Synergistic Effect of New ZrNi₅/Nb₂O₅ Catalytic Agent on Storage Behavior of Nanocrystalline MgH₂ Powders

Abstract

Due to its availability and high storage capacity, Mg is an ideal material in hydrogen storage applications. In practice, doping Mg/MgH₂ with catalyst(s) is necessary in enhancing the de/rehydrogenation kinetics and minimizing both of decomposition temperature and its related apparent activation energy. The present study proposed a new heterogeneous catalytic agent that consisted of intermetallic compound (ZrNi₅)/metal oxide (Nb₂O₅) binary system for using with different concentrations (5−30 wt%) to improve MgH₂. Doping MgH₂ powders with low concentration (5, 7, 10 wt%) of this new catalytic system led to superior absorption/desorption kinetics, being indexed by the short time that is required to absorb/desorb 4.2−5.6 wt% H₂ within 200 s to 300 s. Increasing the doping dose to 15–30 wt% led to better kinetic effect but a significant decrease in the hydrogen storage capacity was seen. The dependent of apparent activation energy and decomposition temperature of MgH₂ on the concentration of ZrNi₅/Nb₂O₅ has been investigated. They tended to be linearly decreased with increasing the catalyst concentrations. The results elucidated the crucial role of catalytic additives on the disintegration of MgH₂ into ultrafine powders (196 nm to 364 nm diameter). The formation of such nanoparticles enhance the hydrogen diffusion and shorten the time that is required for the hydrogenation/dehydrogenation process. Moreover, this refractory catalytic system acted as a grain growth inhibitor, in which Mg/MgH₂ powders maintained their submicron level during the cycle-life-test that was extended to 100 h at 200 °C.

1. Introduction

Magnesium hydride (MgH₂) is a well known solid-sate hydrogen storage material with promising fuel-cell and mobile applications [1,2,3,4,5,6]. This low-cost energy material shows high volumetric (110 g/L) and gravimetric (7.6 wt%) storage capacity [7]. MgH₂ has a high enthalpy change of formation (−75 kJ/mol) and it tends to be decomposed at high temperature (above 300 °C), with very slow hydrogenation/dehydrogenation kinetics, in spite of these desired properties [8,9,10].

Likewise, catalysts are essential substances that are used for doping Mg/MgH₂ materials to modify their behaviors upon deducing the apparent activation energy between reactants (Mg + 2H) and product (MgH₂) in many chemical and oil refining applications. Accordingly, the potential energy barrier gap is decreased. Within the last three decades, MgH₂ was targeted by a great number of catalysts that were employed to enhance its behavior. Pure metals, metal alloys and intermetallic compounds, metallic glasses, metal oxides, carbides, and fluorides are some examples of catalysts that were devoted to improve MgH₂ [5,7,8,9,10].

Among this long list of catalysts, Nb₂O₅ powders have received great attention due to its capability of enhancing the hydrogenation/dehydrogenation kinetics of MgH₂ and the reduction the activation energy of decomposition [11,12,13]. It has been reported that doping MgH₂ with 5 wt% Nb₂O₅ led to improving the absorption/desorption kinetics to uptake and releasing 6 wt% H₂ at 300 °C within 78 s and 300 s, respectively [14]. Hanada et al. reported that the catalyzation of MgH₂ with 14 wt% Nb₂O₅ led to minimizing the hydrogenation temperature (100 ^oC) that is required to absorb 4 wt% H₂ under 5 bar within 30 min [14]. Decomposition of MgH₂/14 wt% Nb₂O₅ required 7 min to desorb 4 wt% H₂ under 0.1 bar/240 °C [14].

Apart from the catalytic metal oxide group, Dehouche et al. published very interesting work related to doping MgH₂ powders with different intermetallic compounds of the Zr_100-xNi_x system [15]. They reported that doping MgH₂ powders with eutectoid Zr₄₇Ni₅₃ resulted in the fastest desorption time and the highest initial desorption rate at 250 °C. In addition, they pointed out that the composition of the dispersed Zr_100-xNi_x catalysts had strong influence on the amount of accumulated hydrogen and the desorption rate of Mg-nanocomposite [15].

The present work aims to study the catalytic effect of x-wt% (y-ZrNi₅/y-Nb₂O₅), x; 5, 7, 10; 15; and, 30, abrasive nanopowders on the behaviors of MgH₂, as indexed by (i) thermal stability, (ii) apparent activation energy of decomposition, (iii) storage capacity, (iv) de/rehydrogenation kinetics, and (v) cycle-life-time. Moreover, the role of abrasive additives on particle refinement and their influences on hydrogen uptake/release processes were investigated.

