or Me (with R2 ) H).2b Therefore, a search for improved
reaction conditions was initiated, and we describe herein a
new cobalt-based catalyst system that exhibits broader scope
for amide-tethered substrates and allows products to be
isolated with higher yields.
cyclizations5b,c also proved ineffective, providing complex
mixtures (entries 4 and 5). In light of recent reports of
organometallic reagents with â-hydrogen-containing alkyl
groups being utilized as stoichiometric reductants for a
variety of transition-metal-catalyzed reductive couplings,6 we
examined Et3B6b-e and Et2Zn6a,b in our reaction. In the
presence of 5 mol % of Co(acac)2 hydrate (degree of
hydration ∼ 2), Et3B resulted in no reaction (entry 6), but
we were delighted to observe that the more reactive Et2Zn
led to the formation of 2a in 89% yield with none of the
side product 3 being observed (entry 7). No reaction occurs
in the absence of Co(acac)2‚2H2O.
Initial investigations began with the cyclization of cin-
namic amide 1a (Table 1). Application of our previously
Table 1. Survey of Reaction Conditions
With effective conditions identified, the scope of the
process was next explored (Table 2). Substrates containing
a wide range of substitution at both the R,â-unsaturated
amide and the ketone7 underwent cyclization to give 4-hy-
droxypiperidin-2-one products in generally excellent yields
and high diastereoselectivities8 (entries 1-12). It should be
noted that the copper conditions (as in Table 1, entry 1)
proved ineffective in the majority of these examples. In a
number of reactions, incomplete conversions were observed
using Co(acac)2‚2H2O (method A), but the combination of
CoCl2 and the electron-rich phosphine Cy2PPh (method B)
was found to provide good results in these cases (entries
10-12). The reaction could also be applied to the formation
of pyrrolidin-2-ones (entries 13-15), though with somewhat
lower yields and diastereoselectivities. Although 5 mol %
of the cobalt source was employed for convenience in these
experiments, the reaction is tolerant of lower catalyst
loadings. For example, on a 5 mmol scale, substrate 1a
underwent cyclization using 0.5 mol % of Co(acac)2‚2H2O
to provide 2a in 79% yield (entry 2).
temp
(°C)
conv
entry
1
reagents
solvent
THF
(%)a 2a/3a
Cu(OAc)2‚H2O, DPPF,
TMDS (1 equiv)
Cu(OAc)2‚H2O, rac-
BINAP, TMDS (1 equiv)
Cu(OAc)2‚H2O, DPPF,
PhSiH3 (1 equiv)
Co(dpm)2,
PhSiH3 (1 equiv)
Co(dpm)2,
PhSiH3 (1 equiv)
Co(acac)2‚2H2O,
Et3B (2 equiv)
rt
77
34:66
39:61
na
2
3
4
5
6
7
THF
THF
rt
rt
87
<5
CH2Cl2 rt
ca. 90 nab
DCE
rt to 50 ca. 80 nab
THF/
hexane
THF/
0 to rt <5
na
Co(acac)2‚2H2O,
Et2Zn (2 equiv)
0 to rt >95
>95:5c
hexane
1
a Determined by H NMR analysis of the unpurified reaction mixture.
b A complex mixture containing unidentified side products was obtained,
with only a trace (<10%) of 2a present. c Product 2a was isolated in 89%
yield. PMP ) para-methoxyphenyl, DPPF ) 1,1′-bis(diphenylphosphino)-
ferrocene, TMDS ) 1,1,3,3-tetramethylhydrosiloxane, BINAP ) 2,2′-
bis(diphenylphosphino)-1,1′-binaphthyl, dpm ) dipivaloylmethane, acac )
acetonylacetonate.
Difficulties were encountered when substrates 4a,b con-
taining phenyl ketones were employed; in contrast to methyl
ketones 1l-n (Table 2, entries 13-15), the desired pyrro-
lidin-2-ones were obtained in <20% yield along with
numerous other side products. However, replacement of Et2-
(4) (a) v. Matt, P.; Pfaltz, A. Tetrahedron: Asymmetry 1991, 2, 691-
700. (b) Gieger, C.; Kreitmeier, P.; Reiser, O. AdV. Synth. Catal. 2005,
347, 249-254. (c) Yamada, T.; Ohtsuka, Y.; Ikeno, T. Chem. Lett. 1998,
1129-1130. (d) Ohtsuka, Y.; Ikeno, T.; Yamada, T. Tetrahedron: Asym-
metry 2003, 14, 967-970.
(5) (a) Isayama, S.; Mukaiyama, T. Chem. Lett. 1989, 2005-2008. (b)
Baik, T.-G.; Luis, A.-L. Wang, L.-C.; Krische, M. J. J. Am. Chem. Soc.
2001, 123, 5112-5113. (c) Wang, L.-C.; Jang, H.-Y.; Roh, Y.; Lynch, V.;
Schultz, A. J.; Wang, X.; Krische, M. J. J. Am. Chem. Soc. 2002, 124,
9448-9453.
(6) For selected examples, see: (a) Kimura, M.; Miyachi, A.; Kojima,
K.; Tanaka, S.; Tamaru, Y. J. Am. Chem. Soc. 2004, 126, 14360-14361.
(b) Kimura, M.; Ezoe, A.; Shibata, K.; Tamaru, Y. J. Am. Chem. Soc. 1998,
120, 4033-4034. (c) Molinaro, C.; Jamison, T. F. Angew. Chem., Int. Ed.
2004, 44, 129-132. (d) Molinaro, C.; Jamison, T. F. J. Am. Chem. Soc.
2003, 125, 8076-8077. (e) Miller, K. M.; Huang, W.-S.; Jamison, T. F. J.
Am. Chem. Soc. 2003, 125, 3442-3443. For a review, see: (f) Montgomery,
J. Angew. Chem., Int. Ed. 2004, 43, 3890-3908.
(7) Aldehydes do not serve as competent electrophiles under these
conditions, as they undergo reduction and ethylation instead.
(8) The relative stereochemistries of 2f, 2i, and 2m were confirmed by
X-ray crystallography and matched those of products obtained previously
using copper catalysis (see ref 2). The stereochemistries of the remaining
products were assigned by analogy. See Supporting Information for further
details.
reported copper conditions2 proved ineffective, providing the
desired product 2a but contaminated with significant quanti-
ties of the uncyclized side product 3 along with the starting
material (entry 1). The formation of 3 may be attributed to
conjugate reduction of the R,â-unsaturated amide being slow,
allowing prior reduction of the ketone to become competitive.
Replacement of DPPF with rac-BINAP led to a similar result
(entry 2), whereas use of PhSiH3 in place of TMDS led to
minimal reaction (entry 3). Having obtained no success with
copper-based catalyst systems, our attention turned to the
use of other metals. In conjunction with an appropriate chiral
ligand, the combination of CoCl2 and NaBH4 has proven
useful for the asymmetric conjugate reduction of R,â-
unsaturated amides.4 Unsurprisingly, conditions employing
NaBH4 led to rapid reduction of the ketone of 1a. Conditions
employing cobalt salts that were developed for intermolecular
reductive aldol reactions5a and later extended to aldol
3730
Org. Lett., Vol. 8, No. 17, 2006