Acylation of 2-methoxynaphthalene and isobutylbenzene over zeolite beta

Article abstract of DOI:10.1006/jcat.2000.2855

The acylation of 2-methoxynaphthalene (2MN) and isobutylbenzene using several zeolite beta samples having varius Si/Al ratios and crystal sizes to examine whether external surface sites could be eliminated to enhance the catalyst performance to obtain a viable acylation catalyst for the formation of precursors to the nonsteroidal anti-inflammatory agents naproxen and ibuprofen. Zeolite beta was active for the acylation of 2MN but was not selective to the desired product, 2-acetyl-6-methoxynaphthalene (2,6-AMN). Mild operating conditions (temperature and acylating agent concentration) could be used to obtain reasonable conversions and to limit catalyst deactivation. The other key product, 1-acetyl-2-methoxynaphthalene (1,2-AMN) formed on the external surface of the zeolite, while 2,6-AMN occurred in the zeolite pores. Thus, the selectivity to 2,6-AMN was enhanced on zeolite beta samples having a larger crystal size, on which most of the acid sites could be passivated by coating the crystals with a layer of amorphous silica. The amount of surface coating on the large crystals determined the yield of and the selectivity to 2,6-AMN. Isobutylbenzene was less reactive than 2MN but the desired product, 4-isobutylacetophenone, was always obtained since the isobutyl group provides for the para position being the preferred sites for acylation. For isobutylbenzene, the zeolite external surface contributed significanlty to the formation of 4-isobutylacetophenone. Zeolite beta with a small crystal size was, thus, the most favored catalyst for this reaction.

Full text of DOI:10.1006/jcat.2000.2855

Journal of Catalysis 192, 215–223 (2000)
doi:10.1006/jcat.2855, available online at http://www.idealibrary.com on
Acylation of 2-Methoxynaphthalene and Isobutylbenzene
over Zeolite Beta
1
2
P. Andy, J. Garcia-Martinez, G. Lee, H. Gonzalez, C. W. Jones, and M. E. Davis
Chemical Engineering, California Institute of Technology, Pasadena, California 91125
Received November 15, 1999; revised February 18, 2000; accepted February 18, 2000
The acylation of 2-methoxynaphthalene (2MN) by alu-
The acylation of 2-methoxynaphthalene (2MN) with acetic an- minum chloride was a step in the first large-scale synthe-
hydride is carried out in the liquid phase in the presence of zeo- sis of naproxen (2), and the acylation of isobutylbenzene
lite beta. Two main products, 1-acetyl-2-methoxynaphthalene (1,2-
AMN) and the desired 2-acetyl-6-methoxynaphthalene (2,6-AMN)
are obtained. Variations of reaction temperature, the concentration
of the acylating agent, and the weight of the catalyst are investigated
to find reaction conditions sufficient to provide for good conversion
Zeolites, with their shape-selective properties and good
of 2MN and limited deactivation. Surface-poisoning experiments
show that the formation of 1,2-AMN occurs on the external surface
of the zeolite. When the external surface of zeolite beta is passivated
by coating with amorphous silica, a significant increase in the se-
with hydrogen fluoride is currently used in the synthesis
of ibuprofen (3). Because of current environmental restric-
tions, replacement of conventional homogeneous catalysts
with solid acid catalysts has great industrial importance.
regenerability, have been found to be viable alternatives to
liquid acidsin numerouscases(4–17). The first industrialap-
plication of a zeolite-promoted acylation for the production
lectivity of 2,6-AMN is observed and this result is a clear example of an aromatic ketone was recently reported by researchers
of shape-selective acylation with zeolite catalysts. The acylation of at Rh oˆ ne-Poulenc (11).
isobutylbenzene with acetic anhydride is investigated also. At con-
versions of around 10%, the selectivity to 4-isobutylacetophenone
is excellent (>97%). For the acylation of isobutylbenzene, the exter-
nal surface of the zeolite contributes significantly to the formation
of 4-isobutylacetophenone, so the shape-selective properties of the
The acetylation of 2MN has been investigated over
MCM-41 (18), HY, ZSM-12 and BEA (4, 16), and ZSM-5
and mordenite (5). Two products are usually obtained,
-acetyl-2-methoxynaphthalene (1,2-AMN) and 2-acetyl-
-methoxynaphthalene (2,6-AMN), with the undesired
,2-AMN product generally predominating. Indeed, the
acylation generally occurs at the kinetically favored 1-
position. However, the deacylation of the acyl group has
been observed to give back 2MN (4). These data are con-
sistent with the following reaction scheme:
�
1
6
1
zeolite are not utilized.
�c 2000 Academic Press
Key Words: shape-selective acylation; 2-methoxynaphthalene;
naproxen; isobutylbenzene; ibuprofen; zeolites beta.
1
. INTRODUCTION
Friedel Craft acylation is a useful step for the synthesis
of pharmaceutical compounds such as (S)-naproxen and
ibuprofen (1).
