Improved regioselective mononitration of toluene over ZSM-5 zeolite catalysts

Article abstract of DOI:10.1016/j.jcat.2007.11.001

We show for the first time that the NH⁺₄ exchanged form of ZSM-5 zeolite is as effective as the H⁺ form of the catalyst for the regioselective conversion of toluene to mononitrotoluene (MNT) using 90% nitric acid as the sole nitrating agent. The auto-ionization of the acid in the absence of protons on the zeolite surface is sufficient for the formation of reactive nitronium ions. We also find that the regioselectivity for the formation of p-MNT is substantially increased over both forms of the zeolite catalyst by first intercalating the acid in the zeolite micropores prior to the introduction of toluene. For instance, the p / o isomer ratio is increased from values in the range 1.3-2.3 under conventional batch reaction conditions to values of 1.8-8.9 when the acid is sequestered in the zeolite. The sequestration of nitric acid in both ion exchanged forms of the zeolite confines more of the nitration reaction to the regioselective environment of the micropores and reduces the extent of reaction in homogeneous toluene solution. However, the ammonium form of the zeolite is preferred over the protonated form when the Si/Al ratio of the zeolite allows for the presence of at least one ammonium ion per unit cell (Si/Al ≤ 40), because under these conditions far less benzaldehyde and other undesired reaction products are formed in comparison to the protonated form of the zeolite at the same Si/Al ratios. The ability of NH⁺₄-ZSM-5 derivatives to limit the formation of unwanted toluene oxidation products is not well understood, but the observed selectivity may be related to the replacement of ammonium ions by protons, the formation of ammonium nitrate, and the buffering of nitric acid within the micropores of the zeolite.

Full text of DOI:10.1016/j.jcat.2007.11.001

Journal of Catalysis 253 (2008) 289–294
www.elsevier.com/locate/jcat
Improved regioselective mononitration of toluene over ZSM-5
zeolite catalysts
a
b,∗
c
Seong-Su Kim , Thomas J. Pinnavaia , Reddy Damavarapu
a
Claytec, Inc. East Lansing, MI 48823, USA
b
Department of Chemistry, Michigan State University, East Lansing, MI 48823, USA
c
Energetic Materials Division, U.S. Army ARDEC, Bldg. 3028, Picatinny Arsenal, NJ 07806, USA
Received 13 June 2007; revised 19 August 2007; accepted 3 November 2007
Available online 11 December 2007
Abstract
+
+
We show for the first time that the NH exchanged form of ZSM-5 zeolite is as effective as the H form of the catalyst for the regioselective
4
conversion of toluene to mononitrotoluene (MNT) using 90% nitric acid as the sole nitrating agent. The auto-ionization of the acid in the absence
of protons on the zeolite surface is sufficient for the formation of reactive nitronium ions. We also find that the regioselectivity for the formation
of p-MNT is substantially increased over both forms of the zeolite catalyst by first intercalating the acid in the zeolite micropores prior to
the introduction of toluene. For instance, the p/o isomer ratio is increased from values in the range 1.3–2.3 under conventional batch reaction
conditions to values of 1.8–8.9 when the acid is sequestered in the zeolite. The sequestration of nitric acid in both ion exchanged forms of the zeolite
confines more of the nitration reaction to the regioselective environment of the micropores and reduces the extent of reaction in homogeneous
toluene solution. However, the ammonium form of the zeolite is preferred over the protonated form when the Si/Al ratio of the zeolite allows for
the presence of at least one ammonium ion per unit cell (Si/Al ꢀ 40), because under these conditions far less benzaldehyde and other undesired
+
reaction products are formed in comparison to the protonated form of the zeolite at the same Si/Al ratios. The ability of NH -ZSM-5 derivatives to
4
limit the formation of unwanted toluene oxidation products is not well understood, but the observed selectivity may be related to the replacement
of ammonium ions by protons, the formation of ammonium nitrate, and the buffering of nitric acid within the micropores of the zeolite.
