CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201200721
Mechanistic Insights into the Reduction of Carbon Dioxide
with Silanes over N-Heterocyclic Carbene Catalysts
Siti Nurhanna Riduan, Jackie Y. Ying,* and Yugen Zhang*[a]
A metal-free reduction of carbon dioxide to methanol with N-
heterocyclic carbenes (NHCs) as organocatalysts was achieved,
which was conducted under ambient conditions with hydrosi-
lane as the reducing reagent and the hydrogen source. Experi-
mental results and density functional theory calculations indi-
cated that the overall process involved a three-step cascade re-
action. All three steps in the cascade reaction were found to
Introduction
The rise in CO2 concentration in the atmosphere has prompted
considerable research into CO2 fixation and utilization. Technol-
ogies involving the use of CO2 as a C1 feedstock have seen
major advancement in recent years.[1] Several notable recent
examples include the conversion of CO2 to fuels by chemical[2]
or photochemical[3] means, and CO2 as a comonomer in the
synthesis of polycarbonates[4] and as a starting material for the
synthesis of motifs commonly found in pharmaceuticals.[5]
Since our report of a successful reduction of CO2 to methanol
through an organocatalyzed hydrosilylation process,[2d]
a number of publications have ensued. The use of frustrated
Lewis pairs, first mooted by the research groups of Stephan
and Erker, has been applied to activate H2 heterolytically,[6] and
more recently, frustrated Lewis pairs were shown to reversibly
uptake and release CO2.[7] A recent report on the room-temper-
ature reduction of CO2 to methanol with a frustrated Lewis
pair of aluminum and phosphines with ammonia borane as
a hydride donor yielded up to 51 % methanol, as determined
by NMR spectroscopy.[8] The same reduction with four equiva-
lents of tetramethylpiperidine–boron-based frustrated Lewis
pairs in an atmosphere of H2 resulted in quantitative conver-
sion with methanol isolated in 25 % yield.[9] The same tetrame-
thylpiperidine–boron-based frustrated Lewis pair system was
used in combination with triethylsilane as a reductant to yield
methane.[10] The reduction of CO2 catalyzed by transition-metal
complexes is still widely studied, with a recent report on an ef-
ficient nickel catalyst for the reduction of CO2, with boron as
a sacrificial reductant.[11] These recent reports illustrate that the
[a] S. N. Riduan, Prof. J. Y. Ying, Dr. Y. Zhang
Institute of Bioengineering and Nanotechnology
31 Biopolis Way, The Nanos, 138669 (Singapore)
Fax: (+ 65) 6478 9080
E-mail: jyying@ibn.a-star.edu.sg
ygzhang@ibn.a-star.edu.sg
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201200721.
 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
be exothermic and were catalyzed by NHCs. The rate-deter-
mining step for the process was the first hydrosilylation step,
which produced the formoxysilane intermediate, with an acti-
vation energy barrier of 19.6 kcal mol 1. This observation ex-
plained the high tendency of the system to form the methox-
ide end product, with a hydrogen-transfer yield in excess of
95 %.
catalytic reduction of CO2 to methanol remains a research area
of considerable interest, as it may be a viable alternative for
the industrial synthesis of methanol from CO2. Although such
reactions are still in the research stage, it is important to un-
derstand their underlying mechanisms, so as to bring about
major advances in this field.[12]
Recently, Wang’s group[13] reported a detailed computational
mechanistic study on the N-heterocyclic carbene (NHC) cata-
lyzed CO2 reduction to methanol with hydrosilane that we re-
ported in 2009.[2d] However, we have found that their calculat-
ed mechanism was partially inconsistent with our experimental
results. Herein, we report new insights into the hydrosilylation
of CO2 with NHCs by combining experiments with DFT
calculations.
Results and Discussion
Elucidation of the reaction mechanism
It was earlier communicated that the reduction of CO2 to
a silyl methoxide product proceeded under ambient condi-
tions, without the need for external CO2 overpressure, or ele-
vated temperatures.[2d] This is advantageous as there is negligi-
ble energy input necessary for the process to occur. Our pro-
posed mechanism included the presence of two reaction inter-
mediates, which were observed with mass spectrometry (MS)
and NMR spectroscopy (Scheme 1). We performed NMR spec-
troscopic studies with equivalent amounts of imidazolium car-
boxylate and silane. The transformation from the imidazolium
carboxylate to the formoxysilane intermediate could be clearly
observed. The addition of one equivalent of diphenylsilane to
the reaction ensured the complete conversion of the formoxy-
silane to a bis(silylacetal) intermediate (see the Supporting In-
formation, Figure S1).
