Kinetics and mechanism of the pyridinolysis of diaryl carbonates

Article abstract of DOI:10.1002/poc.1399

The reactions of 4-methylphenyl and 4-chlorophenyf 4-nitrophenyl carbonates (1 and 2, respectively), phenyl, 4-methylphenyl, 4-chlorophenyl, and 4-nitrophenyl 2,4-dinitrophenyl carbonates (3, 4, S, and 6, respectively), and bis(2,4-dinitrophenyl) carbonate (7) with a series of pyridines are studied kinetically at 25.0°C in 44 wt% ethanol-water and an ionic strength of 0.2 M (KCl). The reactions are followed spectrophotometrically and under excess amine pseudo-first-order rate coefficients (k_obs) are found. For all these reactions, plots of k_obs versus free amine concentration at constant pH are linear, the slope (k_N) being independent of pH. The Bronsted-type plots (log k_N vs. pK_a of the conjugate acids of the pyridines) are all biphasic (linear portions at high and low pK _a and a curvature in between). These plots are in accordance with a stepwise mechanism, through a zwitterionic tetrahedral intermediate (T ^±), and a change in the rate-determining step from formation of T^± to its breakdown to products, as the pyridine basicity decreases. Also studied are the effects of the leaving, non-leaving, and electrophilic groups of the substrate, and of the amine nature, on the pK _a⁰ value (value at the center of curvature of the Bronsted-type plots). Copyright

Full text of DOI:10.1002/poc.1399

Research Article
Received: 1 April 2008,
Revised: 30 April 2008,
Accepted: 30 April 2008,
Published online in Wiley InterScience: 22 May 2008
(www.interscience.wiley.com) DOI 10.1002/poc.1399
Kinetics and mechanism of the pyridinolysis
of diaryl carbonates
a
a
a
a
˜
Enrique A. Castro *, Magdalena Acuna , Claudia Soto , Carolina Trujillo ,
a
a
´
´
´
Barbara Vasquez and Jose G. Santos *
The reactions of 4-methylphenyl and 4-chlorophenyl 4-nitrophenyl carbonates (1 and 2, respectively), phenyl,
4-methylphenyl, 4-chlorophenyl, and 4-nitrophenyl 2,4-dinitrophenyl carbonates (3, 4, 5, and 6, respectively), and
bis(2,4-dinitrophenyl) carbonate (7) with a series of pyridines are studied kinetically at 25.0 8C in 44 wt% ethanol–
water and an ionic strength of 0.2 M (KCl). The reactions are followed spectrophotometrically and under excess amine
pseudo-first-order rate coefficients (kobs) are found. For all these reactions, plots of kobs versus free amine
concentration at constant pH are linear, the slope (kN) being independent of pH. The Brønsted-type plots (log kN
vs. pKa of the conjugate acids of the pyridines) are all biphasic (linear portions at high and low pKa and a curvature in
between). These plots are in accordance with a stepwise mechanism, through a zwitterionic tetrahedral intermediate
(Tꢀ), and a change in the rate-determining step from formation of Tꢀ to its breakdown to products, as the pyridine
basicity decreases. Also studied are the effects of the leaving, non-leaving, and electrophilic groups of the substrate,
and of the amine nature, on the pKa0 value (value at the center of curvature of the Brønsted-type plots). Copyright ß
2008 John Wiley & Sons, Ltd.
Supplementary electronic material for this paper is available in Wiley InterScience at http://www.mrw.interscience.wiley.com/
suppmat/0894-3230/suppmat/
Keywords: kinetics; mechanism; pyridinolysis; diaryl carbonates; Brønsted plots
The aminolysis reactions of diaryl carbonates have also been
subjected to kinetic studies,[10–19] although not as extensively as
INTRODUCTION
The kinetics and mechanisms of the aminolysis of alkyl aryl
carbonates are well documented.[1–9] The pyridinolysis reactions
of 4-nitrophenyl methyl and phenyl methyl carbonates in
aqueous solution show linear Brønsted-type plots with slopes
ca. 1, which were explained by stepwise mechanisms, through a
zwitterionic tetrahedral intermediate (Tꢀ), where its breakdown
to products is the rate-determining step.[1,2] The same aminolysis
of 2,4-dinitrophenyl methyl and 2,4,6-trinitrophenyl methyl
carbonates in aqueous solution exhibit biphasic (two linear
portions and a curvature in between) Brønsted-type plots, with
slopes b1 ¼ 0.2 (high pKa) and b2 ¼ 0.8–0.9 (low pKa).[3,4] These
plots were attributed to stepwise mechanisms and a change in
the rate-limiting step, from Tꢀ breakdown to its formation as the
pyridine becomes more basic.[3,4]
The reactions of benzylamines with ethyl X-phenyl carbonates
(X ¼ MeO, Cl, CN, NO2) in acetonitrile show linear Brønsted-type
plots, with slopes b ¼ 1.7–2.3.[6] From these slope values and
other considerations it was deduced that these reactions obey a
stepwise mechanism, with Tꢀ breakdown to products as rate
determining.[6] It is noteworthy that these slopes are larger than
those found in aqueous solution (b ¼ 0.8–1.1) when breakdown
of Tꢀ is the rate-determining step, which could be attributed to
the fact that the pKa values plotted were those in water and not
those in acetonitrile.