©
2007 Elsevier Inc. All rights reserved.
Keywords: Toluene; Nitration; ZSM-5 zeolite; Regioselectivity
1
. Introduction
In efforts to overcome these limitations, various solid acids
have been investigated as possible regioselective catalysts for
the mononitration of toluene [3–9]. While Nafion-H and other
polysulfonic acid resin catalysts reduce the corrosive nature of
nitration reaction mixtures, they do not substantially improve
the regioselectivity for the para isomer [10]. However, acetic
anhydride-CCl4 suspensions of Cu(NO3)2 supported on K-10
montmorillonite clay (Claycop), selectively catalyze the HNO3
conversion of toluene to MNT with an o:m:p isomer distri-
bution of 23:1:76 [11,12]. Analogous improvements in MNT
regioselectivity have been achieved using amorphous alumi-
nosilicates and crystalline zeolite catalysts, including H-Y and
H-ZSM-11 zeolites [13,14] and, especially, H-ZSM-5 zeolite
with a very high Si/Al ratio of 1000 [7,15,16], particularly
when n-propylnitrate is the nitrating agent. The regioselectiv-
Mononitrotoluene (MNT), particularly the para-substituted
isomer, is a highly desired intermediate for the production of
many important aromatic compounds [1,2]. Mixtures of nitric
acid and sulfuric acid typically yield of the ortho isomer in ex-
cess of the para isomer. Although the meta MNT isomer is
produced in minor yield, it is undesirable, particularly if the
MNT is to be used as a precursor to further substitution re-
actions. Still further, the nitration and further oxidation of the
methyl group leads to additional unwanted by-products.
*
Corresponding author.
E-mail address: pinnavai@cem.msu.edu (T.J. Pinnavaia).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.11.001

290
S.-S. Kim et al. / Journal of Catalysis 253 (2008) 289–294
+
dealumination reagent, 30 g of as-received NH -exchanged ze-
4
◦
olite was added to 450 ml of 44 mM organic acid at 100 C. For
dealumination with EDTA, 20 g of zeolite was allowed to react
with 300 ml of 14 mM acid. After a reaction time of 1 h, the
hot suspension was filtered, washed with hot water three times
◦
and oven-dried at 160 C overnight.
2
.3. Catalyst characterization
+
Scheme 1. MNT regioselectivity achievable with NH -ZSM-5 as catalyst.
4
The structural integrity of ZSM-5 catalysts was monitored
ity provided by H-ZSM-5 zeolite is retained using NO2/O3 or
NO2/O2 as the nitrating agent [18,19]. Improved para isomer
selectivity also has been achieved with zeolite Beta as the cata-
lyst [5,17], particularly when nitric acid is used in combination
with acetic anhydride [20]. Independent of the MNT isomer dis-
tribution achieved in each of these nitration systems, unwanted
oxidation by-products such as benzaldehyde invariably reduce
overall yields, especially when nitric acid is the nitrating agent.
We report here a new approach to improving the regiose-
lectivity and overall yields of MNT obtained in the ZSM-5-
catalyzed nitration of toluene with nitric acid as the nitrating
agent. The new process is based in the first instance on the
sequestration of the nitric acid in the micropores of the cata-
lyst prior to the addition of the substrate. This strategy confines
the nitration reaction to the intracrystal pores of the catalyst
and minimizes the nonselective extra-zeolitic transformation of
the substrate. Furthermore, we show for the first time that the
ammonium-exchanged form of the zeolite can be even more
effective than the protonated zeolite in providing improved re-
gioselective conversions and in minimizing unwanted benzalde-
hyde and related alkyl group oxidation products. Under optimal
catalytic conditions the MNT regioselectivity show in Scheme 1
can be achieved at nitric acid conversion of 95%, while main-
taining the fraction of undesired oxidation products to <5%.
by X-ray diffraction using CuKα radiation (λ = 1.542 Å) and a
Rigaku Rotaflex diffractometer equipped with a rotating anode
operated at 45 kV and 100 mA. N2 adsorption and desorption
◦
isotherms were obtained at −196 C on a Micromeritics Tristar
sorptometer in static adsorption mode. Samples were outgassed
−
4
◦
under vacuum (1.3 × 10 Pa) at 200 C for a minimum of 12 h
prior to analysis. BET surface areas were calculated from the
linear portion of the BET plot according to IUPAC recommen-
dations.
2
.4. Catalytic reactions
The catalytic nitration of toluene under conventional batch
reaction conditions is representative of the general methodol-
ogy used in previous studies of the nitration of toluene over
ZSM-5 catalysts [16]. In a typical reaction a 5.0-g quantity of
the desired catalyst was added to a round-bottom flask at room
temperature, and 10 ml of toluene was slowly added into the
◦
flask. The mixture was blended with a spatula and kept at 90 C
for 30 min. Then, 0.34 g of 90% nitric acid was added drop-
wise to the heated mixture. The catalytic reaction was run for
4
h while maintaining the reaction mixture at a temperature be-
◦
tween 90 and 100 C. The reaction mixture was cooled to room
temperature and acetone was added to the reaction mixture to
dissolve the reaction products. The acetone solution was filtered
off and the product distribution in the filtrate was determined by
gas chromatography on a HP 5890 instrument equipped with a
FID detector. A 15-meter SPB-1 capillary column was used for
the analysis of reactant and products. X-ray powder diffraction
patterns for the spent catalysts indicated the retention of crys-
tallinity under reaction conditions.
