Structure-reactivity relationships in the hydrogenation of carbon dioxide with ruthenium complexes bearing pyridinylazolato ligands

Article abstract of DOI:10.1002/chem.201204199

Pyridinylazolato (N-N') ruthenium(II) complexes of the type [(N-N')RuCl(PMe₃)₃] have been obtained in high yields by treating the corresponding functionalised azolylpyridines with [RuCl ₂(PMe₃)₄] in the presence of a base. ¹⁵N NMR spectroscopy was used to elucidate the electronic influence of the substituents attached to the azolyl ring. The findings are in agreement with slight differences in the bond lengths of the ruthenium complexes. Furthermore, the electronic nature of the azolate moiety modulates the catalytic activity of the ruthenium complexes in the hydrogenation of carbon dioxide under supercritical conditions and in the transfer hydrogenation of acetophenone. DFT calculations were performed to shed light on the mechanism of the hydrogenation of carbon dioxide and to clarify the impact of the electronic nature of the pyridinylazolate ligands. Copyright

Full text of DOI:10.1002/chem.201204199

FULL PAPER
DOI: 10.1002/chem.201204199
Structure–Reactivity Relationships in the Hydrogenation of Carbon Dioxide
with Ruthenium Complexes Bearing Pyridinylazolato Ligands
Keven Muller, Yu Sun, Andreas Heimermann, Fabian Menges,
Gereon Niedner-Schatteburg,* Christoph van Wꢀllen,* and Werner R. Thiel*[a]
Abstract: Pyridinylazolato (N–N’) ru-
thenium(II) complexes of the type
[(N–N’)RuCl(PMe3)3] have been ob-
tained in high yields by treating the
corresponding functionalised azolylpyr-
agreement with slight differences in the
bond lengths of the ruthenium com-
plexes. Furthermore, the electronic
nature of the azolate moiety modulates
the catalytic activity of the ruthenium
complexes in the hydrogenation of
carbon dioxide under supercritical con-
ditions and in the transfer hydrogena-
tion of acetophenone. DFT calculations
were performed to shed light on the
mechanism of the hydrogenation of
carbon dioxide and to clarify the
impact of the electronic nature of the
pyridinylazolate ligands.
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
idines with [RuCl2ACHTUNGTRENNUNG(PMe3)4] in the pres-
ence of a base. 15N NMR spectroscopy
was used to elucidate the electronic in-
fluence of the substituents attached to
the azolyl ring. The findings are in
Keywords: carbon dioxide · hydro-
genation · nitrogen heterocycles ·
phosphanes · ruthenium
Introduction
drogen-storage compounds for several reasons: the hydro-
gen content of formic acid is quite high (approx. 4.3 wt%),[5]
formic acid and formates are non-toxic liquids or solids that
facilitate storage and transportation, and even simple palla-
dium on charcoal catalyses the back-reaction to dihydrogen
and carbon dioxide or carbonates. Last but not least, the di-
hydrogen used for the reduction of carbon dioxide is com-
pletely transferred into the formic acid/formate molecule,
which is in contrast, for example, to the formation of meth-
ane from carbon dioxide in which two equivalents of dihy-
drogen end up in water. This makes formic acid interesting
for hydrogen storage for fuel cells running small devices.
It is therefore not surprising that the hydrogenation of
carbon dioxide has been investigated for more than two dec-
ades. The hydrogenation of carbon dioxide over nickel cata-
lysts to yield methane (the Sabatier process), beneficially
carried out with hydrogen from renewable resources, pro-
vides access to a molecule that is widely used as an energy
source but also as a substrate in industry. Although hetero-
geneous copper, nickel and gold catalysts have mainly been
considered to be capable of reducing carbon dioxide to
carbon monoxide, methanol, methane or alkanes,[6] the for-
mation of formic acid by heterogeneous catalysis has recent-
ly been reported. Fachinetti and co-workers described the
formation of formic acid from carbon dioxide over a hetero-
geneous gold catalyst[7] and Ma and Zheng and their co-
workers reported the use of heterogeneous [Ru(OH)3] and
To solve the problems of climate change and declining fossil
resources, investigations into alternative carbon sources are
of the utmost importance. For this purpose, carbon dioxide
is worthy of consideration as an omnipresent, renewable
and inexpensive C1 carbon source.[1] At about 146ꢀ
106 tonnes per year, the production of urea is the most im-
portant industrial process based on carbon dioxide, followed
by the synthesis of salicylic acid (Kolbe–Schmitt reaction)
and the synthesis of poly(propylene carbonate).[2]
During the last decade, methods have been devised to
transform carbon dioxide directly into organic molecules.[3]
Although most effort has been focussed on the hydrogena-
tion of this molecule, it should not be forgotten that dihy-
drogen is currently mainly generated from fossil sources.
This may change in the future when cheap dihydrogen
might become available from chemical or electrochemical
water splitting or other benign processes.
In principle, a series of C1 products, such as formic acid
and formates, formaldehyde, carbon monoxide, methanol or
methane, are accessible from carbon dioxide,[4] although
some of these reactions are thermodynamically unfavoura-
ble. Formic acid and formates are of special interest as hy-
[RuCl2ACTHNUGTRNE(UNG PPh3)], respectively, immobilised on MCM-41 to cat-
[a] Dipl. Chem. K. Muller, Dr. Y. Sun, A. Heimermann, F. Menges,
G. Niedner-Schatteburg, C. van Wꢁllen, Prof. Dr. W. R. Thiel
Fachbereich Chemie, Technische Universitꢂt Kaiserslautern
Erwin-Schrçdinger-Str. 52–54, 67663 Kaiserslautern (Germany)
Fax: (+49)631 2054676
alyse this reaction.[8] It is therefore not surprising that ruthe-
nium compounds are the catalysts of choice in the homoge-
neous hydrogenation of carbon dioxide,[9] although some
quite active rhodium and iridium compounds have also been
reported.[10]
E-mail: thiel@chemie.uni-kl.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201204199.
Chem. Eur. J. 2013, 19, 7825 – 7834
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7825

The pioneering concepts relating to the hydrogenation of
carbon dioxide to formic acid and formates have been sum-
marised in a series of reviews.[4,9,11] These results also in-
spired investigations into the catalytic reduction of carbon
dioxide to methanol with silanes[12] and the development of
catalysts for the splitting of formic acid into carbon dioxide
and hydrogen.[13]
The formation of HCOOH from carbon dioxide and hy-
drogen is an endergonic reaction (CO2 +H2ÐHCOOH;
DG273K =32.9 kJmolꢀ1).[14] Thus, bases are used to shift the
chemical equilibrium to the product by the formation of for-
mates. With the help of NEt3, Inoue et al. succeeded in the
first homogeneous hydrogenation of carbon dioxide cata-
lysed by Wilkinsonꢄs catalyst,[15] which was further improved
in the 1990s by Leitner and co-workers.[16] A landmark in
this field was set by Noyori and co-workers, who used ruthe-
nium(II) complexes such as [RuCl
2ACHTUNGTNERN(UNG PMe3)4] or [RuH2-
ACHTUNGTRENNUNG
bers (TONs) of up to 7200 could be reached due to the high
solubility of H2 in this solvent, and a four-fold increase in
(PMe3)4].[18]
TON was achieved with [RuClACTHNUGRTENNUG(OAc)HCATUNGTRENNGUN
During the last decade iridium catalysts have received in-
creasing interest in the hydrogenation of carbon dioxide. Be-
tween 2007 and 2011, a number of iridium catalysts were re-
ported to exhibit TONs above 200000 in aqueous solution
with KOH as the base,[19] which was surpassed by Nozaki
and co-workers, who reported an iridium pincer complex
reaching a TON of 3.5 million under similar conditions.[20]
Fukuzumi and Fujita and their co-workers showed that this
reaction is possible with iridium at ambient pressure and
temperature.[21] In 2011, Milstein and co-workers reported a
pincer-type dihydrido iron complex that showed good activi-
ty in the hydrogenation of carbon dioxide in the presence of
NaOH.[22]
We have been interested in elucidating the subtle struc-
ture–reactivity relationships in catalysis for quite some
time.[23] Pyrazole-based ligands have turned out to be ideal
motifs for this purpose because the synthetic pathways lead-
ing to these ligands allow the introduction of electron-with-
drawing and -donating substituents without changing the
steric properties of the ligand. With a series of such ligands
in hand, it is therefore possible to gain a deeper insight into
the electronic influence of a ligand on the activity of the cat-
alyst separate from steric considerations. Thus, we report
herein the synthesis and characterisation of ruthenium cata-
Scheme 1. Synthesis of the azolylpyridine ligands 1a–j. Reagents and
conditions: i) HC(OMe)2NMe2, reflux, 6 h; ii) N2H4, EtOH, reflux, 5 h;
iii) MeC(O)R, NaOEt, 0–708C, 20 min; iv) Br2, CH3COOH, RT, 15 min;
v) HNO3, H2SO4, 0–908C, 4 h; vi) HCOOH, 0–258C, 1 h, then reflux, 4 h;
vii) RCOCl, NaOH, DMF, 0–258C, 5 h; viii)>2008C, 5 min.
AHCTUNGTRENNUNG
three-step procedure starting from 2-cyanopyridine.[24] Un-
substituted 2-(1,2,4-triazol-5-yl)pyridine (1g) was obtained
by heating the intermediate picolinimidohydrazide at reflux
in concentrated formic acid.[25]
The direct synthesis of the corresponding ruthenium com-
plexes by simple ligand exchange with the precursor [RuCl2-
AHCTUNTGRENGUN(N PMe3)4] gave only poor yields. Pyrazolylpyridines are
known to undergo intramolecular hydrogen bonding be-
ꢀ
tween the N H group of the pyrazole and the pyridine ni-
trogen atom,[23b,26] which will hinder coordination to the
ruthenium(II) site. Therefore deprotonation of the ligands
prior to the ligand-exchange reaction was performed with a
slight excess of 1,8-diazabicycloundec-7-ene (DBU) as the
base. Heating a solution of the deprotonated ligand in ace-
lysts of the type [(N–N’)RuClACHTNUGTRNEUNG(PMe3)3] and their perform-
ance in the hydrogenation of carbon dioxide.
tonitrile with [RuCl
(by NMR spectroscopy) intense orange to red monochlorido
complexes of the type [(N–N’)RuCl(PMe3)3] (2a–f, N–N’=
2ACHTUNGTERN(NUNG PMe3)4] at reflux gave quantitatively
AHCTUNGTRENNUNG
Results and Discussion
the pyridinylazolato ligand; Scheme 2). Slowly removing the
solvent under vacuum directly led to the formation of single
crystals, which in most cases were suitable for X-ray diffrac-
tion analysis (see below).
The coordination of the azolylpyridine ligands 1a–j by
ruthenium resulted in the expected changes in their
1H NMR spectra. The resonance arising from the proton at
the 1-position (for assignment see Scheme 1) of the pyridin-
Synthesis and characterisation of the ruthenium catalysts:
The pyrazole-based ligands 1a–f used for coordination to
the ruthenium centre were accessed by established proce-
dures following Claisen or Claisen-type condensations and
subsequent ring closing with hydrazine (Scheme 1),[23a,b,e]
whereas the triazolylpyridines 1h–j were synthesised in a
7826
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7825 – 7834

Hydrogenation of Carbon Dioxide
FULL PAPER
Figure 1. 1D 15N HMBC NMR spectra of 1e (left) and 2e (right).
Scheme 2. Synthesis of the ruthenium complexes 2a–j. Reagents and con-
ditions: i) DBU, CH3CN, 808C, 10 min; ii) [RuCl2ACTHUNGTRENUNG(PMe3)4], CH3CN,
reflux, 1 h.
yl ring shifts by 0.6–0.7 ppm to a lower field compared with
the free ligand. All other 1H NMR resonances are only
slightly shifted to a higher field. In accord with a meridional
arrangement of the three PMe3 ligands, the 31P NMR spectra
show the typical AB2 coupling pattern with a triplet for one
equatorial PMe3 moiety at 13–14 ppm and a doublet for the
two chemically identical PMe3 ligands in axial positions at
about ꢀ2 to ꢀ5 ppm.
Although the influence of the azolylpyridine substitution
pattern on the 31P NMR chemical shifts is rather weak,
strong effects were observed in the 15N HMBC NMR experi-
ments (HMBC=heteronuclear multiple bond correlation).
The discussion here shall be limited to a few examples; an
extended set of data is summarised in Table 1. The
Table 1. 15N HMBC NMR data for selected ligands and ruthenium(II)
complexes in CDCl3.[a]
Compound
d [ppm]
1
Figure 2. 2D 15N HMBC NMR spectrum of 1a (f1: H; f2: 15N).
Na.
Nb
Ng
Nd
1a, tautomer A
1a, tautomer B
ꢀ109.29
ꢀ102.17
ꢀ101.73
ꢀ102.82
ꢀ152.85
ꢀ172.48
ꢀ156.54
ꢀ137.02
ꢀ150.40
ꢀ151.35
ꢀ112.71
ꢀ200.04
ꢀ108.03
ꢀ123.15
ꢀ159.38
ꢀ145.09
ꢀ151.47
ꢀ149.09
ꢀ158.13
ꢀ160.17
ꢀ194.44
ꢀ114.13
ꢀ195.20
ꢀ189.84
ꢀ69.19
ꢀ77.37
ꢀ69.13
ꢀ72.04
ꢀ83.14
ꢀ84.56
–
–
–
1e
1h
2a
2c
2e
2 f
2g
2h
ꢀ163.31
for example, there are two tautomers visible in a ratio of ap-
proximately 4:1. The signals marked A in Figure 2 corre-
spond to the thermodynamically most stable form in which
the proton at the nitrogen atom at the 2-position of the pyr-
azole ring interacts with the nitrogen atom of the pyridine
ring.[23b] The signals marked B correspond to the thermody-
namically less stable tautomer in which the proton is located
on the nitrogen at the 1-position of the pyrazole ring.
–
–
–
–
not obs.
not obs.
[a] For the atom numbering, see Scheme 2.
Coordination to the ruthenium(II) site leads to a consider-
able shift in the 15N NMR resonances. For the ruthenium
complex 2e derived from ligand 1e, the resonance of the
pyrazole nitrogen atom Ng at the 2-position is observed at
ꢀ69.13 ppm (Figure 1b). The two other 15N resonances are
shifted upfield by around 100 ppm. This is a general finding
for all the pyridinylpyrazolate complexes. For the triazolyl-
pyrazolate complexes 2g–j, the resonances of the two coor-
dinating nitrogen atoms are detected at about ꢀ150 and
ꢀ160 ppm. The nitrogen atom at the 2-position of the triazo-
late ring gives a peak at around ꢀ85 ppm, whereas the nitro-
gen atom at the 4-position was not observed.
15N HMBC NMR spectrum of ligand 1e (Figure 1a) shows
the three expected resonances for the nitrogen atom Na
(for the atom numbering see Scheme 2) of the pyridine ring
at ꢀ101.73 ppm, which is close to the resonance at
ꢀ108.03 ppm of the nitrogen atom Nb at the 1-position of
the pyrazole ring, and at ꢀ195.20 ppm for the pyrazole ni-
trogen atom Ng at the 2-position of the pyrazole ring. The
situation may become more complicated in some cases as a
result of temperature-dependent tautomerism: in the
15N HMBC NMR spectrum of ligand 1a recorded at 229 K,
Chem. Eur. J. 2013, 19, 7825 – 7834
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7827

G. Niedner-Schatteburg, C. van Wꢁllen, W. R. Thiel et al.
ꢀ
Table 2. Relative intensities[a] of the species observed in the ESI-MS
spectra of the complexes 2a, 2b, 2e and 2 f.