those of alkyl aryl carbonates. The reactions of quinuclidines
with 3- and 4-nitrophenyl phenyl carbonates and 3,4- and
2,4-dinitrophenyl phenyl carbonates were found to be stepwise,
as judged by the biphasic Brønsted-type plots obtained.[10,11] The
reactions of SA amines,[12,13] anilines,[14] and quinuclidines[13]
with 4-methylphenyl and 4-chlorophenyl 4-nitrophenyl carbon-
ates (1 and 2, respectively) were also found to be stepwise on the
basis of Brønsted-type plots, either biphasic or linear with large
slopes.
In contrast, the reactions of SA amines[9] and quinuclidines[15]
with phenyl 2,4-dinitrophenyl carbonate (3), as well as those of
SA amines,[12,16] quinuclidines,[15] and anilines,[16] with 4-
methylphenyl and 4-chlorophenyl 2,4-dinitrophenyl carbonates
(4 and 5, respectively) were claimed to be concerted. Similarly, the
reactions of SA amines with 4-nitrophenyl 2,4-dinitrophenyl
(6) and bis(2,4-dinitrophenyl) (7) carbonates were also found to be
concerted.[17] For all these concerted reactions, linear Brønsted-
type plots with slopes b ¼ 0.4–0.7 were obtained.
On the other hand, the reactions of primary alkyl amines with
4-nitrophenyl phenyl carbonate show a biphasic Brønsted-type
plot with slopes b1 ¼ 0.27 (high pKa) and b2 ¼ 0.99 (low pKa),
In contrast, the reactions of secondary alicyclic (SA) amines
with 2,4-dinitrophenyl methyl and 2,4,6-trinitrophenyl methyl
carbonates in aqueous solution are concerted (single step with
no Tꢀ intermediate), as shown by the linear Brønsted-type plots
obtained, with slopes 0.48 and 0.36, respectively.[8,9]
*
Correspondence to: J. G. Santos, Facultad de Qu´ımica, Pontificia Universidad
´
Catolica de Chile, Casilla 306, Santiago 6094411, Chile.
˜
´
E. A. Castro, M. Acuna, C. Soto, C. Trujillo, B. Vasquez, J. G. Santos
´
Facultad de Qu´ımica, Pontificia Universidad Catolica de Chile, Casilla 306,
Santiago 6094411, Chile
a
J. Phys. Org. Chem. 2008, 21 816–822
Copyright ß 2008 John Wiley & Sons, Ltd.

PYRIDINOLYSIS OF DIARYL CARBONATES
which was attributed to a stepwise mechanism and a change in
the rate-limiting step.[18,19]
As seen above, there are still some uncertainties regarding the
reason why some reactions are stepwise and others concerted.
Furthermore, the effects of the leaving, non-leaving, and
electrophilic (CO vs. CS) groups of the substrate and the amine
nature on the type of mechanism are not clearly understood. In
order to further extend our investigations on the kinetics and
mechanisms of the aminolysis of diaryl carbonates in this work
we report our kinetic results for the reactions of seven diaryl
carbonates (1–7) with a series of seven pyridines (as shown in
structures below). Our main goal is to assess the effect of the
leaving, non-leaving, and electrophilic groups of the substrate,
and of the amine nature, on the kinetics and mechanism. This will
be achieved by comparison of the title reactions between them
and with the aminolysis of similar carbonates. It is of particular
interest to evaluate the influence of the above groups on the pKa
value for the center of the Brønsted curvature (pKa0). The pKa0
value is important because it is a measure of the leaving abilities
from the tetrahedral intermediate of both the amine and the
leaving group of the substrate (as shown below).
RESULTS AND DISCUSSION
The reactions were studied in 44 wt% ethanol–water, at 25.0 8C
and an ionic strength of 0.2 (KCl). Under excess of amine over
the substrate, pseudo-first-order rate coefficients (kobs) were
obtained for all reactions. The experimental conditions of the
reactions and the values of kobs are shown in Tables 1–7 in
Supplementary Materials.
The kinetic law obtained under the reaction conditions is
described by Eqn (1), where P is 4-nitrophenoxide or
2,4-dinitrophenoxide anion and S is the substrate:
d½Pꢁ
¼ kobs½Sꢁ
dt
(1)
Plots of kobs against concentration of free pyridine at constant
pH were linear, in accordance with Eqn (2), where k0 and kN are
the rate coefficients for solvolysis and pyridinolysis of the
substrates, respectively. The values of k0 and kN were obtained as
the intercept and slope, respectively, of plots of Eqn (2), and were
pH-independent:
kobs ¼ k0 þ kN ½free pyridineꢁ
(2)
Table 1 shows the values of pKa of the pyridinium ions and
those of kN for the reactions under study. With these values the
Brønsted-type plots of Figs 1–3 were obtained.