ChemCatChem 2013, 5, 1490 – 1496 1490
CHEMCATCHEM
FULL PAPERS
Scheme 1. Overall reaction scheme for the hydrosilylation of CO2.
DFT calculations of each step
With the experimental results at hand, we proceeded to per-
form DFT calculations to further understand and elucidate the
reaction mechanism. By using diphenylsilane as the silane
starting material, it was found that the reaction proceeds in
a three-step cascade process, with the energy level of each in-
termediate step lower than that of the previous one (Figure 1).
The complete conversion of a diphenylsilane (Ph2SiH2) mole-
cule to a methoxide end product (Ph2HSiOSi(OCH3)Ph2), with
formoxysilane (Ph2SiH(OCHO)) and bis(silylacetal)
(Ph2HSiOCH2OSiHPh2) as intermediates is a highly exothermic
process with a DE value of 79.3 kcal mol 1. These calculations
are in agreement with our experimentally observed results.
Similarly, the three-step cascade reaction was also calculated in
a recent report by Wang et al.[13]
The energy diagrams for each of the steps are shown in
Figure 2. The first step depicted in Figure 2 A is particularly in-
teresting, as it involves the approach of the silane to the imida-
zolium carboxylate (Imes–CO2), forming a five-membered-ring
transition state, and the subsequent hydride transfer to form
the formoxysilane intermediate (1). It is noted that the energy
barrier for this step is relatively large, and the transition state
of the hydride transfer step (TS-1, see Figure S2 in the Support-
ing Information for calculated structures) lies 19.6 kcal mol 1
above Imes–CO2+Ph2SiH2. However, this value is far lower than
that obtained in the calculation by Wang’s group (+ 41.3 kcal
Figure 1. Energy diagram of the CO2 hydrosilylation reaction calculated by use of DFT. The values in black were calculated by us, whereas the values in gray
were calculated by Wang et al.[13]
 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
mol 1) with a four-membered-ring transition state.[13] We pro-
pose that the NHC first activates CO2 to form Imes–CO2. This
initial step was also observed experimentally. In addition, this
step also represents the rate-limiting step in the entire three-
step cascade reaction, as its energy barrier is the highest in
comparison with the rest of the transition states in the subse-
quent steps illustrated in Figure 2 B and C. In contrast, the
second step is energetically undemanding, with a small energy
difference of 9.1 kcal mol 1 between 1·Imes and TS-2. The
adduct 1·Imes includes the coordination of a formoxysilane in-
termediate to the NHC catalyst, whereas TS-2 involves the ap-
proach of a Ph2SiH2 molecule to the adduct, and the subse-
quent hydride transfer from the hydrosilane. The third reaction
step has a small energy barrier (9 kcal mol 1) in the elementary
step for the activation of 2 by NHC, forming transition state
TS-3. The formation of a more stable Si O Si bond structure is
brought about by rearrangement of the Si O bonds in the
transition state. The ensuing intermolecular hydride transfer
then forms the silyl methoxide end product 3, closing the reac-
tion cycle and releasing the NHC catalyst. The calculated struc-
tures of the identified stationary points TS-1 to TS-3 are repre-
sented in Figure S2 in the Supporting Information and were
simplified to a reaction pathway involving only one Si H
bond.
Rate-determining step
The energy barriers of the three-step reaction reported by
Wang’s group are 30.8, 40.7, and 29.8 kcal mol 1, respectively
(Figure 2).[13] In this case, the rate-limiting step would be the
second step in the three-step cascade reaction, contrary to our
proposed mechanism and experimental results. Our experi-
mental observations strongly support the first hydrosilylation
step as the rate-limiting step, that is, with the largest energy
barrier. In this step, the hydrosilane is converted to a formoxysi-
lane in a CO2 atmosphere over the NHC catalyst. The reaction
tends to form the silyl methoxide end product or the bis(silyla-
cetal) intermediate, even in the presence of excess CO2. Simi-
larly, the rate-determining step of hydride transfer from
ChemCatChem 2013, 5, 1490 – 1496 1491
CHEMCATCHEM
FULL PAPERS
Figure 2. A–C) Calculated energy diagrams and related transition states of each reaction
in the three-step cascade reaction, respectively. Transition states TS-1 a, TS-2 a and TS-3 a
were cited from the paper by Wang et al.[13]
a boron reductant was also reported for the tetramethylpiperi-
dine–boron-based frustrated Lewis pair system and other sys-
tems.[3, 9] The formoxysilane intermediate would react rapidly
with excess silane in the reaction mixture and was only ob-
served in the early stages of the reaction during our GC–MS
monitoring studies.[2d] This result was further confirmed by our
NMR spectroscopic studies (Figure 3). Isotopically enriched
13
CO2 (99 atom % 13C) with diphenylsilane and triethylsilane
were employed to further investigate the intervening process-
es of the reactions. The results of 13C NMR spectroscopy indi-
cated that the reaction with diphenylsilane was completed
within hours to form silylmethoxide as the major product (Fig-
ure 3 A and B), whereas the signals related to the formoxysi-
lane and bis(silylacetal) intermediates remained minor. We also
observed that the activation energy barrier to form the for-
moxysilane was even more pronounced for trisubstituted si-
lanes (Figure 3 C and D). Trace amounts of formoxysilane inter-
mediates of triethylsilane, Et3Si(OCHO) (168 ppm),[15, 16] were
only observed after 12 days in a solution of the NHC catalyst
under an atmosphere of 13CO2.