2
. Experimental
2
.1. Materials
Toluene and nitric acid (90 wt%), were purchased from
Aldrich Chemical Co. Benzaldehyde and o-, m-, and p-nitro-
toluene also were obtained from Aldrich and used as references
+
for GC calibration. Four NH -exchanged forms of ZSM-5 ze-
Toluene nitration reactions with remarkably improved re-
gioselectivity were carried out in the presence of zeolite-
sequestered nitric acid. In this procedure 0.34 g of 90% nitric
acid solution was added into a round flask at room temperature
containing 5.0 g of the desired ZSM-5 catalyst. The mixture
was blended with a spatula and equilibrated at room temper-
ature overnight to allow sequestration of the nitric acid in the
micropores of the catalyst. A 10-ml quantity of toluene was
slowly added into the flask and the blended mixture was heated
4
olite having Si/Al ratios of 12, 25, 40, and 140 were purchased
from Zeolyst International.
+
+
2
.2. NH -ZSM-5 and H -ZSM-5 catalysts
4
+
As-received NH -exchanged forms of ZSM-5 zeolite were
4
+
denoted NH -ZSM-5 (x), where the value of x indicates the
4
Si/Al ratio of the framework. The ammonium forms of the zeo-
lite were converted to the protonated derivatives by calcination
◦
in an oil bath to a reaction temperature between 90 and 100 C
◦
+
at 500 C for 5 h and were denoted H -ZSM-5 (x).
for four hours. The cooled reaction mixture was extracted with
acetone, and the product distribution was determined by gas
chromatography. The fraction of toluene converted to products
was determined by GC, and the overall conversion was based
the corresponding amount of nitric consumed as the limiting
reagent.
+
The dealumination of the external surfaces of NH -ZSM-5
4
catalysts with differing Si/Al ratios was accomplished by
reaction with oxalic acid, citric acid, and ethylenediamine
tetraacetic acid (H4EDTA) as complexing agents for aluminum
[
21–23]. In a typical reaction with oxalic and citric acids as the

S.-S. Kim et al. / Journal of Catalysis 253 (2008) 289–294
291
Fig. 2. Nitrogen adsorption–desorption isotherms for ZSM-5 (40) catalysts (a)
+
+
+
as-received NH -exchanged zeolite, (b) calcined H -zeolite, and (c) NH -ex-
4
4
changed zeolite dealuminated by reaction with oxalic acid. The isotherms are
offset by 50 cc/g for clarity.
mination process removes amorphous debris from the external
surfaces of the crystals or, alternatively, improves the packing
of the zeolite crystals and increases in the bulk density of the
compacted powder used to obtain the XRD patterns.
Representative nitrogen adsorption–desorption isotherms
+
are provided in Fig. 2 for ZSM-5 (40) in NH -exchanged and
4
Fig. 1. Powder XRD patterns of ZSM-5 (x) catalysts with initial Si/Al values
+
+
calcined H forms, as well as for the dealuminated NH form
obtained through reaction with oxalic acid. BET surface areas
and pore volumes for as-received NH -ZSM-5 (x) composi-
(
x) equal to (A) 12, (B) 25, (C) 40, and (D) 140. The patterns labeled (a) are for
4
+
as-received NH -exchanged zeolites and those labeled (b) were dealuminated
4
+
(b) reaction with oxalic acid.
4
tions and the corresponding compositions after dealumination
of the external surfaces are compared in Table 1. The surface
areas for the as-received catalysts compare favorably with the
3
. Results
2
values of 400–425 m /g provided by the supplier [24]. Similar
3
.1. ZSM-5 catalysts
+
surface areas values were obtained for H -ZSM-5 (x) com-
+
+
positions obtained through calcination of the NH -exchanged
The as-received NH -ZSM-5 (x) compositions, where x
4
4
precursors. The dealumination of external surfaces by reaction
with carboxylic acids hardly affects the BET surface area and
pore volume of the zeolite, as indicated by the comparison of
textural properties in Table 1.
represents the Si/Al ratio of the framework, were modified in
two ways. The first modification was achieved through calcina-
◦
tion at 500 C to obtain the zeolite in protonated form, denoted
+
H -ZSM-5 (x). The second modification was accomplished
by reaction with H4EDTA, oxalic acid and citric acid as com-
plexants to selectively remove acidic aluminum sites from the
external surfaces of the zeolite crystals without removing alu-
minum from the internal micropores [21–23]. The objective of
the dealumination reactions was to minimize the less regiose-
lective nitration reactions that may occur at the external surfaces
of the zeolite crystals.
