Organic Letters
Letter
a
enantioselectivities (entries 6−9). Using ligand L2 instead of
L1 slightly improved the enantioselectivity of the reaction
(entry 10). When metal precursor Pd2(dba)3 or Pd-
(CH3CN)2Cl2 was used under identical reaction conditions,
significantly inferior effects were discovered (entries 11 and
12). After systematic investigation of commercially available
chiral phosphine ligands, including L3−L6 (for details, see
Table S2), the er value was increased to 92:8 with L3 (entry
13). Finally, when the reaction was performed at 10 °C in a
DCM/TFE mixed solvent (1/1) for 24 h, 2a was afforded with
95:5 er and 99% isolated yield (entry 17).
Table 1. Optimization of Reaction Conditions
The scale-up asymmetric hydrogenation of 1a under
optimized reaction conditions was then attempted. For a 2
mmol reaction scale, the product was obtained with 90% yield
and 90:10 er (entry 18). When the scale was further increased
to 3 mmol, although the substrate was quantitatively converted
to 2a, the er value dropped to 83:17.
With the optimized Pd(OAc)2/(S)-Difluorphos/H+ catalyst
system in hand, the scope and compatibility for the
enantioselective hydrogenation of α,α-difluorobutenoic esters
were explored with 30 bar of H2 in a TFE/DCE mixed solvent
at 10 °C for 24 h (Scheme 2). Various electron-drawing
substituents on the phenyl ring of the substrates, such as
fluoro, chloro, bromo, and trifluoromethyl groups, were well
tolerated and afforded 2b−2d, 2i, 2l, and 2m in high yields
(90−99%) and enantioselectivities (91:9−97:3 er). Difluori-
nated butenoic esters 2e−2h containing electron-rich sub-
stituents, including methyl and methoxy groups, were hydro-
genated with similar activity with slightly lower er values. The
absolute configuration of 2f was determined to be (S) by
single-crystal X-ray diffraction analysis. No significant effect of
the position of the substituent on the phenyl ring was observed
upon such a transformation. Asymmetric hydrogenation of 1n
with a 1-naphthyl group proceeded with high activity and
moderate selectivity. When the corresponding N,N-diethyl
amide of 1a was subjected to the standard reaction conditions,
no conversion of the starting material was observed. α,α-
Difluoro-β-thienyl and -phenylethyl butenoic esters were also
prepared and applied in such asymmetric hydrogenation under
standard reaction conditions, giving the desired products 2p
and 2q in high yields with 53:47 and 84:16 er, respectively.
Trisubstituted substrate 1r (E/Z = 5/1) exhibited no
conversion with this protocol, regardless of whether the
reaction was carried out at 10 or 60 °C.
To understand this palladium-catalyzed asymmetric hydro-
genation of α,α-difluoro-β-arylbutenoates, control experiments
were conducted under standard reaction conditions (Scheme
3). Using the nonfluorinated 1a structure analogues (3 and 4)
as the substrates under otherwise identical reaction conditions
led to the corresponding hydrogenated products in high yield
but almost without enantioselectivity (Scheme 3a). Further-
more, substituting the ethoxycarbonyl group in 1a with a
fluoro atom or phenylacetyl group resulted in a drastic
decrease in enantioselectivity (Scheme 3b,c). These results
indicated the unique role of the difluoromethylene moiety and
implied the complicated structure−enantioselectivity relation-
ship, which was difficult to elucidate at this stage.
conversion
b
c
entry
TM
L
solvent
DCE
DCE
DCE
DCE
DCE
THF
toluene
iPrOH
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE/DCE
TFE/DCE
(%)
er (%)
1
2
3
4
Pd(OAc)2
L1
L1
−
0
80
>99
0
>99
17
14
>99
>99
>99
>99
0
>99
99
60
−
53:47
0
−
88:12
−
−
53:47
89:11
91:9
52:48
−
92:8
0
65:35
58:42
95:5
90:10
Rh(COD)BF4
Ir-(S)-PHOX
Ni(OAc)2·4H2O
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
L1
L1
L1
L1
L1
L1
L2
L2
L2
L3
L4
L5
L6
L3
L3
d
5
d
6
d
7
d
8
Pd(OAc)2
d
9
Pd(OAc)2
Pd(OAc)2
Pd2(dba)3
Pd(CH3CN)2Cl2
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
Pd(OAc)2
d
10
d
11
d
12
d
13
d
14
d
15
d
16
>99
>99 (99)
>99 (90)
d
,
,
e
f
17
18
d
Pd(OAc)2
a
Reaction conditions: 1a (0.2 mmol), transition metal (5 mol %),
ligand (6 mol %), H2 (30 bar), solvent (1 mL), 25 °C, 16 h. DCE =
1,2-dichloroethane; TFE = trifluoroethanol; THF = tetrahydrofuran.
b
Determined by GC analysis using isooctane as the internal standard.
c
Determined by HPLC analysis on a chiral stationary phase using an
d
e
OJ-3 column. PTSA·H2O (25 mol %) was added. Reaction carried
out at 10 °C in DCE/TFE mixed solvent (1/1 mL) for 24 h; isolated
f
yield given in parentheses. 1a (452 mg, 2 mmol), (S)-Difluorphos
(82 mg, 6 mol %), PTSA·H2O (95 mg, 25 mol %), TFE/DCE (5/5
mL), 10 °C, 24 h.
L1, indicating the unbeneficial influence of the difluoro-
methylene moiety on such alkene enantioselective hydro-
genation (entries 1−4). Protonic acid has been reported to
promote the generation of active palladium−hydride species in
palladium-catalyzed hydrogenation and hydrofunctionaliza-
tion.13h Thus, a catalytic amount of p-toluenesulfonic acid
monohydrate (PTSA·H2O) was added to the reaction system
consisting of Pd(OAc)2 and L1, and to our delight, full
conversion and an 88:12 enantiomeric ratio (er) of 2a were
observed (entry 5; for details of the effects of different acids,
see Table S1). Variation of the solvent indicated that
trifluoroethanol (TFE) slightly improved the er value of the
product while other solvents gave lower yields and/or
Inspired by these results and recent elegant work on non-
precious metal-catalyzed asymmetric reduction of function-
alized alkenes,16 we explored the application of a nickel catalyst
in such an enantioselective hydrogenation. After the
investigation of chiral diphosphine ligands and other reaction
parameters, 1a was hydrogenated in quantitative yield and 8:92
B
Org. Lett. XXXX, XXX, XXX−XXX