Microstructure and electrochemical hydrogen storage characteristics of La0.67Mg0.33−xCaxNi2.75Co0.25 (x = 0–0.15) electrode alloys

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 0 5 0 e3 0 5 5

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/he

Microstructure and electrochemical hydrogen storage characteristics of La0.67Mg0.33LxCaxNi2.75Co0.25 (x [ 0e0.15) electrode alloys Zhenwei Dong a,b, Yaoming Wu b, Liqun Ma a,*, Lidong Wang b, Xiaodong Shen a, Limin Wang b,** a b

College of Material Science and Engineering, Nanjing University of Technology, Nanjing 210009, China State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied chemistry, CAS, Changchun 130022, China

article info

abstract

Article history:

The microstructure and electrochemical hydrogen storage characteristics of La0.67Mg0.33
xCax

Received 23 October 2010

Ni2.75Co0.25 (x ¼ 0, 0.05, 0.10 and 0.15) alloys are investigated. The results show that all alloys

Received in revised form

mainly consist of (La, Mg)Ni3 and LaNi5 phases, besides a small amount of (La, Mg)2Ni7 phase.

25 November 2010

The cycle stability (S80) after 80 charge/discharge cycles of all alloy electrodes first increases

Accepted 2 December 2010

from 60.1% (x ¼ 0) to 64.2% (x ¼ 0.05), then decreases to 45.9% (x ¼ 0.15). The high rate dis-

Available online 6 January 2011

chargeability of all alloy electrodes first increases from 52.6% (x ¼ 0) to 61.4% (x ¼ 0.10), then

Keywords:

from 168.2 mU (x ¼ 0) to 125.7 mU (x ¼ 0.10), then increases to 136.6 mU (x ¼ 0.15). All the results

AB3-type alloy

indicate that the substitution of Mg with a certain amount of Ca can improve the overall

Hydrogen storage

electrochemical characteristics.

decreases to 57.2% (x ¼ 0.15). Moreover, the charge-transfer resistance (Rct) first decreases

Electrochemical kinetics

ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.

MH/Ni battery

1.

Introduction

The traditional AB5-type hydrogen storage alloys have been widely used as commercial electrode materials in metal hydride/ nickel (MH/Ni) secondary batteries in last decades. However, this series of alloys will not meet the application need for higher performance of MH/Ni batteries any more in future due to the relatively low discharge capacities (about 300 mAh/g) [1,2]. In recent years, LaeMgeNi system AB3-type hydrogen storage alloys are considered as one of the novel candidates for AB5-type alloys due to their larger discharge capacities [3
5]. Kohno et al. [6] have reported that the maximum discharge capacity of nonstoichiometric La0.7Mg0.3Ni2.8Co0.5 alloy electrode can reach

410 mAh/g, which is about 30% much larger than that of AB5-type alloys. At present, the main problem for AB3-type alloys to be used for industrialization is the serious degradation of discharge capacity during charge/discharge cycles, which is mainly caused by adding the critical element of Mg into the alloys. It is well known that both of Ca and Mg elements belong to the group IIA, and the electro-negativity of Ca is more negative than Mg. Therefore, it is expected that Ca can alleviate the corrosion of Mg and it is significant to study the effect of substituting Mg with Ca in the alloys on overall electrochemical characteristics of AB3-type hydrogen storage alloys.

* Corresponding author. Tel.: þ86 25 83587243; fax: þ86 25 83240205. ** Corresponding author. Tel./fax: þ86 431 85262447. E-mail addresses: [email protected] (L. Ma), [email protected] (L. Wang). 0360-3199/$ e see front matter ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2010.12.008

3051

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 0 5 0 e3 0 5 5

Table 1 e Composition of La0.67Mg0.33LxCaxNi2.75Co0.25 (x [ 0e0.15) alloys detected by ICP-AES analysis. x ¼ ¼ ¼ ¼

0 0.05 0.10 0.15

Mg (mg/g)

Ca (mg/g)

Ni (mg/g)

Co (mg/g)

Composition formula

340.31 330.29 331.55 332.47

27.17 23.62 19.33 14.79

0 7.29 14.51 21.53

581.58 581.52 583.66 576.55

51.03 57.36 51.52 54.72

La0.68Mg0.31Ni2.75Co0.24 La0.66Mg0.27Ca0.05Ni2.75Co0.25 La0.66Mg0.22Ca0.10Ni2.75Co0.24 La0.67Mg0.17Ca0.15Ni2.75Co0.26

In this work, the La0.67Mg0.33Ni2.75Co0.25 alloy is selected as a basic alloy, and the microstructure and electrochemical characteristics of La0.67Mg0.33
xCaxNi2.75Co0.25 (x ¼ 0, 0.05, 0.10 and 0.20) alloys are investigated.

