Catalysis and Downsizing in Mg-Based Hydrogen Storage Materials

Abstract

Magnesium (Mg)-based materials are promising candidates for hydrogen storage due to the low cost, high hydrogen storage capacity and abundant resources of magnesium for the realization of a hydrogen society. However, the sluggish kinetics and strong stability of the metal-hydrogen bonding of Mg-based materials hinder their application, especially for onboard storage. Many researchers are devoted to overcoming these challenges by numerous methods. Here, this review summarizes some advances in the development of Mg-based hydrogen storage materials related to downsizing and catalysis. In particular, the focus is on how downsizing and catalysts affect the hydrogen storage capacity, kinetics and thermodynamics of Mg-based hydrogen storage materials. Finally, the future development and applications of Mg-based hydrogen storage materials is discussed.

1. Introduction

In the past few decades, hydrogen energy has drawn a great deal of attention with the increase in the energy crisis and environmental pollution due to massive and long-term depletion of fossil fuels. Thus, it is of great importance to develop clean and renewable energy as a substitute for fossil fuels [1]. Hydrogen is considered to be a promising alternative energy carrier for future the because of its outstanding advantages, such as it has a wide variety of sources (it is the most abundant element in the universe), high energy density and environmentally friendly emissions (no CO₂ emissions, just H₂O) [2,3,4]. Hydrogen energy systems include production, storage, transportation, utilization of hydrogen, etc. Among these, hydrogen storage is a key technology limiting the commercialization of hydrogen energy. There are three types of methods to store hydrogen, which include compressed gas storage, low-temperature liquid storage and solid-state storage [2,5,6,7,8]. However, poor volumetric hydrogen density (4.4 MJ/L), high cost of the tank and safety issues have limited the wide application of compressed hydrogen storage technology. Although liquid hydrogen storage (8.4 MJ/L) has almost twice the energy content of hydrogen, hydrogen liquidation requires extremely low temperatures (−252 °C) to achieve enough hydrogen capacity, which greatly increases the cost [9,10]. In contrast, solid-state hydrogen storage is thought to be a promising method due to its relatively high energy density, low cost of containment and safety guarantee. Storage of hydrogen in solid-state materials is deemed to be an efficient and safe method. Therefore, one of the hot topics of research is the development of high capacity hydrogen materials and technology. Due to its low price (ca. 3 USD/kg), great abundance and high theoretical hydrogenation capacity (7.7 wt %), magnesium and its alloys are thought to be promising candidates for hydrogen storage materials [11,12,13]. Unfortunately, the problem of high bonding energy Mg-H causes its high dehydrogenation temperature and poor dehydriding kinetics, which have seriously hindered its commercial applications [14,15,16]. In order to improve the hydrogen storage performance of Mg-based materials, numerous efforts have been made to enhance the kinetics and destabilize the magnesium hydride, including downsizing, catalysis, etc. So, in this paper, we present some progress in the development of the Mg-based hydrogen storage materials related to the downsizing and catalysts and focus on how downsizing and catalysts could affect the hydrogen storage capacity, kinetics and thermodynamics of hydrogen storage materials.

2. Downsizing

In order to realize the downsizing of Mg-based materials, many nanotechnologies have been tried to synthesize these materials with a smaller grain size, including the hydrogen plasma metal reaction technique, high pressure reaction ball milling, thin film synthesis, catalyzed chemical solution synthesis, etc. Figure 1, shows that the kinetics can be expressed by the activation energy of reaction (E_a) while the thermodynamics can be characterized by the formation enthalpy (∆H) and entropy (∆S) of metal hydride [17]. Here, we will focus on how the E_a, ∆H and ∆S changes when the size is decreased to nanoscales.

2.1. Effect of Downsizing on Desorption Thermodynamics

There is a great deal of debate about the effect of downsizing on desorption thermodynamics. Whether downsizing influences the desorption thermodynamics is controversial. Thus, we examine the relationship between the desorption enthalpy and the diameter of different materials.

Some researchers consider that thermodynamic destabilization is enhanced by reducing the particle sizes of MgH₂. Theoretical calculations have predicted the size effect of magnesium-base materials. It has been suggested that the thermodynamics properties will be significantly improved and that the adsorption temperature subsequently reduces when the particle size decreases. Chen et al. [18] investigated the hydrogen storage of magnesium nanowires using the calculation of first-principles density functional theory. Results suggested that reducing the size of magnesium and magnesium nanowires could lead to a thermodynamic destabilization. The calculated desorption enthalpy for A2_MgH₂ (Φ0.85 nm) was 34.54 kJ/mol H₂, which is lower than that of A3_MgH₂ with a larger diameter of 1.24 nm (61.86 kJ/mol H₂). Furthermore, the desorption temperature for A2_MgH₂ nanowire is 264.25 K, which is close to ambient temperature.

De Jong et al. [19] used ab initio Hartree-Fock and density functional theory calculations to study the size effect on desorption thermodynamics of magnesium and magnesium hydride. They discovered that the downsizing of MgH₂ and Mg structures had lower desorption energy originating from the change in lattice energy. It was also found that hydrogen desorption energy dramatically decreases when the crystallite size is smaller than ~1.3 nm. For example, the desorption enthalpy for the MgH₂ cluster of 0.9 nm is 63 kJ/mol H₂ and the desorption temperature is only 473 K. In summary, it is clearly seen that magnesium hydride will destabilize significantly, due to the weakened bonding of H with magnesium, when the particle size is less than 2 nm. The above results of the theoretical calculations showed that reducing particle size to less than 2 nm can greatly enhance thermodynamic destabilization.

Besides the proven effect of nanosizing on desorption thermodynamics by computational results, some experimental results also verify the increase in thermodynamic destabilization when the grain size of Mg/MgH₂ is reduced to a few nanometers. Fang et al. [20] synthesized a nanostructured MgH₂-0.1TiH₂ material by milling the mixture of MgH₂ and TiH₂ for 4 h at room temperature under 13.8 MPa hydrogen pressure. MgH₂-0.1TiH₂ in the size range of 5~10 nm was obtained. Results also indicated that the nanostructured MgH₂-0.1TiH₂ (68 kJ/mol H₂) had lower desorption enthalpy than that of MgH₂ (75 kJ/mol H₂). However, in a similar experiment, Shao et al. reported that after milling under 30 MPa hydrogen, the obtained nanostructured MgH₂-0.1TiH₂ showed greatly enhanced kinetics, but the desorption thermodynamics was not changed [21]. Here it should be noted, that when downsizing and additive introduction are employed together in Mg-based materials, it is challenging to distinguish the specific enhancement mechanisms.