In the IR spectra of the azolylpyridines the N H stretch-
ing frequency appears between 3200 and 3100 cmꢀ1. This ab-
sorption is of course absent in the IR spectra of the rutheni-
um complexes. The absorptions of the PMe3 methyl groups
are typically observed at 2974 and 2910 cmꢀ1, and a very
strong signal at 934 cmꢀ1 can be assigned to a scissoring vi-
bration of the methyl groups on the axially coordinated
PMe3 ligands.
Species
Relative intensity [%]
2a
2b
2e
2 f
A
3
5
0
79
8
0
6
5
0
76
8
1
7
11
0
67
8
0
65
15
9
5
1
[MꢀCl+CH3CN]+
[M]+
A
[MꢀPMe3+CH3CN]+
[MꢀPMe3]+
0
0
To clarify the influence of the substituents at the azolyl
units on catalyst stability independent of the influence of
the solvent, some of the ruthenium pyrazolylpyridine com-
plexes (2a, 2b, 2e and 2 f) were investigated by mass spec-
trometry (ESI–MS). There are two ways to generate mono-
cationic species from the neutral precursors. First, a chlorido
ligand can be removed leading to the 16-valence-electron
2
1
1
[a] Missing percentages are due to unassignable species with relative in-
tensities of <1%.
in Figure 3 and selected bond lengths and angles are pre-
sented in Table 3. Directed by the different trans influences
of the pyridine and pyrazolate/triazolate moieties, the com-
plexes all adopt a distorted octahedral coordination environ-
ment with the three trimethylphosphane ligands oriented in
a meridional arrangement and the chloride ligand located
trans to the pyrazolate/triazolate donor. This is in agreement
with the 31P NMR data of the complexes. Depending on the
cation [(N–N’)RuACHTUNGTRENNUNG
an open coordination site for catalysis in solution. Alterna-
tively, an electron might be lost to give [(N–N’)RuCl-
ACHTUNGTRENNUNG
(PMe3)3]+ ([M]+). These processes may be accomplished by
loss of PMe3 and/or uptake of a molecule of acetonitrile
(the solvent). As expected, loss of the chelating pyridylpyra-
zolate was not observed (see the Supporting Information for
the corresponding spectra). The relative intensities of the
different ionisation processes are presented in Table 2.
ꢀ
donor strengths of the azolate and pyridine sites, the Ru N2
ꢀ
distances are approximately 0.13 ꢅ shorter than the Ru N1
distances. There are slight differences in the bond lengths
depending on the nature and substitution pattern of the azo-
late ring: the complexes containing triazolate moieties have
The modes by which the cationic species are formed
differ with the substitution pattern of the ruthenium pyridyl-
pyrazolate complexes. Compound 2 f, equipped with a
strongly electron-withdrawing nitro group at the 4-position
of the pyrazolate ring, mainly undergoes chloride loss fol-
lowed by the addition of acetonitrile. However, a large
number of ionic species are formed in the course of the ESI
ꢀ
longer Ru N2 bonds than those containing pyrazolate rings
and electron-donating substituents lead to a slight elonga-
ꢀ
tion of the Ru N2 bond (in the pyrazolate series:
C6H4OMe>C6H5 >Br>NO2).
process as
a
result of oxidation: [M]+ (15%),
Catalysis: The hydrogenation of carbon dioxide under super-
critical conditions was performed in a stainless-steel auto-
[MꢀPMe3+CH3CN]+ (5%) and [MꢀPMe3]+ (1%). For
compound 2e with a bromo
substituent at the same posi-
Table 3. Characteristic bond lengths [ꢅ] and angles [8] for the ruthenium(II) complexes 2c–i.
2c[a] 2e[a] 2 f[a] 2h[a]
2d 2g
bond lengths [ꢅ]
tion, simple loss of the chlorido
ligand is the dominant process.
The uptake of acetonitrile after
loss of the chlorido ligand
([MꢀCl+CH3CN]+ (11%)) and
oxidation are strongly sup-
2i
ꢀ
Ru1 Cl1
2.4547(8)
2.3372(8)
2.2776(8)
2.3543(9)
2.171(3)
2.031(3)
2.4534(7)
2.4639(5)
2.4387(6)
2.4427(4)
2.4419(7)
2.4574(7)
2.3451(8)
2.2787(9)
2.3453(8)
2.163(2)
2.034(3)
ꢀ
Ru1 P1
2.3451(8)
2.2838(8)
2.3419(7)
2.155(2)
2.039(2)
2.3491(5)
2.2892(5)
2.3425(5)
2.1497(14)
2.0218(15)
2.3553(7)
2.2840(6)
2.3422(6)
2.1386(19)
2.023(2)
2.3565(5)
2.2748(5)
2.3533(5)
2.1543(16)
2.0394(16)
2.3449(8)
2.2807(8)
2.3464(7)
2.168(2)
2.046(2)
ꢀ
Ru1 P2
ꢀ
Ru1 P3
pressed ([M]+
(0%) and
ꢀ
Ru1 N1
[MꢀPMe3+CH3CN]+
(8%)).
ꢀ
Ru1 N2
bond angles [8]
Cl1-Ru1-P1
Cl1-Ru1-P2
Cl1-Ru1-P3
The ESI-MS data for com-
pounds 2a and 2b show that
the cations obtained by dissoci-
ation of the chlorido ligand are
quite stable because the ten-
dency for the cation to take up
acetonitrile is not as pro-
nounced as for 2e.
86.73(3)
97.19(3)
84.78(3)
86.75(3)
96.00(3)
87.46(2)
91.99(6)
169.28(6)
92.08(3)
172.56(3)
87.29(6)
94.51(6)
93.17(3)
171.94(6)
94.59(6)
88.25(6)
90.33(6)
77.45(8)
87.28(2)
95.67(2)
87.64(2)
93.41(4)
170.81(4)
93.74(2)
171.64(2)
86.88(4)
91.71(4)
93.36(2)
170.92(4)
93.52(4)
86.80(4)
92.24(4)
77.40(6)
86.98(2)
95.57(2)
87.42(2)
93.13(5)
169.60(6)
93.55(2)
172.59(2)
87.34(5)
91.60(6)
91.82(2)
171.29(5)
94.81(6)
88.12(5)
93.03(6)
76.50(7)
88.63(2)
94.71(2)
87.42(2)
93.17(4)
170.28(5)
94.50(2)
172.03(2)
87.68(4)
91.88(5)
92.72(2)
171.87(4)
94.93(5)
85.64(4)
90.86(5)
77.15(6)
85.65(3)
97.47(3)
86.60(2)
91.18(7)
168.68(7)
92.01(3)
171.19(3)
87.73(7)
94.28(7)
93.15(3)
171.31(7)
93.85(7)
88.24(7)
92.50(7)
77.51(10)
86.35(2)
95.53(3)
87.74(2)
93.53(5)
170.13(7)
92.59(3)
171.71(3)
86.46(6)
92.13(7)
93.76(3)
170.81(6)
94.28(7)
88.11(5)
92.71(8)
76.64(8)
Cl1-Ru1-N1 94.00(7)
Cl1-Ru1-N2 171.07(7)
P1-Ru1-P2
93.21(3)
P1-Ru1-P3 169.90(3)
P1-Ru1-N1
P1-Ru1-N2
P2-Ru1-P3
87.75(7)
92.84(8)
93.28(3)
A number of the pyrazolyl-
and triazolylpyridine ruthenium
complexes gave crystals suitable
for X-ray structure analysis.
