Effect of cyclodextrin on elimination reactions

Article abstract of DOI:10.1139/v99-085

The reaction of 1-bromo-2-X-2-(Y-phenyl) ethane derivatives (1: X = Y = H; 2: X = Ph, Y = H; 3: X = H, Y = 4-Ac; 4: X = H, Y = 3-NO₂; 5: X = H, Y = 4-NO₂; 6: X = H, Y = 3-Me; 7: X = H, Y = 4-Me) in basic solution was studied, and in most cases, only the elimination product is formed. Only (2-bromo-1-phenylethyl)benzene, 2, yielded significant substitution product, and this yield decreased with the concentration of HO^-. Addition of cyclodextrin (β-CD) diminished (about half for 0.02 M cyclodextrin concentration) the reaction rate of all substrates but 4 and 5. In the latter two cases, the rate rises. The observed rate-constant value at 0.5 M NaOH is 6.78 × 10⁴ s^-1 (at 40°C) and 1.80 × 10^-3 s^-1 (at 25°C) for 4 and 5, respectively. Under the same reaction conditions but with 0.01 M β-CD, the corresponding rates were 7.70 × 10^-4 s^-1 and 5.20 × 10^-3 s^-1. The elimination yield for 2 increased from 64 to 98% when the β-CD changed from zero to 0.02 M at 0.5 M NaHO. Also, there was an increase in the relative elimination products of 20-40% for compounds 6 and 7. The Hammet ρ values were 1.3 and 2.3 for the reaction in pure solvent and in the presence of β-cyclodextrin, indicating an increase in the negative character of the transition state for the reactions in the latter conditions. The results are interpreted in terms of the formation of an inclusion complex whose structure depends on the substrate.

Full text of DOI:10.1139/v99-085

860
Effect of cyclodextrin on elimination reactions
Luis Viola and Rita H. de Rossi
Abstract: The reaction of 1-bromo-2-X-2-(Y-phenyl) ethane derivatives (1: X = Y = H; 2: X = Ph, Y = H; 3: X = H,
Y = 4-Ac; 4: X = H, Y = 3-NO2; 5: X = H, Y = 4-NO2; 6: X = H, Y = 3-Me; 7: X = H, Y = 4-Me) in basic solution
was studied, and in most cases, only the elimination product is formed. Only (2-bromo-1-phenylethyl)benzene, 2,
yielded significant substitution product, and this yield decreased with the concentration of HO–. Addition of
cyclodextrin (β-CD) diminished (about half for 0.02 M cyclodextrin concentration) the reaction rate of all substrates
but 4 and 5. In the latter two cases, the rate rises. The observed rate-constant value at 0.5 M NaOH is 6.78 × 10–4 s–1
(at 40°C) and 1.80 × 10–3 s–1 (at 25°C) for 4 and 5, respectively. Under the same reaction conditions but with 0.01 M
β-CD, the corresponding rates were 7.70 × 10–4 s–1 and 5.20 × 10–3 s–1. The elimination yield for 2 increased from 64
to 98% when the β-CD changed from zero to 0.02 M at 0.5 M NaHO. Also, there was an increase in the relative
elimination products of 20–40% for compounds 6 and 7. The Hammet ρ values were 1.3 and 2.3 for the reaction in
pure solvent and in the presence of β-cyclodextrin, indicating an increase in the negative character of the transition
state for the reactions in the latter conditions. The results are interpreted in terms of the formation of an inclusion
complex whose structure depends on the substrate.
Key words: cyclodextrin, elimination reactions, inhibition, catalysis.
Résumé : On a étudié la réaction de dérivés du 1-bromo-2-X-2-(Y-phényl)éthane (1, X = Y = H; 2, X = Ph, Y = H; 3,
X = H, Y = 4-Ac; 4, X = H, Y = 3-NO2; 5, X = H, Y = 4-NO2; 6, X = H, Y = 3-Me; 7, X = H, Y = 4-Me) en
solution basique et, dans la plupart des cas, il ne se forme que le produit d’élimination. Le (2-bromo-1-
phényléthyl)benzène, 2, est le seul à donner des quantités significatives de produit de substitution et ce rendement
diminue avec une augmentation de la concentration de OH–. L’addition de cyclodextrine (β-CD) diminue (environ de
moitié pour une concentration égale à 0,02 M en cyclodextrine) la vitesse de réaction pour chacun des substrats, à
l’exception des produits 4 et 5. Dans ces deux derniers cas, la vitesse augmente. Les valeurs des constantes de vitesse
observée, pour une concentration de 0,5 M de NaOH, sont de 6,78 × 10–4 s–1 (à 40°C) et de 1,80 × 10–3 s–1 (à 25°C)
pour les composés 4 et 5 respectivement. Dans les mêmes conditions de réaction, mais en présence de 0,01 M de β-
CD, les vitesses correspondantes sont de 7,70 × 10–4 s–1 et 5,20 × 10–3 s–1. Pour une concentration de NaOH égale à
0,5 M, le rendement en élimination pour le produit 2 augmente de 64 à 98% lorsque la concentration en β-CD passe
de zéro à 0,02 M. Avec les composés 6 et 7, on a aussi observé une augmentation des produits d’élimination de 20 à
40%. Les valeurs ρ de Hammett sont égales respectivement à 1,3 et 2,3 pour les réactions dans un solvant pur et celles
en présence de β-cyclodextrine; ces valeurs signalent une augmentation du caractère négatif de l’état de transition pour
les réactions dans ces dernières conditions. On interprète les résultats en fonction de la formation d’un complexe
d’inclusion dont la structure dépend du substrat.
