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 en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 0

Avai lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ier . com/ loca te /he

Electrochemical properties of Mg-based hydrogen storagematerials modified with carbonaceous materials preparedby hydriding combustion synthesis and subsequentmechanical milling (HCS D MM)

Yunfeng Zhu, Wenfeng Zhang, Chen Yang, Liquan Li*

College of Materials Science and Engineering, Nanjing University of Technology, 5 Xinmofan Road, Nanjing, Jiangsu 210009, PR China

a r t i c l e i n f o

Article history:

Received 23 April 2010

Received in revised form

18 June 2010

Accepted 3 July 2010

Available online 1 August 2010

Keywords:

Mg-based hydrogen storage

materials

Hydriding combustion synthesis

Electrochemical properties

Mechanical milling

Carbonaceous materials

* Corresponding author. Tel.: þ86 25 8358725E-mail address: lilq@njut.edu.cn (L. Li).

0360-3199/$ e see front matter ª 2010 Profedoi:10.1016/j.ijhydene.2010.07.031

a b s t r a c t

Mg2Ni-based hydride was prepared by hydriding combustion synthesis (HCS), and subse-

quently modified with various carbonaceous materials including graphite, multi-walled

carbon nanotubes (MWCNTs), carbon aerogels (CAs) and carbon nanofibers (CNFs) by

mechanical milling (MM) for 5 h. The structural properties of the modified hydrides were

characterized by X-ray diffraction (XRD) and scanning electron microscope (SEM). All of the

modified hydrides show amorphous or nanocrystalline-like phases. The hydride modified

with graphite exhibits the most homogenous distribution of particles and the smallest

particle size. The effects of the modifications on electrochemical properties of the hydride

were investigated by galvanostatic charge/discharge, linear polarization, Tafel polariza-

tion, electrochemical impedance spectroscopy and potentiostatic discharge measure-

ments. The results show that the maximum discharge capacity, the high rate

dischargeability (HRD), the exchange current density and the hydrogen diffusion ability of

the hydride modified with the carbonaceous materials are all increased. Especially, the

hydride modified with graphite possesses the highest discharge capacity of 531 mAh/g and

the best electrochemical kinetics property.

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

1. Introduction for the large-scale applications of Ni/MH batteries. Mg-based

Presently, nickelemetal hydride batteries (Ni/MH) have been

extensively applied in everyday life due to their superior

charge/discharge capability and pollution-free nature.

However, unsatisfactory theoretical specific capacities and

high cost of the commercial anode materials (i.e. LaNi5-based

and ZreTi-based hydrogen storage alloys) restrict their further

developments.

Consequently, the development of high specific capacity

and economical metal hydride as anode materials is crucial

5.

ssor T. Nejat Veziroglu. P

hydrogen storage materials have attracted considerable

interests among the most promising candidates in view of

their high theoretical specific capacity, light weight, abundant

resources, low cost, etc [1e4]. However, either poor cycling

stability or low practical discharge capacity hinders their

potential application. In order to overcome the above short-

comings,much research such asmechanical milling (alloying)

[5,6], element substitution [7e9], Mg-based composite prepa-

ration [10,11], surface modification [12,13] and novel prepa-

ration methods [14,15] has been conducted. Mechanical

ublished by Elsevier Ltd. All rights reserved.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 09654

milling/alloying has been widely used to synthesize Mg-based

alloys with amorphous or nanocrystalline microstructure,

which possesses higher discharge capacity than crystalline

alloys [5].

Hydriding combustion synthesis, put forward in 1997 by

Akiyama et al. [16], has been considered as a valuable method

to prepare Mg-based hydrogen storage alloys on account of its

short process time, low energy consumption and high purity

and activity of product [17e22]. By ways of HCS, Mg2Ni alloy

can be directly synthesized from the powder mixture of

magnesium and nickel at 850 K, which is below the melting

point of magnesium. As a consequence, the evaporation of

magnesium can be prevented effectively. Furthermore, the

preparation and hydrogenation of hydrogen storage alloy

proceed in a single process, which avoids the repeated

remelting and activating process as compared with the

conventional ingot melting. Based on the above advantages of

HCS and MM, we have prepared Mg2Ni-based hydrides with

high discharge capacity and activity by the process of

HCS þ MM (MM for only 5 h) previously [23,24].

