For AB₅-type hydrogen storage alloys, the A-side elements are mainly rare-earth elements such as La, Ce, Pr, Nd, etc. and the B-side elements are mainly transitional metals such as Ni, Co, Mn, Al, Fe, Cu etc. The most basic AB₅-type alloy is LaNi₅, the crystal structure of which is shown in Figure 1 . Due to the limitation of the CaCu₅ lattice structure, the theoretical electrochemical hydrogen storage capacity of AB₅-type alloys is limited to 372 mAh·g⁻¹ [29] , and the actual discharge capacity is generally lower than 350 mAh·g⁻¹.

Elemental substitution is often applied to improve the electrochemical performance of AB₅-type hydrogen storage alloys. For example, for A-side elements, Fu et al. [23] partially substituted La for Ce in the La_1−xCe_xNi_4.5Al_0.5 (x = 0 - 0.4) alloys to improve the hydrogen absorption/desorption thermodynamic properties. Results showed that the maximum hydrogen storage capacity (C_max) of the alloys decreased with the increase of Ce content. The absolute values of the enthalpy and entropy changes of the alloys first increased and then decreased. When x = 0.2, the absolute values of the enthalpy change for hydrogen absorption and desorption reached the lowest which were 26.33 kJ·mol⁻¹ and 24.30 kJ·mol⁻¹, respectively. As Co is a high-cost element but indispensable to achieve good cycle life for AB₅-type alloys, many works have been carried out based on low-Co alloys. For example, Cheng et al. [30] partially substituted Y for Ce in the low-Co La_0.90−xCe_0.08Y_xZr_0.02Ni_3.91Co_0.14Mn_0.25Al_0.30 (x = 0 - 0.7) alloys which led to the formation of the Ce₂Ni₇ secondary phase in the original CaCu₅ single-phase structure. Thus, the C_max and hydrogen diffusion of the alloys were improved, and the corrosion resistance was enhanced, but the pulverization resistance was reduced. The x = 0.3 alloy exhibited excellent electrochemical performance, with a C_max of 350.4 mAh·g⁻¹, a capacity retention rate after 100 cycles (S₁₀₀) of 80.15% and a HRD₃₀₀₀ of 73.56%. Kazakov et al. [31] studied three low-Co alloys La_0.8Ce_0.2Ni₄Co_0.4Mn_0.3Al_0.3 (La-alloy), La_0.6Ce_0.2Nd_0.2Ni₄Co_0.4Mn_0.3Al_0.3 (Nd-alloy) and La_0.6Ce_0.2Nd_0.2Ni_3.8Co_0.4Mn_0.3Al_0.3Cr_0.2 (Cr-alloy). It was found that the La-alloy had the highest hydrogen storage capacity and electrochemical discharge capacity, which were 1.23 wt% and 321.1 mAh·g^–1, respectively. The partial substitutions of Nd for La and Cr for Ni improved the cycle life of the alloy electrodes, but slightly reduced the C_max, HRD and hydrogen diffusion kinetics. After 100 charge/discharge cycles at 1C, the capacity retention of the Nd-alloy was as high as 92.2%.

For the modification of B-side elements, Han et al. [32] studied the effects of adjusting the content of Cu and Be elements simultaneously of the La_0.6Ce_0.2Pr_0.05Nd_0.15Ni_3.55Co_0.75−xMn_0.4Al_0.3 (Cu_0.06Be_0.04)_x (x = 0 - 0.75) alloys. It was found that with the increase of Be-Cu content, the HRD of the alloys was improved, and the hydrogen storage capacity first increased and then decreased, with the peak value of 321.9 mAh·g^–1 at x = 0.45. Zhou et al. [33] [34] studied the effects of partial substitution of Mn for Ni in LaNi_5−xMn_x and La_0.78Ce_0.22Ni_4.4−xCo_0.60Mn_x alloys. Results showed that the discharge capacity and cycle stability of the alloys were slightly improved, but the equilibrium pressure was reduced, which had an adverse effect on low-temperature discharge capability and HRD performance. Xu et al. [35] . found that adding appropriate amount of Sn to LaNi₅ alloy could result in higher anisotropic c/a value, reducing the microstrain of the alloys and thus improving the cycle life. The S₁₀₀₀ of the LaNi_4.25Sn_0.75 alloy reached 95.8%. Zhou et al. [36] believe that Al is a very important element for AB₅-type alloys and they showed that although Al-free AB₅-type hydrogen storage alloys may have higher surface catalytic capacity, better low-temperature and HRD performance, the cycle life would be poor. Comparisons between different B-side elements based on the (LaCe)_1.0Ni_3.8(CoMn)_1.05Al_0.05M_0.1 (M = Ni, Al, Fe, Si, Sn, Cu) alloys showed that larger atomic radius led to higher hydrogen storage capacity and hydride thermodynamic stability, but was not favorable for low-temperature performance. The low-temperature discharge ability, HRD and peak power gradually decreased in the order of Ni > Si > Fe > Sn > Cu > Al, but the cycling stability was increased. The (LaCe)_1.0(NiCoMn)_4.85Al_0.05Cu_0.1 alloy exhibited the best overall electrochemical properties.

Stoichiometry modification is also applied to optimize the elemental composition of AB₅-type alloys. Chen et al. [37] studied the effects of over-stoichiometry based on a series of La_0.582Ce_0.191Zr_0.025Sm_0.202(Ni_0.849Co_0.032Mn_0.05Al_0.069)_5+x (x = 0 - 0.47) alloys and found that over-stoichiometry was beneficial to the cycle stability. The x = 0.35 alloy displayed a high S₁₀₀ of 96.71% enabled by the enhanced anti-pulverization resistance and weakened electrode polarization.

The characteristics of elemental compositions can be useful in the prediction of hydride thermodynamic properties. Prompted by the requirement of fast design of novel AB₅-type hydrogen storage alloys for industrialization, Panwar et al. [38] established a structural model of the correlation between the thermodynamic properties and structural parameters of the multi-element AB₅-type alloys. The model connects the atomic radius of the elements with the heat of formation to calculate the “hydride formation enthalpy” of the binary, ternary and multicomponent AB₅-type metal hydrides.

Different treatments such as rapid solidification, annealing etc. have also been used to improve the hydrogen storage properties of AB₅-type alloys. Yao et al. [39] studied the effects of rapid solidification on the structure and hydrogen storage performance of the LaNi_4.5Co_0.25Al_0.25 alloy. It was found that rapid solidification could improve the uniformity of the constituent elements and expand the crystal lattices, thus improving the activation properties, cycle stability and self-discharge performance, but the HRD was jeopardized. Zhu et al. and Kazakov et al. [40] [41] studied the effects of annealing and found that annealing usually do not change the CaCu₅-phase structure of AB₅-type alloys, but could improve the cycling stability. For example, the cycle life of the MlNi_4.57Co_0.17Mn_0.25Al_0.41Y_0.02 and La_0.8Ce_0.2Ni₄Co_0.4Mn_0.3Al_0.3 alloys was improved significantly prolonged by annealing. However, with the increase of eighter annealing temperature or holding time, the C_max and the HRD performance were decreased, and the impact of annealing temperature was stronger than that of annealing time.

Surface treatment improves the surface characteristics which ameliorate the hydrogen storage properties of AB₅-type alloys. Gan et al. [42] modified the surface of the MlNi_4.07Co_0.45Mn_0.38Al_0.31 alloy using the mixed solution of 12 mol·L⁻¹ NaOH + 0.5 mol∙L⁻¹ NH₄F + 0.1 mol·L⁻¹ KBH₄. After the modification, nano sticks were formed on the alloy surface. Al, Mn and Fe elements were partially dissolved, forming a Ni-rich surface layer. Resultantly, The HRD performance of the alloy was improved with a specific capacity of 120.6 mAh·g⁻¹ at the discharge current density of as high as 10 C. Hubkowska et al. [43] synthesized Pd nanoparticles of irregular shape and size on the surface of LaMmNi_4.1Al_0.3Mn_0.4Co_0.45 alloy by using microwave assisted polyol method which not only enhanced the hydrogen absorption/desorption kinetics performance but also spared the electrochemical activation process.

