Rhodium carbene complexes as versatile catalyst precursors for Si-H bond activation

Article abstract of DOI:10.1002/chem.201102197

Rhodium(III) complexes comprising monoanionic C,C,C-tridentate dicarbene ligands activate Si-H bonds and catalyse the hydrolysis of hydrosilanes to form silanols and siloxanes with concomitant release of H₂. In dry MeNO₂, selective formation of siloxanes takes place, while changing conditions to wet THF produces silanols exclusively. Silyl ethers are formed when ROH is used as substrate, thus providing a mild route towards the protection of alcohols with H₂ as the only by-product. With alkynes, comparably fast hydrosilylation takes place, while carbonyl groups are unaffected. Further expansion of the Si-H bond activation to dihydrosilanes afforded silicones and polysilyl ethers. Mechanistic investigations using deuterated silane revealed deuterium incorporation into the abnormal carbene ligand and thus suggests a ligand-assisted mechanism involving heterolytic Si-H bond cleavage. Ligand-assisted Si-H bond cleavage: Rhodium-catalysed Si-H bond activation provides a methodology for the silyl group to be transferred to oxygen centres, thus providing access to siloxanes and silanols (using H ₂O), alkoxysilanes (using ROH) and to polysiloxanes, such as silicones, when using difunctional silanes. The process is pH neutral, avoids sensitive silylating agents and produces H₂ as useful and exclusive side product. Copyright

Full text of DOI:10.1002/chem.201102197

DOI: 10.1002/chem.201102197
À
Rhodium Carbene Complexes as Versatile Catalyst Precursors for Si H
Bond Activation
Anneke Krꢀger and Martin Albrecht*[a]
À
Abstract: Rhodium
comprising monoanionic C,C,C-triden-
tate dicarbene ligands activate Si H
G
providing a mild route towards the pro-
tection of alcohols with H2 as the only
by-product. With alkynes, comparably
fast hydrosilylation takes place, while
carbonyl groups are unaffected. Fur-
ther expansion of the Si H bond acti-
vation to dihydrosilanes afforded sili-
cones and polysilyl ethers. Mechanistic
investigations using deuterated silane
revealed deuterium incorporation into
the abnormal carbene ligand and thus
suggests a ligand-assisted mechanism
bonds and catalyse the hydrolysis of
hydrosilanes to form silanols and silox-
anes with concomitant release of H2. In
dry MeNO2, selective formation of si-
loxanes takes place, while changing
conditions to wet THF produces sila-
nols exclusively. Silyl ethers are formed
when ROH is used as substrate, thus
Keywords: carbenes · ligand coop-
erativity · mesoionic complexes ·
rhodium · silane activation · silox-
anes
À
involving heterolytic Si H bond cleav-
age.
Introduction
sion of hydrosilanes, and the alcoholysis of hydrosilanes to
form silyl ethers has been reviewed.[7] A critical step evi-
The silicon–oxygen linkage is a useful synthon that has
found widespread application in silicon-based polymers
(e.g., in silicones)[1] and in organic synthesis (e.g., as silyl
dently constitutes the cleavage of the Si H bond.[8]
À
Rhodium
N
(Cp*=C5Me5À =pentamethylcyclopentadienyl anion) have
recently shown great promise in the activation of unreactive
bonds.[9] Specifically, Bergman and Brookhart and independ-
ently Carmona and co-workers demonstrated that Cp*–
À
ethers).[2] The formation of the Si O bond has been well es-
À
tablished and classical synthetic routes use either silyl chlor-
ides, silyl triflates (OTf) or alkoxysilanes as starting materi-
als (Scheme 1a and 1b).[3] Common to all these routes is the
rhodiumACTHNUTRGENUG(N III) complexes activate Si H bonds and catalyse
hydrogen-exchange processes in silanes.[10] The tridentate di-
carbene ligand in the rhodiumACTHNUTRGNEUGN(III) complex 1 represents a
surrogate of Cp*, featuring a monoanionic, facially coordi-
nating C,C,C-tridentate mode (Figure 1).[11] Owing to the
Scheme 1. Generic methods for the formation of siloxanes.
high reactivity of the precursor, requiring rigidly anhydrous
conditions. In addition, the formed HCl or HOTf has to be
scavanged by a base, thus inevitably producing stoichiomet-
ric amounts of an undesired salt. The catalytic activation of
Figure 1. Dicarbene rhodiumACHTNUTRGNE(NUG III) complex 1, comprising a facially coordi-
À
nating, formally monoanionic ligand that is isolobal to Cp-type ligands.
air-stable and moisture-insensitive hydrosilanes through Si
H bond cleavage offers an attractive method to circumvent
the use of sensitive presursors and to perform the reaction
with high atom economy (Scheme 1c).
isolobal relationship of this dicarbene ligand with the Cp*
unit, we became interested in probing the catalytic activity
of the air-stable and easily accessible rhodiumACHTNUTRGNE(NUG III) complex
A few transition-metal complexes,[4,5] as well as heteroge-
neous systems,[6] have been reported to catalyse the conver-
1 in bond-activation processes. Herein, we report on the ver-
satility of this and related complexes as catalyst precursors
for the hydrolytic oxidation of hydrosilanes.