3.2. Regioselective nitration reactions
The conventional method used for the ZSM-5-catalyzed
mononitration of toluene combines the proton form of the zeo-
lite, toluene, and 90% nitric acid in a batch reactor and equili-
brating the mixture at elevated temperatures [16]. Using similar
+
+
The modification of NH -ZSM-5 (x) compositions through
batch reactions conditions and H -ZSM-5 (x) with different
4
calcination to form the protonated derivative or through reac-
tion with carboxylic acids to form surface-dealuminated deriv-
atives did not compromise the crystallinity or the textural prop-
erties of the zeolite framework, as judged by powder X-ray
diffraction and nitrogen adsorption studies. Fig. 1 provides the
Si/Al ratios (x), we obtained the conversions and o-, m-, and
p-MNT product distributions provided in Table 2. Included in
the table is the fraction of non-MNT by-products (denoted bp)
formed in the reaction, of which, benzaldehyde is the major
component. We also include in Table 2 for comparison pur-
poses the conversions and product distributions obtained using
the ammonium form of the zeolite in place of the protonated
analog. We note, rather surprisingly, that the as-received am-
monium forms of the catalysts provide conversions and prod-
uct distributions competitive with the protonated forms. Under
equivalent reaction conditions in the absence of a ZSM-5 cat-
alyst, a nitric acid conversion of 50% is achieved, but nearly
+
X-ray powder diffraction patterns for the as-received NH -
4
ZSM-5 (x) compositions with x = 12, 25, 40, 140, along with
the dealuminated analogs obtained by reaction with oxalic acid.
Equivalent powder patterns were obtained for dealuminated
derivatives obtained using H4EDTA and citric acid as dealumi-
nation reagents. The apparent increases in diffraction intensity
for the dealuminated compositions may indicate that the dealu-

292
S.-S. Kim et al. / Journal of Catalysis 253 (2008) 289–294
Table 1
Comparison of surface areas and pore volumes for as-received and dealumi-
Table 3
Product distributions for the nitration of toluene with nitric acid sequestered in
ZSM-5 catalysts
+
a
nated NH -ZSM-5 (x) catalysts
4
Catalyst
Dealumination Surface area Micropore Total pore
Zeolite catalyst
Heat
treatment
Conversion Product mole ratios p/o
2
b
c
agent
(BET, m /g) volume
cc/g)
volume
(cc/g)
(%)
bp:o:m:p
ratio
(
+
◦
H -ZSM-5 (12)
500 C
45
63
90
76
16:11:3:70
11:14:tr:75
13:14:1:61
19:8:2:71
6.4
5.4
4.4
8.9
+
+
◦
NH -ZSM-5 (11.5) None
4
386
412
429
401
0.18
0.19
0.20
0.19
0.20
0.23
0.25
0.21
H -ZSM-5 (25)
500 C
+
◦
+
H -ZSM-5 (40)
500 C
NH -ZSM-5 (25)
4
None
None
+
◦
H -ZSM-5 (140)
500 C
+
NH -ZSM-5 (40)
4
+
a
+
NH -ZSM-5 (12)
4
As-received 25
As-received 65
As-received 69
tr :12:2:86
7.2
3.3
1.8
8.1
NH -ZSM-5 (140) None
4
+
NH -ZSM-5 (25)
4
4:22:1:72
6:33:2:59
17:9:1:73
+
NH -ZSM-5 (11.5) Oxalic acid
4
369
427
433
408
0.17
0.20
0.20
0.19
0.18
0.24
0.25
0.22
+
NH -ZSM-5 (40)
4
+
NH -ZSM-5 (25)
4
Oxalic acid
Oxalic acid
+
NH -ZSM-5 (140) As-received 62
+
4
NH -ZSM-5 (40)
4
+
◦
+
NH -ZSM-5 (12)
4
160 C
58
53
67
56
1:12:2:85
tr:16:tr:84
3:32:tr:65
21:9:tr:69
7.1
5.3
2.0
7.7
NH -ZSM-5 (140) Oxalic acid
4
+
◦
a
◦
NH -ZSM-5 (25)
4
160 C
Samples were out-gassed at 200 C for 20 h prior to the determination of
+
◦
textural properties by nitrogen adsorption.
b
NH -ZSM-5 (40)
4
160 C
+
◦
The micropore volumes were obtained from the nitrogen uptake at P/P =
NH -ZSM-5 (140) 160 C
0
4
0
.40.
a
c
tr indicates trace quantities (<0.5%).