Li et al. [22] fabricated magnesium nanowires with a diameter in the range of 30~170 nm by using a vapor-transport approach and found that the desorption enthalpies for the magnesium nanowire of 30~50 nm, magnesium nanowire of 80~100 nm and rod-like magnesium particles of 150~170 nm were 65.3, 65.9 and 67.2 kJ/mol H₂, respectively, which are all smaller than that of bulk MgH₂ (75 kJ/mol H₂). These experimental findings were in agreement with computational results to some extent.

Although there are many positive results reported in the literature, some researchers have pointed out that downsizing magnesium materials might not change the desorption thermodynamic properties [23]. Theoretically, Goddard et al. [24] carried out simulations focused particularly on the size effect of MgH₂ in the process of hydrogen absorption/desorption. The results revealed that a steep increase in formation heat and a decrease in structural stability as the particle size decreases below 1.0 nm. Moreover, when the size of the nanoparticle is beyond 2 nm, the difference between the nanoparticle and the bulk can be ignored. The results theoretically explain why magnesium nanoparticles with a size of about 100 nm exhibit insignificant change in thermochemical properties. Other experimental findings also confirm that magnesium hydride particles with a diameter of approximately 100 nm show small thermodynamic change when theoretic simulation is carried out.

Vajo et al. [25] prepared a nanostructured magnesium hydride with a diameter of about 100 nm which was incorporated in a mesoporous carbon aerogel. Results revealed that no significant desorption thermodynamic change occurred in the magnesium hydride-filled nickel decorated aerogel. Similarly, the desorption enthalpies for the magnesium hydride incorporated into copper and magnesium hydride-filled aerogel without wetting layers did not change.

Lin et al. [26] synthesized MgH₂-based hydride composites with a crystallite size of about 20 nm from amorphous Mg₈₀Ce₁₀Ni₁₀ alloys via melt-spinning and hydrogenation. It was found that the desorption enthalpy and entropy of MgH₂ were −77.9 ± 0.3 kJ/mol and 139.8 ± 0.5 J/mol, respectively. The thermodynamics data was the same as that of MgH₂ with a size of about 100 nm in their previous study, indicating that the thermodynamics for MgH₂ desorption are not altered when MgH₂ particles are in the range of 20~100 nm.

Shao et al. [27] obtained magnesium ultrafine particles (magnesium UFPs) with an average size of about 300 nm via hydrogen plasma-metal reaction and found that the as-prepared magnesium UFPs showed excellent hydrogen storage performance. The hydrogen absorption content for the Mg UFPs without additives reached 7.59 wt % at 673 K in a few minutes, which was the highest of the studied temperatures. It could be concluded that the value of desorption enthalpy did not change a lot compared to other reported results. The nanostructured Mg₂Ni with diameters in the range of 30~50 nm via hydrogen plasma-metal reaction in 40 bar hydrogen at 623 K reported by Shao et al. [28] also indicated that the formation enthalpy for Mg₂NiH₄ is −66.32 kJ/mol H₂. This is in agreement with the reported results, which suggests that the enthalpy values are almost the same.

As mentioned above, Shao et al. [21] fabricated the nanostructured MgH₂/0.1TiH₂ via ball-milling a mixture of magnesium and titanium powders under an initial hydrogen pressure of 30 MPa. The average crystallite size of the powder after milling was 5 nm. Both the nanostructure and the catalyst TiH₂ may contribute greatly to the improvement in kinetics, but they do not change the desorption thermodynamics. The desorption enthalpy for as-prepared material and commercial magnesium are 77.4 kJ/mol H₂ and 78.5 kJ/mol H₂, respectively. This agrees with the reported data, and suggests that the enthalpy values do not fluctuate much.

Shao et al. [29] also synthesized the nanocrystalline Ti-catalyzed magnesium hydride in the range of 5~100 nm, by a homogeneously catalyzed synthesis method. Equations (1)–(3) show a typical procedure for fabricating Ti-catalyzed magnesium hydride. It was found that both the Ti catalyst and crystalline structure may contribute to the improvement in hydrogen storage kinetics. The Ti-catalyzed MgH₂ can absorb hydrogen under a temperature of 130 °C, and its hydrogen absorption rate at 300 °C is 40 times higher than that of commercial MgH₂ (Figure 2). In the desorption process, the equilibrium pressures for the nano-MgH₂^∗ are 0.0385, 0.0585 and 0.1362 MPa at 258, 272 and 299 °C, respectively. The desorption enthalpy and entropy values of the nano-MgH₂^∗ are 77.7 kJ/mol H₂ and 138.3 J/Kmol H₂, respectively, which implies that thermodynamic properties do not change with nanostructure and catalyst. Thus, it may be concluded that materials with a diameter above 5 nm do not change desorption thermodynamics.

	(1)
	(2)
	(3)

In order to more clearly understand the relationship between the thermodynamic properties and the size of particles, we tried to identify as many Mg-MgH₂ systems for hydrogen storage as possible. The temperature range, van’t Hoff equations, enthalpy and entropy change values of different Mg-MgH₂ systems are listed in

Table 1. The temperature range, van’t Hoff equations, enthalpy and entropy change values of different Mg-based systems.

