Paving the Way to the Fuel of the Future—Nanostructured Complex Hydrides

Abstract

Hydrides have emerged as strong candidates for energy storage applications and their study has attracted wide interest in both the academic and industry sectors. With clear advantages due to the solid-state storage of hydrogen, hydrides and in particular complex hydrides have the ability to tackle environmental pollution by offering the alternative of a clean energy source: hydrogen. However, several drawbacks have detracted this material from going mainstream, and some of these shortcomings have been addressed by nanostructuring/nanoconfinement strategies. With the enhancement of thermodynamic and/or kinetic behavior, nanosized complex hydrides (borohydrides and alanates) have recently conquered new estate in the hydrogen storage field. The current review aims to present the most recent results, many of which illustrate the feasibility of using complex hydrides for the generation of molecular hydrogen in conditions suitable for vehicular and stationary applications. Nanostructuring strategies, either in the pristine or nanoconfined state, coupled with a proper catalyst and the choice of host material can potentially yield a robust nanocomposite to reliably produce H₂ in a reversible manner. The key element to tackle for current and future research efforts remains the reproducible means to store H₂, which will build up towards a viable hydrogen economy goal. The most recent trends and future prospects will be presented herein.

1. Introduction

The EU’s ambition to be the first continent that completely replaces fossil fuels with alternative, renewable energy sources appears to face many difficulties. The related cost and scarcity of resources add to the carbon-neutral goal, making it even harder to identify a sustainable means to produce this energy. Renewable energy sources may be tracked to the action of the sun or wind, producing electricity that can be stored (batteries) or utilized to split H₂O in oxygen and hydrogen. Out of the possible alternatives, hydrogen (H₂) has emerged as the source capable to produce energy in an environmentally friendly way, without releasing CO₂ or NO_x gases that are held responsible for the air pollution, smog and greenhouse effect worldwide. Modern trends led to a differentiation of H₂ systems by their source, and nine color codes are actively used today in this purpose: white (naturally occurring hydrogen), red (high temperature, catalytic water splitting using electricity from nuclear plant), pink (electrolysis of water using energy from a nuclear plant), purple (combined chemo thermal electrolysis of water using nuclear power), turquoise (thermal splitting of methane via pyrolysis, C is removed in solid form), brown/black (hydrogen produced from coal: bituminous—black color, and lignite—grey color; both CO and CO₂ gases are released in the atmosphere adding up to current pollution), gray (fossil fuels/steam methane reforming; CO₂ is released in atmosphere), blue (fossil fuel; CO₂ is captured and stored underground, hence blue stands for a carbon-neutral process to produce H₂) and green (water electrolysis, using renewable electricity; no CO₂ is released, water splitting produces only H₂ and O₂, being compliant with an anticipated zero-carbon policy). Hydrogen is the lightest element known (Z = 1) and the most abundant element in the universe, exhibiting unique properties like high diffusivity, low density, low boiling point (−252.8 °C) and high gravimetric energy density—roughly three times higher than that of gasoline.

This review aims to cover the most significant milestones and recent advances in the field of nano-sized complex hydrides, which regained momentum due to the wide interest in shifting towards a carbon-neutral, green alternative fuel: hydrogen [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34]. Complex hydrides have many added benefits compared to other means of storing hydrogen: solid-state storage, high gravimetric and volumetric density, potentially reversible behavior and competent structural and morphology-tuning via nanostructuring (Figure 1).

Various methods have been employed to store hydrogen throughout the years, which may be divided into physical (carbon-based sorbents, cryo-compressed and liquified hydrogen) and chemical storage (metal hydrides, complex hydrides and chemical hydrides) (Figure 1). Storing H₂ in its molecular state presents a number of precautions and is generally regarded as posing safety risks (extreme temperature and high pressure) in addition to the challengingly-high overall cost, hence the solid-state storage options appear especially appealing by comparison. Hydrogen storage density is crucial for vehicular applications and has become a key requirement of US Department of Energy DOE [35]. The current goal of DOE is to obtain a fuel able to store 5.5 wt% H₂ (1.8 kWh/kg, 1.3 kWh/L, costing $9/kWh) by 2025, and ultimately to achieve a hydrogen storage target of 6.5 wt% (2.2 kWh/kg, 1.7 kWh/L, at a cost of $8/kWh). A sustainable hydrogen energy can indeed be built using hydrogen as an energy carrier, but its storage in the solid state constitutes an obstacle (Figure 2) even today. A comparative overview shown in Figure 2 highlights the evolution of gravimetric hydrogen content from various sources H₂ (1 bar, 0.3 g H₂/L). These include lab cylinders (150 bar, 10 g H₂/L), sorbents (MOFs, <70 g H₂/L), liquid H₂@20K (71 g H₂/L), liquid methane CH₄@112K (106 g H₂/L), water H₂O (111 g H₂/L), liquid ammonia NH₃@240K (121 g H₂/L), ammonia borane NH₃BH₃ (70–150 g H₂/L) and complex and interstitial hydrides (110–150 g H₂/L). With gravimetric capacities of up to 150 g H₂/L, complex hydrides surpass the capacity offered by water (111 g H₂/L) and still represent the materials with the highest potential hydrogen storing capacity today.

Moreover, even some of the most promising materials today, complex hydrides, have yet to overcome poor reversibility, sluggish kinetics and high desorption temperatures, in order to become mainstream energy carriers. With high storage capacities (Be(BH₄)₂: 20.7 wt% hydrogen, theoretical), metal borohydrides have emerged as very attractive candidates for solid-state hydrogen storage. Prime examples include LiBH₄ (18.4 wt%) and Mg(BH₄)₂ (14.8 wt%) [4,9]. Still, several factors plague the wide acceptance of complex hydrides as H₂ carriers: high dehydrogenation enthalpies [36,37], slow kinetics and/or side-reactions possibly leading to loss of boron (as borane B₂H₆ or higher boranes) and with it, the concurrent loss of reversibility. In addition, the theoretical hydrogen densities are rarely reached in practice, because complete dehydrogenation almost never occurs, and instead leads to formation of the corresponding metal hydride. Achieving sustained reversibility of complex hydride systems has been at the forefront of research efforts related to hydrogen storage materials and technologies [38,39,40]. Ongoing trends in hydrogen storage materials have expanded their scope for further use as electrolytes for batteries, harnessing the high ion conductivity exhibited by complex borohydrides—a behavior which can be tuned by various chemical strategies, or with temperature [41]. The molecular dynamics of anions and cations in polyborate salts comprise the reorientation of complex anions and translational diffusion of cations and complex anions and are key elements enabling the use of Li⁺ and Na⁺ closo-borates as solid electrolytes [42]. In this regard, nanotechnology has played a key role, enabling consistent improvements across the board regarding thermodynamic and kinetic behavior of complex hydrides, and the current review aims to overview the most important achievements in the field of nanostructured and nanoconfined complex hydrides.

Recent advances in complex hydride chemistry have been supported by NMR studies [42,43], neutron scattering [44], EM electron microscopy (SEM, (HR)TEM) [45], in-situ XRD characterization [46,47] and thermodynamic data predictions brought about by assessing bond energies in B-H systems occurring during decomposition of borohydride species [48]. Further theoretical considerations in BET-isotherms allow for a better description of the porosity features of nano-sized scaffolds used for complex hydride materials [49]. However, not all characterization techniques are truly non-destructive, and certain adjustments of the parameters used are compulsory when dealing with sensitive materials such as hydrides [50]. A note on the importance of rigorous implementation of a hydrogen storage assessment strategy in energy storage compounds is reinforced by the divergent results reported by various groups on similar materials [51].

2. Classes of Complex Hydride Materials

The elemental abundance and ease of synthesis have polarized research in the area of complex hydrides around light metal borohydrides/tetrahydridoborates and alanates/tetrahydridoaluminates [52]. Moreover, alkali-metal complex hydrides (LiBH₄, NaBH₄, KBH₄) serve as starting points for further borohydride synthesis via metathetic reactions, hence intensive studies have been carried out using these components. The improved kinetic behavior of Mg has motivated research on Mg(BH₄)₂ and Mg(AlH₄)₂ complex hydrides. As proof, no less than six polymorphs of Mg(BH₄)₂ have been reported to date, some with unexpectedly high morphological features (like high porosity, pore volume and surface area, γ-Mg(BH₄)₂) [4]. The synthesis of γ-Mg(BH₄)₂ by desolvation from the ethereal adduct indicates that the study of neutral molecule-stabilized adducts of borohydrides (with Et₂O, THF, NH₃, S(CH₃)₂, etc.) are well worth investigation, potentially yielding new, exciting polymorphs of otherwise known complex hydrides. In fact, ammoniates of common borohydrides have been investigated regarding their behavior in a nanoconfined state. Transition metal (TM) borohydrides and mixed (dual cation/anion) borohydrides have provided new tools to lower decomposition temperatures while featuring enhanced reversible behavior in hydrogenation studies. Lastly, many recent studies were devoted to ammonia borane (NH₃BH₃, AB) nanoconfinement in a variety of porous hosts, with promising results reported.

2.1. Li-Based Systems

LiBH₄ and LiAlH₄ have long been pursued as promising hydrogen storage carriers, therefore the wide research output in the literature contains—at least as a starting point or as an intermediate—a Li-based system.