J. Phys. Org. Chem. 2008, 21 816–822
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

E. A. CASTRO ET AL.
6
4
2
0
3
0
-3
3
4
2
0
0
-3
4
6
8
10
12
4
8
12
pK
a
pKa
Figure 1. Brønsted-type plots obtained for the reactions of pyridines
with 4-methylphenyl 4-nitrophenyl carbonate (1, *) and 4-chlorophenyl
4-nitrophenyl carbonate (2, *), in 44 wt% ethanol–water, at 25.0 8C and
an ionic strength of 0.2 M
Figure 3. Brønsted-type plots obtained for the reactions of pyridines
with 4-nitrophenyl 2,4-dinitrophenyl carbonate (6, *) and bis(2,4-
dinitrophenyl) carbonate (7, *), in 44 wt% ethanol–water, at 25.0 8C
and an ionic strength of 0.2 M
Z
O
4
2
0
X
O
C
O
NO2
+
Py
Y
k1
k-1
-2
4
Z
O-
C
2
0
X
O
O
NO2
Py +
Y
T
-2
4
k2
2
0
Z
O
C
+
X
O
-O
NO2
-2
3
6
9
12
Py +
Y
pKa
Scheme 1.
Figure 2. Brønsted-type plots obtained for the reactions of pyridines
with phenyl 2,4-dinitrophenyl carbonate (3, *), 4-methylphenyl
2,4-dinitrophenyl carbonate (4, *), and 4-chlorophenyl 2,4-dinitrophenyl
carbonate (5, &), in 44 wt% ethanol–water, at 25.0 8C and an ionic
strength of 0.2 M
Brønsted slopes at high and low pKa, respectively, and kN0 and pKa0,
which are the corresponding values at the center of the
curvature. The Brønsted curves were calculated by means of Eqn
(3), using the parameters (obtained by nonlinear least-squares
fitting) summarized in Table 2:
The biphasic Brønsted plots found for the pyridinolysis
of carbonates 1–7 can be explained by the existence of a
zwitterionic tetrahedral intermediate (Tꢀ) on the reaction
pathway and a change in the rate-determining step, from
breakdown to products of Tꢀ (step k2 in Scheme 1) to formation
of Tꢀ (step k1 in Scheme 1), as the amine becomes more
basic.[3,10,11]
The lines in the Brønsted plots of Figs 1–3 were calculated by
means of a semiempirical equation, Eqn (3), based on the
existence of the intermediate Tꢀ on the reaction pathway.[3,10,11]
This equation contains four parameters: b1 and b2, which are the
log ðkN=kN0 Þ ¼ b2ðpKa ꢂ pKa0Þ ꢂ logð1 þ a=2Þ
(3)
log a ¼ ðb2 ꢂ b1Þ ðpKa ꢂ pKa0Þ
The relatively large errors of b1 and pKa0 for the pyridinolysis of
carbonates 1 and 2 can be attributed to the fact that for these
reactions there is only one amine (4-oxypyridine) that clearly
belongs to the region where the formation of the intermediate Tꢀ
(k1 step in Scheme 1) is rate limiting. Namely, for these reactions
there is only one amine whose pKa is much greater than pKa0.
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 816–822

PYRIDINOLYSIS OF DIARYL CARBONATES
Table 2. Parameters in Eqn (3) obtained for the pyridinolysis of diaryl carbonates 1–7
Substrate
log kN0
pKa0
b1
b2
R2
1
2
3
4
5
6
7
2.43 ꢀ 1.17
2.7 ꢀ 1.1
2.8 ꢀ 0.2
2.7 ꢀ 0.1
3.2 ꢀ 0.1
3.6 ꢀ 0.4
4.1 ꢀ 0.4
9.6 ꢀ 1.4
9.5 ꢀ 1.2
8.0 ꢀ 0.3
8.3 ꢀ 0.2
8.3 ꢀ 0.6
7.6 ꢀ 0.8
7.2 ꢀ 0.6
0.42 ꢀ 0.30
0.45 ꢀ 0.25
0.24 ꢀ 0.04
0.23 ꢀ 0.03
0.23 ꢀ 0.08
0.27 ꢀ 0.11
0.22 ꢀ 0.09
1.1 ꢀ 0.1
0.9976
0.9980
0.9996
0.9998
0.9998
0.996
1.15 ꢀ 0.09
0.83 ꢀ 0.03
0.80 ꢀ 0.02
0.82 ꢀ 0.05
0.93 ꢀ 0.07
0.97 ꢀ 0.07
0.997
Taking into account the shape of the Brønsted plots obtained
and the values of slopes, the most likely mechanism for the
reactions under scrutiny is the stepwise process shown in
Scheme 1. In this Scheme X, Y, and Z were defined above and Py
represents a pyridine.
