Kinetic studies on the hydrogenation of carbon dioxide to formic acid using a rhodium complex as catalyst

Article abstract of DOI:10.1002/zaac.200800352

A detailed kinetic analysis has been performed on the hydrogenation of carbon dioxide with the complex [(dppp)2RhH] (dppp₂ 1,3-bis(diphenyl-phosphino)propane in DMSO using stopped-flow techniques. Activation parameters including the activation volume for this reaction as well as for the related reaction of the formation of the dihydrogen complex [(dppp)₂RhH₂]HCO₂ were obtained and thus allowed the formulation of a mechanism for this reaction cycle.

Full text of DOI:10.1002/zaac.200800352

DOI: 10.1002/zaac.200800352
Kinetische Untersuchungen zur Hydrierung von Kohlendioxid zu Ameisensäure
mit einem Rhodiumkomplex als Katalysator
Kinetic Studies on the Hydrogenation of Carbon Dioxide to Formic Acid
using a Rhodium Complex as Catalyst
Jens Dietrich and Siegfried Schindler*
Gießen / Germany, Justus-Liebig-Universität, Institut für Anorganische und Analytische Chemie
Received June 5th, 2008; accepted July 28th, 2008.
Dedicated to Professor Bernt Krebs on the Occasion of his 70th Birthday
Abstract. A detailed kinetic analysis has been performed on the
hydrogenation of carbon dioxide with the complex [(dppp)2RhH]
(dppp ϭ 1,3-bis(diphenyl-phosphino)propane in DMSO using
stopped-flow techniques. Activation parameters including the acti-
vation volume for this reaction as well as for the related reaction
of the formation of the dihydrogen complex [(dppp)2RhH2]HCO2
were obtained and thus allowed the formulation of a mechanism
for this reaction cycle.
Keywords: Rhodium; Carbon dioxide; Formic acid; Hydrogenation
Introduction
tive catalytic species, the unsaturated 14 electron complex
[Rh(dppb)H], is obtained in situ from the reaction of the
dimeric complex with the bidentate phosphane dppb (1,4-
bis(diphenylphosphino)-butane) and dihydrogen. Alterna-
tively different starting materials can be used according to
Scheme 1 [17, 27].
Possible use of carbon dioxide as a C1 building block in
synthetic processes did lead to increased research activity
trying to gain better understanding of the reactions of car-
bon dioxide at transition metal sites [1Ϫ15]. A promising
approach in that regard seems to be the reduction of carbon
dioxide to formic acid and its derivatives [16, 17]. So far,
formic acid is manufactured industrially mainly by the reac-
tion of carbon monoxide with methanol [18]. Alternatively
it would be interesting to reduce carbon dioxide electro-
chemically or photochemically to formate (see references
in [17]).
One approach is the synthesis of formic acid using tran-
sition metal complexes as homogeneous catalysts and a var-
iety of different catalytic systems have been investigated [16,
17, 19Ϫ24]. Of these systems rhodium complexes show a
promising potential, for example the hydrogenation of car-
bon dioxide with [Rh(PMe2Ph)3(nbd)(BF4)] (nbd ϭ nor-
bornadiene) in THF did lead to the formation of 64 mol
HCO2H per day and mol catalyst (small amounts of water
increased the activity of the catalyst) [25]. Larger turn over
numbers were achieved using [Rh(COD)μ-Cl]2 (COD ϭ
1,5-cyclooctadiene) in DMSO/NEt3 mixtures [26]. The ac-
* Prof. Dr. Siegfried Schindler
Institut für Anorganische und Analytische Chemie,
Justus-Liebig-Universität
Scheme 1
Heinrich-Buff-Ring 58
D-35392 Gießen, Germany
Fax: ϩ49-641-9934149
E-mail: siegfried.schindler@chemie.uni-giessen.de
Detailed investigations allowed to postulate a reaction
mechanism for the insertion of carbon dioxide and the for-
mation of formic acid [17, 26Ϫ32]. It was proposed that the
Z. Anorg. Allg. Chem. 2008, 634, 2487Ϫ2494
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2487

J. Dietrich, S. Schindler
pid mixing accessory or a stopped-flow unit (both from Applied
insertion of carbon dioxide is followed by a sequence of
oxidative and reductive elimination steps for the hydrogen
activation pathway [28]. Previously, an alternative pathway
for the hydrogen cleavage was suggested on the basis of ab
initio calculations [31]. Here no change of the oxidation
state of the rhodium centre is required, and the formation
of formic acid is a result of a [2 ϩ 2] addition of hydrogen
on the rhodium formate intermediate obtained after carbon
dioxide insertion [32].
