Hydrogen Production and Storage on a Formic Acid/Bicarbonate Platform using Water-Soluble N-Heterocyclic Carbene Complexes of Late Transition Metals

Article abstract of DOI:10.1002/cssc.201600861

The synthesis and characterization of two water-soluble bis-N-heterocyclic carbene (NHC) complexes of rhodium and iridium is presented. Both compounds are active in H₂ generation from formic acid and in hydrogenation of bicarbonate to formate.

Full text of DOI:10.1002/cssc.201600861

DOI: 10.1002/cssc.201600861
Full Papers
Hydrogen Production and Storage on a Formic Acid/
Bicarbonate Platform using Water-Soluble N-Heterocyclic
Carbene Complexes of Late Transition Metals
+
[a]
+ [b]
[a]
[a]
Dominik Jantke , Lorenz Pardatscher , Markus Drees, Mirza Cokoja,
[a]
[a, b]
Wolfgang A. Herrmann, and Fritz E. Kꢀhn*
The synthesis and characterization of two water-soluble bis-N-
heterocyclic carbene (NHC) complexes of rhodium and iridium
is presented. Both compounds are active in H2 generation
from formic acid and in hydrogenation of bicarbonate to for-
acid was investigated using a Fisher–Porter bottle equipped
with a pressure transducer. Long-term stability for hydrogen
evolution was tested by surveillance of the gas flow rate. The
procedure does not require any additives like amines or inert
gas conditions. Density functional theory calculations in agree-
ment with experimental results suggest a bicarbonate reduc-
tion mechanism involving a second catalyst molecule, which
provides an external hydride acting as reducing agent.
mate. The rhodium derivative is most active in both reactions,
À1
reaching a TOF of 39000 h and a TON of 449000 for H pro-
2
duction. The catalytic hydrogenation reactions were carried
out in an autoclave system and analyzed using the integrated
1
peak areas in the H NMR spectra. Decomposition of formic
Introduction
Dihydrogen (H ) is considered one of the most promising
promising candidate is formic acid, being easily storable and
2
[6]
energy carriers for the future, because water is its only com-
containing 4.4 wt% of hydrogen at ambient conditions.
[
1]
bustion product. Unfortunately, diatomic hydrogen gas does
not exist on earth in significant amounts, despite hydrogen
being the most abundant element in the universe. Therefore,
Formic acid itself can be produced by catalytic CO hydrogena-
2
tion. Subsequent dehydrogenation again gives rise to CO-free
[3]
H . In comparison to formic acid, methanol has a higher hy-
2
H has to be set free from hydrogen-containing molecules that
drogen storage density, of 12.6 wt%, but hydrogen release is
2
[3]
usually display rather high hydrogen–element binding ener-
still challenging.
[
2]
gies. Accordingly, H production on an industrial scale is cur-
The homogeneously catalyzed decomposition of formic acid
to H and CO has been reported by applying iridium, rutheni-
2
rently based on steam reforming of natural gas, which requires
high temperatures and thus has a high energy demand, result-
2
2
[3,7]
um, iron, rhodium, and aluminum complexes.
However,
ing in a CO output as high as for burning natural gas direct-
most catalysts require inert-gas conditions, only work in organ-
ic solvents, or require the addition of expensive additives such
2
[
3]
ly. Efforts have to be made not only to develop sustainable
methods for H production but also to overcome problems
as dimethyloctylamine (DMOA) or NEt in order to reach re-
2
3
[
4]
[3,7]
with storage, transport, and handling of H gas. Up to now,
markable activities.
By now, the most efficient homogene-
2
physical storage of pure H does not meet industrial require-
ous systems working in aqueous solution without need of ad-
ditives are iridium complexes with turnover frequencies (TOFs)
2
ments such as ambient conditions and high power-to-volume
5
À1
6
ratios. Chemical storage potentially allows H to be stored at
and turnover numbers (TONs) in the order of 10 h and 10 ,
2
[8]
high energy density and either to be rereleased as required or
respectively. The rhodium bis-N-heterocyclic carbene (NHC)
complex 1 presented in this work is the first rhodium com-
pound that shows high activity and stability for this reaction
under ambient conditions, in aqueous solution and in the ab-
sence of any additives. This catalytic system might therefore
be considered to be one of the most sustainable, efficient, and
[5]
to be directly converted to electrical energy in fuel cells.