[a] A. Krꢀger, Prof. Dr. M. Albrecht
School of Chemistry & Chemical Biology
University College Dublin
Belfield, Dublin 4 (Ireland)
Fax : (+) 353 1716 2501
E-mail: martin.albrecht@ucd.ie
652
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 652 – 658

FULL PAPER
Results and Discussion
ing the solvent. In wet MeCN, silanol formation was largely
dominant (87:13 and 98:2 ratio, entries 7 and 8), though pro-
longed reaction times were required. When using wet THF,
full conversion exclusively towards the silanol product was
observed within 2 h (entry 9). In the absence of extraneous
H2O, both the rate of the reaction and the selectivity drop-
ped (entry 10). Although initially, silanol is preferentially
produced under these conditions, the siloxane is the pre-
dominant product at elevated conversion. Thus, the catalytic
oxidation of silanes can be effectively switched by changing
the solvent between dry MeNO2 and wet THF, affording
either pure siloxane or the silanol product exclusively,[14]
without derogating reaction rates.
Silanol versus siloxane formation: The activity of complex 1
in Si H bond activation was probed by using PhMe2SiH as
a model substrate. Initial runs in MeNO2 revealed mixtures
of silanol and siloxane[12] and concomitant formation of H2,
indicated by the evolution of gas. Dihydrogen generation
was confirmed by catalytic runs in deuterated CD3NO2 and
À
1
subsequent H NMR spectroscopic analysis, which revealed
a diagnostic singlet at dH =4.53 ppm. Table 1 summarises the
results obtained in the oxidation of dimethylphenylsilane
catalysed by complex 1.
Table 1. Hydrolytic oxidation of dimethylphenylsilane to siloxane and si-
lanol.[a]
Solvent H2O[b]
[mmol] [8C] [h]
T
t
Conversion Siloxane:Silanol
[%]
ACHTUNGTRENNUNG
1[c] MeNO2
MeNO2
3[d] MeNO2
–
–
–
–
103
103
103
80
103
25
80
25
65
65
2
0
–
2
0.5/2
2/27
2/5
0.5/1
2/24 31/>99
2/16 50/>99
2/24 37/>99
0.5/2
0.5/5
88/>99
1/79
29/>99
95/98
>99:1/>99:1
>99:1/>99:1
88:12/68:32
8:92/10:90
6:94/12:88
12:88/13:87
3:97/2:98
4
5
6
MeNO2
MeNO2 6.4
MeNO2 6.4
7
8
9
MeCN
MeCN
THF
2.0
2.0
2.0
–
Figure 2. Reaction profile for the formation of siloxane and silanol in wet
^
*
MeNO2: & silanol,
siloxane,
silane (for conditions, see Table 1,
24/97
3/38
<1:99/<1:99
17:83/84:16
entry 5); prolonged reaction times induce a gradual and partially cata-
lysed transformation of silanol to siloxane.[14]
10
THF
[a] Conditions unless otherwise stated: silane (1.0 mmol) and catalyst
(1 mol%) in the corresponding solvent (5 mL) using mesitylene as inter-
nal standard; product mixtures analysed by GC-MS. [b] Extraneous
water. [c] Control experiment without the addition of catalyst. [d] Cata-
lyst (0.1 mol%) was used.
Time-dependent monitoring of the reaction (Figure 2) did
not reveal any induction period and the kinetic reaction pro-
file was fully reproducible, which is indicative of homogene-
ous catalysis. In support of this notion, catalyst recyling was
demonstrated by repetitive addition of a second and third
equivalent of substrate after 20 and 45 min, respectively;
that is, at time intervals at which conversions were not yet
complete. This cumulative run reached 67% overall conver-
sion 20 min after addition of the last batch and 85% overall
conversion (255 turnover number) after 2.5 h, demonstrating
the living nature of the catalytically active species.
When using essentially dry MeNO2 at reflux temperature,
full conversion of the hydrosilane selectively to the corre-
sponding siloxane PhMe2Si-O-SiPhMe2 was reached within
2 h (entry 2).[13] Reducing the catalyst loading from 1 to
0.1 mol% decreased the reaction rate significantly (79%
after>24 h), even though the selectivity remained high
(entry 3). The reaction temperature affected both the reac-
tion rate and the selectivity. At 808C, only 29% conversion
was reached after 2 h (entry 4, cf. full conversion at reflux)
and 5 h were required to achieve complete silane oxidation.