It has been widely reported that carbonaceous materials,

such as graphite and multi-walled carbon nanotubes are

effective additives for improving the electrochemical proper-

ties of Mg-based hydrogen storage alloys [25e31]. Thus, the

aim of the present work is to investigate the influences of

various carbonaceous additives (including graphite,MWCNTs,

CAs and CNFs) on the structural and electrochemical proper-

ties of the Mg2Ni-based hydride prepared by the process of

HCS þ MM.

20 30 40 50 60 70

♦

♦

♦

♦

♦

♦

♦Mg2NiH0.3Mg2NiH4

Inte

nsity

(a.u

.)

2-Theta (degree)

(a)

(b)

(c)

(d)

(e)

(f)

HCS

HCS+MM

Graphite

MWCNTs

CAs

NiMgNi2

CNFs

Mg(OH)2

Fig. 1 e XRD patterns of the Mg2Ni-based hydride modified

with various carbonaceous additives: (a) HCS product;

(b) HCS D MM product; (c) HCS D MM-Graphite product;

(d) HCS D MM-MWCNTs product; (e) HCS D MM-CAs

product; (f) HCS D MM-CNFs product.

2. Experiment details

The HCS product was prepared from commercial Mg (99.9

mass% inpurity and<150 mminparticle size) andNi (99.7mass

% in purity and 2e3 mm in particle size) powders. The powders

were mixed in 2:1 of Mg/Ni molar ratio by an ultrasonic

homogenizer in acetone for 1 h. After completely dried in air,

the well-mixed powder was placed directly into the synthesis

reactor without compacting. Before heating, the reactor was

evacuated by a rotary pump, then argon at 0.1 MPa was

introduced and the reactorwas evacuated again. This cleaning

procedure was repeated twice to remove any oxygen in the

system. During HCS process, the mixed powder was heated

from room temperature to 850K at the rate of 7 K/min andheld

for 1 h under 1.9 MPa hydrogen pressure. In order to increase

the degree of hydrogenation of Mg2Ni, the samples were hold

at 623K for about 1 h during the cooling process. Subsequently,

the samples were cooled down to room temperature under

hydrogen atmosphere. After that, the HCS product was

mechanically milled with 3 wt% (vs. the HCS product)

commercial graphite, MWCNTs, CAs and CNFs for 5 h at

a speed of 400 rpm and 40:1 in ball-to-powder using a plane-

tary-type ball mill under argon atmosphere, respectively.

All the testing electrodes were prepared as follows: 0.1 g

HCS þ MM product was mixed with 0.4 g carbonyl nickel

powder, and then cold-pressed into a pellet of 10mmdiameter

and about 1 mm thickness under a pressure of 12 MPa. The

electrochemical measurements were performed in 6 M KOH

aqueous solutionusinga three-compartment cell comprisedof

a metal hydride testing electrode, a sintered NiOOH/Ni(OH)2counter electrode and a Hg/HgO reference electrode. The

charge/discharge cycles were carried out with a LAND Battery

Test instrument. As the initial anode are in their hydride state,

all the testing electrodeswere first discharged at 30mA/g up to

�0.6 V (vs. Hg/HgO), and then charged at 300mA/g for 2 h after

resting for 10 min at 30 � 1 �C. The discharge capacities of

electrodes were evaluated by the mass of active substances.

The discharge capacities at different discharge current densi-

ties (100, 200, 400mA/g) weremeasured to investigate the high

rate dischargeability (HRD). Linear polarization, Tafel polari-

zation, electrochemical impedance spectroscopy (EIS) and

potentiostatic discharge of the electrodes were performed at

room temperature on a CHI660C electrochemical workstation

at 50%, 100%, 50% depth of discharge (DOD) and 100% depth of

charge (DOC), respectively. Linear polarization and Tafel

polarization were measured at scanning rates of 0.1 mV/s and

1mV/s from�5 toþ5mVand�300 to 1500mV (vs. open circuit

potential), respectively. The EIS spectra of the electrodes were

obtained in the frequency range of 100 kHze5mHzwith an AC

amplitude of 5mVunder open circuit conditions. The obtained

EIS spectra were analyzed by ZPLOT electrochemical imped-

ance software. As for potentiostatic discharge, electrodeswere

discharge at þ600 mV potential steps for 3600 s.