Similar shapes and slopes to those of the Brønsted plots in
Figs 1–3 were found in the pyridinolysis of 2,4-dinitrophenyl
acetate,[2] 2,4-dinitrophenyl, and 2,4,6-trinitrophenyl methyl carb-
onates,[3,4] and in the SA aminolysis of 4-nitrophenyl methyl
carbonate,[9] and carbonates 1 and 2.[12,13] The Brønsted plots
obtained in the present work are also in accordance with those
found in the quinuclidinolysis of diaryl carbonates[10,11] and in
the reactions of primary amines with 4-nitrophenyl phenyl
carbonate.[18,19] All these reactions were found to be governed by
stepwise mechanisms, through a zwitterionic tetrahedral inter-
mediate.
weakly basic).[10,11] The change of 4-nitrophenoxy by
2,4-dinitrophenoxy in the tetrahedral intermediate increases k2
due to the much lower basicity of 2,4-dinitrphenoxide compared
with 4-nitrophenoxide (the pKa for the corresponding phenols in
water at 258C are 4.1 and 7.1, respectively),[20] which makes the
former a much better nucleofuge.
Also in accordance with the above discussion, and another
proof that the pyridinolysis of carbonates 1–7 are stepwise, is the
fact that the pKa0 values found (Table 2) for the dinitro derivatives
(about 8.3) are lower than those of the corresponding mononitro
carbonates (about 9.5). This is because, according to the
tetrahedral intermediate hypothesis, an equation can be derived,
Eqn (4),[9] which shows that a larger k2/k ratio means a lower
ꢂ1
pKa0 value:
log kꢂ1=k2 ¼ ðb2 ꢂ b1Þ ðpKa0 ꢂ pKaÞ
(4)
According to the k1 and kN values for the reactions of
4-oxypyridine and pyridine, respectively, with carbonates 1, 2, 4,
and 5 and the pKa values of 4-nitrophenol and 2,4-dinitrophenol
(as shown above), the Brønsted slopes for the leaving group (blg)
can be estimated. The resulting values are blg ¼ ꢂ0.1 for k1 (Tꢀ
formation as rate limiting) and ꢂ0.94 for kN (Tꢀ breakdown as
rate determining). These values are also in agreement with those
usually found for the stepwise aminolysis of carbonates and
esters.[10,11,21–26]
Effect of the leaving group
The aim of this section is to assess how the leaving group
(nucleofuge) of the substrate affects the rate constants involved
in the mechanism (Scheme 1) and how this affects the pKa0 value.
For the reactions of 4-oxypiridine with all the carbonates studied
in this work the rate-determining step is the formation of the
zwitterionic tetrahedral intermediate (Tꢀ) and, therefore, kN ¼ k1
in Scheme 1. From a comparison of the k1 values between
carbonates 1 and 4 and between 2 and 5 (Table 1), where there is
Effect of the amine nature
a
change in the leaving group from 4-nitrophenoxy to
2,4-dinitrophenoxy for the same non-leaving group, it can be
observed that the k1 values for the dinitro derivatives (4 and 5)
are approximately two-fold larger than those for the correspond-
ing mononitro derivatives (1 and 2). This result is obviously due to
the presence of a second nitro group in the substrate, which
increases the electron-withdrawing ability from the leaving
group, leading to a more positive carbonyl carbon atom and
favoring, therefore, the 4-oxypyridine attack.
On the other hand, when the second step is rate determining,
kN ¼ k1k2/kꢂ1. Taking as an example the reactions of the
unsubstituted pyridine, Table 1 shows that the kN values for
the reactions of this amine with the dinitro derivatives (4 and 5)
are approximately 600–700-fold greater than those for the
corresponding mononitro derivatives (1 and 2). Taking into
account that the k1 values are only twice as large for the former
Our aim here is to evaluate the effect of the nature of the amine
(i.e., pyridines vs. SA amines) on the pKa0 value and, therefore, on
the kꢂ1/k2 ratio. The reactions of SA amines with carbonates 1
and 2 exhibit biphasic Brønsted plots, with the curvature center
at pKa ¼ pKa0 ¼ 10.5.[12,13] As seen in Table 2 the pKa0 for the
pyridinolysis of 1 and 2 are ca. 9.5. The larger pKa0 value for
the reactions of SA amines can be attributed to the faster
nucleofugality from the intermediate Tꢀ (larger kꢂ1 in Scheme 1)
of an SA amine compared with an isobasic pyridine.[27,28] Since
the k2 value should not be affected by the amine basicity or
nature,[10,11] the larger k values for the SA amines (compared
ꢂ1
with isobasic pyridines) mean a larger pKa0, according to Eqn (4).