In addition, the degradation mechanism has also been explored to improve the cycle life of AB5-type hydrogen storage alloys. For example, Zhu et al. [44] studied the degradation progress of the AB₅-type LaNi_4.75Mn_0.25 alloys based on the structure and hydrogen absorption/desorption data during long-term cycling. It was found that in the temperature range of 343 - 383 K, the PCT curves of the alloys remained a single plateau within 1000 cycles, but the slope of the plateau became steeper and the hydrogen storage capacity decreased. Further study revealed that the degradation phenomenon was caused by the structural evolution, including pulverization and lattice damage. The internal driving force of the structural change was the microstrain generated during hydrogen absorption/desorption. However, the isotropic parameter c/a increased with cycling, reducing the hysteresis and slowing down the degradation process.

RE-Mg-Ni-based hydrogen storage alloys have super-stacking structures consisting of [A₂B₄] and [AB₅] subunits stacking along c-axis. The most common RE-Mg-Ni-based alloys are La-Mg-Ni-based alloys and the ratios between [A₂B₄] and [AB₅] subunits mainly include 1:1, 1:2 and 1:3, which correspond to (La, Mg)Ni₃, (La, Mg)₂Ni₇ and (La, Mg)₅Ni₁₉ phases, respectively. Each of these phases has two crystal types: hexagonal (2H-) type and rhombohedra (3R-) type. The typical superstacking modes are presented in Figure 2 . RE-Mg-Ni-based alloys have the advantages of both high capacity of AB₂-type alloys and easy activation and fast hydrogen absorption/desorption rate of AB₅-type alloys [28] . However, the overall hydrogen storage properties especially cycle life still need to be further improved for practical applications [45] . The cyclic degradation mechanism of RE-Mg-Ni-based alloys was explored based on a LaSmMgNi_4.1 alloy composed of (LaSm)MgNi₄ and LaNi₅ phases [46] . It was found that after absorption/desorbing cycling, the (LaSm)MgNi₄ main phase decomposed from crystalline state to (LaSm) hydride and MgNi-H amorphous state. Moreover, progressive cracking and pulverization of the alloy particles also led to the cycling degradation. To relieve the capacity degradation during cycling, extensive works have been carried out.

Mg is found to be a very important element for stabilizing the structure of RE-Mg-Ni based alloys. Chen et al. [47] have reported that the La_3-xMg_xNi₉ alloys are mainly formed from the bond interactions between La-Ni, Ni-Ni or/and M-Ni, and La-Ni which is the main factor controlling the structural stability of the alloys. Mg substitution for La could enhance the interaction of the La-Ni bond which increases the cycling stability of the alloys. However, high Mg content would also lead to the reduction of cell volume and decrease the hydrogen storage capacity. Wang et al. [48] found that appropriate amount of Mg could promote the formation of Gd₂Co₇ and Ce₂Ni₇ main phases of the La_1−xMg_xNi_3.4Al_0.1 (x = 0.1 - 0.4) alloys. The alloy with x = 0.2 showed good electrochemical performance with the C_max of 357.4 mAh·g^–1, HRD₁₂₀₀ of 60.1% and S₁₀₀ of 74.5%. Moreover, the fine LaNi₅ structure dispersed in the (La,Mg)₂Ni₇ matrix phase could promote the diffusion of hydrogen atoms, thus enhancing the electrochemical properties of the alloys. Dong et al. [49] prepared a series of single-phase La_1−xMg_xNi_2.5Co_0.5 (x = 0 - 0.4) alloys with PuNi₃-type phase structure by annealing and studied the effects of Mg content. It was found that with hydrogen absorption, the hydrides gradually transformed from amorphous state to PuNi₃-type crystal, and the lattice parameters a and c significantly increased, leading to the expansion of cell volume. But the expansion degree decreased with the increase of Mg content which benefits the cycling stability of the alloys.

In addition to Mg, the effects of other A-side elements have also been studied. For example, Zhou et al. [50] partially substituted La with mixed rare earth (MM) for the La_0.8-xMM_xMg_0.2Ni_3.1Co_0.3Al_0.1 (x = 0 - 0.3) alloys, and found that the addition of MM could lead to more uniform phase distribution and promote the formation of the La₂Ni₇ phase with good electrochemical properties. The alloy with x = 0.3 achieved the C_max of 381.2 mAh·g^–1 and HRD₁₅₀₀ of 60.55%. Xu et al. [51] found that the addition of Ce in the La_0.8-xCe_xMg_0.2Ni_3.8 (x = 0 - 0.5) alloys promoted the formation of the A₅B₁₉ phase. The activation property, capacity retention and HRD of alloy electrodes were all enhanced with the increase of Ce content, and the S₁₀₀ reached 86.17% when x = 0.5. Huang et al. [52] studied the effects of Pr substitution for La of the La_0.7−xPr_xZr_0.1Mg_0.2Ni_2.75Co_0.45Fe_0.1Al_0.2 (x = 0 - 0.20) alloys and found that the anti-corrosion effect of Pr enabled the enhancement of the alloys’ cycling stability which S₂₀₀ reached 75.1% for the x = 0.2 alloy. But the cell parameters of the alloy phases decreased, which hindered the diffusion of hydrogen atoms, decreasing the dynamics property. Liu et al. [53] proposed a new strategy to improve the cyclic stability of superlattice hydrogen storage alloys by enhancing the structural stability and oxidation/corrosion resistance by A-side elemental modification. Specifically, they equalized the volumes of [LaMgNi₄] and [LaNi₅] subunits taking the advantages of site occupation tendency of Gd in the La_0.75−xGd_xMg_0.25Ni_3.5 (x = 0 - 0.15) alloys, so as to maintain a stable crystal structure during cycling. In addition, the high electronegativity of Gd element enhanced the oxidation/corrosion resistance of the alloys and the S₁₀₀ of the x = 0.15 alloy reached 88.2%.

For B-side elements, Li et al. [54] studied the effects of partial substitution of Zn for Ni in the La₂Mg(Ni_1−xZn_x)₉ (x = 0.1 - 0.2) alloys, and found that appropriate amount of Zn could significantly improve the cycle stability of the alloys due to the formation of the dense passivation film on the alloy surface. The S₁₀₀ of the La₂Mg(Ni_0.9 Zn_0.1)₉ alloy reached as high as 94%. Wang et al. [55] reported that the Al-substituted La_0.60Nd_0.15Mg_0.25Ni_3.20Al_0.10 alloy exhibited good cycle stability, and the Mn-substituted La_0.60Nd_0.15Mg_0.25Ni_3.20Mn_0.10 alloy was with increased discharge capacity, reduced plateau pressure and enhanced HRD performance. Moreover, Fan et al. [56] studied the influence of non-stoichiometric ratio on the phase composition and electrochemical properties of the of the A₅B₁₉-A₂B₇ double-phase La₄MgNi_x (x = 16, 17, 18) alloys and found that the increase of x value contributed to the formation of A₅B₁₉ phase and the x = 17 alloy was with the highest discharge capacity (388.8 mAh·g^–1) and the best cycle life (S₁₀₀ = 90.1%).

Doping of elements is also companied by the change of elemental composition, but the alloys’ properties are also influenced by the doping method. For example, Zhang et al. [57] doped pure Ni to the LaMgNi alloy by ball milling which not only promoted the amorphization process of the alloy particles but also enhanced the catalytic effect. The hydrogen absorption/desorption kinetics of the Ni-doped LaMgNi + wt%Ni (x = 100, 200) alloys was significantly improved. With increasing ball milling time, the gaseous hydrogen absorption capacity and kinetics of the alloys first increased and then decreased, but the hydrogen desorption kinetics was always increased. Further studies showed that both increasing nickel content and prolonging the ball milling time could significantly reduce the activation energy of the alloy for hydrogen desorption. Zhang et al. [58] synthesized a Y₂O₃@La_1.7Mg_1.3Ni₉ composite by ball milling, and found that Y₂O₃ was coated on the alloy surface which was mostly in amorphous state. The modified alloy exhibited better corrosion resistance, electrochemical hydrogen storage capacity, and cycling stability.

