Metallic Materials for Hydrogen Storage—A Brief Overview

Abstract

The research and development of materials suitable for hydrogen storage has received a great deal of attention worldwide. Due to the safety risks involved in the conventional storage of hydrogen in its gaseous or liquid phase in containers and tanks, development has focused on solid-phase hydrogen storage, including metals. Light metal alloys and high-entropy alloys, which have a high potential for hydrogen absorption/desorption at near-standard ambient conditions, are receiving interest. For the development of these alloys, due to the complexity of their compositions, a computational approach using CALPHAD (Calculation of Phases Diagrams) and machine learning (ML) methods that exploit thermodynamic databases of already-known and experimentally verified systems are being increasingly applied. In order to increase the absorption capacity or to decrease the desorption temperature and to stabilize the phase composition, specific material preparation methods (HEBM—high-energy milling, HPT—high-pressure torsion) referred to as activation must be applied for some alloys.

1. Introduction

Nowadays, due to the global energy crisis and environmental pollution, it is necessary to develop an alternative energy carrier that is environmentally friendly and sustainable [1]. Three types of clean energy that have the advantages of inexhaustibility, non-toxicity and renewable nature—e.g., solar and thermal energy, nuclear fusion energy and hydrogen energy—have come into consideration as potential replacements for conventional fossil fuels [2].

However, due to the intrinsic characteristics of intermittency and randomness, solar or wind power itself can hardly provide steady energy supply and match electricity demand well [3]. Under such circumstances, conversion of the surplus electricity into clean gaseous fuel, e.g., H₂, is promising [4]. Being a well-known clean energy carrier, hydrogen features high energy density, a long storage period and zero carbon emissions, and could play an important role in the future energy mix [4,5]. As a widely adopted criterion for evaluation, DOE (U.S. Department of Energy) targets for gravimetric density and volumetric density of onboard hydrogen storage are set to be 1.8 kWh/kg system (5.5 wt% hydrogen) and 1.3 kWh/L system (0.040 kg hydrogen/L) for 2025, and 2.2 kWh/kg system (6.5 wt% hydrogen) and 1.7 kWh/L system (0.050 kg hydrogen/L) ultimately [4,6].

Hydrogen energy has also attracted special attention thanks to its higher gravimetric energy density compared to any traditional fuel [1]. Commonly, hydrogen can be used in gaseous or liquid phase, which is associated with considerable technical and safety risks. Metal or intermetallic hydrides are considered promising materials in the research and development of cheap and lightweight solid-state hydrogen storage systems with fast kinetics and high capacity. Other candidates include carbon-based nanomaterials, metal-doped carbon-based nanomaterials, MOFs (metal–organic frameworks), covalent organic frameworks, complex chemical hydrides, clathrates, amides and zeolites [7,8,9]. Under various conditions, these nanomaterials can facilitate the storage of hydrogen in the solid phase through chemisorption (binding energy 50 to 100 kJ/mol, for hydrides) or physisorption (binding energy > 10 kJ/mol, for porous materials) (Figure 1). However, very low (cryogenic) temperatures are needed for physisorption. Disadvantages of chemisorption include slow kinetics, poor reversibility and high energy consumption. Physisorption is essential for carbon-based nanomaterials and MOFs, and chemisorption for metal hydrides and complex hydrides [7,9,10].

Carbon-based systems are characterized by low mass density, high surface area and chemical stability; they are cheap and exhibit fast kinetics [7,11]. Thanks to their porosity, MOFs have a large surface area and are rigid, structurally flexible and thermally stable [3,7].

Solid metal-hydride-based storage systems are considered a reasonable solution for safe storage of hydrogen in large quantities and under moderate pressure [6]. So far, a large number of intermetallic hydrides have been studied (e.g., Laves phases AB–FeTi, AB₂–ZrMn₂, ZrV₂, AB₅–LaNi₅, CaNi₅ and A₂B–Mg₂Ni), where element A is a hydride-forming element and element B is a non-hydride-forming element [4,12]. In order to tune the hydrogenation properties, a wide range of stoichiometries were studied where the elements are substituted at the A and B sites [4]. Although it is possible to create different types of hydrides, only some of them are suitable for hydrogen storage. These materials include only those with medium hydrogen affinity (types AB₅, AB₂ and AB) [13].

The energy that can be obtained by burning, e.g., 1 kg of hydrogen is about 2.6 times higher than the energy obtained from 1 kg of gasoline. However, due to the fact that hydrogen, in its liquefied state, is very light (its density is only 70 kg/m³, whereas the density of gasoline is approximately 750 kg/m³), approximately four times the volume of liquefied hydrogen is required to obtain the same amount of energy (e.g., to travel the same distance) as with gasoline [14].

A great effort is being invested in research and development of new materials for solid-state hydrogen storage. Various binary and ternary systems, intermetallic compounds and high-entropy alloys (HEAs) are being studied.

Metal hydrides have great potential for hydrogen storage thanks to their high H₂ storage capacity per unit volume; they are safe, reliable and the stored hydrogen has high purity compared to compressed or liquid hydrogen stored in tanks [15]. So far, only a few of the already large number of materials tested have been able to meet the requirements for solid-state hydrogen storage. The system must guarantee reversibility along with high gravimetric storage capacity, and operate at low temperatures under moderate pressure—all at an affordable cost [6]. Hydrides provide the most compact technology for hydrogen storage, although there are still drawbacks regarding the kinetics and thermodynamics of hydrogenation and dehydrogenation [16].

Much attention has been paid to low-density metal-based hydrides (Mg, Al, Li), which have high gravimetric capacity but require high temperatures for hydrogen desorption (for Mg alloys it is above 210 °C), which prevents their wider use [17,18]. For a long time, magnesium has been considered a potential medium for hydrogen storage as it reacts with hydrogen to form MgH₂ hydride containing 7.6 wt% of hydrogen (H₂), but the resulting hydride is too thermodynamically stable to release hydrogen at low temperatures [14]. Some intermetallic compounds of Mg and rare earth metals (La, Ce) have more favorable properties for hydrogen storage, showing high absorption rates even at room temperature after activation [14]. Vanadium-based solid solution alloys show promising properties for hydrogen storage, but the resulting hydrides are too thermodynamically stable under ambient conditions [17]. The intermetallic compounds LaNi₅ and TiFe have good reversible hydrogen-storage properties under near-ambient conditions, but low gravimetric capacity due to the presence of heavy elements (LaNi₅) or unfavorable thermodynamics (TiFe) [19,20]. Destabilization of the crystal structure is widely used to improve kinetic and thermodynamic properties [21]. The most widespread techniques include mechanical milling (with or without a catalyst and additives), mechanical alloying and microwave or ion irradiation [4,12,22].

The concept of high-entropy alloys (HEAs) arose together with the pioneering works of Cantor et al. [23] and Yeha et al. [24]. HEAs have opened up another area of research in alloy design and applications. HEAs differ from conventional alloys in that their base consists of five or more major alloying elements and they are characterized by a configurational entropy > 1.5R. Alloys containing four or fewer main elements in an equiatomic or nearly equiatomic ratio are called medium-entropy alloys (MEAs) [25]. The same nomenclature is also used for intermetallic compounds, for which high entropy is less likely to be achieved.