The larger k for SA amines indicates that the intermediate Tꢀ
ꢂ1
formed with these amines is destabilized relative to that arising
from isobasic pyridines.
carbonates, it can be deduced that the ratio k2/k
approximately 300-fold greater for the dinitro derivatives. This
should be mainly due to a larger k2 for the latter carbonates since
is
The destabilization of Tꢀ with SA amines can also be confirmed
by comparison of the mechanisms found for the SA aminolysis
and pyridinolysis of carbonates 3–7. The mechanism for the
former aminolysis is concerted,[9,12,16,17] whereas that for the
ꢂ1
these leaving groups should hardly affect k (they are both
ꢂ1
J. Phys. Org. Chem. 2008, 21 816–822
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

E. A. CASTRO ET AL.
latter aminolysis is stepwise (this work). This shows that the
tetrahedral intermediate formed with pyridines is destabilized by
the change of this amine by an isobasic SA amine.
Subtracting Eqn (6) from Eqn (5), Eqn (7) is obtained. This
equation shows the dependence of pKnlg on the kꢂ1/k2 ratio for
the pyridinolysis of carbonates 3–7:
logðkꢂ1=k2Þ ¼ 4:1 ꢂ 0:6 pKN þ 0:03 pKnlg (7)
Equation (7) shows that there is only a marginal effect of the
Influence of the non-leaving group
There is some controversy regarding the influence of the
non-leaving group of the substrate on the kinetics and
mechanism of the aminolysis of esters and carbonates. In some
cases there is a pKa0 shift by the change of non-leaving group,
whereas in other instances there is no shift. The shift of pKa0 with
the change in non-leaving group is important because this is
related to the shift in the kꢂ1/k2 ratio, as shown by Eqn (4). Table 2
shows that the pKa0 values for a given leaving group do not vary
significantly with the variation of the non-leaving group. Thus,
the pKa0 values are 9.6 and 9.5 for the two mononitro derivatives
(carbonates 1 and 2) and ca. 7–8 for the substrates with dinitro
leaving groups (carbonates 3–7). These results are in line with
those found in the SA aminolysis of carbonates 1 and 2, for which
the pKa0 values are 10.5 and 10.6, respectively.[12,13] In other
reactions there is a small influence of the non-leaving group on
the pKa0 value. This is the case of the pyridinolysis of methyl
2,4-dinitrophenyl carbonate[3] and carbonate 3 (this work), where
the change of methoxy by phenoxy as non-leaving group
changes the pKa0 value from 7.8 to 8.0. Similarly, for the
pyridinolysis of ethyl and phenyl 2,4-dinitrophenyl thionocarbo-
nates the pKa0 value changes from 6.8 to 7.0 by the change of
non-leaving group basicity on the k2/k ratio. As this ratio is
ꢂ1
mainly responsible for the pKa0 value (as shown in Eqn (4)), we
conclude that, in the case of the title reactions, the non-leaving
group effect on the pKa0 value is negligible.
In contrast to the above results, a significant influence of the
non-leaving group on the pKa0 value has been found in the
pyridinolysis and SA aminolysis of S-4-nitrophenyl 4-substituted
thiobenzoates[32–34] and in the pyridinolysis of S-2,4-nitrophenyl
4-substituted thiobenzoates.[35] For the SA aminolysis, there is a
pKa0 shift from 9.7 to >10.8 for a change in the non-leaving group
from 4-methylphenyl to 4-nitrophenyl.[32,34] For the reactions
with pyridines, the pKa0 value increases from 9.4 to 10.7 (for the
mononitro derivatives)[33,34] and from 8.5 to 9.9 (for the dinitro
compounds)[35] for the same change of non-leaving group.
The smaller variation of the pKa0 value with the non-leaving
group found for carbonates compared with benzoates could be
due to the fact that for carbonates there is a longer distance
between the substituent in the non-leaving group and the central
carbonyl carbon. Therefore, a smaller influence of the non-leaving
group on the values of kꢂ1 and k2 (and therefore, a smaller effect
on the pKa0 value) should be expected for carbonates, relative to
benzoates. Nevertheless, Um and co-workers[36–38] have found
that for the aminolysis of aryl X-substituted benzoates in 80 mol%
water/ 20 mol% DMSO there is no variation of the pKa0 value with
the X substituent.
ethoxy to phenoxy.[29,30]
A more significant influence of
the non-leaving group on the pKa0 value can be observed by
comparing the pyridinolysis of S-methyl and O-methyl
2,4-dinitrophenyl carbonates: the pKa0 value changes from 7.3
to 7.8.[3,31] Similarly, in the reactions of SA amines with ethyl and
phenyl 2,4-dinitrophenyl thionocarbonates, the pKa0 value
increases from 7.0 to 7.7 by the change of ethoxy to phenoxy
as the non-leaving group.[30]
Influence of the electrophilic group
In order to assess the influence of the electrophilic group (CO vs.
CS) on the kinetics and pKa0 value we can compare the
pyridinolysis reactions of phenyl 2,4-dinitrophenyl thionocarbo-
nate and its carbonate derivative. For the thionocarbonate a
biphasic Brønsted-type plot with pKa0 ¼ 7.0 was obtained.[30] For
the pyridinolysis of the corresponding carbonate derivative
(carbonate 3) the pKa0 value found is 8.0 (this work). This means
that for a given pyridine the kꢂ1/k2 ratio is larger for the
carbonate compared with the thionocarbonate, see Eqn (4).