In addition, the selectivity dropped markedly, yielding a 2:1
ratio of siloxane and silanol. In an attempt to promote sila-
nol formation, the water content in the solvent was in-
Role of the catalyst: The activity of a series of related and
previously described[11,15] carbene rhodium
ACHTNUTRGNEUNG(III) complexes
2–6 was evaluated in order to identify factors that may criti-
cally influence the catalytic performance (Figure 3). Under
standard conditions (dry MeNO2 at reflux temperature), the
related tridentate dicarbene complexes 2 and 3, containing
one or two “normally” bound (i.e., bound through C2 of the
heterocyclic ring; “abnormally” bound carbenes bind
through C4 of the heterocyclic ring) carbene ligands, respec-
tively, also catalysed siloxane formation with high selectivity
(Table 2, entries 2 and 3). The activity of 2 was significantly
lower and long reaction times were required to reach full
conversion. With complex 3, containing two normally bound
carbenes, rates were higher, though not as high as with 1,
reaching 84% conversion after 2 h. In contrast, the normal
dicarbene complex 4 was essentially inactive in the same
À
creased. Upon addition of extraneous H2O (ca. 6 mmol), Si
OH bond formation was dominant, though not exclusive
(entry 5). Moreover, faster reaction rates were observed,
leading to full conversion within 1 h. Upon lowering the
temperature to room temperature, again a loss of activity
was observed, though the siloxane/silanol ratio did not
change significantly compared to the high-temperature reac-
tion (entry 6), suggesting that the quantities of water affect
the selectivity substantially stronger than the reaction tem-
perature. Higher ratios of silanol were obtained upon chang-
Chem. Eur. J. 2012, 18, 652 – 658
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
653

A. Krꢀger and M. Albrecht
induces higher catalytic activity and better selectivity than
the normal mode and the best performance was achieved
with complex 1 as catalyst precursor. Catalytic runs in the
presence of PPh3 (10 mol equiv) gave slightly lower conver-
sions (entry 7), though the reaction solution remained clear
even when all substrate was consumed. In the absence of
phosphine, formation of a dark residue was observed unless
further substrate was added. The hydrolytic oxidation is not
affected by the presence of mercury(0).[17] Addition of Hg0
(0.5 mL, ca. 3000 mol equiv) to the reaction solution after
10 min (42% conversion) did not stop the reaction (entry 8),
even though an instant decomposition reaction was ob-
served, presumably owing to partial rhodiumACTHUNRTGNEUNG(III) reduction
Figure 3. Catalysts employed in the hydrolytic oxidation of silanes.
in the presence of elemental mercury, H2 and HSiR3. Of
note, Rh/C was significantly less active and less selective
(entry 9). The combination of these results,[18] that is, repro-
ducible kinetics, absence of induction time, insensitivity to
Hg0, stabilisation by PPh3 and the isolation of the rhodium
complex after catalytic conversion (see below), all lend
strong support to a homogenously catalyzed process.
Table 2. Hydrolytic oxidation of dimethylphenylsilane to siloxane and si-
lanol.[a]
Rh catalyst
t
Conversion
[%]
Siloxane:Silanol
[h]
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
0.5/2
88/>99
49/96
72/84
3/70
78/96
73/92
82
>99:1
>99:1
91:9/>99:1
>99:1
>99:1
91:9/>99:1
>99:1
>99:1
Further mechanistic insights into the rhodium-mediated
2/22
0.5/2
2/22
0.5/2
0.5/2
2
2
2
À
Si H bond activation were obtained from experiments using
À
the monodeuterated silane Et3Si D as substrate. Upon reac-
tion with complex 1 in the presence of small quantities of
H2O, deuterium exchange into the heterocyclic C5-position
of the carbene ligand was observed. The absence of a reso-
nance in the H NMR spectroscopy at d=7.4 ppm and the
appearance of a signal in the H NMR spectrum exclusively
1+PPh3
1+Hg0
Rh/C[b]
89
41
1
72:28
2
[a] General conditions: Me2PhSiH (1.0 mmol) and rhodium complex
(1 mol%) in MeNO2 (5 mL) at reflux using mesitylene as internal stan-
dard; conversion and selectivity determined by GC-MS. [b] 5% Rh on C
(21 mg) was used.
at this frequency confirmed the selective and complete deut-
eration of the positions adjacent to the rhodium-bound
À
carbon. This observation suggests a ligand-assisted Si D
bond-cleavage process and does not support an outer-sphere
À
time period (entry 4). Lower activity of complex 4 may be
mechanism involving nucleophilic attack of H2O at a Si D
owing to the absence of a Rh Calkyl bond as potential site
s-complex.[19] Instead, oxidative addition of the hydrosilane
À
[16]
À
for 1,2-addition of the polarised Si H substrate. Alterna-
to the metal centre to form a (silyl)ACTHNGUTERNNU(G hydride)rhodium(V)
tively, acetate displacement in 4 may be more demanding
than MeCN substitution in 1 or 3, thus hampering the acti-
vation of the catalyst precursor. The former hypothesis was
probed by using complexes 5 and 6 as catalyst precursors.
These complexes contain a rigidly bidentate-bonding dicar-
bene ligand with a methylene group rather than a trimethy-
lene linker between the heterocycles, thus preventing ligand
complex may occur,[20] presumably facilitated by the high
electron-donor ability of the abnormal carbene ligands
(Scheme 2, route A).[21] Subsequent tautomerisation of the
rhodium hydride to a carbene isomer would induce a sp2 to
À
À
C H bond activation and the formation of a Rh Calkyl bond.
When employing complex 5, featuring an abnormally bound
dicarbene ligand, only slightly lower performance was ob-
served, as compared to the tridenate analogue 1 (entry 5).
Complex 6, which has two normally bound carbene ligands,
was nearly as effective as the abnormal congener and slight-
ly more active than the tridentate complex 3, which also
contains two normally bound carbene ligands (entry 6). Ap-
À
parently, the presence of a Rh Calkyl unit is not essential for
À
Si H bond activation and ensuing catalytic activity. Direct
mechanistic conclusions are difficult to draw from this cata-
lyst evaluation, for example, because complexes 3 and 6
differ both in the formal charge at the rhodium centre, as
well as in the number of weakly bound solvent ligands.