The total pore volumes were obtained from the nitrogen uptake at P/P =
0
0
.98.
Table 4
Product distributions for the nitration of toluene by reaction with nitric acid
sequestered in dealuminated ZSM-5 catalysts
Table 2
Product distributions for the ZSM-5-catalyzed nitration of toluene under con-
ventional batch reaction conditions
Dealuminated
catalyst
+
Dealumination Conversion Product mole ratios p/o
reagent
(%)
bp:o:m:p
ratio
a
Zeolite catalyst
Conversion
%)
Product mole ratios
bp :o:m:p
p/o
b
NH -ZSM-5 (12) H EDTA
54
79
83
65
2:15:2:80
3:24:1:72
5:33:3:59
15:13:1:70
5.3
3.0
1.8
5.4
(
ratio
4
4
+
+
c
NH -ZSM-5 (25) H4EDTA
H -ZSM-5 (12)
43
58
71
60
12:35:2:51
11:32:2:55
7:36:3:54
21:23:2:54
1.5
1.7
1.5
2.3
4
+
NH -ZSM-5 (40) H EDTA
4
4
+
H -ZSM-5 (25)
+
+
H -ZSM-5 (40)
NH -ZSM-5 (140) H EDTA
4
4
+
H -ZSM-5 (140)
+
NH -ZSM-5 (12) Oxalic acid
4
71
90
95
80
1:17:2:80
5:10:1:70
7:35:1:57
18:11:1:71
4.7
4.2
1.6
6.5
+
NH -ZSM-5 (12)
4
33
52
57
55
3:41:4:52
2:34:3:62
8:39:3:51
14:27:1:57
1.3
1.8
1.3
2.1
+
NH ZSM-5 (25)
4
Oxalic acid
+
NH -ZSM-5 (25)
4
+
NH -ZSM-5 (40) Oxalic acid
4
+
NH -ZSM-5 (40)
4
+
NH -ZSM-5 (140) Oxalic acid
4
+
NH -ZSM-5 (140)
4
+
+
NH -XSM-5 (12) Citric acid
4
67
95
95
89
tr:12:2:86
4:18:1:77
7:35:2:56
18:9:2:72
7.2
4.3
1.6
8.0
a
The conversion was based on nitric acid as the limiting reagent.
bp represents by-products other than mononitrotoluene (mainly benzalde-
NH -ZSM-5 (25) Citric acid
4
b
+
NH -ZSM-5 (40) Citric acid
4
hyde); o, m, and p represent the ortho-, meta-, and para-isomers of MNT.
c
+
The number in parenthesis indicates the Si/Al ratio.
NH -ZSM-5 (140) Citric acid
4
one-third of the products are not MNT and include benzalde-
hyde, benzoic acid, dinitrotoluenes, among others.
version of nitric acid with retention of selectivity toward the
para isomer.
As is shown by the data in Table 3, the regioselective ni-
tration of the toluene ring is substantially altered by first se-
questering the nitric acid in the zeolite prior to the addition of
toluene. The improved regioselectivity toward the para isomer
is realized for both the protonated and ammonium form of the
zeolite. Also, the effects of nitric acid sequestration are evident
over the entire range of Si/Al values from x = 12 to 140.
In an effort to evaluate the potential effects of catalytic nitra-
tion at the external surfaces of the zeolite, we sequestered the
nitric acid in zeolite derivatives in which the external surfaces
4. Discussion
+
The reaction of toluene-nitric acid mixtures over H -ZSM-5
and the related pentasil zeolite H -Beta under liquid and vapor
+
phase reaction conditions, respectively, has been shown pre-
viously to improve the regioselective conversion to MNT in
comparison to the conversions achieved under homogeneous re-
action conditions [16,17]. The results of the present work (cf.,
Table 2) verify the improvement in regioselectivity in the pres-
+
+
of the NH -ZSM-5 (x) catalysts were dealuminated by reaction
ence H -ZSM-5 zeolites. Depending on the Si/Al ratio of the
4
with oxalic acid, citric acid, and ethylenediamine tetraacetic
acid (H4EDTA). The results are presented in Table 4. Compar-
ing the results for the dealuminated catalysts in Table 4 with
zeolite, nitric acid conversions of 40–60% are achieved, 7–21%
of the products are unwanted oxidation derivatives, and the p/o
ratio of MNT isomers is in the range 1.5–2.5. A conversion of
50% is observed under equivalent homogeneous reaction con-
ditions, but almost one-third of the products benzaldehyde, ben-
+
those for NH -ZSM-5 (x) compositions in Table 3 that were
4
not dealuminated, we see that dealumination increases the con-

S.-S. Kim et al. / Journal of Catalysis 253 (2008) 289–294
293
zoic acid, dinitrotoluenes, and other non-MNT derivatives and
the p/o MNT isomer ratio is well below 1.5. Clearly, a signif-
icant fraction of the nitration reaction occurs within the shape-
the zeolites, thus ensuring that nitration is confined almost ex-
clusively to the shape-selective environment of the catalyst.
