Organic Letters
Letter
In accordance with our interest in nickel-mediated cross-
coupling reactions,45,46 we were further interested in capital-
izing on the application of Me3SiMgI in enantiospecific cross-
coupling reactions via C(sp3)−O bond cleavage. We were
optimistic that the methodology developed by Shi,17
Jarvo,11,18−21 and others10,47,48 (Scheme 1a) could be
borrowed for the use of Me3SiMgI as the coupling partner;
however, these literature protocols were not amenable to
Me3SiMgI, and traces of the cross-coupled product were
observed (Scheme 1c), highlighting the limitation of these
literature strategies. The commonly employed silylating agent
PhMe2SiBpin13,49 also failed to furnish the desired product. As
a result, we developed an AgF-assisted nickel catalysis that
results in the complete conversion of benzylic methyl ethers.
The methodology was also extended to enantiospecific
silylation.
We commenced our study with 2-(1-methoxyethyl)-
naphthalene 1a and Me3SiMgI 2a [prepared from TMSI and
Mg (stored at room temperature)]. When benzyl methyl ether
1a was subjected to the previous methods,17,18 a minimal
amount of cross-coupled product 3a was observed (Table 1,
entries 1 and 2). Subsequently, a library of nickel catalysts and
ligands were screened, and the selected entries are listed in
Table 1. Both NiBr2·glyme and NiBr2·diglyme were equally
effective, affording the cross-coupled product 3a in 21% and
22% yields, respectively, with traces (<5%) of eliminated and
reduced byproducts 5a and 5b (entries 3 and 4, respectively).
While NiCl2·diglyme afforded product 3a in 12% yield (entry
6), Ni(cod)2 completely shut down the reaction (discussed
below). The use of a bidentate phosphine ligand (dcype) had
no impact on the yield (entry 7). This led us to screen a list of
additives, and the results are listed in Table 1 (entries 8−12).
A remarkable improvement in the yield to 66% was observed
with AgF (2 equiv) being as an additive (entry 8). The other
additives, including AgBr, LiF, and CsF, affoded product 3a in
yields of only 24%, 36%, and 34%, respectively, with a
significant amount of byproducts 5a and 5b. Pleasingly, the
yield soared to 96% when the temperature was decreased to 0
°C (entry 13), possibly slowing the degradation of nickel to
black. The use of less polar Et2O or toluene had a detrimental
effect on the yield (entries 16 and 17). No reaction was
observed when NiF2 was employed in place of NiBr2·diglyme
(entry 18). Decreasing the amount of Me3SiMgI also
decreased the yield (entry 19). Employing stoichiometric
NiBr2·diglyme afforded 3a in 44% yield and 5b in 52% yield
(entry 20) with complete consumption of 1a. Further tuning of
the reaction condition revealed that the reaction could be
carried out with 0.5 equiv of AgF and 10 mol % NiBr2·
diglyme/PCy3 without a significant compromise in the yield
(entry 14).
With the optimized conditions in hand, we moved further to
expand the scope of the substrates. Methyl ethers 1 with
various α-substituents , including alkyl and aryl groups, were
well tolerated; the cyclohexyl (1b), ethyl (1c), phenyl (1d),
and benzyl (1e) derivatives afforded the corresponding cross-
coupled products in very good yields (Table 2). The 1-
naphthyl derivative afforded product 3f in 50% isolated yield,
and the use of sterically bulkier PhMe2SiMgBr and
Ph2MeSiMgBr also afforded the corresponding products 3ga
and 3gb in good yields.50 A broad range of functional groups
were well tolerated, which provides an opportunity for further
derivatization of the cross-coupled products. Aryl ethers
[C(sp2)−O] (1i, 1j, 1t, and 1ac) were intact under the
optimized condition, and the chemoselective cross-coupled
products (3i, 3j, 3t, and 3ac) were isolated in 70%, 86%, 95%,
and 85% yields, respectively. The sensitive ketal group was
stable and afforded coupled product 3n in 52% yield; partial
decomposition of 3n may be at play. Substrates bearing
fluoride (1k), CF3 (1m), TMS (1l), amine (1ab), and amide
(1u) groups were also compatible and afforded the cross-
coupled products in very good yields. Interestingly, boronic
ester 1s afforded cross-coupled product 3s in 78% yield,
although it can provide trialkyl borane with organometallic
reagents. Ketone derivative 1o also afforded coupled product
3o in 46% yield along with traces of an undesired alcohol
byproduct. Allylic ethers were subsequently investigated;51 as
expected, the reactions were efficient, affording the syntheti-
cally versatile allylsilanes 3p−u in excellent yields and E/Z
ratios.
a
Table 1. Optimization of the Reaction Conditions
additive
(equiv)
1a, 5b
b
b
entry
deviation from above
(%)
3a (%)
c
1
Ni(cod)2/rac-BINAP
NiCl2/dppf, 110 °C
NiBr2·glyme
−
−
−
97, 4
62, 20
71, 5
<5
18
21
22, 21
<5
12
20
66
24
36
34
6
96
91, 88
70
31
66
<5
62
44
d
2
3
e
4
none
−
70, 5
5
6
Ni(cod)2
NiCl2·diglyme
−
−
95, <5
82, 0
7
dcype
−
74, 5
8
9
none
none
none
none
none
0 °C
0 °C
0 °C
AgF (2.0)
AgBr (2.0)
LiF (2.0)
CsF (2.0)
MgBr2
AgF (2.0)
AgF (0.5)
AgF (0.2)
AgF (0.5)
AgF (0.5)
−
28, <5
50, 19
55, 10
57, 8
78, 14
ND, <5
<5, <5
20, 7
79, <5
38, <5
96, <5
36, <5
ND, 52
73, <5
10
11
12
13
14
15
16
17
18
19
20
f
Et2O, 0 °C
toluene, 0 °C
NiF2, 0 °C
1.0 equiv of 2a, 0 °C
1 equiv of NiBr2·PCy3, 0 °C
0 °C
AgF (0.5)
−
AgF (0.5)
Given the importance of heteroarenes in pharmaceuticals
and agrochemicals, we subjected pyridine, benzothiophene,
and thiophene derivatives to cross-coupling reactions and
obtained cross-coupled products 3w−z in good yields. It is
worth noting that the allyl ethers (1q and 1r), pyridines (1w
and 1x), and benzothiophene 1y do not require AgF, and it is
expected that substrates with varying Dewar’s resonance
energy (aromaticity)31−33,52−57 may have a profound impact
on the outcome of the reactivity. For example, the empirical
g
21
25
a
Reaction conditions: 0.20 mmol of 1a, 0.30 mmol of Me3SiMgI·
TMEDA 2a (0.35 M in toluene), 20 mol % NiBr2·diglyme, 20 mol %
b
c
PCy3, additive, and 0.15 M THF. GC yield. With 5 mol % Ni(cod)2,
d
10 mol % rac-BINAP, and toluene for 24 h. With 2 mol % NiCl2/
dppf and 2 mol % dppf for 12 h. After 20 h. isolated yield and
repeated ∼10 times throughout the project. 1a was added after the
e
f
g
addition of all other reagents.
1334
Org. Lett. 2021, 23, 1333−1338