Effect of cyclodextrins on electrophilic aromatic bromination in aqueous solution

Article abstract of DOI:10.1071/CH03102

Cyclodextrins act as molecular reactors to change the ratios of the products of reactions of anisole, acetanilide, 3-methylanisole, and 3-methylacetanilide with pyridinium dichlorobromate. With anisole and acetanilide, bromination at the para position is favoured over ortho substitution, and the effect is greatest with α-cyclodextrin. In the reactions of the methylanisole and methylacetanilide, the cyclodextrins afford higher yields of monobrominated products and less of the di- and tribromides, and β-cyclodextrin has the greatest effect. These outcomes can be attributed to inclusion of the substrates within the cyclodextrins restricting access of the reagent adjacent to the methoxy and acetamido groups. The yields of 4-bromoanisole, 4-bromoacetanilide, 4-bromo-3-methylanisole, and 4-bromo-3-methylacetanilide are thus increased from 73 to 94, 55 to 98, 37 to 86, and 39 to 72%, respectively. Perhaps more significantly, the quantities of the corresponding by-products are substantially reduced, from 27 to 6, 45 to 2, 63 to 14, and 61 to 28%. Since the reactions occur readily in water at ambient temperature, the cyclodextrins make them very efficient.

Full text of DOI:10.1071/CH03102

CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2003, 56, 1107–1111
Effect of Cyclodextrins on Electrophilic Aromatic Bromination in
Aqueous Solution*
Paul G. Dumanski,A Christopher J. Easton,A,C Stephen F. LincolnB and Jamie S. SimpsonA
A Research School of Chemistry, Australian National University, Canberra 0200, Australia.
B Department of Chemistry, University of Adelaide, Adelaide 5005, Australia.
C Author to whom correspondence should be addressed (e-mail: easton@rsc.anu.edu.au).
Cyclodextrins act as molecular reactors to change the ratios of the products of reactions of anisole, acetanilide, 3-
methylanisole, and 3-methylacetanilide with pyridinium dichlorobromate.With anisole and acetanilide, bromination
at the para position is favoured over ortho substitution, and the effect is greatest with α-cyclodextrin. In the reactions
of the methylanisole and methylacetanilide, the cyclodextrins afford higher yields of monobrominated products and
less of the di- and tribromides, and β-cyclodextrin has the greatest effect. These outcomes can be attributed to
inclusion of the substrates within the cyclodextrins restricting access of the reagent adjacent to the methoxy and
acetamido groups. The yields of 4-bromoanisole, 4-bromoacetanilide, 4-bromo-3-methylanisole, and 4-bromo-3-
methylacetanilide are thus increased from 73 to 94, 55 to 98, 37 to 86, and 39 to 72%, respectively. Perhaps more
significantly, the quantities of the corresponding by-products are substantially reduced, from 27 to 6, 45 to 2, 63 to
14, and 61 to 28%. Since the reactions occur readily in water at ambient temperature, the cyclodextrins make them
very efficient.
Manuscript received: 23 April 2003.
Final version: 29 May 2003.
OMe
OMe
Introduction
Cl
Cyclodextrins have attracted considerable attention as
enzyme mimics, due to their ability to form inclusion com-
plexes with small organic compounds in water and catalyze
reactions of the included species.[1–5] They have also been
exploited as molecular reactors, where they control the
assembly of reactants to change the outcomes of chemical
transformations.[6–16] Examplesofthelatterincludethecova-
lent attachment of dipolarophiles to cyclodextrins to reverse
the regioselectivity of cycloadditions with nitrile oxides[6,7]
and the development of a urea-linked cyclodextrin dimer to
bias competing reactions to give indigoid dyes.[8]
Probably the most straightforward examples of cyclodex-
trin molecular reactors are those that involve a change in the
regioselectivity of reaction as a result of a substrate being
included in such a way as to restrict access of a reagent. Pio-
neering research in this area by Breslow et al.[9–11] showed
thatcyclodextrinsaltertheregioselectivityofaromaticsubsti-
tution. Hypochlorous acid chlorination of anisole (3a) in the
absence of a cyclodextrin gave the chlorides (1) and (2) (see
Diagram 1), in a ratio of about 2 : 3. When the reactions were
repeated in the presence of either α- or β-cyclodextrin, sub-
stantiallymoreoftheparaisomer(2)wasformed, particularly
in the former case. Breslow et al.[9–11] reasoned that these
chlorinations involve cyclodextrin hypochlorites and that, in
Cl
(1)
(2)
Diagram 1.
