Introduction

Hydrogen storage alloys have attracted extensive attention because of their high energy storage density, long-charge/discharge cyclic life, and good environment compatibility properties [1, 2]. As the secondary Ni/MH batteries negative materials, the AB5-type rare-earth-based hydrogen storage alloys and AB2-type Zr-based or Ti-based alloys have been commercialized successfully [3]. However, the low capacity of AB5-type alloys (the practical value is about 310 mAh g−1) and the slow activation characteristics of AB2-type alloys limit extensive applications [4,5,6]. Kadir et al. [7, 8] developed AB3-type La–Mg–Ni alloys with PuNi3 structure, which exhibited better characteristics than that of the conventional AB5-type alloys. Not long after that, Kohno et al. [9] reported that the maximum discharge capacity of the La0.7Mg0.3Ni2.8Co0.5 alloy reached 410 mAh g−1. Pan and his co-workers [10, 11] also acquired the similar results on the La0.7Mg0.3(Ni0.85Co0.15)3.5 alloy, namely with PuNi3-type structure and the discharge capacity approaching 400 mAh g−1, which exhibits good application prospect. The PuNi3-type structure can be described as the growth along the c-axis of the alternate AB2 and AB5 subunit with 2:1 proportion [9]. Both AB2 and AB5 subunits can absorb hydrogen instead of only one AB5 subunit. Moreover, magnesium hydrogen storage alloy is the most practical option due to its superiority in cost, weight, storage capacity, and material availability [12, 13]. Our previous research indicated that the ratio RE/Mg = 2 exhibited the best performance in RE3 − xMgx(Ni0.7Co0.2Mn0.1)9 alloys [14]. Mg incorporation improved the discharge capacity of La–Mg–Ni-based alloys, and hence, this type alloy electrode become more and more attractive [15].

It is well known that elemental substitution is one of the most effective methods for improving the overall properties of hydrogen storage alloys. The substitution of rare-earth element at A site in La–Mg–Ni-type alloys can improve the electrode properties with a better anti-corrosion, as well as reducing the cost for commercial application. Zhang et al. [16, 17] pointed out that the La–Mg–Ni system alloy exhibited a improved cycle lifetime due to the significant corrosion decreasing caused by the appropriate substitution of Ce for La. Zhai et al. [18, 19]. revealed that the cycle stability and the high-rate dischargeability of the La1 − xPrxMgNi3.6Co0.4 (x = 0–0.4) alloys obviously augmented with increasing Pr content. Ma et al. [20, 21] found that the electrochemical kinetics of the La–Mg–Ni-based hydrogen storage alloys could be effectively improved by partial substitution of Nd for La. Among them, praseodymium element is of great importance for cycle life of rare-earth-based hydrogen storage alloy electrodes.

Moreover, many meaningful research results at B site in the La–Mg–Ni system hydrogen storage alloys have also been done during recent years. Our previous research showed that suitable Mn content could effectively improve the discharge capacity and kinetics performance of hydrogen storage alloys [22]. The substitution of Ni by Mn noticeably decreased the stability of the hydrides against amorphisation on hydrogenation due to the change in crystal-structure type from PuNi3 to CeNi3 [23]. Many efforts have been made to improve the cycling stability of the La–Mg–Ni-base alloys by partial substitution Co for Ni, while the high-rate dischargeability needs to be further improved [24,25,26].

The Al substitution at B site is generally believed to be one of the most effective methods to improve the anti-corrosion ability of La–Mg–Ni-based alloys. Liao et al. [27] reported that Al incorporation decreased the discharge capacity and the high-rate dischargeability, but leads to a significant improvement the cycling stability of the electrode. Liu et al. [28] found that, with the addition of Al, the capacity cycling retention improved markedly due to the alleviated pulverization together with the Al oxide layer on the alloy surface during charge/discharge cycles. Annealing treatment technique promoted the AB5 phase over the A2B7 phase in the La0.7Mg0.3Ni2.8Co0.5 − xAlx alloys, while the annealed samples decreased the gaseous phase hydrogen storage and electrochemical capacities [29]. Guzik et al. [30] revealed that the Al atoms occupied 12% of the 6 h site within the LaNi5 slabs in the La0.77Mg0.23Ni3.41Al0.09 alloy, which absorbed hydrogen when exposed to 10 bar of H2 gas at room temperature. These research results provide important and valuable methods in improving the cycle stability of La–Mg–Ni-based electrode. However, to the best of our knowledge, no detailed work into the evidence for the stability of hydride alloys, and the hydrogen absorption/desorption mechanisms of the alloy electrodes are not entirely clear. Much more deeply study should be done to gain a better understanding the properties correlated with the Al presence.

