Changing mechanisms in the β-cyclodextrin-mediated hydrolysis of phenyl esters of perfluoroalkanoic acids

Article abstract of DOI:10.1021/jo034402b

The rate of hydrolysis of esters CF₃(CF₂) _nCOOPh (1 (n = 1), 2 (n = 2), and 3 (n = 6)) was measured at pH 6.00 and at pH higher than 9.00 in the presence of β-cyclodextrin (β-CD). For compounds 1 and 2 the reaction rate decreases as the β-CD concentration increases, and they show saturation effects at all pH. It is suggested that the substrate forms an inclusion complex with cyclodextrin. Analysis of the rate data allows calculation of the association equilibrium constant, K_CD, the rate constant for the reaction of the included compound, k_c, and K^TS which is the hypothetical association equilibrium constant for the transition state of the cyclodextrin-mediated reaction. The dependence of log K_CD and log K^TS with the number of atoms in the chain is different. We suggest that the reactions of 1 and 2 take place with the perfluorinated alkyl chain included in the cavity, whereas the transition state for the reaction of phenyl trifluoroacetate involves a complex with the aryl ring inside the cavity. At low pH the inhibition comes from the protection of the carbonyl group toward nucleophilic attack by water. In basic pH the reaction of HO^- as an external nucleophile is also inhibited. The cyclodextrin-mediated reaction involves the ionized OH group at the rim of the cyclodextrin cavity with poor efficiency due to an unfavorable orientation of the substrate in the complex. On the other hand, the reaction of compound 3 is strongly accelerated by cyclodextrin because the association of the substrate with cyclodextrin competes with the monomer-aggregate equilibrium and at high enough cyclodextrin concentration the main species present in solution is the complex between 3 and cyclodextrin.

Full text of DOI:10.1021/jo034402b

Ch a n gin g Mech a n ism s in th e â-Cyclod extr in -Med ia ted Hyd r olysis
of P h en yl Ester s of P er flu or oa lk a n oic Acid s
Mariana A. Fern a´ ndez and Rita H. de Rossi*
Instituto de Investigaciones en F ´ı sico-Qu ı´ mica de C o´ rdoba (INFIQC), Facultad de Ciencias Qu ´ı micas,
Departamento de Qu ´ı mica Org a´ nica, Universidad Nacional de C o´ rdoba, Ciudad Universitaria,
5
000, C o´ rdoba, Argentina
ritah@dqo.fcq.unc.edu.ar
Received March 27, 2003
3 2 n
The rate of hydrolysis of esters CF (CF ) COOPh (1 (n ) 1), 2 (n ) 2), and 3 (n ) 6)) was measured
at pH 6.00 and at pH higher than 9.00 in the presence of â-cyclodextrin (â-CD). For compounds 1
and 2 the reaction rate decreases as the â-CD concentration increases, and they show saturation
effects at all pH. It is suggested that the substrate forms an inclusion complex with cyclodextrin.
Analysis of the rate data allows calculation of the association equilibrium constant, KCD, the rate
TS
c
constant for the reaction of the included compound, k , and K which is the hypothetical association
equilibrium constant for the transition state of the cyclodextrin-mediated reaction. The dependence
TS
of log KCD and log K with the number of atoms in the chain is different. We suggest that the
reactions of 1 and 2 take place with the perfluorinated alkyl chain included in the cavity, whereas
the transition state for the reaction of phenyl trifluoroacetate involves a complex with the aryl
ring inside the cavity. At low pH the inhibition comes from the protection of the carbonyl group
-
toward nucleophilic attack by water. In basic pH the reaction of HO as an external nucleophile is
also inhibited. The cyclodextrin-mediated reaction involves the ionized OH group at the rim of the
cyclodextrin cavity with poor efficiency due to an unfavorable orientation of the substrate in the
complex. On the other hand, the reaction of compound 3 is strongly accelerated by cyclodextrin
because the association of the substrate with cyclodextrin competes with the monomer-aggregate
equilibrium and at high enough cyclodextrin concentration the main species present in solution is
the complex between 3 and cyclodextrin.
In tr od u ction
etate esters which are both accelerated. We attributed
the different behavior of amides and esters to different
rate-determining steps in the cyclodextrin-mediated re-
action. Also, some of the differences in behavior may be
attributed to the fact that the amides were derived of
perfluorinated compounds whereas the esters were hy-
drocarbon derived. In fact, comparing the behavior of aryl
acetates and aryl trifluoroacetates it was concluded that
all the reactions of the perfluorinated esters were less
accelerated than their hydrocarbon-derived analogues.
