Hydrogen energy future with formic acid: a renewable chemical hydrogen storage system

Abstract

Formic acid, the simplest carboxylic acid, is found in nature or can be easily synthesized in the laboratory (major by-product of some second generation biorefinery processes); it is also an important chemical due to its myriad applications in pharmaceuticals and industry. In recent years, formic acid has been used as an important fuel either without reformation (in direct formic acid fuel cells, DFAFCs) or with reformation (as a potential chemical hydrogen storage material). Owing to the better efficiency of DFAFCs compared to several other PEMFCs and reversible hydrogen storage systems, formic acid could serve as one of the better fuels for portable devices, vehicles and other energy-related applications in the future. This perspective is focused on recent developments in the use of formic acid as a reversible source for hydrogen storage. Recent developments in this direction will likely give access to a variety of low-cost and highly efficient rechargeable hydrogen fuel cells within the next few years by the use of suitable homogeneous metal complex/heterogeneous metal nanoparticle-based catalysts under ambient reaction conditions. The production of formic acid from atmospheric CO₂ (a greenhouse gas) will decrease the CO₂ content and may be helpful in reducing global warming.

---

Ashish Kumar Singh

Dr. Ashish Kumar Singh was born in Varanasi, U. P., India in 1985. He received his PhD degree in Inorganic Chemistry from Banaras Hindu University, Varanasi, India in 2011. He then worked with Prof. Qiang Xu (AIST) as a JSPS postdoctoral fellow between 2011 and 2013. He is currently working as a DSK postdoctoral fellow in the group of Prof. B. R. Jagirdar at IPC, IISc. He is currently interested in the development of homo-/heterogeneous catalysts for the activation of small molecules for chemical hydrogen storage.

Suryabhan Singh

Dr. Suryabhan Singh was born in Varanasi, U. P., India in 1982. He received his MSc degree in Chemistry in 2007 and his PhD degree in Inorganic Chemistry under the supervision of Prof. S. Bhattacharya from Banaras Hindu University, Varanasi, India in 2013. He is currently working as a DSK postdoctoral fellow in the group of Prof. S. Natarajan at SSCU, IISc. His research work is focused on the development of metal–organic frameworks for catalytic and gas storage applications.

Abhinav Kumar

Dr. Abhinav Kumar was born in Varanasi, U. P., India in 1979. He received his PhD degree in Inorganic Chemistry under the supervision of Prof. N. Singh from Banaras Hindu University, Varanasi, India in 2009. He has been working as an Assistant Professor at the Department of Chemistry, University of Lucknow, India since 2009. His research work is focused on the development of organometallic/coordination compounds for photovoltaic and material applications.

---

1. Introduction

It is well known that non-renewable fossil fuels can be sustained only for next few decades. In the future, there will be a strong demand for safe and renewable energy carriers for transportation and other energy-related applications. Several people are actively working in different fields such as solar energy conversion, lithium ion batteries, geothermal power and nuclear energy, which could deal with the energy problem for a long time. Hydrogen is a promising and widely considered option as an alternative energy feedstock, however, despite extensive efforts by researchers to use hydrogen as a possible energy source, its storage and transportation is a major hurdle to its direct use.^1–12 To use hydrogen as a clean fuel and to overcome the hurdles in its safe and efficient storage, various advanced research approaches for the development of new materials that can store and deliver hydrogen at acceptable rates have been discovered.^1–12 Based on the methods, it can be mainly divided into two ways – either physical or chemical storage.^1–13

For the physical storage method, hydrogen is stored in its diatomic molecular form either in a closed container at high pressure and low temperature, such as using high-pressure tanks or cryo-compression,^14,15 or getting it adsorbed on high surface area materials viz. various carbon materials,^16–19 zeolites,^20,21 clathrate hydrates,^22,23 or recently the most developed or attractive materials – metal–organic frameworks.^12,24–27 However, the hydrogenation/dehydrogenation energy levels of these materials have a large energy gap, hence, they are usually less energy efficient. For the chemical storage method, hydrogen is stored in the chemically bonded form instead of its molecular form. Usually, some suitable molecules having a higher hydrogen content are selected which could release hydrogen efficiently under ambient conditions either via a catalytic or a non-catalytic process. Examples of these molecules include sodium borohydride, ammonia borane, formic acid, hydrous hydrazine, metal hydrides, metal borohydrides, metal amidoborates, etc.^2–10 Among them, formic acid is widely explored as a possible fuel for fuel cells because of its properties such as being non-toxic (although neat formic acid is corrosive and its vapour is harmful), liquid at room temperature, high density (1.22 g cm⁻³) and normal handling conditions (8.4–100.8 °C).^2–11

Formic acid is most commonly found in nature in the bites and stings of insects²⁸ and is the major by-product of petroleum refining (naphtha partial oxidation, methanol carbonylation/methyl formate hydrolysis), biomass processing, and several industrial organic syntheses. Hydrogen is easily produced from electrolysis of water.²⁹ Scientists have been able to produce formic acid by the hydrogenation of CO₂ present in the atmosphere or separated from industrial waste³⁰ using suitable catalysts.^31–34

Formic acid (FA) is considered as one of the most promising materials for hydrogen storage today. Despite the fact that the hydrogen content (4.4 wt%) in FA is less than the target set by the US Department of Energy for 2012 (ref. 35), it surpasses that of most other state-of-the-art storage materials utilized today owing to its simplicity and useable/net capacity.³ Useable or net capacity is defined as the effective hydrogen content that can be recovered in the form of H₂ from a chemical hydrogen storage system.³⁶ The produced H₂ can be utilized for clean electricity production at low temperature with the production of merely H₂O. In fact, the gravimetric energy density of formic acid is 7 times superior to that of commercial lithium ion batteries.^37,38 Besides formic acid, hydrogen generation from other liquid organic molecules, also named as liquid organic hydrogen carriers (LOHCs),^39,40 has been extensively studied e.g. methanol,^41,42 carbazole, cycloalkanes, etc. However, these systems exhibit diverse problems for use as hydrogen storage materials, such as toxicity, cost, limited stability, low dehydrogenation kinetics, and low efficiency of the regeneration processes.

Formic acid has also been considered as a fuel in direct formic acid fuel cells (DFAFCs).⁴³ Chemical processes in DFAFCs involve two-electron direct oxidation of formic acid at the anode and two-electron reduction of O₂ at the cathode (eqn (1)–(3)). However, in addition to the fuel crossover and catalyst deactivation, DFAFCs suffer from more specific detrimental effects: CO from the dehydration of FA through an undesired route (eqn (4)) poisons the catalysts (20 ppm already destroys the fuel cell). Anode:

HCOOH → CO₂ + 2H⁺ + 2e⁻ E⁰ ~ −0.25 V (1)

Cathode:

1/2O₂ + 2H⁺ + 2e⁻ → H₂O E⁰ = 1.23 V (2)

Overall:

HCOOH + 1/2O₂ → CO₂ + H₂O E⁰_Cell ~ 1.48 V (3)

Undesired route:

HCOOH → −CO_ads + H₂O (4)

Sometimes the catalyst's active phase is prone to FA corrosion and FA's hydrophilicity can dehydrate the proton exchange membrane (PEM) and cause increased cell resistance.

In recent years, immense progress has been made and many examples have been reported on the efficient use of formic acid as a renewable source of energy with great efforts from researchers all over the world.^1–10 This perspective provides an overview of catalytic hydrogen release and regeneration/production of formic acid. The first part is focused on the decomposition of formic acid for chemical hydrogen storage using a suitable homo-/heterogeneous catalyst. The second part particularly deals with the production and regeneration of formic acid from various processes such as biomass conversion and hydrogenation of CO₂. The third part gives an overview of the recent developments in the practical set up for formic acid-based “Rechargeable Hydrogen Batteries”. In the last section, we will briefly discuss the current research progress, expected improvements and future outlook in this research area.

2. Formic acid decomposition

The challenge of producing, storing and transporting hydrogen affordably has kept fuel cells from becoming popular. Instead of transporting hydrogen gas, it is more practical to have a hydrogen-containing material or a chemical hydrogen storage material that can be broken down under ambient conditions to generate H₂ gas whenever required.

Formic acid, containing 4.4 wt% hydrogen, can be decomposed by following two principal pathways (eqn (1) and (2)), in which the process producing CO₂ and H₂ (5) is the desired reaction and that producing CO and H₂O (6) is the undesired side reaction.^3–10

HCOOH → H₂ + CO₂ ΔG = −32.9 kJ mol⁻¹ (5)

HCOOH → H₂O + CO ΔG = −28.5 kJ mol⁻¹ (6)

CO-free decomposition of formic acid through pathway 1 is crucial for formic acid-based hydrogen storage.^3–10 The combination of carbon dioxide and formic acid as a hydrogen storage system might act as an elegant and simple concept wherein selective decomposition of formic acid to H₂ and CO₂ and recycling of CO₂ by reduction in the presence of H₂ to formic acid can be achieved. Meanwhile, the abundance of CO₂ on the earth makes it a cheap and readily available chemical. In this case, decreasing CO₂ emissions by reduction using H₂ makes CO₂ itself as a hydrogen carrier. Tremendous research has been done in search of suitable catalysts (homo-/heterogeneous) for selective decomposition of FA. However, achievement of complete selectivity for the decomposition of formic acid through the desired pathway is still a challenging task.

2.1 Homogeneous catalysts

Over the last few years, there has been a remarkable increase in research activities in search of high-performance homogeneous catalysts for hydrogen release from formic acid. Various groups have highlighted the performances of these homogeneous catalysts in excellent review articles.^3–10 A pioneering study by Coffey in 1967 (ref. 44) described the use of soluble Pt, Ru and Ir phosphine complexes for selective decomposition of formic acid to H₂ and CO₂. Among all the complexes, iridium complex IrH₂Cl(PPh₃)₃ gave the highest rate of decomposition. Rh(C₆H₄PPh₂)(PPh₃)₂, an organometallic complex, is active for the decomposition of formic acid.⁴⁵ A platinum dihydride complex catalyzed the reversible formation of carbon dioxide and hydrogen from formic acid. The process was somewhat dependent on the choice of solvent and was promoted by the addition of a small amount of sodium formate.⁴⁶ King and Bhattacharyya observed that nitrate ions promoted the formic acid decomposition reaction catalyzed by a rhodium(III) catalyst.⁴⁷ The reactivity of a hydride and equivalent halide complexes of molybdenum was studied for formic acid decomposition. It was observed that the use of the hydride is important for catalysis, as the equivalent halide complexes were inactive.⁴⁸ Puddephatt and co-workers have studied the detailed mechanism of the formic acid decomposition process over a binuclear, diphosphine-bridged, diruthenium catalyst [Ru₂(μ-CO)(CO)₄(μ-dppm)₂] and characterized the intermediates of the FA decomposition process using X-ray crystallography.^49,50 This complex catalyzes the reversible formation/decomposition of FA.

2.1.1 Noble-metal catalysts. Systematic studies for hydrogen generation by catalytic decomposition of formic acid have been performed by the groups of Beller and Laurenczy.^6,51–53 The majority of the catalysts active for the selective dehydrogenation of formic acid are complexes of ruthenium and iridium. However, recently some catalyst systems, based on non-noble metals such as Fe and Al, have also been reported. Beller and co-workers investigated the decomposition of formic acid with different homogeneous catalysts at 313 K, including metal salts RhCl₃·xH₂O and RuBr₃·xH₂O and precursors [{RuCl₂(p-cymene)}₂], [RuCl₂(PPh₃)₃], [{RuCl₂(benzene)₂}₂], etc., in the presence of amine/phosphine/salts as adducts.^51–53 The activity of these catalysts depends upon the type of adducts and their concentration. High catalytic activity for the decomposition of formic acid/amine adducts of different compositions was achieved using a variety of Ru precursors and phosphine ligands.⁵² With the RuBr₃·xH₂O/PPh₃ catalyst system (with 3.4 equiv. of PPh₃, TOF 3630 h⁻¹ after 20 min) the best activity was observed for hydrogen generation using the 5HCO₂H/2NEt₃ adduct. They have checked the activity for the dehydrogenation of FA in the presence of [{RuCl₂(p-cymene)}₂] and 22 amidine adducts.⁵³ The activity of the catalyst systems depended on the nature of the bases and their ratio to formic acid. For majority of the bases, an increase in concentration improved the catalyst activity. In general, in the presence of tertiary alkyl amines or more basic amidines, higher activities for FA dehydrogenation were achieved. Further, investigation of the effect of different additives revealed that the presence of halide ions also influenced the rate of hydrogen generation significantly. The best result with [RuCl₂(p-cymene)]₂ as the pre-catalyst was obtained by addition of 10 equiv. of KI. This catalyst system has activity >450%, better than [RuCl₂(p-cymene)]₂. They also observed that amine adducts have no significant effect on hydrogen production using the [RuCl₂(benzene)]₂/PPh₃ catalyst system, while with triethylamine (NEt₃) (3 : 4; amine to HCO₂H) or N,N-dimethyl-n-hexylamine (4 : 5), fast hydrogen generation was observed with [RuCl₂(benzene)]₂/dppe (TON = 1644 and 1469 h⁻¹, respectively).

A series of amine-functionalized ionic liquids (ILs) were prepared and used for hydrogen generation by the selective catalytic decomposition of formic acid in the presence of the [{RuCl₂(p-cymene)}₂] catalyst.^54,55 Amongst the investigated ILs, the 1-(2-diisopropylaminoethyl)-3-methylimidazolium chloride–sodium formate (iPr₂NEMimCl–HCOONa) system exhibited high activity (TOF > 600 h⁻¹) at 333 K. However, this process was not recyclable. Wasserscheid and co-workers reported an outstandingly simple, active and recyclable ionic liquid-based system for the catalytic decomposition of formic acid. The most efficient catalytic system was RuCl₃ dissolved in 1-ethyl-2,3-dimethylimidazolium acetate. This catalyst system converted formic acid to hydrogen and carbon dioxide selectively and was recyclable for at least nine cycles without deactivation or change in selectivity.⁵⁶ Decisively, this simple catalytic system exhibits TOFs of 150 h⁻¹ at 353 K and 850 h⁻¹ at 393 K.

Laurenczy and co-workers carried out the decomposition of FA/sodium formate solution using hydrophilic ruthenium-based catalysts, generated from the highly water-soluble ligand meta-trisulfonated triphenylphosphine (TPPTS) with either [Ru(H₂O)₆]²⁺ or, more conveniently, commercially available RuCl₃.^57,58 This catalyst system could operate over a wide range of pressure, under mild conditions, and at a controllable rate without CO contamination. Later, they immobilized this catalyst on various supports using ion exchange, coordination or physical absorption methods, to get advantage of recycling, especially for dilute formic acid solutions, or for mobile, portable applications of heterogenized catalysts.⁵⁹

Wills and co-workers investigated the activity of FA dehydrogenation using several Ru^II and Ru^III catalyst precursors ([Ru₂Cl₂(DMSO)₄], [RuCl₂(NH₃)₆], RuCl₃ and [Ru₂(HCO₂)₂(CO)₄]) in triethylamine at 393 K, explicitly without adding phosphine ligands.⁶⁰ As can be expected for such a high temperature, FA decomposition activities are exceptionally high (up to ca. 1.8 × 10⁴ h⁻¹). Unfortunately, the CO concentration consistently exceeded 200 ppm for all the catalysts tested. The authors suggested the formation of [Ru₂(HCO₂)₂(CO)₄] as the active species common to all the precursors under these reaction conditions. It is interesting to note that during reuse, all the catalysts exhibited a slight enhancement in activity with each run, indicating ongoing formation of the active catalyst species. Later, the same group used their [Ru₂Cl₂(DMSO)₄]/NEt₃ system in an attempt to continuously decompose formic acid at a rate approaching the catalyst's maximum activity without acid accumulation in the system.⁶¹ Of the two concepts tested – one temperature-based and the other using an impedance probe – the latter gave promising results, even though the gas flow decreased slightly over several days.

Homogeneous Ru catalysts based on polydentate tripodal ligands 1,1,1-tris-(diphenylphosphinomethyl)ethane (triphos) and tris-[2-(diphenylphosphino)ethyl]amine (NP₃), which can either be prepared in situ from suitable Ru(III) precursors or as molecular complexes, exhibited moderate to good activity for the selective dehydrogenation of formic acid to CO₂ and H₂.⁶² The complex [Ru(κ³-triphos)(MeCN)₃](OTf)₂ showed superior performance with a TON of 10 000 after 6 h using 0.01 mol% catalyst and allowed recycling up to eight times (0.1 mol% catalyst) with a total TON of 8000 after ca. 14 h of continuous reaction at 353 K in the presence of n-octyldimethylamine (OctNMe₂). Preliminary mechanistic NMR studies using in situ generated [Ru(κ³-triphos)(MeCN)₃](PF₆)₂ and the molecular complex [Ru(κ⁴-NP₃)Cl₂] demonstrated that the NP₃ ligand helps to stabilise Ru-hydrido species, hence there is a subtle difference in activity due to the ligand effects. Later, to clarify the mechanism of the catalytic dehydrogenation of formic acid (whether a metal-centered inner-sphere or a ligand-centered outer-sphere pathway) using the aforementioned complexes, an integrated experimental–theoretical study has been performed (Fig. 1).⁶³ The mechanism depends on the choice of polydentate phosphines, i.e., triphos vs. NP₃, and the nature and number of ancillary ligands (Cl vs. MeCN), resulting in a different number of vacant coordination sites for the activation of catalysts to their active forms. From the mechanistic study, it was concluded that Ru-hydrido vs. Ru-formato species are pivotal to bring about the efficient release of H₂ and CO₂ following either a metal-centered (inner-sphere) or a ligand-centered (outer-sphere) pathway, respectively.

Fig. 1 Schematic reaction pathway for formic acid dehydrogenation using [Ru(κ⁴-NP₃)Cl₂]. Reprinted with permission from ref. 63. Copyright 2013 American Chemical Society.

Enthaler et al. performed a preliminary study on the ruthenium-catalysed decomposition of formic acid to yield hydrogen by using a ruthenium complex-modified polyformamidine network as a solid catalyst.⁶⁴ The polyformamidine acted as a dual support for [RuCl₂(p-cymene)]₂ or [RuCl₂(p-cymene)]₂/PPh₃: on the one hand, as a ligand and on the other hand, as a base for the activation of formic acid. It is noteworthy that the polyformamidine supported [RuCl₂(p-cymene)]₂/PPh₃ catalyst system has a higher activity (TON = 325 for 3 h) than the unsupported [RuCl₂(p-cymene)]₂/PPh₃ with NEt₃ addition (TON = 169 for 3 h).

