Enhancing Hydrogen Storage Properties of MgH2 by Transition Metals and Carbon Materials: A Brief Review

Magnesium hydride (MgH2) has attracted intense attention worldwide as solid state hydrogen storage materials due to its advantages of high hydrogen capacity, good reversibility, and low cost. However, high thermodynamic stability and slow kinetics of MgH2 has limited its practical application. We reviewed the recent development in improving the sorption kinetics of MgH2 and discovered that transition metals and their alloys have been extensively researched to enhance the de/hydrogenation performance of MgH2. In addition, to maintain the cycling property during the de/hydrogenation process, carbon materials (graphene, carbon nanotubes, and other materials) have been proved to possess excellent effect. In this work, we introduce various categories of transition metals and their alloys to MgH2, focusing on their catalytic effect on improving the hydrogen de/absorption performance of MgH2. Besides, carbon materials together with transition metals and their alloys are also summarized in this study, which show better hydrogen storage performance. Finally, the existing problems and challenges of MgH2 as practical hydrogen storage materials are analyzed and possible solutions are also proposed.


INTRODUCTION
Since the industrial revolution, human society is developing rapidly with continuous improvement in technology and rising demand for energy consumption (Pudukudy et al., 2014;He et al., 2016). Unfortunately, fossil fuels, which play dominate role in promoting the development of world, are not renewable and going to be running out in near future. Besides, the severe environmental problems caused by the excessive exploitation and use of fossil fuels, such as the greenhouse effect, ozone layer depletion, acid rains, and pollution, are damaging and threating the ecological balance of the earth. To mitigate the degradation of the earth, various measures have been taken by scientists to explore renewable and clean alternatives to fossil fuels.
Hydrogen, with its safe, high energy density (142 KJ/kg), environment friendliness, convenient and renewability, is proved to be the most promising sustainable and clean energy to replace fossil energy (Cao et al., 2016;Wan et al., 2020). As an energy carrier, hydrogen is abundant on earth and can be produced from any primary energy fuel: coal, oil, nuclear, natural gas, all sorts of renewable energies, and from grid electricity. Hydrogen also has a huge calorific value of energy, which is three times higher than that of petrol (43 MJ/kg) after combustion. Moreover, the dominating combustion product of hydrogen is clean and non-toxic water. Due to above advantages, hydrogen has received extensive attention from researchers worldwide and has made a rapid progress in recent decades (Winter, 2009;Sadhasivam et al., 2017;Peter, 2018). In order to realize the practical application of hydrogen energy, three challenges need to be conquered presciently, which are hydrogen preparation, storage and application. Among which, hydrogen storage has become the bottleneck technology in the wide spread of hydrogen energy (Felderhoff et al., 2007;Yang J. et al., 2010;Pukazhselvan et al., 2012;Kim et al., 2018).
Among different solid-state hydrogen storage materials, magnesium hydride (MgH 2 ) has been much discussed and holds tremendous hope for storing hydrogen (Bogdanović and Spliethoff, 1990;Norberg et al., 2011;. As the sixth abundant metal element in the earth's crust, magnesium is widely distributed in nature. More importantly, MgH 2 has a high gravimetric capacity of 7.6 wt% (volumetric capacity of 110 g/L) and excellent reversibility. However, the practical application of MgH 2 has been hindered by the high desorption temperature and poor hydrogen absorption/desorption kinetics caused by high thermal stability ( H = 76 kJ/mol) and kinetic barrier (Ea =160 kJ/mol) (Webb, 2015;Peng et al., 2017;Zhou et al., 2019a;Jain et al., 2020).

TRANSITION METALS AND THEIR ALLOYS
On the whole, doping transition metals and their alloys into magnesium hydride has been considered as one of the most feasible methods to accelerate the sorption kinetics of MgH 2 . During recent years, numerous transition metals and their alloys have been developed and researched. In this paper, these catalysts are reviewed and classified, presented as monometallic catalysts, binary alloys, ternary and multicomponent alloys and the composites of alloys and carbon materials. Their catalytic effects on hydrogen storage properties of MgH 2 were summarized in Table 1.