A
+
[
a] Dr. D. Jantke, Dr. M. Drees, Dr. M. Cokoja, Prof. Dr. W. A. Herrmann,
Prof. Dr. F. E. Kꢀhn
Department of Chemistry & Catalysis Research Center
Chair of Inorganic Chemistry
[3]
applicable for this reaction.
Technische Universitꢁt Mꢀnchen
Lichtenbergstraße 4, 85747 Garching bei Mꢀnchen (Germany)
Catalytic hydrogenation of CO on the other hand was re-
2
E-mail: fritz.kuehn@ch.tum.de
ported applying ruthenium, iridium, rhodium, iron, and cobalt
complexes, with ruthenium and iridium being the most active
+
[
b] L. Pardatscher, Prof. Dr. F. E. Kꢀhn
Department of Chemistry & Catalysis Research Center, Molecular Catalysis
Technische Universitꢁt Mꢀnchen
Lichtenbergstraße 4, 85747 Garching bei Mꢀnchen (Germany)
[3]
and stabile. Furthermore, highly active bis-NHC ligand com-
plexes of ruthenium and iridium were reported, utilizing rela-
[9]
+
tively harsh reaction conditions (2008C and 60 atm) indeed.
[
] These authors contributed equally to this work.
For catalytic formic acid production, gaseous CO can be re-
Supporting Information for this article can be found under:
http://dx.doi.org/10.1002/cssc.201600861.
2
placed by water-soluble bicarbonate salts, resulting in a much
ChemSusChem 2016, 9, 1 – 7
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[
10]
easier reaction setup. However, examples known to catalyze
fected by the electron density as well as the proximity to the
metal center, resulting in a larger low-field shift of the signals
of 1 compared to the signals of 2. This is especially visible for
the carbene signals in the C NMR spectra (169.8 ppm for 1,
152.4 ppm for 2).
the hydrogenation of bicarbonate without addition of CO gas
2
are rare and the reported catalytic efficiencies are much lower
[
3]
13
than for hydrogenation in the presence of CO2. The most sta-
bile system for bicarbonate hydrogenation was reported apply-
ing cobalt complexes in MeOH with a TON of 3900 and a con-
The catalytic activity of complexes 1 and 2 as well as the
previously reported ruthenium analogue 3 was examined for
[
11]
version of 94%. Additionally, ruthenium and iron systems
[3]
[13]
were proven to catalyze the hydrogenation of bicarbonate.
More recently an iron catalyst reaching a TON of 4600 was re-
the release of hydrogen from formic acid. The widely accept-
ed mechanism for producing dihydrogen from formic acid fol-
[
12]
[15]
ported, albeit in a CO atmosphere.
lows the catalytic pathway reported by Laurenczy et al. The
2
In this work, we present the first rhodium bis-NHC catalyst
for efficient bicarbonate hydrogenation in aqueous solution
without addition of additives, reaching TONs in the same order
catalytic cycle starts with the replacement of a chloride ligand
by formate followed by the decarboxylative formation of
a metal-bound hydride. After protonation of the hydride by
formic acid, dihydrogen leaves the coordination sphere result-
ing in a free coordination site, allowing the formation of the in-
[11]
of magnitude as the aforementioned cobalt complex. Al-
though a number of catalysts has been reported either for
[15]
CO /bicarbonate hydrogenation or formic acid dehydrogena-
itial compound.
At 808C and a catalyst concentration of
2
À1
tion, only few examples are known to catalyze both reactions
1 mm, a TOF of 6700 h using 1 was obtained. Lower activities
[
3]
À1
À1
efficiently. In contrast, rhodium system 1 as well as iridium
complex 2 and ruthenium complex 3 can be applied for either
of these transformations, depending on the reaction condi-
tions. The observed reactivity indicates the potential of the
presented catalyst to carry out a complete cycle of hydrogen
storage and subsequent hydrogen release without need for in-
termediate catalyst recovery.
were observed for 2 and for 3 with TOFs of 50 h and 30 h ,
respectively (Figure 1a). Hence, optimization of the reaction
conditions was carried out using catalyst 1 (Table 1). At 258C,
À1
the reaction proceeds slowly with a TOF of 91 h . Higher tem-
peratures result in a strong increase in activity with TOFs of
À1
À1
À1
700 h (at 408C), 3200 h (at 608C) and 6700 h (at 808C)
(Figure 1b).