the inclusion complexes with the cyclodextrins, the ortho
positions of the anisole (3a) are shielded from chlorination
while the para position is still accessible. A related study
showed selective para chlorination of acetanilide (3b) in the
presence of cyclodextrins[12] but cyclodextrin hypochlorites
were not involved as intermediates in this case. The reac-
tions discussed above were carried out in water, which has
the added advantage of being an environmentally benign sol-
vent. The cyclodextrins facilitate the use of this medium by
increasing the solubility of organic substrates.
Another example of the use of cyclodextrins to influ-
ence aromatic substitution was reported by Komiyama
and Hirai,[13,14] where the regioselectivity of the Reimer–
Tiemann reaction of phenol with chloroform was altered,
again in favour of para substitution. Tee and Bennett[17]
investigated the effects of cyclodextrins on the bromination
of anisole (3a) with bromine/KBr in water, but in that system
∗ Dedicated to Professor John H. Bowie with best wishes on the occasion of his 65th birthday.
© CSIRO 2003
10.1071/CH03102
0004-9425/03/111107

1108
P. G. Dumanksi et al.
Table 1. Products of bromination of compounds (3a)–(3c), (7a),
and (7b)A
key resonances (Table 2) that were assigned based on com-
parison with the spectra of authentic samples in the cases
of compounds (3a)–(3c), (4a), (5a), (5b), (7a), (7b), (11a),
(13a), and (14a), and on data from the literature[19–25] for the
bromides (4b), (4c), (5c), (6a), (6b), (9a), and (9b). Reso-
nances were assigned to the methylacetanilides (11b), (13b),
and (14b) by analogy with the spectra of the corresponding
methylanisoles (11a), (13a), and (14a), and by comparison
of observed chemical shifts and coupling constants with cal-
culated values.[26] An unidentified compound (X) was also
observed in reactions of the methylacetanilide (7b). It seems
likely that this is the product of a secondary process, since it
was not seen during the initial stages of reaction.
Entry Substrate Cyclodextrin Reagent
Product ratios [%]
(CD)
(mol (3) (4) (5) (6)
equiv.)
1
2
3
(3a)
(3b)
(3c)
—
α-CD
β-CD
1.1
1.1
1.1
0 12 73 15
0
0
2 94
6 86
4
8
4
5
6
—
α-CD
β-CD
1.1
1.1
1.1
0 42 55
0 98
0 21 79
3
2
0
0
7
8
9
—
α-CD
β-CD
1.1
1.1
1.1
41 2 57
40 2 58
77 1 22
0
0
0
The reactions of anisole (3a) and acetanilide (3b) with 1.1
equiv. of the brominating agent gave small amounts of the
dibromides (6a) and (6b), respectively, presumably by way of
the corresponding monobromides (4a), (4b), (5a), and (5b)
(Scheme 1). Phenyl acetate (3c) reacted to a much lesser
extent, even after a much longer reaction time, and only the
monobromides (4c) and (5c) were produced, i.e., there was
no evidence of formation of the dibromide (6c). In the reac-
tions of the methylanisole (7a) and methylacetanilide (7b),
unreacted starting materials and the corresponding mono-
bromides (9a) and (9b), dibromides (11a), (11b), (13a), and
(13b), and tribromides (14a), (14b), and the unidentified
product(X)inthecaseoftheacetanilide(7b), accountedforat
least 95 mol-% of the product mixtures. This was determined
by analysis of 1H NMR spectra and particularly through com-
parison of the integrations of all the methyl proton resonances
relative to those of the aromatic protons. While it is conceiv-
able that the monobromides (8a), (8b), (10a), and (10b), and
the dibromides (12a) and (12b) could also have been pro-
duced (Scheme 2), they were each detected at most in only
trace quantities (≤3%).
