Diastereoselectivity in non-aldol aldol reactions: Silyl triflate-promoted Payne rearrangements

Article abstract of DOI:10.1016/S0040-4039(02)01890-7

The non-aldol aldol process exhibits diastereoselectivity such that one isomer of a secondary epoxy silyl ether gives only the expected ketone product while the other affords both ketone and aldehyde. We present evidence for a novel silyl triflate-promoted Payne rearrangement of silyloxy epoxides, e.g. D₃?E₃. Additional examples of the non-aldol aldol process are presented including cyclic examples and one in which an aryl ring participates to give unusual products.

Full text of DOI:10.1016/S0040-4039(02)01890-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 8169–8172
Diastereoselectivity in non-aldol aldol reactions:
silyl triflate-promoted Payne rearrangements
Michael E. Jung* and Alexandra van den Heuvel
Department of Chemistry and Biochemistry, University of California, Los Angeles, 405 Hilgard Avenue, Los Angeles,
CA 90095-1569, USA
Received 13 August 2002; revised 31 August 2002; accepted 3 September 2002
Abstract—The non-aldol aldol process exhibits diastereoselectivity such that one isomer of a secondary epoxy silyl ether gives only
the expected ketone product while the other affords both ketone and aldehyde. We present evidence for a novel silyl
triflate-promoted Payne rearrangement of silyloxy epoxides, e.g. D3lE3. Additional examples of the non-aldol aldol process are
presented including cyclic examples and one in which an aryl ring participates to give unusual products. © 2002 Elsevier Science
Ltd. All rights reserved.
Over the last few years, we have reported1 on the utility
of the rearrangement of optically active tertiary epoxide
primary alcohols and silyl ethers derived from 2-methyl
allylic alcohols via the Sharpless asymmetric epoxida-
tion.2 Treatment of the epoxide prepared from the
E-allylic alcohol with a silyl triflate and a hindered base
gives the syn 2-methyl-3-trialkylsilyloxyaldehydes
whereas similar treatment of those derived from the
Z-isomers affords the anti products. We called this
sequence the ‘non-aldol aldol’ process since aldol prod-
ucts are prepared in a three-step procedure from alde-
hydes with no aldol condensation involvement. We
report herein the extension of this process to the epox-
ides of secondary allylic alcohols which is a much more
complex process, but which showed reasonable
diastereoselectivity. In addition we present evidence for
a silyl triflate-promoted Payne rearrangement of epoxy
silyl ethers and the existence of two distinct epoxide
rearrangement pathways.
31 and 44% yields, respectively.3 Conversion of 3 to the
silyl ether 4 was straightforward while 5 was prepared
by an initial Mitsunobu reaction, hydrolysis and silyla-
tion. Treatment of the diastereomer 4 with a silyl
triflate in the presence of base afforded the expected
non-aldol aldol product 6 in good yield.4 However,
treatment of the diastereomeric epoxy silyl ether 5
under identical conditions gave a mixture of the two
We proposed,
a priori, that one might observe
diastereoselectivity in the rearrangement of the two
diastereomeric tertiary epoxide secondary silyl ethers
since the position of the hydrogen atom that migrates
to the adjacent forming tertiary carbocation center
would be spatially different in the two diastereomers.
This was indeed the case as shown in Scheme 1. The
allylic alcohol 1 (prepared in four steps from butanal)
underwent Sharpless kinetic resolution–epoxidation to
give the optically active alcohol 2 and the epoxide 3 in
* Corresponding author.
Scheme 1.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01890-7

8170
M. E. Jung, A. 6an den Heu6el / Tetrahedron Letters 43 (2002) 8169–8172
rearrangement products, the ketone 6 and the aldehyde
7. We assume that the reason for this diastereoselectiv-
ity involves the difference in the steric environment of
the groups in the likely transition states, A, B1, and B2.
Thus, in A the alignment of the migrating hydrogen
atom causes no severe steric problems, whereas in B1
the same alignment causes the two methyl groups to be
in close proximity and allows for rotation to B2, which
does not have this methyl–methyl interaction, and per-
mits the migration of the alkyl group instead.5
Scheme 4.
D3 had rearranged to the regioisomer E3 (Scheme 4).
This allows the second epoxide rearrangement manifold
(Scheme 3, bottom) to occur (via E1 and E2) and to
afford the other pairs of ketones and aldehydes. As far
as we can tell, this is the first example of a silyloxy
silylated epoxonium ion equilibration, the silyl ana-
logue of the normal base-catalyzed Payne rearrange-
ment, although Lewis acid-catalyzed epoxy alcohol
rearrangements are known.8
Scheme 2.
That this Payne rearrangement was indeed occurring
was shown by treatment (Scheme 5) of the correspond-
The situation is much more complex when a secondary
alcohol with a larger alkyl group is used. For example
when the ethyl-substituted epoxy silyl ether 8 was
treated under the same conditions, a mixture of four
ketones 9a–d and four aldehydes 10a–d was formed
(Scheme 2)!6 We assumed that two ketones and two
aldehydes could be formed by the non–aldol aldol
pathway and a simple epoxide rearrangement in which
either the hydrogen or the propyl group attached to the
epoxide could migrate (via D1 or D2) to the forming
tertiary carbocation center to give, respectively, an
aldehyde or a ketone (Scheme 3, top). However, the
formation of the other four products required that a
silyl triflate-promoted Payne rearrangement7 had
occurred, namely the silyloxy silylated epoxonium ion
Scheme 5.
Scheme 3.

