Highly efficient reversible hydrogenation of carbon dioxide to formates using a ruthenium PNP-pincer catalyst

Article abstract of DOI:10.1002/cctc.201402119

The use of hydrogen as a fuel requires both safe and robust technologies for its storage and transportation. Formic acid (FA) produced by the catalytic hydrogenation of CO₂ is recognized as a potential intermediate H ₂ carrier. Herei

Full text of DOI:10.1002/cctc.201402119

CHEMCATCHEM
COMMUNICATIONS
DOI: 10.1002/cctc.201402119
Highly Efficient Reversible Hydrogenation of Carbon
Dioxide to Formates Using a Ruthenium PNP-Pincer
Catalyst
Georgy A. Filonenko,[a, b] Robbert van Putten,[a] Erik N. Schulpen,[a] Emiel J. M. Hensen,[a, b] and
Evgeny A. Pidko*[a, b]
The use of hydrogen as a fuel requires both safe and robust
technologies for its storage and transportation. Formic acid
(FA) produced by the catalytic hydrogenation of CO2 is recog-
nized as a potential intermediate H2 carrier. Herein, we present
the development of a formate-based H2 storage system that
employs a Ru PNP-pincer catalyst. The high stability of this
system allows cyclic operation with an exceptionally fast load-
ing and liberation of H2. Kinetic studies highlight the crucial
role of the base promoter, which controls the rate-determining
step in FA dehydrogenation and defines the total H2 capacity
attainable from the hydrogenation of CO2. The reported find-
ings show promise for the development of practical technolo-
gies that use formic acid as a hydrogen carrier.
namely, Ir and Ru complexes. The record in CO2 hydrogenation
activity was set in 2009 by Nozaki and co-workers for an iridi-
um pyridine-based PNP-pincer catalyst (Figure 1). This catalyst
allows turnover frequencies (TOFs) as high as 150000 hꢀ1 at
2008C and 50 bar.[6] Fujita et al. introduced a pH-switchable Ir
catalyst that attained very high rates in FA dehydrogenation
(TOF=228000 hꢀ1 at 908C) and, at the same time, showed
substantial activity in CO2 hydrogenation (TOF=53800 hꢀ1 at
808C and 50 bar).[7] Ru catalysts are generally less active. The
highest TOF of 95000 hꢀ1 reported to date was achieved
under supercritical conditions in the presence of ruthenium tri-
Molecular hydrogen is one of the most promising
candidates for the storage of renewable energy. How-
ever, its safe and economical storage and transporta-
tion requires substantial technological advances.
Among the many potential approaches,[1] the chemi-
cal binding of H2 to produce liquid fuels is recog-
nized as a promising solution for efficient and reversi-
ble H2 storage.[2] For example, formic acid (FA) gener-
ated by the catalytic reduction of CO2 is an attractive
liquid hydrogen carrier.[3] Its catalytic decomposition
produces CO-free H2 that can be used directly in fuel
cell applications.[4] If both reactions in Equation (1)
could be performed efficiently in a single vessel
using the same catalyst, one would obtain a formate-
based hydrogen battery as put forward conceptually
by the group of Beller.[5] Its practical implementation
Figure 1. State-of-the-art catalytic systems for reversible CO2 hydrogenation.
requires a step change in the catalytic activity of re-
versible CO2 hydrogenation.
So far, the best performance in the reversible hydrogenation
of CO2 was achieved with homogeneous noble metal catalysts,
methylphosphine complexes and a triethylamine (NEt3) base.[8]
Previously, we demonstrated that a Ru PNP-pincer catalyst 1,[9]
analogous to Nozaki’s system,[6] can hydrogenate CO2 in THF
with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) as a base pro-
[a] G. A. Filonenko, R. van Putten, E. N. Schulpen, Prof. Dr. E. J. M. Hensen,
Dr. E. A. Pidko
moter with a maximal rate of 22000 hꢀ1 [10]
.
