A novel ruthenium catalysed deracemisation of alcohols

Article abstract of DOI:10.1039/b511492f

The deracemisation of alcohols has been achieved with 65-75% ee using a non-selective ruthenium catalysed oxidation to give an intermediate ketone followed by an enantioselective hydrogenation. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b511492f

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A novel ruthenium catalysed deracemisation of alcohols
Gareth R. A. Adair and Jonathan M. J. Williams
Received (in Cambridge, UK) 17th August 2005, Accepted 21st September 2005
First published as an Advance Article on the web 13th October 2005
DOI: 10.1039/b511492f
The deracemisation of alcohols has been achieved with 65–75%
ee using a non-selective ruthenium catalysed oxidation to
give an intermediate ketone followed by an enantioselective
hydrogenation.
The preparation of enantiomerically pure alcohols has been an
important synthetic target for decades. Enantioselective reactions
involving addition of carbon nucleophiles to aldehydes have been a
1
widely used approach, along with the conversion of ketones into
alcohols by hydrogenation, transfer hydrogenation and hydride
2
addition. The preparation of enantiomerically enriched alcohols
Scheme 2 Noyori catalyst 2 in transfer and direct hydrogenation.
has also been achieved by kinetic resolution and dynamic kinetic
3
resolution strategies.
back to the alcohol by direct hydrogenation using the same
catalyst.
Herein, we report the development of a novel deracemisation
strategy for the conversion of a racemic alcohol into an
enantiomerically enriched alcohol using a catalyst (Scheme 1).
Whilst several catalysts have been reported for the racemisation
Preliminary experiments using Noyori catalyst 2 to oxidise
phenethyl alcohol 1 into acetophenone 3 were performed using
acetone as the oxidant and required the use of base, but were
unsuccessful due to the competing aldol reaction of the acetone
employed.
4
of alcohols, the reverse process is more demanding, since
deracemisation is thermodynamically disfavoured. The conversion
of two enantiomers of a compound into a single enantiomer
requires a decrease in entropy. It therefore follows that a catalyst
will be unable to deracemise a compound unless energy can be
provided from some other source.
We therefore decided to use a combination of ruthenium
hydride complex 4 in the presence of (R)-BINAP 5 and (R,R)-
DPEN 6, with the expectation that the dihydride analogue of the
Noyori complex would be formed in situ (Fig. 2). It is known that
ruthenium hydride complexes do not require the addition of base
The Noyori catalyst 2 (Fig. 1) is well known for its ability to
reduce ketones with high enantioselectivity by direct hydrogena-
5
tion. Interestingly, this catalyst has been reported to be rather
7
in order to function as transfer hydrogenation catalysts.
We were encouraged to find that oxidation of phenethyl alcohol
1
occured readily upon treatment with ten equivalents of acetone,
non-selective in the corresponding reduction achieved with transfer
6
hydrogenation (Scheme 2).
using the ruthenium hydride/BINAP/DPEN combination after
heating in toluene at 60 uC for 4 h. As expected, this oxidation
process gave essentially no kinetic resolution, providing a reaction
mixture containing acetophenone 3 in the presence of the ruthe-
nium catalyst and ligands. Pressurisation of this mixture with
hydrogen (10 bar) led to a reduction of the acetophenone 3 back
into phenethyl alcohol 1 after heating at 45 uC for 16 h. However,
phenethyl alcohol 1 was recovered with the (S)-enantiomer
predominating at 55% ee.
We considered the possibility of using catalyst 2 for the
oxidation of a racemic alcohol into a ketone by non-selective
hydrogen transfer, followed by in situ enantioselective reduction
In developing this transformation, we chose to employ
cyclohexanone as the oxidant rather than acetone, because
cyclohexanone is the more powerful oxidant, and therefore
requires fewer equivalents in order to drive the equilibrium
Scheme 1 Deracemisation of alcohol 1
8
forwards. It is important to drive the equilibrium as far towards
Fig. 1 The Noyori hydrogenation catalyst.
Department of Chemistry, University of Bath, Claverton Down, Bath,
UK BA2 7AY. E-mail: ch9graa@bath.ac.uk; chsjmjw@bath.ac.uk
Fig. 2 The components for an in situ dihydride complex.
This journal is ß The Royal Society of Chemistry 2005
5
578 | Chem. Commun., 2005, 5578–5579