Intermetallic HEAs achieve parameters for hydrogen storage comparable to traditional intermetallic compounds (Laves phases LaNi₅, TiFe) [17,26]. So far, no alloy system has been identified with an abnormal ability to store more hydrogen per metal atom than intermetallic compound (H/M) structures. However, some of these alloy systems have very good reversible capacity at ambient temperature without the involvement of the activation process that is necessary for the production of many alloys [17].

Since it is very difficult to achieve a single-phase composition in HEAs and most of the HEAs investigated so far have been multiphase, a lot of effort is focused on designing procedures to suppress the formation of secondary phases, which can be misleading for elucidating the mechanisms of storage properties. So far, it has not been mentioned in the published studies that the multiphase composition of the alloys could significantly improve these properties.

This article reviews recent advances in solid-state hydrogen storage. In the theoretical part, proposals for computational methods using empirical parameters are described, followed by those using advanced computational methods. Basic concepts for the design of HEAs using Hume-Rothery rules and PCT curves describing behavior during hydrogenation are also mentioned. At the end of the article, notes with recommendations for future research and development in the field of hydrogen storage using these materials are summarized.

2. Methods for Designing the Composition of Alloys

Various methods have been proposed for the design of single-phase HEA compositions, based on already-gained experience and research results. The dependence of various parameters according to Hume-Rothery rules is monitored. The use of software technology working with thermodynamic models is known as CALPHAD. Machine learning (ML) uses previously obtained data to predict the hydrogenation properties of newly proposed HEAs. A theoretical procedure using DFT (density functional theory) exists, in which the relationship between structure and properties is monitored, allowing for the prediction of properties of metal hydrides and providing accurate enthalpies of their formation [27].

With the help of the above-mentioned methods, suitable alloys are subsequently designed, which are synthesized using various methods—arc melting, ball milling and reactive milling under different atmospheres (Ar, H₂), which are also used for the so-called activation of the alloy, or high-energy milling (HEBM) and high-pressure torsion (HTP) [28,29]. The proposed compositions can also be synthesized using the LENS (Laser Engineered Net Shaping) method [30] (Figure 2).

In the research and development of alloys for hydrogen storage, the affinity of individual elements to hydrogen (represented in particular by the enthalpy of hydride formation) is an important consideration to take into account when choosing elements in the design of the alloy composition [17,31]. The individual elements can be classified as hydride-forming (A-type elements) or non-hydride-forming (B-type elements); see Figure 3. The hydride-forming elements have lower enthalpy values and a higher propensity to form a hydride phase. In contrast, non-hydride-forming elements have higher enthalpy values, which makes hydride phase-formation more difficult [17]. The affinity of the alloy for hydrogen is, thus, influenced by the specific combination of the elements, which have different enthalpy of hydride formation [17,31,32]. These findings have accompanied the discovery of new compositions of high-entropy alloys suitable for hydrogen storage. Thanks to them, the design of HEAs does not depend solely on trial-and-error experiments, which are extremely demanding both in terms of time and cost. Various computational tools have been developed to improve the design of alloy compositions.

HEA is based on the hypothesis that, by mixing five or more elements in an equiatomic or near-equiatomic ratio (in the range of 5–35 at%), their entropy will be high enough to overcome the enthalpy of formation of compounds (intermetallic compounds) [17,24]. Based on this assumption, the formation of a disordered solid solution is facilitated and the formation of secondary phases should be suppressed [17]. A suitable combination of elements leads to the formation of simple single-phase solid solutions (BCC, FCC) with different atomic sizes, thereby achieving high lattice strain. [15] A similar effect can be achieved in a hexagonal close-packed (HCP) lattice with C14 and C15 Laves phase arrangements, which can be classified as an HEA or high-entropy intermetallic alloy [33].

Alloys with a base cubic-centered (BCC) arrangement are very promising for hydrogen storage because their storage capacity is associated with a less-tight lattice arrangement with more interstitial sites for hydrogen occupancy compared to face-centered cubic (FCC) and HCP structures [15,34]. In addition to the chemical composition of the HEA, other entropic contributions (vibrational, magnetic and electrical) and mixing enthalpies also affect the phase composition and stability of the HEA [17,35].

The behavior of hydrogen absorption can be described by the pressure–composition isotherm, also called the PCT curve (pressure–composition–temperature); (see Figure 2 in Reference [36]). Intermetallic compounds or other large amounts of metals are able to dissolve hydrogen up to a certain value of at%, but then the concentration increases, and in some regions, H-H interaction begins to occur; thus, hydride nuclei referred to as β-phase are formed, which consequently grow. The coexistence of the two phases at equilibrium pressure ${\mathrm{p}}_{\mathrm{eq}}^0$(T) corresponds to the isotherm plateau. The length of the plateau determines how much hydrogen can be reversibly stored with a small change in pressure. In the pure β-phase, the pressure of gaseous H₂ increases sharply with its concentration. At much higher pressures and higher hydrogen concentrations, additional plateaus and additional hydride phases can be found, reflecting the arrangement of additional interstitial locations of different types. The equilibrium pressure increases with temperature, and the plateau is shorter. The two-phase region ends at the critical point T_C; above T_C, it continues with the transition from α-phase to β-phase [14].

The dissolution of hydrogen to form the metal hydride M can be an exothermic or endothermic reversible reaction:

$$\mathrm{M}+\frac{\mathrm{x}}{2}\leftrightarrow {\mathrm{MH}}_{\mathrm{S}}+\Delta \mathrm{H}.$$

(1)

According to the thermodynamic properties of the alloys, the desorption pressure of the metal hydride varies with temperature according to the Van ’t Hoff Equation (2):

$$\ln \left(\frac{{\mathrm{p}}_{\mathrm{eq}}}{{\mathrm{p}}_{\mathrm{eq}}^0}\right)=\frac{\Delta \mathrm{H}}{\mathrm{RT}}-\frac{\Delta \mathrm{S}}{\mathrm{R}},$$

(2)

where ${\mathrm{p}}_{\mathrm{eq}}^0$ is the pressure at which the plateau occurs, ${\mathrm{p}}_{\mathrm{eq}}$ is the equilibrium pressure, T is the absolute temperature, ∆H and ∆S are the enthalpy and entropy changes and R is the universal gas constant.

The progression of the Van ’t Hoff dependence $\ln \left(\frac{{\mathrm{p}}_{\mathrm{eq}}}{{\mathrm{p}}_{\mathrm{eq}}^0}\right)$ on 1/T yields straight lines from which ∆H and ∆S can be determined.

Since it was found that calculations of cohesive properties, and especially enthalpies, of hydride formation are inaccurate, extremely time-consuming and do not correspond to the observed experimental data, empirical and semiempirical models were proposed. These models are based on the size and concentration of interstitial gaps, the minimum distance between two adjacent hydrogen atoms, the exit work, the different electron density and the electron band structure of the base metal [37].