2.

Experimental procedures

2.1. Alloy preparation and crystallographic characteristics analysis The LaeNieCo alloys were prepared by arc-melting under Ar atmosphere, with the constituent metals on a water-cooled copper hearth. The alloys were turned over and melted three times to make them homogeneous. Then parts of the alloys were crushed and grinded to powder for composites and powder sintering. Before powder sintering, the LaeNieCo alloys, CaNi3, MgNi2 and Ni powders (200e300 mesh) at a certain molar ratio were pre-mixed using a planetary ball mill QM-1 SP for 10 min under Ar atmosphere. The ball to powder mass ratio was 5: 1 and the rotation speed was 200 rpm. After the milling, the mixtures were cold pressed into green compacts under a pressure of 20 MPa. The green compacts were then powder sintered under Ar atmosphere for 12 h at 1273 K. Then the alloys were annealed at 1073 K under Ar atmosphere for 48 h. The purity of all the constituent metals was above 99.9%. The La0.67Mg0.33
xCaxNi2.75Co0.25 (x ¼ 0e0.15) hydrogen storage alloys were mechanically crushed into powders of 200e300 mesh in a glove box under a dry Ar atmosphere. XRD measurements were carried out using a Rigaku D/Max-3B diffractometer with Cu Ka radiation at 40 kV and 30 mA. The surface morphology of the alloys was observed using scanning electron microscopy (SEM).

2.2.

0.1 mV/s from
5 to 5 mV (versus open circuit potential) at 50% depth of discharge (DOD). Potential step experiments were performed on the same instrument (using the M352 CorrWare electrochemical/corrosion software) at 100% charge state. Electrochemical impedance spectra (EIS) were obtained in the frequency range from 10 kHz to 1 mHz at 50% DOD.

3.

Results and discussion

3.1.

Alloy composition and microstructures

The low melting Mg and Ca are inevitably lost during the sample preparation process. We try to add a slight excess of Mg and Ca over sample composition to compensate evaporative loss of Mg and Ca. The ICP-AES analysis results of the alloys are given in Table 1. It can be found that the lost weight during sample preparation is almost the same as excess Mg and Ca and the composition of all prepared alloys is almost identical with the original ones. The XRD patterns for La0.67Mg0.33
xCaxNi2.75Co0.25 (x ¼ 0e0.15) hydrogen storage alloys are shown in Fig. 1. The patterns exhibit similar diffraction peaks, and all alloys mainly consist of (La, Mg) Ni3 phase and LaNi5 phase, besides a small amount of (La, Mg)2Ni7 phase, which indicates that the microstructures of the alloys are almost unchanged by increasing Ca content. The lattice parameters and unit cell volumes of (La, Mg)Ni3, (La, Mg)2Ni7 and LaNi5 phases in the alloys are listed in Table 2. It can be found that the a, c and v of (La, Mg)Ni3 and (La, Mg)2Ni7 phases increase while the a, c and v of LaNi5 phase decrease with the increase of x, which is ascribed to that Mg

(La,Mg)Ni3 (La,Mg)2 Ni7 LaNi5

Electrochemical measurements

The testing electrodes were constructed by mixing the composite powders with carbonyl nickel powders at a weight ratio of 1: 5. The powder mixtures were pressed into a small pellet under a pressure of 20 MPa. Electrochemical charge/ discharge testing was carried out at 298 K by using a DC-5 battery testing system. The electrolyte was 6 M KOH aqueous solution. In each charge/discharge cycle test, the negative electrodes were charged for 7 h at 60 mA/g and discharged at 60 mA/g to the cut-off potential of 0.6 V. High rate dischargeability (HRD) was measured by a ratio of the discharge capacity at a given discharge current density (300e1200 mA/g) to the discharge capacity at 60 mA/g. Linear polarization curves were performed on a EG&G PARC’s Model 273 Potentiostat/Galvanostat station by scanning the electrode potential at the rate of

Intensity (Arb.Units)

x x x x

La (mg/g)

x = 0.15

x = 0.10 x = 0.05 x=0

20

30

40

50

60

70

80

2 (Degree) Fig. 1 e XRD patterns of the La0.67Mg0.33LxCaxNi2.75Co0.25 (x [ 0e0.15) alloys.

