C14 Laves Phase Metal Hydride Alloys for Ni/MH Batteries Applications

Abstract

C14 Laves phase alloys play a significant role in improving the performance of nickel/metal hydride batteries, which currently dominate the 1.2 V consumer-type rechargeable battery market and those for hybrid electric vehicles. In the current study, the properties of C14 Laves phase based metal hydride alloys are reviewed in relation to their electrochemical applications. Various preparation methods and failure mechanisms of the C14 Laves phase based metal hydride alloys, and the influence of all elements on the electrochemical performance, are discussed. The contributions of some commonly used constituting elements are compared to performance requirements. The importance of stoichiometry and its impact on electrochemical properties is also included. At the end, a discussion section addresses historical hurdles, previous trials, and future directions for implementing C14 Laves phase based metal hydride alloys in commercial nickel/metal hydride batteries.

1. Introduction

Nickel/metal hydride (Ni/MH) batteries have a wide range of applications including portable consumer electronics [1], transportation [2], and stationary power sources [3,4] as it features benefits such as safe chemistry, sustainable life, a stable raw material supply, high performance, and reasonable cost [5]. The most common metal hydride (MH) alloy used as the active material in negative electrode of Ni/MH batteries is a misch-metal (combination of La, Ce, Pr, and Nd) based AB₅ alloy with a CaCu₅ crystal structure. Recently, a misch-metal and Mg in the superlattice alloy (A₂B₇ type) with its corresponding improvement in the gravimetric energy density has become more popular in consumer and stationary applications [5,6,7,8,9]. Meanwhile, the Laves phase AB₂ MH alloy has also been proposed to improve the energy density of Ni/MH batteries. Although the AB₂ MH alloy has the potential for relatively high capacity potential (440 mAh·g⁻¹ [10,11] when compared to 330 mAh·g⁻¹ from the AB₅ alloy), it suffers from a relatively slow electrochemical reaction rate and a less-desirable cycle life (see comparison in

Table 1. The performance comparison of hydrogen-absorbing alloys with different chemistries [12]. The numbers in parentheses are the nominal B/A ratios. The symbols are ++ (superb), + (good), 0 (acceptable), and − (poor).

Properties	AB2 (2)	AB3 (3)	A2B7 (3.5)	A5B19 (3.8)	AB5 (5)
AB2 number of units	1	1	1	1	0
AB5 number of units	0	1	2	3	1
Electrochemical capacity/weight	++	0	+	0	−
Electrochemical capacity/volume	0	0	+	0	−
Pulverization of alloy, oxidation (corrosion)	+	+	+	−	−
Reversibility of hydrogen absorption/release	−	0	+	0~+	++
Battery Life	−	0	++	0	+

) [12]. A further comparison of the main battery performances between AB₂ and AB₅ MH alloys can be found in previous research (Table 1 in [13]). Numerous papers, including a few reviews [14,15,16,17,18,19,20], have been have examined how to improve the electrochemical properties of AB₂ MH alloys, however, an updated comprehensive review of this subject would be of use. The current work details the alloy preparation methods, performance requirements, failure mechanisms, constituent phases, selection of elements and stoichiometry, as well as the historical challenges and directions for future research.

C14, C15, and C36 Laves phases are composed of the same A₂B₄ unit-block, but with different stacking sequences; C14 has an a-b-a-b stacking sequence, which leads to a hexagonal structure (Figure 1a); C15 has an a-b-c-a-b-c stacking and a cubic structure (Figure 1b); and C36 exhibits a-b-a-c stacking and a rhombohedral (dihexagonal) structure (Figure 1c) [21,22,23]. The C14 and C15 phases are more commonly observed in MH alloys, whereas the C36 phase may exist between the C14 and C15 phases [24] but is difficult to identify using X-ray diffraction (XRD) [25,26]. Alloys with both C14 and C15 crystal structures serve as hydrogen-storage (H-storage) alloys and electrode materials. C15 alloys have better high-rate and low-temperature performance, however the C14 alloys were more widely used earlier when alloy composition design was driven by capacity and cycle stability. An AB₂ MH alloy with a C14-predominated structure was used in the first commercialized electric vehicle, EV1 by General Motor (Detroit, MI, USA). Similar alloys with the same structure were also used in the first-generation Ni/MH batteries made by Gold Peak Industries (Hong Kong, China) and Hitachi-Maxwell (Osaka, Japan). Recently, the electrochemical performances of a state-of-art C14-predominate MH alloy were compared to those from a recently proposed C15-based MH alloy together with a review of the development works done on the C15-based MH alloys [27]. In this paper, we will review only research focused on the C14-predominated MH alloys.

2. Alloys Preparation

Conventionally, when used as the active material in negative electrodes of Ni/MH batteries, the MH alloys are initially prepared by melting and casting. Then, an annealing process is applied to the ingot to improve compositional uniformity and to reduce internal stress. To reach a final particle size of less than 200 mesh (75 µm), most ingots need to be crushed into powder through mechanical grinding or hydrogenation. The powder may require additional surface treatments to enhance its electrochemical properties. Negative electrode fabrication methods have been previously reviewed [13].

2.1. Melting and Casting

A few preparation methods for MH alloys are described in an introductory article [29] that covers conventional vacuum induction melting (VIM) [30], arc melting (AM) [31], centrifugal casting (CC), melt-spinning (MS) [32,33,34], gas atomization (GA) [35,36], and mechanical alloying (MA) [37,38,39]. A vacuum plasma spray (PS) has also been employed for research purposes [40]. While AM only produces small quantity of alloy (5 to 200 g) for laboratory use, VIM can produce ingots ranging from 1 kg to 1 ton (Figure 2). In terms of electrochemical property, no distinct difference is observed between AB₂ MH alloys prepared by VIM and those prepared by AM [41], except for a slightly higher capacity obtained from VIM due to a larger sample/chamber ratio [42]. Samples from different locations of a VIM-process ingot may have different microstructures, but the electrochemical capacities are very similar. For example, the microstructures of different pieces of ingot prepared by a 60-kg VIM furnace are shown in Figure 3. According to the analytic results from x-ray diffraction (X'Pert Pro, Philips, Amsterdam, The Netherlands) and scanning electron microscope (SEM) (JEOL-JSM6320F, JEOL, Tokyo, Japan) with the x-ray energy dispersive spectroscopy capability, this alloy is C14-predominant with C15, TiNi, and families of (Zr, Ni) secondary phases. We obtained this composition by continuing optimization and testing more than 400 alloys. The properties of C14-predominated alloys prepared by VIM and other techniques were also previously compared [43] and are summarized in

Table 2. Comparison between alloy preparation methods. VIM, CC, MS, GA, MA, AM, and PS are short for vacuum induction melting, centrifugal casting, melt spinning, gas atomization, mechanical alloying, arc melting, and plasma spray, respectively.

