Introduction

In the coming decades, society will experience a massive increase in the demand for renewable energy, specifically wind and solar, and to reduce carbon emissions caused by the combustion of fossil fuels1. To provide a reliable energy supply and more specifically to meet peak energy demands in densely populated regions as well as to avoid high electricity cost spikes, efficient ways storing fluctuating solar and wind power in both short and long terms are required. Besides classic mechanical approaches to store electric energy, with hydroelectric dams being the most famous ones2, its conversion to chemical energy is discussed to be a feasible approach3. Here, hydrogen which can be easily produced by water electrolysis stands out as a means of an established commercial technology4,5. However, handling large quantities of hydrogen is troublesome, since the compressed gaseous and liquid H2 requires vessels that can withstand high pressures (700 bar) and/or low temperatures (−253 °C) to achieve considerable hydrogen storage capacity. Such methods lead to high energy costs and require specific materials and equipment despite their good H2 recovery6. Alternatively, chemical hydrogen storage-release methods convert H2 to stable carrier molecules that can be stored and transported at ambient conditions and deliver afterward the stored H2 on demand via dehydrogenation7,8. Such technologies could bridge the production of green H2 from renewable electricity and its utilization in proton-exchange membrane (PEM) fuel cells to regenerate the stored renewable electricity for terminal energy consumption (Fig. 1a).

Fig. 1: Projected sustainable energy utilization based on renewable electricity storage and regeneration bridged by chemical hydrogen storage-release.
figure 1

a Renewable electricity can be converted to chemical fuel H2 via water electrolysis. The resulting H2 is easily transformed into stable chemical H2 carriers for short- and long-term storage and transportation. The stored H2 can be released on request to regenerate the renewable electricity in proton-exchange membrane (PEM) fuel cells. b Schematic illustration of amine-based carbon capture and storage (CCS), carbon capture and utilization (CCU), previously reported H2 storage-release, and the strategy of carbon neutral H2 storage-release based on dual-functional roles of formamides described in this work.

Full size image

Besides the gaseous H2 carriers e.g., ammonia9 and methane10, liquid organic hydrogen carriers (LOHC) offer high reversibility and superior kinetics in (de)hydrogenation, suitable for long distance transport and onboard applications11,12. As well-known examples of C1 compounds13, methane10, methanol14, and formic acid (FA)14,15 have been widely studied concerning hydrogen storage. Compared to ammonia (ΔG0 = +33.3 kJ mol−1), methane (ΔG0 = + 113.6), and methanol (ΔG0 = +8.9 kJ mol−1), formic acid (ΔG0 = −32.9 kJ mol−1) is more thermodynamically favored in H2 production process. Therefore, chemical H2 storage-release cycles applying the H2/CO2-FA system have been well-developed in the past decades by using the greenhouse gas CO216,17. In addition, an intrinsically similar approach including bicarbonate and formate salts has also been investigated in reversible (de)hydrogenation processes (ΔG0 = ±0.7 kJ mol−1)18,19,20. Surprisingly, formamides as another class of easily and commercially available C1 compounds derived from CO2 reduction in the presence of amines have been rarely studied directly in H2 storage-release cycles21,22,23.

