A ritter-type reaction over H-ZSM-5: Synthesis of N-isopropylacrylamide from acrylonitrile and isopropyl alcohol

Article abstract of DOI:10.1006/jcat.2002.3546

Catalytic synthesis of N-isopropylacrylamide from acrylonitrile and isopropyl alcohol has been studied using a variety of solid acids in a solid-liquid reaction system at 423 K. Among typical solid and liquid acids, H-ZSM-5 exhibited an exceptionally high catalytic activity. The activity (per gram) of H-ZSM-5 increased as the Al-content (100Al/(Si + Al)%) increased and then decreased through a maximum at an Al content of 2.63%. The specific activity per one acid site, which was estimated from the initial rate and the acid amount, increased greatly as the Al content decreased, which resembles those of the hydrophobicity and the acid strength. The superiority of H-ZSM-5 activity over other strong and hydrophobic solid acids suggests the importance of the unique pore structure. While H-ZSM-5 deactivated severely during the reaction, the catalytic activities were mostly recovered by calcination at 773 K in air. IR spectroscopy and adsorption measurements revealed that the deactivation of H-ZSM-5 was mainly caused by the formation of a polymer of acrylonitrile on the catalyst surface blocking the micropores, but not limiting desorption of the product from the micropores.

Full text of DOI:10.1006/jcat.2002.3546

Journal of Catalysis 207, 194–201 (2002)
doi:10.1006/jcat.2002.3546, available online at http://www.idealibrary.com on
A Ritter-Type Reaction over H–ZSM-5: Synthesis of
N-Isopropylacrylamide from Acrylonitrile and Isopropyl Alcohol
Xin Chen and Toshio Okuhara1
Graduate School of Environmental Earth Science, Hokkaido University, Sapporo 060-0810, Japan
Received June 11, 2001; revised January 30, 2002; accepted January 30, 2002
Besides H2SO4, BF3 was reported to be effective for
the reaction of benzyl alcohol with various nitriles (3).
In addition, trifluoromethane sulfonic anhydride gave N-
alkylamide from primary alcohols such as 1-pentanol and
acetonitrile (4), while the sulfonic anhydride was added to
alcohol at a 1 : 1 molar ratio. On the other hand, a metal
complex, Pd[(CH3CN)2(PPh3)2](BF4)2, showed activities
for Ritter- type reactions of acrylonitrile or acetonitrile with
tert-butyl alcohol (5). These homogeneous systems have
also the problems of separation and disposal. Thus solid
catalysts in the place of H2SO4 are desirable for environ-
mentally benign catalytic processes.
There are some reports on solid acids effective for Ritter-
type reactions. A sulfonated polymer resin, Nafion–H, was
active for the reaction of acrylonitrile and benzyl alcohol
(6). An acidic Cs salt of H3PW12O40, Cs2.5H0.5PW12O40,
showed a high activity and selectivity for the reac-
tion of acrylonitrile and 1-adamantanol (7). In addi-
tion, some patents described Ritter reactions catalyzed by
Cs2.5H0.5PW12O40 (8), H3PW12O40 (9), or zeolites (10).
However, solid acids effective for the reaction of acryloni-
trile with isopropyl alcohol were not found, except for those
mentioned in our recent preliminary report (11).
Catalytic synthesis of N-isopropylacrylamide from acrylonitrile
and isopropyl alcohol has been studied using a variety of solid acids
in a solid–liquid reaction system at 423 K. Among typical solid and
liquid acids, H–ZSM-5 exhibited an exceptionally high catalytic
activity. The activity (per gram) of H–ZSM-5 increased as the Al-
content (100Al/(Si + Al)%) increased and then decreased through
a maximum at an Al content of 2.63%. The specific activity per
one acid site, which was estimated from the initial rate and the
acid amount, increased greatly as the Al content decreased, which
resembles those of the hydrophobicity and the acid strength. The
superiority of H–ZSM-5 activity over other strong and hydrophobic
solid acids suggests the importance of the unique pore structure.
While H–ZSM-5 deactivated severely during the reaction, the cata-
lytic activities were mostly recovered by calcination at 773 K in
air. IR spectroscopy and adsorption measurements revealed that
the deactivation of H–ZSM-5 was mainly caused by the formation
of a polymer of acrylonitrile on the catalyst surface blocking the
micropores, but not limiting desorption of the product from the
c
micropores. ꢀ 2002 Elsevier Science (USA)
Key Words: Ritter reaction; N-isopropylacryamide; acrylonitrile;
isopropyl alcohol; zeolites.
INTRODUCTION
Here we wish to report on the catalytic reaction of acry-
lonitrile with isopropyl alcohol to N-isopropylacrylamide
over a zeolite, H–ZSM-5, in a solid–liquid reaction sys-
tem. The catalytic activity and selectivity of H–ZSM-5 have
been compared with those of other typical solid acids. The
influence of Al content in H–ZSM-5 on the activity was
examined by using various H–ZSM-5s having different Al
content. Furthermore, the origins for deactivation of H–
ZSM-5 detected during the reaction have been investigated
by IR spectroscopy and adsorption measurement of the
product. The reusability of H–ZSM-5 by calcination has
been examined.
