HOU Zhonghui, YUAN Zeming, FENG Dianchen, SUN Hao, ZHANG Yanghuan*

(1. Department of Functional Material Research, Central Iron and Steel Research Institute, Beijing 100081, China; 2. College of Material and Metallurgy, Inner Mongolia University of Science and Technology, Baotou 014010, China)

Abstract: The PrMg12-type composite alloy of PrMg11Ni + x wt% Ni (x = 100, 200) with an amorphous and nanocrystalline microstructure were synthesized through the mechanical milling. Effects of milling duration and Ni content on the microstructures and electrochemical hydrogen storage performances of the ball-milled alloys were methodically studied. The ball-milled alloys obtain the optimum discharge capacities at the first cycle. Increasing Ni content dramatically enhances the electrochemical property of alloys. Milling time varying may obviously impact the electrochemical performance of these alloys. The discharge capacities show a significant upward trend with milling duration prolonging, but milling for a longer time more than 40 h induces a slight decrease in the discharge capacity of the x = 200 alloy. As milling duration increases, the cycle stability clearly lowers, while it first declines and then augments under the same condition for the x = 200 alloy. The high-rate discharge abilities of the ball-milled alloys show the optimum values with milling time varying.

Key words: PrMg12-type alloy; mechanical milling; nanocrystalline and amorphous; electrochemical performance; electrochemical kinetics

1 Introduction

The excessive consumption of limited fossil fuels greatly accelerates its depletion speed and simultaneously causes a series of environmental problems. Especially, an amazing growth in the rate of global warming and serious air pollution have forced us to make a change from fossil fuels to cleaner and renewable energy sources. Many clean and renewable energy resources, such as solar, wind,ocean, geothermal, hydrogen and others, are viewed as potential alternatives. Some investigation and statistical results showed that a quarter of the world total energy was consumed by transport[1]and about 23% of global CO2emission originates from vehicular waste gases through the combustion of fossil fuels[2]. Recently,an authoritative survey report from the Ministry of Environmental Protection of China declared that automobile exhaust is the main culprit which gives rise to the severe haze in Beijing. Therefore, to widely develop and apply clean energy vehicles is considered to be an acclaimed strategy to reduce both dependence carbon dioxide emissions and on fossil fuels. Therefore,a widespread application of clean energy vehicles, such as hydrogen fuel cell vehicles, hybrid electric vehicle(HEV) and electric vehicle (EV), is considered to be an attractive strategy to reduce both energy consumption and carbon dioxide emissions. With a consideration regarding the convenience for transportation, hydrogen,of all the available clean energy, is regarded as best fuel owing to the characteristics of inexhaustible supply,high energy efficiency and zero-emission of greenhouse gas[3]. An insurmountable problem for the realization of the application target is to develop a practical hydrogen storage system[4]. Among hydrogen storage methods,metal hydride hydrogen storage is regarded as one of the most promising alternatives to satisfy the demands for mobile application[5]. Recently intensive research has been undertaken to explore hydrogen storage mechanisms ground on different metal hydrides. Many metal hydrides have been found and are thought to have potential application prospect. Especially, rare-earthbased AB5and AB2-type alloys have been generally industrialized in USA, Japan and China. Unfortunately,none of which can satisfy the all requirements of the performance proposed by U.S. Department of Energy(DOE) for vehicular applications[6,7]. In terms of hydrogen absorption capacity, Mg-based alloys are deemed to be a preferred candidate for the negative electrodes of Ni-MH batteries in the hydrogen fuelcell vehicles[8,9]. Especially, REMg12-type alloys have captured considerable attention owing to its absorbed gaseous hydrogen quantity of 3.7 wt%-6.0 wt%[10]and theoretical electrochemical discharge capacity over 1 000 mAh·g–1[11]. Nevertheless, some inherent issues of Mg-based alloys, such as relatively high dehydrogenation temperatures, poor hydrogenation/dehydrogenation dynamics and exceedingly low electrochemical discharge capacity at ambient temperature, severely restrict their practical application for the negative electrode materials of Ni-MH battery or hydrogen storage materials for on-board use.Numerous literatures have reported that microstructure modification and alloying are two effective approaches for ameliorating the hydrogen storage performances of Mg-based alloys[12,13]. Particularly, substituting rare earth elements for part of Mg in Mg-based alloys leads to the decreased hydride stability and the facilitated hydrogen desorption[14,15]. Moreover, it was ascertained that the hydrogenation/dehydrogenation dynamics for the Mg-based alloys are highly sensitive to their microstructures[16]. Particularly, reducing the grain sizes of Mg-base alloys far below the micrometer scale can dramatically improve their hydrogenation/dehydrogenation properties[17]. Some techniques,especially mechanical alloying[18], melt spinning[19]and annealing[20], were successfully used to synthesize amorphous and nanocrystalline Mg-based alloys with a highly homogeneous element distribution. Liet al[20]prepared A2B7-type La-Mg-Ni alloys by a method of melt spinning at first and annealing later. This method is beneficial for the formation of single A2B7-type phases inside the alloy which has better cycle stability and higher discharge capacity than other annealed alloys.Huanget al[21]improved the capacity retention rate after 30 cycles of as-milled MgNi alloy from 24.0%to 55.7% by substituting part of Mg by Ti. Huet al[22]prepared CeMg12/Ni composites with a nanocrystalline and amorphous structure by ball milling. The maximal discharge capacity of the composites increased from 46.7 to 1 061.2 mAh/g by increasing Ni content from 50 wt% to 200 wt%.

