Formic acid as a hydrogen storage medium: Ruthenium-catalyzed generation of hydrogen from formic acid in emulsions

Article abstract of DOI:10.1021/cs4007974

Formic acid is decomposed to H₂ and CO₂ in the presence of RuCl₃ and triphenylphosphines in an emulsion. In situ formed ruthenium carbonyls, such as [Ru(HCO₂)₂(CO) ₂(PPh₃)₂] (1), [Ru(CO)₃(PPh ₃)₂] (2), and [Ru₂(HCO₂) ₂(CO)₄(PPh₃)₂] (3), and a large cluster, involving a Ru₁₂ core, were identified and structurally characterized from the reaction mixtures. The catalytic activity of the mono and binuclear complexes was also investigated and it was found that [Ru ₂(HCO₂)₂(CO)₄(PPh₃) ₂] (3) shows high stability even at elevated temperatures and pressures and its activity is 1 order of magnitude lower than those measured for the mononuclear complexes. It was also attempted to use [Ru(HCO ₂)₂(CO)₂(PPh₃)₂] (1) as a catalyst for the hydrogenation of CO₂ to formic acid under neutral conditions. Although the reduction of CO₂ did not take place, the conversion of [Ru(HCO₂)₂(CO)₂(PPh ₃)₂] (1) to an unexpected carbonate, [Ru(CO ₃)(CO)₂(PPh₃)₂]·H ₂O was observed.

Full text of DOI:10.1021/cs4007974

Research Article
pubs.acs.org/acscatalysis
Formic Acid As a Hydrogen Storage Medium: Ruthenium-Catalyzed
Generation of Hydrogen from Formic Acid in Emulsions
Miklos Czaun,* Alain Goeppert, Jotheeswari Kothandaraman, Robert B. May, Ralf Haiges,
G. K. Surya Prakash,* and George A. Olah*
Loker Hydrocarbon Research Institute and Department of Chemistry, University of Southern California, University Park Campus,
Los Angeles, California 90089, United States
S
* Supporting Information
ABSTRACT: Formic acid is decomposed to H2 and CO2 in the presence of
RuCl3 and triphenylphosphines in an emulsion. In situ formed ruthenium
carbonyls, such as [Ru(HCO2)2(CO)2(PPh3)2] (1), [Ru(CO)3(PPh3)2] (2),
and [Ru2(HCO2)2(CO)4(PPh3)2] (3), and a large cluster, involving a Ru12
core, were identified and structurally characterized from the reaction
mixtures. The catalytic activity of the mono and binuclear complexes was
also investigated and it was found that [Ru2(HCO2)2(CO)4(PPh3)2] (3)
shows high stability even at elevated temperatures and pressures and its
activity is 1 order of magnitude lower than those measured for the
mononuclear complexes. It was also attempted to use [Ru-
(HCO2)2(CO)2(PPh3)2] (1) as a catalyst for the hydrogenation of CO2 to
formic acid under neutral conditions. Although the reduction of CO2 did not
take place, the conversion of [Ru(HCO2)2(CO)2(PPh3)2] (1) to an
unexpected carbonate, [Ru(CO3)(CO)2(PPh3)2]·H2O was observed.
KEYWORDS: hydrogen storage, formic acid decomposition, ruthenium chloride, homogeneous catalyst, emulsion
urning of fossil fuels, mainly coal, natural gas, and oil, to
harmful side products, storing and transporting hydrogen in
Bfulfill the ever increasing energy demand of humankind large quantities is quite challenging because of its volatile,
flammable, and explosive nature.6
has resulted in the accumulation of carbon dioxide in the
Numerous substances, for example, ammonia borane,7 alkali
atmosphere where its concentration will exceed 400 ppm this
borohydrides,8 hydrous hydrazine,8 methane, methanol,6 etc.,
year. The global anthropogenic carbon dioxide emissions rose
have been suggested for chemical energy storage in addition to
hydrogen. However, the practical applicability of some of these
substances may be hampered because of their relatively high
price, toxicity, or safety issues.
Formic acid (FA) is a low volatility, low toxicity organic
acid,9 which can be synthesized by decomposition of biomass,
or by catalytic hydrogenation,10−12 and direct electrochemical
reduction13−15 of carbon dioxide. Subsequently, FA can be
decomposed to a mixture of H2 and CO2 resulting in a “carbon
neutral” energy/hydrogen storage system (Figure 1).16−23 The
required CO2 can also be captured from industrial streams or
even directly from air.24−28
The decomposition of FA can follow two main pathways: (a)
decarbonylation resulting in carbon monoxide and water (1:1)
and (b) decarboxylation providing a 1:1 mixture of carbon
dioxide and hydrogen (Scheme 1). The latter mixture can be
used directly as a feedgas for hydrogen/air fuel cells,29,30 and
the hydrogen depleted exhaust may be recycled for FA
synthesis.
by more than 5% in 2010, representing the largest increase per
annum in the last two decades.1 The finite nature of fossil fuels
and the increasingly severe consequences of high carbon
dioxide levels in the atmosphere, such as global warming and
seawater acidification calls for a sustainable energy cycle to
replace our existing one. It can be expected that as a result of
the conscious effort to decrease anthropogenic greenhouse gas
emission, the contribution of renewable energy sources to the
total energy production will be increasing.
Extensive exploitation of renewable energy sources like solar,
wind and hydro in addition to the utilization of nuclear power
will provide enough energy for the future and predictions
indicate that humankind may indeed face energy storage and
transportation problems as well as shortage of carbon sources
rather than an actual energy shortage.2 A more efficient
utilization of our limited fossil fuel sources may delay the
difficulties associated with carbon shortage.3−5
Using electricity from renewable sources to produce
hydrogen by splitting water followed by storage and then
utilization of hydrogen to generate electricity on demand would
be a “clean” approach to store energy since the only product
generated by burning hydrogen is water, an environmentally
benign compound. Although this technology would not emit
Received: September 9, 2013
Revised: November 27, 2013
Published: December 4, 2013
© 2013 American Chemical Society
311
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
in water disfavors its application in aqueous reaction mixtures.
On the other hand, transition metal salts like RuCl3, as well as
FA and sodium formate are mainly soluble in water.
Substitution of the aromatic rings in triphenyl phosphine may
not only change the electron density on the phosphorus atom
but may also modify the solubility of the ligand in organic and
aqueous solutions as it was demonstrated with TPPTS ligand.32
The miscibility of aqueous and organic phases can be
enhanced by the addition of surfactants to form emulsions.
Even though these emulsions may not be stable over the entire
reaction time due to dramatic changes in the composition of
the reaction mixture, the initial presence of an emulsion can
provide the proper medium for the in situ formation of active
catalytic species. The role of such emulsions in the
decomposition of FA has now been explored.
The aqueous solution of FA, sodium formate and RuCl3 was
added dropwise to the vigorously stirred mixture of toluene,
PPh3 and sodium dodecyl sulfate (SDS) followed by sonication
of the formed emulsion for 15 min at room temperature. The
light brown emulsion was then transferred to a high pressure
Monel autoclave and heated to temperatures of 90 to 117 °C.