Properties	VIM	CC	MS	GA	MA	AM	PS
Purpose	Production	Research/Production	Research/Production	Research/Production	Research	Research	Research
Batch Size	1–1000 kg	1–1000 kg	1–200 kg	1–1000 kg	1–1000 g	5–200 g	5–100 g
Equipment Cost	Medium	High	High	High	Low	Low	High
Production Cost	$3 kg−1	$4 kg−1	$4 kg−1	$5 kg−1	$1000 kg−1	$1000–$5000 kg−1	$2000 kg−1
Cooling Rate	100 °C s−1	1 × 103 °C s−1	1 × 106 °C s−1	1 × 104 °C s−1	>1 × 106 °C s−1	500 °C s−1	1 × 104 °C s−1
Micro-structure	Large crystallites	Medium crystallites	Nano-crystallites	Micro-crystallites	Amorphous	Large crystallites	Micro-crystallites
Alloy Discharge Capacity	Normal	Normal	Low	Low	High	Normal	Low
Alloy High-Rate Discharge-ability	Normal	Normal	High	Low	High	Normal	High
Alloy Cycle Stability (Anti-pulverization)	Normal	Better	Excellent	Excellent	Excellent	Normal	Excellent

. Because of the small sample size, use of expensive argon gas, and high utility costs from water chiller and arc power supply, AM is the most expensive method for making MH alloys. Photographs of equipment and different pouring processes are shown in Figure 4 and Figure 5, respectively. SEM surface morphology and cross-section micrographs of products are compared in Figure 6 and Figure 7, respectively. A review of Japanese Patent Applications shows 9 different ingot preparation methods for MH alloys [44]. For MH alloys that are sensitive to oxygen contamination from the reduction of ceramic refractory used as furnace crucible, a skull VIM melting can be used (Figure 8). MA in a shaker mill or an attritor is a popular method for making meta-stable/amorphous alloys not allowed in the phase diagram [45] (Figure 9).

2.2. Particle Size Reduction

Particle size influences capacity and high-rate dischargeability (HRD) behavior, but not activation properties [46]. As such, a proper particle size is required for battery operation. Unlike the conventional jar crusher used for AB₅ MH alloys, the high hardness of C14 MH alloys requires a hydrogenation process to create fissures and cracks for further grinding [47]. A stationary hydrogen reactor for laboratory use (2-kg load), a prototype rotational hydrogen reactor (50 kg), and a mass production stationary hydrogen reactor (5 tons) are shown in Figure 10. Operation details for the stationary and rotational hydrogen reactors can be found in the associated U.S. Patents [48,49]. Further reduction to a desirable size can be achieved by crushing [50], dry ball milling [51], wet ball milling [52], centrifugal grinding, or jet milling [53]. Three different powder fabrication methods—hydrogenation, dry ball milling, and wet ball milling—on a F-treated AB₂ MH alloy, were compared and the third showed the best electrochemical performance [54].

2.3. Annealing

The main purpose of annealing is to improve uniformity in both composition and micro-structure. Annealing is a necessary step to prepare commercial grade AB₅ MH alloys with a long cycle life [55] (Figure 11a,b). In the case of AB₂ MH alloys, non-Laves secondary phases improve the electrochemical properties of an alloy [56] through synergetic effects [57] and are diminished by annealing treatments [58,59,60,61]. Therefore, annealing is not required for AB₂ MH alloys, except for those prepared by GA and with a surface oxide layer, which can be reduced to the metallic state by annealing in hydrogen [62] (Figure 11c). Annealing in Ar was also improved the capacity of GA- [63] and MS- [64] produced AB₂ MH alloys because of the ability to increase the surface crystallinity. The positive contribution of thermal annealing to electrochemical capacity was previously reported by Klein et al. [65].

2.4. Surface Treatment

Various surface treatments are available for AB₂ MH alloys. A fluorination process can improve the surface anti-corrosion capability [66,67,68,69,70,71] and HRD [72], while a hot alkaline bath helps remove the native oxide that forms during the processes and creates a less dense oxide with catalytic Ni, which further increases the surface’s catalytic ability [51,66,73,74] and HRD [72]. Other methods include hot-charging performed at 80 °C [75], surface treatment in NH₄F and NiCl₂ solution [76], NiO coating by the sol-gel method [77], electroless Cu-coating [78], ball-milling with [79,80] or without Ni powder [81,82], ball-milling with La-containing AB₅ or A₂B₇ MH alloys [83], and surface reduction by KBH₄ and NaBH₄ [84].

3. Performance Criteria

The general requirements for an MH alloy that is suitable for use as the negative electrode in Ni/MH batteries include high electronic conductivity, a durable anti-corrosion surface, high H-storage capability, fast bulk hydrogen diffusibility, adequate metal-hydrogen (M-H) bond strength, acceptable cost, small lattice expansion during hydrogenation, and environmental friendliness [85,86]. As the applications of Ni/MH batteries become more versatile, the demands of MH alloys also vary. We classify the prerequisites of specific alloy properties in accordance with application in

Table 3. Performance requirements of MH alloy in various type of applications. The symbols are ++ (highly desirable), + (important), 0 (not critical), and × (unrelated). EV and HEV represent electric vehicle and hybrid electric vehicle, respectively.

Alloy Requirements	High-Energy (EV)	High-Power (HEV)	Stationary General Purpose	Stationary at High Temperature	Stationary at Low Temperature
H-storage Capacity	++	0	+	+	+
H-diffusibility	+	++	+	+	++
Surface Catalysis	+	++	+	++	++
Anti-corrosion	+	0	+	++	+
Equilibrium Pressure	0	++	0	×	++
Pulverization	++	+	0	+	+
Cost	++	+	++	+	+

. For the AB₂ MH alloys, the surface catalytic ability is not as good as those in AB₅ and A₂B₇ MH alloys, because of its lower B/A ratio, which contributes to a lower level of metallic Ni on the surface [87]. But the corrosion product of AB₂ alloy is dissoluble in electrolyte, which is different from the highly packed oxide of rare earth elements (RE) in AB₅ alloys, and makes it a better candidate for high-temperature applications. The small hysteresis found in the pressure-concentration-temperature (PCT) analysis between absorption and desorption isotherm in AB₂ reduces the pulverization rate and helps improve cycle stability [88,89].

