Selective oxidation of benzylic alcohols to aldehydes with metal nitrate reagents catalyzed by BEA zeolites or clays

Article abstract of DOI:10.1039/b106252m

The oxidation of benzylic alcohols to benzaldehydes by bismuth(III) nitrate was investigated in the presence of various acidic aluminosilicates. The solid acid was used either as support for the nitrate or simply added in the reaction medium as a catalyst of the decomposition of Bi nitrate. The reactivity of the clay-supported reagent is controlled by the dispersion of the nitrate. A high dispersion requires both high surface area and high acidity of the support. There is a good correlation in the oxidation of benzyl alcohol - as the model reaction - between the structure and the chemical activity of the two main groups of supports (K10 and KSF vs. KSF/0 and KP10). In supported reagents made with zeolitic materials such as ZSM-5 or BEA the pores are blocked by the salt, therefore the solvent used for their preparation could not be eliminated completely from the solid. Zeolites cannot, therefore, be used to prepare conventional supported reagents, but work remarkably well, especially BEA, as a catalyst for the oxidation of benzyl alcohol or m-phenoxybenzyl alcohol by bismuth nitrate. Similar oxidation results were obtained with iron(III) nitrate.

Full text of DOI:10.1039/b106252m

View Article Online / Journal Homepage / Table of Contents for this issue
Selective oxidation of benzylic alcohols to aldehydes with
metal nitrate reagents catalyzed by BEA zeolites or clays
a
a
b
c
Janos Farkas, Sandor Bekassy,* Janos Madarasz and Franc ¸o is Figueras
´ ´ ´ ´ ´ ´
a
b
c
Department of Organic Chemical Technology, Budapest University of Technology and
Economics, H-1521 Budapest, Hungary. E-mail:bekas@oct.bme.hu; Fax: þ36 1 463 3648
Institute of General and Analytical Chemistry, Budapest University of Technology and
Economics, H-1521 Budapest, Hungary
Institut de Recherches sur la Catalyse du CNRS, 2 Avenue Albert Einstein, F-69626
Villeurbanne, France
Received (in Montpellier, France) 13th July 2001, Accepted 9th January 2002
First published as an Advance Article on the web 20th May 2002
The oxidation of benzylic alcohols to benzaldehydes by bismuth(III) nitrate was investigated in the presence of
various acidic aluminosilicates. The solid acid was used either as support for the nitrate or simply added in the
reaction medium as a catalyst of the decomposition of Bi nitrate. The reactivity of the clay-supported reagent is
controlled by the dispersion of the nitrate. Ahigh dispersion requires both high surface area and high acidity of
the support. There is a good correlation in the oxidation of benzyl alcohol—as the model reaction—between the
structure and the chemical activity of the two main groups of supports (K10 and KSF vs. KSF=0 and KP10).
In supported reagents made with zeolitic materials such as ZSM-5 or BEAthe pores are blocked by the salt,
therefore the solvent used for their preparation could not be eliminated completely from the solid. Zeolites
cannot, therefore, be used to prepare conventional supported reagents, but work remarkably well, especially
BEA, as a catalyst for the oxidation of benzyl alcohol or m-phenoxybenzyl alcohol by bismuth nitrate. Similar
oxidation results were obtained with iron(III) nitrate.
Oxydation selective d’alcools benzyliques en alde
BEA ou des argiles. L’oxydation d’alcools benzyliques en alde
presence de plusieurs alumino-silicates. Les acides solides etaient mis en œuvre soit comme le support du nitrate,
soit additionnes simplement au milieu reactionnel comme catalyseur de decomposition du nitrate. La reactivite
du reactif supporte par une argile est controlee par la dispersion du nitrate. Une bonne dispersion necessite a la
fois une grande surface et une forte acidite du support. Une bonne correlation existe dans les deux groupes de
supports (K10 et KSF vs. KSF=0 et KP10) entre structure et activite pour la reaction modele d’oxydation de
l’alcool benzylique. Sur les supports zeolithiques tels que ZSM-5 ou BEA, les pores sont bloques par le sel ce qui
empeche l’elimination du solvant lors de la preparation du reactif supporte. Ces zeolithes ne peuvent donc pas
tre utilisees pour preparer des reactifs supportes, mais sont d’excellents catalyseurs, surtout la BEA, pour
l’oxydation de l’alcool par le nitrate de bismuth. Des resultats similaires ont ete obtenus avec le nitrate de fer.