The internal temperature of the reactor and the pressure of the
gaseous product was recorded using Lab View and the rate of
decomposition was calculated on the basis of the pressure
versus time profiles (Figure 2) at various temperatures (e.g.,
Figure 1. Carbon neutral energy storage using FA as storage medium.
Scheme 1. Decomposition Pathways of FA: (a)
Decarbonylation and (b) Decarboxylation
Since CO may deactivate the catalysts of fuel cells, the
development of high activity homogeneous catalysts for the
selective decarboxylation of FA suppressing the decarbon-
ylation has received increasing attention.17−19,23
Ruthenium catalysts developed by Beller et al., which are
formed in situ from [RuCl2(benzene)]2 in the presence of 1,2-
bisdiphenylphosphinoethane (dppe) and a variety of amine
additives e.g. Et3N exhibited excellent performance for FA
decomposition. For example, the [RuCl2(benzene)]2/dppe/
dimethyloctylamine (DMOA) system showed one of the
highest activity (TOF = 16000 h−1 60 °C) reported for
hydrogen generation from FA.31
Another outstanding example was reported by Laurenczy et
al. in 2009 using [Ru(H2O)6](tos)2], [Ru(H2O)6](tos)3], or
RuCl3 in an aqueous solution of FA/Na-formate (9:1) (TOF =
460 h−1) and meta-trisulfonated triphenylphosphine (TPPTS)
as catalyst precursors.32
Although these systems showed excellent performance, their
large scale application may bring up feasibility concerns due to
the relatively high cost of the catalyst precursors.
In addition to this, volatile amine additives (e.g., Et3N) may
contaminate the gaseous products, necessitating gas purification
before their utilization in a fuel cell. To avoid this problem,
application of less volatile amines (e.g., HexNMe2) has been
already suggested.16,31
Figure 2. Pressure vs time diagram for the FA decomposition in the
presence of RuCl3−PPh3−SDS: (a) first run and (b) second run. FTIR
spectra of gaseous products (c and d).
For the wide practical application of continuous hydrogen
generation from FA, the development of more cost-effective
systems would however be essential for example (a) by using
less expensive and stable ligands that can be produced on a
large scale or (b) development of active catalysts using
nonprecious metals. The highest activity and productivity
results (TOF of 9425 h−1 at 80 °C) for nonprecious metal-
based catalyst systems were observed using P(CH2CH2PPh2)3
as a ligand and Fe(BF4)2·6H2O as the metal precursor.21
However its long-term stability remains to be studied.
The aim of this work is to gain an insight into the underlying
chemistry of formic acid decomposition in the presence of cost-
effective and commercially available ligands (e.g., PPh3) and
RuCl3, through identification of organometallic species formed
in situ under relatively harsh conditions, such as aqueous acidic
solutions at elevated temperatures.
1.692 × 10−5 M·s−1 at 90 °C, Table 1). To determine the
selectivity of the decomposition, the chemical composition of
the gas mixture was analyzed by gas chromatography using a
thermal conductivity detector (TCD) and in some cases by
Fourier transform infrared spectroscopy (FTIR) as well.
The high selectivity of the reaction was reflected by the lack
of CO peak in the chromatogram as illustrated in Supporting
Information Figure S1a. Using FTIR spectroscopy, a more
sensitive analytical method, only traces of CO were detected in
the gas mixture (Figure 2c and d). After the first run, 97% FA
was added and the rate of decomposition was measured (2.001
× 10−5 M·s−1 at 90 °C). As Figure 2 illustrates, the system
showed a higher reaction rate in the second run due to the
formation of more active species. However, gas chromato-
graphic measurements revealed the accumulation of 0.25% CO
(Supporting Information Figure S1b).
Triphenyl phosphine (PPh3) is a commonly used P-donor
ligand in organometallic chemistry. However, its poor solubility
312
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
Table 1. Reaction Rates of FA Decomposition in the Presence of Various Phosphine Ligands and Composition of Gaseous
Products
b
gas composition (%)
a
phosphine
PPh3
t (°C)
106 × reaction rate (M·s−1
)
H2
CO2
CO
nd
0.25
nd
0.20
nd
c
1
2
3
4
5
6
7
8
9
90
90
16.920
20.010
13.990
7.314
54.20
56.19
54.71
55.70
53.46
51.31
53.60
54.17
54.01
48.67
45.80
43.56
45.29
44.10
46.54
48.69
46.40
45.83
45.99
47.57
d
c
PPh3
P(C6H4−CH3)3
P(C6H4−CH3)3
PPh3·HBr
90
d
c
90
90
30.172
61.243
16.288
4.428
d
c
PPh3·HBr
90
nd
nd
nd
nd
P(C6H4−Cl)3
90
c
e
OPPh3
90
d
e
OPPh3
90
1.096
c
c
10
11
12
P(C6H4−F)3
P(C6H4−F)3
90
0.579
3.76
110
117
0.997
f
no catalyst
0.377
39.39
50.58
10.1
a
[RuCl3] = 2.43 mM, [Phosphine] = 5.0 mM, [FA] = 3.6 M, [HCO2Na] = 0.4 M, [SDS] = 19.65 mM, 25 mL aqueous solution, 5 mL toluene.
Determined by gas chromatography using TCD, nd: no CO could be detected. First run. Second run. No SDS was added. [FA] = 9 ×
b
c
d
e
f
[HCO2Na] = 3.6 M, 25 mL aqueous solution, no catalyst, Table S1 shows the reaction rates at variable temperatures (see Supporting Information).
Figure 3. 13C NMR (a−c) and 31P NMR spectra (d) of crude catalyst mixture and the stucture of the Ru complexes formed in situ (e) (* =
unidentified species).
In another experiment, the emulsion was separated by
centrifugation directly after the first run and the aqueous phase
was extracted with toluene. In order to understand the nature
of the organometallic species formed in situ that may contribute
to the catalytic activity, the identification of Ru-complexes was
in the presence of PPh3 and Et3N. Instead, the authors
d e s c r i b e d t h e f o r m a t i o n o f t h e b i n u c l e a r
[Ru2(HCO2)2(CO)4(PPh3)2] (3) and the tetranuclear
[Ru4(HCO2)4(CO)8(PPh3)2] complex; the latter being a fairly
unstable species.
1
carried out directly from the crude reaction mixture. The H
In our system, the lack of a peak at 5.8 ppm (31P NMR)
revealed the absence of tetranuclear Ru species
[Ru4(HCO2)4(CO)8(PPh3)2]. However, it should be noted
here that 31P NMR measurements of the crude reaction
mixture also revealed the presence of an unidentified
phosphorus containing species with a peak at 24.3 ppm.
The isolated crude catalyst mixture was washed with hexane
and recrystallized from chloroform by pentane diffusion. The
obtained yellow prisms were characterized by various
spectroscopic methods (Supporting Information Figure S2
and Figure 3) and subjected to X-ray crystal structure analysis.
