Cyclodextrin effect on solvolysis of substituted benzoyl chlorides

Article abstract of DOI:10.1039/b400302k

A kinetic study was carried out on the solvolysis of substituted benzoyl chlorides in the presence of α-, β- and γ-CD. Combination of the substituent dependent mechanism for solvolysis of benzoyl chlorides and the complexation ability of the cyclodextrin yields the following experimental behavior: (i) catalysis by β- and γ-CD for solvolysis of electron-attracting substituted benzoyl chlorides due to the reaction with its hydroxyl group C(6); (ii) absence of αCD influence on solvolysis of benzoyl chlorides with electron withdrawing substituents; (iii) inhibition of solvolysis of benzoyl chlorides with electron-donating groups. This behavior is observed for solvolysis of metal/para substituted substrates in the presence of β-CD, solvolysis of meta-substituted benzoyl chlorides in the presence of α-CD and solvolysis of para-substituted benzoyl chlorides in the presence of γ-CD. This decrease in the rate constant is a consequence of the complexation of the substrate in the cyclodextrin cavity and its low solvation ability, causing the rate of solvolysis in its interior to be negligible. (iv) The solvolysis of meta-substituted benzoyl chlorides in the presence of γ-CD yields a new behavior where the reaction of the complexed substrate is not negligible in the interior of the cyclodextrin cavity, which has been interpreted as a consequence of incomplete expulsion of hydration water from its cavity when the complexation takes place, (v) The experimental results obtained in the presence of α-CD show that weta-substituted benzoyl chlorides give rise to host : guest complexes with 1 : 1 stoichiometries, whereas those which are para-substituted cause a 2 : 1 stoichiometry to be formed. This difference in behavior has been interpreted taking into account the size of the different benzoyl chlorides and their accommodation in the α-CD cavity.

Full text of DOI:10.1039/b400302k

View Article Online / Journal Homepage / Table of Contents for this issue
Cyclodextrin effect on solvolysis of substituted benzoyl chlorides
J. Báscuas, L. García-Río* and J. R. Leis
Dpto. Química Física, Facultad de Química, Universidad de Santiago, 15782 Santiago, Spain.
E-mail: qflgr3cn@usc.es
Received 8th January 2004, Accepted 23rd February 2004
First published as an Advance Article on the web 18th March 2004
A kinetic study was carried out on the solvolysis of substituted benzoyl chlorides in the presence of α-, β- and γ-CD.
Combination of the substituent dependent mechanism for solvolysis of benzoyl chlorides and the complexation
ability of the cyclodextrin yields the following experimental behavior: (i) catalysis by β- and γ-CD for solvolysis of
electron-attracting substituted benzoyl chlorides due to the reaction with its hydroxyl group C(6); (ii) absence of α-
CD influence on solvolysis of benzoyl chlorides with electron withdrawing substituents; (iii) inhibition of solvolysis
of benzoyl chlorides with electron-donating groups. This behavior is observed for solvolysis of meta/para substituted
substrates in the presence of β-CD, solvolysis of meta-substituted benzoyl chlorides in the presence of α-CD and
solvolysis of para-substituted benzoyl chlorides in the presence of γ-CD. This decrease in the rate constant is a
consequence of the complexation of the substrate in the cyclodextrin cavity and its low solvation ability, causing the
rate of solvolysis in its interior to be negligible. (iv) The solvolysis of meta-substituted benzoyl chlorides in the presence
of γ-CD yields a new behavior where the reaction of the complexed substrate is not negligible in the interior of the
cyclodextrin cavity, which has been interpreted as a consequence of incomplete expulsion of hydration water from its
cavity when the complexation takes place. (v) The experimental results obtained in the presence of α-CD show that
meta-substituted benzoyl chlorides give rise to host : guest complexes with 1 : 1 stoichiometries, whereas those which
are para-substituted cause a 2 : 1 stoichiometry to be formed. This difference in behavior has been interpreted taking
into account the size of the different benzoyl chlorides and their accommodation in the α-CD cavity.
Introduction
for reaction of the acylium ion and the potential well no longer
exists, although the rate limiting transition state is essentially
the same. Therefore the reaction will follow the path indicated
by b. Electron-withdrawing substituents increase the energy of
both the acylium ion and the transition state for the dissociative
path and, in contrast, stabilize the upper left-hand corner
and the transition state for the associative mechanism. This
could eventually favor a change to an associative path with an
addition–elimination mechanism through a tetrahedral inter-
mediate (path c). However, it is also possible that reaction
mechanisms follow path d if the addition intermediate does not
have a significant lifetime.
The mechanism of solvolysis of benzoyl chlorides is well known
both in water and in different solvents. The acyl group transfer
1
was shown to follow one of three mechanisms: dissociative,
associative, and concerted displacement. These mechanisms are
well-defined, with a clear borderline between them. The term
borderline is meaningful in this context: it refers to a reaction
series in which a small change in the structure of one of the
reactants, or in the reaction conditions, causes a change from a
concerted to a stepwise mechanism, or vice versa. The first order
rate constants for the solvolysis of most substituted benzoyl
halides increase with electron-withdrawing substituents and
Cyclodextrins are cyclic oligomers of α--glucose which are
ϩ
follow a Hammett correlation with a slope of ρ = 1.7 (for
substituted benzoyl fluorides). There is a change to a negative
slope for anisoyl fluoride and for 4-chlorobenzoyl chloride
3
produced by enzymatic degradation of starch and are dough-
nut-shaped molecules formed by six, seven or eight glucose
units (α, β or γ) and are able to form inclusion complexes with a
ϩ
(
with a slope of ρ = Ϫ3.0 for substituted benzoyl chlorides)
4
great variety of compounds. The interior of the macrocyclic
which indicates a change to a different reaction pathway with a
dissociative mechanism.
structure is lined with hydrocarbon and the oxygens linking the
glucose units, resulting in a hydrophobic cavity with a water
compatible exterior. The use of cyclodextrins (CD) as micro-
reactors to perform chemical reactions has attracted the interest
The change from a dissociative to an associative mechanism
can be represented schematically by a More O’Ferrall–Jencks
2
diagram (Scheme 1). The anisoyl chloride reacts through a dis-
5–7
of chemists since the 1960s. Cyclodextrins may show two
sociative mechanism giving rise to an acylium ion intermediate,
which rapidly reacts with the solvent (path a). With benzoyl
chlorides with electron-attracting substituents there is no barrier
4
basic forms of catalysis: “noncovalent” and “covalent”. In the
former case, the cyclodextrin binds to the substrate and pro-
vides an environment for the reaction that is different from
the bulk solvent; in this case, the effect of cyclodextrins arises
from the less polar nature of the cavity (a microsolvent effect)
and/or from the conformational restraints imposed on the sub-
strate by the geometry of inclusion. Another factor that may be
important for noncovalent catalysis is the differential solvation
effects at the interface of the cyclodextrin cavity with the
8
external aqueous environment. Covalent catalysis should
involve covalent interactions between the substrate and func-
tional groups on the cyclodextrin in the rate determining step
of the reaction. For instance, the anion of the cyclodextrin
may act as a basic catalyst towards acidic substrates included in
9
their cavity.