N-isopropylacrylamide is an important monomer for
water-soluble polymers which are applicable to water-
soluble fibers, thickeners, and so forth (1). Ritter and
Minieri (2) first reported five decades ago that N-
alkylation of acrylonitrile with isopropyl alcohol to N-
isopropylacrylamide (Eq. [1]) proceeded in the presence
of excess H2SO4. Although this monomer has commer-
cially been produced by the so-called “Ritter reaction,” this
H2SO4 process contains problems of corrosion of the reac-
tor and a large amount of catalyst waste.
==
–
==
–
–
CH2 CH CN + i-C3H7OH→CH2 CH CONH i-C3H7.
EXPERIMENTAL
[1]
As solid acids, a variety of H–ZSM-5 having different
Al content were used. Here the Al content is defined
as 100 × [Al]/([Al] + [Si])%, where [Al] and [Si] are the
1 To whom correspondence should be addressed. Fax: 81-11-706-4513.
E-mail: oku@ees.hokudai.ac.jp.
194
0021-9517/02 $35.00
c
ꢀ 2002 Elsevier Science (USA)
All rights reserved.

RITTER REACTION OVER H–ZSM-5
195
amounts of Al and Si atoms in the zeolite. A sample of tion of N-isopropylacrylamide was followed at room tem-
H–ZSM-5 having an Al content of 2.63% (Tosoh, HSZ- perature with the GC.
860HOA) was mainly used. Other H–ZSM-5s—JRC-Z5-
Other Measurements
25H (Al content = 7.1%, Reference Catalysts of Catalysis
Society of Japan (abbreviated as RC)), JRC-Z5-70H (Al
The concentrations of acid site were measured by ther-
content = 2.4%, RC), and HSZ (Tosoh, Al content =
mal desorption of NH3 using a TPD apparatus (BEL
2.70%)—were also used. In addition, HSZ870NHA (Al
Japan, Multi-task TPD) (16), where the TPD spectrum of
content = 1.0%, Tosoh), MFI-300 (Al content = 0.7%,
NH3 from H–ZSM-5 (JRC-Z5-25H) was used as a stan-
UOP), MFI-40 (Al content = 4.8%, UOP), and CBV3024
dard, assuming that the concentration of acid amount is
(Al content = 3.3%, Zeolyst Int.) were also utilized.
Besides H–ZSM-5, H–mordenite (HM-20, RC) and HY
(HY4.8, RC) were tested. In addition, SiO2–Al2O3
(SAL-2, Al2O3 = 13 wt%, RC), sulfated zirconia (denoted
to SO42−/ZrO2) which was prepared according to the
literature (12), and Cs2.5H0.5PW12O40 prepared by a titra-
tion method using aqueous solutions of H3PW12O40 and
Cs2CO3 (13) were also compared. Nafion–H (Du Pont, NR-
50), Nafion–SiO2 composite (14) (SAC 13, Nafion 13 wt%,
Du Pont), Aciplex–SiO2 (15) and Amberlyst 15 (Organo)
were examined as polymer resin catalysts. H2SO4 (Wako
Chemicals) and anhydrous H3PW12O40 (Nippon Inorganic
Color and Chemical Co.), which was obtained by evacua-
tion (343 K) of H3PW12O40 13H2O as received, were used
as homogeneous acid catalysts.
1.2 mmol g−1 (17). The surface area was measured by BET
method using an N2 adsorption system (BEL Japan 28SA).
Infrared spectra of H–ZSM-5 before and after the reaction
were taken with an IR spectrometer (Biorad, FTS-7) by
KBr method at room temperature. As standard IR samples,
polyacrylonitrile and poly-N-isopropylacrylamide, which
were kindly supplied from Kohjin Co., were used. The
amounts of carbon and nitrogen atoms remaining on the
used catalysts were determined with an Elemental Analy-
sis System (Yanaco CHN Corder MT-5). For the analysis,
the solid catalysts were separated from the reaction suspen-
sion by centrifugation and then were dried in air at room
temperature.
RESULTS
Catalytic Reactions
The surface areas, pretreatment conditions, and con-
centrations of acid sites are summarized in Table 1. In
Table 2, the yield and selectivity for the reaction of acry-
lonitrile with isopropyl alcohol are summarized. The yield
of N-isopropylacrylamide (abbreviated PAA) is expressed
on the basis of isopropyl alcohol (abbreviated IPA). It is
noted that H–ZSM-5 having an Al content of 2.63% gave
PAA with a yield of 62.2% for 24 h of the reaction. This
catalyst was far superior in yield to the other solid acids
and liquid acids. Besides the zeolite, Cs2.5H0.5PW12O40
and Amberlyst gave PAA, but the yields were only 10%.
H3PW12O40 worked as the liquid acid and gave the yield
of 15.5%. H2SO4 was not effective under these reaction
conditions.
