Mg1.8La0.2Ni−xNi Nanocomposites for Electrochemical Hydrogen

J. Phys. Chem. B 2006, 110, 25769-25774

25769

Mg1.8La0.2Ni-xNi Nanocomposites for Electrochemical Hydrogen Storage Huabin Yang,* Haichang Zhang, Wei Mo, and Zuoxiang Zhou Institute of New Energy Material Chemistry, Nankai UniVersity, Tianjin 300071, China ReceiVed: July 27, 2006; In Final Form: October 15, 2006

Mg1.8La0.2Ni hydrogen storage alloy was ball-milled with Ni powder, leading to the formation of a nanocrystalline and amorphous microstructure with particle sizes less than 50 nm in diameter. Each sample was examined by transmission electron microscopy (TEM), energy-dispersive spectroscopy (EDS), and X-ray diffraction (XRD). This structure was beneficial for the reduction of electrochemical impedance, as well as significant improvement of its discharge capacity, cycle life, and rate capability for electrochemical hydrogen storage in an alkaline solution. When the molar ratio (x) of Ni over Mg1.8La0.2Ni was equal to 2, the dehydriding capacity reached 2.55 wt % from electrochemical pressure-temperature isotherms (P-C-T). It was in good agreement with its initial discharge capacity, 716 mA‚h/[g of (Mg1.8La0.2Ni)], observed from the electrochemical charge and discharge process. After 50 cycles, its discharge capacity still reached 381 mA‚h/[g of (Mg1.8La0.2Ni)]. Further results showed that this composite had a promising high rate capability. At the current density of 1200 mA/g its discharge capacity reached 48% of its initial capacity.

1. Introduction In recent years, extensive research has been concentrated on various potential hydrogen storage materials, such as carbon materials,1,2 metal nitrides,3 metal-organic materials,4 alloys and intermetallic compounds,5,6 and complex metal hydrides.7-12 However, only some metal hydrides and complex metal hydrides5,6,13-19 are considered possible candidates for Ni/MH batteries. Among these metal hydrides, Mg-Ni-based hydrogen storage alloys are recognized as possible candidates for applications in Ni/MH batteries, due to their higher hydriding and dehydriding capacities, lower specific gravities, richer mineral resources, lower cost, etc. The theoretical discharge capacity of Mg2Ni is estimated to be 1000 mA‚h/g, which is almost 2.6 times as high as that of AB5-type (LaNi5, 372 mA‚h/g) alloys and 1.3 times as high as that of AB2-type (ZrV2, 763 mA‚h/g) alloys. However, at present, the Mg2Ni alloy is still unsatisfactory for practical applications in Ni/MH batteries because of two significant disadvantages: (i) poor hydriding and dehydriding properties at ambient temperature, and (ii) easy corroding in alkaline solutions. Much progress has been made since Lei et al.20 reported that amorphous Mg-Ni alloys prepared by ball-milling could reach a high initial discharge capacity (ca. 500 mA‚h/[g of (MgNi)]) in alkaline solutions at room temperature, which presented a possibility for the applications of Mg-Ni-based alloys in Ni/ MH batteries.21-27 Kohno et al.21,22 ball-milled a mixture of ascast Mg2Ni alloy and Ni powder to reach a higher initial discharge capacity (750 mA‚h/[g of (Mg2Ni)]). They also found that some replacement of Mg by Al could improve the electrochemical reversibility of Mg2Ni alloy electrode. Iwakura23,24 et al. reported an even higher discharge capacity of 1082 mA‚h/[g of (Mg2Ni)] by ball-milling Mg2Ni with 70 wt % Ni powder. Kohno25 et al. also found that the discharge capacity of an electrode of the Mg2Ni alloy ball-milled with Ni powder reached 830 mA‚h/[g of (Mg2Ni)]. It is notable that although * To whom correspondence should be addressed. Telephone: +86-2223508405. Fax: +86-22-23502604. E-mail: [email protected].

all the improved alloys mentioned above showed better cycle lives after ball-milling with Ni powder, their cyclicalities were still far from the requirements for the Ni/MH batteries. Their discharge capacity generally degraded to lower than 400 mA‚ h/[g of (Mg2Ni)] after 20 cycles. To further improve the electrochemical hydrogen storage performance of the Mg-Ni-based alloys, an Mg1.8La0.2Ni alloy was developed in our group. Here, the Mg1.8La0.2Ni alloy was chosen as the test alloy because our previous work showed that La replacement of part of Mg could improve the anticorrosion behavior of Mg-Ni-based alloys in alkaline solutions.26 According to our knowledge, few papers on Mg replacement by La have been reported.26,27 Meng et al.27 found that Mg0.95La0.05Ni alloy electrode could retain 220 mA‚h/[g of (Mg0.95La0.05Ni)] up to the 30th cycle. This research is an extension of our previous work26,28 aimed at developing new composites and providing an effective way for their applications in Ni/MH batteries. 2. Experimental Section 2.1. Sample Preparation. A Fritsch planetary ball miller P-6 (Germany) was employed for the preparation of the composites of the Mg1.8La0.2Ni alloy powders (200-300 mesh) and Ni powder (INCO 255) with a ball-to-powder ratio of 20:1 for 30 h at a rotational velocity of 500 rev/min. Here, the Mg1.8La0.2Ni alloy powder was synthesized by ball-milling followed by annealing.29,30 2.2. Structural and Morphological Characterization. The crystal structure and morphology of the as-prepared samples were determined by X-ray diffraction (XRD; Rigaku D/max2500 with Cu KR radiation) and transmission electron microscopy (TEM; FEI Tecnai 20), respectively. 2.3. Electrochemical Hydrogen Storage. The electrochemical performance of each alloy electrode was conducted in a triple-compartment cell. The Mg1.8La0.2Ni-xNi composites were pressed to form pellets at 30 MPa and serve as the working electrodes. An aqueous solution containing 5 mol/L KOH was employed as an electrolytic solution.

10.1021/jp0647838 CCC: $33.50 © 2006 American Chemical Society Published on Web 12/01/2006

25770 J. Phys. Chem. B, Vol. 110, No. 51, 2006

Figure 1. XRD patterns of Mg1.8La0.2Ni-xNi composites. (a) Crystalline Mg1.8La0.2Ni; (b) ball-milled Mg1.8La0.2Ni; (c) ball-milled Mg1.8La0.2Ni-Ni; (d) BALL-milled Mg1.8La0.2Ni-2Ni; (e) ball-milled Mg1.8La0.2Ni-3Ni.

Electrochemical impedance spectroscopy (EIS) was conducted at depth of discharge (DOD) ) 50%. A commercial nickel oxide electrode served as a counter electrode, and an HgO/Hg electrode served as the reference electrode. A Solartron 1287 electrochemical interface (EI), coupled with a Solartron 1250 frequency response analyzer (FRA), was used for the EIS measurements. The ac amplitude was 5 mV, and the frequency range employed was between 104 and 10-2 Hz. Electrochemical pressure-temperature isotherms (P-C-T) and electrochemical cycle lives were examined in unsealed cells. The Mg1.8La0.2Ni-xNi composites served as the negative electrodes, and commercial sintered NiOOH/Ni(OH)2 plates were used as positive electrodes. Each negative electrode was placed in the central compartment, and two pieces of sintered Ni electrodes of the same size were placed on either side. A sulfonated polypropylene nonwoven separator used in commercial batteries was used to separate the positive and negative electrodes. Each negative electrode was charged at 500 mA/g for 2.5 h; after a 5 min rest, it was discharged to 1 V at 50 mA/g. For the electrochemical P-C-T tests, each steady-state discharge voltage was recorded at every 30 min interval. All the tests were performed using a computer-controlled galvanostatic testing system at 25 °C. 3. Results and Discussion 3.1. Phase Structure. Figure 1a shows the XRD pattern of the crystalline Mg1.8La0.2Ni alloy. Two phasessMg2Ni phase and LaNi3 phasesappeared after La replacement.26 Figure 1b-e shows the XRD patterns of the Mg1.8La0.2Ni alloys ball-milled with different amounts of Ni powder. A broadening of diffraction peaks is clearly observed for all the ball-milled composites. Here, 30 h was selected as the ball-milling time because no significant change was observed from the XRD patterns at times even longer than 30 h. After ball-milling for 30 h, some diffraction peaks indexed to the Mg2Ni phase and the LaNi3 phase of the Mg1.8La0.2Ni alloy disappear, while several diffraction peaks indexed to the Mg2Ni phase remain and broaden at 2θ angles of 18-22°, 33-37°, 38-41°, and 4346° (Figure 1b). This indicates that the Mg1.8La0.2Ni alloy is transformed to an amorphous phase from a crystalline phase, to some extent. Moreover, only one peak can be observed for all the composites ball-milled with Ni powder, where crystalline diffraction peaks disappear completely and a single hump

Yang et al. appears at 2θ angles of 40-45°. This profile indicates that the Mg1.8La0.2Ni alloys, after ball-milling with Ni powder, are partly transformed to amorphous phases from the crystalline phases. Kohno et al. reported that an amorphous structure of the Mg2Ni alloy was formed after ball-milling, while Ni still remained in the crystalline state, where a sharp peak indexed to Ni was clearly observed.21 In the present work, no sharp peaks identified as crystalline metallic Ni are observed. However, it can be observed from Figure 1c-e that the peak positions shift from 41.84° for the Mg1.8La0.2Ni-Ni composite to 42.98° for the Mg1.8La0.2Ni-2Ni and Mg1.8La0.2Ni-3Ni composites. It is known that the peak position of Ni should be at 2θ angles of 44-45°. The single hump appearing at 2θ angles of 40-45° suggests that the peak indexed to Ni is overlapped with that of Mg1.8La0.2Ni. This peak position shift implies that Ni does exist although no sharp peaks indexed to Ni are observed from the XRD pattern. 3.2. Morphological and Microstructural Characterization. In order to further clarify the structural change after the ball-milling, the morphology and the microstructure of the Mg1.8La0.2Ni-2Ni composite are examined as shown in Figure 2. The as-prepared composite is up to 50 nm in diameter as shown in Figure 2a. From the high-resolution TEM (HRTEM) image in Figure 2b, it can be seen that this composite has a nanocrystalline and amorphous composite structure, instead of a completely amorphous structure. This is in a good agreement with the result of the ringlike diffraction pattern with spots around it as shown in the inset SAED (selection area electron diffraction) image. It can be clearly seen that these diffraction rings as illustrated beside the inset SAED image correspond to four rings. Here, R1, R2, R3 and R4 are named as the radii of the four rings from the inner to the outer side, respectively. It can be calculated that R12:R22:R32:R42 is equal to 3:4.56:7.32:10.41. This is in good agreement with 3:4:8:11, which indexed to (111), (200), (220), and (311) of the (hkl) values of the Ni, respectively. Therefore, the diffraction rings can be attributed to the existence of the Ni phase, which is in good agreement with the result from the XRD pattern. The element distribution from the energydispersive spectrum (EDS; Figure 2c) shows that the composite consists of Mg, La, Ni, and O elements. The O signals mainly come from the formation of oxides during the preparation process of the composites. However, no oxides are observed in the XRD pattern shown in Figure 1. The Cu and C signals observed in Figure 2c come from the Cu grid and the carbon film supporting the specimen in the TEM observation. The analysis results show that the content of Ni reaches 70.2 atom %. Because the molar ratio (x) of Ni over Mg1.8La0.2Ni is equal to 2, the total content of Ni in the composite is 7/9 (77.8 atom %) and the introduced Ni should be 2/3 (66.7 atom %) theoretically. Thus, the amount of Ni from the EDS analysis should include the Ni contained in Mg1.8La0.2Ni and the introduced Ni. The Ni is highly dispersed throughout the nanocrystalline and amorphous phase, working as catalytic sites for electrochemical hydriding and dehydriding. On the other hand, this highly dispersed Ni acts as a barrier, which prevents the alloy bulk from further corroding during electrochemical cycling. Therefore, the ball-milling leads to the formation of an amorphous and nanocrystalline composite, which is believed to play an important role in the significant improvement of its electrochemical hydrogen storage performance. 3.3. Electrochemical Impedance Spectra. The electrochemical reactions and the proposed equivalent circuit for the frequency response of the electrode have been given in our previous work.26,28,31 The impedance spectra of the electrodes

Mg1.8La0.2Ni-xNi Composites for Hydrogen Storage

J. Phys. Chem. B, Vol. 110, No. 51, 2006 25771

Figure 3. Electrochemical impedance spectroscopy of Mg1.8La0.2NixNi alloy electrodes at DOD ) 50% with the equivalent circuit inset. (a) Crystalline Mg1.8La0.2Ni; (b) ball-milled Mg1.8La0.2Ni; (c) ball-milled Mg1.8La0.2Ni-Ni; (d) ball-milled Mg1.8La0.2Ni-2Ni; (e) ball-milled Mg1.8La0.2Ni-3Ni.

TABLE 1: Simulated Electrochemical Parameters of Mg1.8La0.2Ni-xNi Alloy Electrodes sample

R2, mΩ‚g

R3, mΩ‚g

crystalline Mg1.8La0.2Ni ball-milled Mg1.8La0.2Ni ball-milled Mg1.8La0.2Ni-Ni Ball-milled Mg1.8La0.2Ni-2Ni ball-milled Mg1.8La0.2Ni-3Ni

53 27 20 20 20

123 101 60 31 81

at the open-circuit voltage (OCV) of DOD ) 50% are shown in Figure 3. It can be seen clearly that all the spectra consist of two depressed overlapping semicirclessa small capacitive semicircle at high frequencies and a large capacitive semicircle at middle frequenciessas well as a Warburg linear region at low frequencies. The impedance spectra are thus modeled with the equivalent circuit as shown in the inset of Figure 3. The solution resistance, R1, is used to account for the resistance of ions through electrolyte and separator. Two Randles-Ershler circuits of parallel R-CPE (constant phase element) combinations are used to represent the electrolyte interface reactions. Here, constant phase elements (CPEs) are used in the circuit model since the porosity, roughness, and inhomogeneity of the electrode surface should be considered instead of capacitance (C). Resistor R2 and constant phase element CPE1 correspond to the oxide film R-CPE circuit, which is attributed to a deviation from the relaxation process. It is possible to attribute this depression to the spread in relaxation times of the surface phenomenon. This process is controlled by some structural factors such as the thickness and porosity of the oxide film.28 Resistor R3 corresponds to the charge-transfer resistance of the following electrochemical and chemical reactions across the interfaces:28

M + H2O + e S M‚‚‚Hads + OHand

M‚‚‚Hads + M‚‚‚Hads S H2 + 2M

Figure 2. TEM, HRTEM, and SAED images and EDS spectrum of Mg1.8La0.2Ni-2Ni composite after ball-milling for 30 h. (a) TEM image; (b) HRTEM images with SAED image inset; (c) EDS spectrum.

Constant phase element CPE2 corresponds to the double layer across the interface. The Warburg impedance in the alloy, W1, is proposed to describe the diffusion of hydrogen in the alloy. The component values derived from Figure 3 are summarized in Table 1. It can be seen that for the crystalline Mg1.8La0.2Ni alloy electrode R2 is 53 mΩ‚g, while for the ball-milled one, 27 mΩ‚g is about half that value. This implies that more oxides are formed on the crystalline alloy surface. After the introduction

25772 J. Phys. Chem. B, Vol. 110, No. 51, 2006 of Ni R2 is further reduced to 20 mΩ‚g and keeps the same value for all the Ni-introduced composites. The reduction of R2 is ascribed to the reduced content of the formation of oxides because of the introduction of Ni. For the ball-milled Mg1.8La0.2Ni-xNi composites R2 is independent of the content of Ni. From Table 1, it can be also found that R3 is 123 mΩ‚g for the crystalline alloy electrode, while it is 101 mΩ‚g for the ballmilled one. This is mainly due to the reduction of the particle size after the ball-milling. After the introduction of Ni R3 is reduced to 60, 31, and 81 mΩ‚g for x ) 1, 2, and 3, respectively. The reduction of R3 is mainly attributed to the introduced Ni, which provides the pathway to absorb and desorb hydrogen quickly and easily. However, when x is increased to 3, R3 increases from 31 mΩ‚g (x ) 2) to 81 mΩ‚g (x ) 3). This should be attributed to too high Ni content, leading to an increase of the electrochemical reaction resistance. This phenomenon also happened in other hydrogen storage alloys.14 3.4. Electrochemical P-C-T. The electrochemical P-C-T is obtained at steady state to evaluate the charged and discharged pressures of the hydrogen storage alloy electrodes.32 This method is quite useful to obtain the P-C-T curves of the metal hydride (MH) electrode over a wide pressure range from 102 to 10-8 atm. This technique allows one to evaluate hydrogen desorption behaviors in amorphous alloys.33 In the equilibrium state, the hydrogen partial pressure P(H2) on the electrode can be equal to the hydrogen equilibrium pressure Peq(H) of the MH as follows:

Yang et al.

Figure 4. Electrochemical P-C-T curves of Mg1.8La0.2Ni-xNi alloy electrodes. (a) Crystalline Mg1.8La0.2Ni; (b) ball-milled Mg1.8La0.2Ni; (c) ball-milled Mg1.8La0.2Ni-Ni; (d) ball-milled Mg1.8La0.2Ni-2Ni; (e) ball-milled Mg1.8La0.2Ni-3Ni.

2M + H2 S 2MH Therefore, an equilibrium potential Eeq(H) can be related to the equilibrium hydrogen pressure of alloy according to the Nernst equation. The Eeq(H) value against that of the HgO/Hg reference electrode is expressed as follows:34

Eeq(H) vs Eeq(HgO/Hg) RT R(H2O) ln ) [E°(H) - E°(HgO/Hg)] + 2F R(H2) ) [E°(H) - E°(HgO/Hg)] +

R(H2O) RT ln 2F γ(H2) Peq(H2)

where E°(H) and E°(HgO/Hg) are the standard electrode potentials of the H2O/H couple and the HgO/Hg couple, respectively, R(H2O) is the activity of water, R(H2) is the activity of hydrogen, and γ(H2) is the fugacity coefficient. The equilibrium potential at 20 °C and 6 M KOH under 1 atm is expressed by the following equation, in good agreement with the data from the solid-gas reaction:17,34

Eeq(H) vs Eeq(HgO/Hg) ) -0.9324 - 0.0291 log Peq(H2) Based on the above concept, electrochemical P-C-T curves for all the alloy electrodes for the first dehydrogenation process are shown in Figure 4. It can be seen that the dehydride capacity is only 0.31 wt % for the crystalline alloy, while it is 0.86 wt % for the ball-milled one. After the introduction of Ni, it reaches 1.11, 2.56, and 2.31 wt % corresponding to x ) 1, 2, and 3, respectively. Because the ball-milled composites have nanocrystalline and amorphous structure, there are no clear dehydriding plateaus observed from all the dehydriding curves. The Ni, dispersed throughout the nanocrystalline and amorphous structure, works as catalytic sites for electrochemical hydriding and dehydriding, and acts as a barrier that prevents the alloy

Figure 5. Discharge capacity as a function of cycle number for Mg1.8La0.2Ni-xNi alloy electrodes. (a) Crystalline Mg1.8La0.2Ni; (b) ballmilled Mg1.8La0.2Ni; (c) ball-milled Mg1.8La0.2Ni-Ni; (d) ball-milled Mg1.8La0.2Ni-2Ni; (e) ball-milled Mg1.8La0.2Ni-3Ni.

bulk from further corroding during electrochemical cycling. Therefore, the introduced Ni is effective for the improvement of electrochemical hydrogen storage capacity. 3.5. Electrochemical Hydrogen Storage Performance. Figure 5 shows the cycle lives of the Mg1.8La0.2Ni-xNi (x ) 1, 2, and 3) alloy electrodes including the crystalline and ballmilled Mg1.8La0.2Ni alloys during electrochemical hydrogen storage. Parts I and II, respectively, of Figure 6show the discharge curves for the first and 50 cycles. It can be found that the initial discharge capacity of each electrode is quite different. In the case of the Mg1.8La0.2Ni alloy electrode, whether ball-milled or not, its initial discharge capacity is quite low. However, for the Mg1.8La0.2Ni composites ball-milled with Ni powders, at least 300 mA‚h/[g of (Mg1.8La0.2Ni)] for the initial discharge capacity is observed. This is caused by the different phases according to XRD patterns (Figure 1) and the appearance from TEM and SAED images (Figure 2). For x ) 2, the discharge capacity is 716 mA‚h/[g of (Mg1.8La0.2Ni)], which is nearly equal to 71.6% of the theoretical capacity and much greater than that of the ball-milled Mg1.8La0.2Ni alloy. This discharge capacity, 716 mA‚h/[g of (Mg1.8La0.2Ni)], is in good agreement with the value derived from the electrochemical P-C-T results, 2.55 wt % (Figure 4d). It can be also found that the cycle performance of the Mg1.8La0.2Ni-2Ni composite

Mg1.8La0.2Ni-xNi Composites for Hydrogen Storage

J. Phys. Chem. B, Vol. 110, No. 51, 2006 25773

Figure 7. Rate capability of Mg1.8La0.2Ni-xNi alloy electrodes as a function of discharge current densities. (a) Crystalline Mg1.8La0.2Ni; (b) ball-milled Mg1.8La0.2Ni; (c) ball-milled Mg1.8La0.2Ni-Ni; (d) ballmilled Mg1.8La0.2Ni-2Ni; (e) ball-milled Mg1.8La0.2Ni-3Ni.

Figure 6. Discharge curves of Mg1.8La0.2Ni-xNi alloy electrodes. (I) First cycle; (II) 50th cycle. (a) Crystalline Mg1.8La0.2Ni; (b) ball-milled Mg1.8La0.2Ni; (c) ball-milled Mg1.8La0.2Ni-Ni; (d) ball-milled Mg1.8La0.2Ni-2Ni; (e) ball-milled Mg1.8La0.2Ni-3Ni.

is much better than those of others. For the crystalline alloy (x ) 0), the discharge capacity after 50 cycles is only 10 mA‚h/[g of (Mg1.8La0.2Ni)], and the ball-milled one (x ) 0) also has a very low value (25 mA‚h/[g of (Mg1.8La0.2Ni)]). Although our previous work showed that La replacement of part of Mg could improve the anticorrosion behavior of Mg-Ni-based alloys in alkaline solutions,26 the crystalline alloy (x ) 0) is still easy to corrode in the alkaline solution, which causes the degradation of capacity due to the Mg and La irreversible corrosion. After transforming into an amorphous phase by ball-milling, this alloy (x ) 0) degrades even faster due to the fine particle size. In contrast, for the composites ball-milled with Ni powders, the discharge capacity after 50 cycles is much higher. It is 183 mA‚ h/[g of (Mg1.8La0.2Ni)] for x ) 1, 381 mA‚h/[g of (Mg1.8La0.2Ni)] for x ) 2, and 279 mA‚h/[g of (Mg1.8La0.2Ni)] for x ) 3, respectively. These results indicate that the introduced Ni improves the cyclicality of the Mg-Ni-based alloys. The significant improvements seem to be ascribable to the formation of the nanocrystalline and amorphous composite structure. The decrease of the discharge capacity for up to x ) 3 is due to the too high introduced Ni content. For the composites ball-milled with Ni powders, it can be found that the tendency of the discharge capacity during cycling can be classified into two stages (Figure 5): a fast capacity degradation one during the initial cycles and a slow capacity degradation one during the subsequent cycles. Although the cyclic durability of the Mg1.8La0.2Ni alloy electrode is significantly improved by the ball-milling with Ni powders, the capacity degradation is still fast during initial cycling. After ball-

milling with Ni powders, each alloy is transformed into the nanocrystalline and amorphous composite structure. The fine particles of the composites are more easily corroded in the alkaline solution, and Mg and La on the surface were oxidized, which causes irreversible capacity degradation during cycling. Therefore, the fast stage is observed for the initial cycling. However, the capacity degradation is significantly reduced because the highly dispersed Ni covered on the surface acts as a barrier for preventing the bulk from further corroding, besides acting as the catalyst for the electrochemical hydriding and dehydriding reaction that happened on the surface during the subsequent cycling. Therefore, the slow capacity degradation stage is observed. In addition, La replacement of part of Mg in the Mg-Ni-based alloys does not show remarkable evidence of the significant improvement in the electrochemical cycle lives in this paper, although it is proved that the existence of La restrains the corrosion of Mg, where La prevents some of the Mg from further corroding due to the formation of La oxides/ hydroxides on the surface.28,35 However, it is noted that, although all the improved alloys mentioned in the Introduction show better cycle lives after ball-milling with Ni powder, their discharge capacity generally degrades to less than 400 mA‚h/ [g of (Mg2Ni)] even after 20 cycles. However, in our present work, a higher discharge capacity after 50 cycles is observed. For example, it is 381 mA‚h/[g of (Mg1.8La0.2Ni)] for the Mg1.8La0.2Ni-2Ni composite. Figure 7 shows the rate capability of the Mg1.8La0.2Ni-xNi (x ) 1, 2, and 3) alloy electrodes, including the crystalline and ball-milled Mg1.8La0.2Ni alloys. At lower discharge rates (