 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
Yield of methanol
The tendency of the system towards completion of
the reaction to the silyl methoxide end product was
also observed with regards to the yield of methanol.
The yield of methanol obtained after hydrolysis
would be dependent on the amount of the silyl
methoxide end product in the reaction mixture, and
it was found that a higher yield of methanol from
the reaction was obtained if a limited CO2 feedstock
was used. Yields of up to 90 % were achieved experi-
mentally in 6 h, if an excess of CO2 was introduced
into the reaction system together with a fixed
amount of Ph2SiH2. However, higher yields of more
than 95 % were achieved if a controlled amount of
CO2 (1 equiv.) was introduced (Scheme 2). These ob-
servations support our position that the formoxysi-
lane formation is the rate-limiting step. With a con-
trolled amount of CO2, the reaction would be atom-
efficient, whereas exactly three equivalents of hydride
would be required to completely convert CO2 into
the final methoxide end product, and this was re-
flected in the detected yield of methanol, > 95 %.
With an excess of CO2 introduced with a balloon,
there would be an increased probability of the hy-
drosilane reacting in the initial stages of the reaction,
resulting in a smaller amount of hydride available for
the complete reduction to silyl methoxide. However,
the yield of methanol was not significantly reduced
(from 95 % to 90 %), suggesting that the initial forma-
tion of formoxysilane has a significant energy barrier,
leaving behind considerable amounts of silane to
react with the formoxysilane and bis(silylacetyl) inter-
mediates to form the silylmethoxide end product.
The high selectivity towards the methoxysilane end
Scheme 2. Yield of methanol under different conditions.
product and the corresponding high yield of methanol were
a consequence of the three-step cascade reaction in which the
formoxysilane formation step was the rate-determining step.
Effect of hydrosilanes
Our experimental studies have, thus far, utilized diphenylsilane
as a starting material. It was demonstrated that the hydrosilyla-
tion of CO2 with diphenylsilane occurred smoothly, and com-
plete consumption of the silane was achieved in 6 h.[2d] Differ-
ent silanes were also screened, and it was observed that trisub-
stituted silanes reacted sluggishly or were not at all reactive.