Inorganic Materials Chemistry group
Beller et al. described a highly stable homogeneous system
that consisted of RuH2(dppm)2 (dppm=1,1-bis(diphenylphos-
phino)methane) as the catalyst and trialkylamine as the base
and demonstrated its worth in consecutive H2 storage/release
cycles with only a minor loss of activity.[5] Recently, non-noble
metal homogeneous Fe and Co catalyst have also been shown
to be active in FA dehydrogenation[11] and CO2 hydrogena-
tion.[12] However, these systems typically suffer from low per-
Department of Chemical Engineering and Chemistry
Eindhoven University of Technology
P.O. Box 513, 5600 MB Eindhoven (The Netherlands)
[b] G. A. Filonenko, Prof. Dr. E. J. M. Hensen, Dr. E. A. Pidko
Institute for Complex Molecular Systems (ICMS)
Eindhoven University of Technology
P.O. Box 513, 5600 MB Eindhoven (The Netherlands)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402119.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
formance or the necessity to use expensive chemicals as pro-
moters. Moreover, expensive polydentate phosphine ligands
used in these systems contribute greatly to the catalyst price.
This further impedes the potential commercial benefits of such
homogeneous catalysts even though they are based on earth-
abundant metals. In this study we report the development of
a highly stable temperature-switchable Ru-based system for re-
versible hydrogen storage with CO2, which shows unprece-
dented rates for H2 loading and release under mild conditions
(1, Figure 1).
As the starting point of this study, we evaluated the FA de-
hydrogenation activity of several Ru pincer catalysts known for
their high activity in the hydrogenation and dehydrogenation
of polar substrates[13] (see Section S2 in the Supporting Infor-
mation). The catalytic experiments were performed by using
a batch reactor under the continuous addition of FA. In this
way catalyst activity and stability could be investigated
simultaneously.
Figure 2. Temperature-switchable FA dehydrogenation with 1 in DMF/DBU
(conditions: DMF/DBU=30:5 mL, 1.42 mmol 1).
<6000 hꢀ1) if loaded with approximately 0.8 equivalents of FA
per DBU. After a rapid increase of the reaction temperature to
908C, intense gas evolution started and the reaction proceed-
ed steadily over time until the FA dosage was stopped. The
temperature was then lowered to 658C (508C in the last cycle).
This cycle was repeated four times. In all cycles, a reproducible
catalytic performance was achieved. Furthermore, this experi-
ment shows that the catalytic system is stable under a wide
range of reaction conditions.
The initial tests revealed the superior catalytic performance
of 1 for FA dehydrogenation independent of the reaction
medium. The best performance in terms of catalyst stability
and reaction rate was achieved if the reaction was performed
using DMF as a solvent (Table S1, Supporting Information).
Unlike in other polar solvents (propylene carbonate, ethanol),
no permanent catalyst deactivation was observed in this
medium, even under highly acidic conditions in the presence
of excess FA. In all experiments, the evolved gas was con-
firmed to contain an equimolar mixture of H2 and CO2 with no
detectable traces of CO. The combination of 1 with NEt3 in
DMF results in an exceptionally high rate of FA dehydrogena-
tion of 257000 hꢀ1 at 908C (Table 1). In a separate experiment
The nature of the base promoters had a significant influence
on the dehydrogenation activity of 1. To understand their role
in FA dehydrogenation, we investigated the kinetics of this re-
action in batch mode. The temperature dependence of the gas
evolution (H2/CO2 =1:1) as a function of substrate concentra-
tion (expressed for convenience as the acid-to-amine ratio,
AAR) is displayed in Figure 3. In the presence of NEt3, the reac-
tion rate increased with the consumption of the substrate (to-
wards lower AAR), which evidences the effective inhibition of
the catalytic reaction by the substrate. The activation energy
for FA decomposition in the presence of NEt3 was 74 kJmolꢀ1.
If DBU was used instead of NEt3, the reaction followed first-
order behavior with respect to formate and had an activation
Table 1. Selected results for continuous FA dehydrogenation with 1.
Entry
Solvent
Base
T [8C]
TOF [hꢀ1
]
TON
1
2
3
4
DMF
DMF
DMF
DMF
NEt3
NEt3
NHex3
DBU
65
90
90
90
46495
257000
256000
93100
n.d.[a]
326500
706500
310000
energy of 108 kJmolꢀ1
.
