Mechanism of Homogeneously and Heterogeneously Catalysed Meerwein-Ponndorf-Verley-Oppenauer Reactions for the Racemisation of Secondary Alcohols

Article abstract of DOI:10.1002/chem.200305460

The mechanism of hydrogen transfer from alcohols to ketones, catalysed by lanthanide(III) isopropoxides or zeolite Beta has been studied. For the lanthanide catalysed reactions, (S)-1-phenyl-(1-²H ₁)ethanol and acetophenone were used as case studies to determine the reaction pathway for the hydrogen transfer. Upon complete racemisation all deuterium was present at the 1-position, indicating that the reaction exclusively takes place via a carbon-to-carbon hydrogen transfer. Zeolite Beta with different Si/Al ratios was applied in the racemisation of (S)-1-phenylethanol. In this case the racemisation does not proceed via an oxidation/reduction pathway but via elimination of the hydroxy group and its readdition. This mechanism, however, is not characteristic for all racemisation reactions with zeolite Beta. When 4-tert-butyl cyclohexanone is reduced with this catalyst, a classical MPV reaction takes place exclusively. This demonstrates that zeolite Beta has a substrate dependent reaction pathway.

Full text of DOI:10.1002/chem.200305460

FULL PAPER
Mechanism of Homogeneously and Heterogeneously Catalysed
Meerwein Ponndorf Verley Oppenauer Reactions for the
Racemisation of Secondary Alcohols
Dirk Klomp, Thomas Maschmeyer, Ulf Hanefeld, and Joop A. Peters*[a]
Abstract: The mechanism of hydrogen
transfer from alcohols to ketones, cata-
lysed by lanthanide(iii) isopropoxides
or zeolite Beta has been studied. For
the lanthanide catalysed reactions, (S)-
1-phenyl-(1-2H1)ethanol and acetophe-
none were used as case studies to de-
termine the reaction pathway for the
hydrogen transfer. Upon complete rac-
emisation all deuterium was present at
the 1-position, indicating that the reac-
tion exclusively takes place via
a
misation does not proceed via an oxi-
dation/reduction pathway but via elimi-
nation of the hydroxy group and its re-
addition. This mechanism, however, is
not characteristic for all racemisation
reactions with zeolite Beta. When 4-
tert-butyl cyclohexanone is reduced
with this catalyst, a classical MPV reac-
tion takes place exclusively. This dem-
onstrates that zeolite Beta has a sub-
strate dependent reaction pathway.
carbon-to-carbon hydrogen transfer.
Zeolite Beta with different Si/Al ratios
was applied in the racemisation of (S)-
1-phenylethanol. In this case the race-
Keywords: hydrogen transfer
lanthanides ¥ Meerwein Ponndorf
¥
Verley reduction
¥
Oppenauer
oxidation ¥ zeolites
Introduction
dation. Several reviews on the Meerwein Ponndorf Verley
10]
and Oppenauer (MPVO) reactions have been published.[7
The oxidation of alcohols and the reduction of ketones has
always been a mainstay of organic chemistry. Recently, the
racemisation of alcohols via the corresponding ketone has
attracted considerable attention. This racemisation in combi-
nation with an enzyme catalysed resolution is the backbone
of many dynamic kinetic resolutions.[1]
In 1925, Meerwein and Schmidt reported a mild method
for the reduction of ketones and aldehydes by alcohols with
metal alkoxides as the catalysts.[2] Verley,[3] Ponndorf[4] and
Lund[5] independently investigated the scope of this reac-
tion. Twelve years later, Oppenauer recognised the possibili-
ty to reverse the reaction and utilised it as an oxidation.[6]
Ever since, the Meerwein Ponndorf Verley (MPV) reduc-
tion and the Oppenauer oxidation have been textbook ex-
amples of highly selective and efficient reactions under mild
conditions. Usually, readily available reductants such as eth-
anol or 2-propanol are used as hydrogen donor in the MPV
reduction, whereas oxidants such as acetone or cyclohexa-
none are used as hydrogen acceptor in the Oppenauer oxi-
The mechanism of the MPV reaction is generally believed
to involve a six-membered cyclic transition state in which
both the substrate and the reductant/oxidant are coordinat-
ed to the metal ion (Scheme 1). A similar bimolecular reac-
tion between a coordinated alcohol and a coordinated
ketone has been proposed as well.[11] As an exception, a
single electron transfer pathway was observed for alkali
metal ions catalysed MPV reductions.[12] Metal hydrides are
only formed during hydrogenations catalysed by transition
metals such as ruthenium.[13] In addition, a few examples are
known of MPV alkynylations and cyanations.[14,15] In these
cases, the transfer mechanism is reported to be similar to
the mechanism of the hydride transfer as depicted in
Scheme 1.
Scheme 1. Generally accepted mechanism of the MPV reduction and Op-
penauer oxidation.