ChemCatChem 2013, 5, 1490 – 1496 1492
CHEMCATCHEM
FULL PAPERS www.chemcatchem.org

To further elucidate the reac-

tion mechanism and investigate
the effect of using sterically en-
cumbered silanes, a mixture of
diphenylsilane and phenyldime-
thylsilane was used. A typical
CO2 hydrosilylation reaction was
set up by using phenyldimethyl-
silane as the hydride source
(Scheme 3). Intermediates 5–7
were not observed in GC–MS
analysis; only the starting mate-
rials were found. Whereas the
first step of the hydrosilylation
reaction was energetically unfav-
orable for phenyldimethylsilane,
we believed that this compound
could participate in the subse-
quent steps of the reaction. If
Figure 3. 13C NMR spectra of the NMR tube reaction of 13CO2, silane (diphenylsilane in parts A and B, triethylsilane a mixture of diphenylsilane
in parts C and D), Imes–CO2 catalyst (5 mol %) in [D7]DMF. Spectra A and B show the conversion of 13CO2 (#) to (Ph2SiH2) and phenyldimethylsi-
13
CH2(OSiR3)2 (!) to 13CH3O SiR3 (heart) after 90 min and 24 h, respectively. Spectrum C proves that no reaction
lane (PhMe2SiH; 1:1 molar ratio)
occurred after 60 min. In spectrum D a very weak peak at 168 ppm is seen, indicating that a small amount of the
triethylformoxysilane (†) intermediate was generated after 12 d. was used as the silane feedstock,
diphenylsilane was consumed in

This is illustrated in Figure 3 D, as formoxysilane intermediates

for triethylsilane Et3Si(OCHO) were only observed with NMR
spectroscopy after 12 days. However, the calculated energy
profiles for all silanes used (Table 1) indicated that the reac-

Table 1. Hydrosilylation of CO2 with various silanes over Imes–CO2 1, and

their total energy differences.[a]

Entry Silane Si OCHO DE1 [kcal mol 1] Si OMe DE2 [kcal mol 1]
time [h][b] time [h][c]
1 PhSiH3 <1 12.06 <6 69.98
2 Ph2SiH2 1.5 21.13 6 79.34
3 Ph2MeSiH –[d] 33.58 – 79.18
4 PhMe2SiH – 33.64 – 103.8
5 Et2MeSiH – 29.35 – 83.10
6 Et3SiH – 17.43 – 76.30
7 Ph3SiH – 18.83 – 79.48

[a] Reaction conditions: silane (1 mmol), CO2 balloon, DMF (2 mL), room
temperature. [b] Time for the full consumption of silanes. [c] Time for re-
action completion. [d] Silane was not fully consumed after 3 d.

Figure 4. Calculated energy diagram and transition states of the reaction

tions were thermodynamically favored, which led us to con-
clude that the low activities of trisubstituted silanes were
mainly attributable to steric effects. The energy level of TS-
1 with bulky, trisubstituted silanes was significantly higher
than that with diphenylsilane. For example, the calculated acti-
vation energy barrier for the first-step hydrosilylation with phe-
nyldimethylsilane was 45.2 kcal mol 1 (Figure 4 A), which was
more than twice that of Ph2SiH2 (19.6 kcal mol 1; see Fig-
ure 2 A).
with dimethylphenylsilane.
1 h, whereas most of the phenyldimethylsilane remained in
the reaction mixture. However, a significant amount of phenyl-
dimethylsilane formed intermediate 5, and final products 6–9
were observed in the GC–MS analysis of the reaction mixture.
Similar observations were made if Ph2SiH2 was added first and
left to react for 30 min before the addition of PhMe2SiH (see

 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim ChemCatChem 2013, 5, 1490 – 1496 1493
CHEMCATCHEM
FULL PAPERS
Scheme 3. Products from the reactions with a mixed silane feedstock of Ph2SiH2 and
PhMe2SiH.
Figures S4 and S5 in the Supporting Information). If PhMe2SiH
was added first, followed by the addition of Ph2SiH2 after
30 min, only products 7, 8, and 9 were observed (i.e., 5 and 6
were not found). These experimental results further confirmed
the calculated reaction mechanism, in which the rate-deter-
mining step was the formation of the formoxysilane. Hydrosily-
lation of CO2 to formoxysilane with NHCs did not proceed with
bulky trisubstituted silanes, as exemplified by the reaction with
phenyldimethylsilane (DE = + 45.2 kcal mol 1). However, after
formoxysilane 1 was formed in the reaction system with less
sterically hindered silanes, such as diphenylsilane, it could
react with the bulky phenyldimethylsilane in the second and
third step of the hydrosilylation reaction to yield silylacetal (5)
and silylmethoxide (6) products. This reaction pathway was cal-
culated to be favorable, and the energy difference between
the approach of phenyldimethylsilane to 1·Imes and the subse-
quent hydride transfer (TS-2’) was minimal, yielding the silyla-
cetal 5 (Figure 4 B). The exchange of silane moieties after for-
mation of the disiloxane was also considered in elucidating
our reaction pathway, and silane substitutions with a disiloxane
were not observed in our control reactions of mixing
(Ph2SiH)2O and PhMe2SiH. These experimental observations,
supplemented by DFT calculations, further reinforced our pro-
posed three-step cascade reaction mechanism.
Formation of an NHC–Si adduct
The formation of the NHC–CO2 adduct has been reported and
widely accepted as an intermediate in many reactions.[14, 17]
However, there are still arguments for a favorable interaction
between NHC and the silane, instead of CO2, as the primary ac-
tivation mode in the first elementary step of the NHC-catalyzed
CO2 hydrosilylation reaction.[13, 18, 19] Although thermodynamical-
ly unfavorable, the p–p stacking interactions between the NHC
and the silane could activate the silane by stretching the Si H
bonds, elongating them, and pushing the electron density to-
wards the hydrogen atoms. The formation of an NHC Si bond
 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
has also been proposed as the initial step for cyanosi-
lylation and siloxane polymerization reactions cata-
lyzed by NHCs.[18]
However, in our experiments, we did not observe
any interactions between the NHC and the silanes.
