Active Sites for the Liquid-Phase Beckmann Rearrangement of Cyclohexanone, Acetophenone and Cyclododecanone Oximes, Catalyzed by Beta Zeolites

Article abstract of DOI:10.1006/jcat.1998.2110

The Beckmann rearrangement of oximes with different molecular sizes, i.e. cyclohexanone, cyclododecanone, and acetophenone oximes, has been studied in liquid phase at 130°C over a series of four Beta zeolites differing in the presence or absence of framework Al and internal silanol groups. When the zeolite does not contain framework Al and internal silanols, no appreciable conversion was observed. The catalyst having internal silanol groups but no framework Al exhibits oxime conversion, but the selectivity to the corresponding amide is low in some cases. In the Beta zeolite without silanol groups but containing framework Al, conversion and selectivity were found to be very high. This superior performance of Broensted acid sites, compared to silanol groups, shows that the results reported for the vapor phase reaction cannot be extrapolated when the reaction is performed in liquid phase. Finally, as could be anticipated according to the dimensions of the micropores, it is shown that H-Beta zeolites exhibit a much better catalytic performance than H-ZSM-5 zeolite for larger sized oximes.

Full text of DOI:10.1006/jcat.1998.2110

JOURNAL OF CATALYSIS 177, 267–272 (1998)
ARTICLE NO. CA982110
Active Sites for the Liquid-Phase Beckmann Rearrangement
of Cyclohexanone, Acetophenone and Cyclododecanone
Oximes, Catalyzed by Beta Zeolites
M. A. Camblor, A. Corma,1 H. Garc´ıa, V. Semmer-Herle´dan, and S. Valencia
Instituto de Tecnolog´ıa Qu´ımica, U.P.V.-C.S.I.C, Universidad Polite´cnica de Valencia, Avenida de los Naranjos s/n, 46022 Valencia, Spain
Received January 15, 1998; revised April 9, 1998; accepted April 17, 1998
In order to overcome these problems, a large number
of heterogeneous catalysts, including silica (2), alumina
(3), silica-alumina (4), supported boron oxide (5–7), sup-
ported tantalum oxide (8), titanium silicalites (9), zeolites
(Y (10–12), Beta (13), mordenite (14), pentasil (5–7, 14–
18), SAPO-11 (19), SAPO-34, and mesoporous materials
(MCM-41) (20) were all tested for the vapor-phase rear-
rangement of cyclohexanone oxime. In this process, the re-
action must be conducted at temperatures between 250 and
350ꢀC in order to keep the oxime and productsin the vapour
phase. These high temperatures have a negative effect, de-
creasing the lactam selectivity and leading to a fast deac-
tivation of the catalyst due to coke formation. Hence, the
possibility of using a solid–liquid phase process at moderate
temperatures has to be reconsidered.
A large number of works on the Beckmann rearrange-
ment reaction usingsolid catalystsfocuseson the nature and
strength of the sites responsible for this reaction (21). Ear-
lier papers suggested that the formation of the Bro¨nsted
amide is catalyzed by sites with medium and strong acid
strengths (11, 22). However, later studies, starting with that
of Sato et al. (14, 23), showed that both the catalytic activity
and selectivity of lactam formation increase with increas-
ing Si/Al ratio in H-ZSM-5 zeolite, leading to the use of
molecular sieves with weaker acidities such as SAPO (24).
More recently, it was proposed that the rearrangement of
cyclohexanone oxime is not necessarily catalyzed by acidic
centers but by sites with rather low or no acidity, which
are sufficient to achieve high selectivity at high conversion,
while maintaining the long life of the catalyst (18).
The Beckmann rearrangement of oximes with different molec-
ular sizes, i.e. cyclohexanone, cyclododecanone, and acetophenone
oximes, has been studied in liquid phase at 130ꢀC over a series of
four Beta zeolites differing in the presence or absence of framework
Al and internal silanol groups. When the zeolite does not contain
framework Al and internal silanols, no appreciable conversion was
observed. The catalyst having internal silanol groups but no frame-
work Al exhibits oxime conversion, but the selectivity to the corre-
sponding amide is low in some cases. In the Beta zeolite without
silanol groups but containing framework Al, conversion and selec-
tivity were found to be very high. This superior performance of
Bro¨nsted acid sites, compared to silanol groups, shows that the re-
sults reported for the vapor phase reaction cannot be extrapolated
when the reaction is performed in liquid phase. Finally, as could
be anticipated according to the dimensions of the micropores, it is
shown that H-Beta zeolites exhibit a much better catalytic perfor-
c
mance than H-ZSM-5 zeolite for larger sized oximes.