In aqueous phases hydrophobic forces are assumed to be
responsible for driving a guest into the CD’s hydrophobic
Scheme 1
1
186
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
interior usually forming 1 : 1 host–guest complexes. The micro-
environment around the reactant in the CD cavity is different
from that which prevails in the reaction media. Three different
effects are expected: (i) microsolvent effects, (ii) the protection
of unstable intermediates or products, and (iii) the solubiliz-
ation of poorly water soluble reactants. In addition, conform-
ational effects can also be expected: (iv) control of reactant
conformation, (v) control of the orientation between reactants,
and (vi) control of the size of the molecule.
In this study we have investigated the influence of α-, β-
and γ-cyclodextrin on the solvolysis of substituted benzoyl
chlorides. As mentioned, the solvolysis mechanism of the
benzoyl chlorides is sensitive to the electron-attracting/electron-
donating substituents, and to the solvent in which the reaction
takes place. We have used eleven substituted benzoyl chlorides
in order to cover a reactivity range between a predominantly
Fig. 1 Hammett plot for solvolysis of substituted benzoyl chlorides in
water at 25 ЊC.
dissociative mechanism (4-CH O) and another which is pre-
3
dominantly associative (4-NO ) and is shown in Scheme 2.
2
1
ϩ
with those found previously by Song and Jencks (ρ = Ϫ3 and
ϩ
ρ = 1.7). The minor discrepancy can be attributed to the fact
that in our reaction medium the percentage of acetonitrile is
slightly higher than that used by Song and Jencks (1.5–1.0%
(
v/v) of acetonitrile).
2
.
Reactions in the presence of ꢀ-CD
Fig. 2 shows the experimental behavior observed for the benzoyl
chlorides which undergo solvolysis by means of an associative
mechanism: 4-CF , 3-CF , 3-NO and 4-NO (not shown). The
3
3
2
2
Scheme 2
observed rate constant, kobs, of 4-CF
3
increases along with the
β-CD concentration and tends to reach a maximum value. The
results obtained on studying the solvolysis of 3-CF and 3-NO2
3
Experimental
show that kobs increases linearly together with the β-CD concen-
tration. In both cases the influence of the [β-CD] on kobs should
be a consequence of the reaction between the benzoyl chloride
and the hydroxyl groups of the cyclodextrin.
The β-CD was supplied by Sigma (purity >98%), and the α-CD
and γ-CD by Cyclolab (purity >98%). All were used without
further purification. In preparing the solutions the water con-
tent supplied by the manufacturer was taken into account.
The benzoyl chlorides were provided by Aldrich, all of which
had purities between 97–98% and were used without further
purification. Their solutions were prepared in acetonitrile
(
Aldrich) to prevent it from decomposing too rapidly.
Reaction kinetics were carried out in an Applied Photo-
physics stopped flow spectrophotometer with unequal mixing.
The benzoyl chloride dissolved in dry acetonitrile was placed
in the smaller syringe (0.1 mL). The larger syringe (2.5 mL)
was filled with an aqueous solution of cyclodextrin. The
total acetonitrile concentration was 3.85% (v/v). Solutions of
benzoyl chlorides were freshly prepared in dry acetonitrile at
the appropriate concentration in order to get a final con-
Ϫ4
centration of 1.0 × 10 M. All experiments were carried out
at 25 ЊC. The kinetic traces were fitted with one exponential
equation using the software of the SF apparatus. The wave-
lengths used to monitor the reactions were 300, 270, 300, 290,
Fig. 2 Influence of β-cyclodextrin concentration on the pseudo-first
2
3
4
60, 270, 300, 285, 295, 250 and 260 nm for 4-CH O, 4-CH ,
order rate constant for solvolysis of (᭺) 4-CF
-NO at 25 ЊC.
3
, (᭹) 3-CF and (ᮀ)
3
3
3
3
-CH , 4-H, 3-CH O, 4-Cl, 3-Cl, 3-CF , 4-CF , 3-NO and
2
3
3
3
3
2
-NO2 respectively. All the kinetic experiments could be
The α-CD, β-CD and γ-CD have doughnut-like annular
reproduced within an error margin of 3%.
structures with wide and narrow hydrophilic ends delineated by
O(2)H and O(3)H secondary and O(6)H primary hydroxyl
11–14
Results and discussion
groups.
The pK values assigned to O(2)H and O(3) in α-,
a
β- and γ-CD are pK = 12.33, 12.20 and 12.08 respectively
a
1
.
Reaction in bulk water
and have an important role in the cyclodextrin mediation
of the hydrolysis of a range of guests inside the cyclodextrin
Fig. 1 shows the results obtained on studying the rate of
10
15
solvolysis of substituted benzoyl chlorides in bulk water. In
the Hammett plot we can observe the existence of two very
annulus. The methods that have been developed for the
controlled modification of cyclodextrins exploit the different
reactivity of the C(2), C(3) and C(6) hydroxyl groups. The
primary hydroxyl groups of the cyclodextrins, C(6), are the
ϩ
different slopes: one negative (ρ = Ϫ2.6 ± 0.3) and another
ϩ
positive (ρ = 1.4 ± 0.5). The first corresponds to substrates
16
which undergo solvolysis by an eminently dissociative mechan-
ism, while the behavior exhibited by the benzoyl chlorides with
electron-attracting groups is consistent with an associative
mechanism. The values of the Hammett slopes are consistent
most basic, they are also the most nucleophilic. Direct O-
alkylation of a C(6) hydroxyl group using the corresponding
CD alkoxide is not feasible, as the C(2) and C(3) hydroxyl
groups are more acidic than those at C(6) and they react prefer-
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3
1187

View Article Online
entially. The C(2) groups are also exposed at the wider end
of the cyclodextrin cavity, such that when they deprotonate
under basic conditions, the resultant alkoxides often react
readily and selectively as nucleophiles. That is, whereas the pri-
mary hydroxyl groups are the most nucleophilic under neutral
and acidic conditions, above pH = 10–11 it is the C(2) second-
ary hydroxyl groups which are generally selectively modified
through treatment with an electrophilic reagents. The kinetic
results for solvolysis of benzoyl chlorides, shown in Fig. 2,
occur through a path which is catalyzed by the cyclodextrin.