The crystal structures of the samples were characterized

by X-ray diffraction (XRD) on an ARL X’TRA diffractometer

with Cu Ka radiation. The morphologies of the samples were

analyzed by a JSM-6360LV scanning electron microscopy

(SEM).

3. Results and discussion

3.1. Structural and morphological characteristics

Fig. 1 shows the XRD patterns of the bare (unmodified) and

modified Mg2Ni-based hydrides. It can be seen from Fig. 1(a)

that the HCS product consists of the main phase Mg2NiH4 and

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 en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 0 9655

very small amount of Mg2NiH0.3, MgNi2 and Ni. As the raw

materials Mg and Ni were mixed in molar ratio of 2:1, the

appearance of small peaks of Ni and MgNi2 means the evap-

oration of Mg still takes place, while the appearance of

Mg2NiH0.3 peak indicates incomplete hydrogenation of Mg2Ni

during the HCS process. The phase content of Mg2NiH0.3 is less

than that of the sample in our previous work [24], indicating

that the hydrogenation degree of Mg2Ni is increased in the

current work. Therefore, the electrochemical properties may

be different as compared to the previous work. After

mechanical milling (Fig. 1(b)), the sharp peaks are broadened

and the peak intensities are also decreased, indicating an

increase in intrinsic stress and a decrease in crystallite size.

After modification with carbonaceous materials, the

Mg2NiH4 peaks around 23� even disappear, suggesting that

the carbonaceous additives can facilitate amorphization or

nanocrystallization of the HCS þ MM product effectively.

Meanwhile, the Mg2NiH0.3 peak around 20� appears after

various carbonaceous modifications, indicating part of

Mg2NiH4 was dehydrogenated during the MM process prob-

ably due to the catalytic effect of carbonaceous additives.

The peaks of Ni and MgNi2 become more distinct after

mechanical milling, which can be attributed to the less

brittle nature of Ni and MgNi2 than the Mg2NiH4 hydride,

whose peaks are broadened easily during the MM process. As

seen in Fig. 1 (c)e(f), it can be found that graphite and

MWCNTs additions have more effect on peak broadening of

Ni around 45� than CAs and CNFs additions. Cui et al. [6]

reported the smaller nickel particles might be inlaid to

surface of large magnesium alloy particles by cold welding

through MM. Hence, it can be expected that the nickel

particles with smaller crystallite size were inlaid to the

surface of graphite and MWCNTsmodified hydrides. Besides,

Mg(OH)2 peaks were detected in the modified hydrides due to

exposure of the samples to air, which is also found previ-

ously [20].

The SEM images of the carbonaceous materials are shown

in Fig. 2aed. It can be seen that the microscale graphite flakes

and nanoscale MWCNTs possess homogenous distribution,

while agglomerates are detected in microscale CAs and

nanoscale CNFs. As shown in Fig. 2 e and f, the HCS product is

severely pulverized during intensive ball milling. However,

the particle size of the HCS þ MM product is irregular and

serious agglomerates took place during the unmodified MM

process. After carbonaceous modifications, the particles of

the hydrides become smaller and more homogenous

(Fig. 2gej). It has been reported that lots of carbon accumu-

lates on the surface, and hence lessens the agglomeration

and adhesion of magnesium hydride particles [32]. In addi-

tion, the hydride modified with graphite exhibits the most

uniform distribution and the smallest particle size. No

distinct macro-aggregated graphite flakes are detected,

implying the structure has been broken down adequately

during the MM process. Owing to the low magnifications

employed, no MWCNTs and CNFs traces can be detected in

the hydrides. In general, we believe that the graphite exhibits

the best process control effect. However, considering its

largest initial particle size, this effect still needs to be further

studied.

3.2. Electrochemical properties

Fig. 3 shows the discharge curves of the Mg2Ni-based hydrides

modified with various carbonaceous additives. The discharge

capacities are listed in Table 1. For the modified hydrides, the

discharge potential is shifted to the negative direction and

show clear discharge potential plateaus. This may be caused

by the decrease of particle size of the hydride, since smaller

particles having large surface areas result in a smaller current

density at particle surface [33], which helps to reduce the

electrochemical polarization. For galvanostatic method, the

discharge capacity of the hydride is determined by two

factors. One is the quantity of the active substances that can

release hydrogen (i.e. the nanocrystalline or amorphous phase

in the hydride), and the other is the electrochemical kinetics

of the hydride. As stated above, the modified hydrides have

more amorphous or nanocrystalline phases and crystal

defects, such as dislocations, grain boundaries, etc. The larger

number of interfaces and grain boundaries provide more

efficient active sites for charge-transfer reaction. Besides,

hydrogen may reach the surface more rapidly for a smaller

particle. Therefore, the addition of carbonaceous materials

significantly improves both the amount of active substance

and the electrochemical kinetics, and hence increases the

discharge capacity. It is well-known that these carbonaceous

materials are important hydrogen storage materials.

