Meerwein-Ponndorf-Verley reduction of cycloalkanones over magnesium-aluminium oxide

Article abstract of DOI:10.1039/b200874m

MgO-Al₂O₃ obtained from layered double hydroxide has been studied as a catalyst in the Meerwein-Ponndorf-Verley (MPV) reduction of cycloalkanones and substituted cyclohexanones in the liquid phase. Conversions for cycloalkanones always exceeded 95% and the selectivity was 100% within thefirst 10h of reaction. In the MPV reduction of 4-tert-butylcyclohexanone to 4-tert-butylcyclohexanol a high stereoselectivity (cis:trans ratio > 12) was obtained. This stereoselectivity is explained by the transition-state selectivity imposed by the adsorption complex. For the reduction of cyclohexanone, a recycling test showed that the catalyst can be reused up to four times without losing more than 10% catalytic activity.

Full text of DOI:10.1039/b200874m

View Article Online / Journal Homepage / Table of Contents for this issue
Meerwein–Ponndorf–Verley reduction of cycloalkanones over
magnesium–aluminium oxide
María A. Aramendía, Victoriano Borau, César Jiménez, José M. Marinas, José R. Ruiz and
Francisco J. Urbano
2
Departamento de Química Orgánica, Universidad de Córdoba, Campus de Rabanales,
Edificio C-3, Ctra. Nnal. IV-A, km 396, 14014 Córdoba, Spain
Received (in Cambridge, UK) 23rd January 2002, Accepted 5th April 2002
First published as an Advance Article on the web 2nd May 2002
MgO–Al2O3 obtained from layered double hydroxide has been studied as a catalyst in the Meerwein–Ponndorf–
Verley (MPV) reduction of cycloalkanones and substituted cyclohexanones in the liquid phase. Conversions for
cycloalkanones always exceeded 95% and the selectivity was 100% within the first 10 h of reaction. In the MPV
reduction of 4-tert-butylcyclohexanone to 4-tert-butylcyclohexanol a high stereoselectivity (cis : trans ratio > 12) was
obtained. This stereoselectivity is explained by the transition-state selectivity imposed by the adsorption complex.
For the reduction of cyclohexanone, a recycling test showed that the catalyst can be reused up to four times without
losing more than 10% catalytic activity.
of magnesium and aluminium oxides with surface basic
properties that can be exploited in a variety of base-catalysed
Introduction
A host of practical applications involving the reduction of car-
processes.13–16
bonyl compounds by catalytic hydrogen transfer as a method
for synthesizing alcohols have been reported in recent times.
If the hydrogen donor used is an alcohol (e.g. propan-2-ol),
the hydrogen transfer process is known as the Meerwein–
Experimental
Ponndorf–Verley (MPV) reduction. The MPV process allows
the highly selective reduction of aldehydes and ketones under
mild reaction conditions. Usually, a metal alkoxide such as
aluminium isopropoxide is used as catalyst for the process.
However, the catalyst must be used in a large excess in order to
obtain acceptable yields and, in addition, eliminating the excess
alkoxide once the reaction is complete can be rather difficult.
Catalysts consisting of chelates of metals such as ruthenium,1,2
iridium3 or even rare earth elements such as samarium4 and
plutonium5 have been tested as substitutes for conventional
metal alkoxides; despite their advantages, however, these alter-
native catalysts are still difficult to remove at the end of the
process and can rarely be reused.
In recent years, a variety of acid and basic heterogeneous
catalysts have been successfully used for this purpose which
avoid the need for removal from the reaction mass. Especially
prominent among such catalysts are magnesium oxide6,7 and
zeolites.8,9 Our research group has successfully used the MPV
reduction of aldehydes of various kinds with basic solid cata-
lysts such as magnesium and calcium oxides10,11 and layered
double hydroxides.12
This paper reports the results obtained in the reduction of
cyclic ketones using secondary alcohols as hydrogen donors
and a heterogeneous catalyst prepared from a magnesium–
aluminium layered double hydroxide.
