Montmorillonite Clay Catalysis: Conversion of Methyl Benzoate and NH3 into Benzonitrile and Amides

Article abstract of DOI:10.1006/jcat.1997.1896

The reaction of methyl benzoate [1] with NH₃ was studied over Montmorillonite K10 clay in the temperature range 280-340°C. Benzonitrile [2], benzoic acid [3], benzamide [4], N-methylbenzamide [5], and methanol were formed as products in varying amounts. This reaction was investigated as a function of molar ratio of 1 to NH₃, temperature, and contact time. At 1/NH₃ = 0.5, 320°C, and contact time of 1.96 × 10^-3 h the conversion of 1 was 64.7 % and the selectivities of 2 and 4 + 5 were 59.6 and 20.2, respectively. The clay catalyst was equally active for other aromatic and aliphatic esters. At 340°C and 67.2% conversion of ethyl propionate the selectivity to propionitrile was 73.3 after 4 h reaction. The actual pathways leading to formation of products was studied in detail. ESCA, NH₃ adsorption, pyridine adsorption, and IR studies, in combination with the catalytic investigation show that NH₃ adsorbed on the acidic >Al-OH sites forms ammonium ions [NH⁺₄] which participate in the reaction. The Lewis acid sites are likely to be involved in the formation of 2. The clay catalyst undergoes deactivation on continuous use for several hours [23 h]. The deactivation is mainly due to dealumination by water produced in the reaction and accumulation of organic residue containing an amide group on the surface of the clay. It was also possible to calculate effective active sites on the catalyst surface and then a correlation with the turnover of formation of benzonitrile could be established.

Full text of DOI:10.1006/jcat.1997.1896

JOURNAL OF CATALYSIS 173, 84–94 (1998)
ARTICLE NO. CA971896
Montmorillonite Clay Catalysis: Conversion of Methyl Benzoate
1
and NH into Benzonitrile and Amides
3
2
A. Wali, S. Unnikrishnan, S. Muthukumaru Pillai, V. K. Kaushik, and S. Satish
Research Centre, Indian Petrochemicals Corporation Limited, Baroda 391 346, India
Received September 23, 1996; revised June 21, 1997; accepted July 31, 1997
drides, and catalysts like AlCl3 and ZnCl2 (9). Preparation
The reaction of methyl benzoate [1] with NH3 was stud- of benzonitrile by the catalytic reaction of toluene with NO
ied over Montmorillonite K10 clay in the temperature range has been reported (10). A method for making benzonitrile
�
2
80–340 C. Benzonitrile [2], benzoic acid [3], benzamide [4],
by dehydrogenation of the Diels–Alder adduct of buta-
diene and acrylonitrile is also available (11). Reaction of
ethanol and NH3 to give acetonitrile over SAPO (40 type)
and V–Y zeolites is recently reported (12). It has also been
shown that nitriles can be prepared from carboxylic acids
and NH3 over silica gel (13), ThO2, Al2O3, and zeolite cata-
lysts (14). Application of this reaction is industrially sig-
nificant only when carboxylic acids are cheaper and more
N-methylbenzamide [5], and methanol were formed as products
in varying amounts. This reaction was investigated as a function
of molar ratio of 1 to NH3, temperature, and contact time. At
�
� 3
1
/NH3 = 0.5, 320 C, and contact time of 1.96 � 10 h the conver-
sion of 1 was 64.7% and the selectivities of 2 and 4 + 5 were 59.6 and
2
0.2, respectively. The clay catalyst was equally active for other aro-
�
matic and aliphatic esters. At 340 C and 67.2% conversion of ethyl
propionate the selectivity to propionitrile was 73.3 after 4 h reaction.
The actual pathways leading to formation of products was studied accessible than their nitrile (3).
in detail. ESCA, NH3 adsorption, pyridine adsorption, and IR stud-
In the petrochemicals industry conversion of C8 cut to
ies, in combination with the catalytic investigation show that NH3 dimethyl terephthalate (DMT) produces a sizeable quan-
+
adsorbed on the acidic >Al–OH sites forms ammonium ions [NH ]
4
tity of methyl benzoate as by-product. Conversion of this
ester into benzonitrile with NH3 is an attractive proposition.
Synthesis of organic compounds employing environmen-
tally benign Montmorillonite clay catalyst is an area of our
current interest (15–17). In this study conversion of methyl
benzoate and other esters to nitriles has been carried out
with NH3 on Montmorillonite clay catalyst. Identification
of acid sites involved in the reaction, recycle of the catalyst,
and a possible deactivation pathway of the clay catalyst
have been studied using ESCA and NH3 adsorption.
which participate in the reaction. The Lewis acid sites are likely
to be involved in the formation of 2. The clay catalyst undergoes
deactivation on continuous use for several hours [23 h]. The deac-
tivation is mainly due to dealumination by water produced in the
reaction and accumulation of organic residue containing an amide
group on the surface of the clay. It was also possible to calculate
effective active sites on the catalyst surface and then a correlation
with the turnover of formation of benzonitrile could be established.
�
c 1998 Academic Press
1
. INTRODUCTION
2. EXPERIMENTAL
2
.1. Materials
Synthesis of nitriles is an important catalytic and indus-
trial process (1–3). They are increasingly being used in se-
lective solvent extraction, as reaction media, as intermedi-
ates in preparation of pharmaceuticals (4) and a variety of
commercial amines and amides, as jet-fuel additives, in cot-
ton bleaching baths, etc. The nitriles are produced in high
yield by vapour-phase catalytic ammoxidation of olefins (5),
alkanes (6), and aromatics (7) using multimetal-containing
oxides.