It’s worth noting that as CO2 capturing reagents amines are frequently used in carbon capture and storage (CCS) processes24, and further utilization of the CO2-amine adducts (captured CO2) in subsequent hydrogenation allows to produce renewable fuels and chemicals, so called carbon capture and utilization (CCU)25. As one of the most prominent examples of CCU, the “George Olah Methanol Plant” in Iceland is based on local renewable energy and CO226. Its total electrical energy demand and the overall efficiency reach 9.5 MWh/t methanol and 60%. Notably, such CCU processes also provide feasible approaches for sustainable chemical H2 storage-release applications based on interconversion of CO2 and C1 compounds (e.g., FA; Fig. 1b)8. For example, recently our group developed a reversible H2 storage-release method based on amino acid lysine promoted CO2 capture and its reversible hydrogenation to FA17. On the other hand, FA in the presence of amines could be easily dehydrated to formamides27 which combine the carbon capturing reagent amines with the H2 storage material FA. Therefore, the direct use of formamides as H2 carriers is practically desired due to their dual-functional roles: the structurally incorporated FA is responsible for H2 storage-release, and the built-in amines provide a carbon capture and utilization (CCU) strategy leading to an ultimate carbon neutral H2 storage-release system (Fig. 1b). Compared to the hydrogen contents of FA (4.34 wt%), the ones of various formamides (1.50–3.17 wt%) are slightly lower but still higher than that of common formate salts (1.02–2.85 wt%, Fig. 2a). Bearing in mind that equivalent CO2 is emitted together with H2 in FA dehydrogenation process which generally requires a post carbon capturing process to reduce carbon emissions14, in addition, H2 storage using formate produces bicarbonate salts which could be decomposed to CO2 as frequently reported20. Besides, another H2 storage technology using H2 storage alloys, e.g. magnesium hydrides28, generally represent hydrogen contents of 1–6 wt%. However, their inferior (de)hydrogenation kinetics, life cycle, and harsh operation conditions (up to 500 °C) make them currently not suitable for most of the applications28,29.

Fig. 2: Concept of reversible carbon neutral hydrogen storage-release cycles based on dual-functional roles of formamides.
figure 2

a Hydrogen content of formic acid and its derivatives (wt%, indicated in green color). b Schematic illustration of the concept of pincer-type iron complex catalyzed reversible carbon neutral H2 storage and release based on dual-functional roles of formamides.

Full size image

So far, expensive noble metal-based catalysts still dominate the area of H2 storage and release. Therefore, the search of suitable non-noble metal catalysts and their efficient utilization in reversible H2 storage-release cycles are particularly important. As a class of versatile catalysts, iron-based pincer complexes14,30,31,32,33,34,35,36,37,38,39,40,41,42 have been studied respectively in hydrogenation43,44,45,46,47,48 and dehydrogenation49,50,51, attracting many interests for potentially reversible H2 storage-release applications37,38,52,53,54. Owing to the metal-ligand cooperation effect, tridentate pincer complexes with a nitrogen donor (N-H) offer effective and stable catalysis in both hydrogenation and dehydrogenation steps55,56,57. As representatives, iron pincer complexes are used in CO2 hydrogenation to produce FA (or its formate salts)58,59,60,61,62, formamides27, and methanol63, as well as the H2 production from FA64,65,66,67,68,69,70,71,72 and methanol73. To the best of our knowledge, no single iron catalyst has been reported for combined H2 storage and H2 release cycles yet. On the basis of our interest in developing efficient methodologies for H2 storage and utilization by using non-noble metal catalysts, we describe herein a concept of iron promoted partially reversible carbon neutral H2 storage-release cycles in a single device based on dual-functional roles of formamides.

Results and discussion

Concept of reversible carbon neutral hydrogen storage-release cycles based on dual-functional roles of formamides

The concept of iron catalyzed reversible carbon neutral hydrogen storage-release cycles based on dual-functional roles of formamides is illustrated in Fig. 2b. Following the hydrogen release pathway (indicated in pink color), formamide (F1) is firstly hydrolyzed into formic acid (FA) and corresponding amine (A1), afterward FA participates in the catalytic cycles of dehydrogenation and hydrogenation. Here the mild potentials of (de)hydrogeantion are provided by redox active iron complexes containing non-innocent pincer ligands62,64. It’s worth noting that CO2 by-product is captured in situ and stored in the presence of amine (A1) initially liberated from formamide hydrolysis. Even though the individual steps of formamides hydrolysis, FA (or formates) dehydrogenation and their reverse reactions are known, the presented hydrogen storage-release concept enables the reuse of in situ captured CO2, which allows to (1) retain the hydrogen storage material CO2 in the reaction, therefore, maintain the theoretical hydrogen storage capacity in successive H2 storage-release cycles, (2) avoid undesired carbon release during dehydrogenation processes, and (3) provide superior H2 selectivity/purity compared to other H2 carrier systems. Following the hydrogen storage pathway (indicated in blue color), the stored CO2 can be re-hydrogenated to FA which is then (partially) converted to formamide (F1) via dehydration condensation with corresponding amine (A1). Thanks to the dual-functional roles of formamides, the built-in amine (A1) is beneficial to both H2 storage and H2 release processes by acting as CO2 absorbent, providing a carbon capture and utilization (CCU) strategy to ensure the H2 storage capacity and carbon neutrality of the overall H2 storage-release process.