Layered double hydroxides, also known as hydrotalcite-like
compounds, are anionic clays consisting of brucite-like
[Mg(OH)2] layers where some Mg2ϩ ions are replaced with
Al3ϩ or, in general, a trivalent ion. As a result, the layers
exhibit a charge deficiency that is offset by anions in their
The basic catalyst employed was obtained using a previously
reported method.17 The layered double hydroxide or hydro-
talcite-like compound was prepared from an aqueous solution
containing 0.3 mol of Mg(NO3)2ؒ6H2O and 0.15 mol of
Al(NO3)3ؒ9H2O (Mg : Al ratio = 2). This solution was slowly
added dropwise to 500 ml of an Na2CO3 solution at pH 10 at
60 ЊC with vigorous stirring. The pH was kept constant by
adding appropriate volumes of 1 M NaOH during precipita-
tion. The suspension thus obtained was kept at 80 ЊC for 24 h,
after which the solid was filtered off and washed with 2 l of
de-ionized water. The resulting solid was denoted HT and was
calcined at 500 ЊC for 8 h to obtain a mixture of magnesium
and aluminium oxides that was labelled HT-500 and used as the
catalyst. In all instances, the solid was calcined immediately
prior to use.
In previous work, solids of a similar nature17–19 were charac-
terized in structural, textural and chemical terms, and found to
exhibit surface basic properties.
The catalytic hydrogen transfer reaction was conducted in a
two-necked flask furnished with a condenser and a magnetic
stirrer. The secondary alcohol (0.06 mol) was treated with 0.003
mol of the ketone and the reaction mixture heated at reflux
temperature with stirring (1000 rpm). The reaction was started
by introducing 1 g of freshly calcined catalyst. Aliquots of the
reaction mixture were collected at different times during the
process and analysed by GC–MS using a Hewlett–Packard
HP 5890 GC instrument furnished with a 30 m × 0.32 mm
Supelcowax 10 column and an HP 5971 A MSD detector. The
catalytic activity was calculated from the initial reaction rate
corresponding to the slope of the conversion versus time curve.
The initial reaction rate was calculated at conversion lower than
10% for which the curve is linear.
interlayer spacing. The general formula for this type of
mϪ
compound is [Mg1 Ϫ xAlx(OH)2]xϩ[Ax/m
] ؒnH2O, where A is
the interlayer anion. The calcination of hydrotalcite-like
compounds at temperatures above 400 ЊC produces a mixture
All chemicals used were purchased from Panreac, Merck or
Aldrich and used without further purification.
1122
J. Chem. Soc., Perkin Trans. 2, 2002, 1122–1125
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b200874m

View Article Online
Table 1 Results obtained in the MPV reduction of cyclic ketones with propan-2-ol using catalyst HT-500
Catalytic activity
(mmol cycloalkanol/h)
Ketone
t/h
Yield (%)
Selectivity
Cyclopentanone
Cyclohexanone
Cycloheptanone
10
10
10
94.6
95.0
92.1
100
100
100
0.800
0.719
0.685
Table 2 Results obtained in the MPV reduction of substituted cyclohexanones with propan-2-ol using catalyst HT-500
Catalytic activity
Entry
Ketone
t/h
Yield (%)
Selectivity
cis : trans ratio
(mmol cycloalkanol/h)
1
2
3
4
5
Cyclohexanone
10
24
10
24
24
95.0
97.9
94.7
94.2
42.6
100
100
100
100
100
—
0.719
0.335
0.595
0.430
0.153
4-tert-Butylcyclohexanone
4-Methylcyclohexanone
3-Methylcyclohexanone
2-Methylcyclohexanone
12.83
5.22
4.30
0.95
The study was finished by conducting catalyst reuse tests. The
processes were carried out by filtering the catalyst off after the
reaction and washing it with methanol several times. The cata-
lyst was then calcined at 500 ЊC for 8 h. Fig. 1 shows the results
Results
The MPV reduction of cycloalkanones with propan-2-ol takes
place as depicted in Scheme 1. During the reaction, propan-2-ol
Scheme 1 Reaction scheme for the reduction of cycloalkanones with
propan-2-ol using catalyst HT-500.
is oxidized to acetone and the cyclic ketone is reduced to a
cycloalkanol. Table 1 shows the yield, selectivity and catalytic
activity obtained in the reduction of various cycloalkanones.
As can be seen, the catalyst exhibited similar selectivity and
yield with the three cycloalkanones. In all instances, the cyclo-
alkanol yield after 10 h of reaction exceeded 90% and the selec-
tivity was 100%. The catalytic activity decreased slightly with
increasing ring size, as noted earlier; however, this resulted in no
substantial difference in final yield or selectivity.