However, the abnormal carbene-bonding mode consistently
Scheme 2. Suggested pathway for deuterium incorporation into the NHC
ligand involving either oxidative addition (route A) or ligand-assisted
À
direct Si D bond cleavage (route B).
654
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 652 – 658

À
Si H Bond Activation with Rhodium Carbene Complexes
FULL PAPER
sp3 rehybridisation of the heterocyclic C5 carbon. Alterna-
tively, this carbene intermediate may be accessible through
troducing the tert-butyldimethylsilyl (TBDMS) protecting
group on primary alcohols. The acid- and base-free condi-
tions are particularly useful for highly complex and pH-sen-
sitive substrates. Of note, larger silyl groups are less reactive
and the (triethyl)silyl ethers were obtained only in moderate
yields (entry 6), perhaps owing to sterically disfavoured nu-
cleophilic attack at the rhodium-bound silicon centre (cf.
À
direct heterolytic bond cleavage of the Si D bond in a
ligand-assisted process across the metalla–allyl fragment in a
formally isohypsic process (route B). Hydrolysis then produ-
ces the silanol and H2 (or HD). Both routes imply the active
participation of the carbene ligand, as suggested earlier for
water oxidation and the reaction of abnormal carbene com-
plexes with H+.[15a,22]
À
Si D activation, see above).
À
Hydrosilylation: The catalytic activity of complex 1 in Si H
Silyl ether formation through silane alcoholysis: Reaction of
PhMe2SiH in the presence of ROH instead of extraneous
H2O cleanly afforded the corresponding silyl ether
bond activation was further investigated in the presence of
other functional groups, in particular, because of the recent
success of N-heterocyclic carbene rhodium complexes as cat-
alyst precursors in hydrosilylation reactions.[27] Under reac-
tion conditions as described for entry 2 in Table 1, phenyla-
cetylene was fully hydrosilylated within 30 min. No siloxane
and only trace amounts of silanol (<1%) were observed.
Styrene hydrosilylation was considerably slower (24% after
2 h) and most hydrosilane was converted to siloxane, owing
to competitive reaction with residual H2O. Under the same
reaction conditions no reactivity was observed towards ace-
tophenone and siloxane was the predominant product (80%
conversion after 2 h). The slightly lower rate of siloxane for-
mation as compared to runs in the absence of ketone (cf.
>99% conversion, Table 1, entry 2) may be attributed to
the coordinating ability of acetophenone to the metal
centre. No olefin or ketone reduction owing to the generat-
ed H2 was detected,[28] which is in agreement with the previ-
ously observed low reactivity of complex 1 in hydrogenation
reactions.[11] These results together with those from alcohol
silylation provide the following reactivity preference of hy-
drosilanes when catalysed by complex 1:
(Table 3).[23] Therefore, this rhodium-mediated Si H bond
À
Table 3. Protection of primary and secondary alcohols using dimethyl-
phenylsilane.[a]
Alcohol
Silane
Product
t
Yield
[%]
[h]
1
4
59
85
À
1
H SiMe2Ph
1
3
72
83
2[b]
À
H SiMe2Ph
1
3
34
66
À
3
H SiMe2Ph
À
4
5
iPrOH
H SiMe2Ph
iPrOSiMe2Ph
1
37
1
4
66
>99
À
H SitBuMe2
1
3
17
28
À
6
7
H SiEt3
À
H SiPh3
1
63
This trend suggests the possibility to carry out selective si-
[a] Conditions unless otherwise stated: silane (1.0 mmol), alcohol
(1.0 mmol) and 1 (1 mol%) in MeNO2 (5 mL) at reflux; product yield
and selectivity determined by GC-MS or 1H NMR spectroscopy.
[b] Me2PhSiH (1.5 mmol) was used.
activation constitutes an alternative route to silyl-protected
alcohols,[24,25] avoiding the use of moisture-sensitive R3SiCl
or R3SiOTf, as typically employed for alcohol protection.[26]
The reaction does not require any sacrificial acid or base, as
H2 is the only formed side product. To probe the potential
of this method, a series of different alcohols and hydrosi-
lanes were evaluated for silyl ether formation.