There were no changes in the chemical shift values
of the carbene center in the NMR spectra, even after
a solution of equimolar amounts of the Imes carbene
and diphenylsilane was allowed to stand for 2 weeks.
Interactions between silanes and the carbene center
were also not detected with a less sterically hindered
carbene, such as 1-butyl-3-methylimidazolylidene.
Such observations were also critically discussed by
Fuchter.[19]
Formaldehyde as an intermediate
It was proposed that the bis(silylacetal) intermediate
might undergo direct dissociation into formaldehyde
and disiloxanes as the third elementary step of the
reaction.[13] The dissociation was mentioned to be akin to the
Brook rearrangement, in which an organosilyl group switches
position with a hydroxyl proton to form a silyl ether in the
presence of a base.[20] The resulting formaldehyde was then re-
ported to be able to undergo hydrosilylation with an NHC, cat-
alyzing the addition of the silane to the aldehyde. This pro-
posed mechanism trigged our interest in tracking the formal-
dehyde intermediate during the reaction. However, in our NMR
monitoring experiments, no peaks associated with formalde-
hyde were observed in either 1H or 13C NMR spectra. It is also
known that the Si O bond is very strong and is unlikely to
break under our reported reaction conditions, without the use
of heat or base.[2] Control experiments were performed by
using both paraformaldehyde and formaldehyde (generated
from heating a solution of paraformaldehyde) in place of CO2,
and no intermediates were observed;[21] diphenylsilane starting
materials were recovered after 24 h. These results indicate that
this proposed pathway,[13] which includes the hydrosilylation of
formaldehyde, was not experimentally feasible.
Conclusions
The first N-heterocyclic carbene (NHC) catalyzed CO2 reduction
reaction with hydrosilanes under ambient conditions was ex-
amined in detail. By combining DFT calculations with experi-
mental investigations, several conclusions could be made. The
proposed reaction pathway consists of a three-step cascade re-
action. All three steps involve exothermic processes and are
catalyzed by the NHC catalyst. The primary activation model
involves the reaction between the NHC and CO2, forming the
NHC–CO2 adduct. The first step, which involves the hydrosilyla-
tion of CO2 to form a formoxysilane intermediate, has the high-
est activation energy barrier (DE = 19.6 kcal mol 1), and is es-
tablished as the rate-determining step. As a result, the system
is highly selective towards a methoxide end product, with in
excess of a 95 % hydrogen transfer or carbon yield. It was also
ChemCatChem 2013, 5, 1490 – 1496 1494
CHEMCATCHEM
FULL PAPERS
shown that formaldehyde is unlikely to be an intermediate in
this reaction.
Experimental Section
General information
All solvents and chemicals were used as received from commercial
suppliers, unless otherwise noted. Dry solvents and an argon-gas-
equipped glove box (Innovative Technologies, Inc.) were used for
the reaction setup. Various imidazolium salts and silanes were pur-
chased from Sigma–Aldrich. Imes–CO2 was synthesized according
to the protocol detailed in the literature.[9] CO2, O2, and purified air
were supplied by SOXAL, and 13C-enriched CO2 was purchased
from Sigma–Aldrich. GC–MS analyses were performed on a Shimad-
zu GCMS QP2010 system, and GC analyses were conducted on an
Agilent GC6890N system. 1H and 13C NMR spectra were recorded
on a Bruker AV-400 instrument (400 MHz). Elemental analyses were
performed at the Chemical, Molecular, and Materials Analysis
Centre, National University of Singapore. PA–FTIR spectra were re-
corded on a Digilab FTS 7000 FTIR spectrometer equipped with an
MTEC-300 photoacoustic detector.
DFT calculations
The DFT calculations were performed with the Gaussian 03 soft-
ware.[22] The exchange–correlation function employed was B3LYP,[23]
which included a fraction of Hartree–Fock exchange to reduce the
self-interaction error. In this study, the 6-31(G) basis sets were used.
After fully optimizing the structure of each compound, the total
energy was obtained. For each reaction, the stationary structures
were obtained by optimizing the complex structures along the in-
trinsic reaction coordinate of the reaction pathway. The calcula-
tions were performed with the reactants in a gaseous phase at RT
and ambient pressure, without any interactions with solvent
molecules.
