Solvent effect and proton inventory in the hydrolysis of p-methylphenyl trichloroacetate

Article abstract of DOI:10.1002/poc.1009

Hydrolysis of p-methylphenyl trichloroacetate in water-acetonitrile mixtures was studied as a function of water concentration in the range 5.5-55.5 M. The proton inventory technique, in H₂O-D₂O mixtures, shows, for a value of D atom fraction in the solvent n = 0.5, deviations from the expected value (for a reaction with one proton being transferred) of 7.5 and 12.3%, for experiments in the presence of 16.6 and 33.3 M L₂O (L = H or D), respectively. Theoretical treatment of the data obtained at [L ₂O] = 16.6 M using the Gross-Butler equation are consistent with a cyclic transition-state structure with three protons involved. Conversely, similar experiments in the presence of [L₂O] = 33.3 M show that multiple water molecules are involved in the transition state of the reaction. Copyright

Full text of DOI:10.1002/poc.1009

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2006; 19: 143–147
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.1009
Solvent effect and proton inventory in the hydrolysis
of p-methylphenyl trichloroacetate
´
˜
´
Clea M. L. Frasson, Tiago A. S. Brandao, Cesar Zucco* and Faruk Nome
´
´
Departamento de Quımica, Universidade Federal de Santa Catarina, 88040-900 Florianopolis, SC, Brazil
Received 4 April 2005; revised 7 October 2005; accepted 10 October 2005
ABSTRACT: Hydrolysis of p-methylphenyl trichloroacetate in water–acetonitrile mixtures was studied as a function
of water concentration in the range 5.5–55.5 M. The proton inventory technique, in H2O–D2O mixtures, shows, for a
value of D atom fraction in the solvent n ¼ 0.5, deviations from the expected value (for a reaction with one proton
being transferred) of 7.5 and 12.3%, for experiments in the presence of 16.6 and 33.3 M L2O (L ¼ H or D),
respectively. Theoretical treatment of the data obtained at [L2O] ¼ 16.6 M using the Gross–Butler equation are
consistent with a cyclic transition-state structure with three protons involved. Conversely, similar experiments in the
presence of [L2O] ¼ 33.3 M show that multiple water molecules are involved in the transition state of the reaction.
Copyright # 2006 John Wiley & Sons, Ltd.
Supplementary electronic material for this paper is available in Wiley Interscience at http://www.interscience.
wiley.com/jpages/0894-3230/suppmat/
KEYWORDS: hydrolysis; p-methylphenyl trichloroacetate; transition state; proton inventory; kinetics
INTRODUCTION
A great number of both experimental4–8 and theoreti-
cal8–12 studies have been devoted to the hydrolysis
reactions of aldehydes, ketones, esters and amides.13–17
It has been observed that the nucleophile attack on a
carbonyl group is mediated by water molecules that
previously hydrated the carbonyl group.12 The concerted
reaction involves proton transfers with carbon–nucleo-
phile bond formation avoiding the energetic cost of
desolvation. Engbersen and Engberts13 suggested that
the strongly negative entropy observed in these reactions
indicates that the transition state is highly hydrated
compared with the reagents in the ground state.
In order to examine the effects of the composition of
the solvent in water–acetonitrile mixtures in the structure
of the transition state, we decided to examine the hydro-
lysis of p-methylphenyl trichloroacetate, where the pre-
sence of the ꢀ-chloro substituents should facilitate the
hydration and the reaction should therefore be sensitive to
the composition of the solvent mixture.
Proton transfer is the most common enzyme-catalyzed
reaction, i.e. enzymes can catalyze efficiently proton trans-
fer reactions involving what would otherwise be high-
energy intermediates.1 In this context, there is considerable
interest in the role of hydrogen bonding in view of the fact
that low-barrier hydrogen bonds (LBHB) may play an
important role in the proton transfer mechanism in enzyme
catalysis.2 However, Warshel and Papazyan3 postulated
that LBHB do not offer a catalytic advantage over ordinary
hydrogen bonds. In fact, ordered water molecules seem to
play a critical role in several enzyme-catalyzed processes,
which provide a scaffold to impart proximity, orientation
and nucleophilicity to water molecules.
