Angewandte
Communications
Chemie
Then, ortho-substituted pyridines (S17–S28) were sur-
veyed (Table 2). Again, the tolerance of methoxy (P18), halo
(P19), ester (P20), nitro (P21), heterocycles (P22 and P23),
and unactivated olefins (P25 and P26) implies great potential
for this methodology in applications in medicinal chemistry.
The reactions with ortho-di-substituted pyridines selectively
generated the cis-disubstituted products (P27 and P28) in
high yields. Notably, use of PhNH2 instead of Ph2NH as the
proton donor provided better yields for some ortho-substi-
tuted pyridines (S22, S23, S27, and S28).
Interestingly, when we subjected the para-substituted
pyridines (S29–S37) to these reaction conditions, one
carbon–carbon double bond on the hetero-ring was retained
(Table 2). Furthermore, their reactivities were found to be
sensitive to the electronic properties of their substituents, and
yields were higher for substrates bearing electron-withdraw-
ing groups (S31–S33, S35 and S37) than those with electron-
donating groups (S30 and S36). The nitrile group in S34 was
hydrosilylated under the standard reaction conditions,[18] but
was retained (P34) by using HBpin in place of Ph2SiH2.
Nevertheless, by harnessing the cascade reduction we obtain
a series of diversely functionalized tetrahydropyridines which
are difficult to synthesize by other methods.
The reduction is also reactive with other electron-
deficient N-heteroarenes (Table 3).[19] The reactions of quino-
lines (S38 and S39), isoquinoline (S40), pyrazine (S41),
quinoxaline (S42), and phenanthrolines (S43–S45) produced
the reduced heterocycles in high yields. For these reductions,
switching the proton donor to PhNH2 generally provided
better yields except for the reactions with isoquinoline (S40)
and pyrazine (S41) where Ph2NH was still optimal.
Scheme 2. Mechanistic studies.
two steps. In contrast, when we directly reacted S1 with
decreased loadings of PhMe2SiH and Ph2NH, INT1 was not
observed in the reaction mixture. Instead, the fully reduced
product P1 was obtained in 33% yield after hydrolysis,
together with 56% of the recovered starting material
(Scheme 2b). These results prove that the reduction is
initiated by dearomative hydrosilylation. However, once
formed under the standard reaction conditions, the N-silyl
1,4-dihydropyridine intermediate (INT1) will undergo rapid
transfer hydrogenation to give the final product.
Table 3: Scope with respect to other N-heteroarenes.[a]
The deuterium-labelled hydrosilane (Ph2SiD2) was used in
the reaction with S4 and S31. The deuterium was exclusively
transferred to the C2-, C4-, and C6-positions in the hetero-
cycles (Scheme 2c and d). These results confirm the hydro-
silane as the sole hydride donor in both steps of dearomative
hydrosilylation and transfer hydrogenation (for the overall
reaction mechanism that is proposed based on previous
studies,[9,12,13] see the Supporting Information). Moreover, it
was the same as the observation made by Chang and co-
workers in the study of the cascade hydrosilylation reaction,[9]
the initial dearomative hydrosilylation step proceeds in a 1,2-
addition fashion with para-substituted pyridines. The follow-
ing transfer hydrogenation of the enamine double bond gives
the tetrahydropyridine product.
[a] Unless otherwise specified, all reactions were performed with 5 mol%
B(C6F5)3, 0.2 mmol substrate (S38–S45), 5 equiv of Ph2SiH2 and 4 equiv
of PhNH2 in 0.5 mL toluene at 1108C for 24 h under N2. Yields are those
of the isolated products. [b] Used 10 mol% B(C6F5)3. [c] Used 5 equiv
Ph2SiH2 and 4 equiv Ph2NH. [d] Used 6 equiv of Ph2SiH2 and 5 equiv of
PhNH2. [e] cis/trans=1.7:1.
In summary, we have developed a B(C6F5)3-catalyzed
metal-free pyridine reduction strategy by a cascade process of
dearomative hydrosilylation (or hydroboration) and transfer
hydrogenation. The broad functional-group tolerance pro-
vides easy access to an array of diversely functionalized
piperidines and tetrahydropyridines which are valuable
building blocks in synthesis. Its suitability for use in the
reduction of other N-heteroarenes has also been demon-
strated. Further studies utilizing this cascade reduction in
synthesis are underway in our laboratory.
To study the reaction mechanism, we performed several
experiments (Scheme 2). First, when S1 was treated with
1.2 equivalents of PhMe2SiH in the presence of 10 mol%
B(C6F5)3 at 1108C for 4 hours, the partially reduced N-silyl
1,4-dihydropyridine INT1 was obtained in 55% yield together
with 41% unreacted starting material as determined by NMR
analysis of the reaction mixture (Scheme 2a). Treatment of
this mixture with additional PhMe2SiH, Ph2NH, and fresh
catalyst, led to the fully reduced product in 78% yield over
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!