Photophysics, Leatherhead, England) equipped with
a J&M
TIDAS 16-500 diode array spectrophotometer (J&M, Aalen, Ger-
many). Data fitting was either perfomed with the program Specfit
(Spectrum Software Associates, Marlborough, MA, U.S.A) or Igor
Pro (WaveMetrics, Lake Oswego, USA). The effect of pressure was
measured on a home made high-pressure stopped-flow instrument
or for slow reactions on a Shimadzu 2101 PC UV-vis scanning
spectrophotometer equipped with a high-pressure cell [43, 44].
Furthermore, [(dppp)2RhH] (2) (dppp ϭ 1,3-bis(di-
phenyl-phosphino)propane; in the literature sometimes ab-
breviated as dpp) was synthesized as a model compound for
the catalytic active species [33]. This complex also showed
catalytic activity, but much less compared to [Rh(dppb)H].
This is not surprising because it has two chelating phos-
phane ligands in contrast to one in Rh(dppb)H. Here we
present a detailed kinetic study on the insertion of carbon
dioxide into the rhodium-hydrogen bond of 2 to gain better
understanding of the reactions involved in the catalytic
cycle.
These reactions could become even more important be-
cause in very recent work it was demonstrated that formic
acid might be used as a hydrogen storage system [34, 35].
Thus, with the formation of formic acid from hydrogen and
carbon dioxide and the consecutive decomposition to
hydrogen and carbon dioxide a complete “green” carbon
dioxide reaction cycle could be achieved [35].
Results and Discussion
The hydride complex [(dppp)2RhH] (2) was easily prepared
by treating the chloro complex [(dppp)2Rh]Cl (1) with an
excess of sodium borohydride [36]. 2 has not been charac-
terized crystallographically but crystal structures were
reported for [(dppe)2RhH] (dppe ϭ 1,2-bis(diphenylphos-
phino)ethane) and some other related hydrido-rhodium(I)
complexes [45Ϫ47].
It is well known that 2 reacts with CO2 in DMSO to form
the cationic formate complex [(dppp)2Rh]HCO2 (3) [33].
Further reaction of 3 with H2 leads to the formation of the
bis-hydride complex [(dppp)2RhH2]HCO2, 4, and then with
the loss of formic acid the catalytic cycle shown in Scheme
2 (R ϭ C6H5) is completed (and 2 is reformed). In contrast
no CO2 insertion is observed if [(dppe)2RhH] is used as
starting material [33]. This is not surprising because the
controlling influence of chelating phosphine ligands of the
general formula PR2-(CH2)n-PR2 on the activity and selec-
tivity in different reactions is well documented [17, 30, 48].
More interesting is that the CO2 insertion into 2 could be
observed only in DMSO. In other solvents bicarbonate
complexes were formed irreversibly.
Experimental Section
Materials and Methods
Reagents and solvents used were of commercially available reagent
quality. [(dppp)2Rh]Cl, [(dppp)2RhH] and fac-[Re(bipy)(CO)3H]
were synthesized according to the literature [36, 37]. UV-vis spectra
were measured on a Hewlett Packard spectrophotometer 8452 A.
NMR spectra were recorded on a Bruker DXP 300 AVANCE spec-
trometer.
