Regioselective nitration of deactivated mono-substituted benzenes using acyl nitrates over reusable acidic zeolite catalysts

Article abstract of DOI:10.1007/s10562-009-0258-7

Nitration of benzonitrile was investigated using a nitric acid/acid anhydride/zeolite catalyst system under different reaction conditions. Trifluoroacetic and chloroacetic anhydrides were found to be the most active among the anhydrides tried. Also, zeolites Hβ and Fe³⁺β (Si/ Al = 12.5) were found to be the most active catalysts. For example, nitration of benzonitrile with trifluoroacetyl nitrate under reflux conditions in dichloromethane gave 3- and 4-nitrobenzonitriles in quantitative yield, of which the para-isomer represented 24-28%. The yield of para-isomer was improved to 33% when passivated Hβ was used under similar reaction conditions. This is easily the most para-selective nitration of benzonitrile ever recorded. Also, no ortho-isomer was formed under the conditions tried. The zeolite can be easily recovered, regenerated by heating and reused up to six times to give results similar to those obtained with a fresh sample of the catalyst. The nitration system was applied successfully to a range of deactivated mono-substituted benzenes to give para-isomers in significantly higher proportions than in the corresponding traditional nitration reactions.

Full text of DOI:10.1007/s10562-009-0258-7

Catal Lett (2010) 134:270–278
DOI 10.1007/s10562-009-0258-7
Regioselective Nitration of Deactivated Mono-Substituted
Benzenes Using Acyl Nitrates Over Reusable Acidic Zeolite
Catalysts
•
Keith Smith Mansour D. Ajarim
•
Gamal A. El-Hiti
Received: 2 December 2009 / Accepted: 16 December 2009 / Published online: 5 January 2010
Ó Springer Science+Business Media, LLC 2010
Abstract Nitration of benzonitrile was investigated using
a nitric acid/acid anhydride/zeolite catalyst system under
different reaction conditions. Trifluoroacetic and chloro-
acetic anhydrides were found to be the most active among
the anhydrides tried. Also, zeolites Hb and Fe3?b (Si/
Al = 12.5) were found to be the most active catalysts. For
example, nitration of benzonitrile with trifluoroacetyl
nitrate under reflux conditions in dichloromethane gave 3-
and 4-nitrobenzonitriles in quantitative yield, of which the
para-isomer represented 24–28%. The yield of para-iso-
mer was improved to 33% when passivated Hb was used
under similar reaction conditions. This is easily the most
para-selective nitration of benzonitrile ever recorded. Also,
no ortho-isomer was formed under the conditions tried.
The zeolite can be easily recovered, regenerated by heating
and reused up to six times to give results similar to those
obtained with a fresh sample of the catalyst. The nitration
system was applied successfully to a range of deactivated
mono-substituted benzenes to give para-isomers in sig-
nificantly higher proportions than in the corresponding
traditional nitration reactions.
1 Introduction
Nitration of aromatic compounds is one of the mostimportant
and widely studied chemical reactions [1–5] because nitroa-
romatic compounds are intermediates for the production of
industrially important chemicals such as polyurethanes,
explosives, agrochemicals, pharmaceuticals, fragrances and
dyes [6, 7].
The traditional method for nitration, involving use of a
mixture of nitric and sulfuric acids, though still in wide-
spread commercial use, suffers from many disadvantages,
including the generation of large quantities of corrosive
waste and in many cases the production of mixtures of
isomers in which the most desirable one is not in high
yield. This has stimulated much research aimed at devel-
opment of alternative procedures that generate less waste
and provide product mixtures in which the most desirable
isomer (often the para-isomer) is much more abundant.
In the case of nitration of monosubstituted benzenes
containing an ortho/para-directing substituent such as an
alkyl or halogeno group much progress has been made [8–
17]. For example, we have shown that nitration of toluene
using a stoichiometric mixture of nitric acid and acetic
anhydride over a reusable zeolite Hb catalyst allows pro-
duction of the para-isomer in 80% yield, with the isomeric
nitrotoluenes and easily recoverable acetic acid as the only
by-products [8, 9].
Keywords Nitration Á Deactivated mono-substituted
benzenes Á Acyl nitrate Á Acidic zeolite catalysis Á
Hb Á Regioselectivity
By contrast, nitration of highly deactivated substrates
such as monosubstituted benzenes having substituents such
as nitro, cyano, acyl or sulfonyl groups, has received much
less attention. In such cases the reactions are usually quite
selective for the meta-isomer, but utilize relatively forcing
conditions and a large quantity of sulfuric acid. Under such
conditions, para-isomers are usually produced in very low
yields [1]. For example, nitration of benzonitrile gives
K. Smith (&) Á M. D. Ajarim Á G. A. El-Hiti
School of Chemistry, Cardiff University, Main Building,
Park Place, Cardiff CF10 3AT, UK
e-mail: smithk13@cardiff.ac.uk
G. A. El-Hiti
e-mail: el-hitiga@cardiff.ac.uk
123

Regioselective Nitration of Deactivated Mono-Substituted Benzenes
271
nitrobenzonitriles in which the ortho/meta/para ratio is
17/81/2 [1].