It is known that in the water-catalyzed hydrolysis of
esters the rate constants have a strong dependence on the
water concentration. For example, reaction of p-nitrophe-
nyl trifluoroacetate is thought to proceed with three
protons undergoing bonding changes in an eight-mem-
bered transition state (1).4
RESULTS AND DISCUSSION
The hydrolysis reaction of p-methylphenyl trichloroace-
tate (2) was carried out in aqueous acetonitrile and
resulted in the formation of p-cresol and trichloroacetic
acid (Scheme 1). First-order rate constants for the hydro-
lysis of 2, measured in the range 0.001–0.1 M HCl,
showed that the reaction was independent of acid
concentration and the average value in the plateau
region corresponds to the first-order rate constant of the
´
*Correspondence to: C. Zucco, Departamento de Quımica, Universi-
´
dade Federal de Santa Catarina, Campus Universitario—Trindade,
88040-900 Florianopolis, SC, Brazil. E-mail: czucco@qmc.ufsc.br
´
Contract/grant sponsors: PRONEX; CAPES; CNPq.
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 143–147

144
C. M. L. FRASSON ET AL.
Scheme 1
spontaneous water reaction (k0 ¼ 2.69 ꢀ 10ꢁ 3 sꢁ 1
,
Fig. 1S in the Supplementary Material, available in Wiley
Interscience).
The rate constant for the hydrolysis of 2 depends
strongly on the composition of the solvent and, in
water–acetonitrile mixtures, a 10-fold increase in water
concentration results in an ꢂ2700-fold increase in the
rate constant (Table 1).
The strong dependence of the rate constant on the water
concentration observed in Table 1 is strongly indicative
of a reaction involving several molecules of water
(Scheme 1) and consistent with the equation
Figure 1. Plots of logkobs versus log[H2O] for the hydration
of p-methylphenyl trichloroacetate in H2O–CH3CN mixtures
(filled squares, left axis) and activity coefficients of water in
H2O–CH3CN mixtures (open squares, right axis)
curvature (Fig. 1). Indeed, it is approximately linear in
the water concentration range from 5.5 to ꢂ22 M
(log[H2O] ¼ 1.35), with a slope which indicates a kinetic
order of 1.6 ꢄ 0.1, and this kinetic order is 3.1 ꢄ 0.1 if we
use the activity of water in acetonitrile instead of the
concentration of water (plot not shown).18 Above 33.3 M
(log[H2O] ¼ 1.52), there is a steep increase in slope and,
in the high water content region, the calculated limiting
slope indicates that the kinetic order in relation to water is
8.0 ꢄ 0.4, a result consistent with the increase in hydro-
gen bonds between water molecules provoked by the
decrease in molar fraction of the organic solvent (the
kinetic order is 9.9 ꢄ 0.1 using the activity data).18
The observed reaction order is indicative that at low
water content the reaction behaves as if it were a third-
order reaction in relation to the water concentration and
that, at the limit of high water concentration, multiple
water molecules (probably a considerable part of those
forming the cybotactic region) are participating in the
reaction. Clearly, the observed effects may be interpreted
as reflecting changes in the structure of the solvent
mixture, which in turn will reflect variation of the transi-
tion-state structure with solvent composition in the hy-
drolysis of 2. A similar behavior has been reported for the
hydration reaction of 2,2-dichloro-1-(3-nitrophenyl)etha-
none in H2O–THF mixtures.7
x
kobs ¼ k1½H2Oꢃ
ð1Þ
Hence the logarithmic form of Eqn (1) should allow us to
estimate the order of the reaction in relation to water:
logkobs ¼ logk1 þ xlog½H2Oꢃ
ð2Þ
According to Eqn (2), a plot of logkobs versus log[H2O]
is expected to be linear, with slope x, the apparent order
with respect to water. In most cases, the relation observed
is not linear because of a variety of effects, of which
changes in polarity as a function of the solvent composi-
tion is most evident and, responsible for major effects in
the observed rate constants. As a consequence, the value
of the slope is not expected to be an accurate value for the
order of the reaction, but to give an indication of the
number of water molecules involved in the rate-limiting
transition state. However, experimentally, the plot of the
observed rate constant for the hydrolysis of 2 as a
function of water concentration, shows considerable
Table 1. Rate constants for the hydrolysis of p-methylphe-
nyl trichloroacetate as a function of the concentration of
water in H2O–CH3CN mixtures at 25 ꢅC
In Fig. 1, we also plot the activity coefficients of water
in acetonitrile versus log[H2O] as determined by French
measurement of vapor pressures of acetonitrile–water
mixtures over the whole composition range using the
static method.18 It is clear that there is only a very small
effect on the activity coefficient of water above 33.3 M
and, in this region, as the concentration of acetonitrile in
the mixture becomes higher, the activity coefficient
increases by only about 10% compared with pure water.