However, in present case, this small amount of uncharged

carbonaceous materials can hardly make a contribution to

discharge capacity. Thus, the high discharge capacity must

come from the amorphous or nanocrystalline Mg2Ni-based

hydride phase. Modifications of the hydride with graphite and

MWCNTs result in more remarkable increments of the initial

discharge capacity than those with CAs and CNFs. This may

be due to the differences of the structural and morphological

characteristics mentioned above.

Fig. 4 shows the evolution of the discharge capacity of the

original and modified Mg2Ni-based hydrides with cycle

numbers. The capacity retention rate R8 (C8/Cmax) is also listed

in Table 1. The rapid degradation in discharge capacity of

Mg-based alloys is attributed to the well-known reason of the

oxidation and corrosion of Mg in alkaline solution [34]. As

shown in Fig. 4 and Table 1, themodifications of carbonaceous

additives cannot improve the cycling stability under present

conditions, which does not agree with the results reported by

Iwakura et al. [26] and Guo et al. [28]. There may be

a competitive mechanism on the cycling stability. On one

hand, the defective structures and small particle size, i.e. large

specific surface area, account for serious oxidation of the

modified hydrides, corresponding to a decrease of anti-

corrosion capability, which is proved by the Tafel polarization

result as follows. On the other hand, the carbonaceous

materials can obstruct the formation of Mg(OH)2 during

cycling, which is found by the above authors. In present work,

it should be noted that the content of carbonaceous additives

(3 wt%) are much smaller than those in their works (10 wt%).

We also prepared the samples with increased amount of

graphite additive, and the results show that the cycling

stability was improved. Therefore, the protection effect of

carbonaceous additives in the current work is weak. As

Fig. 2 e Morphologies of the carbonaceous materials and the hydrides observed by SEM: (a) Graphite; (b) MWCNTs; (c) CAs;

(d) CNFs; (e) HCS product; (f) HCS D MM product; (g) HCS D MM-Graphite product; (h) HCS D MM-MWCNTs product;

(i) HCS D MM-CAs product; (j) HCS D MM-CNFs product.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 09656

0 100 200 300 400 5000.6

0.7

0.8

0.9

54321

1 Bare 2 CNFs 3 CAs 4 MWCNTs 5 Graphite

-Pot

entia

l (V

vs. H

g/H

gO)

Discharge capacity (mAh/g)

Fig. 3 e Discharge curves for the Mg2Ni-based hydride

modified with various carbonaceous additives (Discharge

rate: 30 mA/g; the first cycle).

0 1 2 3 4 5 6 7 8 90

100

200

300

400

500

Dis

char

ge c

apac

ity (m

Ah/g

)

Cycle number

Bare Graphite MWCNTs CAs CNFs

Fig. 4 e Discharge capacities as a function of cycle number

for the Mg2Ni-based hydride modified with various

carbonaceous additives.

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 en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 0 9657

a result, the cycling stability of the modified hydride

decreases. The Tafel polarization curves of the hydrides

modified with various carbonaceous materials are shown in

Fig. 5 and the corrosion potential Ecorr are also listed in Table 1.

The result shows that the Ecorr shifts toward the negative

direction, indicating the anti-corrosion capability is impaired

after the modifications.

The high rate dischargeability of the Mg2Ni-based hydride

modified with various carbonaceous additives is shown in

Fig. 6. The HRD is determined as follows:

HRD ¼ Cd

Cd þ C60� 100% (1)

where Cd is the discharge capacity with cut-off potential of

�0.6 V (vs. Hg/HgO) at the discharge current density Id (100,

200, 400 mA/g), C60 is the residual discharge capacity with the

same cut-off potential at the discharge current density

I (60 mA/g) after the alloy electrode has been fully discharged

at the discharge current density Id. It is obvious that the HRD is

improved notably by the carbonaceous modifications, sug-

gesting favorable effect of the modification on electro-

chemical kinetics. Generally speaking, the HRD is determined

by the charge-transfer on the alloy surface and the hydrogen

diffusion inside the alloy [35]. For further understanding the

relationship between and the electrochemical kinetics prop-

erties of the electrodes and the modifications, linear polari-

zation, EIS and potentiostatic discharge measurements were

performed.