4. Failure Mechanism

While the failure mechanism of the AB₅ MH alloy has been studied extensively and reviewed in detail [90], similar studies on AB₂ MH alloys are rare [91,92]. Unlike the surface of AB₅ MH alloys, which are covered by an inert layer of La(OH)₃ after numerous cycles, the AB₂ MH alloy forms soluble complex ions, including HZrO₃⁻, HMnO₂⁻, and AlO₂^−., which migrate through the separator into the positive electrode, causing micro-shorts and particle pulverization at the electrodes, which leads to positive electrode failure. The surface passivation of AB₅ MH alloys causes stable capacity degradation, while the failure at the positive electrode with AB₂ MH alloys creates a pressure increase, cell venting, and a sudden capacity drop near the end of the cycle life (Figure 12). Three AB₂ MH alloy degradation modes (oxidation, pulverization, and amorphization) are discussed in this section.

4.1. Oxidation

The activated (alkaline bath treated or cycled) AB₂ alloy surface consists of a top non-electrochemical reactive ZrO₂ patch, a surface oxide (~200 nm) with embedded metallic Ni/Co-clusters [10,93], and an amorphous buffer layer (~100 nm) underneath [94]. In one example, the Zr/Ti ratios are 8.2, 1.2, 0.8, and 1.9 for the top oxide, supporting oxide, buffer oxide, and bulk, respectively, which indicates a higher leaching rate for Zr compared to that of Ti [94]. The metallic inclusions on the surface of the AB₂ alloy were thought to be vital for the surface electrochemical reaction until it was determined that electrolyte-conducting channels coated with catalysts on the surface of the inner side of the alloy bulk can significantly improve low-temperature performance [95]. The leaching rates of constituent elements from a typical C14 MH alloy are calculated in

Table 4. Calculation of leaching rate in 100 °C/30% KOH of a C14 MH alloy with a composition of Ti₁₂Zr_21.5V_9.5Cr_4.5Mn_13.6Co₂Ni_36.2Sn_0.3Al_0.4. Data are from [87].

Element	Ti	Zr	V	Al	Mn
Amount in alloy (at %)	12.0	21.5	9.5	0.4	13.6
Concentration in solution after 1 h etching (ppm)	0.6	15.6	11.4	7.8	0.4
Concentration in solution after 4 h etching (ppm)	1.0	48.9	33.8	28.1	0.5
Leaching rate after 1 h etching (ppm/at %)	0.05	0.73	1.2	19.5	0.03
Leaching rate after 4 h etching (ppm/at %)	0.08	2.27	3.6	70.2	0.04

and follow the trend of Al > V > Zr > Ti > Mn. Oxidation and pulverization are the major capacity degradation mechanisms for AB₂ MH alloys [96,97].

4.2. Pulverization

Particle pulverization is a key failure mode for AB₂ MH alloys, because it increases the surface area of the electrode, which leads to more severe oxidation [98]. High internal stress exists in AB₂ MH alloy after hydrogenation, and cannot be removed by annealing because of the beneficial multi-phase structure. Thus, the initial pulverization due to the internal stress in AB₂ MH alloys is more severe than that in AB₅ MH alloys. But, because of the synergetic effect between the main storage phase and the catalytic secondary phase, the PCT hysteresis (the main reason for pulverization during hydride/dihydride cycling) in AB₂ is much smaller than that in AB₅ and, thus, a small degree of pulverization during cycling is expected from AB₂. Two SEM micrographs showing the pulverizations after activation (10 cycles) and near the end of cycle life (450 cycles) are presented in Figure 13. Interestingly, a V-free AB₂ MH alloy also shows a higher pulverization rate than a V-containing alloy [92].

4.3. Amorphization

Hydrogen-induced-amorphization (HIA) frequently occurs on AB₂ alloys [99,100], especially those with an atomic radius ratio between A- and B-site atoms (R_A/R_B) > 1.37 [101]. The amorphous alloy may have a lower electrochemical capacity [101,102] but better HRD [103,104] compared to its crystalline counterpart. An amorphous C14 MH alloy went through MA and its capacity decreased from 325 to 25 mAh·g⁻¹ [105]. A similar degradation of capacity caused by amorphization from MS has been previously reported [64].

5. Non-Laves Secondary Phases

During cooling from the liquid, the C14 phase solidifies first, followed by the C15 phase, and then finally a TiNi phase. The TiNi phase further transforms into Zr₇Ni₁₀, Zr₉Ni₁₁, and ZrNi phases via solid-state reactions [106,107,108,109] (Figure 14). When other modifiers are present, additional secondary phases are also formed. These secondary phases may not store large amounts of hydrogen, but they may improve electrochemical performance through the synergetic effect [57].

5.1. TiNi

A TiNi phase is a commonly observed secondary phase in C14 MH alloys. TiNi has a large soluble Zr content [110]. Occasionally, the Zr-content in the TiNi phase can be larger than the Ti-content [111]. A TiNi phase was found to increase the discharge capacity, and to improve the charge retention and cycle life while lowering HRD capability [56]. In addition, it is known to improve low-temperature kinetics [112,113]. The electrochemical performances of TiNi-based MH alloys were previously reported [114,115,116,117].

5.2. Zr_xNi_y

Three Zr-dominated secondary phases are commonly seen in C14 MH alloys, specifically ZrNi [30], Zr₉Ni₁₁ [65,118], and Zr₇Ni₁₀ [118]. ZrNi and Zr₇Ni₁₀ phases are found to increase HRD, but sacrifice capacity and cycle stability, while Zr₉Ni₁₁ does the opposite [56,119]. Zr₇Ni₁₀ gradually shifted to Zr₉Ni₁₁ as secondary phases in C14 MH alloys with increasing V-content [120]. The electrochemical properties of ZrNi [30,121,122,123], Zr₉Ni₁₁ [121,124,125], Zr₇Ni₁₀ [30,121,124,125,126,127], and other Zr₈Ni₂₁ [30,121,124,125,128], Zr₂Ni₇ [30,121,129], and ZrNi₅ [130] based MH alloys are also available.

