Aqueous phase homogeneous formic acid disproportionation into methanol

Article abstract of DOI:10.1039/c6gc03359h

The catalytic activity of a homogeneous iridium complex in formic acid disproportionation into methanol was explored. Formic acid reduction to methanol proceeds efficiently in aqueous media with the methanol yield depending on the nature of the solvent, substrate concentration, applied H₂ pressure and reaction temperature. The methanol yield peaked at 75% when D₂O was used as a solvent at 50°C. Increasing the reaction temperature to 80°C and doubling the substrate concentration led to an improved methanol concentration, even though the corresponding yield dropped to 59%. Initial H₂ pressures further enhanced methanol formation, affording a highly concentrated 9.8 m methanol solution or a TON of 1260 upon in situ catalyst recycling under aerobic conditions. No catalyst deactivation was observed for five cycles.

Full text of DOI:10.1039/c6gc03359h

Green Chemistry
View Article Online
View Journal
PAPER
Aqueous phase homogeneous formic acid
disproportionation into methanol†
Cite this: DOI: 10.1039/c6gc03359h
a
b
c
c
b
a
K. Sordakis, A. Tsurusaki, M. Iguchi, H. Kawanami, Y. Himeda* and G. Laurenczy*
The catalytic activity of a homogeneous iridium complex in formic acid disproportionation into methanol
was explored. Formic acid reduction to methanol proceeds efficiently in aqueous media with the metha-
nol yield depending on the nature of the solvent, substrate concentration, applied H
2
pressure and reac-
tion temperature. The methanol yield peaked at 75% when D O was used as a solvent at 50 °C. Increasing
2
the reaction temperature to 80 °C and doubling the substrate concentration led to an improved methanol
concentration, even though the corresponding yield dropped to 59%. Initial H2 pressures further
enhanced methanol formation, affording a highly concentrated 9.8 m methanol solution or a TON of
Received 6th December 2016,
Accepted 22nd December 2016
DOI: 10.1039/c6gc03359h
1
260 upon in situ catalyst recycling under aerobic conditions. No catalyst deactivation was observed for
www.rsc.org/greenchem
five cycles.
N-methylpyrrolidone at 240 °C, with carbon monoxide,
Introduction
1
1
however, being the primary product. In 2011, Milstein et al.
investigated indirect CO hydrogenation to MeOH in organic
The valorization of carbon dioxide (CO
2
) as a renewable C
1
2
building block is advantageous not only from an environ- solvents, catalyzed by homogeneous ruthenium-based pincer
1
mental point of view but also due to potential substitution of complexes, via utilization of urea derivatives and organic
problematic and/or toxic reactants such as carbon monoxide, carbonates, carbamates and formates as substrates (at 110 °C),
2
phosgene and methane. Carbon dioxide has also been pro- which in certain cases led to the formation of secondary by-
1
2,13
posed for chemical hydrogen storage in liquid organic carriers products.
Independently, Sanford et al. presented a combi-
3
–5
such as formic acid and methanol. At present, CO
2
serves as nation of three catalytic systems, one of which was the afore-
raw material in a limited number of industrial processes such mentioned Milstein complex, which progressively afforded
as the production of urea, salicylic acid, cyclic carbonates and methanol from CO in a multistep-synthesis approach at
2
1
4
polycarbonates and as an additive to carbon monoxide for the 135 °C. Due to partial incompatibility of the catalysts, a
6
synthesis of methanol (MeOH). Furthermore, the transform- spatial separation of the various reaction steps was essential in
ation of 5500 tons per year of CO into methanol in Iceland order to obtain a maximum TON of 21. The same group later
2
constitutes the first demonstration of a scaled-up, “green” developed a ruthenium catalyst in THF, able to reduce CO
2
to
7
alternative to the fossil fuel-based production pathway.
MeOH in basic reaction solutions, which could facilitate the
8
15
Methanol in turn, is a valuable chemical commodity and the CO capture step, at temperatures of up to 155 °C. Ding et al.