2.1. Empirical Approach

So far, most of the HEAs investigated have been multiphase alloys. The design of single-phase HEAs is based on calculation and correlation involving various empirical parameters with respect to Hume-Rothery rules. According to these rules, the formation of substitutional solid solutions is favored for those alloys containing elements with similar atomic size, electronegativity and valence, with identical crystal structures. Based on the alloy composition, the atomic size mismatch (δ), valence electron concentration (VEC), Pauling electronegativity (χ_P) and thermodynamic parameters including the mixing enthalpy (ΔH_mix) and the parameter Ω proposed by Yang and Zang [17,35,38,39] are calculated.

These parameters can be calculated using the following equations:

$$\mathsf{\delta }=\sqrt{\sum {\mathrm{c}}_{\mathrm{i}}{\left(1-\frac{{\mathrm{r}}_{\mathrm{i}}}{\overline{\mathrm{r}}}\right)}^2}\times 100,$$

(3)

$${\mathsf{\delta }}_{\mathsf{\chi }}=\sqrt{\sum {\mathrm{c}}_{\mathrm{i}}{\left(1-\frac{{\mathsf{\chi }}_{\mathrm{i}}}{\stackrel{-}{\mathsf{\chi }}}\right)}^2}\times 100,$$

(4)

$$\mathrm{VEC}=\sum _{\mathrm{i}-1}^{\mathrm{N}}\left\{{\mathrm{c}}_{\mathrm{i}}{\mathrm{VEC}}_{\mathrm{i}}\right\},$$

(5)

$$\Delta {\mathrm{H}}_{\mathrm{mix}}=\sum _{\mathrm{i}<j}4{\mathrm{H}}_{\mathrm{ij}}{\mathrm{c}}_{\mathrm{i}}{\mathrm{c}}_{\mathrm{j}},$$

(6)

$$\mathsf{\Omega }=\frac{{\mathrm{T}}_{\mathrm{m}}\Delta {\mathrm{S}}_{\mathrm{mix}}}{\left|\Delta {\mathrm{H}}_{\mathrm{mix}}\right|},$$

(7)

where ${\mathrm{T}}_{\mathrm{m}}={\sum }_{\mathrm{i}-1}^{\mathrm{n}}{\mathrm{c}}_{\mathrm{i}\left(\mathrm{Tm}\right)}{}_{\mathrm{i}},\Delta {\mathrm{S}}_{\mathrm{mix}}=-\mathrm{R}\sum {\mathrm{c}}_{\mathrm{i}}{\mathrm{lnc}}_{\mathrm{i}},$ r_i is atomic radius, χ_i is electronegativity, VEC_i is the valence electron concentration of element i, $\overline{\mathrm{r}}=\sum {\mathrm{c}}_{\mathrm{i}}{\mathrm{r}}_{\mathrm{i}}$ is the average atomic radius, $\overline{\mathsf{\chi }}=\sum {\mathrm{c}}_{\mathrm{i}}{\mathsf{\chi }}_{\mathrm{i}}$ is average electronegativity, c_i and c_j are atomic fractions of elements i and j, H_ij is enthalpy of mixing of elements i and j at equimolar concentration in regular binary solutions, (T_m)_i is the melting temperature of element i and R is the universal gas constant [17,35,40].

Let us have a closer look at the thermodynamic parameter Ω defined in Equation (7), which can be interpreted as a way of visualizing the dominant term in the Gibbs free energy [38,39]. When Ω < 1, the mixing enthalpy dominates the Gibbs free energy and tends to stabilize intermetallic compounds and ordered phases. On the other hand, when Ω > 1, the Gibbs free energy is more influenced by the entropy term and tends to form a solid solution [38]. Yang and Zhang [38,39] analyzed several alloys and found that, when Ω > 1.1 and δ < 6.6%, the alloys tend to form single-phase solid solutions. Guo et al. [38,41] used the VEC, defined by Equation (4), as a criterion to predict the crystal structure of an HEA. Alloys with VEC < 6.87 tend to form BCC solid solutions, while FCC solid solutions are expected for VEC > 8 [38,41]. Nygård et al. [42] in their study claim that VEC is positively related to lattice-volume expansion, of which the destabilizing effect on hydrides could be used to tailor the sorption properties of HEAs. The VECs (6.0–6.4) were derived with an assumption of hydrogen desorption from the material in question below 100 °C and at room temperature. However, their work also showed that increasing the VEC above 5 causes a loss of hydrogen-storage capacity, so a compromise must be made when designing a HEA material according to the VEC.

The empirical approach is relatively simple and straightforward, but it does not consider and compare the enthalpies for formation of different possible phases [17].

Edalati et al. [43], in their study on VEC criteria, additionally performed SW calculations for the design of a TiZrCrMnFeNi alloy. They found that the alloy contained 95 wt% of Laves phase C14, which was able to absorb 1.7 wt% hydrogen at room temperature and with fast kinetics. Florian et al. [33] found that the main phase (C-14), which is common under non-equilibrium preparation conditions (e.g., arc melting or LENS), can be correctly predicted for a TiZrNbCrFe alloy only when C-15 and C-36 phases are excluded in CALPHAD calculations [4]. Based on these findings, Edalati et al. [43] and Floriano et al. [26,33] proposed three criteria for investigating high-entropy materials suitable for hydrogen storage at room temperature [16]:

• Selection of the AB₂ or AB system, where A represents hydride-forming elements such as Mg, Ti, Zr, V, Nb, etc. and B represents elements with low chemical affinity for hydrogen such as Cr, Mn, Fe, Co, Ni, etc.;
• Valence electron concentration (VEC) 6.4–6.5;
• The stability of the Laves phase, which should be investigated by thermodynamic calculations using the CALPHAD method (phase diagram calculation).

Although these proposed criteria have been successfully applied to the design of hydrogen-storage materials at room temperature, there have been some less-successful attempts to design HEAs with higher A/B ratios using Laves or BCC structures [15,33].

2.2. Semiempirical Approach

Considering the high number of theoretically possible compositions for HEAs and their unexplored properties, some design methods based on thermodynamics using semiempirical rules, ab initio calculations, machine learning and CALPHAD (Calculation of Phases Diagrams) for the formation and stability of phases in different HEAs have been used to provide reasonable solutions [33].

The CALPHAD method is a computational tool for creating phase diagrams based on thermodynamic models. These models are based on binary and ternary systems, and the reliability of this method is higher when data of alloys contained in the database are used for the interpolation. For quaternary, quinary and higher-order compositions, extrapolation is generally required, which reduces the accuracy of the predictions but still provides satisfactory results [17,35]. This method has limited predictive ability when other unknown intermetallic phases are present, as they are outside the ranges fitted by polynomial functions [17,44]. Another limitation of this method consists of the fact that it is based on thermodynamic equilibrium, and thus it may lack the prediction of metastable phases, which are often present in the most successful alloys [17,44].

