Organometallics
Communication
best of our knowledge, has not been reported for base-
transition-metal catalysis using Co, Fe, or Mn.19
for the reaction (Table 1, entries 5 and 7−11). It is noteworthy
that KOtBu, NaOtBu, and KHMDS led to comparable yields
(Table 1, entries 5, 7, and 8). Toluene and a temperature of
110 °C proved to be more suitable (Table 1, entries 5 and 12−
16). It is worth mentioning that, at a higher temperature of 120
°C, a diminished yield resulted with more unidentified side
products being observed (Table 1, entry 13). Pleasingly, a very
good 75% isolated yield resulted in the 1 mmol scale reaction
(Table 1, entry 5c). A mercury test indicated a homogeneous
reaction (Table 1, entry 5d).
Scheme 1. Switchable β-Alkylation of Secondary Alcohols
with Primary Alcohols Catalyzed by a Cobalt Complex
With the optimized reaction conditions on hand, we then
investigated the reaction scope in terms of the substrates. First,
the scope of the primary alcohols was explored in the
alkylation of 1-phenylethanol. To our delight, para-substituted
benzyl alcohols with electron-donating or -withdrawing groups
reacted smoothly (Table 2, 3b−h). Notably, this method was
also applicable to some challenging substrates such as those
bearing −Br or −CF3 functional groups (Table 2, 3g,h),
furnishing moderate to good yields. Meta-substituted sub-
strates delivered the corresponding alcohol products in
moderate to very good yields (Table 2, 3i−l). It is noteworthy
that the sterically hindered ortho-substituted 2-methylbenzyl
alcohol afforded a very good 85% yield (Table 2, 3m).
However, 2-fluorobenzyl alcohol only resulted in a 19% yield,
probably due to the coordination of the −F functionality to the
catalyst, diminishing its reactivity. Aliphatic primary alcohols
were amenable to this method as well (Table 2, 3r−t). Next, a
series of secondary alcohols were examined to couple benzyl
alcohol. Pleasingly, for aromatic secondary alcohols, −OMe,
−F, −CF3, and naphthalene functional groups were well
tolerated in this method (Table 2, 3u−z). Challenging
aliphatic secondary alcohols were alkylated with benzyl
alcohols leading to satisfactory yields, albeit under relatively
harsher conditions (Table 2, 3aa−ac). This is likely due to the
difficulty in the dehydrogenation of aliphatic secondary
alcohols.15 Notably, aliphatic secondary alcohols also coupled
with aliphatic primary alcohols by this method. 2-Hexanol
reacted with 1-hexanol, affording 5-dodecanol in a good 66%
yield. (Table 2, 3ad).
Next, we carried out a preliminary mechanistic study to
understand the reaction and elucidate the selectivity switching.
In our prior work, ketone is the major product via the ADC
process.18 It is rational to surmise that the selective alcohol
formation may originate from further reduction of the in situ
formed ketone. To examine this hypothesis, we performed the
transfer hydrogenation of 3-phenylpropiophenone (4a) with 3
equiv of 1a as the hydrogen source. We found that 80 mol % of
KOtBu alone can efficiently reduce 4a to 3a to furnish a 90%
yield, supporting a base-mediated Meerwein−Ponndorf−
Verley (MPV) pathway.20,21 However, a drastically lower
KOtBu loading of 2.5 mol % also leads to a 71% yield of 3a.18
An interesting question arises: why is excess base still required in
the alcohol synthesis by the β-alkylation of secondary alcohols with
primary alcohols?We then performed the same reaction in the
presence of 5 mol % of A and 90 mol % of KOtBu.22
Interestingly, a lower 78% yield of 3a was observed in
comparison to that with 80 mol % KOtBu alone, suggesting
that the Co species may have a negative effect on the
hydrogenation of 4a in the presence of excess base. As A has
been recognized to be an efficient precatalyst for the
dehydrogenation of secondary alcohols from our prior
work,15 the alcohol products could be dehydrogenated back
to the ketones. In addition, in the β-alkylation of secondary
At the outset, 1-phenylethanol (1a) and benzyl alcohol (2a)
were selected as the model substrates. Interestingly, it was
revealed that an excess amount of base (110 mol %) was
required for the selective formation of 1,3-diphenylpropan-1-ol
(3a) (Table 1, entries 1−4), which was in stark contrast to the
a b
,
Table 1. Optimization of the Reaction Conditions
base
A
entry
base
(mol %)
(mol %)
solvent
toluene
yield (%)
1
2
3
4
5
6
7
8
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
NaOtBu
KHMDS
KOH
K2CO3
Cs2CO3
KOtBu
KOtBu
KOtBu
KOtBu
KOtBu
0
30
80
2.5
2.5
2.5
2.5
0.7
0
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
31
53
73
110
110
110
110
110
110
110
110
110
110
110
110
110
c
d
80, 75, 79
8
80
77
26
0
9
10
11
12
13
14
15
16
0
e
67
f
71
1,4-dioxane 12
THF
benzene
5
68
a
Reaction conditions unless specified otherwise: 1-phenylethan-1-ol
(0.375 mmol), benzyl alcohol (0.25 mmol), A, base, and solvent (1.5
mL) were heated in a closed 15 mL reaction tube for 24 h under N2.
b
NMR yield using 1,3,5-trimethoxybenzene as internal standard.
c
d
Isolated yield on a 1 mmol scale. Mercury (125 mg) was added to
e
f
the reaction mixture. 100 °C. 120 °C.
ketone case, where only a catalytic amount of base (7.5 mol %)
was used.18 Both A and base are essential for this trans-
formation (Table 1, entries 1 and 6). Significantly, only 0.7
mol % of A was sufficient to mediate the alcohol-forming
reaction (Table 1, entry 5). The ketone side product 3-
phenylpropiophenone (4a) was detected with only 8% yield.
To our surprise, a slightly lower yield (73%) resulted using 2.5
mol % of A, concurrent with the generation of unknown side
products (Table 1, entry 4). Strong bases are more appropriate
1208
Organometallics 2021, 40, 1207−1212