In the present work, part of Mg in the PrMg12-type alloy was substituted by Ni. The amorphous and nanocrystalline composite alloys of PrMg11Ni withxwt% Ni (wherex= 100, 200) were synthesized via the mechanical alloying process. And the influences of Ni content and milling duration on the structure and electrochemical hydrogen storage performance of the alloys have been studied.

2 Experimental

The raw materials for preparing the experimental alloy were weighted following the mass ratio of PrMg11Ni. In view of melting loss, Pr and Mg were over weighted of 5% and 8% respectively. The raw materials with purity above 99.9% were provided from CISRI Corporation. The PrMg11Ni alloy was prepared through a vacuum induction furnace filled with a 0.04 MPa helium atmosphere pressure to protect Mg from volatilization. The molten alloy was poured into a copper mould cooled and a cast ingot was obtained.Subsequently, the as-cast alloys were ground into powders of about 50 μm diameter and mix it with carbonyl nickel powder (alloy-to-nickel powder ratio 1:1 and 1:2) which is performed in a glove box. Then the mixtures of powder were ground in a planetary-type mill and under Ar atmosphere to protect the powder from oxidation in the ball milling process. Mixtures of powders and Ni–Cr stainless steel balls (powder-toball ratio 1:35) were jointly loaded into Ni–Cr stainless steel cans. And the mill was set to 135 rpm for 5, 10,20, 40, and 60 h of milling.

The phase microstructures of these as-cast and ball-milled hydrogen storage alloys were subjected to D/max-2400 X-ray diffraction (XRD) analysis in a diffractometer with Cu-Ka1radiation filtered by graphite, and the experimental parameters of which are 10 °·min–1, 40 kV, and 160 mA respectively. A Philips Quanta 400 SEM connected with an energy dispersive spectrometer (EDS) was utilized to observe the morphologies. The powder samples of these ball-milled alloys were investigated by 200 kV JEM-2100F highresolution transmission electron microscope (HRTEM),along with their crystalline states identified through electron diffraction (ED).

Mixtures of carbonyl nickel powders and alloy powders (nickel-to-alloy power ratio 4:1) were made into round electrode pellet of 15 mm diameter by the cold-press process at a 35 MPa pressure whose gross weight is 1 g. The primary role of the carbonyl nickel powder is conductive agent and binder. The electrochemical kinetics were measured at 303 K through a tri-electrode open cell including a sintered Ni(OH)2/NiOOH counter electrode, a working electrode, namely, the metal hydride electrode, and a Hg/HgO reference electrode, which were all dipped in an 6 M KOH electrolyte. And the discharge voltage means the voltage between the reference and negative electrodes. In every cycle, the alloy electrode was firstly charged and then discharged to the cut-off voltage of – 0.500 V at the same constant current density after 15 min intervals.