NMR spectra of the crude extract showed the presence of
unreacted FA and Na-formate, ruthenium coordinated formate,
the mixture of PPh3 and OPPh3. 13C NMR studies (Figure
3) revealed the presence of [Ru(HCO2)2(CO)2(PPh3)2] (1)
−
(196.7 ppm (t, 11.2 Hz, CO), 167.5 ppm (s, HCO2 )),
[Ru(CO)3(PPh3)2] (2) (208.0 ppm (t, 16.1 Hz, CO)), and
[Ru2(HCO2)2(CO)4(PPh3)2] (3) complex (204.7 ppm, (t, 4.1
−
Hz, CO), 176.3 ppm (t, 8.2 Hz, HCO2 )).
31P NMR measurements (Figure 3d) also confirmed the
presence of complex [Ru2(HCO2)2(CO)4(PPh3)2] (3), [Ru-
(HCO2)2(CO)2(PPh3)2] (1), and [Ru(CO)3(PPh3)2] (2) at
12.6, 31.2, and 55.6 ppm, respectively. Interestingly, Wills et al.
have not reported the presence of 1 and 2 in the reaction
between FA and RuCl3, [RuCl2(DMSO)4] or [RuCl2(NH3)6]
1
The H, 13C, and 31P NMR measurements revealed that
these prisms were complex 1. The FTIR spectrum of the yellow
prisms collected in carbon tetrachloride exhibits intense bands
at 2052.8, 1991.1, 1957.0 cm−1 due to the stretching vibration
313
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
Table 2. Crystallographic Parameters for [Ru(HCO2)2(CO)2(PPh3)2]·2CHCl3, [Ru(CO)3(PPh3)2]·toluene, and
[Ru12C30H14Na2O50·6(C18H15OP)·2(C7H8)·4(H2O)]
1·2CHCl3
2·toluene
4
formula
MW (g/mol)
space group
a (Å)
C42H34Cl6O6P2Ru
1010.40
C42.20H34Cl0.30O2.71P2Ru
C76H66NaO30P3Ru6
2181.61
757.84
P1
̅
9.915(2)
P21/c
P1
̅
17.7301(6)
16.3958(9)
b (Å)
11.919(2)
12.3659(4)
17.5531(9)
c (Å)
20.194(4)
18.3431(6)
17.6055(9)
α (deg)
β (deg)
γ (deg)
V (Å3)
97.563(3)
90
101.4600(10)
113.7560(10)
107.5850(10)
4111.9(4)
91.490(3)
102.9990(10)
112.378(3)
90
2180.0(8)
3918.6(2)
Z
2
4
2
T (K)
133(2)
100(2)
100.(2)
λ (Å)
0.71073
0.71073
0.71073
Dcalcd (g/cm3)
1.539
1.285
1.762
μ (mm−1
)
0.847
0.554
1.220
crystal size (mm)
0.15 × 0.15 × 0.15
13889
0.32 × 0.24 × 0.15
94892
0.570 × 0.080 × 0.080
99952
reflections collected
independent reflections
goodness-of-fit on F2
Final R indices [I > 2σ(I)]
R indices (all data)
9554 [R(int) = 0.0344]
1.047
13990 [R(int) = 0.0365]
1.132
24604 [R(int) = 0.0275]
1.088
R1 = 0.0447, wR2 = 0.0802
R1 = 0.0720, wR2 = 0.1004
R1 = 0.0566, wR2 = 0.1579
R1 = 0.0739, wR2 = 0.1704
R1 = 0.0279, wR2 = 0.0602
R1 = 0.0391, wR2 = 0.0648
Figure 4. Crystal structure of complexes 1, 2, and 3. Hydrogen atoms and the solvent molecules have been omitted for clarity. Thermal ellipsoids are
drawn at the 50% probability level.
of coordinated CO ligands. Furthermore, the bands at 1606.6
and 1300.6 cm−1 can be attributed to the asymmetric and
symmetric carbonyl stretching vibration of unidentate formate
ligand, respectively (Supporting Information Figure S2). These
values are in good agreement with those reported by Johnson et
al. for the same compound (1).33
The X-ray crystal structure of the obtained yellow-colored
prisms revealed that the mononuclear hexacoordinated Ru(II)
compound [Ru(HCO2)2(CO)2(PPh3)2] crystallizes together
with two molecules of chloroform solvent in the triclinic space
occupying the axial positions and the formate and carbonyl
ligands arranged cis in the equatorial positions. The Ru atom
lies in the equatorial plane formed by the oxygen atoms O(3)
and O(5) of the formate ligands and the carbon atoms C(1)
and C(2) of the carbonyl ligands.
Although it has been reported by Johnson et al.33 that the
reaction between [Ru(CO)3(PPh3)2] and formic acid under
toluene reflux results in the formation of 1, the same complex
has never been synthesized directly in a reaction between
RuCl3, PPh3 and formic acid in which the acid acts as both,
carbonylating as well as reducing agent. Moreover, the X-ray
crystal structure of [Ru(HCO2)2(CO)2(PPh3)2] had not been
established until our current studies.
group P1 with two symmetry related units in the unit cell (Z =
̅
2).
Further crystallographic parameters are listed in Table 2. The
crystal structure of 1 is depicted in Figure 4. In this compound,
the ruthenium atom is hexa-coordinated by two triphenylphos-
phine, two carbonyl and two formate ligands. The arrangement
is derived from an octahedron with both PPh3 ligands
After isolating [Ru(HCO2)2(CO)2(PPh3)2] from the chloro-
form solution by pentane diffusion, the supernatant was
evaporated and redissolved in toluene. Hexane was diffused
into the toluene solution and stored at 0 °C for a week giving
314
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
a
Scheme 2. Formation of Complexes 1−4
a
Reaction conditions: (i) FA, Na-formate, PPh3, toluene, SDS, 90 °C; (ii) FA, Na-formate, O = PPh3, toluene, 90 °C.
two different types of crystals: yellow plates and yellow blocks.
The yellow plates were found suitable for single crystal X-ray
diffraction and the structure determination revealed the
binuclear complex [Ru2(HCO2)2(CO)4(PPh3)2] with two
bridging formato ligands (Figure 4). The determined crystallo-
graphic parameters of the binuclear ruthenium complex
(Supporting Information Table S2) are in good agreement
with those reported by Wills et al.34
The yellow block-like crystals were identified by single crystal
X-ray diffraction as [Ru(CO)3(PPh3)2] (2), which crystallizes
with one molecule of toluene solvent in the asymmetric unit.
The crystal structure of the compound is depicted in Figure 4.
The basic geometry of [Ru(CO)3(PPh3)2] is derived from a
bipyramidal ligand arrangement around the Ru center with the
two PPh3 groups occupying the axial and the carbonyls in the
equatorial positions. The determined X-ray crystal structure of
[Ru(CO)3(PPh3)2]·toluene suffers from several disorders (see
Figure 5. Reaction rates of FA decomposition using triphenylphos-
phine and derivatives. Conditions: see in Table 1
in the Supporting Information).