(7) (9) (11) (13) (14)
10
11
12
(7a)
(7a)
(7a)
(7b)
(7b)
—
α-CD
β-CD
0.6
0.6
0.6
50 23 12 14
1
1
0
40 42
36 59
8
3
9
2
13
14
15
—
α-CD
β-CD
1.1
1.1
1.1
23 37 19 20
12 46 18 21
1
1
0
4 86
5
5
16
—
2.2
0 30 29 34
7
(7) (9) (11) (13) (14) (X)
17
18
19
—
α-CD
β-CD
0.6
0.6
0.6
51 36
43 48
44 48
8
6
5
5
3
3
0
0
0
0
0
0
20
21
22
—
α-CD
β-CD
1.1
1.1
1.1
12 39
5 64
3 72
9
9
7
7
6
4
20 13
13
1
3
13
A Reaction with pyridinium dichlorobromate in water at room temper-
ature. Reaction time was 1 h for (3a), (3b), (7a), and (7b), and 16 h
for (3c).
the oligosaccharides did not alter the regioselectivity. Instead,
they retarded the reaction rate, due to inclusion of both the
substrate and the brominating agent. In light of the lack of
regiocontrol in this system, the recent account of the use
of pyridinium dichlorobromate for aromatic bromination in
aqueous methanol[18] prompted us to investigate the effect of
employing cyclodextrins with this reagent. As reactants for
the study, we chose anisole (3a) and acetanilide (3b), in order
to make direct comparisons with the chlorination of those
compounds. Phenyl acetate (3c), and the methyl-substituted
anisole (7a) and acetanilide (7b) were also selected as sub-
strates, on the basis that the ester (3c) is less reactive, and the
ether (7a) and amide (7b) more reactive, towards aromatic
substitution.
In the reactions carried out in the absence of a cyclodex-
trin, the formation of only small amounts of the dibromides
(6a) and (6b) from anisole (3a) and acetanilide (3b), respec-
tively, relative to the yields of the corresponding mono-
bromides (4a), (4b), (5a), and (5b) (Table 1, entries 1 and 4),
indicates that the bromo substituents of (4a), (4b), (5a), and
(5b) reduce the reactivity of these systems towards further
aromatic substitution.This deactivation is typical of halogens
and is greatest with the acetanilides (3b)–(5b). By contrast,
the yields of the dibromides (11a), (11b), (13a), and (13b)
from the methylanisole (7a) and methylacetanilide (7b) are
more substantial (Table 1, entries 10, 13, 16, 17, and 20),
particularly with the anisole (7a), where monobromination
appears to activate the system to further reaction. Presum-
ably, the combination of methoxy and methyl substituents
perturbs the balance of resonance and inductive effects nor-
mally seen with a bromo group. Significant quantities of (8a),
(8b), (10a), (10b), (12a), and (12b) do not build up in the
reactions of (7a) and (7b). This is probably due largely to the
selectivity of bromination para to a methoxy or acetamido
group, but another contributing factor may be that, like (9a)
and (9b), the bromides (8a), (8b), (10a), (10b), (12a), and
(12b) are unusually reactive, and once formed react further to
give (11a), (11b), (13a), (13b), (14a), and (14b), respectively.
