The use of supported zinc bromide for the fast and selective bromination of aromatic substrates

Article abstract of DOI:10.1021/op9800156

Zinc bromide supported on acid activated montmorillonite (K-10) or mesoporous silica (100 A) is a fast, selective catalyst for the para-bromination of activated and mildly deactivated aromatic substrates. The optimum loading of zinc bromide on K-10 is 1.25 mmol/g and 1.75 mmol/g on the higher surface area silica (100 A). Thermal activation of these catalysts at 200 °C results in optimum activity and selectivity. Also, system optimisation has allowed harmful chlorinated solvents to be replaced by less damaging hydrocarbon solvents.

Full text of DOI:10.1021/op9800156

Organic Process Research & Development 1998, 2, 245−249
The Use of Supported Zinc Bromide for the Fast and Selective Bromination of
Aromatic Substrates
Joanne C. Ross, James H. Clark,* Duncan J. Macquarrie, Simon J. Barlow, and Tony W. Bastock‡
†
,†
†
‡,§
Department of Chemistry, UniVersity of York, Heslington, York YO1 5DD, UK, and Contract Catalysts Ltd.,
Knowsley Industrial Park, Prescot, Merseyside L34 9HY UK
Abstract:
Scheme 1. General bromination reaction
Zinc bromide supported on acid activated montmorillonite (K-
1
0) or mesoporous silica (100 Å) is a fast, selective catalyst for
the para-bromination of activated and mildly deactivated
aromatic substrates. The optimum loading of zinc bromide on
K-10 is 1.25 mmol/g and 1.75 mmol/g on the higher surface
area silica (100 Å). Thermal activation of these catalysts at
2
00 °C results in optimum activity and selectivity. Also, system
and the large quantities of catalyst required for good
optimisation has allowed harmful chlorinated solvents to be
replaced by less damaging hydrocarbon solvents.
12,13
regioselectivity.
Other alternatives to the traditional
14,15
Lewis acids include the use of BrF.
Supported reagents have great potential as environmen-
tally friendly alternatives to the more wasteful traditional
catalysts. Supported reagents usually have large surface
areas and are often layered or porous. These properties offer
a large number of catalytic sites within a small quantity of
catalyst, even if loadings are low. We have discovered a
fast, effective method for the bromination of activated and
moderately deactivated aromatic substrates by using sup-
ported zinc bromide as a catalyst (Scheme 1). This paper
discusses the optimisation of the system with regards to
catalyst loading, thermal activation, support, and solvent
effects.
Introduction
Bromoaromatics are widely used as intermediates in the
manufacture of pharmaceuticals, agrochemicals, and other
speciality chemical products. The bromination of aromatic
compounds by electrophilic substitution has been extensively
investigated in the past. Traditional methods use catalysts
such as ferric and aluminium halides, but there are many
disadvantages with these processes such as difficulties in
separation, the production of byproducts from side reactions,
and the lack of regioselectivity. Clays have been established
as solid acid catalysts for over 60 years, initially being used
1
6
1,2
3
as catalysts for petroleum refining. Improvements have been
Results and Discussion
made to these catalysts with the development of the more
robust pillared clays. More recently, advances towards clean
To optimise the catalytic system, the bromination of
bromobenzene was used as a model reaction. The catalytic
activity and selectivity of supported zinc bromide can be
influenced by various parameters such as zinc loading,
solvent, support, and thermal activation.
Effect of Loading. Loading is known to have a major
effect on catalyst activity; therefore, it is very important to
ensure that the optimum loading is used. Underloading can
result in the need for excess catalyst quantities thus decreas-
ing the efficiency of the catalyst. Overloading increases the
risk of leaching and the presence of unreactive, excess
reagent can result in loss of catalyst activity.
4
synthetic methods have been made with the introduction of
5
6
7-13
silica, alumina, and microporous zeolites as catalysts.
These newer processes are a step towards waste minimisation
but still have disadvantages with regards to cost efficiency
†
University of York.
Contract Catalysts Ltd.
Current address: Glycosynth, 14 Craven Ct., Winwic Quay, Warrington,
‡
§
Cheshire WA2 8QU, UK.
(1) Olah, G. A. Acc. Chem. Res. 1971, 4, 240.
(2) Pearson, D. E.; Pope, H. W.; Hargrove, W. W.; Stamper, W. E. J. Chem.