Fukuzumi et al. reported the dehydrogenation of FA, catalyzed by a water-soluble Rh catalyst, [Rh^III(Cp*)(bpy)(H₂O)](SO₄) (Cp* = pentamethylcyclopentadienyl, bpy = 2,2′-bipyridine) in aqueous solution at room temperature.⁶⁵ Similarly, an iridium catalyst [Ir^III(Cp*)(dhbpy)(H₂O)](SO₄) (dhbpy = 4,4′-dihydroxy-2,2′-bipyridine) was reported by Himeda for CO-free FA dehydrogenation.⁶⁶ High catalytic activity (TOF = 14 000 h⁻¹ at 363 K) without deterioration of the catalyst in continuous runs was observed for this catalytic system. They also demonstrated that the heteronuclear iridium–ruthenium complex [Ir^III(Cp*)(H₂O)(bpm)Ru^II(bpy)₂](SO₄)₂ {bpm = 2,2′-bipyrimidine} is a highly active catalyst for hydrogen generation in an aqueous solution under ambient conditions, giving a TOF of about 426 h⁻¹.⁶⁷

In 2013, a new bisMETAMORPhos (MMP) ligand and its iridium complex (Ir-MMP), in which the ligand is in the dianionic state, were reported by Reek and co-workers (Scheme 1).⁶⁸ The anionic form of the ligand MMP functions as an internal base, hence this catalyst system is active for FA dehydrogenation in “base-free” conditions. Base-free dehydrogenation of formic acid is important as it can act as a convenient carrier for hydrogen storage, because it increases the hydrogen content from 2.3 wt% in typical HCOOH/NEt₃ 5 : 2 mixtures to 4.4 wt% in pure HCOOH. The Ir-MMP catalyst not only operates under such base-free conditions, but also produces clean, CO-free dihydrogen and is very robust and active (base free, TOF of 3092 h⁻¹ in toluene). As such, the cooperative catalysis concept, employing a bifunctional ligand, may hold promise for the efficient and effective (storage and) release of H₂ ideally generated from a renewable source, such as solar energy.

Scheme 1 Structure of bisMETAMORPhos (MMP) ligand and its Ir complex (Ir-MMP). (Where R = 4-t-butylphenyl).

2.1.2 Non-noble metal catalysts. The first non-noble metal-based homogeneous catalyst system for hydrogen generation from formic acid was reported by Beller, Ludwig and co-workers. Boddien et al. reported FA dehydrogenation using a catalyst formed in situ from inexpensive Fe₃(CO)₁₂, 2,2′:6′2′′-terpyridine or 1,10-phenanthroline, and triphenylphosphine, under visible light irradiation at room temperature.⁶⁹ Depending on the kind of N-ligands, significant catalyst turnover numbers (>100) and turnover frequencies (up to 200 h⁻¹) were observed. Experimental (NMR, IR studies) and theoretical (DFT) studies of iron complexes, which are formed in situ under reaction conditions, confirmed that PPh₃ plays an active role in the catalytic cycle and that N-ligands enhance the stability of the system. It is shown that the reaction mechanism includes iron hydride species which are generated exclusively under irradiation with visible light.⁶⁹

Later, they demonstrated another in situ generated iron-based system as a highly active catalyst (TON up to 100 000 and TOF nearly 10 000 h⁻¹) for the CO-free decomposition of formic acid in a common organic solvent (propylene carbonate) without any further additives or light (Fig. 2 and Table 1).⁷⁰ The catalyst can be formed in situ from Fe(BF₄)₂·6H₂O and the tetradentate phosphine ligand tris[(2-diphenylphosphino)ethyl] phosphine (PP₃) under the reaction conditions or can be added to the reaction mixture in a pre-synthesized form as [FeH(PP₃)]⁺. During the process, the iron cation is permanently coordinated to four phosphorus centers of PP₃, while the remaining two coordination sites are occupied by the HCOOH substrate and/or product-derived species during the catalytic cycle. Spectroscopic studies and density functional theory calculations suggested two possible pathways for H₂ release from HCOOH, both of which went through a common Fe-hydride species, [FeH(PP₃)]⁺.⁷⁰

Fig. 2 Summary of the activation and deactivation pathways of the Fe(BF₄)₂·6H₂O/PP₃ catalyst system as well as the proposed species formed based on the results of spectroscopic analyses. Free enthalpies calculated at the B3PW91/6-31G (including polarization functions except those of hydrogen) and B3PW/6-311G (including polarization functions also for hydrogen; value underlined) levels of theory. The relative enthalpies of the reaction of the interconversion are given in kJ mol⁻¹. It is assumed that iron retains the formal oxidation state +2. Reprinted with permission from ref. 71. Copyright 2014 Wiley-VCH.

Table 1 Selective dehydrogenation of FA in the presence of in situ generated iron/non-noble catalysts.^a (Adopted with permission from ref. 71)

Metal precursor	T (K)	V 2h (mL)	TON2h	Yield (%)	CO (ppm)	Ref.
		V 3h (mL)	TON3h
a Reaction conditions: 5.3 μmol of metal precursor (100 ppm) plus 10.6 μmol of PP3 (1) were added to 2 mL of FA and 5 mL of PC at 313 or 333 K. The volume of gas was measured with an automatic gas burette and analyzed by GC (H2/CO2, 1 : 1). b Traces of H2 detected. c No H2 detected. All values given are corrected by the value of a blank reaction without a catalyst as the reference. Values are an average of at least two experiments and have an error within 10%.
Fe(BF4)2·6H2O	313	333	1279	100	<1	70
		505	1942
[Fe(acac)2]	313	315	1217	100	<1	71
		486	1879
[Fe(acac)3]	313	324	1253	100	<1	71
		503	1943
Fe(ClO4)2·xH2O	313	258	997	100	<1	71
		388	1500
Fe(ClO4)3·xH2O	313	240	928	100	<1	71
		367	1418
Fe(OAc)2	313	245	945	98	70	71
		489	1889
[Fe3(CO)12]	333	84	325	—	1120	71
		131	505
[Fe(CO)3COT]	333	7.8	30	—	450	71
		33	128
FeCl2^b	333	0.4	1.4	—	<10	71
		0.8	3.0
FeCl3^c	333	0	—	—	<10	71
		0	—
Co(BF4)2·6H2O	333	27	103	—	<10	71
		51	197
[Mn(acac)2]^b	333	0.15	0.6	—	<10	71
		0.23	0.9
Fe(BF4)2·6H2O	333	1583	6119	89	<1	71
		2101	8117

---

Recently, they have extensively studied the iron-catalyzed dehydrogenation of formic acid both experimentally and mechanistically.⁷¹ The active catalysts were generated in situ from different cationic Fe^II/Fe^III precursors and PP₃ (Table 1). These catalysts are active at amine-free and ambient conditions, and the activity of these catalysts was highly dependent on the nature of solvent used, the presence of halide ions, the water content, and the ligand-to-metal ratio. The optimal catalytic performance was achieved by using [FeH(PP₃)]BF₄/PP₃ in propylene carbonate in the presence of traces of water. With the exception of fluoride, the presence of halide ions in solution inhibited the catalytic activity. The in situ transmission FTIR measurements revealed the formation of an active iron formate species (evidenced by the band observed at 1543 cm⁻¹), which could be correlated with the evolution of gas. This active species was deactivated in the presence of chloride ions due to the formation of a chloro species.

Milstein and co-workers reported the iron-based PNP pincer complex [(^tBu-PNP)Fe(H)₂(CO)] (^tBu-PNP = 2,6-bis(di-tert-butylphosphinomethyl)pyridine) as an efficient and selective catalyst for the decomposition of formic acid to carbon dioxide and hydrogen at 313 K in the presence of trialkylamines with turnover numbers of up to 100 000.⁷² From experimental studies, it is observed that the catalytic process proceeds via protonation of the iron dihydride catalyst, followed by dihydrogen liberation which led to an unsaturated species that is transformed into a hydrido–formate complex. Regeneration of the iron dihydride catalyst is finally achieved by CO₂ elimination. This step is predicted to proceed through a novel, non-classical intramolecular β-H elimination according to DFT calculations.

In another recent work, Bielinski et al. reported a homogeneous iron catalyst that gives approximately 1 000 000 turnovers for FA dehydrogenation, when used with a Lewis acid co-catalyst.⁷³ To date, this is the highest turnover number reported for a first-row transition metal catalyst. Based on the preliminary studies, they suggested that the Lewis acid assisted in the decarboxylation of a key iron formate intermediate, as shown in Scheme 2. They have studied the promotion of catalysis by different Lewis acids and observed that the highest TON and TOF were achieved with alkali or alkaline earth metal salt co-catalysts. Importantly, the enhancement of activity is associated with the chemical affinity for carboxylate. The best results were obtained for LiBF₄ containing an alkali metal ion (Li⁺) and a weakly coordinating ligand (BF₄⁻).

Scheme 2 Proposed pathway for decarboxylation of Fe catalysts in the absence (a) and presence (b) of a Lewis acid. Adopted with permission from ref. 73. Copyright 2014 American Chemical Society.

Complete FA dehydrogenation was achieved using 0.01 mol% catalyst and 10 mol% Lewis acid (NaCl, NaBF₄, or LiBF₄). Among them, LiBF₄ gave the fastest time for completion of reaction. In fact, with catalyst loadings as low as 0.0001 mol%, a TON of 983 642 and a TOF of 196 728 h⁻¹ were obtained. From this study, they concluded that this Lewis acid promotion was general for other catalysts used for FA dehydrogenation, as well as the reverse process, i.e. CO₂ hydrogenation.

Myers and Berben reported the first molecular aluminium complexes of the bis(imino)pyridine ligand, (PhI₂P²⁻)Al(THF)X (where X = H or CH₃) for the selective dehydrogenation of formic acid to H₂ and CO₂ with an initial turnover frequency of 5200 turnovers per hour.⁷⁴ Low-temperature reactions show that the reaction of the Al-hydride complex with HCOOH afforded a complex that was protonated three times: twice on the PhI₂P²⁻ ligand and once to liberate H₂ or CH₄ from the Al-hydride or Al-methyl, respectively. In the absence of protons, insertion of CO₂ into the Al-hydride bond is facile and produces an Al-formate. Upon addition of protons, liberation of CO₂ from the Al-formate complex affords an Al-hydride. Deuterium labelling studies and the solvent dependence of the reaction indicated that outer sphere β-hydride abstraction supported by metal–ligand cooperative hydrogen bonding is the likely mechanism for C–H bond cleavage.

Based on previous reports, the most commonly observed mechanism (it may not be applicable for all the catalyst systems) for the catalytic decomposition of formic acid using homogeneous catalysts could be given in Scheme 3. These reactions have the following steps: (1) deprotonation of formic acid to formate either by itself or by addition of base, (2) formation of a metal formate complex, (3) the most important and rate-determining step is β-hydride elimination of CO₂, and finally (4) elimination of CO₂ and H₂.

Scheme 3 A general mechanism for catalytic decomposition of formic acid.

As β-hydride elimination is the rate-determining step, the catalyst with the tendency to stabilize structure 3 has better activity for formic acid dehydrogenation. This goal could be achieved not only by modifying the metal centre/geometry of metal centre/ligand geometry of homogeneous catalysts, but also by using different kinds of bases as well as other additives, e.g. Lewis acids, solvents (ionic liquids), etc. In some cases, base-free dehydrogenation was achieved because of the presence of basic sites in the catalyst itself or basic ionic liquid. Further, the search for suitable reaction conditions and homogeneous catalysts based on non-noble metals for selective dehydrogenation of formic acid is still in progress.

2.2 Heterogeneous catalysts

Initially, the decomposition of formic acid using heterogeneous catalysts has been mainly performed in the gas phase¹⁰ using a metal,^75–78 metal oxides^79–84 and metal supported on oxides or carbon catalysts.^85–93 A gas phase reaction requires the introduction of an inert carrier gas to dilute formic acid below its saturated vapor pressure or heating above the normal boiling point of formic acid (>373 K). However, researchers had also studied hydrogen generation from formic acid decomposition in the liquid phase or at low temperature. Table 2 summarizes selected heterogeneous catalysts for the decomposition of aqueous formic acid to hydrogen and carbon dioxide. Most of the heterogeneous catalysts are supported mono-/bi-/trimetallic nanoparticles with Pd as the major component.

Table 2 Selected heterogeneous catalysts for the decomposition of aqueous formic acid

Catalyst	TOF (h^−1)	T (K)	Ref.
GN = graphene nanosheets, Fc BR = functionalized basic resin, CB = carbon black, N–mrGO = nitrogen-doped mildly reduced graphene oxide, CN = nitrogen doped carbon, mpg-C3N4 = mesoporous graphitic carbon nitride.
PdAu/C	27	365	94
PdAu/C–CeO2	227	365	94
PdAu@Au	—	365	95
Pd–Au–Dy/C	269	365	98
Pd–Au–Eu/C	387	365	98
Pd–Au–Ho/C	224	365	98
PtRuBiOx	312	353	99
Ag@Pd/C (core–shell)	125	293	100
	626	363	100
Pd/C	64	298	102
Au@Pd/N–mrGO	89.1	298	103
Ag0.1Pd0.9/rGO	105.2	298	104
Co0.30Au035Pd0.35	80	298	105
Ni0.40Au0.15Pd0.45/C	12.4	298	106
CoAuPd/DNA–rGO	85	298	107
Pd@SiO2 or Pd/SiO2	—	363	108
Pd@CN (N-doped carbon)	49.8	288	109
Pd/–N(CH3)2 Fc BR	820	348	110
Pd/MSC-30	2623	323	111
PdAu/MIL-101	—	363	112
AgPd(Ag80Pd20)/MIL-101	848	353	113
Pd–NH2-MIL-125	214	305	114
PdNi@Pd/GNs–CB	577	300	115
PdNi/GNs–CB	529	300	115
Au/Al2O3	—	353	89
Au@SiO2	—	363	116
Au/ZrO2	1590	323	117
1% Pt/C–NF	0.09 (s^−1)	373	118
Pd–Ni–Ag/C	85	323	119
Pd/C	304	298	120
Pd–B/C	1184	298	120
Pd/mpg-C3N4	144	298	121

---

Xing and co-workers found that Pd–Au/C and Pd–Au@Au/C have a unique characteristic of evolving high-quality hydrogen from the catalytic decomposition of liquid formic acid in the presence of sodium formate (HCOONa) as a base at a convenient temperature.^94,95 The order of catalytic activity based on TOF values is Pd/C < Pd–Cu/C < Pd–Ag/C < Pd–Au/C. The higher catalytic activities of bimetallic Pd–Ag/C and Pd–Au/C catalysts were attributed to the higher tolerance of Ag and Au to CO poisoning. The catalytic activity was further improved by the addition of CeO₂(H₂O)_x because CeO₂ produces cationic palladium species, which showed high activity in CO oxidation⁹⁶ and methanol decomposition.⁹⁷ Another reason was that CeO₂(H₂O)_x on the Pd surface can induce the decomposition of formic acid in a more efficient route, in which fewer poisoning intermediates would be produced.⁹⁴

Later, they reported a novel Pd–Au bimetallic catalyst with a Pd–Au@Au core–shell nanostructure supported on carbon. This was synthesized by a simultaneous reduction method without using any stabilizer and was successfully used as a catalyst for hydrogen generation by formic acid decomposition. The catalyst exhibited high activity, selectivity and stability at a low temperature and the catalytic performance was much better than that of the corresponding monometallic catalysts. Especially, the reformed gas from formic acid decomposition contained only 30 ppm of CO and can be used directly to feed the fuel cells.⁹⁵ Inspired by the promotion effect of CeO₂ on the catalytic activities of Pd–Au/C and Pd–Ag/C catalysts, they extended their investigations on other rare earth elements (Dy, Eu, and Ho) and obtained TOFs of 269 ± 202, 387 ± 292, 224 ± 73, and 45 ± 11 h⁻¹ for Pd–Au–Dy/C, Pd–Au–Eu/C, Pd–Au–Ho/C and Pd–Au/C catalysts, respectively, at 365 K.⁹⁸ In addition, these catalysts were active even at room temperature temporarily and above 325 K steadily. All the rare earth metal oxides that promoted the catalytic activity of Pd–Au/C catalysts showed lower activation energies for decomposition of formic acid than Pd–Au/C and Pd–Au–Eu/C (84.2 ± 7.4 kJ mol⁻¹). The metal/metal oxide catalysts composed of platinum, ruthenium and bismuth, denoted as PtRuBiO_x, catalyzed the selective dehydrogenation of FA in water at nearly ambient temperature.⁹⁹ The observed activation energy was 37.3 kJ mol⁻¹ and the TOF was estimated to be 312 h⁻¹ in the first hour of the decomposition.

In an effort, Tsang and co-workers prepared various core–shell nanoparticles having an inner core of a metal element and an external shell of palladium.^100,101 Among all metals tested, the highest activity towards FA decomposition at room temperature is obtained for Ag@Pd NPs (diameter 8 nm) with the thinnest continuous Pd shell (1–2 atomic layers), and the corresponding Ag/Pd alloy and pure Pd catalysts showed very low activity.^100,101 The turnover frequencies per surface Pd site were comparable to homogeneous catalysts: 125 h⁻¹ at 293 K and 252 h⁻¹ at 323 K. At 293 K, an equimolar mixture of hydrogen and CO₂ was continuously produced without any trace of CO; on the other hand, CO was detected at temperatures higher than 323 K. Furthermore, theoretical calculations showed a strong correlation between the catalytic activity and the work function of the metal core: the largest net difference with the work function of the Pd shell led to the highest adsorption energy by charge transfer from the core to the shell, and hence to the best possible activity of the resulting bimetallic structure for formic acid decomposition. The very short range of this so-called “ligand” electronic effect between the two metals explained why the highest performance was achieved for the thinnest Pd layer. The nanomaterial interface definitely plays a key role in catalysis.

Yan and co-workers have extensively studied the catalytic dehydrogenation of formic acid using mono-/bi-/trimetallic nanoparticle-based catalysts supported on different carbon materials.¹⁰² A low cost Pd/C catalyst, synthesized in situ with citric acid, has been reported for highly efficient CO-free hydrogen generation from formic acid/sodium formate aqueous solution at room temperature (Fig. 3).¹⁰² The presence of citric acid during the formation and growth of the Pd nanoparticles on carbon drastically enhanced the catalytic properties of the resulting Pd/C at room temperature (conversion efficiency 85% in 160 min and turnover frequency 64 mol_H2 mol_catalyst⁻¹ h⁻¹).

Fig. 3 Schematic presentation of Pd nanoparticle formation over carbon surface. Reprinted with permission from ref. 102. Copyright 2012 Nature Publishing group.

Ultrafine (1.8 nm) and well dispersed Au@Pd core–shell nanoclusters on nitrogen-doped mildly reduced graphene oxide (Au@Pd/N–mrGO), synthesized by a green and facile strategy, without any surfactant and additional reducing agent, exhibited much greater activity than its alloy or monometallic counterparts towards hydrogen generation from formic acid aqueous solution without using any additive at room temperature.¹⁰³ During the synthesis, N–mrGO acts as both a reducing agent and a support by taking advantage of its moderate reducing and high dispersing capabilities. In another report, they found that Ag_0.1Pd_0.9 nanoparticles assembled on reduced graphene oxide (Ag_0.1Pd_0.9/rGO) could be used as an efficient catalyst for formic acid dehydrogenation reaction with 100% H₂ selectivity and exceedingly high activity at room temperature under ambient conditions.¹⁰⁴

Yan and co-workers also reported the Co_0.30Au₀₃₅Pd_0.35 nano-alloy supported on carbon as a stable, low-cost, and highly efficient catalyst for CO-free hydrogen generation from formic acid dehydrogenation at room temperature.¹⁰⁵ The elevated stability of Co⁰ in the protective nano-alloy structure makes its first application in FA dehydrogenation successful. More interestingly, the prepared Co–Au–Pd/C catalyst with a lower consumption of noble metals exhibits 100% H₂ selectivity, the highest activity, and excellent stability toward H₂ generation from FA without any additive at 298 K. The catalytic activities of Co_0.30Au_0.35Pd_0.35/C together with its mono-metallic (Pd/C, Au/C, and Co/C) and bi-metallic (Au_0.50Pd_0.50/C, Co_0.30Pd_0.70/C, and Co_0.30Au_0.70/C) counterparts for H₂ generation from FA decomposition at 298 K in ambient atmosphere are presented in Fig. 4. The as-prepared Co_0.30Au_0.35Pd_0.35/C exhibited a much better activity than the mono- and bi-metallic catalysts synthesized by the same method. In a similar way, carbon supported highly homogeneous trimetallic NiAuPd alloy nanoparticles (Ni_0.40Au_0.15Pd_0.45/C) have also been employed as a catalyst for the selective dehydrogenation of formic acid.¹⁰⁶ This catalyst also exhibited high activity and 100% selectivity for dehydrogenation of FA without any additives at room temperature. The catalytic activity of Ni_0.40Au_0.15Pd_0.45/C was much higher than its monometallic (Pd/C, Ni/C, Au/C) bimetallic counterparts (Ni_0.40Pd_0.60/C, Au_0.25Pd_0.75/C, Ni_0.40Au_0.60/C) as well as physical mixtures of Ni/C, Au/C and Pd/C (Ni : Au : Pd = 0.40 : 0.15 : 0.45) under similar reaction conditions.

