High atomic yield bromine-less benzylic bromination

Ramon Mestres* and Jesús Palenzuela

Departament de Química Orgànica, Universitat de València, 46100 Burjassot, València,

Spain. E-mail: ramon.mestres@uv.es

Received 27th April 2002

First published as an Advance Article on the web 27th June 2002

A two-phase mixture (sodium bromide, aqueous hydrogen peroxide/carbon tetrachloride or chloroform) under
visible light provides a simple and convenient system for benzylic bromination of toluenes. A high atomic yield
for bromine atoms is attained. Substitution of the chlorinated solvents by other more environmentally benign
organic solvents has been attempted and good results are obtained for methyl pivalate.

Introduction convenient in the view of the above precedents and because

water is the only concomitant product (eqn. (1)–(3)).
Molecular bromine in preparative chemistry is a serious cause
of concern, due to its toxicity and corroding properties. In order 2 Br2 + H2O2 + 2 H+ = Br2 + 2 H2O (1)
to circumvent this drawback, generation of bromine in situ by
oxidation of hydrogen bromide or an alkali-metal bromide by ArCH3 + Br2 = ArCH2Br + Br2 + H+ (2)
potassium permanganate, sodium bromate, hydrogen peroxide,
or other reagents has been performed as far back as in the 19th ArCH3 + Br2 + H2O2 + H+ = ArCH2Br + 2 H2O (3)
century, although these methods have not found general Further, when the substitution reaction is taken into account, the
application.1–3 whole conversion should be possible by consuming a single
Further, some of the most significant synthetic reactions of mole of bromide and one mole of acid. The photochemical
molecular bromine, namely the bromination of aromatic nuclei, bromination of a number of substituted toluenes by a bromide–
ketones, carboxylic acids or allylic or benzylic sites occur under hydrogen peroxide reagent under acidic conditions was thus
low atomic yield for bromine. Indeed, only one out of two studied (Scheme 1). The reactions were carried out initially in
bromine atoms of molecular bromine becomes part of the carbon tetrachloride, the solvent commonly used up now for
product, the other atom becomes the corroding hydrogen molecular bromine or NBS brominations. Other chlorinated
bromide, a substance which must be neutralized before it is solvents and other non-chlorinated solvents were then as-
discarded.4 Brominations may thus be qualified as environmen- sayed.
tally unfriendly reactions.5
The Wohl–Ziegler method of bromination with the easy to
handle NBS has become for many years the reagent of choice
for allylic and benzylic brominations.6,7 Direct use of molecular
bromine in the laboratory is avoided and no hydrogen bromide
is generated. However, the atomic yield is still poor and
succinimide is obtained as a concomitant product which,
although not harmful, must be either disposed or recycled to
NBS, and very frequently complicates cristallizations for
isolation and purification of products.
Combined bromide–hydrogen peroxide systems have been
used recently for addition to unsaturated double or triple
bonds,8–10 aromatic electrophilic substitution,8,9,11 synthesis of Scheme 1
amine tribromides,12 or for oxidation of alcohols, aldehydes,9 or
sulfides.13–15 Aromatic side-chain halogenations in the pres-
ence of metal oxide catalysts have been reported as well.16,17
Interestingly, significant improvements of both electrophilic Green Context
aromatic and photochemical side-chain brominations by bro-
The use of bromine is fraught with problems of handling
mine in the presence of zeolites have been recently de-
and the low atom efficiency of substitution reactions (50%
scribed.18
In the course of our studies for the synthesis of some benzyl maximum with bromide by-poduct). For these reasons work
ether dendritic and polymeric structures,19 we wished to has been carried out using bromide and an oxidant to
improve the benzylic bromination of protected methylphenols. generate bromine (or electrophilic Br species) for electro-
Although bromination of 4-methylbenzoyloxybenzene 1a with philic aromatic brominations. Here, bromide is used in
NBS occurred satisfactorily in general terms, we needed a conjunction with light to carry out side-chain brominations
procedure more suitable for a multigram scale. On the other of benzylic systems, extending the concept to radical-type
hand, the urgent need for development of safe and environmen- reactions. Importantly, a non-halogenated solvent has also
tally benign methods encouraged us to search for the possibility been found for these reactions which typically use CCl4 or
of a convenient benzylic bromination through oxidative genera- CHCl3. DJM
tion of bromine in situ. Use of hydrogen peroxide seemed very