´
hydes par des nitrates me
´
talliques catalyse
´
e par des ze
´
olithes
´
hydes par le nitrate de bismuth a e
´
te tudie
´
e
´
´
e en
´
´
´
´
´
´
´
´
´
´
´
´
`
´
´
´
`
´
´
ˆ
´
´
´
´
eˆ
´
´
´
´
´
´
´
The selective oxidation of alcohols into aldehydes is a simple
reaction that is difficult to achieve on the industrial scale,
though can be carried out with tert-butyl hydroperoxide and
support allows them to be used as selective, safe and eco-
friendly reagents. This process has been described first with
acid-treated montmorillontes (for instance K10, KSF from
Sud Chemie) and has found industrial applications. It can be
¨
remarked, however, that the scale-up of supported reactants is
1
5,6
Cr-pillared clays. Other solutions have been reported using
2
hydrogen peroxide and titanium zeolites or vanadium sili-
3
4
cates or cobalt(II) acetate as catalyst. TS1 gives a high
selectivity to aldehyde but the yield is limited to 12–13% since
hydrogen peroxide has to be introduced in very small amounts
difficult. In this process the oxidant is the nitrate ion, according
7
to the detailed mechanism reported by Cornelis and Laszlo,
´
and the preparation of the supported reagent consists basically
in the deposition of metallic nitrates onto dealuminated clays
using a solution of the salt in acetone, producing clayfen and
2
in order to avoid secondary oxidation of the product. This
low yield requires huge recycling of the substrate. In the case of
vanadium silicate catalysts, the conversions obtained with an
7
,8
claycop, for example.
excess of H
2
O
2
are limited to <20% at most, whatever the
It appears, however, that the support is not chemically inert
in this process and the physico-chemical characterization of
these solids by IR spectroscopy, thermal analysis and X-ray
diffraction led to the conclusion that these supported nitrates
3
structure of the alochol. Using cobalt acetate the oxidation of
benzyl alcohol leads exclusively to benzaldehyde but the effi-
ciency of H O is poor because a significant amount of oxygen
2 2
is liberated too.
4
9
are in fact hydrated salts, nearly amorphous due to the small
These yields could be accepted for bulk chemicals but are
too low for fine chemistry, and alternative solutions have been
proposed based on the use of metallic nitrates, which give
high yields and allow eventual leaching of the Cr catalysts to
be avoided. The deposition of these nitrates on an appropriate
size of the crystals. Indeed, the acidic support acted as a cat-
alyst for the decomposition of the nitrate and could be used as
a usual catalyst, that is added as a component in the slurry
containing the reagents and the nitrates. This catalytic tech-
nique, earlier denominated as formation of the reagent in situ,
7
50
New J. Chem., 2002, 26, 750–754
DOI: 10.1039/b106252m
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2002

View Article Online
worked remarkably well with several nitrates and clays, these
Table 2 Characteristics of the K clays
1
0
latter being repeatedly used as normal catalysts. The above
catalytic technique was applied also for the selective oxidation
of the fragile m-phenoxybenzyl alcohol to the corresponding
Surface
area=m g
Average
pore size=nm
Brꢀnsted acidity=
arb. units
2
À1
Sample
1
0
aldehyde. Cornelis and Laszlo have pointed out that zeolites
´
7
K10
229
169
9
5.6
7.1
5.0
7.4
0.33
0.49
0.59
1.03
cannot be used to make supported reagents. Since zeolites
are very common solid acids, it appeared worthwhile to
understand the reason for this failure and to investigate their
possible application in the catalytic method.