The above results are in agreement with those obtained in
similar reactions. For those of primary amines with 4-nitrophenyl
thionobenzoate a biphasic Brønsted plot with pKa0 ¼ 9.2 was
found.[39] For the same aminolysis of the corresponding benzoate
a linear Brønsted plot with slope 0.9 was obtained, which
indicates that the decomposition of the intermediate Tꢀ to
products is rate limiting for the whole amine series (pKa ranging
from 5.7 to 10.9).[39] This means that for the reactions of
4-nitrophenyl benzoate, pKa0 > 10.9. The same results were found
for the reactions of primary alkyl amines with 4-nitrophenyl
phenyl carbonate and its analogous thionocarbonate.[19] The pKa0
values for these reactions are 9.5 and 8.5, respectively.[19] Namely,
there is an increase of one pKa unit by the replacement of CS by
CO. An increase in the pKa0 value was also found in the
pyridinolysis of 4-nitrophenyl benzoate (pKa0 ca. 12) and its thiono
analog (pKa0 ¼ 9.3).[37] Similarly, for the reactions of SA amines
with 4-nitrophenyl thiolacetate the values of kꢂ1 and k2, and also
the values of the kꢂ1/k2 ratio, are larger than those for the same
aminolysis of 4-nitrophenyl dithioacetate.[40] According to Eqn (4)
In order to evaluate the influence of the non-leaving group of
the substrate on the kinetics of the title reactions, a multi-
parametric study was undertaken. With the experimental kN
values obtained for the reactions of carbonates 3–7 with
pyridines with pKa greater than pKa0, where kN ¼ k1, Eqn (5) was
obtained (n ¼ 15, R2 ¼ 0.964). In this equation pKN and pKnlg are
the pKa of the conjugate acids of the nucleophiles and
non-leaving groups, respectively:
log k1 ¼ 2:8 þ 0:29 pKN ꢂ 0:20 pKnlg
(5)
On the other hand, using the experimental kN values for the
reactions of carbonates 3–7 with pyridines with pKa lower than
pKa0, where the k2 step in Scheme 1 is rate determining and
kN ¼ k1k2/kꢂ1, Eqn (6) was obtained (n ¼ 14, R2 ¼ 0.996):
log kN ¼ ꢂ1:3 þ 0:89 pKN ꢂ 0:23 pKnlg
(6)
The coefficients of pKN (bN) in both equations are in
accordance with those found in the aminolysis of carbonates,
esters, and their thio derivatives when formation and breakdown
of the intermediate Tꢀ, respectively, are the rate-limiting
steps.[27,28] On the other hand, the coefficients of pKnlg (bnlg
)
for k1 and kN, Eqns (5) and (6), are closely similar, indicating that
the charge development (from reactant to transition state) on the
oxygen of the non-leaving group is similar for rate determining
formation and breakdown of Tꢀ.
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 816–822

PYRIDINOLYSIS OF DIARYL CARBONATES
this indicates a larger pKa0 value for the aminolysis of the
thiolacetate. The same conclusion arises from a comparison
between the SA aminolysis of 4-nitrophenyl acetate and
4-nitrophenyl thionoacetate.[41] The larger kꢂ1/k2 values for the
acetate indicate that the pKa0 value increases by the change of CS
by CO as the electrophilic group.
The influence of the electrophilic group on the reactivity
toward pyridines can also be evaluated. For pyridines
of pKa > pKa0, for which kN ¼ k1, the k1 values are greater for
carbonates than thionocarbonates (for example for a hypothe-
tical pyridine with pKa ¼ 10 carbonate 3 is 36 times more reactive
than the corresponding thionocarbonate).[30] This shows that CO
is a better electrophile toward pyridines than CS. The same was
found in the pyridinolysis of alkyl aryl thionocarbonates and their
corresponding carbonates: the k1 values are larger for the latter
compounds compared with the former.[29]
and the pH was maintained either by partial protonation of the
pyridines or by means of an external buffer.
Pseudo-first-order rate coefficients (kobs) were found through-
out and determined by means of the spectrophotometer kinetic
software for first-order reactions. The experimental conditions of
the reactions and the values of kobs are shown in Tables 1–7 in
Supplementary Materials.
Product studies
In the reactions of carbonates 1 and 2, 4-nitrophenoxide ion was
found as one of the products of the reactions; in those of
carbonates 3–7, 2,4-dinitrophenoxide anion was obtained as one
of the products. This was achieved by a comparison of the UV–Vis
spectra after completion of the reactions with those of authentic
samples under the same experimental conditions.