Soc. 1958, 23, 1412.
(3) Smith, K. Solid Supports and Catalysts in Organic Synthesis; Ellis Harwood
Ltd.: Chichester, UK, 1992.
2
For this study, various loadings of ZnBr on silica (100
(4) Barrer, R. M.; MacLeod, D. M. Trans. Faraday Soc. 1955, 51, 1290.
(5) Yaroslavsky, C. Tetrahedron Lett. 1974, 38, 3395.
(6) Ranu, B. C.; Sarkar, D. C.; Chakraborty, R. Synth. Commun. 1992, 22, 1095.
(7) Wortel, Th. M.; Oudijn, D.; Vleugel, C. J.; Roelofsen, D. P.; van Bekkum,
H. J. Catal. 1979, 60, 110.
Å) (silizib) and K-10 (clayzib) were used in the bromination
of bromobenzene. For the model reaction studies, carbon
tetrachloride was used as the reaction solvent. The optimum
loading for the K-10 analogue was found to be significantly
(
8) Zabicky, J.; Eger, I.; Galun, A.; Mhasalkar, M. Zeolites 1987, 7, 499-502.
(9) de la Vega, F.; Sasson, Y. Zeolites 1989, 9, 418.
(
(
(
(
10) de la Vega, F.; Sasson, Y. Zeolites 1991, 11, 617.
11) de la Vega, F.; Sasson, Y. Zeolites 1993, 13, 341.
12) Smith, K.; Bahzad, D. J. Chem. Soc., Chem. Commun. 1996, 467.
13) Ratnasamy, P.; Singh, A. P.; Sharma, S. Appl. Catal. A Gen. 1996, 135,
(14) Rozen, S.; Brand, M. J. Chem. Soc., Chem. Commun. 1987, 752.
(15) Rozen, S.; Brand, M.; Lidor, R. J. Org. Chem. 1988, 53, 5545.
(16) Clark, J. H.; Ross, J. C.; Macquarrie, D. J.; Bastock, T. W.; Barlow, S. J.
J. Chem. Soc., Chem. Commun. 1997, 1203.
2
5.
S1083-6160(98)00015-2 CCC: $15.00 © 1998 American Chemical Society and Royal Society of Chemistry
Published on Web 05/21/1998
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
245

Table 1. Effect of Catalyst Support
conversion of
bromobenzeneb
(GC %)
selectivity
(p/o ratio)
supporta
only
ZnBr
2
0.0
9.7
85.0
2.3
Tonsil 13
3.6
7.7
K-10
Peruvian montmorillonite
acid-activated Peruvian
montmorillonite
silica (100 Å)
silica (60 Å)
alumina (acidic)
alumina (neutral)
HMS (16 Å)
HMS (21 Å)
HMS (24 Å)
HMS (35 Å)
Aerosil
92.3
9.2
88.8
90.7
87.1
1.0
77.0
91.7
88.1
49.7
45.6
0.0
7.8
6.6
7.3
Figure 1. Effect of loading on catalyst activity.
5.4
5.3
5.0
5.8
6.4
zirconia
a
5 °C, 4 h, hexane, 1.0 mmol/g ZnBr
2
loading. b Reaction without a catalyst
2
results in only a 20% conversion of bromobenzene in 24 h with a p/o ratio of
.1.
4
suitable supports that give good selectivities and conversions.
1
8
The use of hexagonal mesoporous silicas gives reasonable
conversions but surprisingly low selectivities. The regular
pores and channels within the structure of the support must
not be small enough to provide shape selectivity as found
with microporous zeolites. Using the nonporous silica
Aerosil as a support results in reasonable but lower conver-
sions and selectivities than the mesoporous silica (100 Å).
This is evidence that the reaction can occur on the surface
of the catalyst as well as in the mesopores.
Figure 2. Effect of loading on surface area.
17
lower than that for the higher surface area silica gel. K-10
has an optimum loading at 1.25 mmol/g, whereas the silica
1
2
(100 Å) optimum loading is 1.75 mmol/g (Figure 1).