[a] not determined. Conditions: 1.42 mmol catalyst 1, 35 mL total volume,
base 33.5 mmol (nitrogen based).
The pronounced difference in the kinetics of FA dehydro-
genation for DBU and NEt3 suggests that the base is directly
involved in the elementary steps of the catalytic reaction. We
propose that the rate-determining step (RDS) is different for
the NEt3 and DBU cases. A mechanistic analysis is needed to
trace the origin of this phenomenon. Proposed catalytic cycles
for the dehydrogenation of FA and hydrogenation of CO2 are
shown in Scheme 1.[3,10,11] For both reactions, the key role of
the dihydrido complex 2 has been postulated earlier.[13] The
first step of the dehydrogenation reaction (blue arrows in
Scheme 1) is base-promoted H2 formation through the recom-
bination of the hydride ligand and a proton from BH+ (in
which B is a base). H2 release from the 2+ꢀH2 s-complex yields
the cationic complex 2+, which is converted to 3 by the coor-
dination of the formate anion. The catalytic cycle is closed by
the rearrangement of 3 into a transient species 3*, from which
CO2 evolves to result in the initial complex 2.
with low catalyst loading (0.35 mmol), a turnover number (TON)
of 1063000 was reached within several hours (Figure S10). The
loss of amine during operation at 908C represents a major
drawback of this system. However, it can be remediated by
using the less volatile trihexylamine (NHex3) base (entry 3,
Table 1). To the best of our knowledge, these results represent
the highest FA dehydrogenation activity reported to date.
If the non-nucleophilic DBU base is used in continuous FA
dehydrogenation, 1 shows a marginal activity in DMF at 658C,
whereas a high rate of over 90000 hꢀ1 and TON values of
310000 are achieved at 908C (entry 4, Table 1). This behavior
points to the possibility to control H2 evolution by varying the
reaction temperature in a relatively narrow range. The results
of a variable temperature experiment confirm this hypothesis
(Figure 2). Indeed, the catalyst showed a very low activity (TOF
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
rate. The activation energy determined in DMF/NEt3
(74 kJmolꢀ1) is in good agreement with the DFT-computed
value for the transformation of 3 to 2 (75 kJmolꢀ1).[10] If the
stronger DBU base is employed, the reaction rate is controlled
by the rate of the initial base-assisted H2 formation from 2 and
protonated DBU (2!2+, Scheme 1). The CꢀH bond cleavage
step has only a small, if any, influence on the overall kinetics.
This proposal is supported by the very similar and small kinetic
isotope effects for the dehydrogenation of both DCOOD (1.48)
and HCOOD (1.42).
In the hydrogenation reaction, CO2 insertion (2!3*) is
postulated to be the RDS.[15] This implies that the base primari-
ly influences the reaction thermodynamics. In this way, the
base determines the product yield and thus the H2 storage ca-
pacity. This is in agreement with the results of CO2 hydrogena-
tion experiments with 1 in DMF. The strength of the base pro-
moter does not influence the reaction rate (entries 1 and 2,
Table 2), but it has a substantial effect on the final AAR. In the
Figure 3. Kinetic traces for FA dehydrogenation with 1 in DMF at different
temperatures in the presence of a) NEt3 and b) DBU. Conditions: 30 mL DMF,
9.43 mL NEt3, and 0.356 mmol 1 (a) or 5 mL DBU and 4.63 mmol 1 (b).
Table 2. Catalytic CO2 hydrogenation with 1.
Entry Solvent Base Catalyst [mmol] T [8C] P [bar] TOF[a] [hꢀ1
]
AAR
H2/CO2
1
2
3
4
5
6
DMF
DMF
DMF
DMF
DMF
DMF
NEt3 1.42
DBU 1.42
DBU 0.356
DBU 0.178
DBU 1.42
DBU 1.42
65
65
65
120
90
65
20:20
20:20
30:10
30:10 1100000
2.5:2.5
2.5:2.5
34000
36000
65000
0.26
1.6
1
1.1
1.18
1
60000
7900
[a] See Supporting Information for kinetic data. Conditions: 30 mL sol-
vent, 33.4 mmol base (nitrogen based).
presence of NEt3, CO2 hydrogenation at 658C yields a relatively
low AAR of 0.26. This implies that more than 70% of the base
is not used for H2 storage. Consistent with the findings of
Beller and co-workers,[5] a decrease in the reaction temperature
increases the AAR up to 1 at room temperature (Figure 4a).