Montmorillonite K10 clay was purchased from Fluka
AG, Switzerland. It has the weight-percent composi-
tion of SiO2 = 70, Al2O3 = 15, Fe2O3 = 1.5, CaO = 2.5,
MgO = 3, Na2O = 0.5, K2O = 1.5, and H2O = 6. It has bulk
2
density = 300 � 30 g/l and surface area = 200 � 10 m /g.
+
Na -Montmorillonite was prepared from K10 clay by ion
exchange method (18). Nb2O5 was purchased from Loba
Chemicals, Bombay. Methyl benzoate (94% pure) and
methyl p-toluate of IPCL, Baroda were used. Ethyl ben-
zoate was supplied by High purity chemicals, New Delhi.
Ethyl propionate and ethyl phenylacetate were prepared
from the corresponding acids and ethanol by extractive
esterification (19). Ammonia [99.8% purity] supplied in
Dehydration of carboxamides to nitriles is also known
using dehydrating agents (8) like PCl5, SOCl2, acid anhy-
1
2
IPCL Communication No. 327.
Fax: (91)-0265-372098.
84
0021-9517/98 $25.00
Copyright �c 1998 by Academic Press
All rights of reproduction in any form reserved.

MONTMORILLONITE CLAY CATALYSTS
85
cylinder by GSFC, Baroda was used as such. The HZSM-5 2.3. Acidity Measurement
sample was synthesized by hydrothermal crystallization
The acidity of the clay catalyst was determined indepen-
dently by means of NH3 and pyridine adsorption. A known
method and characterized (20). HY zeolite was synthesized
+
from NaY zeolite (IPCL, Baroda) by NH4 exchange fol-
weight (0.1 g) of the sample was placed in a microreactor
�
lowed by calcination at 550 C. The final HY zeolite has a
�
(
Catalytic Data System, Baroda) and heated at 250 C for
frame work Si/Al of 3.
1
h. Known pulses (0.5 ml) of NH3/pyridine (1 �l) were
The powder form of the catalyst material was converted
into extrudates of 1.5 mm diameter by 2–3 mm length (pre-
flown onto the clay sample along with the diluant He gas.
The unadsorbed NH3 or pyridine was estimated by thermal
conductivity detector. Difference between the fed and un-
absorbed NH3 or pyridine gives the acidic sites on which
NH3 or pyridine are adsorbed. Pyridine adsorptions were
�
dominantly 3 mm long) and then calcined at 250 C for 2 h.
The selectivity to amide formation is independent of the
size of the extrudates. On the other hand, the selectivity to
benzonitrile is maximum in the size of the extrudates used
�
carried out in the temperature range 280–340 C.
(
2–3 mm).
2
.4. IR Measurements
2
.2. Reaction Procedure
Infrared spectra were recorded on Nicolet-550 FT-IR
spectrometer equipped with OMNIC Nicolet FTIR soft-
ware.
The reaction was carried out in a tubular, down-flow re-
actor (20 mm ID); 2 g of Montmorillonite K10 was loaded
�
into the reactor. The temperature was set at 320 C. Distilled
1
g of Montmorillonite K10 clay was activated in the tem-
methyl benzoate was fed (3 ml/h; 0.0234 mol/h) from the
top using a Sage Syringe pump. Equimolar quantity of NH3
600ml/h;0.0234mol/h), measured through a gasflowmeter
�
perature range 280–340 C and itsIR spectrum wasrecorded
(
A). This sample was then treated with 1.5 ml of pyridine
(
for 2 h. Then the excess base was removed under vacuum
for 5 h and the pyridine adsorbed IR spectrum (B) was
recorded. The concentrations of Brønsted and Lewis sites
were calculated after substracting spectrum A from B.
and N2 gas (1200 ml/h) were flown on the catalyst bed. The
product mixture coming out of the reactor was condensed
�
by circulating chilled water (5 C). Samples were collected
at 1 h intervals. The products were characterised by GC-MS
1
(
HP 5890) and H NMR (Jeol JNM FX100). The course of
2
.5. ESCA Measurements
the reaction was monitored by GC using 2 m 1.5% OV-17
column in a Shimadzu 15A GC fitted with flame ionization
detector. The conversion and selectivity were determined
in the following way:
A Vacuum Generators ESCALAB MK II spectrometer
with a MgK� radiation source (1253.6 eV) was used for the
recording of XPS spectra. The anode was operated at 10 kV
and 10 mA and the vacuum in the analysis chamber was
�
8
Conversion of methyl benzoate (mol% )
maintained at better than 5 � 10 m bar. The spectrometer
wascalibrated usingCu (2p3/2) and Ag(3d5/2) photoelectron
lines at 932.7 and 368.3 eV, respectively (21, 22). Samples
were placed onto double side adhesive tape and mounted
on sample holders. These samples were kept overnight in
the preparation chamber before they were transferred to
the analysischamber ofthe spectrometer for analysis. The Si
Moles of converted methyl benzoate � 100
=
Moles of methyl benzoate fed
Selectivity to benzonitrile (mol% )
Moles of benzonitrile formed � 100
=
Moles of converted methyl benzoate
(
2p) photoelectron line at 103.0 eV was used as an internal
Converted methyl benzoate
reference (23–25) for the correction of the charging effect.
Data were collected and analysed on Apple IIe computer
interfacing the spectrometer.
=
methyl benzoate fed
unreactedmethyl benzoate collected.
Turnover number (TON) of 2
Moles of 2 formed
Estimated Lewis acid sites on the catalyst
�
3
. RESULTS AND DISCUSSION
=
The reaction of methyl benzoate [1] with NH3 over Mont-
morillonite K10 gave benzonitrile [2], benzoic acid [3],
benzamide [4], N-methylbenzamide [5], and methanol as
products
Turnover number (TON) of (4 + 5) =
Moles of (4 + 5) formed
Estimated total Brønsted & Lewis acid sites on the catalyst
Turnover frequency (TOF) = TON/Time (s).