314 Green Chemistry, 2002, 4, 314–316 DOI: 10.1039/b203055a

This journal is © The Royal Society of Chemistry 2002
Results and discussion Similarly, changes in the concentration or nature of the acid, the
cation (Na or K) of the bromide salt, or the order of addition of
For comparison purposes, brominations of 4-benzoyloxyto- reagents had no effect.
luene 1a with NBS and with molecular bromine in carbon A rough estimation of the efficiency of the bromination is
tetrachloride were carried out first. Thus, bromination with one given by the bromination yield (Table 1). This is obtained here
equivalent of NBS in the presence of benzoyl peroxide, by summing the amounts of bromine found in the reaction
ilumination with visible light and heating under reflux, mixture, as given by the product ratios for the mono- and di-
according to the typical procedure, gave a 84+16 mixture of brominated compounds 2 and 3, corrected by the crude yield.
mono- and di-bromination compounds, 2a and 3a, with the Good bromination yields were obtained with only a slight
starting material left as a trace amount. With a slight excess (1.1 excess (1.1) of bromide salt. According to eqns. (1)–(3)
molar equivalents) of bromine relative to the substituted toluene bromination could be complete with one molar equivalent of
1a, the red colour of the solution faded in less than 10 min and acid, but the reaction became slower and afforded much lower
the resulting mixture (91% isolated crude yield) contained halogenation (Table 1, entry 5) on use of 1.1 equivalents of acid.
compounds 2a (72%) and 3a (25%), and a very small amount of A large excess of acid seems thus to have a beneficial effect and
the starting toluene 1a (3%). When the bromination was carried most brominations were carried out in the presence of 3
out with slightly more than half an equivalent (0.55) of bromine equivalents of acid. The amount of hydrogen peroxide could be
in the presence of hydrogen peroxide (1.5 equivalents), progress reduced to 1.1 equivalents without a significant detriment on
of the reaction was much slower, a red to orange colour change bromination yields. However, most experiments were carried
being observed after 8 h. Quenching of the reaction after 20 h out with a larger excess of hydrogen peroxide.
gave a crude mixture (85%) of compounds 1a, 2a and 3a in a Chloroform as solvent gave a satisfactory result (Table 1,
ratio 18, 74 and 8%, respectively. The introduction of a second entry 10), but poor bromination yields were obtained with
atom of bromine is in keeping with the results of Smith et al. on methylene dichloride (entry 9).
the bromination of ethyl 4-methylbenzoate,18 and is probably The general applicability of this procedure was confirmed
due to the rate constants for the dissociation of the C–H bonds when bromination of other substituted toluenes 1b–g was
of the starting toluene and of the monobromo derivative 1 being assayed (Table 2). No significant differences were observed in
of a similar order. proportions of the mixtures resulting as function of the electron-
Brominations with the bromide–hydrogen peroxide system donating or electron-withdrawing character of the substituents
were carried out at room temperature in an open vessel provided in toluenes 1b–f, although crude yields for the chloro com-
with a cooling condenser and under illumination with an pounds 1b and 1c were lower and thus, bromination yields were
incandescent light bulb. Sodium (occasionally potassium) also low. For p-cresol 1g side-chain bromination did not occur
bromide, hydrochloric or sulfuric acid and 30% aqueous to a significant extent, electrophilic aromatic substitution
hydrogen peroxide, the concentration of which was determined instead being observed. This finding is in keeping with earlier
volumetrically with KMnO4, were employed. On mixing of the results of oxybromination of strongly activated benzenes.8–10
reagents and the toluene 1, a red colour developed, which Methylene dichloride and chloroform were also assayed for p-
slowly faded in ca. 4–6 h. However, no systematic control of the nitro- and p-methoxycarbonyl toluenes 1e and 1f.
reaction time was conducted, as most experiments were carried Non-chlorinated solvents were examined with substrate 1a in
out overnight (20 h). The simple work-up procedure for most the hope of improving the environmental aspects of the
toluenes 1 consisted in thorough removal of peroxides from the procedure (Table 3). Hexane was not convenient due to the low
organic layer by aqueous sodium hydrogensulfite and evapora- solubility of the substrate 1a. Benzene gave fairly good results,
tion of the solvent to give a crude material, whose composition but esters, namely ethyl and isopropyl acetate or methyl
was established by 1H NMR spectroscopy. Work-up for isobutyrate, or tert-butyl methyl ether gave poor conversions.
bromination of toluene 1d differed in that evaporation of the Methyl pivalate gave a satisfactory bromination yield, although
solvent was not carried out, as remaining unreacted toluene halogenation products were accompanied by 4-benzoylox-
might be partly evaporated along with the solvent. Only the ybenzoic acid, which might result from substitution and
starting material 1, and the mono- and di-bromo compounds 2 oxidation of monobrominated 2a or from hydrolysis of 3a and
and 3 (Scheme 1) were found in the crude mixtures. oxidation of the resulting benzaldehyde. Higher solubility of
Study of the effect of conditions were carried out with water in this solvent is probably the origin of this exceptional
4-benzoyloxytoluene 1a as substrate, due to the easy isolation of behaviour. A similar very small amount of carboxylic acid was
the resulting reaction mixture. It had been ascertained pre- observed in the halogenation of p-nitrotoluene 1e in carbon
viously that no substitution reaction took place in the absence of tetrachloride.
photochemical activation and that the alkali-metal bromide In conclusion, we have found a very simple method for side-
could be added either as a solid or previously dissolved in the chain bromination of methylbenzenes which should be easily
aqueous acid with no significant change in the results. scaled-up.