The results from all previous investigations suggest that ze-
olite beta may be an interesting catalyst for the production
of 2,6-AMN since it was speculated that its formation is
1
Present address: Departamento de Quimica, Inorganica, Universidad
de Alicante, Alicante, 03080, Spain.
2
To whom correspondence should be addressed. Chemical Engineer-
�
favored inside the pores of BEA (4). Moreover, molecu-
ing 210-41, California Institute of Technology, Pasadena, CA 91125. Fax:
(
626) 568-8743. E-mail: mdavis@cheme.caltech.edu.
lar modeling has been investigated over mordenite, L, and
215
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

2
16
ANDY ET AL.
�
BEA (19). It has been shown that the energy barrier for in 3 ml of hexane to which 0.1 ml of TEOS was added.
diffusion of 2,6-AMN is smaller than those of the other iso- Stirring was continued for 3 h, 1 h, or 30 min, after which the
mers in the case of zeolite beta, which is in agreement with zeolite was recovered by filtration and washed with hexane.
observed catalytic properties (4). However, due to the con- This method has been shown to coat the surface of zeolite
tribution of acid sites on the external surface of the crystals, beta with silica (23). The samples are identified by the letter
a large amount of 1,2-AMN product is formed (16). In fact, “p” and followed by the time of coating, e.g., BEA1p(3h)
this reaction has been used as a model reaction to charac- denotes the BEA1 sample that was contacted with TEOS
terize the external surface of zeolite beta crystals (20). In for 3 h.
contrast, the acylation of isobutylbenzene over zeolite beta
has been reported by researchers at Rh oˆ ne-Poulenc to be 2.3. Characterization
selective in the production of 4-isobutylacetophenone, the
desired product (21).
X-ray powder diffraction (XRD) patterns were collected
on a Scintag XDS 2000 diffractometer using CuK� radia-
Here, the acylation of 2MN and isobutylbenzene with ze-
tion. Nitrogen adsorption isotherms were obtained at 77 K
olite beta is investigated. The goal of this work is to test
using an Omnisorp 100 sorption apparatus. Adsorption ca-
whether the external surface sites can be eliminated to
pacities of vapor-phase cyclohexane were obtained on a
improve the performance of the catalyst to obtain a vi-
McBain-Bakr balance. The micropore volume was deter-
able acylation catalyst for the formation of precursors to
mined at P/P = 0.25. Prior to all adsorption experiments,
0
the nonsteroidal anti-inflammatory agents naproxen and
ibuprofen.
�
the samples were degassed under vacuum at 250 C for 2 h.
Scanning electron micrographs (SEM) were recorded on a
Camscan Series 2-LV scanning electron microscope. X-ray
2
. EXPERIMENTAL SECTION
photoelectron spectroscopy (XPS) experiments were con-
ducted on a Kratos AXIS-HS spectrometer. The X-ray
source was monochromatized AlK� at 1478 eV. Elemental
2
.1. Materials
Numerous samples of zeolite beta with different Si/Al analyses were performed by Galbraith Laboratories Inc.,
ratios and crystal sizes were used in this study for the acy- Knoxville, TN. The increases in weight of the used catalysts
lation of 2MN and isobutylbenzene. The sample denoted were obtained on a Dupont Instruments 951 thermogravi-
BEA1 was obtained from PQ and was ion exchanged with metric analyzer (TGA). The samples were heated in air to
�
�
�
a 1.0 M NH4Cl solution for 24 h at � 80 C (repeated three 800 C and maintained for 1 h at 800 C. The extraction of
�
times) with subsequent calcination to 550 C in air to gen- the carbonaceous compounds deposited in the zeolite was
erate the proton form. BEA2 and BEA3 were synthesized carried out according to the procedure described by
in fluoride media by modifying the method reported by Guisnet and Magnoux (24).
Camblor et al. (22). Tetraethylammonium fluoride (12.13 g,
8
1.3 mmol, Aldrich) and aluminum nitrate nanohydrate 2.4. Catalyst Testing
(
(
0.92 g, 2.45 mmol, Aldrich) were dissolved in water
13.58 g, 75.44 mmol). To this solution was added tetraethy-
All chemicals were purchased from Aldrich Chemical
Co. Only the acetic anhydride was purified (reflux in the
presence of magnesium for 5 days and then distilled) before
use.