Nickel (Ni)
Monometallic catalysts, especially transition metals (Ershova et al., 2008;Gasan et al., 2012;El-Eskandarany et al., 2016;Tanniru et al., 2020), have shown great catalytic impact on improving the hydrogen storage properties of MgH 2 . Among all the transition metals studied in recent years, nickel has been the mostly adopted catalysts for MgH 2 . As early as 2005, Hanada et al. (2005) mixed purchased MgH 2 powder with metal Ni by ball milling to get the MgH 2 +nano-Ni composite. Through the thermal desorption mass spectra (TDMS), they found that the hydrogen desorption peak of the Ni doped composite decreased to 260 • C, which was much lower than that of pure MgH 2 (370 • C). Although the superior catalytic effect of Ni nanoparticles was confirmed, other factors such as particle size and catalyst amount were also widely researched lately. Xie et al. (2009) studied the hydrogen storage kinetics of the MgH 2 nanoparticles doped with different concentration of Ni nanoparticles. The DSC curves depicted that the MgH 2 +10 wt% nano-Ni composite could desorb 6.1 wt% hydrogen within 10 min at 250 • C. The desorption rate of MgH 2 +nano-Ni composite increased obviously with the increasing amount of catalyst. However, the activation energy of desorption could not be further lowered when the amount of Ni exceeded a certain value by using Kissinger equation. It was concluded that the catalytic effect of Ni could further be increased by reducing the particle size of catalyst and maintaining the hydrogen storage capacity at the same time . Yang W. N. et al. (2010) investigated the size effect of Ni particles on the hydrogen desorption of MgH 2 . The results showed that the MgH 2 mixed with only 2 at% of fine Ni particles rapidly desorbed hydrogen from 200 • C and almost 6.5 wt% hydrogen could be released when the temperature rose to 340 • C. Nevertheless, DSC curves showed that the peak temperature of the MgH 2 + 2Ni 90 mixture is around 280 • C, which was only about10 • C lower than those of the MgH 2 + 2Ni 200 and the MgH 2 + 2Ni 100 composites. They finally concluded that the site density of the catalyst over the MgH 2 particles but not the particle size was the key factor to improve the hydrogen adsorption kinetics of MgH 2 after comparing with other references.

Titanium (Ti)
In comparison with nickel, titanium has also been demonstrated as a good catalyst for MgH 2 . In 1999, Liang et al. (1999a) studied the catalytic mechanism of titanium mainly through XRD results. During the synthesis process of the MgH 2 +5at%Ti composite via mechanical milling, a very stable TiH 2 phase was formed by reaction of MgH 2 with Ti. Interestingly, TiH 2 could be obtained after desorption, which suggests that no decomposition of TiH 2 phase occurred under the mild desorption condition of 300 • C. The desorption curves showed that MgH 2 −5at%Ti composites could desorb hydrogen completely within 1,000 s at 250 • C while the ball-milled MgH 2 released no hydrogen under the same conditions. A lot of work has been done by researchers to further study the de/hydrogenation kinetics and microstructure of MgH 2 -Ti composites. Wang et al. (2015) also prepared the MgH 2 -Ti composite by ball milling, and found that the initial dehydrogenation temperature of the composite to be 257 • C, which was 51 • C lower than that of pure MgH 2 . The hydrogen capacity could reach 6.18 wt% at the same time.
Compared to the sluggish desorption kinetics of pure MgH 2 , the Ea for the MgH 2 -Ti sample was 103.9 kJ mol −1 , about 35.8% lower than that of pure MgH 2 (161.3 kJ mol −1 ). By analyzing the mechanism it was depicted that elemental Ti reacted with MgH 2 and formed active TiH 1.971 during ball-milling, which acted as active species throughout the desorption process. Shao et al. (2011) mainly researched nanostructured Ti-catalyzed MgH 2 for hydrogen storage. Through EDX measurements, it was discovered that Ti covered the MgH 2 surface. The DTA curve showed that decomposition of Ti-catalyzed MgH 2 sample started from below 300 • C, which was about 130 • C lower than that of the commercial MgH 2 sample. Hydrogen absorption kinetics of MgH 2 -Ti sample were also investigated and the dehydrogenated sample could absorb 6 mass% hydrogen in <1 h at 257 • C, while the commercial MgH 2 needed an absorption time of about 3 h to reach a capacity of 6 mass% even at 350 • C.

Iron (Fe)
As the most common metal element in life, Fe has been widely concerned and studied in recent years. Bassetti et al. (2005) mixed different concentration values of Fe with MgH 2 by ball milling to explore its catalytic effect. The Mg 2 FeH 6 phase could be detected when the ball to powder ratio rose to 20:1. They also concluded that the optimum catalyst concentration was around 10 wt% and lower values seemed to be insufficient to avoid the presence of poorly catalyzed regions. DSC curves revealed that about 5 wt% of hydrogen could be released in 600 s at 300 • C. Besides the desorption property, the cycling performance and the nanosized Fe were further studied. Montone et al. (2012) explored the cycling properties of MgH 2 -Fe nanocomposite in 47 cycles at 300 • C. Cycling results demonstrated that the maximum storage capacity and the rate of sorption remained stable after the first 10 cycles. They also discovered that the major effect of cycling on particle morphology was the progressive extraction of Mg from the MgO shell surrounding the powder particles. In our previous study (Zhang et al., 2019b), Fe nanosheets obtained by wet-chemical ball milling were introduced in MgH 2 for the first time. The MgH 2 +5 wt% nano-Fe composite began to release and absorb hydrogen at 182.1 and 75 • C, respectively. Moreover, the dehydrogenated composite could absorb 6 wt% H 2 within just 10 min at 200 • C. Cycling experiment depicted that the hydrogen capacity of MgH 2 +5 wt% nano-Fe composite could still maintain at about 5 wt% after 50 cycles. During the first cycle, it could been seen from microstructure pictures that the Fe nanosheets on the surface of MgH 2 turned to be numerous ultrafine nanoparticles, which could provide more active sites in the following cycling. Further calculation results revealed that the addition of Fe could greatly weaken the Mg-H interaction strength, facilitating the dehydrogenation of MgH 2 (Figure 1).

Other Monometallic Catalysts
Besides Ni, Ti, and Fe, other monometallic catalysts have also been developed to improve the hydrogen storage  (Figure 2). All these composites could release hydrogen at a low temperature of 225 • C, which was much lower than that of prepared MgH 2 . Gasan et al. (2012) studied the impacts of 5 wt% of additives (V and Nb) on the hydrogen desorption temperature of MgH 2 . XRD results demonstrated that the addition of V powders had a significant impact on the transformation of Mg into the MgO for the amount of MgO in MgH 2 -V system was higher than other systems, relevant samples were studied by others and this phenomenon needs to be further researched. Also, SEM images verified that the mean particle size of composites was decreased by mechanical milling to micro scale. DSC tests showed that the addition of 5 wt% additives reduced hydrogen desorption temperatures of MgH 2 by about 40-50 • C. Liang et al. (1999b) presented the hydrogen storage properties of MgH 2 +V composite prepared by ball milling. The MgH 2 +5 at% V composite could desorb hydrogen at 200 • C and reabsorb hydrogen rapidly even at room temperature, the activation energy of hydrogen desorption was decreased to 62 kJ mol −1 .

Zr-Based Binary Alloys
Recently, many papers reported an interesting strategy for improving hydrogen storage performance of MgH 2 by using intermetallic compounds of transition metals as catalyst. The corresponding results showed that the absorption/desorption properties of modified MgH 2 systems should be evidently enhanced by the synergetic effects of both phases. Following above idea, we successfully synthesized ZrMn 2 nanoparticles ranging from 100 to 300 nm by a facile wet chemical method to explore their catalytic effect on enhancing the hydrogen storage properties of MgH 2 . The MgH 2 +10 wt% nano-ZrMn 2 composite began to release hydrogen from 181.9 • C and 6.7 wt% hydrogen could be discharged in 5 min at 300 • C. Isothermal absorption measurements represented that the dehydrogenated sample could absorb 5.3 wt% hydrogen within 10 min at 100 • C under 3 MPa hydrogen pressure. Based on the XRD, TEM and calculation results, the ZrMn 2 nanoparticles were distributed well on the surface of MgH 2 and helped weaken Mg-H bond strength, which resulted in the enhanced hydrogen storage performance of MgH 2 (Figure 3). With the above research experience, we also studied the catalytic effect of ZrCo nanosheets (Zhang et al., 2019c). The modified MgH 2 composite could desorb almost 6.3 wt% H 2 within 5 min at 300 • C and took up 4.4 wt% H 2 under 3 Mpa hydrogen pressure in 10 min even at 120 • C for doping 10 wt% ZrCo nanosheets. The homogenously distributed ZrCo served as "hydrogen pump" to promote the de/composition of H 2 molecules, which was the key to decrease the de/hydrogenation temperature of MgH 2 . Cycling performance (10 cycles) revealed that there was an apparent reduction in hydrogen storage capacity. In comparison to our studies, the MgH 2 -ZrNi 5 nanocomposite system possessed more excellent hydrogen absorption/desorption performance without serious degradation after 600 complete cycles (El-Eskandarany et al., 2017). The prepared MgH 2 -10 wt% ZrNi 5 sample required 1 and 10 min to absorb and discharge 5.3 wt% H 2 at 275 • C, respectively. Based on the FE-SEM and XRD results, nano-scaled ZrNi 5 grains were uniformly distributed into the MgH 2 matrix and ZrNi 5 particles could create a network of micro channel that extending into the body of metal hydride to provide hydrogen diffusers during the de/hydrogenation processes.