View Article Online
combination of ruthenium hydride complex 4 with (R)-BINAP 5
6 6
and (R,R)-DPEN 6 in C D revealed several peaks in the hydride
region, suggesting that several hydride complexes were present.
However, when we used this combination in a simple hydrogena-
tion of acetophenone 3 we were able to obtain phenethyl
alcohol with 80% ee, somewhat higher than that obtained in the
deracemisation procedure. Ruthenium catalyst 4 by itself was a
very poor catalyst for direct hydrogenation, as was a combination
of catalyst 4 with (R)-BINAP, providing less than 5% conversion
of ketone into alcohol. Catalyst 4 in combination with (R,R)-
DPEN provided a 12% conversion with a modest 27% ee.
It seems reasonable to suggest that there is more than one
ruthenium hydride species present in the reaction mixture, but that
the Ru/BINAP/DPEN complex is the most reactive and the most
selective during the reduction process.
Scheme 3 Conditions (i) Ru(PPh
R,R)-DPEN (2 mol%), cyclohexanone (2 equiv.), solvent, 60 uC, 4 h.
Then pressurisation with H (10 bar), 45 uC, 16 h.
3 4 2
) H (2 mol%), (R)-BINAP (2 mol%),
(
2
In summary, we have developed a procedure for deracemisation
of alcohols based on the Noyori catalyst. The reactions depend on
a non-selective oxidation step followed by a selective hydrogena-
tion step. Studies are now in progress to understand the nature of
the catalyst more fully.
We wish to thank the EPSRC for funding a studentship through
the doctoral training account (to GRAA).
Notes and references
Fig. 3 Alcohols successfully deracemised.
1
2
K. Soai and S. Niwa, Chem. Rev., 1992, 92, 833.
W. Tang and X. Zhang, Chem. Rev., 2003, 103, 3029; M. J. Palmer and
M. Wills, Tetrahedron: Asymmetry, 1999, 10, 2045.
the ketone as possible, since any racemic alcohol remaining at
this stage will compromise the enantioselectivity of the alcohol
recovered at the end of the reaction.
3
4
5
S. D. Bull and D. E. J. E. Robinson, Tetrahedron: Asymmetry, 2003, 14,
1407; H. Pellisseier, Tetrahedron, 2003, 59, 8291.
B. Mart ´ı n-Matute, M. Edin, K. Bog a´ r, F. B. Kaynak and J.-E. B a¨ ckvall,
J. Am. Chem. Soc., 2005, 127, 8817.
R. Noyori and T. Ohkuma, Angew. Chem., Int. Ed., 2001, 40, 40;
T. Ohkuma, H. Ooka, S. Hashiguchi, T. Ikariya and R. Noyori, J. Am.
Chem. Soc., 1995, 117, 2675.
V. Rautenstrauch, X. Hoang-Cong, R. Churland, K. Abdur-Rashid
and R. H. Morris, Chem.–Eur. J., 2003, 9, 4954.
J.-E. B a¨ ckvall, J. Organomet. Chem., 2002, 652, 105.
H. Adkins, R. M. Elofson, A. G. Rossow and C. C. Robinson, J. Am.
Chem. Soc., 1949, 71, 3622.
A representative experimental procedure is as follows: 1-phenyl-1-
pentanol (328 mg, 2 mmol), Ru(PPh ) H 4 (46 mg, 0.04 mmol),
3 4 2
(R)-BINAP 5 (25 mg, 0.04 mmol), (R,R)-DPEN 6 (8.4 mg, 0.04 mmol)
As shown in Scheme 3, repeating the reaction using only two
equivalents of cyclohexanone in place of ten equivalents of acetone
provided the recovered alcohol with 65% ee. Changing the solvent
from toluene provided slightly higher enantioselectivity in the case
of dioxane, tert-butanol and THF.
6
7
8
Application of this deracemisation procedure to other alcohols
was also successful. For the other alcohols, we chose to use three
equivalents of cyclohexanone, simply to ensure that the initial
oxidation to ketone was essentially driven to completion. The
alcohols shown in Fig. 3 were recovered in good yields and good
enantioselectivities, using THF as the solvent. Phenethyl alcohol
was isolated by column chromatography, whilst the other alcohols
9
and cyclohexanone (588 mg, 6 mmol) were heated at 60 uC in THF
(2 mL) for 4 h. After pressurisation with hydrogen (10 bar), heating was
continued at 40 uC for 16 h. After removal of solvent under reduced
pressure, the alcohol was re-isolated by Kugelrohr distillation (304 mg,
9
were isolated by distillation. Enantiomeric excess was determined
93% yield, 75% ee).
10
by HPLC or GC.
10 The enantiomeric excess of 1-phenylethanol was determined using chiral
GC (Supelco BETA-DEX 110, 60 m 6 0.25 mm). 1-Phenylpropanol
and 1-phenylpentanol were anaylsed by HPLC using a Daicel Chiracel
OD-H column (hexane : isopropanol, 97 : 3), and a-methyl-1-
naphthalene methanol with a Daicel Chiracel OB-H column (hexane :
isopropanol, 90 : 10).
The isolated yields were consistently good, and whilst the
enantiomeric excesses were good, they are lower than those
obtained by direct hydrogenation of the ketone with the pre-
4
1
formed Noyori catalyst 2. Analysis of the H NMR spectrum of a
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 5578–5579 | 5579