Fig. 4 Gas generation by decomposition of FA (0.5 M, 10 mL) versus time in the presence of a) Co_0.30Au_0.35Pd_0.35/C, b) Co/C, c) Au/C, d) Co_0.30Au_0.70/C, e) Pd/C, f) Co_0.30Pd_0.70/C and g) Au_0.50Pd_0.50/C (nmetal/nFA = 0.02) at 298 K in ambient atmosphere. Reprinted with permission from ref. 105. Copyright 2013 Wiley-VCH.

They also reported a DNA-directed and facile approach to synthesize the Co–Au–Pd/DNA–rGO composite (Co : Au : Pd = 1 : 1 : 1).¹⁰⁷ The catalytic performance of the Co–Au–Pd/DNA–rGO composite toward the dehydrogenation of FA without any additives at room temperature was compared with the Co–Au–Pd/rGO composite and Co–Au–Pd NPs. The Co–Au–Pd/DNA–rGO composite exhibited the highest activity compared to the other catalysts and the order of activity is Co–Au–Pd < Co–Au–Pd/rGO < Co–Au–Pd/DNA–rGO. Additionally, no gas was generated from aqueous FA solution in the presence of DNA or the DNA–GO composite, thereby suggesting that DNA and DNA–GO serve as a template and a support, respectively, for the growth of Co–Au–Pd NPs, and not as catalysts for FA dehydrogenation.

We have prepared a high-performance palladium silica nanosphere (20–35 nm) catalyst using Pd(NH₃)₄Cl₂ as a precursor in a polyoxyethylene-nonylphenyl ether/cyclohexane reversed micelle system followed by NaBH₄ reduction.¹⁰⁸ Pd NPs supported on silica nanospheres (Pd/SiO₂) were prepared by the conventional impregnation of the Pd(NH₃)₄Cl₂ precursor to silica nanospheres, which were prepared using a similar reversed micelle system without a Pd precursor, followed by NaBH₄ reduction. The as-synthesized Pd@SiO₂ and Pd/SiO₂ catalysts have high catalytic activities as compared to Pd NPs supported on commercial silica at convenient temperature. Remarkably, it was observed that the interactions of the Pd nanoparticles with the surface groups of the silica support are important for the catalytic performance. The strong cooperative effects from surface molecular groups of silica on the metal surface, that is, a strong metal–molecular support interaction (SMMSI), might bring new opportunities in the development of high-performance heterogeneous catalysts from inactive or less active metal catalyst systems.

Cai et al. synthesized Pd nanoparticles immobilized on mesoporous carbon nitride (Pd@CN) as a highly efficient Mott–Schottky photocatalyst for dehydrogenation of formic acid.¹⁰⁹ The exceptional catalytic performance of this catalyst was due to the enhanced electron enrichment of the Pd nanoparticles through charge transfer operating at the interface of the Mott–Schottky contact. Under similar operating conditions, the catalytic performance of Pd@CN is orders of magnitude higher than that of similar Pd@carbon catalysts. Nanostructured carbon nitride acted both as a stabilizer and a semiconducting support for coupling of metal NPs to form the required rectifying Mott–Schottky nano-heterojunctions.

A basic resin bearing –N(CH₃)₂ functional group within its macro-reticular structure served as an efficient organic support for the active Pd nanoparticles responsible for the production of high-quality H₂via formic acid (HCOOH) decomposition at convenient temperature (Fig. 5).¹¹⁰ Physicochemical characterization as well as the kinetic isotope effect (KIE) revealed that not only the formation of small Pd NPs but also the cooperative action by the –N(CH₃)₂ groups within the resins play crucial roles in achieving efficient catalytic performances. In addition to advantages such as simple workup procedure, additive-free, and superior catalytic activity compared with conventional inorganic supports, this catalytic system can suppress unfavorable CO formation of <5 ppm, which makes it an ideal hydrogen vector in terms of potential industrial application for PEMFCs. Moreover, the basic resin support also provided a Pd–Ag nanocatalyst from an aqueous solution of a mixture of PdCl₂ and AgNO₃. The catalytic activities for H₂ production from formic acid decomposition were strongly dependent on the presence of Ag atoms and the Pd–Ag catalyst was found to perform significantly better than the pure Pd and Ag catalysts.

Fig. 5 Model of Pd nanoparticles supported on basic resin bearing –N(CH₃)₂ functional groups within its macroreticular structure. Reprinted with permission from ref. 110. Copyright 2013 American Chemical Society.

Highly dispersed Pd nanoparticles (NPs) deposited on nanoporous carbon MSC-30 have been synthesized using a sodium hydroxide-assisted reduction approach.¹¹¹ The use of NaOH during the formation and growth of the particles resulted in well-dispersed ultrafine Pd NPs on MSC-30. The combination of metal–support interaction and high dispersion of NPs significantly enhanced the performance of the resulting catalyst. This catalyst exhibited 100% H₂ selectivity and very high activity (TOF = 2623 h⁻¹) for heterogeneously catalyzed decomposition of FA at 323 K. The activity of the prepared catalyst was comparable to those acquired from the most active homogeneous catalysts as well as heterogeneous catalysts under ambient conditions.

Xu and co-workers reported the synthesis of bimetallic Au–Pd NPs immobilized in the mesoporous metal–organic framework (MOF) MIL-101 (pore size = 2.9–3.4 nm and window size = 1.2–1.4 nm) as efficient catalysts for decomposition of formic acid towards generation of hydrogen.¹¹² Because of its ordered pore structure, small window and hybrid pore surface, MIL-101 facilitates the encapsulation of metal NPs and the adsorption of formic acid inside the pores. MIL-101 has coordinatively unsaturated Cr³⁺ centres. In order to improve the interactions between the metal precursors and the MIL-101 support, ED-MIL-101 has been prepared by grafting MIL-101 with the electron-rich functional group ethylenediamine (ED). ED-MIL-101 exhibited improved immobilization of small metal NPs. Among the resulting bimetallic Au–Pd NPs immobilized in the MOFs, Au–Pd/MIL-101 and Au–Pd/ED-MIL-101 represent the first highly active MOF-immobilized metal catalysts for the complete conversion of formic acid to hydrogen at a convenient temperature.

Dai et al. investigated the activity of bimetallic Ag–Pd nanoparticles immobilized into MIL-101 for catalytic dehydrogenation of formic acid.¹¹³ Among all the AgPd@MIL-101 catalysts with different compositions of Ag and Pd (Ag@MIL-101, Ag₇₈Pd₂₂@MIL-101, Ag₆₃Pd₃₇@MIL-101, Ag₄₈Pd₅₂@MIL-101, Ag₃₅Pd₆₅@MIL-101, Ag₂₀Pd₈₀@MIL-101 and Pd@MIL-101), the Ag₂₀Pd₈₀@MIL-101 catalyst exhibited the highest catalytic activity for the conversion of formic acid to high-quality hydrogen at 353 K with a TOF value of 848 h⁻¹, which is among the highest values reported at 353 K (Fig. 6).

Fig. 6 Hydrogen generation from HCOOH in the presence of different catalysts: Ag₂₀Pd₈₀@MIL-101; Ag₃₅Pd₆₅@MIL-101; Ag₄₈Pd₅₂@MIL-101; Ag₆₃Pd₃₇@MIL-101; Ag₇₈Pd₂₂@MIL-101; Ag@MIL-101; Pd@MIL-101 at 353 K. Reprinted with permission from ref. 113. Copyright 2013 Royal Society of Chemistry.

In another report, Yamashita and co-workers used the photocatalytically active metal–organic framework MIL-125 and its amine-functionalized equivalent NH₂–MIL-125 to immobilize Pd NPs by photo-assisted and ion exchange deposition methods.¹¹⁴ The Pd–NH₂–MIL-125 catalyst showed high catalytic activity for H₂ generation from FA at ambient temperature (TOF = 214 h⁻¹ at 305 K) in comparison with Pd–MIL-125 and other titanium-based porous materials. The main factors responsible for the high catalytic performance were the basic functionalization of the MOF and small NP sizes. In addition, the photo-assisted deposition method was found to be a more effective method for producing small and highly dispersed NPs within the MOF structure.

Zhang and co-workers synthesized well-dispersed Pd–Ni nanocatalysts grown on a graphene nanosheet–carbon black (GNs–CB) composite support to combine the advantages of GNs and CB.¹¹⁵ Unexpectedly, Pd–Ni NCs loaded on GNs–CB exhibited higher catalytic performance for FA decomposition at room temperature in aqueous media than on Pd or Ni alone. Furthermore, Pd is employed to replace the surface Ni of Pd–Ni NCs toward a novel Pd–Ni@Pd catalyst to further improve the catalytic activity and stability. The use of GNs–CB as a new kind of carbon support to disperse, anchor, and further promote nanocatalysts with active components would undoubtedly assist long-term endeavours to further optimize and enhance the catalytic efficacy of catalysts in the development of FA as a hydrogen-storage material.

There are few reports for formic acid decomposition by Au NP-based heterogeneous catalysts. The first report on Au NP-based heterogeneous catalysts for formic acid dehydrogenation in the gas phase at ambient temperatures has been reported by Ojeda and Iglesia in 2009.⁸⁹ Well-dispersed Au species immobilized on Al₂O₃ catalyzed the decomposition of HCOOH with metal-time yields (rates per Au atom) larger than Pt clusters supported on Al₂O₃ at 353 K. HCOOH dehydrogenation proceeds via either a H-assisted formate decomposition mechanism or sequential cleavage of O–H and C–H bonds and H-atom recombination on isolated sites.

The first work on liquid phase dehydrogenation of formic acid using a monometallic Au NP-based catalyst was reported by Xu and co-workers.¹¹⁶ Monometallic gold nanoparticles encapsulated in amine functionalized silica nanospheres acted as a high-performance catalyst for hydrogen generation from aqueous formic acid. It was observed that the unsupported or silica-supported gold NPs without the amine functional group are inactive for FA dehydrogenation reaction. Hence, the presence of amine in the silica sphere is important for the activity of the gold nanoparticles due to the strong metal–molecular support interaction (SMMSI).

Bi et al. demonstrated the mild and selective dehydrogenation of FA/amine mixtures using ultradispersed sub-nanometric gold NPs on acid-tolerant ZrO₂ as catalysts.¹¹⁷ The catalytic reactions proceed efficiently and selectively under ambient conditions, without the generation of any unwanted by-product, such as CO, and with high TOFs/TONs. A unimolecular mechanism involving unique amine-assisted formate decomposition at the Au–ZrO₂ interface is supported by the exclusive formation of HD and the primary kinetic isotope effects measured from HCOOD or DCOOH dehydrogenation (Fig. 7). The exceptional activity of sub-nanometric Au toward FA activation promises a new area of gold research by fine-tuning of the dispersed Au clusters at the sub-nano level.

Fig. 7 Possible reaction pathway for hydrogen evolution from the 5 : 2 FA–NEt₃ azeotropic mixture system over the Au/ZrO₂ NC catalyst. Reprinted with permission from ref. 117. Copyright 2012 American Chemical Society.

The activities of Pt catalysts on carbon nanofibers with different nitrogen contents were compared for hydrogen production by formic acid decomposition.¹¹⁸ The catalysts contained a fraction of Pt clusters with a mean size of 1.0–2.3 nm and possibly a considerable fraction of Pt clusters with a diameter of less than 0.75 nm which were invisible by transmission electron microscopy (TEM) observation. The activities of N-doped catalysts with low Pt contents (≤1 wt%) were 10 times higher than the activities of undoped catalysts. The N-doped catalysts demonstrated an improved selectivity to hydrogen and an increased resistance to CO inhibition. However, they were inactive for ethylene hydrogenation. These results are explained by the presence of electron-deficient, two-dimensional sub-nm sized Pt clusters stabilized by pyridyl nitrogen on vacancy sites. In accordance, the Pt-4f_7/2 binding energies measured by X-ray photoelectron spectroscopy were 0.6 eV higher for the N-doped samples than for the undoped ones.

Yurderi et al. synthesized trimetallic Pd–Ni–Ag nanoparticles with different ratios of metals as well as its bi-metallic (Pd–Ni, Ni–Ag and Pd–Ag) and mono-metallic (Pd, Ni and Ag) counterparts supported on activated carbon by simple and reproducible wet impregnation followed by a simultaneous reduction method without using any stabilizer at room temperature.¹¹⁹ All the prepared composites were employed as heterogeneous catalysts in the catalytic decomposition of formic acid under mild reaction conditions. It was found that Pd–Ni–Ag/C can catalyze the dehydrogenation of formic acid with high selectivity (~100%) and activity (TOF = 85 h⁻¹) at 323 K. More importantly, the exceptional stability of Pd–Ni–Ag nanoparticles against agglomeration, leaching and CO poisoning makes Pd–Ni–Ag/C a reusable catalyst for formic acid dehydrogenation. The Pd–Ni–Ag/C catalyst retained almost its inherent activity (>94%) even at 5th reuse in the decomposition of formic acid with high selectivity (~100%) under ambient conditions. The quantitative kinetic studies depending on the catalyst [Pd–Ni–Ag], substrate [FA] and promoter [SF] concentrations revealed that the Pd–Ni–Ag/C catalyzed dehydrogenation of FA is a first-order reaction with respect to the catalyst concentration, half-order with respect to the promoter concentration, and with respect to the substrate concentration it appeared to be zero-order when [FA] < 0.140 M and half-order at higher concentrations. The Pd–Ni–Ag/C catalyzed dehydrogenation of FA at different temperatures provided the activation parameters (E_a = 20.5 kJ mol⁻¹, ΔH^≠ = 17.8 kJ mol⁻¹, ΔS^≠ = −190 J mol⁻¹ K⁻¹) and suggested the associative mechanism for the Pd–Ni–Ag/C catalyzed dehydrogenation of FA.

Recently, a potential boron-doped Pd nanocatalyst (Pd–B/C) was reported to enhance hydrogen generation at room temperature from aqueous formic acid–formate solutions at a record high rate.¹²⁰ Pd/C and Pd–B/C catalysts with a Pd loading of 5 wt% were synthesized through wet chemical reduction of NaBH₄ and dimethylamine borane (DMAB), respectively. To exclude any boron doping, HCOONa was also used as the reducing agent to synthesize Pd/C–HCOONa. Catalytic studies suggested only a slightly higher gas production rate on Pd/C–NaBH₄ than on Pd/C–HCOONa (Fig. 8). In contrast, the gas production rate was nearly doubled on Pd–B/C as compared to the other two catalysts at a given mass of Pd. The present Pd–B/C catalyst showed superior catalytic activity, as compared to the TOF values reported so far for various Pd-based catalysts at room temperature.

Fig. 8 Time-course of reforming gas generation from 10 mL solution of 1.1 M FA + 0.8 M SF in the presence of 100 mg of Pd based/C (5 wt% Pd) catalysts at 303 K under ambient atmosphere. Reprinted with permission from ref. 120. Copyright 2014 American Chemical Society.

Highly sensitive attenuated total reflection infrared spectroscopy (ATR-IR) was used to monitor dynamically the interfacial species on Pd–B/C or Pd/C in FA-SF solutions to provide a molecular-level understanding of the different activities of the examined catalysts. ATR-IR spectroscopy revealed that the superior activity of Pd–B/C is related to an apparently impeded CO_ad (adsorbed CO) accumulation on the surface of the Pd–B/C catalyst. This work demonstrated that developing new anti-CO poisoning catalysts coupled with a sensitive interfacial analysis is an effective way toward rational design of cost-effective catalysts for better hydrogen energy exploitation.

Reversible, carbon dioxide-mediated chemical hydrogen storage was first demonstrated using a heterogeneous Pd catalyst supported on mesoporous graphitic carbon nitride (Pd/mpg-C₃N₄).¹²¹ The Pd nanoparticles were uniformly dispersed onto mpg-C₃N₄ with an average size of 1.7 nm without any agglomeration and further exhibited superior activity for the dehydrogenation of formic acid with a turnover frequency of 144 h⁻¹ even in the absence of external bases at room temperature. From DFT studies, it was confirmed that basic sites located at the mpg-C₃N₄ support play synergetic roles in stabilizing the reduced Pd nanoparticles without any surfactant as well as in initiating H₂ release by deprotonation of formic acid. These potential interactions were further confirmed by X-ray absorption near edge structure (XANES). Along with dehydrogenation, Pd/mpg-C₃N₄ was also proved to catalyze the regeneration of formic acid via CO₂ hydrogenation. The governing factors for CO₂ hydrogenation were further elucidated to increase the quantity of the desired formic acid with high selectivity.

Beller and co-workers carried out a detailed theoretical study of formic acid decomposition on various metal surfaces {such as Ni(111), Ni(211), Pd(111), Pd(211), Pt(111) and Pt(211)}, metal oxides (such as TiO₂, MgO, ZnO and NiO) and molybdenum carbide (β-Mo₂C(101)).^122–125 A common pathway observed for the selective decomposition of formic acid to CO₂ and H₂ is shown in Scheme 4. The formate route (HCOOH → HCOO + H) plays the dominant role in formic acid decomposition, and the rate-determining step is the dissociation of formate into surface CO₂ and H (HCOO → CO₂ + H). The most stable adsorption configuration of formate (HCO₂) has a bidentate bridging structure with its two oxygen atoms (C–O) at atop sites.

Scheme 4 General mechanism of formic acid decomposition over various surfaces through formate route.

Further work is required to develop an efficient system for hydrogen release from formic acid using a heterogeneous catalyst owing to selectivity problems induced by the high reaction temperatures for heterogeneous catalytic reactions.

