Study on the hydrogen storage and electrochemical properties of Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys

Journal of Alloys and Compounds 623 (2015) 311–316

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom

Study on the hydrogen storage and electrochemical properties of Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys Lan Zhiqiang, Peng Wenqi, Fu Shuying, Wei Wenlou, Wei Ningyan, Guo Jin ⇑ Guangxi Experiment Centre of Science and Technology, College of Physics Science and Technology, Guangxi University, Nanning 530004, China

a r t i c l e

i n f o

Article history: Received 24 June 2014 Received in revised form 9 October 2014 Accepted 20 October 2014 Available online 31 October 2014 Keywords: Hydrogen storage alloy Hydrogen storage property Electrochemical property

a b s t r a c t Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys were prepared by the magnetic induction melting method. The influence of Al content on the hydrogen storage and electrochemical properties of the alloy was investigated. The results show that the hydrogen storage capacity is gradually reduced as the Al content increases. The (La,Pr,Nd)Ni5 cell volume and the change of enthalpy also decrease as Al is added. Although the discharge capacity decreases with increasing Al content, the addition of Al can reduce the stability of Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) hydride and improve the performance of hydrogen desorption thermodynamics. For the alloy electrode without Al, the maximum discharge capacity (Cmax) and retention discharge capacity after 100 charge–discharge cycles (C100) is 385 mA h g1 and 202 mA h g1, respectively. For the alloy electrode with x = 0.4, while Cmax is only 323 mA h g1, C100 is 273 mA h g1, which is much higher than that of the alloy without Al. The addition of Al can improve the charge– discharge cycle lifetime effectively and can increase the limiting current IL. The kinetic performance of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrode can also be improved by increasing the Al content. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction Re–Mg–Ni-based hydrogen storage alloys have been attracted great attention as a negative electrode material due to their high energy density, long cycle life and high charge/discharge capacity [1–9]. For example, the effects of substitution of R(R = Zr or Nd) for La on the structure and electrochemical properties of the La–Mg–Ni hydrogen storage alloys have been investigated by Pan et al. [2,3]. In their studies, the electrochemical kinetics of the electrodes can be effectively improved by partial substitution of R(R = Zr or Nd) for La, in spite of the maximum discharge capacity of the alloy electrodes monotonously decreased as increasing the R(R = Zr or Nd) content. Liu et al. [4] have also investigated systematically the cycling behavior of the La0.7Mg0.3Ni2.65xCo0.75Mn0.1Alx (x = 0, 0.3) alloy electrodes, and they found that the formation of a dense Al oxide film during cycling was one of important factors for the improvement of the cycling stability of the La–Mg–Ni-based alloy electrodes with Al. Al is a valuable element in improving the cycle stability of the La–Mg–Ni-based alloys with the amount of Co reduced [5]. Balogun et al. [7] investigated that LaNi4.2Co0.3 Mn0.3Al0.2, a single-phase alloy with hexagonal CaCu5 type ⇑ Corresponding author. E-mail address: [email protected] (G. Jin). http://dx.doi.org/10.1016/j.jallcom.2014.10.097 0925-8388/Ó 2014 Elsevier B.V. All rights reserved.

structure, presented the maximum discharge capacity of 330.4 mA h g1. Wang et al. [9] developed a Co-free La0.8Mg0.2Ni3.4Al0.1 alloy and found that the discharge capacity increased to 391.8 mA h g1 after surface was modified with polyaniline, compared to 382.5 mA h g1 for the bare alloys. Although the La–Mg–Ni-based alloy has a high hydrogen storage capacity, the cyclic durability of this type alloy is not so satisfied for practical applications. Co is a key element for improving charge–discharge cyclic stability in La–Mg–Ni-based alloys [10], but the charge– discharge capacity and the kinetic properties are degraded as the Co content increases [11]. Therefore, how to improve the comprehensive electrochemical properties of low-Co Re–Mg–Ni-based hydrogen storage alloys is a topic worth investigating. Gao et al. [12] researched the La0.8xGd0.2MgxNi3.1Co0.3Al0.1(x = 0.1–0.5) alloys and found that the chemical composition and crystalline phase structure of these alloys closely depended on the Mg content (x). It showed that the cycle stability was improved dramatically for x = 0.15. However, both the cyclic stability and discharge capacity of the alloy electrodes decrease as the Mg content further increases. Chen et al. [13] investigated electrochemical properties of La0.78Ce0.22Ni3.73Mn0.30Al0.17FexCo0.8x (x = 0, 0.2, 0.5, 0.8) alloys and found that the discharge capacity and high rate discharge ability were weaken as increasing Fe content. Here, in order to investigate the effect of Al element on the comprehensive properties of