Hydrosilylation of CO2
The imidazolium salt (0.25 mmol) and sodium hydride (0.25 mmol)
were dissolved in DMF (0.5 mL) in a crimp-top vial and stirred for
a minimum of 30 min for the carbene to be generated
(0.5 mmol mL 1). The solution was then centrifuged so that the in-
organic salts resulting from deprotonation would settle at the
bottom of the vial. An aliquot of 0.2 mL of the carbene solution
was transferred into a fresh vial, and the solvent (1.8 mL) was intro-
duced. The vial was sealed, and CO2 was introduced into the vial
through a balloon. The reaction was allowed to stir for 10 min,
after which the silane (1 mmol) was introduced. An internal stan-
dard of mesitylene was added (0.5 mmol). Aliquots of the reaction
mixture were withdrawn after specified reaction times and diluted
with methylene chloride before the GC–MS analysis.
For conversion studies, a GC calibration curve was constructed
with mesitylene and various concentrations of diphenylsilane. Ali-
quots were drawn from the reaction mixture at hourly intervals,
and diluted with methylene chloride before the GC analysis.
Hydrolysis reaction to release methanol
To produce methanol through hydrolysis of the reaction mixture,
the reaction was quenched after 18 h by adding 2 equivalents of
an NaOH/H2O solution. The reaction mixture was stirred for anoth-
 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
er 24 h before an aliquot of isopropanol was added as an internal
standard. An aliquot of 1 mL was removed from the sample and di-
luted with dichloromethane before the resulting mixture was sub-
jected to GC analysis with a Quadrex BTR-CW column (Carbowax
polyethylene glycol bonded phase).
NMR tube reaction and analysis
1,3-Bis(2,4,6-trimethylphenyl)imidazolium carboxylate (Imes–CO2)
was synthesized according the literature method,[14] and a stock so-
lution of Imes–CO2 (0.05 mmol mL 1) was prepared in [D7]DMF. An
aliquot corresponding to 0.01 equivalents of the catalyst was trans-
ferred into an NMR tube, and [D7]DMF (0.5 mL) was added. Silane
(0.1 equiv.) was subsequently added; the tube was sealed, and
then evacuated and refilled with 13CO2 with two freeze–pump–
thaw cycles. The reaction was monitored by 13C decoupled and
coupled NMR spectroscopy.
For reactions performed at stoichiometric ratios, one equivalent of
silane was added to a solution of imidazolium carboxylate in an
NMR tube with a resealable screw-top cap. Additional aliquots of
the silane were subsequently added in a glove box.
Reaction with a controlled amount of CO2
The reaction mixture was prepared according to the above proce-
dure, in which CO2 (15 mL) was introduced to the reaction mixture
with a gas-tight syringe. The reaction mixture was allowed to stir
for 10 min before diphenylsilane was introduced. After all the CO2
was consumed, an argon-filled balloon was introduced to ensure
that the system was closed. Aliquots of the reaction mixture were
withdrawn at hourly intervals and subjected to GC–MS analysis.
Acknowledgements
This work was supported by the Institute of Bioengineering and
Nanotechnology (Biomedical Research Council, Agency for Sci-
ence, Technology and Research, Singapore).
Keywords: carbenes · carbon dioxide fixation · density
functional calculations · hydrosilylation · reduction
[1] a) K. Huang, C.-L. Sun, Z.-J. Shi, Chem. Soc. Rev. 2011, 40, 2435 – 2452;
b) G. A. Olah, G. K. S. Prakash, A. Goeppert, J. Am. Chem. Soc. 2011, 133,
12881 – 12898; c) M. Cokoja, C. Bruckmeier, B. Reiger, W. A. Hermann,
F. E. Kuhn, Angew. Chem. 2011, 123, 8662 – 8690; Angew. Chem. Int. Ed.
2011, 50, 8510 – 8537; d) S. N. Riduan, Y. G. Zhang, Dalton Trans. 2010,
39, 3347 – 3357; e) T. Sakakura, J.-C. Choi, H. Yasuda, Chem. Rev. 2007,
107, 2365 – 2387; f) Y. G. Zhang, S. N. Riduan, Angew. Chem. 2011, 123,
6334 – 6336; Angew. Chem. Int. Ed. 2011, 50, 6210 – 6212.
