechanism
from stepwise to concerted.
2b
was 9.5. Correction of this value for the change of
solvent from water to acetonitrile, through an empirical
2c
equation, gives a pK
a
value of 17.9. Therefore, the center
of the Brønsted curvature for these reactions in aceto-
The pyridinolysis of 4-nitrophenyl benzoate in water
shows a linear Brønsted-type plot with slope 1.1, consis-
tent with a stepwise mechanism where breakdown of the
0
a
nitrile should be pK > 17.9, a value that is much larger
than that obtained for the pyridinolysis of thiolbenzoate
0
a
1
in aqueous ethanol (pK ) 9.7, this work). This can be
(
9
intermediate T is rate-determining. This result is in
0
accounted for by the facts that (i) benzylamines seem to
be better nucleofuges than isobasic pyridines, (ii) the
value of k-1 is much larger in acetonitrile than in water,
and (iii) the value of k is not much affected by the change
These three factors should make the k-1
agreement with the curved Brønsted plot, with pK )
.7, obtained for the pyridinolysis of thiolbenzoate 1 in
a
25
9
2
6
aqueous ethanol (this work), on the following grounds.
The value of k-1 should not vary significantly in these
two solvents because both are highly protic and polar.
2
of solvent.1
k
3b,26
/
0
2
ratio and also the pK value (see eq 4) larger for the
a
On the other hand, the value of k
2
should be larger for
reactions in acetonitrile,
the reaction of thiolbenzoate 1 since 4-nitrobenzenethi-
1
1b
olate (pK
nucleofuge from T than 4-nitrophenoxide (pK
conjugate acid ) 7.1)21 from the corresponding intermedi-
ate. This is also reflected in the larger k values shown
by the reactions of the thiolbenzoate. These two effects
should render the ratio k-1/k larger for the reaction of
-nitrophenyl benzoate in water (relative to that for the
thiolbenzoate in aqueous ethanol), and according to eq
a
of its conjugate acid ) 4.6) should be a better
(
of its
Ack n ow led gm en t. We thank FONDECYT of Chile
(project 1020524) for financial assistance to this work.
R.A. thanks “Direcci o´ n de Investigaci o´ n y Postgrado de
Pontificia Universidad Cat o´ lica de Chile” (DIPUC) for
a Doctoral scholarship.
a
N
2
4
Su p p or tin g In for m a tion Ava ila ble: Tables S1-S3 con-
taining the individual values of kobsd, amine concentrations,
pH and free amine fractions, and other experimental condi-
tions, for the pyridinolysis of thiolbenzoates 1-3. This material
is available free of charge via the Internet at http://pubs.acs.org.
0
a
4
, the pK value should be larger for the reaction of the
benzoate, as found.
The reactions of S-4-nitrophenyl 4-X-thiobenzoates (X
)
2
H, Cl, NO ) with anilines in methanol were found to
be concerted, namely, the hypothetical intermediate T
2
a
(
J O049260F
formed in these reactions is either very unstable or
nonexistent.23 This is in contrast to the results of the
present work since the two solvents involved, methanol
and aqueous ethanol, are very similar and the stabiliza-
(
24) Castro, E. A.; Leandro, L.; Millan, P.; Santos, J . G. J . Org.
Chem. 1999, 64, 1953.
25) The relative nucleofugality of benzylamines and pyridines has
(
not been quantified, to our knowledge. Nevertheless, it is known that
primary amines such as anilines are better nucleofuges than isobasic
pyridines, so it is likely that benzylamines would be better leaving
groups than isobasic pyridines.
(26) Um, I.-H.; Seok, J .-A.; Kim, H.-T.; Bae, S.-K. J . Org. Chem.
2003, 68, 7742.
2
4
(
22) J ensen, J . L.; J encks, W. P. J . Am. Chem. Soc. 1979, 101, 1476.
Douglas, K. T.; Alborz, M. J . Chem. Soc., Chem. Commun. 1981, 551.
23) J encks, W. P. Chem. Soc. Rev. 1981, 10, 345. Williams, A. Chem.
Soc. Rev. 1994, 23, 93.
(
5
404 J . Org. Chem., Vol. 69, No. 16, 2004