Microstructure and properties study of Mg2-xYxNi0.9Co0.1 (x = 0, 0.1, 0.2, 0.3) hydrogen storage alloys

To enhance the hydrogen storage performance of Mg-Ni system alloys, multi-elemental alloys incorporating Y element, namely Mg2-xYxNi0.9Co0.1 (x = 0, 0.1, 0.2, 0.3), were synthesized through ball milling and sintering. The microstructures of Mg2-xYxNi0.9Co0.1 (x = 0, 0.1, 0.2, 0.3) alloys were characterized using XRD and SEM/EDS techniques, and the hydrogen storage properties of Mg2Ni0.9Co0.1 and Mg1.7Y0.3Ni0.9Co0.1 alloys were evaluated via the Sieverts method. At a sintering temperature of 500 °C, the Y element existed in the form of Y/Y2O3 phases and displayed no reactivity with other alloy constituents. The addition of Y enhanced the activation performance of Mg-Ni system alloys, because it takes 2 times for Mg1.7Y0.3Ni0.9Co0.1 alloy to complete activation, while Mg2Ni0.9Co0.1 needs 3, albeit causing a slight reduction from 3.6 wt% to 3.2 wt% in the hydrogen storage capacity when Y replaced Mg. The enthalpy of hydrogen desorption of Mg1.7Y0.3Ni0.9Co0.1 alloy was 57.4 kJ mol−1 H2, which was significantly lower than that of Mg2Ni0.9Co0.1 (68.0 kJ mol−1 H2) and Mg2Ni alloy (64.4 kJ mol−1 H2), indicating improved thermodynamic properties. Moreover, the apparent activation energy of Mg2Ni0.9Co0.1 (71.48 kJ mol−1 H2) was lower than that of Mg1.7Y0.3Ni0.9Co0.1 (83.62 kJ mol−1 H2), implying that the addition of Y reduced the kinetic properties.


Introduction
Magnesium-based hydrogen storage materials are recognized as highly promising due to their lightweight nature, high hydrogen storage capacity (up to 7.6 wt% of pure Mg), abundant resources, and affordability [1][2][3][4][5]. These materials can be utilized as hydrogen sources for fuel cell auxiliary power devices in fuel cell vehicles [6]. However, their high hydrogen desorption temperature and poor hydrogen absorption/desorption kinetics pose limitations to their further development [4,7,8]. Therefore, improving the thermodynamic and kinetic properties of Mg-based materials for hydrogen absorption and desorption has become a critical research topic. While the A 2 B type Mg 2 Ni alloys exhibit significant enhancements in the thermodynamic and kinetic properties of hydrogen absorption and desorption compared to pure Mg, the enthalpy of hydrogen desorption and the activation energy for dehydrogenation of Mg 2 Ni alloy were 64.5 kJ mol −1 and 80 kJ mol −1 respectively, when that of pure Mg were 70-75 kJ mol −1 and 130-140 kJ mol −1 , there is still room for further improvement, particularly in practical applications [9]. Furthermore, the activation process of magnesium-based alloys is also a challenging aspect [10].
Currently, methods to enhance the hydrogen storage performance of Mg-Ni system alloys include alloying [11], nanocrystallization [12], catalyst addition [13], and surface modification [14]. Alloying aims to improve the thermodynamics and kinetics of hydrogen absorption and desorption by incorporating new alloying elements into Mg-Ni system alloys to introduce new phases. Other methods primarily focus on treating existing alloys to create favorable conditions for hydrogen absorption and desorption. Clearly, alloying is the main development direction for Mg-Ni system alloys in order to develop new alloys with superior hydrogen storage properties.
The inclusion of transition elements instead of Ni has been shown to enhance the hydrogen storage properties, particularly the kinetic properties, of Mg-Ni alloys [15]. For instance, in the solution-sintered Mg 2 Ni 1x Ti x (x = 0, 0.08, 0.12, 0.16) alloys [16], the solid solution of Ti in the Mg 2 Ni phase and defects in the preparation process improved the dynamic properties, leading to a reduction in the activation energy for dehydrogenation to 59 kJ mol −1 for the Mg 2 Ni 0.88 Ti 0.12 alloy. Zhang Y et al prepared Mg 2 Ni 0.92 M 0.08 (M = Ti, V, Fe, Si) alloys by solution sintering [17]. The addition of element M enhanced the dynamic properties of the alloy, resulting in a decrease in the activation energy for the dehydrogenation reaction to 42-67.1 kJ mol −1 . Furthermore, nanostructured Mg 2 Ni 1-x Co x (x = 0.05, 0.1) compounds were obtained through a hydrogen plasma-metal reaction [18], and the activation energy of dehydrogenation reaction decreased to 49∼64.3 kJ mol −1 , demonstrating superior sorption kinetics.
Additionally, the incorporation of rare earth elements can also improve the hydrogen storage performance of Mg-Ni system alloys. Induction melting was used to prepare Mg 80 Ni 20 Y x (x = 0-7) alloys [19]. As the Y content increased, the enthalpy of hydrogen desorption decreased gradually from 62.3 kJ mol −1 to 59.8 kJ mol −1 . Similar trends were observed in Mg 10 Ni x Mm (x = 1, 2, 3) alloys and Mg 23-x La x Ni 10 (x = 0, 1, 2) alloys, showcasing the influence of rare earth elements Mm and La on the hydrogen storage properties of Mg-Ni system alloys [20,21].
However, the impact of a single element addition on the hydrogen storage performance of Mg-Ni system alloys is limited due to the improvement mechanism. Hence, exploring the effects of different elements on the microstructure and hydrogen storage properties of Mg-Ni system alloys is an important direction for development [22]. In this regard, we synthesized Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) quaternary alloys by incorporating Y into the Mg 2 Ni 0.9 Co 0.1 ternary alloy. The influence of Y addition on the microstructure and hydrogen storage properties of Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) was investigated. This study aims to enhance our understanding of multi-element alloying and provide guidance for the development of Mg-Ni system alloys with superior hydrogen storage performance.