Table 1 e The electrochemical properties of the Mg2Ni-based h

Modifications Cmax (mAh/g) R8 (%) Ecorr (

Bare 339 13.9 �0.86

Graphite 531 5.4 �0.87

MWCNTs 510 4.1 �0.87

CAs 441 4.4 �0.94

CNFs 416 4.3 �0.91

The exchange current density I0, which is the rate of the

charge-transfer reaction, can be obtained according to the

following formula [36]:

I0 ¼ RTIFh

(2)

where R is the gas constant, T is the absolute temperature (K),

I is the applied current density (mA/g), F is the Faraday

constant and h is the total overpotential (mV). The linear

polarization curves for the modified hydrides are shown in

Fig. 7 and the calculated I0 values are also listed in Table 1. It

can be seen that the value of I0 is increased significantly after

the modifications. This result certifies that the rate of charge-

transfer reaction on the alloy surface is greatly improved,

hence reducing the overpotential during the discharge

process and leading to an increase of discharge capacity of the

electrode. It has been reported that the surface Ni/Mg ratio of

MgNi alloy increases after graphite modification [25].

Furthermore, as stated above, there might have some

unreacted nanocrystalline nickel particles inlaid on the

hydride surface. These nickel particles are beneficial to

enhance the charge-transfer rate on the hydride surface due

to the good electrocatalytic activity of nickel in alkaline elec-

trolyte. The hydride modified with graphite possesses the

largest I0 due to its structural characteristic, for small particles

can enhance the reactivity of the electrode desirably [37].

The EIS spectra and corresponding equivalent circuit of the

modified hydrides are presented in Fig. 8. The spectra are

composed of two semicircles and a straight line. Based on the

ydride modified with various carbonaceous additives.

V) I0 (mA/g) Rct (mU) D/a2 (�10�5 s�1)

6 36.7 493.1 1.3

2 66.7 212.2 2.2

5 60.0 342.4 1.9

2 42.3 410.9 1.8

8 52.0 391.0 1.6

-6 -4 -2 0 2 4 6

-10

-5

0

5

10 Bare Graphite MWCNTs CAs CNFs

Cur

rent

den

sity

(mA/

g)

Overpotential (mV)

Fig. 7 e Linear polarization curves of the Mg2Ni-based

hydride modified with various carbonaceous additives

(scan rate: 0.1 mV/s).

0.6

-1.2 -0.9 -0.6 -0.3 0.0 0.3 0.6 0.9-6

-5

-4

-3

-2

-1

0

-0.94 -0.92 -0.90 -0.88 -0.86-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

CAs

CNFs

MWCNTs

Graphite

Bare

log i (

A/g)

Potential (V, vs Hg/HgO)

Fig. 5 e Tafel polarization curves of the Mg2Ni-based

hydride modified with various carbonaceous additives

(scan rate: 1 mV/s).

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 09658

EIS model of hydrogen storage alloy electrode [38], the high

frequency semicircle, the lower frequency semicircle and the

low frequency straight line are related to the contact resis-

tance (Rc) between the current collector and the alloy pellet,

the charge-transfer reaction resistance (Rct) and the Warburg

impedance (W ), respectively. The values of Rct estimated

according to the equivalent circuit are also listed in Table 1. It

is obvious that the charge-transfer resistance is decreased

after the modifications, which is consistent with the results

obtained by linear polarization. Besides, judging from the

radius of semicircle in the high frequency, we find the modi-

fications also decrease the contact resistance Rc, which is

helpful to improve the discharge capacity as well.

To investigate the hydrogen diffusion behavior in the

hydrides, potentiostatic discharge measurements of the

hydrides were performed and the result is shown in Fig. 9. At

the initial stage of discharging (<500 s), the current declines

sharply. After a time response of about 2000 s, the current

decreases slowly in a linear fashion. In the linear region, the

hydrogen diffusion in the bulk of hydride controls the

100 200 300 400

20

40

60

80

100

HR

D (

%)

Discharge current density (mA/g)

Bare Graphite MWCNTs CAs CNFs

Fig. 6 e High rate dischargeability of the Mg2Ni-based

hydride modified with various carbonaceous additives.

electrode process and the hydrogen diffusion ability D/a2 can

be calculated according to the following equation [39]:

log i ¼ log

�� 6FD

da2ðC0 � CsÞ

�� p2

2:303Da2t (3)

where i (A/g) is the current density, F is Faraday constant,

D (cm2/s) is the hydrogen diffusion coefficient, d (g/cm3) is the

density of the hydrogen storage alloy, a (cm) is the alloy

particle radius, C0 (mol/cm3) is the initial hydrogen concen-

tration in the bulk of the alloy, Cs (mol/cm3) is the hydrogen

concentration on the surface of the alloy particles, and t (s) is

the discharge time. Since the average size of the hydride

particles are not similar for the different samples, it is

reasonable to useD/a2 to evaluate the discharge kinetics of the

electrode contributed by hydrogen diffusion. The values of

D/a2 for the samples were determined from the slopes of

0.8 1.2 1.6 2.0 2.40.0

0.2

0.4

-Zimag ( Ω

)

Zreal (Ω)

Bare Graphite MWCNTs CAs CNFs

CPE2Rs

Rc

CPE1

Rct W

o

Fig. 8 e Electrochemical impedance spectra and

corresponding equivalent circuit of the Mg2Ni-based

hydride modified with various carbonaceous additives.

Fig. 9 e Potentiostatic discharge curves of the Mg2Ni-based

hydride modified with various carbonaceous additives.

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 en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 0 9659

linear current responses and listed in Table 1. It is found that

the value of D/a2 is increased after themodifications. Graphite

addition leads to the largest D/a2 value, indicating that the

hydride modified with graphite possesses the best hydrogen

diffusion ability.

In summary, both exchange current density I0 and D/a2

value are enhanced by the carbonaceous modifications. As

HRD is determined by both of the hydrogen diffusion in the

alloy and charge-transfer on the alloy surface, it is reasonable

that the electrochemical kinetics property is improved after

the modifications. For the hydride modified with graphite, it

has the largest I0 and D/a2 value, and hence possesses the best

high rate dischargeability.

4. Conclusions

The structural and electrochemical properties of the Mg2Ni-

based hydride modified with various carbonaceous additives

were investigated systematically. The XRD and SEM results

reveal that additions of carbonaceous materials can signifi-

cantly facilitate amorphization or nanocrystallization and the

decrease of particle size of the hydride. Electrochemical

measurements indicate that the carbonaceous modifications

improve both the discharge capacity and electrochemical

kinetics. More specifically, the maximum discharge capacity,

the high rate dischargeability, the exchange current density I0and the D/a2 value are all increased after the modifications,

while the electrochemical reaction resistance Rct decreases.

Furthermore, the hydride modified with graphite exhibits the

highest discharge capacity (531 mAh/g), the largest D/a2 value

(2.2 � 10�5 s�1), I0 (66.7 mA/g) and the smallest Rct (212.2 mU).

Acknowledgements

This work is supported by National Natural Science Founda-

tion of China (Nos. 50871052, 50601014), the National Hi-Tech

Research and Development Program of China (863 Program)

(No. 2007AA05Z110), the PhD Programs Foundation ofMinistry

of Education of China (No. 20093221110008), Natural Science

Foundation of Jiangsu Province (No. BK2009361) and Key

Laboratory of Inorganic and Related New Composite Materials

of Jiangsu Province.

r e f e r e n c e s

[1] Peng B, Chen J. Functional materials with high-efficiencyenergy storage and conversion for batteries and fuel cells.Coordin Chem Rev 2009;253:2805e13.

[2] Chen P, Zhu M. Recent progress in hydrogen storage. MaterToday 2008;11:36e43.

[3] Pan HG, Ma S, Shen J, Tan JJ, Deng JL, Gao MX. Effect of thesubstitution of PR for LA on the microstructure andelectrochemical properties of La0.7 � xPrxMg0.3Ni2.45Co0.75

Mn0.1Al0.2 (x ¼ 0.0e0.3) hydrogen storage electrode alloys. IntJ Hydrogen Energy 2007;32:2949e56.