The difference in behavior is mainly due to an inefficient
stabilization of the transition state by â-CD for the
reactions of the perfluorinated esters. Interest in the
reactivity of perfluorinated compounds has grown in
recent years because of the important applications of this
The effect of cyclodextrins on the basic cleavage of aryl
esters of alkanoic acids of different chain length has been
widely studied.1 The mechanism of â-cyclodextrin (â-
CD)-mediated hydrolysis of esters involves the formation
of an inclusion complex between the substrate and the
-5
6
cyclodextrin. For the reactions accelerated by cyclodex-
trin, one of the secondary OH groups at the rim of cy-
clodextrin reacts as a nucleophile with the carbonyl car-
bon of the ester displacing the leaving group and giving
the acylated cyclodextrin. The same mechanism is re-
sponsible for the rate acceleration observed in amide
hydrolysis. However, we found some important differ-
ences in the behavior of phenyl alkanoates and some am-
ides derived from perfluorinated acids. For instance, the
hydrolysis reaction of trifluoroacetanilide and m-nitro-
8,9
type of compounds. On the other hand, their interaction
with â-CD has some interesting peculiarities, and several
mechanism have been determined for the cyclodextrin-
mediated hydrolysis reactions. In addition, the kinetics
7
trifluoroacetanilide was inhibited by â-cyclodextrin, con-
trasting with the reactions of p- and m-nitrophenylac-
1
0,11
of that reactions are also dependent on the pH.
In
*
Corresponding author.
1) Breslow, R.; Trainor, G.; Ueno, A. J . Am. Chem. Soc. 1983, 105,
739.
2) Van Etten, R. L.; Sebastian, J . F.; Clowes, G. A.; Bender, M. L.
J . Am. Chem. Soc. 1967, 89, 3242.
the presence of â-CD, the rate increases in a nonlinear
fashion at constant pH above pH ) 8 but it decreases at
(
2
(
(8) Barthelemy, P.; Tomao, V.; Selb, J .; Chaudier, Y.; Pucci, B.
Langmuir 2002, 18, 2557.
(9) Eastoe, J .; Paul, A.; Downer, A. Langmuir 2002, 18, 3014.
(10) Fern a´ ndez, M. A.; de Rossi, R. H. J . Org. Chem. 1997, 62, 7554.
(11) Fern a´ ndez, M. A.; de Rossi, R. H.; Cervell o´ , E.; J aime, C. J .
Org. Chem. 2001, 66, 4399.
(3) Tee, O. S.; Takasaki, B. K. Can. J . Chem. 1985, 63, 3540.
(4) Tee, O. S.; Mazza, C.; Du, X.-X. J . Org. Chem. 1990, 55, 3603.
(5) Menger, F. M.; Ladika, M. J . Am. Chem. Soc. 1987, 109, 3145.
(6) Komiyama, M.; Bender, M. L. J . Am. Chem. Soc. 1978, 100, 4576.
(7) Granados, A.; de Rossi, R. H. J . Org. Chem. 1993, 58, 1771.
1
0.1021/jo034402b CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/05/2003
J . Org. Chem. 2003, 68, 6887-6893
6887

Fern a´ ndez and de Rossi
pH ) 6. The behavior in alkaline pH is attributed to the
nucleophilic reaction of one of the OH groups of the host,
while the inhibition observed at low pH is ascribed to an
unfavorable microenvironment for the reaction and pro-
tection of the carbonyl group from nucleophilic attack by
water.
We report here a study of the hydrolysis reaction of
phenyl pentafluoropropanoate (1), phenyl heptafluorobu-
tanoate (2), and phenyl perfluorooctanoate (3). This study
was carried out because the effect of the change in the
length of the alkyl chain was shown to give a great deal
of information regarding the nature of the complexes
involved in the reactions of aryl esters of alkanoates.12
It turns out that reactions of 1 and 2 are inhibited by
the presence of cyclodextrin at low and high pH as well.
This behavior contrasts with the results corresponding
to the hydrocarbon-derived esters and with the behavior
of aryl trifluoroacetate. For the later compound, the
reactions are inhibited by â-CD at pH < 7 but accelerated
at pH > 9.10 On the other hand, reactions of compound 3
are accelerated even at very low cyclodextrin concentra-
F IGURE 1. Effect of â-CD on the rate of hydrolysis of 1 at
tion. Due to the high hydrophobic character of the -CF
2
-
4
pH ) 6.00 and buffer concentration 0.01 M. [1]
o
) 1 × 10
groups,13 compound 3 has a high tendency to associate
in water solutions and the association with cyclodextrin
breaks down these aggregates, giving rise to the observed
increase in rate. This result is remarkable because it
points to the importance of â-CD as aggregate breaker
in aqueous solution, an effect that is very important in
regard to the reactivity of organic compounds in water.
In recent years, there has been growing interest in the
use of water as a solvent.14
M; µ ) 0.2 M; temp. 25.1 °C. Solvent contains 3.8% MeCN.
The line was calculated using eq 2 with a ) 5.68, b ) 723,
and c ) 0.452.
(Supporting Information, Tables S1 and S2). The rate
-
4
constant of the reaction was measured at 1.0 × 10 and
-
4
1.8 × 10 M substrate concentration for 1 and at 5 ×
-
5
-4
10 and 1 × 10 M for 2 at pH ) 6.00. The rate
constants were independent of the substrate concentra-
tion, indicating that aggregation of the substrates can
be discounted under these reaction conditions.
Resu lts
The effect of â-CD concentration on reaction rate was
determined at two buffer concentrations (Supporting In-
formation, Tables S3 and S4). Only for substrate 1 at pH
The rate constants for the hydrolysis of perfluoroal-
kanoates 1-3 were determined at pH 6.00 and at
alkaline pH (eq 1).
)
6.00 there were some differences in the rates measured
at 0.01 and 0.1 M buffer, although in all cases the rate
decreases with increasing concentration of â-CD and
shows a saturation effect. (Figure 1 is representative).