Despite the substantial reactivity in the H2 evolution reaction
of the DMF/NEt3 combination at room temperature, the low H2
capacity of the amine-based system at catalytically relevant
temperatures represents a major drawback.
Scheme 1. Proposed catalytic cycles for the hydrogenation of CO2 (red) and
FA dehydrogenation (blue) with catalyst 1.
H2 formation and CO2 release are candidate RDSs. The low
basicity of NEt3 facilitates H2 recombination. The reaction rate
in this case is controlled by the CꢀH cleavage step (3!2),
which does not NEt3 involve directly.[14] Consistent with this,
we observe a marginal effect of the substitution of HCOOH for
HCOOD (kinetic isotope effect=1.04). On the contrary, the use
of DCOOD led to a strong kinetic isotope effect of 3.47, which
indicates that the CꢀH cleavage step determines the reaction
Figure 4. Pressure and temperature dependence of the AAR for a) NEt3 and
b) DBU bases.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
A much higher FA yield can be obtained in the presence of
DBU. A reaction at 658C leads to an AAR of 1.6 with a TOF of
36000 hꢀ1. In agreement with previous reports,[13b,16] the activi-
ty can be further enhanced at higher H2 partial pressures
(entry 3, Table 2). At a H2/CO2 molar ratio of 3:1 (Ptotal =40 bar),
a TOF of 65000 hꢀ1 was obtained at 658C, which increases to
1100000 hꢀ1 if the reaction is performed at 1208C (entry 4,
Table 2). This rate is nearly one order of magnitude higher than
the current record value reported for Nozaki’s Ir PNP-pincer
catalyst, which operates at a higher temperature and pressur-
e.[6a] In this high-temperature experiment, the kinetic trace fol-
lowed first-order behavior with respect to product formation
(Figure S11).
(40 bar) and low-pressure (5 bar) loading procedures, respec-
tively. The complete H2 liberation time did not exceed 1 h, at
which gas evolution peaked at over 160 mLminꢀ1, which corre-
sponds to TOF values higher than 150000 hꢀ1. H2/CO2 charging
times were less than 3 h, which depends on the temperature
program. The cyclic operation was performed without the ad-
dition of extra base between the cycles, which was necessary
for the stable performance of the amine-based system report-
ed previously.[5] No catalyst deactivation was observed in the
course of these experiments.
To conclude, we report the development of a highly active
system for the reversible hydrogenation of CO2. To the best of
our knowledge, the catalytic activities obtained for CO2 hydro-
genation and FA dehydrogenation reactions are the highest re-
ported to date. If used in combination with DBU, catalyst 1
allows the control of the hydrogen liberation activity in
a narrow temperature interval. Our results point to the key
role of the base strength to determine the H2 capacity of the
system under catalytically relevant conditions. Namely, strong
bases are required to generate high AARs at elevated tempera-
ture if the reaction times are short. Base strength was found to
affect the nature of the RDS in FA dehydrogenation. Although
the CꢀH cleavage step controls the rate in the presence of
weak bases, the initial H2 recombination determines the rate if
the reaction is performed in the presence of strong bases. The
reported system holds promise for the development of a practi-
cal H2 storage technology based on the reversible catalytic hy-
drogenation of CO2.