2.1.1. LiBH₄

LiBH₄, the lightest borohydride known, has a very high theoretical storage capacity (18.4 wt%), and yet a high practical storage capacity (13.8 wt%) (Equation (1)) [53,54,55,56]. There are still thermodynamic and kinetic aspects that need to be overcome, and in this respect nanoconfinement became a proverbial tool providing overall noteworthy improvements of nanoconfined LiBH₄ compared to neat lithium borohydride [57]. For instance, the onset desorption temperature (reported for bulk LiBH₄ in the range 370–380 °C) is drastically reduced by ~300 °C through nanoconfinement in CNTs (2 nm; t_onset = 75 °C), or in Cu-MOF (0.9 nm; t_onset = 75 °C) [57]. Interestingly, ¹¹B NMR shifts resulted from first principles calculations could shed more light into mechanistic decomposition pathways of nanoconfined LiBH₄ [58]. Recent synthetic methods aim towards faster and nanosized-oriented syntheses of complex light borohydrides, either by using a surfactant approach [59], or by room temperature precipitation [60]. The effect of dopants (E = Na, K, Al, F, Cl) on hydrogen release energy profiles from E-doped LiBH₄ was modelled using DFT and showed that H₂ release is favored from two neighboring [BH₄⁻] groups rather than a single borohydride anion, and that tensile strain can weaken interaction B-H, effectively facilitating H₂ production [61]. On a similar note, ensemble machine learning protocol recently produced a predictive framework EoE with high accuracy (demonstrated on 3LiBH₄ + AlF₃ + 0.2TiF₃ system, whose predictive behavior resembles well the actual curves for 1st and 2nd dehydrogenation), enabling performance prediction of LiBH₄ with bi-component catalysts [62]. Many research efforts tackle nanoconfined LiBH₄ from a Li-ion conductivity perspective, with direct application in solid electrolytes for batteries [63,64,65,66,67,68,69,70,71,72,73,74,75]. For instance, de Jongh et al. reported nanoconfinement of LiBH₄ in ordered mesoporous silica (42 vol.% silica) with exceptionally high lithium mobility, reaching Li⁺ conductivity of 0.1 mS cm⁻¹ at room temperature [76]. The thermal decomposition pathway of lithium borohydride entails a phase transition at 108–112 °C from o-LiBH₄ to h-LiBH₄ (with high Li⁺ ion conductivity), melting at 275 °C and a possibly stepwise decomposition (Li₂B₁₂H₁₂ as the intermediate) in the range 400–600 °C to finally produce LiH, B and H₂. [1,3,4]:

$$\mathrm{o}-{\mathrm{LiBH}}_{4 \left(\mathrm{s}\right)}\stackrel{108-112 °\mathrm{C}}{\to }\mathrm{h}-{\mathrm{LiBH}}_{4 \left(\mathrm{s}\right)}\stackrel{275 °\mathrm{C}}{\to }{\mathrm{LiBH}}_{4 \left(\mathrm{l}\right)}\stackrel{400-600 °\mathrm{C}}{\to }{ \mathrm{LiH}}_{\left(\mathrm{s}\right)}+{\mathrm{B}}_{\left(\mathrm{s}\right)}+\frac{3}{2}{\mathrm{H}}_2\uparrow$$

(1)

Rehydrogenation requires extreme conditions, an aspect which is in part responsible for the decreased hydrogen storage in subsequent release/uptake cycles (Equation (2)).

$${\mathrm{LiH}}_{\left(\mathrm{s}\right)}+{\mathrm{B}}_{\left(\mathrm{s}\right)}+\frac{3}{2}{\mathrm{H}}_2 \stackrel{~700 °\mathrm{C}, 200{ \mathrm{atm} \mathrm{H}}_2}{\to }{\mathrm{LiBH}}_{4 \left(\mathrm{s}\right)}$$

(2)

Nanoconfinement has led to important achievements in improving reversible behavior of LiBH₄, and a wide range of hosts have been investigated. Li-insertion was reported by Ngene et al. when destabilization of LiBH₄ and LiAlH₄ was achieved due to confinement in HSAG (high surface area graphite) leading to nanostructured complex hydrides able to fully dehydrogenate at 400 °C (Li forms instead of the LiH, which was surprising considering the high thermal stability of LiH which decomposes in bulk at ~900 °C) [77]. The reasoning for such improvement for the Li-intercalation and de-intercalation into turbostratic C nanoscaffolds, forming LiC_x intermediates which were identified by ⁷Li-MAS-NMR (MAS-magic angle spinning) [77].

Ngene et al. have used high surface area graphite (HSAG) after degassing and reducing (10%H₂/Ar, 650 °C, 5 h) as a scaffold for the inclusion of LiBH₄. Desorption was carried out at 400 °C under 1.1 bar H₂ and showed different decomposition for fully confined LiBH₄:

$$12{\mathrm{LiBH}}_4 +60\mathrm{C} \leftrightarrow 10{\mathrm{LiC}}_6+{\mathrm{Li}}_2{\mathrm{B}}_{12}{\mathrm{H}}_{12}+18{\mathrm{H}}_2$$

(3)

In contrast to the nano-sized pathway (3), bulk LiBH₄ follows the full decomposition pathway (4).

LiBH₄ ↔ Li + B + 2H₂

(4)

Rehydrogenation was possible under mild conditions (325 °C, 50 bar H₂). Additionally, LiC_x was identified as a product instead of the usual LiH, due to Li-intercallation in graphitic carbon nanoscaffolds [77].

Using hollow carbon nanospheres, Lai et al. achieved a reduction in plateau pressure upon nanoconfinement of LiBH₄ by solvent impregnation in a 30 wt% loading LiBH₄@HCNWs composite (2–15 nm PSD). The dehydrogenation process had an early onset (t_{onset,release} = 50 °C), which peaked at ~100 °C, although the overall reversible hydrogen content was very modest at ~0.3 wt% [78].

A further modification of carbon-based scaffold was engineered by Gasnier et al., who used TM NP–decorated, N–doped graphene–rich aerogels of high textural properties: V_meso,in = 0.75 cm³/g before decoration and V_meso,f = 0.58 cm³/g after NP decoration (25 wt%, ~5% pore filling with TM = Ni/Co/NiCo) [79]. Metal confinement is more efficient in bigger mesopores, as confirmed by PSD data. A high gravimetric content of 12.2 wt% (des., 400 °C) was recorded for 30 wt%LiBH₄-Co:GN nanocomposite. In the second cycle, the capacity decreased in the following order Ni > NiCo > Co regarding H₂ wt% capacity, which again confirms Ni to be a superior catalyst (Reactions (5) and (6)). Potential physisorption by stacked graphene layers promoted by Ni [0]-desorption starts at ~100 °C [79].

LiBH₄ + Co ↔ CoB + LiH + 1.5H₂

(5)

LiBH₄ + 2Ni ↔ Ni₂B + LiH + 1.5H₂

(6)

HNCs (hollow carbon nanospheres) used for LiBH₄ nanoconfinement afforded a modest ~0.7 wt% partial reversibility, which was achieved at 300 °C and 6 MPa H₂ [80]. The desorption temperature was as low as 100 °C, and solvent infiltration (partially reversible) was found to be superior to melt infiltration (no reversibility for melt–impregnated M(BH₄)_n-HCNs, at 350 °C, 6 MPa H₂). M(BH₄)_x@HCNs nanocomposites had V_pore = 0.03 − 0.12 cm³/g and S_BET = 13.2 − 86.3 m²/g, and Ca(BH₄)₂-based composite showed the best textural parameters. Nanoconfinement seems to reroute dehydrogenation pathway from general Reaction (1) to (7).

$${\mathrm{LiBH}}_{4 \left(\mathrm{s}\right)}\stackrel{\Delta }{\to }\frac{5}{6}{ \mathrm{LiH}}_{\left(\mathrm{s}\right)}+\frac{1}{12}{ \mathrm{Li}}_2{\mathrm{B}}_{12}{\mathrm{H}}_{12 \left(\mathrm{s}\right)}+\frac{13}{12}{ \mathrm{H}}_2\uparrow$$

(7)

LiBH₄ was ground and then melt impregnated into mesoporous carbon hollow spheres (MCHSs) to afford LiBH₄@MCHSs composite/pellets [81]. Wu et al. achieved physical confinement of lithium borohydride in double layers of the carbon nanobowls (DLCB). The overall decomposition conforms to Reaction (1). The recorded improvement was quantified by a strong decrease in activation energy, from 177.1 kJ/mol (neat LiBH₄) to 121.4 kJ/mol (LiBH₄@DLCB-2 nanocomposite), featuring a high volumetric hydrogen density of 82.4 g H₂/L [81].

On the other hand, a similar reduction of E_a was achieved by Zhou et al. who melt-impregnated (MI) LiBH₄ into activated charcoal (AC), reducing the activation energy for desorption to E_a,des = 121.1 kJ/mol for LiBH₄/AC-MI samples that were able to sustain a reversible 6 wt% reversible hydrogen storage content [82].

Employing a high borohydride loading of 70 wt% in porous hollow carbon nanospheres (PHCNSs), Wang et al. achieved a 4.8 wt% reversible H₂ storage, owed in part to the LiBH₄–C surface interactions. The melt infiltration protocol for the most promising sample 60 wt%LiBH₄-40 wt%PHCNSs consisted of rather typical parameters (300 °C, 30 min, 100 bar H₂) [83].

When a carbon wrapped Fe₃O₄ host was loaded with 60 wt% LiBH₄, very promising composite materials 6LiBH₄@4p-Fe₃O₄@C were produced, with an early onset of 175 °C (7.8 wt% H₂, 350 °C, 30 min.) [84]. The activation energy was further reduced compared to previously discussed examples, to E_a,des = 106.2 kJ/mol. Interestingly, although a part of the scaffold plays a sacrificial role (as it reacts with the borohydride source according to Reaction (8)), the formation of iron boride species (Fe₂B, FeB) is regarded as highly beneficial for the reversibility of the system [84].

Fe₃O₄ + 4LiBH₄ → Li₃BO₃ + Fe₂B + FeB + B + 8H₂↑

(8)

Sitthiwet et al. compacted LiBH₄ and electrospun nanofibers of polyacrylonitrile (PAN) catalyzed by Ti(OⁱPr)₄, under high pressure (868 MPa), achieving ~5.2 wt% storage capacity for the resulting composite (ACNF-Ti) [85]. The high pressure applied during the compaction process produced rupture in ACNF-Ti, and subsequent agglomeration of hydrides during cycling (reduced performance in time); using a lower compaction pressure (434 MPa) led to superior kinetic and reversible behavior [85].