At 1008C the limiting pressure of the reaction vessel (see Ex-
perimental Section for more details) of 10 bar was reached in
Results and Discussion
Previously, we reported the synthesis of highly water-soluble
ruthenium and osmium complexes bearing sulfonated NHC li-
gands and were able to apply a similar procedure for the syn-
thesis of the analogous rhodium (1) and iridium (2) deriva-
[
13]
tives. 1 and 2 are obtained in good yields (1: 93%, 2: 90%)
in a one-pot procedure using the respective imidazolium salts,
silver(I)oxide and the corresponding metal precursors
(
(
Scheme 1). The complexes are highly soluble in water
À1
~250 gL ), moderately soluble in MeOH and DMSO, and in-
soluble in THF and less polar solvents. Coupling of the methyl-
1
ene-bridge hydrogen atoms is observed in the H NMR spectra
of 1 and 2 as two doublets with a coupling constant of
2
J(H-H)=13 Hz, indicating the formation of bis-NHC com-
[
13–14]
plexes.
The chemical shifts of the signals are strongly af-
Figure 1. Pressure evolution curves for different reaction conditions.
(
a) c(catalyst)=1 mm, T=808C, c(formate)=3m, pH 2; black line: Rh com-
plex 1, dotted line: Ir complex 2, dashed line: Ru complex 3. (b) c(cata-
lyst)=1 mm, c(formate)=3m, pH 2; dashed line: T=808C, dotted line:
T=608C, black line: T=408C, long dashed line: T=258C. (c) c(forma-
te)=3m, pH 2; dashed line: c(catalyst)=1 mm, T=808C, dotted line: c(cata-
lyst)=0.1 mm, T=1008C, black line: c(catalyst)=0.1 mm, T=808C, long
dashed line: c(catalyst)=0.01 mm, T=1008C. (d) c(catalyst)=0.1 mm,
T=808C; black line: pH 2, dashed line: pH 3, dotted line: pH 4. (e) c(cata-
lyst)=0.001 mm, c(formate)=3m, pH 2, T=1008C; (f) continuous reaction:
Scheme 1. Synthesis of sulfonated NHC complexes of Rh (1) and Ir (2) and
the previously described Ru complex 3.
[
13]
À1
formic acid feed of 3.4 mLmin , T=808C, c(catalyst)=2 mm, pH 2.
&
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Table 1. Catalytic decomposition of formic acid to H
lysts 1, 2, and 3.
2
and CO
2
using cata-
Table 2. Hydrogenation of bicarbonate to formate applying 1, 2 and 3.
[
a]
[b]
Entry Catalyst c(catalyst) [mM] T [8C] t [h] TON
Conv. [%]
[
a]
À1 [b]
Entry Catalyst c(catalyst) [mM] pH T [8C] p [bar]
TOF [h ]
1
2
3
4
5
6
7
8
9
1
2
3
1
1
1
1
1
1
1
1
1
1
1
1
0.2
0.02
0.2
0.2
1
100
100
100
80
100
100
100
120
100
100
100
24
24
24
24
24
24
24
24
72
72
72
1300
1000
1200
200
2300
500
1000
1300
1500
3600
700
64
54
62
10
24
0.5
10
13
78
36
0.7
1
2
3
4
5
6
7
8
9
0
1
2
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
3
4
2
2
2
2
25
80
80
40
60
80
80
80
80
0.50
0.14
0.08
3.26
1880 s
870 s
5.62
1.24
0.32
90
50
30
700
3200
6700
8900
2300
1000
17300
27000
37100
[
c]
0.1
0.1
0.1
0.1
0.1
0.01
0.001
10
11
0.2
0.02
1
90
3480 s
2060 s
2.30
1
1
100
100
100
3
Reaction conditions: 2m KHCO solution (1 or 5 mL, pH 8) containing the
1
1
2
3
[
c]
[d]
desired amount of catalyst were added to an autoclave system, heated to
the target temperature and pressurized with hydrogen (50 bar). After the
experiment the system was cooled to RT and the pressure was released.