In this paper, a series of alloys with chemical composition Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) (Re: La-rich mischmetal) are prepared. The effect of Al substituting for Ni on the phase structure, hydrogen storage, and electrochemical properties of the series alloys are studied.

Experimental

Bulk metals (La, Pr, Mg, Ni, Co, Mn, and Al) from Aladdin were used as raw materials. The purity of all starting metals was above 99.9 wt%. The as-cast alloys were prepared by induction levitation melting on a water-cooled copper under argon atmosphere according to the stoichiometric composition Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04), where Re denotes La0.85Pr0.15. The 10 wt% excess of Mg and 5 wt% of Mn over composition was needed to compensate for evaporative loss under preparation progress. The alloys were turned over and re-melted thrice to ensure good homogeneity. The obtained ingots were then mechanically crushed and ground into the fine powder of 200 mesh size for the next experimentation. The microstructures of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) powers were characterized at room temperature, using a Rigaku D/Max 2500 V diffractometer with Cu Kα radiation and a graphite monochromator operated at 40 kV and 200 mA, scanned with the rate of 6° min−1 in the 2θ range of 20–80°. The lattice parameters and cell volumes were analyzed using software Jade 5.0.

In studying the hydrogen storage properties, the pressure–composition–isotherms (P–C–T) curves of the alloys were measured with an automatic Sieverts-type apparatus under the pressure range of 10−3–20 atm at given temperature. About 1.2 g alloy powder with particle size below 200 mesh was used for the test. For activation, the alloy powder was loaded in the clean reaction tube and exposed to the hydrogen atmosphere at room temperature to get rid of the air. Hydriding was conducted under 20 atm H2 pressure for several minutes, and subsequent dehydriding was carried out by evacuating and heating the hydrided powder up to 300 °C. After repeated the hydriding/dehydriding process for three times, the P–C–T data were collected.

The hydride electrodes were prepared by mixing 0.3 g alloy powder (200 mesh) with 1.2 g carbonyl nickel powder. The mixture was then cold-pressed under a pressure of 20 MPa into a pallet of 15 mm in diameter and 1 mm in thickness. The prepared pellet, together with a Ni strip soldered on the surface, was inserted into folded current collector Ni foam, and the rims of the wrapped pellet were firmly spot-welded to keep the electrochemical contact well between the pellet, Ni strip, and Ni foam. The electrodes were immersed in 6 M KOH solution for 24 h to keep wet thoroughly before the tests. Electrochemical measurements were performed at 303 K in a standard three-electrode open cell consisting of a working electrode (hydride electrode), a sintered Ni(OH)2/NiOOH counter electrode with excess capacity, and a Hg/HgO reference electrode immersed in 6 M KOH electrolyte. The emphasis of these measurements was on the electrochemical properties of the working electrode, and thus, the discharge capacity of the counter electrode was designed to be much higher than that of the hydride electrode. Electrochemical tests were carried out on DC-5 automatic battery testing instrument. Each electrode was charged at 100 mA g−1 for 5 h followed by a 15 min break, and then discharged at 60 mA g−1 to a cut-off cell voltage of − 0.5 V versus the Hg/HgO reference electrode.

The cycle stability of the alloy electrodes was conducted by the charge/discharge process after 80 cycles and the cycle life was defined as the following equation:

$$S_{80} = C_{80} /C_{\hbox{max} } \times 100\% ,$$

wherein C80 was the discharge capacity at the 80th charge/discharge cycles and Cmax was the maximum discharge capacity of the alloy electrodes. To investigate the discharge voltage characteristic of the alloy electrodes, the measurements were conducted after the alloy electrode was first completely activated.