Acylation reactions were carried out in a sealed batch
reactor under argon equipped with a magnetic stirrer
lorthosilicate (TEOS) (25.45 g, 122.2 mmol, Aldrich) drop-
wise with stirring. The solution was allowed to stir for
2
removed in vacuo. Additional water was added and the
evaporation procedure was repeated. The gel obtained and
4 h, after which the alcohol generated by hydrolysis was
(
600 rpm) and heated under autogenous pressure in an oil
5
0 mg of zeolite beta seed crystals were transferred to a
�
bath at 100 C with acetic anhydride as the acylating reagent
Teflon-lined autoclave and heated under autogenous pres-
sure at 140 C for 7 days in rotary mode (BEA2) or in static
mode (BEA3). The samples were recovered by filtration,
washed with water, and dried in air overnight at 100 C. The
samples were calcined in air using the following tempera-
ture program: heat to 175 C at 1 C/min and hold at 175 C
for 2 h; heat to 550 C at 1 C/min and hold at 550 C for 6 h.
and 1,2-dichloroethane as the solvent. Prior to use, 0.10 g or
�
�
0
.25 g of catalyst was calcined in situ in air to 550 C. 2MN or
isobutylbenzene (1.1 mmol), acetic anhydride (1.1 mmol),
and an internal standard (1,3,5-tri-tert-butylbenzene or do-
decane) were dissolved in 0.860 ml of solvent and added to
the vessel. Aliquots of the reaction mixture were sampled
and analyzed by gas chromatography equipped with a FID
detector (HP 5890 II) on a 30-m HP-5 column.
�
�
�
�
�
�
�
2
.2. Postsynthetic Modifications
The products were identified by comparing their re-
Modifications of the zeolite beta crystals were performed tention time to commercial (2,6-AMN) or synthesized
using the following method: 0.25 g of HBEA (BEA1, (1,2-AMN) compounds or by GC-MS analysis (other acy-
BEA2, or BEA3) were stirred under an argon atmosphere lated compounds, diacylated compounds). The error in

ACYLATION OVER ZEOLITE BETA
217
the experimental results was found to be in a range of
TABLE 2
�
3–4% .
Micropore Volume of Zeolite Beta Samples
At the end of the reaction, the sample was recovered
3
by filtration, washed with dichloromethane, and dried at
Micropore volume (cm /g)
room temperature, in vacuo. Then, the sample was regen-
erated under dried air to 550 C or characterized bydifferent
methods.
The external surface activity of the zeolite was mea-
sured using the conversion of the bulky allyl 3-5-di-tert-
butylphenyl ether (23, 25). This ether is too large to en-
ter the pores of zeolite beta and is transformed into
Sample
BEA1
BEA2
BEA3
BEA1p(3h)
BEA2p(3h)
BEA3p(3h)
Nitrogen adsorption
Cyclohexane adsorption
�
0.24
0.19
0.19
0.20
0.18
0.18
0.20
0.21
0.22
0.22
0.21
0.22
2
-allylphenol that is then cyclized into 2-methyldihydro-
benzofuran:
tion isotherm of small zeolite beta crystals it is impossible to
know where the onset of extracrystalline adsorption occurs.
Thus, for the case of BEA1, it is likely the volume deter-
mined with nitrogen is slightly overestimated as it also takes
into account a part of the extracrystalline volume of this ze-
olite. Thus, the cyclohexane adsorption values are the most
reliable and show that there are no decreases in the intraze-
olitic pore spaces of the passivated zeolites as they have the
same micropore volumes as their parent zeolites.
Figure 1 shows the total conversion of allyl 3-5-di-tert-
butylphenyl ether over several zeolite beta samples. As
previously shown (23, 25), the conversion of the probe
molecule is significantly higher on zeolite BEA1 with small
crystals than on BEA2 and BEA3 (BEA2 and BEA3 have
larger crystals and thus less acid sites on the external sur-
face than BEA1). Moreover, the conversion is lower on the
passivated samples BEA2p(3h) and BEA3p(1h) than on
the parent zeolites BEA2 and BEA3. These results show
that the coating of zeolite with a medium crystal size and in
particular a large crystal size is very efficient. Almost all of
the acid sites on the external surface are passivated by the
amorphous silica layer.
The reaction was carried out in a batch reactor under
argon at 100 C with 0.10 g of activated zeolite, 0.4 mmol
of allyl 3,5-di-tert-butylphenyl ether, and 2 ml of 1,2-
dichloroethane. Aliquots of the reaction mixture were sam-
pled and analyzed by gas chromatography on a 30-m HP-5
column.
�
3
. RESULTS AND DISCUSSION
3.1. Characterization of Zeolite Beta Samples
Table 1 lists the Si/Al ratio and the crystal size of the sam-
ples BEA1, BEA2, and BEA3. The materials synthesized in
fluoride media, BEA2 and BEA3, have larger crystal sizes
than the commercial sample. BEA2, obtained under stirred
conditions, has a smaller crystal size than BEA3, which was
prepared at static conditions. All of these samples have an
�
XRD pattern typical of highly crystalline BEA.
The micropore volume of the parent zeolites BEA1,
BEA2, and BEA3 and the passivated samples BEA1p(3h),
BEA2p(3h), and BEA3p(3h) are given in Table 2. Except
for BEA1, the micropore volumes determined with nitro-
gen and cyclohexane are similar. From the nitrogen adsorp-
TABLE 1
Properties of Zeolite Beta Samples
Sample
Si/Ala
Crystal size (�m)b
BEA1
BEA2
BEA3
12
40
40
<1
2.5
9
FIG. 1. Conversion of allyl 3-5-di-tert-butylphenyl ether versus reac-
tion time over zeolite BEA1, BEA2, BEA3, BEA2p(3h), and BEA3p(1h)
a
�
As determined by elemental analysis.