System	Temperature Range (°C)	Van’t Hoff Equation	∆H (kJ/mol H2)	∆S J/(K mol H2)	Reference
MgH2-0.1TiH2 by UHEHP	240–290	logP(bar) = −3560.6/T + 6.6302	68.2	127	[20]
MgH2/0.1TiH2 by ball milling	269–301	ln(P/0.1 MPa) = −9308.8/T + 16.539	77.4	137.5	[21]
Commercial MgH2	330–370	ln(P/0.1 MPa) = −9445.1/T + 16.844	78.5	140
Magnesium Ultrafine Particles	350–400	ln(P/bar) = −9604/T + 16.93	79.8	140.8	[27]
Magnesium hydride	314–576	InfH2 = −2139.2/T + 3.88	73.1	135.8	[32]
MgH2-NiS	350–400	lnP = −9061.7/T + 18.5192	75.34	153.97	[33]
Unalloyed Magnesium Metal	277–427	ln(P/0.1MPa) = −8419.5/T + 15.155	70	126	[34]
Ti-catalyzed MgH2	258–299	ln(P/0.1 MPa) = −9345.7/T + 16.635	77.7	138.3	[29]
MgH2/CMK3-20	300–350	ln(P/0.1 MPa) = −6300.2/T + 8.580	52.38	71.33	[35]
MgH2/CMK3-90	308–350	ln(P/0.1 MPa) = −8620.4/T + 13.453	71.67	111.85
MgH2–nfTa2O5	275–325	ln(P/0.1 MPa) = −9141.2/T + 15.510	76	129	[36]
rGO–Mg	275–325	ln(P/0.1 MPa) = −8347.4/T + 15.191	69.4	126.3	[37]
Ni-doped rGO–Mg	250–300	ln(P/0.1 MPa) = −8046.7/T + 14.674	66.9	122
Pure MgH2	250–370	log(P/0.1 MPa) = −3893.8/T + 7.055	74.6	135.1	Sandia National Lab database

. Figure 3 shows the van’t Hoff plots of desorption reaction for different Mg-MgH₂ systems. In Table 1, it can be seen that commercial magnesium hydride has a ∆H of 78.5 kJ/mol H₂ and ∆S of 140 J/Kmol H_2. When we compare these values to those reported by other systems for MgH₂/0.1TiH₂ , for example, by ball milling (∆H of 77.4 kJ/mol H₂ and ∆S of 137.5 J/Kmol H₂), magnesium ultrafine particles (∆H of 79.8 kJ/mol H₂ and ∆S of 140.8 J/Kmol H₂), magnesium hydride (enthalpy change of ∆H kJ/mol H₂ and ∆S of 135.8 J/Kmol H₂), and Ti-catalyzed MgH₂ (∆H of 77.5 kJ/mol H₂ and ∆S of 138.3 J/Kmol H₂), we can further conclude that there is no apparent change of thermodynamic properties in desorption reactions when downsizing the diameter from 5 to 300 nm. This is also in agreement with Wagemans et al. [19] findings in their quantum chemical study.

Based on the above-mentioned results and considering the error margins of the experiments, it can be concluded that the thermodynamic properties do not change with nanostructure in the diameter range of 5–300 nm. According to the Gibbs free energy of the hydriding reaction equation, ∆G = RT lnP_eq/P_H2, it is found that the hydriding reaction is thermodynamically favored when P_H2 > P_eq, ΔG < 0 [30]. Since, P_eq depends upon the temperature, selecting a different temperature range would affect the P_eq and then the thermodynamics, which may not be affected by the downsizing. Recently, Shao et al. [31] found that the reaction pathway of the decomposition process is changed from two steps to three steps at 633 K, which indicates that we can indirectly change the thermodynamics properties by controlling the reaction pathway through catalysis design.

2.2. Effect of Downsizing on Kinetics

Kinetics is another issue that must be addressed before the studied hydrogen storage materials are widely used. It is known that fast kinetics shorten the time required to absorb/release hydrogen. Therefore, good kinetics is essential for hydrogen storage. It is reported that reducing the size of hydrogen storage materials could lead to a remarkable change in hydrogen absorption/desorption kinetics [12,38,39,40,41]. It is well proven that small particles of hydrides have an influence on hydrogen absorption/desorption kinetic properties, mainly due to shortened hydrogen diffusion and dissociation pathways, and enlarged surface free energy when the size decreases to below a few nanometers [42,43]. Numerous methods are applied to synthesize the hydrogen storage nanostructured materials, including ball milling, the hydrogen plasma metal reaction, nano-confinement, thin film synthesis, and catalyzed solution synthesis. Many experiments show that absorption and desorption kinetics can be enhanced by simply decreasing the size of the materials.

Shao et al. [44,45] synthesized six different magnesium based materials (325 mesh Mg, single crystal Mg, Mg nanoparticles, nanocrystalline Mg₅₀Co₅₀ BCC alloy and Mg thin film samples) to investigate their hydrogen absorption kinetics. Figure 4, shows that the grain sizes are in the range of 40~50 µm for 325 mesh Mg sample particles, and 100~700 nm for the Mg nanoparticles. The BCC Mg₅₀Co₅₀ alloy shows a particle size of about 1 to 3 µm with a few nanometers in crystallite size. For the 84 nm-thick Mg thin film capped by 10 nm Pd layer, the crystallite size is around 50~100 nm. Hydrogen absorption kinetics of the as-prepared materials have been obtained and it has been concluded that the absorption kinetics are significantly in nanoscale and micrometer scale material. For example, Mg nanoparticles have better absorption kinetics than 325 mesh Mg due to their smaller grain size. As a result, there is more chance and more surface area available for hydrogen to react with Mg materials. Besides, the hydrogen diffusion distance decreases in Mg nanoparticles. In addition, the nanostructured Mg₅₀Co₅₀ BCC alloy could absorb hydrogen at near room temperature (303 K) under 3.3 MPa hydrogen, and even at 258 K, achieving a hydrogen capacity of up to 2.65 wt % under hydrogen pressure of 8 MPa (Figure 5). This is the lowest absorption temperature reported so far for Mg-based materials. On the contrary, the single crystal Mg with a millimeter scale does not obtain any hydrogen absorption ability below 573 K. Thus, downsizing the diameter can enhance the hydrogen absorption/desorption kinetics properties.

3. Catalysis

The natural kinetics properties of pure magnesium are very slow, therefore high pressure and high temperatures are required to improve the kinetics. When it comes to thermodynamics, magnesium hydride needs even higher temperature to desorb hydrogen [30,46,47]. In addition, the activation process of hydrogenation and dehydrogenation requires high temperatures of approximately 350 °C under hydrogen pressure to 70 atm without any additives [48]. In order to accelerate the kinetics of hydrogen absorption and desorption, numerous studies have been done to improve these performances. Catalytic additives (like transition metal, metal oxide, halides, etc.) are combined with Mg-based materials to enhance the kinetics of absorption and desorption.