A PARSTAT 2273 electrochemical workstation was used to measure the EIS at different temperature.The fresh electrodes was rested for 2 h after being first fully charged to make the open circuit potential stable. At a 50% depth of discharge (DOD), the EIS of the alloy electrodes were tested by 60 points per decade seconds in the 10 kHz - 5 MHz frequency range at 5 mV of the amplitude of signal potentiostatic or galvanostatic measurements. The potentiodynamic polarization curves were measured at the potential range of – 1.2 to + 1.0 V (vsHg/HgO) with a scan rate of 5 mV s–1. The potentiostatic discharge of laboratorial electrodes in the fully charged stage was measured using CorrWare electrochemistry corrosion software at 500 mV potential steps for 5 000 s.

3 Results and discussion

3.1 Microstructure characteristics

The microstructure characteristics and phase components for the as-cast and ball-milled composite alloys of PrMg11Ni withxwt% Ni (wherex= 0, 100,200) are subjected to X-ray diffraction (XRD) analysis,as presented in Fig.1. The as-cast PrMg11Ni alloy contains a major phase PrMg12and a secondary phase Mg2Ni. The diffraction peaks of the PrMg12and Mg2Ni phases almost disappear after mechanical milling,showing the crystalline structure has transformed to an amorphous and nanocrystalline structure. Besides, the alloy diffraction peaks merged and broadened with the increase of milling duration, which may be ascribed to the lattice stress and expansion caused by that hydrogen atoms enter the lattice interstitials of the alloys.Prolonged milling time also causes the alloy grain sizes to considerably decrease and the markedly increasing degree of disorder in the microstructure, showing that abundant internal energy stored. It leads to the nonstabilization of lattice, generating the ultrafine grain or even an amorphous structure. Moreover, it is observed that, for the same milling time, increasing Ni content makes the intensity of diffraction peaks evidently lower and its width clearly increase, indicating that increasing Ni content is conductive to the glass formation of the alloy. As considered by Abdellaouiet al[23], increasing Ni content can lower the activation energy for the crystalline phase transforming to the amorphous phase in the REMg12alloys.

Fig.1 XRD profiles of as-cast and milled PrMg11Ni + x wt% Ni (x= 0, 100, 200) alloys: (a) x = 100; (b) x = 200

Fig.2 demonstrates ED patterns and HRTEM micrographs of the as-milled PrMg11Ni +xwt% Ni(wherex= 100, 200) alloys. It is evident that, although dispersed strongly in alcohol, the alloy particles obviously agglomerate together. The average sizes of the alloy particles measured by linear intercept method range from 30 to 50 nm. The unexpected fact shows without sharp decrease in the particle sizes of the alloys with milling duration increasing from 10 to 40 h, the particle sizes of the alloys do not always decrease with the extension of milling time. The locally magnified micrographs of the as-milled alloys exhibit an amorphous and nanocrystalline microstructure, and the quantity of amorphous phase in the as-milled alloys shows a visibly upward trend when milling time and Ni content increase, supported by analyzing Debye-Scherrer rings. Some crystal defects are also generated during the mechanical alloying process, such as subgrains and grain boundaries and sub-grains. As milling duration varies from 10 to 40 h, there is a dramatical decrease in the grain size and an obvious increase in the microstructure for thex= 200 alloy.

3.2 Discharge capacity, potential characteristics and activation capability

Fig.2 HRTEM micrographs and ED patterns of the as-milled PrMg11Ni + x wt% Ni ( x= 100, 200) alloys: (a) and (d) x = 100 alloy milled for 10 h; (b) and (e) x = 200 alloy milled for 10 h; (c) and (f) x = 200 alloy milled for 40 h

It is mandatory for the alloy electrode of Ni-MH battery to be activated easily. Generally speaking, the activation capability was determined by the number of charge-discharge cycles when it reaches the greatest discharge capacity after repeating charge-discharge process at a constant current density. The fewer the number of charging-discharging cycle is, the better the activation performance will be. Fig.3 describes the discharge capacity of the ball-milled alloys as functions of cycle number. The ball-milled alloys exhibit remarkable activation capability, obtaining the optimum discharge capacities at the first chargedischarge cycle without needing any activation. The excellent activation capability of the tested alloys is most probably due to the following two aspects.Firstly, during mechanical milling, repeated highenergy impacts results in alloy particle fracturing,which breaks the surface oxide layers and expose fresh surfaces of the alloys to electrolyte, improving the activation performance. Secondly, added Ni powder in the process of milling probably forms a mechanical coated layer on the particle surfaces of the alloys,further increasing the activation capability.