The present work demonstrates that sequestration of nitric
acid in ZSM-5 prior to the addition of toluene also is effective in
promoting MNT formation within the intracrystal pores of the
zeolites. The dramatic improvement in product selectivity that
occurs when toluene is added to zeolite-intercalated nitric acid
indicates that toluene addition helps to keep the nitric acid in the
sequestered state and prevents the reaction system from rapidly
re-equilibrating to a mixture of zeolite and a solution of nitric
acid in toluene. That is, the binding of toluene limits diffusion
of nitric acid out of the micropores and confines the reaction
largely to the shape-selective pores of the zeolite framework.
Returning to the product distributions in Table 3, we see that
+
selective micropores of the H -ZSM-5 zeolite.
The selection of the protonated form of the zeolites in earlier
studies presumably was motivated by the expectation that the
strong Brönsted acidity in the micropores promotes formation
of electrophilic nitronium ions through the equilibrium shown
in Eq. (1). However, as shown by the data in Table 2 for reaction
under conventional batch reaction conditions,
+
+
H + HNO3 ꢁ H2O + NO2
(1)
the HNO3 conversions and p/o-MNT isomer ratios obtained
using weakly acidic ammonium ion exchanged forms of ZSM-5
catalysts are comparable to those observed for the analogous
protonated forms of the zeolites. Thus, the self-ionization of ni-
+
for the H -ZSM-5 catalysts the pre-intercalation of nitric acid
in the zeolite framework does not substantially reduce the for-
mation of undesired benzaldehyde in comparison to the fraction
of such products generated under conventional batch reaction
conditions (cf., Table 2). Nevertheless, nitric acid sequestration
+
tric acid in the micropores of NH -ZSM-5 according to Eq. (2)
4
is sufficient to provide conversions comparable to those ob-
+
served for H -ZSM-5.
+
in H -ZSM-5 (x) compositions greatly increases the regiose-
+
−
2
HNO3 ꢁ H2O + NO + NO .
(2)
2
3
lectivity for MNT formation, as indicated by the substantially
improved p/o MNT ratios. Although pre-intercalation of ni-
tric acid confines much of the nitrating agent to the micropores
of the protonated zeolite and promotes intra-crystal formation
of MNT, the protonated zeolite also permits the formation of
the undesired oxidation products within the framework pores.
+
One especially notable property of the H -ZSM-5 (x) and
NH -ZSM-5 (x) zeolite families is the substantially lower frac-
+
4
tion of undesired benzaldehyde and related oxidation products
formed at Si/Al ratios below x = 140. For all framework Si/Al
ratios investigated, the o/p MNT isomer ratio is in the range
+
1
.3–2.3 and less than 5% of the product is in the form of
However, the NH -exchanged forms of the zeolite at Si/Al ra-
4
m-MNT (cf., Table 2). Nevertheless, under conventional batch
reaction conditions, homogeneous reaction in toluene solution
plays an important role in limiting the product selectivity, as
revealed below.
tios ꢀ40 not only promote intracrystal formation of the para
MNT isomer, but also greatly reduce the fraction of benzalde-
hyde and other undesired oxidation products formed in the re-
+
+
action. Unlike the H -ZSM-5 (x) compositions, NH -ZSM-5
4
+
We further investigated the nitration of toluene over H - and
(x) catalysts with x ꢀ 40 are effective in reducing benzalde-
hyde formation both under conventional batch reaction condi-
tions (cf., Table 2) and under conditions where nitric acid is
pre-intercalated in the zeolite framework (cf., Table 3). Pre-
+
NH -ZSM-5 (x) catalysts by first intercalating the nitric acid
4
into the framework micropores of the zeolites prior to the ad-
dition of toluene. Our rationale for doing so was based on the
expectation that the sequestration of the nitrating agent within
the zeolite micropores would minimize the less-selective reac-
tions occurring in homogeneous solution and thereby enhance
the regioselective conversion of toluene to the para-MNT iso-
mer. As shown by the product distributions in Table 3, the p/o
MNT isomer ratios indeed are generally increased to values in
◦
drying the zeolite at 160 C further reduces the fraction of ben-
zaldehyde formed under intercalated reaction conditions (cf.,
Table 3).