3.1. The Doping of Transition Metals

The doping of transition metals can be an effective way to enhance the kinetics of hydrogenation/dehydrogenation. Many researchers have doped different transition metals to improve the kinetics. Hanada et al. [49] prepared magnesium hydride doped with different 3d-transition metals (Fe, Co., Ni, and Cu) by a mechanical ball milling method, and found that all doped MgH₂ composites showed better hydrogen desorption properties than that of pure MgH₂. They also reported that the doping of Ni performed best among these doped MgH₂ materials. The superior hydrogen desorption properties of Ni doped MgH₂ are mainly ascribed to the fact that the Mg₂Ni phase is generated in the boundary between the MgH₂ phase and Ni catalyst after hydrogenation. In addition, the decrease in activation energy for hydrogen desorption after doping by Ni also explains the phenomenon.

Liang et al. [50] fabricated the composites of MgH₂ + 5at%Tm (Tm = Ti, V, Mn, Fe, Ni) to investigate the hydrogen desorption of the mixture. Results showed that all the doped composites presented better absorption kinetics than that of the undoped MgH₂. Moreover, the catalytic effect of the five metals (Ti, V, Mn, Fe, Ni) is different. The composites containing Ti showed the best absorption kinetics and MgH₂ + 5 at%V exhibited the most rapid desorption kinetics at low temperature among these metal doped composites. In order to further investigate the catalytic effect of vanadium, they ball-milled the mixture of magnesium hydride and 15 wt % vanadium. The enhanced kinetics are attributed to the vanadium hydride working as a hydrogen pump, the specific area, smaller particle size and the introduction of defects [51].

Cui et al. [52] fabricated nanostructured Mg-based materials coated by multi-valence Ti-based catalysts by the reaction of magnesium powder in THF solution with TiCl₃. It was found that the Mg-Ti with a crystal size of ~10 nm significantly reduced hydrogen absorption and desorption temperatures. The coated magnesium begins to release H₂ at about 175 °C and releases 5 wt % H₂ within 15 min at 250 °C. The dehydrogenation entropy also changes from 130.5 J K⁻¹ mol⁻¹ H₂ to 136.1 J K⁻¹ mol⁻¹ H₂. The schematic diagram of the catalytic mechanism is presented in Figure 6. The authors believe that the improvements may originate from the multiple valence Ti acting as the intermediate for electron transfers between Mg²⁺ and H⁻, which significantly contributed to better dehydrogenation.

Lu et al. [53] obtained core-shell structured Mg@TM (TM = Co., V) composites via an approach combining the arc plasma method and electroless plating. Results showed that the hydrogen absorption/desorption enthalpies for Mg@Co.@V (−70.02/74.83 kJ/mol H₂) were both lower than those for the Mg@Co. (73.25/81.47) and Mg@V (−73.91/79.77) samples. In addition, the E_a value is 73.22 kJ/mol H₂ for Mg@Co., 86.30 kJ/mol H₂ for Mg@V and 67.66 kJ/mol H₂ for Mg@Co.@V, which are all lower than that of the pure magnesium ultrafine powder (118.20 kJ/mol H₂). These enhancements may be attributed to the core-shell microstructures and catalytic effects of both V and Co.

Zou et al. [54] prepared the Mg-Transition metal (TM)-La (TM = Ti, Fe, Ni) ternary composites via arc plasma evaporation and carefully investigated the hydrogen sorption properties of the as-prepared composites, leading to significantly improvement in the kinetics of hydrogen absorption and reduction in the temperature of hydrogen desorption. The improvement was attributed to catalytic effect of both Mg₂Ni and La₂O₃.

Magnesium with 10 wt % of metal nanoparticles (Fe, Co., Ni, Cu and Zn) was ball-milled at a rotation of 300 rpm for 4 h under the inert atmosphere of argon by Yu et al. [55]. They confirmed that there was no new phase formation between magnesium hydride and the introduced Fe, Co., Ni metals even after several cycles of hydrogenation/dehydrogenation. In addition, the introduction of metal particles into MgH₂ resulted in lower hydrogen desorption temperatures than that of pure magnesium hydride. Also the nickel-contained magnesium hydride showed the best result with the lowest desorption temperature centered at 203 °C, among these as-prepared materials. Interestingly, the added Zn and Cu metal act as an inhibitor without catalyst in the hydrogenation process because they do not improve the hydrogen absorption kinetics of MgH₂-10-Zn^nano and MgH₂-10-Cu^nano.

Kuji et al. [56] prepared ternary Mg-Tm-V (Tm = Ni, Co., Cu) alloys by mechanical alloying of powder mixture with a milling time from 5 to 25 h at the rate of 12.3 rev/s. The results showed that the hydrogen storage capacities for Mg_1.0Ni_1.0V_1.0 was 2.3 wt %, which was followed by Mg_1.0Co_1.0V_1.0 (1.44 wt %) and Mg_1.0Cu_1.0V_1.0 (0.95 wt %) alloys at 298 K under the pressure of 3 MPa. In addition, all the as-prepared materials were mechanically alloyed to the BCC structure.

Xie et al. [57] synthesized different amounts of nickel doped magnesium hydrides by ball milling at the rate of 300 rpm for 2 h under 2 bar hydrogen pressure. The results of XRD and electron microscopy indicated that the as-prepared MgH₂ + Ni materials show almost the same particle size. Besides, the obtained MgH₂ + 10 wt %Ni desorbed 6.1 wt % hydrogen in 10 min at 523 K under an initial hydrogen pressure of about 0.01 bar, followed by MgH₂ + 25 wt %Ni, MgH₂ + 50 wt %Ni, MgH₂ + 80 wt %Ni and MgH₂. Xie et al. identify and discuss the four effects of the doping of nickel on the dehydrogenation of MgH₂ as follows: the decomposition of hydride phase to form magnesium phase, the diffusion of hydrogen atoms, the combination of hydrogen atoms on the surface, and the desorption of hydrogen molecules. They conclude that the introduction of nickel is mainly to accelerate the combination of hydrogen atoms on the surface of MgH₂.