Kinetic Measurements
DMSO (99.9 %) or deutero DMSO (both from Aldrich) for the
kinetic measurements were either used without further purification
or distilled under vacuum from molecular sieves. Solutions of com-
plexes for the kinetic measurements were prepared in a glove box
(Braun, Garching, Germany; dioxygen and water content less than
1 ppm) and transferred in syringes to the instruments. CO2- or H2-
saturated solutions were prepared by bubbling high-quality CO2
(Linde, Wiesbaden, Germany, 4.5 or 4.8) or H2 (Linde, Wiesbaden,
Germany, 5.0) through DMSO for 15 minutes. For the solubility
of carbon dioxide in DMSO at 25 °C slightly different values are
found in the literature [38, 39]; we used 0.132 M [39]. Solubilities
in acetonitrile (0.28 M) [40] and acetone (0.28 M) [41] were taken
from the literature. The solubility of H2 at 25 °C ϭ 1.05 (1.08) M
[42]. Dilution was accomplished by mixing the gas solutions with
argon saturated solvents. The reactions were studied under pseudo
first order conditions ([CO2] or [H2] >> [complex]) and time re-
solved spectra of the reactions studied were recorded either in a
Hewlett Packard spectrophotometer 8452 A with the help of a ra-
Scheme 2
The reactions in Scheme 2 can be followed by 1H-NMR-
spectroscopy: the signal at δ ϭ Ϫ10.51 ppm of the hydride
decreases while the signal at δ ϭ 8.55 ppm of the formate
hydrogen increases; replacing the carbon dioxide by dihy-
drogen is accompanied by the typical signal for the cation
2488
www.zaac.wiley-vch.de
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494

Hydrierung von Kohlendioxid zu Ameisensäure
Figure 2 Concentration dependence of kobs at 30 °C.
Figure 1 Time resolved spectra for the reaction of 2 (0.32 mM)
with CO2 (66 mM at 35.0 °C , total time ϭ 10000 s, Δt ϭ 200 s
(insert: absorbance vs. time trace at 410 nm).
Table 1 Kinetic data for the reaction of 2 with CO2 under high
pressure.
[(dppp)2RhH2]ϩ at δ ϭ Ϫ8,66 ppm. However, for a detailed
kinetic study the NMR-analysis has the disadvantage that
it is necessary to mix the reactants prior to the measure-
ments and furthermore, that it is very difficult to vary the
CO2 or H2 concentration under these conditions. Therefore,
we chose to follow the reaction by UV-vis spectroscopy.
[2]/mM
[CO2]/M
p/MPa
T/°C
kobs/sϪ110Ϫ4
0.08
0.08
0.08
0.08
0.08
0.066
0.066
0.066
0.066
0.066
10
25
50
75
35
35
35
35
35
3.33 0.01
4.05 0.01
5.56 0.05
7.57 0.09
9.72 0.06
100
CO2 Insertion
The featureless UV-Vis spectrum of 2 changes during
the reaction with CO2 (pseudo first order conditions:
[CO2] >> [2]) to the more pronounced spectrum of the for-
mate complex 3, characterized by an absorbance maximum
at 410 nm as shown in Figure 1.
An isosbestic point was observed at 310 nm and ab-
sorbance vs. time traces could be fitted to a single ex-
ponential function (see insert in Fig. 1) accounting for one
observable reaction step with 2 in first order in the rate law.
Because the reaction was investigated using pseudo-first or-
der conditions ([CO2] >>[2]) it can be described by the sim-
ple first order rate law (Equ. (1)):
Ϫd[2]/dt ϭ kobs x [2]
(1)
Figure 3 Pressure dependence of kobs at 35 °C and [CO2] ϭ
0.066 M.
When the observed rate constants (kobs) at different tem-
peratures (from 25 to 40 °C) were plotted vs. the concen-
tration of CO2 a linear dependence with an intercept was
observed (Figure 2) indicating the reversibility of the reac-
tion (see below). Unfortunately, the error of the obtained
values is quite large, a consequence of the measurement
setup and the long reaction times (several hours); CO2 dif-
fuses out of the lines and sealings of the rapid mixing
accessory, leading to inaccurate CO2 concentrations. From
an Eyring plot (using the slope of the concentration depen-
dencies at different temperatures) the activation parameters
for the reaction of 2 with CO2 were estimated to approxi-
To gain a more accurate insight into this reaction high-
pressure measurements were performed where this problem
does not arise. Here it was observed that the reaction be-
came faster when the pressure was increased. The activation
volume was calculated from a plot of ln(kobs) vs. pressure
to Ϫ30 1 cm3/mol (see Table 1 and Figure 3).
It is interesting at this point to compare the CO2-inser-
tion reaction into the RhϪH bond of 2 with a related reac-
tion. Sullivan and Meyer previously investigated the reac-
tion of the rhenium hydride complex [fac-Re(bipy)(CO)3H]
(5) with CO2 in different solvents according to the following
mately: ΔH# ഠ 74
23 JmolϪ1KϪ1
7 kJmolϪ1 and ΔS# ഠ Ϫ70
.