ꢁ 1998 Academic
Press
INTRODUCTION
The Beckmann rearrangement, i.e. the transposition of
ketoximes into amides, has been extensively studied for
many years. The most important industrial examples are
the conversion of cyclohexanone and cyclododecanone
oximes into the corresponding lactams which are the pre-
cursors for the fabrication of Nylon 6 and 12, respectively.
With the classical, large-scale method for synthesizing
ε-caprolactam, the rearrangement of cyclohexanone oxime
is carried out in the liquid phase using concentrated sulfuric
acid (1). Although this procedure is very convenient from
a chemical point of view, the large amount of ammonium
sulfate formed during the subsequent neutralization of the
oleum, the use of large amounts of fuming sulfuric acid, and
the corresponding problem of corrosion make this process
environmentally questionable.
In our work, we carried out the Beckmann rearrange-
ment of oximes of different molecular sizes in a solid–liquid
reaction system. We used a series of Beta zeolites as cata-
lysts in which the concentrations of external silanols, inter-
nal silanols, and framework Al were varied. The influence
of the structure of the zeolite and the molecular size of
the oxime were determined by comparing the results of ex-
periments conducted in the presence of Beta catalysts with
those using H-ZSM-5.
1 To whom correspondence should be addressed. E-mail: acorma@
itq.upv.es.
267
0021-9517/98 $25.00
c
Copyright ꢁ 1998 by Academic Press
All rights of reproduction in any form reserved.

268
CAMBLOR ET AL.
calcined at 580ꢀC for 3 h. The Beta-D was dealuminated
TABLE 1
with concentrated HNO3 under reflux conditions (25) prior
to calcination. The nomenclature and details of the synthe-
sis of each sample are listed in Table 1, and some relevant
properties are summarized in Table 2. The Si/Al ratio of
Beta zeolites was determined by chemical analysis. Crys-
tallinity of the samples was measured by XRD using beta-
ND as the reference material. The presence of the silanol
groups was established by IR (Nicolet 710 FT spectropho-
tometer) using self-supported wafers (ꢃ10 mg, compressed
at 2 Ton ꢄ cmꢂ2) after outgassing at 400ꢀC under 10ꢂ2 Pa
for 16 h. The acidity of the zeolites was determined by
the standard pyridine adsorption–desorption method. Solid
state 29Si NMR spectra were recorded on a Varioan Unity
VXR-400 WB spectrometer at 79.5 MHz, a spinning speed
of SkHz, a recycle delay of 20 s and using a 55.4ꢀ pube of
4.0 ꢀs.
Nomenclature and Details of Synthesis of the Beta Samples
Duration of
Sample
Gel composition
synthesis (days)
Beta-ND
(Ref. (24))
SiO2; 0.54 TEAOH;
0.54 HF; 7.25 H2O
2.7
Beta-D
Al2O3; 1600 SiO2; 864 TEAOH;
10400 H2O; Beta-seedsa
(3% of the total SiO2)
18
H-Beta-ND
(Ref. (25))
Al2O3; 50 SiO2; 29 TEAOH;
29 HF; 362.5 H2O
5.5
16
H-Beta-D
(Ref. (23))
Al2O3; 100 SiO2;
54 TEAOH; 1500 H2O
a Gel composition: Al2O3; 50 SiO2; 28 TEAOH; 325 H2O; syn-
thesis duration: 3 days (23).
EXPERIMENTAL
Reagents
Catalysts
Oximes were synthezised from the corresponding ketone
by reacting it with hydroxylamine hydrochloride in a mix-
ture of ethanol and pyridine at 85ꢀC according to the gen-
eralmethod reported in the literature (26). The oximeswere
recrystallised from ethanol before use. Their identity and
purity were confirmed by gas chromatography (GC) using a
Hewlett-Packard 5890 instrument and a 25-m capillary col-
umn of crosslinked 5% phenylmethylsilicone and GC-MS
using a Hewlett-Pakard 5988A spectrometer containing the
same column type as used in the GC experiments.