A synergy nanocatalyst-nanoconfinement was reported when LiBH₄ (5–10 nm) was introduced in Graphene/Ni nanocrystals (2–4 nm) due to in-situ generation of both LiBH₄ (from the reaction of LiH and C₆H₁₅NBH₃ (9)), and Ni (from the reduction reaction of (C₅H₅)₂Ni (10)) [86].

$$\mathrm{LiH}+{\mathrm{C}}_6{\mathrm{H}}_{15}{\mathrm{NBH}}_3 \to {\mathrm{LiBH}}_4\cdot {\mathrm{C}}_6{\mathrm{H}}_{15}\mathrm{N} \stackrel{\Delta ,\mathrm{vacuum}}{\to }{\mathrm{LiBH}}_4+{\mathrm{C}}_6{\mathrm{H}}_{15}\mathrm{N}$$

(9)

(C₅H₅)₂Ni + H₂ → Ni + 2C₅H₆

(10)

This strategy led to 9.2 wt% hydrogen storage, reversible under moderate conditions (300 °C), confirmed to be non-degrading even after the 100th cycle [86].

Exploring the role of scaffold porosity, Martínez et al. synthesized mesoporous carbon of various PSD (6, 10, 15, 25 nm) and used them to melt impregnate LiBH₄ up to 90 vol% (10, 30, 50, 70, and 90 vol %; 300 °C, 60 bar H₂, 30 min.) in a two-step impregnation procedure. It was observed that smaller pores fill faster, hence the need for a two-step procedure, and that destabilization of LiBH₄ was achieved in the obtained composite, with LiH playing a destabilizing role. However, the reversibility of the system was rather poor [87].

Attempts to prepare nanosized LiBH₄ (50–60 nm) by a single-pot solvothermal process were successful and afforded up to 12.1 wt% reversible behavior at 400 °C. Important thermodynamic parameters were also recorded: E_a.des = 147 kJ/mol (17% improvement over pristine) and E_a,rehyd = 81 kJ/mol (21% reduction relative to pristine) [88].

In a carefully–designed scaffold based on TiO₂-catalyzed carbon, CNT@PC@TiO₂ (CNT, carbon nanotubes; PC, porous carbon) and a high loading of LiBH₄ (60 wt%) produced composite LiBH₄:CNT@PC@TiO₂, releasing 17.7 wt% hydrogen (500 °C). The CNTs facilitate the heat transfer, while the high release of H₂ can be associated with a synergetic effect nanoconfinement—catalysis—surface interaction—thermal transfer [89].

Although moisture sensitive, LiBH₄ was produced by a freeze-drying technique as a monohydride LiBH₄·H₂O that released 10 wt% from 50 °C to 70 °C. The very low onset and dehydrogenation peak were probably due to the strong affinity of H⁺ in the H₂O ligand and H⁻ in the BH₄ group [90]. However, the system LiBH₄.H₂O provides one-way desorption, as reversibility was not achieved (even under 350 bar H₂) [90].

Various other attempts at catalyzing LiBH₄ decomposition have been made. For instance, TiO-catalyzed (using Ti(OEt)₄ precatalyst) lithium borohydride led to 9 wt% reversible hydrogen storage with enhanced kinetics and reversibility (74.4% H₂ storage capacity retention, 10th cycle). The catalyst seems to also inhibit Li₂B₁₂H₁₂ formation [91].

CoNiB-NPs loaded in carbon aerogels (CA) with pore size ~10 nm provided unexpected thermodynamic enhancements by a synergy nanoconfinement-catalysis, affording LiH desorption at 400 °C vs. 950 °C in bulk (14.5 wt% H₂ released vs. 13.8 maximum for reaction step (1)). Metallic Li(0) was confirmed in desorption material (400 °C) by a 54.1 eV peak in XPS spectrum [92].

Suwarno et al. conducted a thermodynamic investigation of LiBH₄ in nanoporous silica and carbon scaffolds by melt infiltration (300 °C, 5°/min, 25 min), after 10 min pre-mixing LiBH₄ with scaffold for proper contact. Two types of LiBH₄ mobile phases by quasi-elastic neutron scattering were observed, and the interfacial layer thickness was concluded to be an essential factor in hydrogen mobility. The interface interaction borohydride-scaffold was also quantified: 0.053 J/m² (with SiO₂) and 0.033 J/m² (with C) [93].

Other scaffolds have been explored as well. Wang el al. employed Ce₂S₃ as additive for lithium borohydride, with LiBH₄ + 20 wt% Ce₃S₃ showing favorable rehydrogenation (up to the fourth cycle; E_a was reduced from 181.80 kJ/mol for pristine LiBH₄ to 151.82 kJ/mol for LiBH₄ + 20 wt% Ce₃S₃) [94]. The reactivity of Ce₂S₃ with LiBH₄, quantifiable via Reaction (11), afforded Li₂S and CeB₆ with co-catalytic dehydrogenation effects.

2LiBH₄ + Ce₂S₃ → Li₂S + 2CeS + 2B + 4H₂

(11)

Activated carbon nanofibers (ACNF) with very high surface areas (S_BET = 2752 m²/g) were used by Plerdsranoy et al. to melt impregnate LiBH₄ (pre-milled in SPEX Mill, 1 h, ball-to-powder BPR 30:1) [95]. The impregnation occurred in stages, hence prolonged time or multiple impregnations could lead to better results, an observation that could be extended to other melt impregnations as well.

The layered structure of 2D Ti₃C₂ utilized as hosts can bypass particle growth/agglomeration, providing excellent destabilization of LiBH₄ [96]. Among the four compositions studied, the activation energy of LiBH₄@xTi₃C₂, corresponding to mass ratios 2:1, 1:1, 1:2 and 1:3, was dependent on the heating profiles: E_a,1 = 98.27 kJ/mol (12 °C/min) and E_a,2 = 94.44 kJ/mol (8 °C/min).

Zero-valent metals like Ni(0) have showed many times more favorable activity in hydrogenation studies. A report from Chen et al. describes the synthesis of nanoporous nickel by dealloying Mn₇₀Ni₃₀ alloy in aq. (NH₄)₂SO₄. LiBH₄ was introduced by wet impregnation, and rehydrogenation was achieved at 450 °C and 8 MPa H₂ [97].

When introduced via ball milling in a Fe₃O₄@rGO support (one-step hydrothermal route), LiBH₄ produced nanocomposites LiBH₄-Fe₃O₄@rGO with ~3.7 wt% hydrogen storage, and a reduced activation energy E_a = 79.78 kJ/mol [98].

Graphene in a mesoporous resorcinol–formaldehyde matrix was used to melt impregnate LiBH₄ (30% pore volume filling), and enhanced [BH₄⁻] mobility under nanoconfined state lowered peak temperatures compared to bulk counterpart. However, the poor reversibility recorded was due to LiH ejection from the pores during cycling [99].

CeF₃-catalyzed activated carbon produced by the ball milling of components was utilized for loading LiBH₄ (90 vol% of scaffold porosity) [100]. The wt% H₂ released by the composite LiBH₄-AC-CeF₃ at 350 °C was 288 times higher than that of neat LiBH₄, and a highly reversible 8.1 wt% was achieved in this catalyzed scaffold.

Wang et al. incorporated LiBH₄ into modified carbon nanotubes: SWCNTs, andMWCNTs by ball milling, thus introducing local disorder and defects in CNTs. As a result, all incorporated LiBH₄ was decomposed after 3 h at 500 °C [101].

Other fluorides were also explored as active catalysts, like NbF₅ in mesoporous carbon MC via grinding (10 min), followed by LiBH₄ infiltration MC-NbF₅ (30 min, 300 °C, 140 bar H₂, 85% vol% optimum filling) [102]. This approach led to a considerable reduction in apparent activation energy to E_a = 97.8 kJ/mol for LiBH₄@MC-NbF₅. Interestingly, nanoconfined samples do not exhibit a melting transition, in line with the disordered state of LiBH₄. The synergistic role of nanoconfinement and nanocatalysis afforded a desorption that was 3.2 times faster in MC-NbF₅ compared to neat MC (Reaction (12)).

2LiBH₄ + NbH_x → 2LiH + NbB₂ + (3 + x/2) H_2; ΔG = −113.97 kJ

(12)

Dolotko et al. utilized SiS₂ scaffolds to produce x LiBH₄–SiS₂ (x = 2–8) composites. Among them, six LiBH₄—SiS₂ was found to be the most promising material. The mixed borohydride decomposed according to (13), affording a reversible H₂ capacity of 2.4 wt% [103].

Li₂SiS₂(BH₄)₂ → 2B + Li₂S + 0.5Si + 0.5SiS₂ + 4H₂; ΔH = 32.5 kJ/mol H₂

(13)

Nanoporous Mg scaffold was employed by Sofianos et al. for the melt-infiltration of LiBH₄, according to the reaction Sequence (14)–(16).

NaH + MgH₂ → NaMgH₃ → Na(l) + Mg + 3/2H₂

(14)

2LiBH₄ + Mg → 2LiH + MgB₂ + 3H₂

(15)

LiH + Mg → LiMg + 1/2H₂

(16)

Notably, Mg did not form MgH₂ during LiBH₄ melt-impregnation (300 °C, 60 bar H₂) [104].

When LiBH₄ was nanoconfined in AlN functionalized with O-atoms containing grafting groups, a frustrated Lewis pair (FLP)–like interaction occurs H^δ+…H^δ−, which in turn enhances the H₂ release [105]. The AlN@LiBH_{4 (}3:2 wt. ratio) yields the most complete desorption (2.8H/1LiBH₄), while the formed borate Li₃BO₃ facilitates the decomposition and formation of [BH₄]⁻ (Reaction (17)).