Results and Discussion
Each of the substrates (3a)–(3c), (7a), and (7b) (1.3 mM)
was treated with pyridinium dichlorobromate (0.6, 1.1, and/or
2.2 mol equiv.), in water containing methanol (1% v/v). The
reactions were carried out at room temperature, for either 1 or
16 h, in the absence of a cyclodextrin and in the presence of
either α- or β-cyclodextrin (10 mol equiv.).After work-up, the
crude product mixtures were analyzed using 1H NMR spec-
troscopy. The ratios of the components present are shown
in Table 1. These were determined through integration of

Effect of Cyclodextrins on Electrophilic Aromatic Bromination in Aqueous Solutions
1109
Table 2. NMR signals used for determining product ratios
Spectroscopic dataA
7.22 (2 H, m, H3, H5), 6.91–6.80 (3 H, m, H2, H4, H6), 3.77 (3 H, s, OCH3)
Compound Solvent
Reference
B
B
(3a)
(3b)
CDCl3
[D6]DMSO 7.54 (2 H, d, J 7.5, H2, H6), 7.26 (2 H, t, J 7.5, H3, H5), 7.00 (1 H, t, J 7.5, H4), 2.02 (3 H, s,
NHCOCH3)
B
B
(3c)
(4a)
CDCl3
CDCl3
7.38 (2 H, m, H3, H5), 7.24 (1 H, m, H4), 7.09 (2 H, m, H2, H6), 2.28 (3 H, s, COCH3)
7.54 (1 H, dd, J 8.0 and 1.5, H6), 7.27 (1 H, ddd, J 8.0, 7.5 and 1.5, H5), 6.90 (1 H, dd, J 8.5 and
1.5, H3), 6.84 (1 H, ddd, J 8.5, 7.5 and 1.5, H4), 3.89 (3 H, s, OCH3)
(4b)
(4c)
[D6]DMSO 7.61 (1 H, d, J 7.5 and 1.5, H6), 7.33 (1 H, td, J 7.5 and 1.5, H5), 7.10 (1 H, td, J 7.5 and 1.5, H4),
2.06 (3 H, s, NHCOCH3)
[19]
[20]
CDCl3
7.60 (1 H, dd, J 8.5 and 1.5, H6), 7.32 (1 H, ddd, J 8.0, 7.5 and 1.5, H5), 7.13 (1 H, m, H3), 7.12
(1 H, m, H4), 2.34 (3 H, s, COCH3)
B
B
(5a)
(5b)
(5c)
(6a)
(6b)
(7a)
CDCl3
6.70 (2 H, d, J 9.0, H2, H6), 7.29 (2 H, d, J 9.0, H3, H5), 3.79 (3 H, s, OCH3)
[D6]DMSO 7.53 (2 H, d, J 9.0, H3, H5), 7.43 (2 H, d, J 9.0, H2, H6), 2.02 (3 H, s, NHCOCH3)
CDCl3
CDCl3
7.46 (2 H, d, J 9.0, H3, H5), 6.96 (2 H, d, J 9.0, H2, H6), 2.27 (3 H, s, COCH3)
7.65 (1 H, d, J 2.5, H3), 3.87 (3 H, s, OCH3)
[21]
[22]
[D6]DMSO 7.72 (1 H, d, J 2.1, H3), 2.11 (3 H, s, NHCOCH3)
[23]
B
CD3OD
7.11 (1 H, t, J 8.0, H5), 6.71 (2 H, m, H4, H6), 6.66 (1 H, m, H2), 3.73 (3 H, s, OCH3),
2.28 (3 H, s, ArCH3)
B
(7b)
(9a)
CD3OD
CD3OD
7.35 (1 H, br s, H2), 6.90 (1 H, br d, J 8.0, H4), 7.15 (1 H, t, J 8.0, H5), 7.30 (1 H, br d, J 8.0, H6)
7.35 (1 H, d, J 8.5, H5), 6.83 (1 H, d, J 3.0, H2), 6.63 (1 H, dd, J 8.5 and 3.0, H6), 3.74 (3 H, s, OCH3), [24]
2.32 (3 H, s, ArCH3)
(9b)
CD3OD
CD3OD
CD3OD
CD3OD
CD3OD
CD3OD
CD3OD
CD3OD
7.47 (1 H, d, J 2.5, H2), 7.41 (1 H, d, J 8.5, H5), 7.28 (1 H, dd, J 8.5 and 2.5, H6)
7.49 (1 H, d, J 9.0, H5), 6.80 (1 H, d, J 9.0, H6)
7.53 (1 H, d, J 8.5, H5)
7.60 (1 H, s, H6), 6.69 (1 H, s, H3)
7.60 (1 H, br s, H6), 7.76 (1 H, s, H3)
7.79 (1 H, s, H5)
[25]
B
(11a)
(11b)
(13a)
(13b)
(14a)
(14b)
(X)
[26]
B
[26]
B
7.87 (1 H, s, H5)
7.12 (1 H, ddq, J 8.5, 3.0 and 0.5), 7.68 (1 H, d, J 8.5)
[26]
—
A Discrete signals were not observed for all resonances; those which could not be unambiguously assigned are not shown.