It is important to note that very small changes in activity
coefficients are observed in the same region where the
[H2O] (M)
kobs (10ꢁ3 sꢁ1
)
5.50
11.0
16.6
22.4
27.5
33.3
38.9
43.7
50.1
55.0
0.0670
0.174
0.367
0.630
1.91
3.03
8.91
22.3
70.6
184
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 143–147

SOLVENT EFFECT IN HYDROLYSIS OF p-METHYLPHENYL CHLOROACETATE
145
Table 2. Observed rate constants for the hydration of p-
methylphenyl trichloroacetate in DCl/D2O–H2O/CH3CN at
different deuterium molar fractions n
increase in concentration of water has a strong effect on
the observed rate constant. The observed experimental
results could be explained according to a ‘microdomains’
model, where the limit of water concentration at which
acetonitrile can be accommodated within the cavities of
ordinary water is ꢂ37 M (log[H2O] ꢆ 1.56).19,20 Below
this limit there are two microdomains, the first highly
structured consisting predominantly of coordinated water
molecules and the second relatively disordered contain-
ing mostly acetonitrile and hydrogen-bonded water mo-
lecules.20,21 Indeed, for log[H2O] < 1.35 the solvent
mixture is dominated by the formation of acetonitrile–
water clusters and, as a consequence, in this acetonitrile-
rich region, the rate constant decreases linearly with
decreasing water concentration. In the water-rich region,
the decrease in water concentration is accompanied by a
sharp decrease in kobs and, most probably, the observed
effect is related to the effect of enhanced cooperative
water–water hydrogen bonding.22,23 In the intermediate
region (log[H2O] from 1.52 to 1.35), the statistical weight
of each type of microdomain is changing and the ob-
served dependence of the rate constant for the hydrolysis
reactions follows the change in composition of the
solvent.20
n
knobs (10ꢁ4 sꢁ1
)
[L2O] ¼ 16.6 M
[L2O] ¼ 33.3 M
0
3.010
2.700
2.520
2.340
2.080
1.880
1.700
1.540
1.320
1.160
1.056
28.04
25.72
22.74
20.56
18.28
16.25
14.97
12.61
11.13
10.42
9.00
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
L2O–CH3CN mixtures (Table 2). The total water con-
centration ([L2O], where L ¼ H or D) was maintained
constant at 16.6 and 33.3 M to obtain information about
the number of protons involved in the transition states of
both regions of the water order plot (Fig. 1).
Since the slopes of plots of logkobs versus log[H2O]
may be affected by the non-ideal behavior of the solvent
mixture,19,24 we applied the proton inventory technique
in order to gain further insight into the structure of the
transition state for the hydrolysis of 2 in H2O–CH3CN
mixtures. With this technique it is possible to obtain
information about the number of hydrogen atoms that are
effectively transferred in the rate-determining step of the
reaction.