[a]D. Klomp, Prof. Dr. T. Maschmeyer, Dr. U. Hanefeld, Dr. J. A. Peters
Laboratory of Applied Organic Chemistry and Catalysis
Delft University of Technology
In the MPV reduction, the ketone or the aldehyde coordi-
nates to the metal centre of the catalyst, which causes the
activation of the double bond (Scheme 1). A hydride shift
takes place and the alkoxide is released as ketone or alde-
Julianalaan 136
2628 BL Delft (The Netherlands)
Fax : (+31)-15-2784289
E-mail: J.A.Peters@TNW.TUDelft.NL
2088
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200305460
Chem. Eur. J. 2004, 10, 2088 2093

2088 2093
hyde. Alkoxides can exchange with alcohols present in solu-
tion, so an equilibrium of various metal alkoxides may exist
during the reaction.
metals. Several examples are known and the reduction or
24]
oxidation can be performed either in the gas phase[22
or
29]
in solution.[25
In the latter case stereoselectivity has also
In 1953, deuterium tracer studies of aluminium(iii) iso-
propoxide catalysed reactions were reported that supported
the mechanism proposed.[16] However, the experimental
methodology available at that time was rather inaccurate.
The authors estimated the error margin in their procedure
for the deuterium determination (falling-drop method[17]) to
be 10%.
Traditionally, MPVO reactions are performed with stoi-
chiometric amounts of AlIII alkoxides. The reactions were
improved considerably by the introduction of LnIII alkoxides
as the catalyst.[18] With these catalysts, only catalytic
amounts of this metal ion are required. Recently, much
higher reactivities for aluminium catalysed MPVO reactions
have been achieved with dinuclear AlIII complexes[19] and
with AlIII alkoxides generated in situ.[20] Most likely, the rel-
atively high activity of the LnIII ions can be ascribed to the
high ligand exchange rates with these ions.
Recently, doubts were raised whether the mechanism pro-
ceeds exclusively via a carbon-to-carbon hydrogen transfer
(Scheme 2, mechanism A).[21] During racemisation of (S)-1-
phenyl-(1-2H1)ethanol with samarium(iii) or aluminium(iii)
isopropoxide in the presence of acetophenone, part of the
deuterium atoms disappeared from the 1-position. This was
suggested to be due to the occurrence of a second pathway
involving an oxygen-to-carbon hydrogen transfer (Scheme 2,
mechanism B).
been reported. It was proposed that the reaction proceeds
within the pores of the zeolite via the same mechanism as in
the homogeneous case (Scheme 1). However, no conclusive
experimental evidence was given. The heterogeneous cata-
lysts have the advantage that the work-up of the reaction
mixture is easier. Furthermore, they are less water sensitive
than the homogeneous MPVO catalysts. The MPVO reac-
tions can be exploited for the racemisation of alcohols via
Oppenauer oxidation of the alcohol followed by non-selec-
tive reduction of the resulting ketone by the Meerwein
Ponndorf Verley reaction.[30] Obviously, racemisation reac-
tions can be performed using relatively small amounts of a
ketone. Zeolite Beta has also been used for the racemisation
of secondary phenylic alcohols in a dynamic kinetic resolu-
tion, however, in this case water elimination/addition via a
carbenium ion is involved rather than a redox mechanism.[31]
Here, we report on a study of the racemisation of alcohols
using both homogeneously and heterogeneously catalysed
MPVO reactions. The reaction mechanisms of these reac-
tions were established using deuterium and 17O labelling
techniques.
Results and Discussion
The mechanism of the MPVO reactions was studied by
monitoring the racemisation of (S)-1-phenylethanol [(S)-1]
in the presence of the hydrogen acceptor acetophenone (2)
or acetone either catalysed by a lanthanide(iii) isopropoxide
or zeolite Beta. The fate of the various hydrogen atoms
A different class of catalysts for the MPVO reaction con-
sists of zeolite Beta and its derivatives containing other
2
during these reactions was traced by means of H-substitu-
tion on the tertiary carbons of 1 and of the isopropoxide
ligand of the complex. Furthermore, some reactions were
performed with 1, in which the methyl group was labeled
Scheme 2. Two possible pathways for the hydrogen transfer.
2
with H.
Synthesis of isotopically substituted compounds: (S)-1-
phenyl-(1-2H1)ethanol ((S)-[D1]-1) was prepared by reduc-
tion of acetophenone (2) with LiAlD4 followed by a kinetic
resolution of the product alcohol 1 utilising Candida antarc-
tica lipase B (CAL-B) and isopropenyl acetate. The enantio-
pure alcohol (S)-[D1]-1 and acetate (R)-[D1]-3 were ob-
tained in good yields (both 39% in two steps) and excellent
optical purity (both ee>99%) (Scheme 3).
Abstract in Dutch: Het mechanisme van de waterstofover-
dracht van alcoholen naar ketonen, gekatalyseerd door lan-
thanide(iii) isopropoxides of zeoliet Beta, werd onderzocht.