The high surface area of most supported reagents is
generally considered to be one of their assets. Loading has
been found to have an effect on the overall surface area of
the catalyst. The surface area of the catalyst decreases with
increasing loading (Figure 2). This is expected as the zinc
bromide is filling up the pores of the support. For clayzib,
the surface area levels off at loadings greater than 1.5 mmol/
g, presumably because the excess zinc bromide is clustering
on the external surface. Therefore, a maximum in activity
and selectivity is achieved when there is an even distribution
of the zinc bromide within the support resulting in the
optimum number of active catalytic sites.
Effect of Support. Supports can be chosen for specific
reactions depending on their surface area, acidity, and
porosity. Particle size and crystallinity can also have a
significant effect on reactivity. The actual molecular struc-
ture of the support is a very important factor as regular pore
sizes and channel structures will often lead to shape
selectivity in the reaction. This minimises the waste products
that are often produced with more traditional catalysts.
The material used to support the zinc bromide has an
effect on the relative rate and selectivity of the reaction
Effect of Solvent. While carbon tetrachloride has com-
monly been used as a solvent in many halogenation reactions,
environmental legislation will preclude its use in industry.
Furthermore, the role of solvent can be very significant in
19,20
reactions catalysed by supported reagents.
Therefore, we
have carried out a study on a variety of solvents.
For the chlorinated solvents (Table 2), there is a correla-
tion between reaction rate and polarity (i.e., carbon tetra-
chloride is the least polar and, therefore, the fastest). This
results in the expected inverse correlation with selectivity
(i.e., the most polar is the most selective). For the alkane
solvents (even less polar, so therefore even less selective),
the reaction rates and selectivities are very similar and appear
to be independent of chain length. It is also interesting to
note that the reaction in neat bromobenzene is faster and
therefore less selective. This is probably due to a concentra-
tion effect. From these results, it was decided that hexane
would be the best solvent to use for all future studies.
Effect of Zinc Counterion. To study the effects of the
zinc salt counterion, various zinc (II) salts were supported
on silica (100 Å) (Table 3) and K-10 (Table 4) and used in
the bromination of bromobenzene.
(Table 1). In the absence of support, zinc bromide does not
catalyse the reaction. The use of raw montmorillonite (i.e.,
before acid activation) as a support results in low conversions
and poor selectivities. After acid activation of the mont-
morillonite support, the reaction results in good conversions
and selectivities. Silica (100 Å) and acidic alumina are also
The best activity and selectivity for the bromination of
bromobenzene is achieved when using either zinc(II) chloride
(
(
18) Tanev, P. T.; Pinnavaia, T. J. Science 1995, 267, 865.
19) Macquarrie, D. J.; Jackson, D. B J. Chem. Soc., Chem. Commun. 1997,
1781.
(17) Clark, J. H.; Cullen, S. R.; Barlow S. J.; Bastock, T. W. J. Chem. Soc.,
Perkin Trans. 2 1994, 1117.
(20) Macquarrie, D. J. J. Chem. Soc., Chem. Commun. 1997, 601.
246
•
Vol. 2, No. 4, 1998 / Organic Process Research & Development

Table 2. Effect of Reaction Solvent21
conversion of
bromobenzene
selectivity
(p/o ratio)
solventa
(GC %)
bromobenzene
100.0
32.8
43.8
55.9
79.9
77.2
78.9
78.4
9.2
13.3
12.7
12.0
11.3
10.5
10.6
10.5
CH
CHCl
CCl
2
Cl
3
2
4
pentane
hexane
octane
decane
a
2
5 °C, 1 h, 1.75 mmol/g ZnBr
2
on silica (100 Å).
Figure 3. Effect of thermal activation on activity of 1.0 mmol/g
clayzib.
Table 3. Silica (100 Å) as Support
conversion of
bromobenzene
(GC %)
time
(h)
selectivity
(p/o ratio)
zinc salta
Zn(SO
Zn(CH
Zn(NO
ZnBr
4
)‚7H
CO )‚2H
‚6H
2
O
7
7
7
2
3
7.3
11.1
44.9
97.8
90.4
3.6
6.5
8.3
9.9
10.7
3
2
2
O
3
)
2
2
O
2
ZnCl
2
a
2
5 °C, hexane, 1.75 mmol/g loading.
Table 4. K-10 as Support
conversion of
bromobenzene
(GC %)
time
(h)
selectivity
(p/o ratio)
zinc salta
Figure 4. Effect of thermal activation on activity of 1.75
mmol/g silizib.
