Metallic glassy Zr70Ni20Pd10 powders for improving the hydrogenation/dehydrogenation behavior of MgH2

Because of its low density, storage of hydrogen in the gaseous and liquids states possess technical and economic challenges. One practical solution for utilizing hydrogen in vehicles with proton-exchange fuel cells membranes is storing hydrogen in metal hydrides. Magnesium hydride (MgH2) remains the best hydrogen storage material due to its high hydrogen capacity and low cost of production. Due to its high activation energy and poor hydrogen sorption/desorption kinetics at moderate temperatures, the pure form of MgH2 is usually mechanically treated by high-energy ball mills and catalyzed with different types of catalysts. These steps are necessary for destabilizing MgH2 to enhance its kinetics behaviors. In the present work, we used a small mole fractions (5 wt.%) of metallic glassy of Zr70Ni20Pd10 powders as a new enhancement agent to improve its hydrogenation/dehydrogenation behaviors of MgH2. This short-range ordered material led to lower the decomposition temperature of MgH2 and its activation energy by about 121 °C and 51 kJ/mol, respectively. Complete hydrogenation/dehydrogenation processes were successfully achieved to charge/discharge about 6 wt.%H2 at 100 °C/200 °C within 1.18 min/3.8 min, respectively. In addition, this new nanocomposite system shows high performance of achieving continuous 100 hydrogen charging/discharging cycles without degradation.

for example refs 16 and 18). Severe plastic deformation, using cold rolling 20 , equal channel angular pressing, and high-pressure torsion multiple forging or cyclic, channel die compression techniques (a summary of these techniques are summarized in ref. 21) are the other options used for refining the coarse MgH 2 grains to obtain micro-scaled structured grains. It has been pointed out by Dufour and Huot 22 that as-cold rolled Mg-Pd alloy shows fast kinetics when compared with as-ball milled samples. They attributed this improvement seen in the bulk cold-rolled samples to its resistance to the air contamination 22 .
The second strategy used for enhancing the kinetics behavior of MgH 2 depends on doping the metal hydride powders with selected catalysts/nanocatalysts to form ultrafine nanocomposite powders with advanced hydrogenation/dehydrogenation properties. Huge number of nanocatlysts such as pure metals 23 , intermetallic compounds 24,25 , metal oxides 26 , metal/metal oxide nanocomposite 27 , metal crbides 28 , metal chlorides 29 . Recently, two review articles have been published discussing the role of doping MgH 2 particles with catalytic agents on improving the kinetic behavior and cycle-life time 30,31 .
Almost all of these reported systems when compared with pure MgH 2 nanocrystalline powders, they significant usage merits, indexed by lower hydrogen sorption/desorption temperatures and faster hydrogenation/dehydrogenation kinetics,. Moreover, some of these nanocomposite systems such as MgH 2 /7 Mn 3.6 Ti 2. 4 25 , MgH 2 /5Ni/5Nb 2 O 5 27 , MgH 2 /5TiC 28 , and MgH 2 /10 big-cube Zr 2 Ni 32 powders have shown high performance cyclability for achieving complete 1000, 600, 696 and 2546 complete cycles at hydrogenation/dehydrogenation temperature in the range between 250-275 °C without serious hydrogen storage degradations. In addition, these interesting systems have shown fast absorption/desorption kinetics at relatively low temperature (250-275 °C) ranging between 41 to 120 s/121 to 613 s, respectively with reasonable hydrogen storage capacity ranging between 5.1 to 5.3 wt.%. It is believed that high energy ball milling MgH 2 with the nanocatalyst powder of carbides, oxides, and intermetallic abrasive powders lead to fast grain refining of the MgH 2 upon releasing the crystalline stored energy, leading to refine the MgH 2 grains along their grain boundaries and to produce fine grains. Such fine grains with their short-distance grain boundaries always facilitate short diffusion path, leading to fast diffusion of the hydrogen atoms 33,34 .
A very attractive and unique approach used to improve the kinetics behaviors of MgH 2 without using expensive metallic catalysts has been proposed by Jeon et al. in 2011 35 . In their novel process, Mg nanocrystals (NCs) were embedded into polymer (PMMA) matrix in an inert gas atmosphere to form air-stable Mg NCs/PMMA nanocomposite material. Encapsulation the nanosized Mg (~5 nm in diameter) in a polymer with selective gas permeability, protecting the NCs from O 2 and H 2 O. This contamination-free system enjoys high density of hydrogen (6 wt% of Mg) and rapid kinetics (loading in <30 min at 200 °C).