2
9
basis of a proposed methanol economy.
followed a method similar to that of Milstein, only using ethyl-
Nevertheless, the development of more environmentally ene carbonate in THF instead of dimethyl carbonate as a sub-
1
6
friendly and energy efficient processes remains one of the strate. Direct CO hydrogenation to MeOH in the presence of
2
main challenges the current “CO
to overcome. The group of Sasaki firstly studied this reaction and Leitner, who applied an in situ formed ruthenium phos-
2
-to-methanol” research has an alcohol and an acid additive was reported by Klankermayer
1
0
1
7
in the presence of Ru (CO)
and potassium iodide in phine catalyst in THF at 140 °C. The identification of several
catalytic species and their deactivation pathways allowed for
3
12
optimization of the system, making the presence of excess
1
8
19
a
alcohol redundant. Recently Prakash, Olah et al. converted
Institute of Chemical Sciences and Engineering, École Polytechnique Fédérale de
Lausanne, Switzerland. E-mail: gabor.laurenczy@epfl.ch
1
5
CO
solvent and in the presence of pentaethylenehexamine at
55 °C. In this case, the novelty was related to the utilization
of CO captured by bubbling synthetic air through the basic
2
to methanol with a known Ru catalyst, in triglyme as a
b
National Institute of Advanced Industrial Science and Technology, Tsukuba Central
5
c
, Ibaraki 305-8565, Japan. E-mail: himeda.y@aist.go.jp
1
National Institute of Advanced Industrial Science and Technology, 4-2-1 Nigatake,
Miyagino, Sendai, Miyagi, 983-8551, Japan
2
solution for 64 h. We have recently reported direct CO conver-
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
2
c6gc03359h
sion to methanol in aqueous solution and at ambient tempera-
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
ture, catalyzed by the homogeneous iridium complex [Cp*Ir
(
4
4,4′dhbp)(H O)](SO ) (1, Cp* = pentamethylcyclopentadienyl,
,4′dhbp = 4,4′-dihydroxy-2,2′-bipyridine). It was found that
2 4
2
0
the transformation proceeded via the disproportionation of
transiently formed formic acid.
To date, only a limited number of publications report on
formic acid disproportionation. Miller and Goldberg firstly
obtained methanol through the aforementioned reaction in
the presence of [Cp*Ir(bpy)(H
2 2
O)](OTf ) (bpy = 2,2′-bipyridine,
OTf = trifluoromethanesulfonate), albeit with low selectiv-
2
1
ities. The authors observed that more concentrated FA solu-
tions or initial H pressures led to enhanced MeOH formation,
Scheme 1 Methanol production from formic acid disproportionation
catalyzed by homogeneous iridium complex (1).
2
with the latter being practically independent of the solution
pH. Upon optimization of the reaction conditions, a maximum
MeOH selectivity of 12% (or 17.5 mM MeOH) and a TON of 70
were attained at 60 °C for a low FA conversion of just 3%. A
higher TON of 156 with a reduced MeOH selectivity of 7% was
reached at 80 °C, translating into a 39 mM MeOH solution or a
TOF of 6.5 h− . The group of Cantat investigated FA dispropor-
tionation with catalysts comprised in situ from [Ru(1,5′-
2
0
2
In our previous report where CO served as substrate, the
final methanol yield was dictated by that of formic acid
obtained from CO hydrogenation, which is subject to thermo-
dynamic limitations. Herein, we would like to further focus
on the optimization studies for the formic acid disproportiona-
tion reaction, considering that the formic acid substrate can
2
2
6
1
2
2
cyclooctadiene)(methylallyl)
though triphenylphosphine,
ethane, P(CH CH PPh ) and CH C(CH OPPh ) led to negli-
2
] and a phosphine ligand. Even
2
also be obtained from renewable sources other than CO , e.g.
1,2-bis(diphenyl-phosphino)
2
7
biomass (Scheme 1).
2
2
2 3
3
2
2 3
gible methanol formation, employing the triphos ligand CH
3
C
(
CH PPh showed more promising results. In the latter case,
2
2 3
)
the formic acid concentration, reaction temperature and
nature of the solvent significantly affected the final methanol
Results and discussion
yields. Ruthenium(II) carbonyl complexes with a coordinated Complex (1) catalyzes formic acid disproportionation into
triphos ligand, bearing a hydride and a chloro moiety or two methanol in aqueous sulfuric acid, with the selectivity depend-
hydrides formed in chloroform and benzene (or toluene) ing on the substrate and H
respectively, were identified as non-active species for FA dis- applied H pressure and the reaction temperature. When a
proportionation. The best MeOH yield of 50% was obtained 5 m (10.0 mmol FA) aqueous formic acid solution was heated
from the disproportionation of 2.4 mmol FA in THF at 150 °C, at 50 °C in 2.5 m H SO under isochoric conditions (i.e. at con-
2 4
SO concentrations, the evolved/
20
2
2
4
with a catalyst loading of 0.6 mol% and in the presence of stant volume, resulting in a pressure build-up due to FA de-
methanesulfonic acid. In this case, 0.4 mmol MeOH were pro- hydrogenation and disproportionation), a methanol yield of
duced (TON = 26, 1.3 M) within 1 h. More recently, the same 59% along with 98% FA conversion were attained (Table 1,
authors have studied metal-free disproportionation of formate entry 1). The amount of formic acid that did not undergo dis-
anions into methanol derivatives in the presence of stoichio- proportionation was dehydrogenated (i.e. 39%). When the reac-
2
3
metric amounts of dialkylborane reagents.