The method of melt-spinning was adopted to fabricate the nano-crystalline magnesium-rich magnesium-nickel-yttrium (Mg-Ni-Y) ternary alloys (Mg₈₀Ni₁₀Y₁₀ and Mg₉₀Ni₅Y₅) under argon atmospheres by Kalinichenka et al. [58]. They investigated the crystal structure, crystallization behavior and cyclic hydrogenation/dehydrogenation properties of the as-prepared materials. It was found that both Mg-Ni-Y ternary alloys mainly contain Mg(Ni,Y) nanocrystals with an average size in the range of 5~20 nm embedded in an amorphous matrix. Also, the Mg₈₀Ni₁₀Y₁₀ alloy presents higher thermal stability and a faster hydrogenation rate than that of the Mg₉₀Ni₅Y₅ alloy, which may be due to the higher content of yttrium and nickel. In addition, the two activated alloys can absorb almost 5.3 wt % hydrogen at a temperature of 280 °C and under the pressure of 20 bar H₂.

Cho et al. [37] reported the 3d transition metals (Ti, Cr, Mn, Fe, Co. and Ni) doped magnesium crystals encapsulated by rGO layers by a solution based, one-pot synthesis method. Based on the TEM image of the Ni-doped rGO–Mg nanocomposite, the diameter of magnesium crystals is about 3.56 nm. In Figure 7c, the XRD data shows a clear Mg crystalline structure without oxidation, which is evidence that the Mg is successfully encapsulated by molecular-sieving rGO layers (Figure 7a). It was also found that the dopants, Ni, Cr and Mn show preferable absorption rate when compared with other dopants during the first absorption (Figure 7d). In addition, Ni-doped Mg crystals embedded in rGO layers (Ni-doped rGO–Mg) can absorb 6.5 wt % hydrogen and most absorbed hydrogen of about 90% was completed within 2.5 min at 200 °C. When it comes to desorption, the absorbed hydrogen can completely desorb at 300 °C, of which 90% was dehydrogenated within 4.6 min. In summary, the as-prepared Ni-doped rGO–Mg present excellent hydrogen storage properties owing to the synergistic effect of nanosizing, rGO encapsulation and Ni doping. These three complementary functional materials will provide a novel way to make metal hydride.

Shao et al. [59] ball-milled a mixture of magnesium powder and cobalt powder at the rotation speed of 200 rpm for varying times from 0.5 h to 400 h under 0.1 MPa argon and then investigated the phase, morphology, hydrogen storage properties and mechanisms of nanostructured Mg₅₀Co₅₀ materials. It was found that the particle size reduces within 100 h and is not further changed after ball milling. In addition, the morphologies of magnesium based materials are transformed from the Co. phase with FCC structure, Co. phases with HCP and FCC together, then FCC Co. and partial bcc Mg-Co., into totally Mg₅₀Co₅₀ BCC structure with an average grain size of 1~5 nm. Moreover, the as-prepared Mg₅₀Co₅₀ BCC alloy can absorb hydrogen of 2.67 wt % at 258 K under a hydrogen pressure of 8 MPa and still retain the BCC structure after hydrogenation. It is known that the temperature of 258 K is the lowest reported so far for magnesium based materials to absorb hydrogen. Using the hydrogen plasma metal reaction, some doped magnesium materials (Mg₂Ni, Mg₂Co and Mg₂Cu) with an average particle size in the range of 50~200 nm were obtained and it was found that the doping of Ni, Co. and Cu show excellent hydrogen storage properties [60,61,62]. The study also reported the nanostructured Mg-Ni system alloys via the mechanical alloying method under 0.1 MPa argon at a rotation speed of 200 rpm for different time periods. Among these as-prepared Mg-Ni system alloys, the Mg₆₀Ni₄₀, Mg₅₀Ni₅₀, Mg₄₀N₆₀ and Mg₃₃Ni₆₇ alloys with a particle size of 1~3 nm show a BCC structure while the Mg₃₀Ni₇₀ of 5~10 nm show a FCC structure. In addition, the Mg₅₀Ni₅₀ BCC alloy that can absorb 1.85 wt % at 373 K in 7 MPa hydrogen shows the best hydrogen absorption kinetics and absorption efficiency due to the nanostructure and BCC phase when compared with other alloys [63].

Zhang et al. [64] proposed a novel strategy named microencapsulated nano-confinement for the preparation of nearly monodispersed nano-Mg₂NiH₄ anchored onto a graphene sheet surface via hydriding chemical vapor deposition (HCVD), which is presented in Figure 8. The as-prepared material possesses ultrahigh structural stability and superior desorption kinetics. The MgO coating layer with a thickness of about 3 nm efficiently separates the nanoparticles from each other to prevent aggregation during hydrogen absorption/desorption cycles, leading to excellent thermal and mechanical stability. More interestingly, the MgO layer shows superior gas-selective permeability to prevent further oxidation of Mg₂NiH₄ meanwhile accessible for hydrogen absorption/desorption (Figure 8b). As a result, an extremely low activation energy of 31.2 kJ mol⁻¹ for the dehydrogenation reaction is achieved. The study provides alternative insights into designing nano-sized metal hydrides with both excellent hydrogen storage activity and thermal/mechanical stability, exempting surface modification by agents.

In conclusion, the transition metal added to the magnesium hydride can improve the kinetics of hydrogenation/dehydrogenation, which originates from the introduced catalytic effect, such as the hydrogen pump, the surface/boundary reaction, electron transfer, etc. These ideas provide an insight into making metal hydride.

3.2. Introduction of Metal Oxides

Metal oxides are also valid catalysts to improve the absorption/desorption kinetics of MgH₂. Metal oxides (Sc₂O₃, TiO₂, V₂O₅, Cr₂O₃, Mn₂O₃, Fe₃O₄, CuO, Al₂O₃ and SiO₂) were employed to enhance the hydrogen absorption/desorption kinetics by Oelerich et al. [65]. The initial magnesium hydride was pre-milled for 20 h and then a further 100 h after adding the different metal oxides. Results showed that the addition of metal oxides can lead to a notable enhancement of both absorption and desorption kinetics. Furthermore, among these metal oxides, the MgH₂ + 5 mol% Cr₂O₃ composite obtained the fastest hydrogen absorption as it achieved a capacity of 4.7 wt % within 2 min at 300 °C, while the additives of V₂O₅ and Fe₃O₄ showed the most rapid desorption of hydrogen. The excellent kinetics may originate from the local electronic structure of the catalysts and a very high defect density introduced by high energy ball milling.