In the last strategy employed for enhancing the hydrogenation/dehydrogenation properties of MgH 2 is to melt pure bulk Mg with selected alloying elements such as Ni, Pd, and Nd to obtain less stable binary, ternary or multicomponent Mg-based alloy systems with lower heat of formation (Δ H for ). In many cases, the synthesized alloying Mg-based alloy systems do not show attractive properties when compared with MgH 2 . For example, alloying Mg with Ni to form binary Mg 2 Ni lowers the Δ H for of the metal hydride phase (Mg 2 NiH 6 ) to − 64.5 kJmol −1 instead of − 74.5 kJmol −1 for pure MgH 2 . However, the system shows a dramatic degredation in its hydrogen storage capacity (3.5 wt.%) with no significant decreasing in the decomposition temperature 36 . In 2001, Yamada et al. 37 proposed Mg-rich (90 at.%) systems of Mg-Pd, Mg-Nd and Mg-Pd-Nd to replace the traditional Mg 2 Ni alloy. They reported that Mg-based system containing Pd show at 300 °C PCTs curves with three plateau-like regions and hydrogen absorbency of 5 wt.%. They pointed out that the hydrogenation/dehydrogenation of Mg-Nd system was influenced by a catalytic effect of the formed NdH 2.5 and NdH 3 phases that assisted hydriding and dehydriding of the Mg matrix. For Mg-Pd system, they investigated that disproportional reaction of Mg 6 Pd to form Mg 5 Pd 2 and MgH 2 retarded the overall reaction kinetics. In spite of their efforts dedicated to introduce a new Mg-based alloy system with advanced kinetics behavior, their proposed Mg 89 Pd 7 Nd 4 alloy required 50 min to absorb about 4 wt.% H 2 at 300 °C. In addition, the dehydrogenation process of this system required 150 min at the same operating temperature to be completely achieved 37 42 . They claimed that addition of Ti and in-situ synthesized MgF 2 improved the kinetics and the introduction of In as well as Al imparted enhanced thermodynamics to the Mg 85 In 5 Al 5 Ti 5 system 42 . In their study, the dehydrogenation enthalpy change and activation energy of this new system were 65.2 kJ/(mol H 2 ) and 125.2 kJ/ mol, respectively. The hydrogen storage capacity this system measured at 30 bar/340 °C was 5.5 wt.%.
Apart from such a long list of traditional catalysts used for improving the MgH 2 , here we show, for the first time the effect of employing a small mole fractions (5 wt.%) of Zr 70 Ni 20 Pd 10 metallic glassy powders on destabilizing the MgH 2 and improving its kinetics. Theas-synthesized nanocomposite MgH 2 /5 wt.% Zr 70 Ni 20 Pd 10 powders, which shows high density of hydrogen, possess advanced hydrogenation/dehydrogenation processes taking place at low temperature and very short of time.

Results
Structure and morphology. Amorphous Zr 70 Ni 30 Pd 10 powders. Figure 1(a) shows the X-ray diffraction (XRD) pattern of material powders at the initial stage of ball milling (0 h). The powders composed of polycrystalline mixture of hcp-Zr (70 at.%), fcc-Ni (20 at.%), and fcc-Pd (10 at.%) with large particle size, as indicated by the sharp Bragg-peaks displayed in Fig. 1(a). When the mixed powders subjected to a continuous ball milling for 25 h at an argon gas atmosphere, using a high-energy cryo-mill operated under a flow of liquid nitrogen, all of these elemental sharp Bragg peaks were disappeared and replaced by a halo-diffuse pattern of an amorphous phase, as shown in Fig. 1(b). We should emphases that milling the powders under liquid nitrogen is a necessary step used to overcome the expected agglomeration of the powders. Moreover, cryo-milling can also overcome the stacking of the powders on the milling tools that always lead to the formation of a heterogeneous glassy alloy. The cryo-ball milling process is shown in a video clip in the Supplementary Material-1-Vidio.