The highest tion was repeated in deuterated water under otherwise identical
methanol selectivity of 53% (for max. 80% substrate conver- conditions, an improvement of 15% or a methanol yield of 68%
sion) was reached when a bis(formoxy)borate derivative was was obtained (Table 1, entry 2). We attributed this improve-
heated at 130 °C in acetonitrile in the presence of 1 equivalent ment to a decreased rate of the competing formic acid de-
2
8
of (dibenzo-18-crown-6) ether. In this case, bis(formoxy)borate hydrogenation reaction as a result of a kinetic isotope effect,
had been pre-synthesized from a 9-borabicyclo[3.3.1]nonane as previously observed by Miller and co-workers for a similar
2
1
dimer, formic acid and anhydrous sodium formate in THF. iridium catalyst. The formic acid dehydrogenation and dis-
Neary and Parkin reported CpMo(CO) complex in proportionation reactions produce 2 and 2/3 mol of gas per
benzene, that catalyzed FA disproportionation with a lower mol of the converted substrate, respectively. Therefore, the less
a
3
H
2
4
MeOH selectivity of 21% (or 24 mM MeOH) at 100 °C, while pronounced pressure increase when D
the selectivities obtained by Kubiak with [Cp*Ir(R-bpy)Cl]Cl solvent instead of H O, was further indicative of restrained FA
R-bpy = 4,4′-di-R-2,2′-bipyridine; R = CF , H, Me, tBu, OMe) dehydrogenation (Fig. 1a). In agreement with our previous
2
O was utilized as a
2
(
3
2
5
20
type catalysts were limited to below 2%. The methanol investigations, sulfuric acid concentrations over 2.5 m in
selectivity peaked at 1.7% for the iridium complex bearing an O had no favoring effect on the final methanol yields
unsubstituted bipyridine ligand which led the authors to con- (Table 1, entries 2–5).
D
2
clude that the electron donating ability of the ligand could not
We subsequently examined the influence of FA concen-
be directly correlated with the exhibited activity in formic acid tration on its disproportionation reaction in the presence of
disproportionation.
0.5 equivalents of sulfuric acid. Higher FA concentrations
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
a
Table 1 Optimization of the formic acid disproportionation reaction
b
b
Entry
[FA] (m)
2 4
[H SO ] (m)
T (°C)
Solv.
FA conv. (%)
t (h)
MeOH yield (%)
[MeOH]fin. (m)
c
1
5
5
5
5
6.5
7.5
7.5
5
2.5
2.5
3.0
3.5
3.25
3.75
3.75
2.5
50
50
50
50
50
50
50
80
80
H
2
O
O
O
O
O
O
O
98 ± 1
98 ± 1
96 ± 2
96 ± 2
98 ± 1
80 ± 1
98 ± 1
99 ± 1
99 ± 1
72
96
59 ± 1
68 ± 1
68 ± 2
68 ± 1
73 ± 2
74 ± 1
75 ± 1
24 ± 2
59 ± 1
0.96 ± 0.02
1.11 ± 0.02
1.09 ± 0.03
1.09 ± 0.02
1.55 ± 0.04
1.48 ± 0.02
1.84 ± 0.02
0.40 ± 0.03
1.95 ± 0.03
2
3
4
5
6
D
2
D
2
D
2
D
2
D
2
D
2
120
120
240
288
600
20
d
7
8
9
H
H
2
O
O
10
5.0
2
48
a
13
Reaction conditions: ncat = 15.9 μmol (0.159 mol%), msolv. = 2.0 g. FA conversions and MeOH yields were calculated by quantitative C NMR
3
0
spectroscopy, using acetonitrile as an internal standard. The difference between these values corresponds to FA dehydrogenation (apart from
MeOH, H /CO and traces of CH OOCH were produced). CH OOCH was not taken into account (its yield was in all cases below 2%). Data
taken from ref. 20. Experiment performed once, estimated error given.