Following the concept vide supra, both formamide hydrolysis as well as formamide formation were investigated. Thus, initially the hydrolysis process was performed under alkaline conditions74,75 and a proportional relationship between the base (KOH) loading and FA yields was found (Figs. S1–2). Accordingly, equimolar ratio of base to formamide is necessary to provide a sufficient amount of H2 carrier for the subsequent H2 storage-release cycles. Afterwards, the reaction between different amines and FA to produce formamides was examined (Figs. S3–4)27. Interestingly, in this latter condensation process piperazine (A3) gave a much better yield of the corresponding formamides (22%) compared to morpholine (A1, 1%) and piperidine (A2, 1%) under typical reaction conditions used for catalysis (90 °C, 12 h). Obviously, using longer reaction time (72 h) and higher temperature (140 °C) allows to increase the amount of formamide products (A1 47%, A2 14%, A3 46%). Overall, the hydrolysis of formamides to FA and amines is more favored under alkaline condition, than its reverse dehydration condensation.

Next, the CO2 capture effect of those amines was also investigated (Figs. S5-7). Under CO2 pressure (2 bar, 30 min.), both bicarbonate and carbamate species of the corresponding amines were obtained as products in the following order: piperidine (A2, 69%), piperazine (A3, 55%), and morpholine (A1, 42%). These results can be well explained by the reported pKa values of the three amines: A2 (11.22) > A3 (9.73) > A1 (8.36)76,77. Under direct air capture conditions (air flow 1.8 L min−1, ca. 400 ppm CO2, 36 h), piperazine (A3) led to the highest yield of the corresponding carbamate species (32%) compared to piperidine (A2, 15%) and morpholine (A1, 8%). This is attributed to the stronger hydrogen bonding in piperazine (A3) compared to the other two amines78. After all, these results demonstrate the good carbon capture ability of amines A1, A2, and A3, especially with CO2 concentration at ppm level.

Catalytic hydrogen production based on formamides

Representative non-noble metal pincer complexes (Fig. 3a) were utilized as catalysts in hydrogen production process starting from formamides and the results are summarized in Fig. 3b. Iron pincer-complexes Fe-1 and Fe-2, which were used in formic acid dehydrogenation64, led to the best yields of H2 99% and 89% (Figs. S8–9), respectively. Other tested catalysts based on Mn, Co and Mo gave significantly lower H2 yields (up to 43%). In the absence of external base, no H2 was produced. Drastically decreased H2 yields (29% and 37%) were observed after changing the base from KOH to amino acids lysine (Lys) and arginine (Arg), which were recently disclosed for reversible H2 storage-release involving CO2 hydrogenation17. Indeed, utilizing stoichiometric amounts of Lys or Arg gave only trace amount of FA due to slower formamide (F1) hydrolysis (1–2% yields, Fig. S2). For hydrogen production also the nature of formamides was examined. Notably, inexpensive and industrially available simple formamides i.e., methanamide (MA) and dimethylformamide (DMF) gave also good H2 yields (78% and 80%, respectively). However, due to practical considerations, e.g. ammonia and dimethylamine are highly volatile and difficult to handle, we utilized their heavier congeners. As the best candidates, N-formylmorpholine (F1) and 1,4-diformylpiperazine (F4) led to quantitative yields of H2, while N-formylpiperidine (F2) and 1-formylpiperazine (F3) gave 69% and 87% H2 yields, respectively.