Mots clés : cyclodextrine, réactions d’élimination, inhibition, catalyse.
[Traduit par la Rédaction] Viola and de Rossi 867
Introduction
They have a cavity formed by two rings of CH group of the
glucose units and one ring of the glycosidic oxygens. The
secondary OH groups are located in the wider rim of the
cavity and the primary in the smaller one. Since the cavity is
relatively apolar compared with water, it provides a favor-
able environment for the location of many organic com-
pounds, and the formation of inclusion complexes leads to
many interesting changes in the properties of the included
compound which result in changes in their reactivity (2).
One of the properties of cyclodextrins that has been ex-
ploited considerably is the possibility of induction of selec-
tivity in reactions that lead to more than one product.
Inclusion of the substrate and (or) the reagent in the cavity
leads to different distribution of products when the reactions
are compared with the same reaction in the bulk solvent. For
example, the iodination of phenol (3) and the photochemical
rearrangement of phenyl acetate give an increase in the
para/ortho and ortho/para ratio of products, respectively, in
Cyclodextrins are cyclic oligomers of α-D-glucose that are
produced by enzymatic degradation of starch (1). The more
common cyclodextrins have six, seven, and eight glucose
units and are called α, β, and γ cyclodextrins, respectively.
Received November 4, 1998.
This paper is dedicated to Jerry Kresge in recognition of his
many achievements in chemistry.
L. Viola and R.H. de Rossi.1 Instituto de Investigaciones en
Físico Química de Córdoba, Facultad de Ciencias Químicas,
Departamento de Química Orgánica, Universidad Nacional de
Córdoba, Ciudad Universitaria, 5000 Córdoba, Argentina.
1Author to whom correspondence may be addressed.
Telephone: 54-351-4334170. Fax: 54-351-4333030.
e-mail: ritah@dqo.fcq.unc.edu.ar
Can. J. Chem. 77: 860–867 (1999)
© 1999 NRC Canada

Viola and de Rossi
861
Fig. 1. kobs vs. β-CD concentration for the reaction of 1. [HO] =
0.5 M. Temperature, 40°C; solvent, ethanol/water (1:9 v/v); ionic
strength, 0.5 M (NaCl as compensating electrolyte). The line was
calculated fitting the data to eq. [3].
the presence of β-CD (4). More recent examples that are im-
portant for synthetic applications can be found in the litera-
ture (5). The changes in product selectivity may be the result
of a different stabilization of the transition state for the for-
mation of the various products.
The elimination mechanism of 2-bromoethylbenzenes is
well established, and it proceeds by an E2 type of mecha-
nism. The transition state for this mechanism has severe
stereoelectronic requirements; moreover, competition by
substitution often occurs (6). We considered of interest to
study this type of reactions in the presence of cyclodextrin
to determine the effect of inclusion on the reaction rates and
the elimination/substitution ratio. Thus, we studied the reac-
tion of compounds 1–7 in basic water solution and the effect
of cyclodextrins on their relative yield of elimination and
substitution in the cases were the two reactions compete.
Compounds with substituents at 3- or 4-position were cho-
sen because it is known that the orientation of the ring in the
inclusion complex is quite dependent on the position of the
substituent on the ring (7).
Results
(2-Bromoethyl)benzene (1)
In ethanol/water 1:9 v/v, the only product formed was sty-
rene, and the observed pseudo-first-order rate constant is lin-
early dependent of HO– concentration (Table S1).2 The slope
of a plot of kobs vs. HO– concentration (not shown) gave the
second-order rate constant 2.49 × 10–4 M–1 s–1 in good
agreement with the literature values, i.e., 3.01 × 10–4 M–1 s–1
(11). At constant HO– concentration, the reaction was inhib-
ited by the addition of β-CD and showed a saturation effect
(Fig. 1). GC analysis of the products showed that the major
product formed was styrene, although a small (<5%) amount
of the substitution product was also formed but no signifi-
cant changes in the relative yield of the elimination and sub-
stitution products could be detected.