This behavior means that the inclusion of the benzoyl chloride
in the cavity of the cyclodextrin must take place with the
carbonyl group in the narrow side of the cavity, as shown in
Scheme 3.
to a combination of changes in the hydrophobic character of
the substrates and to differences in their size. The first factor
may be the reason for the difference in behavior between the
4-CF and the substituted benzoyl chlorides with nitro groups,
3
while the difference in size allows us to explain the difference
in behavior between the 4-CF and the 3-CF . The existing
3
3
studies in the literature show that the complexation constants
of alcohols by cyclodextrins correlate with the partition co-
20,21
efficients between diethyl ether and water.
These observ-
ations are reasonable since, the binding of guests to cyclo-
dextrins is largely governed by their size and hydrophobicity.
No data are available for the partition coefficients between
water and 1-octanol for the benzoyl chlorides. Likewise the
rapidity of the solvolysis reaction makes it impossible to be
experimentally determined. We can compare the values of the
22
partition coefficients, log P, for substituted phenylacetates.
The obtained values, log P = 1.28 and 1.38 for the substituted
esters with 3-NO and 4-NO groups, are clearly lower than
2
2
those obtained for esters substituted with 4-Cl (log P = 2.35)
and 4-CH (log P = 2.28) groups. These values show that the
3
hydrophobic character of the phenylacetates decreases as a
consequence of the introduction of the nitro groups. What is
more significant is the comparison of the incorporation con-
stants of the benzoyl chlorides from the continuous medium
23
to the interface of AOT/isooctane/water microemulsions. The
incorporation equilibrium constants must increase as the
hydrophobic character of the substrates decreases. In the case
of the benzoyl chlorides it can be noted that the incorporation
equilibrium constants are approximately Koi = 5 for all the
substituents and increase to values of K = 121 ± 9 and K =
Scheme 3
oi
oi
From this kinetic scheme we can obtain the following rate
equation:
1
64 ± 20 for 3-NO and 4-NO respectively. This behavior
2
2
is clear evidence of the reduction in the hydrophobic character
as a consequence of the nitro substituents. This reduction in
the hydrophobic character is a satisfactory explanation for the
decrease in the value of the equilibrium constant for the
formation of the inclusion complex with the β-CD.
(
1)
X
w
X
where k
and kCD are the rate constants of the solvolysis of
The substituted benzoyl chlorides with the 3-CF and 4-CF
3 3
the X-substituted benzoyl chlorides in bulk water and in the
inclusion complex with the cyclodextrin. KCD is the equilibrium
groups do not display hydrophobocity, as is the case for the
nitro substituents. In fact, the equilibrium constant of the
X
constant of the cyclodextrin–benzoyl chloride X-substituted
complex. The experimental results can be fitted perfectly to
formation of the inclusion constant between the 4-CF and
3
the β-CD presents a value which is entirely compatible with the
other benzoyl chlorides (see Table 1). Likewise, the values of the
distribution constants for the benzoyl chlorides with the groups
4-CF
3
equation 1, giving the corresponding values of K CD = (180 ±
Ϫ1
4-CF
3
Ϫ2 Ϫ1
4
0) M and k CD = 9.62 × 10 s . The rate constant for the
solvolytic process catalyzed by the β-CD is approximately three
3-CF and 4-CF between the continuous medium and the
3 3
times greater than the value obtained in bulk water. This result
interface of AOT/isooctane/water microemulsions are similar
to those presented by the benzoyl chlorides with substituents
is similar to that found for the solvolysis of 4-NO in water and
2
17
in mixtures of methanol and water. The rate of solvolysis
increases together with the percentage of methanol in the
3-CH
decrease in the equilibrium constant for the formation of the
3-CF inclusion complex with the β-CD we must take into
account geometric considerations. Simple MM calculations
3 3 3 3
, 4-CH , 3-CH O, 4-CH O, etc. Therefore, to explain the
Ϫ1
reaction medium until a maximum value of 0.153 s is reached
3
for 70% (v/v) of methanol and water. Subsequent increases in
the percentage of methanol cause a decrease in the value of the
reaction rate constant.
allow us to evaluate the size of the 3-CF
molecule (Scheme 4).
3
The results in Fig. 2 also show that kobs increases linearly with
the β-CD concentration for solvolysis of 3-CF and 3-NO .
3
2
NMR studies indicate that in nitrophenols the nitro group is
18
found in the interior of the cavity, which is consistent with the
mode of complexation shown in Scheme 3. This behavior can
X
be explained by equation 1, considering that 1 ӷ KCD[β-CD].
In this case equation 1 is reduced to:
X
w
X
X
kobs = k
ϩ kCD
K
CD[β-CD]
(2)
which satisfies the experimental behavior observed. Table 1
Scheme 4
X
X
shows the values of the KCD and kCD parameters obtained from
the application of equations 1–4 to the solvolysis of substituted
Given that the diameter of the β-CD cavity oscillates between
6.0–6.5 Å, it is obvious that the 3-CF molecule is too big to be
19
benzoyl chlorides in the presence of β-CD.
3
The difference between the behavior observed for the 4-CF3
kobs increase on increasing [β-CD] to a limiting value) and the
fully included in the β-CD cavity. This difficulty gives rise to the
kinetic behavior observed in Fig. 2.