Fig. 3: Catalytic hydrogen production from formamides.
figure 3

a Non-noble metal-based pincer complexes utilized in this study. b Comparison of activity under various conditions towards catalytic hydrogen production. Standard conditions: N-formylmorpholine (F1, 10 mmol), KOH (10 mmol), catalyst (5 μmol, 500 pm), THF/H2O (5/5 mL), 90 °C, 16 h. *1,4-Diformylpiperazine (F4, 5.0 mmol) was used. Yields are based on formyl group in formamides. The dotted lines serve as guides to the eye.

Full size image

The base (KOH) loading in catalytic dehydrogenation process was then investigated: in the presence of 25, 50, 75 mol% of KOH, partial H2 yields (37-61%) and lower H2 selectivity were observed (Fig. S10). In the absence of KOH, no conversion of formamide (F1) occurred as indicated by NMR monitoring on the reaction mixture (Fig. S11). Lewis acids are known to assist dehydrogenation processes catalyzed by iron pincer catalysts64. However, inferior H2 yield (85%) and selectivity (92.5%) were observed in the presence of 10 mol% LiBF4 compared to the standard conditions (Figs. S10 and S34). Changing THF to other organic solvents, i.e., 2-methyl THF (2-MTHF), ethanol, triglyme, 1,4-dioxane, and DMSO, H2 were observed in 47-74% yields. Using water as sole solvent or under neat conditions, no hydrogen was found due to the low solubility of the catalyst. Decreased H2 yield (87%) was observed by lowering the reaction temperature to 80 °C, while elevated temperature (100 °C) did not promote the reaction but resulted in increased CO concentration (14 ppm; Fig. S41). In all other cases using Fe-1 complex and formamides, CO was not detected by gas chromatography (below the CO quantification limit of 10 ppm).

Comparison on different hydrogen carrier systems in catalytic hydrogen production

Under the optimal conditions, the here presented system utilizing formamides is superior regarding both the H2 productivity and selectivity compared to other H2 carriers i.e., formic acid (FA) and potassium formate (PF; Fig. 3b). Specifically, replacing formamide F1 with FA and amine A1, decreased H2 yields (78%) were observed with H2 selectivity of 100% (in the presence of KOH) and 74.9% (in the absence of KOH). Notably, 80 ppm CO were detected in the H2 storage system of FA and A1 (Fig. S27). Loading FA with KOH, further decreased H2 yield (67%) was observed, while in the presence of FA only, no dehydrogenation occurred. On the other hand, starting from potassium formate (PF), H2 was obtained in yields of 86% (in the presence of A1) and 66% (in the absence of A1).

Catalytic hydrogen storage in formates and formamides

Next, the process of H2 storage in formates and formamides was investigated by using hydrogenation of CO2 or potassium bicarbonate in the presence of amines as model reactions (Fig. 4a)27,62,79. In general, the hydrogenation of CO2 or potassium bicarbonate in the presence of morpholine (A1), piperidine (A2), and piperazine (A3) gave good total yields of formates and formamides (82–100%). Specifically, morpholine (A1) and piperidine (A2) led to comparable results regarding the yields of formates (90-97%) and formamides (2–6%), while piperazine (A3) resulted in a significantly higher amount of formamide product (31%) using CO2 as carbon source (Fig. 4b, left side, Figs. S50-52). It’s worth noting that the amine promoted CO2 capture product carbamate was formed as minor species in bicarbonate hydrogenation reaction (Fig. S51), thereby avoiding the release of free carbon dioxide even under basic conditions.

