.
Angewandte
Communications
the fluoro-, chloro-, bromo-, acetamido-, carbomethoxy-,
cyano-, and nitro-substituted compounds which were all
successfully hydrogenated.
which is known to give full conversion with catalyst C in
CH2Cl2 (Scheme 4A).[12] In pure methanol some reactivity
remained, but conversions were significantly lower than in
CH2Cl2 and the ee value fell from 99% to 11%. Addition of
DIPEA quenched the reaction almost completely. Notably,
when a 1:1 mixture of (E)-3-phenylbut-2-enenitrile (3a) with
1 was subjected to these reaction conditions, the unsaturated
nitrile was fully hydrogenated, while 1 did not react (Sche-
me 4B). Apparently, DIPEA enhances the reactivity of the
The b,b-dialkyl-substituted acrylonitriles 3q and 3r
exhibited lower reactivity and required 2 mol% of catalyst
and longer reaction time for full conversion at 08C. The
enantioselectivities were also lower than in the hydrogenation
of the phenyl methyl analogue 3a. Interestingly, the cis-
isomer 3r afforded a higher ee value than the trans-isomer 3q,
contrary to the results obtained with 3a and 3b. The best
results were achieved in the hydrogenation of 3r at 08C,
which led to the saturated nitrile with 82% ee and full
conversion. At À158C the ee value was increased to 88%, but
conversion reached only 51%.
=
catalyst towards the electrophilic C C bond of an a,b-
unsaturated nitrile while inhibiting hydrogenation of more-
[12]
=
electron-rich, less polarized C C bonds.
Thus, selective
hydrogenation of a,b-unsaturated nitriles containing addi-
=
tional C C bonds becomes possible, as demonstrated by the
=
We next investigated to what extent the catalyst loading
and the amount of DIPEA could be reduced without affecting
the enantioselectivity and conversion. In methanol containing
0.23 equivalents of DIPEA as little as 0.05 mol% of catalyst
was necessary to achieve 99% to full conversion within
18 hours. With the exception of 3d (91% ee), the enantiose-
lectivities for 3a, 3g, 3i, 3j, 3n, and 3o remained at the same
levels as those observed in reactions with 1 mol% of catalyst.
At longer reaction times of 112 hours, 96.5% conversion with
95% ee was observed for 3a on a 1.5 mmol scale using only
0.005 mol% catalyst (corresponding to 19300 turnovers). In
methanol with 1 mol% catalyst, the amount of DIPEA could
be reduced to 0.1 equivalents without affecting the ee value
and conversion. In this solvent, even in the absence of
DIPEA, the reaction went to completion, but the ee value
dropped from 95% to 86% (see the Supporting Information,
Table 2).
reduction of 3s (Scheme 4C). The cyano-substituted C C
bond was almost fully hydrogenated to afford 93% of the
monohydrogenation product 5s with 96% ee, and only 6% of
the fully hydrogenated product 5s’. The high preference for
reduction of the trisubstituted C C bond is remarkable in
view of the very high reactivity of terminal C C bonds under
=
=
normal hydrogenation conditions.[11]
Obviously, this base-modified catalyst system must oper-
ate through a different mechanism than under standard base-
free conditions. In view of the pronounced Brønsted acidity of
cationic iridium hydride complexes,[13] a deprotonated neutral
iridium(I) monohydride may be postulated as a reactive
intermediate which is generated by deprotonation of a dihy-
dride complex such as 4 (Scheme 1). A neutral hydride
complex would be expected to be less electrophilic and,
consequently, to release a bound nitrile more easily, thus
opening up a free coordination site, which is required for the
reaction. Moreover, the hydride would be more nucleophilic
than hydrides in a cationic dihydride complex, thus facilitat-
We wondered how standard substrates lacking a nitrile
group behaved under these reaction conditions in methanol
with and without addition of DIPEA. We therefore studied
the hydrogenation of the typical alkene 1 shown in Scheme 4,
=
ing hydride transfer to the electrophilic C C bond of an a,b-
unsaturated nitrile. However, experimental evidence for such
a mechanism is difficult to obtain because of the high
reactivity and sensitivity of iridium hydride species,[14] and
attempts to characterize intermediates other than
(Scheme 1) were unsuccessful so far.
4
Considering the dramatic influence of an added external
base, we wondered whether replacement of the BArF ion by
a basic weakly coordinating anion would bring about a similar
effect.
Indeed complex F, having the sterically hindered 2,6-
ditert-butyl-4-nitrophenolate anion, showed high catalytic
activity in the hydrogenation of 3a in CH2Cl2 and afforded
essentially the same conversion and enantioselectivity as the
corresponding BArF salt E in the presence of DIPEA
(Table 1). High conversion was achieved down to 0.2 mol%
catalyst loading. When the catalyst loading was further
reduced, the ee value remained high, however, conversions
were much lower than that with catalyst E and 0.23 equiv-
alents of DIPEA. This can be explained by degradation of the
phenolate complex F, and was observed over time in solution.
Although the maximum achievable turnover numbers are
lower than those with the DIPEA-modified catalyst systems,
the phenolate complex F offers the advantage to carry out
hydrogenations without addition of excess base under nearly
neutral conditions.
Scheme 4. Reactivity and selectivity studies of a,b unsaturated nitriles
and other alkenes.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 8668 –8671