3052

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 0 5 0 e3 0 5 5

Table 2 e Characteristics of the phases in La0.67Mg0.33LxCaxNi2.75Co0.25 (x [ 0e0.15) alloys. x

Phase

x¼0

Lattice parameter (Ǻ)

(La, Mg)Ni3 LaNi5 (La, Mg)2Ni7 (La, Mg)Ni3 LaNi5 (La, Mg)2Ni7 (La, Mg)Ni3 LaNi5 (La, Mg)2Ni7 (La, Mg)Ni3 LaNi5 (La, Mg)2Ni7

x ¼ 0.05

x ¼ 0.10

x ¼ 0.15

a

c

5.085 5.011 5.031 5.091 5.008 5.039 5.096 4.997 5.046 5.099 4.991 5.058

24.278 3.981 24.137 24.473 3.976 24.212 24.641 3.954 24.304 24.757 3.938 24.385

˚ ) can exist can only exist in AB3-type phase, whereas Ca (1.74 A ˚ ) or AB3-type phase to randomly in AB5-type replace La (1.88 A ˚ ), respectively [7]. Accordingly, the paramreplace Mg (1.60 A eters of (La, Mg)Ni3 phase increase and the parameters of LaNi5 phase decrease by increasing Ca content. The abundances of the phases are calculated by Rietveld refinement and also listed in Table 2. It can be found that the abundances of (La, Mg)Ni3 phase decrease while the abundances of (La, Mg)2Ni7 and LaNi5 phases increase, which indicates that Ca is helpful to increase the abundances of (La, Mg)2Ni7 and LaNi5 phases.

3.2. Polarization current density I0 and hydrogen diffusion coefficient D

60 x=0 x = 0.05 x = 0.10 x = 0.15

Current density (mA/g)

20

Phase abundance (wt.%)

545.3 88.9 531.1 548.4 88.5 533.9 552.1 87.8 536.1 555.8 87.4 538.8

83.31 15.60 1.09 80.66 17.37 1.97 78.01 19.18 2.81 73.91 22.08 4.01

RT FRP

(1)

where R is the gas constant, T is the absolute temperature, F is the Faraday constant, and Rp is the polarization resistance. The I0 values obtained from Eq. (1) are tabulated in Table 3. The I0 of the alloy electrodes first increases from 195.7 mA/g (x ¼ 0) to 223.9 mA/g (x ¼ 0.10), then decrease to 208.1 mA/g (x ¼ 0.15). The potential step curves of the alloy electrodes are shown in Fig. 3. The hydrogen diffusion coefficient D can be calculated by the following equation [12,13]:

LogðiÞ ¼ Log

Fig. 2 exhibits the linear polarization curves of the alloy electrodes at 50% DOD. It is found that there is a good linear dependence between the current density and overpotential within a small overpotential range (5 mV). Based on the measured curves, values of polarization current density I0 and polarization resistance Rp are calculated by the following formula [8
11]:

40

I0 ¼

Cell volume (Ǻ3)

  2    6FDðc0
cs Þ p t D
2 da 2:303 a2

(2)

where D is the hydrogen diffusion coefficient (cm2/s), a is the radius of the spherical particle (cm), i is the diffusion current density (A/g), c0 is the initial hydrogen concentration in the bulk electrode (mol/cm3), cs is the hydrogen concentration on the surface of alloy particles (mol/cm3), d is the density of the hydrogen storage materials (g/cm3), and t is the discharge time. The SEM image of La0.67Mg0.23Ca0.10Ni2.75Co0.25 as a representative alloy is represented in Fig. 4. It can be seen that the alloy particles are pulverized normally to about 50 mm. Assuming that the average particle radius is 50 mm, the hydrogen diffusion coefficient D in the bulk electrodes is estimated by Eq. (2) and also listed in Table 3. It is shown that the D values of the alloy electrodes are almost unchanged (10.5  10
11e11.4  10
11 cm2/s) with the increase of Ca content.

0 -20 Table 3 e Summary of electrochemical characteristics for the alloy electrodes at 298 K.