2 Experimental
Alternative systems developed recently for nitration of
deactivated aromatics include nitric acid in the presence of
P2O5 supported on silica gel [18], nitrogen dioxide at low
temperatures in the presence of ozone [19], N2O5 in the
presence of Fe(acac)3 as a catalyst [20], urea nitrate and
nitrourea [21], nitroguanidine and ethylene glycol dinitrate
in concentrated acid [22], vanadium(V) oxytrinitrate in
dichloromethane at room temperature [23], and nitric acid
in the presence of MCM-41-supported metal bis[(perfluo-
roalkyl)sulfonyl]imides in dichloroethane [24]. However,
such reagents still give mainly the meta-products and
moreover, ortho-isomers are often formed in relatively
high proportions. None of these methods show significant
improvement in the proportion of the para-isomers.
It is well recognised that zeolites and other solid cata-
lysts can play an important role in the development of
greener organic syntheses through their abilities to act as
recyclable heterogeneous catalysts, support reagents,
entrain by-products, avoid aqueous work-ups and enhance
product selectivities [25–32]. For example, we have shown
that zeolites or other solids can have advantages in alkyl-
ation [33], acylation [34, 35], methanesulfonylation [36],
bromination [37] and chlorination [38], as well as nitration
[8–13] of aromatic compounds.
2.1 Materials
Chemicals and solvents were purchased from Aldrich
Chemical Company and used without further purification.
Nitric acid (100%) was purchased from BDH Laboratory
Supplies. Commercial zeolites were purchased from
Aldrich Chemical Company or Zeolyst International. All
zeolite catalysts were freshly calcined at 550 °C for a
minimum of 6 h prior to use.
2.2 Analysis and Characterisation of the Products
Product mixtures from the nitration reactions of deactivated
mono-substituted benzenes were subjected to gas chroma-
tography using a Shimadzu Gas Chromatograph fitted with a
ZEBRON ZB-5 (5% phenyl polysiloxane) 30 m length
column. The GC conditions used for analysis were: 70 °C for
1 min, ramped to 250 °C at 20 °C/min and held for 3 min.
The injection temperature was 250 °C and the detection
temperature 250 °C. Tetradecane was used as a GC standard.
Methyl benzoate and ethyl benzoate were prepared
according to the literature procedure [40], by reaction of
benzoic acid with the appropriate alcohol in the presence of
sulfuric acid as a catalyst, and their structures were con-
firmed by NMR spectroscopy.
Indeed, we have previously shown that nitration of
deactivated substrates using a system comprising nitric
acid, trifluoroacetic anhydride (sometimes with added
acetic anhydride) and zeolite Hb at low temperature
(-10 °C) gives the corresponding nitro products with
higher proportions of the para-isomers, in one case up to as
high as 19% [39], but with a low overall yield. Also, under
some of the conditions tried there was almost no ortho-
isomer. We were interested to see if a process could be
devised that would give even higher yields of para-isomers
for such highly deactivated substrates. Therefore, we
decided to undertake a detailed study of nitration of ben-
zonitrile and other deactivated aromatics over zeolites by
use of acid anhydrides and nitric acid, which are assumed
to produce acyl nitrates in situ (Scheme 1).
GC was the method of choice, unless otherwise indi-
cated, to quantify the product mixtures and to determine
the yield (%) and the total mass balance.
2.3 Typical Experimental Procedure for Zeolite
Cation-Exchange
The standard procedure for cation-exchange involved cal-
cination of supplied Hb (5 g) at 550 °C to remove the
template, then stirring in a refluxing aqueous solution of
the corresponding metal chloride (1 M, 50 mL) for 1 h,
filtration, washing with water (50 mL), repetition of the
whole exchange process then washing with water until
halide-free before calcination at 550 °C.
Such nitrating agents would not react readily with the
substrates in the absence of a catalyst, so that the influence
of the catalyst would be critical. We now report our
findings.
2.4 Procedure for the Surface Passivation of Zeolite
Passivation of external surface of zeolite b was performed
at room temperature by treatment of calcined zeolite b with
an excess of pure trimethylchlorosilane in dichloromethane
(DCM). The mixture was stirred for 5 min and solvent and
excess reagent was removed by a dry nitrogen stream. The
zeolite was dried in an air oven at 120 °C for 6 h then
crushed to a powder. No attempt was made to measure the
internal and external acidity of the zeolite after the silani-
sation process.
(RCO)2O
+
HNO3
RC(O)ONO2
+
RCO2H
R = Me, Et, ClCH2, Cl2CH, CF3
Scheme 1 Generation acyl nitrates from reaction of nitric acid with
acid anhydrides
123

272
K. Smith et al.
2.5 Typical Experimental Procedure for the Nitration
of Deactivated Aromatic Compounds Using Nitric
acid, Trifluoroacetic Anhydride and Zeolite Hb
Catalyst
para-, a higher ortho/para ratio was expected. The low
yield of nitrobenzonitriles (NBNs) prompted us to inves-
tigate the reaction further in an attempt to find conditions
under which the yield could be improved.
The progress of the reaction was tested by conducting
identical reactions over different reaction times from 5 min
to 24 h. After 5 min the yield of 2 was 4% and there was
just a trace of 3, but by 30 min the yield of 2 was 17% and
that of 3 was 4%. The product yields remained constant
thereafter. Clearly, the reaction is quite rapid under these
conditions, with 21% conversion within 30 min, but the
reaction stops for some reason at this point.