The proton inventory technique is based on the fact that
the solvent isotope effect, k0/kn, for a variety of proton-
transfer reactions in H2O–D2O mixtures is a function of
n, the atom fraction25 of deuterium in the mixture. This
happens because not all the exchangeable protons caus-
ing the isotope effect have the same composition as the
bulk solvent. Equation (3), known as the Gross–Butler
equation, shows the theoretical treatment of this phenom-
enon for rate processes:
The results are consistent with a normal isotope effect
with values for KH O=KD O of 2.85 and 3.11, at [L2O]
2
2
16.6 and 33.3 M, respectively, and they are similar to
those for the reaction of other activated aryl esters.14,26 In
both cases a downward curvature is obtained when the
observed rate constants are plotted against the molar
fraction of deuterium (Fig. 2). The degree of curvature
depends on the increase in the molar concentration of
L2O in the mixture, showing, for a value of n ¼ 0.5,
deviations from the expected value (for a reaction with
one proton being transferred) of 7.5 and 12.3% for
[L2O] ¼ 16.6 and 33.3 M, respectively. Thus, an increase
in the molar concentration of L2O in the solvent mixture
"
,
#
v
v
ꢂ
ꢃ
Y
Y
ꢀ
ꢁ
kn ¼ k0
1 ꢁ n þ nꢁTi
1 ꢁ n þ nꢁRj
ð3Þ
i
j
where kn and k0 are the rate constants for a solvent whose
D atom fraction is equal to n and for a pure protiated
solvent, respectively, ꢁ are the fractionation factors for
every proton which exchanges with the solvent during the
process under study and the superscripts R and T refer to
reagents and transition state, respectively; n is the atom
fraction of deuterium in the solvent and varies from 0
(pure protiated solvent) to 1 (pure deuterated solvent).
The rates constants for the hydrolysis reaction of
2 were determined in different molar fractions of
Figure 2. Rate constants for the hydrolysis of p-methylphe-
nyl trichloroacetate in L2O–CH3CN mixtures as a function of
the deuterium molar fraction, n, for [L2O] ¼ 16.6 and 33.3 M.
Theoretical Gross–Butler curves correspond to three protons
(&) and multiprotons (*) intervening in the transition state
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 143–147

146
C. M. L. FRASSON ET AL.
causes an increase in the number of protons that take part
in the transition state. The experimental data shown in
Fig. 2 are consistent with three or more protons being
involved in the process. The arguments follow.
Theoretical treatment of the data (Fig. 2) was per-
formed using the Gross–Butler equation for a reaction
where there is no contribution of the reagent to the
observed isotope effect:25
to that observed by Venkatasubban et al. for p-nitrophe-
nyl trifluoroacetate, a fractionation factor of 0.697 ꢄ
0.005 for a three-proton model.4
ꢀ
ꢁ
m
knobs ¼ k0obs 1 ꢁ n þ nꢁT
ð4Þ
where m corresponds to the number of protons that
contributed to the observed isotope effect.
Structure 3 is consistent with the observed effect and
we can see that the three hydrogens Ha are most likely
responsible for the global isotope effect, since the hydro-
gens Hb contribute little or nothing, since they are
expected to have a fractionation factor of 1.
Thus, the solid lines in Fig. 2 represent the best fits
obtained for the experimental data assuming three pro-
tons and multiprotons participating in the transition state
for [L2O] ¼ 16.6 and 33.3 M, respectively.
Since the value of kH O=kD O is 2.85, at [L2O] ¼ 16.6, it
2
2
According to the Gross–Butler theory, the equations
is very difficult in this particular case to distinguish a two-
proton from a three-proton mechanism. This experimen-
tal difficulty is related to the relatively small global
isotope effect and to the associated precision necessary
to distinguish between these mechanistic possibilities.
Despite this fact, we believe that the choice of an eight-
membered transition state, rather than a six-membered
transition state involving two water molecules, is appro-
priate given that the former has bond angles that accom-
modate approximately linear hydrogen bonds.4,27
ꢀ
ꢁ
1=3
ðkn=k0Þ ¼ 1 þ ꢁT ꢁ 1 n
ð5Þ
ꢀ
ꢁ
lnðkn=k0Þ ¼ ꢁm 1 ꢁ ꢁT
n
ð6Þ
are adequate for reactions where the numbers of protons
participating in the rate-limiting step of the reaction are 3
and infinity, respectively.25 Hence the linear plot ob-
served for the experimental data at [L2O] ¼ 16.6 and
33.3 M indicates that the number of protons participating
in the reaction increases from 3, for [L2O] ¼ 16.6 M, to a
multiproton reaction at water concentration 33.3 M (cor-
relation coefficients >0.99, Fig. 3).
The water-assistance mechanism in ester hydrolysis
has a substantial advantage in comparison with a single
water attack. Molecular orbital calculations carried out
by Hori et al.10 for the oxygen exchange accompanying
the alkaline hydrolysis of methyl acetate showed that
the activation barrier for the water-assisted mechanism
is lower [3.8 kcal molꢁ1 (1 kcal ¼ 4.184 kJ)] than that
of the hydroxide attack without water assistance
(21.8 kcal molꢁ1). In addition, results reported by
Guthrie12 have shown that a water-assisted mechanism
involving proton transfer from water to carbonyl oxygen
in a cyclic transition state avoids the cost of desolvation
of the negative charge generated by the nucleophilic
attack and, therefore, favors a concerted mechanism.
The data for hydrolysis at [L2O] ¼ 33.3 M are indica-
tive of a highly hydrated transition state. Structure 4 is
consistent with a large number of water molecules
participating in the transition state.
The slopes of the lines in Fig. 3, based on a three-
proton and multi-proton model, with each proton having
an identical fractionation factor, are 0.707 ꢄ 0.003 and
0.869 ꢄ 0.001, respectively. The first value is very similar
Figure 3. Plots of ln(kn/k0) or (kn/k0)1/3 for the hydrolysis of
p-methylphenyl trichloroacetate in L2O–CH3CN mixtures as
a function of the deuterium molar fraction, n, for
[L2O] ¼ 16.6 (&) and 33.3 M (*)
Cyclic transition-state structures are frequently postu-
lated in the literature,4–6,9,10 including for proton transfer
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 143–147

SOLVENT EFFECT IN HYDROLYSIS OF p-METHYLPHENYL CHLOROACETATE
147
in water12,28 and acid-12 and base-catalyzed10–12 hydro-
lysis of esters. In pure water or in water-rich solvent
mixtures, water-to-water hydrogen bonds predominate
and a multiple number of protons participate in the
transition state for the hydrolysis of p-methylphenyl
trichloroacetate. As this structure is disrupted by the
addition of the organic cosolvent, the number of protons
participating in the transition state gradually decreases,
reaching the minimum value of three, which corresponds
to the structure 3. Clearly, in aqueous solutions and in
mixtures of water and organic solvents the role of
hydrogen bonding is fundamental. It is difficult to distin-
guish between the contribution of low-barrier hydrogen
bonds and ordinary hydrogen bonds in terms of catalytic
advantage.2 However, it seems that in this particular
reaction, the highly hydrogen-bonded water structure
typical of the water-rich region provides a reaction cage
which favors proximity, orientation and, as a conse-
quence, increases the rate of the reaction.
program; correlation coefficients, ꢂ, were >0.999 for all
kinetic runs and, between replicates, the standard devia-
tion for the first-order rate constants was always <2%.
The pH was maintained at 2.0 with HCl and the solutions
were prepared immediately before use, except in the
study of the solvent isotope effect, where the solutions
were left standing for 15 h at room temperature in order to
allow the H/D balances to be reached.
Acknowledgements
We are indebted to PRONEX, CAPES and CNPq for
financial support.
REFERENCES
1. Kirby AJ. Acc. Chem. Res. 1997; 30: 290–296.
2. Northrop DB. Acc. Chem. Res. 2001; 34: 790–797.
3. Warshel A, Papazyan A. Proc. Natl. Acad. Sci. USA 1996; 93:
13665–13670.
4. Venkatasubban KS, Bush M, Ross E, Schultz M, Garza O. J. Org.
Chem. 1998; 63: 6115–6118.
EXPERIMENTAL
5. Bell RP, Critchlow J. Proc. R. Soc. London, Ser. A 1971; 325: 35–
55.
6. Bell RP, Sorensen PE. J. Chem. Soc., Perkin Trans. 2, 1972; 1740–
1743.