As determined by SEM.
(0.10 g of catalyst, 100 C, 0.4 mmol of allyl 3-5-di-tert-butylphenyl ether,
b
2 ml of 1,2-dichloroethane).

2
18
ANDY ET AL.
�
FIG. 2. Conversion of 2MN (a) and selectivity of acylated products (b) versus reaction time over zeolite BEA1 (0.10 g of catalyst, 100 C, 2MN : acetic
anhydride, 1 : 1 and 1.1 mmol, 0.860 ml of solvent).
3
.2. Acylation of 2MN with Acetic Anhydride
residue contents. With the initial rate data from Fig. 4, the
apparent activation energies are estimated to be 21 kJ/mol
for 1,2-AMN and 45 kJ/mol for 2,6-AMN.
over Zeolite Beta
3
.2.1. Investigation of Reaction Conditions
for the Formation of 2,6-AMN
Effect of acetic anhydride (Ac2O) to 2MN molar ratio.
The ratio of acetic anhydride to 2MN was varied while a
constant amount of2MN and totalreaction solution volume
was maintained. At early reaction times, the concentration
of the acylating agent has practically no influence on the
conversion of 1,2-AMN and 2,6-AMN. Thus, the reaction
order is close to zero order. At longer reaction time, the ef-
fect of the ratio of acetic anhydride to 2MN is different for
the formation of 1,2-AMN and 2,6-AMN (Fig. 5). At longer
reaction times, the kinetic product (1,2-AMN) is favored
when the ratio of Ac2O/2MN increases and the conversion
to 2,6-AMN is a maximum for a ratio equal to 2. More-
over, it must be noted that in the presence of excess acy-
lating agent (2MN/acetic anhydride > 1), small amounts of
Effect of reaction time. Figure 2a shows the total con-
version of 2MN and the yield of reaction products versus re-
action time over zeolite BEA1. 2MN is mainly transformed
into 1,2-AMN and 2,6-AMN. Small amounts of the other
monoacylated methoxynaphthalene isomers are produced.
As the reaction progresses, traces of diacylated compounds
are formed. The conversion of 2MN is 39% at 2 h and then
slowly levels off to a final value of 62% at 24 h. Initially,
,2-AMN is formed with a slight selectivity over 2,6-AMN.
With longer reaction times, the opposite selectivity is ob-
served (Fig. 2b).
1
Effect of catalyst weight. The reaction was conducted at
four different weights of catalyst from 30 to 140 mg. The
increases in conversion of 2MN and yields of 1,2-AMN and
2
,6-AMN (at 4 min of reaction) are linear when the weight is
increased from 30 to 100 mg (Fig. 3). Above this weight, the
conversions no longer increase with catalyst mass (Fig. 3).
Also, the initial selectivity for 2,6-AMN is higher when the
zeolite weight decreases. However, after 24 h of reaction,
the selectivity is independent of the weight of the catalyst.
Effect of reaction temperature. The reaction was inves-
�
tigated at 70, 100, and 120 C. The yield (Fig. 4a) and the
selectivity of 2,6-AMN increase significantly when the tem-
perature increases. The yield of 1,2-AMN increases when
�
the temperature is elevated from 70 to 100 C. However, at
1
�
20 C, a decrease in the yield of 1,2-AMN is observed with
reaction time and may be due to the protiodeacylation of
the acyl group (4) (Fig. 4b). The increase in the temperature
leads also to catalyst samples becoming darker in color af-
FIG. 3. Total conversion, yield of 1,2-AMN and yield of 2,6-AMN (at
min of reaction) versus weight of zeolite BEA1 (100 C, 2MN : acetic
�
4
ter reaction and likely indicates increasing carbonaceous anhydride, 1 : 1 and 1.1 mmol, 0.860 ml of solvent).

ACYLATION OVER ZEOLITE BETA
219
FIG. 4. Yield of 2,6-AMN (a) and 1,2-AMN (b) versus reaction time over zeolite BEA1 as a function of temperature (0.10 g of catalyst, 2MN : acetic
anhydride, 1 : 1 and 1.1 mmol, 0.860 ml of solvent).
diacylated compounds are formed and the catalyst rapidly reaction is higher in the presence of acetic acid (10.3% in-
turns into a dark solid that is indicative of a large formation stead of 7.1% ). This could be due to a strong adsorption of
of heavy products (the increase in catalyst weight after 24 h acetic acid on the internal and external sites of the zeolite.
of reaction is 11.7% as opposed to 7.6% at a ratio of 1).