Quantities are recorded in the footnotes to the appropriate
tables or text. All reactions were carried out in a 250 mL
two-necked round bottomed flask equipped with a water
condenser, fitted with a calcium chloride tube, and a
magnetic stirrer. In a typical experiment, a mixture of
zeolite Hb (Si/Al = 12.5; 0.10–2.0 g), nitric acid (100%;
0.80 mL, 19 mmol), trifluoroacetic anhydride (3.50 mL,
25 mmol) and the deactivated aromatic compound
(9.5 mmol) in DCM (20 mL) was heated under reflux for
the appropriate reaction time. At the end of the reaction
period, the bulk sample was filtered and the catalyst was
washed with analytical grade acetone (3 9 20 mL).
Tetradecane (0.0600 g) was added as a GC standard and
the solution was made up to 100 mL with acetone. The
mixture was analysed by gas chromatography and the
yields of all identified components were calculated.
A similar series of reactions was conducted in refluxing
DCM. The results obtained are recorded in Table 1.
As the results in Table 1 indicate, for reaction times up
to 4 h the selectivity of the reaction remained more or less
constant (2:3 ca. 4:1), but thereafter the rate of reaction was
very low and virtually all of the additional product was 2. It
seemed likely that this very slow reaction producing almost
exclusively 2 was the intrinsic reaction between 1 and a
nitrating agent (chloroacetyl nitrate or nitric acid) without
intervention of the zeolite, implying that the activity of the
zeolite was lost after a certain reaction time. It is unlikely
that the zeolite was destroyed more quickly and at lower
conversion at room temperature than at ca. 40 °C (reflux-
ing DCM), but the loss of activity could be consistent with
equilibrium formation of an inactive complex between the
zeolite and the chloroacetic acid produced in the reaction
(Scheme 3).
The yields of 3-nitrobenzonitrile (3-NBN), 4-nitro-
benzonitrile (4-NBN) and other products are recorded in
tables. Results from variation of the procedure and use of
different substrates are also recorded in tables. Details of
the conditions are recorded in footnotes to the tables.
3 Results and Discussion
At the higher temperature the equilibrium position
would lie further to the side of the uncomplexed zeolite,
allowing both a more rapid reaction and a higher conver-
sion before the rate of the catalysed reaction was reduced
virtually to zero.
Initially, we attempted nitration of benzonitrile (1;
9.5 mmol) with HNO3 (19 mmol) and chloroacetic anhy-
dride (19 mmol), over zeolite Hb (HBEA, Si/Al = 12.5;
2 g) in dichloromethane (DCM, 20 mL) at room tempera-
ture for 2 h. The reaction produced only 3-nitrobenzonitrile
(2) and 4-nitrobenzonitrile (3), with no ortho-nitrated
product (Scheme 2).
The combined yield of nitrobenzonitriles 2 and 3
was 21% and the 2/3 ratio was 17/4, with ca. 79% of 1
remaining. The absence of ortho-isomer was interesting.
Since the cyano group is linear and provides little steric
hindrance to ortho-nitration, and since traditional nitration
produces significantly more ortho-nitrated product than
Table 1 Nitration of benzonitrile (1) in DCM using (ClCH2CO)2O/
HNO3/Hb for various reaction times under reflux conditions
Reaction time (h)
Yields (%)a
1
2
3
0.1
0.5
2
80
58
32
25
22
20
16
35
54
56
60
63
4
7
15
18
17
17
CN
CN
CN
4
8
(RCO)2O, HNO3, zeolite
solvent
+
24
NO2
A mixture of Hb (Si/Al = 12.5; 2.00 g), nitric acid (0.80 mL,
19 mmol), chloroacetic anhydride (3.30 g, 19 mmol) and 1 (1.03 g,
9.5 mmol) in DCM (20 mL) was refluxed for the appropriate reaction
time
NO2
2
3
1
Scheme 2 Nitration of benzonitrile using a HNO3/(RCO)2O/zeolite
system
a
Yields calculated by quantitative GC using tetradecane as standard
123

Regioselective Nitration of Deactivated Mono-Substituted Benzenes
273
Zeolite
+
RCO2H
Zeolite . RCO2H
Table 3 Nitration of benzonitrile (1) over various zeolite catalysts
according to Scheme 2 under reflux conditions
Scheme 3 Possible deactivation of zeolite by complexation
Catalyst (Si/Al ratio)
Yields (%)a
1
2
3
In order to gain further insight into this possibility, the
reaction was carried out with different quantities of zeolite.
The results obtained are recorded in Table 2.
No catalyst
Hb (12.5)
100
25
–
–
56
46
62
14
27
–
18
11
12
4
The results (Table 2) showed that for amounts of cata-
lyst between 0.5 and 2.0 g the yield of nitrobenzonitriles
(NBNs) obtained after 4 h increased with the amount of
catalyst, though not in a linear way. The result with 3 g of
zeolite probably reflects difficulty in stirring the mixture
efficiently. Qualitatively the results with 0.5–2.0 g of
zeolite were consistent with the possible existence of an
equilibrium of the type depicted in Scheme 3, but without
an independent measure of the quantity of at least one of
the species involved it was not possible to draw any
quantitative conclusions. Therefore, attention was next
turned to attempting to find alternative ways to improve the
yield and proportion of the para-isomer.