Results and discussion

Phase structure

Figure 1 shows the XRD patterns of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) hydrogen storage alloys. It can be seen that the main phases in all of the alloys can be indexed as the characteristic peaks of the LaNi5 phase with CaCu5-type structure, and the (La, Mg)2Ni7 phase with Ce2Ni7-type structure, along with minor LaNi2 phase with MgCu2-type structure. The phase composition of the series alloys remains almost unchanged after partial substitution of Al for Ni. The similar results have been obtained in AB3-type La2Mg(Ni1−xAlx)9(x = 0–0.05) alloys [27]. That leads us to believe that Al atoms occupy the Ni site during substitution. The lattice parameter and the unit cell volume of all the phases in the alloys calculated by the Jade software are listed in Table 1. Owning to the fact that the atomic radius of Al (0.182 nm) is larger than that of Ni (0.162 nm), the parameters would be expected to be increased with the increasing x. However, the present results reveal that, with the progressively increasing substitution of Al for Ni in the alloys, both a parameters and c parameters of the LaNi5 phase and the LaNi2 phase increase, while a and c parameters of the (La,Mg)2Ni7 phase decrease.

Fig. 1
figure 1

X-ray diffraction patterns of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04)alloys

Full size image
Table 1 Crystal lattice parameter of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04)alloys
Full size table

To distinctly manifest the influence of Al on the lattice parameters of phases in the alloys, the cell volume dependence of Al content is plotted in Fig. 2. It can be seen that, with the increasing x, the cell volume of both LaNi5 phase and LaNi2 phase in the alloys increases almost linearly, whereas the cell volume of (La,Mg)2Ni7 phase shrinks slightly. That means, the partial substitution of Al for Ni has only enlarged the structure of LaNi5 phase and LaNi2 phase in the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys. In La–Mg–Ni-based alloys, Al element tended to enter LaNi5 phase rather than (La,Mg)2Ni7 phase during Al incorporation [28]. Obtained results reveal that Al atoms have a strong preference and tend to enter to the LaNi5 and LaNi2 structure. The slight shrinking of (La,Mg)2Ni7 phase is attributed to the lattice distortion during the Al substitution.

Fig. 2
figure 2

Variation of cell volume in different phases of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04)alloys

Full size image

P–C– isotherms

To investigate the effect of Al content on the hydrogen absorption/desorption properties, the P–C–T curves of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys at different temperatures are plotted in Fig. 3a–e. From these curves, the maximum hydrogen storage capacity (MHmax, under 20 atm) can be obtained, and the results measured at 303 K are also listed in Table 2. It can be seen that, at given temperature, MHmax of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys goes down continually with the increasing x. For example, at 303 K, with the substitution of Al for Ni, MHmax of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 alloys is 1.16, 1.10, 1.06, 1.04, and 0.99 wt% for x = 0, 0.01, 0.02, 0.03, and 0.04, respectively. In La–Mg–Ni system alloys, the La2Ni7-type phase plays a dominant role in hydrogen storage capacity of the alloys [31]. The change of hydrogen storage capacity for Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 alloys results from the relative variation of the phase abundance of the (La, Mg)2Ni7 phase and LaNi5 phase [24].

Fig. 3
figure 3

P–C–T curves of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04)alloys at different temperature: a x = 0, b x = 0.01, c x = 0.02, d x = 0.03, and e x = 0.04

Full size image
Table 2 Hydrogen storage characters of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 alloy (303 K)
Full size table

The gaseous phase hydrogen storage properties of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 alloys at 303 K are displayed in Fig. 4. It can be found that the slopes of P–C–T curves are very steep and there is almost no plateau in the absorption/desorption process during the substitution Al for Ni. There are three possible reasons which lead to the steep plateau: (1) A certain portion of the crystalline alloy changes to the amorphous state upon hydriding [27];(2) The loss of Mg volatilization during melting of the alloys also results in the steep plateau [32], which can be improved by adding excess Mg to the synthesis; (3) The alloys possess multi-phase structure, and the compounding of the differing absorption/desorption pressures of their components leads to the steep plateau [33]. In the present study, because the Mg content is excessive during melting, we think that the steep plateau is attributed to the above (1) or/and (3) reasons.