Ti-Based Binary Alloys
El nearly spherical shape with particle size ranging between 100 and 320 nm. DSC analysis presented that this composite could absorb/desorb 5.1 wt% hydrogen within 100 and 400 s at 225 • C, respectively. For cycling performance, no obvious degradation in storage capacity was found during the long cyclic-life-time (600 h). FE-SEM micrographs highlighted that TiMn 2 particles could prevent a serious growth of Mg/MgH 2 grains, which led to reduced hydrogen uptake/release kinetics. Neto et al. (2017) doped TiFe compound into MgH 2 and concluded that a fine dispersion could be achieved by increasing milling time or using higher energy ball mill. To attain the best hydrogen kinetics, the sample prepared in the planetary mill for 36 h was the optimum selection and the MgH 2 + 40 wt%TiFe sample milled for 36 h could release about 3 wt% hydrogen within the first hour. Other Ti-based binary alloys such as TiAl and TiNb have also shown excellent enhancement for catalyzing MgH 2 (Zhou et al., 2013). Thermo gravimetric analysis and pressure composition temperature (PCT) isothermal tests showed that the MgH 2 -TiAl and MgH 2 -TiNb samples began to desorb hydrogen below 250 • C, and the addition of TiAl or TiNb could make the dehydrogenated sample take up hydrogen even at room temperature. In terms of the dehydrogenation reaction, the Mg-H bond would be destabilized by doping with TM elements, which could be confirmed in the theoretical model. The TiAl catalyst illustrated the most effective impact on reducing the activation energy to 65 kJ/mol −1 among the Ti-based catalysts. Nevertheless, the Ti intermetallic catalyst did not change the thermodynamic equilibrium pressure of MgH 2 .

La-Based Binary Alloys
Rare earth elements, especially lanthanides, are considered as one of the most promising catalysts because of their high activity. Many researchers have used La-based binary alloys as catalyst dopants in MgH 2 to explore the resulting catalytic effects. In 2000, Liang et al. (2000) studied MgH 2 -LaNi 5 composite and found that Mg, LaH 3 , Mg 2 NiH 4 were formed during the milling process. The first desorption of mechanically milled MgH 2 -30 wt% LaNi 5 could release about 4 wt% hydrogen within 150 s at 300 • C and the dehydrogenated sample could absorb 3.7 wt% hydrogen in 2,000 s at room temperature. In order to understand the cycling properties of the composite, SEM images manifested that no apparent change in particle size was observed after 20 absorption and desorption cycles. To systematically study the MgH 2 -LaNi 5 composite, MgH 2 with different amount of LaNi 5 were synthesized by Fu et al. (2008). XRD patterns illustrated that the extended milling time of 40 h caused an additional decrease of peak intensity for the materials containing 5 and 15 wt% LaNi 5 , which could be ascribed to the brittleness of LaNi 5 . Further kinetics results showed that the influence of LaNi 5 on absorption kinetics was more pronounced at lower temperatures. Additionally, other La-based binary alloys could also improve the hydrogen storage properties of MgH 2 .