[a] Turnover Number after the given reaction time. [b] The conversion
was determined by adding 200 mL of the reaction solution to 400 mL D O
and 20 mL DMSO as external standard. [c] NaHCO was used as bicarbon-
3.10
39200
[
a] Pressure after 1 h or the time needed to reach 10 bar, after which the
reaction was stopped. [b] The TOF was averaged after 10 and 30 min; for
the calculation of the TOF see the Supporting Information. [c] Pressure
after 50 h. [d] The TOF was averaged after 1, 3, and 6 h; for the calcula-
tion of the TOF see the Supporting Information.
2
3
ate source.
less than 5 min. As it was not possible to determine a reliable
TOF due to the short reaction time at this high temperature,
Table 2. At 1008C and using 1 mm of 1, a conversion of 64% is
reached within 24 h. Application of catalyst 2 and 3 leads to
slightly lower conversions of 54% and 62%, respectively. Fur-
ther investigations and optimization of the reaction conditions
were therefore carried out using catalyst 1 (Table 2). First, the
reaction time was increased to 72 h, resulting in a TON of 1500
and a yield of 78%. A TON of 3600 is obtained by decreasing
the catalyst concentration to 0.2 mm. Decreasing the tempera-
ture to 808C as well as increasing it to 1208C results in lower
conversions of 10% and 13% within 24 h, respectively. Further
reduction of the catalyst concentration to 0.02 mm at 1008C
results in a significantly lower TON of 700 and a negligible con-
version of 0.5% after 24 h and 0.7% after 72 h. Replacing
KHCO with NaHCO as bicarbonate source reduces the conver-
À1
the catalyst concentration was reduced, and TOFs of 27000 h
À1
(
0.1 mm) and 37 100 h (0.01 mm) were achieved. A further re-
duction of the catalyst concentration to 0.001 mm at 1008C re-
À1
sulted in the highest measured TOF of 39200 h (Figure 1e).
By extending the reaction time to 50 h at the same reaction
conditions a TON of 449000 was obtained, which is among the
highest TONs ever observed for this reaction without use of
any additives. Using a catalyst concentration of 0.1 mm it was
further determined that the reaction proceeds fastest at a pH
value of 2 (808C), whereas at higher pH values TOFs of
À1
À1
2
300 h (pH 3) and 1000 h (pH 4) were obtained. This might
be attributable to the easier protonation of the intermediate
hydride complex at lower pH values. All results for the catalytic
decomposition of formic acid to H and CO are summarized in
3
3
sion due to a lower solubility of NaHCO in water and thus
3
lower availability of bicarbonate. Application of a catalyst con-
centration of 0.2 mm at 1008C leads to a conversion of 10%
within 24 h (Table 2). The relatively high conversion rates of
1 are quite surprising as this catalyst displays only one vacant
coordination site. Pꢂpai et al. reported theoretical investiga-
tions of a possible reaction mechanism, involving coordination
2
2
Table 1. In order to examine the long-term stability of catalyst
, a continuous reaction at 808C and a catalyst concentration
1
of 2 mm was carried out. For this purpose, formic acid was
added continuously to the reaction mixture. The averaged gas
À1
À1
flow rate of 70.2 mLmin corresponds to a TOF of 4300 h ,
[16]
which was maintained by addition of formic acid with a rate of
of both bicarbonate and hydride to one metal center. Upon
À1
3
.3 mLh (Figure 1 f). Over a period of 72 h no decrease in ac-
loss of hydroxide, the formed CO inserts into the MÀH bond.
2
tivity was observed. Moreover, this experiment proves the pro-
duction of two gaseous molecules, being H and CO . Decom-
This insertion was also postulated by Hull et al., who reduced
[8a]
gaseous CO to formate.
2
2
2
position of formate to H O and CO, which could potentially
However, this mechanism requires the presence of two
vacant coordination sites in the employed catalyst. For com-
plex 1, this second vacant coordination site is unavailable due
to the strong coordination and high stability of a) the 1,2,3,4,5-
pentamethylcyclopentadienyl (Cp*) moiety, and b) the chelat-
ing bis-NHC ligand. Therefore, theoretical calculations explain-
ing the catalytic activity with respect to the structural proper-
ties of 1 were carried out. Using a B3LYP/6-31G* level of
2
take place, would result in a decreased gas flow rate and an in-
crease in the level of liquid.
