Chemistry Letters Vol.33, No.10 (2004)
1289
the increase in the lactam yield to 80% and a reduction in the
amount of benzamide as a hydrolysis product to 0.04% of the
solvent were achieved since the water existing in the reaction
system could be effectively minimized. Thermogravimetric
analysis indicated that the amount of adsorbing water on the cat-
ꢁ
ꢂ1
alyst after heating at 350 C was 0.39 mmol g , corresponding
ꢁ
to the H2O/Al ratio of 0.10. However, the treatment at 400 C
resulted in a decrease in the selectivity to lactam although the
conversion was totally recovered.
a
b
c
H
H
O
O
H
H
d
e
Al
O
Al
O
400 °C
f
g
h
Scheme 1. Reversibly chemical adsorption of water on a Lewis
acid site.
2320
2300
2280
Wavenumber/ cm
2260
2240
2220
2200
-
1
The energetically preferential adsorption of water on Lewis
acid sites should cause the retardation of oxime hydrolysis and
Figure 1. FTIR spectra of HUSY (7) (a) after adsorption of
PhCN at room temperature and evacuation at (b) room temper-
ature, (c) 100, (d) 150, (e) 200, (f) 300, (g) 400, and (h) 500 C.
4
the improvement of lactam selectivity. This type of chemical
adsorption on aluminum and other cations was believed to gen-
erate the hydroxylated species and protons on basic framework
ꢁ
1h, some Lewis acid sites could hold the PhCN adsorbing even
ꢁ
at evacuation temperatures higher than 400 C.
5
oxygens (Scheme 1, the forward reaction). To completely elim-
inate the water and regain Lewis acidity, some energy is of
necessity.
When the water-adsorbed catalyst was heated at temperature
It is likely that PhCN adsorbs on Lewis acid sites forming
more stable intermediate species than the PhCN protonated on
Brꢀnsted acid sites. This may be suggested by the relatively
large blue shift caused by the former species for about 45
cm from the PhCN condensed phase at 2229 cm , indicating
that the CꢃN bond was more strengthened. Hence, probably
when the catalyst was allowed to contact with PhCN in the ab-
sence of relatively strong bases, both adsorbing species formed
on the acid sites. The oxime added afterwards could replace es-
sentially the protonated PhCN species, and the rearrangement to
the lactam was catalysed by weak Brꢀnsted acidity. In contrast,
the more stabilized species on Lewis acid sites remained and
prevented the oxime from hydrolysis.
ꢁ
as high as 400 C, desorption of molecular water and corre-
spondingly dissociated species occurred extensively, recovering
substantially Lewis acid sites (Scheme 1, the reverse reaction).
Moreover, it is common to consider that these sites are the sim-
ilar ones originally existing in the parent HUSY catalyst since
the comparable results were obtained over the catalysts before
and after pre-adsorption of water followed by heat treatment be-
ꢂ1
ꢂ1
ꢁ
tween 400 and 500 C. However, when the temperature was fur-
ꢁ
ther increased to 900 C, the lactam yield was severely reduced
whereas the formation of cyclohexanone was enhanced. This re-
sult should be ascribed to a loss of Brꢀnsted acid sites, as the ac-
tive sites for the rearrangement, and a generation of Lewis acid
The advantageous procedures for improving the lactam se-
lectivity in the liquid-phase Beckmann rearrangement of cyclo-
hexanone oxime over HUSY (7) catalyst were presented. A high
selectivity up to 97% can be achieved by pre-adsorption of water
6
sites, which are the hydrolysis active centers.
A remarkable improvement of the lactam selectivity was al-
so achieved by suspending the catalyst in PhCN prior to the ad-
dition of oxime (Table 1). Despite the decrease in the oxime con-
version due to an obstruction of the PhCN molecules initially ex-
isting on the acid sites, the cyclohexanone formation was retard-
ed with a slight increase in the lactam yield.
ꢁ
on the catalyst followed by heating at 150–350 C or contacting
the catalyst with PhCN before the addition of oxime.
References
1
A. Aucejo, M. C. Burguet, A. Corma, and V. Forn e´ s,
Appl. Catal., 22, 187 (1986); H. Sato, K. Hirose, M. Kitamura,
and Y. Nakamura, Stud. Surf. Sci. Catal., 49, 1213 (1989);
J. R o¨ seler, G. Heitmann, and W. F. H o¨ lderich, Appl. Catal.,
A, 144, 319 (1996); L. X. Dai, Y. Iwaki, K. Koyama, and
T. Tatsumi, Appl. Surf. Sci., 121/122, 335 (1997).
M. A. Camblor, A. Corma, H. Garc ´ı a, V. Semmer-Herl e´ dan,
and S. Valencia, J. Catal., 177, 267 (1998); Y.-M. Chung
and H.-K. Rhee, J. Mol. Catal. A: Chem., 175, 249 (2001).
H. Ichihashi, M. Ishida, A. Shiga, M. Kitamura, T. Suzuki, K.
Suenobu, and K. Sugita, Catal. Surv. Jpn., 7, 261 (2003).
C. Ngamcharussrivichai, P. Wu, and T. Tatsumi, unpublished
results.
Figure 1 shows the FTIR spectra of HUSY (7) catalyst after
ꢁ
pre-treating in situ at 500 C followed by PhCN adsorption at
room temperature and further evacuation at various tempera-
tures. At room temperature, there were at least 5 absorption
bands of CꢃN stretching found at 2230, 2240, 2253, 2272,
ꢂ1
2
and 2284 cm which respectively associated with silanol
groups, weak and strong Brꢀnsted acid sites, and weak and
4
ꢁ
strong Lewis acid sites. The evacuation up to 200 C mainly re-
moved PhCN adsorbing on silanol groups and on Brꢀnsted sites
3
4
5
while most of PhCN on Lewis sites was still intact (Figures 1c–
ꢁ
1e). Elevating the temperatures to 300 C reduced almost entire-
ly the bands relating to Brꢀnsted acid sites with a large decrease
in the bands corresponding to Lewis acid sites (Figure 1f). These
results indicate that PhCN interact with Lewis acid sites more
strongly than with Brꢀnsted ones. As revealed in Figures 1g,
J. W. Ward, J. Catal., 19, 348 (1970); E. Borello, G. D. Gatta,
B. Fubini, C. Morterra, and G. Venturello, J. Catal., 35, 1
(1974).
6
J. W. Ward, J. Catal., 9, 225 (1967).
Published on the web (Advance View) September 4, 2004; DOI 10.1246/cl.2004.1288