Concerning kinetic resolution by the Sharpless asymmetric dihydroxylation reaction

Article abstract of DOI:10.1039/a906423k

The transition state for the product-determining step in the Sharpless asymmetric dihydroxlation reaction is not product-like, and effective kinetic resolution can occur when one face of a chiral alkene is hindered.

Full text of DOI:10.1039/a906423k

Concerning kinetic resolution by the Sharpless asymmetric dihydroxylation
reaction
Hamish S. Christie, David P. G. Hamon* and Kellie L. Tuck
Chemistry Department, University of Adelaide, S.A. 5005, Australia. E-mail: dhamon@chemistry.adelaide.edu.au
Received (in Cambridge, UK) 6th August 1999, Accepted 31st August 1999
The transition state for the product-determining step in the
Sharpless asymmetric dihydroxlation reaction is not prod-
uct-like, and effective kinetic resolution can occur when one
face of a chiral alkene is hindered.
the major diol was then determined as already described. In two
separate experiments, conducted at 0 °C, the reactions were
stopped at 15 and 94% conversion respectively. The ee for the
remaining alkene, for the former reaction, was determined as
~ 4% and, for the latter, as ~ 86%. This computes to an E of ~ 2
in each case. This value of E is too small for effective kinetic
resolution. Clearly the TS of the product-determining step is not
product-like.
It can be reasoned that, if the product-determining step in the
AD reaction is reactant-like and if the alkene is highly
enantiofacially directing, then a kinetic resolution should be
effective when there is considerable difference in the ease of
approach to the two faces of a chiral alkene.11 Such an effect has
been demonstrated12 recently for one molecule, which displays
axial chirality. Recently we completed13 an asymmetric synthe-
sis of grandisol in which the key step is the kinetic resolution of
the primary allylic alcohol 4 by a Sharpless AE reaction. The
Sharpless asymmetric dihydroxlation (AD)1 and asymmetric
epoxidation (AE)2 reactions have proved to be very effective
means whereby asymmetry can be introduced into molecules
starting from prochiral alkenes and allylic alcohols respectively.
The AE reaction has also proved useful for effecting kinetic
resolution of chiral allylic alcohols. However, with only a few
exceptions, the AD reaction has not been effective for carrying
out kinetic resolutions and the reasons for this are not well
understood.1
It is known3 that the AD reaction of 1-phenylcyclohexene
proceeds with high asymmetric induction. Therefore, reaction
of these reagents with 1-phenyl-4-tert-butylcyclohexene 1
would also be expected to proceed with high asymmetric
induction but two diastereomeric products should result. For
instance with AD-mix-b the enantioenriched diastereomers 2
and 3 should form (Scheme 1). If these molecules exist
predominantly in the expected chair conformers, the diaster-
eomer 2 has the phenyl group in an equatorial position whereas
the phenyl group in the other diastereomer 3 has an axial
orientation. The diastereomer 3 will be several kJ mol21 higher
in energy than the diastereomer 2.4 Therefore, if the transition
state (TS) for the product-determining step in the AD reaction is
product-like, there would be a considerable difference in the
rate of formation of these two diastereomers, and an effective
kinetic resolution should be possible.
The AD reaction on the alkene 1 was performed following
Sharpless’s recommended procedure5 for tri-substituted alkenes
(AD-mix-b, MsNH2, ButOH–H2O, 0 °C). Equal quantities of
the diols 2 and 3,6 eluting in that order, were obtained, in 84%
yield after ‘flash’ silica gel chromatography. The enantiomeric
purity of each diol was determined by the use of NMR chiral
shift reagent experiments. The diol 2 from the AD reaction had
> 99% ee7 and the diol 3 had > 95% ee.8 The racemic diols
were obtained using the Sharpless method with quinuclidine as
ligand to give the racemic diols in a ratio of ~ 4.5+1.
key feature in that resolution is that the molecule presents both
an open convex face and a more-hindered, concave face to the
reagents. The reaction proceeds to give products of attack on
one face only, the convex face. This system appeared suitable to
pursue the study of the kinetic resolution by the AD reaction.
Indeed, the alkene 5 appears ideal for this study. Control of
product formation in the Sharpless AD reaction should be
determined mainly by the enantiofacial selectivity of the styrene
moiety. The concave face of this molecule is hindered, so there
should be a considerable difference in the rate of reaction of the
enantiomers of this molecule and kinetic resolution should be
effective.
Reaction of the bicycloketone 614 with PhMgBr and Et2O at
0 °C gave, in 88% yield, a single racemic alcohol, mp 56–58 °C,
presumed to be the diastereomer 7 (Scheme 2). Dehydration of
this alcohol (CH2Cl2, Et3N, MsCl, 0 °C) gave as an oil the
alkene 5 (81%).15 A kinetic resolution of this substrate was
examined in two separate experiments with 26 and 50% oxidant
respectively. However, because the alkene 5 is somewhat
unstable, in each case the ees for both the diol (presumed to be
enantiomer 8, based on the application of Sharpless’ mnemonic)
and the recovered alkene 5 were determined, after separation, in
the following way. The diol 8 was converted to the mono
Mosher ester derivative [(+)-MTPA, CH2Cl2, DCC, DMAP]
and the ee determined.16 The alkene was converted to the same
diol (presumed to be mainly 9) by the achiral dihydroxylation
The kinetic resolution was also studied under the normal
conditions for the AD reaction and the relative rates for the
enantiomers (E) were determined following a literature proce-
dure9 which is related to the relationship derived by Kagan.10
The percentage conversion of the racemic substrate was
followed by NMR spectroscopy. The remaining alkene was
separated from the diols 2 and 3. The ee of the remaining alkene
1 was determined by conversion, using the achiral dihydroxyla-
tion reaction, into the same diols 2 and 3. The enantiopurity of
Scheme 1
Scheme 2
Chem. Commun., 1999, 1989–1990
This journal is © The Royal Society of Chemistry 1999
1989