M. E. Jung, A. 6an den Heu6el / Tetrahedron Letters 43 (2002) 8169–8172
8171
ing propyl substituted silyl ether 11, which afforded a
much simpler mixture of the two ketones 12ab and two
aldehydes 13ab due to the symmetry of the system.9
and thus proceeds via a double inversion pathway.
Evidence for this mechanism comes from the formation
of a small amount of the tetralin bis-silyl ether 24.
Presumably the formation of 24 and 22 both proceed
via the reactive intermediate F in which the phenyl ring
has opened the silylated epoxide. Loss of a proton
would afford 24 while hydride transfer from the silyl-
oxymethyl unit would afford 22.
We have also examined cases with different alkyl
groups at the b-position of the epoxide with unusual
results. Thus, the ethyl-substituted epoxy silyl ether 14
affords the expected non-aldol aldol product 15a and
the aldehyde 16 in an 8:1 ratio in good yield (Scheme
6). The structure of the product was proven by conver-
sion into the known hydroxy ketone 15b.10 Cyclic
epoxy silyl ethers also undergo the non-aldol aldol
rearrangement without difficulty. For example, the
epoxy silyl ether derived from cyclohexene–methanol 17
afforded the b-silyloxy cyclohexane–carboxaldehyde 18
in excellent yield and good ee (as determined by the
optical rotation of the alcohol 19 prepared by reduction
of 18).
Thus, the non-aldol aldol process proceeds in a
diastereoselective manner with one isomer giving only
the expected ketone product and the other affording
both ketone and aldehyde via competing hydride and
alkyl group migrations. In addition we have presented
evidence for a novel silyl triflate-promoted Payne rear-
rangement of silyloxy epoxides. Finally, we have shown
the versatility of the non-aldol aldol reaction in several
different substrates, including an electrophilic
cyclization.11
Acknowledgements
We thank the National Institutes of Health (CA72684)
for generous financial support and the National Science
Foundation under equipment grant CHE-9974928.
References
Scheme 6.
1. (a) Jung, M. E.; D’Amico, D. C. J. Am. Chem. Soc. 1993,
115, 12208; (b) Jung, M. E.; D’Amico, D. C. J. Am.
Chem. Soc. 1997, 119, 12150; (c) Jung, M. E.; Karama,
U.; Marquez, R. J. Org. Chem. 1999, 64, 663; (d) Jung,
M. E.; Marquez, R. Tetrahedron Lett. 1999, 40, 3129; (e)
Jung, M. E.; Lee, W. S.; Sun, D. Org. Lett. 1999, 1, 307;
(f) Jung, M. E.; Sun, D. Tetrahedron Lett. 1999, 40, 8343;
(g) Jung, M. E.; Marquez, R. Org. Lett. 2000, 2, 1669; (h)
Jung, M. E.; Lee, C. P. Tetrahedron Lett. 2000, 41, 9719;
(i) Jung, M. E.; Lee, C. P. Org. Lett. 2001, 3, 333.
2. (a) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.;
Masamune, H.; Sharpless, K. B. J. Am. Chem. Soc. 1987,
109, 5765; (b) Rossiter, B. E. In Asymmetric Synthesis;
Morrison, J. D., Ed.; Academic Press: New York, 1985;
Vol. 5, pp. 193–243; (c) Finn, M. G.; Sharpless, K. B. In
Asymmetric Synthesis; Morrison, J. D., Ed.; Academic
Press: New York, 1985; Vol. 5, pp. 247–308.
Finally, we have shown that treatment of a substrate 20
with a phenethyl substituent affords a 3:1 mixture of
two ketones, namely the product of the normal non-
aldol aldol rearrangement 21 and also the product with
retained stereochemistry 22 (along with a trace amount
of the aldehyde 23) (Scheme 7). This is the first case of
a non-aldol aldol rearrangement with retention of
configuration and was a worrisome result until we
realized that is almost certainly due to the participation
of the phenyl group in opening the activated epoxide
3. The optical purity of the two diastereomers was deter-
mined to be about 80% by use of the Alexakis method:
Alexakis, A.; Mutti, S.; Mangeney, P. J. Org. Chem.
1992, 57, 1224.
4. The optical purity of the product 6 was identical to that
of the starting epoxide indicating
a stereospecific
rearrangement.
5. The exact ratio of products 6 and 7 of the rearrangement
of 5 is somewhat dependent on conditions and the silyl
triflate used, with trimethylsilyl and triethylsilyl triflates
(TMSOTf and TESOTf) giving about a 2:1 ratio and
t-butyldimethylsilyl triflate (TBSOTf) affording about a
1:2 to 1:1 ratio, alkenes and other rearrangement prod-
ucts (vide infra) are formed in all cases.
Scheme 7.

