Dynamic kinetic asymmetric synthesis of β-aminoalcohols from racemic epoxides in cyclodextrin complexes under solid state conditions

Article abstract of DOI:10.1039/b004643o

It has been shown for the first time that enantiopure β-aminoalcohols can be prepared from racemic epoxides by dynamic kinetic resolution involving enantio-differentiating racemisation in cyclodextrin complexes under solid state conditions.

Full text of DOI:10.1039/b004643o

Dynamic kinetic asymmetric synthesis of b-aminoalcohols from racemic
epoxides in cyclodextrin complexes under solid state conditions†
L. Rajender Reddy, N. Bhanumathi and K. Rama Rao*
Organic Chemistry Division-I, Indian Institute of Chemical Technology, Hyderabad-500 007, India.
E-mail: ramaraok@iict.ap.nic.in
Received (in Cambridge, UK) 12th June 2000, Accepted 16th October 2000
First published as an Advance Article on the web 9th November 2000
It has been shown for the first time that enantiopure b-
aminoalcohols can be prepared from racemic epoxides by
dynamic kinetic resolution involving enantio-differentiating
racemisation in cyclodextrin complexes under solid state
conditions.
The novel phenomenon of converting racemic substrates into a
single enantiomer of the product by dynamic kinetic resolution
via racemisation of the substrates has been a formidable
challenge in asymmetric synthesis.1 Recently, it has been
receiving increasing attention and attempts are being made to
achieve every useful asymmetric synthesis by dynamic kinetic
resolution since it overcomes the severe limitation of the
conventional kinetic resolution where the maximum yield of
one stereoisomer of the starting material or product is only 50%.
But, so far there is no report on the dynamic kinetic asymmetric
synthesis of b-aminoalcohols of great significance from the
easily accessible and inexpensive racemic epoxides. The b-
Scheme 1
aminoalcohol moiety is present in many natural products and
drugs and acts as an intermediate in various asymmetric
transformations.2 The approaches hitherto reported from the
h). The time taken was always dependent on the frequency of
racemic epoxides involve only kinetic resolution.3 Further to
our studies on cyclodextrins as chiral templates for the
enantioselective synthesis of a variety of chiral building
blocks,4 we report herein the first biomimetic approach for the
synthesis of a single enantiomer of b-aminoalcohols from the
corresponding racemic epoxides by dynamic kinetic resolution
involving enantio-differentiating racemisation in cyclodextrin
complexes.5
mixing. However, in the case of amine 2, it was added in excess
due to its volatility and was added intermittently during the
course of mixing. The resulting aminoalcohols 4 and 5 were
extracted with ethyl acetate and purified through the formation
followed by release of their hydrochlorides. The isolated yield
of the product was always in the range 70–79% and the
enantioselectivity was excellent in some of the compounds
(Table 1). Compounds 4b and 5b have shown 100% ee followed
by compound 4c (ee 89%) and 4d (ee 85%). The enantiomeric
excesses (ee) of the products (4 and 5) were determined by
chiral HPLC analysis. These aminoalcohols 4 and 5 have been
shown to have R configuration by comparison of the sign of
rotation with those of the known compounds.7 However, these
reactions when carried out using a-cyclodextrin inclusion
complexes, gave enantioselectivities that were far lower than
those obtained with b-cyclodextrin. This may be due to
ineffective complexation.
The cyclodextrins are cyclic oligosaccharides possessing
hydrophobic cavities and they mimic enzymes in their capa-
bility to bind substrates selectively and catalyse chemical
reactions by supramolecular catalysis involving reversible
formation of host–guest complexes with substrates through
non-covalent bonding. Apart from this, with the cyclodextrin
cavity being chiral it can induce asymmetric reactions.6 It
requires the following criteria to be fulfilled to ensure rigidity
for chiral recognition by cyclodextrins: (i) a phenyl ring which
can fit into the cyclodextrin cavity to form an inclusion
complex, (ii) a functional group at the stereogenic center to form
a strong interaction at the cyclodextrin cavity entrance. Since
aryloxyepoxides 1 fulfil these criteria, they have been chosen as
substrates.
The formation of enantiopure b-aminoalcohols (4 and 5)
from the racemic epoxides 1 may be postulated as follows:
Table 1 Solid state asymmetric synthesis of b-aminoalcohols
The inclusion complexes of aryloxyepoxides 1 with b-
cyclodextrin (b-CD) were prepared as described earlier.4 The
reaction of these cyclodextrin complexes of epoxides 1 with
amines 2 or 3 (Scheme 1) when carried out in water as the
reaction medium, yielded aminoalcohols that were nearly
racemic (ee 2%). This may be due to free movement of the guest
molecule in solution. Hence, it was planned to carry out these
reactions in the solid state where the movement of the guest
molecule would be restricted, resulting in better chiral recog-
nition.
An intimate mixture of the epoxide 1–b-cyclodextrin com-
plex (1+1) and the amine 2 or 3 was mixed in equimolar
amounts in an agate mortar using a pestle and the mixing
continued until the starting epoxide disappeared on TLC (3–12
Entry
Product
ee (%)a
Yield (%)b
1
2
3
4
5
6
7
8
4a
4b
4c
4d
5a
5b
5c
5d
73.8
100
89
85
1.4
100
0.8
3.5
79
75
72
70
74
70
76
73
a Determined by HPLC analysis with the chiral column ‘Diacel Chiralcel
OD’ (0.46 cm f 3 25 cm) using hexane: propan-2-ol: diethylamine
(80+20+0.1) as eluent at a flow rate of 0.5 ml min21 using UV detection
(254 nm). b Isolated yields.
DOI: 10.1039/b004643o
Chem. Commun., 2000, 2321–2322
This journal is © The Royal Society of Chemistry 2000
2321