Furthermore, 1–3 were tested for the catalytic hydrogena-
tion of bicarbonate to formate. The complexes were dissolved
in a 2m bicarbonate solution (pH 8), heated to the desired
temperature, and pressurized with 50 bar hydrogen in an auto-
clave vessel. The resulting TONs and conversions are listed in
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
theory, the lower TON and lower averaged TOF observed at
008C with increased conversion and thus higher formate con-
relative to the catalyst, ligand exchange processes are sup-
pressed and the reaction slows down. Calculations for two pos-
sible mechanisms involving a) an external hydride, or b) molec-
ular dihydrogen as reducing agent were carried out using
a simplified model system (Scheme 3). The hydride (a) was in-
terpreted simplified as a free hydride in the applied density
functional theory (DFT) model system and was found to inter-
act with the most protic H-atom of the complex at the methyl-
ene bridge of the bis-NHC ligand. The transfer of this hydride
to the C-atom of bicarbonate leading to its reduction has to
1
centration can be explained by ligand exchange processes at
the metal complexes (Scheme 2). In a first step, the chloride
À1
ligand is exchanged by bicarbonate with DG=7.7 kcalmol
(1 kcal=4.184 kJ), which is then reduced to formate and sub-
sequently replaced by bicarbonate.
This ligand exchange is also endergonic, with
À1
DG=6.0 kcalmol . With increasing concentration of formate
#
À1
overcome a transition state with DG =41.9 kcalmol to pro-
À1
duce a metal-bound ortho-formate with DG=15.5 kcalmol .
According to the relatively high energy barrier, the reduction
of the C-atom might be the rate-determining step. The final
transition state can be described as the elimination of the hy-
#
À1
À1
droxide with DG =21.2 kcalmol (barrier of 5.7 kcalmol )
À1
and gives the product complex with DG=À15.6 kcalmol
with respect to the starting material. The similar reaction utiliz-
ing molecular hydrogen (b) as the reducing agent follows
a higher first transition state (due to the breaking of the HÀH
#
À1
bond) of DG =75.1 kcalmol . The resulting metal bound
ortho-formic acid complex is not favored due to its higher en-
À1
dergonic nature (DG=25.9 kcalmol ) with respect to its
anionic analogue. Elimination of water over a transition state
À1
of DG=44.0 kcalmol gives the formate complex with an
À1
overall DG=À2.5 kcalmol .
Based on our calculations and the experimental observation
that the activity of the system is strongly reduced at very low
catalyst concentration (Table 2), it can be assumed that a reac-
tion mechanism involving a second catalyst molecule provid-
Scheme 2. Ligand exchange processes during bicarbonate reduction.
À1
R=CH
2
CH
2
SO
3
. DG values are in kcalmol
.
Scheme 3. Reduction of bicarbonate to formate with a) an external hydride (blue) and b) molecular dihydrogen (red). R=CH
kcalmol . Each line has its own absolute zero point.
2
CH
2
SO
3
. DG values are in
À1
&
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
ing the external hydride takes place. The formation of the anal-
ogous metal-centered ruthenium hydride complex under the
and the [Cp*MCl
2
18]
]
2
precursors were prepared according to litera-
[
ture procedures. All reactions with air-sensitive compounds were
performed under an argon atmosphere using standard Schlenk
techniques. NMR spectra were recorded on a Bruker Avance DPX
[
17]
experimental conditions was previously reported. The overall
reaction mechanism for the reduction of bicarbonate to for-
mate with 1 is depicted in Scheme 4.
4
00. Chemical shifts are given in ppm and were referenced to the
13
residual proton resonance or to the natural abundance C reso-
nance of the solvent, respectively. Proton and carbon assignments
were confirmed using two-dimensional NMR spectroscopy (COSY
and HMQC). Elemental analyses were obtained from the Microana-
lytical Laboratory of Technical University of Munich. ESI–MS spectra
were recorded on a ThermoElectron LCQ classic instrument.