312

L. Zhiqiang et al. / Journal of Alloys and Compounds 623 (2015) 311–316

the low-Co Re–Mg–Ni alloy, the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0– 0.4) alloys were prepared by magnetic levitation melting under argon atmosphere, and then hydrogen storage and electrochemical properties were discussed in detail. 2. Experimental The Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys were prepared by magnetic levitation melting under an argon atmosphere. The La-rich mischmetal (Mm) used in this work contained La85.2%, Ce3.3%, Pr2.6% and Nd8.9%. The purity of La, Ce, Pr, Nd, Mg, Ni and Co was above 99.9 wt%. The ingots were turned over and remelted three times to ensure good homogeneity, and then part of the alloys were mechanically crushed and ground into 200 mesh powders for the experiments. The crystal structures of the samples were characterized using a Rigaku D/max 2500 V diffractometer. For scanning electron microscope (SEM) analysis, the sample was encapsulated in epoxy resin for polishing, and then the polished surface was etched with a 60% HF aqueous solution. The morphology and elemental composition of the ingot samples were investigated using an SU-8020/X-MAX80 Field Emission Scanning Microscope equipped with an Energy Dispersive Spectrometer (EDS). All the electrodes were prepared by mixing the alloy powder with carbonyl nickel powder in a weight ratio of 1:4 and then cold-pressed under a pressure of 20 MPa into a pellet of 10 mm diameter and about 1 mm thickness. Electrochemical measurements were performed at room temperature in a tri-electrode open cell. NiOOH/Ni(OH)2 was adopted as the counter electrode, the Hg/HgO was adopted as a reference electrode and 6 mol L1 KOH solution was used as the electrolyte. For activation and charge/discharge cycling, all the alloy electrodes were measured by a DC-5 battery testing instrument at a 100 mA g1 charge rate for 5 h, followed by a 10 min rest and then discharged at an 80 mA g1 discharge rate to a cut-off cell potential of 0.5 V relative to the Hg/HgO reference electrode. The electrochemical characteristics of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0– 0.4) alloy electrodes were investigated in the following procedures. The high rate discharge (HRD) was tested on an automatic Arbin-BT2000 battery-testing instrument. The linear polarization and the Tafel polarization were carried out on the GAMRY Corrosion Electrochemical Measurement System. A scanning rate of 10 mV s1 was used from 0.3 to 1.0 V for the Tafel curve. The hydrogen diffusion coefficients were measured by means of the constant potential step discharge technique. The test electrodes in a fully charged state were discharged at a constant potential step of 600 mV for 7200 s.

3. Results and discussion 3.1. Microstructure Fig. 1 shows the XRD patterns of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys. It can be seen that all of the alloys mainly contain (La,Pr,Nd)Ni5 with a hexagonal CaCu5-type structure and LaMg2Ni9 with a PuNi3-type structure phase. Another CaCu5-type structure LaAlNi4 phase appears as the Al content increases. The lattice parameters of (La,Pr,Nd)Ni5, LaMg2Ni9 and LaAlNi4, calculated from the XRD data by Jade 6.0 software, are listed in Table 1.