[2] a) T. Matsuo, H. Kawaguchi, J. Am. Chem. Soc. 2006, 128, 12362 – 12363;
b) C. A. Huff, M. S. Sanford, J. Am. Chem. Soc. 2011, 133, 18122 – 18125;
c) E. Balaraman, C. Gunanathan, J. Zhang, L. J. W. Shimon, D. Milstein,
Nat. Chem. 2011, 3, 609 – 614; d) S. N. Riduan, Y. G. Zhang, J. Y. Ying,
Angew. Chem. 2009, 121, 3372 – 3375; Angew. Chem. Int. Ed. 2009, 48,
3322 – 3325; e) L. Gu, Y. G. Zhang, J. Am. Chem. Soc. 2010, 132, 914 –
915.
[3] a) E. Barton Cole, P. S. Lakkaraju, D. M. Rampulla, A. J. Morris, E. Abelev,
A. B. Bocarsly, J. Am. Chem. Soc. 2010, 132, 11539 – 11551; b) A. J. Morris,
R. T. McGibbon, A. B. Bocarsly, ChemSusChem 2011, 4, 191 – 196.
[4] For reviews, see a) D. J. Darensbourg, Chem. Rev. 2007, 107, 2388 – 2410;
b) G. W. Coates, D. R. Moore, Angew. Chem. 2004, 116, 6784 – 6806;
ChemCatChem 2013, 5, 1490 – 1496 1495
CHEMCATCHEM
FULL PAPERS
Angew. Chem. Int. Ed. 2004, 43, 6618 – 6639; c) D. J. Darensbourg, Inorg.
Chem. 2010, 49, 10765 – 10780.
[5] a) J. Takaya, S. Tadami, K. Ukai, N. Iwasawa, Org. Lett. 2008, 10, 2697 –
2700; b) T. Ohishi, M. Nishiura, Z. Hou, Angew. Chem. 2008, 120, 5876 –
5879; Angew. Chem. Int. Ed. 2008, 47, 5792 – 5795; c) K. Kobayashi, Y.
Kondo, Org. Lett. 2009, 11, 2035 – 2037; d) J. Takaya, K. Sasano, N. Iwasa-
wa, Org. Lett. 2011, 13, 1698 – 1701; e) T. Fujihara, T. Xu, K. Semba, J.
Terao, Y. Tsuji, Angew. Chem. 2011, 123, 543 – 547; Angew. Chem. Int. Ed.
2011, 50, 523 – 527; f) S. Li, W. Xuan, S. Ma, Angew. Chem. 2011, 123,
2626 – 2630; Angew. Chem. Int. Ed. 2011, 50, 2578 – 2582; g) D. Yu, Y. G.
Zhang, Proc. Natl. Acad. Sci. USA 2010, 107, 20184 – 20189; h) I. I. F. Boo-
gaerts, G. C. Fortman, M. R. L. Furst, C. S. J. Cazin, S. P. Nolan, Angew.
Chem. 2010, 122, 8856 – 8859; Angew. Chem. Int. Ed. 2010, 49, 8674 –
8677; i) L. Zhang, J. Cheng, T. Oishi, Z. Hou, Angew. Chem. 2010, 122,
8852 – 8855; Angew. Chem. Int. Ed. 2010, 49, 8670 – 8673; j) H. Mizuno, J.
Takaya, N. Iwasawa, J. Am. Chem. Soc. 2011, 133, 1251 – 1253; k) C. Das
Neves Gomes, O. Jacquet, C. Villiers, P. Thuery, M. Ephritikhine, T. Cantat,
Angew. Chem. 2012, 124, 191 – 194; Angew. Chem. Int. Ed. 2012, 51,
187 – 190; l) D. Yu, Y. G. Zhang, Green Chem. 2011, 13, 1275 – 1279; m) D.
Yu, M. X. Tan, Y. G. Zhang, Adv. Synth. Catal. 2012, 354, 969 – 974; n) S.
Nurhanna Riduan, J. Y. Ying, Y. G. Zhang, Org. Lett. 2012, 14, 1780 – 1783.
[6] a) G. C. Welch, R. R. San Juan, J. D. Masuda, D. W. Stephan, Science 2006,
42, 4793 – 4795; b) P. A. Chase, G. C. Welch, T. Jurca, D. W. Stephan,
Angew. Chem. 2007, 119, 8196 – 8199; Angew. Chem. Int. Ed. 2007, 46,
8050 – 8053; c) P. Spies, S. Schwendemann, S. Lange, G. Kehr, R. Frçhlich,
G. Erker, Angew. Chem. 2008, 120, 7654 – 7657; Angew. Chem. Int. Ed.
2008, 47, 7543 – 7546; d) P. A. Chase, D. W. Stephan, Angew. Chem.