Collman and Roper35 reported the first synthesis of
[Ru(CO)3(PPh3)2] from [RuCl2(CO)2(PPh3)2] by reaction
with zinc dust. In our reaction, [Ru(CO)3(PPh3)2] may be
formed via [RuCl2(CO)2(PPh3)2] since the latter complex can
be a potential intermediate of the reaction between RuCl3, CO
and PPh3 under reductive conditions35 as described in Scheme
2.
decarboxylation of FA. A blank experiment was carried out
without adding any catalyst to make an adequate comparison
on the selectivity of the tested catalysts shown in Table 1. As
expected, both the rate of the decomposition and the selectivity
(rr = 0.377 × 10−6 M s−1 at 117 °C, CO% = 10.1) suffered
from the absence of catalyst giving higher CO content than the
one measured in the presence of the least selective catalyst
precursor, tris(4-fluorophenyl)phosphine, (CO% = 3.76). It can
be concluded that the application of a less active catalyst with
long reaction times favors the accumulation of CO in the gas
mixture.
To explore the robustness of the RuCl3−phosphine system
in an emulsion, triphenylphosphine oxide (OPPh3) was used
as a ligand to model practical conditions (e.g., PPh3 is
contaminated with OPPh3 impurity, the FA feed contains
dissolved oxygen, or both). Although the reaction rate in the
presence of OPPh3 was relatively low (4.43 × 10−6 M s−1) at
90 °C the selectivity of the reaction was high, giving a mixture
of H2 and CO2 (Table 1) with no traces of CO.
Study of the Performance of RuCl3 for FA Decom-
position in the Presence of Triphenylphosphine
Derivatives. Performance of catalysts formed from RuCl3
and ortho- and para-substituted triphenylphosphines, triphe-
nylphosphonium bromide and triphenylphosphine oxide
(Supporting Information Figure S4) have also been investigated
for the decarboxylation of FA in an emulsion. Comparison of
reaction rates obtained using P-donor ligands is illustrated in
Figure 5 and also given in Table 1.
Apparently, the reaction rate is independent of the electronic
effects of the substituents since similar values were obtained for
PPh3, tris(4-chlorophenyl)phosphine and tris(2-tolyl)-
phosphine (16.92 × 10−6, 16.29 × 10−6, 13.99 × 10−6 M s−1,
respectively). Instead, it can be suggested that the solubility of
different ligands and the complexes formed in situ in the
aqueous phase affects the catalytic activity. This is also
supported by the higher activity of the RuCl3−PPh3·HBr
system (30.17 × 10−6 M s−1), where phosphine salt has a
relatively good solubility in water compared with nonionic
phosphine ligands. On the other hand, RuCl3/tris(4-
fluorophenyl)phosphine remained practically inactive even at
It is very unlikely that OPPh3 can be converted into PPh3
even under reductive conditions and that Ru−PPh3 complexes
similar to the ones that were detected when PPh3 was used
directly as a ligand can be formed. In the reaction between
RuCl3 and OPPh3 in the presence of FA and sodium formate
in the toluene/water biphasic system the formation of
[Ru4(CO)12H4]18 (toluene phase) and orange-red block-like
crystals in the aqueous layer were observed.
110 °C as reflected in its low reaction rate (0.997 × 10−6
M
On the basis of the 31P NMR results, the presence of any
phosphine ligands in the orange-red compound was ruled out
(except for OPPh3) while signals at 181.4 and 162.3 ppm in
the 13C NMR spectrum can be assigned to carbonyl and
carbonate ligands, respectively (Supporting Information Figure
S5). IR measurements showed five adsorption bands in the
s−1). This system also showed poor selectivity, resulting in a gas
mixture containing 3.8% CO.
Thermal decomposition of FA to H2O and CO (Scheme 1a)
can also take place as a side reaction, resulting in higher CO
content in the presence of a catalyst which is less active for the
315
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
Figure 6. The core structure of the Ru cluster in the crystal structure of ([Ru12C30H14Na2O50 6(C18H15OP) 2(C7H8) 4(H2O)] (4). (a) Hydrogen
atoms, water, toluene, triphenylphosphine oxide, and carbon monoxide molecules have been omitted for clarity. (b) Hydrogen atoms, water, toluene,
and triphenylphosphine oxide have been omitted for clarity.
Table 3. Comparison of Reaction Rates and Selectivity Using RuCl3/PPh3 in the Presence of Different Surfactants
b
gas composition (%)
surfactant
t (°C)
106 × reaction rate (M·s−1
)
H2
CO2
CO
nd
a
a
a
c
1
2
3
4
5
SDS
100
100
100
100
100
69.14
26.22
9.58
54.20
54.49
55.80
52.23
54.91
45.80
45.44
43.98
47.47
44.95
TBAB
MiTMAB
0.07
0.22
0.30
0.15
9.90
d
14.09
a
[RuCl3] = 2.43 mM, [Phosphine] = 5.0 mM, [FA] = 3.6 M, [HCO2Na] = 0.4 M, [surfactant] = 19.65 mM, 25 mL aqueous solution, 5 mL toluene.
Determined by gas chromatography using TCD. nd: no CO could be detected. No toluene and surfactant was added. No surfactant was added.
b
c
d
region from 2051 to 1922 cm−1 (carbonyl ligands) and also a
broad adsorption at 1540 cm−1 (carbonato) (Supporting
Information Figure S6).
It should be noted that it was very difficult to properly
determine the composition of the cluster by single crystal X-ray
diffraction data due to the problem of determining hydrogen
positions in the electron density map in the presence of several
Ru atoms. The refinement of the structure initially led to a
cluster containing six molecules of H2O and resulted in a
composition [Ru6(CO3)3(CO)12(H2O)6][Na2(H2O)2]. The
resulting final electron density map of the Ru cluster showed
the presence of residual electron density (1.8 eÅ−3) above the
triangular Ru3 faces of the cluster which could be attributed to
hydride ligands. However, it was not possible to obtain any
evidence for the presence of the face-capping hydride ligand by
1H NMR spectroscopy. On the other hand, charge-balance
considerations for the twelve-membered cluster resulted in a
double positive charge for a cluster with the Ru12C30H16Na2O50
that was not being compensated by two negative charges. Close
inspection of all hydrogen atoms that were described as water
molecules revealed that all would take part in hydrogen
bonding. However, it is entirely possible that two of the water
molecules are better described as OH groups which might even
be disordered in a way that the hydrogen atom of the group is
distributed between the two positions, thus simulating the
appearance of a water molecule. In order to maintain charge
neutrality in the total composition of the compound, we favor
the removal of two H+ from two water molecules that can form
Ru−O−Na bridges and the cluster can be described as
[Ru6(CO3)3(CO)12(OH)2(H2O)4][Na2(H2O)2]. However,
more data (neutron diffraction) would be necessary in order
to fully determine the nature and composition of this cluster.