(
other benzoyl chlorides with electron-attracting groups (linear
dependence of kobs on [β-CD]) can be attributed fundamentally
Fig. 3 shows the influence of the β-CD concentration on
kobs for the solvolysis of 4-CH O and 3-CH O. In these cases,
3
3
1
188
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3

View Article Online
Table 1 Values of the kinetic parameters obtained for the solvolysis of benzoyl chlorides in the presence of cyclodextrins
β-CD
α-CD
γ-CD
X
w
Ϫ1
X
Ϫ1
X
Ϫ1
X
Ϫ1
X
Ϫ1
X
Ϫ1
X
Ϫ1
k
/s
k
CD/s
k
CD/M
k
1:1/M
k
2:1/M
k
CD/s
k
CD/M
Ϫ2
Ϫ2
Ϫ2
Ϫ2
Ϫ2
a
a
4
3
4
3
3
4
3
4
3
4
4
-NO2
-NO2
-CF3
-CF3
-Cl
7.90 × 10
3.63 × 10
3.56 × 10
3.10 × 10
4.90 × 10
0.189
0.540
1.14
2.50
6.76
1.87
10.7a
7.9
—
—
1.9 ± 0.1
—
—
18 ± 2
—
—
15 ± 3
—
3 ± 1
—
24 ± 3
20 ± 1
0.119 ± 0.008
(3.5 ± 0.1) × 10
—
(4 ± 1) × 10
(9.7 ± 0.5) × 10
33 ± 12
29 ± 5
10 ± 2
25 ± 8
76 ± 4
64 ± 12
78 ± 4
52 ± 3
116 ± 8
79 ± 19
88 ± 14
a
Ϫ2
Ϫ3
0.233
9.62 × 10
0.226
Ϫ2
180 ± 40
a
a
Ϫ3
7.9
—
—
—
—
—
—
—
—
140 ± 10
225 ± 2
406 ± 68
270 ± 11
568 ± 20
523 ± 16
510 ± 15
12.2 ± 0.7
31 ± 5
20 ± 1
11 ± 1
15 ± 1
20 ± 2
17.6 ± 0.8
-Cl
—
Ϫ3
-CH O
(9 ± 1) × 10
3
-H
-CH3
-CH3
—
0.37 ± 0.03
—
—
-CH O
58.2
3
a
Maximum values, see text.
Fig. 3 Influence of the β-CD concentration on the observed rate constant for different benzoyl chlorides: (᭹) kobs/50 for 4-CH O and (᭺) for kobs
3
3
-CH O. Data for the left figure were fitted according to equation 3 and for the right figure according to equation 4.
3
as occurs with 4-CH , 3-CH , 4-H, 4-Cl and 3-Cl, it is possible
inhibition of approximately 190 times in the transition from
34
3
3
to observe a decrease in the observed rate constant, kobs, as the
β-CD concentration increases. This behavior can be easily
explained by recourse to the formation of an inclusion complex
between the benzoyl chloride and the cyclodextrin, as shown in
Scheme 3. The possibility of a reaction between the complexed
benzoyl chloride and the cyclodextrin hydroxyl groups can
be discarded for those substrates which undergo solvolysis
fundamentally by means of a dissociative mechanism. There-
fore, in this case the reaction path within the inclusion complex
should be the expulsion of the leaving group assisted by the
cyclodextrin.
water to mixtures of methanol and water of 90% (v/v). This
difference in reactivity increases if we consider 4-CH O, a
3
benzoyl chloride which is nearer to the dissociative mechanism,
where kobs decreases by approximately 2900 times in the transi-
tion from bulk water to a mixture of methanol and water of
35
95% (v/v).
Given that the solvation ability of the interior of the cyclo-
dextrin is minimal, it is to be expected that the rate constant
X
k
CD would be much lower than the corresponding rate con-
stant in bulk water, in such a way as to confirm the inequality
X
w
X
X
of k
ӷ kCD
K
CD[β-CD]. Equation 1 can then be rewritten thus:
The properties of the water inside the cavity as well as in close
proximity to the cavity are quite different from those of bulk
20,24–26
(3)
water.
The polarity of the interior of the β-CD cavity
27
has been studied using various techniques. Van Etten showed
that the ultraviolet absorption spectrum of 4-tert-butylphenol
in an aqueous solution of α-CD closely matches its spectrum
in dioxane. Uno et al. concluded, on the basis of blue shifts in
the spectra of amine N-oxides in the presence of cyclodextrins,
that the cavity environment is like methanol or ethanol, depend-
ing upon the probe. Studies using fluorescence spectroscopy
show that the cyclodextrin cavity has a polarity in the vicinity
Fig. 3 shows the fit of equation 3 in its present form, or by
28
means of its linearization (equation 4).
(
4)
2
9–32
of alcohols.
Obtained values for the E polarity parameter
From the fitting procedure we can obtain the values of the
complexation equilibrium constants of the corresponding
benzoyl chlorides by the β-CD.
T
show that it has similar values to those of tert-butyl alcohol
33
or ethylene glycol. However it can be seen that the estimated
polarity of the cyclodextrin cavity depends on the compound
used to probe it.
The difference in behavior obtained between the benzoyl
chlorides shown in Fig. 2 and those of Fig. 3 is due to the dif-
ferent mechanism whereby the reaction takes place: the electron-
donating substituents produce a dissociative mechanism where
the reaction rate in the interior of the cyclodextrin is much
lower than that which is obtained in bulk water and, therefore,
we can observe an inhibiting effect derived from the addition
of β-CD to the reaction medium. The results of Fig. 2 show the
catalytic effect exerted by the β-CD on the solvolysis rate of
In spite of the fact that there is no uniformity, it can be noted
that the interior of the cyclodextrin cavity has a much lower
polarity than that of pure water. The solvolysis reactions
of benzoyl chlorides which occur by means of a dissociative
mechanism are very sensitive to the solvent polarity. In fact,
for the solvolysis of 4-H, kobs decreases as the percentage
of methanol in the reaction medium increases, causing an
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3
1189

View Article Online
the benzoyl chlorides with electron-attracting substituents.
The difference in behavior is due to the fact that the hydroxyl
groups of the cyclodextrin can act as efficient nucleophiles if
the reaction takes place through an associative mechanism.
3
.
Solvolytic reactions in the presence of ꢁ-CD
The behavior observed on studying the solvolysis of benzoyl
chlorides in the presence of α-CD is qualitatively and quanti-
tatively different from that which is found with β-CD. The rate
constants of solvolysis of the benzoyl chlorides 3-CH O, 3-CH3
3
and 3-Cl decrease as a consequence of the formation of the
inclusion complex with the α-CD (see Fig. 4 for 3-CH O). The
3
experimental results have been fitted satisfactorily to equations
3
and 4, giving the kinetic parameters which are shown in
Table 1. The magnitude of the inhibition due to the presence
of α-CD is not as great as with β-CD which is attributed to the
lower equilibrium constants of formation of the inclusion
complexes between α-CD and the benzoyl chlorides. This
result is widely documented in the literature showing that
complexation of substituted cyclohexanes with α- and β-CD
Fig. 5 Influence of α-CD concentration on the observed rate constant
for solvolysis of (᭺) 4-NO , (᭹) 3-NO and (ᮀ) 3-CF at 25 ЊC.