Voor de lanthanide gekatalyseerde reacties is gebruik ge-
maakt van (S)-1-fenyl-(1-2H1)ethanol en acetofenon om het
reactiepad van deze overdracht vast te stellen. Na volledige
racemisatie bevond alle deuterium zich op de 1-positie. Dit
toont aan dat de overdracht van waterstof alleen van koolstof
naar koolstof plaatsvindt. In de zeoliet Beta gekatalyseerde
racemisaties van (S)-1-fenylethanol is gewerkt met verschil-
lende Si/Al verhoudingen in de zeoliet. Het bleek dat de reac-
tie niet verloopt volgens een oxidatie/reductie mechanisme,
maar dat er een eliminatie en readditie van de hydroxygroep
plaatsvindt. Dit mechanisme is echter niet kenmerkend voor
alle racemisaties met zeoliet Beta. Gedurende de reductie van
4-tert-butylcyclohexanon, een stof die minder makkelijk de-
protoneert, vindt een klassieke MPV reactie plaats. Het reac-
tiepad dat wordt gevolgd tijdens de reductie door zeoliet Beta
is dus substraatafhankelijk.
1-Phenyl-(2-2H3)ethanol ([D3]-1) was synthesised via a
straightforward base catalysed H/D exchange reaction of
acetophenone (2) in D2O. Subsequent NaBH4 reduction
gave the desired racemic compound [D3]-1, which was sub-
mitted to a CAL-B catalysed kinetic resolution (Scheme 3).
The alcohol (S)-[D3]-1 and acetate (R)-[D3]-3 were obtained
in three steps in 38 and 39% yield, respectively, and excel-
lent optical purity (both ee >99%).
Deuterated catalysts were obtained by ligand exchange
reactions of samarium(iii) isopropoxide (4) with the isopro-
panol labeled with 2H at the appropriate position(s)
(Scheme 4).
2089
Chem. Eur. J. 2004, 10, 2088 2093
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

J. A. Peters et al.
FULL PAPER
of 1, deuterium present on the 2-position of the isopropox-
ide groups is transferred to the 1-position of the alcohol.
Complete scrambling of the deuterium atoms over the terti-
ary carbons of the isopropoxy groups of the catalyst and the
1-position of phenylethanol would have resulted in 23% of
2H on the latter position. This confirms that the racemisa-
tion takes place exclusively via the mechanism depicted in
Scheme 1. Remarkably, a substantial amount of deuterium
was found in the methyl group of phenylethanol in each of
the experiments. This can be ascribed to H/D exchange be-
tween the methyl groups of acetophenone (formed by Op-
penauer oxidation of 1) and those of acetone (formed by ox-
idation of the isopropoxy groups of the catalyst) via keto
enol equilibria, under influence of the alkaline conditions in
the reaction mixture.
Analoguous experiments were performed using [D1]-4.
The 2H distribution at the 1-position in the products was
similar to that with the completely deuterated catalyst [D7]-
4. The methyl group, however, did not show any incorpora-
tion of deuterium, supporting the keto enol equilibria pro-
posed above for the H/D exchange between acetone and
acetophenone. This also rules out that any D from CD3 is
transferred to the 1-position of the alcohol.
Scheme 3. Synthesis of enantiopure deuterium substituted alcohols.
A reaction performed without hydrogen acceptor (ace-
tone or acetophenone) did not show any activity. This indi-
cates that the reaction has to proceed via the reduction of a
ketone or aldehyde.
Scheme 4. Synthesis of the deuterated samarium(iii) catalysts through
All results described above strongly support that the
MPVO reaction exclusively proceeds via mechanism A
(Scheme 2). This is also in agreement with results reported
by P‡mies and B‰ckvall,[21] who observed 24% loss of deu-
terium upon racemisation of (S)-1-phenyl-(1-2H1)ethanol
using 10 mol% of a Sm-isopropoxide catalyst without deute-
rium. This is very close to the theoretical loss of 23% of
deuterium in phenylethanol upon complete scrambling as
mentioned above. The reason for the deuterium loss in the
product, is the ligand exchange that takes place in the reac-
tion mixture, during which the isopropoxy groups of the cat-
alyst will exchange against the reactant alcohol (Scheme 5).
ligand exchange.
Catalytic reactions with samarium(iii) isopropoxide: The
racemisations of (S)-[D1]-1 and (S)-1 were performed with
an equimolar amount of acetophenone (2) and 10 mol% of
[D7]-4 or [D1]-4 as the catalyst. THF was used as a solvent
to enable the direct comparison with the results published
by P‡mies and B‰ckvall.[21] In addition, some experiments
were performed with heptane as the solvent, which has the
advantage that it does not coordinate to the LnIII ion. The
Table 1. Deuterium contents after complete racemisation of (S)-1 and
(S)-[D1]-1.[a]
Entry
Alcohol
Catalyst
Solvent
% 2H at % 2H at Me[b]
1-position[b]
1
2
3
4
(S)-[D1]-1
(S)-1
(S)-[D1]-1
(S)-1
[D7]-4
[D7]-4
[D7]-4
[D7]-4
THF
THF
heptane
heptane
>99
15
16
17
20
11
>99
15
[a]Alcohol (0.5 mmol),
2 (0.5 mmol), catalyst (0.05 mmol), solvent
(0.75 mL), T=708C, 18 h. [b]Calculated from NMR and MS.
reactions were run for 18 h (Table 1).