C, H, and N analyses were carried out using the Elemental
analyser (Carlo Erba, EA-1106).

8
6
WALI ET AL.
in varying amounts. The reaction was run on the extruded
form of clay catalyst as the flow of reactants was smoother
through such a catalyst bed than through the powder form
of catalyst. Further, it is observed that Si/Al of 4.27 in the
powder clay surface is changed to 3.15 in the extrudates.
This has a beneficial effect on the reaction.
3
.1. Effect of Reaction Time
On fresh Montmorillonite K10 catalyst the conversion of
methyl benzoate is high and the activity slowly decreases
with the increase in time on stream. Figure 1 shows the
typical selectivity and turnover frequencies (TOF) for ben-
zonitrile [2] and the amides [4, 5].
3
.2. Effect of Reaction Temperature
Figure 2 shows the results of the reaction on Montmo-
�
rillonite K10 in the temperature range 250–350 C in 6 h
run. It was observed that the reaction was unsuccessful at
�
temperatures below 250 C. The level of methyl benzoate
conversion increased with increasing reaction temperature.
Selectivity to benzonitrile slowly increases with temper-
ature and after 320 C it remains constant. Formation of
FIG. 2. Effect of reaction temperature: 1 :NH3 = 1 : 1 (molar); N2 =
0 ml; 2 g clay. (᭪) Conversion of 1; (᭝) selectivity to 2 (᭡); (Ճ) selectivity
to 4 + 5 (᭿).
2
�
amides [4, 5] remains steady throughout the temperature
�
range 280 to 340 C. TOFs of both 2 and 4 + 5 increase with
�
.4 kcal/mol, respectively. At 320 C and 63% conversion of
9
the increase in temperature. However, nitrile formation is
higher at 340 C relative to the amides. From the Arrhenius
plot the apparent energy of activation (E ) of benzonitrile
and the amide formation were calculated to be 11.3 and
�
methyl benzoate the selectivity to benzonitrile and amides
are 48.1 and 24.1, respectively. For the subsequent study the
#
a
�
reaction temperature was fixed at 320 C.
3
.3. Reaction of Different Esters with Ammonia
The activity of the clay catalyst was examined in the re-
action of NH3 with various esters (Table 2). Most of the
esters gave the corresponding nitriles, amides, and alky-
lated amides. The amides were not obtained in the case of
ethyl propionate. This is the case with ethyl acetate also as
only acetonitrile was formed. This suggests that the con-
version of the intermediate amide into nitrile is quite fast
with aliphatic esters. Ethyl phenylacetate gave the highest
selectivity to nitrile among the various esters tried.
3
.4. Reaction of Methyl Benzoate [1] and NH3
over Various Catalysts
The conversion of 1 (Table 1) and selectivity to 2, 4, and 5
over different catalysts are depicted in Table 1. All the cata-
lysts were activated at the reaction temperature in a flow of
N2 for 2–3 h prior to startup of the reaction. The reactivity
of Montmorillonite K10 in this reaction leading to the for-
mation of 2, 4, and 5 is quite high. On the HZSM-5 catalyst,
even though conversion of 1 is high, the selectivity to 2 is
+
lower than on the clay catalyst. Na -Montmorillonite K10,
�
as expected, offers low reactivity and selectivity to 2 and
amides. The commercial ammoxidation catalyst has bare
minimum activity for the formation of the nitrile. On Nb2O5
FIG. 1. Effect of reaction time: 1 :NH3 = 1 : 1 (molar); 320 C; 2 g clay;
N2 = 20 ml/min. (᭪) Conversion of 1; (᭝�) selectivity to 2 (᭹); (Ճ) selec-
tivity to [4 + 5] (᭿).

MONTMORILLONITE CLAY CATALYSTS
87
TABLE 1
TABLE 3
Reaction of 1 and NH3 over Different Catalystsa
Effect of Molar Ratio of Methyl Benzoate to NH3 (1/NH3)a
b
Percent
conv.
of 1
Selectivity to
Selectivity to
Percent conv.b
1
/NH3
Temp. Time on
�
No.
(molar ratio)
1
2
(4 + 5)
Catalyst
( C) stream (h)
2
(4 + 5)
1
2
3
4
1
0.5
>1
0.2
62.8
64.7
61.4
67.0
48.1
59.6
19.6
44.8
24.1
20.2
27.7
14.7
Montmorillonite K10
320
1
6
6
6
6
6
5
66.7 52.5
62.8 48.1
45.0 30.7
60.4 22.4
23.8 12.0
23.5 14.0
18
b
b
b
b
b
b
24.1
19.4
28.8
—
—
—
+
Na -Montmorillonite K10
320
320
320
350
H-ZSM5
Nb2O5
Nb2O5
Commercial ammoxidation 320
catalyst
a
�
Montmorillonite K10 2g; 320 C; N2 = 30 ml/min.
b
Average values of 6 h run.
39.0
3.0
H-Montmorillonite K10
320
1
72.6 50.7
69.1 33.2
93.1 57.3
86.5 48.6
7.1
19.4
3.3
5b
selectivity to 4 and 5 is increased when 1/NH3 > 1. This
�
-Al2O3
320
1
5b
suggests that at higher concentration of NH , MeOH, and
11.3
3
H2O formed in the reaction are quickly removed from
the active sites, driving the reaction towards the formation
of 2.
a
Catalyst = 2 g; 1 :NH3 = 1 : 1 (molar ratio); N2 = 30 ml/min.
b
Integral value.
interestingly, the amides are not formed, even though the
selectivityto nitrile ispoor. Thisdifference can be attributed
to the difference in the acidity of the catalysts.