The molecular structures of
P2-Ru1-N1 168.80(7)
P2-Ru1-N2
P3-Ru1-N1
P3-Ru1-N2
N1-Ru1-N2
91.74(7)
87.39(7)
94.69(8)
77.07(10)
compounds 2c–i are presented [a] There are two independent molecules in the unit cell, the data of only one is presented here.
7828
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7825 – 7834

Hydrogenation of Carbon Dioxide
FULL PAPER
The TONs and TOFs of the
complexes studied in this work
are presented in Table 4. Com-
pared with catalyst 2a (Table 4,
entry 1) with a completely un-
substituted ligand, catalyst 2b,
which bears an electron-donat-
ing methyl substituent (+I
effect) at the 5-position of the
pyrazole ring, is clearly less
active (Table 4, entry 2). The in-
troduction of substituents with
a p system in conjugation with
the pyrazole moiety further
lowers the catalytic activity (2c
and 2d, Table 4, entries 3 and
4). It was therefore expected
that
electron-withdrawing
groups would lead to a better
performance. However, this is
not true for substituents at the
4-position: the catalysts 2e and
2 f (Table 4, entries 5 and 6),
which carry electron-withdraw-
ing substituents at this site,
show slightly lower activities
than 2a. The trend is similar for
the triazole series but the per-
formances of ruthenium com-
plexes with the same substitu-
tion pattern as the pyrazole
series are generally higher
(compare catalysts 2a–c with
catalysts 2g, 2h and 2j). It
seems that ligands equipped
with an electron-withdrawing
substituent show better catalyt-
ic activity. For comparison, the
phosphane complexes [RuCl2-
Figure 3. Molecular structures of the ruthenium(II) complexes 2c–i in the solid state. Two crystallographically
independent molecules were found in the unit cells of 2c, 2e, 2 f and 2h, only one of which is shown here. The
ethyl group of compound 2i is disordered over two positions.
Table 4. Catalyst screening for the hydrogenation of carbon dioxide.[a]
clave suitable for high pressures. To avoid the need for a
compressor to attain the desired carbon dioxide pressure, a
method using solid carbon dioxide was established and is de-
scribed in the experimental part of this paper. Often an
amine base is required to stabilise the generated formic
acid. The resulting salts can be cleaved at high temperatures
allowing the reuse of the amine after distillation of the raw
product.[27] Under the given reaction conditions, 1,8-
Catalyst
n
N
nACTHNGUTRENN(UG HCOOH)/nACHTGNUTRENNUNG
A
U
]
1
2
3
4
5
6
7
8
9
2a
2b
2c
2d
2e
2 f
2g
2h
2i
44
34
34
30
40
41
49
38
45
40
0.67
0.51
0.50
0.46
0.60
0.63
0.74
0.57
0.68
0.61
0.27
1.17
4300 1080
3200
2980
3000
800
740
750
4000 1000
4200 1050
4800 1200
diazabicycloACHTUNGTRENNUNG[5.4.0]undec-7-ene (DBU) gave the best per-
3700
920
formance compared with other amine bases, such as the fre-
quently used triethylamine.[18] It is well established in carbon
dioxide hydrogenation catalysis that sub-stoichiometric
amounts of a proton source promote the production of
formic acid. C6F5OH, with a pKa on the aqueous scale
below that of the protonated amine, was reported to provide
the best results.[18] Therefore we chose this additive.
4450 1110
10 2j
11
12
3670
1520
920
380
A
2A
C
2A(PMe3)4] 77
7600 1900
[a] Reagents and conditions: CO2 (20 g), p(H2, 293 K)=70 bar, ptotal
AHCTUNGTRENNUNG
-
Chem. Eur. J. 2013, 19, 7825 – 7834
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7829

G. Niedner-Schatteburg, C. van Wꢁllen, W. R. Thiel et al.
A
E
and energies for all the calculated structures are presented
in the Supporting Information. Note that the high partial
pressures (concentrations) of carbon dioxide and dihydrogen
were taken into account in the calculations of the free en-
thalpies (Gibbs free energies).
The hydrogenation cycle, shown in Scheme 3 exemplarily
for catalyst 2a, starts with the 16-valence-electron complex
[(N–N’)Ru(H)ACTHUNTRGNE(UNG PMe3)2] (A). Owing to the weaker trans in-
though the activity of the triphenylphosphane complex is re-
duced by a factor of about 2.8 compared with 2a, the tri-
ACHTUNGTRENNUNGmethylphosphane derivative shows an activity increased by
a factor of 1.7. We are currently carrying out quantum
chemical calculations that might help to differentiate be-
tween the electronic and steric influences of the substituents
on the nitrogen donor and the phosphanes.
To gain more experimental data on the electronic effects
of the pyridinylazolate ligands, the catalysts were also tested
in the transfer hydrogenation of acetophenone with 2-propa-
nol as the hydrogen source (Table 5). The complexes with
Table 5. Catalyst screening for the transfer hydrogenation of acetophe-
none with 2-propanol.[a]
Catalyst
Yield [%]
After 4 h
After 24 h
1
2
3
4
5
6
7
8
9
2a
2b
2c
2d
2e
2 f
2g
2h
2i
18
1
19
41
4
11
16
21
23
32
54
44
48
4
84
82
23
57
48
54
56
69
86
99
10
11
12
2j
G
2A
2A(PMe3)4]
C
CHTUNGTRENNUNG
[a] Reagents and conditions: 2-propanol (15 mL), catalyst (0.01 mmol),
tBuOK (0.25 mmol), acetophenone (5.0 mmol), T=808C.
Scheme 3. Proposed mechanism for the hydrogenation of carbon dioxide
with (pyridinylazolato)ruthenium catalysts. The transition states have
been omitted for clarity.
substituents possessing an aromatic substituent at the 5-posi-
tion (2c and 2d for the pyrazole series and 2j for the tria-
zole series) gave the best yields after 24 h. For comparison,
the brominated and nitrated pyrazole-based systems 2e and
2 f showed lower activity, as did the methylated species 2b,
which yielded almost no product.
fluence of pyridine versus azolato ligands, the hydrido
ligand is located trans to the pyridine unit. The energies cal-
culated for the analogous reaction mechanism starting with
the hydrido ligand trans to the azolato ligand are systemati-
cally higher than the energies determined for the structures
shown in Scheme 3. At several points of the mechanism, the
site of reactivity could change from trans to pyridine to
trans to the azolato unit and vice versa. This behaviour was
calculated for a series of steps for catalyst 2e; it was found
that the barriers to such isomerisation reactions are quite
high (see the Supporting Information). Figure 4 shows the
calculated free enthalpies for all intermediates and transi-
Quantum chemical calculations: To gain a better under-
standing of the electronic effects of the pyrazole-based lig-
ACHTUNGTRENNUNGands on the catalytic performance of the ruthenium com-
plexes, quantum chemical calculations were carried out.
There is broad agreement in the literature that hydrido
ruthenium complexes, either used directly or formed in situ,
are the active species in the hydrogenation of carbon diox-
ide to formic acid and its derivatives.[20,28] A complete calcu-
tion states generated from [(N–N’)Ru(H)ACHTNURTGNEUNG(PMe3)2] (A) bear-
lation of the mechanism based on the [Ru(H)
G
ing the hydrido ligand trans to the pyridine unit. Depending
on the nature of the pyridylazolato ligand, the coordination
of carbon dioxide is either slightly endergonic, for ligands
2e and 2 f bearing an electron-withdrawing group, or exer-
gonic, for ligands 2a and 2g. Carbon dioxide coordinates to
the ruthenium(II) site in intermediate B in a h2-coordination
lyst was carried out by Sakaki and co-workers.[29] They
found that the insertion of carbon dioxide into the Ru H
bond is the rate-determining step. The activation barrier for
this process decreases with increasing polarity of the solvent.
