Results and Discussion
tolerated on the aromatic moiety (Table 1, entries 1–8), in-
cluding ortho subsituents, which had not been previously re-
ported for this type of substrate (Table 1, entries 2 and 3).
Both electron-rich and electron-poor aromatic substrates
gave high yields and enantioselectivities (Table 1, entries 6
and 7). However, the 3-chloro substrate 1i gave low enan-
tioselectivity and low yield (Table 1, entry 9), which is sur-
prising because the 3-bromo substrate 1d gave notably
higher yield and enantioselectivity (Table 1, entry 4). Heter-
oaromatics also performed poorly, as demonstrated by thio-
phene 1j, which gave both low diastereoselectivity and
enantioselectivity (Table 1, entry 10). Despite these draw-
backs, the procedure remains useful for most aromatic sub-
strates, because it is both operationally simple with no re-
quirement for rigorous degassing procedures and it uses
only the reducing agent and iPrOH as the solvent.
ATH through DKR with TEAF: Our investigations of the
DKR of a-amido-b-ketoesters under transfer-hydrogenation
conditions began by subjecting b-ketoester 1a to TEAF (5:2
azeotrope) in the presence of [RuCl2ACHTNUTRGNE(NUG p-cymene)]2 and the
(S,S)-N-(p-toluenesulfonyl)-1,2-diphenylethanediamine
((S,S)-TsDPEN) ligand of Noyori and co-workers.[18] As pre-
vious reports had indicated that the major diastereomer was
the syn isomer,[17] we were surprised to obtain the anti-b-hy-
droxy-a-amido ester 2a in good yield and with good diaste-
reoselectivity but low enantioselectivity.[19,20] On the basis
that the synthesis of anti-b-hydroxy-a-amido esters by using
ATH accompanied by DKR was previously unexplored, we
opted to further optimize this reaction. Toward this end,
a series of vicinal amino alcohol ligands was screened in the
reaction in the hope that they would impart a higher enan-
tioselectivity.[21] The optimal conditions used the less steri-
ATH through DKR by using sodium formate in emulsions:
Although the traditional conditions for ATH involve iPrOH
or triethylammonium formate azeotropes as reducing agents
in a variety of organic solvents, the performance of ATH in
water or other environmentally benign reaction media has
recently received more attention.[23] This is in part due to
the considerable interest in greener reactions conditions, but
it is also attributable to the particular effects of using water
as the reaction medium,[24] including increased reaction
rates.[25] However, many of the methodologies focus on the
reduction of aromatic ketones or quinolines and, to our
knowledge, there have been no disclosures of DKR com-
bined with ATH in aqueous media. We were interested in
improving upon our previous conditions for the synthesis of
anti-b-hydroxy-a-amino esters, specifically in decreasing the
reaction time, decreasing the catalyst loading, and widening
the substrate scope, so we thought to examine aqueous reac-
tion media for this transformation.
cally encumbered [RuCl2ACHTUNGTRENNUNG
(benzene)]2 with (1S,2S)-2-(benzy-
lamino)-1,2-diphenylethanol
((S,S)-BnDPAE) as the
ligand.[22]
With these conditions in hand, we then surveyed a variety
of aromatic ketone substrates, 1a–j, to determine the scope
of the reaction. A wide variety of substitution patterns were
Table 1. Exploration of substrate scope.[a]
To this end, we began by testing the reported conditions
for the ATH of aromatic ketones in water on substrate
1a.[25] However, this resulted only in aggregation of the
starting material and ligand in the aqueous media. In order
to circumvent this, we explored the use of a CH2Cl2–water
emulsion.[26] This yielded the desired product in 12 h in good
yield and with good enantioselectivity, but lower diastereo-
selectivity (Table 2, entry 1). Encouraged by the significantly
decreased reaction time (relative to 5–7 days, Table 1), we
then wanted to improve the diastereoselectivity. It has been
shown that the pH value can affect ATH.[27] The emulsion
conditions in entry 1 in Table 2 have pH 7 and the previous
conditions with TEAF, which led to excellent diastereoselec-
tivity, proceeded at pH 4. Therefore, we first tried lowering
the pH value. However, only traces of the desired product
were observed at pH 5 (Table 2, entry 2). Also, an increase
in the amount of Et3N to five equivalents, effectively raising
the pH value to 9, led to no increase in diastereoselectivity
(Table 2, entry 3). Thus, we tried cooling the reaction in
order to improve the diastereoselectivity (Table 2, entry 4).
This lengthened the reaction time to three days but led to
excellent yield and diastereo- and enantioselectivities.[28]
Entry
1/2
Yield[b]
[%]
e.r.[c]
1
2
3
4
5
6
7
8
9
a
b
c
d
e
f
g
h
i
95
83
76
94
81
89
95
93
69
75[d]
97:3
99:1
99:1
96:4
98:2
96:4
97:3
95:5
66:34
76:24
10
j
[a] Reactions performed by heating [RuCl2ACTHNUTRGNEUNG(benzene)]2 (0.10 equiv) and
(S,S)-BnDPAE (0.2 equiv) in 2-propanol (c=0.1m) at 808C for 1 h. After
cooling, the catalyst was then added to the b-ketoester 1 (1 equiv) with
HCOOH/Et3N (5:2) complex (c=0.2m), and the mixture was stirred for
5–7 days at ambient temperature. Only a single diastereomer was visible
in the 1H NMR spectrum of the crude reaction mixture. [b] Yield of iso-
lated product. [c] Determined by chiral HPLC. [d] Isolated as an 80:20
mixture of diastereomers.
&
2
&
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!