b
c
2 2 3 3
d
2
9
dehydrogenation rate. This yield was constant throughout the
reaction (Table 1, entries 6 and 7) and translated into a final
MeOH concentration of 1.84 m and a TON of 240. Prolonged
heating was necessary under these conditions to obtain full FA
−1
conversion, resulting in a low average TOF of 0.4 h . The
beneficial effect of a higher substrate concentration was more
pronounced in H O; by doubling both the FA and sulfuric acid
2
amounts at 80 °C, the MeOH yield increased by 145%, from
24% to 59% (Table 1, entries 8 and 9). Owing to the elevated
reaction temperature the average TOF improved over ten-fold to
−1
5
.2 h producing a 1.95 m MeOH solution (TON = 250). In the
absence of external H pressure, three molecules of formic acid
were necessary for the formation of one equivalent of methanol
eqn (1), Scheme 1) hence the resulting yields were calculated
2
(
accordingly (Table 1, entries 1–9).
We have previously reported that a preceding H pressuriz-
2
ation step of the reaction mixture further promoted methanol
formation. We reasoned that H
2
gas tentatively stabilized the
2
formed Ir–(η -H ) structures that were implicated in FA dis-
2
2
0
proportionation. We therefore examined the final methanol
concentrations attained as a function of the initial H pressure
in D O and H O as solvents at two different temperatures. In
2
2
2
these cases, the contribution of direct formic acid hydrogen-
ation towards methanol in parallel with FA disproportionation
had to be considered and therefore the final methanol concen-
trations rather than the MeOH yields were used for compari-
2
son (Table 2). In H O at 80 °C and in the absence of external
H gas, 24% of the FA substrate was disproportionated to yield
2
Fig. 1 (a) Pressure profile for FA disproportionation/dehydrogenation
a 0.4 m methanol solution, while the remaining 75% was con-
reaction in H
H
2
O (black curve) and D
COOH, 2.5 m H SO , ncat = 15.9 μmol, msolv. = 2.0 g. (b) Pressure
increase and final methanol concentrations obtained from FA conver- yield compared to that obtained at 50 °C (i.e. 59%) was related
2
O (grey curve) at 50 °C. 10.0 mmol
13
2 2
verted to H and CO (Table 1, entry 8). The lower methanol
2
4
sion as a function of initial H
filled squares). Reaction conditions: 10.0 mmol
SO , ncat = 15.9 μmol (0.159 mol%), msolv. = 2.0 g, and T = 50 °C.
Reproducibility within ±5% range.
2 2 2
pressure in D O (open squares) and H O
to the exothermic character of the disproportionation reaction.
The aforementioned 0.4 m MeOH concentration was continu-
ously improved to 1.12 m, through a stepwise pre-pressuriz-
13
(
H COOH, 2.5 m
H
2
4
ation with 10, 20, 40 and 200 bar H
This increase was more pronounced at lower H
resulted in slightly improved methanol yields of up to 75% gradually leveled off; changing the initial H
2
(Table 2, entries 1–4).
2
pressures and
pressure from
2
(
Table 1, entries 5 and 7), presumably because of the elevated 10 to 20 bar resulting in an over 60% higher MeOH concen-
pressure that developed inside the reaction vessel (35 and tration (Table 2, entries 1 and 2), while between 40 and 200
2
0 bar for 7.5 and 5 m FA respectively) or due to a lower FA bar H
2
the MeOH increase was limited to below 30% (Table 2,
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
Table 2 Final methanol concentrations obtained from formic acid entries 3 and 4). A similar approach was followed in D
O at
2
a
under various reaction conditions
5
0 °C. In this case the initial H pressures and the final MeOH
2
concentrations showed an almost linear dependence in the
range between 10–50 bar H , yielding a maximum MeOH con-
b
b
T
(°C)
Initial PH2
(bar)
FA conv.
(%)
t
(h)
[MeOH]fin.
(m)
2
Entry
Solv.
centration of 1.82 m (Fig. 1b and Table 2, entries 5–8) and an
1
2
3
4
5
6
7
8
9
80
80
80
80
50
50
50
50
50
50
H
H
H
H
2
2
2
2
O
O
O
O
O
O
O
O
10
20
40
200
10
20
30
50
50
99 ± 1
99 ± 1
99 ± 1
99 ± 1
98 ± 1
90 ± 3
98 ± 1
98 ± 1
98 ± 1
99 ± 1
20
20
20
20
96
96
96
96
72
72
0.65 ± 0.02
0.82 ± 0.02
0.87 ± 0.02
1.12 ± 0.05
1.31 ± 0.02
1.34 ± 0.02
1.65 ± 0.03
1.82 ± 0.03
1.22 ± 0.02
0.95 ± 0.02
−1
average TOF of 2.4 h . As a comparison, a MeOH concen-
tration of 1.22 m obtained in H
reaction conditions was approximately 33% inferior (Table 2,
2
entries 8 and 9). When 50 bar N were employed instead of H ,
2
O under otherwise identical
D
D
D
D
2
2
2
2
2
the final MeOH concentration was similar to that obtained
without any initial pressurization (Table 2, entry 10 and
Table 1, entry 1).