With the exceptional performance of 1 in DMF/DBU we
were able to hydrogenate CO2 at a relatively low pressure (en-
tries 5 and 6, Table 2). A high reaction rate of 60000 hꢀ1 was
obtained at a total pressure of only 5 bar of an equimolar
H2/CO2 mixture at 908C. This is superior to the TOF achieved
by the Fujita Ir catalyst[7] (Figure 1) at a comparable tempera-
ture of 808C and a 10-fold higher pressure. Unlike the DMF/
NEt3 system, the increase of the reaction temperature leads to
the stabilization of a higher AAR in DMF/DBU (Figure 4). The
H2 capacity of the DMF/DBU medium also depends strongly on
the total pressure. A stepwise increase of the pressure from 5
to 40 bar led to higher formate concentrations. A maximum
AAR of 2.1 was obtained at 908C and 40 bar. These data exem-
plify the crucial difference between DBU- and NR3-based sys-
tems. The high capacity of the latter can only be achieved at
low temperatures at the expense of the reaction rate. With
a molar volume similar to that of NEt3, DBU offers a nearly six-
fold higher FA loading capacity at 658C.
Experimental Section
All manipulations unless otherwise stated were performed using
Schlenk techniques. Catalytic hydrogenation tests and cycling ex-
periments were performed by using a 100 mL stainless-steel auto-
clave equipped with a gas compensation device and flowmeter for
evolved gas detection. CO2 hydrogenation was performed at a con-
stant pressure. Samples were withdrawn by using a dip tube instal-
lation and analyzed immediately by HPLC and GC with flame ioni-
zation detection (FID). The reaction was triggered by the addition
of the catalyst to a pressurized and preheated vessel. TOF values
were determined at the initial reaction stage if possible.
To further investigate the possibility of cyclic operation with
1 in DMF/DBU, we performed a series of hydrogen storage–re-
lease cycles over a week. Alternating high- and low-pressure
loading procedures were employed to evaluate the sensitivity
of the system to the variation of the operating conditions. The
results of these measurements are summarized in Figure 5. In
all cycles, the evolved H2/CO2 gas volumes were consistent
with AARs measured by direct sampling of the reaction mix-
ture. AAR values of 1.6 and 1.1 were observed for high-
Dehydrogenation reactions were performed by using a double-
lined glass reactor and syringe pump for FA supply. Gas evolution
was analyzed by using a Bronkhorst flowmeter or a foam flowme-
ter. The gas composition was analyzed by GC with thermal conduc-
tivity detection (TCD) and verified to be H2/CO2 =1:1 with no de-
tectible traces of CO. TOF values were estimated from the gas evo-
lution rate for both gas-detection methods.
Full experimental procedures for hydrogenation, dehydrogenation,
continuous storage, and activation energy determination are given
in the Supporting Information.
Acknowledgements
Figure 5. Total gas evolution in the H2 storage–release cycles with 1 in DMF/
DBU. Storage was performed at 658C under 40 bar (dark bars) and 5 bar
(light bars) H2/CO2 =1:1. Release was performed after decompression of the
system at 658C followed by heating to 908C (conditions: DMF/DBU=
30:5 mL, 1.42 mmol 1).
E.A.P. gratefully acknowledges the Technology Foundation STW
and the Netherlands Organization for Scientific Research (NWO)
for his personal VENI grant.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
[8] P. Munshi, A. D. Main, J. C. Linehan, C.-C. Tai, P. G. Jessop, J. Am. Chem.
Soc. 2002, 124, 7963–7971.
Keywords: homogeneous catalysis
·
hydrogenation
·
[9] B. Gnanaprakasam, J. Zhang, D. Milstein, Angew. Chem. Int. Ed. 2010, 49,
1468–1471; Angew. Chem. 2010, 122, 1510–1513.
hydrogen · reaction mechanisms · ruthenium
[10] G. A. Filonenko, M. P. Conley, C. Copꢃret, M. Lutz, E. J. M. Hensen, E. A.
Pidko, ACS Catal. 2013, 3, 2522–2526.
[11] T. Zell, B. Butschke, Y. Ben-David, D. Milstein, Chem. Eur. J. 2013, 19,
8068–8072.
[12] M. S. Jeletic, M. T. Mock, A. M. Appel, J. C. Linehan, J. Am. Chem. Soc.
2013, 135, 11533–11536.
[13] a) X. Yang, ACS Catal. 2011, 1, 849–854; b) A. Urakawa, F. Jutz, G. Lau-
renczy, A. Baiker, Chem. Eur. J. 2007, 13, 3886–3899.
