Mechanically-Induced Catalyzation of MgH₂ Powders with Zr₂Ni-Ball Milling Media

Abstract

Magnesium hydride (MgH₂) holds immense promises as a cost-effective hydrogen storage material that shows excellent storage capacity suitable for fuel cell applications. Due to its slow hydrogen charging/discharging kinetics and high apparent activation energy of decomposition, MgH₂ is usually doped with one or more catalytic agents to improve its storage capacity. So often, milling the metal hydride with proper amounts of catalyst leads to heterogeneous distribution of the catalytic agent(s) in MgH₂ matrix. The present work proposes a cost-effective process for doping Mg powders with Zr₂Ni particles upon ball milling the powders with Zr₂Ni-balls milling media under pressurized hydrogen. Fine Zr₂Ni particles were gradually eroded from the balls and homogeneously embedded into the milled powders upon increasing the ball milling time. As a result, these fine hard intermetallic particles acted as micro-milling media and leading to the reduction the Mg/MgH₂ powders. Meanwhile, Zr₂Ni eroded particles possessed excellent heterogeneous catalytic effect for improving the hydrogenation/dehydrogenation kinetics of MgH₂. This is implied by the short time required to absorb (425 s)/desorb (700 s) 6.2 wt% H₂ at 200 °C and 225 °C, respectively. The as-milled MgH₂ with Zr₂Ni balls possessed excellent cyclability, indexed by achieving continuous 646 cycles in 985.5 h (~1.5 cycle per hour) without serious degradation.

1. Introduction

Hydrogen storage nanocrystalline and nanocomposite materials are current area of tremendous research interest in chemistry, energy, materials science and the environment. Reference to its natural abundance, low cost, high gravimetric (7.60 wt%) and volumetric (110 g/L) hydrogen storage capacities, magnesium (Mg) metal has been receiving great attention for four decades as a promising hydrogen storage material. Reactive ball milling (RBM), which was investigated in the beginning of the 1990s [1,2] is a cost effective process, employed for conducting gas–solid reactions at ambient temperature, starting from pure metal and desired gas. Since then, RBM has been intensively used to produce large amounts of high quality MgH₂ nanocrystalline powders, starting from Mg metal and hydrogen gas [3,4,5].

In contrast to the many attractive properties found in MgH₂ material, it is seldomly used in desired fuel cell application without intensive treatment [6,7,8,9,10]. This is attributed to high activation energy, high decomposition temperature and slow absorption/desorption kinetics [11,12,13,14]. Severe plastic deformation (SPD) technique [15], including high energy ball milling [16], cold rolling (CRing) [17,18], high pressure torsion (HPT) [19] and equal channel angular pressing (ECAP) [20] were employed to improve the hydrogen storage properties of MgH₂. Apart from the mechanical treatment approach, catalyzation MgH₂ powders with proper volume fractions of catalytic agents [21], including pure metals [22], intermetallic compounds [23,24,25,26], metal-oxides [27,28], metal/metal oxide composites [5,29], -carbides [30,31], metallic glassy alloys [32,33,34] and chlorides [35] have shown outstanding effects on enhancing the hydrogen storage properties of MgH₂ and infer its apparent activation energy of decomposition [36]. More recently, self-assembled MgH₂ and metal borohydride nanoparticles on graphene technique has shown great enhancement of the behavior of the metal hydride phase [37,38,39].

1.1. Mechanically-Induced Catalyzation (MIC) of Mg/MgH₂ Powders

MIC catalyzation refers to doping Mg/MgH₂ powders with desired catalytic agent(s), using high-energy ball milling operated under inert or hydrogen gas for at least 10 h. Like any other chemical processing, the activity of the catalyst is expected to be linearly increased with increasing its amount added to MgH₂. Efficiency of the catalytic agent active sites is not only depending on the amount used, but it is mainly affected by the degree of their dispersion homogeneity in the matrix (Mg/MgH₂). In many cases, pure metal and metal powder alloys added to MgH₂ and then milling are affected by the cold welding taken place during the milling process. Accordingly, they tend to be agglomerated to form large powder particles and are heterogeneously distributed in a few zones in the Mg/MgH₂ matrix. Such heterogeneous distribution of catalytic agents leads to produce either rich or poor active site Mg/MgH₂ powders.