Fig. 4
figure 4

P–C–T curves of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys at 303 K

Full size image

The plateau pressure shown in this work is defined as the mean of the mid-hydrogen storage capacity from the P–C–T curves. The relation between the composition and the plateau pressure of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04) alloys can be obtained from Fig. 4, and is given in Table 2. As we all know, there is a hysteresis loop between hydrogen absorption and desorption due to the pressure discrepancy, which can be characterized by a hysteresis factor, calculated using Hf = ln(Pa/Pd), where Pa and Pd stand for the hydrogen absorption and desorption plateau pressure, respectively. As can be seen in Table 2, with the increasing Al content, Hf becomes larger, implying more energy loss happened during at least one of the absorption and desorption processes. Energy loss results in the increase of the irreversibility between hydrogen absorption and desorption. That is, the elevating portion of the hydrogen in the alloys cannot be released with the increasing Al content, which is consistent with the results reported in the literature [27]. This phenomenon is disadvantageous to the practice application. Moreover, Hf can also be used to predict the pulverization rate of the alloy during charge/discharge cycling [34]. The cycling stability will be discussed latter.

Thermodynamic property

To further understand the hydride properties of composites, the thermodynamics parameters for the hydrogenation of these composites are determined from Van’t Hoff equation lnP = ΔH/RT − ΔS/T. From the slope and intercept of Van’t Hoff plot, as shown in Fig. 5, the enthalpy change ΔH is calculated to be − 24.82, − 26.47, − 31.30, − 34.76, and − 36.49 kJ/mol H2, and the entropy change ΔS is found to be − 87.35, − 94.23, − 113.70, − 128.02, and − 137.02 J/mol H2K for x = 0, 0.01, 0.02, 0.03, and 0.04, respectively. These values are similar or comparable to the commercial AB5 and AB2 MH alloys [35, 36]. ΔH is very important for evaluating the utilization of hydrogen storage alloys. Metal hydride with larger ΔH absolute value needs higher temperature to desorb hydrogen, suggesting that the hydride is more stable [37]. Metal hydride with strong thermodynamic stability is believed to be detrimental to hydrogen desorption. In the present case, the absolute values of ΔH become markedly higher with increasing Al content, suggesting that the stability of the hydrides increases as Al is added. Hence, the dehydriding process becomes more difficult and the reversible hydrogen storage capacity decreases. This result is consistent with Hf mentioned above. The other parameter ΔS is used to describe the disorder degree of the hydride alloys; the larger − ΔS means that the hydride of the alloys is more obvious deviate to the perfect ordered structure, indicating that the filling of an ordered MH structure is not completed [36, 38]. Judging from the experimental data, − ΔS values of the alloys upgrade distinctly with the increase of Al content. We can deduce that the enhancing Al content causes an incompletion in the formation of a complete ordered metal hydride structure. Therefore, with increasing x, the disordered degree of the hydride alloys increases, which yields the lower hydrogen storage capacity.

Fig. 5
figure 5

Van’t Hoff plots for hydrogen desorption of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04)alloys

Full size image

Activation and cyclic stability

The activation performance and the maximum discharge capacities (Cmax) of the electrodes are measured at the condition of the current density of 60 mA g−1. Figure 6 illustrates the discharge capacities of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04) electrode alloys which vary with cycle numbers; Cmax is also listed in Table 2. It is found that all of the investigated alloys can easily activate and reach the maximum discharge capacity at the initial charge/discharge process, which is favorable for their commercial application. No noticeable differences can be observed in the activation performance of these electrodes, which is attributed to the same multi-phase composition of these alloys [39]. Figure 6 also shows that, with increasing Al content, Cmax of the alloys monotonously decreases from 328.7 mAh g−1 (x = 0) to 312.7 mAh g−1 (x = 0.04). Ni is well known as an indispensable element in hydrogen storage electrode alloys due to their high electrocatalytic [40]. The decline of Cmax should be partly attributed to the reduced Ni content. Moreover, compared to Ni, Al has a smaller electronegativity, which leads to a relative smaller repulsion to the electron accompanying an adsorbed hydrogen atom. Similar results were obtained for the substitution of Mn for Ni [32]. Furthermore, it can be clearly seen that the changing trend of the discharge capacity is good agreement with that of the hydrogen storage capacity with substitution Al for Ni, verifying the charge/discharge progress in electrochemical reaction corresponding to the absorption/desorption progress in gas–solid reactions.