Treatment methods such as rapid quenching and annealing have also been widely investigated to improve the hydrogen storage properties of RE-Mg-Ni-based alloys. For example, Hu et al. and Luo et al. [21] [59] compared the as-cast alloy and those after rapid quenching. Results showed that the as-cast alloy was composed of multiphase structure, while those after rapid quenching contained a large number of amorphous nanocrystalline structure whose content increased with the increase of solidification rate. Rapid quenching could reduce the activation energy of the alloy surface and enhance the diffusion ability of hydrogen atoms, so as to improve the electrochemical kinetics performance of the alloys. Li et al. [60] studied the microstructure evolution of the La_2-xMg_xNi₇ (x = 0.3, 0.4) alloys after annealing treatment and found that the characters of the annealed microstructure originated from that of the as-cast microstructure, indicating the heredity effect between the as-cast and annealed microstructure. Moreover, the abundance of the (La, Mg)₂Ni₇ main phase increased in sacrifice of the (La, Mg)Ni₂ and LaNi₅ phases after annealing. Chen et al. [61] reported that annealing unified the elemental distribution of the LaMgNi_3.9Mn_0.2 alloy whose structure was mainly columnar crystal. They also found heredity effect for the annealed alloy from the as-cast one. Moreover, with the increase of annealing time, the LaMg(NiMn)₄ phase content decreased and the (La, Mg)(NiMn)₅ phase content increased. Deng et al. [62] demonstrated that annealing treatment improved the cycling stability of the (Pr_0.1Nd_0.1Y_0.6Sm_0.1Gd_0.1)_0.2Mg_0.17Ni_3.1Co_0.3Al_0.1 alloy and the S₁₀₀ increased from 69% of the as-cast alloy to 83.5% of the annealed alloy. Moreover, they also found that the controlling factor of the HRD altered from hydrogen diffusion to surface charge transfer. Jiao et al. [63] found that annealing could eliminate the AB₅ secondary phase and increase the total content of superlattice phases in the La_0.75Mg_0.25Ni_3.5 alloy, thus improving the discharge capacity. Meanwhile, the homogenization and stress reduction of the (La, Mg)₂Ni₇ and (La, Mg)₅Ni₁₉ superlattice phases of the annealed alloy could relive the pulverization and oxidation of the alloy particles during charge/discharge cycling, and significantly improve the cycling stability of the alloy electrodes. The C_max of La_0.75Mg_0.25Ni_3.5 alloy electrode was 367 mAh·g^–1, and the S₁₀₀ was 81.47%. However, due to the decrease of defects and grain boundaries, the HRD performance of the alloy was decreased. Similarly, Young et al. [64] also reported that annealing increased the abundance of the A₂B₇ main phase and decreased the that of the AB₅ secondary phase of the (Nd_0.87Mg_0.12Zr_0.01) (Ni_0.952Al_0.046Co_0.002)_3.74 alloy. Moreover, the cell volume of the main phase became smaller, and the elemental distribution became more uniform. Both the hydrogen storage capacity and the electrochemical discharge capacity of the alloy were improved by annealing, and the slope factor and the hysteresis were also reduced which enhanced the activation, HRD, and cycle life of the alloy. They also found that with the increase of annealing temperature, the AB₅ phase abundance decreased, which further improved the hydrogen storage capacity and cycle life of the alloy. Increasing the annealing time led to the increase of the A₂B₇ phase abundance, which improved the discharge capacity and cycle life of the alloy, but the HRD was reduced due to the decrease of AB₅ catalytic phase.

La-Y-Ni-based hydrogen storage alloys have similar super-stacking structures to La-Mg-Ni-based alloys. The stoichiometries of this alloy system include AB₃-type, A₂B₇-type and A₅B₁₉-type. The preparation process of La-Y-Ni-based alloys is simpler than that of La-Mg-Ni-based alloys enabled by the possible exclusion of Mg element with high vapor pressure. He et al. [65] studied the phase transition process and hydrogen storage properties of different types of La-Y-Ni-based alloys with various [AB₅]/[A₂B₄] ratios. They reported that the transformation process could be summarized as A₂B₄ + AB₅ → AB₃ → A₂B₇ → A₅B₁₉. Usually, with the increase of the [AB₅]/[A₂B₄] ratio, the hydrogen absorption plateau pressure increases (LaY₂Ni₉ < La₂Y₄Ni₂₁ < La₅Y₁₀Ni₅₇), the stability of the corresponding hydrides decreases, the maximum discharge capacity decreases, and the cycling stability increases. Among the LaY₂Ni₉, La₂Y₄Ni₂₁ and La₅Y₁₀Ni₅₇ alloys, the La₂Y₄Ni₂₁ alloy exhibited the best HRD performance, and the solid hydrogen storage capacity was also high (1.59 wt% at 313 K). Yan et al. [66] also compared different types of LaY₂Ni_8.2Mn_0.5Al_0.3 (AB₃), LaY₂Ni_9.7Mn_0.5Al_0.3 (A₂B₇) and LaY₂Ni_10.6Mn_0.5A1_0.3 (A₅B₁₉) alloys and found that under the same temperature, the alloys’ plateau pressure was in the order of A₅B₁₉ > A₂B₇ > AB₃. The highest hydrogen storage capacity was obtained for the AB₃-type LaY₂Ni_9.7Mn_0.5Al_0.3 alloy which was 1.48 wt% at 313K, and the maximum discharge capacity of the alloy electrode was 385.7 mAh·g^–1. All the three alloys showed good cycling stability. The S₃₀₀ was above 74%, but the HRD performance was lower than that of AB₅-type alloys.

Elemental composition has significant influence on the structure and hydrogen storage properties of La-Y-Ni-based alloys. Y for La-Y-Ni-based alloys resembles Mg for La-Mg-Ni-based alloys and the amount of Y can have great impact on the comprehensive electrochemical performance. Zhao et al. [67] found that the main phase of La_1−xY_xNi_3.25Mn_0.15Al_0.1 (x = 0 - 1) alloy was of Ce₂Ni₇ phase. With the increase of Y content, the Ce₂Ni₇ phase abundance first increased and then decreased. Moreover, the higher the Ce₂Ni₇ phase abundance was, the higher the HRD₉₀₀ value was. When x = 0 - 0.25, the alloys had no plateau in the PCT curves and were easy to become amorphous during hydrogen absorption/desorption. When x ≥ 0.50, hydrogen-induced amorphization was effectively inhibited and a single hydrogen absorption/discharge plateau was obtained. The pressure range of the hydrogen absorption plateau was 0.026 - 0.097 MPa, and the maximum hydrogen storage capacity was 1.418 - 1.48 wt%. Liu et al. [68] found that the capacity and HRD performance of the La_3−xY_xNi_9.7Mn_0.5Al_0.3 (x = 1 - 2.5) alloys also increased with the increase of Y content. The highest C_max and HRD₁₂₀₀ reached 383.8 mAh·g^–1 and 83.67%, respectively at x = 2.5. Guo et al. [69] compared the effect of Y and Mg in the AB₃-type LaY_2−xMg_xNi₉ (x  =  0 - 1.00) La-Y-Ni-based alloys, and found that the addition of Mg decreased the LaY₂Ni₉ phase abundance while increased the (La, Y)₂Ni₇ phase abundance, which improved the alloys’ cycling stability by slowing down pulverization, but reduced the maximum discharge capacity. TEM results further verified that Mg substitution could inhibit the hydrogen-induced amorphization, benefiting the cycling stability of the alloys. The LaY_1.25Mg_0.75Ni₉ alloy showed good overall electrochemical performance with the C_max of 308.4 mAh·g⁻¹ and S₁₀₀ of 93.7%.

The effects of A-side rare-earth elements have also been widely studied for La-Y-Ni-based alloys. Zhao et al. [70] reported that the addition of Nd in the La-Y-Ni-based La_0.4−xNd_xY_0.6Ni_3.52Mn_0.18Al_0.1 (x = 0 - 0.4) alloys could improve the cycling stability and HRD performance. Nd enables the formation of small and dense rare-earth oxide on the alloy surface which inhibits the dissolution of active materials such as Y, Mn and Al, and thus significantly improves the cycling stability of the alloy electrodes. The x = 0.4 alloy showed the best electrochemical performance with the C_max of 382.8 mAh·g⁻¹, S₁₀₀ of 93.7%, and HRD₉₀₀ of 87.8%. By comparing the LaSm_0.3Y_1.7Ni_9.7Mn_0.5Al_0.3 and LaY₂Ni_9.7Mn_0.5Al_0.3 alloys, Guo et al. [71] found that partial substitution of Sm for Y could enhance the corrosion resistance and anti-pulverization ability of the alloys, resulting in the improvement of cycle life. Yang et al. [72] compared the microstructure and electrochemical properties of the La_0.63X_0.2Mg_0.17Ni_3.1Co_0.3Al_0.1 (x = Pr, Nd, Y, Sm and Gd) alloys. It was found that the x = Y alloy contained the highest amount of Ce₂Ni₇ phase (93.3 wt%) and superior electrochemical discharge capacity (404.4 mAh·g⁻¹) and capacity retention (S₁₀₀ = 93.50%).