General synthetic procedure
The methylene-bridged ligand (1.0 eq) was dissolved in 20 mL of
degassed H O, Ag O (1.0 eq) was added, and the suspension was
2
2
stirred at 508C under protection from light. After 90 min NaCl (1.0
eq) was added and the reaction mixture was stirred for another
1
5 min. A solution of [Cp*MCl ] (0.5 eq) in 15 mL of a degassed
2 2
H O/DMSO mixture (1:1) was added to the suspension. After sol-
2
vent removal in vacuo at 508C the residue was extracted with
MeOH (2ꢃ10 mL). The combined extracts were filtered over silica
and the metal complexes were precipitated upon addition of Et O
2
(
50 mL). The product was obtained by washing the solid with Et O
2
(
1
2
3ꢃ10 mL) and drying in vacuo.
1
3
: H NMR (400 MHz, [D ]MeOD, 300 K): d=7.58 (d, J(H-H)=2.1 Hz,
4
3
2
H, NCHCHN), 7.49 (d, J(H-H)=2.1 Hz, 2H, NCHCHN), 6.32 (d, J(H-
2
H)=13.2 Hz, 1H NCHH’N), 5.67 (d, J(H-H)=13.2 Hz, 1H, NCHH’N),
4
3
.39 (t, J(H-H)=8.1 Hz, 4H, NCH CH ), 2.94–2.87 (m, 4H,
2 2
CH CH SO ), 2.42–2.27 (m, 2H, CH CH CH SO ), 2.24–2.12 (m, 2H,
2
2
3
2
2
2
3
Scheme 4. Proposed overall mechanism for the reduction of bicarbonate to
formate.
13
CH CH CH SO ) 1.75 ppm (s, 15H, Cp(CH ) ); C NMR (101 MHz,
2
2
2
3
3 5
1
[
(
D ]MeOD, 300 K): d=169.8 (d, J(Rh-C)=50.8 Hz, C Rh), 124.0
4
I
m
1
C H), 123.3 (C H), 101.3 (d, J(Rh-C)=5.4 Hz, C CH ), 62.9
Im Im Cp 3
(
9
NCH N), 50.6 (NCH CH ), 49.8 (CH CH SO ), 27.8 (CH CH CH ),
2 2 2 2 2 3 2 2 2
Conclusions
À
.8 ppm (C CH ); MS (ESI) m/z (%): 663.4 ([MÀNa] ) (100), 1351.3
Cp
3
À
(
[2MÀNa] )
(13);
elemental
analysis:
calcd
(%)
for
In this work the synthesis of highly water-soluble bis-NHC com-
plexes of rhodium, iridium, and ruthenium is presented. These
complexes are shown to be efficient catalysts for the reduction
of bicarbonate to formate as well as for the hydrogen genera-
tion from formic acid. Using the rhodium derivative, a high
TOF and one of the highest reported TONs for formic acid de-
composition are obtained under ambient conditions and with-
out application of any additives. In bicarbonate reduction, the
C H ClN NaO S RhCl·5H O: C 35.55, H 5.58, N 7.21, S 8.36; found:
23
33
4
6
2
2
C 35.38, H 5.26, N 7.12, S 8.36.
2
2
H)=13.1 Hz, 1H, NCHH’N), 5.60 (d, J(H-H)=13.1 Hz, 1H NCHH’N),
4.32 (t, J(H-H)=8.1 Hz, 4H, NCH CH ), 2.93–2.86 (m, 4H,
CH
1
3
: H NMR (400 MHz, [D ]MeOD, 300 K): d=7.53 (d, J(H-H)=2.1 Hz,
4
3
2
H, NCHCHN), 7.50 (d, J(H-H)=2.1 Hz, 2H, NCHCHN), 6.24 (d, J(H-
2
3
2
2
CH SO ), 2.43–2.32 (m, 2H, CH CH CH SO ), 2.23–2.13 (m, 2H,
2 3 2 2 2 3
2
13
CH CH CH SO ) 1.82 ppm (s, 15H, Cp(CH ) ); C NMR (101 MHz,
2
2
2
3
3 5
[
(
(
(
D ]MeOD, 300 K): d=152.4 (C Ir), 122.9 (C H), 122.5 (C H), 94.6
C CH ), 63.1 (NCH N), 50.2 (NCH CH ), 49.8 (CH CH SO ), 28.2
4 Im Im Im
(optimized) TON of the rhodium derivative is in the range of
Cp
3
2
2
2
2
2
3
the best catalysts reported so far without additionally pressur-
À
CH CH CH ), 9.5 ppm (C CH ); MS (ESI) m/z (%): 753.4 ([MÀNa] )
2
2
2
Cp
3
izing the reaction vessel with CO . The high activities for both
À
À
2
100), 1507.4 ([2MÀ2Na+H] ) (4) 1529.4 ([2MÀNa] ) (9); elemental
processes, hydrogen storage and release, potentially enable
a total hydrogen storage cycle. The high catalyst stability and
analysis: calcd (%) for C H ClN NaO S IrCl·4H O: C 32.56, H 4.87,
23
33
4
6
2
2
N 6.60, S 7.56; found: C 32.76, H 4.44, N 6.47, S 7.53.