Zn(SO
Zn(CH
Zn(NO
ZnBr
4
)‚7H
CO )‚2H
‚6H
2
O
7
7
7
3
3
10.6
14.6
67.1
94.2
88.8
3.8
6.7
7.9
9.2
10.1
For this investigation, both silizib and clayzib were
activated at a range of different temperatures (100-500 °C).
They were then used in the bromobenzene model reaction
system.
For clayzib, low temperatures of activation, result in a
low conversion of bromobenzene (Figure 3). This is
probably due to the presence of relatively large quantities
of water in the catalyst. There is a maximum in the catalyst
activity and selectivity at temperatures between 200 and 250
3
2
2
O
3
)
2
2
O
2
ZnCl
2
a
2
5 °C, hexane, 1.0 mmol/g loading.
or bromide. Of the other salts, zinc(II) nitrate is the best,
but still has low conversion and selectivity even after 7 h.
The zinc acetate and sulfate both result in very poor
conversions and low selectivities. This effect might be
related to poor dispersion of the zinc ions on the catalyst
surface. These trends are independent of support and are
very similar to previous results found for the alkylation of
benzenes using clayzic.22
Effect of Thermal Activation. Pretreatment is an
important step in the preparation of supported reagents. It
has been shown on many occasions that a low level of water
is beneficial to the activity of supported reagents. The most
common form of treatment is thermal activation which
controls the quantity of water present in the reagent although
other, more profound changes cannot be precluded. The
quantity of water required depends entirely on the reaction
being carried out and can be modified accordingly.
°C. Above these temperatures, the activity begins to decrease
again and may be due to degradation of the clay structure at
high temperatures. Similar behaviour has been seen in the
2
3,24
cases of clayzic and claycuc.
Silizib has a similar initial profile to clayzib with low
conversions at low temperatures (Figure 4). Again there is
a peak in activity and selectivity at about 200 °C. Above
this temperature, the activity and selectivity remains fairly
constant. It is not until activation temperatures reach 500
°
C that a decrease in activity is observed. This is due to the
25
greater thermal stability of silica with respect to clay.
Thermal activation of the catalysts result in changes in
surface area. The surface area of silica slightly increases
(23) Clark, J. H.; Cullen, S. R.; Barlow, S. J.; Bastock, T. W. J. Chem. Soc.,
Perkin Trans. 2 1994, 411.
(
21) Cyclohexane and cyclooctane resulted in similar selectivities but lower
activities than the straight chain alkanes. The reason for this is as yet
unexplained.
(24) Barlow, S. J.; Clark, J. H.; Darby, M. R.; Kybett, A. P.; Landon, P.; Martin,
K. J. Chem. Res. 1991, 74.
(25) Vansant, E. F.; Van der Voort, P.; Vrancken, K. C. Studies in Surface Science
and Catalysis. Vol. 93-Characterisation and Chemical Modification of the
Silica Surface; Elsevier: Oxford, 1995.
(22) Clark, J. H.; Kybett, A. P.; Macquarrie, D. J.; Barlow, S. J.; Landon, P. J.
Chem. Soc., Chem. Commun. 1989, 1353.
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
247

Table 5. Bromination of Other Substrates
conversion of
time bromobenzene selectivity
substratea
catalyst
(min)
(GC %)
(p/o ratio)
t-BuPh
t-BuPh
none
2
2
5
5
2
2
5
5
15
15
80
80
0.0
100.0
0.0
68.5
0.8
100.0
0.0
75.0
0.0
ZnBr
none
ZnBr
none
ZnBr
none
ZnBr
none
ZnBr
none
ZnBr
2
-silica
2
-silica
2
-silica
2
-silica
2
-silica
2
-silica
∞
PhC
PhC
2
H
H
5
2
5
2.4
2.0
2.7
PhCH
PhCH
3
3
C
C
6
H
H
6
6
6
PhF
PhF
PhCl
PhCl
Figure 5. Temperature vs surface area for silica and 1.75
mmol/g silizib.
78.9
0.0
92.9
86.6
10.0
a
25 °C, hexane, 1.75 mmol/g loading.