Conversion of 1,2-AMN and 2,6-AMN. Deacylation ex-
Additionally, dealumination of the zeolite is observed. The
periments with 1,2-AMN and 2,6-AMN were carried out
Si/Al ratio of BEA1 is 12, and after 24 h of reaction, the
Si/Al ratio reaches 14.2 and 16.3 for a ratio of acetic anhy-
acid (AMN/acetic acid = 1) (Table 3). The deacylation of
dride to 2MN equal to 1 and 4, respectively. The increase
in dealumination with the acetic anhydride/2MN ratio is
likely due to the increased amounts of acetic acid formed
from elevated conversions (acetic acid is known to dealu-
minate zeolite beta (26)).
�
for 2 h at 100 C in dichloroethane in the presence of acetic
1
,2-AMN leads to the formation of 2MN (19.2% ) and 2,6-
AMN (0.6% ). However, the protiodeacylation of the acyl
group at the 1-position is limited in the presence of acetic
anhydride (1,2-AMN/acetic acid/acetic anhydride = 1). The
deacylation reaction essentially does not occur with the
Effect of presence of acetic acid. The addition of acetic thermodynamically more stable 2,6-AMN (Table 3).
Deacylation experiments with 1,2-AMN and 2,6-AMN
acid (2MN/acetic anhydride/acetic acid = 1) leads to a sig-
nificant decrease in the yield of 1,2-AMN and 2,6-AMN in the absence of the acylating agent were also carried out
�
(
Fig. 6) but a small increase in the selectivity of 2,6-AMN. at 100 C in dichloroethane. After 2 h, no reaction of 1,2-
Moreover, the increase in weight of the catalyst after 24 h of AMN is observed. However, after 24 h, 44% of 1,2-AMN
FIG. 5. Yield of 1,2-AMN (a) and 2,6-AMN (b) versus reaction time over zeolite BEA1 as a function of acetic anhydride to 2MN molar ratio
�
(
0.10 g of catalyst, 100 C, 2MN = 1.1 mmol).

2
20
ANDY ET AL.
�
FIG. 6. Yield of 1,2-AMN (a) and 2,6-AMN (b) versus reaction time over zeolite BEA1. Without acetic acid, 0.10 g of catalyst, 100 C, 2MN : acetic
�
anhydride, 1 : 1 and 1.1 mmol, 0.860 ml of solvent. With acetic acid, 0.10 g of catalyst, 100 C, 2MN : acetic anhydride : acetic acid, 1 :1 :1 and 1.1 mmol,
0
.860 ml of solvent.
is transformed and gives 20.8% of 2MN and 19.2% of 2,6- 3.2.2. Contribution of the External Surface
AMN whereas no transformation of 2,6-AMN is observed
at the same conditions.
These results are consistent with the reaction network
proposed by various authors (4, 18). That is, 1,2-AMN is
deacylated to give back 2MN that can then be acylated to
form 2,6-AMN.
External surface-poisoning experiments. The transfor-
mation of 2MN was carried out in the presence of 2,4,6-tri-
tert-butylpyridine (5 mol% of 2MN) or triphenylphosphine
5 wt% of catalyst) for 2 h with BEA1. These molecules are
(
too large to enter the zeolite pores and can be adsorbed on
the external acid sites. In the absence of poison, the total
conversion is 39.2% and the selectivity to 2,6-AMN is 48% .
In the presence of poison, a significant decrease in the total
conversion occurs; e.g., the conversion is 7.5% in the pres-
ence of 2,4,6-tri-tert-butylpyridine and an increase in the
selectivity of 2,6-AMN is observed, e.g., 75% in the pres-
ence of 2,4,6-tri-tert-butylpyridine. These data suggest that
a large amount of 1,2-AMN is formed on the external sur-
face and that the selective acylation of 2MN into 2,6-AMN
occurs inside the zeolite pores.
Composition of the organic compounds retained on the
zeolite. The composition of carbonaceous residue from
BEA1 was determined from a 400-mg sample exposed to
�
the reaction at 100 C for 24 h. A large part of the retained
compounds are extracted in methylene chloride by simple
soxhlet treatment of the sample (5.6 wt% of the catalyst).
The extract contains mainly 2,6-AMN and diacylated prod-
ucts that are not observed in the reaction solution (Table 4).
Moreover, the 2,6-AMN-to-1,2-AMN ratio is higher in the
extract (2.3) than in the reaction solution (1.5).
Effect of crystal size. Three zeolite beta samples with
A less significant part of the retained compounds is not
collected by extraction and can be recovered in a hydroflu-
oric acid solution that dissolves the zeolite. More than 95%
of the compounds are diacylated molecules (Table 4) that
may be formed by the acylation of 2,6-AMN.
varying crystal size were investigated for the acylation of
2
MN to test for the effects of reaction at the external sur-
face: BEA1 with a small crystal size, BEA2 with an inter-
mediate crystal size, and BEA3 with a large crystal size.