Barkhordarian and collaborators reported several metal oxide catalysts (Nb₂O₅, Fe₃O₄, et al.) doped MgH₂. The Nb₂O₅ doped sample presented the fastest hydrogen sorption kinetics among these metal oxides [66]. 6.9 wt % of hydrogen was absorbed within 60 s and the absorbed 6.9 wt % of hydrogen was fully desorbed in only 140 s. The results of the desorption rate for different metal oxides are listed in Figure 9. It is clearly seen that the Nb₂O₅ dopant possessed the fastest desorption rate compared with that of other dopants, yielding a desorption rate of 0.011 wt %/s. It is suggested that metals with multiple valences and the catalytic effect of electronic exchange reacting with hydrogen molecules are able to accelerate the gas-solid reaction. Similar findings also were reported by Ichikawa et al. [67]. In the literature, the effect of adding 1 mol% niobium oxide into the MgH₂ system by ball-milling for 20 h at 400 rpm was investigated. It was also found that the MgH₂ + 1 mol% Nb₂O₅ quickly absorbs hydrogen at ambient temperature even when the hydrogen pressure is lower than 0.1 MPa. At 150 and 250 °C, more than 5.0 wt % hydrogen gas is absorbed within 30 s and its final capacity reached 5.7 wt %. We can see that the amount of additive Nb₂O₅ is small. So, Aguey-Zinsou et al. [68] prepared magnesium hydride doped by a large amount of Nb₂O₅ (17 wt %). They found that magnesium hydride milled for 200 h has faster hydrogen absorption-desorption kinetics than that for 2 h and magnesium hydride doped by Nb₂O₅ is further improved. They considered that the further enhanced kinetics for magnesium doped by Nb₂O₅ can be ascribed to the effect of Nb₂O₅ which acts as a lubricant, dispersing and cracking agent. Conceição et al. [69] also investigated the hydrogen storage properties of Nb₂O₅ combined with magnesium hydride by ball-milling the mixture at the rate of 300 rpm for 24 h under H₂. They found that the MgH₂ + 5 wt % synthesized-Nb₂O₅ can absorb 5.2 wt % of hydrogen within 1.3 min and desorb almost 6.0 wt % in 5.8 min at the temperature of 300 °C, which are the best results among the as-prepared materials. In addition, they point out that the surface area and the amount of Nb₂O₅ greatly affect the hydrogen storage properties. It is easy to conclude that the additive Nb₂O₅ may contribute to superior enhanced absorption/desorption kinetics in Mg-based materials.

Kumar et al. [36] ball-milled the mixture of magnesium hydride and 5 wt % nano-flakes of Ta₂O₅ (nfTa₂O₅) for 2 h to investigate the effect of the additive on kinetics and thermodynamics of absorption and desorption. Results showed that the MgH₂–nfTa₂O₅ could absorb hydrogen at a temperature of 290 K and absorb more than 5 wt % hydrogen within 10 min at 373 K. The enthalpy values of hydrogen absorption and desorption were calculated to be 80 ± 2 and 76 ± 3 kJ/mol, respectively, which indicated that there is no change in thermodynamics for magnesium hydride doped by nfTa₂O₅. In summary, the nfTa₂O₅ dopant can enhance the kinetics without altering the thermodynamics.

Mustafa et al. [70] reported the hydrogen storage properties of the ball-milled MgH₂ + 5 wt % CeO₂ at a speed of 400 rpm in an argon atmosphere. The results showed that the hydrogenation rate of CeO₂ doped MgH₂ composites was faster than pristine magnesium hydride. In addition, about 3.6 wt % hydrogen is released within 30 min for MgH₂ + 5 wt % CeO₂ composites while less than 1.0 wt % hydrogen is released for pure MgH₂ (Figure 10). Moreover, the activation energies of the CeO₂ doped MgH₂ composite and pure MgH₂ were determined to be 133.62 and 108.65 kJ/mol, respectively, which indicated that the introduction of CeO₂ could lead to a great improvement in hydrogen desorption. That is, the enhanced performance of hydrogen storage is due to the formation of CeH₂ and CeO₂ species.

Recently, Lin et al. [71] reported a novel method to fabricate a composite of CeH_2.73/CeO₂ nanoparticles in Mg-based hydrides from hydrogenation and oxidation upon an amorphous Mg₈₀Ce₁₀Ni₁₀ alloy. Interestingly, there was a spontaneous hydrogen release effect at the CeH_2.73/CeO₂ interface, which leads to a more dramatic increase of catalysis than either CeH_2.73 alone or the CeO₂ catalyst.

Milosevic et al. [72] fabricated two doped MgH₂ composites with different contents (5 and 15 wt %) of VO₂ with particle size in the range of 0.6~100 μm via mechanical milling. The 5 wt % VO₂ doped MgH₂ released 4.9 wt % hydrogen after about 120 s at 350 °C in the pressure of 1 bar while the other reached a capacity of 4.3 wt % after 85 s at 350 °C in the pressure of 0.7 bar for the first desorption cycle. In addition, the activation energies of desorption for MgH₂ + 5 wt % VO₂ and MgH₂ + 15 wt % VO₂ composites were 54 ± 5 and 65 ± 5 kJ/mol H₂, respectively. The two values decrease significantly when compared with that of the pure MgH₂ (161 kJ/mol H₂), suggesting fast hydrogen desorption kinetics. The improved performance may be attributed to structure defects, numbers of vacancies and the presence of VO₂/VH₂.

Cabo et al. [73] ball-milled a mixture of magnesium hydride and 5 wt % mesoporous oxides (Co₃O₄, NiCo₂O₄ and NiO) synthesized by multi-step nano-casting at a rate of 360 rpm under an argon atmosphere for 24 h. They found that the values of desorption temperature and desorption enthalpy for the NiO-doped magnesium hydride is the lowest, followed by NiCo₂O₄, Co₃O₄ and pure MgH₂. When it comes to absorption/desorption rate, the introduction of NiCo₂O₄ shows the fastest absorption rate, while the NiO-doped magnesium hydride shows the highest desorption rate.