2LiBH₄ + 6[HO-] → 2[BO₃]³⁻+ 2LiH + 6H₂

(17)

The carbon replica of 2D–ordered mesoporous silicfa MSU-H was loaded by up to 40 wt% LiBH₄ by the incipient wetness method (0.1 M LiBH₄ in tert–butyl–methyl-ether TBME, in 10–40 impregnation-evaporation steps) [106]. However, at lower loading of 8 wt%, LiBH₄ remains completely confined into C-MSU-H mesopores (no XRD peaks belonging to borohydride phase). Palade et al. reported the partial rehydrogenation of LiBH₄-C-MSU-H nanocomposite at 400 °C (100 bar H₂, 2 h) [106]. The same group reported impregnation of LiBH₄ into Mo:MSU-H catalyzed siloxanic materials close to melting (~270–280 °C) exhibiting partially reversible behavior [107].

Vellingiri et al. explored the role of LiB(OH)₄, Li₂CO₃ and LiBO₂ on LiBH₄ @ SWCNTs and concluded that LiBO₂ significantly enhances the dehydrogenation and rehydrogenation, in part due to H⁺ and H⁻ coupling between in situ formed Li⁺[B(OH)₄]⁻, Li₂⁺[CO₃]²⁻ and Li⁺[BH₄]⁻ [108].

Finally, aluminum derived from AlH₃ was used as a scaffold for LiBH₄ producing LiBH₄/Al composite by BM (ball milling). Reversibility was found to depend on the growth of reaction products LiH, AlB₂ and LiAl on the surface of Al* [109] (

Table 1. LiBH₄-based nanostructured systems and specific details regarding hydrogenation storage capacity.

Host Type	Catalyst/Active Intermediate; H2 Storing Composite	wt% H2	Ref.
High surface area graphite HSAG, carbon aerogels CA	LiCx (7Li NMR), Li2B12H12 (11B NMR)	18.5	[77]
Hollow carbon nanospheres—modified by removal of carboxyl/ketone groups (HCNWs)	B, Li2B8H8 and Li2B12H12 (11B NMR)	~0.3 (reversible (0–2 wt% rev. in similar scaffolds)	[78]
N-Doped Graphene-Rich Aerogels Decorated with Nickel and Cobalt Nanoparticles	Ni, Co NPs generate Ni2B and CoB intermediates	8 wt% H2 at 325 °C(Co-NPs); 8 wt% H2 (Ni-NPs decorated CA, after rehydrogenation at 400 °C)	[79]
HNCs spheres (hollow carbon nanospheres)	Li2B12H12 (by FTIR 2470 cm−1 and 756 cm−1)	0.65–0.47 wt% (LiBH4@HNCs)	[80]
Mesoporous carbon hollow spheres (MCHSs) with diameter 300 nm, and nanochannels 2–8 nm	Li2B12H12; No B2H6 in desorption (TPD data).	9.5 wt% DCLB-3 (70 wt% LiBH4, 30 wt% MCHSs); 10.9 wt% DCLB-2 (20 wt% MCHSs). 8.5 wt% reversible capacity (300 °C)	[81]
Activated charcoal (AC)	LiBH4/AC nanocomposite showed no B2H6 release (MS data)	Tonset,des = 190 °C; 13.6 wt% (400 °C); 6 wt% reversible (350 °C, 6 MPa)	[82]
Porous Hollow Carbon Nanospheres (PHCNSs)	Li2B12H12 (FTIR spectra, vibration at ~2490 cm−1)—weak, retarded by nanoconfined systems xLiBH4@yPHCNS (x,y)∈{(5,5);(6,4);(7,3);(8,2)}	8.1 wt% (350 °C,25 min; tonset,des = 200 °C), 4.8 wt% reversible (5th cycle)	[83]
Carbon wrapped ultrafine Fe3O4 skeleton (p-Fe3O4@C)	Li3BO3, FeB, Fe2B and B proposed as intermediates; xLiBH4@yp-Fe3O4@C	7.8 wt% (350 °C, 30 min; tonset = 175 °C, tpeak = 337 °C); 6.2 wt% reversible (20th cycle); 79.4 g/L volumetric hydrogen density	[84]
Electrospun nanofibers of polyacrylonitrile (PAN)-titanium (IV) isopropoxide composite (ACNF-Ti)	TiO2; LiBH4-ACNF-Ti compacted (868 MPa)	5–5.2 wt% (~75% of theoretical for 1:1 LiBH4:ACNF-T; tdes = 352–359 °C)	[85]
LiBH4 (5–10 nm)/Graphene/Ni nanocrystals (2–4 nm)	Li2B12H12, B2H6—are avoided; Li (XPS) slows kinetics above 315 °C, hence 300 °C chosen as maximum; nano-LiBH4/10Ni@20G	9.2 wt% (reversible, 300 °C, 100th cycle); 11.6 wt% (600 °C, tonset = 130 °C).	[86]
Carbon matrix prepared by resorcinol-formaldehyde method (4 types mesopores: 6, 10, 15, 25 nm; 0.82 cm3 = Vt,pores)	LiH + B—negatively impact reversible behavior; LiH-catalytic destabilization of LiBH4	10 wt% (S10/40); poor reversibility; ~12.5 wt% (50% LiBH4 loading)	[87]
LiBH4 50–60 nm prepared/no support	close contact LiH and B crucial for reversibility	12.1 wt% of reversible (400 °C, tonset,des~190 °C; tonset,rehyd~165 °C, 100 bar H2)	[88]
Core–shell structure of CNT@ porous carbon @TiO2 hybrid	TiO2; LiTiO2 and TiB2 destabilize LiBH4	17.7 wt% (500 °C); 7.3 wt% H2 (60 min, 320 °C); 5.1 wt% (20th cycle)	[89]
Nanosheet-like LiBH4·H2O (20–30 nm thick) by freeze-drying technique	H+ from H2O ligand; LiBO2 identified	10 wt% (at 70 °C; tonset,des = 50 °C)	[90]
LiBH4 heated with Ti(OEt)4 (precatalyst)	TiO in-situ introduced into LiBH4, yielding LiBH4-0.06TiO. Li3BO3 and TiH2 produced can act as catalysts for LiBH4	9 wt% (400 °C, 30 min); 9 wt% (reversible, 500 °C, 50 bar H2, tonset,rehyd = 150 °C).	[91]
Carbon aerogels (CA): 9.7 nm, 0.843 cm3/g, 147.1 m2/g	CoNiB-NPs loaded into carbon aerogels (CA), forming LiBH4@CA@CoNiB	15.9 wt% (600 °C); 14.5 wt% (400 °C, 300 min.; tonset = 192 °C, tpeak = 320 °C); 9.33 wt% (350 °C, 30 min)	[92]
Nanoporous silica (3.3, 3.6, 4.2, 4.4, 5.9, 6.1 nm) and carbon scaffolds (3.1, 3.8, 5.6, 20 nm) (comparison)	LiBH4@SiO2 (SBA-15); LiBH4@C	–; (thermodynamc investigation study-phase transition shift in solid-solid phase transition)	[93]
Ce2S3 scaffolds (solvothermal method, 100–200 nm diameter)	Li2S and CeB6 have co-catalytic effects (rehydrogenation) (Reaction (11)). LiBH4 + 20 wt% Ce3S3	4.0 wt% (3000 s, 400 °C; tonset,dehyd = 250 °C); rehydrogenation (400 °C, 5 MPa H2)	[94]
Activated carbon nanofibers (ACNFtt, tt = activation time during heating, 15–75; SBET = 2752 m2/g, Vtot (2.17 mL/g).	B2H6 is suppressed; C surface shows catalytic effect. Li2B12H12 formed (lowers H2 wt% during cycling).	11.7 wt% (1st), 7.1 wt% (2nd cycle); 81% of theoretical H2 capacity (tonset = 125 °C, tpeak = 170 °C)	[95]
Ti2C3 MXene	MXene catalyst; LiBH4@2Ti3C2 hybrid investigated showed best results. Ti(0)//Ti(II)/ Ti(III)/Ti(IV)-TiO2 complex redox speciation of Ti with possible catalytic role (XPS data)	9.6 wt% (380 °C, 1 h; tonset = 172.6 °C); partial reversibility: 6.5 wt%—2nd cycle; 5.5 wt%—3rd cycle (300 °C, 95 bar H2; 48% capacity degradation at 3rd cycle)	[96]
Ni(0)	nanoporous Ni-based alloy (np-Ni); 1 LiBH4/5 np-Ni	11.9 wt% (400 °C; tonset = 70 °C); 8.2 wt% (2nd cycle, tonset = 60 °C).	[97]
Porous graphene support	Fe3O4 nanoclusters; Fe3O4@rGO destabilizer and catalyst precursor; Li3BO3 catalyst formed in situ. LiBH4–20 wt% Fe3O4@rGO investigated.	3.36 wt% (400 °C, 1000s; tonset = 74 °C); 5.74 wt% uptake (400 °C, 5 MPa H2); 3.73 wt% reversible (5th cycle)	[98]
Graphene in a mesoporous resorcinol–formaldehyde matrix (600 m2/g, 6.1 nm, 1.53 cm3/g)	B2H6 avoided (IR); graphene@MC catalytic effect	13 wt% (400 °C; tonset = 253 °C); 6 wt% (2nd cycle); reversible system (rehydrogenation at 400 °C, 5 h, 60 bar H2)	[99]
Activated carbon	LiBH4-AC-CeF3; CeF3—catalyst	13.1 wt% (LiBH4-AC; 8.1 wt% reversible–4th cycle); 12.8 wt% (LiBH4-AC-CeF3; 9.3 wt% reversible–4th cycle) (tonset = 160 °C)	[100]
Modified carbon nanotubes: SWCNTs, MWCNTs	CNTs; LiBH4@SWCNT, LiBH4@SWCNTs (BM), LiBH4@MWCNTs, and LiBH4@MWCNTs (BM)	11 wt% (LiBH4@ MWCNTs, 450 °C); 11 wt% (LiBH4@MWCNTs (BM), 400 °C)	[101]
Highly ordered mesoporous carbon, C-SBA-15 replica, 1321 m2/g; 1.25 cm3/g (with NbF5 NPs)	MC-NbF5 (catalytic effect of 10 wt% precatalyst NbF5); presumed active catalysts: Nb2O5, NbHx, NbB2 (Reaction (12)). LiBH4@MC-NbF5 system investigated	6.52 wt% (200 °C) for LiBH4@MC-NbF5. tonset = 150 °C (LiBH4@MC-NbF5), 205 °C (LiBH4@MC), 282 °C (LiBH4-NbF5-BM); 10.65 wt% rehydrogenation (mild conditions: 200 °C, 60 bar H2).	[102]
SiS2	x LiBH4–SiS2 (x = 2–8) investigated. No B2H6 during release. First principles calculations identified various polymorphs, and point to decomposition Reaction (13).	For x = 6: 8.2 wt% (tonset = 92 °C), 2.4 wt% reversible. For x = 2: 4.3 wt% (385 °C, tonset = 88 °C), 1.5 wt% rehydrog (385 °C, 160 bar H2)	[103]
Porous Mg scaffold (sintering a NaMgH3 pellet sintered at 450 °C under dynamic vacuum, removing molten Na) (14)	Mg-porous (26 m2/g; 1.25 cm3/g); Reactions (15) and (16); MgB2 as destabilizing agent for 2LiBH4-Mg Investigated materials: Mgporous- x wt% LiBH4 (x = 12.78–35.05)	4.81 wt% (2LiBH4:Mg); 7.12 wt% (PMg33); 9.69/9.74 wt% (2 LiBH4:MgH2 powder/pellet)	[104]
Surface-Modified AlN	AlN@LiBH4; AlN synthesized by alcoholysis of LiBH4 (by freeze-drying, AlN contains O–H and C–H groups (Equation (17)), best catalytic effect). x AlBH4: y AlN; (x,y)∈{(1,1);(3,2);(2,1);(4,1)}. Li2B12H12 forms only during 1st cycle, then it decomposes due to AlN. Li3BO3 in situ with potential catalytic effect (IR data: 746, 1265, 1315 cm−1).	5.2 wt% (1:1); 6.8 wt% (3:2); 7.7 wt% (2:1); 8.3 wt% (4:1) for LiBH4-AlN composites. Rehydrogenation at 400 °C, 10 MPa, 24 h, with capacity loss (3.7 wt%). 6.1 wt%—optimized 2:1 composite, stable capacity (400 °C, 10 MPa H2).	[105]
C:MSU-H 2D ordered mesoporous carbon replica	C:MSU-H pores; three loadings studied: 8, 20 and 40 wt% LiBH4@C-MSU-H	1.01 wt% (325 °C, tonset = 150 °C; 92% of theoretical maximum for 8 wt% LiBH4-MSU-H); 2.7 wt% (375 °C, 96.4% of theoretical maximum for 20 wt% LiBH4-C-MSU-H); 0.79 wt% (2nd cycle, 200 °C onset, 8 wt% LiBH4-MSU-H)	[106]
Mo:MSU-H	Molibdate precursor decomposes thermally at 550 °C (4 h) to give active MoO3 catalyst. Mass ration LiBH4:Mo-MSU-H = 5:1 (~70% host pore filling)	11.2 wt% H2 (5.2 wt% rehydrogenation 450 °C, 80 bar H2)	[107]
Single-Walled Carbon Nanotubes (SWCNTs)	LiB(OH)4, Li2CO3 and LiBO2	4.0 wt% (153–368 °C, SWLiB-A); 4.3 wt% (108–433 °C, SWLiB-N); partical reversibility (100–150 °C, 5–10 bar H2)	[108]
Al Derived from AlH3 (Al*) shows superior hydrogenation effect compared to commercial Al.	LiBH4/Al* composite (Equations (18) and (19)): 2LiBH4 + Al → 2LiH + AlB2 + 3H2 (18) LiH + Al → LiAl + 1/2H2 (19). B2H6 and an intriguing “Li-Al-B-H” phase detected.	6.2 wt% (LiBH4/Al*); 5.5 wt% (LiBH4/Al). Reversibility (5.5 wt%) at 400 °C under 8 MPa H2 (Equation (20)): LiH + LiAl + AlB2 + 7/2H2 ↔ 2LiBH4 + 2Al. (20)	[109]