TABLE 4
TABLE 3
Composition of the Organic Compounds Retained
on the Zeolite BEA1
Conversion of 1,2-AMN and 2,6-AMN over Zeolite BEA1
Reactant
Conv. (% )
2MN (% )
1,2-AMN (% )
2,6-AMN (% )
Distribution (wt% )
1
,2-AMNa
19.8
5.1
0.3
19.2
4.2
0.2
—
—
0.1
0.6
0.5
—
wt% a
2MN
1,2-AMN
2,6-AMN
others AMN
di-AMN
b
1
,2-AMN
a
b
2
,6-AMN
5.6
0
9
—
18
—
41
trace
4
—
28
>95
c
.7
a
�
.10 g of catalyst, 100 C, 2 h, 1,2-AMN(or 2,6-AMN) : acetic acid, 1 : 1
0
a
and 1.1 mmol, 0.860 ml of solvent.
Weight percentage of extract on the catalyst.
Before extraction.
After extraction.
b
�
b
c
0
.10 g of catalyst, 100 C, 2 h, 1,2-AMN : acetic anhydride : acetic acid,
1
:1 :1 and 1.1 mmol, 0.860 ml of solvent.

ACYLATION OVER ZEOLITE BETA
221
�
FIG. 7. Yield of 1,2-AMN (a) and 2,6-AMN (b) versus reaction time over zeolite BEA1, BEA2, and BEA3 (0.10 g of catalyst, 100 C, 2MN : acetic
anhydride, 1 : 1 and 1.1 mmol, 0.860 ml of solvent).
BEA1 has a Si/Al ratio equal to 12 while BEA2 and BEA3 formation are observed with BEA2p(3h) and, in particular,
both have Si/Al = 40 (Table 1).
BEA3p(3h). Since the cyclohexane adsorption experiments
The total conversion of 2MN is nearly the same using show that the micropore volume is the same for BEA2,
BEA1, BEA2, and BEA3 (data not shown) and suggests BEA2p(3h), BEA3, and BEA3p(3h), it is likely that the
that the reaction occurs close to the external surface of decrease in the rate of 2,6-AMN formation is due to the
the zeolite. The yield of 1,2-AMN is higher for the zeolite additional diffusional barrier imposed by the surface layer
with the lower Si/Al ratio and the smaller crystals (Fig. 7a) of amorphous silica.
while the yield of 2,6-AMN is higher on the zeolites with a
To test further the effects of surface passivation, the time
larger crystal size (BEA2 and BEA3, Fig. 7b). These results of coating of the largest crystals (BEA3) was decreased
show that the shape-selective acylation at the 6-position from 3 to 1 h (BEA3p(1h)) and to 30 min (BEA3p(30min)).
occurs preferentially in the interior of the crystals as the The conversions to 1,2-AMN and 2,6-AMN on BEA3p(3h),
relative number of external acid sites are decreased with BEA3p(1h), and BEA3p(30min) are compared to those
increasing crystal size. To test further this premise, attempts from BEA3 in Fig. 8. The yield of 1,2-AMN is dramati-
to passivate the external crystal surfaces were explored.
cally suppressed on the passivated zeolites and, in particu-
lar, on BEA3p(3h) with the largest external surface cover-
ing. However, the yield of 2,6-AMN is significantly higher
for BEA3p(30min) and BEA3p(1h) than for BEA3p(3h)
Passivation of external surface. The activity and selec-
tivity of the passivated zeolite beta samples BEA1p(3h),
BEA2p(3h), and BEA3p(3h) are compared to those of the
parent zeolites BEA1, BEA2, and BEA3 and the data are
given in Table 5. Compared to the parent zeolites, the pas-
sivation of the external surface by coating the crystal with
amorphous silica leads, after 2 h of reaction, to a decrease in
the total conversion but an increase in the selectivity of 2,6-
AMN (Table 5). This phenomena is more pronounced as
the size of the crystal increases. Results from passivated and
nonpassivated BEA2 and BEA3 at similar conversions are
provided to compare selectivities. At near isoconversion,
the yield of 2,6-AMN is higher on BEA3p(3h) (7.9% ) than
on BEA3 (3.4% ) and the yield of 1,2-AMN is dramatically
suppressed on the passivated sample (Table 5).
TABLE 5
Acylation of 2MN Catalyzed by Zeolite Beta Samples (The Effect
of Surface Modification on Reaction Behavior Are Given; 0.10 g of
�
Catalyst, 100 C, 2 h, 2MN : Acetic Anhydride, 1 : 1 and 1.1 mmol,
0
.860 ml of Solvent)
Yield of
Yield of
Yield of
Selectivity
Conv. 1,2-AMN 2,6-AMN other AMN of 2,6-AMN
Sample
BEA1
(% )
39.2
(% )
(% )
(% )
(% )
18.5
15.7
18.8
19.2
1.9
1.9
48
52
BEA1p(3h) 36.8
BEA2
46.8
30.9
17
12.2
5.7
27.4
17.4
16.8
2.1
1.2
1.1
58
56
71
a
Kunkeler et al. (23) have shown that the coverage of the BEA2
external surface by amorphous silica is significantly bet- BEA2p(3h) 23.6
ter on larger crystals and the reaction data shown in Fig. 1 BEA3
35.5
6.3
8.5
12.1
2.6
0.4
21.5
3.4
7.9
1.7
0.2
0.2
60
55
92
b
BEA3
confirm these results. Moreover, XPS analysis shows that
BEA3p(3h)
the Si/Al ratio of BEA3 increases from 37 to 62 after the
passivation treatment. Thus, it is clear that 1,2-AMN is
formed on the external acid sites. Low rates of 2,6-AMN
a
At 30 min of reaction.