The percentage-conversion of acrylonitrile (abbreviated
AN) and IPA are also provided in Table 2. The ratio in
the conversion of IPA (or the yield of PAA) to that of AN
should be 5, if the reaction took place without the side reac-
tions, considering the molar ratio (=5) of AN to IPA in the
reactant mixture. In addition, the yield of PAA should be
equal to the conversion of IPA. However, the conversion
of AN was higher than the value of the yield divided by
5, and the conversion of IPA was higher than the yield of
PAA over H–ZSM-5. Side reactions such as dehydration of
IPA to propene and/or diidopropyl ether (abbreviated as
PE) took place. PE was formed, but the selectivity to PE
was only 4% (Table 2). Propene was probably produced in
the gas phase, while the gas phase was not quantitatively
analyzed in the present study.
The reaction was ordinarily performed in a glass auto-
clave at 423 K in an N2 atmosphere using acrylonitrile,
10 cm3 (150 mmol), isopropyl alcohol, 2.2 cm3 (30 mmol),
and 1 g of catalyst (11). Since acrylonitrile is a carcinogenic
substance, scrupulous care should be taken in handling this.
Hydroquinone (0.05 g) was added as an inhibitor of poly-
merization. As an internal standard, n-dodecane (0.25 cm3)
was added to the suspension. Prior to use, the catalyst was
pretreated at elevated temperatures (the conditions are de-
scribed in Table 1). Water as an impurity in acrylonitrile
was removed with a Molecular Sieve 5A calcined at 573 K
in air. The air in the reactor was replaced with nitrogen, and
then the reaction was performed at 423 K for a certain time
with vigorous stirring. Usually the pressure in the reactor
increased to about 3 atm after the heating of the reactor.
After the reaction, the reactant solution was cooled to room
temperature, and then the products were analyzed by gas
chromatography using an FID-GC (Shimadzu, GC 14B)
equipped with a column of Carbowax 300M Chromosorb
WAW (2 m).
Adsorption of Products
The adsorption of the product, N-isopropylacrylamide,
was measured in a sealed glass tube (Pyrex glass, 20 cm3)
at room temperature. After H–ZSM-5 in the sealed tube
was pretreated at 373 K for 3 h in a vacuum, solution
of N-isopropylacrylamide in 1,3,5-trimethylbenzene was
introduced into the reactor. The change in the concentra-

196
CHEN AND OKUHARA
TABLE 1
The Amount of Acid, Pretreatment Conditions, and Surface Area
Concentrations of acid
Pretreatment
conditions
Surface area
Catalyst
[Si]/[Al] ratio
sites (mmol · g−1
)
(m2 · g−1
)
Solids
Cs2.5H0.5PW12O40
HY
H–mordenite
0.15a(0.063)b
2.60c
423 K, 3 h in He
673 K, 3 h in air
673 K, 3 h in air
773 K, 3 h in air
773 K, 3 h in air
773 K, 3 h in air
773 K, 3 h in air
773 K, 5 h in air
773 K, 5 h in air
773 K, 3 h in air
773 K, 3 h in air
773 K, 5 h in air
773 K, 5 h in air
130
740
371
560
125
379
414
458
514
410
400
367
418
1.40c
SiO2–Al2O3
0.35c
SO24−/ZrO2
0.20c
H–ZSM-5 (Al = 7.10%)
H–ZSM-5 (Al = 4.80%)
H–ZSM-5 (Al = 3.30%)
H–ZSM-5 (Al = 2.70%)
H–ZSM-5 (Al = 2.63%)
H–ZSM-5 (Al = 2.40%)
H–ZSM-5 (Al = 1.00%)
H–ZSM-5 (Al = 0.70%)
[13]
[20]
[30]
[36]
[37]
[40]
[95]
[150]
1.20c
0.77c
0.65c
0.37c
0.31c
0.37c
0.20c
0.11c
Organic resins
Amberlyst 15
Nafion–H
Nafion–SiO2
Aciplex–SiO2
4.70d
0.80d
0.12d
0.46d
as received
as received
as received
as received
50
0.02
340
1.3
Liquids
H3PW12O40
H2SO4
1.0a
423 K, 3 h in He
as received
—
—
20.4a
a From chemical formula.
b Amount of surface.
c From TPD of NH3.
d Ion exchange capacity.
TABLE 2
Yield and Selectivity of N-Isopropylacrylamide from Acrylonitrile and Isopropyl Alcohol over Acid Catalyst
Conversionc (%)
Selectivityd (mol %)
Yielda
(%)
Catalyst
TONb
AN
IPA
PAA
AA
PE
Solid acids
H–ZSM-5 (Al = 2.63%)
62.2
7.4
1.5
2.8
7.7
10.0
10.0
3.0
6.0
6.5
60.2
0.8
0.3
23.5
9.7
29.4
12.5
6.7
21.5
24.6
33.3
20.5
17.6
86.2
37.0
41.3
28.6
40.0
50.0
83.3
60.0
57.5
48.3
93.3
52.6
68.8
72.7
54.8
74.7
68.7
59.4
46.9
34.4
3.1
7.9
12.5
0.0
4.7
5.1
10.7
12.5
16.7
12.4
3.6
39.5
18.7
27.3
40.5
20.1
20.6
28.1
36.4
53.2
HY
H–mordenite
SiO2–Al2O3
SO24−/ZrO2
Cs2.5H0.5PW12O40
Amberlyst 15
Nafion–SiO2
Nafion–H
2.0
11.5
47.6
0.6
8.0
2.1
Aciplex–SiO2
4.2
Liquid acids
H3PW12O40
H2SO4
15.5
4.8
4.7
0.4
15.1
29.4
63.8
65.5
47.4
27.5
10.5
39.2
42.1
33.3
e
Note. Reaction conditions: catalyst, 1.0 g; acrylonitrile, 150 mmol; isopropyl alcohol, 30 mmol; 423 K for 24 h.