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
2489

J. Dietrich, S. Schindler
equation (the much faster photochemical reaction is not
discussed in here) [37, 49]:
Table 2a Kinetic data for the reaction of [fac-Re(bipy)(CO)3H]
with CO2 in DMSO.
Solvent
[5]/mM
[CO2]/M
T/°C
k/molϪ1иsϪ110Ϫ2
[fac-Re(bipy)(CO)3H] ϩ CO2 Ǟ [fac-Re(bipy)(CO)3O2CH]
(2)
DMSO
DMSO
DMSO
DMSO
DMSO
0.55
0.55
0.55
0.55
0.55
0.66
0.66
0.66
0.66
0.66
20
25
30
35
40
3.3 0,1
5.1 0.1
6.1 0.1
8.7 0.1
11.0 0.1
They found that the reaction is irreversible and follows a
simple second order rate law:
Ϫd[5]/dt ϭ k [5] [CO2]
(3)
The activation parameters ΔH# ϭ 53.6 4 kJmolϪ1 and
ΔS# ϭ Ϫ138 29 JmolϪ1KϪ1 were calculated from the tem-
perature dependence of this reaction in THF [49].
Table 2b Kinetic data for the pressure dependence of the reaction
of 5 with CO2 in DMSO (T ϭ 35 °C), acetonitrile (T ϭ 35 °C) and
acetone (T ϭ 20 °C).
We could confirm the results of Meyer and Sullivan in
our kinetic analysis of this reaction (the second order rate
constant for the measurement in acetonitrile must be cor-
rected because the solubility of carbon dioxide in aceto-
nitrile is higher than reported Ϫ0.28 M [40] instead of
0.14 M). Additionally, we studied the reaction in DMSO, a
solvent that has not been used by the authors, however by
us in our investigations on the insertion of carbon dioxide
into a RhϪH bond in 2 described above. Time resolved
spectra of the reaction of 5 with CO2 in DMSO are shown
in Figure 4 (insert shows absorbance time trace and fit at
410 nm) and data are reported in Table 2a. From an Eyring
plot the activation parameters in DMSO were calculated to:
Solventa)
[5]/mM
[CO2]/M
p/MPa
kobs/sϪ1
DMSO
DMSO
DMSO
DMSO
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.66
0.66
0.66
0.66
0.66
0.14
0.14
0.14
0.14
0.14
0.14
0.14
0.14
25
50
75
100
125
25
50
75
100
10
25
50
75
(1.2 0.1) x 10Ϫ2
(1.5 0.1) x 10Ϫ2
(1.8 0.1) x 10Ϫ2
(2.1 0.1) x 10Ϫ2
(2.6 0.1) x 10Ϫ2
(0.96 0.01) x 10Ϫ2
(1.26 0.04) x 10Ϫ2
(1.6 0.1) x 10Ϫ2
(2.0 0.1) x 10Ϫ2
(3.7 0.4) x 10Ϫ4
4.4 0.4) x 10Ϫ4
(5.6 0.1) x 10Ϫ4
(7.1 0.1) x 10Ϫ2
DMSO
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
Acetone
Acetone
Acetone
Acetone
0.6
0.25
0.25
0.25
0.25
These results indicate a general mechanism for the inser-
tion of CO2 into a metal-hydrogen, metal-carbon or metal-
oxygen bond. The clean second order of these reactions
with a large negative activation volume (or activation en-
tropy) account for an associate mechanism with a transition
state that could be formulated as shown in Figure 5 and
that is in accord with theoretical calculations [31, 32].
Figure 4 Time resolved spectra for the reaction of 5 (5.5 x
10Ϫ4 M) with CO2 (0.066 M] at 35.0 °C , total time ϭ 1000 s, Δt ϭ
10 s (insert absorbance vs. time trace and data fit at 410 nm).
Figure 5 Possible transition state for the CO2 insertion reaction.
ΔH# ϭ 44 3 kJmolϪ1 and ΔS# ϭ Ϫ123 10 JmolϪ1KϪ1
Kinetic measurements under pressure did lead to
ΔV# ϭ Ϫ19 1 cm3/mol (data are reported in Table 2b)
in DMSO. In acetone and acetonitrile similar values were
obtained: ΔV# ϭ Ϫ23,5 2 cm3/mol (acetone) and
ΔV# ϭ Ϫ24 1 cm3/mol (acetonitrile). Clearly the acti-
vation volume for this reaction has a quite negative value,
similar to the reaction of 2 with CO2 described above.