Applying an empirical approach, using parameter and thermodynamic calculations by the CALPHAD method can reasonably predict the phase composition and stability of various HEAs [38,39,40,41,45,46]. However, the CALPHAD method is mainly limited by the available databases, which contain data of alkali metals or alkaline earth metals, but the combinations with transition metals are not available yet [38].

Another method is machine learning (ML), that works with large amounts of data, analyzes it, puts it into context and then evaluates it. The method stores the resulting algorithms in models, which are subsequently used to solve similar problems.

For the search for new material compositions, a data-driven approach with machine/statistical learning models and a sufficiently large and reliable database is used [47,48]. As HEA compositions require a large number of materials for synthesis, this model, which does not explicitly depend on the crystal structure, can rapidly predict the thermodynamics of hydride formation and greatly facilitate the rapid discovery of new suitable candidates for HEA hydrides [47,48]. The predictive power in the broader compositional space of HEA could lead to the discovery of materials that simultaneously exhibit thermodynamic destabilization without sacrificing capacity [42,47]. A screening procedure DFT to determine the ΔH of HEA hydrides was computationally processed, and it confirmed the destabilization trend predicted by ML. The experiments performed simultaneously provided a more atomistic view of the hydridation process [47]. Thus, the continuous collection (experimental or computational) of thermodynamic data for HEA hydrides and its storage in centralized repositories will lead to the continuous improvement of thermodynamic models using ML [47,49].

Hu et al. [1] investigated the TiZrVMoNb alloy in their work. Using the DFT method, they predicted the hydrogen site occupancy, the hydrogen-induced phase transformation, the threshold hydrogen content for the phase transformation, the hydrogen content in the phase that can manifest itself during the phase transformation and the maximum hydrogen storage capacity. From the simulation results, they found that the material had a relatively large storage capacity of 2.65 wt% and reduced thermal stability compared to the TiZrHfMoNb alloy [1,4].

This procedure of HEA alloy design based on theoretical calculations was successfully applied to TiZrNbFeNi [33], TiZrCrMnFeNi [26], TiZrCrMnFeNi and TiVCrMnFeCo [43] alloys, using a combination of theoretical design and mechanical synthesis with the help of HPT.

3. Absorption Properties of Selected HEAs

Since HEAs are mainly composed of transition metals with relatively high molecular weight, some of which do not react with hydrogen, the hydrogen-storage capacity of these alloys is limited. Although some light elements that absorb hydrogen, such as Mg [15,50,51], Al [52] or Sc [4,53], are introduced into HEAs, efforts to significantly increase the hydrogen-storage capacity of the respective alloys have not been successful due to the complex interactions among the alloying elements. At the same time, works were published that considered the occupation of individual lattice positions by hydrogen atoms and monitored the stoichiometry of hydrides, including a proposal based on the calculated enthalpies of their formation ΔH [25,54]. [H/M]_max in HEAs is generally thought to be dominated by the affinity between the alloy elements and hydrogen [54].

The storage capability of an HEA can be greatly enhanced by modifying its chemical composition, which leads to a broad family of materials capable of reversibly absorbing hydrogen at both ambient and elevated temperatures. For HEAs containing elements from the transition metal group, alloys capable of storing more than two hydrogen atoms per metal atom (H/M ratio exceeding 2) have not yet been reported, except for the work of Sahlberg et al. [17,55], where a very high H/M ratio of 2.5 was measured during hydrogen absorption measurements, corresponding to 2.7 wt% H₂ at 299 °C (±1 °C) and 53 bar (5.3 MPa) hydrogenation. Subsequent hydrogenation was found to be a one-step mechanism from BCC to BCT, as evident from the corresponding PCT curve with one pressure plateau at 0.1 bar (0.01 MPa) H₂ at 299 °C [55]. They found that the high storage capacity of TiZrVHfNb was due to the large lattice deformation, which is very favorable for hydrogen absorption as it rearranges and occupies all tetrahedral and some octahedral interstitial sites in the hydride structure of BCT (Figure 4) [6,17,55].

However, even with other potentially interesting HEAs (TiVZrNbTa [56], TiVZrNb [57]), a one-step, reversible transition from BCC alloys to BCT/FC was found, regardless of the production process [57]. This facile transformation may be the cause of favorable reversible hydrogenation properties. Such a simple transition indicates a good reversibility of the hydrogenation reaction and is important for practical use.

However, other studies have not been able to reproduce such high H/M ratios in the same alloy, nor has any trend been observed between the VEC atom size mismatch and hydrogen storage properties [6,17]. Zepon et al. [51] found that during the hydrogenation process of MgZrTiFe_0.5Co_0.5Ni_0.5, hydrogen atoms occupied first the octahedral sites in the BCC structure, and only then occupied the tetrahedral sites in the FCC structure, which was transformed from the BCC structure. Karlsson et al. [58] studied the hydrogenation mechanism in TiZrVHfNb by in situ and ex situ experiments using neutron diffraction and found that the hydrogen occupancy ratio of tetrahedral sites to octahedral sites was 53:47 at 500 °C, and changed to 92.9:5.2 at room temperature.

Zlotea et al. [59] reported that a two-phase transformation was induced during the hydrogenation process of TiZrHfTaNb. It changed from the initial BCC structure to a monohydride with a BCT structure and then to a dihydride with an FCC structure at low pressure (about 6 bar, i.e., 600 kPa). They found out that hydrogen atoms first occupied octahedral sites in the BCT structure (monohydride) and tetrahedral sites in the FCC structure (dihydride) [59]. Hu et al. in their study [1] stated that it was necessary to further discuss the occupation of sites for hydrogen atoms in HEA, and the necessity of detailed theoretical investigation on the site preferences for hydrogen atoms combined with phase transformation during hydrogen absorption in HEA, in order to gain more fundamental insight into hydrogen storage in HEA.

Ek et al. [60] investigated the hydrogen sorption properties of a total of 21 HEAs in the TiVZrNbHf system with different elemental ternary, quaternary and quinary compositions. The alloys exhibited very slow absorption kinetics. A total of seventeen compositions had a single-phase BCC structure and one composition had a single-phase hexagonal structure. Out of these alloys, 15 exhibited the presence of either FCC or BCT metal hydrides with maximum H/M ratios close to 2. They further found out that large amounts of Zr and Hf determined the formation of BCT metal hydrides over FCC. In their conclusions, they confirmed the theory that control over the VEC and χ_p parameters is essential for the design of HEAs for hydrogen-storage applications.

The difference between these phase transformations (one-step or two-step) is attributed to different lattice deformations, which are denoted by the parameter δ [59]. A high value of this parameter is considered to cause a one-step transformation [4].

Westlake [1,61] considered a size criterion for the occupancy of interatomic hydrogen positions and suggested a gap value for co-existing types of interstitial sites > 0.4 Å. At a distance of two hydrogen atoms (>2.1 Å), it is possible to partially fill even seemingly less-suitable places with hydrogen. For AB₂ alloys with C14 or C15 Laves structures, this principle was verified for the determination of hydride stoichiometry. It turns out that this approach for HEAs should have been used to understand and increase the H storage capacity. Strozi et al. [38] compared hydrogen absorption, hydride structure, mean enthalpy of hydrogen solution and mean enthalpy of hydride formation for selected HEAs. The results showed that only in systems with a negative value for the enthalpy of hydrogen solution and hydride formation is it possible to observe a relatively high capacity for hydrogen storage. The MgVAlCrNi system prepared by the authors cannot be considered promising, since, except for V, the hydrogen solution enthalpy values are positive.