).

2.1.2. LiAlH₄

A closely-related complex hydride is LiAlH₄ has for a while also polarized the research of hydrogen storage materials. It stands, along with LiBH₄ and LiNH₂, as some of the most promising complex hydrides featuring high hydrogen gravimetric content (

Table 2. LiAlH₄-based nanostructured systems and details regarding hydrogenation storage capacity.

Host Type	Catalyst/Active Intermediate; H2 Storing Composite	wt% H2	Ref.
Graphite	LiCx (instead of LiH); LiAlH4/HSAG nanocomposite (NC-HSAG)	10% wt% (300 °C, LiAlH4/HSAG)	[77]
High surface area graphite (HSAG)	Li3AlH6 active intermediate; no other catalyst was used	0.6 wt% partially reversible storage at 300 °C (30 min) (tonset = 135 °C, 7 MPa H2)	[116]
NiCo2O4@rGO	NiCo2O4, rGO	6.28 wt% (tonset = 62.7 °C); 4.0 wt% (isothermal 150 °C, 20 min.)	[117]
N-doped CMK-3 carbon (NCMK-3)	N-sites in LiAlH4@NCMK-3; LiAlH4@CMK-3 synthesized for comparison.	~1.1 wt% (240 °C, tonset = 126 °C); >80% reversibility of LiAlH4 (50 °C, 100 MPa H2)	[118]
(2D) layered Ti3C2	catalytic effect of 2D Ti3C2 in LiAlH4 + 5 wt% Ti3C2, which forms under ball milling active Ti(0) and Ti(+3) catalytic sites (XPS data)	6.5 wt% (58.6 °C onset); 5.5 wt% (200 °C, 35 min); 3.9 wt% (120 °C, 40 min)	[119]
-	SrFe12O19 addition., with active catalyst formed LiFeO2/Sr-phases	5.54 wt% (130 °C, 20 min) for LiAlH4 + 10 wt% SrFe12O19	[120]
-	ZrC powder, x mol% ZrC-doped LiAlH4 (x = 1,2,5,10).	6.61 wt% (for 1 mol%-doped LiAlH4); 5.62 wt% (145 °C, 180 min); 5.48 wt% (130 °C, 180 min); 4.08 wt% (115 °C, 180 min).	[121]
-	2LiBH4-M (M = Al, LiAlH4, Li3AlH6)	8 wt% (2LiBH4-LiAlH4, 150 min)	[122]
-	Fe, LiFeO2 and amorphous Ba or Ba-containing species (XRD, after desorption at 250 °C); LiAlH4-10 wt% BaFe12O19	4.2 wt% (90 °C, 2.5 h); ~6.0 wt% (250 °C, tonset = 95 °C)	[123]
-	K2NbF7 as precatalyst for in situ formed NbF4, LiF, and K-containing species; LiAlH4 + 10 wt% K2NbF7	~ 6.2 wt% (250ׄ°C, tonset ~75 °C); 3.2 wt% (120 min, 90 °C, kinetic study)	[124]
Nickel-Containing Porous Carbon Sheets (Ni-PCS)	Ni, Cporous; LiAlH4-5 wt% Ni-PCS (11.6 µm)	8.14 wt% (LiAlH4 + PCS); 7.97 wt% (LiAlH4 + Ni-PCS)	[125]
TiO2/Hierarchically Porous Carbon	TiO2/Cnanoporous; LAH-TiO+/HPC with postulated role of defect redox sites Ti4+/Ti3+/Ti2+ in 0.62/0.22/0.16 atomic ration (XPS).	6.2 wt% H2 (60 min, 160 °C); 4.3 wt% (40 min, 130 °C; tonset = 64 °C, tpeak = 115 °C); partial rehydrogenation possible (300 °C, 4 MPa)	[126]
Hollow carbon nanospheres (HCNs)	LiAlH4@HCNs	5.8 wt%/expected 2.3 wt% (solvent traces; sharp onset 146 °C; full conversion to LiH, 90 min). 0.37 wt% reversible rehydrogenation to LiAlH4 (150 °C, 8 MPa H2; vs 2.4 wt% max theoretical for used loading).	[127]
-	LiAlH4.xMe2O; adduct formation excludes Li3AlH6 or other intermediates.	N/A; proof-of-concept regarding rehydrogenation potential of LiAlH4.	[128]
-	LiAlH4 · 4THF regeneration; Ti-catalyzed Al (TiCl3).	4.5 wt& regeneration at 398K using Al* [Al(Ti)- 2 mol% Ti]	[129]
-	Surfactant used as stabilizer for NP size manipulation.	–; desolvation from LiAlH4-X-N (X-solvent; N = 0.1, 1, 5, 10)	[59]
FGi (Fluorographite)	LiF (and possibly LiAlF4), Al4C3—active catalyst(s), generated in LiAlH4-xFGi composites	6.25 wt%, with 5.7 wt% (ultra-fast: seconds, tonset = 61.2 °C, 65 °C main step) for LiAlH4-40FGi.	[130]
h-BN	LiAlH4/x wt% h-BN (x = 0,4,14,40). Composites. h-BN has a catalytic effect on Li3AlH6 decomposition (second step desorption).	7.6 wt% (LiAlH4/4 wt% h-BN), 6.8 wt% (LiAlH4/14 wt% h-BN), 4.7 wt% (LiAlH4/40 wt% h-BN).	[131]

). With a melting temperature of 150–175 °C (endothermic process), LiAlH₄ can, in a three-step process, store up to 10.6 wt% and 96.7 g/L hydrogen when heated to 400–440 °C, which is still too high for stationary or mobile applications. In fact, the attainable hydrogen capacity is 7.9 wt% because only the two decomposition steps (Reactions (21) and (22)) of the total of three are thermodynamically accessible under acceptable conditions, and this represents another limitation of LiAlH₄. Moreover, the individual steps in the decomposition process are predicted to be reversible at an unreasonably high H₂ pressure (1000 MPa), which have rendered LiAlH₄ a “one-off” energy storage material. Reviews concerning alanate-based systems and in particular LiAlH₄ have emerged in the literature and discuss various thermodynamic and kinetic aspects of alanate-based systems [110,111,112,113]. The alkaline and alkaline-earth aluminum hydrides experience different crystal structure evolution when heated: alkali tetrahydridoaluminates follow a transformation from [AlH₄]⁻ tetrahedra to isolated [AlH₆]³⁻ octahedra, while alkali-earth counterparts have not been completely investigated. Yet, they decompose at slightly lower temperatures, showing various intermediate structures with chains of corner-shared octahedra, possibly due to their higher coordination number [114]. The catalyst scope utilized for LiAlH₄ has been covered by recent reports [115]. Of particular interest remain destabilization strategies, aimed at lowering energetic barriers and thus lowering the temperature of individual de/re-hydrogenation steps. Nano-sizing alanate systems is an effective means to achieve this goal, and recent advances are summarized in Table 2. Leaching of active hydrogen storage material or of decomposition products outside of nano-porosity of the host can be of concern and could be a reason for degrading performance of complex hydrides with cycling.