2. Results and Discussion

2.1. Crystal Structure

The X-ray diffraction (XRD) patterns for the starting materials of MgH₂ and ZrNi₅ powders that were obtained after 200 h and 10 h of ball milling are displayed in Figure 1a,b, respectively. MgH₂ powders consisted of β-MgH₂ (PDF file #: 00-012-0697), as indicated by Bragg peaks that appeared at 2θ of 27.97°, 35.78°, 39.73°, and 54.73° (Figure 1a). The XRD pattern of ZrNi₅ milled for 10 h (Figure 1b) displays Miller-indexed Bragg peaks that matched well with the reported ZrNi₅-Friauf-Laves phase (PDF#: 00-037-0924). The broadening that was seen in the diffracted lines was attributed to the grain refinement of the powders during the high-energy ball milling process. Figure 1c presents the XRD pattern of MgH₂ that was doped with a mixture of (2.5 ZrNi₅ + 2.5 Nb₂O₅) powders and then ball-milled for 50 h of ball milling. The Bragg peaks corresponding to MgH₂ and the catalytic agent (ZrNi₅ and Nb₂O₅) became broad, suggesting the formation of nanocomposite powder particles. Neither the reacted phase, such as ZrH₂, NbH, nor foreign phases could be obtained after this stage of milling, as elucidated in Figure 1c.

The high-resolution transmission electron microscope (HRTEM) image of the MgH₂-5 wt% (2.5 ZrNi₅/2.5 Nb₂O₅) nanocomposite powder that was obtained after the 50 h of milling is displayed in Figure 1d. In general, the sample that was composited of ultrafine nanocrystals of MgH₂, ZrNi₅, and Nb₂O₃ revealed lattice fringes seen from different selected zones in Figure 1d, revealing the existence of single crystalline phases in the sample. The lattice fringes were regularly separated with d spacing of 0.189, 0.234, and 0.205 nm, which agrees well with the (222), (220), and (311) lattice index of ZrNi₅, respectively. The Moiré-like fringes that are displayed in Figure 1d with a d spacing of 0.197, 0.176, and 0.241 nm match well with the (002), (380), and (181) lattice index of Nb₂O₅ crystals, respectively. The d spacing of 3 selected β-MgH₂ crystals oriented to (101), (200), and (110) were 0.274, 0.226, and 0.307 nm, respectively, being in excellent agreement with the reported data (PDF file # 00-012-0697). The d spacing of 0.195 nm (Figure 1d) was related to γ-MgH₂ (211) crystal. All of the lattice fringes that were taken from different 25 zones for three individual samples strongly manifested the existence of three phases of pure MgH₂, ZrNi₅, and Nb₂O₅ crystals. The existence of other reacted phases, such as Mg₂Ni, Zr_100-xNi_x, Nb, NbO, NbH, and Nb crystals in this sample that were obtained after 50 h of milling could not be detected under the current resolution of the FE-HRTEM (point resolution is 0.19 nm with beam diameter of 2.5 nm) used in the present study.

2.2. EDS Analysis

Local analysis was performed on the powders of the final product while using energy dispersive x-ray spectroscopy (EDS)/HRTEM technique. Figure 2a displays dark field image (DFI) of MgH₂/5 wt% (2.5 ZrNi₅/2.5 Nb₂O₅) nanocomposite powder that was obtained after milling for 50 h. The nanocomposite aggregate was virtually divided into 14 zones (indexed in Figure 2a by circular symbols) for conducting EDS analysis, while using a beam of 2 nm in diameter.

Table 1. Elemental Analysis Corresponding to the Indexed Zones Shown in Figure 2.

Zone	Content (wt%)
	Mg	ZrNi5	Nb2O5
1	94.80	2.53	2.49
2	95.02	2.52	2.46
3	94.95	2.54	2.51
4	94.99	2.53	2.48
5	94.99	2.49	2.52
6	94.02	2.51	2.47
7	95.01	2.48	2.51
8	95.02	2.49	2.49
9	95.01	2.47	2.52
10	94.95	2.52	2.53
11	95.04	2.48	2.48
12	95.01	2.52	2.47
13	95.0	2.47	2.53
14	94.94	2.56	2.50

lists the corresponding EDS results of elemental analysis for the selected zones shown in Figure 2. Based on the EDS analysis, it is concluded that the composition of the end-product for the nanocomposite powders that were obtained after this stage of milling is very homogeneous and does not seriously fluctuate in composition beyond the nanolevel, as shown in Table 1.