the ability to perform a H -storage cycle without intermediate
2
catalyst recovery could make the protocol highly sustainable.
The use of water as solvent, the absence of additives such as
amines, and the mild reaction conditions lead to a very envi-
ronmentally friendly hydrogen storage system.
Catalytic Reactions
Closed-vessel formic acid decomposition: Fisher–Porter bottle
1
0 mL of formate buffer with the desired concentration and pH
value were charged into a Fisher–Porter bottle, which was then
closed and fitted with a pressure transducer. After heating to the
desired temperature and reaching constant pressure for equilibra-
tion, the bottle was reopened and the catalyst was added as
a highly concentrated aqueous solution. The bottle was closed and
the pressure increase was observed. The calculations of the TOFs
and TONs are described in the Supporting Information.
Experimental Section
General Remarks
Commercially available precursors and reagents were used without
purification unless noted otherwise. The methylene bridged ligand
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Open-vessel formic acid decomposition :volumetric gas mea-
surement
[4] U. Eberle, M. Felderhoff, F. Schꢀth, Angew. Chem. Int. Ed. 2009, 48,
6608–6630; Angew. Chem. 2009, 121, 6732–6757.
[
5] S. Enthaler, ChemSusChem 2008, 1, 801–804.
The desired amount of formate buffer was charged into a multi-
necked flask, which was equipped with a reflux condenser and
a septum. A gas volume measurement system ADM1000 (Agilent
Technologies) was attached on top of the reflux condenser. As
[6] a) S. Fukuzumi, Eur. J. Inorg. Chem. 2008, 1351–1362; b) R. Williams,
R. S. Crandall, A. Bloom, Appl. Phys. Lett. 1978, 33, 381–383; c) B. Loges,
A. Boddien, F. Gꢄrtner, H. Junge, M. Beller, Top. Catal. 2010, 53, 902–
914.
[
7] M. Cokoja, C. Bruckmeier, B. Rieger, W. A. Herrmann, F. E. Kꢀhn, Angew.
Chem. Int. Ed. 2011, 50, 8510–8537; Angew. Chem. 2011, 123, 8662–
À1
soon as the gas flow had decreased to 0 mLmin the equilibration
was finished, and the catalyst was added through the septum as
a highly concentrated aqueous solution. The course of the reaction
was monitored by surveillance of the gas flow. In order to maintain
a constant formic acid concentration, formic acid was added con-
tinuously through the septum using a syringe pump.
8
690.
[
8] a) 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; b) W.-H.
Wang, M. Z. Ertem, S. Xu, N. Onishi, Y. Manaka, Y. Suna, H. Kambayashi,
J. T. Muckerman, E. Fujita, Y. Himeda, ACS Catal. 2015, 5, 5496–5504.
9] A. Azua, S. Sanz, E. Peris, Chem. Eur. J. 2011, 17, 3963–3967.
[
[
10] a) A. Boddien, F. Gꢄrtner, 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) J. Elek, L. Nadasdi, G. Papp, G.