When using one equivalent of dimethylphenylsilane, pri-
mary alcohols, such as benzylalcohol and 2-phenylethanol,
were readily silylated and after 4 h about 85% conversion
to the corresponding silyl ether was reached (entries 1 and
2). Secondary alcohols were converted much slower (en-
tries 3 and 4) and competitive reaction with residual H2O
was deduced from the substantial amounts of siloxane and
silanol detected after 4 h (GC-MS). Variation of the silane
from PhMe2SiH to tBuMe2SiH gave better yields and quan-
titative silyl ether formation was achieved within 4 h
(entry 5). This method thus represents an attractive substi-
tute to the standard procedure (DMF, imidazole)[26] for in-
lylations. To verify this hypothesis, substrates with different
functional groups were reacted in the presence of stoichio-
metric quantities of hydrosilane and catalytic amounts of
complex 1. Substrates containing a primary alcohol and a
carbonyl function gave silyl ethers selectively and did not
affect the ketone moiety (Table 4, entry 1). When using an
olefin-substituted alcohol, such as allylalcohol, the selectivi-
ty was slightly lower and additional hydrosilylation of the
C=C double bond was observed, affording C,O-disilylated
propanol as a minor side product (entry 2). In contrast,
propargylic alcohol underwent selective hydrosilylation,
while silyl ether formation was completely suppressed
(entry 3). Benzoic acid was only poorly silylated and silox-
ane and silanol formation was predominant, indicating that
carboxylic acids are substantially less suitable than H2O for
silane oxidation using complex 1 (entry 4). Accordingly,
complex 1 represents a rare[29] catalyst precursor that allows
alcohols to be selectively silylated in the presence of carbon-
yl groups and C=C double bonds.[30] With functionalised al-
Chem. Eur. J. 2012, 18, 652 – 658
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
655

A. Krꢀger and M. Albrecht
Table 4. Silylation of functionalised substrates.[a]
mode) showed oligomeric fractions containing up to nine Si-
MePhO fragments. While larger oligomers and polymers
may not be sufficiently volatile to be detected by this tech-
nique, these results unambiguously confirm the potential of
Substrate
Silylation
product
t
Yield
[%]
Silanol,
siloxane[%]
[h]
1
2
48
57
20, 13
23, 18
1
À
complex 1 to catalyse multiple Si O bond formation at the
2
3
1
75+12[b]
<1, 7
same structural unit. These complexes may thus find appli-
cation in silicone synthesis, in particular, when acid- and
base-free conditions are sought. The process developed here
avoids the build-up of strong acids or the use of a base, as
typically encountered when employing more common start-
ing materials, such as R2SiCl2 or the corresponding triflates.
Moreover, polymerisation of dihydrosilanes with water is
highly atom economic and produces H2 as a neutral, gaseous
and potentially useful side product.
1
100
<1, <1
1
3
8
13
41, 34
25, 57
4
[a] General conditions: Me2PhSiH (1.0 mmol), substrate (1.0 mmol) and
1 (1 mol%) in MeNO2 (5 mL) at 1038C; conversion and selectivity deter-
mined by GC-MS. [b] Minor product from subsequent olefin hydrosilya-
À
À
tion (PhMe2Si C3H6 OSiMe2Ph).
kynes, however, exclusive hydrosilylation without affecting
the hydroxy functionality was noted.[31]
Conclusion
À
Polymerisations: Expansion of the silane substrate to silanes
Rhodium
(III)–dicarbene complexes activate Si H bonds
À
with two Si H bonds provided access to polysiloxanes and
and catalyse the hydrolytic oxidation of hydrosilanes to
form silanols and siloxanes, with H2 as the only side product.
The product selctivity was adjusted by changing the reaction
conditions and can be directed to either exclusive silanol
formation or to selective siloxane production. Mechanistic
investigations suggest that abnormally C4-bound carbenes
induce a slightly higher catalytic activity than the normal
C2-bound carbenes, and that these abnormal carbenes are
involved in what may be best described as a ligand-assisted
silicones through mild hydrolytic oxidation. Upon using 1,4-
bis(dimethylhydrosilyl)benzene as substrate, polymerisation
occurred readily, indicated by the appearance of an off-
white solid (31%) and a colourless oil (14%, yields based
on mass balance, Scheme 3). While low solubility properties
À
heterolytic Si H/D bond-cleavage process. The catalytic
system has been successfully applied to the silylation of vari-
ous alcohols, thus providing a mild alternative for the pro-
tection of alcohols by silyl groups. Selectivity tests towards
different functional groups revealed a low reactivity to
olefin hydrosilylation and no reactivity towards ketones,
thus enabling silyl ether formation with functionalised alco-
hols. The innocence of the formed H2 as side product and
the insensitivity of the silane reagent towards moisture and
air may provide great advantages in organic syntheses. In
addition, siloxane formation has been expanded to polymer-
isations, thus disclosing a mild method for the production of
polysiloxanes and silicones. The tolerance of complex 1 to-
wards a variety of functional groups may become particular-
ly relevant when preparing custom-tailored silicones, as the
use of pre-functionalised dihydrosilane monomers may be
more attractive than post-functionalisation of polymeric ma-
terials.
Scheme 3. Catalytic formation of polysiloxane (top) and silicone
(bottom) from hydrosilane precursors.
prevented any analysis of the solid fraction in solution, IR
spectroscopy indicated the expected Si–O–Si stretching fre-
quency at n˜ =1072 cmÀ1 and also bands around n˜ =825 cmÀ1,
[32]
À
which were assigned to terminal Si OH units. NMR anal-
ysis of the oil showed multiple resonances around d=7.53
and 0.34 ppm, attributed to the phenyl and methyl protons,
respectively, and also a small septet owing to the silyl-bound
proton (d=4.43 ppm), pointing towards oligomeric rather
than polymeric siloxane fractions.