B Spectrum of an authentic sample.
cyclodextrins on chlorination reported previously.[9–12] In the
case of anisole (3a), the cyclodextrins also limit the extent of
formationofthedibromide(6a), andagainitisα-cyclodextrin
that has the greatest effect. It seems likely that this is mainly
due to the cyclodextrins retarding further bromination of
the monobromide (5a) by blocking the 2-position. The net
result of these effects is an increase in the yields of the
monobromides (5a) and (5b), from 73 to 94, and 55 to 98%,
respectively, and a substantial decrease in the quantity of the
corresponding by-products, from 27 to 6, and 45 to 2%. The
cyclodextrins do not alter the regioselectivity of bromination
of phenyl acetate (3c), although β-cyclodextrin decreases the
extent of reaction (Table 1, entries 7–9). In the reactions of the
methylanisole (7a) and methylacetanilide (7b), the cyclodex-
trins increase the yields of the monobromides (9a) and (9b),
probably by limiting the subsequent reactions of (9a) and (9b)
to give the dibromides (11a), (11b), (13a), and (13b) and tri-
bromides (14a) and (14b), as well as by decreasing the extent
of reaction by way of the monobromides (8a), (8b), (10a),
(10b), (12a), and (12b), to give (11a), (11b), (13a), (13b),
(14a), and (14b). Again the effect of the cyclodextrins is to
prevent bromination adjacent to the methoxy and acetamido
groups, in a similar manner to that seen with anisole (3a) and
acetanilide (3b), but in the methylated systems β-cyclodextrin
has the greatest effect. When 1.1 mol equiv. of the bromi-
nating agent was used with β-cyclodextrin, the yields of the
monobromides (9a) and (9b) increased from 37 to 86, and 39
to 72%, respectively, and there was a substantial decrease in
R
(3)
a–c
a–c
R
R
Br
(4)
Br
a,b
a,b
(5)
R
Br
a R = OMe
b R = NHAc
c R = OAc
Br
(6)
Scheme 1.
In the reactions of anisole (3a) and acetanilide (3b), both
α- and β-cyclodextrin change the ratios of formation of the
monobrominated products (4a), (4b), (5a), and (5b) in favour
of the para-substituted isomers (5a) and (5b) (Table 1, entries
1–6). The effect is greatest with α-cyclodextrin. It is thus
apparent that the cyclodextrins limit ortho bromination of the
substrates (3a) and (3b), presumably through the formation of
inclusion complexes which restrict access of the brominating
agent, in a manner that is directly analogous to the effect of

1110
P. G. Dumanksi et al.
R
C5-H
C3-H
(7)
R
R
R
o
p
7.10
7.15
7.20
7.25
7.30
7.35
7.40
7.45
Br
Br
(8)
(10)
Br
(9)
m
R
R
4.04 4.00 3.96 3.92 3.88 3.84 3.80 3.76
F1 (ppm)
R
Br
Br
Br
Br
Fig. 1. A portion of the ROESY spectrum (500 MHz) recorded for a
solution of anisole (3a) (10 mM) and α-cyclodextrin (10 mM) in D2O,
showing interactions between resonances of the cyclodextrin (x axis)
and those of the anisole (3a) (y axis).
Br
Br
(12)
R
(11)
(13)
C5-H
C3-H
Br
Br
a R = OMe
b R = NHAc
Br
(14)
o
7.10
p
Scheme 2.
7.15
7.20
7.25
7.30
7.35
7.40
the quantity of the corresponding by-products, from 63 to 14,
and 61 to 28%.
It is not obvious why α-cyclodextrin has the greatest
effect on the bromination of anisole (3a) and acetanilide
(3b), yet the reactions of the methylated substrates (7a) and
(7b) are most affected by β-cyclodextrin. The outcomes will
depend on the extent of inclusion of the reagent, pyridinium
dichlorobromate, as well as the substrates (3a), (3b), (7a),
and (7b), and the relative reactivity of the free and included
species. The systems are further complicated by possible
complexation of the primary products such as the mono-
bromides (8a), (8b), (9a), (9b), (10a), and (10b), affecting
their subsequent reactions. Nevertheless, it seems likely that
a major contributing factor to the regiocontrol provided by
the cyclodextrins derives from complexation of the sub-
strates (3a), (3b), (7a), and (7b) in such a way as to restrict
access of the reagent adjacent to the methoxy and acetamido
substituents. Such shielding is reflected in rotating-frame
Overhauser enhancement spectroscopy (ROESY) spectra of
mixtures of anisole (3a) and α- and β-cyclodextrin (Figs. 1
and 2, respectively), which both show nuclear Overhauser
effects (NOEs) between the resonances of the ortho hydro-
gen atoms of the substrate (3a) and the cyclodextrin C3–H
and C5–H resonances. ROESY spectra of mixtures of the
m
7.45
4.04 4.00 3.96 3.92 3.88 3.84 3.80 3.76
F1 (ppm)
Fig. 2. A portion of the ROESY spectrum (500 MHz) recorded for
a solution of anisole (3a) (2 mM) and β-cyclodextrin (2 mM) in D2O,
showing interactions between resonances of the cyclodextrin (x axis)
and those of the anisole (3a) (y axis).
