Organic Letters
Letter
as a gas or through the sublimation of dry ice, formic acid is
readily available as a solution. Thus we attempted to employ
formic acid as an alternative and user-friendly hydroxymethyl
source for the hydroxymethylation of alkynes. More
importantly, as an excellent protic output buffer that gently
releases protons, the formic acid meanwhile may be a proton
source to self-regulate the chemoselectivity of alkyne
hydroxymethylation to produce distinct alcohols (Figure 1d),
thus avoiding the addition of an extra proton.
(entry 8). The effect of temperature on the divergence of
products brought great enlightenment to the subsequent
regulation of the chemoselectivity of this reaction.
With the optimized conditions in hand, we then evaluated
the scope of the direct hydroxymethylation of alkynes with
formic acid. As shown in Scheme 1, a variety of internal
Scheme 1. Direct Hydroxymethylation of Various Alkynes
with Formic Acid
We initiated our investigation for the reactivity of direct
alkyne hydroxymethylation by treating 1,2-diphenylacetylene
(1a) and formic acid (1.5 equiv to 1a) with triethoxysilane,
using Cu(OAc)2 coupled to a range of rigid biphosphine
ligands as catalysts (Table 1, entries 1−4). Although the
Table 1. Optimization of the Direct Hydroxymethylation of
a
Alkynes with Formic Acid
b
entry
ligand
temp (°C)
yield (%)
21
56
c
1
2
3
4
5
L1
L2
L3
L4
L5
L4
L4
L4
55
55
55
55
55
55
90
rt
aromatic alkynes with electron-donating and mildly electron-
withdrawing substituents all reacted smoothly to give the
corresponding trisubstituted allylic alcohols 2b−2f with
complete E-stereoselectivity. There was a slight decrease in
the reactivity of meta- or ortho-substituent diphenyl alkynes,
presumably due to the steric hindrance (2g, 2h). The
bithiophenyl alkyne was also reactive, delivering the (Z)-2,3-
di(thiophen-2-yl)prop-2-en-1-ol (2i) in 71% yield. Remark-
ably, a high regioselectivity was attained in the hydroxyme-
thylation of cyclopropylphenylethyne, affording the major
product 2j. In addition, in the investigation of scope with
respect to simple terminal alkynes, anti-Markovnikov regiose-
lectivity was observed, giving rise to linear allylic alcohols 2k
and 2l in mild yields. In addition, a satisfactory yield (65%) of
2a was obtained when the direct hydroxymethylation of 1a was
performed on a 6 mmol scale. For more information on the
optimization and large-scale experiment, see the Supporting
Considering that an elevated temperature favors the
protonation of the vinylcopper species to produce an alkene,
as shown in the optimized results of the direct alkyne
hydroxymethylation, we then explored the reductive hydrox-
ymethylation of 1a by increasing the temperature to 90 °C as
well as the amount of formic acid to 3 equiv. It was
disappointing that only 1,2-diphenylethene, 1,2-diphenyl-
ethane, and a trace of 2a were obtained. But when the
terminal alkyne of 4-ethynyl-1,1′-biphenyl (3a) was introduced
instead of 1a, fortunately, a 78% yield of 2-([1,1′-biphenyl]-4-
yl)propan-1-ol (4a) was generated coupled to a small amount
of 4-vinyl-1,1′-biphenyl and 4-ethanyl-1,1′-biphenyl. We
attribute the poor reactivity of 1a to the impediment of the
hydroxymethylation of its semireduction product, that is, 1,2-
d
91 (87)
51
43
38
15
e
6
7
8
a
Reactions were conducted in a 10 mL Schlenk tube. Conditions: 1a
(0.0712 g, 0.4 mmol), Cu(OAc)2 (0.0072 g, 0.040 mmol), L (0.044
mmol), CsOAc (0.0039 g, 0.02 mmol), HCOOH (0.6 mmol, 23 μL),
(EtO)3SiH (4.5 mL, 2.4 mmol), cyclohexane (2 mL), 55 °C, 24 h.
b
1
Determined by the H NMR technique using 1,1,2,2-tetrachloro-
c
d
ethane as an internal standard. Not detected. Isolated yield.
e
(EtO)3SiH (1.6 mmol).
Xantphos (L1)- and DTBM-MeO-BIPHEP (L2)-based copper
catalysts were positive toward the generation of (E)-2,3-
diphenylprop-2-en-1-ol (2a), Cu(OAc)2 combined with
DTBM-SEGPHOS (L4) was found to be the most efficient,
affording 2a as a complete E-type geometric isomer, as
1
determined by H NMR analysis (entry 4). The investigation
of the alkyl biphosphine ligand showed that (S,S)-Ph-PBE
(L5) could have a moderate effect on the generation of 2a
(entry 5). A sharp decrease in the yield of 2a was observed
when the amount of triethoxysilane was reduced (entry 6),
probably because the comparable hydrosilylation of formic
acid consumed a certain amount of hydrosilane. Further
optimization showed that an elevated temperature went against
the generation of allylic alcohols but was beneficial to the
protonation of the vinylcopper species to produce 1,2-
diphenylethene (entry 7). When the temperature was lowered
to room conditions, a poor conversion of 1a was obtained
4998
Org. Lett. 2021, 23, 4997−5001