2
2
3
the ability of the medium to solvate the leaving group, it is to be
expected that kobs would display a very small decrease when the
substrate is incorporated into the α-CD cavity.
36,37
indicates deeper penetration into the wider β-CD cavity.
The experimental behavior obtained for solvolysis of para-
substituted benzoyl chlorides cannot be explained on the basis
of the mechanism proposed in Scheme 3 and equations 1–4.
Fig. 6 shows by way of example how the experimental
results obtained for the solvolysis of 4-CH O and 4-CF are not
3
3
justifiable on the basis of equation 4. As Fig. 6 shows there
exists a non-linear dependence of 1/kobs on the α-CD concen-
tration. This behavior is consistent with a mechanism whereby
a cyclodextrin:benzoyl chloride complex is formed with a 2 : 1
stoichiometry, as shown in Scheme 5.
Fig. 4 Influence of α-CD concentration on solvolysis of (᭺) 3-CH O
3
at 25 ЊC. Experimental results were fitted to equation 3 (᭺) and 4 (᭹).
The complexation constants for the benzoyl chlorides by α-
CD are approximately 20 times lower than the corresponding
complexation constants with β-CD. Therefore, we can conclude
that the complexation constants of 3-NO , 4-NO and 3-CF
2
2
3
X
must be less than one. Thus, the inequality 1 ӷ KCD[α-CD]
should be observable for all the experimental conditions used
(
it is important to point out that [α-CD] ≤ 0.12 M). Likewise
Fig. 6 Influence of α-CD on (᭺) 1/kobs for solvolysis of 4-CF
and (᭹)
3
X
1
00/kobs for solvolysis of 4-CH O at 25 ЊC.
the rate constant for the cyclodextrin catalyzed path, kCD, must
be only slightly greater than k
inequality k
3
X
w
, in such a way as to confirm the
X
w
X
X
From this complexation scheme we can obtain the following
rate equation:
ӷ kCDKCD[α-CD]. This should result in a situation
where the observed rate constant would be independent of the
X
α-CD concentration, kobs = k
w
. This behavior is shown experi-
mentally for the benzoyl chlorides 3-NO , 4-NO and 3-CF , as
can be observed in Fig. 5.
2
2
3
(
5)
More surprising behavior has been observed when studying
the solvolysis of para substituted benzoyl chlorides (4-CH O,
3
X
X
where k1:1 and k2:1 are the solvolysis rate constants of the
4
-CH , 4-H, 4-Cl and 4-CF ). In all cases we can observe
3
3
cyclodextrin : benzoyl chloride complexes with stoichiometries
: 1 and 2 : 1. Due to the low solvation ability of the cyclo-
dextrin cavity, it is anticipated that these rate constants would
w
be very small in comparison with k , allowing equation 5 to be
that the presence of α-CD causes a reduction in the value of
1
kobs. This behavior is also manifested by 4-CF , a substrate
3
which undergoes more rapid solvolysis in the presence of β-CD.
It is important to point out that the rates of solvolysis of
X
rewritten thus:
4
-CH O, 4-CH and 4-Cl decrease approximately 8 times when
3 3
the α-CD concentration increases up to [α-CD] = 0.108 M,
while the rate of solvolysis of 4-CF decreases only 1.6 times for
3
(
6)
the same interval of cyclodextrin concentrations. This behavior
is a consequence of the different mechanisms which operate in
the solvolysis of 4-CH O, 4-CH , 4-H and 4-Cl (dissociative
3
3
mechanism) and 4-CF (associative mechanism). Given that
the rate of the associative mechanism is not very sensitive to
3
(7)
1
190
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3

View Article Online
Scheme 5
Equation 7 predicts the existence of a linear and quadratic
insertion of a methylene into the aliphatic chain of phenethyl-
dependence of 1/kobs on α-CD concentration, as has been shown
experimentally in Fig. 6. From the fitting of experimental results
to equation 7 we can obtain the parameters shown in Table 1.
It is necessary to justify the reason why the para substituted
benzoyl chlorides form inclusion complexes with 2 : 1 stoichio-
metries, while those which are substituted in the meta position
form complexes with 1 : 1 stoichiometries. To this end we invoke
geometric considerations regarding the α-CD cavity and the
size of the substituted benzoyl chloride. Scheme 6 shows the
sizes of 3-Cl and the α-CD cavity. As we can observe, the diam-
eter of the α-CD cavity is insufficient to fully accommodate the
aromatic ring of the benzoyl chloride 3-Cl. It is important to
note that the size of 3-Cl is the smaller of the meta substituted
benzoyl chlorides. The geometrical restrictions do not allow
complexation with a second molecule of α-CD due to the fact
that it is impossible to include the aromatic ring in the cavity.
amine, giving 3-phenylpropylamine, leads to an enhancement in
affinity toward β-CD, but the affinity increase is much smaller
when α-CD is the host, since the extra methylene added is
38
believed to stay outside the α-CD cavity. Phenethylamine and
2-methoxyphenethylamine exhibit similar affinities toward α-
CD. However, phenethylamine gives a much higher equilibrium
constant for complexation with β-CD than 2-methoxyphene-
thylamine. A plausible explanation for this contrasting result is
that the benzene ring cannot penetrate deeply into the α-CD
cavity and the 2-methoxy group is left outside, even after com-
plexation, and in this position it is unable to significantly affect
the overall complexation thermodynamics. When the system
is switched to the larger β-CD, the cavity may permit deeper
penetration of the benzene ring, ironically resulting in a con-
siderable decrease in complex stability owing to the severe steric
hindrance caused by the 2-methoxy group. This again leads to
the conclusion that the α-CD cavity is too small to accom-
modate the whole benzene ring. Therefore, part of the molecule
will remain outside the α-CD cavity and participate in a second
complexation, forming an inclusion complex with 2 : 1
stoichiometry.