H
H
2
O
O
10
2
2
50 (N )
1
3
The time course of a 5 m H COOH/D
2
O disproportiona-
a
13
2 4
Reaction conditions: 10 mmol H COOH, 2.5 m H SO , ncat =
5.9 μmol (0.159 mol%), msolv. = 2.0 g. FA conversions and final MeOH
concentrations were calculated by quantitative
tion/hydrogenation reaction in the presence of 50 bar initial
1
13
1
3
2
H pressure at 50 °C was followed by C NMR spectroscopy
C NMR spec-
3
0
troscopy, using acetonitrile as an internal standard. Apart from and is depicted in Scheme 2a. Formic acid was completely
b
1
3
13
13
13
MeOH, H
2 2 3
/CO and traces of CH OOCH were produced. Minor
CH OOCH quantities were not taken into account.
converted into CH OH, CO2 and CH OO CH within
3
3
3
90–100 h, resulting in a moderate pressure increase of approxi-
mately 7 bar. The final concentration of methyl formate, pro-
duced from the esterification reaction between formic acid
and methanol, was negligible due to the reversibility of its
13
Scheme 2 (a) Aqueous FA conversion to methanol observed by 100 MHz C NMR spectroscopy. Spectra recorded every 10 h. (b) Pressure profiles
st
nd
rd
th
th
for consecutive formic acid decomposition reactions. 1 (black), 2 (orange), 3 (green), 4 (yellow) and 5 (blue) cycle. Reaction conditions:
1
1
3
2 4 2
0.0 mmol H COOH, 2.5 m H SO , P(H )init. = 50 bar, ncat = 15.9 μmol (0.159 mol%), mD2O = 2.0 g, and T = 50 °C.
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
formation reaction. Since the iridium complex (1) catalyzed
fast H /D exchange, all isotopologues of methanol (and of MeOH yields for FA disproportionation were always ≤100%.
FA) were formed. Under the aforementioned reaction con- However, under 50 bar H pressure, yields over 100% (calcu-
With initial H
2
pressures below 50 bar, the calculated
+
+
28
2
ditions, the catalyst could be recycled in situ several times lated from eqn (1), Scheme 1) were obtained. We therefore
without exclusion of air. During each run, 10.0 mmol reasoned that direct formic acid hydrogenation to methanol
1
3
H COOH were introduced to the reaction solution, followed occurred under elevated H
2
pressures, in parallel with FA dis-
by pressurization with 50 bar H and heating at 50 °C for 96 h. proportionation. Nevertheless, we could also establish a direct
2
In total, five consecutive reactions were performed without relationship between the applied H pressure and the conver-
2
removal of the catalyst, resulting in a 9.8 m MeOH solution sion of complex (1) into a hydride moiety (1″), which was the
(
i.e. 19.6 mmol MeOH) or a TON of 1260, reproducible main species observed during methanol formation in the
pressure profiles (Scheme 2b) and an overall 99% FA conver- presence of H SO .
2
4
sion. The total required time was 480 h which translated into
an average TOF of 2.6 h . Considering that under identical formed a bright yellow reaction mixture from which immediate
reaction conditions a single catalytic cycle resulted in a 1.82 m gas evolution occurred. In the H NMR spectrum two hydride
Addition of formic acid to an aqueous solution of (1)
−
1
1
MeOH solution (Table 2, entry 8), the stability of the catalyst signals at −11.8 (1′) and −18.1 ppm (1″) appeared, with the
was demonstrated. In addition, the reaction solutions retained former known to be the monohydridic form of (1) and active
3
1
a
clear, yellow colour throughout experiments (catalyst in transfer hydrogenation (Scheme 3a). These species could
decomposition is accompanied by a change of solution colour also be identified via their distinct Cp* resonances, whose
to brown and concomitant precipitate formation).
chemical shifts varied depending on the solution pH. Upon
1
Scheme 3 (a) Partial H NMR spectrum of species (1’) and (1’’) formed from (1) in a 2 m FA/D
2
O solution, showing the Cp* and hydride region.