3. Formic acid production

Formic acid is an important platform chemical for the synthesis of diverse organic substances. In nature, FA is found in the venom of ants²⁸ and is present as a component of the atmosphere due primarily to forest emissions.^126,127 Chemically, it can be prepared by various processes. The most common industrial process is the combination of methanol and carbon monoxide in the presence of a strong base at elevated pressure (40 atm) and temperature (353 K) for the preparation of methyl formate, followed by hydrolysis of methyl formate to formic acid.¹²⁸

CH₃OH + CO → CH₃COOH (7)

CH₃COOH + H₂O → HCO₂H + CH₃OH (8)

The established BASF technology is based on the use of methanol and CO.¹²⁹

CH₃OH + CO → CH₃OCHO (9)

CH₃OCHO + H₂O → CH₃OH + HCO₂H (10)

Formally, the net reaction is

H₂O + CO → HCO₂H (11)

An issue is the separation of formic acid from the reaction mixture. Other methods for the synthesis of FA are as a by-product of acetic acid production, CO₂ hydrogenation, oxidation of biomass, and biosynthesis by reduction of carbon dioxide catalyzed by the enzyme formate dehydrogenase.^130,131 In this section, we will discuss in detail the different methods for formic acid production with major focus on CO₂ hydrogenation. Carbon dioxide, a major by-product of FA decomposition either chemically or electrochemically and the last product of any organic compound burnt in air, is a major greenhouse gas. Utilization of this greenhouse gas as a source of formic acid is not only a smart step to establish FA as a renewable energy but may also help to reduce environmental deterioration.

3.1 Oxidation of biomass

Production of formic acid from biomass offers a mild and sustainable way to produce H₂.¹³² Many researchers have studied the production of FA from biomasses (carbohydrates) using thermal cracking,¹³³ SCW/H₂O₂,¹³⁴ Fe₂(SO₄)₃/O₂ or CuSO₄/O₂ (ref. 135 and 136), OH⁻/H₂O₂ (ref. 137), etc. Wasserscheid and co-workers had performed a systematic study for the selective oxidation of carbohydrate-based biomass to formic acid and CO₂ using a Keggin-type H₅PV₂Mo₁₀O₄₀ polyoxometalate as the catalyst under varying reaction conditions with different catalyst promoters.¹³⁸ In their first report, several water-soluble carbohydrate-based biomasses had been fully and selectively converted to formic acid and CO₂ by oxidation with molecular oxygen in aqueous solution under very mild conditions.¹³⁸ Using glucose as a substrate, their catalyst loading and temperature optimization (353 K, 30 bar O₂, 26 h) studies indicated that the characteristic FA yield at full conversion was 49 ± 2%. Interestingly, biogenic waste products, such as glycerol (from bio-diesel production) and xylan (from the hemicellulosic waste streams of paper manufacturing), were oxidized to FA in 40% and 33% yield, respectively. Using this catalyst, even complex biomass mixtures, such as popular wood sawdust, were transformed to formic acid, giving 19 wt% yields (11% based on the carbon atoms in the feedstock) under non-optimized conditions.

Later, to optimize the reaction conditions for water-insoluble biomass conversion to FA and CO₂, different additives (catalyst promoters) such as CaCl₂, H₃BO₃, ZnCl₂, LiCl, benzene/toluene/methane/trifluoromethane/camphor/xylene/chlorobenzene/p-toluenesulphonic acid, etc. were used with the polyoxometalate catalyst H₅PV₂Mo₁₀O₄₀ in water using oxygen (30 bar) as the oxidising agent at 363 K and the reaction was monitored for 24 h.¹³⁹ Among all additives, p-toluenesulfonic acid acted as the best additive for the transformation of feedstock like wood, waste paper, or even cyanobacteria to formic acid and CO₂ as the sole products. The produced FA can be separated from the aqueous reaction mixture by a simple extraction process and can be regarded as a hydrogen equivalent produced from biomass in a very robust process. The reaction achieved up to 53% yield of formic acid for xylan (hemicellulose) as the feedstock within 24 h at 363 K. Using this promoter, even third generation wet biomass such as algae could be efficiently converted to FA and CO₂, offering a very interesting path to the energetic use of these side products for future biodiesel production plants from algae.

In another work, Wasserscheid and co-workers modified a Keggin-type polyoxometalate catalyst and synthesized a series of different heteropolyacids (HPAs) of the general type H_3+n[PV_nMo_12−nO₄₀] (n = 0–6) (Scheme 5).¹⁴⁰ To confirm the reactivity, all the POM complexes were tested for the conversion of glucose to FA using oxygen (at 30 bar) as the oxidising agent. The higher V-substituted complexes (n = 2–6) showed superior activities to the lower V-substituted complexes (n = 0–1). Applying the optimized polyoxometalate catalyst system H₈[PV₅Mo₇O₄₀] (HPA-5), a total FA yield (with respect to carbon in the biogenic feedstock) of 60% for glucose within 8 h of reaction time and 28% for cellulose within 24 h of reaction time could be achieved at 363 K. The trend of the activity was explained on the basis of optical and electrochemical measurements which concluded that the formation of the pervanadyl species in the higher V-substituted system was responsible for the high activity of these HPA complexes. The mechanism comprises two alternative ways for the dissociation of HPA-n. On the one hand, before the oxidation reaction starts, the fully oxidized HPA-n with a reduction degree of m = 0 dissociates to free pervanadyl and residual degenerated lacunary species. On the other hand, the fully reduced HPA-n after substrate oxidation with a reduction degree of m = n dissociates to vanadyl (VO²⁺) and lacunary species.

Scheme 5 Dissociation/re-oxidation equilibria for HPA-n in aqueous solution.

In an acidic solution, the free vanadyl (VO²⁺) ion cannot be oxidised by molecular oxygen. Instead, the V⁴⁺ formed in the reaction returns into the heteropolyanion either by an electron transfer or by association with the lacunary species (HPA-n-1). Consequently, V⁴⁺ is re-oxidised by oxygen inside HPA-n.

Li et al. also used the same catalyst and added H₂SO₄ as an additive to oxidize cellulose to obtain FA with 28% yield at 443 K and cellulose was completely converted in 9 h. The increase in temperature and acidity (by adding H₂SO₄) increased the FA yield, and the time was shortened to 9 h. Under modified reaction conditions, the conversion of glucose to formic acid could be achieved up to 52% yield within 3 h using 5 mol% H₅PV₂Mo₁₀O₄₀ at 373 K using air as the oxidant.¹⁴¹ Han et al. prepared three phosphovanadomolybdic acids with different contents of vanadium, namely H₄PVMo₁₁O₄₀, H₅PV₂Mo₁₀O₄₀ and H₆PMo₉V₃O₄₀, and three Keggin-type HPA catalysts including two V-free HPAs (H₃PW₁₂O₄₀ and H₃PMo₁₂O₄₀) and one phosphovanadotungstic acid (H₅PV₂W₁₂O₄₀), and evaluated their catalytic performance for the conversion of cellulose.¹⁴² Although a relatively low temperature (~373 K) was used to convert the soluble carbohydrates, a higher reaction temperature (>423 K) was essential for the effective conversion of water-insoluble biomass (e.g. cellulose) with HPAs. H₄PVMo₁₁O₄₀ acts as the most efficient catalyst capable of converting varieties of biomass-derived substrates to formic acid and acetic acid with high selectivity in aqueous medium and oxygen atmosphere. Under optimized reaction conditions, H₄PVMo₁₁O₄₀ gave an exceptionally high yield of formic acid (67.8%) from cellulose.

Recently, Liu et al. observed that heteropolyanion-based ionic liquids with –SO₃H functionalized cations and PMo₁₁VO₄₀⁴⁻ anions showed enhanced activity for catalysing cellulose conversion to FA in the presence of oxygen in water compared to H₄PMo₁₁VO₄₀.¹⁴³ Investigation of the possible pathway of FA production from cellulose indicated that heteropolyanion-based ionic liquids served as bifunctional catalysts with cations catalysing cellulose hydrolysis to glucose and anions catalysing glucose oxidation to FA. Optimization with different as-prepared ionic liquid catalysts resulted in high efficiency for FA production from cellulose conversion, giving a high FA yield of over 50% and a high FA concentration of ~10% in aqueous solution at 453 K and in an oxygen pressure of 1.0 MPa.

Marsh and co-workers developed a catalytic process for efficient production of FA from common carbohydrates via VO₂⁺ formed by dissolving sodium metavanadate in acidic water. The hydrolysis reaction decomposed the polymerized structures of polysaccharides to produce monosaccharides, which were readily oxidized to FA under catalysis of VO₂⁺. Typically, the molar yields of formic acid were 64.9% from cellulose and 63.5% from xylan (hemicellulose) at 433 K.¹⁴⁴ Wang and co-workers used the cheaper and less toxic vanadium salt VOSO₄ as a high-performance catalyst for the transformations of glucose and cellulose into either formic acid or lactic acid (LA) by simply tuning the reaction atmosphere from O₂ to N₂. The catalytic reaction proceeds through the isomerization of glucose to fructose, which subsequently underwent retro-aldol fragmentation to two trioses, that is, glyceraldehyde and 1,3-dihydroxyacetone. The isomerization of these trioses under N₂ led to the formation of lactic acid. The oxidative cleavage of C–C bonds in the intermediates caused by the redox conversion of VO₂⁺/VO²⁺ under aerobic conditions, giving formic acid and CO₂. Addition of an alcohol suppresses the formation of CO₂ and enhances the formic acid yield significantly to 70–75%.¹⁴⁵

From all the observations by different groups, it can be concluded that biomass could be converted to formic acid by vanadyl species (VO₂⁺/VO²⁺) or heteropolyacids (HPAs) containing vanadyl ions. These catalysts work as bifunctional catalysts for the conversion of polysaccharides to monosaccharides as well as the selective conversion of the generated monosaccharides to formic acid and CO₂ under acidic conditions using air or oxygen as the oxidant.

Besides vanadyl-containing catalysts, some people reported the production of FA from biomass in the presence of other catalysts and some additives.^146–148 Gao et al. successfully produced FA with 22% yield from pretreated cellulose under hydrothermal conditions at 483 K for 30 h.¹⁴⁶ A high yield of 75% FA was produced by Jin et al. at 523 K using 120% H₂O₂ from glucose.¹⁴⁷ Choudhary et al. designed a Cu catalyst suitable for the conversion of biomass to lactic acid (LA) and formic acid.¹⁴⁸ They synthesized a magnesia supported copper catalyst by a hydrothermal methodology using CTAB as the capping agent (denoted as Cu–CTAB/MgO). Interestingly, this catalyst boosted the yields of LA and FA dramatically from sugars with decreasing reaction temperature from 523 K to 393 K. Glucose can be converted to lactic acid or formic acid depending on different additives. Lactic acid with a yield of 70% was produced in the presence of NaOH, while FA was produced in the presence of H₂O₂; both were achieved from glucose at 393 K in water using the Cu-CTAB/MgO catalyst, which could be recycled without any significant loss of activity. The catalyst was also found to exhibit excellent activity for the transformation of other sugars.

Despite all efforts, complete and selective conversion of biomass to formic acid and CO₂ is still a challenging task and further work is required to replace the harsh reaction conditions with mild conditions.

3.2 CO₂ hydrogenation

Energy is the most indispensable element of our daily lives, which comes mostly from fossil fuels. Apart from the limited source of these non-renewable fossil fuels, our dependency on them has caused serious environmental concerns, such as greenhouse gases (CO₂) as well as other toxic pollutants. The best way to overcome this problem is to use CO₂ to make high value chemicals. Extensive research has been carried out on the utilization of carbon dioxide to prepare high value chemicals such as formic acid, formaldehyde, methanol, methane, and other long chain organics, and in synthetic chemistry to obtain different products at ambient conditions.^129,149 The conversion of CO₂ to useful chemicals is of significant importance not only due to reductions in greenhouse gases but also to build an energy infrastructure that uses methanol or formic acid etc.^70,150–153 Unlike conversion to formic acid, conversion of CO₂ to formaldehyde, methanol or methane requires harsh conditions which may deactivate the catalysts; hence, these reduction processes are economically unfavourable.

The hydrogenation of CO₂ to formic acid is an important step to establish formic acid as a reversible hydrogen storage material. Several research works have been done in this field and those works are summarized in several excellent review articles as well as book chapters.^{10,32,70,149–161} Hence, we present here only the basic concept of the conversion process and summarize only some important recent research on this topic.

3.2.1. Homogenous catalysts. Pioneer work on the catalytic hydrogenation of CO₂ to formic acid was reported by Inoue et al. in 1976.¹⁶² The reaction was accelerated by adding small amounts of H₂O because the ΔG value is negative in aqueous medium while positive in gaseous phase. Hence, the reaction is favourable in aqueous medium. In addition, the use of bases enhances the catalytic activity because basic additives, such as NaOH, NaHCO₃, Na₂CO₃ and amines, can absorb the generated proton and make the reaction thermodynamically more favourable (eqn (12)–(14)).

CO₂(g) + H₂(g) → HCOOH (l) ΔG = 32.9 kJ mol⁻¹ (12)

CO₂(aq) + H₂(aq) → HCOOH (aq) ΔG = −4.0 kJ mol⁻¹ (13)

CO₂(aq) + H₂(aq) + NH₃(aq) → HCOO⁻ (aq) + NH₄⁺(aq) ΔG = −9.5 kJ mol⁻¹ (14)

Most of the complexes used to catalyze CO₂ hydrogenation are based on group VIII transition metals such as Pd, Ni, Rh, Ru, and Ir. Among these catalysts, Rh, Ru, and most recently Ir complexes were found to be the most effective. Recently, the groups of Himeda and Fujita summarized the homogeneous hydrogenation of CO₂ to formate and methanol as an alternative to photo- and electrochemical CO₂ reduction with major emphasis on a half-sandwich iridium catalyst based on proton-responsive ligands.¹⁶¹

Yang et al. studied the reaction mechanisms for the hydrogenation of carbon dioxide catalyzed by PNP-ligated (PNP = 2,6-bis(di-isopropylphosphinomethyl)pyridine) metal pincer complexes, (PNP)MH_n (M = Ir, Fe and Co), computationally by using density functional theory (DFT).¹⁶³ Iridium is a recently reported high efficiency catalyst for the formation of formic acid from H₂ and CO₂.¹⁶⁴ Fe and Co metal complexes could be used as low-cost non-noble metal catalysts for CO₂ reduction.

Using the rhodium-phosphine complex, [RhCl(mtppms)₃] (mtppms = monosulfonated triphenylphosphine), as a catalyst, free formic acid was produced in the hydrogenation of carbon dioxide dissolved in aqueous sodium formate solutions under H₂ and CO₂ pressure with a water-soluble concentration of sodium formate.¹⁶⁵ The total gas pressure and the pressure ratio of H₂ to CO₂ were the most important factors for the production of HCOOH. Up to 0.13 M concentration of HCOOH was achieved, while there was negligible formic acid production in the absence of sodium formate.

Himeda et al. reported the interconversion between formic acid and H₂/CO₂ using half-sandwich rhodium and ruthenium catalysts with 4,4′-dihydroxy-2,2′-bipyridine (DHBP).¹⁶⁶ In the hydrogenation of CO₂/bicarbonate to formate under basic conditions, activations of the catalysts were caused by the electronic effect of oxyanions generated by the deprotonation of the hydroxyl group. The turnover frequencies of these catalysts increased 65- and 8-fold, respectively, compared to the corresponding unsubstituted bipyridine complexes. The rhodium–DHBP catalyst showed high activity without CO contamination in a relatively wide pH range. These catalytic systems are extremely useful for H₂ storage materials by combining CO₂ hydrogenation with formic acid decomposition.

Maenaka et al. synthesized the [C, N] cyclometalated organoiridium complex, [Ir^III(Cp*)(4-(1H-pyrazol-1-yl-κN²)benzoic acid-κC³)(H₂O)]₂SO₄, as an efficient catalyst for interconversion between hydrogen and formic acid in water at ambient temperature and pressure for both directions depending on pH.¹⁶⁷ The hydrogenation of carbon dioxide occurs in the presence of a catalytic amount of the Ir complex under an atmospheric pressure of H₂ and CO₂ in weakly basic water (pH 7.5) at room temperature, whereas formic acid efficiently decomposes to afford H₂ and CO₂ in acidic water (pH 2.8).

Wesselbaum et al. presented a new idea that allows the continuous-flow hydrogenation of supercritical CO₂ (scCO₂) with integrated product separation from an immobilized catalyst and a stabilizing base to produce pure formic acid in a single processing unit.¹⁶⁸ The concept is based on a biphasic reaction system consisting of scCO₂ as the mobile phase and an ionic liquid as the stationary phase containing the catalyst and a nonvolatile base (Scheme 6). The pure HCO₂H that is free of any cross-contamination can be obtained without interrupting the working catalyst inside the reactor by this method.

Scheme 6 Process for the direct continuous-flow hydrogenation of CO₂ to free formic acid based on a two-phase system with scCO₂ as the extractive mobile phase and an ionic liquid (IL) as the stationary phase containing the catalyst and the stabilizing base. Reprinted with permission from ref. 168. Copyright 2012 Wiley-VCH.

Huff et al. have synthesized the ruthenium pincer complex Ru(PNN)CO(H) (PNN = 6-(di-tert-butylphosphinomethylene)-2-(N,N-diethylaminomethyl)-1,6-dihydropyridine) which is an efficient catalyst for the hydrogenation of carbon dioxide to formate. Stoichiometric studies are presented that support the feasibility of individual steps in a proposed catalytic cycle for this transformation.¹⁶⁹ They explored the influence of base and solvents on the catalytic performance. Under optimized conditions (using diglyme as the solvent and potassium carbonate as the base) up to 23 000 turnovers of formate and a turnover frequency of up to 2200 h⁻¹ were achieved.

The activity of homogeneous ruthenium(II)-1,2-bis(diphenylphosphino)ethane as a catalyst was studied for the equilibrium position in formic acid/amine–CO₂ systems as a function of pressure and temperature under isochoric conditions.¹⁷⁰ It was found that the Ru catalyst was active for both, that is, in formic acid cleavage producing pure hydrogen and CO₂, and in carbon dioxide hydrogenation under basic conditions. High yields of formic acid dehydrogenation into H₂ and CO₂ are favoured by low gas pressures and/or high temperatures, and H₂ uptake is possible at elevated H₂–CO₂ pressures.

Recently, the direct hydrogenation of CO₂ into formic acid using a homogeneous ruthenium catalyst has been studied by Moret et al. in aqueous solution and in dimethyl sulphoxide (DMSO) without any additives.¹⁷¹ In water, at 313 K, 0.2 M formic acid can be obtained under 200 bar, however, in DMSO the same catalyst affords 1.9 M formic acid. In both solvents, the catalyst can be reused multiple times without a decrease in activity. Worldwide demand for formic acid continues to grow, especially in the context of a renewable energy hydrogen carrier, and its production from CO₂ without a base, via direct catalytic carbon dioxide hydrogenation, is considerably more sustainable than the existing routes.

Drake et al. have investigated the role of various common inorganic salts in increasing the productivity for the hydrogenation of CO₂ to formic acid in the presence of various catalysts.¹⁷² Among the various salts, KHCO₃ acted as the best promoter for the hydrogenation of CO₂. The selectivity for formic acid production catalyzed by RuCl₂(PPh₃)(p-cymene) could be increased by 84% to 140% depending on the amount of KHCO₃. The effect was general for other noble-metal catalysts and for one of the most efficient non-noble-metal hydrogenation catalysts shown in Fig. 9. Mechanistic experiments revealed that the reaction likely proceeds via the formation of a metal–carbonate species.

Fig. 9 Catalysts used for hydrogenation of CO₂. Reprinted with permission from ref. 172. Copyright 2013 American Chemical Society.

