Organic Letters
Letter
and 4). The hydrogenation reaction also can be performed in
methanol, giving a full conversion and 90% ee (Table 1, entry
5). While in isopropanol the conversion of reaction is low (see
Table 1, entry 6, 48%, 92% ee). Using NaOtBu, KOH, and
NaOH as the base gave comparable enantioselectivity, albeit
the conversion of reaction was reduced slightly (Table 1,
entries 7−9), but no hydrogenation occurred by employing
weak base K2CO3 (Table 1, entry 10). Screening of the
catalysts revealed that the catalyst (S)-1a having a 3-methyl
group on the pyridine ring gave the best results (see Table 1,
entries 3 and 11−13). The catalyst loading can be reduced to
0.02 mol % (S/C = 5000), affording a complete reaction within
3 h and high enantioselectivity (>99% ee; see Table 1, entry
14). When the catalyst loading was reduced to 0.01 mol % (S/
C = 10 000), the hydrogenation requires higher hydrogen
pressure and longer reaction time (15 atm, 12 h, entry 15).
Under the optimal reaction conditions, a wide range of δ-
aryl-δ-ketoacids 2a−2k were hydrogenated, and the results are
shown in Table 2. The electron property and the position of
the substituent at the phenyl ring of substrates 2a−2k have no
significant effect on both yield and enantioselectivity of the
reaction. The hydrogenations completed within 2 h and the
enantioselectivities were between 96% and >99.9% ee. In
addition, we also studied the hydrogenation of γ-aryl-γ-
ketoacids 2l−2p under the same reaction conditions. The
reactions were completed within 2 h and provided the
hydrogenation products 4l−4p in 90%−98% yield with 98%
ee. Note that the hydrogenation product (R)-4p is a key
intermediate for the synthesis of sertraline,1a which is a chiral
drug widely used for the treatment of depression and anxiety-
related disorders. We also tested γ/δ-alkyl substituted γ/δ-
ketoacids such as 2q and 2r with a methyl group, but low to
moderate enantioselectivities were observed (Table 2, entries
17 and 18). In addition, the asymmetric hydrogenation of γ/δ-
aryl substituted γ/δ-keto-amides, such as 5a and 5b, performed
Scheme 1. Asymmetric Hydrogenation of δ-Ketoesters and
γ- or δ-Ketoacids with Ir-SpiroPAP
reaction conditions, the hydrogenation of acid group of
substrates can be avoided. The γ- and δ-lactones were obtained
by acidifying γ- and δ-hydroxylacids (Scheme 1b).
We initially performed the hydrogenation of δ-ketoacid 2a
under the following conditions: 1 equiv KOtBu (ratio of base
to substrate, B/S = 1) and 0.1 mol % catalyst Ir-SpiroPAP
((S)-1a) in ethanol at room temperature and 10 atm of H2.
Only a trace amount of hydrogenation product 3a was
obtained after 2 h (Table 1, entry 1). An investigation on the
effect of the amount of base to the reaction showed that a
slight excess of base (B/S = 1.1) was required for complete
conversion of 2a (Table 1, entry 2). The enantiomeric excess
(ee) of the hydrogenation product 3a was analyzed after
converting to lactone 4a by treatment with trifluoroacetic acid
(94% ee). Further increasing the amount of KOtBu to B/S 1.2
and 1.4, the ee value of the product 4a was increased to >99%
ee with no effect on the rate of the reaction (Table 1, entries 3
a
Table 1. Asymmetric Hydrogenation of 5-Oxo-5-phenylpentanoic acid (2a). Optimizing the Reaction Conditions
b
c
d
entry
(S)-1
base
B/S
solvent
time (h)
conversion (%)
enantiomeric excess, ee (%)
e
1
2
3
4
5
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1c
1d
1a
1a
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
NaOtBu
KOH
NaOH
K2CO3
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
1.0
1.1
1.2
1.4
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
EtOH
EtOH
EtOH
EtOH
MeOH
iPrOH
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
2
2
2
2
2
trace
100
100
100
100
48
ND
94
>99
>99
90
6
2
92
7
8
9
3
3
3
3
2
3
2
3
95
97
87
0
100
100
100
100
100
99
99
99
ND
98
99
97
>99
>99
e
10
11
12
13
14
15
f
g
12
a
Reaction conditions, unless otherwise noted: 1.0 mmol scale, (S)-1 (0.1 mol %, S/C = 1000), solvent (3.0 mL), room temperature (25−30 °C),
b
c
d
1
10 atm H2. B/S = base/substrate. Determined by H NMR spectroscopy. Determined by HPLC analysis on a Chiralcel AD-H column of the
corresponding lactone (R)-4a, the absolute configuration was determined by comparison the optical rotation with literature values. ND = not
determined. 0.02 mol % of (S)-1a. 0.01 mol % of (S)-1a, 15 atm H2.
e
f
g
B
Org. Lett. XXXX, XXX, XXX−XXX