The image of a high-resolution transmission electron microscope (HRTEM) of the powders obtained after ball milling for 25 h (Fig. 1(f)) shows a maze-like structure without indication to the precipitation of unprocessed crystals of the starting materials ( Fig. 1(g)), suggesting the formation of a single amorphous phase. The nanobeam diffraction pattern (NBDP) taken from the center of the image shows a clear halo-diffraction pattern, typically to The as-prepared amorphous alloy powders possess excellent morphological properties, indicated by a narrow particle size distribution laid in the range between 78-103 nm with a spherical-like morphology, as displayed in the image of field-emission scanning electron microscope (FE-SEM) displayed in Fig. 2(a). The surface area measurements, using Brunauer-Emmett Teller (BET) approach, indicated that these ultrafine amorphous nanopowder particles enjoy a high surface area of about 26.45 m 2 /g. MgH 2 powders. MgH 2 powders were synthesized by reactive ball milling (RBM) of pure Mg powders, using a high-energy ball mill operated under 50 bar of a hydrogen gas pressure. Supplementary Material-3(a-c)) shows the photos of the vial and milling media (a) and (b) the set up performed to charge the vial with 50 bar of hydrogen gas. The photo in (c) presents the complete set up of gas-temperature-monitoring system (GST) prior to start the RBM experiment for preparing of MgH 2 powders, using a high-energy planetary ball used in the present study. The XRD pattern of starting powders shows sharp Bragg-diffraction peaks related to hcp-Mg, as shown in Fig. 1(c). After 6 h of RBM time, a new set of Bragg-peaks was obtained, whereas all of the Bragg-lines corresponding to hcp-Mg are disappeared, indicating the formation of a new phase ( Fig. 1(d)). The analysis of these new Bragg-peaks indicated the formation of polycrystalline mixture of γ -MgH 2 and β -MgH 2 phases with orthorhombic and tetragonal structures, respectively. After this stage of ball milling, the as-synthesized MgH 2 powders consisted of large-grains, ranging between ~60-220 nm in diameter, with irregular morphological characterizations, as presented shown in the dark field image (DFI) presented in Fig. 1(i). The Miller-indexed selected area diffraction pattern (SADP) shown in Fig. 1(j) shows a continuous Debye-Scherrer rings related to a tetragonal phase (β -MgH 2 ) overlapped with a metastable orthorhombic γ -MgH 2 phase. At this TEM resolution, unprocessed Mg crystals could not be detected, indicating the completion of the RBM process for formation of MgH 2 powders. The FE-SEM image for a typical aggregated MgH 2 powder particle is shown in Fig. 2(b). The agglomerated powder, which was coated with a thick MgO resulted during SEM sample preparations ( Fig. 2(b)), had an ellipsoidal-like morphology with a particle size of about 1.5 μ m in diameter. The surface area of as-synthesized MgH 2 powders after this ball milling time (6 h) was 7.30 m 2 /g.  as-synthesized metal hydrides were doped with the synthetic amorphous alloy powders and high-energy ball milled for different RBM time under 50 bar of hydrogen gas pressure. A the early stage of ball milling (6 h) the composite powders consisted of heterogeneous structure of micro-scaled powders of MgH 2 , and nano-scaled amorphous powders, as elucidated in Fig. 2(c). It is worth mentioning that at this stage of milling, the powder's chemical composition widely varied from particle to particle and within an individual particle. The field-emission bright field image (FE-BFI) of the MgH 2 -rich particle obtained after 6 h of milling is shown in Fig. 3(a). These selected aggregated particles were mainly composed of MgH 2 grains (~35 nm in diameter), as suggested by the SADP (Fig. 3(c)) taken from region I shown in Fig. 3(a)). The composite powders obtained after 20 h showed a different feature, as elucidated in Fig. 3(b). The featureless fine structure of the amorphous powders, which have a very high surface area, became a metallic host matrix wherein the MgH 2 were embedded to form a typical composite powders, as shown in Fig. 3(b). The SADP corresponding to zone II ( Fig. 3(b) shows an overlap between the two MgH 2 phases (tetragonal (β ) and orthorhombic (γ )) and the halo-amorphous pattern displayed in 3(d). The SADP displayed in 3(d) shows a continuous Debye-ring diffraction pattern of β -and γ -MgH 2 phases with the absence of sharp spots. This implies the formation of a nanocrystalline MgH 2 grains embedded into the amorphous matrix. However, the "guest phase" of MgH 2 grains that became somewhat finer (as a result of further milling) in sizes (ranging between 8-32 nm), were still heterogeneously distributed into the metallic matrix, as indicated by the dark regions presented in Fig. 3(b).
In order to ensure the homogeneous distribution of MgH 2 grains into the amorphous matrix, the composite powders were furtherly ball milled for 50 h. The XRD pattern of nanocrystalline-MgH 2 /5 wt.% amorphous Zr 70 Ni 30 Pd 10 composite powders obtained after the end of processing time (50 h) is presented in Fig. 1(e). The primary-and secondary haloes become very broad without indication to the existence of medium-or long-range ordered structure, as shown in Fig. 1(e). Moreover, the Bragg-diffraction peaks related to MgH 2 (γ and β phases) show significant broadening ( Fig. 1(e)), indicating the effect of RBM time on grain refining and formation of nanocrystallites. Those Bragg peaks shown in Fig. 1(e), which are related to fcc-MgO phase, came from the oxidation of the powder surfaces during preparation the XRD sample outside the helium-atmosphere glove box. The nanocomposite powders of this end-product comprised of ultrafine particles, laid in the range between 50 nm up 420 nm in diameter ( Fig. 2(d)) and having a surface area of 16.80 m 2 /g.
