CHEMCATCHEM
COMMUNICATIONS
direct reduction of the benzaldehyde (entry 1). By changing
the silane from PMHS to TMDS (2.5 equiv.), the selectivity in-
creased to 96% in favour of amine 3. At 708C in THF, the selec-
tivity towards the formation of the amine increased for both
PMHS and TMDS (respectively 94 and 97%, entries 3 and 4).
On decreasing the temperature to 508C, the imine intermedi-
ate 2 was obtained as the main product (81%, entry 5). We
have also performed the reaction in the absence of 4 ꢀ molec-
ular sieves: the selectivity decreased and a 23:77 molar ratio
mixture of benzyl alcohol 1 and amine 3 was obtained
(entry 6). The quantity of TMDS could be decreased to
1.5 equiv. without any alteration of the selectivity (entry 7). In
contrast, if the catalytic loading of nickel acetate was lowered
to 2.5 mol%, a 1:1 mixture of the imine 2 and the amine 3 was
obtained (entry 9). Finally, if the quantity of aniline was low-
ered from 1.5 to 1 equiv., 8% of benzyl alcohol 1 was detected,
which showed that the formation of the imine was in competi-
tion with the reduction of the aldehyde (entry 10). To reduce
the formation of the undesired alcohol but only use a stoichio-
metric amount of p-anisidine, a sequential procedure was
used; the silane was added after 8 h of reaction and the de-
sired amine then obtained quantitatively (entry 11).
Table 2. Scope
of
the
reductive
amination
reaction.[a]
Entry[a]
Product
R
Yield[b,c]
[%]
1
OMe
H
>97 (97)
91 (90)
93 (79)
2[d]
3[e]
F
4
OMe
H
>97 (97)
>95 (81)
82 (82)
5[d,f]
6[e]
F
7[d,e]
OH
87 (80)
91 (46)
>97 (62)
41
8[d,e]
CO2Me
NHCOMe
CN
9[d,e]
10[d,e]
11[d,e]
NO2
14
12
13
14
H
>97 (95)
91 (89)
91 (85)
OMe
CF3
With the optimised conditions in hand (1.1 equiv. amine,
1 equiv. aldehyde, 708C, 8 h, then 1.5 equiv. TMDS, 708C, 16 h),
we focused on the scope of the reaction. At first, various
amines were tested with aromatic aldehydes such as benzalde-
hyde and p-anisaldehyde (Table 2, entries 1–6): with the acti-
vated p-methoxyaniline, the corresponding amines were ob-
tained with 97% isolated yields. In contrast, with non-activated
aniline, the reduction of the imine required a longer reaction
time of 40 h to reach completion and, with the electron-defi-
cient 4-fluoroaniline, 4 equiv. of TMDS were needed to obtain
the secondary amine in good yields (79–82%). The electronics
of the amine partners seemed to have a strong influence on
the reaction because, with an electron-rich amine, the reaction
was performed in short time, whereas the presence of elec-
tron-withdrawing group induced longer reaction times (8 vs.
40 h). The group tolerance was also studied on the aldehyde
partner by reaction with 1.1 equiv. of p-anisidine. Thus, the tol-
erance towards acidic phenol moieties was also demonstrated,
as was the selectivity towards esters and amides functional
groups; the resulting reactions led to high conversions of
amines (87–97%) and isolated amines with non-altered func-
tional groups isolated in moderate to good yields (46–80%;
entries 7–9). In contrast, with p-cyano- and p-nitro-benzalde-
hyde derivatives, the conversions towards the desired amine
were low or moderate (41 and 14%, respectively).
15[e]
16[e]
H
OMe
>97 (97)
>97 (97)
17
18[c]
CH2Ph
CH3À(CH2)11À
>97 (71)
>97 (67)
19[e]
20[e]
Ph
>97
>97 (97)
CH3À(CH2)11À
21
22[d,h]
p-OMeÀC6H4
0[g]
58
PhCH2ÀCH2
[a] Typical conditions: To a solution of Ni(OAc)2 (5 mol%) and 4 ꢀ molecu-
lar sieves (400 mg), PCy3 (10 mol%), amine (1.1 mmol, 1.1 equiv.) and al-
dehyde (1 mmol) were added in that order and stirred at 708C for 8 h.
Then TMDS (1.5 mmol, 1.5 equiv.) was added and the reaction stirred for
16 h, after which hydrolysis was performed by adding MeOH (2 mL) and
NaOH (2 mL), followed by stirring overnight. [b] Yields were determined
by using 1H NMR spectroscopy. [c] Isolated yields after purification by
column chromatography in parentheses. [d] t=40 h instead of 16 h,
[e] 4 equiv. TMDS, [f] 1.5 equiv. amine. [g] p-Methoxybenzyl alcohol was
observed. [h] t=48 h instead of 8 h.
If using secondary amine such as the dipropylamine, the re-
action was more difficult to perform. If reacting with p-methox-
ybenzaldehyde, no resulting product of reductive amination
was detected and the sole product obtained was the corre-
sponding alcohol (entry 21). Notably, the dipropylamine could
be condensed with the hydrocinnamaldehyde with moderate
conversion (58%; entry 22). The transformation of ketones
under similar reductive amination conditions was unsuccessful,
with only low amounts of amine obtained. In comparison to
other transition metal catalysts (iron[5l] and zinc[5h]), the activity
of this nickel catalytic system was competitive in the reductive
amination of aldehydes with siloxanes as the reducing reagent.
To try to develop better insights into the reaction conditions,
several stoichiometric experiments were conducted in order to
identify which type of active catalysts could be generated
The reaction of aldehyde with benzylamine proceeded well
and the products were isolated with 85–95% yields, irrespec-
tive of the nature of the substituent on the aromatic ring of
the aldehyde (entries 12–14). If the reaction was performed
with aliphatic aldehydes such as hydrocinnamaldehyde and oc-
tanal, the corresponding secondary amines were obtained in
good to excellent conversions and yields (entries 17–20). Nota-
bly, aliphatic amines such as dodecylamine were also suitable
for the reductive amination (97% yield, entries 15, 16, 18 and
20).
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 2861 – 2864 2862