Organometallics
Communication
a
Our group has focused on designing new chiral anionic
ligands to develop the first-row transition metal catalyzed
enantioconvergent C(sp3)−C cross-coupling of racemic alkyl
halides with diverse nucleophiles.6 During this course, we
found that a chiral multidentate anionic N,N,P-ligand could
tune the chiral environment of the copper catalyst to realize
the enantiocontrol over the in situ generated prochiral alkyl
radical species.6 We speculated that such a multidentate
anionic ligand might coordinate strongly with the cobalt
catalyst to realize the enantioconvergent C(sp3)−C cross-
coupling of racemic alkyl halides. The success of this strategy
would open new vistas for the development of cobalt-catalyzed
enantioconvergent cross-coupling. Given the importance of
chiral 1,1-diarylalkanes in drug discovery,12 we herein report
the first application of quinine-derived N,N,P-ligand in the
cobalt-catalyzed enantioconvergent Negishi C(sp3)−C(sp2)
cross-coupling of racemic benzyl chlorides with arylzinc
reagents. Notably, this is also the first time that the more
stable alkyl chlorides rather than bromides,11 are used in the
cobalt-catalyzed enantioconvergent C(sp3)−C(sp2) cross-cou-
pling.
In order to verify our hypothesis, we first investigated the
cross-coupling between racemic 3-(1-chloroethyl)benzonitrile
1a and 4-methoxyarylzinc reagent 2a in the presence of
catalytic amount of CoBr2. The initial attempts showed that
the reaction proceeded smoothly to afford coupling product 3a
without the addition of any ligand in THF even at 0 °C, which
supported the background Negishi cross-coupling (Table 1,
entry 1). We then lowered down the reaction temperature to
−20 °C and found that the background reaction could be
inhibited (Table 1, entry 2). As such, we examined chiral
bisoxazoline ligand L1 which was previously utilized in cobalt-
catalyzed asymmetric cross-coupling11 and found that L1 was
not effective for the reaction of benzyl chlorides 1a. We then
investigated the effect of our developed multidentate electron-
rich N,N,P-ligand L2,6b and found that it not only greatly
promoted the reaction but also afforded the coupling product
3a in 50% yield with an enantiomeric ratio (e.r.) of 77:23
(Table 1, entry 4). Encouraged by this result, we systematically
screened the reaction parameters. We first modified the
structure of ligands with different steric and electronic
properties at different positions of the P-aryl ring (Table 1,
entries 5−11). The screening results indicated that the
electron-withdrawing substituents on 3,5-position of the P-
aryl ring (L3) provided 3a with a lower e.r. than the electron-
donating groups (L4), albeit with a similar yield (Table 1,
entries 5 and 6). While the reaction with phenyl-substituted
ligand L5 provided 3a with a comparable enantioselectivity, it
provided the product in a much lower yield than that of L4
(Table 1, entry 7). The electron-donating substituents on the
para position of the P-aryl ring (L6) gave 3a in 65% yield with
a 79:21 e.r. (Table 1, entry 8). Although the para phenyl-
substituted ligand L7 gave a similar yield with L6, the
enantiomeric ratio is lower (Table 1, entry 9). The ortho-
substituent of the P-aryl ring (L8) has a deleterious effect on
the reaction efficiency, suggesting that the steric environment
on the cobalt is crucial to the cross-coupling reaction (Table 1,
entry 10). The reaction with a naphthyl-substituted ligand L9
gave a poor result as well (Table 1, entry 11). Considering the
reaction efficiency and the enantioselectivity, we chose L6 as
the best ligand for the subsequent reaction optimization. The
results showed that 3a was generated in 90% yield with a 77:23
e.r. at 0 °C, and this retention of e.r. revealed that the chiral
Table 1. Screening of Reaction Conditions
b
c
entry
[Co]
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
CoBr2
L*
solvent
T/°C
yield
e.r.
1
2
3
4
5
6
7
8
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
toluene
DCM
DMF
EtOAc
0
−20
−20
−20
−20
−20
−20
−20
−20
−20
−20
0
rt
0
0
0
0
0
0
75%
0%
L1
L2
L3
L4
L5
L6
L7
L8
L9
L6
L6
L6
L6
L6
L6
L6
L6
15%
50%
41%
35%
19%
65%
63%
trace
21%
90%
45%
95%
93%
99%
76%
87%
98%
50:50
77:23
54:46
82:18
80:20
79:21
70:30
9
10
11
12
13
14
15
16
17
18
19
57:43
77:23
50:50
76:24
75:25
75:25
74:26
69:31
74:26
CoBr2
Co(salen)
Co(PPh3)2Cl2
CoBr2
CoBr2
CoBr2
CoBr2
a
Reaction conditions: ( )-1a (0.05 mmol), 2a (0.15 mmol), [Co]
(10 mol %), and L* (15 mol %), in solvent (0.50 mL) for 72 h under
argon. Yield was based on H NMR analysis of the crude product
b
1
c
using 1,3,5-trimethoxybenzene as an internal standard. The e.r.
values were determined by HPLC analysis.
N,N,P-ligand/cobalt catalysis could suppress the background
reaction and promote the enantioconvergent process (Table 1,
entries 1 and 12). However, the reaction afforded 3a with no
e.r. at room temperature (Table 1, entry 13). Further screening
of different cobalt salts and solvents showed that the reaction
provided the best yield with a slightly decreased e.r. in toluene
(Table 1, entries 14−19). Finally, we identified the optimal
reaction conditions as follows: The reaction of 1a and 2a in a
molar ratio of 1:3 in the presence of 10 mol % of CoBr2 and 15
mol % of L6 in toluene afforded 3a in 99% yield, with 75:25
e.r. at 0 °C.
With the optimal reaction conditions established, we then
examined the scope of benzylic chlorides (Table 2). The
reaction of benzyl chloride without any substituent on the aryl
ring provided 3b in 61% yield with 71.5:28.5 e.r.. The chlorine-
substituted substrates both afforded desired products 3c and
3d in good yields, and the meta-substituted one gave a higher
yield and enantioselectivity than the para-substituted one. The
fluorine-substituted substrates provided 3e and 3f in a similar
yield and e.r. The substrate with a trifluoromethyl group
provided 3g with the best enantioselectivity (80:20 e.r.). To
investigate the functionality tolerance, we studied the cross-
coupling reaction of carbonyl-substituted benzyl chloride, the
reaction delivered product 3h as well, albeit with a lower yield
(40%). In addition, the substrates with an ester substituent
2216
Organometallics 2021, 40, 2215−2219