The HRTEM image taken near the edge of MgH 2 /5 wt.% amorphous Zr 70 Ni 30 Pd 10 composite particle obtained after 50 h of RBM time is shown in Fig. 3(e) together with the corresponding NBDP ( Fig. 3(f)). Overall, the composite powders obtained after this stage of milling consisted of continuous amorphous matrix (maze-like morphology shown in Fig. 3(e)) hosting ultrafine nanoclusters (~4 nm in diameter) of order-structure (related  (Table 1). STEM-BFI and -DFI images for the end-product nanocomposite powders (50 h of ball milling) are displated in (g,h), respectively. The X-ray-elemental mapping for Mg, Zr, Ni and Pd corresponding to (g) are presented in (i-l), respectively.
to MgH 2 ), as indexed by the green arrows labels shown in Fig. 3(e). It is worth mentioning that the MgH 2 grains were distributed into the metal amorphous matrix in a segregation fashion with the absence of agglomerates or aggregated grains, as displayed in Fig. 3(e). The NBDP (Fig. 3(f)) corresponding to the white circular lable shown in Fig. 3(e) shows halo-diffraction pattern related to amorphous Zr 70 Ni 30 Pd 10 coexisted with spot-like pattern came from nanocrystalline γ -and β -MgH 2 phases oriented in different axial directions.
In order to investigate the elemental chemical composition of the nanocomposite powders obtained after 50 h of ball milling, the powders were subjected to intensive local EDS analysis, using a beam focus of 5 nm. The red-circular zones labeled by Roman numerals (I to X) in Fig. 3(e) refer to the EDS selected zones for performing the compositional analysis. Table 1 shows the detailed of these analyses in weight percent (wt.%). From these analyses, one can say that the composition of the nanocomposite powders varying from 72.6 to 99.2 wt.% Mg. This corresponding to 0.8 to 27.4 wt.% Zr 70 Ni 30 Pd 10 . More information about the distribution of MgH 2 in the amorphous matrix composite EDS mapping approach was performed. The STEM-BFI and the corresponding STEM-DFI of a selected nanocomposite MgH 2 /5 wt.% -amorphous Zr 70 Ni 30 Pd 10 powder particle are shown in Fig. 3(g,h), respectively. The powder has a nearly spherical morphology (Fig. 3(g)) and containing fine lenses of less than 10 nm in diameter homogeneously distributed into the fine structured powder particle, as shown in Fig. 3(h). The fine spherical lenses were corresponding to Mg (Fig. 3(i)), whereas the elemental composition of the particle related to the elemental Zr, Ni and Pd, as indicated in the elemental mapping presented in Fig. 3(j-l), respectively.  Fig. 4(a). The first event is an endothermic reaction refers to the glass transition temperature (T g ) related to the form of glassy phase, whereas the second event takes place through a sharp exothermic peak related to the crystallization of the metallic glassy phase, as shown in Fig. 4(a). The differences between the crystallization temperature (T x ) and T g refers to the supercooled liquid region (Δ T x ; T x -T g = 61 K).
The DSC scan of as-synthesized MgH 2 powders obtained after 6 h of RBM (before mixing with the metallic glassy powders) reveals shoulder-like endothermic peaks, as shown in Fig. 4(b). The metastable phase of γ -MgH 2 tends to decompose at lower temperature (Tγ -dec = 709 K), when compared with the decomposition temperature of β -MgH 2 (Tβ -dec = 733 K), as displayed in Fig. 4(b). Both Tγ -dec and Tβ -dec shifted to the low temperature side upon ball milling with metallic glassy Zr 70 Ni 30 Pd 10 powders for 10 h, and recorded to be 691 K and 709 K, respectively (Fig. 4(c)). Significant decreasing in the decomposition temperature (616 K) for MgH 2 is realized of the sample mixed with 5 wt.% Zr 70 Ni 30 Pd 10 powders and milled for 50 h, as displayed in Fig. 4(d). This implies the outstanding effect of metallic glassy and the RBM time for destabilizing the MgH 2 phase.
The activation energy for MgH 2 powders obtained after 6 h of RBM time and nanocomposite MgH 2 /5 wt.% amorphous Zr 70 Ni 30 Pd 10 powders obtained after 50 h of the ball milling time were investigated by DSC analysis conducted with different heating rates (k) of 5, 10, 20, 30, and 40 °C/min and shown in Fig. 4(c,d), respectively. All the scans for both material powders revealed single endothermic events related to the decomposition of MgH 2 . While the peak height increased proportionally with the increasing the heating rates, the peak temperatures (T p ) were significantly shifted to the higher temperature side upon increasing the heating rates from 5 °C/min to 40 °C/ min, as shown in Fig. 4(c,d).