For the effects of B-side elements on the structure and hydrogen storage properties of La-Y-Ni-based alloys, Al and Mn are the mostly studied. For example, Li et al. [73] found that Al addition could improve the cycle life of the LaY_1.9Ni_10.2−xAl_xMn_0.5 (x = 0 - 0.6) alloys but the function was limited as Al could not inhibit the pulverization of the alloy powder. Xiong et al. [74] found that partial substitution of Mn for Ni in the LaY₂Ni_10.5−xMn_x (x = 0 - 2.0) alloys increased the Ce₂Ni₇ phase abundance and expanded the cell volumes of the main phases, leading to the decrease in the hydrogen absorption/desorption equilibrium pressure. The C_max and HRD first increased and then decreased, and the x = 0.5 alloy exhibited the highest hydrogen storage capacity of 1.40 wt% with the corresponding discharge capacity of 392.9 mAh·g^–1. However, Yu et al. found that the increase of Mn content of the La_5.42Y_18.50Ni_76.08−xMn_x (x = 0 - 6) alloys reduced the Y₂Ni₇ and La₂Ni₇ phase contents while increased the LaY₂Ni₉ phase abundance, inhibiting the hydrogen induced amorphization and increasing the C_max of the alloys. Zhou et al. [75] found that with the increase of Mn content, the Ce₂Ni₇ phase abundance of the La_1.3Ce_0.5Y_4.2Ni_20−xMn_xAl (x = 0 - 0.7) alloys increased, leading to the increase of C_max. However, the decrease of the Gd₂Co₇ phase resulted in the decrease of the low-temperature discharge capacity from 286.6 mAh·g^–1 (x = 0) to 64.7 mAh·g^–1 (x = 0.7) at –30˚C. In addition, with the increase of Mn content, the kinetics performance of the alloys was also decreased significantly.

Annealing treatment is a very effective method to optimize the hydrogen storage performance of La-Y-Ni-based alloys and thus has been widely studied. It is generally recognized that long-time annealing promotes microstructure homogenization, increases the main phase abundance, and thus improves the alloys’ hydrogen storage performance [76] - [80] . For example, Wang et al. found that annealing increased the content of the Ce₂Ni₇ main phase of the La_0.33Y_0.67Ni_3.25Mn_0.15Al_0.1 alloy [76] . Li et al. [77] found that annealing the La_0.4Y_0.6Ni_3.52Mn_0.18Al_0.1 alloy at 1173 - 1373 K increased the Ce₅Co₁₉ phase abundance which improved the discharge capacity, HRD performance and cycling stability of the alloy electrode. When the annealing temperature was 1273 K, the phase abundance of the Ce₅Co₁₉ phase reached the highest so as the electrochemical performance of the alloy. Guo et al. [78] found that with the increase of annealing temperature (1073 - 1373 K), the C_max, HRD performance and cycling stability of the LaY₂Ni_9.7Mn_0.5Al_0.3 alloy electrode first increased and then decreased which was consistent with that of the main Ce₂Ni₇ phase abundance. Li et al. [79] found that as the annealing temperature increased from 1173 K to 1373 K, the Ce₅Co₁₉ phase abundance of the La_0.35Y_0.65Ni_3.5Mn_0.2Al_0.1 alloy first increased and then decreased. The HRD₉₀₀ and D₀ also showed a positive correlation with the Ce₅Co₁₉ phase abundance. Moreover, when the annealing temperature was 1273 K, the corrosion current density was the least, and the cycle stability was the best. Zhou et al. [80] found that with the increase of annealing temperature (1098 - 1323 K), the low temperature performance of the La_1.9Y_4.1Ni_20.8Mn_0.2Al alloy first increased and then decreased. At the annealing temperature of 1148 K, the alloy obtained modified H₂ channel, high dehydrogenation plateau pressure, good oxidation/corrosion resistance and good kinetics characteristics. The C_max at low temperature (243 K) was 298.6 mAh·g^–1, and the S₁₀₀ reached 80.5%.

Through long-time annealing at appropriate temperatures, single-phase La-Y-Ni-based alloys could be obtained. Zhao et al. [81] prepared a single-phase LaY₂Ni_10.5 alloy with a rhombohedral Gd₂Co₇ structure by annealing at 1000˚C and a hexagonal Ce₂Ni₇ structure by annealing at 1100˚C. A double Gd₂Co₇-Ce₂Ni₇ phase structure was obtained when the temperature was between 1000˚C and 1100˚C. In other words, the rhombohedral structure is more stable than the hexagonal structure at low temperature. Based on the single-phase alloys, further study was carried out which showed that Y atom preferentially replaces La in the [A₂B₄] subunit of the hexagonal and rhombohedral superlattice structures. Compared with the rhombohedral phase, the hexagonal phase has a relatively small [A₂B₄] subunit volume, and thus showed higher structural stability during the hydrogen absorption/desorption cycling. Compared with multiphase alloys, single-phase alloys displayed better capacity, rate performance and cycle stability. The same group also obtained a single-phase La_0.33Y_0.67Ni_3.25Mn_0.15Al_0.1 alloy with a 2H-Pr₅Co₁₉ structure by annealing the alloy at 1100˚C [82] . They reported that Y preferentially occupies the [A₂B₄] subunit, Mn preferentially occupies the [AB₅] subunit, and Al locates at the boundaries between [A₂B₄] and [AB₅] subunits. Reduced volume mismatch between [AB₅]-1, [AB₅]-2 and [A₂B₄] subunits could improve the electrochemical performance of the alloys. The C_max was 368.8 mAh·g^–1, the S₂₀₀ was 73.4%, and the HRD₁₅₀₀ was 58.6% for the single-phase La_0.33Y_0.67Ni_3.25Mn_0.15Al_0.1 alloy. Xing et al. [83] compared the alloy treated with annealing with rapid quenching, and found that annealing treatment improved the C_max and HRD performance, while decreased the gaseous hydrogen absorption/desorption cycling stability. The rapid quenched alloy exhibited excellent cycling stability due to its good anti-pulverization resistance. After 30 cycles of gaseous hydrogen absorption/desorption, the particle size retention rate the rapid quenched La₄MgNi₁₉ alloy reached 98.6%.

Moreover, considering that carbon has good electrocatalytic activity, electrical conductivity and corrosion resistance, Wu et al. [84] coated the (LaSmY) (NiMnAl)_3.5 alloy with different contents (0.1 wt% - 1.0 wt%) of nano-carbons by low-temperature sintering with sucrose as the carbon source. Results showed that the C_max, HRD and cycle stability of the alloy electrodes first increased and then decreased with the increase of carbon content. The electrochemical performance of the alloy electrode with carbon content of 0.3 wt% was the best with the C_max of 359.0 mAh·g^–1, the S₃₀₀ of 80.01%, and the HRD₁₂₀₀ of 74.39%. The kinetics results showed that carbon coating could improve the electrocatalytic activity and conductivity of the alloy electrodes

Typical Zr-based AB₂-type hydrogen storage alloys include Zr-V, Zr-Cr, Zr-Ni and Zr-Mn alloys. Among these types of alloys, ZrCr2, ZrMn2, and ZrNi2 have MgZn2-type structure (hexagonal structure), and ZrV2 has MgCu2-type structure (Cubic structure) as shown in Figure 4(a) , Figure 4(b) . Zr-V-based AB₂-type alloys usually have fast hydrogen absorption/desorption rate, but are difficult to be prepared and have high hydrogen desorption residue; Zr-Cr-based alloys can form stable hydride and have long cycle life, but are difficult to be activated. Zr-Ni-based alloys have high hydrogen storage capacity (2.0 wt% H2) and a stable structure, but are poor in hydrogen absorption/desorption reversibility. Zr-Mn-based alloys can absorb hydrogen under the pressure below 1 bar and room temperature, but are with high cost [8] [119] .