3
.6. Effect of Contact Time
The results (Table 4) show how the contact time affects
the conversion of 1 and selectivity to the products. As ex-
pected both conversion of 1 and selectivity to all the prod-
ucts increase with the increase in contact time. However,
the yield of product is less at the highest contact time, sug-
gesting triggering off some other side reaction.
3
.5. Effect of Molar Ratio of 1 to NH3
Various molar ratios of 1 to NH3 were flown over the clay
catalyst and the results are given in Table 3. Data collected
over a 6 h period shows this ratio does not influence the
overall conversion of 1. Selectivity to 2 is high when the
ratio of 1 to NH3 has been 0.5. On the other hand, the
3
.7. Reaction Pathways
In order to establish the actual pathways leading to the
formation of products, various combinations of substrates,
products, and catalyst were studied (Table 5). The products
2, 3, 4, and 5 are formed only when methyl benzoate and
NH3 are reacted over the Montmorillonite K10 at elevated
temperature. A one-to-one weight ratio of benzamide [4]
TABLE 2
Reaction of Various Esters with NH3 over Montmorillonite K10a
Percent
Selectivity to
Time on
stream
conv. of
ester
�
and the clay was mixed and heated at 320 C in the flow re-
Ester
nitrile
amides
actor under a flow of N2. The solid product recovered from
the reactor after 2 h shows the formation of benzoic acid
Methyl benzoate
6 h
62.8
63.3
63.6
48.1
48.5
49.7
24.1
23.5
22.6
5
4
h
h
Ethyl benzoate
6 h
34.2
39.0
37.2
50.0
40.3
38.0
20.1
17.0
15.0
TABLE 4
5
4
h
h
Effect of Contact Timea
Methyl p-toluateb
Ethyl phenylacetate
Ethyl propionate
5h
51.6
52.4
38.9
40.3
23.4
23.7
c
d
Selectivity to (yield )
Contact timeb
(h)
Percent conv.c
4
h
No.
1
2
(4 + 5)
5 h
59.5
60.5
66.6
67.2
33.3
32.8
4
h
1
2
3
4
0.00598
0.00196
0.00098
0.00049
79.5
64.7
52.4
35.6
45.7[0.0110]
59.6[0.0285]
49.7[0.0313]
35.4[0.0363]
18.7[0.0038]
20.2[0.0097]
13.9[0.0087]
7.9[0.0081]
6 h
58.4
57.4
59.2
44.5
46.1
43.6
—
—
—
5
4
h
h
Ethyl propionatec
4 h
67.2
73.3
—
a
b
c
d
�
3
20 C; Montmorillonite K10; 1 :NH3 = 1 :2.
a
�
Clay = 2 g; ester : NH3 = 1 : 1 (molar ratio); 320 C; N2 = 30 ml/min.
Contact time = [Volume ofclaycatalyst � time]/[Volume ofgasflown].
b
c
Heptane is used as diluent.
Average values of 6 h run.
�
At 340 C.
Yield = moles of product formed per gram of catalyst in 5 h run.

8
8
WALI ET AL.
TABLE 5
Reactivities of Substrates/Product on the Clay Catalyst
No.
Substratea
Catalyst
Products
1
2
3
4
5
6
Methyl benzoate
Methyl benzoate
—
Montmorillonite K10
—
—
—
—
Methyl benzoate + NH3
b
b
Methyl benzoate + NH3 Montmorillonite K10
2, 3, 4, 5
3
Ammonium
propionate
—
Benzamide
Montmorillonite K10
Montmorillonite K10
Propionic acid + NH3
7
Ammonium
propionate
Montmorillonite K10c
a
b
c
�
N2 = 30 ml/min; 320 C.
�
Catalyst = 2 g; 1 :NH3 = 1 : 1 (molar ratio); N2 = 30 ml/min; 320 C.
�
Catalyst = 2 g; N2 = 30 ml/min; 320 C.
Conversion of esters to amides and N-substituted amides
and the dehydration of amide to nitrile are known to be acid
[
3]. The most probable reaction here is the interaction of 4 catalyzed reactions (26). Therefore it is evident that the clay
with water molecules trapped inside the layers of Montmo- has acid sites that convert the ester to amide, N-substituted
rillonite clay. The benzoic acid formed in this reaction was amide, and nitrile. The origin of these sites was explored to
�
reacted with NH3 over clay at 320 C but part of it sublimed understand the actual mechanism of the reaction.
and choked the reactor. Hence, another acid, namely pro-
pionic acid, was allowed to react with NH3 over the catalyst. 3.8. Acidity of the Clay
�
Only ammonium propionate was formed at 320 C over the
The Montmorillonite K10 clay exhibits activity at tem-
peratures above 250 C. So the actual acidic sites that are
clay catalyst. In another run 10 wt% of methanol was fed,
along with methyl benzoate [1] and NH3, over Montmoril-
lonite K10 and the selectivity of the products was compared
�
responsible for the reaction were estimated by NH3 ad-
�
sorption at 250 C and the values are given in Table 6.
with the selectivity of products from methyl benzoate [1]
Among all the clay materials examined, Montmorillonite
and NH3 feed. On the addition of methanol, the selectivity
K10 has the maximum acidity. The acidity decreases af-
to 4 and 2 is drastically reduced. When this feed was re-
placed by methyl benzoate [1] and NH3 the selectivity to
and 2 increased. This clearly proves that the first step of
the reaction of methyl benzoate and NH3 involves the for-
mation of benzamide [4] and methanol in an equilibrium
reaction.
�
ter the 23 h run. The clay catalysts steamed at 320 C
+
and Na -Montmorillonite K10 have low acidity. The used
Na -Montmorillonite clay is a relatively poor solid acid
4
+
catalyst.