Entries 1 and 3 show complete retention of deuterium at
the a-position with respect to the hydroxyl group. This
strongly supports a direct transfer of deuterium from carbon
to carbon (Scheme 2, mechanism A) and excludes the occur-
rence of a reduction via transfer of H/D from the alcohol
function to the carbonyl carbon atom (Scheme 2, mecha-
nism B). Similar experiments with unlabeled starting com-
pound (entries 2 and 4) show that during the racemisation
Scheme 5. Ligand exchange and subsequent reactions in the MPVO cycle.
2090
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2088 2093

Meerwein Ponndorf Verley Oppenauer Reactions
2088 2093
Catalytic reactions with zeolite H-Beta: Zeolite Beta is also
known to oxidise alcohols and to reduce ketones in a similar
29]
fashion as the homogeneous LnIII catalysts.[25
some mechanistic studies are reported.
Herein,
Zeolite H-Beta, activated at 4008C, was used as catalyst
and acetone was applied as hydrogen acceptor. The experi-
ments were carried out at 508C, in order to minimise the
loss of acetone from the reaction mixture. The reactions
were terminated after one hour, after which the enantiomer-
ic excess and the mass balance were determined (Figure 1).
Zeolite Beta catalysed the racemisation, even in the ab-
sence of an oxidant. This is in contrast to the findings with
samarium(iii) isopropoxide. In most cases, this racemisation
reaction was even faster than in the presence of acetone.
Furthermore, when the reaction was performed in the pres-
ence of acetone no acetophenone could be detected in the
reaction mixture. Acetophenone would be the reaction in-
termediate when the racemisation proceeds via an oxida-
tion/reduction of the alcohol and, therefore, it can be con-
cluded that the reaction proceeds via a different pathway.
The incomplete mass balance as determined by GC analysis
also supports the presence of another pathway differing
from the redox reactions. The phenomena observed can be
rationalised by the occurrence of an elimination/addition
mechanism in which the hydroxy group is eliminated from
the alcohol, forming a carbocation intermediate. This inter-
mediate gives upon rehydration phenylethanol. However,
instead of an SN1 reaction an E1 elimination takes place and
styrene is formed. Indeed, when a small quantity of H217O
was added to the reaction mixture, 17O was found to be in-
corporated into the produced racemic alcohol. The occur-
rence of the carbocation and subsequent formation of sty-
rene also explains the incomplete mass balance, since sty-
rene can polymerise and the polymer escapes detection by
GC.
Figure 1. Zeolite Beta catalysed racemisations of (S)-1-phenylethanol in
*
the presence and in the absence of acetone. & ee, with 1 equiv acetone;
&
ee, without acetone; mass balance, with 1 equiv acetone; // mass bal-
ance, without acetone. [a]Determined by chiral GC.
Scheme 6. The reduction of 4-tert-butylcyclohexanone and the oxidation
of 4-tert-butylcyclohexanol.
For both catalysts complete conversion of 6 was observed
after 18 h. From Table 2 it is evident that the deuterium that
was present at the 2-position of isopropanol is fully transfer-
red to the 2-position of the product alcohol; this indicates a
hydrogen transfer according to Scheme 2, mechanism A, in-
dependent of the activation temperature. Furthermore, no
loss in the mass balance was observed and no deuterium
was detected at other positions of the product, proving that
here exclusively a direct carbon-to-carbon hydrogen transfer
takes place (see Scheme 2, mechanism A).
The occurrence of a carbocation intermediate was con-
firmed by reactions of styrene and (S)-1-phenylethanol (1)
with one equivalent of water in the presence of zeolite Beta
(Si/Al=16). Styrene showed a very slow reaction to 1-phe-
nylethanol, only 6% of it was formed after 20 h. During the
same reaction time, however, (S)-1-phenylethanol was com-
pletely racemised; in this case styrene could be detected by
GC and as expected the mass balance was incomplete.
As can be seen in Figure 1 the acidity of the zeolite,
which depends on the Si/Al ratio, determines the racemisa-
tion rate. The highest rate was obtained with a zeolite with
Si/Al=16.
Table 2. Deuterium contents of 7 after reduction of 6.[a]
Entry
Ketone
Catalyst[b]
Solvent
% 2H at
1-position[c]
1
2
6
6
Beta 16 (400)
Beta 16 (600)
2-(2-2H1)propanol
2-(2-2H1)propanol
>99
>99
[a]Ketone (2.0 mmol), catalyst (30 mg), solvent 3 mL, T=508C, 18 h.
[b]The first number is the Si/Al ratio, the number in parentheses is the
activation temperature in 8C, activation overnight. Prepared according to
van Bekkum et al.[35] [c]Calculated from NMR and MS, 3 5% of the
trans isomer was found, also with >99% 2H at the 1-position.
Since phenylethanol is very susceptible to dehydration,
zeolite Beta catalysed reactions were also performed with 4-
tert-butylcyclohexanone (6).