Ti-Based Ternary Alloys
Based on the great improvement of the binary alloy, ternary alloys which replace part of the binary alloy with another transition metal have also been concerned in recent years (Hu et al., 2004;Shahi et al., 2013;El-Eskandarany, 2016;Lu et al., 2018). For instance, Zhou et al. (2013) doped TiVMn alloy into MgH 2 to study its hydrogen storage performance. On the contrast with other Ti-based binary alloys catalysts, the dehydrogenation kinetics of the MgH 2 -TiVMn composite was much better (Figure 4). Moreover, PCT curves also depicted that the addition of TiVMn exhibited the best catalytic effect, which could release more hydrogen under the same condition. The dehydrogenation activation energy was calculated to be 85.2 kJ/mol·H 2 by OFW model, which was much lower than that of pure MgH 2 . Khodaparast and Rajabi (2015) prepared the MgH 2 +5 at% Ti-Mn-Cr sample by milling the Ti-Mn-Cr alloy produced by melt spinning method with pure MgH 2 . When Ti-Mn-Cr was doped into MgH 2 , the dehydrogenation temperature of the composite reduced from 399 to 345 • C, much lower than that of prepared MgH 2 under the same conditions. Mahmoudi et al. (2011) prepared MgH 2 -5 at% TiCr 1.2 Fe 0.6 composites at the nanoscale. In comparison to pure MgH 2 , the initial desorption temperature of the MgH 2 -5 at% TiCr 1.2 Fe 0.6 sample decreased to 241 • C and almost 5 wt% hydrogen could be obtained at 300 • C. Further, XRD and TEM studies stated that the interface of the TiCr 1.2 Fe 0.6 alloy with magnesium also acted as active sites for nucleation of the hydride phase, thereby decreasing the nucleation barrier and enhancing the dehydrogenation property.

Non-Ti-Based Ternary Alloys
Besides Ti-based alloys, other alloys formed by various single transition metals also illustrated their remarkable effect on improving hydrogen storage performance of MgH 2 . Agarwal et al. (2009) studied the catalytic effect of ZrCrNi alloy on hydrogenation properties of MgH 2 . The ZrCrNi alloy was prepared by melting the three pure metals in an arc furnace and then milling with MgH 2 for 5 h in a SPEX 8,000 mixer-miller to receive the MgH 2 -10 wt% ZrCrNi sample. They also performed 20 cycles of de/hydrogenation to explore the stabilization of kinetics and the achievement of hydrogen capacity. In aspect of de/hydrogenation performance, the composite could quickly desorb and absorb about 90% of its maximum hydrogen capacity within 7 min at 300 • C after the 20th cycle. XRD and SEM patterns demonstrated that there was no other phases formed during milling and cycling. Also, the alloy was homogeneously dispersed in the MgH 2 /Mg matrix. To improve hydrogen desorption properties of MgH 2 , Motavalli and Rajabi (2014) prepared the MgH 2 -5 at% Ni 3 FeMn sample by mechanical milling, where the Ni 3 FeMn catalyst was in two states: as-cast and melt-spun ribbon. DTA curves clarified that 30 h mechanically alloyed catalysts in both states could significantly decrease the desorption temperature. MgH 2 -Ni 3 FeMn melt-spun composite could discharge H 2 in lower temperature due to the ability to improve particle size refinement of MgH 2 and a more pronounced homogeneous distribution of the alloyed elements. The MgH 2 -5 at% Ni 3 FeMn melt-spun ribbon composite could release 3.39 wt% hydrogen within 1,000 s at 340 • C. Zhou et al. (2019b) doped purchased VTiCr into MgH 2 and demonstrated a reversible capacity of 4 wt% H 2 between 150 and 350 • C for the MgH 2 -VTiCr composite. Besides, the dehydrogenated sample could absorb hydrogen in a low hydrogen pressure of 0.04-0.4 bar. The VTiCr catalyst was uniformly dispersed on the surface of MgH 2 matrix. VTiCr was deemed as a strong catalyst that provided not only excellent catalytic effect but also offer effective cyclic stability in the sense that the reaction kinetics still remained stable after the 10 cycles.