Ngene et al. synthesized LiAlH₄/HSAG composites by wet impregnation of LiAlH₄ (0.8 g LiBH₄ in 1 mL dried THF) in graphite, followed by drying under a dynamic vacuum (24 h, 35 °C). The nanocomposite LiAlH₄/HSAG released H₂ in three distinct steps, according to Equation (21): 150–175 °C; (22): 180–220 °C; and (23): 400–420 °C [77].

3LiAlH₄ → Li₃AlH₆ + 2Al + 3H₂ 5.3 wt% H₂

(21)

Li₃AlH₆ → 3LiH + Al + 3/2 H₂ 2.6 wt% H₂

(22)

3LiH + 16C → 3LiC₆ + 3/2H₂ 3.1 wt% H₂

(23)

LiH peaks disappeared from XRD diffractogram, hence full decomposition of LiH to LiC₆ took place (at 300 °C). Non-porous supports (NPG, non-porous graphite) only favor reactions (21) and (22), to stop decomposition at the LiH stage, whereas porous graphite pushed the reaction forward to yield LiC₆ as the final Li-containing product [77].

When nanoconfined in high surface area graphite (HSAG), LiAlH₄ was studied by Wang et al. to provide reasoning for a different (de)hydrogenation pathway for LiAlH₄ (nanoconfined) vs the known dehydrogenation pathway for LiAlH₄ (bulk), following the set of reactions (21), (22) and (24) [116].

LiH + Al → LiAl + ½H₂; 2.6 wt% H₂; ΔH = 140 kJ/mol H₂

(24)

Considering the loading of active alanate species in graphite support, the 0.6 wt% reversible storage accounts for ~30% reversibility of the LiAlH₄ in the LiBH₄@HSAG nanocomposite (3–15 nm). The rehydrogenated material could release H₂ from 100 °C, providing evidence of reversibility via Li₃AlH₆ rehydrogenation—made possible by the nanostructured reaction participants [116].

Xia et al. used NiCo₂O₄@rGO supports to prepare (LiAlH₄ + 7 wt% NiCo₂O₄@rGO) by ball milling [117]. LiAlH₄ (ball-milled) released H₂ with an early onset of 105.5 °C, while (LiAlH₄ + 7 wt% rGO) started to release H₂ at 108.3 °C. Adsorption energy of [AlH₄⁻] at NiCo₂O₄ facilitated dehydrogenation reaction. The active catalyst was synthesized by a co-precipitation technique, from Ni²⁺ and Co²⁺ sources assisted by urea decomposition in water (NH₄⁺, CO₂, HO⁻, Equation (25)):

2NiCo₂(OH)_2x(CO₃²⁻)_(2−x) · nH₂O + O₂ → 2NiCo₂O₄ + (2 − x)CO₂ + (2x + n) H₂O

(25)

Switching to a doped carbon-based support, Cho et al. prepared N-doped CMK-3 carbon (NCMK-3) for LiAlH₄ confinement by solvent infiltration [118]. The LiAlH₄ was freshly recrystallized from Et₂O before impregnation procedure, and the nitrogen N–sites (NCMK-3) were found to be critical for providing binding anchors (AIMD simulations) for metal cation Li⁺, a coordination mode responsible for hypothesized Li-N bond formation. Additionally, no reversibility was observed with undoped CMK-3. Li₃AlH₆ formation is suppressed by combined nanoconfinement and N-doping of the support, providing “thermodynamic stabilization” of metastable metal hydrides. Li₃AlH₆ phase is destabilized by nanosizing (does not form), and an overall Reaction (26) was proposed, supported by in-situ SAXS and WAXS data [118].

LiAlH₄ → LiH + Al+ 3/2H₂

(26)

Two dimensional layered materials of MXene type have also been investigated as supports for LiAlH₄ confinement. LiBH₄ and Ti₃C₂ were mixed by ball milling (Ar, 10 h) and five compositions were investigated in LiAlH₄ + x wt% Ti₃C₂ system (x = 1, 3, 5, 10, and 15). Ti₃C₂ was found responsible for lowered Al-H bond energy in LiAlH₄ and interfacial charge transfer/dehybridization of Al-H, which should facilitate the hydrogen release: E_a,1 = 79.81 kJ/mol (31.3% improvement over pristine LiAlH₄) and E_a,2 = 99.68 kJ/mol (25.1% lower compared to bulk LiAlH₄) [119].

Additives like SrFe₁₂O₁₉ were found to be beneficial for LiAlH₄ dehydrogenation, providing important thermodynamic improvements associated with the first step (21) and second step (22): ΔE_a,1 = 27 kJ/mol and ΔE_a,1 = 15 kJ/mol [120]. The final state of alanate following dehydrogenation comprises LiH and Al phases, as confirmed by XRD data. The dehydrogenation pathway is otherwise unchanged by catalyst addition. The maximum amount of H₂ released was 6.75 wt% and 6.5 wt% (10 wt% and 20 wt% SrFe₁₂O₁₉ doping) [120].

Li et al. introduced ZrC in LiBH₄ by ball milling (Spex, BPR = 20:1, 1200 rpm, 60 min), without ZrC reacting with LiAlH₄ during this process (XRD diffractogram) [121]. Even hand shaking mixing for 30 min caused catalytic effect to manifest when 5 mol% ZrC doping was used. Thermodynamic enhancements and increased surface defects triggered a lowering of E_a for elementary steps for 10 mol% ZrC-doped LiAlH₄ sample, translating in lower onset dehydrogenation temperatures of 85.3 °C (first step, Equation (21), Δ = 90.7 °C relative to pristine) and 148.4 °C (second step, Equation (22), Δ = 57.8 °C relative to pristine). However, reversibility was not achieved (250 °C, 8 MPa H₂) [121].

Other Al-based additives M (M = Al, LiAlH₄, Li₃AlH₆) provided thermodynamic enhancements in 2LiAlH₄-M composites [122]. Dehydrogenation temperature decreases in the order LiBH₄ (469 °C) > 2LiBH₄-Al (445 °C) > 2LiBH₄-LiAlH₄ (435 °C) > 2LiBH₄-Li₃AlH₆ (416 °C). The most promising material among investigated samples was 2LiBH₄-Li₃AlH₆ composite, which released 9.1 wt% H₂ in 150 min. (Figure 3)

A heavier hexaferrite, that of barium BaFe₁₂O₁₉, was used by Sazelee et al. as a catalyst for LiAlH₄ dehydrogenation [123]. Important thermodynamic improvements were registered over the pristine variant of the alanate: E_a1 = 71 kJ/mol (ΔE_a,1 = 32 kJ/mol) and E_a2 = 90 kJ/mol (ΔE_a,2 = 22 kJ/mol) associated with the first (Equation (21)) and second dehydrogenation step (Equation (22)). BaFe₁₂O₁₉ catalyst was introduced via ball milling, a process that reduced LiAlH₄ particle size (from 10–50 µm-as received, to sub-µm), creating high surface defects and grain boundaries [123].

K₂NbF₇ was introduced as an additive in LiAlH4, using planetary ball milling [124]. Using a 10 wt% loading, the nanocomposite LiAlH₄ and 10 wt% K₂NbF₇ showed improved thermodynamics with t₁ = 90 °C, E_a,1 = 80 kJ/mol, ΔE_a,1 = 24 kJ/mol (21) and t₂ = 149 °C; and E_a,2 = 86 kJ/mol and ΔE_a,2 = 26 kJ/mol (22). Desorption from catalyzed sample was 30 times faster than that of pure LiAlH₄ [124].

Yang et al. have investigated three alanates: LiAlH₄, NaAlH₄ and Mg(AlH₄)₂ and drew a relationship between the catalytic effect and the cation’s electronegativity. A stronger catalytic effect (as seen for [Ni]) was recorded when the cation had a lower electronegativity [125]. For instance, the dehydrogenation temperature of NaAlH₄ decreased from 140 to 111 °C. The reduced form of the catalyst, [Ni], was prepared from NiCl₂ and ethylene glycol by sonication and thermal treatment, and was further incorporated in the alanates by ball milling for 90 min. using a BPR = 40:1. The obtained LiAlH₄-5 wt% Ni-PCS composite suppressed the expansion of alanate during decomposition, in addition to the confirmed catalytic activity of [Ni] [125].