Hb (150)
42
Hb (300)
27
HY (12.5)
HY (28)
80
70
2
Ferrierite
100
98
–
H-Mordenite (20)
HZSM-5 (30)
–
–
100
–
–
A mixture of zeolite (2.00 g), nitric acid (0.80 mL, 19 mmol), chlo-
roacetic anhydride (3.30 g, 19 mmol) and 1 (1.03 g, 9.5 mmol) in
DCM (20 mL) was refluxed for 4 h
a
Yields calculated by quantitative GC using tetradecane as standard
Several different acidic zeolites (2 g for 9.5 mmol of 1)
were screened for efficacy in a 4 h reaction. In order to
illustrate the effect of zeolite on yield, the reaction was also
attempted without zeolite. The results obtained are recor-
ded in Table 3.
dimensional channels (zeolites b and Y) showed significant
conversions of benzonitrile into nitrobenzonitriles.
The Si/Al ratio of the zeolites used influenced the
selectivity of the reactions, with the samples having lower
Si/Al ratios giving higher proportions of 4-substituted
product than those with higher Si/Al ratios. Thus, when
zeolite HY (Si/Al = 12.5) was used as a catalyst the 3:2
ratio was 4:14. However, over zeolite HY (Si/Al = 28) the
reaction was much less para-selective, giving a ratio of
2:27. Similarly, with zeolite Hb the 3:2 ratio varied from
18:56 over Hb (Si/Al = 12.5) to 12:62 over Hb (Si/Al =
300). None of the zeolites tried gave a better conversion
than the one originally studied and further studies therefore
concentrated on the use of zeolite Hb (Si/Al = 12.5).
In order to test the effect of solvent in the reaction,
reactions were carried out in several different refluxing
solvents. The results are recorded in Table 4.
In the absence of any catalyst, no reaction occurred.
Also, when the reaction was carried out over zeolite
H-ZSM-5 (a medium pore zeolite), no NBNs were formed,
presumably because the pores are not large enough to
accommodate the reacting species and the number of acidic
sites on the surface is low. The results were very similar
when ferrierite and mordenite were used as catalysts.
Ferrierite and mordenite, although they have relatively
large entry ports to the channels, contain one-dimensional
channel systems, which restricts diffusion within the
channels and inhibits interaction between species. By
contrast, all samples of large pore zeolites with three-
Table 2 Nitration of benzonitrile (1) over various quantities of
zeolite Hb
Table 4 Nitration of benzonitrile (1) according to Scheme 2 over
zeolite Hb (Si/Al = 12.5) in different refluxing solvents
Hb (Si/Al = 12.5; g)
Yields (%)a
Solvent
Reflux temperature (°C)
Yields (%)a
1
2
3
1
2
3
–
100
46
–
–
0.5
1
44
48
56
35
11
14
18
11
Dichloromethane
Chloroform
40
61
83
56
25
37
70
56
48
24
–
18
14
8
37
25
55
2
1,2-Dichloroethane
Acetone
3
100
–
A mixture of Hb (Si/Al = 12.5), nitric acid (0.80 mL, 19 mmol),
chloroacetic anhydride (3.30 g, 19 mmol) and 1 (1.03 g, 9.5 mmol) in
DCM (20 mL) was refluxed for 4 h
A mixture of Hb (Si/Al = 12.5, 2.00 g), nitric acid (0.80 mL,
19 mmol), chloroacetic anhydride (3.30 g, 19 mmol) and 1 (1.03 g,
9.5 mmol) in solvent (20 mL) was refluxed for 4 h
a
a
Yields calculated by quantitative GC using tetradecane as standard
Yields calculated by quantitative GC using tetradecane as standard
123

274
K. Smith et al.
No NBNs were formed when acetone was used as sol-
nitrates in situ (Scheme 1) and then used for nitration of
benzonitrile according to Scheme 2. The results are
recorded in Table 6.
vent. However, for chlorinated hydrocarbon solvents the
reaction proceeded and the para/ortho ratio was similar for
all such solvents. Reflux temperature seemed to be the
major factor influencing the yield of the NBNs. The yields
of 2 and 3 were highest when the reaction was carried out
in the lowest boiling solvent (DCM) and lowest in high-
boiling dichloroethane (DCE). It seemed likely that the
chloroacetyl nitrate intermediate was breaking down or
escaping from the reaction mixture at elevated tempera-
tures. In order to test this possibility, mixtures containing
solvent (chloroform or DCE), zeolite, nitric acid and
chloroacetic anhydride, in the same proportions as used for
the reactions represented in Table 4, but without benzoni-
trile, were heated at reflux for 4 h, after which benzonitrile
was added and the mixtures were then refluxed for a further
4 h. For the case of DCE, no NBNs were formed and 100%
of the benzonitrile remained. In the case of chloroform, just
17% of 3-NBN and 6% of 4-NBN were formed and 76% of
benzonitrile remained. This clearly supports the view that
the active reagent is lost from the mixtures during heating.
It seemed that the best solvent was DCM and had already
been used in the earlier studies.
Clearly, acetyl and propionyl nitrates were not reactive
enough for such a deactivated substrate. The results with
dichloroacetic anhydride were unusual—the para/meta
ratio was relatively high (0.37), but there was also a sig-
nificant quantity of the ortho-isomer.