a Yield = 100 × (mol of N-isopropylacrylamide)/(mol of isopropyl alcohol added initially).
b Number of N-isopropylacrylamide formed divided by the number of protons on the surface.
c The conversion is defined as (mol of acrylonitrile or isopropyl alcohol disappeared)/(mol of acrylonitrile or isopropyl alcohol present initially).
Acrylonitrile and isopropyl alcohol are abbreviated AN and IPA.
d Selectivityto N-isopropylacrylamideisdefinedas100×(mol of N-isopropylacrylamide)/[(mol of N-isopropylacrylamide) + (mol of acrylamide) +
(mol of diisopropyl ether)]. N-isopropylacrylamide, acrylamide, and diisopropyl ether are abbreviated PAA, AA, and PE, respectively.
e 0.18 g of the catalyst was used.

RITTER REACTION OVER H–ZSM-5
197
TABLE 3
Yield and Selectivity of N-Isopropylacrylamide from Acrylonitrile and Isopropyl Alcohol over H–ZSM-5
Conversionb (%)
Selectivityb (%)
AA
Amountc (mmol · g−1
)
Ala
(%)
Yield
(%)
TONb
AN
IPA
PAA
PE
C
N
7.10
4.80
3.30
2.70
2.63
2.40
1.00
0.70
18.0
32.9
33.3
43.3
52.2
42.5
38.3
23.6
4.5
12.8
15.4
35.1
50.5
34.5
57.5
64.4
23.5
25.0
23.5
20.0
11.1
6.5
70.3
76.9
70.0
78.6
71.0
68.8
65.0
33.3
82.5
86.8
88.5
89.0
92.0
87.9
90.9
88.3
15.8
9.0
7.1
5.5
4.6
5.3
4.7
4.8
1.7
4.2
4.4
5.5
3.4
6.8
4.4
6.9
—
23.6
10.8
11.8
10.9
—
—
4.8
2.2
2.3
1.7
—
7.8
6.2
—
11.5
—
1.8
Note. Reaction conditions: catalyst, 1.0 g; acrylonitrile, 150 mmol; isopropyl alcohol, 30 mmol; 423 K for 6 h.
a Al content.
b Yield, TON, conversion, and selectivity are the same as those in Table 2.
c Amounts of C and N atoms remaining on the surface measured by elemental analysis.
In addition, the conversions of AN higher than the value
Figure 1 shows the time courses of the reaction of AN
of the PAA yield divided by 5 might be brought about by withIPAat423KoverH–ZSM-5swithdifferentAlcontent.
the polymerization of AN, hydration of AN to acrylamide As shown in Fig. 1, the reaction took place rapidly at the
(abbreviated AA), and so forth. Since the selectivity to AA initial stage over all these H–ZSM-5s, but the reaction rates
was 4%, the polymerization of AN might be appreciable. became lower gradually. Figure 2 presents time courses of
Similar trends were observed over the other solid acids, as the accumulation of N atoms during the reaction. There are
shown in Table 2.
trends showing that the accumulated amount of N atom
The conversion of AN over H–ZSM-5 (Al = 2.63%), increased as the Al content increased.
23.5%, can be divided into those to PAA (62.2/5 = 12.5%),
Figure 3 shows the acid amount (Fig. 3A), the initial rates
AA (about 2%), and the polymer of AN (about 9%). of the reaction, and N atom accumulation (Fig. 3B) as a
In the cases of Amberlyst, Nafion, Nafion–SiO2, and function of the Al content. The acid amounts were esti-
Cs2.5H0.5PW12O40, the conversions of IPA were above 50%, mated from the TPD spectra of NH3. Both the initial rates
while the yields were less than 10%, suggesting that de- were obtained from the slopes of the curves of Figs. 1 and
hydration of IPA to propene (including the polymers) is 2. The acid amount increased nearly linearly with the Al
significant. While H–mordenite, Nafion–SiO2, and H2SO4 content (Fig. 3A). As Fig. 3B shows, the rate of N atom ac-
gave the low yields, the conversions of AN were more than cumulation correlates well with the Al content, that is, the
29%, suggesting the preferential occurrence of the poly-
merization of AN. It should be emphasized that the selec-
14
tivity to PAA was much higher (∼94%) over H–ZSM-5
(Al = 2.63%) (Table 2). The TON, which is defined as the
12
number of PAA molecules produced divided by the num-
ber of acid sites on the surface, reached 60 over H–ZSM-
5 (Al = 2.63%) at 24 h. This shows that the reaction pro-
ceeded catalytically on H–ZSM-5. Furthermore, it is noted
that H–ZSM-5 (Al content = 2.63%) gave the highest TON
among these solid acids.