Multicomponent Alloys
As mentioned above, single transition metals and their binary and ternary alloys have shown great catalytic effects on MgH 2 -based systems (Haghparast and Rajabi, 2015;El-Eskandarany et al., 2018). Further, studies about multicomponent alloys were also stated, Yu et al. (2010) found that the addition of Ti 0.4 Cr 0.15 Mn 0.15 V 0.3 alloy could apparently improve the de/absorption properties of MgH 2 . The MgH 2 -Ti 0.4 Cr 0.15 Mn 0.15 V 0.3 composite began to release hydrogen at 255 • C and reached its peak at 294 • C, which was much lower than that of unanalyzed MgH 2 . Besides, the dehydrogenated sample could absorb 3.1 wt% H 2 in 500 min even at 29 • C. The cycling results manifested that the dehydrogenation rate increased slowly in the first 20 cycles and then remained stable after 20 cycles. SEM and TEM techniques showed that the Ti 0.4 Cr 0.15 Mn 0.15 V 0.3 alloy hydride nanoparticles were well-distributed on the surface of MgH 2 . Meena et al. (2018) found that MgH 2 could desorb H 2 even at 180 • C with the addition of 50 wt% NiMn 9.3 Al 4.0 Co 14.1 Fe 3.6 alloy. Compared to as-milled MgH 2 sample, the Ea of this composite was lower by about 46.56 kJ/mol. Haghparast and Rajabi (2015) studied the de/hydrogenation kinetics of MgH 2 -TiCrMn 0.4 Fe 0.4 V 0.2 composite and found that the dehydrogenation temperature of modified MgH 2 decreased to 378 • C, which was lower than that of as-received MgH 2 (421 • C). V 45 Zr 20 Ni 20 Cu 10 Al 3 Pd 2 powders were doped into MgH 2 by El-Eskandarany et al. (2018) and found that the desorption temperature of MgH 2 -10 wt% V 45 Zr 20 Ni 20 Cu 10 Al 3 Pd 2 powders was 308.9 • C, which was 116 • C lower than that of pure MgH 2 .This prepared nanocomposite possessed superior de/hydrogenation kinetics at relatively low temperature (180 • C), absorbing and desorbing 5.5 wt% H 2 within 200 s.

Alloys and Graphene
All above catalytic materials have shown remarkable improvement on the hydrogen storage performance of MgH 2 , however, stable cycling performance is still the bottleneck for realizing the practical application of MgH 2 . Carbon materials such as graphene and carbon nanotubes, were widely researched and lots of studies have proven that carbon materials are helpful in preserving stable cycling properties (Xia et al., 2015). Hudson et al. (2016) reported that graphene together with Fe nanoclusters could enhance the hydrogen sorption kinetics of MgH 2 . From the TPD and DSC curves, the peak temperature of desorption for MgH 2 +5wt% Fe@G was 281.7 • C, lower than that of exhibited peak ball-milled MgH 2 . In addition, the activation energy of MgH 2 +5wt% Fe@G composite was reduced to 119.1 kJ/mol (24% lower than that of ball-milled MgH 2 ). Furthermore, TEM confirmed that the grain size of MgH 2 increased only 15 nm after 6 cycles, displaying a low grain growth rate during cycling due to the addition of graphene. Density functional theory calculations demonstrated that the defect in graphene and the presence of iron clusters at the defect site of graphene played important role in desorbing hydrogen. Ji et al. (2020) prepared FeNi nanoparticles dispersed on reduced graphene oxide nanosheets (FeNi/rGO) and then found that this catalyst played a vital role in improving the hydrogen storage performance of MgH 2 .
The MgH 2 -FeNi/rGO sample started to release hydrogen at 230 • C and the dehydrogenated sample could absorb 5.4 wt% within 20 min at 125 • C. Further investigations proved that FeNi nanoparticles were well distributed on the MgH 2 surface in the nanoscale range ( Figure 5). More importantly, cycling tests exhibited that 6.9 wt% hydrogen capacity was maintained even after 50 cycles. Singh et al. (2017) investigated the catalytic effect of FeCoNi@GS on hydrogen sorption of MgH 2 . The onset desorption temperature of this sample was around 255, 25 • C lower than that of FeCoNi catalyzed MgH 2 . The FeCoNi@GS remained stable even after 24 cycles with FeCoNi particles uniformly distributed on the surface of GS.

Alloys and Carbon Nanotubes
Carbon nanotubes (CNTs), were widely researched in every field for its small particle size and great microstructure (Luo et al., 2007;Gao et al., 2019;Liu M. et al., 2019 composite started to release hydrogen at 230, 45 • C lower than that of MgH 2 -FeCl 3 . Moreover, the MgH 2 -FeCl 3 /CNT sample could desorb more hydrogen than that of MgH 2 -FeCl 3 under the same isothermal condition. SEM images confirmed that the CNT was not destroyed after the short milling process and indicated that the sample with CNT appeared to have less agglomeration. It was believed that the presence of the unique structure of the CNTs played a critical role in the improvement of hydrogen storage properties in the MgH 2 -FeCl 3 /CNTs composite. In our recent investigation , CNTs combined with Zr 0.4 Ti 0.6 Co nanosheets was adopt to strengthen the hydrogen storage properties of MgH 2 . With the addition of Zr 0.4 Ti 0.6 Co sheets, the sorption kinetics were evidently improved while hydrogen capacity was slowly decreasing. Meanwhile, the MgH 2 -Zr 0.4 Ti 0.6 Co/CNTs exhibited no reduction in cycling performance even after 10 cycles after doping CNTs (Figure 6). Deeper structure investigation revealed that particle size of MgH 2 -Zr 0.4 Ti 0.6 Co/CNTs was almost unchanged, contributing to the stable cycling performance.