Scheme 2
The fact that the epoxide 1 isolated from the cyclodextrin
complex is racemic and the yields of the aminoalcohols 4 and 5
were always more than 50%, suggests that kinetic resolution is
not operating under these conditions as then only a maximum of
50% conversion can be expected. Hence, in the present
investigation, to get a single enantiomer of the product (4 or 5)
from the racemic starting epoxide 1, it requires interconversion
of one of the enantiomers of the epoxide 1 (Scheme 2). This is
possible through racemisation which is controlled by entropy
effects.8 Racemisation can take place in the present case by the
interconversion of the epoxide enantiomer facilitated under
solid state conditions. Though racemisation is quite a slow
process in the absence of suitable driving mechanism i.e. the
external amine in the present case, the racemisation of the b-CD
complexes of the individual R and S epoxides of 1a (R = H) has
been attempted by grinding them intimately for 5 h and
analysing by chiral HPLC.9 It is observed that racemisation
does indeed take place only with S-enantiomer to an extent of
2% as seen by the change in %ee. This reaction also further
substantiates the process of dynamic kinetic resolution i.e. as
the R-enantiomer of the epoxide (1) in the CD complex reacts
with the amine, the S-epoxide gets converted to the R-epoxide to
take the reaction forward, leading to only the R-enantiomer of
the product (4).
Thus, when one of the enantiomeric forms of the epoxide 1 in
the b-cyclodextrin cavity, due to its favourable geometry, is
captured selectively by the external amine (2 or 3), the
phenomenon of dynamic kinetic resolution sets in under the
reaction conditions. Hence, by dynamic kinetic resolution
through racemisation of the starting epoxide, it is possible to get
enantiomerically pure aminoalcohols (4 and 5). This has also
been confirmed from the individual experiments utilizing either
R or S epoxides (Scheme 3). Both R and S epoxides gave the
aminoalcohol (4a) mainly as the R enantiomer (90% ee by
HPLC). This gives further evidence to show that racemisation
of the starting epoxide and dynamic kinetic resolution are
operating under the reaction conditions to give a single
enantiomer of the product.
Scheme 3
enantio-differentiating racemisation in cyclodextrin complexes
under solid state reaction conditions.
We thank the Director, Dr K. V. Raghavan for constant
encouragement and support during the course of this work and
CSIR, New Delhi, India for the award of research fellowship to
LRR.
Notes and references
† IICT Communication No. 4579.
1 (a) R. S. Ward, Tetrahedron Asymmetry, 1996, 1475; (b) F. Toda and K.
Tanaka, Chem. Lett., 1983, 661.
2 (a) Pharmacological Basis of Therapeutics, ed. L. S. Goodman and A.
Gilman, 6th edn., MacMillan, New York, 1980; (b) A. S. Rao, S. K.
Paknikar and J. G. Kirtane, Tetrahedron, 1983, 39, 2323; (c) Bo. Lamn,
K. Ankener and M. Frantsi, Acta Chem. Scand., 1987, B41, 202; (d) C. K.
Juslen, M. Kuhn, H. R. Loosli, T. J. Petcher, H. P. Weber and A. V.
Wartburg, Tetrahedron Lett., 1976, 4147; (e) J. S. Zewig, J. L. Luche, E.
Barreiro and P. Crabbe, Tetrahedron Lett., 1975, 2355.
3 (a) N. Matsuo and N. Ohno, Tetrahedron Lett., 1985, 26, 5533; (b) H. S.
Bevinakatti and A. A. Banerji, J. Org. Chem., 1991, 56, 5372; (c) W. E.
Lander and G. M. Whitesides, J. Am. Chem. Soc., 1984, 106, 7250; (d)
J. F. Larrow, S. E. Schaus and E. N. Jacobsen, J. Am. Chem. Soc., 1996,
118, 7420.
4 (a) K. Rama Rao and P. B. Sattur, J. Chem. Soc., Chem. Commun., 1989,
342; (b) K. Rama Rao and H. M. Sampth Kumar, Synth. Commun., 1993,
23, 1877.
5 K. Rama Rao, N. Bhanumathi and L. Rajender Reddy, Indian Pat. Appl.
No. 127/DEL/2000, US Pat. Appl. No. 09/535 384, 2000.
6 K. Takahashi, Organic Reactions Mediated by Cyclodextrins, Chem.
Rev., 1998, 98, 2013.
7 H. Takahashi, S. Sakuraba, H. Takeda and K. Achiwa, J. Am. Chem. Soc.,
1990, 112, 5876.
8 (a) Y. Izumi and A. Tai, Stereodifferentiating reactions, the nature of
asymmetric reactions, Academic Press, New York, 1977, p. 169–177; (b)
E. L. Eliel and S. H. Wilen, Stereochemistry of Organic Compounds,
John Wiley and Sons, Inc., New York, 1994, 424.
Thus, in conclusion, it has been shown for the first time that
enantiopure b-aminoalcohols of high potential can be made
from the easily accessible and inexpensive racemic epoxides.
This can be achieved by dynamic kinetic resolution involving
9 We would like to thank one of the referees who has suggested this set of
experiments which has given further proof to our results.
2322
Chem. Commun., 2000, 2321–2322