On the other hand, both chloroacetyl and trifluoroacetyl
nitrates produced only 2 and 3, with trifluoroacetyl nitrate
giving a higher yield than chloroacetyl nitrate but with a
slightly lower para/meta ratio (0.28; cf. 0.32 for chloro-
acetyl nitrate).
Trifluoroacetic anhydride is intrinsically more reactive
than chloroacetic anhydride, so a higher reaction rate might
be expected. However, this alone does not explain why a
higher yield was obtained, since deactivation of the zeolite
rather than lack of intrinsic reactivity appeared to be the
limiting factor with chloroacetyl nitrate. Perhaps the
greater volatility of trifluoroacetic acid, resulting in a sig-
nificant proportion of it being in the vapour phase under the
reaction conditions, and a lower tendency for it to complex
the zeolite, allow a greater amount of free zeolite
(Scheme 3) to remain to catalyse the reaction. If so, it
might also be possible to use a smaller quantity of zeolite
and still obtain a high yield of product. Therefore, a series
of reactions was conducted over various quantities of Hb.
In order to avoid any possibility that hydrolysis of some of
the anhydride would lead to a lower proportion of trifluo-
roacetyl nitrate, the quantity of anhydride was raised to
25 mmol (for 19 mmol of HNO3). The results are shown in
Table 7.
The effect of quantity of chloroacetic anhydride is
shown in Table 5. There was a trend to higher overall
product yields with increasing anhydride amount, but the
amount of para-isomer remained constant. Perhaps, if an
equilibrium such as that shown in Scheme 3 is operating,
after a certain conversion level the activity of the zeolite is
effectively eliminated so that any further reaction proceeds
in a manner more like traditional nitration.
Despite all the different parameters varied, the maxi-
mum yield of para-nitrobenzonitrile achieved using chlo-
roacetic anhydride was 18%. Therefore, attention was
focused on the effect of other anhydrides. Various anhy-
drides with different reactivity (acetic, propionic, chloro-
acetic, dichloroacetic and trifluoroacetic anhydrides) were
reacted with nitric acid to produce the corresponding acyl
The results clearly indicated that nitration of benzoni-
trile under these conditions could be catalysed with as little
as 0.10 g of Hb. The reaction was nearly complete (96%)
and the yield of 4-nitrobenzonitrile was 24%, with a
Table 6 Nitration of benzonitrile (1) according to Scheme 2 over
zeolite Hb (Si/Al = 12.5) using various acid anhydrides
Table 5 Nitration of benzonitrile (1) with various quantities of
chloroacetic anhydride
Acid anhydride
Yields (%)a
1
2
3
Other
Chloroacetic anhydride (mmol)
Yields (%)a
Acetic
100
72
25
34
4
–
–
–
–
1
2
3
Propionic
19
20
25
30
25
19
–
56
60
81
80
18
18
18
18
Chloroacetic
Dichloroacetic
Trifluoroacetic
56
41
75
18
15
21
10b
–
A mixture of Hb (Si/Al = 12.5, 2.00 g), nitric acid (0.80 mL,
19 mmol), acid anhydride (19 mmol) and 1 (1.03 g, 9.5 mmol) in
DCM (20 mL) was refluxed for 4 h
A mixture of Hb (Si/Al = 12.5, 2.00 g), nitric acid (0.80 mL,
19 mmol), chloroacetic anhydride and 1 (1.03 g, 9.5 mmol) in DCM
(20 mL) was refluxed for 4 h
a
Yields calculated by quantitative GC using tetradecane as standard
a
b
Yields calculated by quantitative GC using tetradecane as standard
2-Nitrobenzonitrile was produced in 10% yield
123

Regioselective Nitration of Deactivated Mono-Substituted Benzenes
275
Table 7 Nitration of benzonitrile (1) using TFAA over various
quantities of zeolite Hb
Table 8 Nitration of benzonitrile (1) according to Scheme 2 over
various cation-exchanged zeolites b
Hb (Si/Al = 12.5; g)
Yields (%)a
Entry
Zeolite
Yields (%)a
1
2
3
1
2
3
–
100
4
–
–
1
2
H?b
K?b
–
90
2
76
7
24
–
0.10
0.25
0.50
2.0
4.0
72
77
77
76
77
24
23
23
24
23
–
3
Zn2?b
Cd2?b
Hg2?b
Al3?b
Fe3?b
In3?b
La3?b
Ce3?b
80
78
34
77
17
22
11
22
–
4
–
–
5
54
2
–
6
7
–
77 (73)
85
24 (28)
15
A mixture of Hb (Si/Al = 12.5), nitric acid (0.80 mL, 19 mmol),
TFAA (3.50 mL, 25 mmol) and 1 (1.03 g, 9.5 mmol) in DCM
(20 mL) was refluxed for 4 h
8
–
9
–
87
14
a
Yields calculated by quantitative GC using tetradecane as standard
10
–
85
15
A mixture of zeolite (2.00 g), nitric acid (0.80 mL, 19 mmol), TFAA
(3.50 mL, 25 mmol) and 1 (1.03 g, 9.5 mmol) in DCM (20 mL) was
refluxed for 4 h
para/meta ratio of 0.33. There was no advantage in using
the zeolite in more than 0.25 g quantity for the particular
conditions chosen, since neither the yield nor the selec-
tivity of the reaction was influenced by the amount of the
zeolite beyond that.