It is remarkable that the Beta zeolite catalyses both the
elimination/addition and the MPVO reaction. In both cases
the zeolite is activated at 4008C in order to expel water and
to increase the number of Lewis acid sites. Apparently, with
exactly the same catalyst it is possible to follow two differ-
ent pathways. The elimination is most likely catalysed by
the Br˘nsted acid sites whereas the MPV reduction is car-
ried out at the Lewis acid sites. Corma et al. showed that
with Beta resembling zeolites, like Sn-Beta, where only
Lewis acid sites are available in the catalyst, it is possible to
reduce 2 without any elimination.[29]
From previous work it is known that zeolite Beta prefera-
bly catalyses the reduction towards the cis-4-tert-butyl cyclo-
28]
hexanol (7).[25 In oxidations cis-isomer 7 is converted into
ketone 6, whereas the other isomer of the alcohol is rather
unreactive. Therefore, we monitored the reduction of the
ketone rather than the epimerisation of the two alcohols
(Scheme 6). Zeolite Beta, activated at different tempera-
tures, was applied since it has been shown that the activation
temperature plays a crucial role in the catalytic activity.[24,25]
2091
Chem. Eur. J. 2004, 10, 2088 2093
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

J. A. Peters et al.
FULL PAPER
wax 52 CB, on column injector at 608C, a Flame Ionisation Detector at
2508C and nitrogen as carrier gas. Retention times (min) (0 2 min:
1008C; 2 18 min: 58CminÀ1; 18 20 min: 1808C): 1,3,5-triisopropylben-
zene (10.7), 4-tert-butylcyclohexanone (15.5), cis-4-tert-butylcyclohexanol
(16.4), trans-4-tert-butylcyclohexanol (17.5). NMR spectra were recorded
on a Varian VXR-400S or a Varian Unity Inova-300 spectrometer at
258C. Chemical shifts are reported in ppm with TMS as an internal stan-
dard (d=0 ppm). The deuterium content of the isotopically substituted
Conclusion
The MPVO reaction catalysed by lanthanide(iii) isopropox-
ides proceeds exclusively via a hydride transfer from the
carbon at the a-position with respect to the alcohol function
to the carbonyl carbon of the ketone or the aldehyde. Here,
for the first time the mechanism was proven in a reliable
and undisputable way using deuterium substituted alcohols
and catalysts. Furthermore, it was observed that a keto enol
tautomerisation takes place during the reaction through
which hydrogen atoms other than in the 1-position can be
exchanged.
The racemisation of (S)-1 catalysed by zeolite Beta pro-
ceeds, (predominantly) via an elimination/addition mecha-
nism. 17O studies support this observation. Furthermore, the
rate of racemisation and the extent of oligomerisation in the
reaction is determined by the acidity of the zeolite and the
number of active sites. On the other hand, zeolite Beta pos-
sesses the capacity to reduce ketones also via a regular re-
duction pathway as is shown by the reduction of 7. This is
summarised in Scheme 7. The preference for either of the
pathways in racemisations depends on the substrate used
and is not influenced by the activation temperature and
therefore the number of Lewis acid sites of zeolite Beta.
2
compounds was calculated from the 1H- and H NMR spectra. Mass spec-
tra were recorded with a VG SE spectrometer at 70 eV. Elemental analy-
sis ICP-OES was performed with Perkin Elmer Optima 4300DV. Immo-
bilised Candida antarctica Lipase B (CAL-B) as Chirazyme L2, c-f. C2,
lyo was a gift from Roche Diagnostics. The enzyme activity was deter-
mined following a standard procedure.[32] Prior to use, the enzyme was
dried overnight under vacuum over silica in a desiccator. For column
chromatography Fluka silica gel 60 was used and Merck aluminium
sheets with silica gel 60 F254 were used for TLC. Elution was carried out
with mixtures of petroleum ether 40 658C (PE) and diethyl ether
(Et2O).
1-Phenyl-(1-2H1)ethanol ([D1]-1): At À208C, lithium aluminium deuter-
ide (5 g, 119 mmol) was dissolved in diethyl ether (80 mL). Freshly distil-
led acetophenone (2) (22.6 mL, 192 mmol) was added dropwise to the
stirred mixture. The reaction mixture was allowed to warm to room tem-
perature after which the stirring was continued for 2 h. The temperature
was lowered again to À208C. Subsequently, water (5 mL), 15% aqueous
NaOH solution (5 mL) and water (15 mL) were carefully added.[33] The
solids were filtered off and washed thoroughly with diethyl ether. Distil-
lation under reduced pressure yielded [D1]-1 as a colourless oil (20.8 g,
169 mmol, >99% deuterated, 88%). B.p. 988C at 30 mbar; 1H NMR
(300 MHz, CDCl3): d=1.49 (t, J=0.90 Hz, 3H, CH3), 1.84 (s, 1H, OH),
7.24 7.39 (m, 5H, ar-H).
(S)-1-Phenyl-(1-2H1)ethanol ((S)-[D1]-1) and (R)-(1-2H1)-1-methylbenzyl
acetate ((R)-[D1]-3): [D1]-1 (2.0 g, 16.3 mmol) and isopropenyl acetate
(1.9 mL, 16.3 mmol) were dissolved in toluene (20 mL). The temperature
was raised to 608C and CAL-B (0.4 g, 1.5 kU) was added to the mixture.