So far, most studies on HEAs for hydrogen storage have focused on transition-metal-based alloys with heavy elements. Lightweight HEAs are a group of alloys consisting in part of one or more light elements (Al, Mg, Li, Ti, Si) incorporated into a solid solution so that a high H/M ratio is maintained, and the molar mass reduced, thereby increasing the gravimetric capacity. One of the most promising groups of materials are Mg-based light metals.

Kao et al. [54] synthesized a CoFeMnTi_xV_yZr_z HEA and found that the hydrogen storage capacity can be optimized by adjusting the Ti, V and Zr content, while the addition of these elements results in the expansion of the crystal lattice and increases the interstitial space.

Witman et al. [47], using the ML model, significantly reduced the time and financial demands of the first-principles modeling and its subsequent experimental validation. They focused on hydride HEAs and used ML to verify their stability in a large HEA space. Based on the targeted thermodynamic properties and the stability and density of the alloy phase, they performed a targeted synthesis of several new hydrides that exhibited significant destabilization (70× increase in equilibrium pressure, 20 kJ/mol H₂ decrease in desorption enthalpy) relative to the reference HEA hydride, TiVZrNbHfH_x, to experimentally validate their predictions.

An overview of selected HEA alloys, their hydrogenation properties, methods of synthesis and processing proposed by different authors is given in

Table A1. Summary of different BCC HEAs and their hydrogen storage properties.

Normalized Chemical Composition Ordered by Atomic Number	Synthesis and Processing	Alloy Phase	Hydride Phase	Structural Transf. upon Hydrogenation	H2 Absorp. Capacity (wt%)	H/M	H2 Absorp. Kinetics	Hydride Decompos. Onset/Peak Temperatures (K)	Ref.
Ti0.2Zr0.2Nb0.4Hf0.2	Arc melting	BCC	FCC	1 step	1.12	—	—	656/—	[2]
Ti0.2Zr0.2Nb0.3Mo0.1Hf0.2	Arc melting	BCC	FCC	1 step	1.54	—	—	605/—	[2]
Ti0.2Zr0.2Nb0.2Mo0.2Hf0.2	Arc melting	BCC	FCC	1 step	1.18	—	—	575/—	[2]
Ti0.2Zr0.2Nb0.1Mo0.3Hf0.2	Arc melting	BCC	BCT	1 step	1.40	—	—	437/—	[2]
Ti0.2Zr0.2Mo0.4Hf0.2	Arc melting	BCC	BCT	1 step	0.92	—	—	441/—	[2]
Ti0.2V0.2Zr0.2Nb0.2Hf0.2	Arc melting	BCC	BCT	—	2.1 (573 K) 2.2 (473 K)	1.94 (573 K) 2 (473 K)	1.7 wt% in 300 s (573 K, 2 MPa	623/—	[11]
Ti0.25V0.25Zr0.25Nb0.25	Arc melting	BCC	FCC	1 step	—	1.98 (293 K)	—	~573/—	[60]
Ti0.22V0.22Zr0.22Nb0.11Hf0.22	Arc melting	BCC	BCT	—	—	1.82 (293 K)	—	~573/—	[60]
Ti0.22V0.22Zr0.11Nb0.22Hf0.22	Arc melting	BCC	FCC	—	—	1.99 (293 K)	—	~593/—	[60]
Ti0.22V0.22Zr0.22Nb0.22Hf0.11	Arc melting	BCC	FCC	—	—	2.00 (293 K)	—	~593/—	[60]
Ti0.22V0.11Zr0.22Nb0.22Hf0.22	Arc melting	BCC	FCC	1 step	—	1.96 (293 K)	—	~573/—	[60]
Ti0.11V0.22Zr0.22Nb0.22Hf0.22	Arc melting	BCC	FCC	—	—	1.97 (273 K)	—	~573/—	[60]
Ti0.2V0.2Zr0.2Nb0.2Hf0.2	Arc melting	BCC	FCC	1 step	—	1.99 (293 K)		~593/—	[60]
Ti0.25V0.25Zr0.25Hf0.25	Arc melting	BCC	Phase separation	—	—	—	—	—	[60]
Ti0.25V0.25Nb0.25Hf0.25	Arc melting	BCC	FCC	—	—	1.99 (293 K)	—	~593/—	[60]
Ti0.25V0.25Nb0.25Hf0.25	Arc melting	BCC	BCT	—	—	1.98 (293 K)	—	~623/—	[60]
V0.25Zr0.25Nb0.25Hf0.25	Arc melting	BCC (major) Unknown (minor)	Phase separation	—	—	—	—	—	[60]
Ti0.2V0.2Zr0.2Nb0.2Hf0.2	Arc melting	BCC	BCT	1 step	2.7 (573 K)	2.5 (573 K)	—	~473/~673	[70]
Ti0.2V0.2Zr0.2Nb0.2Hf0.2	Arc melting (followed by ball milling)	BCC	FCC (293 K) BCT (723 K)	1 step	1.8	1.9 (562 K)	—	—	[58]
Ti0.2Zr0.2Nb0.2Hf0.2Ta0.2	Arc melting (homogenized by induction heating)	BCC	FCC	2 step	—	~2.0 (573 K)	—	~593/~648	[59]
* Ti0.2V0.2Zr0.2Nb0.2Mo0.2	LENS—300 W	BCC (major) NbTi4 (minor)	FCC (TiHx) BCC (NbH0.4)	—	2.3 (323 K) 1.78 (673 K) after activation	—	2.3 wt% in 1380 s (303 K, 8.5 MPa H2)	—	[30]
** Ti0.2V0.2Zr0.2Nb0.2Mo0.2	LENS—1000 W (3×)	BCC (major) Zr-rich (Ppt)	BCC (major) Zr-rich (Ppt)	—	0.59 (323 K) 0.61 (673 K) after activation	—	0.59 wt% in 1380 s (303 K, 8.5 MPa H2)	—	[30]
Ti0.25V0.25Zr0.25Nb0.25	Arc melting	BCC	FCC	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.24V0.24Zr0.28Nb0.24	Arc melting	BCC	FCC	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.22V0.22Zr0.33Nb0.22	Arc melting	BCC	FCC	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.21V0.21Zr0.37Nb0.21	Arc melting	BCC	FCC	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.2V0.2Zr0.4Nb0.2	Arc melting	BCC	FCC	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.25V0.25Zr0.04Nb0.25Ta0.21	Arc melting	BCC	FCC (major) BCT (minor)	—	—	~1.9	—	—	[65]
Ti0.25V0.25Zr0.125Nb0.25Ta0.125	Arc melting	BCC	FCC (major) BCC (minor)	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.25V0.25Zr0.19Nb0.25Ta0.06	Arc melting	BCC	FCC (major) BCC (minor)	1 step (phase separation upon 1 cycle)	—	~1.9	—	—	[65]
Ti0.25V0.25Nb0.25Ta0.25	Arc melting	BCC	FCC (major) BCT (minor)	—	—	1.9	—	~498/—	[65]
* (VFe)60 (TiCrCo)40- xZrx	Arc melting	BCC (major) Laves phase C14 (minor) CeO2 (minor) FCC (minor)	—	2 steps	3.5 (298 K)				[64]
Ti0.25V0.25Cr0.25Mo0.25	Arc melting	BCC	BCC	1 step	—	∼0.75	—	∼523/—	[42]
Ti0.2V0.2Cr0.2Nb0.2Ta0.2	Arc melting	BCC	FCC (major) BCC (minor)	1 step	—	∼1.9	—	473 (1.Max) ~556 (2.Max)/—	[42]
Ti0.2V0.2Zr0.2Nb0.2Hf0.2	Arc melting	BCC	FCC	1 step	—	∼1.9		553 (1.Max) ~666 (2.Max)/—	[42]
Ti0.25V0.25Nb0.25Hf0.25	Arc melting	BCC	FCC	1 step	—	∼2	—	553 (1.Max) ~648 (2.Max)/—	[42]
** Ti0.25V0.25Nb0.25Ta0.25	Arc melting	BCC	FCC (major) BCC (minor)	1 step	—	∼1.9	—	~503 (1.Max) ~602 (2.Max)/—	[42]
*** Ti0.25V0.25Cr0.25Nb0.25Mo 0.25	Arc melting	BCC	FCC (major) BCC (minor)	1 step	—	∼2	—	473 (1. Max) ~556 (2.Max)/—	[42]
Ti0.25Zr0.25Nb0.25Hf0.25	Arc melting	BCC	BCT	1 step	—	∼2	—	553 (1.Max) ~694 (2.Max)/—	[42]