Other researchers have employed Ti–based catalysts to enhance dehydrogenation kinetics of alanates [126]. Zhao et al. prepared, by a one–step solvent method, the active catalyst TiO₂/Hierarchically Porous Carbon (HPC), which showed evenly distributed TiO₂ NPs (~10 nm) on honeycomb hollow hemispheres (600 nm diameter). By utilizing a solvent infiltration process, LiAlH₄ was incorporated into TiO₂/HPC support, at various proposed loadings (29, 37, 45 and 55 wt%). The authors observed a synergy TiO₂—HPC, as their composite TiO₂/HPC exhibits superior de-/rehydrogenation catalytic activity than either TiO₂ or HPC. The composition 37 wt% LiAlH₄—25 wt% TiO₂/38 wt% HPC showed the lowest dehydrogenation temperature of 64 °C. The activation energies decreased in the order 63.5 ± 0.4 kJ/mol (HPC) > 57.8 ± 2.1 kJ/mol (TiO₂) > 47.1± 3.5 kJ/mol (TiO₂/HPC), confirming the TiO₂/HPC as the most effective catalyst among the studied examples, and justifying the proposed synergy. The dehydrogenated composite comprises of Al and LiH, without complex alanates (LiAlH₄ or Li₃AlH₆), while the broadened XPS spectra of Ti 2p were suggestive of multiple oxidation states of Ti, which promotes de/rehydrogenation performance of the nanoscaffold [126].

Employing a solvent impregnation method, Pratthana et al. infiltrated LiAlH₄ 1.0 M (Et₂O solution) in hollow carbon nanospheres (HCNs) to prepare nanoconfined LiAlH₄@HCNs [127]. The choice of solvent (Et₂O) was motivated by the better wettability of Et₂O solution surface tension of 0.0165 J·m⁻², compared to HCNs. As a result, the altered thermodynamic pathway was recorded, as described by the suggested Reaction (27) describing also a partial decomposition of LiH.

2LiAlH₄ → 2LiH + 2Al + 3 H₂ → 2LiAl + 1/2 H₂

(27)

The altered mechanism proposed by Equation (27) was consistent with experimental findings showing that only LiH and Al were present in the sample at 150 °C, and thus (27) represents an alternate dehydrogenation pathway [127]. Additionally, the missing intermediate Li₃AlH₆ containing the complex anion [AlH₆]³⁻ may be destabilized at the nanoscale and could split into two H- and [AlH₄]⁻ due to the Jahn-Teller distortion effect [127].

On a related note, adducts of type LiAlH₄.xMe₂O also excluded formation of Li₃AlH₆ [128]. When regeneration of LiAlH4 was solvent-mediated, the reaction was tracked by in situ ²⁷Al and ⁷Li NMR, confirming regeneration at 0 °C (Reaction (28)). The formation of adduct was proven crucial for reversibility. The solvate formation strategy proposed by Humphries et al. was constructed on previous reports where [Al(Ti)] alloying (room temperature, 13 bar H₂) or LiAlH₄.4THF adduct (120 °C, 350 bar H₂) showed regeneration to be feasible [128].

$$\mathrm{LiH}+\mathrm{Al}+3/2{\mathrm{H}}_2 \stackrel{{\mathrm{Me}}_2\mathrm{O}}{\leftrightarrow } {\mathrm{LiAlH}}_4·x{\mathrm{Me}}_2\mathrm{O}$$

(28)

Using TiCl₃ as the catalyst, Graetz et al. synthesized LiAlH₄-2 mol% TiCl₃ composites that achieved regeneration of LiAlH₄ via LiAlH₄ · 4THF, using solvent (THF) adducts from LiH and Ti-catalyzed Al desolvation (Reaction (29)) [129].

$$\mathrm{LiH}+\mathrm{Al}+3/2{\mathrm{H}}_2 \stackrel{\mathrm{THF}}{\leftrightarrow } {\mathrm{LiAlH}}_4·4\mathrm{THF}$$

(29)

In an attempt to obtain alanate NPs, Pratthana et al. used various surfactants for size tuning in the range 10–200 nm [59]. Hexylamine (HXA, 99%), dodecylamine (DDA, ≥99%), octadecylamine (ODA), ≥99%), heptanethiol (HTT, 98%), dodecanethiol (DDT, ≥98%), octadecanethiol (ODT, 98%), tetra-n-butylammonium bromide (TBAB, ≥98%), tetra-n-octylammonium bromide (TOAB, 98%), tetra-n-decylammonium bromide (TDAB, ≥99%) and tridecylic acid (TDA) were investigated. Surfactants have enabled high steric hindrance that can restrict alanate particle size. However, nanosizing and destabilization (even in 2–5 nm porosity), does not enable H₂ release below 100 °C. Among grafted functionalities, the thiol (-SH) and amino (-NH₂) functional groups lowered the H₂ release temperature. LiAlH₄ can reduce various groups (-COOH, -SH), reacting even with resulted –OH groups (Reactions (30)–(32)).

LiAlH₄ + RCOOH → RCH₂OH + LiOH + AlH₃

(30)

LiAlH₄ + 4RSH → LiAlH_4−n(SR)_n + nH₂

(31)

4LiAlH₄ + 12ROH → LiAlH₄ + 3LiAl(OR)₄ + 12H₂

(32)

A fluorographite scaffold (FGi contains 62 wt% F) was used to produce nanocomposites LiAlH₄-xFGi (x = 0, 20, 30 and 40 wt%) by ball milling for 2 h with BPR 40:1 [130]. It was observed that LiAlF₄ may potentially decrease the hydrogen storage capacity of LiAlH₄@FGi, whereas using a 20 wt% FGi loading (t_onset = 103.2 °C) was not enough to stimulate a fast reaction. LiAlH₄.30FGi released 3.23 wt% H₂ at 62.7 °C in seconds. LiAlF₄ also seems to prevail in LiAlH₄.30FGi according to XRD data. Notably, the carbide species Al₄C₃ forms in sample LiAlH₄-40FGi according to Equation (33) [130].

4LiAlH₄ + 4CF → 4LiF + Al₄C₃ + C + 8H₂

(33)

Reaction (33) also reduces E_a from 11.39 kJ/mol to −178.54 kJ/mol, enabling full dehydrogenation at 65 °C [130].

Finally, a hexagonal boron nitride (h-BN) host was used by Nakagawa et al. to produce LiAlH₄/h-BN nanocomposites [131] by ball milling. The role of h-BN was compared to the desorption properties of LiAlH₄/X (X = graphite, LiCl and LiI) composites, and a maximum storage capacity of 7.6 wt% was reported for LiAlH₄/4 wt% h-BN. The Li-ion conductivity enhancement was also suggested for the obtained composites [131].

2.1.3. Li₃AlH₆

Li₃AlH₆ was proposed and confirmed as a key intermediate in thermal decomposition of LiAlH₄ (reactions (21) and (22)). Reports have described the overall improvement in the dehydrogenation of alanates when Li₃AlH₆ was destabilized through catalysis or nanoconfinement, in such a way that the first and second dehydrogenation steps collapsed into a single hydrogen release event [77,116,118,122,123,126,128,131]. In some cases, the intermediacy of Li₃AlH₆ was excluded based on experimental diffraction data [118,128]. In fact, when adducts of LiAlH₄ with various ethers (Me₂O, Et₂O and THF, etc.) were employed to guide a regeneration route of spent alanates, typically, Li₃AlH₆ was excluded from the plausible intermediates [128]. The same is the case when N-doped ordered mesoporous carbon was used as the host, and the formation of Li₃AlH₆ was completely suppressed [118].

However, the synergistic effects of reactive hydride composites considerably improve the behavior of Li₃AlH₆. 2LiBH₄-Li₃AlH₆ composite released 9.1 wt% H₂ in 150 min [122]. Li et al. showed that among investigated composited 2LiAlB₄ + M (M = Al, LiAlH₄ and Li₃AlH₆), the sample 2LiBH₄-Li₃AlH₆ exhibited the best results in hydrogenation studies [122]. The synthesis of Li₃AlH₆ was performed by milling LiH and LiAlH₄ (Reaction (34)) for many hours, with a complete reaction being achieved by the 50th h mark, although clear formation of Li₃AlH₆ was confirmed by XRD even after 20 h of milling (Figure 3) [122].

$$2\mathrm{LiH}+{\mathrm{LiAlH}}_4 \to { \mathrm{Li}}_3{\mathrm{AlH}}_6$$

(34)

By the end of the hydrogen release programme of 2LiBH₄ + Li₃AlH₆, the majority of the phases were identified as AlB₂ and LiH, along with an apparently extraneous peak at about 2θ = 50°, which was not observed from 2LiBH₄ + Al composite dehydrogenation products. The reversibility of the nanocomposite 2LiBH₄ + Li₃AlH₆ still requires further research data [122].

On a similar note, but with a slight variation of the borohydride component, the NaBH₄ + Li₃AlH₆ nanocomposite was studied by Yahya et al. [132] who pointed to thermodynamic destabilization, with best results when a 1:1 molar ratio of complex hydrides was used. Li₃AlH₆ was synthesized by the authors within 12 h using the same Reaction (34). The composite was able to dehydrogenate in two stages (170 °C, Li₃AlH₆ and 400 °C, NaBH₄), and could reabsorb 6.1 wt% at 420 °C and 30 atm H₂ in 60 min [132]. Improved dehydrogenation behavior was due to the formation of Na, Al and AlB₂ which can act as catalysts for dehydrogenation steps. Indeed, the role of Li₃AlH₆ seems to be the destabilization of alkali metal borohydride [132]. It could be seen as a reservoir for active [Al] which results from the decomposition at 170 °C (Reaction (22)).

Indeed, this may react with NaBH₄ according to Reaction (35) to produce the final dehydrogenation phases consisting of Na, AlB₂ and remaining LiH (from Li₃AlH₆ decomposition).

2NaBH₄ + Al → 2Na + AlB₂ + 4H₂

(35)

In contrast to previous rehydrogenation attempts in LiAlH₄ system where 100 atm H₂ and temperatures up to 260 °C were applied, the report from Yahya et al. shows temperature (420 °C) and destabilization (with Li₃AlH₆) to be key factors in achieving reversibility [132].