5.3. V-Based bcc Solid Solution

In the MH alloys with a high V-content, a V-based body-centered-cubic (bcc) is commonly observed [131,132,133]. This phase is unlikely to corrode and remains at the surface after activation [72]. As the V-content and the B/A ratio increases, the microstructure of the alloy will change from a predominated Laves phase into a Laves-phase-related bcc two-phase solid solution [134]. The electrochemical storage of the two-phase alloy is higher, but the HRD suffers in the Laves-phase alloy [135].

5.4. ZrO₂

ZrO₂ is formed during the melting stage and acts as an oxygen scavenger [136], which is ordinarily seen in C14 MH alloys with higher Zr/Ti ratios [111]. It forms a protective barrier against oxidation [72], but does not form hydride under normal operating conditions in batteries [137].

5.5. Other Secondary Phases

Other secondary phases can be detected using non-transition metal modifiers, such as RE. The solubility of RE in the C14 phase is extremely low and the addition of Y, La, Ce, Nd results in the formation of YNi [138], LaNi [139], CeNi [140], and NdNi [141] phases, respectively. These RE-Ni phases are detrimental to the electrochemical performance of MH alloys [112]. Like RE-Ni, the ScNi secondary phase increased discharge capacity, but weakened cycle stability [142]. A combination of hypo-stoichiometry and the addition of RE can boost low-temperature alloy performance [113]. High Mg-content secondary phases have been detected and facilitate the activation process [143,144]. A ZrC phase was formed with a C-addition and improves charge retention but deteriorates cycle life, HRD, and low-temperature performance. Zr₂Ni₂Sn [56,136] and Zr (Ni, Mn) Sn_0.35 [145] phases can be found in alloys with a relatively high Sn-content and are detrimental to the alloys’ electrochemical performance. The structure and properties of Zr₂Ni₂Sn have been reported [146].

6. Selections of Element

AB₂ alloys have a greater compatibility of constituent elements, compared to that of AB₅ MH alloys, because of their superb solubility of components and great number of available phases. The compatibility of substituting elements makes AB₂ a better candidate to meet various application demands. Laves phase alloys are composed of A-site and B-site atoms (Figure 1). The A-site atom is usually larger than the B-site atom, but tends to shrink as the electronegativity of the B-site atom increases, resulting in electron transfer. The ideal R_A/R_B is 1.225 [147].

6.1. A-Site Element

For electrochemical applications, Ni, with an atomic radius of 1.377 Å in the Laves phase, is the major B-site element [148], which limits the choice of A-site element to those with an atomic radius of approximately 1.68 Å. Some physical properties of the A-site atoms are summarized in

Table 5. Properties of A-site elements in AB₂ alloys. IMC denotes intermetallic compound.

Properties	Ti	Zr	Hf	Nb	Pd
Atomic Number	22	40	72	41	46
Atomic Radius in AB2	1.614	1.771	1.743	1.625	1.521
Electronegativity	1.54	1.33	1.30	1.60	2.20
Earth Crust Abundance (%)	0.66	0.013	3.3 × 10−4	1.7 × 10−3	6 × 10−7
Melting Temperature (°C)	1668	1855	2150	2468	1555
ΔHh (kJ·mol H2−1)	−136	−164	−161	−83	−41
Number of IMCs with Ni	3	8	8	3	0

. Ti and Zr are the most frequently used A-site elements in Laves phase MH alloys because of their availability. Zr has the strongest M-H bond (lowest heat of hydride formation [149]) and the adjustment of Zr/Ti content is the most effective method to change M-H bond strength.

6.1.1. Titanium

Ti results in weaker M-H bond strength, higher PCT equilibrium pressure [65], and improved HRD performance [41,150,151]. Ti forms an inert layer of TiO₂ and impedes activation [31,152], which improves cycle stability [153] and causes capacity degradation [154]. However, another group reported that Ti is beneficial for improving activation [155]. Ti-content impacts the discharge capacity in two competing ways: weaker M-H bonds lead to a lower capacity, but higher plateau pressure also improves the reversibility and increases the discharge capacity. Previous research has documented that increasing Ti-content has led to increasing [156,157,158,159], decreasing [155,158,160], and unchanged [150] electrochemical discharge capacities. A high Ti-content also contributes to a higher abundance of bcc phase [161] and a higher C14 abundance [162] because of its higher chemical potential [163].

6.1.2. Zirconium

To maintain stoichiometry in comparison to AB₂, the sum of Ti- and Zr-contents is approximately 33 at %. Therefore, Zr has the opposite effects of Ti. A higher Zr-content will stabilize the hydride [163] and lower HRD [150]. The bulk oxide of Zr is usually formed during solidification and has a higher solubility in KOH solution than TiO₂ [87]. A dense layer of ZrO₂ formed on the surface of the Zr particles during powder processing (grinding, sifting, and packaging), and impedes the activation process [152,164,165].

6.1.3. Hafnium

Not many studies have examined the electrochemical properties of Hf- substituted AB₂ MH alloys [42,160]. An XRD study confirms that Hf and Ti share the A-site and V and Ni are found in the B-site [166]. Hf-substitution yields a slightly smaller unit cell [167], which is unexpected because of its relatively large atomic radius (Table 5). The reduced lattice constant c/a ratio with the addition of Hf [167] predicts a lower pulverization rate. Partially substituting Hf for Zr results in a lower electrochemical capacity, but a higher HRD [160].

6.1.4. Niobium

Nb occupying the A-site was confirmed by a Rietveld analysis with XRD data from a substituted ZrCr₂ alloy [168]. Partial replacement of Zr by Nb leads to a smaller C14 unit cell volume and, consequently, a higher PCT plateau pressure [169]. It increases the C14 phase abundance and reduces lattice constants [167]. Like Hf, Nb also contributes to a reduced lattice constant c/a ratio [167] leading to a smaller pulverization rate. The NiNb secondary phase is a poor catalyst/H-storage material and, remarkably, reduces the surface exchange current [169]. Like Hf, Nb-substitution deteriorates capacity [42], but enhances HRD [160].