Study of the Effect of Surfactants on the Reaction
Rate and on the Formation of Ru Species. A series of
control experiments were carried out to investigate the effect of
surfactants on the reaction rate and the formation of Ru species.
Although, the lack of toluene and surfactant did not result in
The orange-red blocks were subjected to single crystal X-ray
diffraction, which showed the presence of a large unexpected
Ru carbonato cluster incorporating twelve ruthenium atoms
([Ru12C30H14Na2O50·6(C18H15OP)·2(C7H8)·4(H2O)] (4),
unit cell volume = 4111.9(4) Å3). Crystallographic parameters
of the complex are summarized in Table 2. The structure is
quite unique because it consist of two identical cluster units
[Ru6(CO3)3(CO)12(OH)(H2O)2] that are bridged by a
[Na2(H2O)2] unit (Figure 6).
The asymmetric unit of the structure contains just one of
these Ru6 cluster units while the second, symmetry-related unit
is generated by the symmetry transformation (−x + 2, −y + 1,
−z + 1). The entire unit contains a total of six bridging
carbonato ligands that are connected to four Ru atoms each. In
all cases two oxygen atoms (O1 and O3) from each of the six
carbonate groups coordinate a single Ru atom each, while the
third oxygen atom (O2) coordinates two Ru atoms. Carbon−
oxygen distances involving oxygen atoms that coordinate just
one Ru atom range between 1.265(3) and 1.282(2) Å and are
significantly shorter than the ones involving an oxygen atom
that coordinates two Ru atoms (1.299(3)−1.304(3) Å). The
O−C−O bond angles around C1 of the carbonate group range
between 117.9(2)° and 122.1(2)° and deviate only little from
the ideal value of 120°.
In the Ru3 triangle, the bond angles range between
59.380(6)° and 60.442(7)° resulting in an equilateral triangular
geometry. In all cases two Ru atoms of the triangle are bridged
by one oxygen atom of a carbonato groups. In addition, each of
the Ru atoms of the triangle is coordinated by two terminal CO
ligands and another Ru atom that is not involved in the Ru3
triangle.
316
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
Table 4. Comparison of Reaction Rates and Selectivity Using Catalyst 1−3
gas composition (%)
CO2
a
catalyst
t (°C)
106 × reaction rate (M·s−1
)
H2
CO
1
2
3
1
2
3
115
115
115
1.210
2.060
0.703
46.54
55.31
54.23
50.74
42.95
37.12
2.72
1.74
8.66
a
[Catalyst] = 0.424 mM, [FA] = 3.6 M, [HCO2Na] = 0.4 M, [SDS] = 19.65 mM, 25 mL aqueous solution, 5 mL toluene.
complete inhibition, a significantly lower reaction rate (0.990 ×
10−5 M·s−1 at 100 °C, Table 3) and selectivity (CO = 0.3%)
could be observed. Experiments were also carried out in the
absence of surfactant in a water−toluene biphasic system giving
a slightly higher reaction rate (1.409 × 10−5 M·s−1 at 100 °C)
than the one observed in the aqueous phase. The selectivity of
the decomposition reaction also increased with only 0.15% CO
present in the gaseous phase. Interestingly, in the water−
toluene biphasic system, [Ru2(HCO2)2(CO)4(PPh3)2] complex
was formed with no traces of any other Ru complexes detected
by 31P NMR measurements except for a minor non-
characterized species at 13.1 ppm (Supporting Information
Figure S7). It also demonstrates the synthetic importance of the
method, namely, using FA as both reducing agent and carbonyl
source in the preparation of organometallic compounds.
The utilization of an alkylammonium salts such as
tetrabutylammonium bromide (TBAB) as a detergent, instead
of SDS was also investigated. The reaction rate of the
decomposition in the presence TBAB was 2.622 × 10−5 M·
s−1 (rr = 6.914 × 10−5 M·s−1 at 100 °C using SDS). On the
basis of the 31P NMR spectrum, [Ru2(HCO2)2(CO)4(PPh3)2]
was formed in high yield in addition to some [Ru-
(HCO2)2(CO)2(PPh3)2] and negligible amounts of [Ru-
(CO)3(PPh3)2] (Supporting Information Figure S8).
Using myristyl trimethylammonium bromide (MiTMAB)
resulted in a slightly lower reaction rate (0.958 × 10−5 M·s−1 at
100 °C) than the one measured in the absence of surfactant
and toluene (0.990 × 10−5 M·s−1). Although 31P NMR
investigation of the freshly prepared sample revealed the initial
presence of [Ru(CO)3 (PPh3 )2 ], one day later
[Ru2(HCO2)2(CO)4(PPh3)2] was the only Ru-phosphine
species present in significant concentration (Supporting
Information Figure S9).
It is worth noting that similar to the “ligand free”
decomposition of FA in the presence of RuCl3,18 no
decomposition reaction takes place using any of the
triphenylphosphine derivatives as ligands under atmospheric
pressure regardless of any other reaction parameters, most
probably due to low partial pressure of CO that is essential for
the formation of the active catalysts.
Authentic [Ru(CO)3(PPh3)2] (2) was synthesized according
to a literature method,36 recrystallized from the mixture of THF
and heptanes (1:1) and used as a catalyst for FA decomposition
in emulsion (see Experimental Section). Analysis of the product
gases (H2% = 55.31, CO2% = 42.95, CO% = 1.74) and the
observed reaction rate (2.060 × 10−6 M/s at 115 °C) suggested
that complex 2 contributes to the catalytic activity to a certain
degree but does not give high selectivity. In addition, the 31P
NMR study of the crude reaction mixture (Supporting
Information Figure S10) revealed the transformation of 2
i n t o [ R u ( H C O 2 ) 2 ( C O ) 2 ( P P h 3 ) 2 ] ( 1 ) a n d
[Ru2(HCO2)2(CO)4(PPh3)2] (3).
[Ru2(HCO2)2(CO)4(PPh3)2] (3) was synthesized from
RuCl3 in the presence of FA and Na-formate in water/toluene
biphasic system in the absence of detergent and the crude
material was recrystallized from CHCl3 by hexane diffusion as it
was described earlier. It is very unlikely that complex 3 is an
active species (or its direct precursor) in the FA decomposition
as reflected by the low catalytic activity and poor selectivity
(7.027 × 10−7 M/s at 115 °C, H2%= 54.23, CO2%= 37.12, CO
%= 8.66) which we observed using 3 as a catalyst in emulsion.