Substrate 1 is very insoluble in water, so dioxane/water
1:9 was used as solvent. Under all our reaction conditions,
the only product formed was styrene. The addition of 1.5 ×
10–2 M of β-CD produced no changes in the rate. The ob-
served rate constants at 0.5 M OH– are 1.27 × 10–4 and 1.24
× 10–4 s–1, respectively. The fact that the rate was not af-
fected by β-CD can be attributed to competitive binding by
the solvent. Since it is known that dioxane forms inclusion
complexes with β-CD (KAss = 9 M–1) (8), and the association
constant of 1 with β-CD should be in the order of 200 M–1
(see below), only 10% of the substrate would be bound at
the maximum β-CD used.
Ethanol also binds to β-CD but the association equilib-
rium constant is lower, i.e., 1.2 M–1 (9); besides, its smaller
size may allow the co-inclusion of ethanol and the substrate
as was found in other systems (10).
At constant β-CD concentration (0.01 M), there was linear
dependence of kobs on HO–, and the slope was 1.49 × 10–4
M–1 s–1, which is about 60% of the value in the absence of
2 Copies of material on deposit may be purchased from: The Depository of Unpublished Data, Document Delivery, CISTI, National Research
Council Canada, Ottawa K1A 0S2, Canada.
© 1999 NRC Canada

862
Can. J. Chem. Vol. 77, 1999
Fig. 2. kobs vs. β-CD concentration for the reaction of 2. [HO] =
0.5 M. Temperature, 40°C; solvent, ethanol/water (1:9 v/v); ionic
strength, 0.5 M (NaCl as compensating electrolyte). The line was
calculated fitting the data to eq. [3].
1-(2-Bromoethyl)-4-acetylbenzene (3)
The second-order rate constant obtained from the slope of
the linear plot (not shown) of kobs vs. HO– concentration
(Table S1)2 is 3.3 × 10–3 M–1 s–1. There is a modest (-20%)
but out of experimental error decrease in the observed rate
in the presence of β-CD. The ratio of elimination/substitu-
tion products is not affected (Table S2).2
1-(2-Bromoethyl)-3-nitrobenzene (4) and 1-(2-
bromoethyl)-4- nitrobenzene (5)
The reactions of 4 were measured at 40°C and those of 5
at 25°C. There is linear dependence of the observed rate
constant with the HO– concentration (Table S1)2 from which
a second-order rate constant of 1.33 × 10–3 and (3.65 ± 0.17)
× 10–3 M–1 s–1 are calculated, respectively, for 4 and 5. The
second-order rate constants for 5 is somewhat higher than
the reported values 2.61 × 10–3 (12) and 3.47 × 10–3 M–1 s–1
(11). The discrepancy may be attributed to the presence of
ethanol.
In contrast with the behavior of the substrates described
before (1–3), the addition of cyclodextrin produces an in-
crease in the observed rate constant for both substrates, al-
though for 4 it is at most 10%, with α- or β-cyclodextrin at
0.01 M (Table S2).2 On the other hand, the increase is more
significant for 5, and saturation effect is observed at three
concentrations of HO– (Fig. 3). The relative increase in rate
is higher at lower HO– concentration, i.e., the ratio of the
rate at β-CD = 0.015 M and zero is 3.2, 12.7, and 22.2 for
HO– concentrations 0.49, 0.04, and 0.01 M, respectively. In
all cases, only the elimination product is formed.
Table 1. Effect of β-CD on the observed rate constant and the
elimination yield for the reaction of 2 in basic solution.a
β-CD (10–3 M)
kobs (10–5 M)
Elimination yield (%)
0
1.0
3.0
5.0
24.8
20.3
17.9
14.6
14.2
13.0 (26.2)b
13.2
64
75
79
78
85
79
83
98
At constant β-CD concentration, there is nonlinear de-
pendence of the observed rate constant for 5 with HO– con-
centration (Fig. 4).
7.0
10.0
15.0
20.0
The addition of 3 × 10–2 M concentration of α- or γ-
cyclodextrin with HO– = 0.5 M gives rate constants 3.28 ×
10–3 and 3.08 × 10–3 s–1, indicating that both receptors accel-
erate the reaction in approximately the same amount but less
than β-CD, since the value of the observed rate constant in
the presence of 0.015 M β-CD is 5.75 × 10–3 s–1.
11.7
aTemperature, 25°C; solvent, ethanol/water (1:9 v/v); ionic strength, 0.5
M (NaCl as compensating electrolyte) unless otherwise indicated; NaOH,
0.5 M. Substrate concentration 2 × 10–5 M.
bSoluble starch instead of cyclodextrin.
1-(2-Bromoethyl)-3-methylbenzene (6) and 1-(2-
bromoethyl)-4-methylbenzene (7)
β-CD. On the other hand, the addition of α-CD did not af-
fect the rate in the range 1 × 10–4–2.5 × 10–2 M.