In the region of pKa < pKa0 (where Tꢀ breakdown to products is
rate determining), carbonate 3 is slightly more reactive than the
corresponding thionocarbonate.[30] As in this region kN ¼ k1k2/
CONCLUSIONS
k
and k1 is much greater for carbonates (as shown above), it
ꢂ1
From the title reactions the following conclusions can be drawn:
(i) The biphasic Brønsted-type plots (log kN vs. pKa of pyridinium
ions) obtained, the magnitude of the slopes, and the pKa values
found for the center of the break (pKa0) are in accordance with
those for a stepwise mechanism. (ii) An important influence of the
leaving group on the pKa0 value and only a marginal effect of the
non-leaving group were found. (iii) SA amines shift the pKa0 to
higher values, relative to pyridines, due to the faster nucleofug-
ality of an SA amine from the intermediate Tꢀ compared with that
of an isobasic pyridine. (iv) The change of CS by CO as the
electrophilic group also enlarges the pKa0 value.
follows that k2/k should be greater for the thionocarbonate.
ꢂ1
This is in accordance with the larger kꢂ1/k2 values found in the
pyridinolysis of alkyl aryl carbonates relative to those in the same
aminolysis of the corresponding thionocarbonates.[29] These
results are also in agreement with the larger pK0 value (larger k
/
ꢂ1
a
k2 ratio) found for the pyridinolysis of carbonate 3 (this work)
compared with that obtained for the same aminolysis of phenyl
[30]
2,4-dinitrophenyl thionocarbonate.
EXPERIMENTAL
Materials
Acknowledgements
The series of pyridines were purified as reported.[42,43] Carbonates
1,[12] 2,[13] 3,[9] 4,[12] 5,[16] and 7[44] were prepared as described.
Carbonate 6 was prepared as follows: to 2.0 g (10.8 mmol) of
4-nitrophenyl chloroformate in 40 ml of anhydrous acetonitrile, a
mixture of 1 ml of pyridine and 1.685 g (10.8 mmol) of
2,4-dinitrophenol in acetonitrile was added at room temperature
and left 3 h with stirring. The mixture was washed three times
with cold water, dried with magnesium sulfate and the solvent
evaporated off. The solid obtained was recrystallized from a
dichloromethane–hexane mixture yielding a solid that melted at
134–135 8C, and was identified as follows: NMR-1H (400 MHz,
CDCl3) d (ppm): 7.52 (d, 2H, J ¼ 9.2 Hz); 7.70 (d, 1H, J ¼ 8.8 Hz); 8.35
We thank MECESUP of Chile (Projects PUC-0004 and RED QUI-
MICA UCH-01), FONDECYT of Chile (Projects 1020538 and
1060593) for financial assistance.
REFERENCES
[1] P. M. Bond, R. B. Moodie, J. Chem. Soc. Perkin Trans. 1976, 679–682.
[2] E. A. Castro, M. Freudenberg, J. Org. Chem. 1980, 45, 906–910.
[3] E. A. Castro, F. J. Gil, J. Am. Chem. Soc. 1977, 99, 7611–7612.
˜
[4] E. A. Castro, F. Ibanez, S. Lagos, M. Schick, J. G. Santos, J. Org. Chem.
1992, 57, 2694–2699.
(d, 2H, J ¼ 9.2 Hz); 8.63 (dd, 1H, J ¼ 2.8 Hz, J ¼ 9.2 Hz); 9.09 (d, 1H,
[5] T. H. Fife, J. E. C. Hutchins, J. Am. Chem. Soc. 1981, 103, 4194–4199.
[6] H. J. Koh, J. W. Lee, H. W. Lee, I. Lee, Can. J. Chem. 1998, 76, 710–716.
[7] P. Tundo, L. Rossi, A. Loris, J. Org. Chem. 2005, 70, 2219–2224.
[8] E. A. Castro, M. Cubillos, J. G. Santos, J. Org. Chem. 2001, 66,
6000–6003.
13
J ¼ 2.8 Hz). NMR- C (400 MHz, CDCl3) d (ppm): 121.71 (C2 ,C6 );
0
0
0
0
0
122.30 (C3); 125.60 (C3 ,C5 ); 126.19 (C6); 129.75 (C5); 141.10 (C2);
0
—
145.89 (C ); 146.25 (C ); 147.85 (C ); 149.39 (C O); 154.71 (C ).
—
1
4
4
1
Anal. Calcd. for C13H7N3O9: C, 44.71; H, 2.02; N, 12.03. Found: C,
44.71; H, 1.84; N, 11.98.
´
[9] E. A. Castro, M. Aliaga, P. Campodonico, J. G. Santos, J. Org. Chem.
2002, 67, 8911–8916.
[10] M. J. Gresser, W. P. Jencks, J. Am. Chem. Soc. 1977, 99, 6963–6970.
[11] M. J. Gresser, W. P. Jencks, J. Am. Chem. Soc. 1977, 99, 6970–6980.
´
[12] E. A. Castro, M. Andujar, P. Campodonico, J. G. Santos, Int. J. Chem.
Kinet. 2002, 34, 309–315.