-40

S80 HRD1200 I0 Rct Cmax (mAh/g) (%) (%) (mA/g) (mU)

x

-60 -6

-4

-2

0

2

4

6

Overpotential (mV) Fig. 2 e Linear polarization patterns of the alloy electrodes at 50% DOD.

x x x x

¼0 ¼ 0.05 ¼ 0.10 ¼ 0.15

388.1 371.5 369.8 317.2

60.1 64.2 52.6 45.9

52.6 56.3 61.4 57.2

195.7 205.0 223.9 208.1

168.2 144.0 125.7 136.6

D (cm2/s) 11.4 10.9 10.8 10.5

   

10
11 10
11 10
11 10
11

3053

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 0 5 0 e3 0 5 5

100 x= 0 x = 0.05 x = 0.10 x = 0.15

-1.0

Log i (i (A/g))

High rate dischargeability (%)

-0.5

-1.5 -2.0 -2.5 -3.0

90 80 70 60 50

x= 0 x = 0.05 x = 0.10 x = 0.15

40 -3.5 0

500

1000

1500

2000

2500

300

3000

Time (s)

High rate dischargeability

Fig. 5 represents the relationship between the high rate dischargeability (HRD) and discharge current density of La0.67Mg0.33-xCaxNi2.75Co0.25 (x ¼ 0e0.15) alloy electrodes. The HRD is defined and calculated according to the following formulation [14]: HRDð%Þ ¼

Cd  100 C60

900

1200

Fig. 5 e High rate dischargeability of the alloy electrodes at 298 K.

Fig. 3 e Potential step curves of the alloy electrodes at 298 K.

3.3.

600

Discharge current density (mA/g)

(3)

where Cd is the discharge capacity of alloy electrode at current density to the cut-off potential of 0.6 V, C60 is the discharge capacity at the discharge current density of 100 mA/g to 0.6 V. The HRDs of the alloy electrodes at discharge current density of 1200 mA/g are listed in Table 3. It can be seen that the

HRD1200 first increases from 52.6% (x ¼ 0) to 61.4% (x ¼ 0.10), then decreases to 57.2% (x ¼ 0.15), which is ascribed to that the mass fraction of Ni on alloy surface increases with the dissolution of Ca and the electro-catalytic Ni can promote the diffusion of hydrogen. However, excessive Ca leads to producing Ca(OH)2 that deposits on alloy surface and prevents the charge-transfer reaction of hydrogen. Therefore, the HRD of the alloy electrodes first increases and then decreases by increasing Ca content. The HRD1200 as a function of polarization current density I0 is shown in Fig. 6. It can be found that the HRD1200 is a linear function with I0. Iwakura et al. [15] have reported that a linear dependence of HRD on polarization current density is observed when electrochemical reaction on alloy surface is the rate-determining factor. Otherwise, HRD is constant when diffusion rate of hydrogen in alloy bulk is the rate-determining factor. Accordingly, it is concluded that the HRD1200 is essentially controlled by charge-transfer reaction of hydrogen on alloy surface at the discharge current density of 1200 mA/g.

High rate dischargeability (%)

62

60

58

56

54

52 190

200

210

220

230

Polarization current density, I0 (mA/g) Fig. 4 e SEM image of La0.67Mg0.23Ca0.10Ni2.75Co0.25 alloy particles.

Fig. 6 e HRD1200 as a function of I0 of the alloy electrodes at 298 K.

3054

Discharge capacity and cycle stability

0.5

As shown in Table 3, the maximum discharge capacity (Cmax) of La0.67Mg0.33
xCaxNi2.75Co0.25 (x ¼ 0e0.15) alloy electrodes decreases from 388.1 mAh/g (x ¼ 0) to 317.2 mAh/g (x ¼ 0.15), which is ascribed to that the abundance of (La, Mg)Ni3 phase decreases while the abundance of LaNi5 phase increases with the increase of Ca content, and it is believed that the maximum discharge capacity of (La, Mg)Ni3 phase is larger than that of LaNi5 phase. Therefore, the maximum discharge capacity of the alloy electrodes decreases by decreasing the abundance of (La, Mg)Ni3 phase. The relationship between the discharge capacity and cycle number of the alloy electrodes is represented in Fig. 7. The cycle stability (S80) after 80 charge/discharge cycles is listed in Table 3. It can be seen that the S80 of the alloy electrodes first increases from 60.1% (x ¼ 0) to 64.2% (x ¼ 0.05), then decreases to 45.9% (x ¼ 0.15). The cycle behavior of the LaeMgeCaeNieCo alloy electrodes can be classified into three stages: the corroding stage of Ca, Mg and La, respectively. Ca is first corroded due to its most negative electro-negativity, which prevents Mg and La from corrosion and results in the improvement of cycle stability. However, excessive Ca content (x  0.10) can accelerate the corrosion rate of the alloy electrodes and lead to decreasing the cycle stability. Accordingly, the cycle stability of the alloy electrodes first increases and then decreases by increasing Ca.