Different metal oxides (Cr₂O₃, TiO₂, Fe₃O₄, Fe₂O₃, In₂O₃ and ZnO) and commercial magnesium hydride powder were ball-milled in a Uniball-5 mill for 20 h by Polanski et al. [74]. The catalytic effects of these metal oxides introduced in the magnesium hydride on the hydrogen sorption properties were also investigated. It is found that these six additives doped in magnesium hydride present various behaviors. The Cr₂O₃ doped and TiO₂ doped greatly improve the hydrogenation process. More importantly, the Cr₂O₃ doped absorbed hydrogen to nearly full capacity within 2 min. The Fe₃O₄ and Fe₂O₃ doped showed a negligible influence while the In₂O₃ and ZnO doped seem to inhibit hydrogenation. Besides, all the metal oxides studied except for In₂O₃ improve the dehydriding properties with respect to the desorption kinetics. It is clearly seen that the In₂O₃ dopant seems to inhibit the desorption reaction after 30 min when compared with the magnesium hydride without any additives. Concerning the decomposition enthalpy, the study found that this may not be affected by the introduction of metal oxide during the second cycle while it has a smaller value during the first cycle due to a high numbers of defects.

Chen et al. [75] prepared MgH₂ and ZrO₂ composites in molar fractions of 99:1 and 95:5 by ball milling the mixture for 20 h under the pressure of 1 atm Ar atmosphere. They found that the grain sizes range for MgH₂-ZrO₂ with molar ratios of 99:1 and 95:5 were 80~200 nm and 60~100 nm, respectively. Besides, the hydrogen storage capacity of MgH₂ and ZrO₂ composites with a molar ratio of 95:5 is 6.75 wt % at 423 K, and is still 4.0 wt % at 298 K. In addition, the activation energy for the hydrogenation reaction of MgH₂-ZrO₂ composites with a molar ratio of 95:5 is 13.05 kJ/mol H₂. The improved performance of as-prepared materials was attributed to the grain refining effect induced by the Zr catalyst.

In conclusion, metal oxides can be a good alternative to introduce into Mg-based materials to improve the hydrogen storage kinetics.

3.3. Halides Doped

It is reported that the introduction of some halides into Mg-based material can have an influence on hydrogen storage properties.

Ivanov et al. [76] investigated the effects of different inorganic salts (NaF, NaCl, MgF₂ and CrCl₃) ball-milled with Mg-based materials, on the hydrogenation and dehydrogenation properties. It was found that the introduction of inorganic salts not only boost the metal powdering, but also modify the metal surface. Besides, doping of different inorganic salts shows various reaction kinetics, especially for the first cycle of hydrogenation. Only the Mg-NaCl composite showed a sigmoid kinetic curve at first hydriding but without an induction period contrary to other as-milled composites, which may be due to their natures.

Jin et al. [77] fabricated a series of 1~10 mol% transition metal fluorides (NiF₂, TiF₃, VF₄, NbF₅, ZrF₄, CrF₂, FeF₂, CuF₂, CeF₃, YF₃) doped magnesium hydride to study their hydrogenation and dehydrogenation properties. After milling for 15 min, they found that it is these hydride phases formed by the reaction between MgH₂ or metal–hydrogen solid solutions that improve the hydrogenation kinetics rather than the fluorides. A similar result was also reported by Wang et al. [78] for the doping of Ti-based materials.

Danaie et al. [79] used transmission electron microscopy analysis to investigate the TiF₃ doped magnesium hydride. Results showed that the catalyzed TiF₃ doped magnesium hydride has excellent hydrogen storage kinetics at various temperatures. In addition, the catalyzed TiF₃ doped magnesium hydride could be cycled many times with no significant degradation and both the hydride and the metal phase can be found during the absorption and desorption processes, which is very different from the pure magnesium hydride. Moreover, the number of hydride nuclei does not increase too much in the process of absorption while the introduction of TiF₃ can dramatically increase the number of the newly formed magnesium crystallites. Besides, the doped TiF₃ also boost extensive twinning in the hydride phase.

Lin et al. [80] ball-milled the mixture of magnesium hydride and cerium fluorides (CeF₃ and CeF₄) at the rate of 400 rpm for 4 h. It was found that the introduction of CeF₃ present almost the same values of hydrogen desorption temperature and activation energy as those of the ball-milled magnesium hydride, while the doping of CeF₄ decreased significantly. The reduced hydrogen desorption temperature and activation energy of CeF₄ doped magnesium hydride may be due to the easy electron transfer induced from the high valence Ce-cation and the formation of the F-containing Ce-F-Mg species on the CeF₄/MgH₂ interface.

Ismail [81] used 10 wt % LaCl₃ ball-milled with magnesium hydride to study the hydrogen storage properties. He found that the introduction of LaCl₃ decreased the desorption temperature (50 °C less) and the activation energy of dehydrogenation (23 kJ/mol less) compared with as-milled pure magnesium hydride. When it comes to absorption and desorption kinetics, the absorption capacities for the doped and pure magnesium hydride are 5.1 wt % and 3.8 wt % in 2 min at 300 °C. Besides, 4.2 wt % hydrogen is released for the doped magnesium hydride while only 0.2 wt % for the undoped magnesium hydride. In summary, the doping of LaCl₃ can enhance the kinetics of hydrogen storage due to the catalytic effect of the La-Mg alloy and the MgCl₂ formed during the heating process.

A series of halide additives (NbF₅, NbCl₅, TaF₅, ZrF₄, ZrCl₄, CeF₃, CdF₂, CdCl₂, TiCl₃, TiF₃, BaCl₂, BaF₂, NaF, VCl₃, FeF₃, FeF₂, CrCl₂, CrF₂, CrCl₃) doped MgH₂ were investigated by Malka et al. [82]. The results showed that the magnesium hydride and halide additive composites present a strong catalytic effect on the magnesium hydride desorption process. Additionally, the fluorides exhibit a better catalytic effect when compared with chlorides, which is in agreement with the results reported by Ma et al. [83]. It was also found that the desorption temperatures for halides of higher oxidation states are much lower than that of lower oxidation states and the halides from group IV and V are the best catalysts among the studied materials.

To sum up, doping with halides can improve the kinetics by decreasing the desorption temperature and the activation energy. However, the hydrogen storage capacity is reduced owing to the introduction of the above additives which normally do not absorb hydrogen.