The H-ZSM-5 used was a commercial sample produced
by P.Q. Industries (CBV8020). All the Beta zeolite samples
were synthesized hydrothermallyat 140ꢀC, usingtetraethyl-
ammonium (TEA+) as a structure-directing agent in the
absence of alkali cations. The reagents used were metal
Al (Meck, 99.9% ), alkali-free TEAOH (35 wt% aqueous
solution, Aldrich, Na < 2 ppm, K < 0.5 ppm), HF (Merck,
48% ), and amorphous silica (Aerosil 200, Degussa) for the
H-Beta-D sample and tetraethylorthosilicate (Merck) for
the other samples. The crystallizations were carried out
in PTFE-lined stainless steel 60-ml autoclaves under ro-
tation (60 rpm). After quenching at the appropriate crys-
Reaction Procedure
A solution of oxime (100 mg) and decane (50 mg), as
tallization time, the solids were separated by centrifugation the internal standard, in chlorobenzene (20 ml, Aldrich,
(16,000 rpm; 90 min) in the case of syntheses in OHꢂ me- used as received) as the solvent was heated at 130ꢀC. The
dia and by filtration in the case of Fꢂ media. They were catalyst (100 mg), previously calcined at 500ꢀC overnight,
extensively washed with distilled water, dried at 100ꢀC, and was added to the solution. The resulting suspension was
TABLE 2
Relevant Characteristics of the Zeolites Used in This Work
Acidity (ꢀmol py)c
Crystal
Bro¨nsted
623 K
Lewis
623 K
Presence
sizesa
Crystallinityb
(% )
Sample
Si/Al
37
>10000
>10000
>4000
34
of silanols
(ꢀm)
523 K
673 K
523 K
673 K
H-ZSM-5
Beta-ND-1
Beta-ND-2
Beta-D
H-Beta-ND
H-Beta-D
High
—
100
100
64
93
63
39
26
6
5
5
4
d
d
d
d
d
d
6–8
0.5–1.0
0.2
0.5
0.1
d
d
d
d
d
d
d
d
d
d
d
d
High
Low
High
24
34
10
23
5
10
14
29
11
30
8
23
31
a Derived from SEM images.
b Of calcined samples.
c From extinction coef. by Emeis (32).
d No pyridine was detected by IR.

LIQUID-PHASE BECKMANN REARRANGEMENT
269
stirred magnetically and refluxed at the reaction temper-
ature. The course of the reaction was followed at times
ranging from 0.1 to 24 h by GC. At the end of the re-
action, the catalyst was filtered and washed with CH2Cl2
and the organic solutions concentrated in vacuo, weighed,
and analyzed by GC-MS. The zeolite was submitted to con-
tinuous solid–liquid extraction with CH2Cl2 using micro-
Soxhlet equipment. After removing the solvent, the residue
was weighed and analyzed by GC-MS. The total amount of
recovered material accounted for more than 90% of the
starting oxime, and no significant differences were found
between the relative percentages of each compound pres-
ent in the reaction liquid phase and the zeolite-extracted
residue.
RESULTS AND DISCUSSION
In order to determine the activity and selectivity of
silanols compared to bridging hydroxyl groups, a series
of five zeolites was chosen to be used as catalysts.
First, totally siliceous, highly crystalline Beta zeolites
(Beta-ND) were prepared by the novel fluoride method
(27, 28). The Beta-ND samples contain only external
silanol groups as possible catalytic sites (see 29SiMAS
NMR below). A second all-silica Beta sample (Beta-D)
was prepared by dealumination (25) of a zeolite with
a Si/Al ratio of 800 obtained directly by synthesis us-
ing a previously described synthesis procedure (29). The
lower crystallinity of Beta-D compared to Beta-ND can
be quantitatively assessed by powder XRD. Figure 1
shows the corresponding diffractograms after calcination
obtained exactly under the same conditions. Before calci-
nation, Beta-ND and Beta-D were both highly crystalline.