Complex 1 was also used to produce silicones under com-
parably mild conditions.[3,33] When using MePhSiH2 as a sub-
strate (Scheme 3), an oily fraction was recovered in essen-
tially quantitative yield. Analysis by NMR and IR spectros-
copy revealed similar features as those described above, that
is, multiple resonances in the aromatic (d=7.77–7.19 ppm)
Experimental Section
General comments: The complexes 1–6 were prepared according to pub-
lished procedures.[11,15] All other reagents were commercially available
and were used as received. MeNO2 (puriss., absolute, over molecular
sieves, ꢀ98.5%), MeCN (HPLC grade, ꢀ99.9%) and THF (absolute,
over molecular sieves, ꢀ99.5%) were purchased from Sigma–Aldrich
and used without further purification. All NMR spectra were recorded at
258C on Varian spectrometers operating at 400 MHz (1H NMR) and
100 MHz (13C{1H} NMR). Chemical shifts (d in ppm, coupling constants
1
and aliphatic (d=0.59–0.03 ppm) H NMR region, and Si–
O–Si stretching frequency in the n˜ =1099–1016 cmÀ1 range.
Moreover, the appearance of weak IR bands around n˜ =
2171–2127 cmÀ1 suggests the presence of terminal Si H
À
groups.[32] Most instructively, ESI-MS analysis (positive
656
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 652 – 658

À
Si H Bond Activation with Rhodium Carbene Complexes
FULL PAPER
J in Hz) were referenced to residual solvent resonances. Assignments are
based on homo- and heteronuclear shift correlation spectroscopy. GC-
MS and HRMS were performed by the Analytical Laboratory at Univer-
sity College Dublin, Ireland.
Feldman, T. D. Tilley, Organometallics 2004, 23, 2488–2502; c) P.
Sangtrirutnugul, T. D. Tilley, Organometallics 2007, 26, 5557–5568.
[9] C. S. Wei, C. A. Jimꢂnez-Hoyos, M. F. Videa, J. F. Hartwig, M. B.
Hall, J. Am. Chem. Soc. 2010, 132, 3078–3091.
[10] a) F. L. Taw, R. G. Bergman, M. Brookhart, Organometallics 2004,
23, 886–890; b) J. Campos, A. C. Esqueda, J. Lꢃpez-Serrano, L.
Sꢄnchez, F. P. Cossio, A. de Cozar, E. ꢅlvarez, C. Maya, E. Carmo-
na, J. Am. Chem. Soc. 2010, 132, 16765–16767.
Procedures for the formation of siloxanes, silanols and silyl ethers: In a
typical reaction, MePhSiH2 (136 mg, 1.0 mmol) and mesitylene (120 mg,
1.0 mmol) as internal standard were dissolved in MeNO2 (2 mL) and
added to a solution of complex 1 (7.3 mg, 1.0 mol%) in MeNO2 (3 mL).
[11] A. Krꢀger, A. Neels, M. Albrecht, Chem. Commun. 2010, 46, 315–
317.
[12] a) Confirmed by comparison of GC-MS and NMR analyses with the
values published in references [6a] and [12b]; b) K. Rꢀhlmann, H.
Grosse-Ruyken, D. Scheller, U. Scheim, J. Organomet. Chem. 1988,
356, 39–47.
[13] a) The oxygen source required for siloxane generation from hydrosi-
lanes using the rhodium dicarbene complexes most presumably is re-
sidual H2O, which was identified by Karl Fischer titrations. Addition
of extraneous H2O accelerates the rate of silane oxidation substan-
tially, however, this change also switches the product selectivity
from siloxanes to silanols (cf. Table 1). No support was obtained for
the active participation of O2, as suggested earlier in reference
[13b], to promote the rate of silane oxidation and catalyst regenera-
tion. No significant variation of reaction rates was observed when
carrying out the reaction either under an inert atmosphere of N2 or
under an atmosphere of pure O2 (1 bar), and reactions were thus
The mixture was immediately immersed into
a preheated oil bath
(1108C). Aliquots were taken at regular intervals, treated with Et2O, fil-
tered over Celite and subsequently analysed by GC-MS. The product
identity was confirmed by the pertinent mass spectrometric and 1H and
13C NMR spectroscopic data. For the formation of silyl ethers and sila-
nols, ROH and H2O, respectively, were added to the solution of the
silane and mesitylene before addition to the complex. Aliquots were ana-
lysed by NMR spectroscopy after careful solvent evaporation.
Typical procedure for polymerisation experiments: Complex 1 (7.3 mg, 1
mol%) was dissolved in MeNO2 (3 mL). MePhSiH2 (122 mg, 1.0 mmol)
or 1,4-bis(dimethylsilyl)benzene (194 mg, 1.0 mmol) was dissolved in
MeNO2 (3 mL) and added to the reaction mixture, which was immediate-
ly placed in a preheated oil bath (1108C) and refluxed for 17 h. The re-
sulting suspension was filtered, thus giving an off-white solid (<5 mg,
from MePhSiH2; 65 mg, 31%, from 1,4-bis(dimethylsilyl)benzene), which
was analysed by IR spectroscopy. The filtrate was added to Et2O, filtered
and all volatiles were removed in vacuo. The residual oil (131 mg, 96%,
from MePhSiH2; 30 mg, 14%, from 1,4-bis(dimethylsilyl)benzene) was
analysed by MS and NMR spectroscopy.
À
typically run under aerobic conditions. Further confirmation of O
H bond activation to generate siloxanes was obtained from reactions
using ROH instead of H2O as nucleophile, see below; b) K. Mori,
M. Tano, T. Mizugaki, K. Ebitani, K. Kaneda, New J. Chem. 2002,
26, 1536–1538.