2008, 120, 7543 – 7547; Angew. Chem. Int. Ed. 2008, 47, 7433 – 7437; e) V.
Sumerin, F. Schulz, M. Nieger, M. Leskel, T. Repo, B. Rieger, Angew.
Chem. 2008, 120, 6090 – 6092; Angew. Chem. Int. Ed. 2008, 47, 6001 –
6003.
[7] C. M. Mçmming, E. Otten, G. Kehr, R. Frçhlich, S. Grimme, D. W. Stephan,
G. Erker, Angew. Chem. 2009, 121, 6770 – 6773; Angew. Chem. Int. Ed.
2009, 48, 6643 – 6646.
[8] G. Mnard, D. W. Stephan, J. Am. Chem. Soc. 2010, 132, 1796 – 1797.
[9] A. E. Ashley, A. L. Thompson, D. O’Hare, Angew. Chem. 2009, 121,
10023 – 10027; Angew. Chem. Int. Ed. 2009, 48, 9839 – 9843.
[10] A. Berkefeld, W. E. Piers, M. Parvez, J. Am. Chem. Soc. 2010, 132, 10660 –
10661.
[11] a) S. Charraborty, J. Zhang, J. A. Krause, H. Guan, J. Am. Chem. Soc. 2010,
132, 8872 – 8873; b) F. Huang, C. Zhang, J. Zhang, Z.-X. Wang, H. Guan,
Inorg. Chem. 2011, 50, 3816 – 3825.
[12] a) O. Jacquet, C. D. N. Gomes, M. Ephritikhine, T. Cantat, J. Am. Chem.
Soc. 2012, 134, 2934 – 2937; b) C. Das Neves Gomes, O. Jacquet, C. Villi-
 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
ers, P. Thuery, M. Ephritikhine, T. Cantat, Angew. Chem. 2012, 124, 191 –
194; Angew. Chem. Int. Ed. 2012, 51, 187 – 190.
[13] F. Huang, G. Lu, L. Zhao, H. Li, Z.-X. Wang, J. Am. Chem. Soc. 2010, 132,
12388 – 12396.
[14] H. A. Duong, T. N. Tekavec, A. M. Arif, J. Louie, Chem. Commun. 2004,
112 – 113.
[15] A. Jansen, H. Gçrls, S. Pitter, Organometallics 2000, 19, 135 – 138.
[16] A. Jansen, S. Pitter, J. Mol. Catal. A: Chem. 2004, 217, 41 – 45.
[17] a) H. Zhou, W. Z. Zhang, C. H. Liu, J. P. Qu, X. B. Lu, J. Org. Chem. 2008,
73, 8039 – 8044; b) Y. Kayaki, M. Yamamoto, T. Ikariya, Angew. Chem.
2009, 121, 4258 – 4261; Angew. Chem. Int. Ed. 2009, 48, 4194 – 4197; c) H.
Zhou, Y. M. Wang, W. Z. Zhang, J. P. Qu, X. B. Lu, Green Chem. 2011, 13,
644 – 650; d) Y. G. Zhang, J. Y. G. Chan, Energy Environ. Sci. 2010, 3, 408 –
417.
[18] a) J. J. Song, F. Gallow, J. T. Reeves, Z. Tan, N. K. Yee, C. H. Senanayake, J.
Org. Chem. 2006, 71, 1273 – 1276; b) J. Raynaud, A. Ciolino, A. Baceiredo,
M. Destarac, F. Bonette, T. Kato, Y. Gnanou, D. Taton, Angew. Chem.
2008, 120, 5470 – 5473; Angew. Chem. Int. Ed. 2008, 47, 5390 – 5393; c) J.
Raynaud, Y. Gnanou, D. Taton, Macromolecules 2009, 42, 5996 – 6005.
[19] M. J. Fuchter, Chem. Eur. J. 2010, 16, 12286 – 12294.
[20] A. G. Brook, Acc. Chem. Res. 1974, 7, 77 – 84.
[21] M. X. Tan, Y. G. Zhang, J. Y. Ying, Adv. Synth. Catal. 2009, 351, 1390 –
1394.
[22] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsu-
ji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Na-
kajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Och-
terski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
[23] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648; b) C. T. Lee, W. T. Yang,
R. G. Parr, Phys. Rev. B 1988, 37, 785 – 789.
Received: October 15, 2012
Published online on March 6, 2013
ChemCatChem 2013, 5, 1490 – 1496 1496