The lower crystallinity of Beta-D after the calcination step
is the consequence of the presence of lattice defects and/
FIG. 2. 29Si NMR of some of the catalysts used in this work.
or disruptions that introduces a significant population of
internal silanol groups. As can be seen from the 29Si NMR
spectra shown in Fig. 2, Beta-ND (only Beta-ND-1 shown)
shows a very well resolved spectrum and the absence of a
peak at yields lower than ꢂ105 ppm. This indicates that the
Beta-ND sample does not have structural defects, and con-
sequently, that external silanol groups only are present in
these catalysts. On the other hand, the spectrum of Beta-
D is poorly resolved, and its deconvolution shows a very
broad signal centered at ꢂ103 ppm, indicative of defects
which are associated with internal silanol groups. Other ex-
perimental evidence for the presence of this second type of
silanol group is the OH stretching region of the IR spectrum
of Beta-D (Fig. 3). The external silanols are indicated by
the presence of a relatively sharp band at 3745 cmꢂ1, while
FIG. 1. XRD patterns of calcined Beta samples.

270
CAMBLOR ET AL.
Other products, such as aniline or 5-cyanopent-1-ene, which
were observed in other studies and arise from dehydro-
genation and fragmentation, were not detected under our
reaction conditions.
The results in terms of conversion, selectivity, and ini-
tial reaction rate (r0), measured for the liquid-phase rear-
rangement of cyclohexanone, acetophenone, and cyclodo-
decanone oximes at 130ꢀC over the different catalysts, are
summarized in Tables 3 to 5. Initial reaction rates were
obtained from the slope of the time-conversion plot at con-
versions below 10% .
It has been reported that silicalite (an all-silica pentasil
structure isomorphic to ZSM-5) is the most selective cata-
lyst for the vapor-phase Beckmann rearrangement of the
cyclohexanone oxime (16). According to this, the vapor-
phase results can be simply extrapolated to the liquid-phase
reaction; it is expected that the best Beta zeolite catalysts
will be the pure silica samples (Beta ND and Beta D). How-
ever, as can be seen in Tables 3 to 5, the three oximes are
practically not transformed on the Beta-ND which have
only external silanol groups, regardless of the crystallite
size of the sample. The sample Beta-D, containing inter-
nal silanols, behaves differently. Beta-D gives moderate
conversions, albeit the selectivity to the rearrangement was
low for cyclododecanone and acetophenone oximes. Thus,
we conclude that internal silanol groups have some activity
in the liquid-phase transformation of oximes.
FIG. 3. IR spectrum of Beta samples.
the internal silanols are associated with a less well-defined
broad absorption band at 3550–3700 cmꢂ1. The broadness
and low frequency of this band are indicative of intense
hydrogen bridge interactions in the internal silanol in con-
trast to the much sharper band associated with the external
silanols.
From the catalytic point of view, it is anticipated that the
presence of internal silanols can influence the activity of the
zeolite in two different ways: (i) by having a different acid
strength than the external silanols and/or (ii) by increasing
the polarity of the internal voids and, thus, enhancing the
absorption of the oxime inside the pores. With regard to the
relative hydrophillicity/hydrophobicity of samples Beta-D
and Beta-ND (point (ii)), it has been shown that, due to
the remarkable hydrophobicity of Beta-ND it adsorbs 150
times more n-hexane than H2O (27).
TABLE 3
The two other Beta zeolites studied here (see Table 2) are
similar to Beta-ND and Beta-D but contain Al (H-Beta-
ND and H-Beta-D). Since the presence of framework Al
introduces Bro¨nsted acid sites into these solids, it should be
possible to compare the activity of external silanols, internal
Beckmann Rearrangement of the Cyclohexanone Oxime 1a
Selectivities for
2a and/or 3a
(% )
Initial rate
r0 ꢄ 103
Conversion
(% )
Sample
Aerosil
(mol hꢂ1 gꢂ1
)
ꢅ
ꢅ
silanols, and bridging hydroxyls Si(OH)Al .