6.1.5. Palladium

Pd is smaller than both Ti and Zr, and when it enters the A-site, it reduces the unit cell of the Laves phases, which results in a decrease in discharge capacity. However, it also enhances the surface catalytic ability and improves both HRD and low-temperature performance for the alloy [170]. Improvement in HRD of C14 alloys was also reported by Yang, Ovshinsky, and their coworkers [171,172].

6.1.6. Scandium

Use of Sc, another expensive additive, in AB₂ MH alloy is scarce. The gaseous phase H-storages of C15/C36 ScM₂ (M = Fe, Co, and Ni) were reported [173] with an electrochemical study of a C15 ScNiCo_0.2Mn_0.5Cr_0.2 alloy [174]. The partial replacement by Zr and Y in Sc-based C15 showed cycle stability improvement with the corresponding sacrifice in discharge capacity [142]. The addition of Sc increased the abundance of the ZrNi secondary phase, improved the activation and capacity, but resulted in a trade-off in HRD [175].

6.2. B-Site Element

Ni is the most widely used B-site element in alkaline battery applications. However, Ti, Zr, and their mixture do not form Laves phase intermetallic compound (IMC) with Ni [176]. Therefore, B-site substitutions, mainly from the first row of transition metals—because of their availability and light weight—are essential for the stability of Laves phase IMC. Both the electron affinity and the atomic radius decrease as the atomic number of the modifier atom increases, which reduces the average metal-hydrogen bond strength, increases plateau pressure, and decreases H-storage capacity. With different solubility and chemical properties related to corrosion, these modifier elements are crucial for engineering various electrochemical properties and will be reviewed in the following sections. Various properties of the first-row transition metals are summarized in

Table 6. Properties of first-row transition metal elements as B-site atoms in AB₂ alloys.

Properties	V	Cr	Mn	Fe	Co	Ni	Cu	Zn
Atomic Number	23	24	25	26	27	28	29	30
Atomic Radius in AB2	1.491	1.423	1.428	1.411	1.385	1.377	1.413	1.538
Electronegativity	1.63	1.66	1.55	1.83	1.88	1.91	1.90	1.65
Earth Crust Abundance (%)	0.019	0.014	0.11	6.3	0.003	0.009	0.007	0.008
Melting Temperature (°C)	1890	1857	1245	1535	1495	1453	1083	420
ΔHh (kJ·mol H2‒1)	−35	−8	−8	10	15	−3	20	8
Number of IMCs with Ti	0	1	1	2	4	3	5	7

. The atomic radius of the element decreases as the atomic number increases until Ni and then increases from Cu and Zn. The number of IMC with Ti also increases roughly with increasing atomic number.

6.2.1. Vanadium

V is the only B-site element that brings high H-storage capacity [17]. It results in a more stable hydride, more disorder, but also decreases peak power and charge retention [177]. The high leaching rate of V is the main cause of poor charge retention in conventional V-containing MH alloys [87,178]. The high electronic conductivity of V also improves the activation property of alloys [179]. A balance of V-content is needed to optimize battery performance [180]. V-free MH C14 MH alloys were designed to address the charge retention issue, but did so at the expense of cycle stability [144,178]. The large atomic radius of V enlarges the alloy unit cell and reduces the volume expansion during hydrogenation, which in turn reduces PCT hysteresis and pulverization tendency [181], which results in a shorter cycle life [42]. Reports of V lowering capacity are also available [42].

6.2.2. Chromium

The addition of Cr in C14 MH alloys improves cycle stability [11,131,182,183,184,185] and charge retention [11,184,186] but decreases the HRD [11,186,187], capacity [131], and activation tendency [11]. Cr can suppress segregation of Ti, Zr, and V on the surface of MH alloys [188], retarding the oxidation of V by forming a V-based bcc solid solution secondary phase [131,189], and slowing down pulverization [190]. Retarding HRD by adding Cr was further attributed to poor bulk hydrogen diffusibility and surface exchange current [133]. Compared to Co, Cr promotes a dendritic grain structure that is beneficial for cycle stability [118]. Its capability of reducing PCT hysteresis also contributes to a longer cycle life [11].

6.2.3. Manganese

Partial replacement of Ni by Mn increases capacity [17,186], facilitating the formation process [186], but deteriorating the cycle life of the battery because of its poor oxidation resistance [186,191] and results in micro short-circuits [192]. A high Mn-content in an alloy reduces the cycle life and charge retention because of poor oxidation resistance in the KOH electrolyte [11]. A low Mn-content provides improves activation, capacity, and HRD [193]. Therefore, a careful balance between the content of Mn- and other B-elements is necessary to optimize battery performance [11,42]. In a separate report, Mn was found to increase the discharge capacity and exchange current density while hindering HRD and cycle stability [194]. Mn facilitates homogenizing the chemical composition of different constituent phases in a multi-phase alloy system [11].

6.2.4. Iron

In V-containing C14 alloys, the addition of Fe facilitates activation, increases the discharge capacity and surface reaction area, decreases HRD [131] and hydrogen diffusibility, and impairs low-temperature performance [195]. However, reports also suggest that Fe contributes to a low capacity [42] and higher cycle life [196,197]. In a V-free C14 alloy, the addition of Fe improves low-temperature performance, but hinders cycle life and charge retention [144]. The solubility of Fe in Zr₇Ni₁₀ and Zr₉Ni₁₁ phases is lower than that in TiNi phases, which explains the promotion of TiNi phase by Fe-addition [195].

6.2.5. Cobalt

Co promotes an equiaxial grain structure that improves surface reaction kinetics, activation, and capacity, but deteriorates cycle stability [196]. An optimal Co-content in a C14 MH alloy of 1.5 at % provides the best performance in terms of formation, cycle life, and charge retention, but exhibits worse specific power and low-temperature performance compared to other compositions [111]. Compared to other modifiers, such as Fe, Cu, Mo, and Al, Co-substituted C14 MH alloys show a relatively higher discharge capacity [120,198,199], good cycle performance [120,193,199], and poor HRD [152]. Reports on Co-substituted C14 MH alloys with increased plateau pressure [200], a reduced capacity [200], and a lower cycle life [42] are also available.

6.2.6. Nickel

Ni is the most efficient catalytic B-site element because of the formation of metallic clusters embedded in the surface oxide after activation [87,201,202]. The introduction of Ni greatly improves electrochemical behavior [203], HRD [11], low-temperature performance [11,202], surface exchange current [204], and cycle stability [186], but impedes charge retention [11].