1.2. Ball-Milling Media as a Source of Catalysts

In 2016, El-Eskandarany and his team replaced tool steel balls with Ni-balls milling media for RBM process of Mg powders under pressurized hydrogen gas [40]. They pointed out that Mg/MgH₂ powders were homogeneously catalyzed with eroded Ni fine particles upon ball–powder–ball collisions [41]. This approach inhibited the agglomeration of metallic catalytic agent and ensured the formation of very uniform nanocomposite powders.

In the present work, we attempted to investigate the possibility of doping Mg powders with tetragonal-Zr₂Ni, using Zr₂Ni balls-milling media as a source of solid catalysts. The catalytic effect on the hydrogen storage properties, indexed by hydrogenation/dehydrogenation kinetics measured at different RBM times and temperatures were studied. Moreover, the effect of RBM time on decomposition temperature and apparent activation energy of MgH₂ were investigated. In addition, the cyclability of MgH₂ milled with Zr₂Ni balls is reported.

2. Results and Discussion

2.1. Hydrogenation Kinetics of MgH₂ during the RBM Process

Effect of Milling Media (MM)

For the purpose of the present work, an arc melting technique was used to fabricate Zr₂Ni master alloy buttons, started from pure Zr and Ni bulk metals. The materials were placed in a Cu-hearth cooled by water, where the melting process (Figure 1a,b) was conducted for six times to ensure homogeneity of the product. The fabricated Zr₂Ni master alloys had nearly spherical-like shape ~10 mm in diameter (Figure 1c). These spherical-balls were used as milling media (MM) for conducting RBM of Mg powders under pressurized (50 bar) H₂ (Figure 1d). Details of these experiments are described in the Materials and Methods section.

In all experiments, high-pressure milling vial equipped with a gas-temperature-monitoring (GTM) system was used to monitor the progress of the mechanically induced gas–solid-reaction taking place between Mg and H₂ during different stages of RBM time. This was indicated by the pressure change (desorption) and time required for absorption at a specific RBM time. In addition, the GTM system allowed us to realize the effect of MM (tool steel and Zr₂Ni balls) on the mechanically-induced hydrogenation process.

Figure 2a elucidates the effect of RBM time on the vial’s pressure during RBM of Mg under 5000 kPa of H₂, using MM-to-P as 40:1. At the first stage of RBM, the vial’s pressure slightly dropped from 5000 kPa to ~4900 kPa after 3.2 h (Figure 2a). At this early stage of RBM, the powders were subjected to severe lattice imperfections due to the effect of ball–powder–ball collisions (Figure 2b). Meanwhile, the hydrogen pressures inside the vials employed tool steel and Zr₂Ni as MM were 4799 and 4391 kPa, respectively (Figure 2a).

After 6 h, this drastic decrease in pressure upon using Zr₂Ni MM is believed to be related to three factors; (i) a greater number of active Mg surfaces were created and guttered (absorbed) much more hydrogen when compared with the powders milled with tool steel MM, (ii) a certain amount of hard Zr₂Ni fine powders came from corroded Zr₂Ni balls and introduced to Mg powders may have played a vital catalytic role for splitting hydrogen molecules into atoms and (iii) the presence of such refractory Zr₂Ni hard powders in the Mg/MgH₂ powder matrix played the role of micro-media, in which they led to accelerate reduction of powder sizes into the sub-micro level.

In the intermediate stage of RBM, the sample milled with Zr₂Ni MM slightly guttered much more hydrogen upon milling for 10 h, as indicated by the drop in vial pressure (4269 kPa), as shown in Figure 2a. After this stage, the vial’s pressure of Mg milled with tool steel balls reached almost the same value (4269 kPa), which suggested that both samples milled with different MM have absorbed almost the same amount of hydrogen (Figure 2a). During the last stage of RBM (14 to 20 h) the slight decrease in the vial’s pressure was monitored (~4000 kPa) for both samples, which suggested completion of the gas–solid-reaction process, as displayed in Figure 2a.