For compound 1 at pH ) 6.00, the observed rate con-
stants at each buffer concentration were fitted to eq 2
where a and b are adjustable parameters, c is the ob-
served rate constant in the absence of â-CD, which was
determined independently, and [â-CD] represents the
molar concentration of â-CD.
c + a[â-CD]
kobs
)
(2)
1
+ b[â-CD]
The results obtained for substrates 1 and 2 will be
described first because they are qualitatively similar,
while those of 3 will be described later since they present
some peculiarities.
In general, the ester reactivity diminished when the
chain length increases. The hydrolysis rate constants of
substrates 1 and 2 were measured as a function of pH at
pH 6.00, 9.92, and 10.50. At each pH several concentra-
tions of buffer were used. A very weak catalysis by the
buffer was observed only for compound 1 at pH 6.00
Using parameters a and b calculated at each buffer
concentration, the values of the rate constants were cal-
culated at the same concentration of â-CD. These values
were then plotted against buffer concentration, and from
the intercepts of the plots the rate constants extrapolated
to zero buffer concentrations were obtained. In all other
reactions the complete set of data obtained at the dif-
ferent buffer concentrations was used and fitted to eq 2.
The fact that the reactions at pH > 9.00 were inhibited
by â-CD was surprising because the rate of the hydro-
carbon-derived esters of the same chain length are ac-
celerated by â-CD and the rate of the shorter members
of the series, trifluoroacetate, is also accelerated at pH
> 9.
(
(
12) Tee, O. S.; Du, X.-X. J . Am. Chem. Soc. 1992, 114, 620.
13) Shinoda, K.; Hato, M.; Hayashi, T. J . Phys. Chem. 1972, 76,
9
09.
(14) Lindstr o¨ m, U. M. Chem. Rev. 2002, 102, 2751.
6
888 J . Org. Chem., Vol. 68, No. 18, 2003

â-Cyclodextrin-Mediated Hydrolysis of Phenyl Esters
TABLE 1. Obser ved Ra te Con sta n ts for th e Hyd r olysis
a
of 3 a s a F u n ction of Su bstr a te Con cen tr a tion
λ ) 242 nm
λ ) 269.6 nm
3],10- M k1, 10-4 s-1 k2, 10-4 s-1 k1, 10-4 s-1 k2, 10-4 s-1
5
[
pH ) 6.0b
d
0
0
1
1
2
3
4
7
.84
.94
.79
.81
.43
.30
.87
.38
1.2 ( 0.1
3.32 ( 0.08
3.22 ( 0.06
1.87 ( 0.07
1.76 ( 0.05
1.22 ( 0.06c
d
d
d
1.99 ( 0.05
0.94 ( 0.02
0.99 ( 0.02
21 ( 2c
0.61 ( 0.02 33 ( 6
c
c
c
c
c
c
12.7 ( 0.7 0.51 ( 0.07 15 ( 1
12.8 ( 0.5 0.43 ( 0.01 11.9 ( 0.4
0.89 ( 0.03
c
c
c
c
0.51 ( 0.01c
0.34 ( 0.03c
c
c
10.8 ( 0.4 0.29 ( 0.01
pH ) 9.9e
8.9 ( 0.4
1
2
8
.11
.20
.63
900 ( 200f
f
700 ( 300
160 ( 50f
a
T ) (25.1 ( 0.1)°C; µ ) 0.2 M; solvent contains MeCN ) 3.8%;
b
c
buffer ) 0.01 M. Buffer Na2HPO4/NaH2PO4. The values were
obtained by fitting the absorbance vs time data to the following
equation: A ) A∞ + a1e
exponential equation. Buffer NaHCO3/Na2CO3. Each value cor-
responds to the average of at least 10 determinations, and the
errors are standard deviation.
-
k1t
-k2t d
+ a2e
.
The data fit to single-
e
f
The results for compound 3 showed some special
features not found with the substrates of shorter chain
length. The hydrolysis of this compound was studied vs
substrate concentration, and it was observed that the rate
constant changes with the concentration of the substrate.
The kinetic measurements at pH 6.00 were done follow-
ing the appearance of phenol at its wavelength maxi-
mum, 269.6 nm, and others where there was a significant
change with time, i.e., 242 nm. The absorbance vs time
data gave a very good fit to a single-exponential equation
F IGURE 2. Absorbance vs time for the reaction of phenyl
-
5
perfluorooctanoate (4.87 × 10 M) at pH ) 6.00 and buffer
concentration 0.01 M in water with 3.8% MeCN and ionic
strength 0.2 M, (25.1 ( 0.1)° C.
-6
mum of the product and at other wavelengths no mea-
surable changes were observed.16
In the presence of â-CD, the hydrolysis rate constant
of 3 increases considerably, for instance, from 0.007 to
for concentrations of 3 in the range of 8.4 × 10 to 1.81
-
5
×
10 M. However, the rate constant decreased when
the concentration of the substrate increased, and analysis
of the data at different wavelengths gave different values.
