Table 2: Study of addition reactions of 3 to 5 (Scheme 3).
ical autocatalytic effect of the product can become beneficial
in the catalytic, enantioselective synthesis of important
pharmaceutical intermediates. The possibility of the practical
application of this autocatalysis concept, which is otherwise
largely of academic significance, is to the best of our
knowledge unprecedented.
Entry
4
(S)-6
[mol%]
T
[8C]
t
[h]
Product (S)-6
yield [%][a] ee [%][b]
[mol%]
1[c,d]
2[c,d]
3[c]
24
–
18
18
–
40
40
40
40
7
2
6
2
53[f]
20
24
30
30
58(38)[e]
51(36)[e]
49(30)[e]
76(70)
91(88)
94(92)
4[c,g]
In summary, we have documented a catalytic enantiose-
lective process for the production of a key precursor to
efavirenz. Key to the development was the use of the product
as an inherent part of the catalytic system. This novel
approach towards zinc acetylide addition to ketones expands
the arsenal of existing autocatalytic transformations and also
poses general challenging questions about the productꢀs role
in asymmetric reactions. In a manner that is complementary
to Soaiꢀs involving autocatalysis, the strategy described herein
has demonstrated the conversion of a catalytic, enantioselec-
tive transformation, which is poorly selective even in the
presence of an external chiral catalyst, into a highly selective
process. It also provides an example wherein the presence of
the product as well as a second chiral catalyst can work in
synergy to generate a catalytic and enantioselective reaction
process. Following the strategy outlined, we believe the
manufacturing cost of efavirenz could be substantially
reduced in comparison to this of the existing stoichiometric
process. There has been considerable public debate on the
affordability of medicines; in this respect the approach we
describe may provide wider access to therapy to patients
worldwide. Additionally, we believe that this is fertile
territory for future explorations in the field.
[a] Yields after subtraction of the initially added product are given in
parentheses. [b] ee values were determined by HPLC with a Daicel
Chiralpak AD-H column, hexanes/iPrOH 85:15, 1 mLminÀ1. Values
corrected for initially added (S)-6 ligand are given in parentheses (see
Ref. [15]). [c] Performed with (1R,2S)-4 unless otherwise stated, 2 mmol
scale, THF(major component)/Tol/hexanes, 0.24 equiv of Et2Zn,
0.9 equiv of nHexLi, 2 equiv of 3. [d] 0.48 equiv of Et2Zn. [e] Yield of
isolated product after purification by chromatography. [f] Product was
not isolated; yield was calculated by HPLC analysis of the reaction
aliquot. [g] Performed with (1S,2R)-4.
but no ligand 4, the alcohol adduct (S)-6 was obtained in 58%
yield and 76% ee (Table 2, entry 2).[22] This finding is intrigu-
ing, because, unlike (S)-2, the product as a chiral controlling
group is dominant over ligand 4 in promoting the reaction and
dictating the configuration of the newly formed product (S)-6
(compare with Table 1, entry 6). In a third experiment, when
the addition was conducted in the presence of both ligand 4
and the autocatalyst product ((S)-6/(1R,2S)-4 0.3:0.18), (S)-6
adduct was isolated in 51% yield and 91% ee (Table 2,
entry 3).[23]
Finally, a fourth experiment was carried out to examine
whether there was a matched and mismatched pair of
combinations involving (1S,2R)-4, (1R,2S)-4, and (S)-6.
Surprisingly, the analysis of the reactions indicated enantio- Experimental Section
General experimental procedure and characterization of the prop-
selective formation of (S)-6 regardless of the amino alcohol
ligand employed (Table 2, entries 3 and 4).[24] This dominant
effect of product as an autocatalyst which dictates the
absolute configuration of the adduct 6 was not detected in
the case of (S)-2 (Table 1, entry 7). Consequently, the config-
uration of the product autocatalyst is crucial for asymmetric
induction in this transformation.[25] Although the enantiose-
lectivity and yield were only moderate, this system exhibits
the very first example of an asymmetric autocatalytic
alkynylation of a ketone, where the product constitutes the
only source of chiral catalyst (Table 2, entry 2).
The origin of the enantiofacial bias in asymmetric zinc
alkylation reactions mediated by rigid bidentate b-amino-
alcohols is well described by the Noyori–Kitamura five-ring
chelate model.[26] The enantioselective addition of lithium
acetylide–ephedrate to the PMB-protected aminoketone 1
(PMB = p-methoxybenzyl) was accounted for by the active
2:2 cubic tetramer, which was well characterized by spectro-
scopic and structural methods.[27] By contrast, reliable mech-
anistic studies on 1,2-alkynylzinc additions, in particular
catalytic reactions, are lacking. We have obtained a crystal
structure of a complex formed by Et2Zn, 4, and 6, which is
competent as a catalyst, albeit with reduced selectivity, and is
instructive to examine (see the Supporting Information).
However, at present a detailed construct to rationalize the
autocatalytic effects we observed must await additional in-
depth mechanistic studies. Nonetheless, the phenomenolog-
argylic alcohols, as well as spectroscopic data of the discussed
compounds can be found in the Supporting Information. CCDC-
652045 and 652046 contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Received: October 25, 2010
Published online: March 1, 2011
Keywords: asymmetric catalysis · autocatalysis · efavirenz ·
.
tertiary alkynols · zinc acetylides
[1] a) P. G. Cozzi, R. Hilgraf, N. Zimmermann, Eur. J. Org. Chem.
Soai, T. Kawasaki, I. Sato in The Chemistry of Organozinc
Compounds, Part 2 (Eds.: Z. Rappoport, I. Marek), Wiley,
Chichester, 2006, pp. 555 – 593.
[2] For selective recent achievements in zinc acetylide additions to
4940 – 4941; d) B. M. Trost, A. H. Weiss, A. Jacobi von Wangelin,
g) M. Li, X.-Z. Zhu, K. Yuan, B.-X. Cao, X.- L. Hou,
Angew. Chem. Int. Ed. 2011, 50, 2957 –2961
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim