Effect of Substitutional Elements on the Thermodynamic and Electrochemical Properties of Mechanically Alloyed La1.5Mg0.5Ni7−xMx alloys (M = Al, Mn)

The A2B7-type La-Mg-Ni-M-based (M = Al, Mn) intermetallic compounds were produced by mechanical alloying and annealing. The thermodynamic and electrochemical properties of these materials were studied. The nickel substitution by aluminum and manganese in the La-Mg-Ni system improves the kinetics of hydrogen absorption. The hydrogen desorption capacity of Mn substituted compounds is improved significantly, and it reaches the value of 1.79 wt.% at 303 K when the composition is La1.5Mg0.5Ni6.80Mn0.20. On the other hand, the La1.5Mg0.5Ni6.85Al0.15 shows a much higher reversible electrochemical capacity than the La1.5Mg0.5Ni7 materials at the 50th cycle. The electrochemical discharge capacity stability increases with the increasing value of Al and Mn up to x = 0.2 and 0.3, respectively. Additionally, a reduction in the discharge capacity was measured for the Al and Mn content above x = 0.25 and 0.5, respectively. From the practical aspect, only La1.5Mg0.5Ni6.80Mn0.20 has a potential in the application as a hydrogen storage material.


Introduction
The A 2 B 7 -type intermetallic compounds have attracted attention as the new generation of hydrogen storage materials [1][2][3][4]. From the scientific point of view, a lot of efforts to improve the thermodynamic and electrochemical properties of the La-Mg-Ni-type materials have been undertaken. For example, modification of their chemical composition, microstructure evolution, heat and surface treatments, etc. [5][6][7][8].
The La-Ni-type alloys can be produced by the arc or induction melting in the high purity argon gas [22]. It is important to note, that modification of the hydrogen storage properties of these phases can be achieved via mechanical alloying (MA) [7,16,[19][20][21]. Lately, the influence of magnesium content on the properties of the La 2−x Mg x Ni 7 system was investigated in detail [7,9,23]. The content of Mg in the (La,Mg) 2 Ni 7 system influenced their final phase composition. For example, the resulting main three phases (La,Mg) 2 Ni 7 , (La,Mg)Ni 3 , and LaNi 5 were detected, for x = 0.48-0.5, x = 0.6, and x > 0.48 in the La 2−x Mg x Ni 7 alloy [23]. The existence of Mg in the La 2 Ni 7 alloy anticipates the pulverization of their hydrides.

Materials and Methods
The nanostructured La 1.5 Mg 0.5 Ni 7−x M x (M = Al (0 ≤ x ≤ 0.25), Mn (0 ≤ x ≤ 0.5)) compounds were synthesized by mechanical alloying in a high purity argon atmosphere (Table 1). Mechanical alloying was carried out using a 8000 SPEX mixer mill (SPEX SamplePrep, Metuchen, NJ, USA) with milling frequency of 875 Hz, employing a weight ratio of hard steel balls to powder weight ratio of 4.25:1 at ambient temperature for 48 h in a continuous mode. The following metals were used: La powders-grated from rod (Alfa Aesar, 99.9%), Mg powder (Alfa Aesar, 325 mesh, 99.8%), Ni powder (Aldrich, 5 µm, 99.99%), Al powder (Aldrich, 200 mesh, 99%), and Mn powder (Aldrich, 325 mesh, ≥99%). The elemental powders were weighed, blended, and poured into a round bottom stainless vial (35 mL) in a glove box (Labmaster 130) filled with automatically controlled argon atmosphere (O 2 ≤ 2 ppm and H 2 O ≤ 1 ppm) to obtain the materials. A composition of starting materials mixture was based on the stoichiometry of an "ideal" reaction. However, due to oxidation of La and Mg, the content of these elements was increased by 8 wt.%. The amount of La and Mg extra addition (8 wt.%) was determined during our basic research (not shown here), to obtain after the MA process, materials with a chemical composition as close as possible to the stoichiometry of an "ideal" reaction. The La 1.5 Mg 0.5 Ni 7−x M x (Al; x = 0, 0.10, 0.15, 0,20, 0.25 and Mn; x = 0, 0.2, 0.3, 0.4, 0.5) powders synthesized by MA were finally heat-treated in 1123 K for 0.5 h in a high purity argon and subsequently cooled in air. For this treatment, the powder (5 g of each composition) was closed under the argon in a quartz tube with a volume of approximately 4 cm 3 .
The phase analysis crystal structure of synthesized powders was investigated at room temperature by the XRD method (Panalytical, Empyrean model, Almelo, the Netherlands) with CuKα 1 (λ = 1.54056 Å) radiation. The phase quantitative analysis was based on the line profile analysis of the XRD powder patterns realized with the X'Pert High Score Plus software (Tables 2 and 3). The Williamson-Hall (W-H) analysis method was used to study crystallite sizes based on the diffraction pattern of the obtained mechanically alloyed powders. Table 1. Chemical compositions of output powders needed to get the following La-Mg-Ni-M (M = Al, Mn) compounds after all stages of specimen preparation by the application of a SPEX 8000 mixer mill (total weight of milling powders-5 g, the ball to powder mass ratio-4.25:1, milling time-48 h).

Compound
Chemical Pressure-composition isotherms were determined by a Sievert PCI apparatus (Particulate Systems, HPVA 200 model, Norcross, GA, USA). The concentration of the absorbed hydrogen was calculated based on the hydrogen pressure changes measured in the reaction chamber during the tests. The mass of the sample for each measurement cycle was approx. 0.6 g. The investigations of the hydrogen absorption kinetics were carried out at 303 K and under 3 MPa (hydrogen pressure) in the first, second, and third cycle. Each measurement was finished after obtaining the equilibrium hydrogen pressure-the change of pressure did not exceed 200 Pa within 5 min. After each cycle, the samples were degassed at the temperature of 673 K and in a vacuum. The pressure-composition-isotherm (PCI) curves were obtained in the subsequent cycle after the measurements of the kinetics at the same temperature in the hydrogen pressure range up to approx. 7 MPa. The hydrogen absorption and desorption cycles that occurred during the measurements of the kinetics acted as the activation process. The hydrogen content in the samples was obtained by measuring pressures at constant volumes.
The electrochemical studies were done at room temperature in a three-electrode open cell. The material electrodes in a pellet form (d = 8 mm) consisted of the powder mixture of the synthesized material (0.4 g) and carbonyl nickel (0.04 g). A full description of the electrochemical studies is included in our previous work [7,20]. The electrodes were charged and discharged at a current of 40 mA g −1 and the cut-off voltage was −0.7 V vs. the reference Hg/HgO electrode.

Sample Phase Composition
The series of compounds with the nominal composition La 1.  Tables 2  and 3). In La 1.5 Mg 0.5 Ni 6.8 Al 0.2 , MgO is present. The abovementioned oxide phases like to be formed during the MA process. No pure La, Mg, Ni, Al, or Mn elements are found in the collected XRD patterns. The mean crystallite sizes of produced powders were 37-46 nm according to the XRD analysis. The published data for 2H-type (P63/mmc) and 3R-type (R3-m) of the La 1.5 Mg 0.5 Ni 7 phase were used as input data for refinements of our XRD patterns [7]. The analysis results are summarized in Tables 2 and 3. The graphical representation of the line profile analysis of La 1.5 Mg 0.5 Ni 6.8 Al 0.2 and La 1.5 Mg 0.5 Ni 6.8 Mn 0.2 are presented in Figure 1. The content of the A 2 B 7 -type phase increased from 88.9% in La 1.5 Mg 0.5 Ni 7 to 99.8% and 97.8% in La 1.5 Mg 0.5 Ni 6.8 Al 0.2 and La 1.5 Mg 0.5 Ni 6.8 Mn 0.2 , respectively. However, a reduction in the abundance of the A 2 B 7 -type phase was observed for the Al and Mn concentration above x = 0.15 and 0.3, respectively.  Recently, the influence of the Ni replacement with Mn on the phase composition of the ReNi2.6−xMnxCo0.9 (x = 0.0, 0.225, 0.45, 0.675, 0.90) alloys synthesized by induction melting were studied [27]. In these alloys, (La,Ce)2Ni7 phase, (Pr,Ce)Co3 phase, and (La,Pr)Ni5 phase were the main phases.
All alloys absorb hydrogen at 303 K. The shift from the α-solid solution to the β-hydride phase is observed. The absorption plateau pressure for the La1.5Mg0.5Ni7−xMx (M = Al and Mn) is much lower than that for La1.5Mg0.5Ni7. The Al or Mn contents in the studied system influenced the hydrogen sorption pressure. Due to the higher stability of the La1.5Mg0.5Ni7−xMx (M = Al and Mn) hydrides, a decrease of the sorption pressure was measured for higher Al or Mn contents in the La1.5Mg0.5Ni7−x system. Previously, it was shown that the plateau pressure during the sorption increases with the increasing Mg content in La2−xMgxNi7 [7]. Additionally, the substitution of La by Mg results in the Recently, the influence of the Ni replacement with Mn on the phase composition of the ReNi 2.6−x Mn x Co 0.9 (x = 0.0, 0.225, 0.45, 0.675, 0.90) alloys synthesized by induction melting were studied [27]. In these alloys, (La,Ce) 2 Ni 7 phase, (Pr,Ce)Co 3 phase, and (La,Pr)Ni 5 phase were the main phases. Independently, the effects of the partial Ni replacement by Al on the microstructures of the as-cast and rapidly quenched La 0.7 Mg 0.3 Ni 2.55−x Co 0.45 Al x (x = 0, 0.1, 0.2, 0.3, 0.4) were studied [28]. Both, the as-cast and quenched samples consist of multiple phases including (La,Mg)Ni 3 , LaNi 5 , and LaNi 2 . In general, the synthesis method of the La 1.5 Mg 0.5 Ni 7 -type hydrogen storage alloys influences strongly their final phase composition.

Thermodynamic Properties
The hydrogen absorption-desorption behavior of La 1.5 Mg 0.5 Ni 7−x M x (M = Al (0 ≤ x ≤ 0.25), Mn (0 ≤ x ≤ 0.5)) are outlined in Tables 4 and 5. Figure 2 shows a correlation between the PCI data measured for the synthesized materials.
All alloys absorb hydrogen at 303 K. The shift from the α-solid solution to the β-hydride phase is observed. The absorption plateau pressure for the La 1.5 Mg 0.5 Ni 7−x M x (M = Al and Mn) is much lower than that for La 1.5 Mg 0.5 Ni 7 . The Al or Mn contents in the studied system influenced the hydrogen sorption pressure. Due to the higher stability of the La 1.5 Mg 0.5 Ni 7−x M x (M = Al and Mn) hydrides, a decrease of the sorption pressure was measured for higher Al or Mn contents in the La 1.5 Mg 0.5 Ni 7−x system. Previously, it was shown that the plateau pressure during the sorption increases with the increasing Mg content in La 2−x Mg x Ni 7 [7]. Additionally, the substitution of La by Mg results in the stability decrease of (La,Mg) 2 Ni 7 hydrides [35]. It is important to note, that the hydrogen sorption plateau pressure close to the atmospheric one was obtained in La 2−x Mg x Ni 7 for x = 0.25 and 0.5 [7].
Independently, Mani et al. observed that plateau pressure on the hydrogen sorption in the La-Mg-Ni-based systems decreased with the incorporation not only of Al, Mn but also of V, Cu, Fe, and Co [36]. The replacement of nickel by Mn, Al, and Co in La 1.5 Mg 0.5 Ni 7−x M x has been studied to find the optimum concentration of the substituting elements that would ensure the corrosion resistance and high hydrogen capacity.   Table 6 shows the time-capacity kinetics curves for hydrogen absorption for La 1.5 Mg 0.5 Ni 7 and La 1.5 Mg 0.5 Ni 7−x M x . The data were obtained at 303 K for the first three cycles. To reach the maximum hydrogen storage capacity and best kinetic properties the activation process of the samples was performed (see Table 6). It is important to note, that the chemical modification of Ni by Al or Mn in this study affected the kinetics of the hydrogen sorption. The La 1.5 Mg 0.5 Ni 7 absorbs 95% of the maximum hydrogen volume in 301 min, while the best Al-containing sample needs only 4 min (x = 0.1). After chemical modification of the alloy by Al or Mn substitution, the maximum hydrogen storage capacity was increased ( Figure 2). The highest value of 1.79 wt.% was obtained for the La 1.5 Mg 0.5 Ni 6,80 Mn 0.20 .

Electrochemical Properties
The  Tables 7 and 8. It is important to note, that the electrodes show the maximum capacities in the 3rd cycle. For the La 1.5 Mg 0.5 Ni 6.85 Al 0.15 electrode, the highest obtained discharge capacity was 328 mAh/g. The discharge capacities of all the studied La 1.5 Mg 0.5 Ni 7−x M x electrode materials degraded during the charge-discharge cycling, most likely to partial oxidation of the electrode materials or the formation of the stable hydride phases. The origin of this behavior could be the formation of the Mg(OH) 2 and La(OH) 3 layers on the surface. These layers decrease the surface electrocatalytic activity and prevent hydrogen diffusion into the electrodes. The pulverization of the electrodes during the hydrogenation and dehydrogenation cycles also influence the electrochemical properties.
The cycle stability of the La 1.5 Mg 0.5 Ni 7−x Al x (Al (x = 0.10, 0.15, 0.20) and Mn (x = 0.2, 0.3) electrodes increases. Up to now, the best cycle stability was observed in La 2−x Mg x Ni 7 [7]. The phase composition of this material influenced strongly its electrochemical properties. The major phase in the LaMgNi 7 sample is LaNi 5 [7]. This material has a higher electrochemical cycle stability in comparison with (La, Mg) 2 Ni 7 [37]. On the other hand, La 1.5 Mg 0.5 Ni 7−x Al 0.2 has the best capacity retaining rate after the 50th cycle. It is important to note, that the partial substitution of nickel with aluminum or manganese resulted in the increased cycle stability of the MH x alloy electrodes. Additionally, these chemical modifications also influenced the kinetics of the hydrogen sorption reducing the time of the hydrogenation process.  30 and C 50 are the discharge capacities after 30th and 50th charge/discharge cycles, C max is the maximum discharge capacity.
* C30 and C50 are the discharge capacities after the 30th and 50th charge/discharge cycles, Cmax is the maximum discharge capacity.
The discharge capacities of all the studied La1.5Mg0.5Ni7−xMx electrode materials degraded during the charge-discharge cycling, most likely to partial oxidation of the electrode materials or the formation of the stable hydride phases. The origin of this behavior could be the formation of the Mg(OH)2 and La(OH)3 layers on the surface. These layers decrease the surface electrocatalytic activity and prevent hydrogen diffusion into the electrodes. The pulverization of the electrodes during the hydrogenation and dehydrogenation cycles also influence the electrochemical properties. (a) (b) Figure 3. Discharge capacities as a function of cycle number of (a) La1.5Mg0.5Ni7−xAlx and (b) La1.5Mg0.5Ni7−xMnx. 3) electrodes increases. Up to now, the best cycle stability was observed in La2−xMgxNi7 [7]. The phase composition of this material influenced strongly its electrochemical properties. The major phase in the LaMgNi7 sample is LaNi5 [7]. This material has a higher electrochemical cycle stability in comparison with (La, Mg)2Ni7 [37]. On the other hand, La1.5Mg0.5Ni7−xAl0.2 has the best capacity retaining rate after the 50th cycle. It is important to note, that the partial substitution of nickel with aluminum or manganese resulted in the increased cycle stability of the MHx alloy electrodes. Additionally, these chemical modifications also influenced the kinetics of the hydrogen sorption reducing the time of the hydrogenation process.

Discussion
Recently, the research was directed to the new generation of hydrogen storage (La, Mg)2Ni7 materials [2,15,19,24]. These hydrogen storage phases could replace the poor cycle stability of the ZrV2-and LaNi5-type hydrides [15,38]. Many different ways of synthesis of nanostructured hydrogen