The low catalytic activity is in accordance with the similar
finding of Wills et al.34 Intact complex 3 could be recovered
from the toluene phase after the decomposition reaction,
demonstrating the high stability (and most likely low activity)
of the complex. Supporting Information Figures S11 and S12
show the comparison of 31P and 13C NMR spectra of 3,
respectively, before and after heating with FA. [Ru-
(HCO2)2(CO)2(PPh3)2] (1) was synthesized from [Ru-
(CO)3(PPh3)2] (2) by refluxing in FA following a literature
method.33 Utilization of 1 as a catalyst resulted in moderate
activity (1.210 × 10−6 M/s at at 115 °C) and selectivity H2% =
46.54, CO2% = 50.74, CO% = 2.72). Supporting Information
Figure S13 shows complex 1 was converted to the binuclear
[Ru2(HCO2)2(CO)4(PPh3)2] (3) upon treating with FA.
One can conclude that isolated and structurally characterized
Ru complexes (1−3) formed in situ from RuCl3 and PPh3 in
the presence of FA do not contribute significantly to the overall
catalytic activity. Regardless of their low catalytic activity, the
one pot preparation of 1−4 from FA without using explosive or
toxic gases, such as H2 and CO, provides a very attractive
approach to the synthesis of Ru carbonyls.
Comparison of reaction rates and CO contents led us
conclude that the addition of surfactants, particularly SDS,
significantly increased the activity and the selectivity of the FA
decarboxylation and also affected the in situ formation of Ru-
carbonyl complexes.
Study of FA Decomposition in the Presence of
Complexes 1, 2, and 3. Questions arise whether the isolated
Ru complexes 1−3 are active catalysts for FA decomposition or
the resting states of the active species or simply inactive
complexes representing “dead ends” in the catalytic cycle? In
order to answer these questions, authentic complexes 1−3 were
synthesized and used as catalyst for the decarboxylation of FA
(Table 4).
Conversion of [Ru(HCO2)2(CO)2(PPh3)2] to a Ruthe-
nium Carbonato Complex. Many examples support that a
catalyst active for the decomposition of FA may also catalyze
the reverse reaction, namely, the formation of FA from CO2
and H2 under the reaction conditions.29,37−39 This could lead
to a CO2 neutral hydrogen storage system.40 Usually, the
addition of bases favors the FA formation (because of the
formation of formates) while acidic conditions favor the
decomposition. However it must be added here that, from a
practical point of view, shifting the unfavorable equilibrium of
CO2 hydrogenation to FA by addition of bases necessitates a
further purification step to recover FA from formate salts. On
317
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
the other hand, only a few examples for the reduction of CO2
to FA in the absence of base have been reported.37 There has
also been a very promising progress in isolating FA from amine
adducts while minimizing the catalyst loss by using a two phase
liquid−liquid system (diol/[NHex3·HCO2H]−NHex3) as the
reaction media.41 Alternatively, the formate−bicarbonate
equilibrium (HCO2Na + H2O ↔ NaHCO3 + H2) can also
be considered to design a carbon-neutral charge−discharge
device to store energy.40,42,43
We attempted to hydrogenate CO2 in the presence of
[Ru(HCO2)2(CO)2(PPh3)2] (1) under neutral conditions. The
toluene solution of [Ru(HCO2)2(CO)2(PPh3)2] was pressur-
ized with the 1:3 mixture of CO2 and H2 (460 psi) and the
temperature was maintained at 50 °C for 24 h.
Monitoring the pressure of the gas phase showed no
consumption of CO2−H2 mixture. Furthermore, no FA or any
other product of CO2 reduction could be detected in the
reaction mixture. However, the disappearance of the intensive
peak at 1611.5 cm−1 in the IR spectrum of the crude reaction
mixture assigned to monodentate formate ligands in [Ru-
(HCO2)2(CO)2(PPh3)2] and the shift of carbonyl peaks
(2053.1 and 1992 cm−1) indicate the formation of a different
Ru-carbonyl complex (Scheme 3). Yellow blocks grew from the
Figure 7. Crystal structure of [Ru(CO3)(CO)2(PPh3)2]·H2O. Hydro-
gen atoms have been omitted for clarity. Thermal ellipsoids are drawn
at the 50% probability level.
the reported catalysts were enhanced by adding surfactants,
especially sodium dodecyl sulfate to the toluene/water biphasic
system. Ruthenium carbonyls, such as [Ru-
(HCO2)2(CO)2(PPh3)2] (1), [Ru(CO)3(PPh3)2] (2), and
[Ru2(HCO2)2(CO)4(PPh3)2] (3), have been isolated from
the crude reaction mixtures and structurally characterized. In
separate reactions, authentic 1, 2, and 3 were synthesized and
then used as catalysts for the decarboxylation of formic acid.
Based on the presented reaction rates and gas composition
data, it can be concluded that these complexes contribute to the
overall performance only to a small degree and other
substances may account for the observed catalytic activity and
selectivity. Despite the low catalytic activity of 1, 2, and 3, the
formation of these complexes directly in a reaction between
RuCl3 and formic acid in the presence of PPh3 shows an elegant
way for the synthesis of ruthenium formato carbonyls without
using explosive and toxic gases such as hydrogen and carbon
monoxide. It was also found that the reaction between RuCl3
and OPPh3 in the presence of FA and Na-formate resulted in
the formation of a novel high nuclearity ruthenium cluster
[Ru12C30H14Na2O50·6(C18H15OP)·2(C7H8)·4(H2O)] (4). In
the course of the formation of above ruthenium carbonyls FA
acted as both reducing and carbonylating agent.
Scheme 3. Conversion of [Ru(HCO2)2(CO)2(PPh3)2] to
[Ru(CO3)(CO)2(PPh3)2]·H2O
golden yellow solution at room temperature after 3 days. FTIR
spectrum of the crystalline material in KBr showed two
intensive peaks at 2045.6 and 1982.5 cm−1 attributed to
carbonyl ligands and three peaks at 1651.3, 1626.7, and 1237.6
cm−1 because of bidentate carbonate ligand (Supporting
Information Figure S14). High resolution mass spectrometry
in a solution of acetonitrile and methanol (1:1) showed a peak
at 743.0694 amu which corresponds to MH+ (where M =
[Ru(CO3)(CO)2(PPh3)2]). Single crystals grown from toluene
were suitable for X-ray diffraction. The structure of complex
[Ru(CO3)(CO)2(PPh3)2]·H2O is shown in Figure 7.
In accordance with the FTIR spectrum, the X-ray structure
determination revealed that the monodentate formate ligand
converted to a bidentate carbonate and a water molecule
connected to the third oxygen atom of the carbonate ligand by
hydrogen bonds. The crystallographic parameters are in good
agreement with those available in the literature.44 It was
reported by Calderazzo et al.44 that [Ru(CO3)(CO)2(PPh3)2]·
H2O can be prepared from a N,N-diisopropylcarbamato
ruthenium derivative [Ru(O2CNiPr2)2(CO)2(PPh3)2] by con-
trolled hydrolysis. However, to the best of our knowledge this is
the first example of the synthesis of [Ru(CO3)(CO)2(PPh3)2]·
H2O from a ruthenium bisformate.