Fig. 4: Catalytic hydrogen storage in formates and formamides.
figure 4

a Catalytic hydrogen storage process via hydrogenation of CO2 or potassium bicarbonate in the presence of amines. b Left side: comparison of activity towards hydrogen storage with different amines (A1, A2, and A3) and carbon sources (CO2 and KHCO3). Standard conditions: amine (10 mmol), CO2 (20 bar) or KHCO3 (10 mmol), Fe-1 (5 μmol, 500 ppm), H2 (60 bar), THF/H2O (5/5 mL), 90 °C, 12 h. Right side: variation of reaction parameters in hydrogenation of KHCO3 with morpholine. Standard conditions: morpholine (A1, 10 mmol), KHCO3 (10 mmol), Fe-1 (5 μmol, 500 ppm), H2 (60 bar), THF/H2O (5/5 mL), 90 °C, 12 h. Yields are based on amine. The dotted lines serve as guides to the eye.

Full size image

Afterwards, variation of reaction parameters in the hydrogenation step using bicarbonate was performed in the presence of morpholine (A1, Fig. 4b, right side, Fig. S53). Reducing the H2 pressure from 60 bar stepwise to 40, 20, and 10 bar, total yields of formates and formamides decreased from 99 to 43%. Moreover, lowering the reaction temperature from 90 °C to 60, 50, and 40 °C, no obvious loss of hydrogen storage capacity was observed, while further decrease to 30 °C, drastically dropped the formate yield to 39%. Further, time dependent product generation of hydrogen storage and release reactions catalyzed by Fe-1 was investigated (Table S1). Lower total yields of formates and formamides were obtained in 3 and 6 h (66% and 87%, respectively) in hydrogenation reactions with morpholine (A1) and CO2. On the other hand, performing the dehydrogenation reactions with N-formylmorpholine (F1) in shorter reaction times led to decreased H2 yields (29% in 4 h and 49% in 8 h). These results demonstrate that long reaction times are indeed required.

Promoting effect of amines in hydrogen storage and release processes

Next, we explored the promoting effect of seven additional amines in formate dehydrogenation and bicarbonate hydrogenation in more detail (Fig. 5). In addition to A1, A2, and A3, classical amines which are widely utilized in CO2 hydrogenation and corresponding dehydrogenation processes were tested (Fig. 5a). In hydrogen production reactions (Fig. 5b), the presence of amines A1, A2, and A3 gave high H2 yields (up to 92%) and selectivity (up to 100%) compared to the one without amine (56% yield and 95% selectivity). Trials with other amines i.e., diazabicycloundecene (DBU)52,80,81, diazabicyclooctane (DABCO), trihexylamine (THA)81, and dimethyloctylamine (DMOA)82,83 resulted in moderate H2 yields (55% to 76%). However, no H2 was produced by using tetramethylguanidine (TMG). Interestingly, the two basic amino acids Lys and Arg led to H2 in 87% and 90% yields, respectively17,20,84.

Fig. 5: Comparison between selected amine promotors in hydrogen storage and release reactions.
figure 5

a Chemical structures of selected amine promotors utilized in formate dehydrogenation and bicarbonate hydrogenation. b Hydrogen production from formate in the presence of various amines. Standard conditions: KHCO2 (10 mmol), amine (10 mmol), Fe-1 (5 μmol), THF/H2O (5/5 mL), 90 °C, 16 h. Yields are based on KHCO2. c Hydrogenation of bicarbonate in the presence of various amines. Standard conditions: KHCO3 (10 mmol), amine (10 mmol), Fe-1 (5 μmol, 500 ppm), THF/H2O (5/5 mL), H2 (60 bar), 90 °C, 12 h. Yields are based on KHCO3.

