Angewandte
Chemie
strates was observed, thus yielding the corresponding
Me2CHCH2CH2CH3 (10c) and Ph2CH2 (11c, Table 1). The
reduction of 10a required heating to 508C for 24 hours,
whereas 11a was reduced after five hours at room temper-
nation of 10a and 11a in the presence of two equivalents of
silane, thus affording quantitative formation of 10c and 11c,
respectively.
The alkyl-substituted dicationic analogues [(SIMe-
s)Et2PF]2+ (6) and [(SIMes)Me2PF]2+ (7) were prepared in
a fashion analogous to 5. To this end, the chlorophosphines
R2PCl (R = Et, Me) were reacted with 1,3-bis(2,4,6-trime-
thylphenyl)imidazol-2-ylidine (SIMes) in the presence of
[Et3Si][B(C6F5)4]·2(C7H8),[21] thus yielding the phosphenium
cation [(SIMes)R2P][B(C6F5)4]. Subsequent oxidation by
XeF2, followed by fluoride abstraction gave 6 and 7 in good
overall yields.
Table 1: Deoxygenation/hydrosilylation of the alkyl- and aryl-substituted
ketones 10a and 11a, respectively.
The compounds 6 and 7 also proved to be effective
catalysts for 10a and 11a, thus yielding the corresponding
products 10c and 11c quantitatively (Table 1). In a similar
vein, oxidation of [(SIMes)Ph2P][B(C6F5)4] with SO2Cl2 and
subsequent chloride abstraction afforded the salt
[(SIMes)Ph2PCl][B(C6F5)4]2 (8), which also proved to be
effective for the deoxygenation of 10a and 11a. It is important
to point out that while the monocationic species
[(SIMes)Ph2PO][B(C6F5)4] (9) is ultimately observed as
a catalyst degradation product, independent synthesis con-
firmed it is completely ineffective as a catalyst. In addition,
several commercially available silanes were screened for the
hydrodeoxygenation of 11a (see the Supporting Informa-
tion), and Et3SiH was selected as the hydride source for the
catalysis described herein.
Catalyst
10a
11a
(mol%)[a]
T [8C]
Conv. [%][b]
T [8C]
Conv. [%][b]
B(C6F5)3 (1.0)
B(C6F5)3 (5.0)
B(C6F5)3 (10)
1 (1.0)
2 (1.0)
3 (1.0)
4 (1.0)
5 (1.0)
6 (1.0)
7 (1.0)
RT
50
50
50
50
50
50
50
50
50
50
50
>99 (10b)
>99 (10b)
>99 (10b)
>99 (10c)
>99 (10c)
>99 (10b)
0
>99 (10c)
>99 (10c)
>99 (10c)
>99 (10c)
0
RT
50
50
RT
RT
50
RT
RT
RT
RT
RT
RT
>99 (11b)
91/9 (11b/11c)
80/20 (11b/11c)
>99 (11c)
>99 (11c)
15 (11b)/85 (11c)
0
>99 (11c)
>99 (11c)
>99 (11c)
>99 (11c)
0
8 (1.0)
9 (1.0)
These results establish that EPCs are highly effective
catalysts for deoxygenation of ketones. To probe the general-
ity of this observation, the EPCs 1 and 5 were used as catalysts
for the deoxygenation of a series of ketones (Tables 2 and 3).
Para-fluoro-substituted benzophenone (12a), as well as para-
bromo- (13a), ortho-chloro- (14a), and ortho-methyl-substi-
tuted derivatives (15a) were converted in high yields into the
respective deoxygenated products 12c–15c by both catalysts
within five hours at ambient temperature. The para-methoxy-
substituted benzophenone (16a) is selectively deoxygenated
in the presence of 2.1 equivalents of Et3SiH, thus yielding the
compound 16c. By using longer reaction times and 3.2 equiv-
alents of silane, the reduction proceeds further with methoxy
ether cleavage, thus affording PhCH2(C6H4OSiEt3) (16d).
Acetophenone (17a) is quantitatively reduced to ethyl
benzene (17c), while the deactivated a-CF3-acetophenone
derivative 18a is selectively hydrosilylated by either 1 and 5
with Et3SiH in a 1:1 stoichiometry to give the silyl ether 18b in
high yield. Interestingly, increasing the catalyst loading to
5.0 mol% and the use of seven equivalents of Et3SiH effects
hydrodefluorination,[2,5] thus yielding ethyl benzene. Finally,
the ketone 19a is quantitatively converted into 19c with 2.1
equivalents Et3SiH. It is interesting to note that for ketones
11a–17a and 19a, reactions with Et3SiH in a 1:1 stoichiometry
gave 1:1 mixtures of the starting material and the correspond-
ing deoxygenation products 11c–17c and 19c. The respective
hydrosilylated intermediates were not observed.
[a] Et3SiH (2.1 equiv) was added to a solution of the catalyst (1 mol%) in
CD2Cl2 and then the substrate (1.0 equiv) was added. [b] Determined by
1H NMR spectroscopy.
ature. The corresponding reaction using 10a and Et3SiH in
a 1:1 ratio gave selective formation of the hydrosilylated
intermediate Me2CHCH(OSiEt3)CH2CH3 (10b), whereas
a 1:1 ratio of 11a and Et3SiH gave 50% conversion to 11c
with 50% of unreacted 11a.
We note that hydrosilylation of ketones has been pre-
viously reported using the Lewis acid catalyst B(C6F5)3.[18] For
comparative purposes, the substrates were treated with silane
in the presence of the catalyst B(C6F5)3 (Table 1). Quantita-
tive hydrosilylation of 10a, thus yielding exclusively 10b, was
observed. In the case of 11a, 1.0 mol% of B(C6F5)3 effected
hydrosilylation of 11a, while 5 mol% B(C6F5)3 gave a 91:9
mixture of 11b and 11c, and 10 mol% B(C6F5)3 gave an 80:20
mixture of 11b and 11c. The data demonstrate that with
B(C6F5)3 as the catalyst, significantly increased catalyst
loading gave poor yields of the deoxygenated product 11c
compared to EPCs. The impact of reduced Lewis acidity of
the EPC was also probed. Whereas 2[6] effects complete
deoxygenation of 10a and 11a, the catalyst 3[19] effects only
the hydrosilylation of substrate 10a to 10b. Nonetheless, 3
does catalyze the reduction of 11a, thus affording 11c in 85%
yield. Further reduction of the Lewis acidity, for example in
the phosphonium ion [Ph3PF]+ (4),[10,20] eliminated catalytic
activity completely and neither hydrosilylation nor deoxyge-
nation was observed. The previously reported dicationic
phosphonium ion 5[9] is an effective catalyst for the deoxyge-
In contrast to aryl-substituted ketones, the dialkyl ketones
20a–27a are selectively hydrosilylated to give 20b–27b at
room temperature after one hour in the presence of either
1 or 5 as the catalyst (Table 3). Longer reaction times (24 h)
and heating of the reaction mixtures to 508C results in
Angew. Chem. Int. Ed. 2015, 54, 8250 –8254
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8251