Accurate determination of the extent to which an S(E)2' reaction of an allylsilane is anti

Article abstract of DOI:10.1016/S0040-4039(00)60115-6

The E and Z allylsilanes E-1 and Z-1 have been prepared essentially 100% geometrically and enantiomerically pure; their S(E)2' reactions with adamantyl chloride are highly, but not completely, stereospecific.

Full text of DOI:10.1016/S0040-4039(00)60115-6

Letters, Vol. 33, No. 31,
Printed in Great Britain
44794482, 1992
Press
Accurate Determination of the Extent to which an
Reaction of an Allylsilane is anti
Michael J. C. Buckle, Ian Fleming* and Salvador Gil
University Chemical Laboratory,
Road, Cambridge CB2
The E and
pure; their
allylsilanes E-l and Z-l have been prepared essentially 100% geometrically and
reactions with
chloride are highly, but not completely,
It has been well established, principally by
other constraints, the
certainly very high, probably
Wetter2 and
that, in the absence of
reaction of allylsilanes is stereospecifically anri. The degree of stereospecificity is
but it was not possible in any of their work to measure more accurately
than that how completely stereospecific the reactions were, largely because the allylsilanes that they used were,
necessarily at that time, less than 100% homochiral (optically pure). With new methods available for the
synthesis of allylsilanes, we have now examined more accurately than has been possible in the past the degree
of stereospecificity of the electrophilic substitution reaction E-l 2, which we offer as a paradigm, using an
E-l that is essentially 100% homochiral.
a n t i
10
Cl
(0.1
-78
E-l
30 min
ee
E
50%
In order to set up the stereogenic centre we tried several combinations of nucleophile, chiral auxiliary, and
double bond geometry. We obtained the highest level of stereochemical purity
carbon by the conjugate addition of iithium dimethyl-cuprate to the silicon-containing E-substrate 3 based on
Oppolzer’s sultam derived from Unfortunately, the major product 4 was the more soluble, and
at the silicon-bearing
recrystallisation was not an efficient method for removing even the 1% of diastereoisomer 5 that was present.
We resorted therefore to a simple device that was guaranteed to work-we removed the chiral auxiliary and
replaced it with its enantiomer
6 derived from (-)-camphor.5 The major product 7 was now, inevitably, the less
soluble diastereoisomer, and one recrystallisation served to remove all but 0.2% of the minor component 8.
This procedure is not to be recommended in synthesis, because of the inevitable losses, nor would it be needed.
It served us well here, because the overall yield was not our primary concern.
For good measure, we
the sultam 7 three times, at which point we could not detect, by
careful GC analysis, any
of the diastereoisomer 8. We removed the chiral auxiliary with magnesium
4479

4 4 8 0
benzyloxide, and used the benzyl ester 9 in our allylsiiane synthesis, 9
the highly diastereoselective aldol reactions of enolates and stereospecific decarboxylative eliminations.7
Since the E-l and its Z isomer Z-l will give opposite enantiomers on electrophilic substitution, it
10
1 1
E-l or Z-l, based on
was necessary to be as thorough in removing each from the other as we had been in setting up the
bearing stereogenic centre in the first place. We achieved this using repeated column chromatography on silica
gel heavily impregnated with silver nitrate. After this procedure, the allylsilanes E-l and Z-l were both
geometrically pure, with
of the other present in each, as determined by careful GC analysis.
more
soluble
99
.
.
1
less
soluble
89%
7
de
82%
0
Ph
9
96%
1 0 7 8 %
0
Me0
A
1.
2.
M e 0
z - 1
E -l
E:Z
reflux, 2 h
11 73%
1.
2.
Py
+
E -l
reflux, 5 h
z - 1
We carried out the reaction E-l
2 twice, using adamantyl chloride as a representative simple
and a catalytic amount of titanium tetrachloride at -78
and Z products 2, which were present in a ratio of
with the same result each time. We
using the same silver
separated the
impregnated column, obtaining each free of the other
as determined yet again by careful GC
analysis. We measured the enantiomeric purity of both alkenes by ozonolysis, followed by reduction with
sodium borohydride, and derivatisation with Mosher’s acid
We were unable to use GC analysis at this stage,

4481
but the
12 and 13 to within 1%. We find that the major product
minor product E-2 is present as a 90: 10 mixture of enantiomers. Clearly the degree of stereospecificity is, as
very high, but is also measurably incomplete, in agreement with our and those of
and
spectra allowed us to measure the diastereoisomeric ratio of the final products
is enantiomerically pure but that the
that, when other stereochemical constraints are present, the extent to which allylsilane
reactions are
is easily eroded.
1 .
-78
2 min
0 “C, 1 h
2.
3 .
acid
DCC. DMAP
0
3 h
2
1 3
50% 90: 10 from E-2 derived
E-l
60%
73%
from Z-2 derived from E-l
from E-2 derived from Z-1
We offer two simplified explanations for why the Z product should be formed with higher enantiomeric
purity. Attack on the allylsilane E-l in a conformation close to 15 may take place on the lower surface more
selectively than attack takes place on the upper surface of the alternative conformation 14, because the lower
+
Ad+
0 %
6 0 %
4
%
H
Ad+
Ad+
1
4
1 5
surface of 15 is occupied by a hydrogen atom, whereas the upper surface of 14 is occupied by a methyl group.
This argument assumes that of the product is formed by attack taking place in conformation 14, and that
E
of the Z product is formed by attack taking place in conformation 15; in other words, there is no rotation about
the
bond in the intermediate cations before the silyl group is plucked off by a nucleophile, presumably
chloride ion. Alternatively, the intermediate cation 16, produced by attack on the lower surface of the
conformation 15, may change its conformation, by rotation about the bond 16 before the silyl
group is lost, to a greater extent than the intermediate 18, produced by attack on the upper surface of
conformation 14, changes its conformation, because the lowering of energy is greater in the former case.
We have carried out one rather inconclusive experiment to try to find out which of these explanations is
the more plausible. We repeated the
allylsilane E-l. In each of these reactions, we obtained only the E-product,
of This result is consistent with electrophilic attack taking place only in conformation 19 and not
reaction three times using the Z-allylsilane Z-l in place of the
which proved to be a mixture