[14] S. Oldenhof, B. de Bruin, M. Lutz, M. A. Siegler, F. W. Patureau, J. I. van
der Vlugt, J. N. H. Reek, Chem. Eur. J. 2013, 19, 11507–11511.
[15] a) Y.-Y. Ohnishi, T. Matsunaga, Y. Nakao, H. Sato, S. Sakaki, J. Am. Chem.
Soc. 2005, 127, 4021–4032; b) Y. Musashi, S. Sakaki, J. Am. Chem. Soc.
2000, 122, 3867–3877.
[1] a) U. Eberle, M. Felderhoff, F. Schꢁth, Angew. Chem. Int. Ed. 2009, 48,
6608–6630; Angew. Chem. 2009, 121, 6732–6757; b) J. Yang, A. Sudik,
C. Wolverton, D. J. Siegel, Chem. Soc. Rev. 2010, 39, 656–675.
[2] M. Yadav, Q. Xu, Energy Environ. Sci. 2012, 5, 9698–9725.
[3] a) A. Boddien, F. Gaertner, C. Federsel, P. Sponholz, D. Mellmann, R. Jack-
stell, H. Junge, M. Beller, Angew. Chem. Int. Ed. 2011, 50, 6411–6414;
Angew. Chem. 2011, 123, 6535–6538; b) A. Boddien, F. Gaertner, D. Mell-
mann, P. Sponholz, H. Junge, G. Laurenczy, M. Beller, Chimia 2011, 65,
214–218; c) S. Enthaler, ChemSusChem 2008, 1, 801–804; d) F. Joꢂ,
ChemSusChem 2008, 1, 805–808; e) C. Fellay, P. J. Dyson, G. Laurenczy,
Angew. Chem. Int. Ed. 2008, 47, 3966–3968; Angew. Chem. 2008, 120,
4030–4032.
[4] B. Loges, A. Boddien, H. Junge, M. Beller, Angew. Chem. Int. Ed. 2008, 47,
3962–3965; Angew. Chem. 2008, 120, 4026–4029.
[16] a) C. A. Huff, M. S. Sanford, ACS Catal. 2013, 3, 2412–2416; b) J. L.
Drake, C. M. Manna, J. A. Byers, Organometallics 2013, 32, 6891–6894.
[5] A. Boddien, C. Federsel, P. Sponholz, D. Mellmann, R. Jackstell, H. Junge,
G. Laurenczy, M. Beller, Energy Environ. Sci. 2012, 5, 8907–8911.
[6] a) R. Tanaka, M. Yamashita, K. Nozaki, J. Am. Chem. Soc. 2009, 131,
14168–14169; b) R. Tanaka, M. Yamashita, L. W. Chung, K. Morokuma, K.
Nozaki, Organometallics 2011, 30, 6742–6750.
Received: March 10, 2014
[7] J. F. Hull, Y. Himeda, W.-H. Wang, B. Hashiguchi, R. Periana, D. J. Szalda,
J. T. Muckerman, E. Fujita, Nat. Chem. 2012, 4, 383–388.
Published online on && &&, 0000
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
5
&
ÞÞ
These are not the final page numbers!

COMMUNICATIONS
G. A. Filonenko, R. van Putten,
E. N. Schulpen, E. J. M. Hensen,
E. A. Pidko*
Like pinning down an ant: An Ru PNP
catalyst provides unprecedented rates
of CO2 hydrogenation up to 1100000
h
ꢀ1 (turnover number >200000) and
&& – &&
excellent activity in the dehydrogena-
tion of formic acid in DMF if used with
the non-nucleophilic 1,8-diazabicyclo-
[5.4.0]undec-7-ene base. Unlike amine-
based systems, this medium allows us
to achieve high formate capacity to-
gether with rapid H2 charging and
release.
Highly Efficient Reversible
Hydrogenation of Carbon Dioxide to
Formates Using a Ruthenium PNP-
Pincer Catalyst
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 5
&
6
&
ÞÞ
These are not the final page numbers!