2.3. Morphology

Figure 3a displays the relationship between the average particle size of nanocomposite powders that were obtained after 50 h of milling and the concentration (wt%) of (x ZrNi₅/x Nb₂O₅) catalytic agent, in the range between 0 wt% to 30 wt%. MgH₂ powders, which were obtained after 200 h of reactive ball milling (RBM) (before doping) composited of irregular cluster particles that were agglomerated together due to Van der Wales electrostatic attraction to form large aggregates, as shown in Figure 3b. The powder particles that were distributed in a wide range had an average particle size of 1480 nm, as presented in Figure 3a.

When the powders were catalyzed with 5 wt%(2.5 ZrNi₅/2.5 Nb₂O₅) and milled for 50 h, finer spherical particles were obtained (Figure 3c). In addition, they were distributed in a narrow range to provide an average particle size of 364 nm, as shown in Figure 3a. We should emphasize that the ZrNi₅/Nb₂O₅ hard powders played a micro-milling agent role during the milling process. This nanosized lubricant abrasive agent penetrated the rather soft MgH₂ powder matrix and occupied their catalytic sites onto/into the domain powders (Figure 2). The further increasing of catalytic dose to 7, 10, and 15 wt% led to monotonical decreasing on the sizes of nanocomposite powders to 318, 306, and 284 nm, as presented in Figure 3a.

The effect of hard ZrNi₅/Nb₂O₅ catalysts became pronounced upon increasing the dose to 20 wt% (10 ZrNi₅/10 Nb₂O₅) and 30 wt% (15 ZrNi₅/15 Nb₂O₅). The drastic decreasing of the particle sizes that reached to 203 nm and 196 nm, respectively, implied this (Figure 3a). In addition, the nanocomposite powders had almost spherical-like morphology with minimal agglomeration, as displayed in Figure 3d,e. Using such high molar fraction of catalytic agents led to an increase in the abrasive shear forces that were applied to MgH₂ powders when the effect of applied forces that were generated by the milling balls (10 mm in diameter) on such ultrafine powders (~ 200 nm) were deduced. Based on the TEM and scanning electron microscope (SEM) results of the present study, it is concluded that increasing the weight percentage of (x ZrNi₅/x Nb₂O₅) against MgH₂ powders had a beneficial effect on the fabrication of ultrafine nanocomposite powders with uniform spherical shapes (Figure 3c−e) and uniform distribution (Figure 1d).

2.4. Thermal Analysis

High-pressure differential scanning calorimetry (HP-DSC) and normal DSC were performed at different heating rates in order to investigate the apparent activation energy (E_a) of the hydrogenation and dehydrogenation process, respectively.

2.4.1. High-Pressure Differential Scanning Calorimetry

The HP-DSC traces that were conducted under 30 bar of hydrogen gas atmosphere with different heating rates of 10, 11, 12, 14, 15 °C/min for the nanocomposite MgH₂/5wt% (2.5 ZrNi₅/2.5 Nb₂O₅) powders that were obtained after 50 h of milling are shown together in Figure 4a,b. Obviously shown in the figure, all of the scans revealed exothermic peaks (low temperature peaks) that were related to the hydrogenation reaction taking place between hydrogen and metallic Mg. The peak temperature of exothermic peaks appeared in the temperature range between 450 °C and 482 ^oC, depending on the applied heating rates (Figure 4b). The XRD analysis (not shown here) of the powders that were taken after these exothermic peaks confirmed the formation of MgH₂ phase coexisted ZrNi₅ and Nb₂O₅ phases without evidence of the formation of any other reacted phase(s).

The second peaks appeared at a higher temperature side (above 712 °C) that was related to endothermic events, referred to as the decomposition of MgH₂ phase (Figure 4a). Where the peak intensity of the exothermic and endothermic peaks increased proportionally with increasing the heating rates, the peak temperatures (T_p) were significantly shifted to the higher temperature side upon increasing the heating rates from 10 °C/min to 15 °C/min, as shown in Figure 4a. The improved hydrogenation kinetics of the nanocomposite MgH₂/5 wt% (2.5 ZrNi₅ + 2.5 Nb₂O₅) powders was investigated by calculating the apparent E_a, while using Arrhenius approach. The value E_a of the hydrogenation reaction was determined by measuring the absorption temperature (T_p) peaks with different heating rates (k) and then plotting the ln(k) versus 1/T_p, as shown in Figure 4c. The best fit for the results was calculated by the least-squares method. It follows from Figure 4c that all data points lay closely on the same straight line.

The E_a of hydrogenation is obtained from the slope of line (−E/R, where R is the gas constant) and found to be 23.61 kJ/mol. This value, which is far below the one (65 kJ/mol) that was calculated for pure MgH₂ powders [16], indicates a significant improvement on the hydrogenation kinetics upon doping MgH₂ with the present catalytic agent of (ZrNi₅/Nb₂O₅) powders.

2.4.2. Normal-Pressure Differential Scanning Calorimetry

The thermal stabilities and the apparent E_a of decomposition for nanocomposite MgH₂/y wt% (x ZrNi₅/x Nb₂O₅) nanocomposite powders that were obtained after different milling time were investigated with the DSC approach under 1 bar of helium gas atmospheric pressure at different heating rates. Figure 5 shows the DSC thermograms that were conducted at heating rates of 5, 10, 20, 30, and 40 °C/min for the nanocomposite MgH₂/5 wt% (2.5 ZrNi₅/2.5 Nb₂O₅) powders that were obtained after 12.5 h (Figure 5a) and 50 h (Figure 5b) of the milling time. At specific heating rates, the peak temperature of decomposition was significantly decreased with increasing the milling time (Figure 5a,b).

The tendency of temperature decrease can be realized from Figure 3, which elucidates the effect of milling time and the additives on powder refining. The destabilization of MgH₂ upon doping with the hard spherical powders of (ZrNi₅/Nb₂O₅) and milling for long time (50 h) led to a drastic decreasing in the particle size of MgH₂ (Figure 3a), since such abrasive catalysts acted as micro-scaled grinding media to generate lattice defects and imperfections. The existence of these defects raised the free energy of the system, making its boundaries accessible. This is indicated by the significant decreasing of Ea with increasing the milling time (Figure 6a).

MgH₂/5 wt%(2.5 ZrNi₅ + 2.5 Nb₂O₅) powders that were obtained after 12.5 h of milling had Ea of 165.91 kJ/mol (Figure 6a). Further milling time lead to a drastic decrease of Ea value that reached 92.95 kJ/mol after 50 h, as illustrated in Figure 6a. Figure 6b summarizes the concentration effect of (ZrNi₅/Nb₂O₅) on the E_a and decomposition temperature (Temp.^decom) for the nanocomposite powders obtained after 50 h of milling. Both of the parameters tended to monotonically decrease with increasing the milling time. For example, pure MgH₂ that was obtained upon RBM for 200 h had E_a and Temp.^decom of 133.64 kJ/mol and 721 K, respectivelyy (Figure 6b).

Doping the hydride phase with 5 wt% (2.5 ZrNi₅ + 2.5 Nb₂O₅) led to a significant decrease on E_a and Temp.^decom to reach a lower values of 92.98 kJ/mol and 783 K, respectively. These values were continuously decreased with increasing the dose of cataysts to 7 wt% (88.27 kJ/mol and 475 K), 10 wt% (85.07 kJ/mol and 469 K), and 15 wt% (81.69 kJ/mol and 458 K), as eludicated in Figure 6b. A further increase of the catalytic dose had a beneficial effect on decreasing the E_a and Temp.^decom to 77.25 kJ.mol^-1/446 K (20 wt%) and 75.66 kJ.mol⁻¹/441 K (30 wt%), as displayed in Figure 6b. The Ea of decomposition of nanocomposite MgH₂/y wt%(x ZrNi₅ + x Nb₂O₅) systems had lower E_a when compared with different MgH₂ catalyzed systems, as shown in

Table 2. Apparent Activation energy (Ea) of Decomposition for MgH₂ Doped with Selected Catalysts.

Catalytic Agents	Ea (kJ/mol)	References
Present study Pure MgH2, 200 h of RBM	113.67	-
Present study 5 wt% (2.5 ZrNi5 + 2.5 Nb2O5)	92.98	-
VnbO5	99	[17]
Nb2O5	102	[18]
MnFe2O4	108	[19]
Co2NiO	118	[20]
CeO2	109	[21]
TiO2	111	[22]
SrFe12O19	114	[23]
V	119	[24]
FeCl3	130	[25]
Cr2O3	86	[26]

. In contrast, the E_a of the present system shows higher values when compared with MgH₂ doped with Cr₂O₃, as presented in Table 2.

2.5. Hydrogenation/Dehydrogenation Kinetics

2.5.1. Pure MgH₂ Nanocrystalline Powders

The hydrogenation/dehydrogenation kinetics of pure MgH₂ nanocrystalline powders that were obtained after 200 h of RBM are displayed in Figure 7a,b, respectively. The hydrogenation kinetic was measured under 10 bar of H₂, where the kinetic of dehydrogenation was investigated under 200 mbar of hydrogen. Both measurements were obtained at applied temperatures of 200 °C and 300 °C.

The fine MgH₂ powders (Figure 3b) that were obtained after such long-term of milling (200 h) were capable of absorbing 2.57 wt% H₂ in 300 s at rather low temperature (200 °C), as presented in Figure 7a. The absorption kinetic was increased with increasing the temperature to 300 °C, as indicated by the larger molar fraction of H₂ (4.72 wt%) that absorbed at the same length of time (300 s). Increasing the applied time 2000 s) increased the hydrogen amount that was absorbed by the system to reach to 6.1 wt% and 6.69 wt% at 200 °C and 300 °C, respectively (Figure 7a). Further improving on the hydrogen storage capacity (6.28 wt%) was attained with increasing the absorption time to 2500 s for the sample that was measured at 200 °C, as displayed in Figure 7a. Comparing the hydrogenation kinetics of as-prepared MgH₂ nanocrystalline powders of this study with several of the reported values in the litrature (

Table 3. Hydrogen storage properties for different types of MgH₂ powder.

Material Type	Way of Preparations	H2 Storage Capacity (wt%)	Hydrogenation/Dehydrogenation Kinetics	Ref.
MgH2 of the present work	Reactive ball milling of pure Mg powders under 50 H2 bar, 200 h	6.69 −6.52	Absorption: 300 °C/10 bar/2000 s Desorption: 300 °C/0.2 bar/8000 s
Commercial Mg powders	As-received	0 1.4	Absorption: 300 °C/10 bar/120 h     400 °C/10 bar/120 min	[28]
Nanocrystalline MgH2	Ball milling under Ar gas for 20 h	6.0	Absorption: 300 °C/120 min	[28]
Ultrafine Mg nanoparticles	Hydrogen-plasms-metal reaction under a mixture of 70% Ar + 30% H2 at 25 V/300A.	7.5	Absorption: 300 °C/40 bar/30 min	[35]

) led us to realize that long term milling always leads us to deducing the large aggregated powders into finer particles (Figure 3b).

Applying long milling time was essential in increasing the lattice imperfections and mechanical deformation that were intoduced to the milled powders, and hence deduced the kinetics barrier (inset of Figure 7a). It is worth mentioing that longer milling time is responsible for modifying the thermodynamic barrier of the system upon the formation of significant volume fraction of γ-MgH₂ metastable phase (inset of Figure 7a), as suggested by many authors [27,28,29,30,31,32,33,34]. From Table 3, it can be summarized that the starting MgH₂ powders that were prepared in the present work had better hydrogenation kinetics and higher storage capacity when compared to the other reported ones that were prepared with shorter ball milling time.

In contrast to the rather fast hydrogenation kinetic, the MgH₂ powders obtained after 200 h of RBM revealed very slow dehydrogenation kinetic, as shown in Figure 7b. At 200 °C, the powders released a very small amount of its hydrogen storage capacity (−0.18 wt%), even after 12,000 s (200 min), as displayed in Figure 7b. Increasing the applied temperature led to enhancing the desorption kinetics, as indicated by the rather short time that is required to release −6.36 wt% H₂ within 4000 s (66.67 min). At this temperature, the hydrogen released tended to be saturated at −6.52 wt%, as indexed in Figure 7b.

2.5.2. Nanocomposite MgH₂/y-wt% (x-ZrNi₅/x-Nb₂O₅) Powders

The catalyzation step was necessary to improve both the hydrogenation and dehydrogenation kinetics of as-prepared MgH₂ nanocrystalline powders that were obtained after 200 h of RBM time. Binary (x-ZrNi₅/x-Nb₂O₅) catalytic agent was selected with different concentrations to study their synergetic effect on enhancing the storage properties of MgH₂ powders. For all compositions, the MgH₂ powders were doped with the proper dose of the catalytic binary system and experienced 50 h of high energy ball milling to ensure the formation of homogeneous powders that were closed to the starting nominal composition. In the present study, six batches of nanocomposite powders were prepared under the same conditions.

The samples can be classified into four batches, composited of low, medium, and high concentrations of catalysts.

Table 4. Batches of as-prepared nanocomposite powders with different catalytic concentrations.

Sample	Concentrations (wt%)
	MgH2	ZrNi5	Nb2O5
1	95	2.5	2.5
2	93	3.5	3.5
3	90	5	5
4	85	7.5	7.5

lists the code number of each sample and the corresponding catalytic concentrations.

The kinetics of hydrogenation and dehydrogenation for four samples corresponding to Batch #I (Table 4) are displayed in Figure 8a–c, respectively. In general, the hydrogenation kinetics that were measured under 10 bar at 175 °C of all the catalyzed samples with different concentrations were very fast absorption kinetics (Figure 8a) when compared with the undoped MgH₂ powders (Figure 7a and Table 3). Kinetically, they behave almost same, however they significantly differed in their hydrogen storage capacity, as shown in Figure 8a,b. The hydrogen storage degradations shown in samples 2, 3, and 4 were attributed to the large molar fraction of heavy-molecular mass of the catalytic agent that was used for doping MgH₂ powders (Table 4). Based on these results, it can be concluded that MgH₂/5 wt% (2.5 ZrNi₅/2.5 Nb₂O₅) nanocomposite system combined excellent hydrogen storage properties, as indicated by the fast absorption of 5 wt% within 50 s, as presented in Figure 8a,b. This nanocomposite system saturated at 5.6 wt% H₂, after 300 s (5 min), is shown in Figure 8a,b.

The dehydrogenation kinetics of samples #1, #2, #3, and #4 are displayed in Figure 8c. All of the measurements were conducted at 200 °C under 200 mbar of H₂ pressure. The samples showed significant differences in their basic hydrogen storage properties, depending on the concentrations of catalytic agent that was added to each of them. For example, samples #1 and #2 showed excellent kinetics, implied by desorbing −2.26 and −2.15 wt% H₂ within very short time (50 s), as shown in Figure 8c. The hydrogen that was released for these two samples was saturated after 200 s at −5.49 and −5.27 wt% (Figure 8c). The low-magnification SEM- secondary electron detector (BED) micrograph that was taken for sample 1 after compilation of dehydrogenation test is presented in Figure 9a. Obviously presented, the fine catalytic (ZrNi₅/Nb₂O₅) nanoparticles that are indicated by the yellow arrows shown in Figure 9a are homogeneously adhered on each individual Mg particle to form uniform nanocomposite powders. This homogeneous distribution was confirmed by scanning transmission electron microscope (STEM)/dark field image (DFI) beyond the nanoscale level. ZrNi₅/Nb₂O₅ nanoparticles occupied their catalytic sites on the surface of Mg powder matrix in a homogeneous fashion without agglomeration, as indexed in Figure 9b. Such homogeneous distribution of ZrNi₅/Nb₂O₅ nanoparticles among the powder surfaces of Mg/MgH₂ that are shown in the present work is believed to be the major factor leading to superior improvement on hydrogen uptake/release kinetics.

In contrast to samples #1 and #2, the dehydrogenation kinetics that are related to samples #3 and #4 showed unexpected slow kinetics behaviors with a serious decrease of their storage capacity, as can be seen in Figure 8c. Based on several morphological examinations of these two samples, we believe that using large molar fractions (10 wt% and 15 wt%) of catalysts led to sealing/plugging the “gateway pores” existing on outer surface of MgH₂ powders and hence prevented a fast release of hydrogen. This assumption was based on FE-SEM observations (Figure 9c) of sample #4-powders that were obtained after 300 s of desorption time (Figure 8b). The overall Mg/MgH₂ powder matrix hosted a large volume fraction of ZrNi₅/Nb₂O₅ nanoparticles that tended to adhere on the outermost shell of the matrix powders in heterogeneous distribution fashion (Figure 9a). Most of the pores that formed on the powder surface during the hydrogenation process were “sealed” by these fine ZrNi₅/Nb₂O₅ nanoparticles to create a hydrogen releasing barrier. Accordingly, the gas was trapped inside the core of Mg powders. More study is required to investigate the drawback of using “overdose” catalytic agents on the kinetics behavior and storage capacity of metal hydride powders. Figure 9b presents the STEM/DFI of this sample. In the figure, a huge amount of nano-spherical ZrNi₅/Nb₂O₅ particles were adhered on Mg/MgH₂ powders, handicapping hydrogen release.

