direct aldol reactions of the R-phenylsulfanyl aldehydes 1-3
with ketones as donors under the List-Barbas conditions1a
(Scheme 1). Moreover, following our current research interest
tion products (see the Supporting Information). For the study
of the acid-catalyzed ring expansion of 1a-d and 2a,b, we
used as a model compound the aldols 1a and 2a whose
absolute configuration (R) was known by X-ray analysis (see
the Supporting Information). A preliminary attempt, in
different acid conditions, led only to the corresponding
dehydration products by an elimination reaction. For this
reason, we reduced 1a and 2a to the corresponding optically
pure syn- and anti-diols 4 and 9 using either the Prasad’s
reduction protocol10 or NaBH4 without chelating Lewis acid
and then separating the two diastereoisomers by column
chromatography.
Scheme 1
After screening different acids, we treated the cyclopropyl
diols 4 with SnCl4 in CH2Cl2 at rt, and unexpectedly, we
did not isolate the corresponding cyclobutanones. As a matter
of fact, (1R,3S)-anti-4 gave a mixture of two cis-fused (see
Supporting Information) cyclobutyl tetrahydrofuran diaste-
reoisomers (1R,3S,5S)-6 and (1S,3S,5R)-6 in a 14:86 ratio,
while the (1R,3R)-syn-4 led to the corresponding enantiomers
(1S,3R,5R)-8 and (1R,3R,5S)-8 in a 60:40 ratio.
We suppose that the ring expansion could occur through
two concurrent paths: (a) formation of an episulfonium ion
(path “a”) that allows preservation of the stereochemical
integrity11 of the migrating terminus through two consecutive
inversions during the ring expansion to the cyclobutylth-
ionium ion 5 or 7, that are finally captured by the oxygen
atom of the 3-hydroxyl group; (b) intermediacy of a
carbocation (path “b”) that causes a loss of stereochemical
integrity of the migrating terminus and consequent formation
of both 5 and 5′ or 7 and 7′ (Scheme 2).
in the preparation of carbocyclic7 and heterocyclic8 com-
pounds using cyclopropyl or cyclobutyl derivatives, we were
also interested in the study of the acid-catalyzed ring
expansion of the corresponding cyclopropyl and cyclobutyl
aldols, obtainable from 1 and 2, hopefully in an enantioen-
riched form.
The reaction of 1-3 with different ketones in the presence
of proline led to the corresponding aldol products 1a-d,
2a,b, and 3a,b in moderate to good yields and with ee values
in the range of 95-99% (Scheme 1).
The chiral discrimination increased with the increasing
bulkiness of the ring sizes (Table 1, entries 1-6), while a
Table 1. Direct Asymmetric Aldol Reactions Catalyzed by
L-proline.
entry aldol
n
R
R′
time (h) yielda (%) eeb (%)
If we make the reasonable hypothesis that the stereochem-
istry of the carbon carrying the 3-hydroxy group in both the
diols 4 is preserved, it is evident that a different amount of
epimerization of the migrating terminus is occurring in the
two isomers, during the ring expansion. This fact should be
a consequence of the different importance of the two possible
reaction paths for the two diols 4.
In Scheme 2, it appears that the reactions following path
“a” with the anti-diol 4 should lead to the corresponding
tetrahydrofuran (1S,3S,5R)-6 with the two large group in a
trans position, while the syn-diol 4 should lead to the tetra-
hydrofuran (1S,3R,5R)-8 with the two large substituents in
a less stable cis position. This fact could imply that path “a”
will be more important in the reaction of the anti-diol 4 rather
than in the syn-diol 4, with a consequent high level of
preservation of the stereochemistry of the migrating terminus.
Therefore, the reaction of the syn-diol 4 will have a higher
percentage of the ionic mechanism (path “b”) with major
erosion of the stereochemical integrity of the migrating
terminus.
1
2
3
4
5
6
7
8
1a
2a
3a
1b
2b
3b
1c
1d
1
2
3
1
2
3
1
1
H
H
H
H
H
H
H
Me
Me
Me
Et
Et
Et
16
16
24
24
24
48
48
16
80
95
98
>99
96
99
>99
96
88
56c
51c
44c
21c
Pr
21c
-(CH2)3-
syn: 29
anti: 51
>99
89
a Isolated yields after column chromatography. b The ee was determined
by chiral-phase HPLC analysis. c Moderate yield due to low conversion.
very small amount of condensation product was isolated only
in the case of entry 1. The preferential formation of the aldol
rather than the condensation products was very likely dictated
from the steric hindrance of the phenylsulfanyl group that
could render more difficult the Mannich-elimination se-
quence.9 In fact, the same reaction with acetone and the
corresponding cycloalkyl carboxaldehydes, lacking the R-phen-
ylsulfanyl group, led exclusively or mainly to the condensa-
Interestingly, when we treated the diols (1R,3R)-syn-9 and
(1R,3S)-anti-9 with SnCl4 we did not obtain the correspond-
ing cyclopentanones5 but the spiro-cyclobutyl tetrahydro-
furans (6R,8S)-syn-10 and (6S,8S)-trans-10, diastereoiso-
merically pure by stereospecific [1,2]-PhS migration.
(6) (a) Bernard, A. M.; Cadoni, E.; Frongia, A.; Secci, F.; Piras, P. P.
Org. Lett. 2002, 4, 2565. (b) Alberti, G.; Bernard, A. M.; Frongia, A.; Piras,
P. P.; Secci, F.; Spiga, M. Synlett 2006, 2242.
(7) Bernard, A. M.; Frongia, A.; Secci, F.; Piras, P. P. J. Chem. Soc.,
Chem. Commun. 2005, 3853.
(8) Bernard, A. M.; Frongia, A.; Secci, F.; Piras, P. P.; Spiga, M. Org.
Lett. 2005, 7, 4565.
(9) (a) Kreher, U. P.; Rosamilia, A. E.; Raston, C. L.; Scott, J. L.; Strass,
C. R. Org. Lett. 2003, 5, 3107. (b) Wang, W.; Mei, Y.; Li, H.; Wang, J.
Org. Lett. 2005, 7, 601.
(10) Chen, K.-M.; Gunderson, K. G.; Hardtmann, G. E.; Prasad, K.;
Repic, O.; Shapiro, M. J. Chem. Lett. 1987, 1923.
(11) Hiroi, K.; Ogata, T. Chem. Lett. 1990, 527.
542
Org. Lett., Vol. 9, No. 3, 2007