6.2.7. Copper

In a typical C14 MH alloy, the addition of Cu, with is high degree of pulverization, facilitates activation processes and improves discharge capacity and low-temperature performance, but decreases HRD because of lower hydrogen diffusibility [205]. Cu has also been reported to impair cycle stability [206] and lower capacity [42].

6.2.8. Zinc

Zn has an extremely low melting point compared to its neighbors’ in the periodic table (Table 6) and is unable to remain in the C14 alloy at a melting point above 1350 °C. Therefore, ZnNi [207] and ZnCu [208] IMCs were used as raw materials to reduce Zn loss from evaporation. 1 at % Zn results in an increase in discharge capacities, a slight decrease in HRD, an easier activation process, and a satisfactory low-temperature performance [207].

6.2.9. Second Row Transition Metals (Mo)

Mo is the only element in the second row of transition metals that is used as a modifier in the C14 MH alloys, as it is accessible and not very toxic. According to our recent study, partial substitution of Co by Mo in a C14 MH alloy increases both the bulk diffusibility of hydrogen and surface catalytic ability, which leads to improved HRD and low-temperature performance [209]. Mo also improves charge retention and cycle stability, but yields a slightly lower capacity [209]. The cycle stability improvement of C14 MH alloy using Mo was also reported [42]. However, reports on the negative impact of Mo on the capacity, cycle life, and activation of C14 MH alloy are also available, presumably due to lack of Cr or Co in the composition [197,199]. Partially replacing Cr with Mo increases the gaseous phase H-storage capacity [210,211], ease of activation, and HRD [212].

6.2.10. Third Row Transition Metals (W, Pt)

W and Pt are the only two elements in the third row of the transition metals that are used because of concerns with availability and weight. W improves corrosion resistance and self-discharge of C14 MH alloys [213]. The addition of Pt in a C14 ZrCrNi alloy with an annealing treatment improved HRD and capacity because of the high catalytic activity of Pt for hydrogen electrosorption processes [214].

6.2.11. Group 13 Elements (B, Al)

B exhibits excellent solubility in V-free C14 MH alloys and contributes to a lower capacity, a high HRD, and improved superior low-temperature performance [144]. In V-containing alloys, B increased cycle stability [215] and bulk diffusion [216], however it also results in decreased capacity and HRD [216]. Reports on the positive contributions of B to HRD and low-temperature performance in V-containing C14 alloys are also available [217]. Ball milling C14 alloys with B show improved cycle stability at the cost of a lowered capacity [218].

Al is the most frequently used non-transition metal in Laves phase MH alloys. Al use has been shown to strongly correlate with a high capacity and a good HRD performance [219]. Al additives, together with Co, positively contribute to activation, charge retention, HRD, and low-temperature performances [136,220]. In the gaseous phase, Al was also shown to reduce PCT hysteresis, which in turn retards pulverization [200]. This also contributes to a lower flammability, which enhances safety in powder handling [221]. Even with the highest leaching out rate observed in Table 4, Al has been reported to improve the corrosion resistance of alloys, leading to better charge retention performance [164,213,222].

6.2.12. Group 14 Elements (C, Si, Ge, Sn)

In V-free C14 alloys, the addition of C promotes the formation of a ZrC secondary phase and improves charge retention, but also deteriorates low-temperature performance, cycle life, and HRD [144]. In a Ti-only C14 alloy, adding C promoted the formation of a TiC secondary phase, which facilitated the activation and decreased the gaseous phase H-storage capacity [223]. In one Zr-based C14 MH alloy, C improved cycle stability [17].

The addition of Si to C14 MH alloys improved the self-discharge [17], high-temperature discharge performances [17,224,225], low-temperature performance [226], and cycle stability [17,226] with trade-offs for lower capacity caused by an inert layer of SiO₂ on the surface of the alloy [224,225]. Si-incorporated C15 MH alloys’ improved low-temperature performance is attributed to a layer of Ni₂O₃ that formed on the metal surface and can be detected by transmission electron microscopy (TEM, CM200/FEG, Philips, Amsterdam, The Netherlands) [227]. Si also contributes to a lower flammability that enhances safety in powder handling [221]. Both Si and Ge additives to ZrCr₂ MH alloy reduce gaseous phase H-storage capacity [228].

Zr is widely used in nuclear power plants because of its stability from a low neutron scattering cross-section. The costly separation process from Hf (neutron absorber) makes pure Zr metal expensive [229]. Future cost reduction of Zr is expected as the demand of a Hf-containing Zr for battery applications (Hf impurities do not matter since Hf is also a hydride former) increases. Currently, a large amount of Zircaloy (a Sn-Zr alloy [230]) scrap is available from demolished nuclear power plants and the addition of Sn is essential for cost reduction in raw materials. According to our own study, a small amount of Sn (0.1 at %) can improve activation, HRD, charge retention, and low-temperature performances at the expense of capacity and cycle life loss [219,231]. Sn also promotes the formation of detrimental secondary phases. Thus, the content of Sn should be limited [145].

6.2.13. Group 16 Elements (O, S, Se)

The only report on the addition of O, S, and Se to C14 MH alloys compares gaseous phase characteristics and shows the same order of effectiveness (S > Se > O) in raising plateau pressure and PCT hysteresis, increasing storage capacity, and reducing the PCT slope factor [232].

6.3. Secondary Phase Promoter

The addition of elements with large atomic radii, but extremely low solubility in Ti-Zr based AB₂ MH alloys, in the Laves phase will form secondary phases either in an element form or a Ni-containing alloy form. These secondary phases may play an important role in the electrochemical performance of the host alloy.

6.3.1. Group 1 Elements (Li, K)

Li has an extremely low melting point and very high vapor pressure at liquid temperatures for C14 MH alloys, therefore adding Li at the last stage and in the mold results in failure. However, Li-incorporated AB₅ MH alloys with their lower melting temperature, were successfully prepared by AM [233], VIM [234], and a diffusion method [235]. Li did promote unidentified secondary phases, based on XRD analysis [235]. KB was added in a Ti₂Ni based MH alloy as a source of K and showed improved cycle stability [215].