4
Solvolytic reactions in the presence of ꢂ-CD
On studying the influence of γ-CD on the solvolytic rate of
benzoyl chlorides, we must distinguish three very different types
Scheme 6
of behavior. The para substituted benzoyl chlorides: 4-CH O,
3
4
-CH , 4-H, 4-Cl and 4-CF , show a decrease in k as the
3 3 obs
Scheme 6 also shows the dimensions of the para substituted
benzoyl chloride. As can be observed, the size of 4-Cl can be
fitted to the size of the α-CD cavity. Therefore the benzoyl
chlorides substituted in the para position can be included in
the cavity of α-CD through the aromatic ring or the carbonyl
group, and can form inclusion complexes with 2 : 1 stoichio-
metries, as shown in Scheme 5. It is important to note the
difference in behavior found between α-CD and β-CD: the
para-substituted benzoyl chlorides form complexes with 2 : 1
stoichiometry with α-CD but only with 1 : 1 stoichiometry with
β-CD. Once again, the reason must lie in the geometric restric-
tions of the cavity of the cyclodextrins. In both cases the depth
of the cavity is the same: 7.80 Å. Therefore it is to be expected
that the para-substituted benzoyl chlorides form inclusion
complexes with the same stoichiometry in both cases. However,
in the case of α-CD the smaller diameter of the cavity (4.7–5.3 Å)
causes the penetration of the benzoyl chloride in the cavity to
be smaller than in the case of β-CD which has a larger cavity
diameter (6.0–6.5 Å). This difference means that part of the
benzoyl chloride molecule is outside the α-CD cavity and, con-
sequently, available for complexation with a second cyclo-
dextrin molecule.
concentration of γ-CD increases (see Fig. 7 by way of example
for 4-Cl). The experimental results can be explained quanti-
tatively by the mechanism proposed in Scheme 3 and equations
3
–4. Table 1 shows the values of the kinetic parameters
obtained. Fig. 8 shows the behavior obtained on studying the
solvolysis of 4-NO and 3-NO . In this case we can observe a
catalytic effect due to the reaction of the complexed substrate
2
2
This result is consistent with existing findings in the literature
which show that usually guest molecules carrying a phenyl
moiety exhibit stronger affinities to β-CD than to α-CD. The
Fig. 7 Influence of γ-CD concentration on solvolysis of (᭺) 4-Cl
at 25 ЊC. Experimental results were fitted to equation 3 (᭺) and 4 (᭹).
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3
1191

View Article Online
It is important to analyze the difference in behavior shown
by the different substrates. The solvolysis reactions of the
substituted benzoyl chlorides with the most strongly electron-
attracting groups: 4-NO2 and 3-NO , occur through an
2
eminently associative mechanism. Both reactions are catalyzed
by γ-CD due to the reaction with the primary hydroxyl group
of γ-CD. This behavior requires that the geometry of the
inclusion complex be like that which is shown in Scheme 3. The
solvolysis of the other benzoyl chlorides does not present a
catalytic effect as a consequence of the addition of γ-CD, and
we can observe in all cases a reduction in the reaction rate as
the cyclodextrin concentration increases. However, we must
distinguish between the behavior presented by the meta and
para substituted substrates.
The cyclodextrins crystallize from water as hydrates of
variable composition. α-CD is usually encountered as the hexa-
Fig. 8 Influence of γ-CD concentration on solvolysis of (᭹) 4-NO2
and (᭺) 3-NO at 25 ЊC. Experimental results were fitted to equation 1.
2
39,40
hydrate,
β-CD exists as the undecahydrate and as the
41
with the hydroxyl groups of the cyclodextrin. Fitting experi-
dodecahydrate. γ-CD is sometimes described as an octa-
4
-NO
2
mental results to equation 1, we obtain the values of KCD
=
hydrate, but it can crystallize with from 7 to 18 molecules of
Ϫ1
4-NO
2
Ϫ1
42–46
(
33 ± 12) M and kCD = (0.119 ± 0.009) s . The obtained
water.
It is not known exactly how many water molecules (h)
4
-NO
2
value for kCD is slightly greater than the value obtained in bulk
exist in the cavity of natural cyclodextrins, nor how many water
molecules (i) are released from the cavity upon complexation,
but for α-CD, these numbers (h and i) are estimated to be 2 or
4
-NO
2
Ϫ2 Ϫ1
water, k
3
(
w
= 7.90 × 10 s . Likewise for the solvolysis of
3-NO
2
Ϫ1
3-NO
2
-NO we obtain values of KCD = (29 ± 5) M and kCD
=
2
Ϫ2 Ϫ1
3-NO
2
24,47,48
41
3.53 ± 0.03) × 10 s . The obtained value for kCD is practic-
3 and ≤2, respectively.
In the case of β-CD, six to seven
3-NO
2
Ϫ2
ally the same as that obtained in bulk water, k
w
= 3.50 × 10
water molecules are distributed within the cavity, even though
the cavity is big enough to accommodate up to 11 of them.
Taking into account the originally included or interacting water
molecules, the 1 : 1 complexation reaction of a guest (G) with a
cyclodextrin host (H) may be written as follows:
Ϫ1
4-NO
2
3-NO
2
s . The values of kCD and kCD are compatible with those
17
obtained in mixtures of methanol : water: the rate constant
increases together with the percentage of methanol, passes
through a maximum for a percentage of methanol of around
7
0% (v/v) and then decreases.
The solvolysis of other benzoyl chlorides substituted in the
G ؒ gH O ϩ H ؒ hH O
2
2
meta position: 3-CH , 3-CH O, 3-Cl and 3-CF , shows clearly
H ؒ G ؒ (g ϩ h Ϫ i)H O ϩ iH O (9)
2 2
3
3
3
differentiated experimental behavior. The rate constant
decreases as the concentration of γ-CD increases, however,
this experimental behavior cannot be justified on the basis
of equations 3–4. Fig. 9 shows a plot of 1/kobs against [γ-CD]
indicating that the behavior deviates from the linear depend-
ence predicted by equation 4. The experimental results should
fit the mechanism shown in Scheme 3 where the substrate
complexed with γ-CD can undergo a solvolysis reaction. On the
basis of equation 1 we can obtain the following equation:
where g represents the number of water molecules interacting
with the free guest, h the number of tightly bound hydration
water molecules inside the free cyclodextrin cavity, and i the net
displacement of water upon complexation. There is no accurate
information available concerning the values of g, h and i in
solution. However it seems clear than the inclusion of a benzoyl
chloride in the cavity of a γ-CD should not displace all the
existing water molecules in its interior. The more accurate
the adjustment of the guest in the cavity of the cyclodextrin, the
greater the number of displaced water molecules in its interior.