Figures in parentheses are for similar reactions, but with 4 g of zeolite
Fe3?b
a
Yields calculated by quantitative GC using tetradecane as an
internal standard
We further investigated the effect of reaction time in the
presence of only 0.1 g of the catalyst. Nitrobenzonitriles
were obtained in 59% yield (46% of 2 and 13% of 3) after
30 min and in 95% yield (72% of 2 and 23% of 3) after 2 h.
Thereafter the remaining 1 was converted very slowly,
eventually giving a mixture of 2 (75%) and 3 (24%) after 6 h.
Samples of zeolite Hb were recovered from the above
reactions, regenerated by calcination at 550 °C and reused
in similar nitration reactions. The yields of 2 and 3
obtained showed that the zeolite could be reused at least
five times with no reduction in either the yields or
selectivity.
yields with the smaller divalent cations could reflect easier
entry to the pore network. The lower yield with Hg2?b
would be consistent with the larger cation causing greater
constriction to access to the pore network. There was also a
slight increase in para-selectivity going from Zn2?b to
Hg2?b as the cations get larger, but at the expense of
overall yield.
All trivalent cations (Entries 6–10) produced quantita-
tive yields of nitrobenzonitriles of which the 4-nitro-
benzonitrile yield was in the range of 14–24%. Fe3?b and
Al3?b gave the best para-selectivities among this group of
zeolites. Furthermore, 4-NBN (3) was obtained in 28%
yield when 4.0 g of Fe3?b zeolite was used compared with
24% when 2 g was used, while use of only 0.1 g of Fe3?b
resulted in 22% of 3, 58% of 2 and 21% unreacted 1.
Therefore, Fe3?b behaved in a manner broadly similar to
that of Hb.
Unfortunately, none of the ion-exchanged zeolites
resulted in a significantly higher proportion of 3 than with
Hb, so attention was turned to the possibility that surface-
passivated zeolite would offer further improvement by
eliminating reaction catalysed at the external surface of the
zeolite and by introducing further constraints on access to
the pores.
The maximum yield of 4-nitrobenzonitrile (3) obtained
thus far was in the region of 24% and it seemed likely that
this was the maximum that could be achieved using Hb as
the catalyst. Therefore, a range of cation-exchanged b
zeolites was prepared for testing as catalysts. For these tests
the conditions used were reflux for 4 h with zeolite (2.0 g),
HNO3 (19 mmol) and TFAA (25 mmol). The results are
reported in Table 8.
The results recorded in Table 8 showed some interesting
features. For example, the yield of nitrobenzonitriles was
only 7%, with no 4-NBN formed, when K?b was used as
catalyst (Entry 2). K? is a large univalent cation and may
block entry to the pores of the zeolite, spoiling its potential
as a catalyst. Also, K?b is not very acidic. The small
amount of nitrobenzonitrile produced may have arisen by
reaction at the small number of sites on the external surface,
thereby reducing the tendency to produce para-isomer.
For the divalent cation-exchanged zeolites the yield of
NBNs was high in the case of Zn2?b and Cd2?b (Entries 3
and 4) but only moderate with Hg2?b (Entry 5). The higher
Samples of Hb and Fe3?b zeolites (each with Si/Al =
12.5) were treated with chlorotrimethylsilane in DCM and
then dried at 120 °C. A series of nitration experiments was
then conducted with various quantities of these passivated
b zeolites over a 2 h period at reflux. The results are
recorded in Table 9.
123

276
K. Smith et al.
R
R
R
R
Table 9 Nitration of benzonitrile (1) according to Scheme 2 over
various quantities of passivated zeolite
NO2
TFAA, HNO3, zeolite
DCM, reflux, 2 h
+
+
Passivated zeolite (g)
Yields (%)a
NO2
NO2
7
1
2
3
4
5
6
Hb (0.10)
40
4
45
65
64
67
67
60
70
71
73
73
15
30
31
33
33
20
24
24
26
28
Scheme 4 Nitration of deactivated aromatics 4 using a HNO3/TFAA/
zeolite system
Hb (0.25)
Hb (0.50)
5
Hb (1.00)
–
Table 10 Nitration of deactivated aromatics 4 using nitric acid and
TFAA over zeolite Hb according to Scheme 4 and comparison with
use of ‘mixed acids’
Hb (2.00)
–
Fe3?b (0.10)
Fe3?b (0.25)
Fe3?b (0.50)
Fe3?b (1.00)
Fe3?b (2.00)
22
6
Entry
R
Yield (%)a
4
5
6
7
2
1
2
3
4
5
6
7
8
9
10
CN
– (17)
– (6)
76 (80)
92 (91)
76 (72)
67 (75)b
52 (27)c
26 (41)
75 (72)
85 (69)
77 (78)
75 (74)
24 (2)
7 (2)
–
NO2
A mixture of passivated zeolite, nitric acid (19 mmol, 0.8 mL), tri-
fluoroacetic anhydride (25 mmol, 3.5 mL) and 1 (9.5 mmol) in DCM
(20 mL) was refluxed for 2 h
CHO
10 (22)
– (–)b
14 (1)
15 (–)b
30 (–)c
36 (26)
21 (1)
10 (3)
12 (3)
13 (2)
COMe
COBu
COBut
CO2H
CO2Me
CO2Et
CO2Bu
10 (18)c
38 (33)
– (27)
4 (20)
8 (20)
10 (20)
a
Yields calculated by quantitative GC using tetradecane as an
internal standard
With either of the catalysts the reaction was more or less
complete within the 2 h reflux period provided at least
0.25 g of the passivated catalyst was used. In the case of
passivated Fe3?b there was only a small increase in the
proportion of 3, but with passivated Hb the proportion rose
to over 30%. With 1 g of passivated Hb under these con-
ditions the reaction was complete, giving 67% of 2 and
33% of 3, which is easily the highest proportion of 3 ever
achieved in nitration of benzonitrile.