3.4. Other Additives

Besides the above-mentioned additives, some other additives have also been investigated, such as sulfide [33,84,85,86], and nitride [87,88,89,90]. Jia et al. [85] fabricated MoS₂ and then investigated the effects of MoS₂ additive on the hydrogen storage properties. It was found that the hydrogen storage capacity of MgH₂ doped by MoS₂ (MgH₂-MoS₂) was 3.01 wt %, which is higher than that of pure MgH₂ (1.88 wt %) under the same conditions. In addition, the hydrogen absorption/desorption for MgH₂-MoS₂ was also improved. When it comes to hydrogen absorption kinetics, the uptake time for reaching 90% of the maximum hydrogen storage capacity of MgH₂-MoS₂ and pure MgH₂ was 72 min and 13 min, respectively. This means that the doping of MoS₂ is effective in ameliorating the hydrogen absorption kinetics of MgH₂. The e enhancement of hydrogen absorption and desorption kinetics was ascribed to the catalytic effect of MgS and Mo or MgO and Mo. Zhang et al. [86] ball-milled a mixture of MgH₂ and 20 wt % Fe₃S₄ to investigate the effects of Fe₃S₄ on hydrogen storage properties. Results showed that the MgH₂ added by Fe₃S₄ presented three times higher hydrogen storage capacity and a two times faster hydrogen desorption rate than that of pure MgH₂. In addition, the introduction of Fe₃S₄ decreased the dehydrogenation temperature by 90 K in comparison with the pure MgH₂. The improvement in hydrogen storage performance may be attributed to the catalytic effect of the new formation of Fe and MgS. Xie et al. [33] prepared flower-like NiS particles and subsequently ball-milled with Mg to study hydrogen storage behaviors. It quickly absorbed 3.5 wt % hydrogen within 10 min and desorbed 3.1 wt % hydrogen within 10 min at 573 K. The activation energies of hydrogen absorption and hydrogen desorption decreased to 45.45 and 64.71 kJ/mol, respectively. The authors thought that the enhancement of hydrogen storage properties may be attributed to the synergistic catalytic effects of the in situ formed MgS, Ni and Mg₂Ni multiple-phase catalysts during the hydrogenation/dehydrogenation process.

Besides the introduction of sulfides, nitrides have also been added into the Mg/MgH₂ system to improve the hydrogen storage performance. Different with previous discussed additives, here the nitrides contributed to hydrogen absorption/desorption capacity. Zhang and collaborators ball-milled the MgH₂ and Li₃N mixture with a ratio of 1:1. They found that the MgH₂-Li₃N had a capacity of about 3.2 wt % under a pressure of 10 MPa with an onset absorption temperature of 403 K [88]. Based on their previous work, Zhang et al. [89] ball-milled different ratios of MgH₂ and Li₃N mixture (Li₃N:MgH₂ = 0.5, 1, 2) under the same conditions to obtain a high amount of LiMgN and study the hydrogen storage performances. It was found that the hydrogen storage capacity and dehydrogenation storage capacity were both advanced by increasing the content of Li₃N. Among the three obtained samples, the MgH₂-2Li₃N absorbed almost 7.3 wt % hydrogen at 523 K under the hydrogen pressure of 10 MPa and released 3.3 wt % hydrogen at 623 K under 0.01 MPa. The existence of LiMgN and excess of Li₃N may explain why the hydrogen storage performance of ball-milled MgH₂-Li₃N was improved. Wang et al. [90] fabricated the Ni₃N@N-doped carbon (Ni₃N@NC) with a core-shell structure and subsequently ball-milled the MgH₂ and as-obtained Ni₃N@NC with different ratios to investigate the differences in hydrogen storage performance. Results showed that the MgH₂-5 wt % Ni₃N@NC composites presented the best hydrogen storage performances among the studied samples. The MgH₂-5 wt % Ni₃N@NC composites desorbed 6.0 wt % hydrogen (only 0.2 wt % hydrogen for the pure MgH₂) within 20 min at 598 K and absorbed 6.1 wt % hydrogen within 100 s (30 min for pure MgH₂) at 573 K for rehydrogenation under the same conditions. In summary based on the above-mentioned reports, the sulfides and nitrides additives introduced in the Mg/MgH₂ systems greatly enhanced hydrogen storage performance.

4. Summary and Outlook

Hydrogen storage is a key challenge which needs to be resolved in order to move towards a hydrogen society. Mg-based materials are thought to be promising candidates owing to their many advantages, which include low price, great abundance and theoretically high hydrogenation capacity. However, poor kinetics and high desorption temperature hinder their application as hydrogen storage materials. For onboard storage, it is expected that effective Mg-based materials can be synthesized to enhance the kinetics and decrease the desorption temperature for one bar hydrogen equilibrium pressure below 100 °C. Tremendous progress in the kinetics enhancement and thermodynamic tailoring of downsized Mg/MgH₂ or those doped by various additives (transition metals, metal oxides, halides, etc.) has been made. The positive effect of downsizing and catalysts on kinetics is widely accepted among researchers. However, for the downsizing effect on the absorption and desorption thermodynamics of Mg-MgH₂, the hydrogen storage community reports different results. We believe some of the observed thermodynamics change by plotting the van’t Hoff equations at different temperatures, may be due to factors such as the different equilibrium parameters used for identifying equilibrium states and deriving equilibrium pressure values during PCT measurements, or different kinetic performance of the samples which may affect the equilibrium pressures when using the same equilibrium state parameters. Nevertheless, these developed materials are still not yet satisfactory for commercialization. Nanostructured materials appear to have some aggregation problems during cycling, and this needs to be solved. It has been reported that Mg-based materials obtained by nano-confinement, which can block grain growth and retain the nanostructure of materials, may improve the kinetics of absorption/desorption, and also may somehow change thermodynamic destabilization [8,91,92]. Emerging ideas, such as new composite materials, metastable alloys, geometrical storage materials, as well as nano-confinement technology, may give us some new directions to develop Mg-based hydrogen storage materials in the future. On the other hand, Mg-based nanomaterials combined with solid oxide fuel cells for stationary energy storage have been proposed, and this makes it possible to apply Mg-based hydrogen storage materials to large-scale storage with a working temperature for storage materials higher than 250 °C [43]. In this case, thermodynamic change is no longer a problem.