+
The role of NH4 exchange cations in reducing benzalde-
hyde formation is not clearly understood. It may be that the
ammonium ion mediates the acid–base equilibria in the zeolite
micropores. The need for the presence of an ammonium ion in
+
+
the range 4.4–8.9 and 1.8–7.7 for the H - and NH -exchanged
4
+
each unit cell is indicated in part by the ineffective role of NH -
forms of the zeolite, respectively. To our knowledge these are
the highest p/o ratios yet observed for the mononitration of
toluene using nitric acid as the nitrating agent in the absence of
a nitronium ion promoter.
4
ZSM-5 (x) with a Si/Al ratio of x = 140 (cf., Tables 2 and 3). At
this catalyst composition, half the unit cells in the zeolite crystal
lack ammonium cations. For Si/Al ratios ꢀ40, however, there is
on-average more than one ammonium ion per T96O192 unit cell
Para/ortho MNT isomer ratios up to 19 at MNT yields of
(
T = Si or Al). It is likely that some of the ammonium ions in
1
5–50% have been reported for the nitration of toluene over
+
+
NH -ZSM-5 (x) catalysts are replaced by protons, or even ni-
H -ZSM-5 using n-propylnitrate as the nitrating agent [7].
Also, quantitative conversions of toluene to MNT with p/o
ratios in the range 1.0–5.4 have been achieved using the pro-
tonated form of zeolite beta as the catalyst and acetylnitrate
formed in situ from nitric acid and acetic anhydride [5]. Al-
though far more costly than nitric acid alone, n-propylnitrate
and acetylnitrate offer certain advantages as nitrating agents.
For these latter nitrating agents, reactive nitronium ions form
only when the reagents encounter protons in the micropores of
4
tronium ions, leading to the formation of ammonium nitrate in
the micropores and a reduction in the fraction of free nitric acid
+
in the pores in comparison to H -ZSM-5 (x) catalysts. Indeed,
+
the ratio of HNO3/NH4 is >1.0 under all reaction conditions
+
except for NH -ZSM-5 (x) with a Si/Al ratio of x = 12, where
4
+
the HNO /NH ratio is ∼0.80. But the product selectivity of
3
4
the ammonium exchange form is always superior to the selec-
tivity observed for the fully protonated zeolite, so long as the

294
S.-S. Kim et al. / Journal of Catalysis 253 (2008) 289–294
nitric acid is sequester in the micropores of a ZSM-5 frame-
work with a Si/Al ratio ꢀ40. This observation suggests that
the ammonium ions may buffer the intercalated nitric acid by
forming ammonium nitrate in the micropores and protons on
the exchange sites of the zeolite framework.
mononitration reaction to the shape-selective micropores of the
zeolite and thereby substantially improves the regioselectivity
toward the para isomer of mononitrotoluene (MNT). Although
the MNT regioselectivity is similar for the two exchanged forms
+
+
of the zeolite, the NH4 form is preferred over the H form
because it dramatically reduces the amount of toluene that is ox-
idized to benzaldehyde and other unwanted oxidation products.
At Si/Al ratios corresponding to the presence of more than one
ammonium ion per unit cell (i.e., at Si/Al ratio ꢀ40), the para
position of toluene is preferentially nitrated by the sequestered
nitric acid with little or no oxidation of the methyl group. How-
ever, at Si/Al ratios corresponding to less than one ammonium
ion per unit cell (e.g., Si/Al ratio = 140), the methyl posi-
tion of toluene undergoes oxidation, leading to the formation
of substantial fraction of benzaldehyde as an undesired reaction
product.
As indicated by the data in Table 4, the dealuminaton of the
+
external surfaces of NH -ZSM-5 (x) zeolites by H4EDTA, cit-
4
ric acid or oxalic acid did not appreciably affect the product
distributions in comparison to the initial catalysts. This result
suggests that potentially less-selective reactions on the exter-
nal surfaces of the zeolite play little or no role in the observed
product distributions. However, the dealumination typically im-
proves the reactivity substantially, as indicated by the increases
in conversion (cf., Tables 3 and 4). Also, the increases in cat-
alytic activity parallel the increases in the intensities of the
Bragg reflections in the X-ray powder diffraction pattern of the
catalysts after dealumination. Thus, the role of the weak acids
may indeed be to remove amorphous debris from the external
surfaces of the zeolite crystals.