other substrates (3b), (7a), and (7b) with cyclodextrins also
show NOEs between the resonances of the substrate and
cyclodextrin hydrogen atoms. These show that the substrates
are included in the cyclodextrins in aqueous solution but fur-
ther interpretation of the data is impractical in these cases
due to overlapping substrate resonances and changes in the
chemical shifts of the substrate resonances induced by the
cyclodextrins.
In summary, both α- and β-cyclodextrin affect the regio-
selectivity of bromination of anisole (3a) and acetanilide

Effect of Cyclodextrins on Electrophilic Aromatic Bromination in Aqueous Solutions
1111
(3b), and the methylated analogues (7a) and (7b), to increase
the yields of the corresponding bromides (4a), (4b), (9a),
and (9b). A corollary of this is that the reactions are fur-
ther improved by the substantial reduction in the yields of
the by-products. Since the brominations occur readily in
water at ambient temperature, and they require only stoichio-
metric quantities of reagents, the cyclodextrins make them
very efficient chemical transformations.
shown in Table 1, as determined by integration of key resonances that
are listed in Table 2.
Acknowledgments
The authors gratefully acknowledge the support of the
Australian Research Council and the award of an APA(I)
Scholarship to P.G.D.
References
Experimental
[1] R. L. VanEtten, G. A. Clowes, J. F. Sebastian, M. L. Bender,
J. Am. Chem. Soc. 1967, 89, 3253.
[2] R. L. VanEtten, J. F. Sebastian, G. A. Clowes, M. L. Bender,
J. Am. Chem. Soc. 1967, 89, 3242.
[3] C. J. Easton, S. F. Lincoln, Modified Cyclodextrins. Scaffolds
and Templates for Supramolecular Chemistry 1999 (Imperial
College Press: London).
[4] R. Breslow, S. D. Dong, Chem. Rev. 1998, 98, 1997.
[5] L. Barr, C. J. Easton, K. Lee, S. F. Lincoln, J. S. Simpson,
Tetrahedron Lett. 2002, 43, 7797.
[6] A. G. Meyer, C. J. Easton, S. F. Lincoln, G. W. Simpson, Chem.
Commun. 1997, 1517.
[7] A. G. Meyer, C. J. Easton, S. F. Lincoln, G. W. Simpson, J. Org.
Chem. 1998, 63, 9069.
[8] C. J. Easton, J. B. Harper, S. F. Lincoln, New J. Chem. 1998,
22, 1163.
[9] R. Breslow, P. Campbell, J. Am. Chem. Soc. 1969, 91, 3085.
[10] R. Breslow, P. Campbell, Bioorg. Chem. 1971, 1, 140.
[11] R. Breslow, H. Kohn, B. Siegel, Tetrahedron Lett. 1976, 1645.
[12] R. Chênevert, G. Ampleman, Can. J. Chem. 1987, 65, 307.
[13] M. Komiyama, H. Hirai, J. Am. Chem. Soc. 1983, 105, 2018.
[14] M. Komiyama, H. Hirai, J. Am. Chem. Soc. 1984, 106, 174.
[15] R. Breslow, J. Desper, Y. Huang, Tetrahedron Lett. 1996, 37,
2541.
[16] J. Yang, R. Breslow, Angew. Chem. Int. Ed. 2000, 39, 2692.
[17] O. S. Tee, J. M. Bennett, Can. J. Chem. 1984, 62, 1585.
[18] H. A. Muathen, Synthesis 2002, 169.