The reaction was monitored by chiral GC. After 2 h the enzyme was re-
moved by filtration and the solution was concentrated in vacuo. The two
products were separated by column chromatography (PE/Et2O 4:1, after
the acetate was eluted: PE/Et2O 1:1). After removal of the solvents alco-
hol (S)-[D1]-1 (0.84 g, 6.8 mmol, ee>99%, >99% deuterated, 42%) and
acetate (R)-[D3]-3 (1.18 g, 7.1 mmol, ee>99%, >99% deuterated, 44%)
were obtained. The 1H NMR spectrum of the alcohol was identical to
that of [D1]-1. (R)-[D3]-3: 1H NMR (300 MHz, CDCl3): d=1.53 (t, J=
0.82, 3H, CD-CH3), 2.06 (s, 3H, C(O)CH3), 7.23 7.39 (m, 5H, ar-H).
(S)-1-Phenyl-(2-2H3)ethanol ((S)-[D3]-1) and (R)-1-(2H3)methylbenzyl
acetate ((R)-[D3]-3): Acetophenone (2) (2.0 mL, 17.1 mmol), D2O
(20 mL, 1.0 mol) and a catalytic amount of potassium carbonate were
stirred vigorously overnight at 1008C. The mixture was cooled to room
temperature and NaBH4 (324 mg, 8.6 mmol) was added. After 3 h, the re-
action mixture was extracted with diethyl ether (3î20 mL), the com-
bined organic layers were dried over MgSO4 and concentrated in vacuo.
The residue was dissolved in toluene (20 mL). CAL-B (206 mg, 0.7 kU)
and isopropenyl acetate (2.0 mL, 18.2 mmol) were added and the temper-
ature was raised to 508C. The reaction was followed by chiral GC. After
3 h the enzyme was filtered off and the solution was concentrated in
vacuo. The products were purified as described above yielding, after
evaporation of the solvents, the desired alcohol (S)-[D3]-1 (811 mg,
6.49 mmol, ee 99%, 97% deuterated, 38% (over three steps)) and ace-
tate (R)-[D3]-3 (1.11 g, 6.65 mmol, ee 99%, 96% deuterated, 39% (over
three steps)). (S)-1-phenyl-(2-2H3)ethanol: 1H NMR (300 MHz, [D8]diox-
ane): d=3.80 (s, 1H, CH), 4.77 (d, J=2.7 Hz, 1H, OH), 7.19 7.40 (m,
Scheme 7. The two possible reaction pathways for zeolite Beta.
Experimental Section
All experiments were performed in dried glassware under a nitrogen at-
mosphere unless stated otherwise. All chemicals were purchased from
Aldrich. Anhydrous solvents and solids were used as received, the other
liquids were dried and distilled prior to use. The zeolites with a Si/Al
ratio of 16 and 86 were commercially available from CU Chemie,
Uetikon. The ee values were determined and reactions were followed by
gas chromatography by using a Hewlett Packard 5890 Series II gas chro-
matograph, equipped with a 40 mî0.25 mm chiral column Chiraldex B-
PH (b-cyclodextrins permethylated hydroxypropyl), split injector (1:100)
at 2208C, a Flame Ionisation Detector at 2508C and using He as carrier
gas. In the catalytic reactions 1,3,5-triisopropylbenzene was used as inter-
nal standard. Retention times (min) at 1208C isotherm: 1,3,5-triisopro-
pylbenzene (11.0), acetophenone (11.5), (S)-1-methylbenzyl acetate
(14.5), (R)-1-methylbenzyl acetate (15.0), (R)-1-phenylethanol (23.0),
(S)-1-phenylethanol (23.5). Non-enantiomeric compounds were analysed
by a Varian Star 3400 CX, equipped with a 50 mî0.53 mm column CP
5H, Ar-H); 2H NMR (46 MHz, dioxane):
d =1.29 (CD3); (R)-1-
1
(2H3)methylbenzyl acetate: H NMR (300 MHz, [D8]dioxane): d=2.03 (s,
3H, C(O)-CH3), 5.86 (s, 1H, CH), 7.23 7.40 (m, 5H, Ar-H); 2H NMR
(46 MHz, dioxane) d=1.41 (CD3).
2-(2-2H1)Propanol: This alcohol was synthesised in an analogous way as
[D1]-1, using acetone (17.6 mL, 238 mmol), LiAlD4 (5 g, 119 mmol) in di-
ethyl ether (50 mL).[33,34] Distillation (828C/atmospheric pressure) yielded
the desired alcohol (12.2 g, 101 mmol, >99% deuterated, 85%).
1H NMR (300 MHz, [D8]dioxane): d=1.20 (t, J=0.9 Hz, 6H, 2îCH3),
2.08 (s, 1H, OH).
2092
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2088 2093

Meerwein Ponndorf Verley Oppenauer Reactions
2088 2093
Samarium(iii) (2H7)isopropoxide ([D7]-4): Samarium(iii) isopropoxide
(1 g, 30.5 mmol) in 2-(2H8)propanol (10 mL, 131 mmol) was stirred over-
night. After settling of the solids, the solution was transferred into anoth-
Acknowledgement
U.H. thanks the Royal Netherlands Academy of Arts and Sciences
(KNAW) for a fellowship. The authors gratefully acknowledge Roche Di-
agnostics Penzberg (Dr. W. Tischer) for the generous gift of the enzyme
(CAL-B, Chirazyme L-2, c.f., C2, Lyo). We thank Kristina Djanashvili
and Anton van Estrik for measuring the NMR-spectra, Adrie Knol-Kalk-
man for the mass-spectra and Dr. Gˆran Verspui (Avantium Technolo-
gies) for stimulating discussions.