O solution under variable H pressure. (c) Quantification of con-
pressure derived from integration of quantitative H NMR spectra in a 0.6 m H SO /D O solution (black
O solution (green squares), a 2.5 m H SO /D O solution (red squares) and in a 2.5 m H SO /H
cat = 15.9 μmol, msolv. = 2.0 g, T = 20 °C.
1
(
b) Partial H NMR spectrum (Cp* region) of species (1) and (1’’) in a 2.5 m H
version of (1) to (1’’) under variable H
squares), a 1.5 m H SO /D
2
SO
4
/D
2
2
1
2
2
4
2
2
4
2
2
4
2
2
4
2
O solution (blue squares).
n
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
addition of 50 mol% sulfuric acid (relative to FA) the solution’s rather low, indicating that if FAL was an actual intermediate in
colour became pale yellow, gas evolution ceased, the resonance the formic acid disproportionation reaction, it would have
1
3
13
attributed to (1′) disappeared and that of (1″) significantly been detected by C NMR spectroscopy considering that
decreased. Hydride (1″) was however reformed following press- enriched substrate was always employed.
C
urization with H , which was found to enhance methanol
Methanol is typically added to formalin as a stabilizing
2
formation. We therefore concluded that (1″) was relevant to FA agent. Consequently, FAL is present in the form of methylene
disproportionation. This species featured a characteristic Cp* glycol, poly(oxymethylene) glycols, hemiformal and poly(oxy-
peak which was distinct from that of the starting compound methylene) hemiformals produced via oligomerization reac-
1
3
and could therefore be used for quantitative determination of tions. The resulting C NMR spectrum of formalin is therefore
the corresponding ratios (Scheme 3b). It was found that the complex, with signals between 80 and 100 ppm as well as
3
2
nature of the solvent (H O or D O) had no effect on the between 55 and 60 ppm. Even though traces of methanol
2
2
formation of (1″) for a given sulfuric acid concentration were present in the substrate formalin, the increasing metha-
(
Scheme 3c, red and blue squares). However, an increase in nol concentration could be clearly observed in consecutive
1
3
the latter resulted in a reduced conversion of (1) to (1″)
C NMR spectra under our reaction conditions.
(Scheme 3c, red, green and black squares). Under our typical
reaction conditions (i.e. in 2.5 m aqueous H SO and under 50
2
4
bar H ) that afforded a 9.8 m MeOH solution over five consecu-
tive runs, approximately 30% of catalyst (1) was transformed to
2
Conclusions
(
1″) at the beginning of the reaction (Scheme 3c, red squares). We have demonstrated that methanol can be produced from
We have previously hypothesized that (1′) was protonated in formic acid disproportionation with unprecedented yields of
the presence of sulfuric acid to yield (1″) that most likely up to 75% in D O (under similar conditions the highest
2
2
22
2 2
bears a labile η -H ligand. However, our efforts to elucidate reported value is 50.2% ). In the absence of initial H
the nature of (1″) by multinuclear NMR spectroscopy and pressure, formic acid alone acts as the hydrogen donor yield-
mass spectrometry did not provide unequivocal proof of its ing up to 1.84 m MeOH after a single reaction at 50 °C. Higher
structure.
MeOH concentrations can be reached in the presence of initial
Throughout FA disproportionation reactions, formaldehyde H gas, which appears to have a dual functionality; it addition-
2
(FAL) was never detected as an intermediate. Nevertheless, we ally enables direct formic acid hydrogenation into methanol
examined the activity of catalyst (1) towards formalin (37 wt% and acts as a stabilizing force for certain iridium-hydride/
FAL in water, Scheme 4). When formalin was heated at 60 °C dihydrogen structures that are proposed to be imperative for
in the presence of 2.5 m H SO , slow methanol and methyl methanol formation. The advantages of the iridium complex
2 4
formate formation occurred. Initially formic acid was produced (1) compared to previously reported formic acid disproportio-
presumably via formaldehyde dehydrogenation, which then nation catalysts are related both to the higher methanol yields
underwent disproportionation into methanol. The latter could attained and the milder reaction conditions, such as lower
also be produced via transfer hydrogenation of formaldehyde, temperature and the utilization of water as the solvent. We
where both formaldehyde and formic acid acted as hydrogen obtained a maximum TON of 1260, corresponding to a 9.8 m
donors. However, the rate of formaldehyde transformation was MeOH solution from a total of five catalytic cycles (or four
in situ recyclings). Under these conditions, catalyst (1) did not
exhibit a loss of activity during recycling under aerobic con-
2
1
ditions. In comparison to Miller’s studies,
we reached
similar activities in terms of TOF values, however with a 6-fold
improvement of the MeOH yield and complete FA conversion
2
2
(
vs. 12%). Cantat and co-workers on the other hand reported
a 50.2% MeOH yield, albeit in an organic solvent (THF) with a
catalyst loading of 0.6 mol% and at a significantly higher
temperature of 150 °C. In summary, this work could contribute
to the development of a long-term, non-fossil-based methanol
production process employing “green” formic acid, which
should in turn be obtained from sources such as CO
biomass.