Conclusions
Zinc bromide supported on either acid-activated mont-
morillonite or silica (100 Å) is a fast, selective catalyst for
aromatic bromination. The optimum loading of zinc bromide
is 1.75 mmol/g on silica (100 Å) and 1.25 mmol/g on the
acid-activated montmorillonite. Dichloromethane, when used
as a solvent, gives the highest selectivity but the lowest
reaction rate. Thermal activation temperatures of 200 °C
are also required for optimum catalyst activity and selectivity.
Figure 6. Temperature vs surface area for K-10 and 1.0
mmol/g clayzib.
with increasing activation temperature (Figure 5). The
surface area of silizib follows the same pattern but at lower
values. The low surface areas are probably due to, first, the
zinc bromide blocking up the pores and, second, the presence
of water within the pores (attracted by the zinc).
Experimental Section
General. Specific surface area data were determined by
the adsorption of dinitrogen at 77 K (calculated by applica-
tion of the Brunauer, Emmett, and Teller (BET) isotherm
using a Coulter SA3100 analyser). Gas chromatography
samples were recorded using a Hewlet Packard HP6890 gas
chromatograph with HP5 capillary column. Mass spectra
were obtained using a Varian 3400CX gas chromatograph
with DB5 capillary column interfaced to a Finnigan Mat
Magnum mass spectrometer.
Catalyst Preparation. Zinc bromide (10 mmol) was
dissolved in methanol (150 mL). Support (10 g) was added,
and the mixture was briefly stirred. The solvent was then
slowly removed by rotary evaporation at 35 °C for 1 h. All
catalysts were activated at 200 °C in a variable-temperature
Gallenkamp oven unless otherwise stated.
General Bromination Procedure. The catalyst (0.6 g),
solvent (15 mL), and substrate (10 mmol) were placed in a
two-necked 25 mL round-bottomed flask (blacked out from
light), and the mixture was magnetically stirred at 700 rpm.
Bromine (10 mmol) was added, and the reaction was carried
out at 25 °C. Samples (0.2 mL) were taken at regular
intervals and quenched with a saturated solution of sodium
thiosulfate (1 mL). The organic layer was then extracted
with diethyl ether (3 mL) and analysed by GC.
Materials. Chromatography grade silica Keiselgel 100
was purchased from Merck, and the acid-activated mont-
morillonite, K-10, was obtained from S u¨ d Chemie. Zinc
bromide was purchased from Aldrich at 98% purity and
stored under a dry atmosphere. The solvents hexane (HPLC
grade), methanol (AR grade), diethyl ether (AR grade),
Again, the surface area of K-10 slightly increases up to
2
00 °C and then levels out (Figure 6). These results are
26
indicative of an aged clay. The surface area of clayzib,
however, increases quite dramatically with increasing activa-
tion temperature. This is consistent with previous results
found for clayzic, where the surface area increases linearly
with the activation temperature.1
7
Other Substrates. This optimised system can be applied
to the bromination of other substrates (Table 5). The
catalytic system works best for activated and moderately
deactivated substrates. Strongly deactivated substrates are
not brominated with these systems.
For all of the substrates, the reaction rate is dramatically
increased with the introduction of the catalyst. Chloroben-
zene has similar selectivity to bromobenzene, whereas
fluorobenzene and tert-butylbenzene give extremely high
para/ortho ratios. The very selective nature of tert-butyl-
benzene is due to the steric hindrance of the tert-butyl
functional group. However, for fluorobenzene, the fluorine
group has high electronegativity which results in a powerful
inductive effect and the enhanced deactivation of the ortho-
position.
(
26) Rhodes, C. N. The Characterisation of Acid-Treated Montmorillonite Clays
as Catalysts and as Supports for Lewis Acid Catalysts; Leeds Metropolitan
University: Leeds, UK, 1994.
248
•
Vol. 2, No. 4, 1998 / Organic Process Research & Development

dichloromethane (HPLC grade), chloroform (AR grade),
cyclohexane (AR grade), and carbon tetrachloride (AR grade)
were all from Fischer Scientific. All other reagents were
from Aldrich.
Fellowship to J.H.C.), and the Royal Society (for a University
Research Fellowship to D.J.M.) for financial support and
other members of the Envirocats and York Green Chemistry
Groups for helpful discussions.
Acknowledgment
We thank Contract Chemicals Ltd., the EPSRC and the
Royal Academy of Engineering (for a Clean Technology
Received for review February 9, 1998.
OP9800156
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
249