Fig. 1 Reuse of the catalyst in the MPV reduction of cyclohexanone
with propan-2-ol: (a) variation of the catalytic activity (in mmol
cyclohexanol/h); (b) percentage variation of the cyclohexanol yield.
obtained. As can be seen, the yield after 10 h of reaction was
scarcely diminished—even after three reuses of the catalyst.
The only apparent effect was a decrease in the catalytic activity.
Let us now consider substituted cyclohexanones, which also
react as depicted in Scheme 1. Table 2 shows the yield and
selectivity for substituted cyclohexanols, as well as the catalytic
activity obtained. The table contains an additional column
listing the cis : trans isomer ratio obtained at the end of the
reaction. Based on these data, the yield for 4-methylcyclo-
hexanone was similar to that for cyclohexanone, even though
the catalytic activity was considerably lower in the former case.
On the other hand, with a much bulkier substituent (e.g. tert-
butyl) at position 4 on the ring, the catalytic activity was nearly
halved relative to cyclohexanone, so the reaction took 24 h to
obtain a similar conversion. The insertion of a methyl group at
position 3 gave rise to results similar to those obtained with a
tert-butyl substituent at position 4. Finally, the presence of
a methyl group at position 2 in the cyclohexanone resulted in
a dramatic decrease in catalytic activity, which was nearly five
times lower than that with unsubstituted cyclohexanone; as a
result, the conversion at 24 h was very low (only 42.6%).
One other important aspect of the reduction of substituted
cyclohexanones is the presence of stereoisomers in the reduc-
tion products. With a methyl substituent, the stereoselectivity
of the process was higher with the group at position 4 than
at position 2. It should be noted that the bulkier tert-butyl
substituent resulted in higher stereoselectivity.
Discussion
In previous work,10,12 we accounted for the catalytic activity
results obtained in the MPV reduction of various aldehydes
with propan-2-ol with a mechanism similar to that reported by
Ivanov et al.,20 based on which the hydrogen transfer occurs
via a concerted process that involves the formation of six-
membered cyclic intermediate with the alcohol and the car-
bonyl compound adsorbed on the catalyst surface (Scheme 2).
Scheme 2 Reaction mechanism for the MPV reaction studied.
The rate-determining step of the process must be related to the
interaction of the alcohol with the acid–base sites, which causes
its dissociation to the corresponding alkoxide. Also, carbonyl
compounds are known to interact with acid and basic sites on
solid surfaces to give condensation reactions. In the proposed
The reduction of cyclohexanone was also performed using
alternative secondary alcohols as hydrogen donors. As can be
seen from Table 3, the yields were excellent. The selectivity was
100% in all instances.
J. Chem. Soc., Perkin Trans. 2, 2002, 1122–1125
1123

View Article Online
Table 3 Results obtained in the MPV reduction of cyclohexanone with alkan-2-ols using catalyst HT-500
Catalytic activity
(mmol alcohol/h)
Entry
Alcohol
T /ЊC
t/h
Yield (%)
1
2
3
4
5
6
Hexan-2-ol
Pentan-2-ol
Butan-2-ol
Propan-2-ol
Hexan-3-ol
Pentan-3-ol
136
118
98
82
135
115
7
8
8.5
10
7
8
97.9
98.0
96.1
95.0
96.2
96.8
3.240
2.249
1.559
0.719
2.744
2.030
scheme, the isopropoxide formed from propan-2-ol transfers a
hydride ion that attacks the carbon atom in the carbonyl group.
The synchronous process illustrated in Scheme 2 yields a new
alcohol and a new ketone as the result of the hydrogen transfer.
This mechanism is quite consistent with the results obtained
in the reduction of cyclic ketones (Table 1). In fact, increasing
the ring size in the ketone from five to seven atoms should result
in no significant changes in the complex that is adsorbed on the
catalyst other than steric hindrance; based on the results, the
effect must be insubstantial as it only caused a slight decrease in
catalytic activity and the yield at 10 h was close to 95% in the
three cases. On the other hand, the substituted cyclohexanones
exhibit marked differences as regards the adsorbed complex
that result in substantial differences in both catalytic activity
and yield (Table 2). In addition, the process produces stereo-
isomers. However, as shown below, the proposed model also
accounts for these results.