10
8
6
Table 3 gives the catalytic data on the reaction of AN with
IPA over various H–ZSM-5s having different Al content.
The data were collected at 6 h and 423 K. The yield of
PAA depended greatly on the Al content of H–ZSM-5. The
yield once increased as the Al content increased and then
decreased through a maximum at 2.63% of the Al content.
As shown in Table 3, the remaining amounts of nitrogen
atoms after the reaction (6 h) became larger at higher Al
4
2
0
0
1
2
3
4
5
6
7
Time / h
FIG. 1. Time courses of reaction of acrylonitrile and isopropyl alco-
hol over various H–ZSM-5s. Al content: (ꢀ) 3.30%, (ꢁ) 2.63%, and (ꢂ)
1.00%. Reaction conditions: catalyst, 1 g; temperature, 423 K. Acryloni-
trile, 10 cm3 (150 mmol), isopropyl alcohol, 2.2 cm3 (30 mmol), and hy-
droquinone (0.05 g) as an inhibitor of polymerization and n-dodecane
(0.25 cm3) as an internal standard were mixed.
contents. The largest amount of the N atom (4.8 mmol g−1
)
corresponds to only about 3% of the amount AN presented
initially. On the other hand, the remaining carbon atoms for
all cases were about five times those of the N atoms.

198
CHEN AND OKUHARA
6
5
4
3
2
1
0
70
60
50
40
30
20
10
0
0
1
2
3
4
5
6
7
6
0
1
2
3
4
5
100 Al/(Al+Si) / %
Time / h
FIG. 4. Specific rate as a function of the Al content of H–ZSM-5.
FIG. 2. Time courses of the accumulation of N atoms. The reaction
conditions are the same as in Fig. 1. Al content: (ꢀ) 3.30%, (ꢁ) 2.63%,
and (ꢁ) 1.00%.
lyst without any treatment, the reaction hardly occurred;
the yield was only 3% after 6 h. On the other hand, when
the used catalyst was calcined at 773 K for 5 h in air, the
catalyst became active again. The obtained yield and the se-
lectivity to PAA are plotted against the number of repeated
run. It should be emphasized that the decrease in the yield
with the run number was slight and the selectivity gradually
increased with the run number.
acid amount. On the other hand, the initial rate for the for-
mation of PAN became a near-saturated value or became
lower with a further increase in the Al content.
Figure 4 provides the dependence of the specific rate on
the Al content. It is clear that the specific activity decreased
to about one-sixth as the Al content increased from 0.7 to
9.8%.
Changes in the surface area and pore volume of H–ZSM-
5 (Al content = 2.63%) with the reaction and the calcina-
tion were measured. The total surface area of the H–ZSM-5
dropped greatly, from 410 to 16 m2 g−1, after the first reac-
tion. By the calcination of the used catalyst at 773 K in air,
the surface area almost recovered.
In Fig. 6, infrared spectra of used H–ZSM-5 (Al con
tent = 2.63%) and the polymers of the product and reactant
aregiven, togetherwiththespectrumofH–ZSM-5onwhich
the product PAA was adsorbed. The used H–ZSM-5 which
was dried at 373 K showed several peaks consisting of the
Reusability of H–ZSM-5 (Al content = 2.63%) is given
inFig. 5. Whenthereactionwasrepeatedovertheusedcata-
1.0
A
0.8
0.6
0.4
0.2
0
100
80
60
40
20
0
100
80
60
40
20
0
B
4
24
3
2
1
16
8
0
1
2
3
4
5
6
0
0
0
1
2
3
4
5
Number of run
100 Al/(Al+Si) / %
FIG. 5. Changes in yields and selectivities for the reaction of acryloni-
trile and isopropyl alcohol as a function of the number of repeated runs.
Between the runs, the catalyst was calcined at 773 K in air for 5 h. The
reaction conditions are the same as those in the caption to Fig. 1.
FIG. 3. Effects of Al content of H–ZSM-5 on the acid amount, the
initial rate, and the rate of N atom accumulation. The reaction conditions
are the same as in Fig. 1. The acid amounts were determined by NH3 TPD.

RITTER REACTION OVER H–ZSM-5
199
first group (1460 cm−1, 1544 cm−1 (NH), 1626 cm−1 (C C),
==
2.0
1.5
1.0
0.5
and 1657 cm−1 (C O)), the second characteristic peak at
a
b
==
2244cm−1 (CN), andthethirdgroup(2870cm−1, 2937cm−1
,
and 2970 cm−1), together with the broad peak around
3400 cm−1 due to the OH region. Since the peak of the CN
group in AN appears at 2230 cm−1, the peak at 2244 cm−1
observedforPAN(Fig. 6d)canbedistinguishedfromthatof
AN. The peak at 2244 cm−1 of the used H–ZSM-5 (Fig. 6a)
is thus assigned to the PAN formed on the catalyst. On
the other hand, the peaks of 1544 cm−1 and 2970 cm−1
were not detected for PAN but were found for poly-N-
isopropylacrylamide (PPAA) or PAA adsorbed on H–
ZSM-5 (Figs. 6b and 6c). This shows that PPAA and/or PAA
were present on the surface of H–ZSM-5 after the reaction.