Our DFT study allowed elucidation of the mechanism of hy-
drogenation for four of the hydrido ruthenium catalysts (2a,
2e, 2 f and 2g) bearing different pyridinylazolate ligands.
The energetics of different substitution patterns are present-
ed below and details concerning the calculations are sum-
marised in the Experimental Section. The atom coordinates
ꢀ ꢀ
mode and adopts a bent structure with an O C O angle of
around 1408. With an activation barrier of around 25–
30 kJmolꢀ1, the hydrido ligand is transferred to the carbon
atom via a four-membered-ring transition state leading to an
O,H-coordinated formato ligand (C). Although the free en-
7830
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7825 – 7834

Hydrogenation of Carbon Dioxide
FULL PAPER
Figure 4. Calculated profiles of the free reaction enthalpies for the hydrogenation of carbon dioxide with the catalysts 2a (black), 2e (red), 2 f (green)
and 2g (blue). For the structures A–G, see Scheme 3.
thalpies of the transition state TSBC are largely independent
of the substitution pattern of the azolate moiety, there is a
clear influence on the structure C. This fact is even more
pronounced in the subsequent transition state TSCD for the
opening up of a free coordination site for the incoming dihy-
drogen; the ruthenium centre changes from an 18- to a 16-
valence-electron state. TSCD is the point of highest enthalpy
in the whole catalytic cycle (for the final step, see below) for
all ligand substitution patterns calculated. The absolute bar-
riers differ by about 15 kJmolꢀ1, although the differences in
the barriers relative to structure C are less pronounced. The
overall reaction C!D is slightly exergonic except for cata-
lyst 2g, the only system bearing a pyridinyltriazolato ligand.
This step is followed by the strongly exergonic addition of
dihydrogen (ca. ꢀ40 kJmolꢀ1), which leads to the 18-va-
lence-electron system E. For the subsequent transfer of a
hydrogen atom to the formate anion, this ligand has to un-
gonic by about 55–65 kJmolꢀ1. In this step, ligands equipped
with an electron-withdrawing substituent have a lower barri-
er. The overall free enthalpy of the reaction is around
+55 kJmolꢀ1. In practical applications, an ammonium for-
mate is formed instead by addition of a base (here DBU).
This lowers the total enthalpy considerably and makes the
whole reaction sequence thermoneutral or slightly exother-
mic.
According to the DFT calculations, the detachment of the
formate hydrogen atom from the ruthenium site leading to a
16-valence-electron intermediate is the rate-determining
step. The calculated barriers show a clear dependence on
the ligand structure (DGRACHTNUGTRNEUNG(TSCD): 2a<2e<2 f<2g). In the
pyrazolylpyridine series (2a, 2e, 2 f), the barrier rises with
increasing electron-withdrawing effect of the substituent
(N<Br<NO2). This is not reflected in the observed differ-
ences in catalytic activity. The pyridinyltriazolato catalyst 2g
in reality shows a performance even better than 2a even
though its barrier for TSCD is higher. The structural effects
observed in the catalytic hydrogenation of carbon dioxide
must therefore not only be related to the electronic impact
of the substituents but also be due to further interactions
with the solvent and/or additives present in the catalytic re-
action. As mentioned above, the dissociation of the formate
ꢀ
dergo a rotation around the C O bond to bring the oxygen
atom into close proximity to the dihydrogen ligand. The ro-
tation is hindered by a barrier of about 17 kJmolꢀ1, almost
independent of the nature of the pyridinylazolato ligand,
and the final product F is stabilised by around 20 kJmolꢀ1
with respect to the starting structure E. There is again
almost no influence of the ligand substitution pattern on the
reaction enthalpy DGR(E)ꢀDGR(F). The following transfer
of a hydrogen atom proceeds via a late six-membered-ring
transition state. This reaction is endergonic by about
45 kJmolꢀ1 and the reaction barrier is around 38 kJmolꢀ1
and thus lower than the free enthalpy of the reaction. This
is due to the fact that the electronic energies and not the
free enthalpies were optimised, which excludes entropic ef-
fects. The final dissociation of formic acid from the inter-
mediate G to regenerate the hydrido complex A is ender-
from the ruthenium site leading to DBUH CHTNUGTRNENUG
+A(HCOOꢀ) is a
crucial step in which the interaction of the bulky DBU with
intermediate G is dependent on the size of the chelating
ligand.
Chem. Eur. J. 2013, 19, 7825 – 7834
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7831

G. Niedner-Schatteburg, C. van Wꢁllen, W. R. Thiel et al.
C23H37ClN3P3Ru·ACTHNUTRGNEUNG(CH3CN): C 47.96, H 6.44, N 8.95; found: C 48.05, H
6.60, N 9.08.
Conclusion
Chlorido[3-(4-methoxyphenyl)-5-(pyridin-2-yl)pyrazolato]tris(trimethyl-
phosphine)ruthenium(II) (2d): Yield: 30%. Elemental analysis calcd (%)
for C24H39ClN3OP3Ru: C 46.87, H 6.39, N 6.83; found: C 46.70, H 6.28, N
6.73.
Azolylpyridines coordinate to [RuCl2ACHTNUGTRENUNG(PMe3)4] in the pres-
ence of a base to give the corresponding pyridinylazolato
complexes [(N–N’)RuCl(PMe3)3] in which the chloride
AHCTUNGTRENNUNG
ligand is trans to the azolato site. The ligands and ruthenium
complexes were completely characterised by a number of
techniques, including 15N NMR spectroscopy. ESI mass spec-
trometry showed that the ruthenium complexes follow dif-
ferent ionisation and fragmentation pathways depending on
the substitution pattern of the azolato moiety. These differ-
ences were also found in the catalytic hydrogenation of
carbon dioxide to formates under supercritical conditions.
[4-Bromo-5-(pyridin-2-yl)pyrazolato]tris(trimethylphosphine)chloridoru-
thenium(II) (2e): Yield: 65%. Elemental analysis calcd (%) for
C17H32BrClN3P3Ru: C 34.74, H 5.49, N 7.15; found: C 34.89, H 5.64, N
7.40.
Chlorido[4-nitro-5-(pyridin-2-yl)pyrazolato]tris(trimethylphosphine)ru-
thenium(II) (2 f): Yield: 53%. Elemental analysis calcd (%) for
C17H32ClN4O2P3Ru: C 36.86, H 5.82, N 10.11; found: C 36.87, H 5.95, N
10.22.
Chlorido[5-(pyridin-2-yl)-1,2,4-triazolato]tris(trimethylphosphine)ruthe-
nium(II) (2g): Yield: 37%. Elemental analysis calcd (%) for
C16H32ClN4P3Ru: C 37.69, H 6.33, N 10.99; found: C 37.60, H 6.20, N
11.10.
Compared with a standard system such as [RuCl2ACTHNUTRGNEN(UG PMe)4],
the ruthenium complexes developed in this work showed
good activities that varied with the substituents on the che-
lating nitrogen ligand. In general, the unsubstituted ligands
gave the best performances with the triazolato system being
even more active than the pyrazolato complex. DFT calcula-
tions performed to ascertain the reaction mechanism al-
lowed the electronic influence of the substituents to be clari-
fied; the results showed good correlation with the catalytic
activities observed for the pyridinylpyrazolato series.