At 4 min of reaction.
b

2
22
ANDY ET AL.
FIG. 8. Yield of 1,2-AMN (a) and 2,6-AMN (b) versus reaction time over zeolite BEA3, BEA3p(30 min), BEA3p(1 h), and BEA3p(3 h) (0.10 g
�
of catalyst, 100 C, 2MN : acetic anhydride, 1 : 1 and 1.1 mmol, 0.860 ml of solvent).
and is consistent with enhanced diffusional limitations as AMN is similar on the fresh zeolite and on the regenerated
the thickness of the amorphous layer is increased.
samples. However, an increase in the selectivityto 1,2-AMN
The regeneration of BEA3p(3h) was investigated. As can occurs when the catalyst is regenerated (Fig. 9b). The se-
be seen from the data illustrated in Fig. 9a, the yield of 2,6- lectivities to 1,2-AMN from the regenerated samples cor-
respond to the values obtained on BEA3p(1h). These data
suggest that, on the sample passivated for 3 h (BEA3p(3h)),
the amorphous silica is sintering during regeneration and
making surface acid sites available.
3
.3. Acylation of Isobutylbenzene with Acetic
Anhydride over Zeolite Beta
Figure 10 shows the total conversion of isobutylbenzene
and the yield of reaction products versus reaction time over
BEA1. The major product is 4-isobutylacetophenone. A
small amount of 2-isobutylacetophenone is observed. The
selectivity of 4-isobutylacetophenone is excellent. How-
ever, the activity of the catalyst is low. The turnover num-
ber at 4 min of reaction is equal to 0.3 for the acylation of
FIG. 9. Yield of 2,6-AMN (a) and 1,2-AMN (b) versus total conver-
FIG. 10. Conversion of isobutylbenzene versus reaction time over ze-
olite BEA1 (0.10 g of catalyst, 100 C, isobutylbenzene : acetic anhydride,
1 : 1 and 0.8 mmol, 1 ml of solvent).
�
sion over zeolite BEA3p(3h) as a function of the number of regeneration
�
(
100 C, 2MN : acetic anhydride, 1 : 1).

ACYLATION OVER ZEOLITE BETA
223
TABLE 6
ACKNOWLEDGMENTS
Acylation of Isobutylbenzene Catalyzed by Zeolite Beta Sam-
ples (0.25 g of Catalyst, 100 C, 11 h, Isobutylbenzene : Acetic An-
hydride, 1 : 1 and 1.1 mmol, 0.860 ml of Solvent)
Financial support of the Albemarle Corp. is gratefully acknowledged.
The authors thank Drs. Sheila Yeh and Stacey Zones from Chevron for
performing the XPS analysis. Javier Garcia Martinez thanks MEC for a
FPI fellowship.
�
Selectivity of
Sample
Conv. (% )
4-isobutylacetophenone (% )
REFERENCES
BEA1
BEA2
BEA3
16.6
10
11
>99
>99
97
1
2
. van Bekkum, H., Hoefnagel, A. J., van Koten, M. A., Gunnewegh,
E. A., and Kouwenhoven, H. W., in “Zeolites and Microporous Crys-
tals” (T. Hattori and T. Yashima, Eds.), Stud. Surf. Sci. Catal., Vol. 83,
p. 379. Elsevier, Amsterdam, 1994.
. Harrington, P. J., and Lodewijk, E., Org. Process Res. Develop. 1, 72
(
1997).
isobutylbenzene whereas it is equal to 15.8 for the acylation
of 2MN.
3
4
. Sheldon, R. A., Chem. Ind. 7, 903 (1992).
. Harvey, G., and Mader, G., Collect. Czech. Chem. Commun. 57, 862
The conversion of isobutylbenzene and the selectivity
of 4-isobutylbenzene over BEA1, BEA2, and BEA3 are
presented in Table 6. The zeolite with the lower Si/Al ra-
tio and the smallest particle size is the more active ze-
olite for the transformation of isobutylbenzene into 4-
isobutylacetophenone. In this case, a high external surface
has no effect on the reaction selectivity as the acylation does
not require the shape-selectivity properties of the zeolite.
Moreover, small crystals, as in the case of BEA1, are prefer-
able for obtaining a high number of exposed external sites
and a shorter intracrystalline diffusion path length.
(
1992).
5
. Yadav, G. D., and Krishnan, M. S., in “Recent Advances in Basic and
Applied Aspects of Industrial Catalysis” (T. S. R. Prasada Rao and
G. Murali Dhar, Eds.), Stud. Surf. Sci. Catal., Vol. 113, p. 259. Elsevier,
Amsterdam, 1998.