Adsorptions of product molecules were examined on
H–ZSM-5 (Al content = 2.63%). Figure 7 shows the time
courses of the adsorption of PAA at 298 K on H–
ZSM-5 (Al content = 2.63%). When the solution of 1,3,5-
trimethylbenzene containing PAA (0.41 mol dm−3) was
used, about 1.5 mmol g−1 of PAA was rapidly adsorbed
on H–ZSM-5 (Fig. 7a). When AN (5 cm3) was further
added to the 1,3,5-trimethylbenzene solution, most of the
PAA adsorbed in the micropores of H–ZSM-5 was des-
orbed rapidly at 298 K. It was confirmed that the ad-
sorption amount was only 0.2 mmol g−1 on H–ZSM-5
(not shown), when AN was used as the solvent instead of
1,3,5-trimethylbenzene.
0
0
2
4
6
8
10
12
Time / h
FIG. 7. Adsorption and desorption of N-isopropylacrylamide on and
from H–ZSM-5 (Al content = 2.63%). (a) Fresh H–ZSM-5 (0.72 g) was
added to a solution of N-isopropylacrylamide, 2.06 mmol (0.233 g), and
1,3,5-trimethylbenzene (5 cm3) at room temperature. (b) Following (a),
acrylonitrile (5 cm3) was added to the solution.
Influence of the polymer of AN formed on H–ZSM-5 on
the activity and selectivity was examined. After H–ZSM-5
(Al content = 2.63%, 1 g) was treated with AN (150 mmol)
at 423 K for 6 h, the H–ZSM-5 was separated and then
evacuated at 373 K for 3 h. The AN-treated H–ZSM-5 gave
only a 7.1% yield at 6 h, which is one-sixth that of the fresh
H–ZSM-5 (Table 3) under the same reaction conditions.
When excess IPA (10 cm3, 136 mmol) was added to the
H–ZSM-5 (1 g) at 423 K for 6 h in the presence of dode-
cane (0.25 cm3) and hydroquinone (0.05 g), the conversion
of IPA was 62.4%, and the conversion to diisopropyl ether
was 44%. After the used catalyst was separated by cen-
trifugation and dried at 373 K, the elemental analysis of the
used catalyst showed that the amount of remaining carbon
atoms was 5.98 mmol g−1. The IPA-tretated H–ZSM-5 gave
a yield for PAA of 20%, which is about half that of the fresh
catalyst. The surface area and the pore volume decreased
to 40% that of the fresh catalyst.
1544
2937
1626
0.5
2970
1460
1657
2870
2244
a
b
DISCUSSION
c
Catalytic Property of H–ZSM-5
In a commercial process, N-isopropylacrylamide has
beensynthesizedwithexcessH2SO4. AsdescribedintheIn-
troduction, the synthetic process using a solid acid catalyst
is desirable from the standpoint of it being an environmen-
tally benign process. The data in Table 2 demonstrate that
d
4000 3500 3000 2500 2000 1500 1000
Wavenumbers / cm-1
FIG. 6. Infrared spectra of used H–ZSM-5 catalysts. (a) The used H–ZSM-5 exhibited an exceptionally high catalytic perfor-
H–ZSM-5 (Al content = 2.63%) for the reaction of acrylonitrile and iso-
propyl alcohol at 423 K, (b) H–ZSM-5 (Al content = 2.63%) on which
N-isopropylacrylamide was adsorbed, (c) poly-N-isopropylacrylamide
(PPAA), and (d) polyacrylonitrile (PAN). The adsorption of N-
isopropylacrylamide was carried out at room temperature for 5 h in 1,3,5-
trimethylbenzene.
mance for the reaction of AN (acrylonitrile) with isopropyl
alcohol (IPA). This reaction proceeded on acidic sites,
while the reaction mechanism has not been fully elucidated.
Luzgin and Stepanov (18) studied the reaction of acetoni-
trile with tert-butyl alcohol over H–ZSM-5 by means of

200
CHEN AND OKUHARA
solid-state NMR. They claimed that an N-alkylnitrilium and 0.04 of the relative pressure of water was compara-
cation is formed as an intermediate by the attack of a tert- ble to that of H–ZSM-5 (Al content = 2.63%), indicat-
butyl cation on the CN bond on the surface. Thus one can ing similar hydrophobicity. It was already reported by
consider that the acid strength and amount of a catalyst are us that for catalytic synthesis of N-adamantylacrylamide
primarily important for the catalytic performance.