As shown in Table 1, the cell volumes of the (La,Pr,Nd)Ni5 and LaMg2Ni9 phase decrease as Al is added. The cell volume of the (La,Pr,Nd)Ni5 phase decreases from 86.98 Å3 to 86.11 Å3 as the x value increases from 0.0 to 0.4, while the cell volume of the LaAlNi4 phase first decreases from 90.14 Å3 to 90.07 Å3 as the x value increases from 0.1 to 0.3, and then increases to 90.19 Å3 as the x value further increases to 0.4. Fig. 2a shows the SEM image for the as-cast Mm0.75Mg0.25Ni3.5 Co0.2Al0.1 alloy as a representative example of Mm0.75Mg0.25Ni3.5 Co0.2Alx (x = 0.0–0.4) alloys. The element composition of the alloy is identified by EDS, and the result is illustrated in Fig. 2b and c. EDS (Fig. 2b) analysis for region A (the dark region in Fig. 2a) shows an atomic ratio of 8.86La, 16.5 Mg and 68.45Ni, which is close to the ideal atomic ratio of 8.33La, 16.67 Mg and 75Ni for the LaMg2Ni9 phase. According to EDS and XRD analysis, region A should be the LaMg2Ni9 phase. Region B (the gray region in Fig. 2a) corresponds to the CaCu5-type structure of the (La,Pr,Nd)Ni5 and LaAlNi4 phases. The abundances of all elements in the alloy agree well with the expected values. In terms of the phase structure of the alloy, LaMg2Ni9 is one of the three main phases in the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0– 0.4) alloys. However, the hydrogen storage capacity of LaMg2Ni9 is only about 0.33 wt% H/M [14], which is much lower than that of the CaCu5-type structure. Therefore, only the (La,Pr,Nd)Ni5 and LaAlNi4 phases are considered as the main hydrogen storage phase in the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys. 3.2. Hydrogen storage properties The PCT curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys at different temperatures are shown in Fig. 3. It can be seen that the hydrogen storage capacity and the plateau pressure of the alloys decrease with increasing Al content. It is well known that hydrogen atoms mainly occupy empty sites in hydrogen storage in a bulk alloy. Therefore, the shrinkage of the (La,Pr,Nd)Ni5 cell volume may result in a decrease in the hydrogen storage capacity of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy. According to the PCI curves shown in Fig. 3, and based on the Van’t Hoff equation ln P ¼ DRTH  DRS, the Van’t Hoff plots are shown in Fig. 4. The change of enthalpy, DH, is calculated from the linear slope in Fig. 4 and is given in Table 2. The results show that DH is 25.17 kJ mol1, 23.38 kJ mol1, 19.68 kJ mol1, 22.18 kJ mol1 and 21.12 kJ mol1 for x = 0.0, 0.1, 0.2, 0.3 and 0.4, respectively. It is interesting to note that the value of DH decreases as Al is added, which indicates that the addition of Al reduces the stability of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) hydride. This means that Table 1 The lattice parameters of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys.

Fig. 1. XRD patterns of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys.

Cell volume (Å3)

Samples

Phase

Lattice parameter a (Å)

c (Å)