KP10
KSF
KSF=0
117
Iron(III) and bismuth(III) nitrate are the most effective
among the metal nitrates supported on K10. In the reactions
with supported iron(III) nitrate, designated as Fe=K10, a part
of the solid remained stuck on the walls of the flask, therefore
this reagent was not suitable for more precise evaluation. This
problem was not observed with Bi=K10, therefore bismuth(III)
nitrate was deposited on different supports and the structure-
activity correlation of the individual reagents was investigated
in the oxidation of benzylic alcohols for which clayfen and
continued. The dry precipitate was powdered. The nitrate:
support ratio was then 10 mmol nitrate (ion) to 1.8 g support.
Reaction conditions
The oxidation of benzyl alcohol was used as a model.
7
claycop were particularly active. Most bismuth compounds
are relatively non-toxic, readily available at low cost and are
1
1
fairly insensitive to small amounts of water. Acomparison
was made also by the same catalytic systems using several
aluminosilicates including ZSM-5 and BEAzeolites. The main
difference between zeolites and clays is in the size of the pores:
while acid-treated clays are mesoporous (average pore size 4.4
to 7 nm), zeolites are microporous with pore sizes of 0.6 nm for
ZSM-5 and about 0.7 nm for BEA.
It was carried out in a batch reactor and monitored by gas
chromatography using a column consisting of Celite impreg-
nated by a mixture of PEG 1500, SE-30 and THEED in a 2 : 2 : 1
ratio.
As a reaction procedure either (a) or (b) was applied: (a)
.27 g of previously prepared supported reagent was added to
2
Experimental
3
.08 g (1.04 cm , 10 mmol) of alcohol, diluted in 50 cm
3
1
benzene as solvent, at room temperature, followed by heating
Materials
ꢀ
À
to 60 C as reaction temperature—the molar ratio of NO :
3
alcohol ¼ 0.67 : 1, corresponding to the stoichiometry of the
reaction; (b) in the catalytic (in situ) method, the amounts
corresponding to those in method (a) [support 1.2 g, hydrated
metal nitrate (1.08 g, 2.22 mmol) and alcohol] were added to
the benzene solvent at room temperature. The mixture was
then heated to the reaction temperature.
¨
Several members of the K-series from Sud Chemie (Germany)
were used as supports or catalysts: KSF, KSF=0, KP10 and
K10 while HZSM-5 is an industrial solid from MOL Co.
Hungary). BEAzeolite was prepared according to the pro-
(
cedure of Martens et al. with a Si=Al ratio of 7.8. The zeolite
1
2
À1
was converted to a solid acid (acidity: 0.46 meq g ) by
exchange with NH Cl (1 M) cations followed by calcination in
3
For the oxidation of m-phenoxybenzyl alcohol, 2 g (1.74 cm ,
4
ꢀ
10 mmol) alcohol was used under the same conditions. The
reaction was monitored by HPLC (C18 reversed phase col-
umn, methanol : water ¼ 65 : 35 v=v as eluent, UV detection at
oxygen at 350 C for 3 h. The structure of the supported
reagents was characterized by X-ray diffraction and thermal
analysis. The characteristics of the zeolite materials and clays
are reported in Tables 1 and 2, respectively.
2
54 nm).
Structural characterization
Preparation of the supported reagents
TG-DTG-DTAcurves were recorded using a MOM OD-2
derivatograph. A240 mg sample was charged in a platinum
3
Bi(NO ) Á5H O (13.5 g) was added to acetone (187 cm ) in a
3
3
2
ꢀ
À1
0
1
.5 L evaporating flask. The mixture was stirred vigorously for
5 min until complete dissolution of the crystals of the
crucible. The heating rate was 5 C min for the metal nitrates,
À1
ꢀ
either pure or supported, and 10 C min for the supports,
in flowing air atmosphere.