References
We note that the improved MNT regioselectivity and re-
duced yields of undesired oxidation products observed in this
work were achieved under conditions where the volume of 90%
nitric acid used for nitration represented approximately 30% of
the framework pore volume of the zeolite catalyst (∼0.19 cc/g,
cf., Table 1). Increasing the amount of nitric acid beyond the
pore volume of the zeolite compromised both the regioselec-
tivity for MNT, as well as the yield of unwanted oxidation
[1] K. Schofield, Aromatic Nitration, Cambridge Univ. Press, Cambridge,
1980.
[2] L. Delaude, P. Laszlo, K. Smith, Acc. Chem. Rec. 26 (1993) 607.
[3] L.V. Malysheva, E.A. Paukshtis, K.G. Ione, Catal. Rev. Sci. Eng. 37 (1995)
1
79.
4] J.M. Smith, L. Liu, D.E. Resasco, in: C.H. Bartholomew, G.A. Fuentes
Eds.), Catalyst Deactivation, Elsevier, Amsterdam, 1997, p. 199.
[5] K. Smith, A. Musson, G.A. DeBoos, J. Org. Chem. 63 (1998) 8448.
[
(
[
[
[
6] B.M. Choudary, M. Sateesh, M.L. Kantam, K.K. Rao, K.V.R. Prasad,
K.V. Raghavan, J.A.R.P. Sarma, Chem. Commun. (2000) 25.
7] T.J. Kwok, K. Jayasuriya, R. Damavarapu, B.W. Brodman, J. Org. Chem.
+
products. For instance, for the citric acid dealuminated NH -
4
ZSM-5 (12) catalyst in Table 4, the conversion is 67%, less than
59 (1994) 4939.
0
.5% of the product is benzaldehyde, and the p/o isomer ratio
8] J.A.R. Rodrigues, A.P. Oliveira, P.J.S. Moran, Synth. Commun. 29 (1999)
2169.
of MNT is 7.2 under conditions where the volume of nitric acid
used in the nitration is 30% of the micropore volume. Increas-
ing the volume of nitric acid to 120% of the micropore volume
of the same catalyst results in approximately the same conver-
sion (64%) but 6% of the product is benzaldehyde and the p/o
isomer ratio is decreased to 1.8.
[9] O.L. Wright, J. Teipel, D. Thoennes, J. Org. Chem. 30 (1965) 1301.
10] G.A. Olah, R. Malhotra, S.C. Narang, J. Org. Chem. 43 (1978) 4628.
11] P. Laszlo, J. Vandormael, Chem. Lett. (1988) 1843.
[
[
[
12] A. Cornelis, L. Delaude, A. Gerstmans, P. Laszlo, Tetrahedron Lett. 29
(1988) 5657.
[
13] S.M. Nagy, K.A. Yarovoy, M.M. Shakirov, V.G. Shubin, L.A. Vostrikova,
K.G. Ione, J. Mol. Catal. 64 (1991) L31.
14] S.M. Nagi, E.A. Zubkov, V.G. Shubin, Bull. Acad. Sci. USSR Chem. Ser.
The limited volume of nitric acid that can be sequestered in
+
[
NH -ZSM-5 micropores restricts the amount of nitric acid and
4
(1989) 1780.
toluene that can be processed per unit weight of catalyst. For
this reason it may not be practical to use batch reaction condi-
tions for the large scale production of MNT over this zeolite.
As in the case of most catalytic processes, creative process de-
[
[
[
15] R. Damavarapu, K. Jayasuriya, T.J. Kwok, U.S. patent 5977418, 1999.
16] K. Jayasuriya, R. Damavarapu, U.S. patent 5946638, 1999.
17] S.P. Dagade, S.B. Waghmode, V.S. Kadam, M.K. Dongare, Appl. Catal. A
Gen. 226 (2002) 49.
[18] K. Smith, T. Gibbins, R.W. Millar, R.P. Claridge, J. Chem. Soc. Perkin
Trans. 1 (2000) 2753.
19] X.H. Peng, H. Suzuki, C.X. Lu, Tetrahedron Lett. 42 (2001) 4357.
20] S. Bernasconi, G.D. Pirngruber, R. Prins, J. Catal. 224 (2004) 297.
21] M. Muller, G. Harvey, R. Prins, Micropor. Mesopor. Mater. 34 (2000) 135.
22] G.T. Kerr, J. Phys. Chem. 72 (1968) 2594.
+
sign will be needed to realize the full potential of NH -ZSM-5
4
catalysts for the practical mononitration of toluene.
[
[
[
[
[
5
. Conclusions
The pre-intercalation of nitric acid in the micropores of
23] R. Szostak, Stud. Surf. Sci. Catal. 58 (1991) 153.
[24] http://www.zeolyst.com/html/zsm5.asp.
+
+
NH and H exchanged forms of ZSM-5 zeolite confines the
4