2
or
Experimental
General
Scheme 4 Aqueous formaldehyde transformation to methanol in the
presence of catalyst (1). Reaction conditions: 2 g formalin (37 wt% FAL),
13
Catalyst (1) was synthesized according to previously published
2 4
2.5 m H SO , ncat = 15.9 μmol, and T = 60 °C. 100 MHz C NMR spectra
3
3
recorded every 1.5 h. Red spectrum recorded after 30 h.
procedures. All other chemicals were commercial products
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
and were used without further purification. All experiments
were prepared without precluding air and repeated at least
twice. Concentrations are given in terms of molality (m), i.e.
References
1 M. Aresta, Carbon Dioxide as Chemical Feedstock, Wiley-
VCH, Weinheim, 2010.
2 F. Shi, Y. Deng, T. SiMa, J. Peng, Y. Gu and B. Qiao, Angew.
Chem., Int. Ed., 2003, 42, 3257–3260.
1
13
mol per kgsolvent (solvent: H
spectra were recorded on Bruker Avance DRX 400 MHz, Avance
00 MHz and Avance II 800 MHz spectrometers and processed
2 2
O or D O). H and C NMR
2
with TopSpin and MestReNova. Hydride-containing species
were determined by H NMR spectroscopy using a standard
3 M. Grasemann and G. Laurenczy, Energy Environ. Sci., 2012,
5, 8171–8181.
1
water suppression pulse sequence tuned to enhance the
upfield region.
4 A. F. Dalebrook, W. Gan, M. Grasemann, S. Moret and
G. Laurenczy, Chem. Commun., 2013, 49, 8735–8751.
5
6
7
8
E. Alberico and M. Nielsen, Chem. Commun., 2015, 51,
6714–6725.
Q. Liu, L. Wu, R. Jackstell and M. Beller, Nat. Commun.,
Formic acid disproportionation/hydrogenation
In a typical experiment catalyst (1) was added to an aqueous
2
015, 6, DOI: 10.1038/ncomms6933.
Carbon Recycling International, CRI – Carbon Recycl. Int.,
016.
H SO solution, followed by formic acid. The experiments were
2
4
performed either in sapphire NMR tubes or in Parr autoclaves
for PH , init. = 200 bar). In certain cases the solution was pre-
2
(
2
M. Bertau, H. Offermanns, L. Plass, F. Schmidt and
H.-J. Wernicke, Methanol: The Basic Chemical and Energy
Feedstock of the Future, Springer, 2014.
G. A. Olah, G. K. S. Prakash and A. Goeppert, Beyond Oil
and Gas: The Methanol Economy, Wiley-VCH Verlag GmbH &
Co. KGaA, Weinheim, 2nd edn, 2009.
pressurized with H2 at ambient temperature. The reactions
were followed by monitoring the pressure increase due to gas
1
3
evolution as a function of time or by C NMR spectroscopy
for sapphire NMR tubes). At the end of the reaction the sap-
9
(
phire NMR tubes were cooled and carefully depressurized to
avoid loss of organic volatiles along with the gases. An internal
1
1
1
1
1
1
1
1
1
0 J. Klankermayer, S. Wesselbaum, K. Beydoun and
W. Leitner, Angew. Chem., Int. Ed., 2016, 55, 7296–7343.
1 K. Tominaga, Y. Sasaki, T. Watanabe and M. Saito, Bull.
Chem. Soc. Jpn., 1995, 68, 2837–2842.
2 E. Balaraman, Y. Ben-David and D. Milstein, Angew. Chem.,
Int. Ed., 2011, 50, 11702–11705.
3 E. Balaraman, C. Gunanathan, J. Zhang, L. J. W. Shimon
and D. Milstein, Nat. Chem., 2011, 3, 609–614.