Chlorido[3-methyl-5-(pyridin-2-yl)-1,2,4-triazolato]tris(trimethylphos-
ACHUTNGRENpNUG hine)AHCTUNGTRENNrGUN uthenium(II) (2h): Yield: 39%. Elemental analysis calcd (%) for
C17H34ClN4P3Ru: C 38.97, H 6.54, N 10.69; found: C 38.54, H 6.79, N
10.81.
Chlorido[3-ethyl-5-(pyridin-2-yl)-1,2,4-triazolato]tris(trimethylphos-
ACHUTNGRENpNUG hine)AHCTUNGTRENNrGUN uthenium(II) (2i): Yield: 41%. Elemental analysis calcd (%) for
C18H36ClN4P3Ru: C 40.19, H 6.75, N 10.41; found: C 40.08, H 6.66, N
10.34.
Chlorido[3-phenyl-5-(pyridin-2-yl)-1,2,4-triazolato]tris(trimethylphos-
ACHUTNGRENpNUG hine)AHCTUNGTRENNrGUN uthenium(II) (2j): Yield: 24%. Elemental analysis calcd (%) for
C22H36ClN4P3Ru·ACTHNUTRGNENUG(CH3CN): C 45.97, H 6.27, N 11.17; found: C 46.10, H
6.20, N 11.30.
Catalytic hydrogenation of carbon dioxide: The reactions were carried
out in a stainless-steel autoclave (100 mL, Berghoff) suitable for pres-
sures of up to 200 bar and equipped with a Teflon lining and an internal
thermometer. DBU (10 g, 65 mmol), pentafluorophenol (20 mg,
0.14 mmol), dry, solid carbon dioxide (20 g) and the catalyst (0.01 mmol)
were introduced into the autoclave, which was closed and pressurised
with hydrogen (70 bar). Then the autoclave was placed in a preheated
aluminium block. After 15 min of heating the required inner temperature
of 1008C and an overall pressure of 170 bar were reached and maintained
for 4 h. The mixture was constantly stirred at 300 rpm. Then the auto-
clave was rapidly cooled to 58C in an ice bath, depressurised, opened
Experimental Section
General: All manipulations were carried out under an atmosphere of pu-
rified argon by using standard Schlenk techniques. Elemental analyses
were carried out at the Department of Chemistry, TU Kaiserslautern. IR
spectra were recorded with a Perkin-Elmer Spectrum 100 FT-UATR-IR
spectrometer equipped with a Diamond/ZnSe plate. NMR spectra were
recorded with Bruker DPX 400 and Avance 600 spectrometers. ESI-MS
spectra were recorded on a modified Bruker amaZonSL mass spectrome-
ter. The spectroscopic data of all compounds synthesised in this work are
reported in the Supporting Information. The pyridylpyrazole and -tria-
zole based ligands were prepared following previous reports in the litera-
ture.[23a–c,24,25] Ligand 1d was synthesised following the same process as
1
and the reaction mixture analysed by H NMR spectroscopy.
Catalytic transfer hydrogenation of acetophenone with 2-propanol: These
reactions were carried out in a two-necked flask (50 mL) equipped with a
reflux condenser and a Quickfit septum adapter. In this flask, the catalyst
(0.01 mmol) was dissolved in 2-propanol (15 mL) and heated to 808C.
KOtBu (26.8 mg, 0.25 mmol) and acetophenone (0.6 g, 5.0 mmol) were
then added to the mixture. Samples were taken every hour with a PE sy-
ringe and analysed by GC-MS.
ESI-MS: Sample solutions at concentrations of about 10ꢀ3 m were pre-
pared under oxygen-free conditions in LC-MS-grade acetonitrile and
stored at room temperature for some time to allow equilibration. The ion
source was used in the positive electrospray ionisation mode. A scan
speed of 32500 m/zsꢀ1 was used in ultra-scan mode (0.3 FWHM/m/z) and
the scan range was at least 70–800 m/z. Sample solutions were continu-
ously infused into the ESI chamber by means of a syringe pump at a flow
rate of 2 mLminꢀ1. Nitrogen was used as drying gas at a flow rate of
3.0 Lminꢀ1 at 2208C. The solutions were sprayed at a nebuliser pressure
of 4 psi (280 mbar) and the electrospray needle was held at 4.5 kV. In-
struments were controlled with the BrukerTrapControl 7.0 software and
data analysis was performed by using the Bruker Data Analysis 4.0 soft-
ware.
used for 1c. The ruthenium precursors [RuCl
(PMe3)4] were obtained from RuCl3 (Strem) according to literature pro-
cedures.[30,31]
2ACHTUNGTRNE(UNNG PPh3)3] and [RuCl2-
ACHTUNGTRENNUNG
General procedure for the synthesis of the ruthenium complexes 2a–j:
1,8-Diazabicycloundec-7-ene (DBU; 60 mL) was added to a solution of
the appropriate pyrazolyl- or triazolylpyridine 1a–j (0.4 mmol) in aceto-
nitrile (10 mL) and the mixture was heated at reflux for 10 min. After
cooling to room temperature, [RuCl2ACHTUNTGRNEUNG(PMe3)4] (190.4 mg, 0.4 mmol) was
added and the mixture was heated at reflux for 1 h . Then the solvent
was reduced to around 3 mL. After several hours, yellow to orange crys-
tals of the desired ruthenium complexes were separated from the liquid
and recrystallised from acetonitrile. All yields given below are for the re-
crystallised samples. The conversions were generally almost quantitative,
however, the DBUH+Clꢀ salts had to be separated.
Chlorido[5-(pyridin-2-yl)pyrazolato]tris(trimethylphosphine)rutheniu-
m(II) (2a): Yield: 23%. Elemental analysis calcd (%) for
C17H33ClN3P3Ru: C 40.12, H 6.54, N 8.26; found: C 40.02, H 6.43, N 8.30.
Chlorido[3-methyl-5-(pyridin-2-yl)pyrazolato]tris(trimethylphosphine)ru-
thenium(II) (2b): Yield: 22%. Elemental analysis calcd (%) for
C18H35ClN3P3Ru·ACHTUNGTRENNUNG(CH3CN): C 42.59, H 6.79, N 9.93; found: C 42.95, H
6.67, N 9.15.
X-ray structure analyses: Crystal data and refinement parameters for the
ruthenium(II) complexes 2c–i are presented in Table 6. The structures
were solved by direct methods (SIR92[32]), completed by subsequent dif-
ference Fourier syntheses and refined by full-matrix least-squares proce-
dures.[33] Semi-empirical absorption corrections (Multiscan) were carried
Chlorido[3-phenyl-5-(pyridin-2-yl)pyrazolato]tris(trimethylphosphine)ru-
thenium(II) (2c): Yield: 32%. Elemental analysis calcd (%) for
7832
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7825 – 7834

Hydrogenation of Carbon Dioxide
FULL PAPER
Table 6. Crystallographic data and parameters for data collection and refinement.