Similarly the influence of the mixtures of 1, 4, NH3 and 1,
NH3 feeds on Montmorillonite clay were compared. Here
the addition of 4 suppresses the selectivity to 2 because, as in
the case of the methanol addition run, the forward reaction
of 1 and NH3 leading to 4 and the methanol is suppressed.
When the mixture comprising 1, 4, and NH3 is replaced
by 1 + NH3 feed, the selectivity to 2 is not increased. This
study suggests that both 1 and 4 compete for the same sites
on the clay surface. As the NH3 adsorbed on clay cannot
react with 4, as it does with 1, the active sites are not re-
generated. When benzamide [4] and methanol are heated
TABLE 6
Estimation of Acid Sites of Montmorillonite Clays by NH3
�
Adsorption at 250 C
NH3 adsorbed
(mmole)/g � cat
No.
Clay catalysts
Naturea
1
2
Montmorillonite K10
Montmorillonite K10
Fresh
0.0476
0.0330
After 23 h run
and regenerated
Steamed for
6 h at 320 C
Steamed for
b
3
4
Montmorillonite K10
Montmorillonite K10
0.0320
0.0319
�
�
at 190 C in the presence of the clay catalyst in an autoclave
for about 2 h benzonitrile [2], N-methylbenzamide [5], and
�
10 h at 320 C
1
are detected in the organic fraction. This further estab-
5
6
Na-Montmorillonite K10 Fresh
Na-Montmorillonite K10 After 6 h run and
0.0270
0.0121
lishes that the formation of 4 from 1 is reversible in nature.
However, the same mixture on heating at 140 C for 14 h
b
�
regenerated
does not give benzonitrile.
Based on the above study the reaction pathways involved
in this reaction can be arrived as
a
All are used in extruded form.
The catalysts were regenerated by heating at 320 C under the flow
b
�
of air.

MONTMORILLONITE CLAY CATALYSTS
89
TABLE 7
Influence of Reaction Time on Selectivity to 2, 4, 5a
Selectivity to
Time on
Percent
stream, h
conv. of 1
2
4
5
1
2
3
66.7
64.0
60.0
52.5
54.1
48.9
3.6
4.0
4.9
14.0
16.5
18.9
a
Montmorillonite K10 = 2 g; 1 :NH3 = 1 : 1 (molar ratio); N2 = 30 ml/
�
min; 320 C.
3
.9. Turnover Frequency and Performance
of Catalyst with Time on Stream
As the acidity values of the different clay catalysts are dif-
ferent the actual evaluation of the catalyst would be to study
turnover frequency (TOF) of products (number of moles of
product formed per mole of acidic site of catalyst per sec-
ond) at different periods of reaction. Acidity of the clay is
a key factor for the reaction. NH3 adsorbtion provides the
total concentration of acid sites. It is essential to find among
Brønsted and Lewisacid siteswhich type ofacid sitesare im-
portant for formation of 2, 4, and 5. So H-Montmorillonite
K10 clay and � -Al2O3 have been examined as catalysts
FIG. 3. IR Spectrum of adsorbed pyridine on Montmorillonite K10
clay heated at 280 C.
�
(
Table 1). High conversions of 1 are observed on � -Al2O3
and H-Montmorillonite clay. But the selectivity to 2 is simi- of Brønsted acid sites with the increase in temperature is
lar on these two catalysts. H-Montmorillonite clay is a very known to occur due to polarisation and dissociation of wa-
strong solid acid (15), compared to Montmorillonite K10 ter (15, 29). When water is removed newer Lewis acid sites
clay, and has higher Brønsted acid sites (27). So very high are also exposed and activated.
acidic sites are not suitable for the formation of 2, compared
Figure 4 clearly shows that total TON in a 6 h run of both
to acid sitesavailable on Al2O3. Thissuggeststhe most prob- nitrile and amides increases with the reaction temperature,
able sites in the reaction are Lewis acid sites. High selectiv- whereas Fig. 2 shows the selectivity of the products is almost
�
ity to 4 + 5 is observed on Montmorillonite K10, HZSM-5, constant after 320 C. The efficiency and stability of the clay
and H-Montmorillonite K10 clay. Among the two amides, catalyst was studied by following the TON of benzonitrile
selectivity to 5 is always higher than 4 (Table 7). This sug- at every hour of the reaction (Fig. 5). There is a gradual fall
gests the possible involvement of very strong Brønsted acid in the efficiency of the clay catalyst with time on stream.
sites in the formation of 5.
This reveals that deactivation of the catalyst occurred dur-
Pyridine adsorption was carried out on Montmorillonite ing the course of reaction. It is known that methanol has an
�
K10 clay in the temperature range 280–340 C and the acid affinity for the clay surface (30). So it was thought that the
sites were estimated (Table 8) using IR absorbance values
�
1
at 1542 and 1444 cm for the Brønsted and Lewis acid sites,
respectively. Figure 3 shows the adsorbed pyridine species
TABLE 8
�
Estimation of Acid Sites on Montmorillonite K10 Clay
at Different Temperatures
on the clay heated at 280 C. Pyridine adsorption can distin-
guish between the Brønsted and Lewis acid sites (28–29).
The characteristic bands of pyridine at 1639, 1542, 1490 cm
are observed and indicate the presence of Brønsted acid
sites. Bands at 1600, 1490, 1444 cm which correspond to
pyridine adsorbed on Lewis acid sites are also observed.
Ratios of Brønsted acid to Lewis acid sites are calculated
from the clay samples activated at different temperatures
Brønsted acid sitesa
Pyridine adsorbed/g � cat
�
�
1
No.
Temp. ( C)
(mmole)
Lewis acid sites
1
2
3
4
280
300
320
340
0.0683
0.0730
0.0774
0.0825
0.067
0.037
0.049
0.096
(
Table 8). The ratios of Brønsted to Lewis acid sites, as
well as the amount of sites increase with the increase in the
a
Relative proportion of Brønsted and Lewis acid sites is found from
�
1
clay activation temperature. The increase in concentration IR peaks at 1542 and 1444 cm
.

9
0
WALI ET AL.
by the methanol extraction was not successful. So after an
8 h run the flow of 1 and NH3 was suspended. The catalyst
1
was purged with N2 and then in a flow of air it was heated
�
at 320 C for 4 h. Subsequently this catalyst was used for
the conversion of ester to nitrile. In this method also the
catalyst behaved in a way similar to that in the previous
treatment even though the initial activity at 19 h was high.
The catalyst, after the reaction, was dried and weighed and
showed 8% weight increase. The same catalyst after calcina-
�
tion in air at 320 C shows that organic residue equivalent to
8
wt% increase is removed. This also shows that the above
two methods of regeneration of catalyst did not work with
Montmorillonite clay catalyst.
Therefore, in order to understand the nature of surface-
active sites of the catalyst, the deactivation phenomenon,
and the mechanism of reaction, the clay samples were ex-
amined by ESCA.
3
.10. ESCA Results
The XPS survey spectra of Montmorillonite K10 and its
�
extrudates calcined at 320 C are shown in Fig. 6. An im-
provement in surface acidity (Si/Al: 4.27 → 3.15, Table 9)
of extrudates suitable for reaction indicates enrichment of
aluminum atoms on the surface during the extrusion pro-
FIG. 4. Turnover number (integral value) vs reaction temperature.
Reaction conditions are the same as in Fig. 2: (᭪) 2; (Ճ) (4 + 5).
deactivated catalyst could be regenerated by refluxing with cess. Such changes in (Si/Al) values on the surface and in
methanol. After 13 h continuous run the catalyst was taken bulk of thermally treated zeolites have been reported by
out, refluxed with methanol, dried under vacuum, and then several authors (25, 31, 32).
used for the conversion of esters to nitrile. An initial in-
3
During the reaction of methyl benzoate with NH the
crease in activity of the catalyst was observed. However, latter is adsorbed on acidic sites of extrudates, resulting in
�
1
h after that the catalyst was behaving the same way as an increase in Si/Al atomic ratio (3.56) up to 280 C with
before. This suggests that the regeneration of the catalyst approximately 22% acidic sites covered with basic nitrogen
FIG. 5. Efficiency of Montmorillonite clay catalyst with time on stream. Reaction conditions are the same as in Fig. 1.

MONTMORILLONITE CLAY CATALYSTS
91
FIG. 6. XPS survey spectra of Montmorillonite (a) powder and (b) extrudates.
�
compounds. As the reaction temperature is increased to in the reaction which increases initially up to 320 C and
�
�
�
3
20 C and then to 340 C, weakly bonded nitrogen and its then becomes constant above 320 C.
compounds desorb from the surface, resulting in a decrease In order to study deactivation behaviour of the catalyst,
in Si/Al values to 3.31 and 3.17, respectively. However, the the Si/Al atomic ratio on the surface were estimated for
�
amount of the nitrogen bonded to aluminum atoms on the catalyst runs for 6, 13, and 23 h at 320 C. An increase in
surface was found to be constant (� 15% ). This behaviour of Si/Al ratio from 3.15 to 3.31, 3.41, and then to 3.62 indicates
the catalyst can be correlated with the selectivity of nitriles surface dealumination and, hence, decrease in acidity. It can
also be seen from Fig. 7 that a prolonged (23 h) exposure
of the catalyst to reactants results in a large accumulation
of carbon and nitrogen compounds on the surface.
The broadening in N(1S) photoelectron lines of used
catalyst shown in Figs. 8b and 8c indicate the presence of
more than one nitrogen species on its surface. An active
TABLE 9
Surface Quantitative Results for Extrudates after Various
Treatments (Atomic Ratio)
catalyst (6 h run) has N(1S) binding energy close to the
NH3 adsorbed in fresh catalyst and deactivated catalyst has
Sample
Si/Al
N/Al
Na/Al
Powder Montmorillonite
Extrudates—fresh
Extrudates
4.27
3.15
—
—
—
—
N(1S) binding energy close to the amide (–CONH ) func-
tionalities. In fact the binding energy values and line shape
2
data of N(1S) photoelectron lines suggest the formation
�
�
2
3
3
3
3
80 C after 6 h run
3.56
3.31
3.17
3.41
3.62
3.67
0.22
0.15
0.15
—
—
—
—
—
+
of ammonium ion (NH ) (33) and amide (CONH2) func-
20 C after 6 h run
4
+
�
tionalities (34) during the reaction; NH ions participate in
40 C after 6 h run
4
�
20 C after 13 h run
Not estimated
the reaction while the presence of an amide on the surface
leads to deactivation of the catalyst. The catalyst regener-
�
a
20 C after 23 h run
0.87
�
0.56a
Extrudates—regenerated at 340 C
for 6 h after 23 h run at 320 C
�
�
ated at 340 C after a 23 h run at 320 C gives N/Al = 0.56
with N(1S) binding energy of 400 eV which is charactristic
of benzamide. Thus accumulation of amides on the surface
and decrease in surface acidity of the catalyst are the main
reasons of catalyst deactivation.
�
Na exchanged extrudates, fresh
3.31
3.68
3.44
—
—
—
0.23
0.38
—
�
used (320 C after 6 h run)
Fresh extrudates dealuminated
�
with steam (320 C for 6 h)
Fresh extrudates dealuminated
3.47
—
—
+
In Na -Montmorillonite (extrudates), it is presumed
�
with steam (320 C for 10 h)
that sodium atoms are exchanged with the acidic sites
of Montmorillonite. For these extrudates the estimated
a
N(1S) B.E. = 400 eV.

9
2
WALI ET AL.