4 C. A. Huff and M. S. Sanford, J. Am. Chem. Soc., 2011, 133,
standard was added (CH CN) and the MeOH and FA concen-
3
1
3
trations were determined from quantitative C NMR spec-
troscopy (see ESI† for further explanations). For the recycling
experiments, at the end of the catalytic reaction the sapphire
tube was cooled, carefully depressurized and subsequently
1
3
1
2
0.0 mmol H COOH and 50 bar H were added to initiate a
new cycle (without quantitative analysis). When the desired
number of cycles was completed, an internal standard was
introduced and the final solution was analyzed as described
above. In the case of autoclaves, the reactions were prepared in
an identical manner, while heating was achieved with an oil
bath that was pre-set to the desired temperature. The final
FA/MeOH concentrations were determined by quantitative
18122–18125.
5 N. M. Rezayee, C. A. Huff and M. S. Sanford, J. Am. Chem.
Soc., 2015, 137, 1028–1031.
6 Z. Han, L. Rong, J. Wu, L. Zhang, Z. Wang and K. Ding,
Angew. Chem., Int. Ed., 2012, 51, 13041–13045.
7 S. Wesselbaum, T. vom Stein, J. Klankermayer and
W. Leitner, Angew. Chem., Int. Ed., 2012, 51, 7499–7502.
8 S. Wesselbaum, V. Moha, M. Meuresch, S. Brosinski,
K. M. Thenert, J. Kothe, T. vom Stein, U. Englert,
M. Hölscher, J. Klankermayer and W. Leitner, Chem. Sci.,
1
3
C NMR measurements in the final solutions, with an added
internal standard (CH CN). In the absence of external H gas,
3
2
the resulting methanol yields were calculated as three times
the moles of MeOH produced divided by the amount of FA
consumed.
2
015, 6, 693–704.
Formaldehyde transformation to methanol
1
9 J. Kothandaraman, A. Goeppert, M. Czaun, G. A. Olah and
Catalyst (1) was added to a sapphire NMR tube, containing 2 g
G. K. S. Prakash, J. Am. Chem. Soc., 2016, 138, 778–781.
of formalin and a given amount of sulfuric acid. The solution 20 K. Sordakis, A. Tsurusaki, M. Iguchi, H. Kawanami,
was heated inside the spectrometer and the reaction followed
by C NMR spectroscopy at regular time intervals.
Y. Himeda and G. Laurenczy, Chem. – Eur. J., 2016, 22,
15605–15608.
1
3
2
2
1 A. J. M. Miller, D. M. Heinekey, J. M. Mayer and
K. I. Goldberg, Angew. Chem., Int. Ed., 2013, 52, 3981–
3984.
Acknowledgements
2 S. Savourey, G. Lefèvre, J.-C. Berthet, P. Thuéry, C. Genre
and T. Cantat, Angew. Chem., Int. Ed., 2014, 53, 10466–
10470.
K. S. and G. L. are grateful to the Swiss National Science
Foundation and EPFL for financial support. Dr P. Miéville is
thanked for support with NMR techniques. Y. H. thanks 23 C. Chauvier, P. Thuéry and T. Cantat, Chem. Sci., 2016, 7,
the Japan Science and Technology Agency (JST), ACT-C for
financial support.
5680–5685.
24 M. C. Neary and G. Parkin, Chem. Sci., 2015, 6, 1859–1865.
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
2
2
2
2
5 A. F. Sasayama, C. E. Moore and C. P. Kubiak, Dalton 30 S. Moret, P. J. Dyson and G. Laurenczy, Dalton Trans., 2013,
Trans., 2016, 45, 2436–2439. 42, 4353–4356.
6 S. Moret, P. J. Dyson and G. Laurenczy, Nat. Commun., 31 T. Abura, S. Ogo, Y. Watanabe and S. Fukuzumi, J. Am.
014, 5, 4017. Chem. Soc., 2003, 125, 4149–4154.
7 J. Reichert, B. Brunner, A. Jess, P. Wasserscheid and 32 M. Maiwald, H. H. Fischer, M. Ott, R. Peschla, C. Kuhnert,
2
J. Albert, Energy Environ. Sci., 2015, 8, 2985–2990.
C. G. Kreiter, G. Maurer and H. Hasse, Ind. Eng. Chem. Res.,
2003, 42, 259–266.
8 W.-H. Wang, J. F. Hull, J. T. Muckerman, E. Fujita,
T. Hirose and Y. Himeda, Chem. – Eur. J., 2012, 18, 9397– 33 Y. Himeda, N. Onozawa-Komatsuzaki, S. Miyazawa,
9
404.
H. Sugihara, T. Hirose and K. Kasuga, Chem. – Eur. J., 2008,
14, 11076–11081.
2
9 Y. Himeda, Green Chem., 2009, 11, 2018–2022.
Green Chem.
This journal is © The Royal Society of Chemistry 2017