-
1
0
.074 s at pH 6.00 with substrate concentration of 2.4
This is a clear indication that more than one kinetic
-
5
-4
process is taking place.15
× 10
M and â-CD concentration of 4.4 × 10 M and
-
3
-5
5.1 × 10 M, respectively (Table 2). With this increment
a maximum value is reached, but in the presence of
further addition of â-CD, the rate constant diminishes.
Similar behavior is observed at pH 6.00 and 9.90. At
substrate concentrations where the kinetics fit to a
At concentrations higher than 2.43 × 10 M the
reaction no longer fits to a single-exponential equation
and two processes could be observed at several wave-
lengths. At 242 nm, for instance, an initial rise at very
short times followed by a decay was observed. The first
process became more visible as the substrate concentra-
tion increases (Table 1 and Figure 2).
-
5
double-exponential equation (g2.4 × 10 M) in the
absence of â-CD, addition of â-CD at concentrations g2
-
3
-
5
-5
× 10 M change the shape of the curve and it fits to a
At two key concentrations (1.0 × 10 and 2.5 × 10
single exponential (Figure 3).
M) where the observations are different (single- and
double-exponential fits) we did measurements of hydroly-
sis rate when the buffer concentrations are modified to
analyze its effect. There was no effect on the hydrolysis
rate constant associated with the buffer concentration
whatever substrate concentration was used.
Discu ssion
The reactivity of perfluorinated esters diminishes when
the chain becomes longer in a manner similar to that
observed with hydrocarbon-derived esters. For instance,
the hydrolysis rate constant of m-nitrophenyl esters at
At pH 9.9 the hydrolysis rate constants were measured
for several substrate concentrations at λ ) 269 nm, and
the data fit well to a single exponential, but the rate
constant decreased when the concentration was raised
-
1
pH 11.7 is 0.086, 0.040, and 0.029 s
for acetate,
4
propionate, and butanoate, respectively. For perfluori-
nated esters, the values at pH 9.9 are 71.6, 36, and 26.5
-
5
-5
from 1.1 × 10 to 8.6 × 10 M (Table 1).
-
1
s
for the phenyl perfluoroacetate, pentafluoropro-
In this case it was not possible to do measurements at
other wavelengths because the amplitude of the mea-
sured process was very small at the wavelength maxi-
panoate, and heptafluorobutanoate, respectively. In the
(16) It should be noticed that at pH 6.00, the reactions were
measured in a conventional spectrophotometer in 5 cm path cells,
whereas at pH 9, the measurements were carried out in a stopped flow
apparatus where the optical pathlength is only 1 cm.
(
15) Bernasconi, C. F. Relaxation Kinetics; Academic Press: New
York, 1976; p 142.
J . Org. Chem, Vol. 68, No. 18, 2003 6889

Fern a´ ndez and de Rossi
TABLE 2. Effect of th e Con cen tr a tion of â-CD on th e Ra te Con sta n t (s-1) for th e Hyd r olysis of 3a
[
3], 10-
2.198
pH ) 6.00c
5
M
â-CD], 10-
3
M
1.085
1.790b
2.435
3.014
7.535
[
0
0
1
2
3
5
7
.44
.90
.28
.43
.71
.07
.68
0.017 ( 0.002
0.06 ( 0.02
0.007 ( 0.002
0.021 ( 0.002
0.030 ( 0.007
0.047 ( 0.008
0.063 ( 0.004
0.074 ( 0.006
0.067 ( 0.004
0.06 ( 0.01
0.03 ( 0.01
0.043 ( 0.009
0.10 ( 0.04
0.060 ( 0.009
0.09 ( 0.01
0.070 ( 0.006
0.060 ( 0.005
0.052 ( 0.007
0.042 ( 0.004
0.06 ( 0.02
0.04 ( 0.01
1
1
1.5
pH ) 9.90d
0
0
1
2
5
.17
11 ( 3e
19 ( 3e
24 ( 4
25 ( 3
20 ( 2
16 ( 2
20 ( 12e
17 ( 5
17 ( 6
e
.35
.30
.76
.12
29 ( 6
25 ( 7
24 ( 4
e
18 ( 7
e
24 ( 4
13 ( 4
0.03
20 ( 3
16 ( 2
15.3(0.5
a
T ) (25.1 ( 0.1) °C; µ ) 0.2 M; MeCN ) 3.8%; [buffer] ) 0.096 M unless otherwise noted; each value is the average of at least 10
determinations, and the error represents standard deviation of the mean. [Buffer] ) 0.046 M. c Buffer is NaH2PO4/Na2HPO4. Buffer
is NaHCO3/Na2CO3. Calculated neglecting the first 5-10% of the reaction.
b
d
e
effect expected for the replacement of a â-hydrogen by a
methyl group. It is remarkable that the rate ratios for
trifluoroacetate/perfluoropropionate and perfluorpropi-
onate/perfluorobutanoate, which are 2.00 and 1.36, re-
spectively, are the same as those for the hydrogenated
esters despite the fact that the volume of a CF
about 1.5 larger than that of a CH
group.17 It appears
that the increase in steric effect is compensated by the
3
group is
3
3
increase in electron-withdrawing effect when a CF group
replaces one fluorine atom in the R position.