In addition, for the sake of comparison, we also included H-ZSM5
5
67
0
>98 3a
95 2a
—
—
1.6
—
Beta-ND-1
the activity of an H-ZSM-5 zeolite sample, having a similar
Si/Al ratio as the other Al-Beta zeolites (Table 2). Using
these two structures, we can now discuss the influence of
zeolite pore size on the reactivity of oximes of different
kinetic diameters.
During the reaction (the Beckmann rearrangement) the
only compounds observed, besides the starting material
(1a-c), were the corresponding amide (2a-c), and the parent
ketone (3a-c).
Beta-ND-2
Beta-D
H-Beta-ND
H-Beta-D
Amorphous
SiO2
0.2
38
54
68
0.1
80 3a
>98 2a
>98 2a
>98 2a
72 2a
—
3.1
4.8
8.6
—
Note. Total conversion and selectivities for ε-caprolactam 2a and cyclo-
hexanone 3a measured for 24 h during the reaction; initial reaction rates
r0 are given.

LIQUID-PHASE BECKMANN REARRANGEMENT
271
tive and the selectivity for the Beckmann rearrangement is
very high. This is a clear indication that bridging hydroxyl
groups are active and selective for carrying out the trans-
formation of oximes. It can be argued that the active sites
are not the bridging hydroxyl groups but rather the inter-
nal silanols, generated when some of the framework Al is
extracted from the framework positions during calcination.
While it is true that some framework dealumination oc-
curred during the activation of H-Beta-ND, thus creating
internal silanols, the number of internal silanols formed in
H-Beta-ND is much lower than that of sample Beta-D (see
Fig. 2). Nevertheless, the activity of the former is higher,
indicating the positive role of the Bro¨nsted acid sites associ-
ated with framework Al. This observation is also supported
by the results obtained with H-Beta-D, which gives a higher
catalytic activity than Beta-D zeolite.
TABLE 4
Beckmann Rearrangement of the Acetophenone Oxime 1b
Selectivities for
2b and/or 3b
(% )
Initial rate
Conversion
(% )
r0 ꢄ 103
Sample
(mol hꢂ1 gꢂ1
)
H-ZSM5
76 2b
24 3b
—
45
0
1.1
—
Beta-ND
Beta-D
92 2b
50
67
97
8
3b
1.4
0.9
2.9
H-Beta-ND
H-Beta-D
>98 2b
>98 2b
Note. Conversion total and selectivities for acetanilide 2b and acetophe-
none 3b measured for 24 h during the reaction; initial reaction rates r0 are
given.
Thus, it appears that, while internal silanol groups are
active and selective in most cases, Bro¨nsted acid sites asso-
ciated with bridging hydroxyls are even more active, ex-
hibiting the highest selectivity for the formation of the
amide. From these observations and taking into account
that external silanol groups are much less active and se-
lectively catalyze the hydrolysis to the ketone, but not
the Beckmann rearrangement, we assume that the inter-
nal silanol groups are more acidic than the external silanols
(30). This may be caused by the hydrogen bonding interac-
tions of the former, as well as by the fact that the reactant
is already partially activated when it penetrates the pores
of the zeolite, owing to the strong electric fields on the
electronic confinement occurring inside the zeolite pores
(31).
If this is so, then one can conclude that nonacidic silanols
preferentially catalyze the hydrolysis of the oxime, while
weakly acidic and/or internal silanols can even catalyze the
Beckmann rearrangement. Our results show that Bro¨nsted
hydroxyl groups, associated with framework Al, are highly
active and selective sites for the formation of amides in the
liquid phase. When the reaction takes place in the vapor
phase, it may be the case that, at the higher reaction tem-
peratures required, these stronger acid sites become rapidly
poisoned and that most of the final observations are the re-
sult of the internal silanol groups.