Furthermore, it is very similar to the activation volume of
ΔV# ഠ Ϫ25 cm3/mol for the insertion of CO2 into a
WϪCH3-bond [50].
.
Back Reaction
As discussed above, the intercept observed in the plot of
kobs vs. the concentration of CO2 indicates that the reaction
of 2 with CO2 is reversible following the rate law:
Ϫ
Ϫd[2]/dt ϭ k1 [2] [CO2] ϩ kϪ1[3] [HCO2
]
(4)
Therefore, we tried to measure the reaction under
pseudo-first order conditions both for the forward and
backward reaction by adding a tenfold excess of sodium
2490
www.zaac.wiley-vch.de
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494

Hydrierung von Kohlendioxid zu Ameisensäure
formate to the reactants. Surprisingly, we observed that the
forward reaction became faster under these conditions by a
factor of 5 to 10. It is not quite clear what causes this effect.
For the insertion of CO2 into the tungsten-alkyl bond of
[PPN][CH3W(CO)5] it was found that the presence of
anions did increase the reaction rate as well by a factor of
5 [50]. Strong ion pairing was discussed in this case. While
we cannot exclude the effects of ions in our case we think
it is more likely that the reaction is accelerated by the for-
mation of formic acid as discussed below.
During the reaction the absorbance first decreases and
than increases again (see insert in Figure 7). The final spec-
trum is identical to the spectrum of 3, the final spectrum
obtained in Figure 1. The absorbance vs. time traces could
be fitted to the sum of two exponentials (data and fit are
shown in the insert in Figure 7).
To avoid this problem we decided to investigate the reac-
tion of 1 with sodium formate according to the following
equation.
In an NMR experiment we could clearly see the forma-
tion of the hydride signal at Ϫ10.5 ppm while the formate
hydrogen signal at 8.5 ppm decreased. The UV-vis spectrum
of 1 is very similar to 3 and time resolved spectra of its
reaction with sodium formate under pseudo-first order con-
ditions are shown in Figure 6. The reaction is very slow
and the data could not be fitted to the sum of exponential
functions. Obviously, several reaction steps are involved
here that could not be separated. Therefore, it was not pos-
sible to describe the back reaction of 2 with CO2 in more
detail.
Figure 7 Time resolved spectra for the reaction of 2 (0.0005 M)
with HCOOH (0.125 M) at 20.0 °C, total time ϭ 20 s, Δt1
(0Ϫ2 s) ϭ 20 ms and Δt2 (2Ϫ20 s) ϭ 120 ms.
When the observed rate constants (kobs) for the first step
(decrease of absorbance) at different temperatures (from
20.0 to 35.0 °C) were plotted vs. the concentration of
HCOOH a linear dependence with an intercept was ob-
served (Table 3 and Figure 8). From the slopes second order
rate constants could be calculated (Table 3). The intercept
again indicates the reversibility of the reaction (Table 3).
From an Eyring plot the activation parameters for the
forward reaction were calculated to ΔH# ϭ 23.9
0.3 kJmolϪ1 and ΔS# ϭ Ϫ139 2 JmolϪ1KϪ1 and for the
backward reaction to ΔH# ϭ 61 5 kJmolϪ1 and ΔS# ϭ
Ϫ58 16 JmolϪ1KϪ1
.
The calculated rate constants for the second reaction step
(increase of absorbance) are independent of the acid con-
centration (Figure 9 and Table 4).
This reaction is the loss of dihydrogen, the back reaction
of 4 to 3. The activation parameters were calculated to
ΔH# ϭ 75 1 kJmolϪ1 and ΔS# ϭ Ϫ3 2 JmolϪ1KϪ1
.
Figure 6 Time resolved spectra for the reaction of 1 (0.0005 M)
with NaO2CH (0.005 M) at 35.0 °C , total time ϭ 7.3 h, Δt ϭ
22 min.
Reaction of 1 with dihydrogen
While it was not possible to directly measure the reaction
of 3 with dihydrogen to form 4 the analogous reaction of 1
with dihydrogen could be investigated using stopped-flow
techniques. Time resolved spectra of this quite fast reaction
are shown in Figure 10 (insert shows an absorbance time
trace and fit at 410 nm).