A mixture of Hb (Si/Al = 12.5, 0.25 g), nitric acid (0.80 mL,
19 mmol), TFAA (3.50 mL, 25 mmol) and 4 (9.5 mmol) in DCM
(20 mL) was refluxed for 4 h
Figures in parentheses are for similar reactions using mixed acids. To
a stirred mixture of HNO3 (9.5 mmol) and H2SO4 (19 mmol) the
appropriate substrate (9.5 mmol) was added and the mixture stirred
for 10 min at 0 °C
At this point it was appropriate to investigate the scope
of the new process with other deactivated mono-substituted
benzenes. Therefore, a series of substrates was subjected to
nitration in three different ways—using a traditional mixed
acid procedure to provide a baseline for ‘‘normal’’ nitra-
tion, using the TFAA/nitric acid/zeolite Hb approach
(Scheme 4) and using the latter approach with passivated
zeolite Hb. The results are shown in Table 10 (mixed acid
and zeolite Hb approaches) and 11 (using passivated Hb).
Tables 10 and 11 show that nitration of deactivated aro-
matics 4 was general and gave quantitative yields of nitro
products in most cases. The results also revealed that both
types of zeolite Hb, but especially the passivated one,
enhanced the para-selectivity in all such nitration reactions.
Nitration of benzonitrile, butyl phenyl ketone, tert-butyl
phenyl ketone and benzoic acid gave the highest proportions
of para-isomers, several being over 30%. Furthermore, in all
cases except one (tert-butyl phenyl ketone) the ortho-prod-
ucts were obtained in lower proportions than by traditional
nitration, in several cases the amount being so low as to be
unmeasurable. The proportion of the meta-isomer was in
most cases not very different than that produced in the tra-
ditional nitration reaction, the increase in the amount of
a
Yields normally calculated by quantitative GC using tetradecane as
standard
b
1
Yields calculated by H NMR spectroscopy
c
Reaction was carried out using only nitric acid rather than mixed
acid; starting material (ca. 55%) remained. A similar reaction with
mixed acid was aggressive and none of the expected products was
detected by GC
para-isomer having been more or less matched by a decrease
in the amount of the ortho-isomer.
The case of tert-butyl phenyl ketone was an exception to
the general trend. As for the other substrates the proportion of
para-isomer had increased for the zeolite-catalysed process,
but uniquely the proportion of ortho-isomer had also
increased and in consequence the meta-isomer had become
the minor product. The product proportions with this sub-
strate are unusual even for traditional nitration of an acyl-
benzene and this has been attributed to the bulk of the
tert-butyl group forcing the carbonyl group to twist out of
alignment with the benzene ring [41, 42]. It is possible that
within the confines of the zeolite pores the twisting out of
alignment is further enforced. This substrate would warrant
further investigation.
123

Regioselective Nitration of Deactivated Mono-Substituted Benzenes
277
previously from deactivated substrates. They demonstrate
how careful choice of catalyst, reagent, solvent and condi-
tions can have a powerful effect in improving para-selec-
tivity even with substrates that normally give very little
para-product.
Table 11 Nitration of deactivated aromatics 4 using nitric acid and
TFAA over passivated zeolite Hb according to Scheme 4
Entry
R
Yield (%)a
5
6
7
1
2
CN
–
67
90
73
70b
50
26
78
85
78
72
33
8
Acknowledgments We thank Cardiff University and the Saudi
Government for financial support.
