Organic Letters
Letter
Aliphatic primary alcohols 6h and 6i and aliphatic secondary
alcohol 5n were also amenable to this method.
which was consistent with the reported literature.4 The result
suggests a cross-coupling reaction occurred in a selective
manner with substituted cyclohexanol. Gratifyingly, the scope
of the methodology was expanded to a variety of other higher
analogues of cyclic alcohols such as cycloheptanol and
cyclooctanol. Finally, the highly challenging acyclic aliphatic
secondary alcohol was found to be amenable to this method,
giving 9cf−df, 9lf, and 9cg in good yields. Synthetic
application of the catalytic protocol was demonstrated by
reacting cholesterol with 5c and derivative of ( ) tocopherol
(6j) with 5a. These reactions can also run on gram scale,
highlighting the practical utility.
We next explored the catalytic cross-coupling of two
different secondary alcohols to synthesize β-disubstituted
ketones. The major challenge is to overcome the unwanted
self-coupled aldol product, while selectivity is also another
issue.4 Investigation started with the reaction of 5c and
cyclohexanol in the presence of Ru-1 (2 mol %) and KOtBu
(20 mol %) in toluene at 150 °C (see Table S2). To our
delight, we observed the desired product 9ca in 83% yield after
12 h. Notably, no self-coupled product or fully hydrogenated
product 9ca′ was found. Employing Ru-2 under similar
conditions led to a better yield of 9ca, 89% (Table S2, entry
2). Thus, the scope of the reaction was pursued with Ru-2. As
displayed in Scheme 5, a wide variety of secondary alcohols
To validate the homogeneous catalytic system for this C-
alkylation, the mercury drop test was conducted, showing no
inhibition of the reaction or reduction of the product yield
(Scheme S5). The liberation of H2 in ADC reaction was
probed by the Pd/C catalyzed hydrogenation of styrene
(Scheme S6). The impact of the NH functionality was
examined by using complex Ru-2Me as the precatalyst, leading
to lower yields of 7ca and 9ca and thereby indicating the
importance of the NH functionality through a metal−ligand
tion). To gain more insight into the active ruthenium species
[Ru(III) or Ru(II)] in the catalytic cycle, a titration
experiment was performed by adding a base into precatalyst
Ru-2 and monitoring the EPR spectra. Upon loading the base,
the paramagnetic signal of ruthenium(III) slowly disappeared
and formed EPR-silent diamagnetic Ru(II) species by using 4
equiv of KOtBu (Figure S3). It infers that 4 mol % of KOtBu
(compared to Ru-2) is necessary for activation of precatalyst.
As the resulting active species (I) is diamagnetic, the NMR
analysis was executed in CD3OD (see the Supporting
Information). For further clarity, the experiment has been
performed in the presence of PPh3. A singlet at 28.5 ppm was
observed in 31P{1H} NMR, indicating the phosphine ligand is
coordinated to the ruthenium(II) center. Besides, species I was
also confirmed by mass spectrometry analysis (see the
we can reasonably conclude that an in situ generated
ruthenium(II) species was the active species in this catalytic
cycle. The resulting intermediate I can react with alcohol 5c to
provide reactive ruthenium hydride species II, which was
supported by mass spectrometry analysis showed in Figure S7.
Despite several attempts, we failed to isolate metal hydride
species under the reaction condition. Next, in the presence of
KOtBu, the base mediated cross-aldol condensation of 5c′ with
6a′ (or with 5a′ and 8a′) afforded α,β-unsaturated ketone 7ca-
I (or 9aa-I), suggesting the formation of a ketone or aldehyde
in situ as the intermediates. The reaction of 7ca-I with 6a
under optimized condition afforded desired ketone 7ca in
good yield. Likewise, β-branched ketone 9aa was obtained by
Similar results were observed with molecular H2 in its place of
alcohol (See ESI). These data suggest that in both cases, α,β-
unsaturated ketone (7ca-I or 9aa-I) is the reaction
intermediate and the reaction likely proceeds via a ruthenium
hydride species, either derived from dehydrogenation of
alcohol or molecular H2. Furthermore, deuterium labeling
experiment was performed in both cross-coupling reactions to
validate the involvement of hydrogen-borrowing methodology.
Under the standard conditions, the reaction of 5c-D with 6a or
with 8a yielded expected H/D scrambled product 7ca-D or
9ca-D in 85% and 78% yields, respectively (Scheme S15). The
Scheme 5. Ruthenium-Catalyzed Selective Cross-Coupling
a
of Secondary Alcohols
a
General conditions: 1-arylethanol (1 mmol), secondary alcohol (1
mmol), catalyst Ru-2 (2 mol %), KOtBu (20 mol %), and toluene (2
mL) were heated at 150 °C under air for 12 h. Isolated yield. Yield
in gram-scale reaction after 24 h. Aliphatic secondary alcohol (6
b
c
d
mmol) was used.
was examined to synthesize β-disubstituted ketones. An array
of diverse functionalities such as 4-methyl, 4-tert-butyl, 4-
methoxy, 4-dimethylamine, and 4-phenyl on the 1-phenyl-
ethanol (5b−f) were efficiently reacted with 8a to obtain β-
disubstituted ketones 9ba−fa in good to excellent yields (78−
90%). Likewise, meta-substituted 5g also produced the
corresponding ketone in high yield. Remarkably, sterically
hindered ortho-substituted substrate was well tolerated as
witnessed from the 82% yield of 2-cyclohexyl-1-(o-tolyl)-
ethanone (9ha). However, chloro-substituted aromatic secon-
dary alcohol showed diminished activity. It was interesting to
note that substitution on a cyclohexyl ring such as 4-tert-
butylcyclohexanol 8b (or 4-methyl-butylcyclohexanol 8c) gave
a mixture of diastereomers with a diastereomeric ratio of 81:19
as determined from the 1H NMR of the reaction mixture. The
major isomer is 1,4-cis conformation of the cyclohexyl ring,
871
Org. Lett. 2021, 23, 869−875