Alloys and Other Carbon Materials
Apart from carbon materials mentioned above, other carbon materials also have distinguished effect on ameliorating the de/hydrogenation kinetics of MgH 2 (Xia et al., 2018). An et al. (2014) reported that the one-dimensional porous Ni@C nanorods modified MgH 2 performed an excellent hydrogen storage properties. The addition of Ni@C decreased the onset temperature of MgH 2 to 175 • C. Cycling results illustrated no significant loss of hydrogen storage capacity and the MgH 2 -5 wt% Ni@C composite had favorable cycle stability. Chen et al. (2018) reported the mesoporous carbon CMK-3 performed well in enhancing the hydrogen storage properties of MgH 2 . The onset desorption temperature of MgH 2 -10 wt% Ni/CMK-3 was 170 • C lower than that of pure MgH 2 (above 350 • C) and the sample could discharge 6 wt% H 2 even at 295 • C. The more fascinating fact was that 3.9 wt% hydrogen was absorbed at 55 • C for MgH 2 -Ni/CMK-3 composite. The sample maintained nearly 90.8% of the original de/hydrogenation capacity when cycled for 10 times, indicating that MgH 2 -Ni/CMK-3 had a good cycle stability. Wang et al. (2018) combined graphene oxide-based porous carbon (GC) and TiCl 3 to improve the reversible kinetics of MgH 2 . The MgH 2 /GC-TiCl 3 composite could reversibly deliver about 7.6 wt% hydrogen at 300 • C within 9 min and the average dehydrogenation rate was several times faster than that of the single catalytic MgH 2 system. Concerning cycling property, the capacity of the MgH 2 /GC-TiCl 3 sample was also stable with slower kinetics, owing to the nanoconfinement effect of the ball-milled GC. In a word, graphite and carbon with their derivatives could mainly improve the cycling performance, which results in remarkably enhanced the hydrogen storage properties of MgH 2 .

CONCLUSIONS AND PERSPECTIVES
To realize the practical application of hydrogen energy, numerous effects still need to be carried out in the coming future. For hydrogen storage materials, magnesium hydride is generally believed as a promising material due to its natural abundance, excellent reversibility, light weight and efficient cost. Among the methods investigated, the transition metals have demonstrated excellent catalytic effect on improving the hydrogen storage properties of MgH 2 . Further studies about alloys based on transition metals are demonstrated to be more effective than the single metal counterparts. In our recent studies, Zr-based alloys and Fe-based alloys were successfully prepared and confirmed to striking improve the de/hydrogenation performance of MgH 2 . Although the transition metals and their alloys have shown superior enhancement on the de/absorption performance of MgH 2 , maintaining good cyclic performance is still a challenge for MgH 2 -based systems. A large number of experiments indicated that carbon materials show excellent effect on maintaining good hydrogen absorption and desorption performance. Our group also demonstrated that carbon nanotubes and reduced graphene oxide together with transition metal alloys can improve the de/hydrogenation kinetics of MgH 2 while maintain stable cycling properties at the same time. From above review on literature and our own work, we propose the following strategy to further enhance the hydrogen storage properties of MgH 2 : (1) regulate the components of transition metal alloys to its best catalytic effect, (2) make the particle size of the alloys as small as possible, (3) combine alloys and carbon materials to synthetically improve the hydrogen storage properties of MgH 2 . In summary, nanoscale transition metal alloys together with carbon materials would be a promising catalyst for realizing the practical application of MgH 2 .

AUTHOR CONTRIBUTIONS
LZ, SS, and JX contributed conception and design of the study. ZS wrote the first draft of the manuscript. XL, FN, and NY wrote sections of the manuscript. All authors contributed to manuscript revision, read and approved the submitted version.