2c
2d
2e
2 f
2g
2h
2i
empir. formula
C23H37ClN3P3Ru C24H39ClN3OP3Ru C17H32BrClN3P3Ru C17H32ClN4O2P3Ru C16H32ClN4P3Ru C17H34ClN4P3Ru C18H36ClN4P3Ru
Mr
584.99
615.01
587.80
553.90
509.89
523.91
537.94
crystal size [mm] 0.47ꢀ0.19ꢀ0.17 0.25ꢀ0.23ꢀ0.18
0.19ꢀ0.17ꢀ0.16
150(2)
0.39ꢀ0.16ꢀ0.16
150(2)
0.22ꢀ0.14ꢀ0.13 0.27ꢀ0.05ꢀ0.04 0.44ꢀ0.16ꢀ0.14
T [K]
150(2)
150(2)
150(2)
150(2)
150(2)
l [ꢅ]
1.54184
triclinic
P1
1.54184
orthorhombic
P212121
10.7125(1)
11.4555(1)
23.6612(3)
90
0.71073
1.54184
orthorhombic
Pca21
18.4387(2)
8.5206(1)
30.7663(3)
90
1.54184
orthorhombic
Pbcn
22.8818(1)
13.1475(1)
15.0458(1)
90
1.54184
orthorhombic
Pca21
18.7398(2)
8.6339(1)
29.7087(3)
90
1.54184
monoclinic
P21/c
9.0284(1)
11.2453(2)
24.5818(4)
90
94.271(1)
90
2488.79(7)
4
1.436
7.990
3.61–62.64
20341
3987
crystal system
space group
a [ꢅ]
b [ꢅ]
c [ꢅ]
triclinic
¯
¯
P1
9.0227(3)
18.3572(6)
18.4391(6)
103.315(3)
103.116(3)
103.215(3)
2764.44(18)
4
1.406
7.233
3.06–62.68
21155
8820
11.3465(3)
15.2651(4)
15.8486(4)
108.613(2)
105.962(2)
98.240(2)
2419.40(11)
4
1.614
2.616
2.70–32.56
28909
15667
a [8]
b [8]
90
90
90
90
90
90
90
90
g [8]
V [ꢅ3]
Z
2903.63(5)
4
1.407
6.941
3.74–62.64
20995
4866.66(9)
8
1.522
8.313
4.80–62.66
39369
4526.35(5)
8
1.496
8.754
3.86–62.64
32904
4806.80(9)
8
1.448
8.259
2.97–62.66
35767
1calcd. [gcmꢀ3
]
m [mmꢀ1
]
q range [o]
reflns coll.
indep. reflns
4606
7649
3617
7394
[Rint=0.0314]
data/restr./param. 8820/0/577
[Rint=0.0293]
4606/0/308
0.0205, 0.0494
[Rint=0.0243]
15567/0/487
0.0238, 0.0430
[Rint=0.0283]
7649/1/524
0.0170, 0.0441
[Rint=0.0309]
3617/0/236
0.0199, 0.0519
[Rint=0.0340]
7394/1/490
0.0184, 0.0427
[Rint=0.0250]
3987/3/265
0.0270, 0.0674
final R indices
[I>2s(I)][a]
R indices
0.0320, 0.0874
ACHTUNGTRENNUNG
0.0350, 0.0890
0.0212, 0.0497
0.0384, 0.0441
0.0172, 0.0441
0.0207, 0.0524
0.0191, 0.0429
0.0284, 0.0681
ACHTUNGTRENNUNG(all data)
GooF[b]
1.058
1.038
0.875
1.064
1.120
1.009
1.042
Flack parameter
–
0.009(7)
–
0.270(4)
0.00079(2)
0.387/ꢀ0.453
0.569(5)
0.943/ꢀ0.259
–
D1max
/
[eꢅꢀ3
]
0.989/ꢀ0.742
0.166/ꢀ0.398
0.726/ꢀ0.670
0.383/ꢀ0.360
0.817/ꢀ0.756
min
2
2
2
[a] R1=Sj jFo jꢀjFc j j/SjFo j, wR2=[Sw
N
(Fo ꢀFc2)2/
N
.
out.[34] All non-hydrogen atoms were refined by using anisotropic dis-
placement parameters. All hydrogen atoms were placed in calculated po-
sitions and refined by using a riding model.
differences for the corresponding association steps. Therefore a much
higher partial pressure (estimated for the reaction conditions: 70 bar for
dihydrogen, 100 bar for carbon dioxide) was used to calculate the Gibbs
free energies for these two compounds.
CCDC-930861 (2c), CCDC-930862 (2d), CCDC-930863 (2e), CCDC-
930864 (2 f), CCDC-930865 (2g), and CCDC-930866 (2h), CCDC-930867
(2i) contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge from The Cambridge Crystal-
lographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgements
Quantum chemical calculations: All DFT calculations were performed
with the B3LYP functional[35] with the TZVP basis set[36] for all atoms
except ruthenium, for which a scalar relativistic effective core potential
replacing 28 core electrons[37] was used together with a def-TZVP[38] va-
lence basis set. Geometry optimisations for gas-phase species were per-
formed with a combination of TURBOMOLE[39] and the Gaussian 03
program package[40] (by using the EXTERNAL interface of the latter)
starting from the X-ray structures of the precatalysts. All minima and
transition states were characterised by frequency calculations, transition
states having exactly one negative Hessian eigenvalue. Intrinsic reaction
coordinate (IRC) calculations were carried out to verify that the transi-
tion states really connect the two minima with which they are associated.
In a second series of calculations, all the minima and transition states
found so far were reoptimised by using the Gaussian 09[41] program pack-
age and solvent effects were also taken into account by using the
COSMO[42] model with a dielectric constant e=1.3 used for supercritical
carbon dioxide, taken from Monte–Carlo molecular simulations.[43] The
Gibbs free energies were calculated by using the Gaussian 09 package for
all species in the catalytic cycle, as well as for dihydrogen and carbon di-
oxide, by using the harmonic approximation for the vibrational part, the
rigid rotor approximation for the rotational part and the free particle
model for the translational part of the partition function. A temperature
of 373 K and a standard partial pressure of about 1 bar (1 atm) were used
for ruthenium-containing species in the catalytic cycle. The concentra-
tions of dihydrogen and carbon dioxide were much higher under the re-
action conditions, which significantly influences the Gibbs free-energy
The authors thank the Deutsche Forschungsgemeinschaft (ERA Chemis-
try and SFB/TRR 88) and the National Research Fund (FNR), Luxem-
bourg, for financial support of this work.
[1] Carbon Dioxide as Chemical Feedstock (Ed.: M. Aresta), Wiley-
VCH, Weinheim, 2010.
[2] A. Bazzanella, D. Krꢂmer, M. Peters, Nachr. Chem. 2010, 58, 1226–
1230.
[3] T. Sakakura, J.-C. Choi, H. Yasuda, Chem. Rev. 2007, 107, 2365–
2387.
[4] P. G. Jessop, T. Ikariya, R. Noyori, Chem. Rev. 1995, 95, 259–272.
[5] a) F. Joꢆ, ChemSusChem 2008, 1, 805–808; b) S. Enthaler, J. von
Langermann, T. Schmidt, Energy Environ. Sci. 2010, 3, 1207–1217;
c) B. Loges, A. Boddien, F. Gꢂrtner, H. Junge, M. Beller, Top. Catal.
2010, 53, 902–914; d) M. Grasemann, G. Laurenczy, Energy Envi-
ron. Sci. 2012, 5, 8171–8181.
[6] F. C. Meunier, Angew. Chem. 2011, 123, 4139–4140; Angew. Chem.
Int. Ed. 2011, 50, 4053–4054.
[7] D. Preti, C. Resta, S. Squarcialupi, G. Fachinetti, Angew. Chem.
2011, 123, 12759–12762; Angew. Chem. Int. Ed. 2011, 50, 12551–
12554.
[8] a) C. Hao, S. Wang, M. Li, L. Kang, X. Ma, Catal. Today 2011, 160,
184–190; b) Y. Yu, J. Fei, Y. Zhang, X. Zheng, J. Fuel Chem. Tech-
nol. 2006, 34, 700–705.
Chem. Eur. J. 2013, 19, 7825 – 7834
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7833

G. Niedner-Schatteburg, C. van Wꢁllen, W. R. Thiel et al.
[9] P. G. Jessop, F. Joꢆ, C.-C. Tai, Coord. Chem. Rev. 2004, 248, 2425–
2442.
[10] C. Federsel, R. Jackstell, M. Beller, Angew. Chem. 2010, 122, 6392–
6395; Angew. Chem. Int. Ed. 2010, 49, 6254–6257.