Aiming at the above problems, many studies have been carried to improve the overall hydrogen storage properties of Zr-based AB₂-type alloys, among which elemental modification is one of the most common and effective methods. For example, for the modification of A-side elements, Matsuyama et al. [120] studied the effects of partial substitution of Ti for Zr in the Zr_1−xTi_xNi alloys (0.05 ≤ x ≤ 0.5). The alloys exhibited single discharge plateau for 0 ≤ x ≤ 0.2 while double discharge plateaus for x ≥ 0.3. The discharge plateau potentials shifted towards negative value with increasing Ti content. The HRD and cycling performance of the alloys were improved with higher Ti content. Wan et al. [121] studied the Zr_1−xTi_xLa_0.03Ni_1.2Mn_0.7V_0.12Fe_0.12 (x = 0.12 - 0.22) alloys, and found that the x = 0.12 alloy exhibited excellent reversible discharge capacity of 466 mAh·g⁻¹, and the x = 0.22 alloy had good cycling stability and large current discharge capability with a S₅₀₀ of 71% and a HRD₄₀₀ of 71%. They also studied the function of rare-earth elements on Zr-based Ti_0.2Zr_0.8La_0–0.05Ni_1.2Mn_0.7V_0.12Fe_0.12 alloys [122] . La could improve the activation performance of the alloys due to the catalytic effect of the LaNi hydride, but it also caused a decrease in the discharge capacity and cycling stability. The Ti_0.2Zr_0.8La_0.03Ni_1.2Mn_0.7V_0.12Fe_0.12 alloy exhibited a C_max of 420 mA·g⁻¹, a HRD_0.71 of 79 % and a S₅₀₀ of 63%. Leng et al. [123] tried to change the ratio of Zr content from x = 3 to x = 9 for the Zr_xV₅Fe (x = 3 - 9) alloys and found that with the increase of Zr content, the α-Zr phase abundance increased and the C15-ZrV₂ phase abundance decreased with the decrease of the hydrogen absorption plateau pressure. The C15-ZrV₂ phase in the Zr₇V₅Fe alloy displayed the lowest hydrogen absorption plateau pressure at room temperature, and the hydrogen absorption kinetics curves at 623 K indicated that the Zr₇V₅Fe alloy with the smallest average particle size and the largest phase boundary area exhibited the fastest hydrogen absorption reaction kinetics.

For the modification of B-side elements, Yao et al. [124] studied the partial substitution of V for Mn in the ZrMn_2−xV_x (x = 0 - 0.8) alloys and found that the lattice parameters of the ZrMn₂ phase increased with increasing V content, resulting in a decrease in the dehydrogenation equilibrium pressure from 2.931 bar (ZrMn₂) to 0.080 bar (ZrMn_1.2V_0.6) at 200˚C, with a corresponding enthalpy change increasing from 43.29 kJ·mol⁻¹ H₂ to 60.38 kJ·mol⁻¹. The ZrMn_1.2V_0.6 alloy exhibited excellent cycling stability with a stable hydrogen storge capacity of 1.63 wt% during 20 cycles. Wu et al. [125] studied the effects of partial substitution of V for Ni in the ZrMgNi_4−xV_x (x = 0 - 2) alloys and found that the reversible hydrogen storage capacity gradually increased with increasing V content. Under 4 MPa H₂ pressure and 300 K, the ZrMgNi₂V₂ alloy absorbed 1.8 wt% H₂ in about 2 h without complicated activation process and reversibly desorbed the hydrogen in about 30 min at 473 K. Erika et al. [126] partially substituted Mo for Cr in the ZrCr_1-xNiMo_x (x = 0 - 0.6) alloys and found that Mo could increase the Ni content in the secondary phase (Zr₇Ni₁₀ relative to Zr₉Ni₁₁), which improved the HRD performance. Luo et al. [127] reported that partial substitution Mo for Co in the ZrCo_1−xMo_x (x = 0 - 0.2) alloys could significantly improve the initial activation behavior and the anti-disproportionation property, but the hydrogen storage capacity was decreased from 1.893 wt% (x = 0) to 1.66 wt% (x = 0.2), the plateau region of the P-C-T curves was shortened and the hydrogen desorption equilibrium pressure was decreased. Wu et al. [128] studied the effects of Ni addition in the Zr(V_1−xNi_x)₂ (x = 0.02 - 0.25) alloys and reported that with the increase of Ni content, the hydrogen absorption capacity decreased and the equilibrium pressure increased. The Zr(V_0.05Ni_0.95)₂ alloy exhibited good cycling stability at 823 K. Tu et al. [129] partially substituted Mn with Fe in the ZrFe_1.95−xMn_xV_0.10 (x = 0 - 0.15) alloys and found that the increase of Mn content improved the activation properties of the alloys and reduced the hysteresis, but the plateau slope increased and the hydrogen storage capacity first increased and then decreased. At 243 K, the ZrFe_1.85Mn_0.10V_0.10 alloy displayed the highest hydrogen storage capacity of 1.69 wt% and the fastest hydrogen absorption rate with the t_0.9 of 46 s. Ai et al. [130] studied the Zr_56.97V_35.85Fe_7.18−xCr_x (x = 0, 3.59, 7.18) alloys and revealed that with the increase of Cr content, the diffusion activation energy was reduced due to the increase of the cell volume of the C15-ZrV₂ phase. The hydrogen absorption capacity of the alloy was increased and the hydrogen absorption kinetics performance improved. The t_0.9 decreased from 90 s (x = 0) to 25 s (x = 7.18). Qin et al. [131] investigated the functions of Al addition in the ZrFe_2−xAl_x (x = 0.1, 0.2) alloys and found that the alloys maintained high hydrogen storage capacity due to the presence of the Zr₂Fe impurity phase, which was inconsistent with a previous study that Al alloying caused a sharp decline in the hydrogen storage capacity.

Some studies also explored the phase function on the hydrogen storage properties of Zr-based AB₂-type alloys. Yao et al. [132] reported that the formation of metastable phases in the phase transition reaction of homogeneous structure could optimize the cyclic stability of this alloy system. It was found that the phase transformation of the Zr_0.8Ti_0.2Co alloy during hydrogen absorption/desorption was from Zr_0.8Ti_0.2Co₄Zr_0.8Ti_0.21CoH₃ (heterogeneous structural phase transition) to Zr_0.8Ti_0.2Co₄Zr_0.8Ti_0.2CoH_1.4 (homogeneous structural phase transition). Therefore, the stable capacity was provided by the metastable phase Zr_0.8Ti_0.2CoH_1.4. By studying the co-substitution of Nb and Ni the in Zr_1−xNb_xCo_1−yNi_y (x, y = 0 - 0.3) alloys, Yao et al. [133] proposed that the phase transition reaction of the thermodynamically homogeneous structure is an effective way to avoid disproportionation. Among them, the Zr_0.8Nb_0.2Co_0.8Ni_0.2 alloy exhibited an ultra-long cycle life with the S₁₀₀ of 97.6%, as well as a high hydrogen storage capacity of 2.42 H (f.u.), which was the result of the synergistic effect of the homogeneous structure phase transition and the H-ordered migration mechanism (lamellar and linear de-embedding). Yu et al. [134] found that the main hydrogen-absorbing phases of the Zr₅₇V₃₆Fe₇ alloy were ZrV₂ and α-Zr phases; The hydride was with high stability, resulting in the difficulty for hydrogen desorption. But the alloy was with good pulverization resistance and air poisoning resistance. After being exposed in air for 2 h, both the hydrogen absorption rate and the hydrogen absorption capacity were decreased after 3 cycles. However, both the hydrogen absorption rate and the hydrogen absorption capacity could be largely recovered after reactivation for 1 h by pumping at 450˚C. It has also been pointed out that the electrochemical properties of AB₂-type alloys depend not only on the phase composition but also on the amount of a certain specific phase [122] . By studying the ZrNi_1.2Mn_0.5Cr_0.2V_0.1 alloy treated with different cooling rates, Solonin et al. [122] reported that the alloys with high content of Zr₇Ni₁₀ phase could be activated faster, those with high C15 and C14 phase content displayed higher maximum discharge capacity, while those with lower Zr₇Ni₁₀ phase content showed better cycling stability.