A plot of the observed rate constants kobs vs. [H2] at
25 °C was linear with an intercept. The intercept of
0.31 0.04 sϪ1 is the rate constant for the back reaction of
Reaction of 2 with formic acid
The reaction of formic acid with 2 is the back reaction of
the final step in the catalytic cycle shown in Scheme 2. Here
an oxidative addition of Hϩ on the Rh(I) complex takes
place and a Rh(III) dihydrido complex is formed. The dihy-
dride can be detected by NMR-spectroscopy. Again the re-
action can be followed by UV-vis spectroscopy, using the
stopped-flow technique (Figure 7).
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
2491

J. Dietrich, S. Schindler
Table 3 Kinetic data for the reaction of 2 (5 x 10Ϫ4 M) with
Table 4 Kinetic data for the reaction of 2 with HCOOH (second
HCOOH.
reaction step).
[HCOOH]/M
T/°C
kobs/sϪ1
T/°C
k/s
0.025
0.025
0.025
0.025
0.05
0.05
0.05
20
25
30
35
20
25
30
35
20
25
30
35
20
25
30
35
20
25
30
35
20
25
30
35
0.49 0.02
0.61 0.01
0.82 0.01
1.10 0.03
0.90 0.01
1.12 0.02
1.33 0.02
1.66 0.04
1.10 0.07
1.36 0.01
1.60 0.01
1.90 0.02
1.33 0.08
1.6 0.08
1.9 0.2
20
25
30
35
40
0.21 0.01
0.35 0.03
0.58 0.03
0.98 0.05
1.57 0.09
0.05
0.062
0.062
0.062
0.062
0.075
0.075
0.075
0.075
0.10
0.10
0.10
0.10
0.125
0.125
0.125
0.125
2.3 0.2
1.8 0.1
2.2 0.1
2.5 0.2
3.2 0.2
2.14 0.02
2.60 0.02
3.20 0.04
3.85 0.03
T/°Ca)
20
25
30
35
k1 / MϪ1иsϪ1
16.7 0.3
20.2 0.5
23.9 0.7
kϪ1 / sϪ1
0.07 0.02
0.11 0.04
0.15 0.06
0.3 0.1
Figure 9 Plot of kobs vs. [HCOOH] (reaction 2).
28
1
Figure 10 Time resolved spectra for the reaction of 1 (0.0005) with
dihydrogen (0.44 M at 25.0 °C in DMSO, total time 8.2 s, Δt ϭ
99 ms.
Figure 8 Plot of kobs vs. [HCOOH] (reaction 1).
[(dppp)2RhH2]Cl to 1 and H2. This value is identical in the
range of the errors to the obtained rate constant for the
reaction of 4 to 3 and H2 measured during the reaction of
2 with formic acid (Table 4; 25 °C). Thus we could confirm
our mechanistic scheme in an excellent way by different
kinetic measurements of the individual reaction steps.
From Specfit we could obtain all second order rate
constants for the forward reaction of 1 with H2 to
[(dppp)2RhH2]Cl (Table 5). The activation parameters were
calculated from the according Eyring plot to ΔH#
ϭ
27 3 kJmolϪ1 and ΔS# ϭ Ϫ150 10 JmolϪ1KϪ1
.
2492
www.zaac.wiley-vch.de
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494

Hydrierung von Kohlendioxid zu Ameisensäure
[4] A. Behr, Carbon Dioxide Activation by Metal Complexes,
VCH, Weinheim, 1988.
[5] A. Behr, Angew. Chem. 1988, 100, 681Ϫ698; Angew. Chem.
Int. Ed. Engl. 1988, 27, 661.
[6] J. Costamagna, G. Ferraudi, V. J. Canales, Coord. Chem. Rev.
1996, 148, 221Ϫ248.
[7] D. J. Darensbourg, M. W. Holtcamp, Coord. Chem. Rev. 1996,
153, 155Ϫ174.
Table 5 Kinetic data for the reaction of 1 with dihydrogen.
[H2] /M
T / °C
k /(molϪ1иsϪ1
)
0.11
0.22
0.33
0.44
0.44
0.44
0.44
0.66
25
25
25
20
25
30
35
25
2.1 0.1
1.81 0.05
1.87 0.05
1.40 0.03
1.86 0.06
2.10 0.04
2.57 0.06
1.83 0.09
[8] P. G. Jessop, T. Ikariya, R. Noyori, Chem. Rev. 1995, 95,
259Ϫ272.