1
er flask. This solution was concentrated under vacuum. For H NMR, the
catalyst was dissolved in [D]trifluoroacetic acid to which a drop of D2O
was added to increase solubility. Dioxane was added as internal standard.
1H NMR (300 MHz, [D1]TFA): d=0.80 1.00 (m, CH3). Integration of the
1H NMR spectrum showed the deuterium content to be 87%.
Samarium(iii) (2-2H1)isopropoxide ([D1]-4): This catalyst was prepared in
a similar manner as above, with 2-(2-2H1)propanol (10 mL, 128 mmol).
From NMR the deuterium contents was determined to be 85%.
General procedure for the synthesis of zeolite Beta:[35] Several zeolites
have been prepared with different Si/Al ratios. The amounts used are
given after the general procedure.
[1]H. Pellissier, Tetrahedron 2003, 59, 8291 8327.
[2]H. Meerwein, R. Schmidt, Justus Liebigs Ann. Chem. 1925, 444,
221 238.
[3]A. Verley, Bull. Soc. Chim. Fr. 1925, 37, 537 542.
[4]W. Ponndorf, Angew. Chem. 1926, 38, 138 143.
[5]H. Lund, Ber. Dtsch. Chem. Ges. 1937, 70, 1520 1525.
[6]R. V. Oppenauer, Recl. Trav. Chim. Pays-Bas 1937, 56, 137 144.
[7]A. L. Wilds, Org. React. 1944, 2, 178.
[8]C. Djerassi, Org. React. 1953, 6, 207.
[9]J. F. de Graauw, J. A. Peters, H. van Bekkum, J. Huskens, Synthesis
1994, 1007 1018.
[10]K. Nishide, M. Node, Chirality 2002, 14, 759 767.
[11]W. N. Moulton, R. E. van Atta, R. R. Ruch, J. Org. Chem. 1961, 26,
290 292.
[12]E. C. Ashby, J. N. Argyropoulos, J. Org. Chem. 1986, 51, 3593 3597.
[13]S. Muthukumaru Pillai, S. Vancheesan, J. Rajaram, J. C. Kuriacose, J.
Mol. Catal. 1983, 20, 169 173.
[14]T. Ooi, T. Miura, K. Maruoka, J. Am. Chem. Soc. 1998, 120, 10790
10791.
[15]T. Ooi, T. Miura, K. Takaya, H. Ichikawa, K. Maruoka, Tetrahedron
2001, 57, 867 873.
NaAlO2 and a 35% TEAOH solution in water were mixed and stirred
for 15 min in a Teflon insert for an autoclave. To this, LUDOX HS-40
was added. A thick gel formed which was homogenised manually. The
insert was put into the autoclave and then the mixture was heated at
1708C for four days. The autoclave was quickly cooled with water and
the white powder obtained was centrifuged, washed with water three
times and then dried in air overnight.
The zeolite was placed in an oven and heated to 5508C at a rate of
18CminÀ1. It was kept at that temperature for 15 h. After cooling, Na+
was exchanged by H+ by stirring the zeolite three times for 48 h with a
0.1m NH4NO3 solution (250 mL). The zeolite obtained was calcined at
4508C for 15 h to yield a white powder. The Si/Al ratio was determined
by elemental analysis and calculated from the 29Si NMR spectrum.[36]
Zeolite Beta (Si/Al=6): NaAlO2 (2.12 g, 25.9 mmol), 35% TEAOH in
water (8 mL, 19.5 mmol) and LUDOX HS-40 (15 mL). Zeolite Beta (Si/
Al=25) NaAlO2 (0.53 g, 6.5 mmol), 35% TEAOH in water (8 mL,
19.5 mmol) and LUDOX HS-40 (15 mL).
General procedure for the samarium(iii) catalysed MPVO racemisation:
Samarium(iii) isopropoxide (0.050 mmol) was dissolved in the solvent of
choice (Table 1) (0.75 mL), acetophenone (2) (58.3 mL, 0.50 mmol) and
(S)-1-phenylethanol ((S)-1) (60.7 mL, 0.50 mmol) were added and the
temperature was raised to 708C, after which the solution was stirred
overnight for 18 h towards complete racemisation. At regular time inter-
vals, samples of 20 mL were taken, which were analysed by chiral GC.
[16]E. D. Williams, K. A. Krieger, A. R. Day, J. Am. Chem. Soc. 1953,
75, 2404 2407.
[17]M. Cohn, Preparation and measurement of isotopic tracers (Ed.: J.
Edwards), Ann Arbor, Michigan, 1947, p. 51.
[18]J. L. Namy, J. Souppe, J. Collin, H. B. Kagan, J. Org. Chem. 1984, 49,
2045 2049.
