2
G. Wienhöfer et al. / Journal of Organometallic Chemistry xxx (2013) 1e4
2
. Experimental
Table 2
Iron-catalyzed selective reduction of cinnamaldehyde: influence of solvents.a
Fe(BF
4 2 2 3
) 6H O/ PP
Fe(BF
4
)
2
$6H
2
O (0.7 mg; 0.002 mmol) and tris[2-(diphenyl-
CH PPh ; tetraphos] (1.4 mg;
.002 mmol) are placed in a Schlenk-tube under argon atmosphere.
mL dry tetrahydrofurane is added and the purple solution is
L; 0.5 mmol) and 100 L n-
O
OH
HCO
2
H, solvent
phosphino)-ethyl]phosphine [P(CH
2
2
2 3
)
0
1
Entry
Solvent
Conv.b
Yieldb
[%]
Selec.b
[%]
stirred for 2 min. Cinnamaldehyde (63
hexadecane as an internal GC-standard are injected and a sample is
taken for GC-analysis. The solution is heated to 60 C and the re-
action starts by addition of 1.1 equiv formic acid (22 mL; 0.55 mmol).
After 2 h, a second sample is taken for GC-analysis and conversion
and yield are determined by comparison with authentic samples.
For the isolation, the reaction is scaled up by a factor of 20. When
the reaction is completed, the reaction solution is diluted with a
mixture of n-hexane and ethyl acetate (3:1), filtered through a plug
of silica and the solvent removed in vacuum.
m
m
[%]
ꢀ
1
2
3
4
5
6
7
Toluene
MeOH
i-PrOH
THF
EtOH
2-Me-THF
t-AmOH
e
e
e
70
25
31
48
9
5
7
16
31
16
9
64
99
33
99
99
14
14
a
Reactions conditions: 0.5 mmol cinnamaldehyde, 0.75 mol% [FeF(PP
mL solvent, 2 equiv FA, 40 C, 2 h.
Determined by GC using n-hexadecane as an internal standard.
3
)][BF
4
],
ꢀ
1
b
3
. Results and discussion
Variation of the catalyst concentration (Table 3, entries 4e7)
demonstrated that 0.5 mol% of the catalyst is sufficient to maintain
complete conversion and quantitative yield.
At the start of this project, we used cinnamaldehyde as the
benchmark substrate for the reduction of unsaturated aldehydes.
The catalytic experiments were performed in the presence of a
Best results were obtained applying one equivalent of formic
acid (Table 3, entries 8e11). This finding is in contrast to our pre-
vious reductions of nitroarenes, where an excess of reducing agent
was required due to partly decomposition of formic acid into
hydrogen and carbon dioxide [16]. Apparently, under the present
conditions such decomposition of formic acid does not take place
and no increase of pressure is observed. Using lower amounts of
acid led to no complete conversion (Table 3, entry 8).
The highest yield was obtained using a catalyst loading of 0.4 mol
% and a slight excess of 1.1 equivalents of formic acid (Table 3, entry
12). Notably, the reaction time can be significantly reduced at higher
catalyst loading. Hence, applying 2 mol% of catalyst complete con-
version was obtained within only 10 min (Table 3, entry 13).
Next, we investigated the general applicability of our catalyst
combination of Fe(BF)
ethyl]phosphine [P(CH
4
$6H
2
O and tris[2-(diphenyl-phosphino)-
; (PP )] using formic acid (FA) as
2
CH PPh
2
2
)
3
3
reducing agent (Table 1). The single components of the catalyst
system did not show any reactivity at all while the combination of
ligand and cationic iron species exhibited good conversion (Table 1,
3
entries 2e4). Adding the ligand PP to the iron salt the complex
[
FeF(PP )][BF ] is formed in situ which is indicated by a change of
3
4
the colorless solution into deep purple. Next, the defined complex
was applied under the same reaction conditions. As expected,
reactivity comparable to the in situ system was obtained (Table 1,
entry 5) [19]. Remarkably, in all the reactions the conjugated CC-
double bond is not attacked. Decarbonylation, Tishchenko-type
dimerization or aldol condensation is also not observed and full
selectivity towards the cinnamyl alcohol is obtained.
system. First, we focused on the reactivity of different a,b-unsatu-
rated aldehydes (Table 4). Cinnamaldehydes bearing substituents in
Next, we investigated the influence of different solvents on the
benchmark reaction. While in toluene no reactivity is observed,
applying different alcohols and ether gave high to moderate con-
versions (Table 2). However, in alcohols the major products were
the undesired hemiacetal and acetal. Here, cinnamic alcohol was
obtained only in minor amount. Tetrahydrofurane led to the high-
est yields and therefore was chosen as the solvent for further re-
actions. As shown in Table 3, the reaction temperature has a major
influence on the reactivity of the system. While the increase from
Table 3
Optimization of selected reaction parameters for the selective reduction of cinna-
maldehyde.
a
Fe(BF
4
)
2
6H
2 3
O/ PP
O
OH
HCO
2
H, THF
Conv.b
[%]
Yieldb
[%]
Selec.b
[%]
Entry
Catalyst loading
[
mol%]
ꢀ
ꢀ
2
0
C to 40 C raised the conversion from 11% to 31%, a further
1c
0.75
0.75
0.75
0.2
0.3
0.4
0.5
0.3
0.3
0.3
0.3
0.4
2.0
11
31
100
20
31
40
100
75
97
83
60
11
31
>99
20
31
40
>99
75
97
83
60
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
ꢀ
increase to 60 C led to full conversion (Table 3, entries 1e3).
d
2
3
4
5
6
7
Table 1
Iron-catalyzed transfer hydrogenation of cinnamaldehyde: comparison of the in
a
situ-generated system and the defined iron-tetraphos complex.
PPh
2
Fe(BF
4
)
2
6H
2
O/ PP
3
e
O
OH
8
9
3
PP =
P
PPh2
f
HCO
2
H
g
PPh
2
10
h
1
1
1
2
i
j
100
100
>99
>99
Conv.b
Yieldb
Selec.b
13
Entry
Catalyst
a
ꢀ
[
%]
[%]
[%]
Reactions conditions: 0.5 mmol cinnamaldehyde, 1 mL THF, 2 equiv FA, 60 C, 2 h.
b
c
d
e
f
Determined by GC using n-hexadecane as an internal standard.
1
2
3
4
5
e
e
e
e
ꢀ
2
4
0
1
1
0
0
C.
C.
Fe(BF
PP
Fe(BF
4
)
2
2
$6H
$6H
2
O
e
e
e
ꢀ
3
e
e
e
.75 equiv FA.
equiv FA.
.25 equiv FA.
4
)
2
O/PP
3
31
30
31
30
>99
>99
3 4
[FeF(PP )][BF ]
g
h
i
a
Reactions conditions: 0.5 mmol cinnamaldehyde, 0.75 mol% catalyst, 1 mL THF,
1.5 equiv FA.
1.1 equiv FA.
10 min.
ꢀ
2
equiv FA, 40 C, 2 h.
b
j
Determined by GC using n-hexadecane as an internal standard.
Please cite this article in press as: G. Wienhöfer, et al., Journal of Organometallic Chemistry (2013), http://dx.doi.org/10.1016/
j.jorganchem.2013.06.010