[14] A gradual increase of the siloxane/silanol ratio was observed over
time. For example, when the product mixture obtained from wet
MeNO2 (entry 5) was kept at reflux for 1h after reaching full con-
version, the siloxane/silanol ratio increased from initially 10:90 to
15:85 (Figure 2). This instability of the product mixture points to-
wards a secondary process involving silanol to siloxane transforma-
tion. Reference experiments confirmed that in refluxing MeNO2 si-
lanol undergoes condensation to form disiloxane (60% conversion
after 2 h). The presence of 1 accelerates this process with 80% of si-
lanol converted to siloxane after 2 h. Reactions aimed at producing
silanols in high selectivity should therefore be stopped after reach-
ing full conversion, as longer reaction times inherently derogate se-
lectivity.
Acknowledgements
We thank A. Coburn and K. Conboy for analytical assistance and the
Swiss National Science Foundation, the Alfred Werner Foundation, and
Sasol Ltd for financial support of this work.
[1] a) ACS Symposium Series, Vol. 838, Synthesis and Properties of Sili-
cones and Silicone-Modified Materials (Eds.: S. J. Clarson, J. J. Fitz-
gerald, M. J. Owen, S. D. Smith, M. E. van Dyke), Oxford University
Press, Washington DC, 2003; b) E. Pouget, J. Tonnar, P. Lucas, P. La-
croix-Desmazes, F. Ganachaud, B. Boutevin, Chem. Rev. 2010, 110,
1233–1277.
[2] For silanols in cross coupling, see: a) S. E. Denmark, C. S. Regens,
Acc. Chem. Res. 2008, 41, 1486–1499; b) M. N. Missaghi, J. M. Gal-
loway, H. H. Kung, Organometallics 2010, 29, 3769–3779.
[3] a) M. A. Brook, Silicon in Organic, Organometallic, and Polymer
Synthesis, Wiley, New York, 2000; b) V. Chandrasekhar, R. Boo-
mishankar, S. Nagendran, Chem. Rev. 2004, 104, 5847–5910.
[4] a) S. Chang, E. Sharrer, M. Brookhart, J. Mol. Catal. A: Chem.
1998, 130, 107–119; b) A. Biffis, M, Basato, M. Brichese, L. Ron-
coni, C. Tubaro, A. Zanella, C. Graiff, A. Tiripicchio, Adv. Synth.
Catal. 2007, 349, 2485–2492; c) D. E. Barber, Z. Lu, T. Richardson,
R. H. Crabtree, Inorg. Chem. 1992, 31, 4709–4711; d) L. D. Field,
B. A. Messerle, M. Rehr, L. P. Soler, T. W. Hambley, Organometal-
lics 2003, 22, 2387–2395.
[15] a) L. Yang, A. Krꢀger, A. Neels, M. Albrecht, Organometallics 2008,
27, 3161–3171; b) A. Krꢀger, L. J. L. Hꢆller, H. Mꢀller-Bunz, O.
Serada, A. Neels, S. A. Macgregor, M. Albrecht, Dalton Trans. 2011,
40, 9911–9920.
À
À
[16] For examples of related Si H bond addition across a M L bond,
see: P. J. Tiong, A. Nova, E. Clot, P. Mountford, Chem. Commun.
2011, 47, 3147–3149.
[17] a) D. R. Anton, R. H. Crabtree, Organometallics 1983, 2, 855–859;
b) G. M. Whitesides, M. Hackett, R. L. Brainard, J. P. P. M. Laval-
leye, A. F. Sowinski, A. N. Izumi, S. S. Moore, D. W. Brown, E. M.
Staudt, Organometallics 1985, 4, 1819–1830.
[18] a) J. A. Widegren, R. G. Finke, J. Mol. Catal. A: Chem. 2003, 198,
317–341; b) C. M. Hagen, J. A. Widegren, P. M. Maitlis, R. G. Finke,
J. Am. Chem. Soc. 2005, 127, 4423–4432.
[19] a) K. A. M. Ampt, S. B. Duckett, R. N. Perutz, Dalton Trans. 2004,
3331–3337; b) M. V. Cꢇmpian, J. L. Harris, N. Jasim, R. N. Perutz,
T. B. Marder, A. C. Whitwood, Organometallics 2006, 25, 5093–
5104; c) R. N. Perutz, S. Sabo-Etienne, Angew. Chem. 2007, 119,
2630–2645; Angew. Chem. Int. Ed. 2007, 46, 2578–2592.
[5] a) E. A. Ison, R. A. Corbin, M. M. Abu-Omar, J. Am. Chem. Soc.
2005, 127, 11938–11939; b) R. A. Corbin, E. A. Ison, M. M. Abu-
Omar, Dalton Trans. 2009, 2850–2855; c) Y. Lee, D. Seomoon, S.
Kim, H. Han, S. Chang, P. H. Lee, J. Org. Chem. 2004, 69, 1741–
1743.
[20] a) S. R. Klei, T. D. Tilley, R. G. Bergman, J. Am. Chem. Soc. 2000,
122, 1816–1817; b) N. Schneider, M. Finger, C. Haferkemper, S. Bel-
lemin-Laponnaz, P. Hofmann, L. H. Gade, Chem. Eur. J. 2009, 15,
11515–11529.