Discussion
Recently, the research was directed to the new generation of hydrogen storage (La, Mg) 2 Ni 7 materials [2,15,19,24]. These hydrogen storage phases could replace the poor cycle stability of the ZrV 2and LaNi 5 -type hydrides [15,38]. Many different ways of synthesis of nanostructured hydrogen storage materials are available [39]. The mechanical processes include mechanical alloying or high energy ball milling. MA is an effective process to produce the (La-Mg) 2 Ni 7 alloys with reduced crystallite sizes and fresh surfaces. MA can improve the kinetics of hydrogen absorption and desorption of the processed materials due to large surface areas and as a consequence short hydrogen diffusion pathways. For example, the TiV alloy synthesized by MA shows a multi-crystalline microstructure [34].
The main purpose of our current study is the synthesis of new (La,Mg) 2 Ni 7 -type hydrogen storage alloys via its chemical modification. The effect of the different metals on the phase compositions as well as thermodynamic and electrochemical properties of this system was studied. In the La-Mg-Ni-type system, various crystalline phases could be formed, among which (La,Mg)Ni 3 , (La,Mg) 2 Ni 7 , and (La,Mg) 5 Ni 19 are observed [40,41]. They are composed of the [A 2 B 4 ] and [AB 5 ] subunits alternatively stacking along the c axis [15]. Studies on the thermodynamic and electrochemical behavior of the La 2 Ni 7 -type compounds show that the additional presence of (La,Mg) 5 Ni 19 or LaNi 5 phase has a positive catalytic effect on the charge-discharge process of this alloy [42]. The transitional metals affected the hydrogen absorption/desorption plateau pressure of hydrogen storage materials and influenced their thermodynamic and electrochemical properties [37,[43][44][45].
In this work, the influence of the Al and Mn concentration in the A 2 B 7 -type (La 1.5 Mg 0.5 Ni 7−x M x (M = Al (0 ≤ x ≤ 0.25), Mn (0 ≤ x ≤ 0.5)) materials, synthesized by MA, on the thermodynamic and electrochemical properties was studied. All of the hydrogen storage materials are composed of the La 2 Ni 7 phase (the hexagonal structure-Ce 2 Ni 7 -type and the rhombohedral structure-Gd 2 Co 7 -type). In La 1.5 Mg 0.5 Ni 6.75 Al 0.25 , the AB 3 -type phase was detected, too. Additionally, traces of the La 2 O 3 phase in some of the La 1.5 Mg 0.5 Ni 7−x M x (Al and Mn) alloys were observed. In the La 1.5 Mg 0.5 Ni 6.8 Al 0.2 alloy, the MgO phase was viewed. The abundance of the La 2 Ni 7 -type phase increased from 88.9% (La 1.5 Mg 0.5 Ni 7 alloy) to 99.8% and 97.8% for La 1.5 Mg 0.5 Ni 6.85 Al 0.15 and La 1.5 Mg 0.5 Ni 6.7 Mn 0.3 , respectively. The maximum value of the La 2 Ni 7 phase is observed in the La 1.5 Mg 0.5 Mg 0.5 Ni 6.85 Al 0.15 alloy ( Table 2).
All studied alloys absorb hydrogen at 303 K. The shift from the α-solid solution to the β-hydride phase is observed. The absorption plateau pressure for the La 1.5 Mg 0.5 Ni 7−x M x (M = Al and Mn) are much lower than for La 1.5 Mg 0.5 Ni 7 . The hydrogen sorption pressure of the studied system depends on the Al or Mn contents. Due to the higher stability of the La 1.5 Mg 0.5 Ni 7−x M x (M = Al and Mn) hydrides, a decrease of the sorption pressure was observed for higher Al or Mn contents in the La 1.5 Mg 0.5 Ni 7−x M x system. The highest value of hydrogen content and the discharge capacity was measured for La 1.5 Mg 0.5 Ni 6.8 Mn 0.2 (1.79 wt.%) and La 1.5 Mg 0.5 Ni 6.85 Al 0.15 (328 mAh/g), respectively. On the other hand, the substitution of Ni with Al or Mn in the MA and annealed La 1.5 Mg 0.5 Ni 7−x M x (Al; x = 0.10, 0.15, 0,20, 0.25, and Mn; x = 0.2, 0.3) improved the cycle stability of the synthesized electrodes. Additionally, the stability of the electrochemical discharge capacity increases with the increasing content of Al and Mn up to x = 0.2 and 0.3, respectively. However, a significant reduction in the discharge capacity was measured for the Al and Mn content above x = 0.25 and 0.5, respectively.
An additional increase of hydrogenation properties of these hydrogen storage materials can be established by encapsulation of alloy particles with thin amorphous nickel coating [19]. Modification of the La 1.5 Mg 0.5 Ni 7 particles with an electroless, 1 m thick Ni-P coating weaken the electrodes corrosion process.

Conclusions
In this work, a series of La 1.5 Mg 0.5 Ni 7−x M x (M = Al (0 ≤ x ≤ 0.25), Mn (0 ≤ x ≤ 0.5)) intermetallics were prepared using mechanical alloying and heat treatment. The effect of the substitution of Ni with Al and Mn on the sample phase compositions, thermodynamic and electrochemical properties of the A 2 B 7 -type La-Mg-Ni-M-based materials was investigated. Partial replacement of Ni with Al and Mn in the La-Mg-Ni alloy improved the hydrogen sorption of this system. For the La 1.5 Mg 0.5 Ni 6.9 Al 0.1 and La 1.5 Mg 0.5 Ni 6.8 Mn 0.2 alloys, the time required to get 95% of the maximum hydrogen capacity at the 3rd cycle decreases to reach 5 and 6 min, respectively. On the other hand, the La 1.5 Mg 0.5 Ni 6.85 Al 0.15 alloy has a high discharge capacity, then that of the La 1.5 Mg 0.5 Ni 7 . Moreover, the change of Ni with Al or Mn in La 1.5 Mg 0.5 Ni 7−x M x also enhanced the stability of the discharge capacity. Generally, the improvement of the properties of the La-Mg-Ni-M-based hydrogen storage alloys discussed in this manuscript is the function of the phase composition and final microstructure of the synthesized hydrogen storage material.

Conflicts of Interest:
The authors declare no conflict of interest.