Table 1 Conditions and results (1H NMR) for bromide/hydrogen peroxide benzylic bromination of toluene 1a (0.25 M) in chlorinated solvents

Reagentsa Product ratio

Bromination
Run H2O2 NaBr H+ Solvent Crude yield 1 2 3 yieldb

1 3 1.5 3c CCl4 95 2.5 80.5 17 108.7

2 3 1.1d 3 CCl4 94 6 84 10 97.7
3 3 1.1 2 CCl4 91 16 76 8 83.7
4 3 1.1 1.5 CCl4 89 15 77 8 81.9
5 3 1.1 1.1c CCl4 87 29 71 tr 61.8
6 2 1.1 3 CCl4 89 16.5 78 5.5 79.2
7 1.5 1.1 3 CCl4 91 20.2 74.3 5.5 77.6
8 1.1 1.1 3 CCl4 88 16 76 8 81
9 1.5 1.1 3 CH2Cl2 86 37 63 tr. 54.2
10 1.5 1.1 3 CHCl3 93 15.5 79.5 5 83.2
a Molar ratio relative to 1. b Bromination yield; estimated: (crude yield/100)[ratio 2 + (2 3 ratio 3)]. c 2 M HCl; other runs, conc. H SO . d KBr.
2 4

Green Chemistry, 2002, 4, 314–316 315

Table 2 Bromide/hydrogen peroxide benzylic bromination of toluenes Crystallization from hexane–benzene gave the title compound
1b–f (0.25 M) in chlorinated solventsa as colourless prisms (2.11 g; 73%), mp 110–111 °C (lit.,20
Product ratio
108–111 °C).
Crude Bromination
Run 1 Solvent yield 1 2 3 yieldb