. Corma, A., Climent, M. J., Garcia, H., and Primo, J., Appl. Catal. 49,
1
. Rohan, D., Canaff, C., Magnoux, P., and Guisnet, M., J. Mol. Catal. A
129, 69 (1998).
. Richard, F., Carreyre, H., and Perot, G., J. Mol. Catal. A 103, 51 (1995).
. Gaare, K., Akporiaye, D., Holm, K., and Skattebol, L., Acta Chem.
Scand. 51, 1229 (1997).
0. Singh, A. P., and Pandey, A. K., J. Mol. Catal. A 123, 141 (1997).
1. Spagnol, M., Gilbert, L., Jacquot, R., Guillot, H., Tirel, P. J., and
Le Govic, A. M., in “Book of Abstracts of 4th International Sympo-
sium on Heterogeneous Catalysts and Fine Chemicals, Basel, Switzer-
land, Sept. 1996”.
6
7
09 (1989).
8
9
1
1
4
. CONCLUSIONS
1
2. Neves, I., Jayat, F., Magnoux, P., Perot, G., Ribeiro, F. R., Gubelmann,
M., and Guisnet, M., J. Mol. Catal. A 93, 169 (1994).
Zeolite beta is active for the acylation of 2MN but is not
selective to the desired product (2,6-AMN). Mild operat-
ing conditions (temperature and concentration of acylating
agent) can be used to obtain reasonable conversions and to
13. Freese, U., Heinrich, F., and Roessner, F., Catal. Today 49, 237 (1999).
4. Bhattacharya, D., Sharma, S., and Singh, A. P., Appl. Catal. A 150, 53
1
(
1997).
limit the deactivation of the catalyst. Surface-poisoning ex- 15. Smith, K., Zhenhua, Z., Delaude, L., and Hodgson, P. K. G., in “Het-
erogeneousCatalysisand Fine ChemicalsIV” (H. U. Blaser, A. Baiker,
and R. Prins, Eds.), Stud. Surf. Sci. Catal., Vol. 108, p. 99. Elsevier,
Amsterdam, 1997.
6. Heinichen, H. K., and Holderich, W. F., J. Catal. 185, 408 (1999).
7. Chiche, B., Finiels, A., Gauthier, C., and Geneste, P., J. Org. Chem. 51,
2128 (1986).
8. Gunnewegh, E. A., Gopie, S. S., and van Bekkum, H., J. Mol. Catal.
A 106, 151 (1996).
periments show that the formation of 1,2-AMN occurs on
the external surface of the zeolite and that the formation
of 2,6-AMN can take place inside the pores of the zeolite.
Therefore, the selectivity to 2,6-AMN is improved on ze-
olite beta samples with a larger crystal size. On the large
crystals, most of the acid sites can be passivated by coat-
ing of the crystals with a layer of amorphous silica. Upon
reduction of surface acid sites, significant increases in the
selectivity for 2,6-AMN are observed. The amount of sur-
face coating on the large crystals determines the yield of
1
1
1
1
2
9. Bharathi, P., Waghmode, S. B., Sivasanker, S., and Vetrivel, R., Bull.
Chem. Soc. Jpn. 72, 2161 (1999).
0. Harvey, G., Binder, G., and Prins, R., in “Catalysis by Microporous
Materials” (H. K. Beyer, H. G. Karge, I. Kiricsi, and J. B. Nagy, Eds.),
Stud. Surf. Sci. Catal., Vol. 94, p. 397. Elsevier, Amsterdam, 1995.
1. Spagnol, M., Gilbert, L., Guillot, H., and Tirel, J. P., PCT Int. Patent
WO 97/48665, 1997.
2
,6-AMN and the selectivity to this product. These data are
2
2
a clear example of shape-selective acylation with zeolites.
Isobutylbenzene is less reactive than 2MN but the de-
sired product, 4-isobutylacetophenone, is always obtained
2. Camblor, M. A., Corma, A., and Valencia, S., J. Mater. Chem. 8, 2137
(
1998).
since the isobutyl group provides for the para position be- 23. Kunkeler, P. J., Moeskops, D., and van Bekkum, H., Micropor. Mater.
1
1, 313 (1997).
ing the preferred sites for acylation. For isobutylbenzene,
the external surface of the zeolite contributes significantly
to the formation of 4-isobutylacetophenone. Thus, zeolite
beta with a small crystal size is the most desired catalyst for
this reaction.
2
2
4. Guisnet, M., and Magnoux, P., Appl. Catal. 54, 1 (1989).
5. Creyghton, E. J., Elings, J. A., Sheldon, R. A., and van Bekkum, H.,
Micropor. Mater. 5, 299 (1996).
6. Derouane, E. G., Dillon, C. J., Bethell, D., and Derouane-Abd Hamid,
S. B., J. Catal. 187, 209 (1999).
2