Fig.3 Evolutions of the discharge capacity of the as-milled PrMg11Ni + x wt% Ni (x = 100, 200) alloys with cycle number: (a) x = 100; (b) x= 200

Fig.4 Discharge potential curves of the as-milled PrMg11Ni + x wt% Ni (x = 100, 200) alloys: (a) x = 100; (b) x = 200

Fig.4 presents the discharge potential curves for the ball-milled alloy electrodes of PrMg11Ni withxwt% Ni (wherex= 100, 200) at a current density of 60 mA·g–1. As a vitally significant performance of an alloy electrode, the discharge potential characteristic connects very directly to the stability of the output power of a battery, which is determined by the potential plateau of the alloy discharge curves. And the longer and more horizontal discharge potential plateau corresponds to the better discharge potential characteristics of the alloy. The mechanical milling highly improves the discharge potential characteristics,lengthening the discharge plateau and enhancing the discharge potential. The discharge potential of an alloy electrode is bound up with the internal resistance of a battery that includes polarization resistance and ohmic internal resistance, as considered by Laiet al[24]. The internal resistance primarily rests with the hydrogen diffusion in the alloy and it reduces with the hydrogen diffusion coefficient increasing. Apparently, it is observed that the variation of milling duration plays an evident role in the discharge capacity of alloys.The discharge capacities of the alloys as functions of milling duration are also presented in Figs.4(a) and 4(b), respectively. Apparently, the discharge capacities markedly increase with milling time extending, but it is noteworthy that the milling for more than 40 h induces a slight decline in the discharge capacity of thex= 200 alloy. By comparing Figs.4 (a) and 4(b), increasing Ni content is found to dramatically enhance electrochemical performances of the as-milled alloys. To be more specific, Ni content increased from 100 wt% to 200 wt% causes the discharge capacity of the alloys milled for 40 h to vary from 141.8 to 1 059.4 mAh·g–1at the discharge current density of 60 mA·g–1. Overall, the mechanical milling dramatically improves the discharge capacities of the experimental alloys. The positive role of mechanical alloying in the discharge capacity is most likely ascribed to forming a nanocrystalline microstructure due to the fact that the grain boundary region has a relatively greater hydrogen concentration than the amorphous region and grain interior region[25]. In addition, it is believed that hydrogen atoms adhere more strongly to the surface of the nanocrystalline nickel than to the polycrystalline nickel, which is very conductive to the hydrogen dissociation reaction and an enhancement in the capacity of the electrode[26]. Regarding the positive effect caused by increasing Ni content on the discharge capacity, it is unquestionably ascribed to its high electrocatalytic activity. Mechanical milling can result in the cold welding of nickel particles and alloy particles, forming the highly dispersed metallic Ni nanocrystals through alloy matrix which reduces the stability of the hydride and improves the discharge capacity of the alloy[27].

3.3 Electrochemical cyclic stability

Fig.5 SEM morphologies and XRD pattern of 10 h-milled PrMg11Ni + 100 wt% Ni alloy before and after electrochemical cycles: (a) before cycling; (b) after cycling; (c) XRD pattern after cycling

Fig.6 Evolution of the capacity retaining rate (Sn) of the as-milled PrMg11Ni + x wt% Ni (x = 100, 200) alloys with cycle number: (a) x =100; (b) x = 200