[9] W. Leitner, Coord. Chem. Rev. 1996, 153, 257Ϫ284.
[10] K. K. Pandey, Coord. Chem. Rev. 1995, 140, 37Ϫ114.
[11] B. P. Sullivan, K. Krist, H. E. Guard, Electrochemical and
Electrocatalytic Reactions of CO2, Elsevier, Amsterdam, 1993.
[12] D. Walther, Coord. Chem. Rev. 1987, 79, 135Ϫ174.
[13] D. Walther, Nachr. Chem. Tech. Lab. 1992, 40, 1214.
[14] X. Yin, J. R. Moss, Coord. Chem. Rev. 1999, 181, 27Ϫ59.
[15] D. Walther, M. Ruben, S. Rau, Coord. Chem. Rev. 1999, 182,
67Ϫ100.
[16] Y. Inoue, H. Izumida, Y. Sasaki, H. Hashimoto, Chem. Lett.
1976, 863Ϫ864.
[17] W. Leitner, Angew. Chem. 1995, 107, 2391Ϫ2405; Angew.
Chem. Int. Ed. Engl. 1995, 34, 2207Ϫ2221.
[18] Ullmann’s Encyclopedia of Industrial Chemistry, 6th Edition,
VCH, Weinheim, 1999.
Conclusions
With our detailed kinetic study we could clearly support
the catalytic cycle shown in Scheme 2. Kinetic data includ-
ing the quite negative activation volume of Ϫ30 cm3/mol
support the formation of the transition state shown in Fig-
ure 5 during the reaction of [(dppp)2RhH] (2) with CO2.
Furthermore, we observed a negative activation volume of
Ϫ19 cm3/mol for a related irreversible CO2 insertion reac-
tion into the rhenium hydride complex [fac-Re(bipy)-
(CO)3H]. These findings are in accord with theoretical cal-
culations performed previously. The reaction of 2 with CO2
is reversible, however our efforts to kinetically analyze this
reaction were unsuccessful. In contrast we could study the
reaction of 2 with formic acid, thus obtaining the dihydrido
complex [(dppp)2RhH2]HCO3 (4), the back reaction in
Scheme 2. As expected this reaction (from 2 to 4) is ac-
companied by a large negative activation entropy of
Ϫ139 JmolϪ1KϪ1 clearly indicating an associative mecha-
nism. The back reaction (from 4 to 2) on the other side
clearly has an activation entropy that is much less negative.
The final step, the irreversible loss of dihydrogen (from 4 to
3) indicates an interchange mechanism because the value
for the activation entropy is close to 0. Because we could
not isolate 3 in pure form we investigated the reaction with
the analogous chloride complex [(dppp)2Rh]Cl (1) instead.
Here again, we observed an associative mechanism for the
reaction of 1 with H2, clearly supported by the large nega-
[19] P. G. Jessop, T. Ikariya, R. Noyori, Nature 1994, 368,
231Ϫ233.
´
´
[20] G. Laurenczy, F. Joo, L. Nadosdi, Inorg. Chem. 2000, 39,
5083Ϫ5088.
[21] P. Munshi, A. D. Main, J. C. Linehan, C.-C. Tai, P. G. Jessop,
J. Am. Chem. Soc. 2002, 124, 7963Ϫ7971.
[22] P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem.
Soc. 1996, 118, 344Ϫ355.
´
´
[23] I. Joszai, F. Joo, J. Mol. Cat. A 2004, 224, 87Ϫ91.
[24] Y. Himeda, Eur. J. Inorg. Chem. 2007, 3927Ϫ3941.
[25] J.-C. Tsai, K. M. Nicholas, J. Am. Chem. Soc. 1992, 114,
5117Ϫ5124.
[26] E. Graf, W. Leitner, J.C.S. Chem. Comm. 1992, 623Ϫ624.
[27] F. Gassner, E. Dinjus, H. Görls, W. Leitner, Organometallics
1996, 15, 2078Ϫ2082.
[28] W. Leitner, E. Dinjus, F. Gaßner, J. Organomet. Chem. 1994,
475, 257.
[29] R. Fornika, H. Görls, B. Seemann, W. Leitner, J. Chem. Soc.,
Chem. Comm. 1995, 1479Ϫ1481.
tive value of Ϫ150 JmolϪ1KϪ1
.