Hsu et al. have reported the use of a Ru catalyst [(PNNP)(acetonitrile)Ru^II] for the efficient and reversible reduction of CO₂ to formic acid in the presence of DBU as a base.¹⁷³ At a catalyst loading of only 0.075 mol%, both the reduction using H₂ pressure and the decomposition of the formic acid adduct under pressure-free conditions were established.

The homogeneously catalyzed hydrogenation of carbon dioxide to formic acid and its derivates has been well studied. Recently, an increase in product formation was achieved by further development of homogeneous catalysts. Currently, iridium-based catalysts offer the highest catalytic activity known in the hydrogenation of carbon dioxide. Here we summarized various catalyst systems for CO₂ hydrogenation to formic acid (Table 3) as reported in a recent book chapter by Behr et al.³⁸

Table 3 Some homogeneous catalysts for selective CO₂ hydrogenation to formic acid and reaction conditions, data obtained from ref. 38

SN	Catalyst	Solvent	Base/additive	T (K)	P CO2/H2 (bar)	Time (h)	TON (h^−1)	TOF (h^−1)
bipy = 2,2′-bipyridine, cod = cyclooctadiene, CYPO = Cy2PCH2CH2OCH3, DBF = dibutylformamide, DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene, dcpb = Cy2P(CH2)4PCy2, dcpe = 1,2-dicyclohexylphosphinoethane, dhbipy = 4,4′-dihydroxy-2,2′-bipyridine, DHphen = 4,7-dihydroxy-1,10-phenanthroline, dippe = 1,2-(diisopropylphosphino)ethane, dppb = 1,4-bis(diphenylphosphino)butane, dppbts = tetrasulfonated-1,4-bis(diphenylphosphino)butane, dppp = 1,3-bis(diphenylphosphino)propane, hfacac = hexafluoroacetylacetonate, mTPPMS = meta-monosulfonated triphenylphosphine, NBD = bicycle[2.2.1]hepta-2,5-diene, NHC = N-heterocyclic carbine, PP3 = [P(CH2CH2PPh2)3], PTA = 1,3,5-triaza-7-phosphaadamantane, TEtA = triethanolamine, Tp = hydrotris(pyrazolyl)borate, TPPMS = monosulfonated triphenylphosphine, TPPTS = trisulfonated triphenylphosphine.
1	[Ru(Cl2bipy)2(H2O)2] [CF3SO3]2	EtOH	NEt3	323	30/30	8	5000	625
2	RuCl2(PMe3)4	scCO2	NEt3	323	120/80	47	7200	n.a.
3	[RuCl(C6Me6)(DHphen)]Cl	H2O	KOH	393	30/30	24	15 400	642
4	RuCl(OAc)(PMe3)4	scCO2	NEt3	323	120/70	0.3	31 700	95 000
5	[RuCl2(mTPPMS)2]2	H2O	NaHCO3	353	35/60	0.03	320	9600
6	RuH2(PMe3)4	scCO2	NEt3	323	120/85	1	1400	1400
7	Ru2(CO)5(dppm)2	Acetone	NEt3	RT	38/38	21	2160	103
8	RuCl3/PPh3	EtOH	NEt3	333	60/60	5	200	40
9	CpRu(CO)(μ-dppm)Mo(CO)2Cp	Benzene	NEt3	393	30/30	45	43	1
10	TpRuH(PPh3)(CH3CN)	THF	NEt3	373	25/25	16	1815	113
11	RuH2(PPh3)4	Benzene	NEt3	RT	25/25	20	87	4
12	RuH2(PPh3)4	Benzene	Na2CO3	373	25/25	4	169	42
13	[(C5H4(CH2)3NMe2)Ru(dppm)]BF4	THF	None	353	40/40	16	8	0.5
14	[RuCl2(CO)2]n	H2O, iPrOH	NEt3	353	27/81	0.3	400	1300
15	RuCl3(aq)/dppbts	H2O	HNMe2	333	25/25	15	4980	n.a.
16	K[RuCl(EDTA-H)Cl]·2H2O	H2O	None	313	17/3	0.5	n.a.	250
17	RuCl2(PTA4)	H2O	NaHCO3	353	0/60	n.a	n.a	807
18	[(Cl2bipy)2Ru(H2O)2][CF3SO3]2	EtOH	NEt3	423	30/30	8	5000	n.a.
19	[(NHC)2RuCl][PF6]	H2O	KOH	473	20/20	75	23 000	n.a.
20	[(C6Me6)Ru(dhbipy)]	H2O	KOH	393	30/30	8	13 620	4400
21	[(dppm)Ru(H)(Cl)]	H2O	NaHCO3	343	0/50	2	1374	687
22	[RuCl2(TPPMS)2]2	H2O	NaHCO3	298	5/35	2	524	262
23	[Cp*Ru(DHphen)Cl]Cl	H2O	KOH	393	30/30	24	15 400	3600
24	[(η^6-arene)Ru(η^2-N, O–L)(H2O)]BF4	H2O	NEt3	373	50/50	10	400	n.a.
25	RuH2(PMe3)4	scCO2	NEt3	323	120/80	3	1900	630
26	RuH2(PPh3)4	DBF	TEtA	323	30/30	1	698	698
27	Rh(hfacac)(dcpb)	DMSO	NEt3	298	20/20	5	n.a.	1335
28	RhCl(TPPTS)3	H2O	NHMe2	RT	20/20	12	3439	n.a.
29	[RhH(cod)]4/dppb	DMSO	NEt3	RT	20/20	18	2200	375
30	RhCl(TPPTS)3	H2O	NHMe2	354	20/20	0.5	n.a.	7260
31	[RhCl(cod)]2/dippe	DMSO	NEt3	297	40 total	18	205	11
32	RhCl(PPh3)3	Benzene	Na2CO3	373	55/60	3	173	58
33	[Rh(η^2-CYPO)2]BPh4	MeOH	NEt3	328	25/25	7	1000	n.a
34	RhCl3/PPh3	H2O	NHMe2	323	10/10	10	2150	215
35	[Rh(NBD)(PMe2Ph)3]BF4	THF	None	313	48/48	48	128	3
36	RhCl(PPh3)3/PPh3	MeOH	NEt3	298	40/20	20	2700	125
37	RhCl3(aq)/dppbts	H2O	HNMe2	333	25/25	15	3910	n.a.
38	[RhCl(cod)]2/dppb	DMSO	NEt3	RT	20/20	22	1150	52
39	RhCl(TPPTS)3	H2O	NEt3	296	20/20	n.a.	n.a.	1364
40	[Cp*Rh(DHphen)Cl]Cl	H2O	KOH	353	20/20	32	2400	270
41	Rh(NO)(dcpe)	n.a.	DBU	323	1.5/1.5	16	106	n.a.
42	[RhCl(mTPPMS)3]	H2O	HCOONa	323	50/50	20	65	n.a.
43	[RhCl(mTPPMS)3]	H2O	CaCO3	323	20/80	24	300	n.a.
44	[Cp*Ir(dhbipy)Cl]Cl	H2O	KOH	393	30/30	57	42 000	19 0000
45	[Cp*Ir(DHphen)Cl]Cl	H2O	KOH	393	30/30	48	222 000	33 000
46	[Cp*IrCl(DHphen)]Cl	H2O	KOH	393	30/30	10	21 000	2100
47	[(PNP)IrH3]	H2O	KOH	393	30/30	48	3 500 000	73 000
48	[IrI2(AcO)(bis-NHC)]	H2O	KOH	473	30/30	75	190 000	n.a.
49	PdCl2	H2O	KOH	433	n.a./110	3	1580	530
50	PdCl2	H2O	KOH	513	40/106	3	340	n.a.
51	Pd(dppe)2	Benzene	NEt3	383	25/25	20	62	3
52	Pd(dppe)2	Benzene	NaOH	RT	24/24	20	17	0.9
53	PdCl2[P(C6H5)3]2	Benzene	NEt3	RT	50/50	n.a.	15	n.a.
54	FeCl3/dcpe	DMSO	DBU	323	60/40	7.5	113	15.1
55	Fe(BF4)2·6H2O, PP3	MeOH	NaHCO3	353	0/60	20	610	n.a.
56	Co(BF4)2·6H2O, PP3	MeOH	NaHCO3	353	0/60	20	3877	n.a
57	trans-[(tBu-PNP)Fe(H2)(CO)]	H2O	NaOH	393	3.33/6.66	5	788	n.a
58	NiCl2(dcpe)	DMSO	DBU	323	160/40	216	4400	20
59	Ni(dppe)2	Benzene	NEt3	RT	25/25	20	7	0.4
60	TiCl4/Mg	THF	None	RT	1/1	24	15	n.a.
61	MoCl3/dcpe	DMSO	DBU	323	60/40	7.5	63	8

---

Some catalysts are found to be active for reversible CO₂ hydrogenation as well as formic acid dehydrogenation by varying the reaction conditions.^164,174 These systems hold promise for the development of a practical H₂ storage technology based on the reversible catalytic hydrogenation of CO₂. Nozaki and co-workers reported a highly active catalytic hydrogenation of carbon dioxide to form formic acid salts and the dehydrogenation of formic acid salt using a PNP-ligated iridium(III) trihydride complex (Fig. 10). The hydrogenation of carbon dioxide using this PNP-ligated iridium(III) trihydride complex afforded potassium formate in aqueous KOH solution with a TOF of 150 000 h⁻¹ and a TON of 3 500 000, both of the values being the highest ever reported.^164,174 In the presence of triethanolamine as the base, both CO₂ hydrogenation (TON 29 000, TOF 14 000 h⁻¹ at 5.0 MPa at 297 K) and FA dehydrogenation (TON 4900, TOF 1000 h⁻¹ at 333 K) proceed at an appreciable rate, showing that the iridium catalyst can be used for reversible hydrogen storage by changing the reaction conditions only.

Fig. 10 Plausible mechanism for hydrogenation of CO₂ using iridium trihydride complex. Reprinted with permission from ref. 174. Copyright 2011 American Chemical Society.

Pidko et al. reported the highly efficient reversible hydrogenation of carbon dioxide to formates using a ruthenium PNP-pincer catalyst.¹⁷⁵ This catalyst has a very high activity for CO₂ hydrogenation and FA dehydrogenation. If used in combination with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), the catalyst allows the control of hydrogen liberation activity in a narrow temperature interval. The H₂ capacity is dependent on the base strength, and strong bases are required to generate a high acid-to-amine ratio at elevated temperature if the reaction times are short.

Besides the several works on noble-metal based catalysts, some non-precious-metal catalysts based on Fe,^70,176–180 Co,^181–183 Ni,^177,184 Cu,¹⁸⁵ Mo,¹⁷⁷ and Ti (ref. 186) have also been investigated. Beller and co-workers reported the efficient, stable, and easy-to-synthesize non-noble metal complex iron(II)-fluoro-tris(2-(diphenylphosphino)phenyl)phosphino]tetrafluoroborate for the reduction of CO₂ and bicarbonates.¹⁷⁹ The catalytic hydrogenation of the bicarbonates proceeded in good yields with high catalyst productivity and activity (TON > 7500, TOF > 750 h⁻¹).

Beller and co-workers successfully utilized the cationic Co(BF₄)₂·6H₂O/PP₃ catalyst system for the conversion of sodium bicarbonate/CO₂ to sodium formate (TON 687 at 353 K and 3877 at 393 K).¹⁸¹ A cobalt-based catalyst system for the production of formate from CO₂ and H₂ under ambient conditions has been synthesized by Jeletic et al.¹⁸² The complex Co(dmpe)₂H (dmpe = 1,2-bis(dimethylphosphino)ethane) catalyzes the hydrogenation of CO₂, with a TOF of 3400 h⁻¹ at room temperature and 1 atm of 1 : 1 CO₂ : H₂ (74 000 h⁻¹ at 20 atm) in THF. These results highlighted the value of the fundamental thermodynamic properties for the rational design of catalysts. Himeda et al. reported new water-soluble pentamethylcyclopentadienyl cobalt(III) complexes with proton-responsive 4,4′- and 6,6′-dihydroxy-2,2′-bipyridine ligands as efficient catalysts for CO₂ hydrogenation to formate in aqueous bicarbonate media at moderate temperature under a total pressure of 4–5 MPa (CO₂ : H₂ 1 : 1).¹⁸³

Several possible mechanisms are conceivable for CO₂ hydrogenation to formic acid which mainly involves the insertion of CO₂ to the M–H bond followed by the elimination of FA. Th insertion of carbon dioxide into the M–H bond is identical for all mechanisms in which carbon dioxide forms a metal formate complex via η²-coordination and migration of hydrogen (Scheme 7).³⁸

Scheme 7 General mechanism for the homogeneously catalyzed hydrogenation of CO₂ to FA.

Theoretically, it is established that the insertion can occur by M–H bond breaking based on the weak H–CO₂ interaction.^187–189 The elimination of formic acid can proceed via reductive elimination of formic acid, addition of hydrogen to the formate complex, direct hydrogenolysis of the M–O bond without previous oxidative addition to hydrogen or hydrolysis of the metal formate.^{38,162,190,191}

3.2.2. Heterogeneous catalysts. Compared to homogeneous catalysts, there are only a few reports on CO₂ hydrogenation to formic acid over heterogeneous catalysts owing to the unfavorable reaction conditions and low chemoselectivity. For the first time, the hydrogenation of CO₂ to formic acid was reported in 1935 using RANEY® nickel as the catalyst under 200–400 bar hydrogen pressure and 353–423 K.¹⁹² However, one equivalent of amine had to be added in order to shift the thermodynamic equilibrium towards the formation of the product formic acid.

Wiener et al. reported the reduction of carbonates to formic acid under mild reaction conditions using Pd/C as a catalyst.¹⁹³ However, the reaction could not achieve completeness due to the chemical equilibrium between carbonate and formate. Preti et al. examined the activity of some heterogeneous catalysts, such as iron powder, RANEY® nickel and Co, Cu, Ru, Rh, Pd, Ag, Ir, Pt and Au black, for the formation of formic acid by CO₂ hydrogenation.¹⁹⁴ Au black catalyses the CO₂ hydrogenation in the presence of neat NEt₃ to form HCOOH/NEt₃ adducts. Activity studies have been extended to commercially available aurolite (Mintek; 1 wt% Au/TiO₂, extrudates). Formic acid can be separated from HCOOH/NEt₃ with the help of a high-boiling point amine (n-C₆H₁₃)₃N. When coupled with the catalytic HCOOH decomposition to CO-free hydrogen together with easily removable and reusable CO₂, these findings complete the chemical loop for the long-sought-after hydrogen storage with CO₂.

Hao et al. reported the preparation and application of heterogeneous supported ruthenium catalysts.¹⁹⁵ The catalysts were active in the synthesis of formic acid from the hydrogenation of carbon dioxide. Abundant hydroxyl groups, which interact with the ruthenium components, play an important role in the catalytic reactions. Highly dispersed ruthenium hydroxide species enhance the hydrogenation of CO₂, while crystalline RuO₂ species, which are formed during the preparation of the catalyst, restrict the production of formic acid. The best activity of ruthenium hydroxide as a catalyst for the hydrogenation of CO₂ to formic acid was achieved over a γ-Al₂O₃ supported 2.0 wt% ruthenium catalyst, which was prepared in a solution of pH 12.8 with NH₃·H₂O as a titration solvent.

A series of catalysts made of ruthenium loaded on γ-Al₂O₃ and γ-Al₂O₃ nanorods were studied for the hydrogenation of CO₂ to formic acid by Preti et al.¹⁹⁶ Among these catalysts, 2% Ru/Al (n) has the highest activity. The excellent catalytic activity is related to the highly dispersed ruthenium species on the surface of the support and abundant hydroxyl groups of the support.

A series of catalysts made of ruthenium loaded on γ-Al₂O₃ nanorods were prepared and their catalytic performances for the hydrogenation of CO₂ to formic acid have been studied by Na Liu et al.¹⁹⁷ The results revealed that the catalytic activity is determined by the structure of the supported ruthenium oxide species. Th optimal activity of the catalyst for the hydrogenation of CO₂ to formic acid is achieved over a γ-Al₂O₃ nanorod supported 2.0 wt% ruthenium catalyst.

Peng et al. theoretically investigated the hydrogenation of CO₂ to formic acid on transition metal surfaces because they exhibit unique reactivity for the hydrogenation reaction. They explored the potential of subsurface hydrogen for hydrogenating carbon dioxide on Ni(110).¹⁹⁸ A heterogeneous, mesoporous silica-tethered iridium catalyst (Ir-PN/SBA-15) containing a bidentate iminophosphine ligand was reported for the hydrogenation of CO₂ to formic acid in aqueous solution.¹⁹⁹ This catalyst is recyclable and exhibits high activities for formic acid production under mild conditions [333 K, 4.0 MPa total pressure (H₂/CO₂ = 1 : 1)].

The reaction mechanisms of the hydrogenation of carbon dioxide to formic acid over a Cu-alkoxide-functionalized metal–organic framework (MOF) have been investigated theoretically.²⁰⁰ The reaction can proceed via two different pathways, namely, concerted and stepwise mechanisms. In the concerted mechanism, the hydrogenation of CO₂ to formic acid occurs in a single step. It requires a high activation energy of 67.2 kcal mol⁻¹. For the stepwise mechanism, the reaction begins with hydrogen atom abstraction by CO₂ to form a formate intermediate. The intermediate then takes another hydrogen atom to form formic acid. The activation energies are calculated to be 24.2 and 18.3 kcal mol⁻¹ for the first and second steps, respectively. Because of the smaller activation barriers associated with this pathway, it therefore seems to be more favoured than the concerted one.

Ruthenium complexes immobilized on amine-functionalized silica have been synthesized by an in situ synthetic approach.²⁰¹ The catalysts exhibit high activity (TOF = 1384 h⁻¹) with complete (100%) selectivity for CO₂ hydrogenation to formic acid. Immobilization of homogeneous catalysts offers practical advantages such as easy separation and recycling. The combination of the basic ionic liquid 1-(N,N-dimethylaminoethyl)-2,3-dimethylimidazolium trifluoromethane-sulfonate ([mammim][OTf]) and a silica-supported ruthenium complex (“Si”–(CH₂)₃–NH(CSCH₃)–RuCl₃–PPh₃) was found to catalyze CO₂ hydrogenation to formic acid with satisfactory activity and selectivity.²⁰² This immobilized catalyst catalyzes the hydrogenation of CO₂, with a maximum TOF of 103 h⁻¹ at a total pressure of 18 bar (H₂ : CO₂ = 1 : 1) without loss of activity over 5 catalytic cycles.

A reversible heterogeneous Pd catalyst (>3 nm) supported on mesoporous graphitic carbon nitride (Pd/mpg-C₃N₄) for the interconversion of CO₂ and formic acid has been synthesized by Lee et al. (Fig. 11).¹²¹ This catalyst catalyzes the formation of formic acid via CO₂ hydrogenation along with dehydrogenation of formic acid with a turnover frequency of 144 h⁻¹ even in the absence of external bases at room temperature. The basic sites of the support could also induce initial interaction with CO₂ for the synthesis of FA. The governing factors for CO₂ hydrogenation are further elucidated to increase the quantity of the desired formic acid with high selectivity.

Fig. 11 HRTEM analyses: (a and b) images of Pd/mpg-C₃N₄, (c) HADDF-STEM image of Pd/mpg-C₃N₄, and (d) particle size distribution of Pd/mpg-C₃N₄. Reprinted with permission from ref. 121. Copyright 2014 Royal Society of Chemistry.

