Table 1. Effect of Catalyst Loading on Autocatalytic
Reduction of 2,6-Diacetylpyridine 1
entrya
2a (mol %)
Zn(OTf)2 (mol %)
eeb (%)
deb (%)
1
2
3
4
5
0
0
10
20
20
10
20
5
10
120
0
0
28
38
40
35
45
25
41
47
a Conversions were determined by 1H NMR and were >80% after 24 h.
b Ee and de were determined by chiral HPLC on a Chiralcel OD-H column
and corrected for the added diol 2.
Figure 1. Effect of additives on the reduction of 2,6-diacetylpy-
ridine with sodium triacetoxyborohydride in the presence of
stoichiometric amounts of zinc trifluoromethanesulfonate in dichlo-
1
romethane. Conversions were determined by H NMR after 24 h.
In a typical reduction experiment, 2,6-diacetylpyridine (1)
(10 mM) was treated with sodium triacetoxyborohydride (20
mM), p-nitrophenol (PNP) (60 mM), (R,R)-2 (20 mol %),
and zinc trifluoromethanesulfonate and stirred for 24 h at
room temperature. The double reduction proceeded to 90%
conversion in dichloromethane, and the diol (2) obtained after
basic workup was determined by chiral HPLC to be
enantiomerically enriched in the (R,R) form (40% ee and
47% de, Table 1, entry 5). The % conversion, ee, and de
were corrected for the added diol 2.
Interestingly, diastereoselectivity was observed even with-
out addition of the chiral diol. When 1 was reduced in the
presence of catalytic amounts of zinc trifluoromethane-
sulfonate, diol 2 was produced in a diastereomerically
enriched form, where the homochiral form ((R,R)- and (S,S)-)
of the diol predominates, but no enantiomeric excess was
observed (Table 1, entries 1 and 2). The amount of zinc in
the system can be reduced to catalytic amounts without
significant change in selectivity (Table 1, entry 4), perhaps
due to the poor solubility of the free zinc in dichloromethane.
The reducing system used in this reaction is generated in
situ by a novel combination of sodium triacetoxyborohydride
and p-nitrophenol. A variety of additives with acidic protons
were investigated for the reduction of 2,6-diacetylpyridine
1, and p-nitrophenol was found to be the most effective
(Figure 1).
(PhOH, phenol; HFP, 1,1,1,3,3,3-hexafluoro-2-propanol; PNP,
p-nitrophenol; PhSH, thiophenol; PFP, pentafluorophenol).
system is insensitive to the presence of aromatic rings but is
affected by the acidity of the additive.
p-Nitrophenol reacts with sodium triacetoxyborohydride
to produce tris(p-nitrophenoxy)borohydride 3 (eq 1). The
complete substitution of the acetate groups with p-nitrophe-
noxy (OPNP) groups to produce the new borohydride species
1
3 is evident through H NMR, where the acetate signal at
1.99 ppm is replaced by the acetic acid signal at 2.08 ppm
after only 5 min at room temperature in dichloromethane
(see Supporting Information). The 11B NMR spectral analysis
in methanol indicates the presence of a single species at (δ
18.6). Alcohols with enhanced acidity have been shown to
react with sodium borohydride to generate tris(alkoxy)-
borohydrides.9 In contrast to the well-known disproportion-
ation of alkoxyborohydrides, these tris(alkoxy)borohydrides,
formed from acidic alcohols, are quite stable and resistant
to any disproportionation.
The reduction worked in the presence of other proton
sources. Thiophenol (PhSH) was found to be only slightly
less effective than p-nitrophenol. The reduction was less
effective with additives such as pentafluorophenol (PFP),
1,1,1,3,3,3-hexafluoro-2-propanol (HFP), and phenol (PhOH).
1,1,1,3,3,3-Hexafluoro-2-propanol (HFP) was just as effec-
tive as phenol (PhOH), which leads us to believe that the
In the absence of p-nitrophenol, no appreciable enantio-
selectivity was observed; therefore, we believe 3 to be
important for the construction of the proposed active catalyst
4 (Scheme 1).
Molecular mechanics computation10 of 1 coordinated to
the proposed active catalyst (R,R)-4 shows that the two Re-
faces of 1 are blocked by the tris-(alkoxy)borohydride groups
of 4 (see Supporting Information for a graphic illustration).
(6) Ohkuma, T.; Koizumi, M.; Yoshida, M.; Noyori, R. Org. Lett. 2000,
2, 1749.
(7) (a) Chen, G. M.; Brown, H. C.; Ramachandran, P. V. J. Org. Chem.
1999, 64, 721. (b) Sablong, R.; Osborn, J. A. Tetrahedron Lett. 1996, 37,
4937. (c) Jiang, Q. Z.; VanPlew, D.; Murtuza, S.; Zhang, X. M. Tetrahedron
Lett. 1996, 37, 797.
(9) Golden, J. H.; Schreier, C.; Singaram, B.; Williamson, S. M. Inorg.
Chem. 1992, 31, 1533.
(8) (a) Le Goanvic, D.; Holler, M.; Pale, P. Tetrahedron: Asymmetry
2002, 13, 119. (b) Brown, H. C.; Chen, G. M.; Ramachandran, P. V.
Chirality 1997, 9, 506.
(10) Molecular mechanics computation was performed using HyperChem
5.0 from HyperCube, Inc., Gainesville, Florida. Octahedral geometry was
assumed for the Zn(II) complex.
3948
Org. Lett., Vol. 5, No. 21, 2003