The reactivity of perfluorooctanoate is significantly
lower than that of the other two shorter esters. This low
reactivity indicates that 3 is aggregated even at the lower
-
6
concentration used (8.4 × 10 M). For hydrocarbon-
derived esters it is very well established that the decrease
in rate of reaction with concentration is due to self-
association of the esters.18 Besides, p-nitrophenyl per-
fluorononanamide, which was shown to be aggregated
-
6
even at 2 × 10 M concentration, is also much less
reactive than p-nitrophenyl acetanilide.19
It is known that cyclodextrins normally accelerate
hydrolysis reactions of esters by the nucleophilic reaction
of their ionized secondary OH groups, which leads to an
acylated cyclodextrin.20 Also, general base catalysis of
water addition has been postulated for the catalysis of
2
1
some ester hydrolysis. However, there are several ex-
amples where â-CD inhibits reactions by a noncovalent
mechanism such as a microsolvent effect. On the other
hand, it can occur that the geometrical requirements for
the inclusion in the cavity impose such conformation to
the substrate that induces an increase or a diminution
on the reaction rate.
F IGURE 3. Absorbance vs time for the reaction of compound
-
5
3
(7.54 × 10 M) at 25.1 ( 0.1 °C, pH 9.90, buffer 0.096 M,
One of the earliest examples of inhibition by cyclodex-
trins is the intramolecular transfer reaction of acyl group
of 2-hydroxymethyl-4-nitrophenyl trimethyl acetate, which
-
4
-2
â-CD ) 3.5 × 10 M (A) and â-CD ) 1.00 × 10 M (B).
2
2
is catalyzed by R-CD but inhibited by â-CD. It is
hydrocarbon-derived esters, the decrease in rate is at-
tributed to the steric effect produced by the replacement
of hydrogen in the R position for a methyl group. The
effect is more important for the change from acetate to
propionate (rate ratio 2.1) than from propionate to
butanoate (rate ratio 1.4), consistent with the smaller
(
(
17) Wilson, L. D.; Verral, R. E. J . Phys. Chem. B 1998, 102, 480.
18) (a) J iang, X.-K.; Hui, Y.-Z.; Fan, W.-Q. J . Am. Chem. Soc. 1984,
1
06, 3839. (b) J iang, X.-K Acc. Chem. Res. 1988, 21, 362.
(
(
19) Granados, A.; de Rossi, R. H. J . Am. Chem. Soc. 1995, 117, 3690.
20) Tee, O. S. Adv. Phys. Org. Chem. 1994, 29, 1.
(21) Komiyama, M.; Inoue, S. Bull. Chem. Soc. J pn. 1980, 53, 3334.
6
890 J . Org. Chem., Vol. 68, No. 18, 2003

â-Cyclodextrin-Mediated Hydrolysis of Phenyl Esters
TABLE 3. Ra te a n d Equ ilibr iu m Con sta n ts for th e Hyd r olysis of P h en yl Tr iflu or oa ceta te a n d Su bstr a tes 1 a n d 2 in th e
P r esen ce of â-CDa
ku, s-
1 b
kc, s-1
kc/ku
k2, M-1 s-1 c
K
CD, 10 M
2
-1 d
KTS, M-1
substrate
pH
phenyl trifluoroacetatee
6.00
2.06
13.0
71.6
0.44
36
0.017 f
27
180
0.0085
24
82
0.015
0.008 f
0.48
2.51
0.019
0.67
0.71
0.051
0.49
0.44
2.2 f
1.2
1.2
1.2
1.05f
256
246
14
361
522
186
1057
1053
3
9
9
.02
.91
3.2 × 10
4
2.2 × 10
1
2
6.00
6.2 ( 0.2
7.3 ( 0.1
5.3 ( 1.4
7.3 ( 2.1
36 ( 12
22 ( 6
g
h
4
9
.92
(1.3 ( 0.4)10
(6 ( 2)104
54 ( 44
1
1
0.49
115
6.00i
.92j
0.29
26.5
133
4
9
13
58.3
(2.8 ( 0.9)10
k
(1.4 ( 0.2)105
0.50
24 ( 3
a
Solvent: water with 3.8% acetonitrile, 25 °C, ionic strength 0.2 M. b Values extrapolated to zero buffer concentration. c Corresponds
d
to parameter a in eq 2, and the errors are standard deviation of the fit. Corresponds to parameter b in eq 2 ,and the error are standard
deviation of the fit. e Data from ref 10. f Data calculated by extrapolation of the data of 1 and 2 vs number of carbons. Values calculated
g
h
applying eq 2 to data obtained at 0.02 and 0.1 M buffer concentrations. Values calculated applying eq 2 to data obtained at 0.02 and
j
0
.08 M buffer concentrations. i Values calculated applying eq 2 to data obtained at 0.1 M buffer concentration. Values calculated applying
k
eq 2 to data obtained at 0.05 and 0.1 M buffer concentrations. Values calculated applying eq 2 to data obtained at 0.08 M buffer
concentration.
suggested that the substrate enters deeper and tighter
in the bigger cavity of â-CD, so it is more difficult for it
to accomplish the geometry of the transition state.
Inhibition was also observed trough a noncovalent
2
3
mechanism for the hydrolysis of benzocaine and of alkyl
esters such as ethyl p-aminobenzoate and m- and o-ethyl
aminobenzoate.24 The authors suggest that the last
substrates undergo total inclusion in the â-CD cavity and
are shielded from nucleophilic attack by the alkoxide ion
of the cyclodextrin and also by hydroxyl ion from solution.