NO2
–
3
CHO
12
b
–
15
17b
33
37
24
10
13
14
4
COMe
COBu
COBut
CO2H
CO2Me
CO2Et
CO2Bu
5
10
37
–
References
6
7
1. Olah GA, Malhotra R, Narang SC (1989) Nitration: methods and
mechanisms. VCH, New York
2. Schofield K (1980) Aromatic nitration. Cambridge University
Press, Cambridge
3. Taylor R (1990) Electrophilic aromatic substitution. Wiley,
Chichester
4. Ingold CK (1969) Structure and mechanism in organic chemistry.
Bell, London
5. de la Mare PB, Ridd JH (1959) Aromatic substitution: nitration
and halogenation. Butterworths, London
6. Gigante B, Prazeres A, Marcelo-Curto M, Cornelis A, Laszlo P
(1995) J Org Chem 60:3445
8
4
9
8
10
11
A mixture of passivated Hb (Si/Al = 12.5, 0.25 g), nitric acid
(0.80 mL, 19 mmol), TFAA (3.50 mL, 25 mmol) and 4 (9.5 mmol)
in DCM (20 mL) was refluxed for 4 h
a
Yields normally calculated by quantitative GC using tetradecane as
standard
b
1
Yields calculated by H NMR spectroscopy
7. Carvalheiro B, Laszlo P, Cornelis A, Marcelo-Curto M (1994)
PCT Inc Appl WO 94/19310
4 Conclusion
8. Smith K, Musson A, DeBoos GA (1998) J Org Chem 63:8448
9. Smith K, Musson A, DeBoos GA (1996) Chem Commun 469
10. Smith K, Almeer S, Black SJ (2000) Chem Commun 1571
11. Smith K, Almeer S, Peters C (2001) Chem Commun 2748
12. Smith K, Almeer S, Black SJ, Peters C (2002) J Mater Chem
12:3285
13. Smith K, Ajarim MD, El-Hiti GA (2009) Topics Catal 52:1696
14. Kalbasi RJ, Ghiaci M, Massah AR (2009) Appl Catal A 353:1
15. Kuba MG, Prins R, Pirngruber GD (2007) Appl Catal A 333:24
16. Kuba MG, Prins R, Pirngruber GD (2007) Appl Catal A 333:78
17. Peng X, Suzuki H, Lu C (2001) Tetrahedron Lett 42:4357
18. Hajipour AR, Ruoho AE (2005) Tetrahedron Lett 46:8307
19. Suzuki H, Murashima T (1994) J Chem Soc Perkin Trans 1:903
20. Bak RR, Smallridge AJ (2001) Tetrahedron Lett 42:6767
21. Almog J, Klein A, Sokol A, Sasson Y, Sonenfeld D, Tamiri T
(2006) Tetrahedron Lett 47:8651
22. Oxley JC, Smith JL, Moran JS, Canino JN, Almog J (2008)
Tetrahedron Lett 49:4449
23. Dove MFA, Manz B, Montgomery J, Pattenden G, Wood SA
(1998) J Chem Soc Perkin Trans 1:1589
24. Yuan Y, Nie J, Zhang Z, Wang S (2005) Appl Catal A Gen
295:170
25. Clark JH (2002) Acc Chem Res 35:791
26. Clark JH (ed) (1994) Catalysis of organic reactions using sup-
ported inorganic reagents. VCH, New York
Proton and cation-exchanged forms of zeolites are able to
catalyse and improve the para-regioselectivity in the
nitration reaction of benzonitrile using a nitric acid/acid
anhydride/zeolite system. No reaction takes place with the
same system in the absence of the zeolite. The catalysed
reaction produces only 3- and 4-nitrobenzonitriles in
quantitative yield, with no 2-nitrobenzonitrile produced in
most cases under the conditions tried. Trifluoroacetic and
chloroacetic anhydrides were found to be the most useful
of the anhydrides tried. Zeolite Hb with a Si/Al ratio of
12.5 is the most active of the catalysts tried and also gives a
high proportion of 4-nitrobenzonitrile. Furthermore, heat-
ing easily regenerates the zeolite, which can be reused at
least 6 times to give results similar to those obtained with a
fresh sample of the catalyst. Passivation of the zeolite
surface by treatment with chlorotrimethylsilane renders the
reaction even more prone to produce the para-isomer,
which is formed in 33% yield, the para/meta ratio
becoming 0.50.
27. Delaude L, Laszlo P, Smith K (1993) Acc Chem Res 26:607
28. Smith K (1992) In: Yoshida Z, Ohshiro Y (eds) New aspects of
organic synthesis II. Kadonsha, Tokyo and VCH, Weinheim, p 43
29. Butters M (1992) In: Smith K (ed) Solid supports and catalysts in
organic synthesis. Ellis Harwood, Chichester, p 130
30. Smith K (1991) In: Scaros MG, Prunier ML (eds) Catalysis of
organic reactions. Marcel Dekker, New York, p 91
31. Smith K, El-Hiti GA (2004) Curr Org Synth 1:253
32. Smith K, El-Hiti GA (2006) Curr Org Chem 10:1603
33. Smith K, Roberts SD, El-Hiti GA (2003) Org Biomol Chem
1:1552
The zeolite-catalysed process is general for nitration of
monosubstituted deactivated aromatic compounds and
gives significantly increased proportions of para-substi-
tuted isomers compared with the results obtained from the
traditional mixed acid method. In most cases the proportion
of ortho-isomer produced is substantially reduced.
Although the reactions described here are not sufficiently
selective to render these reactions suitable for commercial
application, they nevertheless produce substantially higher
proportions of para-nitro compounds than has been possible
34. Smith K, El-Hiti GA, Jayne AJ, Butters M (2003) Org Biomol
Chem 1:1560
123

278
K. Smith et al.
35. Smith K, El-Hiti GA, Jayne AJ, Butters M (2003) Org Biomol
Chem 1:2321
39. Smith K, Gibbins T, Millar RW, Claridge RP (2000) J Chem Soc
Perkin Trans 1:2753
36. Smith K, Ewart GM, El-Hiti GA, Randles KR (2004) Org Biomol
Chem 2:3150
40. Vogel AI (1989) Vogel’s textbook of practical organic chemistry,
5th edn. Longman, Harlow
37. Smith K, El-Hiti GA, Hammond MEW, Bahzad D, Li Z, Siquet C
(2000) J Chem Soc Perkin Trans 1:2745
41. Barker SD, Norris RK, Randles D (1981) Aust J Chem 34:1875
42. Moodie RB (1982) Organic reaction mechanisms 231
38. Smith K, Butters M, Paget WE, Goubet D, Fromentin E, Nay B
(1999) Green Chem 1:83
123