6.3.2. Group 2 Elements (Mg)

Solubility of Mg in the Laves phase is extremely low (0.l–0.3 at % [143]). In V-containing C14 alloys, it forms a cubic secondary phase, which facilitated the formation process but did not improve other electrochemical properties [143]. In V-free C14 alloys, the addition of Mg formed a Mg₂Ni secondary phase and improved low-temperature performance and charge retention, but hampered cycle life and HRD [144]. Ball milling with Mg₂Ni can also introduce Mg into AB₂ MH alloys [236]. In recent years, Mg and La were introduced to AB₂ MH alloys by either remelting [237] or MA [238] with an AB₂/superlattice alloy mixture where superlattice alloys contain both La and Mg.

6.3.3. Rare Earth (RE) Metals

A great number of studies have focused on how to solve the slow activation of RE-incorporated C14MH alloys [67,165,184,239,240,241,242,243]. Low solubility of RE in Zr, Ti-based C14 IMC promotes a RE-Ni (AB) secondary phase and facilitates activation processes [244]. Properties of commonly used REs as modifiers in the AB₂ MH alloy are summarized in

Table 7. Properties of some REs used as additives in AB₂ MH alloys.

Properties	Y	La	Ce	Pr	Nd	Sm	Gd	Yb
Atomic Number	39	57	58	59	60	61	55	70
Price (US$/kg) [249]	35	7	7	85	60	7	55	95
Ionic Radius in Laves (Å) [148]	1.990	3.335	2.017	2.013	2.013	1.990	1.992	1.990
Electronegativity	1.22	1.10	1.12	1.13	1.14	1.17	1.20	1.24
Melting Temperature (°C)	1522	918	798	931	1021	1072	1313	819
Oxidation potential (V)	−2.372	−2.379	−2.335	−2.353	−2.323	−2.304	−2.279	−2.19
Heat of Hydride Formation (kJ·mol H2−1) [250]	−114	−97	−103	−104	−106	−100	−98	−91

. La, Ce, and Sm are the least expensive. In general, the ionic radius, ease of oxidation, and MH-bond strength become smaller when the atomic number increases. Previous studies in a comparison series of RE-doped C14 MH alloys [112,113,138,139,140,141] showed that the RE-Ni phase is beneficial for activation, but detrimental to HRD and low-temperature performance, and that a careful stoichiometric design is needed to balance the abundance of the RE-Ni and TiNi phases [113]. The incorporation of La provides the greatest low-temperature performance, but a contradictory microstructure (an increase in the detrimental LaNi phase and corresponding decrease in the beneficial TiNi phase). Further TEM studies indicated that the Ni-Cr alloy channeled and catalyzed by a single-crystal sheet contributes to improvements in the surface catalytic ability of La-incorporated alloys [95]. Gd remains as a metallic inclusion when melted with C14 MH alloys and improves low-temperature performance, but deteriorates HRD, cycle stability, and charge retention [144]. Instead of using RE, which are nearly insoluble in Laves phase alloys, MA was used and achieved similar effects of capacity and activation [245]. In addition to adding RE in elemental form, RE-containing AB₅ [246,247] and A₂B₇ [248] can be used for similar results.

6.4. Summary of Modifier Studies

Unlike AB₅, AB₂ MH alloys are compatible with a remarkable number of elements. The degree of impact for commonly used elements on MH alloy properties are summarized in

Table 8. Influence of commonly used modifiers to C14 alloy performance. The symbols are ++ (highly beneficial), + (beneficial), 0 (no significant effect), and − (detrimental). Degree of pulverization is the combination of actual measurement [88] and estimation from PCT hysteresis and C14 phase lattice constant ratio a/c. Equilibrium pressure changes are indicated with up or down arrows.

Alloy Requirements	Ti	Zr	V	Cr	Mn	Fe	Co	Ni	Cu	Zn	Al	Si	La
H-storage Capacity	−	++	+	0	0/+	0	+	0	0	0	+	0	0
H-diffusibility	+	−	−	0	0	−	0	0	−	+	+	+	+
Surface Catalysis	−	+	+	−	−	−	0	++	−	+	+	+	++
Anti-corrosion	+	−	−	++	−	0	0	++	0	0	0	0	0
Equilibrium Pressure	↑↑	↓↓↓	↓↓	↓	↓	0	↑	↑	0	0	↑	0	0
Anti-pulverization	+	−	0	−	0	0	0	0	−	−	+	−	0
Cost	0	0	−	+	++	++	−	0	++	++	++	++	+

. Proper selection of elements can be based on the alloy requirements after checking each of the seven rows. Meanwhile, the pros and cons of each element can be easily evaluated in the corresponding column.

7. Stoichiometry

An IMC with a fixed stoichiometry comes with a large negative heat formation [45]. The different energy levels of anti-site and vacancy defects in a specific IMC may change its stoichiometry range [251,252,253]. If the energy for such defects is relatively low, a wide range of compositions (off-stoichiometry) are expected, among which AB₂ Laves phase alloys make up the largest portion in IMC [21]. Compositions of a few IMCs based on Group 4 elements (Ti, Zr, and Hf) as A-site atoms are listed in

Table 9. Solubility (range of A-site atom concentration in at %) of Laves phase IMCs (radius from [148] and solubility from [257]).

IMC	Structure	RA/RB	Solubility (at%)
ZrV2	C15	1.19	33.3
HfV2	C15	1.17	33.5–34.5
TiCr2	C15	1.13	35–37
ZrCr2	C15	1.24	31–36
HfCr2	C15	1.22	33–35
TiMn2	C14	1.13	30–40
ZrMn2	C14	1.24	20.8–40
HfMn2	C14	1.22	25.5–38
TiMn2	C14	1.14	27.5–35.5
ZrMn2	C15	1.26	27.1–34
HfMn2	C15	1.24	32–33.5
TiCo2	C15	1.17	33–33.5
ZrCo2	C15	1.28	27–35
HfCo2	C15	1.26	27–36

and plotted in Figure 15 as a function of R_A/R_B. IMCs with a C14 structure have a wider composition range (more likely to suffer from point defects) and the phase solubility of IMCs with a C15 structure becomes larger as the radii ratio increases. An AB₂ alloy with a B/A ratio below 2.0 is a hypo-stoichiometric IMC. The reason that more hypo-stoichiometric IMCs tend to form as the radii ratio increases is that a relatively small B-atom more rapidly forms a double B in A-site (dumbbell model [254]) and further increases the B/A ratio. The characteristics of a few off-stoichiometric C14 MH alloys were previously compared [45,255,256] and important results are summarized in the next three sections.