Full size image

In the corresponding hydrogen storage process (Fig. 5c, Fig. S54), amines A1, A2, and A3 gave quantitative yields of formates and formamides27, while 64% of formate were obtained in the absence of amine. Moreover, DBU, DABCO85, THA, DMOA, Lys, and Arg led to either lower formate yields (23% to 87%) or even inhibited amide formation. On the other hand, TMG gave nearly quantitative yields of formate and formamide even though it was not active in the H2 production process at all85. As there is no obvious direct correlation of pKa of the applied amine and the storage capacity there will be other factors that potentially influence the system, i.e., solubility and boiling point of amines, hydrogen bonding, steric hindrance, catalyst poisoning etc. After considering the H2 productivity and selectivity in dehydrogenation (Fig. 5b) and total yields of formates and formamides in hydrogenation (Fig. 5c), we concluded that morpholine (A1) and piperazine (A3) are the most suitable amine promoters among all other tested amines. Although formate generation dominates at milder conditions (90 °C, 12 h), formamide yields could be improved at higher temperature and longer reaction time (140 °C, 72 h; Fig. S3), therewith formally clothing the formamide-based hydrogen storage cycle. However, due to practicability milder conditions were employed in subsequent catalytic (de)hydrogenation reactions, as this also allows for efficient and partially reversible H2 storage (Fig. 5b, c).

Carbon neutral hydrogen storage-release cycles based on dual-functional roles of formamides

After having optimized conditions in hand for both elementary steps, (a) H2 release from formamides and (b) corresponding H2 storage process, we turned our attention to the combination of these hydrogenation and dehydrogenation processes in a single device. The overall “carbon neutral” hydrogen cycle was performed in a closed autoclave starting by dehydrogenation of commercially available formamides using the well-designed catalyst Fe-1 (500 ppm) in the presence of KOH in aqueous THF solution (90 °C, 16 h). Afterwards, the reactor was cooled to room temperature (r.t., 25 °C) and the generated hydrogen was released carefully to the manual burettes and analyzed by GC. Then, the reactor was charged with H2 (60 bar) and heated to 90 °C without changing the reaction mixture (H2 storage step). After the hydrogen uptake stopped (12 h), the overpressure of H2 was released at r.t. and the autoclave was subjected once more to the H2 release step (90 °C, 16 h). Following this procedure, 10 H2 storage-release cycles were performed over 20 days (Fig. 6, Figs. S55–63). Notably, during the whole time, only H2 is charged and discharged and the reloading of hydrogen storage material, catalyst, solvents, additives is not necessary. Even though the iron pincer complexes are generally sensitive to air (oxygen), once the H2 storage-release cycles are in operation, the whole system is closed and generally under over-pressure of H2. On the other hand, air has also to be excluded from the system in order to suppress the hydrogen-air explosions (4.0–75.6%v/v of H2 in air).

Fig. 6: Fe promoted partially reversible carbon neutral hydrogen storage-release cycles using formamides.
figure 6

Hydrogen evolution in the storage-release cycles applying formamides. Standard conditions: formamide, KOH (1.0 equiv.), Fe-1 (500 ppm), THF/H2O (5/5 mL), 90 °C. The cycles started from dehydrogenation (16 h), then hydrogenation (12 h, 60 bar of H2) was performed. Yields are based on formyl group in formamides.

Full size image

31P NMR spectra of pre- and post-reaction samples (after 1 cycle) revealed that the original signal of Fe-1 complex (99.6 ppm) was shifted to lower field (114.0 ppm) after the catalytic dehydrogenation reaction (Fig. S64). This signal is assigned to iron pincer derivative I-2 (Fig. 2) and considered as the resting state in (de)hydrogenation reactions64. Besides, only minor species were found in the spectra which might either be the stereoisomers (e.g., trans- and cis-configurations) of the iron pincer complexes or their decomposition products86.

Comparing the different tested formamides, 1,4-diformylpiperazine (F4) resulted in higher H2 selectivity (>99.5%) than N-formylmorpholine (F1, >99.0%) at 10 mmol loading due to the better carbon capture ability of the corresponding amine piperazine (A3) compared to morpholine (A1) especially at low CO2 concentration (Fig. S5). Slightly lower H2 yields were observed with F4 (>70%) compared to F1 (>82%) over 10 H2 charge-discharge cycles, due to the lower hydrogen storage capacity using corresponding amines A3 than A1 (Fig. 4b). To our delight, upscaling reactions applying N-formylmorpholine (F1, 50 mmol) reached 86% H2 yield in the first cycle, even though gradually decreased yields were observed at 70% in the 10th cycle. Overall, H2 can be obtained in more than 70% yield and 99% selectivity in 10 charge-discharge cycles (Table S2). For a direct application of the generated hydrogen in PEM fuel cells and to avoid the poisoning of platinum electrodes87, it is important to note that CO was not detected (below the GC quantification limit of 10 ppm) in the H2 stream. Advantageously, both the hydrogenation and dehydrogenation steps operated at a temperature level of 90 °C, which can be supplied by the waste heat from e.g., PEM fuel cells or hydrogen internal combustion engines88.

In conclusion, we demonstrate partially reversible hydrogen storage-release cycles utilizing formamides. This class of hydrogen storage materials has been largely overlooked despite their attractive physical and chemical properties (inertness, hydrogen content, toxicity, boiling point, etc.). In the presented system, the inherent components of formamides play a dual-functional roles: (a) the formic acid part enables H2 storage and release and (b) the built-in amines provide a carbon capture and utilization (CCU) strategy allowing for an overall “carbon neutral” energy storage system. By using well-designed iron catalyzed hydrogenation and dehydrogenation steps, selective hydrogen formation (CO below detection limit of GC) under mild conditions and high catalyst productivity as well as stability (>20 days) were achieved. To the best of our knowledge, this is also one of the rare examples that an iron based catalytic system allows multiple H2 storage-release cycles in a single device.

Starting from carbon dioxide or bicarbonate in the presence of selected amines, H2 storage proceeded with quantitative total yields of formamides and formates at comparably low temperature (<100 °C). Among the different tested amines, morpholine (A1) and piperazine (A3) exhibited superior behavior in both H2 storage and H2 release processes. The feasibility of combined hydrogenation and dehydrogenation processes in a single device was demonstrated in 10 H2 charge-discharge cycles catalyzed by an iron complex under mild reaction conditions. Advantageously, the presented system is partially reversible and no reloading of hydrogen storage material, catalyst, solvents, additives is necessary during the whole process.

Methods

Calculation of the hydrogen contents (wt%)

The hydrogen contents (wt%) of formic acid, formate salts, and formamides are calculated as follows:

$${{{{{{\rm{wt}}}}}}\%}_{{{{{{\rm{formic\; acid}}}}}}}={{{{{{\rm{M}}}}}}}_{{{{{{\rm{H}}}}}}2}\,/({{{{{{\rm{M}}}}}}}_{{{{{{\rm{formic\; acid}}}}}}})\,\times 100\%$$
(1)
$${{{{{{\rm{wt}}}}}}\%}_{{{{{{\rm{formate\; salt}}}}}}}={{{{{{\rm{M}}}}}}}_{{{{{{\rm{H}}}}}}2}\,/({{{{{{\rm{M}}}}}}}_{{{{{{\rm{formate\; salt}}}}}}}+{{{{{{\rm{M}}}}}}}_{{{{{{\rm{H}}}}}}2{{{{{\rm{O}}}}}}})\times 100\%$$
(2)
$${{{{{{\rm{wt}}}}}}\%}_{{{{{{\rm{formamide}}}}}}}=({{{{{{\rm{M}}}}}}}_{{{{{{\rm{H}}}}}}2}\times {{{{{\rm{N}}}}}})/({{{{{{\rm{M}}}}}}}_{{{{{{\rm{formamide}}}}}}}+{{{{{{\rm{M}}}}}}}_{{{{{{\rm{H}}}}}}2{{{{{\rm{O}}}}}}}\times {{{{{\rm{N}}}}}})\times 100\%$$
(3)

where M is the molecular weight, N is the number of formyl groups per formamide molecule.

Standard procedure for catalytic dehydrogenation starting from formamides

Under an argon atmosphere, N-formylmorpholine (F1, 1 mL, 10 mmol), base (10 mmol), catalyst (5 μmol), THF (5 mL) and H2O (5 mL) were added to a 100 mL autoclave equipped with a magnetic stir bar. Then, the reaction mixture was heated and stirred in a pre-heated oil bath for 16 h. The reactor was cooled to r.t. (25 °C) and the inside pressure was released carefully to the manual burettes. A 5 mL degassed syringe was used to obtain a gas sample analyzed by gas chromatography (GC, CO quantification limit of 10 ppm). Yield of H2 is calculated as follows:

$${{{{{{\rm{Yield}}}}}}}_{{{{{{\rm{H}}}}}}2}=({{{{{\rm{mmol}}}}}}\,{{{{{{\rm{H}}}}}}}_{2})/({{{{{\rm{mmol}}}}}}\,{{{{{\rm{formyl}}}}}}\,{{{{{\rm{group}}}}}}\,{{{{{\rm{in}}}}}}\,{{{{{\rm{formamides}}}}}})\times 100\%$$
(4)

Standard procedure for catalytic hydrogenation of CO2 or bicarbonate

Under an argon atmosphere, amine (10 mmol), CO2 (20 bar) or KHCO3 (1 g, 10 mmol), Fe-1 (2 mg, 5 μmol), THF (5 mL) and H2O (5 mL) were added to a 100 mL autoclave equipped with a magnetic stir bar. After pressurizing the reactor with H2 (60 bar), the reaction mixture was heated and stirred on a pre-heated oil bath for 12 h. Then, the reactor was cooled to r.t. (25 °C) and the overpressure was carefully released. A biphasic reaction mixture was obtained containing a transparent organic upper layer and an aqueous yellow lower layer. Addition of deionized water (ca. 3 mL) to the above mentioned biphasic mixture resulted in a homogeneous solution17. Imidazole (170 mg, 2.5 mmol) was added as an NMR internal standard (I.S.) to the reaction mixture, which was then analyzed by 1H NMR with ca. 0.1 mL D2O to lock the signals. Yields of formate and formamide are calculated as follows:

$${{{{{{\rm{Yield}}}}}}}_{{{{{{\rm{formate}}}}}}}=({{{{{\rm{mmol\; formate}}}}}})/({{{{{\rm{mmol\; amine}}}}}})\times 100\%$$
(5)
$${{{{{{\rm{Yield}}}}}}}_{{{{{{\rm{formamide}}}}}}}({{{{{\rm{mmol\; formamide}}}}}})/({{{{{\rm{mmol\; amine}}}}}})\times 100\%$$
(6)

Standard procedure for catalytic H2 evolution in the H2 storage-release cycles

The H2 storage-release cycles start from the dehydrogenation (H2 release): Fe-1 (2 mg, 5 μmol, 500 ppm), N-formylmorpholine (F1, 1 mL, 10 mmol), KOH (561 mg, 10 mmol), THF (5 mL) and H2O (5 mL) were added to a 100 mL autoclave equipped with a magnetic stir bar. The reaction mixture was then heated and stirred in a pre-heated oil bath at 90 °C for 16 h. The reactor was cooled to r.t. (25 °C) and the stored hydrogen was released carefully to the manual burettes then the content of the gas phase was analyzed with a 5 mL degassed syringe by gas chromatography (GC, CO quantification limit of 10 ppm). The autoclave was then filled with 60 bar of H2, heated and stirred on a pre-heated oil bath at 90 °C for 12 h (H2 storage). After the completion of H2 storage, the reactor was cooled to r.t. (25 °C) and the overpressure was carefully released. Then the autoclave was subjected to the H2 release step once again. Following such process, the H2 evolution in the H2 storage-release cycles were implemented over 20 days. Yield of H2 is calculated according to Eq. (4).