2.2. Crystal Structure

X-ray diffraction (XRD) was employed to study the effect of RBM time on structural changes of Mg powders obtained after 6 h (early stage), 12.5 h and 25 h (intermediate stage) and 50 h (final stage) of RBM. The XRD pattern of the sample obtained after 6 h of milling (Figure 3a) revealed Bragg peaks corresponding to pure hcp-Mg with no evidence of MgH₂ present, as shown in Figure 3a. The XRD pattern of the sample obtained after 12.5 h RBM revealed new sharp Bragg peaks related to β-MgH₂ and γ-MgH₂ overlapped with low intensity peaks corresponding to unreacted hcp-Mg (Figure 3b). The Bragg peaks related to hcp-Mg were hardly seen in the XRD of the sample obtained after 25 h, suggesting the completion of gas–solid-reaction (Figure 3c).

The XRD pattern of the sample obtained after 50 h of RBM time, using Zr₂Ni balls is displayed in Figure 4a. The powders consisted of ultrafine grains of MgH₂, characterized by broad diffracted Bragg lines corresponding to β- and γ-MgH₂ (Figure 4a). After this stage of RBM (50 h) considerable amount of Zr₂Ni (~5 wt%) fine particles were mixed with MgH₂, as indicated by the Zr₂Ni-Bragg peaks overlapped with the metal hydride phase (Figure 4a). These fine particles were introduced to the powders as a result of using Zr₂Ni as milling media. The XRD of Zr₂Ni master alloy used to manufacture the milling media is shown in Figure 4b. All diffracted lines are matching well with the reported phase of tetragonal-Zr₂Ni (PDF# 00-038-1170), as indexed in Figure 4b.

2.3. Local Structure and Compositional Analysis

In order to understand the effect of catalyzation techniques by Zr₂Ni on the behavior of MgH₂ powders, 5 wt% of tetragonal-Zr₂Ni powders were mechanically ball milled with Mg powders under pressurized hydrogen gas for 50 h, using tool steel milling media with MM-to-P as 40:1. The bright field image (BFI) of the sample obtained after 50 h is displayed in Figure 5a. Obviously shown, metallic Zr₂Ni powders tended to agglomerate to form large composite powder of about 3 μm in diameter, as shown in Figure 5a. One drawback of milling Zr₂Ni alloy powder with Mg powder was the tendency of Zr₂Ni to agglomerate and distribute heterogeneously in Mg/MgH₂ powders to form rich/poor catalyzed zone within the particle (Figure 5a).

In contrast, when Zr₂Ni fine-shots were gradually introduced to Mg powders upon using Zr₂Ni milling media, the catalyzation process took place homogeneously, as indicated by the good distribution of Zr₂Ni grains in MgH₂ matrix (Figure 5b).

More information related to the homogeneity of MgH₂/Zr₂Ni nanocomposite powders obtained after 50 h of RBM time, using Zr₂Ni was obtained from elemental analysis performed with scanning transmission electron microscope (STEM)/X-ray elemental mapping (Figure 6). The dark gray zones shown in Figure 6a refer to corroded Zr₂Ni particles (~8 nm in diameter) that came from Zr₂Ni and were introduced to MgH₂ matrix (Figure 6b). These particles were embedded homogeneously in the MgH₂ matrix (Figure 6b,c) to form uniform aggregate powders 110 nm in diameter, as displayed in Figure 6a.

2.4. Thermal Analysis

Differential scanning calorimetry (DSC) was conducted to study the influence of RBM time on decomposition temperature (T^dec) of MgH₂ doped with 5 wt% Zr₂Ni powders, using tool steel MM (Figure 7a). These results were compared with DSC results of MgH₂ powders milled for different RBM times, using Zr₂Ni balls (Figure 7b). In both figures, there are two endothermic events that refer to the decomposition of MgH₂ powders, suggesting that the decomposition process took place through two stages of γ-β-MgH₂ phase transformation followed by decomposition of β-MgH₂ into Mg-metal and hydrogen molecule, as summarized in Equation (1);

$$\mathrm{as}-\mathrm{RBMed} \mathrm{powders} \left(\mathsf{\gamma }/\mathsf{\beta }-{\mathrm{MgH}}_2\right)\to \mathsf{\beta }-{\mathrm{MgH}}_2\to \mathrm{Mg}+2\mathrm{H}.$$

(1)

Temperature of γ-β-phase transformation $\left({\mathrm{T}}_{\mathsf{\gamma }\to \mathsf{\beta }}^{\mathrm{trans}}\right)$ for MgH₂ doped with 5 wt% Zr₂Ni powders, using tool steel MM tended to decrease slightly from 678 K to 672 K after milling for 25 h and 37.5 h, respectively (Figure 7a).

Increasing the RBM time to 50 h and 100 h led to drastic reducing on ${\mathrm{T}}_{\mathsf{\gamma }\to \mathsf{\beta }}^{\mathrm{trans}}$, recorded to be 651 K and 635 K respectively. This significant decrease may be attributed to the reduction of the free energy required to accomplish γ-β-MgH₂ phase transformation that was enhanced by introducing severe lattice imperfection to the powders upon further milling. It can also be attributed to the effect of the Zr₂Ni catalytic agent that became well distributed in the MgH₂ matrix. The γ-MgH₂ phase of the sample obtained after 200 h of RBM time transformed into β-phase at a lower temperature of 627 K, as shown in Figure 7a. Introducing lattice imperfection in the catalyzed MgH₂ with Zr₂Ni powders upon increasing RBM time from 25 h to 50 had obvious effect on decreasing T^dec of the β-MgH₂ phase, as indexed by the decomposition peak temperatures of 695 K, 678 K and 666 K, respectively (Figure 7a).

It should be noted that this two-stage decomposition is not unique for MgH₂ prepared in the present work, but it has been reported by many authors [9,11,13,36].

The ${\mathrm{T}}_{\mathsf{\gamma }\to \mathsf{\beta }}^{\mathrm{trans}}$ of the first endothermic peaks for MgH₂ powders milled with Zr₂Ni balls after 25 h, 37.5 h and 50 h of RBM time (Figure 7b) showed monotonical reduction with increasing RBM time from 533 K, 529 K to 519 K, respectively. Meanwhile, the corresponding T^dec of β-MgH₂ phase retreated from 592 K to the low temperature side of 577 K and 566 K, respectively, as shown in Figure 7b.

The apparent activation energy (E_a) of decomposition for MgH₂ powders milled with Zr₂Ni balls for 50 h was investigated by DSC with heating rates (k) of 5, 10, 20, 30 and 50 °C/min, using the Arrhenius approach in Equation (2)

$${\mathrm{E}}_{\mathrm{a}}=-\mathrm{RTln}\left(\mathrm{k}\right)$$

(2)

where k is a temperature-dependent reaction rate constant, R is the gas constant, and T is the absolute peak temperature of decomposition. All the DSC scans revealed sharp, single endothermic events related to the decomposition of MgH₂ phase, as displayed in Figure 8a. Obviously, the peak height increased proportionally with increasing heating rates, where the peak decomposition temperatures were significantly shifted to the higher temperature side upon increasing the heating rates from 5 °C/min to 40 °C/min, as shown in Figure 8a. Taking trace of the sample conducted at 5 °C/min as a typical example, its corresponding T^dec was 519 K (Figure 8a). This value is far below the value (651 K) belonging to MgH₂ milled with Zr₂Ni powders for 50 h (Figure 7a). It can be concluded that milling MgH₂ with Zr₂Ni balls led to destabilizing the metal hydride phase and decreasing the decomposition temperature by 132 K. Such significant decreasing in T^dec is believed to be related to the homogeneous distribution of the particles from the Zr₂Ni balls and uniformly embedded into MgH₂ matrix, as shown in Figure 5b and Figure 6.

E_a of decomposition of MgH₂ milled with Zr₂Ni balls for 50 h was determined by measuring the decomposition peak temperature corresponding to the different k and then plotting ln(k) versus 1/Tp, as shown in Figure 8b. A best fit for the results was calculated by the least-square method. It follows from Figure 8b that all data points lie closely on the same straight line. E_a of 91.51 kJ/mol, which was obtained from the slope of the line (-E/R) is far below than that one obtained for pure MgH₂ nanocrystalline powders (146.53 kJ/mol [16]) and also less than the one calculated for MgH₂ milled with Zr₂Ni powders for 50 h, using tool steel balls (122.38 kJ/mol). It can be concluded here that using Zr₂Ni balls led to an outstanding improvement in the decomposition kinetics of MgH₂ powders.

2.5. De/Rehydrogenation Behaviors

2.5.1. Pressure-Composition-Temperature (PCT)

The PCT behavior of MgH₂ powders milled with Zr₂Ni balls for 50 h was volumetrically investigated by Sievert’s approach at different temperatures (50 °C to 275 °C) and plotted in Figure 9a. The system revealed excellent capability for absorbing a significant amount of hydrogen (4.23 wt%) at very low temperature (50 °C), as displayed in Figure 9a. Hydrogen absorbed by Mg-active surfaces that were created by high energy milling, using Zr₂Ni balls 5 wt% was increased to higher values of 4.58 wt%, 5 wt% and 5.19 wt% upon increasing temperature to 75 °C, 100 °C and 125 °C, respectively (Figure 9a). At a rather high temperature (150 °C and 200 °C) hydrogen masses absorbed by the system were 5.32 wt% and 6.11 wt% respectively, as elucidated in Figure 9a. At this temperature, the powders corresponding to each applied temperature revealed a single reversible hydrogenation/dehydrogenation cycle (Figure 9a). Further temperature increases (225 °C, 250 °C and 275 °C) led to a slight increase in the hydrogen weight present (6.22 wt%, 6.27 wt% and 6.34 wt%, respectively). The presence of clear hydrogenation plateaus can be seen in the range between 0–5.6 wt% H₂ at the high temperature side ranging between 200 °C to 275 °C, as shown in Figure 9a. In addition, smooth desorption plateaus were characterized in the whole hydrogen concentration range (0–5.8 wt%) for all cycles developed at different applied temperatures, as displayed in Figure 9a.

The SEM micrograph of the sample taken after the last dehydrogenation PCT cycle conducted at 275 °C is displayed in Figure 9b. The powder particles were slightly grown in size to form a large aggregate with an apparent particle size of about 100 μm in diameter (Figure 9b). This limited powder agglomeration was attributed to repetition of the hydrogenation/dehydrogenation cycles at different temperatures. The STEM/dark field (DF) image corresponding to the indexed zone shown in Figure 9b is presented in Figure 9c. After this severe domination under cyclic pressure/temperature modes, the powders still maintained their nano-dimension characterizations, indexed by the presence of nanograined Mg (light gray)/Zr₂Ni (dark gray) composite equiaxed grains with sizes ranging between 8 nm to 57 nm, as displayed in Figure 9c.

2.5.2. Hydrogenation Kinetics

Figure 10a shows the effect of RBM time on the hydrogenation kinetics for MgH₂ powders milled with Zr₂Ni balls. The measurements were conducted at 200 °C under hydrogen pressure of 10 bar. All samples obtained after different RBM times were capable of absorbing hydrogen with different scales of absorption time (Figure 10a). In general, the powder milled with Zr₂Ni balls showed better kinetics when compared with the Mg doped with 5 wt% Zr₂Ni powders milled with tool steel balls for 50 h. This is implied by the very long time required (~36085 s) to absorb 6.2 wt% H₂ at high temperature (350 °C), as displayed in Figure 10b.

The powders obtained after the early stage of RBM (12.5 h) absorbed 1.78, 3.3 and 4.53 wt% H₂ within 25, 100 and 300 s respectively, as displayed in Figure 10c. Limited improvement on the hydrogenation kinetics was obtained with increasing the RBM time to 25 h (intermediate stage), as characterized by the marginal increase in absorbed H₂ (4.88 wt%) obtained after 300 s (Figure 10c). The powders obtained after 37.5 h of RBM exhibited good absorption kinetics, as suggested by the higher hydrogen wt% (5.28) absorbed after only 300 s, as presented in Figure 10c. Towards the end of RBM processing time (50 h), where the nanocomposite powders of this final product had homogeneous composition, they absorbed higher hydrogen wt% of 4.98 and 6.1 within 100 s and 300 s, respectively (Figure 10c).

Hydrogen storage capacity of the samples obtained after the early and intermediate stages of RBM (12.5 h, 25 h and 37.5 h) saturated at ~5.5 wt% after ~700 s of absorption time (Figure 10a), while the final product sample (50 h) saturated at ~6.2 wt% H₂ within 425 s, as presented in Figure 10a. When the absorption kinetics of this system are compared with the other MgH₂-based systems, it can be concluded that the present system is one of those few systems showing fast absorption kinetics at rather low temperature (200 °C) and pressure (10 bar H₂), as presented in Figure 10c.

2.5.3. Dehydrogenation Kinetics

Measuring kinetics of the hydrogen released was attained to complete the picture of MgH₂ powders milled with Zr₂Ni balls kinetics behavior for the samples obtained after different RBM times. Figure 11a demonstrates influence of RBM time on improving the dehydrogenation kinetics for MgH₂ powders milled with Zr₂Ni balls for 12.5, 25, 37.5, and 50 h. The dehydrogenation characteristics during the first few seconds of hydrogen desorption (0 s to 300 s) for these selected samples (Figure 11a) are elucidated in Figure 11b. All measurements were conducted at 225 °C under hydrogen pressure of 200 mbar.

The sample obtained after a few hours of RBM (12.5 h) showed very slow desorption kinetics, indicated by the long time required to discharge 0.2 wt% H₂, as shown in Figure 11b. These kinetics were improved by allowing longer desorption time (800 s), where the hydrogen released from the sample was close to 1.5 wt% (Figure 11a). Increasing RBM time to 25 h led to a moderate improvement, characterized by increasing the wt% of hydrogen (2.83) released after only 300 s, as displayed in Figure 11b. Further hydrogen releasing (~5.7 wt%) was attained with increasing desorption time to 800 s (Figure 11a). Outstanding improvement on the dehydrogenation kinetics was achieved for the samples obtained at the end of RBM processing time, 37.5 h and 50 h where they released 5.82 and 5.97 wt% H₂, after a very short time (300 s), as shown in Figure 11b. They reached very close values of 6.2 wt% H₂ after 700 s, where they saturated at this value even after 800 s of desorption time (Figure 11a).

The crystal structure of the 50-h sample taken after dehydrogenation and hydrogenation kinetics measurements was investigated by XRD technique. The XRD patterns of the sample taken after dehydrogenation are shown in Figure 12a,b. Obviously, the powders consisted of polycrystalline composite of hcp-Mg and tetragonal-Zr₂Ni phases, as shown in Figure 12a. Neither reacted and/or intermediate phase(s) could be detected, as presented in Figure 12b. The hydrogenation process of this sample led to the formation of β-MgH₂ phase coexisting with Zr₂Ni, as displayed in Figure 12c.

2.5.4. Cyclability and Performance

Figure 13a displays the full profile of hydrogenation/dehydrogenation cycle-life-time test of Mg powders milled with Zr₂Ni balls under 50 bar H₂ for 50 h investigated at 225 °C, using 10 bar and 200 mbar H₂ for absorption and desorption, respectively. The sample was firstly subjected to severe heat treatment, conducted at 300 °C under 25 bar of hydrogen for 50 h. This powder activation step is necessary to breakdown the hard MgO layer coating the powder particles and to reduce the oxide phase into Mg metal and H₂O vapor, which is evacuated outside of the reactor. Accordingly, the hydrogen storage capacity of the powders increased to 6.52 wt%, as shown in Figure 13a. During the first 400 h of the test, the sample showed behavior of degradation, characterized by an obvious drop in its storage capacity that reached to ~6 wt% H₂ and kinetics decay, indicated by the rather long time (1 h) to complete 1.3 cycles, as indexed in Figure 13b.