3.5.

Electrochemical impedance spectroscopy

The charge-transfer resistance of the hydrogen storage alloy electrode can be determined by the electrochemical impedance spectra (EIS). The EIS of La0.67Mg0.33
xCaxNi2.75Co0.25 (x ¼ 0e0.15) alloy electrodes at 50% DOD is shown in Fig. 8. It can be seen that each spectroscopy consists of a small semicircle in high frequency region, a large semicircle in low frequency region and a straight line. According to the model proposed by Kuriyama et al. [16], the small semicircle in EIS curve indicates the contact resistance between alloy powder and conductive material, and the large semicircle represents the charge-transfer resistance (Rct) on alloy electrode surface.

Discharge capacity (mAh/g)

400

0.4

-Z" ( /g)

3.4.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 0 5 0 e3 0 5 5

R el

Ccp

C pp

R R ct w

R cp

R pp

C ct

x= 0 x = 0.05 x = 0.10 x = 0.15

0.3

0.2

0.1

0.0 0.0

0.2

0.4

0.6

0.8

1.0

1.2

Z'( /g) Fig. 8 e Electrochemical impedance spectra of the alloy electrodes at 50% DOD.

The EIS data is analyzed using an equivalent circuit represented by the inset in Fig. 8. The Re1 is the electrolyte resistance between metal hydride electrode and reference electrode, the Rcp and Ccp are the contact resistance between alloy particles and current collector in high frequency region, the Rpp and Cpp stand for the contact resistance and capacitance between alloy particles, respectively, the Rct and Cct contribute to the charge-transfer reaction resistance and double layer capacitance in low frequency region, and the Rw is the Warburg impedance. The Rct of the alloy electrodes, summarized in Table 3, first decreases from 168.2 mU (x ¼ 0) to 125.7 mU (x ¼ 0.10), then increases to 136.6 mU (x ¼ 0.15). It can be found that the Rct of La0.67Mg0.23Ca0.10Ni2.75Co0.25 alloy electrode is much smaller than that of other alloys, which illustrates that the charge-transfer reaction on alloy surface becomes faster with the dissolution of Ca in KOH aqueous solution. The result is consistent with the change of HRDs completely and verifies that the effect of charge-transfer reaction of hydrogen on alloy surface is the essentially controlled step once again, which is mainly ascribed to the cooperative effect of the mass fraction of Ni and the deposition of Ca(OH)2 on alloy surface.

300

4. 200 x= 0 x = 0.05 x = 0.10 x = 0.15

100

0 0

10

20

30

40

50

60

70

80

Cycle number (n) Fig. 7 e Cycle stability of the alloy electrodes after 80 charge/discharge cycles.

Conclusion

The La0.67Mg0.33-xCaxNi2.75Co0.25 (x ¼ 0e0.15) hydrogen storage alloys studied in this work mainly consist of (La, Mg)Ni3 and LaNi5 phases, besides a small amount of (La, Mg)2Ni7 phase. The cycle stability after 80 charge/discharge cycles first increases from 60.1% (x ¼ 0) to 64.2% (x ¼ 0.05), then decreases to 45.9% (x ¼ 0.15). The high rate dischargeability of all alloy electrodes first increases from 52.6% (x ¼ 0) to 61.4% (x ¼ 0.10), then decreases to 57.2% (x ¼ 0.15). The charge-transfer resistance first decreases from 168.2 mU (x ¼ 0) to 125.7 mU (x ¼ 0.10), then increases to 136.6 mU (x ¼ 0.15). The entire results indicate that a small amount of Ca can improve the overall electrochemical characteristics of AB3-type hydrogen

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 3 0 5 0 e3 0 5 5

storage alloys. However, excessive Ca leads to accelerating the decline of cycle stability and decreasing the electrochemical kinetics, which is mainly ascribed to the corrosion of Ca in KOH aqueous alkali and deposition of Ca(OH)2 on the alloy surface.

[6]

[7]

Acknowledgements [8]

This work was financially supported by the National High Technology Research and Development Program of China (2007AA03Z228), the Project supported by the Major program for the fundamental research of the Chinese Academy of Sciences (KJCX2-Y1W-H21), the Science and Technology Cooperation Project of Chinese Academy of Sciences and Jilin province (2008syhz0008), the NSFC fund for Creative Research Group (20921002) and the Open Subject of State Key Laboratory of Rare Earth Resource Utilization.