[11] a) P. G. Jessop, The Handbook of Heterogeneous Hydrogenation,
Homogeneous Hydrogenation of Carbon Dioxide, Part III (Eds.:
J. G. de Vries, C. J. Elsevier), Wiley-VCH, Weinheim, 2008, Chap-
ter 17; b) W. Leitner, Angew. Chem. 1995, 107, 2391–2405; Angew.
Chem. Int. Ed. Engl. 1995, 34, 2207–2221.
2480–2487; b) R. A. Jones, F. M. Real, G. Wilkinson, A. M. R.
Galas, M. B. Hursthouse, K. M. A. Malik, J. Chem. Soc. Dalton
Trans. 1980, 511–518.
[32] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. C.
Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr. 1994, 27, 435–
435.
[33] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[34] a) CrysAlisPro, Version 1.171.33.48, Oxford Diffraction Ltd, Oxford,
2009; b) CrysAlisPro, Version 1.171.33.66, Oxford Diffraction Ltd,
Oxford, 2010.
[35] a) P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J.
Phys. Chem. 1994, 98, 11623–11627; b) A. D. Becke, J. Chem. Phys.
1993, 98, 5648–5652; c) A. D. Becke, Phys. Rev. A 1988, 38, 3098–
3100; d) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–
789.
[36] a) A. Schꢂfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100,
5829–5835; b) K. Eichkorn, F. Weigend, O. Treutler, R. Ahlrichs,
Theor. Chem. Acc. 1997, 97, 119–124.
[12] S. Riduan, Y. Zhang, J. Ying, Angew. Chem. 2009, 121, 3372–3375;
Angew. Chem. Int. Ed. 2009, 48, 3322–3325.
[13] a) A. Boddien, B. Loges, H. Junge, M. Beller, ChemSusChem 2008,
1, 751–758; b) S. Enthaler, ChemSusChem 2008, 1, 801–804; c) T. C.
Johnson, D. J. Morris, M. Wills, Chem. Soc. Rev. 2010, 39, 81–88.
[14] W. Wang, S. Wang, X. Ma, J. Gong, Chem. Soc. Rev. 2011, 40, 3703–
3727.
[15] Y. Inoue, H. Izumida, Y. Sasaki, H. Hashimoto, Chem. Lett. 1976,
863–864.
[16] a) E. Graf, W. Leitner, J. Chem. Soc. Chem. Commun. 1992, 623–
624; b) F. Gassner, W. Leitner, J. Chem. Soc. Chem. Commun. 1993,
1465–1466.
[17] a) P. G. Jessop, T. Ikariya, R. Noyori, Nature 1994, 368, 231–233;
b) P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc.
1996, 118, 344–355.
[37] D. Andrae, U. Hꢂußermann, M. Dolg, H. Stoll, H. Preuß, Theor.
Chim. Acta 1990, 77, 123–141.
[38] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–
3305.
[39] TURBOMOLE, V6.4 2012, a development of University of Karls-
ruhe and Forschungszentrum Karlsruhe GmbH, 1989–2007, TUR-
BOMOLE GmbH, since 2007; available from http://www.turbomo-
le.com.
[18] P. Munshi, A. D. Main, J. C. Linehan, C.-C. Tai, P. G. Jessop, J. Am.
Chem. Soc. 2002, 124, 7963–7971.
[19] a) Y. Himeda, N. Onozawa-Komatsuzaki, H. Sugihara, K. Kasuga,
Organometallics 2007, 26, 702–712; b) T. J. Schmeier, G. E. Dober-
einer, R. H. Crabtree, N. Hazari, J. Am. Chem. Soc. 2011, 133,
9274–9277; c) A. Azua, S. Sanz, E. Peris, Chem. Eur. J. 2011, 17,
3963–3967.
[20] R. Tanaka, M. Yamashita, K. Nozaki, J. Am. Chem. Soc. 2009, 131,
14168–14169.
[21] a) Y. Maenaka, T. Suenobu, S. Fukuzumi, Energy Environ. Sci. 2012,
5, 7360–7367; b) 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.
[40] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomer-
y, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyen-
gar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, 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.
Ochterski, 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. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, 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., Wall-
ingford CT, 2004.
[41] Gaussian 09, Revision B.1, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
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.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢇ.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian,
Inc., Wallingford CT, 2009.
[22] R. Langer, Y. Diskin-Posner, G. Leitus, L. J. W. Shimon, Y. Ben-
David, D. Milstein, Angew. Chem. 2011, 123, 10122–10126; Angew.
Chem. Int. Ed. 2011, 50, 9948 –9952.
[23] a) W. R. Thiel, M. Angstl, T. Priermeier, Chem. Ber. 1994, 127,
2373–2377; b) W. R. Thiel, J. Eppinger, Chem. Eur. J. 1997, 3, 696–
705; c) A. Hroch, G. Gemmecker, W. R. Thiel, Eur. J. Inorg. Chem.
2000, 1107–1114; d) D. Zabel, A. Schubert, G. Wolmershꢂuser,
R. L. Jones, Jr., W. R. Thiel, Eur. J. Inorg. Chem. 2008, 3648–3654;
e) K. Muller, A. Schubert, T. Jozak, A. Ahrens-Botzong, V. Schꢁne-
mann, W. R. Thiel, ChemCatChem 2011, 3, 887–892; f) L. Taghiza-
deh Ghoochany, S. Farsadpour, Y. Sun, W. R. Thiel, Eur. J. Inorg.
Chem. 2011, 3431–3437.
[24] E. Orselli, G. S. Kottas, A. E. Konradsson, P. Coppo, R. Frçhlich, L.
De Cola, A. van Dijken, M. Bꢁchel, H. Bçrner, Inorg. Chem. 2007,
46, 11082–11093.
[25] J. Roppe, N. D. Smith, D. Huang, L. Tehrani, B. Wang, J. Anderson,
J. Brodkin, J. Chung, X. Jiang, C. King, B. Munoz, M. A. Varney, P.
Prasit, N. D. P. Cosford, J. Med. Chem. 2004, 47, 4645–4648.
´
[26] V. Vetokhina, K. Dobek, M. Kijak, I. I. Kaminska, K. Muller, W. R.
Thiel, J. Waluk, J. Herbich, ChemPhysChem, 2012, 13, 3661–3671.
[27] a) Z. Zhang, S. Hu, J. Song, W. Li, G. Yang, B. Han, ChemSusChem
2009, 2, 234–238; b) J. Karl, A. Hank, BASF SE, Ludwigshafen,
2007, WO 2006/021411 A1, EP1784375 B1.
[28] a) T.-T. Thai, B. Therrien, G. Sꢁss-Fink, J. Organomet. Chem. 2009,
694, 3973–3981; b) R. Tanaka, M. Yamashita, L. W. Chung, K. Mo-
rokuma, K. Nozaki, Organometallics 2011, 30, 6742–6750.
[29] Y.-Y. Ohnishi, T. Matsunaga, Y. Nakao, H. Sato, S. Sakaki, J. Am.
Chem. Soc. 2005, 127, 4021–4032.
[42] A. Klamt, G. Schꢁꢁrmann, J. Chem. Soc. Perkin Trans. 2 1993, 799–
805.
[43] Y. Zhang, J. Yang, Y.-X. Yu, J. Phys. Chem. B 2005, 109, 13375–
13382.
[30] T. Stephenson, G. Wilkinson, J. Inorg. Nucl. Chem. 1966, 28, 945–
956.
[31] a) R. A. Jones, G. Wilkinson, I. J. Colquohoun, W. McFarlane,
A. M. R. Galas, M. B. Hursthouse, J. Chem. Soc. Dalton Trans. 1980,
Received: November 23, 2012
Revised: February 19, 2013
Published online: April 15, 2013
7834
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 7825 – 7834