Treatments such as annealing, melt-spun, rapid solidification are also very effective to improve the hydrogen storage properties of Zr-based AB₂-type alloys [135] - [139] . For example, Wan et al. [139] reported that annealing treatment increased the maximum discharge capacity of the Zr_0.76Ti_0.24Ni_1.1Mn_0.7V_0.2 alloy from 350 mAh·g⁻¹ to 400 mAh·g⁻¹ and also improved the cycling stability and HRD performance. Lee et al. [136] also found that the ZrV_0.7Mn_0.5Ni_1.2 alloy annealed at 1000˚C for 12 h was with higher discharge capacity and better HRD performance than the as-cast alloy. The improved high-rate discharge performance was due to the increase of Ni content in the matrix phase. Zhang et al. [137] compared the effects of annealing and melt-spun and found that the annealed Zr_0.9Ti_0.4V_1.7 alloy had a high hydrogen storage capacity of 2.83 wt% and fast hydrogen absorption kinetics after one cycle of activation, while the alloy after melt-spun performed better in the initial hydrogen absorption rate due to the increased alloy surface area and the grain refinement. But the amount of hydrogen storage was smaller. Luo et al. [138] compared the effects of different preparation processes on the hydrogen storage properties of the Zr-based Zr₇V₅Fe alloy. It was found that the annealed alloy had the lowest plateau pressure of about 0.02 Pa and the highest hydrogen storage capacity of about 1.4 wt% at 623 K; the melt-spun alloy and the melt-spun + annealed alloy could be activated without incubation, exhibiting excellent activation property; The melt-spun + annealed treated Zr₇V₅Fe alloy had the best hydrogen absorption kinetics property. Wijayanti et al. [139] studied the effect of rapid solidification process on the Ti_0.15Zr_0.85La_0.03V_0.12Mn_0.7Fe_0.12Ni_1.2 alloy. It was found that the grain size of the alloy decreased from the initial 3.5 mm to 250 nm with the increase of the cooling rate. The alloy with cooling rate of 16.5 Hz exhibited the highest discharge capacity of 414 mAh·g⁻¹ due to its optimal two-phase structure containing C15 and C14 phases.

Surface coating is an effective method to improve the anti-poisoning ability of the Zr-based AB₂-type alloys. For example, Zhang et al. [140] found that homogeneous deposition of the Pd-Ag coating on the surface of the ZrV₂, Zr_0.9Ti_0.1V₂ and Zr₅₇V₃₆Fe₇Zr alloys improved the anti-poisoning ability of the alloys in a wide temperature range. the Pd-Ag coating also resulted in the easy activation and accelerated the hydrogenation kinetics of the alloys.

Ti-based AB₂-type hydrogen storage alloys mainly include Ti-Mn, Ti-Cr, and Ti-V alloys. Among these alloys, TiMn₂ and TiCr₂ have MgZn₂-type structure (hexagonal structure), and TiV₂ has MgCu₂-type structure (Cubic structure) as shown in Figure 5 . The hydrogen storage capacity of Ti-based AB₂-type alloys can reach 2.6 wt% under atmospheric pressure and room temperature. Among them, the Ti-Mn alloy can be readily activated under room temperature, but the alloy particles are easily pulverized and thus the cycle stability is poor. The Ti-Cr alloys can also absorb hydrogen readily but has some shortcomings such as high plateau pressure, low hydrogen absorption capacity, slow heat transfer rate and large hysteresis. The Ti-V alloys have high hydrogen storage capacity but poor activation and dehydrogenation properties [26] . Focusing on Ti-based AB₂-type hydrogen storage alloys, various modifications have been carried out including elemental subsitution, annealing, ball milling etc.

For elemental substitution, Han et al. [141] used Zr to partial substitute for Ti in the Ti_7−xZr_xV₅Fe (x = 0 - 2.1) alloys which contained α-Zr and C15-ZrV₂ phases. Withthe increase of Ti content, the C15-ZrV₂ phase abundance first increased and then decreased, while the opposite was true for the α-Zr phase; The plateau pressure of the α-Zr phase increased with the increase of Ti content, while that of the C15-ZrV₂ phase showed the opposite trend. Therefore, the stoichiometric ratio of A/B might play a key role in determining the plateau pressure of the two phases. Zhang et al. [142] found that Zr could refine the nano-eutectic texture of the Ti₄₀Zr_60−xV_x (x = 20 - 30) alloys with ultra-fine nano-eutectic structures of 50 - 500 nm between lamellar layers. The alloys exhibited excellent activation and hydrogenation properties but low reversible hydrogen storage capacity. The Ti₄₀V₃₅Zr₂₅ alloy achieved the highest hydrogen absorption capacity of 2.4 wt% within 10 min under1 MPa H₂ and 200˚C. Cao et al. [142] studied the effects of Mn on the Ti_0.85Zr_0.17Cr_1.2-xMn_xFe_0.7V_0.1 (x = 0 - 0.3) alloys, and it was found that the plateau pressure and hydrogen storage capacity of the alloys increased with increasing Mn content. Moreover, they also found that the effects of Cr addition in the Ti_0.85Zr_0.17Cr_1.0+yMn_0.2Fe_0.7V_0.1−y (y = 0 - 0.10) alloys had similar effects to those of Mn except that the hydrogen storage capacity decreased with decreased slightly. The hydrogen absorption plateau pressure of the Ti_0.85Zr_0.17Cr_1.1Mn_0.2Fe_0.7 alloy was 5.08 MPa under 293 K and the desorption pressure was 24.90 MPa under 363 K. Yan et al. [143] studied the effects of partial substitution of Mo for Cr in the (Ti_0.85Zr_0.15)_1.1Cr_1−xMo_xMn (x = 0.05 - 0.2) alloys. It was found that with the increase of Mo content, both the hydrogen absorption and desorption capacities decreased, and the plateau slope increased. The maximum hydrogen absorption capacity and desorption capacity was obtained for the (Ti_0.85Zr_0.15)_1.1Cr_0.95Mo_0.05Mn alloy which were 1.76 wt% and 1.09 wt%, respectively. Nygard et al. [144] substituted Fe for both Ti and V for the (Ti_0.7V_0.3)_1−zFe_z (z = 0 - 0.3) alloys and found that the increase of Fe content reduced the enthalpy and activation energy contributing to the fast kinetics property. Pineda et al. [145] reported that the addition of Al in the Al_x(TiVNb)_1−x (x = 0.05 - 0.25) alloys decreased the lattice parameters thus reducing the hydrogen storage capacity. The x = 0.05 alloy showed relatively high hydrogen storage capacity (2.96 wt%) and its hydride was less stable. Compared with the ternary TiVNb alloy, it had lower initial hydrogen desorption temperature of about 100˚C and the enhanced reversible hydrogen storage capacity of 1.76 H/M (2.83wt %). Zhou et al. [146] studied a series of low-cost and low-V Ti_0.95Zr_0.05Mn_0.9+xCr_0.9+xV_0.2−2x (x = 0 to 0.02), Ti_0.93Zr_0.07Mn_1.1+yCr_0.7+zV_0.2−y−z (y = 0, 0.05, z = 0.002, 0.05) and Ti_0.93+wZr_0.07Mn_1.15Cr_0.7V_0.15 (w = 0.002, 0.04) alloys. It was found that the decrease of V content led to the increase of plateau pressure but decreased the plateau slope. The addition of super-stoichiometric Ti resulted in the expansion of cell volume and the improvement of hydrogen affinity, which reduced of hydrogen absorption/desorption plateau pressure and hysteresis. The Ti_0.95Zr_0.07Mn_1.15Cr_0.7V_0.15 alloy was with a hydrogen absorption capacity of 1.83 wt% under 3.2 MPa H₂ pressure and 10˚C and a hydrogen desorption capacity of 1.07 wt% at 90˚C. Bing et al. [147] developed a Ti-Zr-Cr-based (Ti_0.8Zr_0.2)_1.1Mn_1.2Cr_0.55Ni_0.2V_0.05 alloy with the inclusion of Mn, Ni and V elements, which displayed good thermodynamic and kinetics properties. The alloy could desorb hydrogen under ambient conditions of 1 - 40 atm and 273 - 333 K. The hydrogen storage capacity of the alloy was 1.82 wt% at 298 K, and the hydrogen absorption and desorption plateau pressure were 10.88 and 4.31 atm, respectively. Wijayanti et al. [148] changed the ratio between A (Ti, Zr) and B (Mn, V, Fe, Ni) components and found that AB_1.9 alloy with substoichiometry had a high discharge capacity of 495 mAh·g^–1. The increase of the B/A ratio significantly increased the hydrogen desorption equilibrium pressure from 0.3 bar (B/A = 1.9) to 0.8 bar (B/A = 2.0) at 293 K. The AB_1.95 alloy exhibited the best activation properties and HRD, and the AB_2.0 alloy exhibited excellent cycling stability and a high hydrogen diffusion coefficient. Khajavi et al. [149] studied the effects of both elemental substitution of Fe for Mn and annealing treatment of the Ti_0.5Zr_0.5(Mn_1−xFe_x)Cr (x = 0 - 0.4) alloys. XRD results showed that neither Fe addition nor annealing treatment changed the multiphase structure of the alloys, and the main phases had the same hexagonal structure. But the lattice parameters significantly decreased with increasing Fe content.