[9]

[10]

[11]

references [12] [1] Li Y, Han SM, Li JH, Zhu XL, Hu L. Study on kinetics and electrochemical properties of Ml0.70Mg0.30(Ni3.95Co0.75Mn0.15 Al0.15)x (x ¼ 0.60, 0.64, 0.68, 0.70, 0.76) alloys. Mater Chem Phys 2008;108:92e6. [2] Chen J, Kuriyama N, Takeshita HT, Tanaka H, Sakai T, Haruta M. Hydrogen storage alloys with PuNi3-Type structure as metal hydride electrodes. Electrochem Solid State Lett 2000;3:249e52. [3] Dong ZW, Wu YM, Ma LQ , Shen XD, Wang LM. Electrochemical properties of (La1
xTix)0.67Mg0.33Ni2.75Co0.25 (x ¼ 0e0.20, at%) hydrogen storage alloys. Mater Res Bull 2010;45:256e61. [4] Kadir K, Sakai T, Uehare I. Structural investigation and hydrogen storage capacity of LaMg2Ni9 and (La0.65Ca0.35)(Mg1.32 Ca0.68)Ni9 of the AB2C9 type structure. J Alloys Compd 2000; 302:112e7. [5] Liu YX, Xu LQ , Jiang WQ , Li GX, Wei WL, Guo J. Effect of substituting Al for Co on the hydrogen-storage performance

[13]

[14]

[15]

[16]

3055

of La0.7Mg0.3Ni2.6AlxCo0.5
x (x ¼ 0.0e0.3) alloys. Int J Hydrogen Energy 2009;34:2986e91. Kohno T, Yoshida H, Kawashima F, Inaba T, Sakai I, Yamamoto M, et al. Hydrogen storage properties of new ternary system alloys: La2MgNi9, La5Mg2Ni23, La3MgNi14. J Alloys Compd 2000;311:L5e7. Zhang P, Liu YN, Tang R, Zhu JW, Wei XD, Liu SS, et al. Effect of Ca on the microstructural and electrochemical properties of La2.3
xCaxMg0.7Ni9 hydrogen storage alloys. Electrochim Acta 2006;51:6400e5. Notten PHL, Hokkeling P. Double-phase hydride forming compoundsda new class of highly electrocatalytic materials. J Electrochem Soc 1991;138:1877e85. Shen XQ , Chen YG, Tao MD, Wu CL, Deng G, Kang ZZ. The structure and 233 K electrochemical properties of La0.8ex NdxMg0.2Ni3.1Co0.25Al0.15 (x ¼ 0.0e0.4) hydrogen storage alloys. Int J Hydrogen Energy 2009;34:2661e9. Pan HG, Ma S, Shen J, Tan JJ, Deng JL, Gao MX. Effect of the substitution of PR for LA on the microstructure and electrochemical properties of La0.7
xPrxMg0.3Ni2.45Co0.75Mn0.1 Al0.2 (x ¼ 0.0e0.3) hydrogen storage electrode alloys. Int J Hydrogen Energy 2007;32:2949e56. Zhang FL, Luo YC, Wang DH, Yan RX, Kang L, Chen JH. Structure and electrochemical properties of La2
xMgxNi7.0 (x ¼ 0.3e0.6) hydrogen storage alloys. J Alloys Compd 2007; 439:181e8. Li Y, Han D, Han SM, Zhu XL, Hu L, Zhang Z, et al. Effect of rare earth elements on electrochemical properties of LaeMgeNiebased hydrogen storage alloys. Int J Hydrogen Energy 2009;34:1399e404. Zheng G, Popov BN, White RE. Hydrogen-atom direct-entry mechanism into metal membranes. J Electrochem Soc 1995; 142:154e6. Zhang XB, Sun DZ, Yin WY, Chai YJ, Zhao MS. Crystallographic and electrochemical characteristics of La0.7 Mg0.3Ni3
x(Al0.5Mo0.5)x (x ¼ 0e0.4) hydrogen storage alloys. Electrochim Acta 2005;50:3407e13. Iwakura C, Oura T, Inoue H, Matsuoka M. Effects of substitution with foreign metals on the crystallographic, thermodynamic and electrochemical properties of AB5etype hydrogen storage alloys. Electrochim Acta 1996;41:117e21. Kuriyama N, Sakai T, Miyamura H, Uehara I, Ishikawa H, Iwasaki T. Electrochemical impedance spectra and deterioration mechanism of metal hydride electrodes. J Electrochem Soc 1992;139:L72e6.