1 1b CCl4 62 13.5 79.5 7 58

2 1c CCl4 60 16 74.5 9.5 56 Acknowledgements
3 1d CCl4 14.5 79.5 5
4 1e CCl4 82 14 79 7 76.2 The present work has been financed by Ministerio de Ciencia y
5 1e CH2Cl2 92 12.5 80.5 7 86.9 Tecnología (Project PPQ2000-0816).
6 1e CHCl3 93 13 78 8.7 88.7
7 1f CCl4 63 15 78 7 57.9
8 1f CH2Cl2 76 57 43 tr. 32.7
9 1f CHCl3 99 13 87 tr. 86.1
a Conditions for all runs: 3 equivalents of 30% H O , 1.1 equivalent of
2 2 References
NaBr, 3 equivalents of conc. H2SO4. b Bromination yield; estimated: (crude
yield/100)[ratio 2 + (2 3 ratio 3)]. 1 Eugen Müller, Methoden der Organischen Chemie (Houben-Weyl),
Band V/4, Georg Thieme Verlag, Stuttgart, 1960, p. 13.
2 A. R. Katritzky, O. Meth-Cohn and C. W. Rees, Comprehensive
Table 3 Conditions and results (1H NMR) for bromide/hydrogen peroxide Organic Functional Group Transformations, Pergamon, Oxford,
benzylic bromination of toluene 1a (0.25 M) in non-halogenated solvents 1995, vol. 2, p. 3.
3 B. M. Trost and I. Fleming, Comprehensive Organic Synthesis,
Product ratio Pergamon, Oxford, 1991, vol. 7, p. 15.
Crude Bromina- 4 J. Clark, Green Chem., 1999, 1, 1.
Runa Solvent yield 1 2 3 tion yield 5 P. T. Anastas and J. C. Warner, Green Chemistry, Theory and
Practice, Oxford University Press, Oxford, 1998, p. 105.
1 Hexane–ethyl acetate 85 63 37 — 31.5 6 Eugen Müller, Methoden der Organischen Chemie (Houben-Weyl),
2 Benzene 91 21 74 5 76.5 Band V/4, Georg Thieme Verlag, Stuttgart, 1960, pp. 221, 341.
3 Ethyl acetate 76 66.5 33.5 — 25.5 7 A. R. Katritzky, O. Meth-Cohn and C. W. Rees, Comprehensive
4 Isopropyl acetate 84 20.4 79.6 tr 66.8 Organic Functional Group Transformations, Pergamon, Oxford,
5 Methyl isobutyrate 97 67 33 — 32 1995, vol. 2, p. 3.
6 Methyl pivalate 84 11 84 5 79 8 N. B. Barhate, A. S. Gajare, R. D. Wakharkar and A. V. Bedekar,
7 Methyl pivalate 78 5.7 85.7 8.6 80 Tetrahedron, 1999, 55, 11127.
8 tert-Butyl methyl ether 89 58 42 tr. 30 9 J. H. Espenson, Z. Zhu and T. H. Zauche, J. Org. Chem., 1999, 64,
a Conditions for all runs: 3 equivalents of 30% H O , 1.1 equivalent of 1191.
2 2
10 B. F. Sels, D. E. DeVos and P. A. Jacobs, J. Am. Chem. Soc., 2001,
NaBr, 3 equivalents of conc. H2SO4.
123, 8350.
11 G. Rothenberg and J. C. Clark, Green Chem., 2000, 2, 248.
12 M. K. Chaudhuri, A. T. Khan, B. K. Patel, D. Dey, W. Kharmawo-
Experimental phlang, T. R. Lakshmiprabha and G. C. Mandal, Tetrahedron Lett.,
1998, 39, 8163.
4-Benzoyloxybenzyl bromide 1a 13 A. Bravo, B. Dordi, F. Fontana and F. Minisci, J. Org. Chem., 2001,
66, 3232.
A mixture of 4-benzoyloxytoluene (2.12 g; 10 mmol), sodium 14 E. Mondal, G. Bose, P. R. Sahu and A. T. Khan, Chem. Lett., 2001,
1158.
bromide (1.13 g; 11 mmol), 30 % hydrogen peroxide (1.70 mL; 15 A. T. Khan, J. Boruwa, E. Mondal and G. Bose, Indian J. Chem. B,
15 mmol), conc. sulfuric acid (1 mL; 15 mmol) and chloroform 2001, 40, 1039.
(40 mL) was stirred in a 100 mL round bottomed flask provided 16 E. Baciocchi and M. Creszenzi, Tetrahedron, 1988, 45, 6525.
with a reflux condenser and under illumination with a 100 W 17 J. Nakayama, T. Kawamura, K. Kuroda and A. Fujita, Tetrahedron
incandescent lamp. The mixture became red and the colour Lett., 1993, 34, 5725.
slowly faded in 2.25 h. The mixture was washed with 40% 18 K. Smith, G. A. El-Hiti, M. E. W. Hammond, D. Bahzad, Z. Li and
C. Siquet, J. Chem. Soc., Perkin Trans. 1, 2000, 2745.
sodium hydrogensulfite (2 3 10 mL), until a negative starch- 19 A. K. Maiti, R. Martínez, R. Mestres and A. Tortajada, Tetrahedron,
iodide test for peroxides, and water and dried. Evaporation of 2001, 57, 3397.
the solvent gave a crude mixture as an oil which solidified (2.71 20 J. H. Barnes, J. Elks, F. F. Stephens and G. J. Waller, J. Chem. Soc.,
g; 93%). The 1H NMR spectrum of a sample was run. 1953, 764.

316 Green Chemistry, 2002, 4, 314–316