Angewandte
Communications
Chemie
Table 1: Cu-catalyzed coupling of 4-chloroanisole with ethyl cyanoacetate
under the assistance of different ligands.[a]
implied that the better performance of L2 might result from
relatively poor electron-donating ability of pyridin-2-yl
methyl group, but not from the additional coordination to
copper. Further evaluation revealed that changing the left
part of the ligand L2 to anthracenyl (L5, AMPO) and pyridin-
2-ylmethyl (L6) also gave 3a in reasonably high yields
(entries 10 and 11), while ligands such as proline and 1,10-
phenanthroline previously reported led to no conversion
(entries 12 and 13).
In view of the above encouraging result, we examined
whether the newly developed catalytic system is applicable
for coupling of aryl bromides at low catalytic loading.
Gratifyingly, under the catalysis of 0.5 mol% CuBr and
1 mol% L2, coupling of 4-bromoanisole and ethyl cyanoace-
tate proceeded smoothly at 808C to afford 3a in 70% yield
(entry 14). Changing the copper salt to CuCl led to the
formation of 3a in a better yield (entry 15). Further attempt
indicated that combination of K3PO4 and ethanol gave a best
result (compare entries 16–18).
We next examined the established conditions with a vari-
ety of (hetero)aryl chlorides and bromides. As summarized in
Table 2, a series of aryl halides (Cl, Br) bearing either
electron-donating or electron-withdrawing groups at the para
and meta-position worked well, producing the corresponding
a-aryl acetonitriles 3b–3j in 70–89% yields. In case of ethyl 4-
bromobenzoate as a substrate, the decarboxylation step
should be carried out under acidic conditions to preserve
the ester group in the structure (3g). Coupling with 2-
chloroanisole gave 3k in a poor yield, although a good result
was observed in case of 2-bromoanisole as a coupling partner.
This result indicated that the steric hindrance significantly
influenced the coupling reaction with aryl chlorides. Further
evidence was observed from the difference in coupling with 2-
chloronathphlene (3p) and 1-chloronathphlene (3o). Addi-
tionally, three trisubstituted and two heterocycle-substituted
aryl halides were applicable, delivering a-aryl acetonitriles
3l–3n and 3q in good yields.
Entry
X
[Cu]
L
Base
Solvent
Yield [%][b]
1
2
3
4
5
6
7
8
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Br
Br
Br
Br
Br
CuI
CuI
CuI
L1
L2
L2
L2
L2
L2
L2
L3
L4
L5
L6
L7
L8
L2
L2
L2
L2
L2
K3PO4
K3PO4
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
EtOH
7
17
50
80
61
28
0
30
58
75
62
0
t-BuONa
t-BuONa
t-BuONa
t-BuONa
NaOH
t-BuONa
t-BuONa
t-BuONa
t-BuONa
t-BuONa
t-BuONa
t-BuONa
t-BuONa
K3PO4
CuBr
CuCl
CuBr
CuBr
CuBr
CuBr
CuBr
CuBr
CuBr
CuBr
CuBr
CuCl
CuCl
CuCl
CuCl
EtOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
EtOH
9
10
11
12
13
14
15
16
17
18
0
70
84
71
87
56
K3PO4
K3PO4
dioxane
Heteroaryl halides are recognized as difficult substrates
under previous Pd- or Cu-catalyzed coupling conditions.[11–13]
Fortunately, our new catalytic system addressed this long-
standing problem. Under the standard conditions, coupling
reaction of a range of heteroaryl halides (Cl, Br) proceeded
smoothly, leading to the formation of a-heteroaryl acetoni-
triles 3r–3ad in good to excellent yields. Noteworthy was that
the common heterocycles such as pyridine (3r, 3s), quinoline
(3t, 3u), isoquinoline (3v), benzothiophene (3w–3y), indole
(3z), benzothizole (3aa), carbazole (3ab), dibenzothiophene
(3ac) and pyrrole (3ad) were conveniently introduced by
employing the corresponding heteroaryl halides. When 2-
bromothiophene was used, coupling reaction took place to
afford ester 4 in 53% yield. After its decarboxylation, we
failed to isolate 2-(thiophen-2-yl)acetonitrile, mainly because
of its low boiling point.
[a] Reaction conditions: for entries 1–13: 1 (1.0 mmol), 2 (1.5 mmol),
[Cu] (0.05 mmol), ligand (0.1 mmol), base (2.5 mmol), solvent (1.5 mL),
1058C, 24 h; then water, 1058C; For entries 14–18: 1 (4.0 mmol), 2
(8.0 mmol), [Cu] (0.5 mol%), ligand (1.0 mol%), base (12.0 mmol),
solvent (4.0 mL), 808C, 24 h, then H2O (1.0 mL), 808C, 12 h. [b] The
1
yield was determined by H NMR analysis of crude products.
K3PO4 in 2-propanol at 1058C. However, low conversion was
observed and the desired a-aryl nitrile 3a was obtained in 7%
yield after decarboxylation. Changing the ligand to N-(2-
methylnaphthalen-1-yl)- N’-(pyridin-2-ylmethyl)oxalamide
(MNPMO, L2) gave an improved result (entry 2). Using this
ligand we further screened several bases, copper salts and
solvents, and found that t-BuONa was the best base (compare
entries 2, 3 and 7), CuBr was the best catalyst precursor
(compare entries 3–5), and 2-propanol was the better solvent
than ethanol. Since L2 showed better activity than L1, we
wondered if its pyridine moiety can provide the additional
coordination to copper, and thereby facilitating the catalytic
cycle. Consequently, L3 and L4, two simple analogues of L2,
were also examined (entries 8 and 9). Their activity trend
Furthermore, we found that the present method could be
utilized for direct functionalization of known drugs bearing an
aryl chloride unit. For example, coupling reaction with
chlorpheniramine and clomipramine afforded 3ae and 3af,
respectively. Additionally, an ester-embodied aryl chloride
was applicable to afford 3ag in 68% yield, while 1-(4-
Angew. Chem. Int. Ed. 2021, 60, 7082 –7086
ꢀ 2020 Wiley-VCH GmbH
7083