EXPERIMENTAL SECTION
■
Typical Procedure for the Decomposition of Formic
Acid in Emulsion in the Presence of RuCl3 and
Triphenylphosphine Catalyst Precursors. Phosphine
ligand (0.015 mmol; e.g., 0.0393 g, PPh3) and 0.590 mmol of
emulsifier (e.g., sodium dodecyl sulfate (0.17 g) were mixed in
5.0 mL of toluene and deaerated. In a separate flask nitrogen
was bubbled through a mixture of 22.5 mL FA (4.0 M) and 2.5
mL HCO2Na (4.0 M) for 15 min. Subsequently 0.0156 g
(0.073 mmol) RuCl3 was added to the aqueous solution. This
mixture was then transferred to the toluene solution under
vigorous stirring and the resulting light brown emulsion was
sonicated under N2 for 15 min. The emulsion was then
transferred to an autoclave in a glovebox. The autoclave was
closed and immersed into a preheated oil bath and heating was
continued until the decomposition reached approximately
CONCLUSIONS
■
Understanding the underlying chemistry of the decomposition
of formic acid by identifying the organometallic species formed
in situ is essential for the development of robust and recyclable
catalysts. We have investigated for the first time the catalytic
decomposition of formic acid to hydrogen and carbon dioxide
using ruthenium trichloride and triphenylphosphines as catalyst
precursors in emulsion and in biphasic (aqueous/organic)
systems. It was found that both the activity and the selectivity of
318
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
100% FA conversion. Temperature of the reaction mixture and
pressure of the product gases were monitored using Lab View
8.6. The autoclave was allowed to cool down and the gas
mixture was vented to a plastic bag from where samples were
taken and quantitatively analyzed by gas chromatography and
FTIR spectroscopy.
for 3 days. The yellow cubes were filtered from the brown
mother liquor and washed with a few drops of toluene and then
with hexane at 0 °C. Yield: 0.086 g (50.3%). [Ru(CO3)-
(CO)2(PPh3)2]·H2O: FTIR (KBr, νmax/cm−1) 2045.6, 1982.5,
1651.3, 1626.7, and 1237.6 cm−1; HRMS in acetonitrile-
methanol (1:1) 743.0694 MH+ (where M = [Ru(CO3)-
(CO)2(PPh3)2]).
The reaction mixture was separated by centrifugation (4400
rpm, 5 min), extracted with toluene three times and the
combined organic layers were dried over anhydrous sodium
sulfate. Evaporation of toluene gave a crystalline yellow solid
and some yellow oil. 13C and 31P NMR measurements of the
crude product were carried out without further purification in
CDCl3. Ru-phosphine species identified:
ASSOCIATED CONTENT
■
S
* Supporting Information
Details for crystal structure determination, crystallographic
parameters, a typical gas chromatogram, FTIR spectra, 1H, 13C,
and 31P NMR spectra are given. This material is available free of
charge via the Internet at http://pubs.acs.org.
[Ru(HCO2)2(CO)2(PPh3)2] (1): 13C NMR (100 MHz,
CDCl3) (196.7 ppm (t, 11.2 Hz, CO), 167.5 ppm (s,
−
HCO2 )); 31P NMR (162 MHz, CDCl3) 31.2 ppm; FTIR
AUTHOR INFORMATION
(CCl4, νmax/cm−1) 2052.8, 1991.1, 1957.0 1606.6 and 1300.6.
■
Corresponding Authors
See the FTIR spectrum of [Ru(HCO2)2(CO)2(PPh3)2] in
Supporting Information Figure S2. [Ru(CO)3(PPh3)2] (2): 13
C
*Fax: (+)1 213 740 5087. E-mail: czaun@usc.edu.
*Fax: (+)1 213 740 5087. E-mail: gprakash@usc.edu.
*Fax: (+)1 213 740 5087. E-mail: olah@usc.edu.
NMR (100 MHz, CDCl3) 208.0 ppm, (t, 16.1 Hz, CO); 31P
NMR (162 MHz, CDCl3) 55.6 ppm, FTIR (KBr, νmax/cm−1)
1895. [Ru2(HCO2)2(CO)4(PPh3)2] (3): 13C NMR (100 MHz,
CDCl3−) 204.7 ppm, (t, 4.1 Hz, CO), 176.3 ppm (t, 8.2 Hz,
HCO2 ); 31P NMR (162 MHz, CDCl3) 12.6 ppm.
Notes
The authors declare no competing financial interest.
Crude reaction mixtures were washed with hexane at room
temperature three times and the remaining crystalline solid was
dissolved in chloroform and hexane was then diffused to the
homogeneous solution. The obtained crystals were suitable for
X-ray diffraction.
Reaction between RuCl3 and OPPh3 in the Presence
of FA and Na−Formate in Aqueous/Toluene. The
solution of 0.0885 g (0.318 mmol) of OPPh3 in 5.0 mL of
toluene was mixed with the aqueous solution of 0.0330 g
(0.154 mmol) of RuCl3 and reacted in a Monel autoclave at
100 °C for 12 h. The reaction mixture was cooled to RT
followed by venting of the gaseous mixture of H2 and CO2. The
yellow-orange toluene phase was separated and the aqueous
layer was extracted with toluene three times. Orange-red
crystals formed from the aqueous phase by next day which were
suitable for X-ray structural determination.
ACKNOWLEDGMENTS
■
Support of our work by the Loker Hydrocarbon Research
Institute and the U.S. Department of Energy is gratefully
acknowledged. We also acknowledge NSF CRIF grant 1048807
for the purchase of a X-ray diffractometer.
REFERENCES
■
(1) Harvey, F. The Guardian 2011, May 29.
(2) Olah, G. A.; Goeppert, A.; Prakash, G. K. S. J. Org. Chem. 2009,
74, 487.
(3) Olah, G. A.; Prakash, G. K. S.; Goeppert, A. J. Am. Chem. Soc.
2011, 133, 12881.
(4) Olah, G. A.; Goeppert, A.; Czaun, M.; Prakash, G. K. S. J. Am.
Chem. Soc. 2012, 135, 648.
(5) Olah, G. A.; Prakash, G. K. S.; Goeppert, A.; Czaun, M.; Mathew,
T. J. Am. Chem. Soc. 2013, 135, 10030.
[Ru12C30H14Na2O50·6(C18H15OP)·2(C7H8)·4(H2O)] (4):
13C NMR (100 MHz, CD3OD) 162.3 ppm (CO3), 181.4
ppm (CO); 31P NMR (162 MHz, CD3OD) 28.8 ppm; FTIR
(KBr, cm−1) 2050.7, 2022.5, 1987.5, 1956.0, 1922.4, 1540.4.
Conversion of [Ru(HCO2)2(CO)2(PPh3)2] to [Ru(CO3)-
(CO)2(PPh3)2]·H2O. Nitrogen was bubbled through a mixture
of 22.5 mL of FA (4.0 M) and 2.5 mL of HCO2Na (4.0 M) for
15 min and then 0.0468 g (0.219 mmol) of RuCl3 was added.