3.3. Photochemical and enzymatic reductions of CO₂ to formic acid

CO₂ fixation by photochemical and enzymatic reduction is an inexpensive, attractive and green method for the synthesis of organic compounds from CO₂ as the starting material. Although several works have been done on CO₂ fixation, the quantum yields and selectivity of the products based on the reduction of CO₂ are still low.^203–206 In artificial photocatalytic CO₂ reduction, the reduction of CO₂ to CO is more favourable than to formic acid or other products. In principle, the synthesis of formic acid by photocatalytic reduction of CO₂ is mainly done by using the formate dehydrogenase (FDH) enzyme which catalyzes the reversible conversion of formic acid to CO₂. NADH (nicotinamide adenine dinucleotide) dependent enzymes can be used to reduce the formed N,N′-alkyl-4,4′- or -2,2′-bipyridinium salt molecules as artificial substrates instead of NADH or NAD⁺. For example, FDH from Saccharomyces cerevisiae is a NADH-dependent enzyme and uses the various reduced forms of the bipyridinium salt molecules as substrates for the conversion of CO₂ to formic acid.

Willner and co-workers reported the enzymatic formic acid synthesis from HCO₃⁻ with FDH and the photoreduction of various 4,4′- or 2,2′-bipyridinium salts by a system containing [Ru(bpy)₃]²⁺ (bpy = 2,2′-bipyridine) as a photosensitizer and mercaptoethanol (RSH) as an electron donor.^207–210 The efficiency of formic acid synthesis depends on the nature of the bipyridinium salts, and the quantum yield of formic acid from HCO₃⁻ is in the range of 0.5–1.6%. Schuchmann et al. reported the discovery of a bacterial hydrogen-dependent carbon dioxide reductase (HDCR) that directly uses H₂ for the interconversion of CO₂ to formate from the acetogenic bacterium Acetobacterium woodii.²¹¹ They targeted this bacterium because it grows with H₂ + CO₂ under ambient conditions using an ancient pathway for CO₂ fixation and energy conservation. It lives under extreme energy limitation and directly uses H₂ for CO₂ reduction which would be beneficial for the cell.

Alissandratos et al. reported the biological conversion of CO₂ and H₂ into formate by utilising engineered whole-cell biocatalysts.²¹²Escherichia coli. JM109(DE3) cells transformed for the overexpression of either native formate dehydrogenase (FDH), the FDH from Clostridium carboxidivorans, or genes from Pyrococcus furiosus and Methanobacterium thermoformicicum predicted to express FDH based on their similarity to known FDH genes were all able to produce levels of formate well above the background, when present with H₂ and CO₂.

Kodaka et al. reported the direct synthesis of formic acid from CO₂ gas with the system consisting of triethanolamine (TEOA), [Ru(bpy)₃]²⁺, various bipyridinium salts (1,1′-dialkyl-4,4′-bipyridinium salts or 1,1′-dialkyl-2,2′-bipyridinium salts) and FDH.²¹³ The effects of the structure and redox potentials of various bipyridinium salts have been investigated for the synthesis of formic acid from CO₂. After 7 h of irradiation, the formation of formic acid in the concentration range of 0.2–0.95 mM was achieved from a sample solution containing 0.5 M TEOA, 0.5 mM [Ru(bpy)₃]²⁺, 3.0 mM bipyridinium salt, and 8 mg of FDH. The best activity for formic acid synthesis from CO₂ was observed by using N,N′-trimethylene-2,2′-bipyridinium salt. The yield of formic acid depends on the redox potentials of the bipyridinium salts.

Amao and Miyatani synthesized formic acid from HCO₃⁻ with FDH and the photoreduction of N,N′-dimethyl-4,4′-bipyridinium (MV²⁺).^214–216 They used a water-soluble zinc porphyrin (zinc tetrakis(4-methylpyridyl)porphyrin, ZnTMPyP) as a photosensitizer. When the sample solution containing 0.3 M TEOA, 9.0 μM ZnTMPyP⁴⁺, 15 mM MV²⁺, 20 units of FDH and 1.0 mM NaHCO₃ was irradiated, 60 μm of formic acid (estimated yield 6%) was produced after 3 h of irradiation. In another report, they also introduced a system for the synthesis of formic acid from CO₂ gas with FDH, MV²⁺, NADPH as an electron relay and Mg chlorophyll-a (MgChl-a) as a visible light photosensitizer (Fig. 12).²¹⁷ After 4 h of irradiation, the formation of 56 μM formic acid was achieved from the sample solution containing NADPH (3.0 mM), MgChl-a (9.0 μM), MV²⁺ (0.1 mM), FDH (5 units), and CO₂ gas (saturated) at 303 K. Further, it has been observed that formic acid was not produced in the absence of any of the components.

Fig. 12 Photochemical CO₂ reduction system that combines the photoreduction of NAD⁺ as an electron relay by the photosensitization of dye molecules (P) and CO₂ reduction with FDH. Reprinted with permission from ref. 217. Copyright 2006 Elsevier.

Sato and co-workers reported the immobilization of a molecular CO₂ reduction catalyst onto a semiconductor surface for engineering hybrid photocatalytic materials.^218–222 The ruthenium catalysts were grafted onto different p-type semiconductors such as N-doped Ta₂O₅, Zn doped InP, etc. via covalent linkages or via a polymerization process. It was reported that the robustness of a hybrid system depends strongly on the choice of grafting linkages. The [Ru(dpbpy)(bpy)(CO)₂]/N–Ta₂O₅ (dpbpy = 4,4′-diphosphonate-2,2′-bipyridine, bpy = 2,2′-bipyridine) system with a phosphate linker displayed a 5 times higher rate of CO₂ reduction to HCOOH (in acetonitrile solution using visible light) compared with the [Ru(dcbpy)(bpy)(CO)₂]/N–Ta₂O₅ (dcbpy: 4,4′-dicarboxy-2,2′-bipyridine, bpy = 2,2′-bipyridine) system with a carbonate linker.^218,221 By using a Deronzier's-type ruthenium bipyridine-based polymeric film catalyst, a robust system adapted for working in aqueous solution was achieved.²²¹

Recently, Ishitani and co-workers investigated the photocatalytic reduction of CO₂ in an aqueous solution using a supramolecular binuclear Ru(II)–Re(I) complex, of which the Ru(II) photosensitizer and Re(I) catalyst units are connected with a bridging ligand (Fig. 13).²²³ Sodium ascorbate was used as an electron donor. Formic acid is the main product of the photochemical reaction in aqueous solution (TON = 25, selectivity = 83%), whereas CO is the main product in DMF/TEOA.

Fig. 13 Reaction pathways of the photochemical reduction of CO₂ to HCOOH using supramolecular Ru(II)–Re(I) bimetallic complex with sodium ascorbate in an aqueous solution. Adopted with permission from ref. 223. Copyright 2015 American Chemical Society.

4. Practical set up for formic acid-based hydrogen storage

A sustainable cycle for energy storage based on formic acid and carbon dioxide can be designed as shown in Fig. 14.³⁴

Fig. 14 A catalytic cycle for hydrogen storage in formic acid.

For energy storage, carbon dioxide is converted to formic acid or a formate derivative either electrochemically^224,225 or by catalytic hydrogenation.^32,156,226 The resulting material is a liquid at ambient conditions, either pure formic acid, an adduct containing formic acid, or an inorganic formate in aqueous solution, and can thus be handled, stored, and transported easily. On the other side of the cycle, energy is released either in a direct formic acid fuel cell, or by selective on-demand decomposition to carbon dioxide and hydrogen, which can be used directly in an appropriate hydrogen oxygen fuel cell.^227–232 If pure hydrogen is required, the gases may also be separated using membrane techniques.²³³ The hydrogen content of pure formic acid is 43 g kg⁻¹ or 52 g L⁻¹; its release from formic acid is thermodynamically downhill by ΔG° = −32.8 kJ mol⁻¹ at room temperature.¹⁵⁶ As early as 1978, the electrochemical reduction of carbon dioxide to formic acid and its subsequent decomposition on Pd/C for energy storage were proposed by Williams et al.²³⁴ Later on, a system for solar energy conversion by reduction of aqueous carbonate was proposed by Halmann in 1983.²³⁵ In 1986, a similar concept was described by Wiener et al. They proposed the use of a Pd/C catalyst to decompose aqueous formate solutions to obtain hydrogen.^236–238 However, both approaches have not led to any application. Two decades later, the research on the use of carbon dioxide for energy storage was resumed, which has been continued in recent years.^{6,34,61,239–242}

To identify the main technological obstacles on the way to an efficient use of FA as a hydrogen source and storage material, a hydrogen generator based on the decomposition of FA over a homogeneous Ru(II) catalyst was designed and built by Laurenczy and co-workers (Fig. 15).⁶ The hydrogen generator successfully met the target power output of 1 kW_el or roughly 30 L min⁻¹ H₂/CO₂ assuming 50% fuel cell efficiency. The system was designed around a 5 L reaction vessel with electrical surface heating elements, containing 11.8 g of RuCl₃·3H₂O (0.045 mol) and 51.2 g of Na₃TPPTS (Na₃TPPTS sodium salt of meta-trisulfonated triphenylphosphine) (0.090 mol) catalyst in 1.5 L of aqueous HCOONa solution (1 M). The free reactor volume (~3.5 L) is used as a buffer in the case of rapid changes in the product flow rate set point as well as for gas–liquid separation and effectively prevents foam or spray from reaching the product gas outlet.

Fig. 15 A hydrogen generator designed and built in laboratories of Laurenczy for a power output of 1 kW_el (30 L min⁻¹ H₂/CO₂): (1) formic acid tank; (2) pump; (3) tube in tube heat exchanger; (4) reactor; (5) heat exchanger; (6 and 7) condensate separators; (8) mass flow controller. Adopted with permission from ref. 6. Copyright 2012 Royal Society of Chemistry.

A standard mass flow controller is utilized to control the product flow rate, while a controlled formic acid feed flow is utilized to maintain the reaction pressure. To minimize water vapour entrainment in the product gas, the system also comprises a product gas cooler with a recycle condensate separator. As the control concept does not comprise the measurement of pH or formic acid concentration in the catalyst solution, the reaction has to run at a large catalyst excess for reasons of process safety. This setup has been used safely for over one year without the need to exchange or replenish the catalyst solution.

In 2013, Beller et al. designed Ru(II)/dppe as the catalyst system for continuous hydrogen production from formic acid (Fig. 16).^243,244 The chemoselectivity is very high, resulting in very low CO contents of less than 10 ppm, which allows for its direct application for hydrogen generated in proton-exchange membrane (PEM) fuel cells. They presented the concept and insight into a setup for the continuous production of hydrogen from FA–amine adducts under mild conditions. This setup allows for a more secure system and a longer continuous hydrogen flow of up to 47 L h⁻¹. The continuous setup consists of four main parts (Fig. 16): a high pressure reactor, an FA dosage system, a gas cleaning unit, and a quantification and analysis part for the resulting gas mixture. The central parts to control the dosage of FA are the pump and the liquid flow meter. The reactor, a stainless-steel autoclave, is connected to a condenser, a gas filter, and an active carbon column. These will remove any traces of discharged liquids or vapors. All components are commercially available. To produce hydrogen on demand, a certain amount of FA is dosed into the reactor containing the catalyst and the amine, and immediately, it is converted to hydrogen and carbon dioxide, in the same way that gasoline is consumed in an Otto engine. This allows for the use of the full hydrogen density of FA of 4.4 wt% (not including the surrounding system) because only the acid has to be stored in the tank.

Fig. 16 Schematic representation of the setup used for continuous hydrogen production. Reprinted with permission from ref. 243. Copyright 2013 Wiley-VCH.

Beller et al. also developed a reversible “hydrogen battery” which allowed reversible fast hydrogen loading (even at RT) and unprecedented unloading (TON for H₂ liberation of 800 000).²⁴⁴ For this purpose, they used the defined [RuH₂(dppm)₂] complex instead of the in situ catalyst system. With this catalyst, hydrogenation of carbon dioxide and hydrogen release from the resulting formic acid were possible up to eight consecutive cycles at room temperature.

The group of Plietker reported a rechargeable H₂ battery based on the principle of the Ru-catalyzed hydrogenation of CO₂ to formic acid (charging process) and the Ru-catalyzed decomposition of formic acid to CO₂ and H₂ (discharging process).²⁴⁵ Both processes are driven by the same catalyst at elevated temperature either under pressure (charging process) or pressure-free conditions (discharging process). Up to five charging–discharging cycles were performed without a decrease in storage capacity. The resulting CO₂/H₂ mixture is free of CO and can be employed directly in fuel-cell technology (Fig. 17).

Fig. 17 Schematic for charging and discharging process developed by Plietker et al. and charging and discharging cycles of the H₂ battery. Reprinted with permission from ref. 245. Copyright 2013 Wiley-VCH.

Some rechargeable hydrogen batteries were developed based on the selective hydrogenation of bicarbonates and the selective dehydrogenation of formates, instead of the hydrogenation of CO₂ and the dehydrogenation of formic acid. Beller and co-workers have presented a ruthenium-based catalyst system ([{RuCl₂(benzene)}₂]/dppm (Ru/P = 1 : 6)) for the selective hydrogenation of bicarbonates and the selective dehydrogenation of formates (Fig. 18).²⁴⁶ It was demonstrated that the two reactions can be coupled leading to a reversible hydrogen-storage system. The hydrogenation of NaHCO₃ to sodium formate was performed in 96% yield at 343 K in water/THF without additional CO₂. The dehydrogenation of sodium formate was achieved with high conversion (>90%) at ambient temperature (303 K). In contrast to the dehydrogenation of formic acid/amine adducts, this process is amine-free.

Fig. 18 Reversible hydrogen storage with a bicarbonate/formate system: catalytic hydrogenation of bicarbonate and hydrogen release from formate with the same Ru/dppm in situ catalyst. Reaction conditions: dehydrogenation (DH): 40 mL of DMF, 10 mL of H₂O, 303 K, 24–48 h; hydrogenation (H): 20 mL of H₂O, 10 mL of THF, 80 bar H₂, 24 h; sequence I: 71 μmol of [{RuCl₂(benzene)}₂], 0.43 mmol of dppm, 39.6 mmol of NaHCO₂; sequence II: 10.4 μmol of [{RuCl₂(benzene)}₂], 41.6 μmol of dppm, 23.8 mmol of NaHCO₃. Reprinted with permission from ref. 246. Copyright 2013 Wiley-VCH.

Cao and co-workers developed the first heterogeneous catalytic system (1 wt% Pd/r-GO, Pd NPs supported on reduced graphene (rGO)) for the efficient reversible dehydrogenation of aqueous formates to bicarbonates.²⁴⁷ The r-GO nanosheets are a distinct and powerful support allowing for the generation of highly strained Pd NPs that enable facile and repetitive formate/bicarbonate interconversion under mild aqueous conditions. By controlling the temperature and the hydrogen pressure, the reversible catalytic transformations can be repeated nearly quantitatively six times with almost no loss of efficiency. This could lead to the construction of a simple, truly rechargeable HCOOK/KHCO₃-based hydrogen storage device.

An ammonium bicarbonate/formate redox equilibrium system has also been demonstrated for reversible hydrogen storage and evolution using the same Pd/AC nanocatalyst by Lin and co-workers.²⁴⁸ The reaction temperature and H₂ pressure are key factors in controlling the hydrogen storage–evolution equilibrium in this system. Up to 96% yield of ammonium formate was achieved when the hydrogenation reaction was carried out at room temperature at an initial hydrogen pressure of 2.75 MPa, whereas nearly 100% hydrogen yield was obtained from the dehydrogenation of ammonium formate at 353 K at an initial nitrogen pressure of 0.1 MPa. Compared to the homogeneous catalytic system, this heterogeneous system has the following advantages: no organic solvents or inorganic additives are needed; high volumetric energy density of hydrogen (stored in concentrated ammonium formate aqueous solutions) is achieved; the solid salts or their aqueous solutions can be easily transported and distributed; and the Pd/AC catalyst is stable, being more easily recycled and handled than homogeneous catalysts.

Further work on the design and testing of the hydrogen battery device based on a reversible catalytic system for formic acid dehydrogenation/hydrogenation of CO₂ as well as dehydrogenation of formate/hydrogenation of bicarbonates is still under way.

5. Conclusions and perspectives

In this perspective, we have focused on the application of formic acid as a renewable hydrogen energy source for the future. In addition, we have tried to focus on formic acid production from various processes and, at the same time, the possibility of its regeneration from greenhouse gas CO₂ has also been explored. We have attempted to present an up-to-date overview of this rapidly growing field of research. This subject area is very active and many papers are published every year (even during the preparation of this perspective) by chemists, physical and materials scientists. Therefore, it is difficult to take into account all the contributions to this field in a limited number of pages. We apologize if some significant contributions have been left out.

In the first part, we have summarized the recent developments in formic acid decomposition reactions catalysed either by homogeneous or heterogeneous catalysts in the presence and absence of base and other additives. For improvement of catalytic performance, it is required to facilitate the activation of the C–H bond of the formate anion which is anticipated to be the rate-limiting step. Majority of homogeneous catalytic systems are based on Ru/Rh/Ir/Fe complexes, whereas Pd is the main constituent of heterogeneous catalysts.

Extensive work on the homogeneous catalytic decomposition of FA has been done by Beller and Leuranczy et al. Thanks to the people working in the field to prepare low-cost and efficient catalysts for hydrogen generation from formic acid. For homogeneous catalysts, β-hydride elimination is the rate-determining step of FA dehydrogenation. Thus, lowering the activation energy of β-hydride elimination is the key to success in enhancing the activity and selectivity of FA dehydrogenation. Scientists achieved this goal by modifying different parameters of the catalyst systems e.g. metal centre, design of ligands, reaction temperature, solvent medium, additives (phosphine/base/salts), etc. One of the recent approaches, as suggested by Chauvier et al., is to use a Lewis acid capable of converting the formate anion to CO₂, with a low activation energy and negligible exergonicity.²⁴⁹

As obvious for most of the heterogeneous catalytic processes, the activity/selectivity of the heterogeneous catalytic FA dehydrogenation process also depends upon the sizes, morphologies, distributions, and surface states of NPs, capping agents, and catalyst supports, for bi-/tri-metallic NP counter elements. Generally, smaller size particles have better catalytic activity because of their large surface area but they easily aggregate due to their high surface energy and so induce low activity. Hence, a suitable capping agent or support is required for the dispersion of NPs to avoid the aggregation without severe deactivation of the surface of NPs by strong affiliation of the capping agent or support. In the FA dehydrogenation process, unsupported Pd catalysts are inactive due to the aggregation of NPs as well as poisoning of the surface by CO formed upon FA decomposition. Good activity with high selectivity for FA dehydrogenation was accomplished by immobilization of Pd NPs on/in carbon/graphene/silica/zirconia/ceria supports, etc. These supports not only facilitate the dispersion of metal NPs but also the catalytic CO oxidation at the metal–support interface, and thus prevent the deactivation of the catalyst by CO poisoning.

Similarly, in bimetallic NPs, the second component such as Au or Ag helps to diminish CO production by FA decomposition. As a result, the reaction proceeds via FA dehydrogenation with enhanced activity and selectivity. To further understand/explain the mechanism involved in formic acid decomposition on the surface, density functional theory approaches have also been adopted.

In the second part, we tried to cover various methods to prepare formic acid with special emphasis on biomass conversion to formic acid as well as CO₂ hydrogenation by different methods. Wasserscheid and co-workers developed a method to obtain formic acid and CO₂ selectively by different types of biomasses either water-soluble or insoluble. By varying the reaction conditions, such as pH, temperature or biomass substrate, they were successful in obtaining formic acid with up to 60% selectivity. We have also discussed CO₂ reduction or hydrogenation to formic acid in detail. CO₂ hydrogenation to FA depends on the initial pressure of the reactant (H₂ and CO₂), solvent, etc. Basic conditions favour the formation of FA in high yield owing to the neutralization of the acid product. Further, the presence of basic sites on supported/heterogeneous catalysts induces initial interaction with CO₂ for the synthesis of FA. Better catalytic performance could be achieved by tuning such interactions.

Some catalyst systems are active for both dehydrogenation of formic acid and hydrogenation of CO₂ by varying the reaction conditions only, thus providing an opportunity for the charge/discharge of hydrogen fuel cells similar to other rechargeable batteries. Beller and others developed few rechargeable hydrogen batteries based on these reversible catalyst systems, as summarized in the third part.

Besides the several advantages, the formic acid-based hydrogen storage system has only 4.4 wt% hydrogen content which gets further decreased by addition of amine as well as solvent to around 1%. To meet practical applications, further work is needed for the development of low-cost, more efficient, reversible catalyst systems. Moreover, the development of photocatalytic systems for both processes (fixation and evolution of hydrogen to/from formic acid), which employ sustainable/renewable energy resources, is highly desirable.

Acknowledgements

The authors thank the reviewers for their valuable suggestions. AK thanks DST for financial support. AKS and SS acknowledge the UGC for the DSK Postdoctoral Fellowship. The authors are thankful to those who have worked in this area; their names can be found in the references section. AKS is grateful to Prof. Q. Xu, AIST, Ikeda, Osaka, for his generous support and guidance during his postdoc in Japan.