In order to discuss the potential role of the external
silanol groups, which did not show any activity on the Beta-
ND sample, cyclohexanone oxime was heated at 130ꢀC in
the presence of a silica (Aerosil, Degussa) with a large sur-
face area (300 m2 ꢄ gꢂ1) and a large population of external
silanols (Fig. 2). The results of this reaction are also in-
cluded in Table 3. It can be seen that, when Aerosil is used
as the catalyst, the oxime conversion and the selectivity to-
wards ε-caprolactam are much lower than that obtained
with Beta-D. Furthermore, an amorphous silica (Basf,
137 m2 ꢆ gꢂ1) was also used as catalyst and practically no
conversion was obtained. This clearly indicates the differ-
ent catalytic behaviour of the external and internal silanols
of the zeolite. The external silanols are mainly responsi-
ble for the hydrolysis of the oxime to give the correspond-
ing ketone, while the internal silanols can catalyze the
Beckmann rearrangement to produce the corresponding
amide. The different catalytic behavior of the two types of
silanols may be related to differences in acidity, as discussed
below.
When framework Al is introduced and Bro¨nsted acid-
ity is developed, the resulting Beta-ND is catalytically ac-
TABLE 5
Beckmann Rearrangement of the Cyclododecanone Oxime 1c
Selectivities for
2c and/or 3c
(% )
Initial rate
r0 ꢄ 103
Influence of the zeolite pore dimensions. It has been
reported that ZSM-5 zeolite is a good catalyst for trans-
forming cyclohexanone oxime into caprolactam under va-
por phase conditions (2, 11). We confirm this fact in solu-
tion since, as can be seen in Table 3, the activity and the
selectivity of ZSM-5 for 2a are comparable to those of H-
Beta-D. However, if we compare the results obtained for
the transposition of acetophenone 1b oximes (Table 4) and
cyclododecanone 1coximes(Table 5), two ketoximeswhose
molecular diameter is greater than that of 1a, then H-ZSM-
5 shows a decrease in activity for the reaction with 1b and is
completely inactive when it comes to converting 1c. Mean-
Conversion
(% )
Sample
(mol hꢂ1 gꢂ1
)
H-ZSM5
Beta-ND
Beta-D
0
0
—
—
—
—
39 2c
38 3c
>98 2c
72 2c
17 3c
46
92
1.4
2.8
H-Beta-ND
H-Beta-D
91
5.3
Note. Conversion total and selectivities for ω-laurinlactam 2c and cy-
clododecanone 3b measured for 24 h during the reaction; initial reaction
rates r0 are given.

272
CAMBLOR ET AL.
while, H-Beta-D performs the Beckmann rearrangement
of 1b and 1c very well.
2. Sato, S., Urabe, K., and Izumi, Y., J. Catal. 102, 98 (1986).
3. Curtin, T., McMonagle, J. B., and Hodnett, B. K., Catal. Lett. 17, 145
(1993).
Since H-ZSM-5 and H-Beta-D have similar Bro¨nsted
acidity (see Table 2), the differences in activity are prob-
ably related to geometrical effects due to the differences
in pore dimensions. Indeed, ZSM-5 zeolite has a pore size
of 0.55 nm, while Beta has a pore size of 0.7 nm. Thus,
the bulkier acetophenone oxime (0.7 nm) is barely able to
diffuse inside the pores; an important contribution of the
sites at the external surface takes place, as can be seen from
the lower selectivity to the amide obtained with this reac-
tant on H-ZSM-5. In the case of cyclododecanone oxime
(0.9 nm), it is already too large to react on H-ZSM-5, and
this zeolite becomes inactive. In the case of Beta, there are
problems related to the diffusion of the cyclododecanone
oxime, and, consequently, the contribution of the external
surface to the total activity should become more important
decreasing therefore the selectivity to the amide (Table 5).
There is no doubt that the selectivity could be improved by
silanizing the external silanol groups.
4. Gutierrez, E., Aznar, A. J., and Ruiz-Hitsky, E., Stud. Surf. Sci. Catal.
59, 539 (1995).
5. Takahashi, T., Ueno, K., and Kai, T., J. Chem. Eng. 69, 1096 (1991).
6. Ro¨seler, J., Heitmann, G., and Ho¨lderich, W. F., Stud. Surf. Sci. Catal.
105, 1173 (1996).
7. Ro¨seler, J., Heitmann, G., and Ho¨lderich, W. F., Appl. Catal. 144, 319
(1996).
8. Ushikudo, T., and Wada, K., J. Catal. 148, 138 (1994).
9. Thangaraj, A., Sivasanker, S., and Ratnasamy, P., J. Catal. 137, 252
(1992).