The reactions of these substrates were studied at 40°C,
there is first-order dependence on the HO– concentration
(Table S1) and the second-order rate constant obtained from
the slope of the linear plot (not shown) is 2.23 × 10–4 and
1.84 × 10–4 M–1 s–1 for 6 and 7, respectively. The addition of
β-CD produces a decrease in the rate and an increase in the
relative elimination yield (see Experimental) of about 20%
for 6 and 40% for 7 at the higher cyclodextrin concentration
used (Table S2). The decrease in the yield of substitution
product was confirmed by HPLC.
(2-Bromoethyl-1-phenylethyl)benzene (2)
This reaction gave substitution as well as elimination
products, although the latter is predominant. The concentra-
tion of the elimination products increased with that of HO–
in agreement with the known effect of base concentration on
the competition between substitution and elimination reac-
tions (6). The observed pseudo-first-order rate is linearly de-
pendent on HO– concentration with a second-order rate
constant of 4.95 × 10–4 M–1 s–1 (Table S1).2 The addition of
β-CD produces a decrease on the observed rate and shows
saturation effect (Fig. 2, Table 1). The yield of the elimina-
tion product increases as the amount of CD increases (Ta-
ble S2).2 At constant β-CD concentration, there is an
increase in the observed rate with HO– concentration, which
is linear but at low concentration the values are somewhat
smaller than expected. The elimination yield is higher than
in the absence of CD and increases as the OH concentration
increases but in a somewhat scattered form (Table S1).2
Discussion
Effect of cyclodextrin on reaction rates
All the reactions studied were affected by the addition of
β-CD. To determine whether the effect was due to a medium
effect, the observed rate constants were determined in the
presence of soluble starch in similar weight concentration as
that of β-CD. In Table 2, the data are collected, and it can be
© 1999 NRC Canada

Viola and de Rossi
863
Fig. 4. Change in kobs with HO– concentration for the reaction of
5. [β-CD] = 0.01 M. Temperature, 25°C; solvent, ethanol/water
(1:9 v/v); ionic strength, 0.5 M (NaCl as compensating
electrolyte). The line was calculated using eq. [9].
Fig. 3. kobs vs. β-CD concentration for the reaction of 5.
Temperature, 25°C; solvent, ethanol/water (1:9 v/v); ionic
strength, 0.5 M (NaCl as compensating electrolyte). [HO] = 0.5
M, A; 0.04 M, B; and 0.01 M, C. The line was calculated
adjusting the data to eq. [4].
complex, there are two possible modes of inclusion, with the
aliphatic chain inside the cavity (Fig. 5A) and with the aro-
matic part of the molecule inside (Fig. 5B).
The pKa of β-CD is 12.2 (13), so we suggest that the reac-
tion takes place as indicated in Scheme 1 where k0, k1, and
k2 represent the reactions of the free substrate, the substrate
complexed with neutral β-CD (CDOH), and the reaction
with anionized β-CD (CDO–), respectively.
The observed rate constant for Scheme 1 is given by
eq. [1] where f represents the fraction of ionized β-CD as de-
fined in eq. [2]. Under the conditions of constant HO– con-
centration, 0.5 M, f ≈ 1, then eq. [1] simplifies to eq. [3]:
k0[HO−] + k1K1[HO−](1 − f)[CD] + k2K2 f[CD]
[1]
[2]
[3]
kobs
f =
kobs
=
1 + K1(1 − f)[CD] + K2 f[CD]
[HO− ]
[HO− ] + Kb
k0[HO− ] + k2K2[CD]
1 + K2[CD]
=
Fitting the experimental data to eq. [3],3 k2 and K2 were
calculated, and the data are collected in Table 3.
The reactions of substrate 5 in the presence of β-CD were
studied at three HO– concentrations, i.e., 0.01, 0.04, and
0.49 M. For each HO– concentration, the observed rate con-
stant was fitted to an equation of the form of eq. [4] with a =
k0[HO–] and b and c adjustable parameters given in eqs. [5]
and [6]. The calculated parameters are given in Table 4.
seen that within experimental error, starch produces no
changes in the observed rate constant, which indicates that
the effect of β-CD should be attributed to specific interac-
tions. Besides, the addition of β-CD produces a decrease in
the absorption at the maximum wavelength for 1-(2-bromo-
ethyl)-4-acetylbenzene and 1-(2-bromoethyl)-4-nitrobenzene.
These results indicate that the substrate may form some type
of inclusion complex with the β-CD. For the 1:1 inclusion
a + b[β −CD]
[4]
kobs =
1 + c[β −CD]
3 Sigmaplot, DOS version 6.0. Handel Scientific. 1986–1993.
© 1999 NRC Canada

864
Can. J. Chem. Vol. 77, 1999
Table 2. Effect of β-CD and soluble starch on the pseudo-first-order observed rate constant for the
reaction of compounds 1–7 with hydroxide ion.a
kobs (10–4 s–1)b
kobs (10–4 s–1)c
kobs
(10–4 s–1)d
CD
Alm
Substrate
1
1.22
2.48
16.2
6.78
18.0
1.14
0.94
0.82
1.30
13.0
7.58
52.0
0.91
0.65
1.26
2.62
2e
3
4
5e
6
21.0
1.18
0.96
7
aTemperature, 40°C; unless otherwise indicated, solvent, ethanol/water (1:9 v/v); ionic strength, 0.5 M (NaCl as
compensating electrolyte); NaOH, 0.5 M.
bObserved rate constant.
cObserved rate constant in the presence of 0.01 M β-CD.
dObserved rate constant in the presence of 11.35 mg/mL of soluble starch.
eTemperature, 25°C.