x = 0.0

(La,Pr,Nd)Ni5 LaMg2Ni9

5.015 4.993

3.993 24.17

86.98 521.9

x = 0.1

(La,Pr,Nd)Ni5 LaAlNi4 LaMg2Ni9

5.024 5.060 4.983

3.991 4.066 24.16

86.55 90.14 519.5

x = 0.2

(La,Pr,Nd)Ni5 LaAlNi4 LaMg2Ni9

5.017 5.060 4.990

3.970 4.062 24.10

86.54 90.08 519.6

x = 0.3

(La,Pr,Nd)Ni5 LaAlNi4 LaMg2Ni9

4.995 5.079 5.000

3.992 4.059 24.01

86.26 90.07 519.9

x = 0.4

(La,Pr,Nd)Ni5 LaAlNi4 LaMg2Ni9

4.996 5.062 4.987

3.983 4.064 24.08

86.11 90.19 518.8

L. Zhiqiang et al. / Journal of Alloys and Compounds 623 (2015) 311–316

313

the metal hydride electrode in alkaline solution and the equilibrium hydrogen pressure can be expressed by E(V vs. Hg/ HgO) = 0.932–0.0296lnP(H2) [15]. Here, E is the potential of the metal hydride alloy electrode and P(H2) is the equilibrium hydrogen pressure. In general, the cell volume is one of the factors that influence the equilibrium hydrogen pressure in hydrogen absorption–desorption processes. Therefore, the decrease of the discharge voltage platform may be attributed to the shrinkage of the (La,Pr,Nd)Ni5 cell volume. In order to further investigate the electrochemical properties of Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes, the charge/ discharge cycle stabilities are tested, and are shown in Fig. 6. The alloy electrodes with added Al are fully activated in two charge– discharge cycles, while the alloy electrode without Al(x = 0) needs five charge–discharge cycles to reach full activation. The result shows that a proper amount of Al can improve the catalytic activity for Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes. This improvement in activation performance may result from the good conductivity of Al. Although the maximum discharge capacity Cmax decreases with increases in Al content, the alloy electrode with added Al shows a higher retention capacity C100 than that for an alloy without Al. For example, for the alloy electrode without Al, Cmax and C100 are 385 mA h g1 and 202 mA h g1, respectively, and for the alloy electrode with x = 0.4, Cmax and C100 are 323 mA h g1 and 273 mA h g1, respectively. Due to the formation of an Al oxide film on the surface of alloy particles in the alkaline electrolyte [16], the formation of La(OH)3 and Mg(OH)2 is restrained. This protects the alloy electrode from further corrosion in the charge–discharge process [4]. These results indicate that the charge–discharge cycle life can be improved effectively by addition of Al element. To investigate the electrochemical kinetic properties of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes, the high rate discharge ability (HRD) and linear polarization curve are obtained. Fig. 7 shows the HRD of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrode. The HRD value is defined and calculated according to the following formulation:

HRDð%Þ ¼

Fig. 2. SEM image and EDS spectrum for the Mm0.75Mg0.25Ni3.5Co0.2Al0.1 alloy (a) SEM image; (b) EDS pattern of region A; (c) EDS pattern of region B.

the performance of hydrogen desorption thermodynamics can be improved. 3.3. Electrochemical properties Fig. 5 shows the discharge curves of the Mm0.75Mg0.25Ni3.5Co0.2 Alx (x = 0.0–0.4) alloy electrodes, and the discharge capacities are listed in Table 3. It can be seen that the maximum discharge capacity (Cmax) decreases from 385 mA h g1 to 323 mA h g1 as the Al content increases from 0.0 to 0.4. The discharge voltage platform is in good agreement with the plateau pressure of the hydrogen absorption/desorption process, as shown in Fig. 3. According to the Nernst equation, the relationship between the potential of

Cd  100% C d þ C 80

ð1Þ

where Cd is the discharge capacity at the discharge current density (Id) and C80 is the residual discharge capacity at I = 80 mA g1. It can be seen from Table 3 that the HRD1200 for the alloy electrodes decreases from 83.4% to 52.4% as the x value increases from 0.0 to 0.4, at a discharge current density of Id = 1200 mA g1. It is well known that the high rate discharge ability of the Ni–MH battery is mainly controlled by the charge transfer rate on the electrode surface and the hydrogen diffusion rate in the electrode bulk [17]. The exchange current density (I0), which is generally used to characterize the charge–discharge reaction, can be calculated by the following equation [18]:

I0 ¼

IRT RT ¼ F g FRP

ð2Þ

where R, T, F, g and RP are the gas constant, absolute temperature, Faraday constant, overpotential and polarization resistance, respectively. Fig. 8 shows the linear polarization curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes. It is found that the polarization current density is almost linearly dependent on an overpotential ranging from 5 mV to +5 mV. The polarization resistance (RP) can be obtained by fitting the slope of the polarization curves. The exchange current density (I0) is obtained from the polarization resistance (RP) and Eq. (2), these being listed in Table 4. It can be seen that the polarization resistance (RP) is reduced with added Al. Because of its good conductivity, the

314

L. Zhiqiang et al. / Journal of Alloys and Compounds 623 (2015) 311–316

Fig. 3. The PCT curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys (a) 303 k; (b) 333 k; (c) 363 k.

Table 2 The dehydrogenation enthalpy of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) hydrides. Samples

x = 0.0

x = 0.1

x = 0.2

x = 0.3

x = 0.4

DHdes/(kJ mol1)

25.17

23.38

19.68

22.18

21.12

Fig. 4. The ln P H2 vs. 1000/T curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys.

addition of Al enhances electron transportation. That is, the addition of Al acts to improve the exchange current density (I0). Fig. 9 shows the Tafel polarization curve of the Mm0.75Mg0.25 Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes at 298 K. During the

Fig. 5. Discharge curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes.

315

L. Zhiqiang et al. / Journal of Alloys and Compounds 623 (2015) 311–316 Table 3 The electrochemical properties of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) electrodes. Samples

Cmax/(mA h g1)

HRD1200/(%)

C100/(mA h g1)

x = 0.0 x = 0.1 x = 0.2 x = 0.3 x = 0.4

385 378 337 334 323

83.36 75.75 70.40 61.05 52.42

202 207 227 250 273

Fig. 8. The linear polarization curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes at 298 K.