Laurenczy, F. Joo, Appl. Catal. 2003, 255, 59–67; c) H. Horvꢂth, G. Lau-
renczy, A. Katho, J. Organomet. Chem. 2004, 689, 1036–1045; d) F. Joꢅ,
G. Laurenczy, L. Nadasdi, J. Elek, Chem. Commun. 1999, 971–972; e) G.
Laurenczy, F. Joo, L. Nadasdi, Inorg. Chem. 2000, 39, 5083–5088.
Hydrogenation of bicarbonate
The hydrogenation of bicarbonate to formate was carried out in an
autoclave system manufactured by Haage. The catalyst was dis-
solved in the desired amount of water and bicarbonate (NaHCO3
or KHCO ) was added to yield a 2m bicarbonate solution. The solu-
3
[
[
[
11] C. Federsel, C. Ziebart, R. Jackstell, W. Baumann, M. Beller, Chem. Eur. J.
tion was transferred into an autoclave, which was then closed and
heated to the desired temperature. When the required tempera-
ture was reached, the autoclave was purged twice with H2 and
then pressurized to 50 bar. As soon as the reaction finished, the au-
toclave was cooled to RT and hydrogen was released. The solution
was removed from the autoclave and an aliquot (200 mL) was
2
012, 18, 72–75.
12] F. Bertini, N. Gorgas, B. Stçger, M. Peruzzini, L. F. Veiros, K. Kirchner, L.
Gonsalvi, ACS Catal. 2016, 6, 2889–2893.
13] D. Jantke, M. Cokoja, A. Pçthig, W. A. Herrmann, F. E. Kꢀhn, Organome-
tallics 2013, 32, 741–744.
[14] a) T. Scherg, S. K. Schneider, G. D. Frey, J. Schwarz, E. Herdtweck, W. A.
Herrmann, Synlett 2006, 2894–2907; b) M. Poyatos, E. Mas-Marza, M.
Sanau, E. Peris, Inorg. Chem. 2004, 43, 1793–1798.
given into a NMR tube, which was filled with 400 mL of D O and
2
2
0 mL of DMSO serving as external standard. The integrated peak
1
[15] C. Fellay, N. Yan, P. J. Dyson, G. Laurenczy, Chem. Eur. J. 2009, 15, 3752–
areas of formate and DMSO in the H NMR spectrum gave the con-
centration of formate in the analyzed solution.
3760.
[
[
16] G. Kovꢂcs, G. Schubert, F. Joo, I. Pꢆpai, Catal. Today 2006, 115, 53–60.
17] D. Jantke, M. Cokoja, M. Drees, W. A. Herrmann, F. E. Kꢀhn, ChemCatCh-
em 2013, 5, 3241–3248.
Keywords: carbenes · hydrogen storage · iridium · rhodium ·
formate splitting
[
18] a) F. Godoy, C. Segarra, M. Poyatos, E. Peris, Organometallics 2011, 30,
684–688; b) C. White, A. Yates, P. M. Maitlis, D. M. Heinekey, Inorg.
Synth. 1992, 29, 228–234.
[
[
[
1] K. Mazloomi, C. Gomes, Renewable Sustainable Energy Rev. 2012, 16,
024–3033.
2] N. Armaroli, V. Balzani, Angew. Chem. Int. Ed. 2007, 46, 52–66; Angew.
Chem. 2007, 119, 52–67.
3] W.-H. Wang, Y. Himeda, J. T. Muckerman, G. F. Manbeck, E. Fujita, Chem.
Rev. 2015, 115, 12936–12973.
3
Received: June 29, 2016
Revised: August 1, 2016
Published online on && &&, 0000
&
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
6
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
Zwei leckerbis’en: Two water-soluble
bis-N-heterocyclic carbene (NHC) com-
plexes of rhodium and iridium are pre-
sented. Both compounds are active in
D. Jantke, L. Pardatscher, M. Drees,
M. Cokoja, W. A. Herrmann, F. E. Kꢀhn*
&& – &&
H generation from formic acid and in
Hydrogen Production and Storage on
a Formic Acid/Bicarbonate Platform
using Water-Soluble N-Heterocyclic
Carbene Complexes of Late Transition
Metals
2
hydrogenation of bicarbonate to for-
mate. Density functional theory calcula-
tions in agreement with experimental
results suggest a bicarbonate reduction
mechanism involving a second catalyst
molecule, which provides an external
hydride acting as reducing agent.
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
7
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