In addition to elemental modification, microstructures and hydrogen storage properties can also be tailored by applying different preparation methods and subsequent treatments. For example, Stepanova et al. [150] prepared the Ti_6.5Al_3.5Mo_1.5Zr_0.3Si alloy with refined the texture using the electron beam melting method, and microhardness of alloy was also found to decrease with increasing beam current, but the hydrogenation process increased the alloy hardness due to the precipitation of the δ-TiH hydride and the redistribution of the alloy elements (formation of intermetallic particles). After the treatment with the beam current of 3 mA and the scanning speed of 150 mm·s^–1, the sample exhibited the highest hydrogen absorption rate of 0.006 wt%·min^–1. Khajavi et al. [151] found that ball milling and cold rolling could recover the alloys after exposion to air. The Ti_0.5Zr_0.5(Mn_1−xFe_x)Cr (x = 0 - 0.4) alloys could not absorb hydrogen after 10 days of air exposure, but could quickly recover the hydrogen absorption and desorption ability after mechanical deformation such as a short period of ball milling or a single pass of cold rolling. Moreover, ball milling could result in fast kinetics but also lead to the loss of hydrogen storage capacity. Comparatively, cold rolling provided faster kinetics and minimal capacity loss and was considered to be an appropriate treatment for the recovery of the alloys after exposion to air. Liu et al. [152] prepared the Cd/Pd particles with special core/shell microstructure by a two-step reduction method which was further ball milled with the Ti₄₉Zr₂₆Ni₂₅ alloy. The Cd/Pd composite coated on the Ti₄₉Zr₂₆Ni₂₅ alloy surface could reduce the charge transfer resistance and accelerate the hydrogen transfer of the Ti₄₉Zr₂₆Ni₂₅ alloy, thus improving the electrochemical performance and reaction kinetics of the alloy electrode. The Ti₄₉Zr₂₆Ni₂₅ + 7 wt% Cd/Pd electrode showed the highest discharge capacity of 272.9 mAh·g^–1 and the Ti₄₉Zr₂₆Ni₂₅ + 5 wt% Cd/Pd electrode showed the best cycling stability.

A typical representative of AB-type hydrogen storage alloys is TiFe alloy, an intermetallic compound being widely studied at present. The structure of TiFe alloy belongs to body centered cubic (BCC) structure as shown in Figure 6 . TiFe and TiFe-based alloys generally have the advantages of high hydrogen storage capacity (1.86 wt%), excellent cycling stability, rapid hydrogen absorption/desorption kinetics in a wide temperature range, and high abundant with low cost [153] - [155] . However, the activation difficulty and large hysteresis during hydrogen absorption/desorption have been limiting the application of this alloy system. To solve the above problems, intensive efforts have been made to understand the hydrogen storage properties and the development methods of this alloy system [9] [154] .

Elemental substitution and addition can moderate the hydrogenation behavior of TiFe-based hydrogen storage alloys by optimizing their microstructure and phase composition. The substitution and addition of rare-earth elements in TiFe-based alloys usually increase the alloy phase boundary, refine the crystal size, shorten the activation incubation period, and significantly improve their activation properties. Zhai et al. [156] found that adding an appropriate amount of La could improve the hydrogen absorption kinetics of the Ti_1−xLa_xFe_0.8Mn_0.2 (x = 0 - 0.09) alloys. The hydrogen storage capacity reached the maximum of 1.633 wt% at 323K when x = 0.01, and the plateau pressure increased with the substitution of La for Ti. Liu et al. [157] reported that the Ti_1.1−xFe_0.6Ni_0.3Zr_0.1Mn_0.2La_x (x = 0 - 0.08) alloys with La addition showed excellent activation performance, and can be fully activated within one cycle Han et al. [158] found that the addition of Y significantly improved the activation properties and hydrogen absorption/desorption kinetics of the Ti_1.1−xZr_0.1Y_xFe_0.6Ni_0.3Mn_0.2 (x = 0 - 0.08) alloys which could be fully activated by hydrogen absorption under 3 MPa H₂ pressure and 150˚CC, and also exhibited a fast hydrogen absorption rate at 10 C. When x = 0.02, the saturated hydrogen absorption rate of the alloy reached 92%, and the maximum hydrogen absorption capacity at 70 °C was 1.704 wt%. Zhang et al. [159] . Partially substituted Sm for Ti in the Ti_1.1−xFe_0.6Ni_0.1Zr_0.1Mn_0.2Sm_x (x = 0 - 0.08) alloys which refined the grain size and thus improved the activation performance and greatly shortened the activation incubation period. All the alloys exhibited good activation property and could be fully activated under room temperature. Particularly, the x = 0.04 and x = 0.08 alloys absorbed hydrogen without incubation time, and the hydrogen storage capacity of the x = 0.02 alloy at 313 K reached 1.438 wt%. Xu et al. [160] reported that the substitution of Pr for Ti in the Ti_1.1−xPr_xFe_0.6Ni_0.3Mn_0.2 (x = 0 - 0.08) alloys greatly improved the cycling performance. When x = 0.08, the S₅₀ of the alloy electrode reached 90.42%. Shang et al. [161] found that the Ti_1.1−xFe_0.7Ni_0.1Zr_0.1Mn_0.1Pr_x (x = 0 - 0.08) alloys with Pr partially substituting for Ti at x = 0.04 and 0.08 displayed good activity property, and could absorb hydrogen without any incubation time. Meanwhile, Pr addition was also beneficial to the nucleation rate and grain refinement, thus enhancing the hydrogenation kinetics of the alloys. At 313 K, the hydrogenation capacity of the x = 0.02 alloy reached a 1.438 wt%.