The reaction of 4-carboxy-2-nitrophenyl acetate at pH
1
1.7 is accelerated by â-CD, but the reaction of the
1
2
propanoate and butanoate derivatives are inhibited.
The cleavage of esters that show saturation kinetics
is usually explained by the reaction of the unbound
substrate and that of the 1:1 substrate-â-CD complex.
This is schematically shown in eqs 3 and 4, where S
represents the substrate and SCD the complex of the
substrate and â-CD and P the reaction products
F IGURE 4. Dependence of the association equilibrium con-
stant of perfluoroesters with â-CD with the number of carbons
in the acyl chain.
ku
9
indicates that the inclusion mode of the substrate is the
same for the three compounds. The slope of this plot has
a value of 0.67.
The association constants of sodium perfluoro al-
kanoates also increase with the chain length; the slope
S
8 P
(3)
(4)
KCD
kc
98 P
S + â-CD y z SCD
The observed rate constant for this system is given by
eq 5, which is of the same mathematical form of eq 2 with
of the plot of log K vs N is 0.58 for salts with 3, 4, 5, and
25
7
carbon atoms in the chain. Linear correlation between
a ) KCD
k
c
, b ) KCD, and c ) k
u
alkyl chain length and binding constant is usual; how-
ever, in general, the sensitivity of K to N is strongly
dependent on the headgroup. The value of the slope of
ku + KCDkc[â-CD]
kobs
)
(5)
4
1
+ KCD[â-CD]
log K vs N for 4-nitro 2-carboxy phenyl alkanoates is 0.2,
2
6
whereas it is 0.1 for p-nitrophenyl alkanoates.
In discussing the effects of â-CD on the cleavage of 1
and 2 we will make use of the kinetic parameters k /k
denotes the limiting
acceleration or retardation due to â-CD, and k is the
second-order rate constant for S + â-CD f P. The value
of k measures the reactivity of â-CD toward different
From the dependence of the observed rate constant on
â-CD concentration, the values of KCD and k can be
c
c
u
,
TS 27
obtained, and these values are collected in Table 3.
The binding of the esters is quite dependent on the
length of the perfluorinated chain (Table 3). For com-
parison, the results of phenyl perfluoroacetate are in-
cluded in Table 3. In Figure 4 it is shown that there is a
linear correlation between log KCD and N, the number of
carbons in the chain. The linear correlation obtained
2 c c u
k (dKCDk ) and K . The ratio k /k
2
2
substrates under the reaction conditions, and so it
measures the selectivity of â-CD for different substrates.
TS
The pseudoequilibrium constant K is the apparent
(
22) Griffiths, D. W.; Bender, M. L. J . Am. Chem. Soc. 1973, 95,
(25) De Lisi, R.; Milioto, S.; De Giacomo, A.; Inglese, A. Langmuir
1999, 15, 5014.
(26) Bonora, G. M.; Fornasier, R.; Scrimin, P.; Tonellato, U. J . Chem.
Soc., Perkin Trans 2 1995, 367.
1
4
679.
(23) Lach, J . L.; Chin, T. F. J . Pharm. Sci. 1964, 53, 924.
24) Chin, T. F.; Chung, P. H.; Lach, J . L. J . Pharm. Sci. 1968, 57,
(
4.
(27) Tee, O. S. Carbohydr. Res. 1989, 192, 181.
J . Org. Chem, Vol. 68, No. 18, 2003 6891

Fern a´ ndez and de Rossi
SCHEME 2. Sch em a tic Rep r esen ta tion of th e
Tr a n sition Sta te for th e Cyclod extr in -Med ia ted
Rea ction in Ba sic Solu tion s
SCHEME 1. Sch em a tic Rep r esen ta tion of th e
Com p lexes w ith th e Ar yl Rin g In sid e (A) th e
Ca vity a n d th e P er flu or in a ted Ch a in In sid e th e
Ca vity (B)
the perfluorinated alkyl chain, it is reasonable to think
that the complex shown in Scheme 1B will predominate
for compounds 1-3.
At low pH the nucleophile is water and it has difficul-
ties reaching the carbonyl group of the included com-
pound. As the chain becomes longer, the carbonyl group
become more exposed to the solvent and is more easily
association constant of the transition state of the â-CD-
mediated reaction (symbolized as TS-CD) into the
transition state of the normal reaction (TS) and â-CD,
eq 6
[
TS-CD]
kcKCD
ku
TS
K
)
)
(6)
[
TS][â-CD]
c u
reached by the nucleophile (k /k increases). In basic
solutions, the cyclodextrin OH groups are partially ion-
ized, and for most of the hydrolysis reactions of esters
that are accelerated by cyclodextrins, the mechanism of
the cyclodextrin-mediated reaction involves the reaction
of one of these ionized OH groups with the substrate
The derivation of eq 6 follows easily from the application
2
8
u c
of transition-state theory to k and k . The usefulness
TS
TS
of K is that log K is directly proportional to the free
energy of stabilization of the transition state due to â-CD
TS
regardless of the mechanism. Thus, variation of log K
(
Scheme 2). The rate of this reaction is strongly depend-
with structure may be used to probe transition-state
structures and as a criterion of mechanism.