10. Landis, P. S., and Venuto, P. B., J. Catal. 6, 245 (1966).
11. Aucejo, A., Bruguet, M. C., Corma, A., and Forne´s, V., Appl. Catal.
22, 187 (1986).
12. Corma, A., Garcia, H., Primo, J., and Sastre, E., Zeolites 11, 593 (1991).
13. Dai, L. X., Hayasaka, R., Iwaki, Y., Koyano, K. A., and Tatsumi, T.,
Chem. Commun., 1071 (1996).
14. Sato, H., Ishii, N., Hirose, K., and Nakamura, S., Stud. Surf. Sci.Catal.
28, 755 (1986).
15. Sato, H., Hirose, K., and Nakamura, S., Stud. Surf. Sci.Catal. 49, 1213
(1989).
16. Sato, H., Hirose, K., and Nakamura, S., Chem. Lett., 1987
(1993).
CONCLUSIONS
17. Yashima, T., Miura, K., and Komastu, T., Stud. Surf. Sci. Catal. 84, 1897
(1994).
We have proven that external silanol groups present in 18. Ho¨lderich, W. F., Stud. Surf. Sci. Catal. 46, 193 (1989).
19. Singh, P. S., Bandyopadhayay, R., Hegde, S. G., and Rao, B. S., Appl.
Catal. 136, 249 (1996).
20. Dai, L. X., Hayasaka, R., Iwaki, Y., and Tatsumi, T., Sogo Shikensho
zeolites may be active when the reaction is carried out at
moderate temperatures in the liquid phase, and thus the
oxime is hydrolyzed to give the corresponding ketone. On
Nenpo (Tokyo Daigaku, Kogakubu) 55, 159 (1996).
the other hand, internal silanols are active and selective for
performing the Beckmann rearrangement (33). This may
be due to a higher acidity of the internal silanols, compared
to the external silanols and/or to the preactivation of the
oxime in the pore of the zeolite. Under the same reaction
conditions, Bro¨nsted acid sites are very active and selective
for the formation of the amide. Finally, properly prepared
Beta zeolite is suitable for carrying out the Beckmann rear-
rangement of large size oximes, such as acetophenone and
cyclododecanone oximes.
21. Reddy, J. S., Ravishankar, R., and Ratnasamy, P., Catal. Lett. 17, 139
(1993).
22. Bell, W. K., and Chang, C. D., Europ. Patent 056 698 (1985).
23. Sato, H., Hirose, K., Ishii, N., and Umada, Y., E.P. 234 088 (1987).
24. Olson, K. D., E.P. 251 168 (1988).
25. Fajula, F., E.P. 0 488 867 A1 (1991).
26. Vogel, L., “A Text Book of Practical Organic Chemistry,” ELBS ed.,
p. 343. Longmann, London, 1971.
27. Camblor, M. A., Corma, A., and Valencia, S., Chem. Commun., 2365
(1996). [Blasco, T., Camblor, M. A., Corma, A., Esteve, P., Guil, J. M.,
Mart´ınez, A., Perdigo´n-Melo´n, J. A., and Valencia, S., J. Phys Chem.
B 102, 75 (1998).]
28. Camblor, M. A., Corma, A., and Valencia, S., manuscript in prepara-
tion.
ACKNOWLEDGMENT
29. Camblor, M. A., Corma, A., Mifsud, A., Pe´rez-Pariente, J., and
Valencia, S., Stud. Surf. Sci. Catal. 105, 341 (1997).
30. Van Santen, R. A., private communication.
31. Corma, A., Zicovich-Wilson, C. M., and Viruela, P., J. Phys. Chem. 98,
10863 (1994).
Financial support by the Spanish MAT97-1016-C02-01 is gratefully
acknowledged.
REFERENCES
32. Emeis, C. A., J. Catal. 141, 347 (1993).
1. Kim, S.-G., Kawakami, T., Ando, T., and Yukawa, Y., Bull. Chem.
Soc. Jpn. 52(4), 1115 (1979).
33. Holderich, W. F., Ro¨seler, J., Heitmann, G., and Liebenz, A. T., Catal.
Today. 37, 353 (1997).