X-Ray powder diffraction patterns were recorded on a HZG-
hydrated metal nitrate. The clay or zeolite sample (15 g) was
added in small amount and stirring continued for another
+
1
suspension under reduced pressure (rotary evaporator) on a
5 min. The solvent was then removed from the resulting
4 type diffractometer, using Cu-K radiation (l ¼ 1.5405 A),
a
Ni filtered, with a scan rate of 1 min
ꢀ
À1
.
ꢀ
ꢀ
water bath, not exceeding 20 C. After the first step of drying,
the dry solid crust adhering to the walls of the flask was flaked
off and crushed with a spatula, and rotary evaporator drying
After a 2 h pretreatment at 300 C in He flow the surface
area and porosity measurements were made by Ar adsorption
ꢀ
at À185 C, following the ASTM D 4222-91 standard.
Table 1 Pore characteristics of zeolites and of supported bismuth
nitrate reagents
Results and discussion
3
Micropore volume=cm g
À1
Comparison of zeolites and clays as supports for solid reagents
Thermoanalytical measurements on the zeolite-supported
ꢀ
Sample
ZSM-5
BEA
reagents show, at different temperatures (159 C for Bi=BEA,
ꢀ
1
46 C for Bi=ZSM-5) and in different intenstities, a very
Zeolite
Bi=zeolite
Bi=zeolite heated to 500 C
0.115
0.030
0.064
0.153
0.021
0.077
exothermic peak in the DTAtrace (Fig. 1), connected with a
weight loss. This effect is so strong in the case of BEAthat
part of the sample was pushed out of the sample holder. The
ꢀ
New J. Chem., 2002, 26, 750–754
751

View Article Online
The K-supported reagents did not show this exothermic
9
,13
decomposition,
presumably because these solids have a
mesoporous structure from which the solvent can evaporate
easily and completely during preparation.
The X-ray diffraction patterns of the K-supported reagents
show that on K10 and KSF—supports having low specific
surface area or low Brꢀnsted acidity (Table 2)—bismuth
nitrate is present in crystalline form [Fig. 3(a)]. On KSF=0 and
KP10, which have higher specific surface area and higher
Brꢀnsted acidity, bismuth nitrate is present in a modified,
partly amorphous phase [Fig. 3(b)]. In this latter case it is
possible that part of the metal nitrate has been hydrolyzed as a
result of its water of crystallization content and the stronger
acidity of the support, producing probably bismuth oxynitrate.
Another difference in the thermoanalytical curves of these
two types of K-supported reagents is that the bismuth nitrate
on K10 and KSF shows a sharp endothermic peak (Fig. 2) at a
ꢀ
temperature (62 and 68 C, respectively) significantly lower
ꢀ
13
than that of bulk bismuth nitrate (75 C). This sharp peak
has been ascribed to the melting of the nitrate, and the
decrease of the melting point was reported in an earlier
1
3
paper. The reagents supported on KSF=0 and KP10 do not
show this endothermic peak probably because some transfor-
mation has already occurred as a result of the higher disper-
sion of the nitrate and the stronger acidity of the support.
Moreover, with KSF=0 the decomposition of the metal nitrate
has already started: during storage at room temperature a
Fig. 1 DTAcurves of zeolite-supported bismuth nitrate reagents.
process is reproducible, only its extent can be different. Such a
phenomena did not occur with clay-supported reagents (Fig. 2).
Areasonable explanation of this difference is that during the
preparation the solvent used could not completely evaporate
from the zeolite structure because of the blockage of the small
pores by solid nitrate. Upon increasing the temperature of the
sample, desorption of the residual amount of acetone occurs
and the organic material burns on the external surface or is
rapidly oxidized by the nitrate. Table 1 clearly shows that the
salt impregnated on the zeolite surface fills the pores but after
ꢀ
heating up to 500 C the pore volume is partly regenerated.
Fig. 3 (a) Comparison of the X-ray diffraction patterns of Bi=KSF
supported reagent (top) with those of Bi(NO
middle) and the KSF clay (bottom). (b) Comparison of the XRD
3
)
3
2
Á5H O (JCPDS-12-148
profile of Bi=KSF=0 supported reagent (top) with the patterns of
Fig. 2 DTAcurves of clay-supported bismuth nitrate reagents.