In the present work, the activation energy (E a ) of dehydrogenation of pure MgH 2 and nanocomposite MgH 2 /5 wt.% amorphous Zr 70 Ni 30 Pd 10 powders was calculated according to the Arrhenius equation: where k is a temperature-dependent reaction rate constant, R is the gas constant, and T is the absolute temperature. The E a values were determined by measuring the decomposition the T p corresponded to the different heating Hydrogenation/dehydrogenation behaviors. Pressure-Composition-Temperature. Figure 5(a,b) shows the pressure-composition-temperature (PCT) curves investigated at 200 °C 350 °C for nanocomposite MgH 2 /5 wt.% amorphous Zr 70 Ni 30 Pd 10 powders obtained after 40 h of ball milling ( Fig. 5(a)) and MgH 2 powders obtained after 6 h of RBM time ( Fig. 5(b)), respectively. For successful hydride formation (complete absorption) of MgH 2 sample, high temperature (350 °C) and pressure (P abs , in the range between 200 mbar to 40 bar) were required to absorb about 4.7 wt.% H 2 , as shown in Fig. 5(b). This indicates that the sample requires the application of higher temperature to get its theoretical hydrogen storage capacity (7.6 wt.% H 2 ). Moreover, the sample showed poor dehydrogenation behavior, indexed by exhibiting a clear pressure hysteresis with significant large gabs (~17 bar) between the pressure needed for absorption (hydride formation), P abs and the pressure required for hydride decomposition, P des , as presented in Fig. 5(b). In addition, pure MgH 2 powders obtained after 6 h of RBM time failed to desorb its stored hydrogen content completely even after 12 h of the desorption time ( Fig. 5(b)). In contrast to the MgH 2 sample, the nanocomposite powders showed an excellent PCT hydrogenation/dehydrogenation curves, indexed by the complete hydrogen sorption/desorption event at lower temperature (200 °C) and pressure (50 mbar to 10 bar), as shown in Fig. 5(a). At such relative low temperature and pressure, the nanocomposite sample reached to a higher value of hydrogen storage capacity (5.8 wt.%) when compared with pure MgH 2 (4.7 wt.% H 2 ). Moreover, the plateau region for the nanocomposite sample was very flat with negligible slope with the absence of multistep hydrogenation/dehydrogenation, as shown in Fig. 5(a). In addition, the sample succeed to achieve complete desorption through an almost flat PCT curve with minimal difference value (~72 mbar) between P abs and P des , as shown in Fig. 5(a). Moreover, our system possessed rather low-pressure plateau (~1.5 bar/200 °C) closed to that value reported for Mg 3 PrNi 0.1 (1.6 bar/297 °C) 38 Mg 80 Ce 18 Ni 2 (3 bar/311 °C) 39 , but it shows lower pressure plateau when compared with mechanically alloyed Mg-5 at.%In (8 bar/350 °C) 40 , Mg-5at.%Sn revealed (~5 bar/323 °C) 41 , and Mg 85 In 5 Al 5 Ti 5 (7 bar/380 °C) 42 . The hydrogen storage capacity of MgH 2 /5 wt.% amorphous Zr 70 Ni 30 Pd 10 system possessed high hydrogen storage capacity (5.8 wt.%), being closed to that one (~5.5 wt.%) reported for both Mg(In) solid solution 40 and Mg-Sn nanocomposite 41 . This storage capacity of our system is far above other MgH 2 -based hydrides, such as Mg 3 PrNi 0.1 (2.5 wt.%) 38   (3.5 wt.%) 39 , however, it is a bit below than the storage capacity reported for MgH 2 catalysed by Ti-based nanocoating (6.05 wt.%) 43 .
Kinetics of absorption. One major problem restricting the potential applications of MgH 2 compound in real fuel cell and energy storage applications is its very slow hydrogen uptake/release kinetics that required the application of a high temperature (above 400 °C) to be enhanced. For example, at 175 °C, the as-synthesized MgH 2 powders obtained after 6 h of RBM (without additives of metallic glassy powders) require 10 min to absorb about 2 wt.% H 2 , as shown in Fig. 6(a). When MgH 2 mixed with the metallic glassy powders and milled for 10 h, a remarkable improvement on the absorption kinetics is achieved, indexed by an increase in the hydrogen amount absorbed within 10 min to 3.89 wt.%, as shown in Fig. 6(a). Increasing the RBM time led to enhance the absorption kinetics, indexed by the absorbed amount hydrogen recorded for the samples obtained after 20 h (5.26 wt.%), and 30 h(5.68 wt.%). As elucidated in Fig. 6(a). The sample obtained after ball milling with the metallic glassy powders for 40 h shows outstanding hydrogenation kinetics, indicated by the very short time (~2 min) required to absorb ~5 wt.%H 2 , as shown in Fig. 6(a). This sample reaches to its maximum storage capacity (~5.8 wt.%H 2 ) after only 9 min, as presented in Fig. 6(a). Figure 6(b) shows the temperature effect, ranging between 100-200 °C on the absorption kinetics of composite MgH 2 /5 wt.% metallic glassy Zr 70 Ni 30 Pd 10 powders obtained after 50 h of RBM time. The relation between the absorbed hydrogen during the first minute of the experiment is shown inset of Fig. 6(b). The fabricated composite powders reveal excellent hydrogenation characteristics, indexed by their high capability of absorbing hydrogen (~4.6 wt.