[19]T. Ooi, T. Miura, Y. Itagaki, H. Ichikawa, K. Maruoka, Synthesis
2002, 279 291.
[20]E. J. Campbell, H. Zhou, S. T. Nguyen, Org. Lett. 2001, 3, 2391 2393.
[21]O. P‡mies, J.-E. B‰ckvall, Chem. Eur. J. 2001, 7, 5052 5058.
[22]J. Shabtai, R. Lazar, E. Biron, J. Mol. Catal. 1984, 27, 35 43.
[23]M. Berkani, J. L. Lemberton, M. Marczewski, G. Perot, Catal. Lett.
1995, 31, 405 410.
[24]M. Huang, P. A. Zielinsky, J. Moulod, S. Kaliaguine, Appl. Catal. A
1994, 118, 33 49.
[25]J. C. van der Waal, E. J. Creyghton, P. J. Kunkeler, H. van Bekkum,
Top. Catal. 1997, 4, 261 268.
[26]E. J. Creyghton, S. D. Ganeshie, R. S. Downing, H. van Bekkum, J.
Mol. Catal. 1997, 115, 457 472.
[27]A. Corma, M. E. Domine, L. Nemeth, S. Valencia, J. Am. Chem.
Soc. 2002, 124, 3194 3195.
[28]M. A. AramendÌa, V. Borau, C. Jimÿnez, J. M. Marinas, J. R. Ruiz,
F. J. Urbano, J. Chem. Soc. Perkin Trans. 2 2002, 1122 1125.
[29]A. Corma, M. E. Domine, S. Valencia, J. Catal. 2003, 215, 294 304.
[30]P. M. Dinh, J. A. Howarth, A. R. Hudnott, J. M. J. Williams, W.
Harris, Tetrahedron Lett. 1996, 37, 7623 7626.
[31]S. Wuyts, K. de Temmerman, D. de Vos, P. Jacobs, Chem. Commun.
2003, 1928 1929.
[32]L. Veum, M. Kuster, S. Telalovic, U. Hanefeld. T. Maschmeyer, Eur.
J. Org. Chem. 2002, 1516 1522.
General procedure for the zeolite Beta catalysed racemisation of (S)-1:
Zeolite Beta was activated at 4008C overnight. This zeolite (20 mg) was
introduced into a Schlenk flask. Toluene (4 mL) was added followed by
(S)-1-phenylethanol ((S)-1) (0.24 mL, 2 mmol). If required, acetone
(0.15 mL, 2 mmol) was added (Figure 1). The mixture was heated to
508C. The reaction was followed by taking 20 mL samples at regular inter-
vals and analysing them with chiral GC.
Zeolite Beta catalysed reaction with styrene: Zeolite Beta (Si/Al=16)
was activated at 4008C overnight. This zeolite (20 mg) was introduced
into a Schlenk flask. Toluene (4 mL) was added followed by styrene
(0.23 mL, 2 mmol), acetone (0.15 mL, 2 mmol) and water (36 mL,
2 mmol). The mixture was heated to 508C. The reaction was followed by
taking 20 mL samples at regular intervals and analysing them with chiral
GC. 6% 1 was observed after 20 h.
As a control experiment, the reaction was also performed using (S)-1
(0.24 mL, 2 mmol) instead of styrene. After 20 h complete racemisation
and formation of styrene was observed.
Zeolite Beta catalysed racemisation of (S)-1 in the presence of H217O:
Zeolite Beta (Si/Al=13.5) was activated at 4008C overnight. This zeolite
(20 mg) was introduced into a Schlenk flask. Toluene (4 mL) was added
followed by (S)-1-phenylethanol ((S)-1) (0.24 mL, 2 mmol) and H217O
(5 mL, 0.26 mmol, degree of 17O labelling: 25.7%). The mixture was
heated to 508C and was analysed by MS after 2 h. It was calculated from
the mass spectrum that the product was 1.0% 17O enriched.
¬
¬
¬
[33]V. M. Mic ovic, M. L. Mihailovic, J. Org. Chem. 1953, 18, 1190 1200.
[34]L. Friedman, A. T. Jurewicz, J. Org. Chem. 1968, 33, 1254 1255.
[35]P. J. Kunkeler, B. J. Zuurdeeg, J. C. van der Waal, J. A. Bokhoven,
D. C. Koningsberger, H. van Bekkum, J. Catal. 1998, 180, 234 244.
[36]G. Engelhardt, D. Michel, High-Resolution Solid-State NMR of sili-
cates and Zeolites, Wiley, Chichester, 1987, pp. 145 154.
General procedure for the zeolite Beta catalysed reduction of 7: Zeolite
Beta was activated at the appropriate temperature overnight (Table 2).
This zeolite (20 mg) was introduced into a Schlenk flask. 2-(2-2H1)Propa-
nol (3 mL) was added followed by 4-tert-butyl cyclohexanone 7 (0.31 g,
2 mmol). The mixture was heated to 508C. The reaction was followed by
taking 20 mL samples at regular intervals and analysing them with achiral
GC.
Received: August 18, 2003
Revised: November 28, 2003 [F5460]
2093
Chem. Eur. J. 2004, 10, 2088 2093
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim