Hydrogen generation from reactions of hydrides with hydrated solids in the solid state

Abstract

The reaction of water with hydrides (hydrolysis reaction) is very attractive for onsite hydrogen generation. Instead of using liquid water like in most cases, we demonstrate that hydrogen generation from hydrolysis reactions can also occur in the solid state. By simply heating mixtures of hydrated solids and hydrides, hydrogen generation is readily achieved through the recombination from the protonic hydrogen in the hydrated solids and the hydridic hydrogen in the hydride. The composites 5CaH₂ + Na₄P₂O₇·10H₂O and NaBH₄ + H₂C₂O₄·2H₂O give attainable gravimetric hydrogen storage capacity of 2.76% at 40 °C and 2.79% at 70 °C with rapid response, respectively. The dehydrogenation temperature can be controlled by the dehydration temperature of the hydrated solids. This innovative hydrogen generation approach provides a temperature activated, easy to control solution for onsite hydrogen generation.

---

1. Introduction

Hydrogen storage is one of the key technologies for hydrogen energy utilization. Currently, it remains highly challenging to develop reversible hydrogen storage systems which can meet all the requirements for broad application.¹ An alternative technology is to generate hydrogen from disposable hydrogen carriers on demand and recovering the hydrogen carriers from the spent fuel.^2,3 This approach is much easier to realize and may have impact in the near term. It is particularly attractive for portable, onsite hydrogen generation when hydrogen refilling infrastructures are not immediately available.

The hydrolysis reaction of hydrides or reactive metals is able to produce hydrogen in mild conditions with highly competitive gravimetric hydrogen storage capacity (GHSC), which is very attractive for onsite hydrogen generation.^4–8 However, in practical applications, hydrogen generation from hydrolysis is more complicated than just mixing the metals/hydrides with water. Almost all the hydrolysis reactions of reactive metals/hydrides are thermodynamically highly favourable, accompanied by substantial exothermic effects. The reaction heat will further accelerate the reaction. Thus, once the reactive metal/hydride and water is thoroughly mixed, the reaction can be easily out of control. In a practical hydrogen generation system, the hydride/metal and water has to be stored separately and allowed to contact in a controlled way. This will require additional storage space and controlling units, which will cause significant loss of the hydrogen storage capacity on a system basis.

Currently, most study on hydrolysis hydrogen generation is focused on optimizing the materials to improve the hydrogen yield and hydrogen generation kinetics. Much less attention is paid to improve the reaction process to facilitate compact system integration and more convenient operation, which is equally important for practical application. It is illustrative to recall the hydrogen generation from thermal decomposition of metal hydrides, in which dehydrogenation will only occur above a certain temperature threshold, i.e. is temperature activated. As a result, hydrogen generation is very easy to control. Moreover, all the materials involved in this case are solids, which is much more convenient to handle.

For hydrogen generation from hydrides, an interesting question is whether it is possible to combine the merits of high GHSC from the hydrolysis approach and the easy operation from the thermal decomposition approach? In this work, we demonstrate that hydrogen generation from the hydrolysis reactions of hydrides can also be carried out in the solid state and become temperature activated. This can be achieved by replacing the liquid water by hydrated solids. Upon heating, the water released from the hydrated solids can react with the hydride to generate hydrogen. The mechanism is just like the hydrolysis reaction, while no liquid state water is involved. Two all solid state hydrogen generation systems provide competitive GHSC above 2.7%, which are highly attractive for onsite hydrogen generation.

2. Experimental

2.1 Sample preparation

The hydrides and the hydrated solids are milled in an agate mortar to fine powders. The powder of the hydrides and the hydrated solids are mixed by gentle hand milling for additional 5 min. The amount of hydrides (CaH₂ or NaBH₄) used is 200 mg. The amount of hydrated solids used is calculated according to the reaction stoichiometry, i.e. to match the amount of the protonic hydrogen in the hydrated solids to the hydridic hydrogen in the hydrides. Each water molecule provides one protonic hydrogen and each H₂C₂O₄ molecule provides two protonic hydrogen atoms. The amount used are: Na₄P₂O₇·10H₂O–423.8 mg, KAl(SO₄)₂·12H₂O–375.6 mg, MgCl₂·6H₂O–321.9 mg, NiSO₄·6H₂O–416.2 mg, ZnSO₄·7H₂O–390.4 mg and H₂C₂O₄·2H₂O–667.2 mg. For the NaBH₄ + H₂C₂O₄·2H₂O combination, the amount of H₂C₂O₄·2H₂O is varied to give optimal gravimetric hydrogen capacity.

2.2 Materials characterization

The materials were characterized by powder X-ray diffraction (XRD, PANalytical X'Pert³ Powder, Cu Kα) and scanning electron microscopy (SEM, Hitachi S4800). The thermal gravimetric analysis was measured by TGA-DSC-DTA (TA, Q600 SDT), the heating rate is 2 K min⁻¹ in nitrogen flow. The DSC was measured by NETZSCH DSC 204 HP. The heating rate was 2 K min⁻¹ in argon flow.

2.3 Measurements of hydrogen generation kinetics

In the isothermal hydrogen generation measurement, the mixture is loaded in a glass reactor with a gas outlet valve for hydrogen generation measurement. An oil bath is heated to the target temperature. The glass reactor is rapidly inserted in to the oil bath. The generated hydrogen is collected and quantified using an inverted water filled measuring cylinder. The volume of H₂ is converted to the moles of H₂ using the ideal gas equation.

In the temperature programmed hydrogen generation measurement, the mixture is loaded in a bubbler like glass reactor. The glass reactor is heated by an oil bath at a constant heating rate of 2 K min⁻¹ in a constant Ar flow of 50 standard cubic centimetres per minute (sccm). The released gas is detected by a residual gas analyser (Omnistar, Pfeiffer Vacuum).

3. Results and discussion

3.1 Hydrogen generation from CaH₂ and hydrated inorganic salts

The general strategy of the hydrogen generation is illustrated in Fig. 1a. The composite contains a metal hydride and a hydrated solid which are homogeneously. Upon heating above a threshold temperature, the metal hydride can react with the water released from the hydrated solid to generate hydrogen.

Fig. 1 (a) Schematic illustration of the solid state hydrolysis reaction for hydrogen generation. (b–d) Elemental mapping in SEM of the 5CaH₂ + Na₄P₂O₇·10H₂O composites.

We first demonstrate this strategy using CaH₂, a highly reactive hydride which reacts vigorously with water. When CaH₂ is mixed with Na₄P₂O₇·10H₂O, a hydrated inorganic salt with high water percentage (40.4%), the X-ray diffraction pattern (Fig. 2a) suggests that both CaH₂ (ICSD 98-026-0873) and Na₄P₂O₇·10H₂O (ICSD 98-001-5389) remain unchanged after mixing. Elemental mapping (Fig. 1b–d) suggests that the two components are macroscopically homogeneously mixed in the composite. The CaH₂ + Na₄P₂O₇·10H₂O composites are stable at room temperature as shown in Fig. 3c (the lower panel). During the temperature programmed hydrogen generation measurement process, hydrogen evolution is not observed by MS detecter before the temperature is up to 40 °C. When the composite is heated to 40 °C, clear hydrogen evolution is observed. More than 70% theoretical hydrogen is released in the first 10 min and the final yield is 92%, corresponding to GHSC of 2.76% (Fig. 2b). After hydrogen release, the products are composed of Ca(OH)₂ (ICSD 98-020-2220) and Na₄P₂O₇ (ICSD 98-001-0370), as suggested by XRD (Fig. 2a). This suggests that the hydrogen generation is from the reaction between CaH₂ and the water in Na₄P₂O₇·10H₂O, which can be described as:

5CaH₂ + Na₄P₂O₇·10H₂O → 5Ca(OH)₂ + Na₄P₂O₇ + 10H₂ (1)

Fig. 2 Structure characterization and dehydrogenation study of the 5CaH₂ + Na₄P₂O₇·10H₂O composites: (a) XRD patterns before and after dehydrogenation, (b) isothermic dehydrogenation kinetics at 40 °C.

Fig. 3 (a) MS analysis of the generated gas for CaH₂ + Na₄P₂O₇·10H₂O. (b) TGA profiles of the hydrated inorganic salts. (c) Differential TGA curves of several hydrated inorganic salts (the upper panel) and the dehydrogenation profiles of CaH₂ + hydrated solids during temperature programmed heating (the lower panel). (d) Plot of the dehydrogenation peak temperature (T_H) with the dehydration peak temperature (T_w).

The above mechanism is in accordance with the dehydration behaviour of Na₄P₂O₇·10H₂O. Thermal gravimetric analysis (TGA, Fig. 3b) suggests that Na₄P₂O₇·10H₂O shows one single step weight loss of 39.5% in the temperature range between 40 and 75 °C, which matches well with theoretical mass percentage of H₂O in Na₄P₂O₇·10H₂O (40.4%). Thus, the hydrogen generation is still due to the hydrolysis reaction of CaH₂, while there is no liquid state water involved. We denote this new hydrogen generation scheme as the solid state hydrolysis (SSH) reaction. Mass spectroscopy (MS) measurement (Fig. 3a) suggests that the released gas is composed of H₂ with only small amount of water vapour without other impurities, which can be used for hydrogen fuel cells directly or with simple dehydration.

Clearly, other hydrated inorganic salts can also be used as the water carriers. Several hydrated salts with high water weight percentage are also tested for the SSH reaction with CaH₂. The results are summarized in Fig. 4 and Table 1. Although the hydrogen generation kinetics and the final yields vary significantly, there is an interesting correlation between the dehydrogenation temperature of the composites and the dehydration temperature of the hydrated salts. The peak dehydration temperature (T_w) is obtained from the derivative of the TGA curves (Fig. 3c, the upper panel). The dehydrogenation peak temperature (T_H) is obtained from the hydrogen release profile during the temperature programed heating process detected by MS (Fig. 3c, the lower panel). There is a clear positive correlation between T_H and T_w, as shown in Fig. 3d. This correlation is more convincingly demonstrated by MgCl₂·6H₂O, which exhibits two well separated dehydration steps before 150 °C. Correspondingly, there are also two dehydrogenation peaks. The above results are fully in agreement with the dehydrogenation mechanism proposed for the SSH reaction, as shown in eqn (1). This also provides the flexibility in tuning the dehydrogenation temperature in the SSH reaction by rational choice of hydrated solids with proper dehydration temperature.

Fig. 4 Isothermic dehydrogenation kinetics curves of Na₄P₂O₇·10H₂O at 40 °C, KAl(SO₄)₂·12H₂O at 50 °C, MgCl₂·6H₂O at 60 °C, ZnSO₄·7H₂O at 60 °C and NiSO₄·6H₂O at 80 °C.

Table 1 Summary of the solid state hydrolysis (SSH) composites studied in this work^a

Composition	Theoretical GHSC (%)	Tw (°C)	ΔwH (kJ mol^−1 H2O)	TH (°C)	Attained GHSC (%)	Enthalpy (kJ mol^−1 H2)
a GHSC: gravimetric hydrogen storage capacity. Tw: the dehydration peak temperature of the hydrated solid. TH: the dehydrogenation peak temperature of the SSH composite. ΔwH: the enthalpy to generate 1 mol H2O from the hydrated solid.
5CaH2 + Na4P2O7·10H2O	3.0	54	37.9	49	2.76	−78.2
6CaH2 + KAl(SO4)2·12H2O	3.3	70	19.5	54	1.91	−96.6
3CaH2 + MgCl2·6H2O	3.6	87	23.6 (ref. 9)	65	1.40	−92.5
3CaH2 + MgCl2·6H2O	3.6	120		100	1.62
3CaH2 + NiSO4·6H2O	3.1	90	36.8	80	1.24	−79.3
3.5CaH2 + ZnSO4·7H2O	3.2	62	13.9	63	1.02	−102.6

---

3.2 Hydrogen generation from NaBH₄ and hydrated oxalic acid: a more general SSH reaction

The SSH reaction for hydrogen generation can be generalized as recombination of the hydridic and protonic hydrogen in the solid state. Clearly, this type of reaction is not limited to CaH₂ + hydrated inorganic salts. Using hydrogen with more protonic characteristic, hydrides difficult to undergo direct hydrolysis can also serve as solid state hydrogen carriers. In this section, we demonstrate a more general SSH reaction using NaBH₄ and hydrated oxalic acid.

Direct hydrolysis of NaBH₄ is rather slow and incomplete.^10,11 In fact, NaBH₄ can form rather stable aqueous solution, particularly in alkaline solution. The H in the BH₄⁻ group shows much less hydridic characteristics compared to that in alkaline or alkaline earth metal hydrides, due to the very similar electronegativity of boron and hydrogen. Consequently, hydrogen with stronger acidity is required to facilitate the hydrogen generation. Demirci et al. demonstrated that hydrolysis of NaBH₄ can be rather complete in aqueous acid solution such as hydrochloride acid and acetic acid.¹² However, this approach has not been widely adopted for onsite hydrogen generation due to the inconvenience in storing and carrying the acid solution.

Oxalic acid is a solid organic acid with high percentage of protonic hydrogen. The solid state composite NaBH₄ + H₂C₂O₄·2H₂O exhibits theoretical GHSC of 4.9%. XRD study shows that the physical mixture of NaBH₄ and H₂C₂O₄·2H₂O (molar ratio 1 : 1) is stable up to 60 °C (Fig. 5a). During the temperature programmed heating process, the composite shows a sharp hydrogen release peak at 65 °C. Rapid dehydrogenation is achieved in isothermic dehydrogenation at 70–80 °C (Fig. 5b and c). The above features are very similar to the CaH₂ + hydrated salt systems. About 55% of theoretical hydrogen is released in about 2 min, corresponding to GHSC of 2.70%. MS analysis shows that the released gas is composed of hydrogen and trace amount of water vapour, without impurities like CO or CO₂ from decomposition of oxalic acid (Fig. 5d). After dehydrogenation, the solid is mainly composed of NaB(C₂O₄)₂ (ICSD 98-028-1622) and the unreacted NaBH₄, as suggested by XRD (Fig. 5a). There is no unreacted H₂C₂O₄·2H₂O detected. The main dehydrogenation reaction, therefore, is given by:

NaBH₄ + 2H₂C₂O₄·2H₂O → NaB(C₂O₄)₂ + 4H₂ + 4H₂O (2)

Fig. 5 XRD patterns (a) and isothermic H₂ generation kinetics (b) of NaBH₄ + H₂C₂O₄·2H₂O at 60, 70, 80 °C. (c) H₂ generation profiles of NaBH₄ + H₂C₂O₄·2H₂O during temperature programmed heating. (d) MS analysis of the generated gas.

The above reaction gives 50% conversion of NaBH₄ in the NaBH₄ + H₂C₂O₄·2H₂O (molar ratio 1 : 1) composite, which is in good agreement with the experimental results (55% conversion of NaBH₄). The additional 5% may be resulted from the water vapour induced dehydrogenation:

NaBH₄ + 4H₂O → NaBO₂ + 4H₂ (3)

The molar ratio of NaBH₄ and H₂C₂O₄·2H₂O in the composite is adjusted to optimize the GHSC. As shown in Fig. 6, increasing the fraction of H₂C₂O₄·2H₂O can increase the conversion percentage of NaBH₄. In the NaBH₄ + 3H₂C₂O₄·2H₂O composite, more than 95% NaBH₄ is converted. However, the GHSC of the composite decreases with the H₂C₂O₄·2H₂O fraction. The NaBH₄ + H₂C₂O₄·2H₂O (molar ratio 1 : 1) gives the highest GHSC of 2.78%.

Fig. 6 (a) Isothermic hydrogen generation kinetics of NaBH₄ + H₂C₂O₄·2H₂O composites with different reactant ratio. (b) The H₂ yield based on NaBH₄ and the system hydrogen capacity of NaBH₄ + H₂C₂O₄·2H₂O composites with different reactant ratio.

3.3 Comparison of the SSH reactions with other hydrogen generation techniques

The SSH reaction brings the hydrolysis of hydrides into the solid state, which provides an innovative solution for single use hydrogen sources. As the SSH reactions only involve solid state materials, the hydrides and hydrated solids can be premixed and stored in sealed packages. Hydrogen can be generated on demand by simply heating the solid composites to a desired temperature, which is very similar to thermal decomposition of metal hydrides and is much easier to operate compared to mixing liquid water with the hydride powder.

Despite of the inevitable loss in GHSC due to using hydrated solids instead of water, the SSH reactions remain provide very competitive GHSC. As shown in Table 1, the theoretical GHSC of several SSH composites well exceeds 3%. The attained GHSC for 5CaH₂ + Na₄P₂O₇·10H₂O and NaBH₄ + H₂C₂O₄·2H₂O reaches 2.76% at 40 °C and 2.78 wt% at 70 °C, respectively. For comparison, the metal hydrides with moderate dehydrogenation temperature typically show theoretical GHSC lower than 2% (LaNi₅H₆: 1.4%,¹³ TiFeH_1.95: 1.8%,¹⁴ TiMn_1.5H_2.4: 1.8%¹⁵).

When volumetric hydrogen storage capacity (VHSC) is also taken into account as a major figure of merit, the SSH reactions become even more competitive compared to direct hydrolysis. For instance, the theoretical VHSC for directly hydrolysis of CaH₂ is 66 g H₂ L⁻¹ (calculated from the total volume of the reactants CaH₂ + 2H₂O). In addition, the volume of the formed Ca(OH)₂ is expanded by 33% compared to that of CaH₂. Therefore, when generating hydrogen by gradually adding water into the CaH₂ container (which is the most common way to realized controlled hydrogen generation), the CaH₂ container should be at least 33% larger than the volume of CaH₂ to accommodate the volume expansion, which gives theoretical VHSC of 58 g L⁻¹ (based on the minimum internal volume of the CaH₂ and water containers). The theoretical VHSC of the SSH mixture 5CaH₂ + Na₄P₂O₇·10H₂O is 54 g H₂ L⁻¹ (density: 1.7 g cm⁻³ for CaH₂ and 1.8 g cm⁻³ for Na₄P₂O₇·10H₂O). There is no need to leave additional space in the container, as the volume is contracted by 27% when it is converted to the spent fuel 5Ca(OH)₂ + Na₄P₂O₇ (density: 2.24 g cm⁻³ for Ca(OH)₂ and 2.53 g cm⁻³ for Na₄P₂O₇). Therefore, the theoretical VHSC of the SSH reaction for CaH₂ is already comparable to that of direct hydrolysis.

For hydrogen generation from hydrolysis of NaBH₄, the benefit of the SSH reactions is more than just excluding the liquid water. Although the theoretical GHSC of NaBH₄ hydrolysis (NaBH₄ + 2H₂O) is as high as 10.8%, direct hydrolysis of NaBH₄ is very slow and incomplete. In most practical applications, NaBH₄ is stored as alkaline solution and the dehydrogenation needs to be catalysed by transition metal catalysts.^16–19 To avoid precipitation of the hydrolysis product NaBO₂ which will passivate the catalysts, the optimized concentration of NaBH₄ solution is around 15% and the corresponding theoretical GHSC is significantly reduced to only 3.2%.²⁰ The solid acid promoted SSH reaction described in this work achieves efficient hydrogen generation from NaBH₄ without using transition metal catalysts, which is very promising for low cost, efficient onsite hydrogen generation from NaBH₄. The attained GHSC 2.78% is very close to the theoretical value of 15% NaBH₄ solution. Aqueous acid solution has been studied to promote hydrogen generation from NaBH₄ solution,^12,21 while the solid state hydrated oxalic acid is clearly much more convenient to handle compared to the acid solution.

An additional advantage of the SSH reactions compared to direct hydrolysis reactions is the lower reaction heat. Hydrolysis of most reactive hydrides or metals is highly exothermic. In the SSH reactions, however, this heavy thermal burden can be effectively alleviated by coupling the endothermic dehydration reaction. Here we illustrate this effect using CaH₂. The dehydrogenation enthalpy Δ_dHH of the SSH reactions of CaH₂ is given by:

Δ_dHH = Δ_HyH(CaH₂) + Δ_wH = −116.05 + Δ_wH (4)

where Δ_HyH(CaH₂) = −116.05 kJ mol⁻¹ H₂ is the enthalpy for dehydrogenation from the hydrolysis of CaH₂ (ref. 22) and Δ_wH is the enthalpy for generation of 1 mol H₂O from the hydrated solids. As shown in Table 1, Δ_wH for hydrated inorganic salts with high water content can be up to near 40 kJ mol⁻¹ H₂O. Although the reaction remains exothermic (see the DSC curves in Fig. S1†), the hydrolysis heat can be significantly reduced. For instance, the 5CaH₂ + Na₄P₂O₇·10H₂O composite shows much lower reaction heat of −78.2 kJ mol⁻¹ H₂.

The SSH reactions in this work are representative examples of how recombination of the hydridic and protonic hydrogen in the solid state can promote hydrogen generation. Many hydrogen storage/generation systems are based on this strategy, such as LiNH₂–LiH,²³ ammonia borane^3,24 and its analogies,²⁵ amides/borohydrides composites^26,27 and ammoniates of borohydrides.^28,29 A recent work on AlCl₃·mNH₃–nLiBH₄ (ref. 30) (m = 3 to 6; n = 3 to 5) is a closer analogy to this work. Except very few examples, most of the above systems have very poor reversibility and practically can only serve as single use hydrogen carriers. Compared to the above systems, the SSH composites are composed of much cheaper materials and thus are more viable for onsite hydrogen generation.

Further improvement of the GHSC from the SSH reactions is possible. As summarized in Table 1, the GHSC of the SSH systems is generally much lower than the theoretical value. A major loss of the GHSC is due to the unreacted water vapour escaped from the system, as demonstrated by the MS analysis (Fig. 3a). By proper design of the reactor to elongate the residence time of the water vapour in the reactor, it is possible to reduce the unreacted water vapour and further enhance the GHSC from the SSH reactions. Moreover, better control of the hydrogen generation rate is also needed to meet the requirement of onsite hydrogen generation, which will be our next research focus.

4. Conclusions

In conclusion, efficient hydrogen generation can be achieved from the solid state reaction between hydrides and hydrated solids through a mechanism similar to the hydrolysis reactions. The dehydrogenation temperature can be tuned by the dehydration temperature of the hydrated solids, allowing temperature activated dehydrogenation. Two SSH composites 5CaH₂ + Na₄P₂O₇·10H₂O and NaBH₄ + H₂C₂O₄·2H₂O exhibit competitive GHSC of 2.76% at 40 °C and 2.78 wt% at 70 °C, respectively. The SSH reactions partly retains the high GHSC of hydrolysis reactions while the dehydrogenation can be conveniently initiated by simple heating, similar to thermal decomposition of metal hydrides, which provides an innovative solution for compact onsite hydrogen generation systems for portable hydrogen fuel cells.

Acknowledgements

The authors acknowledge the financial support from National Science Foundation of China (NSFC, No. U1201241, 11375020, 51431001 and 21321001).

Notes and references

1. J. Yang, A. Sudik, C. Wolverton and D. J. Siegel, Chem. Soc. Rev., 2010, 39, 656–675.
2. J. Graetz, Chem. Soc. Rev., 2009, 38, 73–82.
3. A. Staubitz, A. P. M. Robertson and I. Manners, Chem. Rev., 2010, 110, 4079–4124.
4. V. C. Y. Kong, F. R. Foulkes, D. W. Kirk and J. T. Hinatsu, Int. J. Hydrogen Energy, 1999, 24, 665–675.
5. V. C. Y. Kong, D. W. Kirk, F. R. Foulkes and J. T. Hinatsu, Int. J. Hydrogen Energy, 2003, 28, 205–214.
6. L. Z. Ouyang, J. M. Huang, C. J. Fang, H. Wang, J. W. Liu, Q. A. Zhang, D. L. Sun and M. Zhu, J. Alloys Compd., 2013, 580, S317–S319.
7. L. Ouyang, M. Ma, M. Huang, R. Duan, H. Wang, L. Sun and M. Zhu, Energies, 2015, 8, 4237–4252.
8. D. Strawser, J. Thangavelautham and S. Dubowsky, Int. J. Hydrogen Energy, 2014, 39, 10216–10229.
9. H. Jahn and G. Wolf, J. Solution Chem., 1993, 22, 983–994.
10. R. Retnamma, A. Q. Novais and C. M. Rangel, Int. J. Hydrogen Energy, 2011, 36, 9772–9790.
11. L. Yu and M. A. Matthews, Int. J. Hydrogen Energy, 2011, 36, 7416–7422.
12. O. Akdim, U. B. Demirci and P. Miele, Int. J. Hydrogen Energy, 2009, 34, 7231–7238.
13. G. Liang, J. Huot and R. Schulz, J. Alloys Compd., 2001, 320, 133–139.
14. J. J. Reilly and R. H. Wiswall Jr, Inorg. Chem., 1974, 13, 218–222.
15. T. Gamo, Y. Moriwaki, N. Yanagihara, T. Yamashita and T. Iwaki, Int. J. Hydrogen Energy, 1985, 10, 39–47.
16. S. S. Muir and X. Yao, Int. J. Hydrogen Energy, 2011, 36, 5983–5997.
17. Y. Kojima, K. Suzuki, K. Fukumoto, M. Sasaki, T. Yamamoto, Y. Kawai and H. Hayashi, Int. J. Hydrogen Energy, 2002, 27, 1029–1034.
18. C. Salameh, A. Bruma, S. Malo, U. B. Demirci, P. Miele and S. Bernard, RSC Adv., 2015, 5, 58943–58951.
19. U. B. Demirci and P. Miele, Phys. Chem. Chem. Phys., 2014, 16, 6872–6885.
20. U. B. Demirci, O. Akdim, J. Andrieux, J. Hannauer, R. Chamoun and P. Miele, Fuel Cells, 2010, 10, 335–350.
21. M. Sankaran and S. Vaidyanathan, J. Power Sources, 2009, 187, 216–223.
22. J. A. Dean, Langes Chemistry Handbook 15th, McGraw-Hill Book Co., U.S., 1999.
23. P. Chen, Z. T. Xiong, J. Z. Luo, J. Y. Lin and K. L. Tan, Nature, 2002, 420, 302–304.
24. C. W. Hamilton, R. T. Baker, A. Staubitz and I. Manners, Chem. Soc. Rev., 2009, 38, 279–293.
25. Z. Xiong, C. K. Yong, G. Wu, P. Chen, W. Shaw, A. Karkamkar, T. Autrey, M. O. Jones, S. R. Johnson, P. P. Edwards and W. I. F. David, Nat. Mater., 2008, 7, 138–141.
26. A. Borgschulte, M. O. Jones, E. Callini, B. Probst, S. Kato, A. Zuettel, W. I. F. David and S.-I. Orimo, Energy Environ. Sci., 2012, 5, 6823–6832.
27. J. Wang, H.-W. Li and P. Chen, MRS Bull., 2013, 38, 480–487.
28. F. Yuan, Q. Gu, X. Chen, Y. Tan, Y. Guo and X. Yu, Chem. Mater., 2012, 24, 3370–3379.
29. G. Xia, Q. Gu, Y. Guo and X. Yu, J. Mater. Chem., 2012, 22, 7300–7307.
30. Y. Song, F. Wu, X. Zheng, X. Ma, F. Fang and Y. Guo, Chem. Commun., 2015, 51, 1104–1110.

---

Footnote

† Electronic supplementary information (ESI) available: Details of sample preparation, additional structural and dehydrogenation characterization. See DOI: 10.1039/c6ra04410g

---