%) within a short time (1 min) at low temperature (100-125 °C), as displayed inset of Fig. 6(b). They reached together to their saturation values of 5.6 wt.% H 2 after 6.5 min, as presented in Fig. 6(b). Increasing the applied temperature to 150 °C improves the absorption kinetics, as suggested by the higher hydrogen absorbed (~5.5 wt.%) in 1 min (inset Fig. 6(b)). This sample reached to its saturated capacity value of 5.8 wt.% H 2 after about 1.18 min, as displayed in Fig. 6(b). Significant improvement is achieved at 200 °C when the sample reached to a hydrogen capacity of 5.8 wt.% within 0.75 min (inset Fig. 6(b)), and does not show any degradation upon increasing the absorption time to 10 min, as shown in Fig. 6(b). Figure 6(c) shows the dependence of desorption kinetics for MgH 2 measured at 175 °C on the metallic glassy additive and RBM time. Originally, pure MgH 2 powders obtained after 6 h of RBM time has a poor desorption kinetics, indexed by the low value of hydrogen released(~0.35 wt/%) after 50 min of desorption time (Fig. 6(c)). When MgH 2 mixed with the metallic glassy the powders and milled for 20 h, a better desorption kinetics can be attained, as indicated by a higher H 2 desorbed value (~1 wt.%) obtained after 50 min (Fig. 6(c)). Remarkable improving on the desorbed kinetics is realized for the composite sample obtained after 20 h (4.88 wt.% H 2 /50 min) and 30 h (5.6 wt.% H 2 /26 min), as shown in Fig. 6(c). The composite sample powders obtained after 40 h of RBM time, shows excellent dehydrogenation characteristics, indexed by the very short time (4.6 min) required to release about 5.7 wt.%H 2 , as displayed in Fig. 6(c).

Kinetics of desorption.
The temperature effect on the desorption kinetics for the sample obtained after 50 h of RBM time is shown in Fig. 6(d). The composite powders examined at 125 °C desorbed about 2 wt.% H 2 within 10 min, as shown In Fig. 6(d). Increasing the applied temperature to 150 °C leads to enhance the dehydrogenation kinetic behavior, indicated by the shorter time (6.5 min) to release about 5.5 wt.% H 2 . This value is saturated at 5.6 wt.% H 2 after 9.35 min, as shown in Fig. 6(d). Outstanding enhancement for the desorption kinetics is attended of the sample measured at 200 °C, showing a very short time (3.8 min) needed to release about 5.7 wt.%H 2 (Fig. 6(d)).  Figure 7(a) shows the hydrogen absorbed/desorbed cycles achieved continuously for 100 times at a temperature of 200 °C. It should be emphasized that surface treatment of the powders led to improve its capability of hydrogen absorption, reached to 6.15 wt.%, as shown in Fig. 7(a). No remarkable degradation in the hydrogen storage capacity could be detected even after 100 cycles, as shown in Fig. 7(a). The kinetics of hydrogenation/dehydrogenation remaining constant with nearly constant absorption and desorption values of 6.15 wt.%.
The BFI micrograph of the sample taken after the completion of 100 sorption/desorption cycles conducted at 200 °C is shown in Fig. 7(b). The powder consisting of featureless morphology where numerous ultrafine dark-contrast spherical grains were embedded into the light-gray fine matrix (Fig. 7(b)). It should be notifying that the Mg powders were segregated and fairly distributed into the metallic glassy matrix. These Mg grains maintained their original sizes without severe grain growth even after performing the sample for 100 cycles. The Fast Fourier Transform (FFT) image of the zone indexed in Fig. 7(a) shows halo diffraction rings of an amorphous phase coexisted with diffracted spots corresponding to hcp-Mg crystal oriented to [001], as shown in Fig. 7(c). The HRTEM image of selected Mg grains are shown with a higher magnification in Fig. 7(d). Clear Moiré-like fringes with different interplanar spacing (d) of fine grains (~5 nm in diameter) are shown in Fig. 7(d). These displayed grains having d spacing values of 0.259 nm, 0.188 nm, 0.276 nm and 0.246 nm that match well with the (002), (102), (100) and (101) lattice indexes of hcp-Mg, respectively. The absence of severe grain growth in the Mg grains after completion of cyclic test can be attributed to the surrounded hard metallic glassy matrix that played the role of a grain growth inhibitor. Fig. 7(e) shows a high-resolution STEM-BFI displaying the host metallic glassy matrix and spherical metallic Mg grains. The matrix maintained its fine structure without any evidences of crystallizations or formation of medium-range ordered phase during the hydrogenation/dehydrogenation cycles. The thermal stability of our prepared metallic glassy phase used in the present study and the absence of phase transformations during the hydrogenation/dehydrogenation cycles leads to a sustainable hydrogen storage capacity with constant hydrogen uptake/releasing kinetics, as shown in Fig. 7(a).