The kinetic parameters that characterize the cleavage
of esters mediated by â-CD vary significantly with the
alkyl chain length (Table 3). At pH 6 there is an
30
ent on the orientation of the substrate in the complex.
Trifluoroacetate may have the right orientation in the
complex, and the reaction is effectively accelerated.
For compounds 1 and 2 the predominant complex is
probably that shown in Scheme 2 and the increase in the
length of the chain situates the carbonyl group at a longer
distance from the reactive group of the cyclodextrin. It
is very well established that in intramolecular reactions
the distance and orientation of the reacting groups is very
important.31 We do not think that the main nucleophile
in the cyclodextrin-mediated reaction in basic pH is the
hydroxide ion because in that case the kinetic parameters
for phenyl trifluoroacetate, 1 and 2 should show similar
trends.
c u
important increase in the value of k /k from trifluoro-
acetate to 2 and also an important increase in the
reactivity of the cyclodextrin-mediated reaction. These
systematic changes are due to an increase in the stabi-
lization of the transition state for the cyclodextrin-
TS
mediated reaction (K from 1 to 186) as the alkyl chain
TS
is lengthened by two carbons. The slope of log K vs
number of carbons (N) is 1.1 at pH 6 and 0.46 at pH 9.9,
but the data at pH 9.9 give a low correlation coefficient
(
0.880). On the other hand, the plot of log KCD vs N has
2
The reactions of compound 3 in water occur in more
than one step as is evident in Figure 2 where the
a slope of 0.67 and a good correlation coefficient (r
0
)
TS
.997) (Figure 4). Using the log K at pH 9.90 for
3
2
absorbance at 242 nm increases and then decreases.
trifluoroacetate and 1 the slope is 0.17, whereas with the
data for 1 and 2 the slope is 0.76. These results may
indicate that the transition state for trifluoroacetate is
different than that for the longer chain esters.
There is a good correlation between the maximum
absorption reached and the concentration of the sub-
strate. On the basis of our experience on the behavior of
similar perfluorinated compounds,33 we suggest that the
faster process is associated with the aggregation process
of the substrate and the slower one with the hydrolysis
reaction. The aggregation phenomenon of perfluorinated
compounds is much slower than that corresponding to
hydrocarbon-derived compounds.33 The rate of the first
c u
For the reactions at pH ) 6.00, the plot of k /k vs N is
linear with slope of 0.41. On the other hand, a similar
plot of the data at pH 9.90 is definitively nonlinear. The
data for trifluoroacetate and 1 define a line with slope of
-
0.57, whereas the data for 1 and 2 give a slope of -0.13.
We think that all these results clearly indicate that
different mechanisms operate under a different set of
conditions. The perfluoroalkanoate esters may form at
least two types of complexes with cyclodextrin, namely,
with the aryl ring inside the cavity (Scheme 1A) and/or
with the alkyl chain inside the cavity (Scheme 1B). In
the case of phenyl trifluoroacetate, theoretical calcula-
(29) The difference in steric energy for the complex with the aryl
ring inside and that with the trifluoromethyl inside was calculated as
4
.2 kcal/mol (see ref 11).
(30) The reaction of m-nitro phenyl acetate is more strongly acceler-
ated than that of the para-substituded derivative due to the more
favorable orientation of the substrate in the complex (see ref 20).
(31) (a) Hartwell, E.; Hodgson, D. R. W.; Kirby, A. J . J . Am. Chem.
tions indicate that both types of complexes are of similar
Soc. 2001, 122, 9326. (b) Menger, F. M. Acc. Chem. Res. 1985, 18, 128.
(32) Bernasconi, C. F. Relaxation Kinetics; Academic Press: New
York, 1976; Chapter 9.
energy.29 Taking into account the high hydrophobicity of
(
33) Fern a´ ndez, M. A.; Granados, A. M. ; El Seoud O. A.; de Rossi;
(28) Kurz, J . L. J . Am. Chem. Soc. 1963, 85, 987.
R. H. Langmuir 2002, 18, 8786.
6
892 J . Org. Chem., Vol. 68, No. 18, 2003

â-Cyclodextrin-Mediated Hydrolysis of Phenyl Esters
The â-cyclodextrin36 was used as received, but the purity
was periodically checked by UV spectroscopy; the pure com-
pound does not absorb above 230 nm. The substrates were
prepared by the reaction of phenol with the corresponding acid
chloride following literature methods.37 The product was
obtained after distillation of the remaining acid chloride and
phenol. The products were characterized by IR and mass
spectrometry.
process that we calculated with the two-exponential
fitting of the data has a complex relationship with the
aggregation phenomenon,34 and there are not enough
data to do further calculations. Besides, the nonlinear
fitting is very dependent on the initial parameters used;
therefore, the numbers given for the faster process in
Table 1 are only indicative of the order of magnitude of
the rate of the phenomenon measured.