Table 4 Electrochemical kinetics parameters of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) electrodes. Samples

Rp/(mX)

IL/(mA g1)

I0/(mA g1)

D (1011 cm2 s1)

x = 0.0 x = 0.1 x = 0.2 x = 0.3 x = 0.4

502.5 295.2 431.5 257.0 322.6

378.3 560.7 779.0 1503.7 1056.3

173.2 294.8 205.1 338.7 269.8

9.06 7.16 6.99 6.36 5.63

Fig. 6. The charge–discharge curves of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes.

Fig. 7. High rate discharge ability of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes.

anodic polarization, the maximum anodic current with increasing potential is defined as the limiting current IL. The limiting current IL is mainly controlled by hydrogen atom diffusion inside the alloy electrode during anode polarization. The limiting current IL, obtained from the Tafel polarization curve, is listed in Table 4. As Al content increases, the limiting current IL increases from 378.3 mA g1 (x = 0.0) to 1503.7 mA g1 (x = 0.3), and then decreases to 1056.3 mA g1 (x = 0.4). Obviously, the addition of Al increases the limiting current IL. The electrochemical kinetic performance of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes can be improved by increasing the Al content.

Fig. 9. Tafel polarization curve of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes at 298 K.

In order to investigate hydrogen diffusion in the electrode bulk, the semilogarithmic plots of anodic current versus time response of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes is measured, and shown in Fig. 10. The semilogarithmic plot can be divided into two time regions. At the beginning (t < 2500 s), the current rapidly decreases due to the consumption of hydrogen on the alloy surface. The current then decreases linearly with time after sufficient time has passed (t > 2500 s). The linear decrease is mainly attributed to hydrogen diffusion in the electrode bulk. The diffusion coefficient of the hydrogen atom in the electrode bulk can be calculated by the following formula [19]:

316

L. Zhiqiang et al. / Journal of Alloys and Compounds 623 (2015) 311–316

as Al is added. The shrink of the (La,Pr,Nd)Ni5 cell volume results in a decrease in the hydrogen storage and in the maximum discharge capacity. (2) With the increase of Al content, the discharge capacity of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrode decreases from 385 mA h g1 to 323 mA h g1. The addition of Al reduces the discharge capacity, however, it can improve the cyclic stability significantly. The capacity retention ratio increases from 52.6% (x = 0.0) to 84.6% (x = 0.4) after 100 charge/discharge cycles. (3) The diffusion coefficient of the alloy electrode decreases from 9.06  1011 cm2 s1 (x = 0.0) to 5.63  1011 cm2 s1 (x = 0.4), and the high rate dischargeability HRD1200 for the alloy electrodes decreases from 83.4% (x = 0.0) to 52.4% (x = 0.4) with the increase of Al content. The same change trend for the diffusion coefficient and high rate discharge ability indicates that the high discharge rate is mainly dependent on the hydrogen atom diffusion rate. Fig. 10. Semilogarithminc curves of anodic current vs. time response of the Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloy electrodes at full charged state.

Acknowledgements

logðiÞ ¼ log



6FD 2

da



ðC 0  C s Þ 

p2 D t 2:303a2

ð3Þ

where i, D, C0, Cs, a, d and t denote the diffusion current density (A g1), hydrogen diffusion coefficient (cm2 s1), initial hydrogen concentration in the bulk of the alloy (mol cm3), hydrogen concentration on the surface of the alloy particles (mol cm3), alloy particle radius (cm), density of the hydrogen storage alloy (g cm3) and the discharge time (s), respectively. It is well known that a larger hydrogen diffusion coefficient D means faster diffusion of the hydrogen atom, and the alloy electrode will show better kinetic characteristic. As shown in Table 4, the diffusion coefficient D of the electrode decreases from 9.06  1011 cm2 s1 (x = 0.0) to 5.63  1011 cm2 s1 (x = 0.4). This result shows that the addition of the Al element reduces the efficiency of hydrogen transference in the electrode bulk. An Al-oxide film coating the electrode particle in the alkaline electrolyte prevents the hydrogen atoms from diffusing into the electrode bulk. From a comparison with both D in Table 4 and HRD1200 in Table 3, it can be seen that there is the same change trend for the diffusion coefficient D and the high rate dischargeability HRD1200 with the increase in Al content. It is believed, therefore, that the high discharge rate is dependent on the hydrogen atom diffusion. 4. Conclusions The effects of Al content on the hydrogen storage and electrochemical properties of low-Co Re–Mg–Ni alloys are investigated, and the conclusions are summarized as follows. (1) The Mm0.75Mg0.25Ni3.5Co0.2Alx (x = 0.0–0.4) alloys mainly contain (La,Pr,Nd)Ni5, LaMg2Ni9 and LaAlNi4 phases. The cell volumes of the (La,Pr,Nd)Ni5 and LaMg2Ni9 phases decrease