Discussions
In contrast to the traditional catalyst families (e.g. elemental metals, metal alloys, compounds) used to improve the kinetic behaviors of MgH 2 powders, the present study proposes a metastable metallic glassy Zr 70 Ni 20 Pd 10 alloy nanopowder as a superior enhancer leading to modify the hydrogenation/dehydrogenation properties of MgH 2 . This prepared metallic glassy phase is homogeneous in structure and uniform in composition beyond the atomic level ( Fig. 1(f) Supplementary Fig. S2). Moreover, our fabricated metallic glassy nanopowder possesses a high thermal stability, indicated by the large Δ T x (61 K) and high T x (936 K) values, as shown in Fig. 4(a). Thus, Zr 70 Ni 20 Pd 10 metallic glassy powders did not undergo to any structural changes upon processing MgH 2 /5 wt.% Zr 70 Ni 20 Pd 10 nanocomposite powders at relatively lower temperature of less than 800 K (527 °C).
During the early stage (< 10 h) of ball milling a mixture of MgH 2 and 5wt.% Zr 70 Ni 20 Pd 10 metallic glassy powders, the ball-powder-ball collusions ( Supplementary Fig. S4) led to break down the large MgH 2 powder particles (Supplementary Fig. S5(a,b)) and assisted adherence of the glassy fine powders onto the surface of MgH 2 particles ( Supplementary Fig. S5(c)). During this stage of milling, the hard metallic glassy nanopowders penetrated the oxide layer formed on the surfaces of MgH 2 powders to create micro-holes on their surfaces ( Supplementary Fig. S5(d,e)).
Increasing the ball milling time (10 to 20 h) led to "migrate" a large volume fraction of metallic glassy powders through the cavities and micro-channels created in the body of MgH 2 particles and located their grains and at the  Fig. S5(f)). Such glassy powders, which acted as "micro-grain splitter" led to break up the large MgH 2 grains along their weak grain boundary zones ( Supplementary Fig. S5(g)) and forming finer grains ( Supplementary Fig. S5(h)). Further ball milling time (20-40 h) resulting an increase the volume fractions of MgH 2 fine grains and the number of grain boundaries ( Supplementary Fig. S5(h)). Since the hydrogen diffusion is much faster along the grain boundaries when compared with inside grains, the hydrogenation/ dehydrogenation kinetic behaviors of MgH 2 were gradually improved ( Fig. 6(a,c)) with increasing the number of "liberated" grains ( Supplementary Fig. S5(i)).
During the last stage of ball milling (40-50 h), the role of ball-powder-ball collusions on achieving further refining of MgH 2 powder particles was almost absent since the size of the powders became ultrafine (less than 1 μ m) 8 . Accordingly, the final refining process was attained by the nanosized metallic glassy powders, which were homogeneously distributed within MgH 2 to form typical homogeneous nanocomposite powders ( Supplementary  Fig. S5(j-l)). The end-product of the nanocomposite powders obtained after 50 h ( Supplementary Fig. 5(m)) possessed excellent morphological characteristics, indexed by a homogeneous dispersions of equal nano-sized MgH 2 grains segregated into the metallic glassy powders, without agglomeration. This nanocomposite structured powders facilitated fast hydrogen diffusion, as suggested by Jeon for MgH 2 NCs/PMMA nanocomposite system 35 . During the particle/grain, refining "long-voyage" extended to 50 h of high-energy ball milling, MgH 2 powders were subjected to continuous mechanical deformations created by the impact and shear forces generated from milling, and micro-abrasive milling media of the balls and metallic glassy powders, respectively. These forces were translated into sever plastic deformation, and lattice imperfections created into the MgH 2 powders ( Supplementary Fig. S5(n)).
Based on the results of the present study, high surface area (26.45 m 2 /g) of metallic glassy Zr 70 Ni 20 Pd 10 nano-spherical powders played a superior role for improving both of hydrogenation/dehydrogenation kinetics of MgH 2 powders. The PCT curve of the formed nanocomposite powder in the present work, containing 10wt.% Pd did not multi-steps process suggested by Yamada et al. for Mg-Pd, Mg-Nd, and Mg-Pd-Nd systems 37 .
When our results are compared with the In the present work, the formation of intermediate phases of Mg 2 NiH 6 and Mg 6 Pd upon hydrogenation/dehydrogenation process in temperatures ranging between 125-250 °C could not be detected. This can suggest the absent of the common catalytic role of the metallic glassy powders when mixed with MgH 2 powders. Thus, it can be concluded that the kinetics of MgH 2 powders were enhanced by a drastic grain refining and the formation of large volume fractions of segregated grains that facilitated fast hydrogen diffusion along their numerous number of grain boundaries. This grain refinement was achieved upon using abrasive metallic glassy powders.
In summary, the hydrogenation/dehydrogenation kinetics of MgH 2 powders prepared by reactive ball milling technique was greatly improved upon mechanically-induced doping with a small mole fraction (5 wt.%) of Zr 70 Ni 20 Pd 10 metallic glassy powders. Adding such a metallic metastable phase led to destabilize the MgH 2 and improved its kinetics. The as-synthesized nanocomposite MgH 2 /5 wt.% Zr 70 Ni 20 Pd 10 powders possessed high density of hydrogen and exhibited irreversible hydrogenation/dehydrogenation process taking place at low pressure and temperature with a very short time.