The addition of transitional elements usually induces the formation of secondary phases with catalytic effect on the activation performance of TiFe-based alloys. Lee et al. [162] found a small amount of TiFe₂ phase with C14 Laves structure and Ti₂Fe phase with cubic structure appearing in the Zr- and ZrCr-containing TiFeZr and TiFeZrCr alloys. Because TiFe₂ and Ti₂Fe phases have higher microstrain, more dislocations and smaller grain size than the TiFe main phase, the TiFeZr and TiFeZrCr alloys exhibited easy activation performance under 30 bar H₂ pressure and room temperature. Li et al. [163] reported that partial substitution of Fe for Ni could significantly reduce the minimum activation temperature. Among the TiFe_1−xNi_x (x = 0 - 0.4) alloys, the lowest activation temperature was obtained for the x = 0.4 alloy which was 443 K. Further studies showed that the decrease of activation temperature was attributed to the presence of NiO in the nickel-containing alloys which reduced the compactness of the surface oxide film. Moreover, it was found that the plateau pressure decreased and the hydride stability increased with increasing Ni content. Dematteis et al. [164] studied the effects of microstructure change and secondary phase formation on the activation and kinetics performance of the TiFe alloy by partial substitution of Mn and Ti for Fe. It was found that the secondary phase reacted with hydrogen which enhanced the activation performance and allowed hydrogen absorption under mild conditions. However, the large number of secondary phases reduced the hydrogen storage capacity and reversible capacity of the alloys. A good compromised was reached for the TiFe_0.85Mn_0.05 alloy which displayed the best comprehensive hydrogen absorption/desorption performance with a reversible capacity of 1.63 wt% (under 0.03 - 2.5 MPa H₂ pressure and 25˚C), easy activation (incubating at 25˚C and 2.5 MPa H₂ pressure for 6 hours) and good kinetics performance. The alloy exhibited a fast hydrogen absorption rate with the t₉₀ less than 2 min. Dematteis et al. [165] studied the effects of substitution of Cu for Fe in the TiFe_0.88−xMn_0.02Cu_x (x = 0 - 0.04) alloys. It was found the alloys were of multiphase structure with the secondary phases of β-Ti and Ti₄Fe₂O phases. The secondary phase abundance increased with the addition of Cu content. A small amount of secondary phase was helpful for the hydrogen absorption activation while the reversible capacity was reduced and the activation conditions became more stringent with the further increase of the secondary phase amount. Further study showed that the decrease in the capacity was attributed to the increase in the pressure difference between the first and second plateaus of the intermetallic compounds caused by Cu substitution which decreased the first plateau pressure while increases the second plateau pressure. Ali et al. [166] studied the effects of substituting Cu for Y of the TiFe_0.86Mn_0.1Y_0.1−xCu_x (x = 0.01 - 0.09) alloys. Results showed that with the increase of Y content, the hydrogen storage capacity first increased and then decreased with the maximum value of 1.89 wt% at x = 0.05. Moreover, the plateau pressure and the plateau slope decreased with Cu addition. Li et al. [167] found that the addition of Zr in Ti_1.08Y_0.02Fe_0.8Mn_0.2Zr_x (x = 0 - 0.08) alloys significantly shortened the activation period, and the x = 0.04 alloy could be directly activated without inoculation time. But excessive Zr led to the formation of precipitates, which reduced the hydrogen absorption/desorption capacity of the alloys. Faisal et al. [168] studied the binary TiFe alloy and eight ternary Ti-Fe-V alloys with 3 wt% Ce addition. It was found that the addition of Ce promoted the activation property of the alloys at room temperature by inhibiting the formation of the Ti₄Fe₂O_1−x oxides and adjusting the phase composition. V addition could improve the available hydrogen storage capacity of the alloys. Among the designed Ti-Fe-V series alloys, the Ti₄₆Fe_47.5V_6.5 alloy displayed the best hydrogen storage performance with the available hydrogen capacity of about 1.5 wt% under 1 MPa hydrogen absorption pressure and 0.1 MPa hydrogen desorption pressure at 30˚C. Leng et al. [169] tried to adjust the amount of Co in the TiFe_0.8Mn_0.2Co_x (x = 0 - 0.15) alloys which improved the alloys’ hydrogen storage capacity. With the increase of Co content, the lattice parameters and hydrogen storage capacity of the TiFe phase decreased, but the improvement of the flatness on the hydrogen desorption plateau increased the available hydrogen capacity of the alloys. Further research showed that the increase in the available hydrogen capacity might be owning to the adjustment of the change of the octahedral interstitial environment caused by Mn appearance in the TiFe phase, which helped improve the flatness of α-β desorption plateau. Shang et al. [170] found that the addition of super-stoichiometric Ti in TiFe alloy led to the decrease of hydrogen storage capacity, kinetics performance and dehydrogenation utilization rate of the alloys. Further substitution of Mn for Fe reduced the average grain size, which benefited the activation, improved the hydrogen storage capacity and kinetics performance, and led to a slight increase in the dehydrogenation enthalpy value. The Ti_1.14Fe_0.8Mn_0.2 alloy exhibited the highest hydrogen absorption capacity of 1.764 wt%, the shortest saturated hydrogen absorption time of 300 s, and a high dehydrogenation utilization rate of 96.9%.

As activation is one of the main obstacles faced with TiFe-based alloys, extensive works have been done focusing on ameliorating the activation property of this alloy system. For example, Liu et al. [171] reported that the addition of an appropriate amount of high oxygen could improve the initial hydrogen absorption of TiFe alloy. The TiFe-O_3.78 composite was fully activated after two hydrogen absorption/desorption cycles at room temperature. Further studies showed that high oxygen content would lead to the formation of the Ti₄Fe₂O phase in TiFe alloy, thereby improving the activation kinetics. However, the reduction of the TiFe phase content decreased the hydrogen storage capacity. Doping catalysis is another important means to improve the activation performance of TiFe-based alloys. Alam et al. [172] reported that doping a small amount of metal La into TiFe alloy could effectively reduce the activation incubation period. Under 40 bar hydrogen pressure and room temperature, the first hydrogen absorption capacity reached 1 wt% in less than 5 min. Lv et al. [173] added different amounts of ZrMn₂ to the TiFe alloy to produce TiFe + x wt% ZrMn₂ (x = 2 - 12) composites. The composites were composed of TiFe main phase and Zr-rich and Mn-rich secondary phases. The secondary phase abundance increased with the increase of ZrMn₂ addition. Due to the very fine distribution of the secondary phase, the first hydrogenation kinetics and hydrogen storage capacity increased, and the plateau pressure decreased. Moreover, the synergistic effect of Zr, Mn, V and other elements improved the kinetics properties and hydrogen storage capacity. Patel et al. [174] added Zr, Mn and Zr + Mn to TiFe alloy and it was found that the addition of enabled the activation under 20 bar hydrogen pressure and room temperature, but the kinetics was very slow. The alloy with 2 wt% Zr was not activated. When 4 wt% Zr was added, the alloy absorbed 1.2 wt% hydrogen. However, when Zr and Mn were added together, the alloy with 1 wt% Mn + 2 wt% Zr exhibited better kinetics than the alloy with only Mn or only Zr. The maximum hydrogen storage capacity reached within 7 h, which was also higher (about 1.8 wt%). The composite doped with 4 wt% Zr + 2 wt% Mn exhibited a hydrogen absorption capacity of 2 wt% within 5 hours. Moreover, Patel et al. [175] further added V, Zr + V and Zr + V + Mn into TiFe alloy. it was found that the alloys with (Zr, V) and (Zr, V, Mn) demonstrated rapid activation performance which might be enabled by the existence of the Ti₂Fe-like secondary phase acting as the entrance for hydrogen atoms. However, the TiFe alloy with 2 wt% V could not be activated at room temperature, and the that with 2 wt% Zr alone was ineffective, neither, which meant that there were some synergistic effects when Zr and V were added simultaneously.

Appropriate ball milling process can refine the grain size and even induce an amorphous structure, which generates cracks on the alloy surface and increases the hydrogen diffusion channel. Shang et al. [176] synthesized a Ti_1.04Fe_0.7Ni_0.1Zr_0.1Mn_0.1Pr_0.06 + 10 wt% Ni composite with amorphous structure by ball milling. Owing to the decrease of particle size, grain refinement, increase of crystal defects and change of surface state, the alloy reached the maximum discharge capacity in the first cycle. However, with the increase of ball milling time, the discharge capacity and electrochemical kinetics properties of the composites decreased significantly. Li et al. [177] also found that ball milling treatment led to grain refinement, particle size reduction and disordered structure of the Ti_1.04Fe_0.7Ni_0.1Zr_0.1Mn_0.1Pr_0.06 alloy. The sample milled for 0.5 h showed the best hydrogen absorption and desorption kinetics and hydrogen storage capacity. Under 3 MPa H₂ pressure and 313 K, the hydrogenation capacity reached 1.626 wt% within 2158 s, and the activation did not require incubation time when the temperature increased to 423 K. Similarly, Yuan et al. [178] reported that the activation time of the ball-milled alloy was greatly shortened, and the nano-grain boundaries and the phase boundaries resulted from the secondary phase provided a large number of channels for hydrogen diffusion. At 443 K, the activation time of the ball-milled alloy (0.75 h) was only 4 min. However, due to the increase in the number and density of crystal boundaries, the hydrogen storage capacity decreased significantly from 1.387 wt% to 0.46 wt% at 323 K.

Heat treatment has also been attempted on TiFe-based alloys, but the effects were not so desired due to the decrease of secondary phase content. For example, He et al. [179] compared the as-cast and annealed TiFe-6 wt% ZrCr₂ alloys and found that both alloys were with the TiFe main phase, a small amount of TiFe₂ and Ti₂Fe secondary phases. But the TiFe₂ secondary phase was significantly reduced after annealing. Both the alloys could be hydrogenated under 31 bar hydrogen pressure and room temperature without harsh activation process, but the hydrogenation of the annealed sample required about 40 hours of incubation time. Further studies showed that the secondary phase (TiFe₂) which was reduced by annealing played a key role in the hydrogenation process at room temperature.