from acrylonitrile and 1-adamantanol, Cs2.5H0.5PW12O40
As shown in Table 3 and Fig. 1, the change in the activity was superior to various solid acids, such as H–ZSM-5,
of H–ZSM-5 with the change in Al content was intriguing. mordenite, HY, and so forth, and to liquid acids such as
Since the acid amount increased monotonically as the Al H2SO4 and H3PW12O40 (7). Furthermore, as described
content increased (Fig. 3A), the highest activity of H–ZSM- above, Cs2.5H0.5PW12O40 was more active than H–ZSM-5
5 with an Al content of 2.63% is not due to the acid amount. for the synthesis of N-tert-butylacrylamide from tert-butyl
The acid strength increases as Al content decreases for H– alcohol. Since Cs2.5H0.5PW12O40 has mesopores (width,
ZSM-5 zeolites (19), while Niwa and Katada (17) showed 40 nm) and at the same time micropores having widths
that the difference in acid strength was not significant for Al larger than 0.75 nm (13), there is little restriction of
content between 4.80 and 2.63%. Indeed, the trend in the diffusion of molecules (or products) into or out of the
change in specific activity (Fig. 4) is consistent with that of pores of Cs2.5H0.5PW12O40. Thus it is reasonable that
acid strength, but the significant change in specific activity Cs2.5H0.5PW12O40 has much higher activity for reactions
would not be only due to that of acid strength.
concerning bulky reactants. The lower activities of H–ZSM-
Ammonia TPD (20, 21) and microcalorimetry of NH3 ad- 5 for the above reactions are due to the limitation of the
sorption (13) revealed that H–ZSM-5 (Al content = 4.8%) diffusion of reactant or product molecules or the limitation
was a weaker acid than Cs2.5H0.5PW12O40. Also it was of the formation of the intermediates in the pore.
shown (17) that H–ZSM-5 with an Al content of 2.63% was
In spite of the lower acid strength of H–ZSM-5 than
a weaker acid than H–mordenite. Since the acid strength of Cs2.5H0.5PW12O40 and their similar hydrophobicity,
of H–ZSM-5 used in the present study is not higher than as described above, H–ZSM-5 was more active than
those of the above-mentioned solid acids, the exception- Cs2.5H0.5PW12O40 (Table 2) for the reaction with IPA. The
ally high activity of H–ZSM-5 (Table 2) suggests that the microporous structure of H–ZSM-5 may be another critical
catalytic performance is not determined only by the acid factor governing specific activcity. At present, we speculate
strength.
that the pore structure of H–ZSM-5 may be favorable for
As an important characteristic of H–ZSM-5, hydropho- this reaction. For example, the reactant molecules are likely
bicity should be considered. Some researchers inferred that to be arranged in positions in the constrained pores of H–
the hydrophobicity of H–ZSM-5 becomes high as the Al ZSM-5 (pore width, 0.54 × 0.56 nm) suitable for accelerat-
content decreases (22, 23). Therefore the specific activity ing the reaction.
of H–ZSM-5 might be affected by the hydrophobic circum-
stance of the pores as well as by the acid strength (Fig. 4).
Since isopropyl alcohol is more basic than acrylonitrile, iso-
Catalyst Deactivation and Reusability
propyl alcohol interacts more strongly with the acid sites of
Although the reaction of AN and IPA proceeded cata-
H–ZSM-5, whichbringsabouttheinhibitionofthereaction. lytically on H–ZSM-5, the deactivation of H–ZSM-5 was se-
However, the adsorption of isopropyl alcohol would be ef- vere. The reaction rate became lower at the latter stage of
fectively suppressed due to the increase in the hydropho- the reaction (Fig. 1). The extent of the deactivation seems to
bicity as the Al content decreased. By this the reaction may be nearly independent of the Al content. Plausible reasons
proceed smoothly due to the weakening of the inhibition for the deactivation are blockages of the pores by the poly-
effect of isopropyl alcohol. One can consider that water mers formed from the reactant AN or product PAA and/or
formed during the reaction poisons the catalyst. It is thus by the irreversible adsorption method of product PAA. As
expected that the high hydrophobicity of the H–ZSM-5 described above, the considerable blockage of the pores
with the low Al contents is also favorable to this water- was revealed by the adsorption of N2.
participating reaction. As another possible factor influenc-
As can be seen in Fig. 7, the product PAA can readily
ing the activity and selectivity, the crystallite size should be enter into the pores of H–ZSM-5 even at 298 K. Thus the
considered. However, the effects of the crystallite size of possibility of the blockage by the adsorption of PAA can
H–ZSM-5 with the same Al content were not examined in be excluded. Elemental analysis confirmed that the carbon
the present study; these still remain to be solved.
and nitrogen atoms accumulated appreciably on H–ZSM-
It should be emphasized that the value of TON ob- 5s (Table 3). IR data indicated that PAN was present on
tained for H–ZSM-5 (Al content = 2.63%) was higher the used H–ZSM-5 as well as the polymer of PAA (PPAA)
than that of Cs2.5H0.5PW12O40. Recently, Cs2.5H0.5PW12O40 (Fig. 6). Since the activity of H–ZSM-5 became about one-
was reported to be highly hydrophobic (24). The den- sixth after treatment with only AN, the main species re-
sity of water adsorbed on Cs2.5H0.5PW12O40 at 298 K sponsible for the deactivation is probably PAN. The cross

RITTER REACTION OVER H–ZSM-5
201
section of AN is estimated to be 2.5 × 10−19 m2 from the vative Technology for the Earth, Japan, and by a Grant-in-Aid for Scien-
tific Research from the Ministry of Education, Science, Sports and Culture,
Japan.
molecular weight and the density in liquid state (25). As-
suming that one nitrogen atom corresponds to one unit of
AN monomer in the PAN on the surface, the surface area
of the PAN formed corresponds to about 20 times that of
REFERENCES
the external surface area of H–ZSM-5. Thus the blockage
1. Coover, H. W., Newton, H., and Shearer, N. H., U.S. Patent 2,719,176
of the pores of the PAN is plausible.