This work was supported by the Key Project of Guangxi Experiment Centre of Science and Technology (Grant No. LGZX201202), the National Natural Science Foundation of China (Grant No. 51271061), the Project of the Guangxi Department of Education (Grant No. 2013YB006) and the Key Laboratory of Guangxi for Nonferrous Metals and Materials Processing Technology. References [1] Y.F. Liu, H.G. Pan, M.X. Gao, R. Li, Y.Q. Lei, J. Alloys Comp. 376 (2004) 296–303. [2] H.G. Pan, Y.J. Yue, M.X. Gao, X.F. Wu, N. Chen, Y.Q. Lei, Q.D. Wang, J. Alloys Comp. 397 (2005) 269–275. [3] S. Ma, M.X. Gao, R. Li, H.G. Pan, Y.Q. Lei, J. Alloys Comp. 457 (2008) 457–464. [4] Y.F. Liu, H.G. Pan, M.X. Gao, H. Miao, Y.Q. Lei, Q.D. Wang, Int. J. Hydrogen Energy 33 (2008) 124–133. [5] Y. Zhao, M.X. Gao, Y.F. Liu, Li Huang, H.G. Pan, J. Alloys Comp. 496 (2010) 454– 461. [6] J.D. Wang, S.M. Han, Y. Li, J.J. Liu, L.D. Che, L. Zhang, J.L. Zhang, J. Alloys Comp. 582 (2014) 552–557. [7] M.S. Balogun, Z.M. Wang, H.X. Chen, J.Q. Deng, Q.R. Yao, H.Y. Zhou, Int. J. Hydrogen Energy 38 (2013) 10926–10931. [8] Y. Li, Y. Tao, D.D. Ke, S.Q. Yang, S.M. Han, J. Alloys Comp. 615 (2014) 91–95. [9] B.P. Wang, L.M. Zhao, C.S. Cai, S.X. Wang, Int. J. Hydrogen Energy 39 (2014) 10374–10379. [10] H. Miao, H.G. Pan, S.C. Zhang, N. Chen, R. Li, M.X. Gao, Int. J. Hydrogen Energy 32 (2007) 3387–3394. [11] Z.J. Gao, Y.C. Luo, Z. Lin, R.F. Li, J.Y. Wang, L. Kang, J. Solid State Electrochem. 17 (2013) 727–735. [12] Z.J. Gao, Y.C. Luo, R.F. Li, Z. Lin, L. Kang, J. Power Sources 241 (2013) 509–516. [13] D.L. Chao, C.L. Zhong, Z.W. Ma, F. Yang, Y.C. Wu, D. Zhu, C.L. Wu, Y.G. Chen, Int. J. Hydrogen Energy 37 (2012) 12375–12383. [14] K. Kadir, T. Sakai, I. Uehara, J. Alloys Comp. 302 (2000) 112–117. [15] M. Raju, M.V. Ananth, L. Vijayaraghavan, J. Power Sources 180 (2008) 830–835. [16] H.G. Pan, Y.F. Liu, M.X. Gao, Y.Q. Lei, Q.D. Wang, J. Electrochem. Soc. 152 (2005) A326–A332. [17] C. Iwakura, M. Matsuoka, K. Asai, T. Kohno, J. Power Sources 38 (1992) 335– 343. [18] P.H.L. Notten, P. Hokkeling, J. Electrochem. Soc. 138 (1991) 1877–1885. [19] G. Zheng, B.N. Popov, R.E. White, J. Electrochem. Soc. 142 (1995) 2695–2698.