[13] E. A. Castro, M. Andujar, A. Toro, J. G. Santos, J. Org. Chem. 2003, 68,
Kinetics measurements
These were carried out by means of a HP-8453 diode array
spectrophotometer in 44 wt% ethanol–water, at 25.0 ꢀ 0.1 8C and
an ionic strength of 0.2 M (KCl). The reactions were studied by
monitoring the appearance of 4-nitrophenoxide anion at 400 nm
or 2,4-dinitrophenoxide anion at 360 nm.
All the reactions were examined under excess of amine over
the substrate. The initial substrate concentration was 5 ꢃ 10ꢂ5 M,
3608–3613.
[14] E. A. Castro, M. Gazitu´a, J. G. Santos, J. Org. Chem. 2005, 70,
8088–8092.
[15] E. A. Castro, P. Campodonico, R. Contreras, P. Fuentealba, J. G. Santos,
´
J. R. Leis, L. Garc´ıa-R´ıo, J. A. Saez, L. Domingo, Tetrahedron 2006, 62,
2555–2562.
[16] E. A. Castro, P. Campodonico, A. Toro, J. G. Santos, J. Org. Chem. 2003,
68, 5930–5935.
´
J. Phys. Org. Chem. 2008, 21 816–822
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

E. A. CASTRO ET AL.
´
[17] E. A. Castro, C. Soto, B. Vasquez, J. G. Santos, Arkivoc 2008, x, 151–160.
[32] E. A. Castro, J. Bessolo, R. Aguayo, J. G. Santos, J. Org. Chem. 2003, 68,
8157–8161.
[33] E. A. Castro, M. Vivanco, R. Aguayo, J. G. Santos, J. Org. Chem. 2004, 69,
5399–5404.
[34] E. A. Castro, R. Aguayo, J. Bessolo, J. G. Santos, J. Phys. Org. Chem.
2006, 19, 555–561.
[35] E. A. Castro, R. Aguayo, J. Bessolo, J. G. Santos, J. Org. Chem. 2005, 70,
3530–3536.
[18] I.-H. Um, H.-R. Park, E.-Y. Kim, Bull. Korean Chem. Soc. 2003, 24,
1251–1255.
[19] I.-H. Um, E. Y. Kim, H.-R. Park, S.-E. Jeon, J. Org. Chem. 2006, 71,
2302–2306.
[20] A. Albert, E. P. Serjeant, The Determination of Ionization Constants,
Chapman and Hall, London, 1971, 87.
[21] A. C. Satterthwait, W. P. Jencks, J. Am. Chem. Soc. 1974, 96, 7018–7031.
[22] E. A. Castro, M. Cubillos, J. G. Santos, J. Org. Chem. 1999, 64,
6342–6346.
[36] I.-H. Um, J.-Y. Lee, S.-H. Ko, S.-K. Bae, J. Org. Chem. 2006, 71,
5800–5803.
[23] E. A. Castro, G. Mun˜oz, M. Salas, J. G. Santos, Int. J. Chem. Kinet. 1995,
27, 987–995.
[37] I.-H. Um, S.-J. Hwang, M.-H. Baek, E.-J. Park, J. Org. Chem. 2006, 71,
9191–9197.
[24] H. J. Koh, J.-W. Lee, H. W. Lee, I. Lee, New J. Chem. 1997, 21, 447–451.
[25] H. J. Koh, C. H. Shin, H. W. Lee, I. Lee, J. Chem. Soc. Perkin Trans 1998,
1329–1332.
[38] I.-H. Um, S.-M. Chun, K. Akhtar, Bull. Korean Chem. Soc. 2007, 28,
220–224.
[39] P. Campbell, B. A. Lapinskas, J. Am. Chem. Soc. 1977, 99, 5378–5382.
˜
[40] E. A. Castro, F. Ibanez, J. G. Santos, C. Ureta, J. Org. Chem. 1992, 57,
[26] H. J. Koh, K. L. Han, H. W. Lee, I. Lee, Bull. Korean Chem. Soc. 2002, 23,
715–720.
[27] E. A. Castro, Chem. Rev. 1999, 99, 3505–3524.
[28] E. A. Castro, J. Sulfur Chem. 2007, 28, 407–435.
´
[29] E. A. Castro, M. Cubillos, J. G. Santos, J. Tellez, J. Org. Chem. 1997, 62,
2512–2517.
7024–7028.
[41] E. A. Castro, F. Ibanez, J. G. Santos, C. Ureta, J. Org. Chem. 1993, 58,
˜
4908–4912.
[42] P. M. Bond, E. A. Castro, R. B. Moodie, J. Chem. Soc. Perkin Trans. 1976,
68–72.
[30] E. A. Castro, M. Aliaga, S. Evangelisti, J. G. Santos, J. Org. Chem. 2004,
69, 2411–2416.
[31] E. A. Castro, M. Aliaga, J. G. Santos, J. Org. Chem. 2004, 69, 6711–6714.
[43] E. A. Castro, C. L. Santander, J. Org. Chem. 1985, 50, 3595–3600.
[44] O. E. El Seoud, M. I. El Seoud, J. P. S. Farah, J. Org. Chem. 1997, 62,
5928–5933.
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 816–822