Suggestive evidence that directed reductions of alkoxy
ketones might be possible in this system was first obtained
in studies of the scope of this method in the reduction of
certain â-hydroxy ketones. For example, reduction of the
â-methoxy substrate 1 below (Scheme 1) gave an unusually
to those obtained with the free hydroxy substrate (entry 1).
Thus the methyl ether substrate affords an almost perfect
result: a 99% yield of reduction products is obtained with a
99:1 level of stereoselectivity. It is also clear that there are
marked substituent effects in that the benzyl (Bn) group does
not promote or direct the reaction, and this substrate was
recovered unchanged under conditions where the more
reactive alkoxy substrates are completely consumed. Thus,
both benzyl and TBS groups are uninvolved and are merely
Scheme 1. Reduction of â-Methoxy Ketone 1
“
spectator” groups. The methoxymethyl (MOM), methyl-
thiomethyl (MTM), and methoxyethoxymethyl (MEM) ethers
are all activating and “directing” groups, as is the methyl
ether. It appears that the cutoff between these two types of
alkoxy groups occurs at approximately two contiguous
carbon atoms; thus, the ethoxy substrate is reduced more
slowly than the methoxy derivative. As shown in Table 1,
the ethoxy derivative is only ca. 30% reduced after 18 h at
0
°C, while the methoxy derivative is completely reduced
in 1 h under the same conditions. The stereoselectivity,
however, still significantly favors the anti product. Similarly,
the benzyloxymethyl (BOM) derivative gave a 41% yield
after 18 h at 0 °C with recovered starting material accounting
for the remainder of the material.
low selectivity for this type of reduction, particularly when
compared to the same substrate (3) with methoxy replaced
by hydrogen. The simplest explanation for this result was
that the methoxy group was also directing the reaction, but
in a sense that favored increased formation of the syn-1,3-
diol product.
We have previously attributed the stereochemical outcome
in the reductions of the â-hydroxy ketone substrates as arising
from chelation of samarium in a product-determining sa-
Thus, we have examined a series of â-alkoxy ketones, 5,
in these reductions. The results are summarized in Table 1;
1
marium carbanion. It thus appears that these results can be
taken as an indication of the ability of the alkoxy groups to
participate in this chelation process. It is of considerable
interest to note that this same “cutoff” behavior (i.e., in the
Table 1. Reduction of â-Alkoxy Ketones Using SmI -MeOH
2
series OMe, OEt, OBn) was previously reported in the
3
context of chelation of â-alkoxy aldehydes with TiCl
4
. In
that case, there was a distinct change between methoxy and
ethoxy in terms of the preferred solution structures of the
chelates and their resultant reaction chemistry, although
chelate formation was seen with all three alkoxy groups and
entry
R
time (h)
yield (%)
ratio (anti:syn)
4
TiCl . It is also of interest to note that similar observations
1
2
3
4
5
6
7
8
(a )
(b)
(c)
(d )
(e)
(f)
H
1
1
18
12
1
4
18
12
95
99
30
71
92
87
41
NR
99:1
99:1
87:13
95:5
95:5
91:9
86:14
have been made with respect to the structure of the alcohol
additive used in these reactions. The effectiveness of the
alcohol decreases in the series MeOH, EtOH, i-PrOH,
t-BuOH; with tert-butyl alcohol, no reduction at all is
observed with the hydroxy substrates at 0 °C.
The examples shown in Table 1 are all structurally simple
cases chosen to facilitate comparisons between alkoxy groups
in the absence of complicating and unknown effects due to
substrate structure that could be encountered in more
complex systems. However, applications of this process to
solve problems encountered in certain reductions needed in
ongoing synthetic investigations in our labs have been carried
out in structurally more complex systems with good results.
In this context, the reduction of alcohol 7a and its derivatives
Me
Et
MOM
MTM
MEM
BOM
Bn
(g)
(h )
results for the free hydroxy substrate are also included for
comparison. Stereochemical assignments for the products of
these reactions were made by chemical correlation with
known samples of the anti-1,3-diols.
This usually involved simply the removal of the protecting
group, although in the case of the methyl ether substrate the
assignment was made by methylating the reduction product
and comparing with material obtained by methylation of the
anti-1,3-diol. The structure of this anti diol was originally
assigned by NMR methods and confirmed in the course of
this work by X-ray crystallographic analysis.
7b and 7c are informative and also illustrate the implementa-
tion of a procedure more amenable to large scale reactions
than that previously employed, which is run at 0.1 M.
Reduction of the free alcohol 7a is complicated by an
unexpectedly facile retro-aldol cleavage that results in a low
The results clearly show that the selectivity and chemical
yields obtained with certain â-alkoxy ethers are comparable
(
3) Keck, G. E.; Castellino, S.; Wiley, M. J. Org. Chem. 1986, 51, 8478-
8480.
2308
Org. Lett., Vol. 2, No. 15, 2000
Keck