FIG. 7. XPS survey scans of (a) fresh extrudates and (b) used extrudates (23 h) of Montmorillonite K10.
value of the Si/Al ratio was 3.31 and the Na/Al atomic
ratio on the surface was 0.23 (Table 9). Thus acidity of
these extrudates is lower (Table 6), compared to fresh ex-
trudates, which is also supported by their catalytic data.
+
When Na -Montmorillonite (extrudates) were used in the
reaction, sodium enrichment on the surface takes place
(
Na/Al = 0.38) and, hence, surface acidity further goes
down, resulting in the poor activity of exchanged extru-
dates on use. XPS survey scans of fresh and used Na -
+
Montmorillonite (extrudates) are shown in Fig. 9. The bind-
ing energy of Na(1S) photoelectron line in these extrudates
is 1073 � 0.2 eV (Fig. 10) which is close to the reported
value (35) for NaY zeolite. It is worth mentioning that ni-
trogen species were not detected on the surface of used
+
Na -Montmorillonite (extrudates).
As sodium and nitrogen are adsorbed on acidic sites of
alumina in Montmorillonite, it is possible to correlate the
effective active sites on the surface of clay extrudates to
catalytic activity as shown in Table 10. Thus acidic sites at-
tached to aluminum atoms of clay are responsible for activ-
ity behaviour of the catalyst. Steam dealumination of fresh
extrudates was carried out for 6 and 10 h to see the extent
of aluminum atoms leaching out from the surface of the
silica-alumina framework. An increase in Si/Al ratio from
3
.15 to 3.44 and 3.47 (Table 9) indicates that 8.5% of alu-
minum atoms are leaching out and dealumination is nearly
�
complete in 6 h steaming of extrudates at 320 C. This is sup-
ported by TPD data shown in Table 6 which gives nearly
the same ammonia adsorption values for these two dealu-
minated extrudates.
FIG. 8. N(1S) photoelectron lines of (a) benzamide, (b) extrudates of
Montmorillonite K10 after 23 h run, (c) extrudates after 6 h run, (d) fresh
extrudates after adsorption of NH3.

MONTMORILLONITE CLAY CATALYSTS
93
+
FIG. 9. XPS survey spectra of Na -Montmorillonite K10 (extrudates) in (a) fresh and (b) used stage.
3
.11. Mechanism
morillonite K10 also the predominant formation of 5 sug-
gests it is catalyzed by Brønsted acid sites.
The above studies clearly indicate the acid sites respon-
The deactivation is clearly due to loss of acidity on the
sible for the catalysis are on the aluminium atoms. We have
observed that Brønsted acid sites and Lewis acid sites (pre-
dominantly) are present on Montmorillonite clay. The facile
reaction of 1 and NH3 on � -Al2O3 to give 2 and the similar
trend on clay suggests the participation of Lewis acid sites
in this reaction. It is reported on similar solid acid catalyst
�
surface of clay because of dealumination by water at 320 C.
This loss of aluminum atoms is estimated to be 8.5 wt% ; i.e.,
every one aluminium atom out of 11–12 atoms is leached
out. This results in accumulation of negative charges on
the clay. The amide derivatives can strongly adsorb on the
clay to neutralise the charges. In the process some more
2�
SO4 /ZrO2 dehydration of amide to nitrile is catalyzed by
Lewis acid sites (36). So it is most probable that formation
3+
of 2 is catalyzed by Lewis acid sites (Al ) of clay. Involve-
ment of both Brønsted and Lewis acid sites in the formation
of 4 cannot be ruled out. In presence of H-Montmorillonite
K10 and H-ZSM5 the formation of 2 is decreased with the
increase in formation of 4 and 5 (predominantly). On Mont-
TABLE 10
Correlation of Catalytic Data with Active Sites Estimated
from Surface Spectroscopic Data
Si/Al
atomic
Effective
active
Sample
ratio sites (at% ) Ratioa Activityb Ratioc
1. Fresh
3.15
8.26
86.4
extrudates
1
.24
.40
1.21
1.36
2
3
. Used
3.31
3.31
6.80
6.10
69.7
67.8
extrudates (6 h)
+
. Na exchanged
fresh extru.
1
+
4
. Used Na exch.
3.68
4.50
48.4
extrudates (6 h)
a
Active sites of fresh extrudates/active sites of used extrudates.
Turnover of nitrile per hour activity taken after 1 h stabilization in
b
+
FIG. 10. Na(1S) photoelectron spectrum of Na -Montmorillonite
K10 (extrudates).
case of fresh extrudates.
c
Activity of fresh extrudates/activity of used extrudates.

9
4
WALI ET AL.
acid sites become inaccessible to the reactants for further
reaction to take place.
5. Chauvin, A., and Lefebvre, G., in “Petrochemical Processes,” Chap.
1, Vol. 2, Gulf, Paris, 1989.
1
6. Kim, Y. C., Ueda, W., and Morooka, Y., Appl. Catal. 70, 189 (1991).
7. Denton, W. I., and co-workers, Ind. Eng. Chem. 42, 796 (1950).
8. Friedrich, K., and Wallenfels, K., in “The Chemistry of Cyano Group”
4
. CONCLUSIONS
(
Z. Roppo Port, Ed.), pp. 67–122. Interscience, New York, 1970.
It has been shown that Montmorillonite K10 clay is
9. Norris, J. A., and Sturgis, B. M., J. Am. Chem. Soc. 62, 1413 (1940).
an active catalyst in the production of benzonitrile [2], 10. DuPont De Nemours & Co., Inc., U.S. Patent 2,736,739 (1956).