[21] a) A. R. Chianese, A. Kovacevic, B. M. Zeglis, J. W. Faller, R. H.
Crabtree, Organometallics 2004, 23, 2461–2468; b) M. Heckenroth,
A. Neels, M. G. Garnier, P. Aebi, A. W. Ehlers, M. Albrecht, Chem.
[6] a) T. Mitsudome, S. Arita, H. Mori, T. Mizugaki, K. Jitsukawa, K.
Kaneda, Angew. Chem. 2008, 120, 8056–8058; Angew. Chem. Int.
Ed. 2008, 47, 7938–7940; b) B. P. S. Chauhan, A. Sarkar, M. Chau-
han, A. Roka, Appl. Organomet. Chem. 2009, 23, 385–390.
[7] E. Lukevics, M. Dzintara, J. Organomet. Chem. 1985, 295, 265–315.
[8] For seminal mechanistic work, see: a) S. R. Klei, T. D. Tilley, R. G.
Bergman, Organometallics 2002, 21, 3376–3387; b) L. Turculet, J. D.
Chem. Eur. J. 2012, 18, 652 – 658
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
657

A. Krꢀger and M. Albrecht
Eur. J. 2009, 15, 9375–9386; c) D. G. Gusev, Organometallics 2009,
28, 6458–6461; d) E. Aldeco-Perez, A. J. Rosenthal, B. Donnadieu,
P. Parameswaran, G. Frenking, G. Bertrand, Science 2009, 326, 556–
559.
Curiosities to Efficient Synthetic Tools (Ed.: S. Gonzalez), Royal So-
ciety of Chemistry, Cambridge, 2011, pp. 366–398; for a recent con-
tribution with rhodium(I) complexes with bidentate dicarbene li-
gands, which demonstrate orthogonal selectivity, see: c) S. K. U. Rie-
derer, P. Gigler, M. P. Hçgerl, E. Herdtweck, B. Bechlars, W. A.
Herrmann, F. E. Kꢀhn, Organometallics 2010, 29, 5681–5692.
[28] M. P. Doyle, K. G. High, V. Bagheri, R. J. Pieters, P. J. Lewis, M. M.
Pearson, J. Org. Chem. 1990, 55, 6082–6086.
[22] a) L. Bernet, R. Lalrempuia, W. Ghattas, H. Mueller-Bunz, L.
Vigara, A Llobet, M. Albrecht, Chem. Commun. 2011, 47, 8058–
8060; b) A. Krꢀger, M. Albrecht, Aust. J. Chem. 2011, 64, 1113–
1117.
[23] For related metal-mediated formation of silyl ethers from hydrosi-
lanes, see: a) E. Lukevics, M. Dzintara, J. Organomet. Chem. 1984,
271, 307–317; b) L. Horner, J. Mathias, J. Organomet. Chem. 1985,
282, 175–189; c) K. Yamamoto, M. Takemae, Bull. Acad. Vet. Fr.
Bull. Chem- Soc. Jpn. 1989, 62, 2111–2113.
[24] a) G. A. Molander, S. Swallow, J. Org. Chem. 1994, 59, 7148–7151;
b) J. M. Lassaletta, R. R. Schmidt, Synlett 1995, 925–927.
[25] Silyl ethers are also accessible through hydrosilylation of ketones
and aldehydes, see: a) Coomprehensive Handbook on Hydrosilyla-
tion (Ed.: B. Marciniec), Elsevier, Oxford, 1992; b) H. Nishiyama in
Transition Metals for Organic Synthesis, Vol 2, 2nd ed. (Eds.: M.
Beller, C. Bolm), Wiley-VCH, Weinheim, 2004, pp. 182–191; c) S.
Dꢈez-Gonzꢄlez, S. P. Nolan, Org. Prep. Proc. Int. 2007, 39, 523–559.
[26] T. W. Greene, P. G. M. Wuts, Protective Groups in Organic Synthesis,
3rd ed., Wiley, New York, 1999.
[29] a) X.-L. Luo, R. H. Crabtree, J. Am. Chem. Soc. 1989, 111, 2527–
2535; b) B. T. Gregg, A. R. Cutler, Organometallics 1994, 13, 1039–
1043; c) U. Oehmichen, H. Singer, J. Organomet. Chem. 1983, 243,
199–204.
[30] For a heterogeneous system, see: P. Raffa, C. Evangelisti, G. Vitulli,
P. Salvadori, Tetrahedron Lett. 2008, 49, 3221–3224.
[31] For a complementary system that does not affect triple bonds, see:
C. Lorenz, U. Schubert, Chem. Ber. 1995, 128, 1267–1269.
[32] D. Lin-Vien, N. B. Colthup, W. G. Fateley, J. G. Grasselli, The Hand-
book of Infrared and Raman Characteristic Frequencies of Organic
Molecules, Academic Press, San Diego, 1991.
[33] J. B. Grande, F. Gonzaga, M. A. Brook, Dalton Trans. 2010, 39,
9369–9378.
[27] a) S. Dꢈez-Gonzꢄlez, N. Marion, S. P. Nolan, Chem. Rev. 2009, 109,
3612; b) B. Cetinkaya in N-Heterocyclic Carbenes: From Laboratory
Received: July 18, 2011
Published online: December 8, 2011
658
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 652 – 658