[4] Tian QF, Zhang Y, Sun LX, Xu F, Tan ZC, Yuan HT, et al.Effects of Pd substitution on the electrochemical propertiesof Mg0.9 � xTi0.1PdxNi (x ¼ 0.04e0.1) hydrogen storage alloys.J Power Sources 2006;158:1463e71.

[5] Lei YQ, Wu YM, Yang QM, Wu J, Wang QD. Electrochemical-behavior of some mechanically alloyed MgeNi-basedamorphous hydrogen storage alloys. Z Phys Chem 1994;183:379e84.

[6] Cui N, He P, Luo JL. Magnesium-based hydrogen storagematerials modified by mechanical alloying. Acta Mater 1999;47:3737e43.

[7] Wang MH, Zhang LZ, Zhang Y, Sun LX, Tan ZC, Fen X, et al.Effects of partial substitution by Fe and Co for Ni in theMg1.75Al0.25Ni electrode alloy on their electrochemicalperformances. Int J Hydrogen Energy 2006;31:775e9.

[8] Zhang YH, Li BW, Ren HP, Ma ZH, Guo SH, Wang XL. Anelectrochemical investigation of melt-spun nanocrystallineMg20Ni10 � xCux (x ¼ 0e4) electrode alloys. Int J HydrogenEnergy 2010;35:2385e92.

[9] Anik M, Akay I, Ozdemir G, Baksan B. Electrochemicalhydrogen storage performance of MgeTieZreNi alloys. Int JHydrogen Energy 2009;34:9765e72.

[10] Jiao LF, Yuan HT, Wang YJ, Wang YM. Electrochemicalproperties of magnesium-based hydrogen storage alloysimproved by transition metal boride and silicide additives.Int J Hydrogen Energy 2009;34:1476e82.

[11] Zhang YH, Jiao LF, Wang YJ, Wang QH, Zhang YY, Liu L, et al.Influence of CoSi on the electrochemical hydrogen storageproperties of MgNi alloy. Int J Hydrogen Energy 2008;33:4819e23.

[12] Chen J, Bradhurst DH, Dou SX, Liu HK. The effect of chemicalcoating with Ni on the electrode properties of Mg2Ni alloy.J Alloys Compd 1998;280:290e3.

[13] Anik M, Akay I, Topcu S. Effect of electroless nickel coatingon the electrochemical hydrogen storage characteristics ofAl and Zr including Mg-based alloys. Int J Hydrogen Energy2009;34:5449e57.

[14] Pedneault S, Huot J, Roue L. Nanostructured Mg2Nimaterials prepared by cold rolling and used as negativeelectrode for NieMH batteries. J Power Sources 2008;185:566e9.

[15] Hima Kumar L, Viswanathan B, Srinivasa Murthy S.Hydrogen absorption by Mg2Ni prepared by polyol reduction.J Alloys Compd 2008;461:72e6.

[16] Akiyama T, Isogai H, Yagi J. Hydriding combustion synthesisfor the production of hydrogen storage alloy. J Alloys Compd1997;252:L1e4.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 5 ( 2 0 1 0 ) 9 6 5 3e9 6 6 09660

[17] Li LQ, Akiyama T, Yagi J. Hydrogen storage alloy of Mg2NiH4

hydride produced by hydriding combustion synthesis frompowder of mixture metal. J Alloys Compd 2000;308:98e103.

[18] Li LQ, Akiyama T, Yagi J. Activation behaviors of Mg2NiH4 atdifferent hydrogen pressures in hydriding combustionsynthesis. Int J Hydrogen Energy 2001;26:1035e40.

[19] Zhu YF, Liu YF, Gu H, Li LQ. Structural and hydriding/dehydriding properties of MgeLaeNi-based composites.J Alloys Compd 2009;477:440e4.

[20] Liu DM, Zhu YF, Li LQ. Crystal defect analysis and surfacecharacteristics ofMg2NiH4producedbyhydriding combustionsynthesis. Int J Hydrogen Energy 2007;32:2417e21.

[21] Gu H, Zhu YF, Li LQ. Catalytic mechanism of Nb2O5 and NbF5on the dehydriding property of Mg95Ni5 prepared byhydriding combustion synthesis and mechanical milling. IntJ Hydrogen Energy 2009;34:7707e13.

[22] Liu XF, Zhu YF, Li LQ. Hydrogen storage properties ofMg100 � xNix (x ¼ 5, 11.3, 20, 25) composites prepared byhydriding combustion synthesis followed by mechanicalmilling (HCS þ MM). Intermetallics 2007;15:1582e8.