Methods
Preparation of the metallic glassy powders. Pure Zr (100 μ m, 99% purity), Ni (10 μ m, 99.9% purity) and Pd (10 μ m, 99.5% purity) metal powders provided by Alfa Aesar -USA, were used as starting alloying elements. The powders were balanced and manually mixed inside a helium (He) gas atmosphere (99.99%)-glove box (UNILAB Pro Glove Box Workstation, mBRAUN, Germany) to give the starting charge (1 g) with an average nominal composition of Zr 70 Ni 20 Pd 10 . The powders were then sealed together with five Cr-stainless steel balls (10 mm in diameter) into a FeCr steel vial (20 ml in volume, Retsch, Germany), using a ball-to-powder weight ratio as 50:1. In order to avoid the agglomeration of the metallic powders during the ball milling process, a cryo-mill system provided by Retsch was used. In this experiment, the vial containing the balls and powders was mounted on the cryo-milling system where the process taking place under continuous cooling, using liquid nitrogen flow. The liquid nitrogen circulated through the system and was continually replenished from an auto fill system in the exact amount, which is required to keep the temperature at − 196 °C. This milling process was carried out with a frequency of 25 Hz for 25 h. The end-product obtained after 25 h was discharged in the He-atmosphere glove box.
Preparation of the metal hydride powders. Elemental Mg metal powders (~80 μ m, 99.8% provided by Alfa Aesar -USA), and hydrogen gas (99.999%) were used as starting materials. An amount of 5 g Mg was balanced inside a He gas atmosphere (99.99%) -glove box (UNILAB Pro Glove Box Workstation, mBRAUN, Germany). The powders were then sealed together with twenty five hardened steel balls into a hardened steel vial (150 ml in volume), using a gas-temperature-monitoring system (GST; supplied by evico magnetic, Germany). The ball-to-powder weight ratio was 40:1. The vial was then evacuated to the level of 10 −3 bar before introducing H 2 gas to fill the vial with a pressure of 50 bar. The reactive ball milling (RBM) process was carried out at room temperature, using a high energy ball mill (Planetary Mono Mill PULVERISETTE 6, Fritsch, Germany). After 6 h of RBM time, the powders were discharged from the vial inside the glove box and sealed in Pyrex vails. The as-synthesized MgH 2 powders were then mixed in the glove with the desired weight percentage (5%) of Zr 70 Ni 20 Pd 10 amorphous powders, using an agate mortar and pestle. The mixed powders were charged together with twenty five hardened steel balls into the vial and sealed under He gas atmosphere. The vial was then filled with 50 bar of hydrogen gas atmosphere and mounted on the high-energy ball mill. The milling process was interrupted after selected time (10,20,30,40, and 50 h) and the powders obtained after an individual milling time were completely discharged into 8 Pyrex vails for different analysis. The contamination contents of Fe and Cr of the powders obtained after 50 h of ball milling were Based on the EDS analyses of at least 50-60 powder particles, Scientific RepoRts | 6:26936 | DOI: 10.1038/srep26936 the average contents of Fe and Cr introduced to the powders upon using tempered Cr-steel milling tools were 0.86 and 0.19 wt.%, respectively. XRD and TEM. The crystal structure of all samples was investigated by XRD with CuKα radiation, using 9kW Intelligent X-ray diffraction system, provided by SmartLab-Rigaku, Japan. The local structure of the synthesized material powders was studied by 200 kV-field emission high resolution transmission electron microscopy/scanning transmission electron microscopy (HRTEM/STEM) supplied by JEOL-2100F, Japan, and equipped with Energy-dispersive X-ray spectroscopy (EDS) supplied by Oxford Instruments, UK. Thermal stability. Differential scanning calorimetry (DSC)/differential thermal analysis (DTA) unit, provided by Setaram -France with a heating rate of 20 °C/min was employed to investigate the glass transition temperature, and thermal stability indexed by the crystallization temperature and enthalpy change of crystallization for the metallic glassy powders. Shimadzu Thermal Analysis System/TA-60WS, using differential scanning calorimeter (DSC) was employed to investigate the decomposition temperatures of MgH 2 -based composite powders with a heating rate of 20 °C/min. The activation energy for pure MgH 2 and MgH 2 /5wt.% Zr 70 Ni 20 Pd 10 was investigated, using Arrhenius approach with different heating rates (5,10,20,30,40 °C/min).
The hydrogenation/dehydrogenation behaviors. The hydrogen absorption/desorption kinetics were investigated via Sievert's method, using PCTPro-2000, provided by Setaram Instrumentation, France, under hydrogen gas pressure in the range between 200 mbar to 10 bar. The samples were examined at different temperatures of 100, 125, 150, 175, and 200 °C.