ACS Catalysis
Letter
a
except for P9 (61% yield), which has a strongly electron
withdrawing nitro group, and P11 (70% yield), which has
sterically bulky 2,4-dimethylphenyl groups. Notably, a vinyl
substituent, which is reactive in metal-catalyzed hydro-
genations, survived this borane-catalyzed cascade reduction
(P10). 2-Aryl-3-methylpyridines were also suitable substrates,
generating predominantly cis piperidines P13−P18 in 75−98%
yields. The retention of the electron-rich heteroaromatic
substituents in P19−P21 demonstrates the electronic bias of
this reduction method. The method was also compatible with
pyridines bearing an ester (P22), a phenylamino group (P23),
a methoxy group (P24), a benzyloxy group (P25), a hydroxy
group (P26), or a chlorine atom (P27) directly attached to the
pyridine ring. These results are particularly noteworthy
because these functional groups are very labile under the
influence of a pyridine ring, so much so that the available
methods for pyridine reduction, including methods involving
transition-metal catalysis,3,4 have rarely been studied with these
kinds of substrates. Moreover, a 2-alkyl-3-arylpyridine and a
2,3-dialkylpyridine were also reactive, giving products P28 and
P29, respectively, in good yields with high cis selectivity.
We also extended the reaction to more types of pyridines,
which gave some interesting reduction products (Table 3).
When the substrate had a 2-furyl or 2-thienyl ring at the meta
position, a reductive ring-opening reaction occurred,11
generating a piperidine functionalized with an alcohol (P30)
or a thiol (P31) group, respectively. In addition, 3-fluoro-2-
phenylpyridine showed unique reactivity, affording cyclic imine
P32 in 40% yield. Moreover, when 2,4-disubstituted pyridines
were tested, they gave products either with a retained carbon−
carbon double bond (P33 and P34)8a or with a fully reduced
N-heterocycle (P35), depending on the electronic properties
of the substituent at the 4-position. In addition, the current
method was applicable to 2,6-disubstituted and 2-substituted
pyridines (P36−P39), as well as a number of other N-
To highlight the practical utility of this method, we
performed gram-scale reactions of S1 and S26, which afforded
P1 and P26 in 91% and 71% yields, respectively, with high cis
selectivity (Scheme 2a). Protection of P26 with a tert-
butyloxycarbonyl group gave 2 (94% yield), which is an
intermediate in previously reported syntheses of the neuro-
kinin receptor antagonists L-733060 and CP-99994, as well as
3-hydroxypipecolic acid, a precursor for synthesis of the
antibiotic tetrazomine.12 In addition, 2 could be dehydrated
under Mitsunobu conditions to give a 75% yield of 3, which
reacted with an alkyllithium reagent to give trans-2-aryl-3-
alkylpiperidine 4.13 Alternatively, hydroboration/oxidation of 3
provided trans-3-hydroxy-2-phenylpiperidine 5.14
Table 3. Further Exploration of the Scope
a
Unless otherwise specified, all reactions were performed with 5 mol
% of B(C6F5)3, 0.20 mmol of a pyridine, and 0.24 mmol of HBpin in 2
mL of toluene under 3 MPa of H2 at 80 °C in a 30 mL autoclave;
isolated yields are reported. 10 mol % of B(C6F5)3, 4 MPa of H2, 2
mL of PhCF3, 100 °C. 2.0 equiv of HBpin. 3.0 equiv of Et2SiH2.
b
c
d
yield of P1 after hydrogenation. The formation of M3 in the
absence of H2 might have resulted from a transfer hydro-
genation reaction of M1, with HBpin acting as the hydride
donor and residual H2O as the proton donor.8a In another
experiment, we monitored parallel reactions of pyridine S1
under the standard conditions, stopping the reactions at
various times to evaluate the amounts of substrate,
intermediates, and products (Scheme 3b). In these reactions,
tetrahydropyridine M3 was the only intermediate we observed;
S1 was first converted to M3 (only 2% of the original S1 was
left after 4 h), and then M3 was converted to P1-Bpin at a
slower rate, probably because of the steric bulk of the
tetrasubstituted olefin. Furthermore, running hydroboration
and hydrogenation reactions at the same time inhibited the
formation of byproducts (e.g., M4), probably because of the
kinetics of the coexisting processes.
Because no reaction occurred in the absence of HBpin
(Table 1, entry 16), we investigated the role of this compound.
To this end, a B(C6F5)3-catalyzed reaction of 2,3-diphenylpyr-
idine (S1) was carried out in the absence of H2 (Scheme 3a).15
After reaction at 60 °C for 6 h, 1,4-dihydropyridine M1, 1,2-
dihydropyridine M2, tetrahydropyridine M3, and C3-borylated
tetrahydropyridine16 M4 were obtained in 23%, 5%, 10%, and
2% yields (as determined by NMR spectroscopy), respectively,
along with unchanged S1. When the resulting mixture was
treated with H2 (3 MPa) and additional B(C6F5)3 (5 mol %) at
80 °C for 20 h, P1 was obtained in 88% yield by NMR
spectroscopy with an internal standard. Elevation of the
reaction temperature led to the formation of more M4 (see the
10826
ACS Catal. 2021, 11, 10824−10829