observed diastereoselectivity.8 However, one other sim-
plistic model (not dissimilar to a FelkinꢀAnh model)
which is able to rationalize successfully the experimental
data in this case uses insight obtained from single crystal
X-ray diffraction analysis of R,β-unsaturated ester (()-22.
This revealed a solid state conformation in which the C(4)-
hydrogen atom lies almost perpendicular to the plane
of the R,β-unsaturated system, with the bulky C(4)-
N,N-dibenzylamino substituent occupying the less hindered
“outside” position and the C(4)-methyl group in the more
hindered “inside” position (Figure 1). Conjugate addition
of lithium amide 29 to (()-22 in this conformation would
be predicted to occur on the least hindered face past the
“small” hydrogen substituent to give (()-3,4-syn-32, as
observed experimentally. A similar analysis applied to
R,β-unsaturated esters (()-23ꢀ26 (R=Et,Bn,iBu, CH2OBn)
would also successfully rationalize the observed substrate
diastereofacial control, leading to (()-3,4-syn-33ꢀ36.
However, increased steric bulk of the C(4)-substituents in
R,β-unsaturated esters (()-27 and (()-28 would serve to
disfavor analogous conformations, thereby providing a
rationale for their differing behavior. Presumably, the very
large steric congestion around C(4) in the case of (()-28
(R = iPr) precludes addition of the sterically demanding
lithium amide to C(3) completely.
Scheme 3
a 70% of a sample of 46 in 95:5 dr was also isolated.
(()-47 [and, hence, (()-44, (()-46, and (()-48] are in
accordance with that predicted by the transition state
mnemonic developed by us to rationalize the exceptional
facial bias of this class of lithium amide.9 This reagent
control, when combined with that of the R,β-unsaturated
ester (substrate control: production of the 3,4-syn-diaster-
eoisomer favored), results in very highly selective reac-
tions. These results suggest that R,β-unsaturated esters
(()-22ꢀ26 are viable substrates for our PKR protocol4
(Scheme 3).
The PKR of R,β-unsaturated esters (()-22ꢀ26 using a
50:50 pseudoenantiomeric mixture of lithium amides (R)-
30 (2 equiv) and (S)-31 (2 equiv) was next investigated.
These reactions produced, in each case, a 50:50 mixture of
thecorresponding (RR)-adducts44ꢀ48in g95:5drand the
(RS)-adducts 49ꢀ53 in g95:5 dr. Facile separation and
purification via flash column chromatography allowed
isolation of (RR)-44ꢀ48 in >99:1 dr and 31ꢀ42% yield
and (RS)-49ꢀ53 in >99:1 dr and 40ꢀ43% yield. In each
case, the product of addition of lithium amide (R)-30 was
spectroscopically identical to the major diastereoisomer
formed in the corresponding MKR reaction. Additionally,
the relative configuration within β,γ-diamino ester 49 was
unambiguously established via single crystal X-ray diffrac-
tion analysis, with the absolute (3R,4R,RS)-configuration
being assigned from the known (S)-configuration of the
R-methylbenzyl stereocenter. Given the pseudoenantiomeric
nature of lithium amides (R)-30 and (S)-31, this analysis
Figure 1. Chem 3D representation of the single crystal X-ray
diffraction structure of (()-22 [(S)-enantiomer depicted; se-
lected H atoms are omitted for clarity], and Newman projection
along the C(3)ꢀC(4) bond.
The extent of enantiorecognition between R,β-unsatu-
rated esters (()-22ꢀ26 (which offered high levels of sub-
strate control) and lithium amide (()-30 was next in-
vestigated, with high levels of enantiorecognition being
expected. Indeed, addition of 1.6 equiv of lithium amide
(()-30 to (()-22ꢀ26 gave, in each case, essentially a single
diastereoisomeric product 44ꢀ48 in g95:5 dr, indicating
very high levels of enantiorecognition between substrate
and reagent, and consistent with E g196 in each case.
Purification facilitated isolation of diastereoisomeric-
ally pure (>99:1 dr) samples of (()-44ꢀ48. The relative
(3RS,4RS,RSR)-configurations within (()-45 and (()-47
were unambiguously established by single crystal X-ray
diffraction analyses, and therefore the relative configura-
tions within (()-44, (()-46, and (()-48 were assigned by
analogy. It is notable that the relative configurations of the
C(3)- and C(R)-stereogenic centers within both (()-45 and
(9) Costello, J. F.; Davies, S. G.; Ichihara, O. Tetrahedron: Asym-
metry 1994, 5, 1999. For a review, see: Davies, S. G.; Smith, A. D.; Price,
P. D. Tetrahedron: Asymmetry 2005, 16, 2833.
ꢀ
(10) Hoang, C. T.; Bouillere, F.; Johannesen, S.; Zulauf, A.; Panel,
ꢀ
C.; Pouilhes, A.; Gori, D.; Alezra, V.; Kouklovsky, C. J. Org. Chem.
2009, 74, 4177.
(11) In the case of β,γ-diamino ester 48, the yield of the corresponding
4-aminopyrrolidin-2-one 68 (60% isolated yield) was somewhat com-
promised by the formation of methyl (3S,4R)-3,4-diacetamido-5-hydro-
xypentanoate 69 (25% isolated yield). This presumably arises from
competing lactone formation (rather than lactam formation) from 58
under the reaction conditions, followed by methanolysis upon workup
of the acetylation procedure.
220
Org. Lett., Vol. 14, No. 1, 2012