Preparation of p-methylphenyl trichloroacetate. A solu-
tion of trichloroacetyl chloride (0.1 mol), p-methylphenol
(0.1 mol) and pyridine (0.1 mol) in diethyl ether (250 ml)
was refluxed for 2.5 h. The solvent was evaporated and
the resultant solid was passed through a column of silica
gel HF 60, eluted with hexane–acetone (9:1, v/v). The
solvent was removed and a clear yellow solid was
obtained in 83% yield, m.p. 51 ꢅC.
´
7. Wendhausen Jr. R, Zampiron E, Vianna JJ, Zucco C, Rezende
MC, Nome F. J. Phys. Org. Chem. 1990; 3: 89–94.
8. Bowden K, Fabian WMF, Kollenz G. J. Chem. Soc., Perkin Trans.
2 1997; 547–552.
9. Guthrie JP. J. Am. Chem. Soc. 1996; 118: 12878–12885.
10. Hori K, Hashitani Y, Kaku Y, Ohkubo K. J. Mol. Struct. (Theo-
chem) 1999; 461–462: 589–596.
11. Janoschek R, Fabian WMF. J. Org. Chem. 1999; 64: 3271–3277.
12. Guthrie JP. J. Am. Chem. Soc. 2000; 122: 5529–5538.
13. Engbersen JFJ, Engberts JBFN. J. Am. Chem. Soc. 1975; 97:
1563–1568.
14. Fernandez MA, de Rossi RH. J. Org. Chem. 1999; 64: 6000–6004.
15. Ba-Saif S, Luthra AK, Williams A. J. Am. Chem. Soc. 1989; 111:
2647–2652.
16. Stefanidis D, Cho S, Dhe-Pagamon S, Jencks WP. J. Am. Chem.
Soc. 1993; 115: 1650–1656.
17. Hengge AC, Hess RA. J. Am. Chem. Soc. 1994; 116: 11256–
11263.
18. French HT. J. Chem. Thermodyn. 1987; 19: 1155–1161.
19. Marcus Y, Migron Y. J. Phys. Chem. 1991; 95: 400–406.
20. Easteal AJ, Woof LA. J. Chem. Thermodyn. 1982; 14: 755–762.
21. Gorbunov BZ, Nabrukhin YI. J. Mol. Struct. 1972; 14: 113–116.
22. Engberts JBFN. Pure Appl. Chem. 1982; 54: 1797–1808.
23. Symons MCR. Acc. Chem. Res. 1981; 14: 179–187.
24. El Seoud OA, El Seoud MI, Farah JPS. J. Org. Chem. 1997; 62:
5928–5933.
25. Schowen KBJ. In Transition States of Biochemical Process,
Gandour RD, Schowen RL (eds). Plenum Press: New York,
1978; 225–283.
26. Oakenfull DG, Riley T, Gold V. Chem. Commun. 1966; 385–387.
27. Gandour RD. Tetrahedron Lett. 1974; 295–298.
28. Huskey WP, Warren CT, Hogg JL. J. Org. Chem. 1981; 46: 59–63.
Kinetics. The aqueous solutions were prepared from
deionized, doubly distilled water. Deuterium oxide and
deuterium chloride with a minimum isotopic purity of
99.9 atom-% of deuterium were purchased from Aldrich
and manipulated under a nitrogen atmosphere.
The rate constants of p-methylphenyl trichloroacetate
hydrolysis were followed spectrophotometrically with a
Hewlett-Packard Model 8452 spectrometer equipped
with a thermostated water-jacketed cell holder at
25.0 ꢄ 0.05 ꢅC. In a typical run, the reaction was initiated
by injection of 20 ml of a 10ꢁ2 M stock solution of the
substrate in acetonitrile (Merck, HPLC grade) to 3 ml of
aqueous solvent mixture equilibrated at 25.0 ꢄ 0.05 ꢅC.
The absorbance (A) decay was recorded at 244 nm and
absorbance versus time data were stored directly on a
microcomputer. First-order rate constants, kobs, were
estimated from linear plots of ln(A1 ꢁ At) against time
for at least a 90% reaction using an iterative least-squares
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 143–147