Experiments
The Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) alloys were synthesized through a process of ball milling and sintering. The raw materials employed in the alloy preparation were pure metal powders with a minimum purity of 99.9 wt% and particle sizes not less than 300 mesh. The metal powders were accurately weighed and placed into a vacuum ball milling tank, with a ball-to-material ratio of 20:1. The ball mill was operated at a rotation speed of 200 r min −1 for a total duration of 60 min. Subsequently, 5 g of the mixed alloy powder was weighed in a glove box for tablet pressing, applying a pressure of 15 MPa and maintaining it for 5 min. Finally, the samples were sintered under argon protection, with a sintering temperature of 500°C, a duration of 20 h, and a heating rate of 5°C min −1 . After sintering, the samples were cooled to room temperature in the furnace. Part of the sintered alloy sample was mashed into powder in the glove box and screened through a 200-mesh screen for test analysis.
The phase structure of the alloys was characterized using x-ray diffraction (XRD) with Cu K-α radiation, a characteristic wavelength (λ) of 1.54056 Å, a voltage of 40 kV, a current of 30 mA, and a diffraction angle (2θ) ranging from 10°to 85°. The microstructure of the samples was observed using a Quanta FEG450 field emission scanning electron microscopy (SEM), and the chemical composition of each micro area was analyzed using energy-dispersive x-ray spectroscopy (EDS). The hydrogen storage properties of the samples were evaluated using an automated Sievert's type apparatus.

Microstructure
Adding rare earth elements to Mg-Ni system alloys can reduce the absolute values for enthalpies of hydrogen absorption and desorption of Mg 2 Ni phase and Mg phase, and improve the hydrogen storage properties of the alloys [19,20,23]. Rare earth element RE can exist in different phase forms in Mg-Ni system alloys, which can form not only Mg-RE and RE-Ni binary phase, but also Mg-RE-Ni ternary phase [24][25][26][27], therefore, it is necessary to study the effect of Y on the microstructure of Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) alloys before analyzing that on the hydrogen storage properties.
alloys at about 29.6°, 34.4°and 49.5°at 2θ, which is Y 2 O 3 phase, where oxygen comes from unavoidable air contact. Y element only exists in the form of Y 2 O 3 phase, which indicates that Y does not react with other alloying elements to form intermetallic compounds. With the increase of x, the diffraction peak intensities of 3) alloys gradually increases, indicating that the relative content of Y 2 O 3 phase increase. Besides, with the increase of x, the diffraction peak intensities of Mg 2 Ni decrease gradually, which indicates that the relative content of Mg 2 Ni phase decreases, which is caused by the partial substitution of Y for Mg. In addition, compared with x = 0 alloy, the diffraction peak intensities of Mg and Mg 2 Co phases in x = 0.1, 0.2 and 0.3 alloys decrease due to the partial substitution of Mg by Y, indicating that the contents of Mg and Mg 2 Co phases are reduced.
The BSE images of Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) alloys are showed in figure 2, and the EDS analysis results of each typical region are listed in table 1, based on EDS analysis accuracy, when the element content is less than 1%, it is considered that the element is not contained. It can be seen from figure 2 that all of the BSE images of Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) alloys have gray phases (zone A), bright gray phases (zone B) and black phases (zone C), and there are almost no significant pores in dense samples mainly due to the proper process. The gray phases are the matrixes, the bright gray phases are distributed in the matrixes as small pieces, and the black phases are distributed in the vicinity of the matrix phases as irregular flakes. Combined with the results of EDS analysis (table 1), we can determine that the gray matrix phases, bright gray phases and black phases are Mg 2 Ni phases, Mg 2 Co phases and Mg phases respectively. Besides, in addition to Mg and Ni elements, there are Co element in zone A of each sample, which indicates that a small amount of Co is solidly dissolved into Mg 2 Ni phase. In the BSE images of Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0.1, 0. There are only diffraction peaks of Y 2 O 3 phases in XRD spectrum, but there are no that of Y phases, which may be due to the oxidation of Y phases in the samples tested by XRD.