Bi(NO
3
)
3
2
Á5H O (middle) and the KSF=0 support (bottom).
7
52
New J. Chem., 2002, 26, 750–754

View Article Online
large quantity of nitrous gas was formed. It is interesting to
mention that with KSF, also a strong acid, this decomposition
does not occur: since this solid shows a much lower specific
surface area, this observation suggests that the decomposition
occurs at the interface between the nitrate and the support,
which is small if the crystals of nitrate are large.
Table 5 Yields in m-phenoxybenzaldehyde (%) obtained in the oxida-
tion of m-phenoxybenzyl alcohol with bismuth mitrate, catalyzed by
different aluminosilicates
Reaction time=h
Solid acid
0
.5
1
1.5
2
2.5
3
Oxidation of benzylic alcohols
K10
KP10
KSF
KSF=0
ZSM-5
BEA29
30
23
28
75
22
47
57
37
85
43
65
81
66
94
62
85
87
91
93
89
The two techniques of supported reagent and simple catalysis
have been compared using zeolites and clays for the oxidation
of benzylic alcohols. With clays both the supported reagents
and the catalytic reaction method reach yields above 92%.
Table 3 shows clearly that bismuth nitrate alone hardly reacts
and solid acids are necessary as catalysts for the oxidation. A
good correlation can be found for the mentioned 2 groups of
clay materials between structure and chemical activity in the
oxidation of benzyl alcohol, with both reaction methods.
Tables 3 and 4 show that with K10 the supported reagent as
well as the catalytic system works well and no significant dif-
ference is noticed between the two methods. For the KSF-
based supported reagent, the initial reaction rate is high but
the final conversion is relatively low. In this case the decom-
position of the deposited nitrate is very fast and exothermic
due to the reaction temperature, therefore the reaction is lim-
ited by heat and mass transfer and part of the metal nitrate is
decomposed and does not participate in the oxidation. In the
catalytic reaction technique the bismuth nitrate species occupy
the active sites gradually, the decomposition is much more
uniform and the selectivity toward nitrate is better.
80
87
63
81
95
week of storage. In this case the rate of decomposition of the
nitrate is much higher than the rate of oxidation of benzyl
alcohol.
In the case of zeolites, there is little difference between the
performances of the conventional and the catalytic systems.
Apparently the residual acetone does not play a significant role
in the mechanism, and simply requires more care. There is a
clear difference between the activities of the two zeolites: BEA
is better than ZSM-5, most probably because of its easier
accessibility (larger pore size, very small particles correspond-
ing to a high external surface), and the results with BEAbased
systems are comparable with those of K10 based systems.
10
Corresponding to earlier results, m-phenoxybenzyl alcohol
is well oxidized to aldehyde using different mesoporous clays
as catalysts (Table 5) but it is interesting to notice that this
rather bulky substrate can be readily and selectively oxidized
by also applying the process catalyzed by the microporous
BEA. This clearly evidences that the zeolite acts as a catalyst
for the decomposition of the nitrate. The zeolite-catalyzed
reaction has similar or higher performance than the clay-
catalyzed one also in the oxidation of m-phenoxybenzyl alco-
hol. In this case too, BEAhas higher activity than ZSM-5. The
highest activity of KSF=0 can be explained by the highest
acidity and the greatest pore size of this mesoporous material
In the case of K-supports having the bismuth nitrate in a
partly amorphous form, the initial reaction rate with KP10 is
higher for the conventionally prepared supported reagent and
a strong Brꢀnsted acidity on the support facilitates the nitrate
decomposition. With KSF=0 the supported reagent is much
less active than the catalytic reaction system, since a large
proportion of the bismuth nitrate is decomposed during one
(
Table 2).