The most immediate inference from the data in Table 2 is that
introducing a methyl group onto position 4 in cyclohexanone
(entry 3) seemingly has no decisive influence on the final reac-
tion yield relative to unsubstituted cyclohexanone; in fact, the
yield at 10 h was ca. 95% in both cases. When the substituent at
position 4 is much bulkier (e.g. tert-butyl, entry 2), the catalytic
activity decreases by more than 50% with respect to cyclo-
hexanone. The reaction yield is also strongly affected; in fact,
the reaction must be extended to 24 h to obtain results similar
to those provided by cyclohexanone after 10 h. This suggests
the presence of strong steric hindrance in the adsorbed
complex, the effect obviously being much greater with 4-tert-
butylcyclohexanone than with 4-methylcyclohexanone. In
order to examine further the effect of steric hindrance, we
conducted MPV reductions of 3- and 2-methylcyclohexanone
(entries 4 and 5 in Table 2). A comparison of the results for the
three methyl-substituted cyclohexanones reveals decreased
yield and catalytic activity for the 4-substituted ketone in rela-
tion to the 2-substituted compound; this is consistent with the
proposed mechanism since the presence of a substituent at
position 2 must severely hinder the formation of the adsorbed
complex (so much so that the yield at 24 h was only 42.6%).
As noted earlier, the MPV reduction of substituted cyclo-
hexanones produces stereoisomers of the reaction product.
Accordingly, the next step in this work was to check whether
the proposed mechanism would account for the experimental
stereoselectivity results. Based on the data of Table 2, the 4-
substituted cyclohexanones exhibited the best stereoselectivity
results (particularly that with the bulky tert-butyl group). Based
on the proposed mechanism, 4-tert-butylcyclohexanone can be
adsorbed in the two ways shown in Scheme 3, which lead to
a major isomer (cis) and a minor isomer (trans), respectively.
The adsorbed complex that yields the trans isomer exhibits a
stronger interaction between the methyl groups in propan-2-ol
and the tert-butyl substituent than does the adsorbed complex
leading to the cis isomer. As a result, 4-tert-butylcyclohexanol
is obtained in a high cis : trans ratio. This ratio decreases on
replacing the bulky tert-butyl group with a methyl group, which
substantially decreases steric hindrance (and the cis : trans ratio
for 4-methylcyclohexanol as a result). The cis : trans ratio for
the 3- and 2-substituted methylcyclohexanones is even lower, so
much so that 2-methylcyclohexanol is obtained as a virtually
Scheme 3 Transition states in the formation of cis-4-tert-butyl-
cyclohexanol and trans-4-tert-butylcyclohexanol.
equimolar mixture of both stereoisomers. This is consistent
with the proposed model since the two adsorbed complexes
for 2-methylcyclohexanone are similar and lead to both
stereoisomers.
These results can also be explained from stereochemical
considerations such as the fact that the preference for the
equatorial attack, leading to the axial cis-alcohol, is favoured to
reduce the interaction of the axial hydrogens in positions 3 and
5 with the incoming alcohol, a type of 1,3-diaxial interaction.
The nature (tert-butyl or methyl) and the position (3 or 4) of the
substituent modify the geometry and induce changes in the
selectivity (the percentage of equatorial alcohol changes from 7
to 19%). When the substituent is in position 2 the situation is
completely different, in this case, in the equatorial approach
there is a steric interaction between the incoming alcohol and
the methyl group in the equatorial position, so that both
approaches have similar interactions and the stereoselectivity is
very small. The presence of strong steric requirements in both
axial and equatorial approaches accounts for the low yield
obtained in this case.
As can be seen from Table 3, which shows the results
obtained in the reduction of cyclohexanone with different sec-
ondary alcohols, all provided excellent results as regards both
yield and catalytic activity. In the alkan-2-ol series, conversion
decreased in the following sequence: hexan-2-ol > pentan-2-ol
> butan-2-ol > propan-2-ol. Accordingly, hexan-2-ol would be
the best choice; however, because the reaction temperature
would be nearly 60 ЊC higher than that with propan-2-ol,
additional thermal energy would have to be expended, not only
during the process but also afterwards to remove the excess
alcohol, and the hexan-2-one produced. Similar considerations
apply to alkan-3-ols (hexan-3-ol and pentan-3-ol). As can be
seen, there was no significant difference between the results
provided by secondary alcohols with the OH group at position
2 or 3.