8172
M. E. Jung, A. 6an den Heu6el / Tetrahedron Letters 43 (2002) 8169–8172
6. Again the exact ratios of the eight products 9a–d and
(3R,4S)-4-[(1,1-Dimethyl)ethyldimethylsilyloxy]-3-methyl-
heptan-2-one (6, TBS). 1H NMR (400 MHz, CDCl3) l:
3.91 (m, 1H), 2.55 (qd, 1H, J=4.5, 7.0 Hz), 2.18 (s, 3H),
1.33 (m, 4H), 1.04 (d, 3H, J=7.0 Hz), 0.89 (t, 3H, J=7
Hz), 0.89 (s, 9H), 0.06 (s, 3H), 0.04 (s, 3H). 13C NMR
(100 MHz, CDCl3) l: 211.6, 73.5, 52.0, 36.9, 30.1, 25.9,
10a–d vary depending on the exact conditions and the
silyl triflate used. For example, the ratio of the four
aldehydic products, 9a–d, is approximately 3:1.6:1.5:1
when the TES ether is treated with TES triflate and is
4.3:3.3:3:1 when the TBS ether is treated with TBS triflate
(TBSOTf). The ketone ratios are similarly dependent of
which combination of silyl ether and triflate are used.
7. Payne, G. B. J. Org. Chem. 1962, 27, 3819.
8. See for example: (a) Wen, X.; Norling, H.; Hegedus, L. S.
J. Org. Chem. 2000, 65, 2096; (b) Gillmore, A. T.;
Roberts, S. M.; Hursthouse, M. B.; Abdul Malik, K. M.
Tetrahedron Lett. 1998, 39, 3315; (c) Tius, M. A.; Reddy,
N. K. Tetrahedron Lett. 1991, 32, 3605.
9. The structures of the two ketones were determined by
comparison to authentic samples prepared by the Evans
asymmetric aldol reaction, followed by formation of the
Weinreb amide and addition of alkyl Grignard reagent.
10. (a) Hoffmann, R. W.; Ditrich, K.; Froech, S. Liebigs
Ann. Chem. 1987, 1, 977; (b) Utaha, M.; Onoue, S.;
Takeda, A. Chem. Lett. 1987, 971.
18.8, 18.1, 14.2, 11.5, −4.3, −4.5. IR (neat): 1715 cm−1
.
(5S,6R)-6-(Triethylsilyloxy)-5-methylnonan-4-one (12a).
1H NMR (400 MHz, CDCl3) l: 3.88 (m, 1H), 2.63 (dq,
J=5.3, 7.0 Hz), 2.51 (dt, J=7.4, 17.3 Hz), 2.42 (dt, 1H,
J=7.2, 17.3 Hz), 1.57 (m, 2H), 1.28 (m, 4H), 1.04 (d, 3H,
J=7.0 Hz), 0.96 (t, 9H, J=7.9 Hz), 0.90 (2t, 6H, J=7.4
Hz), 0.60 (q, 6H, J=7.9 Hz). 13C NMR (126 MHz,
CDCl3) l: 213.5, 73.7, 51.5, 44.7, 37.2, 18.7, 16.8, 14.2,
13.8, 12.3, 6.8, 5.0. IR (neat): 1713 cm−1
.
(1R,2S)-2-[(1,1-Dimethyl)ethyldimethylsilyloxy]cyclohex-
1
anecarboxaldehyde (18). H NMR (500 MHz, CDCl3) l:
9.71 (s, 1H), 4.38 (m, 1H), 2.23 (m, 1H), 1.88 (m, 1H),
1.8–1.2 (m, 7H), 0.86 (s, 9H), 0.06 (s, 3H), 0.03 (s, 3H).
13C NMR (126 MHz, CDCl3) l: 205.4, 67.6, 54.4, 33.4,
25.6, 24.0, 21.0, 20.4, 17.9, −4.4, −5.2. IR (neat): 1732
cm−1
.
11. Spectroscopic characterization of selected compounds: