The Mg₂NiH₄ used in this experiment was prepared by a method of hydrogenation combustion synthesis method (HCS), which was carried out in Prof. Yunfeng Zhu's group in Nanjing Tech University (Gu et al., 2009; Zhu et al., 2017). CeH_2.73 (purity >99%) was purchased from Hunan Research Institute of Non-ferrous metals. About 0.1 g CeH_2.73 powder was located in a plastic bottle of 3 ml, in the Ar glove-box. Then it was transferred into the small side box of glove-box. The door of side box was then open to let air in for different time: 0.5, 1.5, 3, 10, and 60 min. The oxidized CeH_2.73 became CeH_2.73/CeO₂ composites, and 2 mol% of CeH_2.73/CeO₂ composites were then doped into Mg₂NiH₄ by ball milling at 400 rpm for 4 h.

The phase structures were analyzed by X-ray diffraction (XRD), and the phase transition during dehydrogenation was analyzed by in situ XRD analysis. The XRD analysis was carried out by an X-ray diffractometer apparatus (UItima IV, Rigaku, Japan) with Cu-K α (λ = 0.15405 nm). The tube voltage and tube current were 40 kV and 40 mA, respectively. Thermogravimetry (TG) and Differential Scanning Calorimetery (DSC) were used to study the hydrogen desorption behaviors at different heating rates on a METTLER TGA/DSC 3⁺ synchronous thermal analyzer. The microstructures of the Mg₂NiH₄-Ce_2.73/CeO₂ composites were observed by backscattered electron imaging using scanning electron microscope (SEM) and transition electronic microscopy (TEM).

Six sets of CeH_2.73/CeO₂ composites prepared by oxidation treatments for different durations were analyzed by XRD. The CeH_2.73/CeO₂ composites are marked as S1, S2, S3, S4, S5, and S6, according to the oxidation duration of CeH_2.73 of 0, 0.5, 1.5, 3, 10, and 60 min, respectively. The diffraction patterns are shown in Figure 1. With the increase of oxidation time, CeH_2.73 gradually transforms into CeO₂. After oxidation of 60 min, no CeH_2.73 has left. The obtained XRD patterns were refined by a Jade 6.0 software and the phase contents were calculated by the K value (RIR) method. The relative contents of the CeO₂ phase as increase of the oxidation time are shown in Figure 2. The oxidation data can be fitted by the Avrami–Erofeev equation deduced from the nucleation and growth process:

where α is the ratio of reacted material to total material, m and B are constants. The fitted m is 1.085, which is very close to 1.07, indicating the oxidation of CeH_2.73 at room temperature is a three-dimensional interface reaction process (Lin et al., 2012b). For the S4 sample, the relative content of CeH_2.73 is 54.2 wt%, and the relative content of CeO₂ phase is 45.8 wt%, indicating almost the same molar ratio of cerium hydride and oxide when the oxidation time is around 3 min. After the oxidation treatments, the samples were doped into Mg₂NiH₄ by ball milling.

In order to understand the morphologies of the CeH_2.73/CeO₂ composite doped Mg₂NiH₄, back scattering scanning electron microscopy (BSEM) was carried out. Because the S1–S4 sample will continue to oxidize in the air, the S5 doped Mg₂NiH₄ sample was selected as the experimental material. The morphologies of the as-prepared CeH_2.73/CeO₂ composite (S5 sample) and the ball-milled Mg₂NiH₄ + S5 sample are shown in Figure 3. It could be clearly seen that the particle size of the as-prepared CeH_2.73/CeO₂ composite is about 200–400 nm. After ball milling, the particle size of the CeH_2.73/CeO₂ composite is greatly reduced to below 30–100 nm. Figure 3B shows that the CeH_2.73/CeO₂ composites, which are brighter particles, are uniformly distributed in the Mg₂NiH₄ matrix. The homogeneous CeH_2.73/CeO₂ composites could be beneficial for catalyzing hydrogen storage properties for the Mg-based materials (Hong et al., 2009; Shao et al., 2011; Li et al., 2018).

The decomposition behaviors of the ball-milled Mg₂NiH₄+CeH_2.73/CeO₂ composites are studied by TG-DSC synchronous thermal analyzer as shown in Figure 4. The DSC study of Mg₂NiH₄ + CeH_2.73/CeO₂ composites are at a heating rate of 20 K/min, showing the dehydrogenation initial temperature of Mg₂NiH₄ decreases after addition of CeH_2.73/CeO₂ composites. Among the six sets of catalysts, the S4 sample obtained by oxidation of CeH_2.73 for 3 min, exhibits the highest catalytic effect on reducing the dehydrogenation temperature of Mg₂NiH₄. The dehydrogenation temperature is decreased to 267°C, which is about 17°C lower than the that of as-milled Mg₂NiH₄ (284°C).

To further elucidate the dehydrogenation activation energy of the Mg₂NiH₄-CeH_2.73/CeO₂ composites, DSC experiments for the Mg₂NiH₄ + CeH_2.73/CeO₂ composites at heating rates of 10 K/min and 50 K/min were carried out (results not shown). The activation energy of the dehydrogenation process was calculated by using the Kissinger's method (Kissinger, 1957),

where E_des is the dehydrogenation activation energy, T_m, β, R, and C are the peak temperature, heating rate of DSC experiments, gas constant and another constant, respectively. The relation between T_m and β is linearly fitted as plotted in Figure 5, and the activation energy was summarized in Table 1. Results show that the S4 sample by oxidation of CeH_2.73 for 3 min exhibits the best catalytic effect on reducing the dehydrogenation activation energy of Mg₂NiH₄ (from 76.3 kJ/mol to 62.6 kJ/mol). Zhang (Zhang et al., 2018) et al. suggested that alloying Mg₂Ni with Ti, V, Fe, or Si could reduce the hydrogenation and dehydrogenation activation energies to 60–70 kJ/mol. Our study suggests that the CeH_2.73/CeO₂ composites are also efficient to reduce the dehydrogenation activation energy of Mg₂NiH₄.

Combining XRD and TG-DSC study, it indicates clearly that the initial dehydrogenation temperature and activation energy of Mg₂NiH₄ are maximally reduced as the cerium hydride and cerium oxide are in the same amount. These results well-accord with our previous finding on the effect of CeH_2.73/CeO₂ composites on the dehydrogenation properties of MgH₂ (Lin et al., 2014), which means the CeH_2.73/CeO₂ composite with the same molar ratio of cerium hydride and oxide could be a good catalyst for the dehydrogenation properties of Mg-based hydrogen storage materials, including both MgH₂ and Mg₂NiH₄.

The morphology and microstructure of Mg₂NiH₄ and as-oxidized 10 min CeH_2.73/CeO₂ composite (S5) are shown in Figure 6a. From the HRTEM images in Figure 6b, it could be found that the composite matrix is Mg₂NiH₄, which lattice fringe corresponds to the lattice plane of (111) and (220). Particle sizes of the CeO₂ nanoparticles are in a large range from 50–200 to 5–10 nm, which have been clearly shown in Figures 6a,d, respectively. Moreover, it could be seen that the in-situ generated CeO₂ particles are embedded homogeneously in the Mg₂NiH₄ matrix after ball milling. The interfaces between Mg₂NiH₄ and CeO₂ composite, as shown in Figures 6c,d, are contacted closely with each other. The homogeneously distributed catalysts can lead to high catalysis effect, and thus be beneficial for the hydrogenation/dehydrogenation properties of Mg₂NiH₄. Because of the too small amount of CeH_2.73, we could not clarify their microstructures in the TEM study.

The phase transition of the Mg₂NiH₄ + S5 sample during dehydrogenation was further studied by in situ XRD. The obtained in situ XRD patterns are shown in Figure 7. The XRD pattern of Mg₂NiH₄ + S5 sample at room temperature contains the diffraction peaks of Mg₂NiH₄ and CeO₂ because the XRD intensity of CeH_2.73 is too weak. The diffraction peaks of CeO₂ do not obviously change during the whole dehydrogenation process, indicating that the CeO₂ phase might act as a catalyst of Mg₂NiH₄ dehydrogenation. When the temperature was raised to 250°C, the diffraction peak of Mg₂NiH_0.3/Mg₂Ni phase appears, which means the dehydrogenation of Mg₂NiH₄ starts. As temperature increases, the intensity of Mg₂NiH_0.3/Mg₂Ni phase gradually enhances, and that of the Mg₂NiH₄ peaks decreases. As temperature reaches about 330°C, the diffraction peaks of CeO₂ remains unchanged and no MgO is found.

The in situ XRD patterns are then refined using Jade 6.0, and the relative contents of phases are calculated by the RIR method. At 270°C, the relative content of Mg₂NiH_0.3 phase is 47.1 wt%, and the relative content of Mg₂NiH₄ phase is 52.9 wt%. After heating to 290°C, the relative content of the two phases changes rapidly, and the relative content of Mg₂NiH_0.3 phase increases to 81.8 wt% and Mg₂NiH₄ The relative content of the phase was reduced to 18.2 wt%. Subsequently, the rate of phase change during the heating process was significantly slowed down. The relative content of Mg₂NiH_0.3 phase was 89.1 wt% at 400°C, and the relative content of Mg₂NiH₄ phase reduces to about 10.9 wt%.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.