1
1
1. Richmond, M. H., Brit. Patent 968752 (1964).
2. Nagaiah, K., Kulkarni, S. J., Subrahmanyam, M., and Rama Rao,
A. V., J. Chem. Soc. Chem. Commun., 273 (1994).
benzamide [4], and N-methyl benzamide [5] from methyl
benzoate on reaction with NH3 in the temperature range
�
2
80–340 C. Various reaction parameters which influence
1
3. Mitchell, J. A., and Reid, E. E., J. Am. Chem. Soc. 53, 321 (1931).
the conversion of ester and the selectivity of the products 14. Nagaiah, K., Kulkarni, S. J., Subrahmanyam, M., and Rama Rao, A. V.,
have been studied. This catalyst is equally active for the
conversion of other esters into corresponding nitriles. From
NH3, as well as pyridine adsorption, the amount of acidic
sites participating in the reaction has been estimated. There
Ind. J. Chem. Tech. 1, 356 (1994).
5. Muthukumaru Pillai, S., and Ravindranathan, M., J. Chem. Soc. Chem.
Commun., 1813 (1994).
6. Muthukumaru Pillai, S., Wali, A., and Satish, S., React. Kinet. Catal.
Lett. 55, 251 (1995).
1
1
is slow and gradual deactivation of the clay catalyst during 17. Wali, A., Muthukumaru Pillai, S., and Satish, S., “Catalysis Present
the reaction.
and Future,” p. 167. Wiley Eastern, New Delhi, 1995. [International
Series on Chemical Technology]
8. Choudary, B. M., Valli, V. L. K., and Durga Prasad, A., Syn. Commun.
ESCA study of the clay extrudates at various stages of
preparation and reaction is carried out. The surface acidity
of the Montmorillonite increases in the extrusion process
because of surface enrichment of aluminum atoms. During
1
1
21, 2007 (1991).
9. Buehler, C. A., and Pearson, D. E. (Eds.), in “Survey of Organic Syn-
theses,” Chap. 14, p. 802. Wiley-Interscience, New York/London, 1970.
the reaction NH3 is adsorbed on the acidic >Al–OH sites 20. Wali, A., Muthukumaru Pillai, S., Unnikrishnan, S., and Satish, S.,
+
J. Mol. Catal. A: Chem. 109, 149 (1996).
and forms ammonium ions (NH ). These participate in the
4
2
2
2
1. Kaushik, V. K., Spectrochim. Acta B 44, 581 (1989).
2. Kaushik, V. K., J. Electron Spectra Related Phenom. 56, 273 (1991).
3. Tkatchenko, O. P., Shpiro, E. S., Wark, M., Schulz-Ekloff, G., and
Jaeger, N. I., J. Chem. Soc. Faraday Trans. 89, 3987 (1993).
reaction. The deactivation of the catalyst was mainly due to
dealumination by water produced in the reaction and ac-
cumulation of organic residue containing amide groups on
the surface of the clay. As a result of this study it is possible 24. Deforsee, C., Delmon, B., and Canesson, P., in “Molecular Sieves II”
(
[
J. R. Katzer, Ed.), p. 76. Am. Chem. Soc., Washington, DC, 1977.
ACS Symposium Series 40]
to calculate effective acidic sites on the clay surface and,
hence, a correlation with the performance of the catalyst is
possible.
2
2
5. Kaushik, V. K., Bhat, S. G. T., and Corbin, D. R., Zeolites 13, 671 (1993).
6. March, J., in “Advances in Organic Chemistry,” pp. 375, 932. Wiley
Eastern, New Delhi, 1986.
2
7. Clark, J. H., Martin, K., Teasdale, A. J., and Barlow, S. J., J. Chem. Soc.
Chem. Commun., 2037 (1995).
ACKNOWLEDGMENTS
2
2
8. Parry, E. P., J. Catal. 2, 371 (1963).
We thank the management of Indian Petrochemicals Corporation
Limited, Baroda, for permission to publish this work. Thanks are due
to Mr. A. B. Parikh and Mr. A. D. Patel for technical support.
9. Laszlo, P. (Ed.), in “Preparative Chemistry Using Supported
Reagents,” p. 456. Academic Press, New York, 1987.
0. Khatib, K., Bottero, J. Y., Pons, C. H., Uriot, J. P., and Anselme, C.,
Clay Miner. 29, 401 (1994).
3
REFERENCES
31. Okamoto, Y., Ogawa, M., Maezawa, A., and Imanaka, T., J. Catal. 112,
27 (1988).
32. Barr, T. L., Appl. Surf. Sci. 15, 1 (1988).
chemical Processes,” p. 108. Kluwer Academic, Boston/London, 1989. 33. Bradley, R. H., Beamson, G., Ling, X., and Sutherland, I., Appl. Surf.
4
1
2
. Matar, S., Mirbach, M. J., and Tayim, H. A., in “Catalysis in Petro-
. Othmer, K., “Encyclopedia of Chemical Technology” (H. F. Mark,
D. F. Othmer, C. G. Overberges, G. T. Seaborg, M. Grayson, and
D. Eckroth, Eds.), Vol. 15, p. 888. Wiley, New York, 1981.
. Lebedev, N. N., in “Chemistry and Technology of Basic Organic and
Petrochemical Synthesis,” Vol. 1. p. 237. Mir, Moscow, 1984.
. Fuhrhop, J., and Penzlin, G., in “Organic Synthesis,” p. 303. VCH,
Weinheim, 1994.
Sci. 72, 273 (1993).
34. Bradley, R. H., Ling, X., Sutherland, I., and Beamson, G., Carbon 32,
185 (1994).
35. Kaushik, V. K., and Ravindranathan, M., Zeolites 12, 415 (1992).
36. Rajdhyaksha, R. A., and Joshi, G. W., in “Heterogeneous Catalysis and
Fine Chemicals II,” p. 479. Elsevier, Amsterdam, 1991. [Stud. Surf. Sci.
Catal. 59 (1991)]
3
4