P h en yl P er flu or op r op a n oa te. IR (KBr, cm-1): 3080,
The presence of â-CD significantly affects the rate of
hydrolysis of 3. For instance, at pH 9.9 with substrate
1798, 1600, 1232, 1170, 833, 761, 502. MS, m/z (rel intensity):
+
2
40 (M , 20), 147 (1), 119 (26), 100 (36), 94 (20), 79 (46), 69
-
5
(70), 45 (100).
concentration 1.1 × 10 M, the rate constant changes
from 9 × 10 to 29 s in the presence of 3.5 × 10
of â-CD. At pH 6.00 the rate increases up to â-CD
P h en yl P er flu or obu ta n oa te. IR (KBr, cm-1): 3074, 1799,
-
2
-1
-4
M
1
2
602, 1223, 1194, 1143, 850, 757, 509. MS, m/z (rel intensity):
90 (M , 7), 169 (10), 150 (11), 119 (27), 100 (35), 94 (20), 79
+
-
3
concentration of 5.07 × 10 M and then there is a slight
decrease (Table 2). We suggest that â-CD forms a complex
with 3 in its monomeric form; therefore, the monomer-
aggregate equilibrium is disrupted due to the low con-
centration of the monomer in solution. The 1:1 complex
of 3 with â-CD reacts at a similar rate as that expected
for 3 in its monomeric form. The slight decrease in the
(
37), 69 (100), 45 (92).
P h en yl P er flu or oocta n oa te. IR (KBr, cm-1): 3070, 1797,
1
593, 1246, 1196, 1156, 847, 747, 533. MS, m/z (rel intensity):
+
+
491 ([M + 1] , 5), 490 (M , 3), 169 (13), 131 (30), 119 (24), 100
(15), 93 (12), 77 (100), 69 (63), 65 (59).
The purity was also controlled comparing the spectrum of
a completely hydrolyzed solution with a solution of the
corresponding phenol under the same conditions.
rate may be attributed to the formation of a 1:2 complex
as observed with hydrocarbon-derived compounds;12 how-
The reactions were followed by measuring the change in
absorbance with time. All reactions of compounds 1 and 2 and
those of compound 3 at alkaline pH were done in a stopped-
flow spectrometer, with unequal mixing, as previously de-
ever, there is not enough data to probe this hypothesis
since solubility reasons precludes the study of reactions
at higher concentrations.
10
scribed. The reactions of compound 3 at pH 6 were measured
In conclusion, formation of inclusion complexes of
esters 1-3 with cyclodextrin leads to different kinds of
effects depending on the substrate and on the pH, and
they are attributed to different mechanisms for the
reactions. For substrates 1 and 2 that are monomers
under the conditions of the study, the reaction is strongly
inhibited in acid and basic pH as well but the mechanism
of the inhibition is different under the two sets of
conditions. At low pH the inhibition comes from protec-
tion of the carbonyl group toward nucleophilic attack by
water. In basic pH the reaction of OH as external
nucleophile is also inhibited. The cyclodextrin-mediated
reaction involves the ionized OH group at the rim of the
cyclodextrin cavity with poor efficiency due to an unfa-
vorable orientation of the substrate in the complex. On
the other hand, the reaction of compound 3 is strongly
accelerated by inclusion in the cavity of cyclodextrin
because it breaks the aggregates of the substrate.
in a conventional UV-vis spectrophotometer. For the kinetic
runs, 0.6 mL of a stock solution of 3 in acetonitrile was injected
into a 5 cm optical pass length quartz cuvette containing 15
mL of a water solution containing all the other ingredients.
All the reactions were carried out at (25.1 ( 0.1) °C, ionic
strength (µ) 0.2 M using NaCl as compensating electrolyte,
and with 3.8% acetonitrile as cosolvent.
Ack n ow led gm en t. This research was supported in
part by the Consejo Nacional de Investigaciones Cien-
t ´ı ficas y T e´ cnicas (CONICET), the Agencia Nacional de
Ciencia y Tecnolog ´ı a (FONCyT), the Agencia C o´ rdoba
Ciencia, SECYT (Universidad Nacional de C o´ rdoba),
and Fundaci o´ n Antorchas.
Su p p or tin g In for m a tion Ava ila ble: Tables S1 and S2
containing the observed rate constant for hydrolysis of sub-
strates 1 and 2 as a function of pH and buffer concentration.
Tables S3 and S4 containing the observed rate constant for 1
and 2, respectively, at different pH, buffer, and â-CD concen-
trations. This material is available free of charge via the
Internet at http://pubs.acs.org.
Exp er im en ta l Section
Aqueous solutions were made up from water purified in a
Millipore apparatus. Acetonitrile, HPLC grade, was used as
received.
J O034402B
The pH measurements were done in a pH meter at con-
trolled temperature and calibrated with buffers prepared in
the laboratory according to the literature.35
(
35) Analytical Chemistry. The Working Tools; Strouts, C. R. N.,
Gilgilan, J . H., Wilson, H. N., Eds.; Oxford University Press: London,
1958, pp 228-233.
(
36) Cyclodextrins were a gift from the pharmaceutical company
(
34) Adenier, A.; Aubard, J .; Schwaller, M.-A J . Phys. Chem. 1992,
Ferromet S. A., Buenos Aires, Argentina.
(37) Clark, R. F.; Simons, J . H. J . Am. Chem. Soc. 1953, 75, 6305.
9
6, 8785.
J . Org. Chem, Vol. 68, No. 18, 2003 6893