Notes and references

1. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353–358.
2. J. Graetz, Chem. Soc. Rev., 2009, 38, 73–82.
3. U. B. Demirci and P. Miele, Energy Environ. Sci., 2011, 4, 3334–3341.
4. B. Loges, A. Boddien, F. Gartner, H. Junge and M. Beller, Top. Catal., 2010, 53, 902–914.
5. P. Makowski, A. Thomas, P. Kuhn and F. Goettman, Energy Environ. Sci., 2009, 2, 480–490.
6. M. Grasemann and G. Laurenczy, Energy Environ. Sci., 2012, 5, 8171–8181.
7. F. Joó, ChemSusChem, 2008, 1, 805–808.
8. A. K. Singh and Q. Xu, ChemCatChem, 2013, 5, 652–676.
9. M. Yadav and Q. Xu, Energy Environ. Sci., 2012, 5, 9698–9725.
10. E. Fujita, J. T. Muckerman and Y. Himeda, Biochim. Biophys. Acta, 2013, 1827, 1031–1038.
11. S. Enthaler, J. von Langermann and T. Schmidt, Energy Environ. Sci., 2010, 3, 1207–1217.
12. S. L. Li and Q. Xu, Energy Environ. Sci., 2013, 6, 1656–1683.
13. A. W. C. van den Berg and C. O. Areán, Chem. Commun., 2008, 668–681.
14. F. Amaseder and G. Krainz, Liquid hydrogen storage systems developed and manufactured for the first time for customer cars. In: SAE paper 2006-01-0432, 2006.
15. S. M. Aceves, F. Espinosa-Loza, E. Ledesma-Orozco, T. O. Ross, A. H. Weisberg, T. C. Brunner and O. Kircher, Int. J. Hydrogen Energy, 2010, 35, 1219–1226.
16. B. Sakintuna and Y. Yürüm, Ind. Eng. Chem. Res., 2005, 44, 2893–2902.
17. A. C. Dillon, K. M. Jones, T. A. Bekkedahl, C. H. Kiang, D. S. Bethune and M. J. Heben, Nature, 1997, 386, 377–379.
18. A. D. Lueking, R. T. Yang, N. M. Rodríguez and R. T. K. Baker, Langmuir, 2004, 20, 714–721.
19. W. Chaikittisilp, K. Ariga and Y. Yamauchi, J. Mater. Chem. A, 2013, 1, 14–19.
20. J. Weitkamp, M. Fritz and S. Ernst, Int. J. Hydrogen Energy, 1995, 20, 967–970.
21. A. Zuttle, Mater. Today, 2003, 6, 24–33.
22. H. Lee, J. W. Lee, D. Y. Kim, J. Park, Y. Seo, H. Zeng, I. L. Moudrakovskl, C. I. Ratcliffe and J. A. Ripmeester, Nature, 2005, 434, 743–746.
23. Y. H. Hu and E. Ruckenstein, Angew. Chem., Int. Ed., 2006, 45, 2011–2013.
24. J. L. C. Roswell and O. M. Yaghi, Angew. Chem., Int. Ed., 2005, 44, 4670–4679.
25. M. P. Suh, H. J. Park, T. K. Prasad and D.-W. Lim, Chem. Rev., 2012, 112, 782–835.
26. M. Dinca and J. R. Long, J. Am. Chem. Soc., 2005, 127, 9376–9377.
27. X. Lin, J. Jia, X. Zhao, K. M. Thomas, A. J. Blake, G. S. Walker, N. R. Champness, P. Hubberstey and M. Schröder, Angew. Chem., Int. Ed., 2006, 45, 7358–7364.
28. D. R. Hoffman, Ant venoms, Curr. Opin. Allergy Clin. Immunol., 2010, 10, 342–346.
29. M. Carmo, D. L. Fritz, J. Mergel and D. Stolten, Int. J. Hydrogen Energy, 2013, 38, 4901–4934.
30. M. E. Boot-Handford, J. C. Abanades, E. J. Anthony, M. J. Blunt, S. Brandani, N. M. Dowell, J. R. Fernández, M.-C. Ferrari, R. Gross, J. P. Hallett, R. S. Haszeldine, P. Heptonstall, A. Lyngfelt, Z. Makuch, E. Mangano, R. T. J. Porter, M. Pourkashanian, G. T. Rochelle, N. Shah, J. G. Yao and P. S. Fennell, Energy Environ. Sci., 2014, 7, 130–189.
31. Y. Yasaka, H. Yoshida, C. Wakai, N. Matubayasi and M. Nakahara, J. Phys. Chem. A, 2006, 110, 11082–11090.
32. P. G. Jessop, F. Joó and C.-C. Tai, Coord. Chem. Rev., 2004, 248, 2425–2442.
33. J. A. Turner, Science, 2004, 305, 972–974.
34. T. C. Johnson, D. J. Morris and M. Wills, Chem. Soc. Rev., 2010, 39, 81–88.
35. Office of Energy Efficiency and Renewable Energy; The Freedom CAR and Fuel Partnership, Targets For Onboard Hydrogen Storage Systems For Light-duty Vehicles, September 2009, http://www1.eere.energy.gov.
36. J. Yang, A. Sudik, C. Wolverton and D. J. Siegel, Chem. Soc. Rev., 2010, 39, 656–675.
37. A. F. Dalebrook, W. Gan, M. Grasemann, S. Moret and G. Laurenczy, Chem. Commun., 2013, 49, 8735–8751.
38. A. Behr and K. Nowakowski, Adv. Inorg. Chem., 2014, 66, 223–258.
39. D. Teichmann, W. Arlt, P. Wasserscheid and R. Freymann, Energy Environ. Sci., 2011, 4, 2767–2773.
40. D. Teichmann, W. Arlt and P. Wasserscheid, Int. J. Hydrogen Energy, 2012, 37, 18118–18132.
41. M. Nielsen, E. Alberico, W. Baumann, H.-J. Drexler, H. Junge, S. Gladiali and M. Beller, Nature, 2013, 495, 85–89.
42. A. Monney, E. Barsch, P. Sponholz, H. Junge, R. Ludwig and M. Beller, Chem. Commun., 2014, 50, 707–709.
43. X. Yu and P. G. Pickup, J. Power Sources, 2008, 182, 124–132.
44. R. S. Coffey, Chem. Commun., 1967, 923–924.
45. S. H. Strauss, K. H. Whitmire and D. F. Shriver, J. Organomet. Chem., 1979, 174, C59–C62.
46. R. S. Paonessa and W. C. Trogler, J. Am. Chem. Soc., 1982, 104, 3529–3530.
47. R. B. King and N. K. Bhattacharyya, Inorg. Chim. Acta, 1995, 237, 65–69.
48. J. H. Shin, D. G. Churchill and G. Parkin, J. Organomet. Chem., 2002, 642, 9–15.
49. Y. Gao, J. Kuncheria, G. P. A. Yap and R. J. Puddephatt, Chem. Commun., 1998, 2365–2366.
50. Y. Gao, J. K. Kuncheria, H. A. Jenkins, R. J. Puddephatt and G. P. A. Yap, J. Chem. Soc., Dalton Trans., 2000, 3212–3217.
51. B. Loges, A. Boddien, H. Junge and M. Beller, Angew. Chem., Int. Ed., 2008, 47, 3962–3965.
52. A. Boddien, B. Loges, H. Junge and M. Beller, ChemSusChem, 2008, 1, 751–758.
53. H. Junge, A. Boddien, F. Capitta, B. Loges, J. R. Noyes, S. Gladiali and M. Beller, Tetrahedron Lett., 2009, 50, 1603–1606.
54. X. Li, X. Ma, F. Shi and Y. Deng, ChemSusChem, 2010, 3, 71–74.
55. X. Li, F. Shi, X. Ma, L. Lu and Y. Deng, J. Fuel Chem. Technol., 2010, 38, 544–553.
56. M. E. M. Berger, D. Assenbaum, N. Taccardi, E. Spiecker and P. Wasserscheid, Green Chem., 2011, 13, 1411–1415.
57. C. Fellay, P. J. Dyson and G. Laurenczy, Angew. Chem., Int. Ed., 2008, 47, 3966–3968.
58. C. Fellay, N. Yan, P. J. Dyson and G. Laurenczy, Chem. – Eur. J., 2009, 15, 3752–3760.
59. W. Gan, P. J. Dyson and G. Laurenczy, React. Kinet. Catal. Lett., 2009, 98, 205–213.
60. D. J. Morris, G. J. Clarkson and M. Wills, Organometallics, 2009, 28, 4133–4140.
61. A. Majewski, D. J. Morris, K. Kendall and M. Wills, ChemSusChem, 2010, 3, 431–434.
62. I. Mellone, M. Peruzzini, L. Rosi, D. Mellmann, H. Junge, M. Beller and L. Gonsalvi, Dalton Trans., 2013, 42, 2495–2501.
63. G. Manca, I. Mellone, F. Bertini, M. Peruzzini, L. Rosi, D. Mellmann, H. Junge, M. Beller, A. Ienco and L. Gonsalvi, Organometallics, 2013, 32, 7053–7064.
64. S. Enthaler, H. Junge, A. Fischer, A. Kammer, S. Krackl and J. D. Epping, Polym. Chem., 2013, 4, 2741–2746.
65. S. Fukuzumi, T. Kobayashi and T. Suenobu, ChemSusChem, 2008, 1, 827–834.
66. Y. Himeda, Green Chem., 2009, 11, 2018–2022.
67. S. Fukuzumi, T. Kobayashi and T. Suenobu, J. Am. Chem. Soc., 2010, 132, 1496–1497.
68. S. Oldenhof, B. de Bruin, M. Lutz, M. A. Siegler, F. W. Patureau, J. I. van der Vlugt and J. N. H. Reek, Chem. – Eur. J., 2013, 19, 11507–11511.
69. A. Boddien, B. Loges, F. Gärtner, C. Torborg, K. Fumino, H. Junge, R. Ludwig and M. Beller, J. Am. Chem. Soc., 2010, 132, 8924–8934.
70. A. Boddien, D. Mellmann, F. Gärtner, R. Jackstell, H. Junge, P. J. Dyson, G. Laurenczy, R. Ludwig and M. Beller, Science, 2011, 333, 1733–1734.
71. D. Mellmann, E. Barsch, M. Bauer, K. Grabow, A. Boddien, A. Kammer, P. Sponholz, U. Bentrup, R. Jackstell, H. Junge, G. Laurenczy, R. Ludwig and M. Beller, Chem. – Eur. J., 2014, 20, 13589–13602.
72. T. Zell, B. Butschke, Y. Ben-David and D. Milstein, Chem. – Eur. J., 2013, 19, 8068–8072.
73. E. A. Bielinski, P. O. Lagaditis, Y. Zhang, B. Q. Mercado, C. Würtele, W. H. Bernskoetter, N. Hazari and S. Schneider, J. Am. Chem. Soc., 2014, 136, 10234–10237.
74. T. W. Myers and L. A. Berben, Chem. Sci., 2014, 5, 2771–2777.
75. K. Hirota, K. Kuwata and Y. Nakai, Bull. Chem. Soc. Jpn., 1958, 31, 861–864.
76. D. H. S. Ying and R. J. Madix, J. Catal., 1980, 61, 48–56.
77. Y. K. Sun, J. J. Vajo, C. Y. Chan and W. H. Weinberg, J. Vac. Sci. Technol., A, 1988, 6, 854–855.
78. X. D. Peng and M. A. Barteau, Catal. Lett., 1991, 7, 395–402.
79. N. Aas, Y. Li and M. Bowker, J. Phys.: Condens. Matter, 1991, 3, S281–S286.
80. P. A. Dilara and J. M. Vohs, J. Phys. Chem., 1993, 97, 12919–12923.
81. V. A. Gercher and D. F. Cox, Surf. Sci., 1994, 312, 106–114.
82. J. Stubenrauch, E. Brosha and J. M. Vohs, Catal. Today, 1996, 28, 431–444.
83. R. Larsson, M. H. Jamroz and M. A. Borowiak, J. Mol. Catal. A: Chem., 1998, 129, 41–51.
84. A. Bandara, J. Kubota and A. Wada, J. Phys. Chem. B, 1997, 101, 361–368.
85. J. Rasko, T. Kecskes and J. Kiss, J. Catal., 2004, 224, 261–268.
86. T. Shido and Y. Iwasawa, J. Catal., 1993, 141, 71–81.
87. G. Jacobs, P. M. Patterson, U. M. Graham, A. C. Crawford and B. H. Davis, Int. J. Hydrogen Energy, 2005, 30, 1265–1276.
88. G. Jacobs, P. M. Patterson, U. M. Graham, A. C. Crawford, A. Dozier and B. H. Davis, J. Catal., 2005, 235, 79–91.
89. M. Ojeda and E. Iglesia, Angew. Chem., Int. Ed., 2009, 48, 4800–4803.
90. D. A. Bulusheva, S. Beloshapkin and J. R. H. Ross, Catal. Today, 2010, 154, 7–12.
91. F. Solymosi, Á. Koós, N. Liliom and I. Ugrai, J. Catal., 2011, 279, 213–219.
92. G. Halasi, G. Schubert and F. Solymosi, Catal. Lett., 2011, 142, 218–223.
93. D. A. Bulushev, L. Jia, S. Beloshapkin and J. R. H. Ross, Chem. Commun., 2012, 48, 4184–4186.
94. X. Zhou, Y. Huang, W. Xing, C. Liu, J. Liao and T. Lu, Chem. Commun., 2008, 3540–3542.
95. Y. Huang, X. Zhou, M. Yin, C. Liu and W. Xing, Chem. Mater., 2010, 22, 5122–5128.
96. G. Glaspell, L. Fuoco and M. S. El-Shall, J. Phys. Chem. B, 2005, 109, 17350–17355.
97. W.-J. Shen and Y. Matsumura, J. Mol. Catal. A: Chem., 2000, 153, 165–168.
98. X. Zhou, Y. Huang, C. Liu, J. Liao, T. Lu and W. Xing, ChemSusChem, 2010, 3, 1379–1382.
99. S. W. Ting, S. Cheng, K. Y. Tsang, N. van der Laak and K. Y. Chan, Chem. Commun., 2009, 7333–7335.
100. K. Tedsree, T. Li, S. Jones, C. W. A. Chan, K. M. K. Yu, P. A. J. Bagot, E. A. Marquis, G. D. W. Smith and S. C. E. Tsang, Nat. Nanotechnol., 2011, 6, 302–307.
101. K. Tedsree, C. W. A. Chan, S. Jones, Q. Cuan, W. K. Li, X. Q. Gong and S. C. E. Tsang, Science, 2011, 332, 224–228.
102. Z.-L. Wang, J.-M. Yan, H.-L. Wang, Y. Ping and Q. Jiang, Sci. Rep., 2012, 2, 598.
103. Z.-L. Wang, J.-M. Yan, H.-L. Wang, Y. Ping and Q. Jiang, J. Mater. Chem. A, 2013, 1, 12721–12725.
104. Y. Ping, J.-M. Yan, Z.-L. Wang, H.-L. Wang and Q. Jiang, J. Mater. Chem. A, 2013, 1, 12188–12191.
105. Z.-L. Wang, J.-M. Yan, Y. Ping, H.-L. Wang, W.-T. Zheng and Q. Jiang, Angew. Chem., Int. Ed., 2013, 52, 4406–4409.
106. Z.-L. Wang, Y. Ping, J.-M. Yan, H.-L. Wang and Q. Jiang, Int. J. Hydrogen Energy, 2014, 39, 4850–4856.
107. Z.-L. Wang, H.-L. Wang, J.-M. Yan, Y. Ping, O. Song-Il, S.-J. Li and Q. Jiang, Chem. Commun., 2014, 50, 2732–2734.
108. M. Yadav, A. K. Singh, N. Tsumori and Q. Xu, J. Mater. Chem., 2012, 22, 19146–19150.
109. Y.-Y. Cai, X.-H. Li, Y.-N. Zhang, X. Wei, K.-X. Wang and J.-S. Chen, Angew. Chem., Int. Ed., 2013, 52, 11822–11825.
110. K. Mori, M. Dojo and H. Yamashita, ACS Catal., 2013, 3, 1114–1119.
111. Q.-L. Zhu, N. Tsumori and Q. Xu, Chem. Sci., 2014, 5, 195–199.
112. X. Gu, Z.-H. Lu, H.-L. Jiang, T. Akita and Q. Xu, J. Am. Chem. Soc., 2011, 133, 11822–11825.
113. H. Dai, N. Cao, L. Yang, J. Su, W. Luo and G. Cheng, J. Mater. Chem. A, 2014, 2, 11060–11064.
114. M. Martis, K. Mori, K. Fujiwara, W.-S. Ahn and H. Yamashita, J. Phys. Chem. C, 2013, 117, 22805–22810.
115. Y.-L. Qin, J. Wang, F.-Z. Meng, L.-M. Wang and X.-B. Zhang, Chem. Commun., 2013, 49, 10028–10030.
116. M. Yadav, T. Akita, N. Tsumori and Q. Xu, J. Mater. Chem., 2012, 22, 12582–12586.
117. Q.-Y. Bi, X.-L. Du, Y.-M. Liu, Y. Cao, H.-Y. He and K.-N. Fan, J. Am. Chem. Soc., 2012, 134, 8926–8933.
118. L. Jia, D. A. Bulushev, O. Y. Podyacheva, A. I. Boronin, L. S. Kibis, E. Y. Gerasimov, S. Beloshapkin, I. A. Seryak, Z. R. Ismagilov and J. R. H. Ross, J. Catal., 2013, 307, 94–102.
119. M. Yurderi, A. Bulut, M. Zahmakiran and M. Kaya, Appl. Catal., B, 2014, 160–161, 514–524.
120. K. Jiang, K. Xu, S.-Z. Zou and W.-B. Cai, J. Am. Chem. Soc., 2014, 136, 4861–4864.
121. J. H. Lee, J. Ryu, J. Y. Kim, S.-W. Nam, J. H. Han, T.-H. Lim, S. Gautam, K. H. Chae and C. W. Yoon, J. Mater. Chem. A, 2014, 2, 9490–9495.
122. Q. Luo, G. Feng, M. Beller and H. Jiao, J. Phys. Chem. C, 2012, 116, 4149–4156.
123. Q. Luo, T. Wang, M. Beller and H. Jiao, J. Mol. Catal. A: Chem., 2013, 379, 169–177.
124. Q. Luo, T. Wang, M. Beller and H. Jiao, J. Power Sources, 2014, 246, 548–555.
125. Q. Luo, M. Beller and H. Jiao, J. Theor. Comput. Chem., 2013, 12, 1330001.
126. http://en.wikipedia.org/wiki/Formic_acid.
127. A new study confirms that nature is responsible for 90% of the Earth's atmospheric acidity – http://wattsupwiththat.com.
128. W. Reutemann and H. Kieczka, Formic Acid, in Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH, Weinheim, 2002.
129. M. Aresta, A. Dibenedetto and A. Angelini, Chem. Rev., 2014, 114, 1709–1742.
130. R. Miyatani and Y. Amao, Biotechnol. Lett., 2002, 24, 1931–1934.
131. R. Miyatani and Y. Amao, J. Mol. Catal. B: Enzym., 2004, 27, 121–125.
132. J. Song, H. Fan, J. Ma and B. Han, Green Chem., 2013, 15, 2619–2635.
133. N. Taccardi, D. Assenbaum, M. E. M. Berger, A. Bosmann, F. Enzenberger, R. Wolfel, S. Neuendorf, V. Goeke, N. Schodel, H. J. Maass, H. Kistenmacher and P. Wasserscheid, Green Chem., 2010, 12, 1150–1156.
134. D. Yu, M. Aihara and J. Antal, Energy Fuels, 1993, 7, 574–577.
135. G. D. Mcginnis, S. E. Prince, C. J. Biermann and J. T. Lowrimore, Carbohydr. Res., 1984, 128, 51–60.
136. G. D. Mcginnis, W. W. Wilson and C. E. Mullen, Ind. Eng. Chem. Prod. Res. Dev., 1983, 22, 352–357.
137. F. M. Jin, J. Yun, G. M. Li, A. Kishita, K. Tohji and H. J. Enomoto, Green Chem., 2008, 10, 612–615.
138. R. Wölfel, N. Taccardi, A. Bösmann and P. Wasserscheid, Green Chem., 2011, 13, 2759–2763.
139. J. Albert, R. Wölfel, A. Bösmann and P. Wasserscheid, Energy Environ. Sci., 2012, 5, 7956–7962.
140. J. Albert, D. Lüders, A. Bösmann, D. M. Guldi and P. Wasserscheid, Green Chem., 2014, 16, 226–237.
141. J. Li, D.-J. Ding, L. Deng, Q.-X. Guo and Y. Fu, ChemSusChem, 2012, 5, 1313–1318.
142. J. Zhang, M. Sun, X. Liu and Y. Han, Catal. Today, 2014, 233, 77–82.
143. J. Xu, H. Zhang, Y. Zhao, Z. Yang, B. Yu, H. Xu and Z. Liu, Green Chem., 2014, 16, 4931–4935.
144. W. Wang, M. Niu, Y. Hou, W. Wu, Z. Liu, Q. Liu, S. Ren and K. N. Marsh, Green Chem., 2014, 16, 2614–2618.
145. Z. Tang, W. Deng, Y. Wang, E. Zhu, X. Wan, Q. Zhang and Y. Wang, ChemSusChem, 2014, 7, 1557–1567.
146. P. Gao, G. Li, F. Yang, X.-N. Lv, H. Fan, L. Meng and X.-Q. Yu, Ind. Crops Prod., 2013, 48, 61–67.
147. F. Jin, J. Yun, G. Li, A. Kishita, K. Tohji and H. Enomoto, Green Chem., 2008, 10, 612–615.
148. H. Choudhary, S. Nishimura and K. Ebitani, Appl. Catal., B, 2015, 162, 1–10.
149. M. Cokoja, C. Bruckmeier, B. Rieger, W. A. Herrmann and F. E. Kühn, Angew. Chem., Int. Ed., 2011, 50, 8510–8537.
150. G. A. Olah, Angew. Chem., Int. Ed., 2005, 44, 2636–2639.
151. G. A. Olah, A. Goeppert and G. K. S. Prakash, Beyond Oil and Gas: The Methanol Economy, Wiley-VCH, 2006.
152. J. Graciani, K. Mudiyanselage, F. Xu, A. E. Baber, J. Evans, S. D. Senanayake, D. J. Stacchiola, P. Liu, J. Hrbek, J. F. Sanz and J. A. Rodriguez, Science, 2014, 345, 546–550.
153. M. Nielsen, E. Alberico, W. Baumann, H.-J. Drexler, H. Junge, S. Gladiali and M. Beller, Nature, 2013, 495, 85–89.
154. W. Leitner, E. Dinjus and F. Gassner, in Aqueous-Phase Organometallic Catalysis, Concepts and Applications, ed. B. Cornils and W. A. Herrmann, Wiley-VCH, Weinheim, 1998, pp. 486–498.
155. Y. Himeda, Eur. J. Inorg. Chem., 2007, 3927–3941.
156. P. G. Jessop, Homogeneous hydrogenation of carbon dioxide, in Handbook of Homogeneous Hydrogenation, ed. J. G. De Vries and C. J., Elsevier, Wiley-VCH, Weinheim, 2007, vol. 1, pp. 489–511.
157. C. Federsel, R. Jackstell and M. Beller, Angew. Chem., Int. Ed., 2010, 49, 6254–6257.
158. W. Wang, S. Wang, X. Ma and J. Gong, Chem. Soc. Rev., 2011, 40, 3703–3727.
159. Recent advances in transition metal-catalysed homogeneous hydrogenation of carbon dioxide in aqueous media in Hydrogenation, Karamé, ed. W.-H. Wang and Y. Himeda, InTech, 2012, ISBN: 978-953-51-0785-9.
160. Y. Izumi, Coord. Chem. Rev., 2013, 257, 171–186.
161. W.-H. Wang, Y. Himeda, J. T. Muckerman, G. F. Manbeck and E. Fujita, Chem. Rev., 2015 DOI:10.1021/acs.chemrev.5b00197.
162. Y. Inoue, H. Izumida, Y. Sasaki and H. Hashimoto, Chem. Lett., 1976, 5, 863–864.
163. X. Yang, ACS Catal., 2011, 1, 849–854.
164. R. Tanaka, M. Yamashita and K. Nozaki, J. Am. Chem. Soc., 2009, 131, 14168–14169.
165. G. Zhao and F. Joó, Catal. Commun., 2011, 14, 74–76.
166. Y. Himeda, S. Miyazawa and T. Hirose, ChemSusChem, 2011, 4, 487–493.
167. Y. Maenaka, T. Suenobu and S. Fukuzumi, Energy Environ. Sci., 2012, 5, 7360–7367.
168. S. Wesselbaum, U. Hintermair and W. Leitner, Angew. Chem., Int. Ed., 2012, 51, 8585–8588.
169. C. A. Huff and M. S. Sanford, ACS Catal., 2013, 3, 2412–2416.
170. K. Sordakis, M. Beller and G. Laurenczy, ChemCatChem, 2014, 6, 96–99.
171. S. Moret, P. J. Dyson and G. Laurenczy, Nat. Commun., 2014, 5, 4017.
172. J. L. Drake, C. M. Manna and J. A. Byers, Organometallics, 2013, 32, 6891–6894.
173. S.-F. Hsu, S. Rommel, P. Eversfield, K. Muller, E. Klemm, W. R. Thiel and B. Plietker, Angew. Chem., Int. Ed., 2014, 53, 7074–7078.
174. R. Tanaka, M. Yamashita, L. W. Chung, K. Morokuma and K. Nozaki, Organometallics, 2011, 30, 6742–6750.
175. G. A. Filonenko, R. V. Putten, E. N. Schulpen, E. J. M. Hensen and E. A. Pidko, ChemCatChem, 2014, 6, 1526–1530.
176. G. O. Evans and C. J. Newell, Inorg. Chim. Acta, 1978, 31, L387–L389.
177. C.-C. Tai, T. Chang, B. Roller and P. G. Jessop, Inorg. Chem., 2003, 42, 7340–7341.
178. S. Enthaler, K. Junge and M. Beller, Angew. Chem., Int. Ed., 2008, 47, 3317–3321.
179. C. Ziebart, C. Federsel, P. Anbarasan, R. Jackstell, W. Baumann, A. Spannenberg and M. Beller, J. Am. Chem. Soc., 2012, 134, 20701–20704.
180. Y. Zhang, A. D. MacIntosh, J. L. Wong, E. A. Bielinski, P. G. Williard, B. Q. Mercado and N. Hazari, Chem. Sci., 2015, 6, 4291–4299.
181. C. Federsel, C. Ziebart, R. Jackstell, W. Baumann and M. Beller, Chem. – Eur. J., 2012, 18, 72–75.
182. M. S. Jeletic, M. T. Mock, A. M. Appel and J. C. Linehan, J. Am. Chem. Soc., 2013, 135, 11533–11536.
183. Y. M. Badiei, W.-H. Wang, J. F. Hull, D. J. Szalda, J. T. Muckerman, Y. Himeda and E. Fujita, Inorg. Chem., 2013, 52, 12576–12586.
184. H.-W. Suh, T. J. Schmeier, N. Hazari, R. A. Kemp and M. K. Takase, Organometallics, 2012, 31, 8225–8236.
185. C. M. Zall, J. C. Linehan and A. M. Appel, ACS Catal., 2015, 5, 5301–5305.
186. B. Jezowska-Trzebiatowska and P. Sobota, J. Organomet. Chem., 1974, 80, C27–C28.
187. Y. Musashi and S. Sakaki, J. Am. Chem. Soc., 2000, 122, 3867–3877.
188. A. Urakawa, M. Iannuzzi, J. Hutter and A. Baiker, Chem. – Eur. J., 2007, 13, 6828–6840.
189. C. Bo and A. Dedieu, Inorg. Chem., 1989, 28, 304–309.
190. P. G. Jessop, T. Ikariya and R. Noyori, Chem. Rev., 1995, 95, 259–272.
191. H. H. Karsch, Chem. Ber., 1977, 110, 2213–2221.
192. M. W. Farlow and H. Adkins, J. Am. Chem. Soc., 1935, 57, 2222–2223.
193. H. Wiener, J. Blum, H. Feilchenfeld, Y. Sasson and N. Zalmanov, J. Catal., 1988, 110, 184–190.
194. D. Preti, C. Resta, S. Squarcialupi and G. Fachinetti, Angew. Chem., Int. Ed., 2011, 50, 12551–12554.
195. C. Hao, S. Wang, M. Li, L. Kang and X. Ma, Catal. Today, 2011, 160, 184–190.
196. D. Preti, S. Squarcialupi and G. Fachinetti, ChemCatChem, 2012, 4, 469–471.
197. N. Liu, J. Lei, M. Y. Li and P. Wang, Adv. Mater. Res., 2014, 283, 881–883.
198. G. Peng, S. J. Sibener, G. C. Schatz and M. Mavrikakis, Surf. Sci., 2012, 606, 1050–1055.
199. Z. Xu, N. D. McNamara, G. T. Neumann, W. F. Schneider and J. C. Hicks, ChemCatChem, 2013, 5, 1769–1771.
200. T. Maihom, S. Wannakao, B. Boekfa and J. Limtrakul, J. Phys. Chem. C, 2013, 117, 17650–17658.
201. Y. Zhang, J. Fei, Y. Yu and X. Zheng, Catal. Commun., 2004, 5, 643–646.
202. Z. Zhang, Y. Xie, W. Li, S. Hu, J. Song, T. Zhiang and B. Han, Angew. Chem., Int. Ed., 2008, 47, 1127–1129.
203. Y. Amao, ChemCatChem, 2011, 3, 458–474.
204. K. R. Thampi, J. Kiwi and M. Grazel, Nature, 1987, 327, 506–508.
205. T. Inoue, A. Fujishima, S. Konishi and K. Honda, Nature, 1979, 277, 637–638.
206. B. Aurian-Blajeni, M. Halmann and J. Manassen, J. Sol. Energy, 1980, 25, 165–170.
207. D. Mandler and I. Willner, Org. Biomol. Chem., 1988, 2, 997–1003.
208. I. Willner and D. Mandler, J. Am. Chem. Soc., 1989, 111, 1330–1336.
209. I. Willner, N. Lapidot, A. Riklin, R. Kasher, E. Zahavy and E. Katz, J. Am. Chem. Soc., 1994, 116, 1428–1441.
210. I. Willner, I. Willner and N. Lapidot, J. Am. Chem. Soc., 1990, 112, 6438–6439.
211. K. Schuchmann and V. Müller, Science, 2013, 342, 1382–1385.
212. A. Alissandratos, H.-K. Kim and C. J. Easton, Bioresour. Technol., 2014, 164, 7–11.
213. M. Kodaka and Y. Kubota, Org. Biomol. Chem., 1999, 2, 891–894.
214. R. Miyatani and Y. Amao, Biotechnol. Lett., 2002, 24, 1931–1934.
215. R. Miyatani and Y. Amao, J. Mol. Catal. B: Enzym., 2004, 27, 121–125.
216. R. Miyatani and Y. Amao, J. Jpn. Pet. Inst., 2004, 47, 27–31.
217. I. Tsujisho, M. Toyoda and Y. Amao, Catal. Commun., 2006, 7, 173–176.
218. S. Sato, T. Morikawa, S. Saeki, T. Kajino and T. Motohiro, Angew. Chem., Int. Ed., 2010, 49, 5101–5105.
219. T. Arai, S. Tajima, S. Sato, K. Uemura, T. Morikawa and T. Kajino, Chem. Commun., 2011, 47, 12664–12666.
220. S. Sato, T. Arai, T. Morikawa, K. Uemura, T. M. Suzuki, H. Tanaka and T. Kajino, J. Am. Chem. Soc., 2011, 133, 15240–15243.
221. T. Arai, S. Sato, K. Uemura, T. Morikawa, T. Kajino and T. Motohiro, Chem. Commun., 2010, 46, 6944–6946.
222. T. M. Suzuki, H. Tanaka, T. Morikawa, M. Iwaki, S. Sato, S. Saeki, M. Inoue, T. Kajino and T. Motohiro, Chem. Commun., 2011, 47, 8673–8675.
223. A. Nakada, K. Koike, T. Nakashima, T. Morimoto and O. Ishitani, Inorg. Chem., 2015, 54, 1800–1807.
224. E. E. Benson, C. P. Kubiak, A. J. Sathrum and J. M. Smieja, Chem. Soc. Rev., 2009, 38, 89–99.
225. H. Arakawa, M. Aresta, J. Armor, M. Barteau, E. Beckman, A. Bell, J. Bercaw, C. Creutz, D. A. Dixon, D. Dixon, K. Domen, D. DuBois, J. Eckert, E. Fujita, D. Gibson, W. Goddard, D. Goodman, J. Keller, G. Kubas, H. Kung, J. Lyons, L. Manzer, T. Marks, K. Morokuma, K. Nicholas, R. Periana, L. Que, J. Rostrup-Nielson, W. Sachtler, L. Schmidt, A. Sen, G. Somorjai, P. Stair, B. Stults and W. Tumas, Chem. Rev., 2001, 101, 953–996.
226. W. Leitner, Angew. Chem., Int. Ed. Engl., 1995, 34, 2207–2221.
227. F. A. de Bruijn, D. C. Papageorgopoulos, E. F. Sitters and G. J. M. Janssen, J. Power Sources, 2002, 110, 117–124.
228. R. K. Ahluwalia and X. Wang, J. Power Sources, 2008, 180, 122–131.
229. A. E. Russell, S. C. Ball, S. Maniguet and D. Thompsett, J. Power Sources, 2007, 171, 72–78.
230. S. Giddey, F. Ciacchi and S. Badwal, Ionics, 2005, 11, 1.
231. T. Tingelöf, L. Hedström, N. Holmström, P. Alvfors and G. Lindbergh, Int. J. Hydrogen Energy, 2008, 33, 2064–2072.
232. J. Larminie and A. Dicks, Fuel cell systems explained, Wiley, Chichester, 2003.
233. P. Bernardo, E. Drioli and G. Golemme, Ind. Eng. Chem. Res., 2009, 48, 4638–4663.
234. R. Williams, R. S. Crandall and A. Bloom, Appl. Phys. Lett., 1978, 33, 381–383.
235. M. Halmann, M. Ulman and B. Aurian-Blajeni, J. Sol. Energy, 1983, 31, 429–431.
236. H. Wiener, Y. Sasson and J. Blum, J. Mol. Catal., 1986, 35, 277–284.
237. B. Zaidman, H. Wiener and Y. Sasson, Int. J. Hydrogen Energy, 1986, 11, 341–347.
238. H. Wiener, B. Zaidman and Y. Sasson, J. Sol. Energy, 1989, 43, 291–296.
239. S. Enthaler, ChemSusChem, 2008, 1, 801–804.
240. F. Joo, ChemCatChem, 2014, 6, 3306–3308.
241. S. Fukuzumi, Eur. J. Inorg. Chem., 2008, 1351–1362.
242. Q. L. Zhu and Q. Xu, Energy Environ. Sci., 2015, 8, 478–512.
243. P. Sponholz, D. Mellmann, H. Junge and M. Beller, ChemSusChem, 2013, 6, 1172–1176.
244. A. Boddien, C. Federsel, P. Sponholz, D. Mellmann, R. Jackstell, H. Junge, G. Laurenczy and M. Beller, Energy Environ. Sci., 2012, 5, 8907–8911.
245. S.-F. Hsu, S. Rommel, P. Eversfield, K. Muller, E. Klemm, W. R. Thiel and B. Plietker, Angew. Chem., Int. Ed., 2014, 53, 7074–7078.
246. A. Boddien, F. Gärtner, C. Federsel, P. Sponholz, D. Mellmann, R. Jackstell, H. Junge and M. Beller, Angew. Chem., Int. Ed., 2011, 50, 6411–6414.
247. Q.-Y. Bi, J.-D. Lin, Y.-M. Liu, X.-L. Du, J.-Q. Wang, H.-Y. He and Y. Cao, Angew. Chem., Int. Ed., 2014, 53, 13583–13587.
248. J. Su, L. Yang, M. Lu and H. Lin, ChemSusChem, 2015, 8, 813–816.
249. C. Chauvier, A. Tlili, C. D. N. Gomes, P. Thuéry and T. Cantat, Chem. Sci., 2015, 6, 2938–2942.

---