Finally, the reuse tests performed revealed that the catalyst
can be recycled at least four times without losing more than
10% catalytic activity. Fig. 1 shows the conversion and catalytic
1124
J. Chem. Soc., Perkin Trans. 2, 2002, 1122–1125

View Article Online
7 J. Kijenski, M. Glinski and J. Czarnecki, J. Chem. Soc., Perkin Trans.
2, 1991, 1695.
8 J. C. Van der Waal, P. J. Kunkeler, K. Tan and H. Van Bekkum,
J. Catal., 1998, 173, 74.
9 F. Quignard, O. Grazziani and A. Choplin, Appl. Catal. A: Gen.,
1999, 182, 29.
activity obtained. In addition to the ease with which the reac-
tion product can be isolated, this process therefore has two
other substantial advantages over homogeneous catalysis: the
ease of recovering the catalyst and the ability to reuse it.
10 M. A. Aramendía, V. Borau, C. Jiménez, J. M. Marinas, J. R. Ruiz
and F. J. Urbano, J. Mol. Catal. A: Chem., 2001, 171, 153.
11 M. A. Aramendía, V. Borau, C. Jiménez, J. M. Marinas, J. R. Ruiz
and F. J. Urbano, J. Colloid Interface Sci., 2001, 238, 385.
12 M. A. Aramendía, V. Borau, C. Jiménez, J. M. Marinas, J. R. Ruiz
and F. J. Urbano, Appl. Catal. A: Gen., 2001, 206, 95.
13 B. M. Choudary, M. Lakshmi Kantam, B. Kavita, Ch. Venkat
Reddy and F. Figueras, Tetrahedron, 2000, 56, 9357.
14 I. Rodriguez, S. Iborra, F. Rey and A. Corma, Appl. Catal. A: Gen.,
2000, 194, 241.
Acknowledgements
The authors gratefully acknowledge funding from the Spanish
Ministry of Science and Technology and the Plan Nacional de
Investigación, Desarrollo e Innovación Tecnológica (Project
BQU-2001–2605), as well as from the Consejería de Educación
y Ciencia de la Junta de Andalucía.
15 M. Sychev, R. Prihod’ko, K. Erdmann, A. Mangel and R. A. van
Santen, Appl. Clay Sci., 2001, 18, 103.
16 M. J. Climent, A. Corma, S. Iborra and J. Primo, J. Catal., 1995,
151, 60.
References
1 D. G. I. Petra, J. N. H. Reek, P. C. J. Schoemaker and P. W. N. M.
Leeuwen, Chem. Commun., 2000, 683.
2 A. Aranyos, G. Csjernyick, K. J. Szabo and J.-E. Bäckvall, Chem.
Commun., 2000, 683.
3 D. Maillard, C. Nguefack, G. Pozzi, S. Quici, B. Balada and
D. Sinon, Tetrahedron: Asymmetry, 2000, 11, 2821.
4 G. A. Molander and J. A. Mckie, J. Am. Chem. Soc., 1993, 115,
5821.
5 B. P. Warner, J. A. D’Alessio, A. N. Morgan, C. J. Burns,
A. R. Schake and J. G. Watkin, Inorg. Chim. Acta, 2000, 309, 45.
6 G. Szöllösi and M. Bartok, Appl. Catal. A: Gen., 1998, 169, 263.
17 M. A. Aramendía, Y. Avilés, V. Borau, J. M. Luque, J. M. Marinas,
J. R. Ruiz and F. J. Urbano, J. Mater. Chem., 1999, 9, 1603.
18 M. A. Aramendía, Y. Avilés, J. A. Benítez, V. Borau, C. Jiménez,
J. M. Marinas, J. R. Ruiz and F. J. Urbano, Microporous
Mesoporous. Mater., 1999, 29, 319.
19 M. A. Aramendía, V. Borau, C. Jiménez, J. M. Marinas, J. R. Ruiz
and F. J. Urbano, Mater. Lett., 2000, 46, 309.
20 V. A. Ivanov, J. Bachelier, F. Audrey and J. C. Lavalley, J. Mol.
Catal., 1994, 91, 45.
J. Chem. Soc., Perkin Trans. 2, 2002, 1122–1125
1125