Water molecules in the α-CD or β-CD cavity cannot saturate
their tetrahedral hydrogen bonding capacity as can those in
the bulk of the solvent. They may therefore be regarded as
molecules of enhanced energy or enthalpy. The cavity of γ-CD
is so wide and it may accommodate so many water molecules
that their properties resemble water molecules in the bulk
solvent. In this way we can consider that the interior of the
γ-CD cavity has a high polarity, which is consistent with an
approximately 10-fold reduction in the solvolysis rate of the
meta-substituted benzoyl chlorides, with respect to bulk water.
The results obtained show that for the para-substituted
benzoyl chlorides no reaction is observable in the interior of
the cyclodextrin cavity. This behavior may be due to a greater
capacity for penetration in the interior of the cavity. This
penetration causes the consequent expulsion of water from the
interior of the cyclodextrin and consequently causes γ-CD to
have a very hydrophobic interior which is unable to solvate the
leaving group for the solvolysis reaction. The meta-substituted
benzoyl chlorides present a lower capacity of penetration due to
the greater steric hindrance. This lower capacity of penetration
translates to a lower degree of water expulsion from the interior
of the γ-CD and consequently a more hydrophilic environment
than in the previous case. The existence of water molecules in
the interior of the γ-CD cavity that coexist with the benzoyl
chloride can facilitate the solvation of the leaving group and,
therefore, facilitate the solvolysis reaction in the interior of the
cyclodextrin.
(
8)
which complies perfectly with the experimental behavior as can
be observed in Fig. 9. Table 1 shows the kinetic parameters
obtained for the solvolysis of 3-CH , 3-CH O, 3-Cl and 3-CF .
3
3
3
Fig. 9 Influence of γ-CD concentration on the observed rate constant
for solvolysis of (᭺) 3-CH and plot of (᭹) 1/k vs. [γ-CD] at 25 ЊC.
3
obs
Experimental results were fitted to equations 1 and 8.
1
192
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3

View Article Online
This variation in the polarity of the γ-CD cavity as the com-
(c) A. Granados and R. H. de Rossi, J. Am. Chem. Soc., 1995, 117,
690; (d ) A. Granados and R. H. de Rossi, J. Org. Chem., 2001, 66,
3
plexating substrates vary becomes evident when comparing
X
X
w
1548; (e) M. A. Fernández and R. H. de Rossi, J. Org. Chem., 2001,
66, 4399.
the values of kCD and k
. The values obtained (Table 1) show
X
X
that k
w
is approximately 10 times lower than kCD for 3-CH . In
3
8
(a) O. S. Tee and J. M. Bennett, J. Am. Chem. Soc., 1988, 110, 269;
view of the previous considerations, we have not been able
(
b) O. S. Tee, Adv. Phys. Org. Chem., 1994, 29, 1.
X
to obtain the value of kCD for 4-CH . However we can estimate
9 V. Daffe and J. Fastrez, J. Chem. Soc., Perkin Trans. 2, 1983, 789.
10 As indicated in the experimental section the kinetic results have been
obtained in the presence of a 3.85% (v/v) of acetonitrile.
11 A. R. Khan, P. Forgo, K. J. Stine and V. T. D’Souza, Chem. Rev.,
1998, 98, 1977.
3
a maximum value which confirms the inequality which is
necessary for the simplification of equation 1 to equation 3,
X
w
X
X
namely k
ӷ kCD
K
CD[γ-CD]. We can thus obtain a maximum
4
-CH
3
Ϫ2 Ϫ1
value of k CD = 6.99 × 10 s . The lower polarity of γ-CD
1
2 C. J. Easton and S. F. Lincoln, Modified Cyclodextrins, Imperial
College Press, London, 1999.
motivated by the complexation of 4-CH causes the value of
3
X
k
CD to be at least 100 times lower than that obtained in bulk
13 W. Saenger, C. Niemann, R. Herbst, W. Hinrichs and T. Steiner, Pure
Appl. Chem., 1993, 65, 809.
4 W. Saenger, J. Jacob, K. Gessler, T. Steiner, D. Hoffmann, H. Sanbe,
K. Koizumi, S. M. Smith and T. Takaha, Chem. Rev., 1998, 98, 1787.
5 (a) R. I. Gelb, L. M. Schwartz, J. J. Bradshaw and D. A. Laufer,
Bioorg. Chem., 1980, 9, 299; (b) R. I. Gelb, L. M. Schwartz and
D. A. Laufer, Bioorg. Chem., 1982, 11, 274.
16 (a) L. D. Melton and K. N. Slessor, Carbohydr. Res., 1971, 18, 29; (b)
S. E. Brown, J. H. Coates, D. R. Coghlan, C. J. Easton, S. J. van Eyk,
W. Janowski, A. Leopore, S. F. Lincoln, Y. Luo, B. L. May,
D. S. Schiesser, P. Wang and M. L. Williams, Aust. J. Chem., 1993,
46, 953.
water, contrasting with the behavior obtained in 3-CH3.
1
It is important to note that this behavior has not been
observed in the presence of α-CD or β-CD. This is due to the
fact that the aromatic rings of the benzoyl chlorides fit better to
the size of the α- and β-CD cavities with the consequent expul-
sion of a greater number of water molecules. This expulsion
causes the interior of the cavities of the complexed α- and β-CD
to be more hydrophobic than in the case of γ-CD and con-
1
X
sequently the rate constant kCD will be very small, confirming
X
w
X
X
the inequality k
ӷ kCD
K
CD[CD].
1
7 T. W. Bentley and R. O. Jones, J. Chem. Soc., Perkin Trans. 2, 1993,
351.
8 H. J. Schneider, F. Hacket and V. Rüdiger, Chem. Rev., 1998, 98,
2
1
Conclusions
1
755.
X
The studies carried out allow us to conclude that the solvolysis
of benzoyl chlorides in the presence of cyclodextrins shows two
clearly differentiated behaviors: when the solvolysis mechanism
occurs through an associative path the presence of cyclo-
dextrins catalyzes the process through the reaction with its
hydroxyl group C(6). The substituted benzoyl chlorides with
electron-donating groups, which undergo solvolysis through a
dissociative path, show a reduction of the rate constant caused
by the presence of the cyclodextrins. This behavior is due to
the complexation of the substrates with the cyclodextrins and
also to the limited value of the rate constant in the cavity of
the cyclodextrin, due to its low capacity to solvate the leaving
19 For 3-CF
3
, 3-NO
2
and 4-NO
2
only a maximum value of KCD is
obtained, due to the fact that it is not possible to obtain it experi-
mentally. The maximum values have been calculated, confirming
X
the inequality 1 ӷ KCD[β-CD] for the maximum concentration of
cyclodextrin used in the kinetic experiments. By using the maximum
X
X
value of KCD a minimum value of kCD can be computed.
20 Y. Matsui and K. Mochida, Bull. Chem. Soc. Jpn., 1979, 52, 2808.
21 Y. Matsui, T. Nishioka and T. Fujita, Top. Curr. Chem., 1985, 128, 61.
2 C. Hansch and A. Leo, Substituent Constant for Correlation Analysis
in Chemistry and Biology, John Wiley & Sons, New York, 1979.
3 L. García-Río, J. R. Leis and J. A. Moreira, J. Am. Chem. Soc., 2000,
2
2
2
1
22, 10325.
4 D. Hallen, A. Schön, I. Shehatta and I. Wadsö, J. Chem. Soc.,
Faraday Trans., 1992, 88, 2859.
Ϫ
group, Cl .
25 W. Saenger, Angew. Chem., Int. Ed. Engl., 1980, 19, 344.
6 I. Sanemasa, T. Osajima and T. Deguchi, Bull. Chem. Soc. Jpn.,
2
The presence of α- and γ-CD causes alterations in behavior
which are related to: (i) the stability of the substrate–cyclo-
dextrin complexes: it causes the solvolysis rate of the benzoyl
1
990, 63, 2814.
2
7 R. L. Van Etten, J. F. Sebastian, G. A. Clowes and M. L. Bender,
J. Am. Chem. Soc., 1967, 89, 3242.
chlorides 4-NO , 3-NO and 3-CF to be insensitive to the
2
2
3
28 B. Uno, N. Kaida, T. Kawakita, K. Kano and T. Kubota, Chem.
Pharm. Bull., 1988, 36, 3753.
presence of α-CD; (ii) the stoichiometry of the complexes,
causing the formation of complexes α-CD : benzoyl chloride
with 1 : 1 stoichiometries for those which are substituted in
the meta position and 2 : 1 for those substituted in the para
position; (iii) the displacement of water in the interior of
the cavity, causing the solvolysis reaction to be detected in the
interior of the cavity of γ-CD when meta-substituted benzoyl
chlorides are used.
29 N. J. Turro, T. Okubo and C. J. Chung, J. Am. Chem. Soc., 1982, 104,
3
953.
3
0 G. S. Cox, N. J. Turro, N. C. Yang and M. J. Chem., J. Am. Chem.
Soc., 1984, 106, 422.
1 S. Hamai, J. Phys. Chem., 1990, 94, 2595.
2 V. Ramamurthy and D. F. Eaton, Acc. Chem. Res., 1988, 21, 300.
33 G. S. Cox, P. J. Hauptmann and N. Turro, J. Photochem. Photobiol.,
1984, 39, 597.
4 T. W. Bentley, G. E. Carter and H. C. Harris, J. Chem. Soc., Perkin
Trans. 2, 1985, 983.
3
3
3
Acknowledgements
35 T. W. Bentley and I. S. Koo, J. Chem. Soc., Perkin Trans. 2, 1989,
385.
36 J. Lehmann, E. Kleinpeter and J. Krechl, J. Inclusion Phenom., 1991,
10, 233.
1
Financial support from the Xunta de Galicia (PGIDT00PX-
I20907PR and PGIDT03-PXIC20905PN) and Ministerio de
Ciencia y Tecnología (Project BQU2002-01184) is gratefully
acknowledged.
3
7 M. V. Rekharsky, F. P. Schwarz, Y. B. Tewari, R. N. Goldberg,
M. Tanaka and Y. Yamashoji, J. Phys. Chem., 1994, 98, 4098.
8 M. V. Rekharsky and Y. Inoue, Chem. Rev., 1998, 98, 1875.
9 P. C. Manor and W. Saenger, J. Am. Chem. Soc., 1974, 96, 3630.
0 K. Lindner and W. Saenger, Acta Crystallogr., Sect. B, 1982, 38, 203.
1 K. Lindner and W. Saenger, Angew. Chem., Int. Ed. Engl., 1978, 17,
694.
3
3
4
4
References
1
2
B. D. Song and W. P. Jencks, J. Am. Chem. Soc., 1989, 111, 8470.
(a) W. P. Jencks, Chem. Rev., 1972, 72, 705; (b) R. A. More O’Ferrall,
J. Chem. Soc. B, 1970, 274.
42 H. Schlenk and D. M. Sand, J. Am. Chem. Soc., 1961, 83, 2312.
43 K. Harata, Chem. Lett., 1984, 641.
3
4
J. Szejtli, Cyclodextrin Technology, Kluwer Academic Publishers,
Dordrecht, 1988.
M. L. Bender and M. Komiyama, Cyclodextrin Chemistry, Springer-
Verlag, New York, 1977.
44 K. Harata, Bull. Chem. Soc. Jpn., 1987, 60, 2763.
45 S. J. Heyes, N. J. Clayden and C. M. Dobson, Carbohydr. Res., 1992,
233, 1.
46 M. G. Usha and R. J. Wittebort, J. Am. Chem. Soc., 1992, 114, 1541.
47 G. Barone, G. Castronuovo, P. Del Vecchio, V. Elia and
M. Muscetta, J. Chem. Soc., Faraday Trans. 1, 1986, 82, 2089.
48 H. Fujiwara, H. Arakawa, S. Murata and Y. Sasaki, Bull. Chem. Soc.
Jpn., 1987, 60, 3891.
5
6
7
K. Takahashi, Chem. Rev., 1998, 98, 2013.
R. Beslow and S. D. Dong, Chem. Rev., 1998, 98, 1997.
(a) A. V. Veglia and R. H. de Rossi, J. Org. Chem., 1988, 53, 5281;
(
b) M. Barra and R. H. de Rossi, J. Org. Chem., 1989, 54, 5020;
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 8 6 – 1 1 9 3
1193