[23] Zhu YF, Wang YC, Li LQ. Electrochemical properties ofMg-based hydrogen storage alloys prepared by hydridingcombustion synthesis and subsequent mechanical milling(HCS þ MM). Int J Hydrogen Energy 2008;33:2965e9.

[24] Zhang WF, Zhu YF, Yang C, Li LQ. Effects of metal additive onelectrochemical performances of Mg-based hydrogen storagematerials prepared by hydriding combustion synthesis andsubsequent mechanical milling (HCS þ MM). Int J HydrogenEnergy; 2010;. doi:10.1016/j.ijhydene. 2009.12.038.

[25] Iwakura C, Inoue H, Zhang SG, Nohara S. A new electrodematerial for nickelemetal hydride batteries: MgNi-graphitecomposites prepared by ball-milling. J Alloys Compd 1999;293e295:653e7.

[26] Iwakura C, Inoue H, Nohara S, Shin-ya R, Kurosaka S,Miyanohara K. Effects of surface and bulk modifications onelectrochemical and physicochemical characteristics ofMgNi alloys. J Alloys Compd 2002;330e332:636e9.

[27] Wu DC, Li L, Liang GY, Guo YL, Wu HB. Improvedelectrochemical properties of amorphous Mg65Ni27La8electrodes: surface modification using graphite. J PowerSources 2009;189:1251e5.

[28] Guo ZP, Huang ZG, Konstantinov K, Liu HK, Dou SX.Electrochemical hydrogen storage properties of

nonstoichiometric amorphous MgNi1 þ x-carbon composites(x ¼ 0.05e0.3). Int J Hydrogen Energy 2006;31:2032e9.

[29] Wang FX, Gao XP, Lu ZW, Ye SH, Qu JQ, Wu F, et al.Electrochemical properties of Mg-based alloys containingcarbon nanotubes. J Alloys Compd 2004;370:326e30.

[30] Wang Y, Deng WQ, Liu XW, Wang SY, Wang X.Electrochemical properties of ball-milled LaMg12eNicomposites containing carbon nanotubes. Int J HydrogenEnergy 2009;34:1444e9.

[31] Gao XP, Wang FX, Liu Y, Pan GL, Li S, Qu JQ, et al.Electrochemical hydrogen discharge properties of MgNi-carbon nanotube composites. J Electrochem Soc 2002;149:A1616e9.

[32] Huang ZG, Guo ZP, Calka A, Wexler D, Liu HK. Effects ofcarbon black, graphite and carbon nanotubes additives onhydrogen storage properties of magnesium. J Alloys Compd2007;427:94e100.

[33] Boonstra AH, Bernards TNM, Lippits GJM. The influence ofoxidation upon the storage capacity of LaNi5 electrodes.J Less-Common Met 1990;159:327e36.

[34] Zhang Y, Lei YQ, Chen LX, Yuan J, Zhang ZH, Wang QD. Theeffect of partial substitution of Zr for Ti on theelectrochemical properties and surface passivation film ofMg35Ti10 � xNi55 (x ¼ 1, 3, 5, 7, 9) electrode alloys. J AlloysCompd 2002;337:296e302.

[35] Feng F, Geng M, Northwood DO. Electrochemical behavior ofintermetallic-based metal hydrides used in Ni/metal hydride(MH) batteries: a review. Int J Hydrogen Energy 2001;26:725e34.

[36] Notten PHL, Hokkeling P. Double-phase hydride formingcompounds: a new class of highly electrocatalytic materials.J Electrochem Soc 1991;138:1877e85.

[37] Ise T, Nurata T, Hirota Y, Nogami M, Nakahori S. The effect ofparticle size on the electrochemical properties of hydrogenabsorbing alloy electrodes. J Alloys Compd 2000;298:310e8.

[38] Kuriyama N, Sakai T, Miyamura H, Uehara I, Ishikawa H,Iwasaki T, et al. Electrochemical impedance anddeterioration behavior of metal hydride electrodes. J AlloysCompd 1993;202:183e97.

[39] Zheng G, Popov BN, White RE. Electrochemicaldetermination of the diffusion coefficient of hydrogenthrough a LaNi4.25Al0.75 electrode in alkaline aqueoussolution. J Electrochem Soc 1995;142:2695e8.