4 A value of 15.9 kJ mol21 was obtained by calculation. Gaussian 94
(revision C), Gaussian, Inc., Pittsburgh, PA., 1995.
5 K. B. Sharpless, W. Amberg. Y. L. Bennani, G. A. Crispino, J. Hartung,
K.-S. Jeong, H.-L. Kwong, K. Morikawa, Z.-M. Wang, D. Xu and X.-L.
Zhang, J. Org. Chem., 1992, 57, 2768.
6 Diols 2 and 3 show CHOH resonances at d 4.02 (dd, J 11, 4 Hz) and 4.38
1
(m, width at half height < 10 Hz), respectively, in the H NMR (after
D2O exchange). These diols have been prepared before in racemic form,
see G. Berti, B. Macchia and F. Macchia, Tetrahedron, 1968, 24,
1755.
7 Chiral shift reagent [Eu(hfc)3] in 15% C6D6–CCl4. The But protons’
resonance due to the minor enantiomer was not visible by proton NMR,
but the 13CH satellite of the But protons’ resonance of the major isomer
was. Addition of 1% of the racemate gave a discernible peak for the
minor enantiomer.
8 Chiral shift reagent [Eu(tfc)3] in 15% C6D6–CCl4. Addition of 5% of the
racemate was required before a discernible peak for the But protons’
resonance of the minor enantiomer could be detected.
9 V. S. Martin, S. S. Woodard, T. Katsuki, Y. Yamada, M. Ikeda and K. B.
Sharpless, J. Am. Chem. Soc., 1981, 103, 6237.
10 G. Balavoine, A. Moradpour and H. B. Kagan, J. Am. Chem. Soc., 1974,
96, 5152.
11 Initially we synthesised (3RS,6RS)-3-tert-butyl-1,6-diphenylcyclohex-
ene but it did not appear to undergo the AD reaction. This synthesis will
be reported in the full paper.
12 J. M. Gardiner, M. Nørret and I. H. Sadler, Chem. Commun., 1996,
2709.
13 D. P. G. Hamon and K. L. Tuck, submitted for publication.
14 R. L. Cargill, J. R. Dalton, G. H. Morton and W. E. Caldwell, Org.
Synth., 1984, 62, 118.
reaction (quinuclidine ligand) and this diol was converted to the
mono Mosher ester derivative from which the ee was obtained.
The ees for the diols in the two reactions were 88 and 80%, and
for the alkenes they were 52 and 80%, respectively. The
enantiomeric ratios for these two reactions calculate to E = 26
and 21, respectively (the same within experimental error),
which translates to an effective kinetic resolution ( > 95% ee at
60% conversion).
In conclusion, we find that the reaction of 1-phenyl-4-tert-
butylcyclohexene with AD-mix-b gives the two diastereomers 2
and 3 in very high enantiopurity. However, the rate difference
between formation of the two enantiomers is only small (E ~ 2)
and kinetic resolution is ineffective. We also find that the
reaction of the alkene 5 with AD-mix-b gives an effective
kinetic resolution (E > 20) and only one diastereomeric diol is
formed. These results are consistent with an early, rather than a
late, transition state for the product-determining step in the
Sharpless AD reaction.
15 The radical inhibitor, 2,6-di-tert-butyl-p-cresol, was added during the
work-up procedure, otherwise much lower yields of this alkene were
obtained.
16 In the 1H NMR spectrum, methoxy resonances at d 3.37 and 3.53 for the
diastereomers produced from the racemic diol showed separation at the
base-line.
Notes and references
1 M. S. VanNieuwenhze and K. B. Sharpless, J. Am. Chem. Soc., 1993,
115, 7864; H. V. Kolb, M. S. VanNieuwenhze and K. B. Sharpless,
Chem. Rev., 1994, 94, 2483 and references therein.
2 R. A. Johnson and K. B. Sharpless, Catalytic Asymmetric Synthesis, ed.
I. Ojima, VCH, New York, 1993, pp. 103–158.
3 S. B. King and K. B. Sharpless, Tetrahedron Lett., 1994, 35, 5611.
Communication 9/06423K
1990
Chem. Commun., 1999, 1989–1990