In order to enhance the cyclability of the system and improve the powders, they were activated again for 25 h at 350 °C under 35 bar of hydrogen. Then, the cyclic test was continued at 225 °C under hydrogenation/dehydrogenation pressure of 10 and 200 mbar, respectively. In the second part of this test (400 h to 800 h), the powders revealed excellent hydrogen storage capacity (~6.5 wt%) that did not decrease upon increasing the cycle-life-time, as shown in Figure 13c. Moreover, the number of cycles achieved per one hour was significantly increased to be 1.8 cycles (Figure 13c). The hydrogen storage capacity maintained its high value of 6.5 wt% during the third stage of cycle-life-time test, extended from 800 h to 980 h, as displayed in Figure 13d. In addition, the gas absorption/desorption kinetics was constant and did not depreciate during this stage of the test, as implied by achieving 2.3 cycles per h (Figure 13d). Figure 13e shows the last five cycles of the test, which were achieved in the range between ~980.5 h to 985.5 h. The results indicated that the system kept its constant value of storage capacity (~6.5 wt% H₂) with acceptable rate of hydrogenation/dehydrogenation that was measured to be one cycle per hour, as presented in Figure 13e.

3. Materials and Methods

3.1. Fabrication of Zr₂Ni Ball Milling Media

Spherical Zr₂Ni balls with ~10 mm in diameter were prepared by the arc melting technique, using 800A-Arc Melter AM 200, provided by Edmund Bühler GmbH, Bodelshausen-Germany. In this experiment bulk Zr (99.5 wt%) and Ni (99.99 wt%) foils supplied by Sigma-Aldrich (St. Louis, Missouri, USA) were balanced to give the nominal atomic composition of tetragonal-Zr₂Ni phase. The starting foils were placed in a Cu-hearth cooled by water, where the melting process was conducted for 5 times to ensure homogeneity of the product. These spherical balls were used as milling media (MM) for conducting reactive ball milling of Mg powders under pressure (50 bar). A part of the fabricated Zr₂Ni balls were crushed down into small pieces (~3 mm to 5 mm), using a 20-ton cold press and then ball milled under argon gas for 100 h, using a high-energy planetary ball mill operated at 250 rpm.

3.2. Preparation of MgH₂ Powders Catalyzed with Zr₂Ni Particles

Pure Mg powders (99.8 wt%, 80 μm in diameter), supplied by Sigma-Aldrich, St. Louis, Missouri, USA were used as starting materials for conducting gas–solid reactions with hydrogen gas (99.99 wt%, provided by local gas company in Kuwait). In the first experimental set, 6 g of the powders were sealed into a tool steel vial with Zr₂Ni balls of 10 mm in diameter inside a He glove box. In the second set, however, Mg powders were mixed with 5 wt% of Zr₂Ni and charged in the vial with 10 mm tool-steel balls. In both experimental sets, the ball-to-powder weight ratio was 47:1. The vials were then pressurized with H₂ gas under 60 bar before it mounted on a planetary type ball mill. The reactive ball milling (RBM) process was started for different RBM times of 6, 12.5, 25, 37.5 and 50 h. After each RBM run, the milled product was completely discharged inside the glove box, where new unprocessed powders were charged in the vial for further milling. This step was necessary to keep the ball-to-powder weight ratio constant.

3.3. Powder Characterizations

3.3.1. XRD, High Resolution Transmission Electron Microscope (HRTEM), STEM/Energy-Dispersive X-ray spectroscopy (EDS) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS)

The general crystal structure of all samples was investigated by X-ray diffraction (XRD) with Cu-Kα radiation, using 9 kW intelligent X-ray diffraction system, provided by SmartLab-Rigaku, Akishima-shi, Tokyo, Japan. The local structure of samples was studied by 2100F-field emission high resolution transmission electron microscope (FE-HRTEM) operated at 200 kV. The microscope was supplied by JEOL, Musashino, Akishima, Tokyo, Japan. Scanning transmission electron microscopy (STEM, -2100F, Tokyo, Japan), which is equipped with energy-dispersive X-ray spectroscopy (EDS) supplied by Oxford Instruments, Abingdon, UK was used to conduct elemental local analysis. In addition to EDS elemental analysis, inductively coupled plasma mass spectrometry (ICP-MS) provided by Shimadzu Scientific Instruments, Saitama, Japan was employed to get elemental analysis by a chemical analytical approach.

3.3.2. SEM

FE-scanning electron microscope (FE-SEM), provided by JEOL, Musashino, Akishima, Tokyo, Japan, operated at 15 kV was used to investigate the morphological characterizations of the powders.

3.3.3. Normal DSC

Shimadzu Thermal Analysis System/TA-60WS, Saitama, Japan, using normal differential scanning calorimeter (DSC) was employed to investigate the decomposition temperatures of MgH₂ powders with a heating rate of 5 °C/min. The apparent activation energy of decomposition (E_a) for the powders obtained after different RBM time were investigated, using the Arrhenius approach with different heating rates (5, 10, 20, 30, 40 °C/min).

3.3.4. De/Rehydrogenation Behaviors

The hydrogen absorption/desorption kinetics were investigated via Sievert’s method, using PCTPro-2000, provided by Setaram Instrumentation, Caluire-et-Cuire, France, under hydrogen gas pressure of 200 mbar (dehydrogenation) and 10 bar (hydrogenation). The samples were examined at different temperatures of 50, 75, 100, 125, 150, 200, 225, 250 and 275 °C. In this analysis, the dosed hydrogen pressure in absorption/desorption was gradually increased/decreased by 1000 mbar until equilibrium pressure reached to 13,000 and 50 mbar, respectively. The PCT absorption/desorption kinetics were fitted in real-time by the software, to determine the sufficient equilibration time (the next point would start when the uptake had relaxed just 99% to asymptote). A minimum time of 30 min per equilibrium point and a maximum timeout of 300 min were set for each kinetic step in both the absorption and desorption isotherms.

4. Conclusions

The aim of this study was focused on preparing homogeneous MgH₂/catalytic agent (tetragonal-Zr₂Ni) nanocomposite powders with excellent hydrogen storage characteristics and cyclic performance. The work has succeeded to introduce a cost effective catalyzation process, using bulk Zr₂Ni balls as milling media. Based on the results driven from the present work it can be concluded that:

• In contrast to traditional doping and milling Mg powders with Zr₂Ni powders using tool steel balls, our proposed in-situ gradual doping Mg was successfully achieved upon using Zr₂Ni-balls milling media ball milling. This new catalyzation process has shown mutually beneficial for overcoming the agglomeration of catalytic agent in Mg matrix.
• During reactive ball milling process, corroded Zr₂Ni fine particles from Zr₂Ni balls, were introduced to Mg matrix and led to accelerate the mechanically induced gas–solid-reaction. These hard particles played a vital micro milling-media role for disintegrating Mg/MgH₂ powders to nanoscale particles after only 50 h of milling.
• Homogeneous distribution of Zr₂Ni in a MgH₂ matrix had the desired effect on decreasing the decomposition temperature of MgH₂ to 519 K with a low value of apparent activation energy of decomposition (91.51 kJ/mol).
• Drastic disintegration of MgH₂ particles conducted by Zr₂Ni particles led to a decrease in the hydrogen diffusion distance and facilitated excellent hydrogenation properties of Mg. This was realized by the powder capability to react with hydrogen at very low (50 to 125 °C) and moderate (125 to 150 °C) temperatures. The hydrogen storage capacity of the powders measured at 50 °C and 150 °C were 4.23 wt% and 5.32 wt%, respectively. At 200 °C, however, storage capacity reached to 6.11 wt%.
• MgH₂/5 wt% Zr₂Ni nanocomposite powders obtained after 50 h of milling with Zr₂Ni balls possessed superior hydrogenation kinetics at 200 °C, characterized by the short time (425 s) required to absorb 6.2 wt% H₂. This came in contrast with Mg sample doped with 5 wt% Zr₂Ni powders and milled for 50 h, using tool steel balls, which absorbed 6.2 wt% H₂ at 350 °C in 36085 s.
• Likewise the excellent characteristics of hydrogenation, the milled samples with Zr₂Ni revealed good dehydrogenation kinetics, indexed by a short time needed (700 s) to release its full storage capacity (6.2 wt% H₂).
• The powders also showed excellent cyclability for achieving a continuous 646 cycles in 985.5 h without severe degradation.
• One merit of this proposed catalyzation approach is the ability of Zr₂Ni balls to be used several times for mechanical doping of MgH₂ and maybe other metal hydride systems.