* Rev. H2 capacity (wt% H/M)—Up to 2.1 wt% (298 K). ** Rev. H2 capacity (wt% H/M)—1.96 wt% (293 K, vac/ 2.3 MPa H2) stable 10 cycles. *** Rev. H2 capacity (wt% H/M)—Fading within five cycles to 0 wt%.

,

Table A2. Summary of different intermetallic HEAs and their hydrogen storage properties.

Normalized Chemical Composition Ordered by Atomic Number	Chemical Composition	Synthesis Method	Alloy Phase	Maximum H2 Storage Capacity (wt% H2)	H2 Absorption Kinetics	H/M	Hydride Decomposition Temperature (K)	Ref.
Ti0.20Fe0.40Ni0.15Zr0.20Nb0.05		Arc melting	Laves phases C14 (major) BCC (minor)	1.38	—	0.95	305 **	[33]
Ti0.20Fe0.20Ni0.20Zr0.20Nb0.20		Arc melting	Laves phases C14 (major) BCC (minor)	1.64	—	1.17	305 **	[33]
Ti0.17V0.17Cr0.17Fe0.17Ni0.17Zr0.17		LENS	Laves phases C14 (major) α-Ti solid solution (minor)	1.81	—	—	323 *	[30]
TixVyMnFeCoZrz	0.5 ≤ x ≤ 2.5 0.4 ≤ y ≤ 3.0 0.4 ≤ z ≤ 3.0	Arc melting	Laves phases C14	0.03–1.80	18 ≤ t0.9 (s) ≤ 1250 (298 K, 0.97 MPa H2)	0.02–1.17	298 *	[54]
Ti0.17Cr0.17Mn0.17Fe0.17Ni0.17Zr0.17		Arc melting + HPT	C14 Laves (major)	1.7	1.6 wt% H2 in 60 s (303 K, 3.9 MPa H2)	1	305 *	[43]

* Desorption performed in Sieverts-type apparatus. ** Desorption performed in Sieverts-type apparatus and not fully dehydrogenated (retained H₂ in material).

and

Table A3. Summary of different lightweight HEAs and their hydrogen-storage properties.

Normalized Chemical Composition Ordered by Atomic Number	Synthesis and Processing	Phases	H2 Storage Capacity (wt%)	H2 Absorp. Kinetics	H/M	Hydride Decompos. Onset/Peak Temperatures (K)	Enthalpy of Hydrogen Solution (kJ mol−1 H)	Ref.
Mg0.20Ti0.20V0.20Cr0.20Fe0.20	Reactive milling – 3.0 MPa H2 – BPR 40:1 – RPM 600 + HPT	BCC (major) MgH2 (minor) Amorphous phase (minor)	0.3 (623 K)	0.2 wt% in 3600 s (303 K, 2 MPa)	—	483/520 (1.Max) 633 (2.Max)	—	[15]
Mg0.20Al0.20V0.20Cr0.20Ni0.20	Mechanical alloying – 0.7 MPa Ar – BPR 20:1 – RPM 600	BCC	—	—	—	—	—	[38]
Mg0.20Al0.20V0.20Cr0.20Ni0.20	Reactive milling – 3.0 MPa H2 – BPR 20:1 – RPM 600 – 24 h	BCC	0.3	—	0.09–0.14	589/650	+12.2	[38]
Mg0.28Al0.19V0.28Cr0.19Ni0.06	Reactive milling – 3.0 MPa H2 – BPR 20:1 – RPM 600 – 24 h	BCC (major) BCC (minor) MgH2 (minor)	0.28–0.41 (calc, based on XRD data)	—	0.11–0.16	—	+9.84	[38]
Mg0.26Al0.31V0.31Cr0.06Ni0.06	Reactive milling – 3.0 MPa H2 – RPM 600 – BPR 20:1 – 24 h	BCC (major) BCC (minor) MgH2 (minor)	—	—	—	—	+8.36	[38]
Mg0.22Ti0.22Cr0.11Mn0.11Ni0.11Nb0.22	Reactive milling – 0.7 MPa Ar – BPR 20:1	BCC (major) Cr (minor) Mn (minor)	—	—	0.8	—	—	[50]
	– RPM 600 – 24 h	Mg (minor)						[50]
Mg0.22Ti0.22Cr0.11Mn0.11Ni0.11Nb0.22	Reactive milling – 3.0 MPa H2 – BPR 20:1 – RPM 600 – 24 h	FCC (major) Cr (minor) Mn (minor) Mn (minor) Mn2NiH4 (minor)	1.6	—	—	493/576 (1. Max) 653 (2. Max)	—	[50]
Mg0.22Ti0.22Fe0.11Co0.11Ni0.11Zr0.22	Reactive milling – 0.7 MPa Ar – BPR 40:1 – RPM 600 – 24 h	BCC	1.2 (623 K)	1.0 wt.% in1800 s (623 K, 2 MPa H2)	0.67	—	−14.4	[51]
Mg0.22Ti0.22Fe0.11Co0.11Ni0.11Zr0.22	Reactive milling – 0.7 MPa H2 – BPR 40:1 – RPM 600 – 24 h	FCC	—	—	—	503/573 (1. Max) 648 (2. Max)	—	[51]
Al0.17Cr0.17Mn0.17Fe0.17Ni0.17W0.17	Mechanical alloying – Ar – BPR 10:1 – RPM 300 – PCA toluene – 20 h	BCC (major) FCC (minor)	0.62 (293 K)	—	—	358/—	+11	[52]
Mg0.10Ti0.30V0.25Zr0.10Nb0.25	Mechanical alloying – Ar – BPR 26:1 – RPM 700 – 2 h	BCC	2.7 (298 K)	2.7 wt% in 60 s (298 K, 2.5 MPa)	1.72	523/563	—	[48]
Mg0.10Ti0.30V0.25Zr0.10Nb0.25	Reactive milling – 7 MPa H2 – BPR 60:1 – RPM 40 – 1 h	FCC	—	—	1.65	—	—	[48]

BRP—ball-to-powder weight ratio; RPM—revolutions per minute.

in the Appendix A.