7.1. Stoichiometric Alloy

The stoichiometric C14 MH alloys with a B/A ratio close to 2.0 have the highest electrochemical capacity and surface exchange current, and the smallest PCT hysteresis (least pulverization) [256] and is the most studied alloy stoichiometry in the field of battery applications.

7.2. Hypo-Stoichiometry

A C14 alloy with a B/A ratio below 2.0 is a hypo-stoichiometric IMC. The relatively low B/A ratio promotes the occurrence of AB phase (typically TiNi with a B2 structure) and improves activation behavior [256]. In a MS + annealing sample, the additional Ti (hypo-stoichiometry) increases both the capacity and cycle life of the electrode [64]. In a gaseous phase study of a TiCr based C14 alloy, a hypo-stoichiometry ratio (B/A of 1.8) gave the highest H-storage capacity [211].

7.3. Hyper-Stoichiometry

A hyper-stoichiometric C14 alloy has a B/A ratio higher than 2.0 and shows a high HRD [151,256], a high open-circuit voltage (from a higher PCT hydrogen equilibrium pressure) [256], and a flat PCT plateau [256]. The hyper-stoichiometric alloy also shows an improved cycle life, due to increased mechanical stability during cycling [185,258]. As x increases in a series of hyper-stoichiometric alloys AB_x (x = 2–6), the bcc phase abundance increases, and results in improvements in capacity, kinetics [259], and cycle stability [260].

8. Discussions

The Ovonic Battery Company (Troy, MI, USA) focused on research about AB₂ MH alloys for Ni/MH battery applications for more than three decades before being acquired by BASF (Ludwigshafen, Germany) in 2012. Throughout the years, we witnessed changing market demands and the corresponding reaction from the research communities. Before the successful debut of the commercialized Ni/MH battery in 1988, two main hurdles of the AB₂ MH alloy were slow activation and poor cycle stability. While the former was addressed by fluorination and hot alkaline baths, the latter was improved substantially by composition modifications, mainly the introduction of Cr and consequent bcc secondary phase. The challenge to the first generation AB₂-made Ni/MH battery was the self-discharge and was solved by the combination of V-free alloys and the use of the sulfonated separator. At that stage, AB₂ MH alloys were successfully used in products made by Hitachi-Maxell (Tokyo, Japan), Gold-peak (Hong Kong, China), and General Motors (Detroit, MI, USA). With the cheap misch-metal (main raw material for the rival AB₅ MH alloy) available from China, AB₂ MH alloys started to phase out from the consumer market at the turn of the century. Later, the upset of RE prices in 2010 created opportunities for AB₂ MH alloys. However, the bar to enter the market has not been raised in favor of AB₂, mostly in the high-rate dischargeability requirement for HEV and power tools. The superlattice alloy used by FDK (Tokyo, Japan), Panasonic (Osaka, Japan), and Kawasaki Heavy Industry (Tokyo, Japan) became a strong competitor with a capacity between AB₂ and AB₅, but better high-rate performance. Mg, an indispensable element in the superlattice alloy, makes it more vulnerable to KOH electrolyte attack, but was mitigated using surfactant additives in the negative electrode paste [44]. The structure of the superlattice allows the removal of Mn and Co from the composition and improves the charge-retention characteristics tremendously [261]. At the current stage, AB₂ has a hard time competing with AB₅ in term of price and superlattice alloys in terms of performance.

However, new opportunities started to appear when the conventional Ni/MH battery based on the RE-based AB₅ or A₂B₇ alloy lost the battle against the rival Li-ion technology. New applications require greater performance improvement in certain areas. For example, a solid-state battery with a thin solid separator [262] and ultra-high-power application with a very thin separator require a spherical MH alloy shape, which can be produced using GA techniques. Superlattice alloy with Mg inside cannot tolerate the Mg-vaporization because of the large surface area of the powder produced in the GA process, whereas GA with an associated annealing process already been successfully developed for AB₂ MH alloy [62]. The battery/fuel cell combination requires a MH alloy operated at intermediate temperature range (200–250 °C) [263]. The other opportunity is in stationary applications that require tolerance in the environment of above 50 °C [3,4,264]. While the RE-containing AB₅ and A₂B₇ MH alloys form an impeccable surface oxide in the high-temperature environment, the transition metal-based AB₂ MH alloys leach out at a higher speed and are totally controllable by the composition re-adjustment. In the case of the new start-stop automobile application, a low-temperature cranking power is desperately needed [265]. There are a few elements contributing to the low-temperature performance of AB₂ MH alloys through different mechanisms, for example, Y for surface area increase, La, Pd and Mo for surface catalytic ability improvement, and Fe for the increase of beneficial TiNi secondary phase. The pouch cell design (like commercially available Li-ion battery for cell phones) of Ni/MH battery operating under a flooded electrolyte configuration requires new MH alloys with a very low plateau pressure and a highly electrochemically catalytic surface [266,267]. The last example is the need for new high-capacity MH alloys that fully utilize the wide voltage window of the newly developed ionic liquid electrolyte [268]. The new chemical environment (electrolyte interface) and lifting of voltage constraint (from competition of hydrogen evolution) present new challenges (i.e., opportunities) for the development of new MH alloys. Although the function of each modifying element in specific areas has been studied and reported in the current work, combinations and the accompanying synergetic effects among these elements have not been reported. Therefore, a review like the current work only provides a guideline for future research. Readers are encouraged to use the results of this study to pursue further performance improvements in C14 Laves phase based MH alloys.

9. Conclusions

The main advantages of C14 Laves phase based metal hydride alloys are higher capacities and flexibilities in composition, stoichiometry, and constituent phases, which allow fine tailoring in the electrochemical performance to meet the demands of different applications. Decades of research in this area have already solved historical hurdles in C14-based alloys, such as slow activation, low cycle life, high self-discharge, and poor low-temperature performance. To face the new challenges from future applications, we have compiled comprehensive comparisons of preparation methods, alloy property requirements, the pros and cons of each constituting/modifying element, and choice of stoichiometry in the current study. Based on these previous works, new directions for additional improvement in the electrochemical performance should focus on the combination of modifier elements with different functions.