[30] E. Graf, W. Leitner, Chem. Ber. 1996, 129, 91Ϫ96.
[31] F. Hutschka, A. Dedieu, W. Leitner, Angew. Chem. 1995, 107,
1905Ϫ1908; Angew. Chem. Int. Ed. Engl. 1995, 34, 1742Ϫ1745.
[32] F. Hutschka, A. Dedieu, M. Eichberger, R. Fornika, W. Le-
itner, J. Am. Chem. Soc. 1997, 119, 4432Ϫ4443.
[33] T. Burgemeister, F. Kastner, W. Leitner, Angew. Chem. 1993,
105, 781Ϫ783; Angew. Chem. Int. Ed. Engl. 1993, 32, 739Ϫ741.
[34] B. Loges, A. Boddien, H. Junge, M. Beller, Angew. Chem.
2008, 120, 4026Ϫ4029; Angew. Chem. Int. Ed. Engl. 2008, 47,
3962Ϫ3965.
Acknowledgements. The authors gratefully acknowledge financial
support from the DFG and the BMBF. Furthermore, we would
like especially to thank Prof. Walter Leitner (RWTH Aachen) for
providing the Rhodium complexes and his assistance with this
work. Also we are thankful to Prof. Rudi van Eldik (University of
Erlangen-Nürnberg) for his support of this project.
[35] C. Fellay, P. J. Dyson, G. Laurenczy, Angew. Chem. 2008, 120,
4030Ϫ4032; Angew. Chem. Int. Ed. Engl. 2008, 47, 3966Ϫ3968.
[36] B. R. James, D. Mahajan, Can. J. Chem. 1979, 57, 180Ϫ187.
[37] B. P. Sullivan, T. J. Meyer, J. Chem. Soc., Chem. Commun.
1984, 1244Ϫ1245.
References
[1] T. Sakakura, J.-C. Choi, H. Yasuda, Chem. Rev. 2007, 107,
2365Ϫ2387.
´
[2] P. G. Jessop, F. Joo, C.-C. Tai, Coord. Chem. Rev. 2004, 248,
[38] A. Gennaro, A. A. Isse, E. Vianello, J. Electroanal. Chem.
1990, 289, 203Ϫ215.
2425Ϫ2442.
[3] W. M. Ayers, in ACS Symposium Series, Vol. 363, ACS, Wash-
ington, DC, 1988.
[39] M. H. Schmidt, G. M. Miskelly, N. S. Lewis, J. Am. Chem.
Soc. 1990, 112, 3420Ϫ3426.
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
2493

J. Dietrich, S. Schindler
[40] E. Fujita, C. Creutz, N. Sutin, D. Szalda, J. Am. Chem. Soc.
1991, 113, 343Ϫ353.
[46] R. G. Ball, B. R. James, D. Majahan, J. Trotter, Inorg. Chem.
1981, 20, 254Ϫ261.
[41] W. Kunerth, Phys. Rev. 1922, 19, 512Ϫ524.
[42] Hydrogen and Deuterium, Vol. 5/6, Pergamon, Oxford, 1981.
[43] C. D. Hubbard, R. van Eldik, Instrum. Sci. Technol. 1995,
22, 1.
[47] L. Dahlenburg, N. Hock, H. Berke, Chem. Ber. 1988, 121, 2083.
[48] W. Leitner, M. Bühl, R. Fornika, C. Six, W. Baumann, E.
Dinjus, M. Kessler, C. Krüger, A. Rufinska, Organometallics
1999, 18, 1196Ϫ1206.
[44] D. Magde, R. van Eldik, in High Pressure Techniques in Chem-
istry and Physics: A Practical Approach (Eds.: W. Holzapfel,
N. Isaacs), Oxford Univ. Press, Oxford, 1997.
[45] S. Chang-Ping, W. Je-Yu, G. He-Fu, X. Wen-Qing, L. Xiu-
Yun, W. Jia-Huai, Chin. J. Struct. Chem. 1990, 9, 301.
[49] B. P. Sullivan, T. J. Meyer, Organometallics 1986, 5,
1500Ϫ1502.
[50] D. J. Darensbourg, R. K. Hanckel, C. G. Bauch, M. Pala, D.
Simmons, J. N. White, J. Am. Chem. Soc. 1985, 107,
7463Ϫ7473.
2494
www.zaac.wiley-vch.de
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2008, 2487Ϫ2494