PPh3 (0.118 g, 0.45 mmol) and 0.51 g (1.770 mmol) of SDS
was dissolved in 5 mL of toluene. The toluene solution was
deaerated and the aqueous solution was added under vigorous
stirring and sonicated under N2 for 15 min. The emulsion was
transferred to an autoclave and pressurized with CO (30 psi).
The reaction mixture was kept at 90 °C for 17 h and then the
organic layer was dried under reduced pressure. (Note: Since
the FTIR analysis of the crude product confirmed the
formation of complex 1, it was used for the next step without
further purification.)
(6) Olah, G. A.; Goeppert, A.; Prakash, G. K. S. Beyond Oil and Gas:
The Methanol Economy, 2nd ed.; Wiley VCH: Weinheim, Germany,
2009.
(7) Conley, B. L.; Guess, D.; Williams, T. J. J. Am. Chem. Soc. 2011,
133, 14212.
(8) Jiang, H. L.; Singh, S. K.; Yan, J. M.; Zhang, X. B.; Xu, Q.
ChemSusChem 2010, 3, 541.
(9) Tanaka, N. Potential for Biomass and Carbon Dioxide Capture and
Storage; International Energy Agency Environmental Projects Ltd.:
London, 2011.
(10) Aresta, M. Carbon Dioxide as Chemical Feedstock; Wiley-VCH:
Weinheim, Germany, 2010.
(11) Federsel, C.; Jackstell, R.; Beller, M. Angew. Chem., Int. Ed. 2010,
49, 6254.
(12) Federsel, C.; Jackstell, R.; Boddien, A.; Laurenczy, G.; Beller, M.
ChemSusChem 2010, 3, 1048.
(13) Agarwal, A. S.; Zhai, Y.; Hill, D.; Sridhar, N. ChemSusChem
2011, 4, 1301.
(14) Prakash, G. K. S.; Viva, F. A.; Olah, G. A. J. Power Sources 2013,
223, 68.
The yellow solid was transferred to an autoclave and
dissolved in 15 mL of toluene and a 1:3 mixture of CO2 and H2
(460 psi) was added. The temperature was maintained at 50 °C
for 24 h. Small brown particles were removed with a syringe
filter and the obtained golden yellow solution was kept in a vial
(15) Keene, F. R. Electrochemical and Electrocatalytic Reactions of
Carbon Dioxide; Elsevier: Amsterdam, 1993.
(16) Boddien, A.; Loges, B.; Junge, H.; Beller, M. ChemSusChem
2008, 1, 751.
(17) Joo, F. ChemSusChem 2008, 1, 805.
319
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320

ACS Catalysis
Research Article
(18) Czaun, M.; Goeppert, A.; May, R.; Haiges, R.; Prakash, G. K. S.;
Olah, G. A. ChemSusChem 2011, 4, 1241.
(19) Loges, B.; Boddien, A.; Gartner, F.; Junge, H.; Beller, M. Top.
Catal. 2010, 53, 902.
(20) Fellay, C.; Dyson, P. J.; Laurenczy, G. Angew. Chem., Int. Ed.
2008, 47, 3966.
(21) Boddien, A.; Mellmann, D.; Gaertner, F.; Jackstell, R.; Junge, H.;
Dyson, P. J.; Laurenczy, G.; Ludwig, R.; Beller, M. Science 2011, 333,
1733.
(22) Grasemann, M.; Laurenczy, G. Energy Environ. Sci. 2012, 5,
8171.
(23) Hull, J. F.; Himeda, Y.; Wang, W. H.; Hashiguchi, B.; Periana,
R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Nat. Chem. 2012, 4, 383.
(24) Czaun, M.; Goeppert, A.; May, R. B.; Peltier, D.; Zhang, H.;
Prakash, G. K. S.; Olah, G. A. J. CO2 Util. 2013, 1, 1.
(25) Goeppert, A.; Meth, S.; Prakash, G. K. S.; Olah, G. A. Energy
Environ. Sci. 2010, 3, 1949.
(26) Meth, S.; Goeppert, A.; Prakash, G. K. S.; Olah, G. A. Energy
Fuels 2012, 26, 3082.
(27) Goeppert, A.; Czaun, M.; May, R. B.; Prakash, G. K. S.; Olah, G.
A.; Narayanan, S. R. J. Am. Chem. Soc. 2011, 133, 20164.
(28) Goeppert, A.; Czaun, M.; Surya Prakash, G. K.; Olah, G. A.
Energy Environ. Sci. 2012, 5, 7833.
(29) Loges, B.; Boddien, A.; Junge, H.; Beller, M. Angew. Chem., Int.
Ed. 2008, 47, 3962.
(30) Shin, J. H.; Yoon, J. H.; Lee, S. H.; Park, T. H. Bioresour.
Technol. 2010, 101, S53.
(31) Sponholz, P.; Mellmann, D.; Junge, H.; Beller, M.
ChemSusChem 2013, 6, 1172.
(32) Laurenczy, G.; Fellay, C.; Yan, N.; Dyson, P. J. Chem.Eur. J.
2009, 15, 3752.
(33) Johnson, B. F. G.; Johnston, R. D.; Lewis, J.; Williams, I. G. J.
Chem. Soc. A 1971, 689.
(34) Morris, D. J.; Clarkson, G. J.; Wills, M. Organometallics 2009, 28,
4133.
(35) Collman, J. P.; Roper, W. R. J. Am. Chem. Soc. 1965, 87, 4008.
(36) Ahmad, N.; Levison, J. J.; Robinson, S. D.; Uttlky, M. F.;
Wonchoba, E. R.; Parshall, G. W. In Inorganic Syntheses; John Wiley &
Sons, Inc.: Hoboken, NJ, 2007; Vol. 15, p 45.
(37) Hayashi, H.; Ogo, S.; Fukuzumi, S. Chem. Commun. 2004, 2714.
(38) Himeda, Y. Green Chem. 2009, 11, 2018.
(39) Gao, Y.; Kuncheria, J. K.; Jenkins, H. A.; Puddephatt, R. J.; Yap,
G. P. A. J. Chem. Soc., Dalton Trans. 2000, 4703.
(40) Boddien, A.; Gaertner, F.; Federsel, C.; Sponholz, P.; Mellmann,
D.; Jackstell, R.; Junge, H.; Beller, M. Angew. Chem., Int. Ed. 2011, 50,
6411.
(41) Schaub, T.; Paciello, R. A. Angew. Chem., Int. Ed. 2011, 50, 7278.
(42) Zaidman, B.; Wiener, H.; Sasson, Y. Int. J. Hydrogen Energy
1986, 11, 341.
(43) Papp, G.; Csorba, J.; Laurenczy, G.; Joo, F. Angew. Chem., Int.
Ed. 2011, 50, 10433.
(44) Dell’Amico, D. B.; Calderazzo, F.; Labella, L.; Marchetti, F. J.
Organomet. Chem. 2000, 596, 144.
320
dx.doi.org/10.1021/cs4007974 | ACS Catal. 2014, 4, 311−320