2.6. Hydrogenation/Dehydrogenation Cycle-Life-Time

The cyclability of MgH₂ that was doped with 5 wt% (2.5 ZrNi₅/2.5Nb₂O₅) powders and milled for 50 h was examined by cycle-life-test for 100 h (Figure 10a). This test was conducted under 10 bar H₂ (hydrogenation)/200 mbar (dehydrogenation) at 200 °C. The sample was first activated under 40 bar of hydrogen at 370 °C for 54 h. This activation process was necessary to breakdown the thin MgO layer that coated the powders to increase the mass molar fraction of fresh Mg metal surfaces against the oxide phase. Accordingly, the hydrogen storage capacity was significantly increased to ~ 6.5 wt%. In addition, this activation process led to creating large volume fractions of pores that were desired for facilitating fast charging/discharging processes.

This new system possessed good performance, as indicated by the large number of achieved hydrogenation/dehydrogenations cycles, reaching to 228, as displayed in Figure 10a. The last 9.5 cycles achieved between a time interval of 95 h to 100 h, as displayed in Figure 10b. Obviously, the sample maintained its fast hydrogen uptake/released kinetics with almost constant storage capacity value in the range between 6.48 wt% to 6.49 wt%, as displayed in Figure 10b.

A low magnification secondary SEM micrograph of the powders that were obtained after completion of the last desorption cycle is presented in Figure 11. The sample consisted of numerous ultrafine clusters of Mg and (ZrNi₅/Nb₂O₅) that were agglomerated together due to Van der Waals attraction to form large aggregates with a porous structure, as shown in Figure 11.

In order to realize the possibility of the formation of an intermediate phase during the cycle-life-test, the XRD of the powder that was taken after compilation of the last dehydrogenation cycle was conducted (Figure 12a). The powders contained large Mg crystals, as indicated by the sharp Bragg peaks shown in Figure 11a. Both of ZrNi₅ and Nb₂O₅ phases revealed their corresponding Bragg lines, as presented in Figure 12a. Neither the MgH₂ phase nor reacted product, such as ZrH₂, NbH, ZrO₂, etc. could be detected. The FE-HRTEM micrograph of the powders taken after the cycle-life-test is shown in Figure 11b. The image displays an aggregated Mg powder, indexed by a Miller fringe image of (101). This particle was adhered by to fine crystals corresponding to ZrNi₅ (311) and Nb₂O₅ (002), as elucidated in Figure 12b.

3. Materials and Methods

3.1. Sample Preparations

3.1.1. MgH₂ Nanocrystalline Powders

In the present study, Mg powders (99.8 wt%, 80 μm in diameter) supplied by Sigma-Aldrich, St. Louis, MI, USA. were used as starting reactant materials with hydrogen gas (99.99 wt%) provided by local company in Kuwait. Seven gram of the powders were sealed into a tool steel vial with 60 tool steel balls of 12 mm in diameter inside the He glove box. The ball-to-powder weight ratio was 50:1. The vial had been charged with H₂ gas under 55 bar before it mountained on planetary type ball mill. The reactive ball milling (RBM) process was started for 200 h to obtain a fully reacted MgH₂ nanocrystalline powders.

3.1.2. Catalysts

• ZrNi₅ Intermetallic Compound

Pure bulk Zr pellets (99 wt%) and Ni foil (99.99 wt%) supplied by Sigma-Aldrich, St. Louis, Missouri, USA were snipped into small pieces and balanced to give the nominal atomic composition of ZrNi₅. The starting materials were etched with 70% H₂O + 30% HCl, rinsed by ethanol, and then dried at 250 °C for 3 h. Arc melter equipment was used in Ti-gettered He atmosphere to fabricate the desired ZrNi₅ alloy, while using water-cooled Cu crucible. The melting process was started by melting Ti and then melting the metallic bulk mixture of Zr and Ni. The metallic button that was obtained after melting was turned over and re-melted six times to ensure the homogeneity. The master alloy was then removed out from the arc melter, washed by acetone and ethanol, and then crushed down into small shots (~18 mm), while using 100 ton cold press. The shots were then crushed down into finer particles (less than 100 μm in diameter) while using a disk mill.

• Preparation of ZrNi₅/Nb₂O₅ Catalysts

Six composition of binary ZrNi₅/Nb₂O₅ catalytic agent systems were prepared to provide the nominal composition (wt%) of 2.5/2.5, 3.5/3.5, 5/5, 7.5/7.5, 10/10, and 15/15. The powders of each patch were individually mixed with low energy ball mill for 25 h, being operated at 100 rpm.

3.1.3. Preparation of MgH₂/y-wt% (x-ZrNi₅/x-Nb₂O₅) Nanocomposite Powders

The as-prepared MgH₂ and binary ZrNi₅/Nb₂O₅ (y) powders were balanced in the He glove box and mixed to give the nominal composition (wt%) of 95/5, 93/7, 90/10, 85/15, 80/20 and 70/30. The mixed powders of each patch (10 g) were charged into the RBM vial with 50 ball of 12 mm in diameter and then pressurized with 50 bar of H₂. The system was then mountained on a high energy-planetary ball mill, where the milling process was taken place for 50 h with a seed of 250 rpm.

3.2. Sample Characterizations

All samples were characterized by means of X-ray diffraction (XRD) (Rigaku smart Lab, Tokyo, Japan) with Cu-Kα radiation, 200-kV field emission high resolution transmission electron microscope (FE-HRTEM) that was equipped with a STEM/EDS system, and FE-scanning electron microscope (FE-SEM). The apparent activation energies of hydrogenation and dehydrogenation were investigated with different heating rates, while using high-pressure differential scanning calorimetry (HP-DSC) (Setaram, France) and normal DSC systems, respectively. The hydrogenation/dehydrogenation kinetics were investigated via Sievert’s method under hydrogen pressure in the range between 200 mbar (desorption) to 10 bar (absorption).

4. Conclusions

In parallel to the new trend of employing advanced carbon-based materials as catalysts for improving hydrogen storage properties [36,37,38], the present work proposed a new efficient catalytic agent of binary ZrNi₅/Nb₂O₅ system that is used to enhance the hydrogen storage property of MgH₂. In this study, the reactive ball milling technique, suing a high-energy ball mill operating under 55 bar of H₂, was employed to prepare MgH₂ nanocrystalline powders, starting from pure Mg powders. On the other hand, the Ti-gettered He atmosphere arc melter prepared the intermetallic compound of ZrNi5. The alloy was crushed down into small pieces with 100-ton cold press. The pieces were then milled by disk mill to obtain finer particles. Binary catalytic agent of ZrNi₅/Nb₂O₅ powders were prepared with a low energy ball mill mixture for 25 h to obtain the nominal composition of 2.5/2.5, 3.5/3.5, 5/5, 7.5/7.5, 10/10, and 15/15. The as-prepared MgH₂ powders were then high-energy ball milled for 50 with the desired ZrNi₅/Nb₂O₅ composition of MgH₂/y-wt% (x-ZrNi₅/x-Nb₂O₅), where y; 95, 93, 85, 80, and 70 wt%.

The results have shown that doping MgH₂ powders with low concentrations (5, 7, 10 wt%) of this new catalytic agent led to superior absorption/desorption kinetics, as indexed by the short time that is required to absorb/desorb 4.2−5.6 wt% H₂ within 200 s to 300 s. When the doping dose was increased to 15−30 wt%, pounced kinetics were attained, however a significant decrease in the hydrogen storage capacity was seen. It is worth mentioning that preparations of ultrafine nanocomposite powders led to enhancing the hydrogen diffusion and shortening the time that is required for hydrogenation/dehydrogenation process. Doping MgH₂ with a hard refractory catalytic system acted as a grain growth inhibitor, in which the Mg/MgH₂ powders maintained their submicron level during the cycle-life-test, being extended to 100 h at 200 °C, with moderate degradation on the gas uptake/release kinetics.

We should emphasize that the formation of intermediate reacted phases, such as Nb_xH, and Nb, could not be observed in the present work upon the hydrogenation process of MgH₂/(ZrNi_5/Nb₂O₅). The formation of such reactive undesired products as intermediate phases may be attributed to the relatively high temperature that is used in the hydrogenation process that usually ranged between 250 °C to 350 °C. Under the application of such high temperature, hydrogen reduced Nb₂O₅ into NbO, Nb, and H₂O (vapor). Metallic Nb preferred to react with hydrogen at such a high temperature to form NbxH. Since very low hydrogenation temperature (200 °C) and moderate hydrogen pressure (10 bar) was applied in the present study, these undesired phases were absent.