(1955).
2. Ritter, J. J., and Minieri, P. P., J. Am. Chem. Soc. 70, 4050 (1948).
3. Firouzabadi, H., Sardarian, A. R., and Badparva, H., Synth. Commun.
24, 601 (1994).
4. Martinez, A. G., Alvarez, R. M., Vilar, E. T., Fraile, A. G., Hanack, M.,
and Subramanian, L. R., Tetrahedron Lett. 30, 581 (1989).
5. Ooi, S., Nomura, M., Matsushita, K., and Inoue, Y., Nippon Kagaku
Kaishi 1100 (1993).
6. Olah, G. A., Yamato, T., Tyer, P. S., Trivedi, N. J., Singh, B. P., and
Parakash, G. K. S., Mater. Chem. Phys. 17, 21 (1987).
7. Matsuda, H., and Okuhara, T., Catal. Lett. 56, 241 (1998).
8. Burrington, J. D., Rhubringht, D. C., and Ramey, C. E., U.S. Patent
5,712,413 (1998).
9. Yamamoto, K., Japanese Patent (Tokkaishou) 50-135018 (1975).
10. Onuma, K., Kaneko, T., and Nakanishi, A., Japanese Patent
(Tokkaishou) 62-201853 (1987).
As described above, the carbon atom detected on the H–
ZSM-5 was about five times that of the N atom (Table 3).
The carbon deposition might be brought about from PAN
(three times that of N atoms = about 5.1 mmol g−1) and
from IPA (6.0 mmol g−1).
The reusability of deactivated H–ZSM-5 was confirmed
(Fig. 5). By calcination at 773 K, the activity of used H–
ZSM-5 was recovered mostly, and at the same time the pore
volume was also restored. These results also support the
idea that the deactivation of the catalyst would be caused
by the blockage of the pores. However, this regeneration
requires high temperature; in other words, this treatment is
costly. An improved process for suppressing deactivation is
necessary for this catalytic system.
11. Chen, X., Matsuda, H., and Okuhara, T., Chem. Lett. 709
(1999).
12. Hino, M., and Arata, K., React. Kinet. Catal. Lett. 19, 101 (1982).
13. Okuhara, T., Nishimura, T., and Misono, M., in “Proceedings, 11th In-
ternation Congress on Catalysis, Baltimore, 1996” (J. W. Hightower,
W. N. Delgass, E. Iglesia, and A. T. Bell, Eds.), p. 581. Elsevier,
Amsterdam, 1996.
14. Harmer, M. A., Farneth, W. E., and Sun, Q., J. Am. Chem. Soc. 118,
7708 (1996).
15. Okuyama, K., Chen, X., Takata, K., Odawara, D., Suzuki, T.,
Nakata, S., and Okuhara, T., Appl. Catal. 190, 253 (2000).
16. Li, L., Yoshinaga, Y., and Okuhara, T., Phys. Chem. Chem. Phys. 1,
4913 (1999).
17. Niwa, M., and Katada, N., Catal. Surv. Jpn. 1, 215 (1997).
18. Luzgin, M. V., and Stepanov, A., Mendeleev Commun. 238 (1996).
19. Gates, B. C., “Catalytic Chemistry.” Wiley, New York, 1992.
20. Nishimura, T., Okuhara, T., and Misono, M., Appl. Catal. 73, L7
(1991).
21. Okuhara, T., Nishimura, T., Watanabe, H., and Misono, M., J. Mol.
Catal. 74, 247 (1992).
22. Olson, D. H., Haag, W. O., and Logo, R. M., J. Catal. 61, 390
(1980).
CONCLUSION
We found that H–ZSM-5 was exceptionally effective for
the N-alklyaltion of acrylonitrile with isopropyl alcohol to
N-isopropylacrylamide in a solid–liquid reaction system.
The microporous structure of H–ZSM-5 may be critical for
this reaction. The Al content affected product yield and the
catalyst having the Al content of 2.63% was most active
among H–ZSM-5s. This unique activity pattern is due to
the hydrophobicity as well as the acid strength. The cata-
lyst deactivation might be caused mainly by the covering
of the surface of the catalyst with polyacrylonitrile formed
during the reaction and blockage of the pores. It should be
emphasized that the catalytic activity was almost recovered
by calcination of the catalyst in air.
23. Nakamoto, H., and Takahashi, H., Zeolites, 2, 67 (1982).
24. Yamada, T., and Okuhara, T., Langmuir 16, 2321 (2000).
25. Emmett, P. H., and Brunauer, S., J. Am. Chem. Soc. 59, 1553
(1937).
ACKNOWLEDGMENTS
The authors thank Mr. Hideho Matsuda for help with product analy-
sis. This study was partly supported by the Research Institute of Inno-