DOI: 10.1002/anie.201105182
Hosomi–Sakurai Reaction
A Catalytic Asymmetric Borono Variant of Hosomi–Sakurai Reactions
with N,O-Aminals**
Yi-Yong Huang, Ananya Chakrabarti, Naohide Morita, Uwe Schneider, and Shu¯ Kobayashi*
Catalytic intermolecular cross-couplings between Csp3 centers,
such as O,O-acetals and N,O-aminals, and allyl species are
challenging, but provide convenient access to various impor-
tant substance classes such as homoallyl ethers and amines.[1]
For these transformations, Hosomi–Sakurai reactions using
allyl silicon-based reagents are generally employed.[2,3] These
carbon–carbon bond formations proceed through either
Lewis acid or Brønsted acid activation of the electrophile to
generate a stabilized carbenium ion intermediate that can
react with a silicon-based nucleophile. However, catalytic
asymmetric Hosomi–Sakurai allylations of Csp3 centers have
Scheme 1. Asymmetric borono variant of Hosomi–Sakurai reactions?
B(pin)=pinacolatoboron, PG=protecting group.
proved to be challenging.[2f,j,4]
Allyl boronates are typically employed for additions to
Csp2 centers,[5] although a few Csp3 Csp3 cross-couplings have
À
been reported.[6,7] These nontoxic reagents are intrinsically
less nucleophilic than silicon-based compounds and have
been neglected in the context of allylation of Csp3 centers.
However, allyl boronates may offer significant advantages
such as superior stability and unique reactivity and selectivity.
During a project initially aimed at the catalytic activation of
allyl boronates for selective C–C coupling with more complex
electrophiles, we observed a peculiar reactivity with Csp3
intermediates such as N,O-aminals. These electrophiles are
abundant in natural products[8a] and play an important role in
organic synthesis.[2f,8b–e] Thus, the development of catalytic
asymmetric carbon–carbon bond formations with N,O-ami-
nals is worthwhile.[1] Intrigued by the unexpected reactivity,
we started more detailed investigations with a view towards
asymmetric catalysis. We report herein an approach to
address the challenge of catalytic asymmetric Hosomi–
Sakurai reactions involving Csp3 centers by employing boro-
nates instead of silicon-based reagents.
Hosomi–Sakurai allylations with silicon-based reagents 4a
and 4b barely proceeded,[9] which stands in sharp contrast to
our earlier study.[6a] The substantially higher reactivity of 2
over 4 under mildly Lewis acidic conditions constitutes a
prerequisite for asymmetric catalysis. We postulated a dual
catalytic activation[6a] of rac-1a and 2 to generate iminium ion
and allyl indium(I) intermediates (Scheme 1), thus we
screened potential indium(I) catalysts bearing chiral counter-
anions[9] rather than chiral ligands.[10] In these experiments the
combination of indium(I) chloride and chiral silver binol
phosphate (R)-5a-Ag[11] was found to be the most promising
chiral catalyst system for the formation of product (R)-3a
(e.r. = 88:12).[9] Also, allyl silane 4a proved to be substantially
less effective than allyl boronate 2 in terms of both reactivity
and selectivity.[9] Thus, these results demonstrate the viability
of our originally envisaged asymmetric concept, in which we
proposed the use of boronates instead of silanes. At this stage,
we explain the success of 2 based on its higher propensity to
undergo transmetalation, thereby forming a more reactive
allyl indium species.
Next, we further optimized the reaction conditions
(Table 1). A screen of silver binol phosphates identified (R)-
5b-Ag as the best chiral source (Table 1, entries 1–5). The use
of an apolar cosolvent (cyclopentylmethyl ether) and a slight
excess of the chiral silver salt further improved the asym-
metric induction even at a lower catalyst loading (Table 1,
entries 6–8). We then conducted several control experiments.
In the absence of indium(I) chloride, (R)-5b-Ag displayed
low reactivity and low asymmetric induction (Table 1,
entry 10). The use of chiral Brønsted acid (R)-5b-H, which
may be generated in situ under the present catalysis con-
ditions, did not lead to any reaction (Table 1, entry 11).[4g,8c]
The combination of indium(I) chloride and (R)-5b-H pro-
vided low asymmetric induction (Table 1, entry 12), thereby
demonstrating that an achiral metal salt and a chiral Brønsted
In an initial screen of various Lewis and Brønsted acids for
the reaction of N,O-aminal rac-1a with allyl boronate 2
indium(I) triflate[6a] was identified as the best catalyst for the
formation of homoallyl amide rac-3a (Scheme 1; R = phenyl,
PG = benzoyl, R’ = methyl).[9] In contrast, the corresponding
[*] Dr. Y.-Y. Huang, Dr. A. Chakrabarti, N. Morita, Dr. U. Schneider,
Prof. Dr. S. Kobayashi
Department of Chemistry
School of Science and Graduate School of Pharmaceutical Sciences
The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033 (Japan)
E-mail: shu_kobayashi@chem.s.u-tokyo.ac.jp
[**] This work was partially supported by a Grant-in-Aid for Scientific
Research from the Japan Society for the Promotion of Science
(JSPS), ERATO (JST), NEDO, and GCOE. We thank Mr. Takeshi
Yamakawa for the preparation of several precursors of chiral silver
binol phosphates (R)-5a–e.
Supporting information for this article is available on the WWW
Angew. Chem. Int. Ed. 2011, 50, 11121 –11124
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11121