Fig. 5. Schematic representation of the inclusion complexes of
compounds 1–7.
Scheme 1.
kobs(1 + K1(1 − f)[CD] + K2 f[CD]) − k0[HO− ]
[5]
[6]
b = (k1K1[HO–](1 – f) + k2K2 f )[CD]
= (k1K1Kb + k2K2) f [CD]
[8]
f
= (k1K1Kb + k2K2)[CD]
c = K1(1 – f) + K2 f
Since K1 and K2 were calculated as indicated above, the left
hand side of eq. [8] can be obtained for each concentration
of HO– using the data at constant β-CD and varying HO–
(Table S1);2 these values were calculated and are not con-
stant as eq. [8] requires but they show a linear dependence
with HO– concentration with slope 3.86 × 10–3 and intercept
4.87 × 10–3 (plot not shown). From the last two values, k3K2
and k1K1Kb + k2K2 can be calculated and are in fair agree-
ment with the values obtained from another set of data,
namely, the change in rate with β-CD at constant HO– (see
above). From the results just discussed, it follows that a
pathway involving the reaction of the substrate complexed
with cyclodextrin anion and HO– (eq. [7]) is playing some
role in the catalyzed pathway. The observed rate constant is
then given by eq. [9]:
The decrease in c as the HO– concentration increases indi-
cates that the association constant of cyclodextrin anion with
the substrate, K2, is smaller than the corresponding value of
the neutral cyclodextrin, K1. Using the value of c at 0.5 M
HO– as K2, K1 can be calculated from the value of c at 0.01
and 0.04 M giving 370 ± 10 M–1. The parameter b divided
by f should be constant (see eq. [5]). It can be seen in the
last column of Table 4 that similar values are obtained at the
two lower concentrations of HO–, but the value increases
considerably at high HO– concentration. This might indicate
that at this high concentration there is some contribution to
the overall catalyzed reaction from a pathway involving the
substrate complexed with cyclodextrin anion and HO–
(eq. [7]).
[9]
kobs
k3
[7]
SCDO– + HO–
→ P
k0[HO− ] + (k1K1Kb + k2K2 + k3K2[HO− ]) f[CD]
1 + K1(1 − f)[CD] + K2[CD] f
=
If reaction [7] takes place, a term k3K2[HO–] should be in-
cluded in parameter c. Using the average value of b/f for the
two lowest concentrations of HO–, namely, 0.769, k3K2 is
calculated as 0.777 M–1 s–1.
The value of k3K2 is of the same order of magnitude as
that corresponding to the reaction involving the cyclodextrin
anion acting as a base, k2K2, so it is likely that k1K1 is also
Equation [1] can be rearranged as shown in eq. [8]:
© 1999 NRC Canada

Viola and de Rossi
Table 3. Rate and equilibrium constants for the substrates 1–7 and transition state and changes in Gibbs free energy.
865
Substrate
k2 (10–4 s–1)
K2 (M–1)
–∆Go (kcal M–1)
K*TS (M–1)
–∆G*o (kcal M–1)
1a
2b
3a
5b
6a
7a
58
184
380
292
156
95
324
351
353
299
283
346
87
277
298
334
384
250
310
100
120
760
67
153
216
658
56
53
261
147
aTemperature, 40°C.
bTemperature, 25°C.
may function as simply basic catalyst towards acidic sub-
strates included in their cavity (18).
Table 4. Parameters of eq. [4] calculated with the data of
compound 5.
Considering the less polar microenvironment of the cavity
of β-CD, and that elimination reactions are faster in less po-
lar solvents, one would expect that all elimination reaction
were catalyzed by cyclodextrin. Contrary to the above ex-
pectations, all the reactions studied except for compounds 4
and 5 are retarded.
b
c
HO– (M)
b/f
0.311 ± 0.034
0.522 ± 0.035
1.167 ± 0.099
304 ± 46
216 ± 21
154 ± 19
0.01
0.04
0.49
0.808
0.731
1.150
There is no correlation between K*TS and KS, which is an
evidence that the transition state of the catalyzed or retarded
reaction does not correspond to the inclusion complex of the
substrate. There are two possible types of 1/1 inclusion com-
plex: one with the aromatic ring inside the cavity and an-
other with the aliphatic chain inside the cavity (Fig. 5).
Since all the compounds have the same aliphatic chain, the
differences in stability of the two types of complexes should
depend on the nature and position of the substituent. The
compounds having nitro and acetyl groups are those that
have the more hydrophilic moieties, therefore, it is more
likely that the major type of interaction is the one repre-
sented by A in Fig. 5.
When the transition state for the included compound is
more stable than that for the free compound, the reaction
will be accelerated by cyclodextrin and vice versa (eq. [10]).
When the stabilization of the included transition state is
higher than that of the substrate, the barrier for the reaction
will be lower and consequently the overall rate faster.
The affinity for the substrate is stronger than that for the
transition state for all the substrates but 5, consequently, the
reaction is catalyzed as shown in Table 3. This can be asso-
ciated with a change in the nature of the transition state for
this compounds. Several authors (19) have proposed that in
these systems the increased withdrawing effect of the
substituent changes the transition state from a classic E2
mechanism towards one with some E1cb character.
The second-order rate constant can be correlated by the
Hammet equation, and a value of 1.3 and 2.3 is obtained for
k0 and k2, respectively, indicating an increase in the car-
banionic character of the transition state for the reactions in
the presence of CD. The mechanism of the catalyzed step is
kinetically indistinguishable from one in which the substrate
complexed with neutral cyclodextrin reacts with HO– (K1k1
step in Scheme 1). Then, the increase in the carbanionic
character of the transition state may indicate that the base,
HO–, is stronger in the cavity. However, for this pathway to
be kinetically significant under our reaction conditions, K1k1
should be about 100 times higher than k2K2, which seems
unlikely considering the values that have been estimated for
these two rate constants for compound 5, so the increase in
similar. Therefore, in the sum the term k1K1Kb is probably
negligible.
Following Kurz treatment (14), which was adapted by Tee
(15) to reactions catalyzed by cyclodextrins, we calculated
the formal equilibrium constant for the association of the
transition state (TS) with β-CD, and the corresponding free
energy using eq. [10]:
[TS.CDO− ]
k2K2
[10] K*TS
=
=
[TS][CDO− ] k0[HO− ]
K*TS is an apparent association constant for the formation of
the transition state of the CD-mediated reaction (symbolized
as TS.CDO–) from the transition state of the normal reaction
(TS) and the β-CD. The logarithm of this quantity is related
to the differences in free energy of the transition state of the
catalyzed and uncatalyzed reactions under standard condi-
tions. The value of K*TS is obtained from experimentally
measured quantities, so it gives a measure of the binding en-
ergy of the transition state to the catalyst for those condi-
tions regardeless of the mechanism. (16) Therefore, K*TS and
its variation can be taken as a criterium of mechanism in
similar way as other kinetic parameters such as Bronsted β
or Hammett ρ values. According to eq. [10], the change in
rate is determined by the strength of binding of the transition
state relative to that of the substrate, i.e., k2͞k0 = K*TS͞K2.
Cyclodextrins may show two basic forms of catalysis:
“noncovalent” and “covalent” (1). In the former case, the
cyclodextrin binds to the substrate (S) and provides an envi-
ronment for the reaction that is different from the bulk sol-
vent; in this case, the effect of CD arises from the less polar
nature of the cavity (a microsolvent effect) and (or) from the
conformational restrains imposed on the substrate by the ge-
ometry of inclusion. Another factor that may be important
for noncovalent catalysis may be differential solvation ef-
fects at the interface of the cyclodextrin cavity with the ex-
ternal aqueous enviroment (17). In the covalent catalysis,
there is some type of covalent interactions between the sub-
strate (S) and functional groups on the CD in the rate-
limiting step of the reaction. For instance, the anion of CD
© 1999 NRC Canada

866
Can. J. Chem. Vol. 77, 1999
carbanionic character of the included transition state may be
attributed to esteric constrain in the cavity for a truly E2 type
of mechanism.
Similar effect was observed by Tagaky and co-workers
(20) for elimination reactions in micelles. They also found
an important increase in the ρ value, which was attributed to
a change of the transition state toward a more negative char-
acter although the results of isotopic exchange indicated that
pared from (2-bromoethyl)benzene and acetyl chloride and
AlCl3 in S2C (22). The purity of the compound was con-
trolled by mass spectrometry, NMR, IR, and thin layer
chromatograpy.
1-(2-Bromoethyl)-3-nitrobenzene was obtained from
(3-nitrophenyl)ethanol (Sigma) and PBr3 in benzene (23).
The product was identified by NMR.
1-(2-Bromoethyl)-4-nitrobenzene was obtained from
(2-bromoethyl)benzene and nitric acid in acetic anhydride as
described in the literature (22). 1-(2-Bromoethyl)-3-methyl-
benzene was obtained according to the literature from
(3-methylphenyl)ethanol and PBr3 in benzene solution (23).
The product had 5% of the starting alcohol as determined by
NMR. 1-(2-Bromoethyl)-4-methylbenzene was obtained as
described in the literature (23) from (4-methylphenyl)etha-
nol and PBr3 in benzene. NMR analysis indicated that it
contains 20% of the starting alcohol.
the mechanism was not E1cb
.
Comparing the values of K2 and K*TS for substrates 6 and 7
(Table 3), which have the same substituent but in different
position, it can be seen that the meta substituted compound
has a significatively lower equilibrium constant than the
para. This result indicates that the inclusion of the substrate
takes place with the aromatic ring inside the cavity, which is
characteristic of benzene derivatives (21). The unfavorable
forces that lead to a weaker interaction of the meta substi-
tuted compound remain in the transition state for the reac-
tion with β-CD because the ratio of K2 and K*TS for 6 and 7
are about the same, namely, 0.36 and 0.37, respectively.
Kinetic procedures
Reactions were initiated by adding 10 µL of a concen-
trated solution of the substrate dissolved in ethanol to 3 mL
of a water solution containing all the other constituents. The
cell was flashed with nitrogen and stopped. The change in
absorption with time was measured at the maximum of the
corresponding styrene derivative, namely, styrene 248 nm,
diphenylethene 247 nm, p-acetylstyrene 282 nm, m-nitrostyrene
241 nm, p-nitrostyrene 308 nm, m-methylstyrene 250 nm,
p-methyl styrene 249 nm. The yield of products for (2-
bromoethyl)- and (2-bromo-1-phenylethyl)benzene was de-
termined from eq. [11]:
Effect of β-CD on the reaction selectivity
In the presence of β-CD, the relative yield of elimination
(see Experimental) increases for substrates 6 and 7 by 20
and 39%, respectively. In the case of substrate 2, the elimi-
nation yield increases from 64 to 98% when β-CD increases
from 0 to 0.02 M at constant HO– concentration (Table S2).
The free activation energy of the elimination reaction for
substrates 1 and 3–5 is lower than that of the substitution re-
action, and it is possible that β-CD changes the relative en-
ergy of these two transition states, however, the change is
not big enough, and the substitution product is not formed in
detectable amount. On the other hand, the compounds hav-
ing electron donor substituents produce a decrease in the rel-
ative acidity of the hydrogen of the β-carbon increasing the
free energy for the elimination reaction, but the substitution
is less affected, therefore, the two reactions compete, and the
effect of β-CD can be experimentally determined.
(A∞− Ao)100
[11] % elimination =
ε[S]o
where A∞ and Ao represent the initial and final absorbance at
the wavelength of the measurement, S represents the initial
substrate concentration, and ε the extinction coefficient of
the product.
For all the other compounds, the relative elimination
yields (REY) was determined by using eq. [12] were the
subscript x and y stand for the reactions in the presence and
absence of cyclodextrin, respectively.
The relative yield of elimination products increase with
the β-CD concentration, which may be a consequence of
microsolvent effect, since elimination reaction are favored
over substitution in solvents of low polarity.
Experimental section
(A∞ − Ao)x[S]oy
%Ex
[12] REY =
=
%Ey (A∞− Ao)y[S]ox
The kinetic measurements were carried out on the thermo-
stated cell of a Beckman 24 spectrophotometer connected to
a personal computer that was used to acquire and process the
kinetic data.
Product analysis
UV-vis spectra were done on a Shimadzu 260 spectro-
photometer. RMN were done on a Bruker AD-C 200.
Water was purified in a Millipore aparatus. Dioxane was
dried with Na metal and distilled before use over lithium
aluminum hydride. Ethanol was distilled and the purity con-
trolled by UV-vis spectrometry. The α- and γ-cyclodextrin
(Aldrich), β-cyclodextrin (Roquette),4 and soluble starch
(Mallinkcrodt) were used as received. Inorganic salts were
analytical grade and (2-bromo-1-phenyl)ethylbenzene were
used as received. 4-Acetyl-1-(2-bromoethyl)benzene was pre-
The reactions of (2-bromoethyl)benzene, 1-(2-bromoethyl)-
3-methylbenzene, and 1-(2-bromoethyl)-4-methylbenzene
were analyzed by HPLC and CGL and the reaction of 1-(2-
bromoethyl)-3-nitrobenzene by H1 NMR using a MCH-5-n
cap column of 150 × 4 mm with UV-vis detection. The sol-
vent was water/methanol 4:6 v/v and 0.5 mL/min flow rate.
Retention times were 5.69, 21.94, and 24.32 for 2-phenyl
ethanol, styrene, and 1, respectively. GC analyses were car-
ried out in a Konik KNK 3000 HRGC with a column CPSIL
19 CV 6 m long and 0.75 mm diameter.
4 α- and β-cyclodextrin were generous gift of Ferromet SA, Buenos Aires, Argentina.
© 1999 NRC Canada

Viola and de Rossi
867
8. M. Barra. Ph.D. thesis. Facultad de Ciencias Químicas,
Universidad Nacional de Córdoba, Cordoba. 1988.
9. M.V. Rekharsky, F.P. Schwarz, Y.B. Tewari, and R.N.
Goldberg. J. Phys. Chem. 98, 10 282 (1994).
10. (a) O.S. Tee and M. Bozzi. J. Am. Chem. Soc. 112, 7815
(1990); (b) O.S. Tee, M. Bozzi, J.J. Hoeven, and T.A. Gadosy.
J. Am. Chem. Soc. 115, 8990 (1993).
Acknowledgements
This research was supported in part by the Consejo
Nacional de Investigaciones Científicas
(CONICET), Consejo de Investigaciones de Córdoba
(CONICOR), and Universidad Nacional de Córdoba,
Secretaría de Ciencia y Técnica (SECYT UNC). Luis Viola
was a grateful recipient of a fellowship from the CONICET.
y
Técnicas
11. S. Alumni and W.P. Jencks. J. Am. Chem. Soc. 102, 2052
(1980).
12. K.A. Wilk. J. Phys. Chem. 95, 3405 (1991).
13. (a) R.L. Van Etten, G.A. Clowes, J.F. Sebastian, and M.L.
Bender. J. Am. Chem. Soc. 89, 3253 (1967); (b) R.I. Gelb,
M.L. Schwartz, J.J. Bradshaw, and D.A. Laufer. Biorg. Chem.
9, 299 (1980).
14. J.L. Kurz. J. Am. Chem. Soc. 85, 987 (1963).
15. (a) O.S. Tee. Carbohydr. Res. 192, 181 (1989); (b) In Ad-
vances in physical organic chemistry. Academic Press Limited,
New York. 1994. pp. 1–85.
16. R.L. Schowen, In Transition stated in biochemical processes.
Edited by R.D. Gandour and R.L. Schowen. Plenun, New
York. Chap. 2.
17. O.S. Tee and J.M. Bennett. J. Am. Chem. Soc 110, 269 (1988).
18. V. Daffe and J. Fastrez. J. Chem. Soc. Perkin Trans 2, 789
(1983).
19. (a) J.R. Gandler, J.W. Storer, and D.A. Ohlberg. J. Am. Chem.
Soc. 112, 7756 (1990); (b) L.F. Blackwell, P.D. Buckley, K.W.
Jolley, and A.K.H. MacGibbon. J. Chem. Soc. Perkin Trans. 2,
169 (1973); (c) S. Alunni, E. Bacciocchi, P. Perrucci, and R.
Ruzziconi. J. Org. Chem. 43, 2414 (1978).
20. Y. Yano, Y. Yosida, A. Kurashima, Y. Tamura, and W. Tagaki.
J. Chem. Soc. Perkin Trans. 2, 1128 (1979).
21. (a) J.A. Hamiltonn and L. Chen. J. Am. Chem. Soc. 110, 5833
(1988); (b) Y. Kotake and E.G. Jansen. J. Am. Chem. Soc. 111,
5138 (1989).
References
1. M.L. Bender and M. Komiyama. Cyclodextrin chemistry.
Springer-Verlag, New York. 1977.
2. (a) W. Saenger. Angew. Chem. Int. Ed. Engl. 19, 344 (1980);
(b) I. Tabushi and Y. Kuroda. In Advances in catalysis. Vol.
32. Edited by D.D. Eley, H. Pines, and P.B. Weisz. Academic
Press, London. 1983. pp. 417–462; (c) D.R. Smith. Chem. Ind.
London. 14 (1994).
3. A. Veglia and R.H. de Rossi. J. Org. Chem. 53, 5281(1988).
4. A.V. Veglia, A.M. Sanchez, and R.H. de Rossi. J. Org. Chem.
55, 4083 (1990).
5. (a) G. Westman, O. Wennerström, and I. Raston. Tetrahedron,
49, 483 (1993); (b) E. Monflier, E. Blouet, Y. Barbaux, and A.
Mortreux. Angew. Chem. Int. Ed. 33, 2100 (1994); (c) A. Guy,
J. Doussot, A. Dallamaggiore, and R. Garreau. Can. J. Chem.
73, 599 (1995); (d) C.J. Easton, C.M. Hughes, E.R.T. Tiekink,
G.P. Savage, and G.W. Simpson. Tetrahedron Lett. 36, 629
(1995).
6. J. March. In Advanced organic chemistry, reactions, mecha-
nisms, and structure. 4th ed. John Wiley and Sons, New York.
1992. p. 1003.
7. (a) M. Sakura, M. Kitagawa, H. Hoshi, Y. Inoue, and R.
Chujo. Carbohydr. Res. 198, 181 (1990); (b) D. Mentzafos,
I.M. Mavridis, G. Le Bas, and G. Tsoucaris. Acta Cryst. B47,
746 (1991).
22. E.L. Foreman and S.M. Mc Elvain. J. Am. Chem. Soc. 62,
1435 (1940).
23. J.H. Speer and A.J. Hill. J. Org. Chem. 139, 139 (1937).
© 1999 NRC Canada