Fig. 6
figure 6

Activation course the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) electrode alloys

Full size image

The cyclic stability of hydrogen storage alloy is a decisive factor for the Ni/MH batteries life, which directly relates to the practical applications. The discharge capacities of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04)alloy electrodes as a function of cycle number are plotted in Fig. 7. The relative results are also summarized in Table 2. It can be seen that, with Al content increasing, the capacity retention (S80) increases from 63.6% (x = 0) to 76.5% (x = 0.04) after 80 charge/discharge cycles, which indicates that the partial substitution of Al for Ni significantly improves the cycling stability of the alloy electrode. However, the present result is poor compared with the results reported by Pan et al. [27] partly because of the existence of Mn element in these alloys.

Fig. 7
figure 7

Cycle life curves of the Re 2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys

Full size image

The relationship between the discharge capacity, the cycle stability, and the Al content in the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04) electrode alloys is also plotted in Fig. 8. In the hydrogen storage alloys, there are two important factors responsible for the degradation of the discharge capacity of the alloy electrodes during charge/discharge cycles: [41, 42] the pulverization of the alloys particles and corrosion/oxidation of the active compositions. On one hand, the improvement of the stability of the alloy electrodes could ascribe to the higher pulverization resistance during cycling [43]. As discussed in P–C-isotherms, Hf raise with the increasing amount of Al substitution, the alloys with larger Hf possess higher rigidity, which lead to increase the anti-pulverization ability for the reduction of the capacity decline rate with cycling. On the other hand, it is well known that Al in alkaline solution is easily oxidized to Al oxide, which protects the alloy electrode from being further corroded/oxidized in KOH solution. Similar results [44, 45] have been reported that the addition of Al in La–Mg–Ni-based alloys can form a thicker oxide film, which effectively shields the alloy from being corroded on the surface contacting the electrolyte. The Al2O3 film can also prevent hydrogen diffusion from the surface to the bulk of alloys, so the kinetic properties of alloys were deteriorated. Therefore, the partial substitution of Al for Ni exhibits a more pronounced effect on the cycle life of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04) alloys.

Fig. 8
figure 8

Relationship between the capacity retention rate (S80) and Al content (x) of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys

Full size image

Discharge characteristic

The discharge voltage characteristic is an important performance of the electrode alloy, which is characterized by the potential plateau of the discharge curve of the alloy. The discharge potential plateau of alloy electrode corresponds to the oxidation process of desorbed hydrogen from metal hydride [46]. Thus, the better discharge characteristic of the alloys is determined by the longer and more horizontal discharge voltage plateau. Figure 9 shows the discharge curves of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloy electrodes at 60 mA g−1 and 303 K after being activated. The discharge potential plateau is between 0.84 and 0.92 V. Obviously, each composite alloy electrode has a longer and horizontal potential plateau, and the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloy electrodes exhibit almost the same discharge potential plateau, which can be attributed to the similar phase component. Moreover, during discharge process, the discharge plateau above 0.84 V maintains the discharge capacity over 90% of Cmax, and then decreases sharply down below 0.84 V, indicating that alloy electrode possesses a high and stable discharge potential. The excellent discharge plateau may be ascribed to the multi-phase phase in the alloy, which increases the phase boundaries to provide more tunnels for the diffusion of hydrogen [47]. Therefore, the diminution of the diffusion resistance of the alloys leads to the high discharge potential.

Fig. 9
figure 9

Discharge voltage curves of the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys

Full size image

Conclusions

The relationship of the structure, hydrogen storage and electrochemical performance for the Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9(x = 0‒0.04) alloys has been investigated. Partial substitution of Al for Ni in metal hydride alloy induced the change of structural parameter. The substitution of Al for Ni of Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys happened in LaNi5 phase and LaNi2 phase prior to (La, Mg)2Ni7 phase. With Al content increasing in the alloys, the hydrogen storage capacity decreased from 1.16 wt% (x = 0) to 0.99 wt% (x = 0.04) at 303 K. The increasing absolute value of enthalpy change ΔH from 24.82 to 36.49 kJ/mol and the entropy change ΔS from 87.35 to 137.02 J/mol revealed that the thermodynamic stability and disordered degree of the hydrides alloys increased with the Al addition. The discharge capacity of the alloys decreased with the increase of Al content, while the cycle stability (S80) improved noticeably, increasing from 63.6% (x = 0) to 76.5% (x = 0.04). The Re2Mg(Ni0.7 − xCo0.2Mn0.1Alx)9 (x = 0‒0.04) alloys have higher discharge potential above 0.84 V, exhibiting high and stable discharge plateau.