Activation properties
Through the above analysis, it is found that Y element exists in the form of Y/Y 2 O 3 phases in all the Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0.1, 0.2, 0.3) alloys, which means that the effect mechanisms of Y addition on hydrogen storage performance of Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0.1, 0.2, 0.3) alloys are the same, since most of the mechanisms are based on phase composition for alloys of the same composition system [11][12][13][14]. Hence, we only tested the hydrogen storage properties of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy as representative, and compared with that of Mg 2 Ni 0.9 Co 0.1 alloy, to discussed the mechanism of Y addition on hydrogen storage performance of the system.
Magnesium-based hydrogen storage alloys are easy to react with oxygen, mainly because Mg is a reactive metal and it is easily oxidized to form an oxide layer on the surface to prevent further oxidation, but the oxide layer also hinders the reaction between alloys and hydrogen [28]. Therefore, the alloys should be activated before use [29]. Figure 3 shows the first three activation curves of Mg 2 Ni 0.9 Co 0.1 and Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloys. The activation temperature was 350°C and the hydrogen ab-/desorption pressures were 3.5/0.004 MPa. It can be seen that the hydrogen absorption of Mg 2 Ni 0.9 Co 0.1 alloy is not saturated in the first activation process, and the hydrogen absorption capacity is 3.01 wt% in 3600 s. During the second and third hydrogen absorption processes, Mg 2 Ni 0.9 Co 0.1 alloy reaches saturation at about 900 s and 270 s, with hydrogen absorption capacities of 3.18 wt% and 3.16 wt%, respectively. Comparing the second and third activation curves of Mg 2 Ni 0.9 Co 0.1 alloy, although the hydrogen absorption rate of the second activation is slightly slower than that of the third activation, the latter part of the two curves almost coincide. Therefore, it can be considered that Mg 2 Ni 0.9 Co 0.1 alloy has been activated after three cycles. During the first activation, Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy is saturated at 1600 s, and the hydrogen absorption capacity is 2.44 wt%. The second and third activation curves of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy almost coincide and reach saturation at about 500s with saturated hydrogen absorption capacity of 2.54wt% and 2.51wt% respectively. It can be considered that Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy can be activated by secondary circulation.  It takes 7 to 10 cycles for Mg 2 Ni alloys prepared by traditional induction melting method to complete activation [29], and it also takes 4 to 5 cycles for Mg 2 Ni alloys prepared by laser sintering [30]. Obviously, the Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy prepared in this paper has good activation properties. The solid solution of Co atoms in Mg 2 Ni phase would cause lattice expansion, the defects occurred during the preparation of the alloy mainly including stacking fault, dislocation and subgrain boundary [16,31], and the formation of Mg 2 Co phase is helpful to the increase of the phase boundary area of alloy. All these factors are conducive to the diffusion of hydrogen atom in the alloy. The addition of Y further improved the activation properties of Mg-Ni system alloys. It is well known that Y is easily oxidized. The Y phase in Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy can prevent the oxidation of Mg and absorb the O element at the interface. The hydrogen storage capacity of Mg 2 Ni 0.9 Co 0.1 alloy is about 3.2 wt%, less than that of Mg 2 Ni (3.6 wt%), which is related to the burning loss of Mg and the formation of Mg 2 Co phase. Compared with Mg 2 Ni 0.9 Co 0.1 alloy, the hydrogen storage capacity of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy further decreased to about 2.5 wt%, which was due to the substitution of Y for Mg. Figure 4 shows the PCT curves of hydrogen desorption of Mg 2 Ni 0.9 Co 0.1 and Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloys at different temperatures. It can be seen from the figure that Mg 2 Ni 0.9 Co 0.1 and Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloys have only one dehydrogenation plateau pressure at three temperatures, corresponding to the dehydrogenation process of Mg 2 Ni hydrides. Both Mg 2 Co phase and Y phase can absorb hydrogen to form hydrides Mg 2 CoH 5 /Mg 3 CoH 5 and YH 2 /YH 3 phases. Under the test conditions, Mg 2 CoH 5 /Mg 3 CoH 5 did not release hydrogen [32,33]. The hydrogen desorption PCT curve of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy has no plateau pressure from YH 3 to YH 2 , which is similar to some Mg-RR-Ni ternary alloys [19,21].    [30]). After Y addition, the dehydrogenation enthalpy of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy decreased significantly and the thermodynamic properties were improved. This is because the transition from YH 3 to YH 2 will cause stress and strain in the alloy during dehydrogenation, thus reducing the stability of Mg 2 Ni hydride [19,35]. Figure 6 shows the dehydrogenation kinetics curves of Mg 2 Ni 0.9 Co 0.   The hydrogen desorption mechanism of hydrogen storage alloys can be explained by JMAK ((Johnson-Mehl-Avrami-Kolmogorov) model based on nucleation and growth process [37,38]:

Kinetic properties
where α is the fraction transformed, k is the reaction rate constant associated with the activation energy, η is the Avrami exponent of reaction order providing some information about the dimensionality of the transformation. Figure 7 shows the piots of ln[-ln(1-α)] versus lnt of Mg 2 Ni 0.9 Co 0.1 and Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloys at different temperatures. It can be seen that ln[-ln (1-α)] has a linear relationship with lnt at a given temperature, R 2 > 0.99, which indicates that JMAK equation can describe the hydrogen evolution process of the alloys. The η values of the two samples are in the range of 0.76∼1.06, which is very close to 1, indicating that the dehydrogenation process of the alloys conforms to the diffusion reaction mechanism.
Furthermore, the apparent activation energies of the alloys can be calculated by Arrhenius equation [39]: where k o is a pre-exponential factor., E A is the apparent activation energy, k is the rate constant, R is the gas constant and T is the temperature. The Arrhenius plots of lnk versus 1000/RT of Mg 2 Ni 0.9 Co 0.1 and Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloys are shown in figure 8. The E A values of Mg 2 Ni 0.9 Co 0.1 and Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloys are 71.48 kJ mol −1 H 2 and 83.62 kJ mol −1 H 2 . In the Mg-RE-Ni ternary alloy system, the addition of RE is beneficial to improve the dynamic properties of Mg-Ni alloys [40][41][42]. While in this study, the apparent activation energy of Mg 2 Ni 0.9 Co 0.1 is lower  than that of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 , which means that the addition of Y decreases the kinetic properties. This may be due to the stress caused by the addition of Y, which reduces the solid solubility of Co atoms in Mg 2 Ni phase.

Conclusion
Mg 2-x Y x Ni 0.9 Co 0.1 (x = 0, 0.1, 0.2, 0.3) alloys were prepared by ball milling and sintering. The effect of Y addition on the microstructure and hydrogen storage properties of the alloys is studied, and the following conclusions are drawn: (1) At the sintering temperature of 500°C, Y element exists in the form of Y/Y 2 O 3 phases and does not react with other alloy elements.
(2) Owing to the existence of the Y phase, it takes 2 cycles for Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy to complete activation, while Mg 2 Ni 0.9 Co 0.1 needs 3, indicating that the addition of Y further improves the activation performance of Mg-Ni system alloys, even though the hydrogen storage capacity decreases slightly from 3.6 wt% to 3.2 wt%.
(3) The enthalpy of hydrogen desorption of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 alloy (57.4 kJ mol −1 H 2 ) is significantly lower than that of Mg 2 Ni 0.9 Co 0.1 (68.0 kJ mol −1 H 2 ) alloy, showing that the addition of Y improves the thermodynamic properties of the alloy.
(4) The apparent activation energy of Mg 2 Ni 0.9 Co 0.1 (71.48 kJ mol −1 H 2 ) is lower than that of Mg 1.7 Y 0.3 Ni 0.9 Co 0.1 (83.62 kJ mol −1 H 2 ), which means that the addition of Y decreases the kinetic properties. This may be due to the stress caused by the addition of Y, which reduces the solid solubility of Co atoms in Mg 2 Ni phases.