Whether an alloy can be used as a negative electrode material in Ni-MH battery goes hand in hand with the cycling stability of an alloy electrode. Various factors can send the discharge capacity of an alloy into recession such as the Mg(OH)2layer formation,amorphous phase crystallization and alloy particle pulverization. Fig.5 compares the alloy’s morphology before and after charge/discharge cycles, which is helpful for investigating the mechanism of the decrease in discharge capacity. Unlike the sample alloy before the cycle, the surface of the alloy after 20 cycles is covered with a layer of loose flocs, as Fig.5 (b) shows.This loose layer was identified as Mg(OH)2by XRD measurements, which encapsulates the alloy particles and is not conducive to their contact and reaction with the electrolyte. As a result, the Mg(OH)2layer leads to the decrease in cycling stability[28]. In addition, the postcycle XRD curve has sharper diffraction peaks than the pre-cycle XRD curve. The sharpened diffraction peaks mean that the electrochemical cycle crystallizes the amorphous phase, which also contributes to the reduction of discharge capacity. Here, capacity retaining rate (Sn) is adopt to symbolize the cyclic stability of the experimental alloy electrode. It can be interpreted asSn=Cn/Cmax× 100%, whereCnis the discharge capacity when the number of the charge-discharge cycle isnat a current density of 300 mA·g–1andCmaxis the maximum discharge capacity, respectively. If an alloy electrode,after being repeated charge-discharge cycling, still has a high capacity retention rate, we consider that it owns superior electrochemical cycling stability and fits for the negative materials in Ni/MH batteries. The cycle number as functions of theSnvalues of the ball-milled PrMg11Ni +xwt% Ni (wherex= 100, 200) alloys are demonstrated in Fig.6. Evidently, theSnvalues rapidly decline with cycle number increasing, exhibiting poor electrochemical cycle stability. To well illustrate the effect of milling duration on the alloy cycling stability,theS20(n= 20) values of the alloys as functions of milling duration are described in Fig.7. It is observed that theS20value for thex= 100 alloy always decreases while that of thex= 200 alloy shows an upward trend after the first decline with milling duration increasing. As is known to all, the rapid degradation in the discharge capacity of Mg-based alloys is mainly because the Mg in the alkaline KOH solution suffers from the severe corrosion and then form Mg(OH)2on the alloy surface[28]. The negative effect of mechanical milling on the alloy cycling stability is convinced to be linked with the nanocrystalline structure and lattice strain produced by mechanical milling owing to inevitable intergranular corrosion. Evidently,prolonging milling time over 20 h is shown to positively affect the cycle stability of thex= 200 alloy,for which two reasons are considered to be primarily responsible. Firstly, increasing Ni content and milling time greatly promotes the formation of an amorphous structure since an alloy has an excellent resistance to corrosion in an amorphous structure, hence improving the cycle stability. Secondly, prolonging milling time significantly improves the interface structure of alloy and Ni particles and probably facilitates to form the particle surfaces coated Ni due to Ni content increasing.That Ni additive can ameliorate the corrosion resistance of an alloy has been reported by numerous literatures.

Fig.7 Evolution of the capacity retaining rate (S20) of the asmilled PrMg11Ni + x wt% Ni (x = 100, 200) alloys with milling time

Fig.8 Evolution of the HRD of the as-milled PrdMg11Ni + x wt% Ni (x = 100, 200) alloys with current density: (a) x = 100; (b) x = 200

3.4 Electrochemical kinetics

To maintain a high discharge capacity even during the charge-discharge cycle with high current density is conductive to the practical application for an alloy electrode of Ni–MH battery, especially power battery. Generally, the electrochemical dynamics of an alloy electrode can be evaluated byHRDthat is interpreted as:HRD=Ci/C60× 100%, whereC60andCiare the optimum discharge capacities when the alloy electrode is charged-discharged at the current densities of 60 andimA·g–1. Fig.8 demonstrates the variations of theHRDs of the ball-milled PrMg11Ni +xwt% Ni(wherex= 100, 200) alloys with current density. TheHRDs conspicuously decrease with current density growing. To facilitate comparison, we take the current density of 300 mA·g–1(i= 300 mA·g–1) as a criterion to establish the relationship between the milling duration andHRDs, as described in Figs.8. TheHRDs show a downward trend after the first rise with milling time extending, and the maximumHRDs of as-milled PrMg11Ni +xwt% Ni (wherex=100, 200) alloys are 82.51% and 86.91% respectively, which are nearly equivalent to the electrochemical dynamics of the rare earth-based AB5-type alloy electrode possessing acknowledged excellent kinetics performance.Meanwhile, thex= 200 alloy is also found to exhibit a higherHRDthan that of thex= 100 alloy for all the milling duration, suggesting that increasing Ni content is beneficial to ameliorate the electrochemical dynamics.

As we all know, theHRDof an alloy electrode turns out to hinge on the diffusion ability of hydrogen atoms and charge-transfer resistance[29]. To disclose the mechanism associated with the electrochemical dynamics of alloys, the roles of Ni content and milling duration on the charge-transfer resistance and hydrogen diffusion ability are studied. The diffusion coefficient of hydrogen atoms can be derived from the semi-logarithmic graphs of anode currentversusthe working time of an alloy electrode, as shown in Fig.9. In consideration of the White’s model[30], the hydrogen diffusion coefficient in the bulk alloy can be derived easily from the slope of the linear portion of the relevant plots in accordance with the following equations:

whereimeans the density of diffusion current (A·g–1),Drepresents the diffusion coefficient of hydrogen atoms (cm2·s–1),dcorresponds to the density of hydrogen storage alloy (g·cm–3),ais the radius of alloy particle (cm),C0refers to the original concentration of hydrogen atom in the bulk alloy (mol·cm–3),Cscorresponds to the hydrogen concentration on the alloy particle surface (mol·cm–3), andtjust means the discharge time (s). TheDvalues of the alloys obtained from Eq.(2) as functions of milling duration are shown in Fig.9. Obviously, theDvalues show a downward trend after the first rise with milling time extending. As an another important parameter relevant to the hydrogen diffusion ability, the limiting current density (IL) is used to evaluate the effect of Ni content and milling duration on the hydrogen diffusion rate of the experimental alloy electrodes. TheILvalue can be derived from the potentiodynamic polarization curve of an alloy electrode, as depicted in Fig.10. As we all see, in all cases, there is an obvious inflection point on each anodic polarization curve, namely a critical value that is referred to as limiting current density (IL). The inflection point means an oxidizing reaction occurring on the surface of the alloy electrode, and then the formed oxidation layer further prohibits hydrogen atoms from penetrating inward[26]. So, theILcan be deemed to be a critical current density for passivating.On the basis of the data in Fig.10, theILvalues of the alloys as functions of milling time are inserted in Figs.10 (a) and 10(b), respectively. By comparing Fig.9 and Fig.10, theDandILvalues as functions of milling duration are found to be very similar, suggesting that these two parameters factually reflect the diffusion ability of hydrogen atoms. A careful observation finds that the milling times corresponding to the maximumDandILvalues are different from those at which the alloys obtain the maximumHRDs, which means that,other than the hydrogen diffusion ability, there are other reasons affecting electrochemical dynamics.

Fig.9 Semilogarithmic curves of anodic current vs time responses of the as-milled PrMg11Ni + x wt% Ni ( x = 100, 200) alloys: (a) x =100; (b) x = 200

Fig.10 Potentiodynamic polarization curves of the as-milled PrMg11Ni + x wt% Ni (x = 100, 200) alloys: (a) x = 100; (b) x = 200

Fig.11 Electrochemical impedance spectra (EIS) of the as-milled PrMg11Ni + 200 wt% Ni alloy and the equivalent circuit

Apparently, it definitely is charge transfer rate. In the case of the charge-transfer rate on the surface of an alloy electrode, it can be evaluated by its electrochemical impedance spectrum (EIS)in accordance with the Kuriyama’s model[31]. As a typical representative, the EIS curves of thex= 200 alloy milled at different times are depicted in Fig.11.Two distorted capacitive loops are observed at high and middle frequency regions separately, along with one line at low frequency region in each EIS, which commendably describes the electrochemical process of an alloy electrode. Meanwhile, the smaller semicircle in the high frequency region displays the contact resistance between the conductive material and alloy powder; the larger one in the middle frequency region corresponds to the charge transfer resistance (Rct) on the alloy surface, whereas the straight line in the low frequency region corresponds to Warburg impendence.With the aid of the equivalent circuit inserted Fig.11,theRctvalues were obtained from the fitting program Z-View. Kuriyamaet alconsidered that theRctvalue depends on both the reactivity of the alloy surface and reaction area. And the electrochemical reactivity on the alloy surface can be evaluated with apparent activation enthalpy ΔrH*, which can be obtained by the following equation:

Fig.12 Electrochemical impedance spectra (EIS) of the as-milled PrMg11Ni + x wt% Ni (x = 100, 200) alloys at various temperature: (a) x = 100; (b) x = 200

whereTis temperature,Ris the gas constant,Rctcorresponds to the charge-transfer resistance for the metal hydride electrodes andC0represents a constant in which the surface area is included. In consideration of the calculation conditions of Eq.(3), the EIS curves for the ball-milled PrMg11Ni +xwt% Ni (wherex=100, 200) alloys at different temperatures (303, 313 and 323 K) were presented in Fig.12. On account of the data in Fig.12, the graphs of ln(T/Rct)vs1/Tcan be plotted by using logarithmic transform of Eq.(3),termed as Kuriyama plots. The Kuriyama plots of the ball-milled alloys are shown in Fig.13, which are found to be nearly linear. From the slopes of the Kuriyama plots, the ΔrH*values can be easily derived. Thus, we establish the relationship between the ΔrH*values of the ball-milled alloys and milling duration, as shown in Figs.13(a) and 13(b). Obviously, the ΔrH*values show an upward trend after the first decline with milling time prolonged. By comparing the change trends of theHRDand ΔrH*value with milling time shown in Figs.8 and 13, it is found that the maximumHRDs of the ball-milled PrMg11Ni +xwt% Ni (wherex= 100,200) alloys correspond to the minimum ΔrH* values,implying that the electrochemical dynamics of the alloys basically hinges on the charge-transfer ability of the alloy electrode surface. So, we can infer that theHRDs of the ball-milled alloys improved by increasing Ni content and milling time are convinced to be ascribed to a decrease in the activation enthalpy ΔrH*.

Generally, the electrochemical charging/discharging process of the electrode in an alkaline solution occurs through the following reaction:

where M represents the hydrogen storage alloy. It indicates that, when the alloy electrode is charged in KOH solution, electrolytic hydrogen atoms on the alloy-electrolyte interface diffuse into bulk alloy, and then store them in the metallic lattice in the form of hydride. In the process of discharging, the hydrogen atoms stored in the bulk alloy diffuse toward the surface where they are oxidized. The hydrogen atoms attaching themselves to the alloy electrode surface are thought to have two places to go, (a) developing into metal hydride through diffusion or (b) integrating together to generate hydrogen molecule. Namely, the introduction of charge current principally lies on the hydrogen diffusion rate within the alloy surface layer,which is just the ratio between the diffusion current and the imposing current. Based on the results presented in Figs.8, 9, 10 and 13, it can be inferred that prolonging milling time plays a positive or negative role in the electrochemical dynamics of the alloys. The positive role is most probably ascribed to the following two aspects: firstly, mechanical milling greatly modifies the surface state of alloy powder, and especially a great number of crystal defects are created, which are very beneficial to accelerate the charge transfer and hydrogen diffusion; secondly, mechanical milling dramatically decreases the sizes of alloy powders and then increases the interface area between the alloy particle surface and the electrolyte. In the case of the negative action generated by milling, it is believed to be associated with the facilitated amorphous phase due to the increase of milling time. An amorphous phase can efficiently prevent the pulverization of alloy in the process of the charge-discharge cycle, which reduce new available surface of alloy electrode, impairing not only the hydrogen diffusion ability but also chargetransfer rate at the interface of alloy and electrolyte[32].

Fig.13 Plots of ln(T/Rct) vs 1/T and evolution of the activation enthalpy ΔrH* values of as-milled alloys

4 Conclusions

a) The nanocrystalline and amorphous PrMg11Ni +xwt% Ni (wherex= 100, 200) alloys were successfully synthesized by mechanical milling. The structural analysis reveals that increasing Ni content is beneficial for the glass formation of the PrMg12-type alloy.

b) Increasing Ni content dramatically enhances the electrochemical capacities of the ball-milled hydrogen storage alloys. Meanwhile, it also markedly ameliorates the electrochemical dynamics performance,for which the activation enthalpy reduced by increasing Ni content is principally responsible.

c) The as-milled alloys exhibit remarkable activation performance and obtain the optimum discharge capacity at first charging and discharging cycles without needing any activation. Overall,prolonging milling duration significantly improves the discharge capacities and discharge potential characteristics of alloys, but it must be mentioned that the milling for more than 40 h brings on a slight reduction in the discharge capacity of thex= 200 alloy.Furthermore, prolonging milling duration also results in an undesired decrease in the chemical cycle stabilities of alloys.

d) The as-milled alloys exhibit superior electrochemical dynamic performance (i e,HRD),and theHRDs have maximum values with milling time varying. Furthermore, kinetics study finds that the maximumHRDs of as-milled alloys just coincide with the minimum ΔrH*values, which implies that the charge transfer ability of alloy electrode surface is a major governing factor affecting the electrochemical dynamics of alloys.