4. Manufacturing Processes (Methods) for Increasing Absorption Capacity

The methods of preparing HEAs can be divided into four categories. Zhang et al. [62] in their work present methods beginning from the liquid state, the solid state and the gaseous state, and also the electrochemical process. For some HEAs, powder metallurgy methods can be used, as Torralba [63] states in his work. Two of these methods are considered for the preparation of HEAs suitable for hydrogen-storage applications. The mechanical alloying (MA) technique, which is solid-state powder processing, includes various variants of ball milling. The method of processing from the liquid state is most often arc melting [62].

Conventional processes such as arc melting are often used to produce HEAs, as they facilitate the characterization of the structure. Arc melting is carried out in a vacuum or in an inert atmosphere. In this process, pure metal elements are melted at high temperatures (over 1000 °C), and are also often re-melted to ensure homogeneity [2,42,43]. Yang et al. [64] investigated the cyclic properties of the HEA (VFe)₆₀(TiCrCo)_40−xZr_x. They synthesized the alloy by arc melting in an Ar protective atmosphere, and then annealed it at 1400 °C for 30 min. The alloy had a major BCC phase with a minor C14 Laves phase and a Zr-based FCC phase. They found out that the addition of zirconium improved the hydrogen absorption kinetics, but the storage capacity decreased with increasing Zr content. As the element Zr can reduce the accumulation of microdeformations in the crystal lattice during hydrogen absorption–desorption cycles, the cyclic properties of (VFe)₆₀(TiCrCo)_40−xZr_x alloys are improved. As the Zr content increases, the rate of decrease in the desorption capacity of (VFe)₆₀(TiCrCo)_40−xZr_x decreases, and the best cyclic properties were obtained when x = 2 with only 4.5 % decrease in capacity after 10 cycles. Nygård et al. [65] investigated the HEA TiVZrNbTa with different ratios of Ti, V, Zr, Nb and Ta. The alloy was synthesized by arc melting. All the obtained alloys had BCC crystal structures and formed metal hydrides with an FCC structure, which had a hydrogen-to-metal ratio close to 2. No correlation between hydrogen-storage capacity and local lattice strain was observed.

Sleiman and Huot [66] synthesized a Ti_0,2V_0,2Zr_0,2Nb_0,2Hf_0,2 alloy using arc melting in an Ar atmosphere (0.7 bar). They monitored the influence of particle size, hydrogenation temperature and hydrogenation pressure on the activation process of a TiVZrHfNb alloy. They found that the activation properties depended on the particle size. Reducing the particle size accelerates the absorption kinetics and shortens the incubation time. At the same time, however, it negatively affects the maximum capacity for hydrogen storage. Particle size plays a crucial role in hydrogenation kinetics [6,17]. Shen et al. [2] found that TiZrHfMoNb has the ability of reversible storage during hydrogen absorption–desorption cycles [1,2]. Nygård et al. [42] reported that TiVCrNb exhibits fast hydrogen absorption kinetics with a hydrogen-storage capacity of 1.96 wt%, and the hydrogen absorption of this HEA does not depend on any complicated activation procedure [1,42].

Hu et al. [53] designed and successfully synthesized (arc melting), based on density functional theory (DFT), the HEA TiZrHfScMo with a BCC structure. Using DFT, they calculated the lattice constant, formation enthalpy, binding energy and electronic properties of the hydrogenated alloy. In their work, they investigated the absorption behavior of the alloy using DFT. The results showed that the octahedral sites in the BCC lattice are first occupied at low hydrogen concentrations. At a hydrogen concentration above 1.8 wt%, the structure transitions from BCC to FCC, and from this instant, tetrahedral sites in FCC are preferentially occupied.

In general, it can be stated that lightweight HEA alloys have a wide range of compositions with different affinities for hydrogen. For example, an alloy of Mg with some transition metals (e.g., Mn, Cr, Ti, Nb) has limited solubility of these elements in solid solution, leading to segregation and formation of multiple phases. Since the light elements have low melting temperatures, they cannot be prepared by conventional melting techniques; instead, they are synthesized using the ball milling technique in a protective argon atmosphere, which has been shown to be an efficient method for the synthesis of supersaturated solid solutions (mostly BCC), especially in the case of an unfavorable mixing enthalpy ΔH_mix [17,38,50]. Alternatively, HEAs have been synthesized by reactive ball milling in a hydrogen atmosphere, leading to the formation of a hydride phase (mostly FCC) [17,38,67]. However, spherical milling can also lead to the formation of a multiphase alloy [17]. In addition, to achieve better activation properties for HEAs, the ball milling method is further supplemented by high-pressure torsion treatment (HPT) [15,16,17].

The HPT method is widely used to generate nano-grains or ultrafine grains in metallic materials. It is an efficient method even for the synthesis of immiscible systems such as Mg-Ti and Mg-Zr with hydrogen-storage capability [16]. There are also attempts to use this method to synthesize non-metallic HEAs (based on oxides, nitrides and possibly hydrides) such as the MgTiVCrFe-H hydride for hydrogen storage [16]. The HPT method has been shown to be suitable not only for the synthesis of HEAs for hydrogen storage—there is an increase in activity and kinetics, which is mainly attributed to the formation of lattice defects in the structure and grain boundaries. These, then, act as fast pathways for the transport of hydrogen from the surface to the whole volume of the material [15,16,17,67].

The method of mechanical alloying—ball milling (BM)—is carried out either in an inert or reactive atmosphere. Zepon et al. [51] prepared a lightweight MgZrTiFe_0.5Co_0.5Ni_0.5 alloy using the high-energy ball milling (HEBM) method. HEA was milled for 24 h at 600 rpm under a pressure of 0.7 MPa in an Ar atmosphere and 3.0 MPa in an H₂ atmosphere.

Reactive ball milling (RBM) processes are used for alloy systems with low alloying affinity (positive mixing enthalpy), which tend to segregate unwanted phases. This method was used by Marques et al. [50] for the synthesis of MgTiNbCr_0.5Mn_0.5Ni_0.5 and Mg_0.68TiNbNi_0.55 alloys. FCC hydride and Mg₂NiH₄ were monitored for this alloy. Montero et al. [57] synthesized a Ti_0.325V_0.275Zr_0.125Nb_0.275 alloy, finding that longer milling times have no benefit and that partial amorphization of the structure occurs.