Similar oxidation results were obtained with iron(III) nitrate,
Table 3 Yields in benzaldehyde (%) obtained in the oxidation of
benzyl alcohol with bismuth nitrate, catalyzed by different aluminosi-
licates (method b)
concerning both the differences of the K-catalysts and the
applicability of the zeolites.
Reaction time=h
In conclusion, the catalytic reaction technique is a very
useful method for the selective oxidation of benzylic alcohols
to aldehydes using metallic nitrates. Since in this procedure the
solid acid catalyzes the decomposition of the metallic nitrate,
any solid acid can be used; such as strong acid K clays and
large pore zeolites like BEA, which gives yields as high as 97%
in aldehyde in short times.
Solid acid
0.5
1
1.5
2
2.5
23
3
None
K10
KP10
KSF
KSF=0
ZSM-5
BEA67
12
—
40
38
41
13
20
81
81
72
65
30
20
23
97
92
86
77
39
96
91
84
48
99
94
88
52
99
92
57
74
89
92
97
Acknowledgements
J. F. thanks the Foundation Varga Jozsef of the Budapest
´
University of Technology and Economics for a grant. The
financial support of the Hungarian Scientific Research Fund
(OTKAT-029193) and the Ministry of Education (FKFP
0402=1997) is also gratefully acknowledged.
Table 4 Yields in benzaldehyde (%) obtained in the oxidation of
benzyl alcohol with conventional supported bismuth nitrate reagents
prepared from different supports (method a)
This publication was prepared within the framework of the
Hungarian-French intergovernmental scientific and technolo-
gical cooperation, OMFB project no. F-13=99.
Reaction time=h
System
0.5
1
1.5
2
2.5
3
Bi=K10
Bi=KP10
Bi=KSF
Bi=KSF=0
Bi=ZSM-5
Bi=BEA32
—
51
13
3
11
67
76
75
56
11
30
91
85
87
71
15
46
97
90
92
77
19
55
95
95
80
22
60
96
81
26
64
References
1
2
R. A. Sheldon, I. W. C. E. Arends and H. E. B. Lempers, Catal.
Today, 1998, 41, 387.
A. Esposito, C. Neri and F. Buonomo, US Pat. 4,480,135, 1984 to
Anic.
New J. Chem., 2002, 26, 750–754
753

View Article Online
3
4
R. Neumann and M. Levin-Elad, Appl. Catal. A, 1995, 122, 85.
A. Puzari, L. Nath and J. B. Baruah, Indian J. Chem., Sect. A:
Inorg., 1998, 37, 723.
P. Laszlo, Science, 1987, 235, 1473.
T. W. Bastock and J. H. Clark, in Speciality Chemicals, ed. B.
Pearson, Elsevier, London New York, 1991, p. 383.
11 (a) J. Reglinski, in Chemistry of Arsenic, Antimony and Bismuth,
ed. N. C. Norman, Blackie Academic and Professional, New
York, 1998, pp. 403–440; (b) H. Suzuki, T. Ikegami and Y.
Matano, Synthesis, 1997, 249.
12 J. A. Martens, J. Perez-Pariente and P. A. Jacobs, in Chemistry and
Reactivity of Organic and Inorganic Constrained Systems, ed. R.
Setton, NATO ASI Ser., Ser. C, Kluwer, Amsterdam, 1986, vol. 165,
pp. 115–129.
5
6
7
8
9
A. Corne
P. Laszlo and A. Corne
S. Bekassy, T. Cseri, G. Kenessey, G. Pokol, K. Tomor and G.
Liptay, J. T herm. Anal., 1993, 40, 1285.
S. Bekassy, T. Cseri, M. Horvath, J. Farkas and F. Figueras, New
J. Chem., 1998, 22, 339.
´
lis and P. Laszlo, Synthesis, 1985, 909.
´
lis, Aldrichim. Acta, 1988, 21, 97.
´
´
13 T. Cseri, S. Bekassy, G. Kenessey, G. Liptay and F. Figueras,
´ ´
T hermochim. Acta, 1996, 288, 137.
1
0
´
´
´
7
54
New J. Chem., 2002, 26, 750–754