[19] E. Kolehmainen, K. Laihia, R. Kauppinen, D. Rasala,A. Puchala,
J. Phys. Org. Chem. 1995, 8, 577.
General
1H NMR spectra were recorded on either a Varian Inova 500 spec-
trometer or a Varian Gemini 300 spectrometer. Spectra were referenced
against tetramethylsilane (δ 0.00 ppm) for CDCl3 solutions, external 3-
(trimethylsilyl)-3,3,2,2-tetradeuteropropionic acid sodium salt for D2O
solutions, or residual protons for other deuterated solvents. ROESY
spectra were recorded on a Varian Inova 500 spectrometer employing a
mixing time of 250 ms. The sample tubes were sealed with RotoTite
valves purchased from Wilmad Glass, and samples were repeatedly
degassed by freeze–pump–thaw cycling before spectra were recorded.
The samples contained either α-cyclodextrin (0.01 M) and anisole
(3a) (0.01 M), or β-cyclodextrin (0.002 M) and anisole (3a) (0.002 M),
in D2O.
α-Cyclodextrin and β-cyclodextrin were generous gifts of Nihon
Shokuhin Kako Co., Japan. They were recrystallized from water and
dried under vacuum over P2O5 to constant weight before use. Water
was purified with a MilliQ water system.
Anisole (3a), phenyl acetate (3c), 4-bromoanisole (5a), 4-
bromoacetanilide (5b), 3-methylanisole (7a), and m-toluidine were
purchased from Sigma–Aldrich. Acetanilide (3b) was bought from
Ajax Chemicals and 2-bromoanisole (4a) was purchased from
Fluka. Pyridinium dichlorobromate was prepared as described by
Muathen.[18] Anisole (3a) was purified by distillation (bp 153–154◦C)
before use. 3-Methylacetanilide (7b) was prepared by acetylating
m-toluidine.[27] Samples of 2,4-dibromo-3-methylanisole (11a), 2,4-
dibromo-5-methylanisole (13a), and 2,4,6-tribromo-3-methylanisole
(14a) were prepared by bromination of 3-methylanisole (7a) with
bromine in acetic acid.[28]
[20] J. A. Donnelly, J. J. Murphy, J. Chem. Soc. C 1970, 2596.
[21] D. S. Wilbur, W. E. Stone, K. W. Anderson, J. Org. Chem. 1983,
48, 1542.
[22] J. R. L. Smith, L. C. McKeer, J. M. Taylor, J. Chem. Soc.,
Perkin Trans. 2 1988, 385.
[23] H. Y. Choi, D. Y. Chi, J. Am. Chem. Soc. 2001, 123, 9202.
[24] M. C. Carreno, J. L. Garcia Ruano, G. Sanz, M. A. Toledo,
A. Urbano, J. Org. Chem. 1995, 60, 5328.
[25] L. M. Gaster, F. D. King, P. A. Wyman, in Int. Patent Appl. WO
96/19477 1996, p. 94 (SmithKline Beecham: Harlow).
[26] HNMR Predictor (Advanced Chemistry Development: Toronto);
http://www2.acdlabs.com/ilab/ (accessed April 2003).
[27] T. W. Greene, P. G. M. Wuts, Protective Groups in Organic
Synthesis, 2nd edn 1991 (John Wiley: New York, NY).
[28] R. D. Haworth, A. Lapworth, J. Chem. Soc. 1923, 123, 2982.
General Bromination Procedure
The aromatic substrate (3a)–(3c), (7a), or (7b) (0.2 mmol) in methanol
(1.5 mL) was added to water (100 mL) that contained either no cyclodex-
trin or α- or β-cyclodextrin (2.0 mmol), and the resulting mixture was
stirred vigorously. Pyridinium dichlorobromate (0.22 mmol) was added
and the total volume of the solution was made up to 150 mL with water.
The resulting mixture was stirred at room temperature, for 1 h with (3a),
(3b), (7a), and (7b), and for 16 h with (3c), and the reaction was then
quenched through the addition of an excess of NaHSO3. The solution
was extracted with ether (2 × 80 mL), and the ether extracts were washed
with water (100 mL), dried, and concentrated under reduced pressure.
The residue was analyzed using 1H NMR spectroscopy. Spectra were
recorded in either CDCl3, CD3OD, or [D6]DMSO, depending on the
solubility of the components. The ratios of the components present are