De Marco et al. [15] synthesized MgVCr and MgVTiCrFe alloys with very fine microstructures by mechanical alloying using high-energy ball milling (HEBM) under a hydrogen atmosphere followed by high-pressure torsion (HPT) to improve activation. The alloys were reactive-milled (RM) for 24–72 h; they were mainly composed of a solid solution of BCC and partly of the β-MgH₂ phase. RM for the MgVCr alloy guarantees an even distribution of elements (Mg, V and Cr) by overcoming the thermodynamic immiscibility among these elements. No significant hydrogen sorption was observed for the main BCC phase; TiH₂ and Mg₂FeH₄ hydrides were observed. This alloy showed a low storage capacity of ~1 wt%. On the other hand, the HPT process led to complete amorphization of the material. Hydrogenation studies showed that HPT can be considered a potential processing method, as the obtained alloy was characterized by better hydrogen absorption kinetics than the untreated sample. However, it should be mentioned that the cyclic sorption and desorption of hydrogen in this alloy resulted in the decomposition of the material into less complex compounds (such as TiFe, TiH, TiCr₂). Dewangan et al. [52] synthesized the HEA AlCrFeMnNiW using HEBM, and its storage capacity was estimated to be 0.616 wt% at room temperature and normal pressure in a single exposure. Sun et al. [68] studied lightweight HEAs and the effect of individual lightweight elements (Al, Ti, Si, Li, Mg) on the phase composition of the alloys, which were produced by casting, the rapid solidification process and mechanical alloying [38,68]. They found that the addition of Mg led to the formation of intermetallic phases during solidification, which could be dissolved in the solid solution by the HEBM process. This process can suppress the formation of other secondary phases, thus allowing the formation of single-phase alloys that could not be produced by convex casting [38,52,69].

Edalati et al. [16] have shown in their studies that another possibility for the production of HEAs was the synthesis of alloys prepared from elemental powders using the high-energy ball milling (HEBM) method under a hydrogen atmosphere [15,28,29]. This process can achieve nanostructured materials with a good level of mixing between elements, β-phase content and improved properties for hydrogen storage. For better activation of the alloys, these samples are subsequently subjected to the HPT process after HEBM (see Figure 5) [15,28,29].

The LENS additive technology method is suitable for use with materials for which it is expected that rapid solidification (10³–10⁶ K/s) will prevent the segregation of undesirable phases. Homogeneous structures can be obtained by using high laser power, powder of suitable quality and suitable processing parameters [4,30]. Kunce et al. [30] synthesized the HEA TiZrNbMoV using additive Laser Engineered Net Shaping (LENS) technology. In their study, they investigated the effect of laser energy on the resulting microstructure and its subsequent applicability for hydrogen storage. The TiZrNbMoV alloy, after synthesis using low laser power, exhibited a biphasic dendritic matrix with some unmelted Mo particles and BCC solid-solution dendrites surrounded by an orthorhombic NbTi₄-type phase. The alloy, which was synthesized using a 1 kW laser, had a multiphase microstructure with a Mo-rich matrix containing Zr-rich precipitates and no dendritic segregation. The maximum hydrogen capacities obtained for this alloy were 0.59 wt% after synthesis and 0.61 wt% after additional heat treatment.

5. Summary

Of the existing candidate materials applicable for safe and effective hydrogen storage, HEAs seem to be the most promising solution due to their characteristics. HEAs’ variability in composition and structure offer high potential for the further improvement of their hydrogen-storage properties. Extensive research in this field is just beginning, but the results of many authors show good progress and indicate directions where research and development should be focused. The important characteristics for successful hydrogen-storage applications are, especially, the capacity of hydrogen absorption, energy and temperatures of hydrogen absorption and desorption, kinetics of these processes and stability of material structure during the hydrogenation and dehydrogenation cycles.

Results of HEA research show that the BCC phase is the most promising among all the reported ones theoretically suitable for hydrogen storage. Synthesis of proper HEAs is considered using parameters based on Hume-Rothery rules, which are atomic size mismatch (δ), valence electron concentration (VEC), Pauling electronegativity (χ_P) and thermodynamic parameters including the mixing enthalpy (ΔH_mix) and the parameter Ω. The studies revealed the necessity of focusing on VEC because it has a very important role in the description of phase formation and their temperature stability. As for the intermetallic-based HEAs, attention should be paid to VEC in the research to describe the stability of hydrides; calculation methods and experiments should be applied in this field to improve knowledge and understanding concerning the effects and conditions for the formation of less-stable hydrides. In order to ensure high absorption capacity for HEAs, it is important to optimize their chemical composition with respect to preservation and stability of the BCC phase. The AB₂ or AB systems, where A represents hydride-forming elements (Mg, Ti, Zr, V, Nb, etc.) and B represents non-hydrogen-forming elements (Cr, Mn, Fe, Co, Ni, etc.), should be focused on and further developed.

The next step to improve the storage capacity of the alloys can be seen in the development of processing technologies that lead to the activation of the crystal lattice, which can increase the capacity of interstitial sites and reduce the energy needed for hydrogenation and dehydrogenation. Most of the studied alloys were synthesized using arc melting. In several studies, RBM, HEBM and HPT were used, and it was mostly their application that led to the improvement of the properties resulting in the homogenization of an alloy’s structure.

Lightweight HEAs based on Mg, Al, Ti and other elements are also an interesting area, but so far little-studied. The HEA Mg_0.10Ti_0.30V_0.25Zr_0.10Nb_0.25 [48], which absorbed 1.7 H/M (2.7 wt.%) at room temperature, seems to be a good candidate for future research.

6. Conclusions

Within the wide range of promising materials suitable for hydrogen storage in atomic form, research has recently focused on high-entropy alloys with a BCC phase due to their composition, which can be modified to a relatively large extent with respect to storage capacity, reversibility, kinetics, thermodynamics and cyclability.

To date, research on HEAs for hydrogen storage is at its early stage and a number of new insights have been gained, but, at the same time, a number of other issues and phenomena are emerging that have not been adequately explored yet. For these reasons, many of the HEAs synthesized so far fall short of the predicted parameters, so it is necessary to get more data to analyze into a central storage system for machine learning, using the results of computational models (CALPHAD) and experiments. Phase composition tuned to ensure the high absorption capacity of HEAs, together with the development of processing technologies, need to be investigated with a focus on improving the hydrogenation and dehydrogenation properties. In the case of lightweight HEAs, the research has to be focused on improving reversible storage capacity and kinetics at ambient temperature and pressure.

Hysteresis of hydrogenation parameters has not been systematically monitored yet for a number of alloys (especially HEAs). It can be influenced by many factors, such as alloy composition, processing temperature, interstitial space size and number of absorption–desorption cycles, among others. Understanding the hysteresis behavior is the key to the exploitation of prospective HEAs for hydrogen storage.