4 4 8 2
in conformation 20, with a
the reaction taking place in the high-energy conformation 20, with the intermediate cation, not implausibly,
changing its conformation completely by rotation about the bond before the loss of the group.
ratio of attack from above to that from below. It is also consistent with 5% of
Because the
ratio in this experiment and the 90: 10 ratio in the experiments on the E-isomer are so similar,
Ad+
0%
H
Ad+
-
-
we are unable to distinguish between these explanations, and must rest on the possibility that a combination of
the two is also possible. In summary, we have confirmed that the
reaction of an allylsilane is highly
stereospecific, and have detected that, nevertheless, it is not completely so.
We thank the Conselleria de Cultura
i Ciencia of the Generalitat Valenciana for a grant (SG), the
SERC for a maintenance award (MJCB) and Nathalie Maugein for assistance with some of the preparative
work.
REFERENCES and NOTES
1.
Hayashi, T.; Konishi, M;
4965; Hayashi, T.; H.; Kumada, M. Tetrahedron
M; Kumada, M. J. Chem. Chem. Commun.
H.; Kumada, M.
Am. Chem.
1982,104, 4962-4963 and 4963-
4605-4606; Hayashi, T.; Konishi,
Hayashi, T.; Kabeta, K.; Yamamoto,
5661-5664; Hayashi, T.; Okamoto, Y; Kabeta,
4224-4226; Hayashi, T.; Konishi, M;
3772-3781; Hayashi, T.; Matsumoto,
T.; Tamao, K.; Kumada, M. Tetrahedron
K.; Hagihara, T.; Kumada, M.
Org. Chem.
Okamoto, Y.; Kabeta, K.; Kumada, M. Org. Chem.
Y.;
Y.
884-685; Hayashi, T.; Matsumoto, Y.;
Y. Chem.
1 9 8 7 , 2 0 3 7 - 2 0 4 0 .
2 .
3 .
Wetter, H.-J; Scherrer, P. Helv. Chim.
Wetter, H.-J; Scherrer, P.; Schweitzer, W. B. Helv. Chim.
Matassa, V. G.; Jenkins, P. R.; Kiimin, A.; Damm, L.; Schreiber, J.; Felix, D.;
118-122. For an exception, however, see:
1985-1989.
E.; Eschenmoser,
A. Israel J. Chem.
321-343.
4 .
5.
Oppolzer, W.; Mills, R. J.; Pachinger, W.; Stevenson, T.
Helv. Chim.
1542-1545.
We checked that this sample of camphor sultam was enantiomerically pure following: Oppolzer, W.;
Vandewalle, M.; Van der Eycken, J.; Vullioud,
C. Tetrahedron
6 .
This is a conservative estimate of the amount of isomer that we could have detected, on the expanded GC
traces, based on comparisons with the amount of various (and unknown) adventitious impurities present
to the easily measurable extent of 0.62%.
7 .
Fleming, I.; Sarkar, A. K. Chem.
Sasaki, T.; Usuki, A.;
Chem. Commun. 1986, 1199-1201.
M.
Org. Chem.
3559-3564.
Dale, J. A.; Dull, D. L.; Mosher, H.
recognition between Mosher’s acid and the enantiomeric alcohols in the corresponding racemic
alcohol.
Org.
2543-2549. There was no appreciable
We assume that the reaction is anti stereospecific, not syn, there being little doubt that it would not be.
This result
was no loss
confirms that the samole of Mosher’s acid was enantiomericallv oure. and that there
purity during the synthesis and degradation, including, most worryingly,
the chromatographic separations, where silver-coordination might have allowed some racemisation.
1 2 .
1 3 .
Au-Yeung, B.-W.; Fleming, I. Tetrahedron, Supplement
Supplement No. 1, 13-24;
Fleming, I.; Williams, R. V. J. Chem.
Trans.1 1981, 684-688; Fleming, I.; Terrett, N. K.
J. Orgunomet. Chem. 1984,264, 99-118.
Young, D.;
W.;
G. Tetrahedron Lett., 1983, 24, 5789-5792;
6 1 2 -6 1 4 ; Orgunometallics 541-547;
G;
W.;
W. J. Org. Chem.
B.; Maynard, I.; Penman, K.
product was just detectable as 0.2% of the mixture, too little to measure its optical purity.
Chem.
Chem. Commun. 1986, 954-955.
1 4 .
15.
The
The three runs gave ratios of enantiomer of
and
(Received in UK 11 May 1992)