2.1.4. LiNH₂

Lithium amide (LiNH₂) was used in the recent past as a component of RHC (reactive hydride composite) mixtures, many of which showed remarkable hydrogen storage properties, especially when coupled with main group borohydrides such as Mg(BH₄)₂. However, it was found that LiNH₂ can also enhance Li-ion conductivity in LiBH₄–LiNH₂/metal oxide nanocomposites [133,134,135]. The main reason for observed enhancement was the partial anion substitution of [BH₄]⁻ with [NH₂]⁻, coupled with nanoconfinement in mesoporous oxide scaffolds, which allowed for a bump in ion conductivity from 10⁻⁸ S cm⁻¹ for neat LiBH₄, to 5 × 10⁻⁴ S cm⁻¹ in LiBH₄–LiNH₂/metal oxide composites [135]. It was also inferred from the same study that the porosity of the host was essential for tuning conductivity features of the composite, even more so than the nature of the scaffold or the chemical interactions scaffold-amide/borohydride [133,134,135,136].

Hydrogen storage studies involving LiNH₂ are rather scarce, but some reports refer to an interaction of LiNH₂ with other complex hydrides. and the observed synergies in hydrogenation investigations [137]. Surprisingly, ball milling LiNH₂ and Mg₂FeH₆ produced by a metathesis reaction Li₄FeH₆, (typically obtained from LiH and Fe at 700 °C and 5.5 GPa H₂) (Reaction (36)).

Mg₂FeH₆ + 4LiNH₂ → Li₄FeH₆ + 2Mg(NH₂)₂, ΔH = −92.8 kJ/mol

(36)

The resulting hydride nanocomposite Mg₂FeH₆-4LiNH₂ was subjected to further de/rehydrogenation studies when ~4.8 wt% H₂ (t_onset = 130 °C) was desorbed and 3.7 wt% H₂ was re-absorbed, with slightly worse kinetics than the known and studied Mg(NH₂)₂-2LiH composite [137] (Figure 4). The regenerated form after rehydrogenation of the spent composite comprises of Mg(NH₂)₂, LiH and Fe.

2.1.5. Li-RHC (Reactive Hydride Composite)

Practical implementation of RHC and of hydrides in general is hampered by the high operating temperature and rather low reversibility exhibited by such hydrogen-storing systems [138]. The behavior of RHC (a combination of hydride materials) can be enhanced by choosing proper selection of metal hydrides and complex hydrides, by significant thermodynamic alteration by metathesis reaction—which in turn will speed up kinetics and the reversibility of the RHC system. Collateral improvements were recorded in Li-ion conductivity studies conducted on nanoconfined RHC composites such as the LiBH₄-LiNH₂ system catalyzed by metal oxide NPs [135,139]. For instance, an enhancement of OMS (ordered mesoporous silica, of 1D type MCM-41 or 2D type SBA-15) afforded Li-ion conductivity 10 times higher for LiBH₄-LiNH₂/MCM-41 (1.16 × 10⁻⁴) compared to the nonconfined composite LiBH₄-LiNH₂ [139].

An overview of components of RHCs (

Table 3. Examples of binary and tertiary RHC with corresponding components.

RHC Type	Components	RHC Type	Components
Binary	Ca(BH4)2 + MgH2	Ternary	Ca(BH4)2 + 2LiBH4 + 2MgH2
	2LiBH4-MgH2		NaAlH4-MgH2-LiBH4
	2LiBH4-2.5Mg2NiH4		Ca(BH4)2-LiBH4-MgH2
	Na3AlH6-3MgH2		LiNH2-MgH2-Ca(BH4)2
	2NaAlH4-Ca(BH4)2		LiAlH4-MgH2-LiBH4
	NaBH4-Li3AlH6		LiBH4-CaH2-MgH2
	Ca(BH4)2-Mg(AlH4)2		MgH2-Na3AlH6-LiBH4
	Ca(BH4)2-LiNH2		LiBH4-Mg(NH2)2-LiH
	NaAlH4-Ca(BH4)2		LiNH2-MgH2-LiBH4
	LiBH4-Mg(BH4)2-(TiF3)		LiNH2-LiH-Mg(BH4)2
	Na3AlH6-LiBH4 –(MgFe2O4)
	MgH2-LiAlH4-(SeFe12O19)
	MgH2-NaAlH4-(TiF3)
Ternary	LiBH4-NaBH4-M(BH4)2 M = Mg, Ca	Quaternary	LiBH4-NaBH4-KBH4-Mg(BH4)2
	LiBH4-KBH4-M(BH4)2 M = Mg, Ca		LiBH4-NaBH4-KBH4-Ca(BH4)2
	LiBH4-Mg(BH4)2-Ca(BH4)2		LiBH4-NaBH4-Mg(BH4)2-Ca(BH4)2
	NaBH4-KBH4-M(BH4)2 M = Mg, Ca		LiBH4-KBH4-Mg(BH4)2-Ca(BH4)2
	MBH4-Mg(BH4)2-Ca(BH4)2 M = Na, K		NaBH4-KBH4-Mg(BH4)2-Ca(BH4)2

) also allowed in-depth analysis of some of the chemical reactions involved in RHC—presented in

Table 4. Examples of proposed chemical reactions in binary and tertiary RHCs.

RHC Type	Chemical Reaction(s) Proposed
Binary	Ca(BH4)2 + MgH2 → CaH2 + MgB2 + 4H2	(37)
	Ca(BH4)2 + MgH2 → 2/3CaH2 + 1/3CaB6 + Mg+ 13/3 H2	(38)
	Ca(BH4)2 + MgH2 → CaH2 + 2B + Mg + 3H2	(39)
	2LiBH4 + MgH2 → 2LiBH4 + Mg + H2	(40)
	2LiBH4 + Mg + H2 → 2LiH + MgB2 + 4H2	(41)
	2LiBH4 + 2.5Mg2NiH4 → 2LiH + MgNi2.5B2 + 4MgH2 + 4H2	(42)
	Na3AlH6 + 3MgH2 → 3NaMgH3 + Al + 3/2 H2	(43)
	2NaAlH4 + Ca(BH4)2 → Ca(AlH4)2 + 2NaBH4	(44)
Tertiary	Ca(BH4)2 + 2LiBH4 + 2MgH2 → 1/3CaH2 + 2/3CaB6 + 2LiH + 2Mg + 26/3 H2	(45)

.

While many research results are not always accompanied by a chemical reaction to account for the hydrogen released, there are some examples that justify the use of binary or tertiary systems (Reactions (37)–(45)) (Table 4). In some cases, like for the system 2NaAlH₄ + Ca(BH₄)₂, the intermediacy of intermetallic components or derived complex hydrides (CaAlH₅, CaH₂, Al₄Ca, Al₂Ca) was inferred [138].

Many of these reactions are chain-reactions (one product, such as Al produced in Reaction (43) reacts with one of the reagents—MgH₂—to produce the intermetallic Mg₁₇Al₁₂) releasing considerable amounts of hydrogen (Reaction (46)).

17MgH₂ + 12Al → Mg₁₇Al₁₂ + 17H₂

(46)

By corroborating reported reaction data with hydrogen storing properties (Table 3 and Table 4), it can be observed that Li-based RHC (Li-RHC) are the most widespread RHC investigated to date [138,140,141,142,143,144]. TEM investigations in LiBH₄-MgH₂ RHC catalyzed by 3TiCl₃·AlCl₃ additives revealed MgB₂ platelets (Reactions (40) and (41)) originating from potential nucleation centers including Mg or TiB₂ and AlB₂. The formation and identification of MgB₂ in the reactive mixture pleads for the kinetic improvement brought about by MgB₂ in the RHC [141]. It was also suggested that using additives providing a small atomistic misfit (1.7% relative to MgB₂) could further enhance the behavior in hydrogenation studies [141].

Perhaps one of the first reactions to meet nanoconfined space, the RHC system LiBH₄-MgH₂, was reported by Nielsen et al. in 2010, when using a carbon aerogel of pore size ~21 nm (~722 m²/g, 1.1 cm³/g) [142]. The mixture 2LiBH₄:1MgH₂ follows the decomposition path described by reactions (40) and (41), that can be summed up by the overall Reaction (47).

2LiBH₄ + MgH₂ → 2LiH + MgB₂ + 4H₂

(47)

The reverse Reaction (47) can be regarded as the starting point for Li-RHC systems investigated by several groups [143,144]. Using high energy ball milling (HEBM), the particle size of Li-RHC could be reduced to 10 to 70 μm (15 m²/g, rounded-platelet morphology) [144]. Comprehensive kinetic modelling was performed by Neves et al. using the best-fitting models (JMAEK-Johnson-Mehl-AvramiErofeyev-Kholmogorov with n = 1 and n = 1.5; contracting area model, contracting volume model), revealing that the limiting step is the movement of the not-hydrogenated/hydrogenated material interface [144]. Further nanostructuring can be achieved by ball milling with additives producing active catalytic species such as Li_xTiO₂ and AlTi, which can hydrogenate within 30 min at 400 °C [144].

Polymeric scaffolds including the adaptive poly(4-methyl-1-pentene) (TPX™ Polymer) were employed as hosts for 2LiH + MgB₂ + 7.5(3TiCl₃·AlCl₃) RHCs, and allowed for higher stability in hydrogen storage studies (variation ~0.005 wt% per cycle) [140]. When the C aerogel scaffold was used by Nielsen et al. for 2 LiBH₄: MgH₂ RHC, no toxic B₂H₆ emissions were detectable [142]. About 55% of the free pore volume of the scaffold was infiltrated by RHC, suggesting successive impregnation cycles could increase overall H₂ wt% storage [142].

Gamba et al. utilized 2LiH + MgB₂/2LiBH₄ + MgH₂ RHC for hydrogen purification under a H₂–CO (0.1 mol%) mixture and CO methanation [143]. Inclusion of only 1 mol% TiO₂ additive mechanical milling (2 h, planetary ball mill, 400 rpm, 0.1 MPa Ar) led to nanocomposites Li-RHC-Ti that afforded a stable 10.1 wt% hydrogen capacity after more than 10 a/d cycles [143].

The same composite 2LiH + MgB₂/2LiBH₄ + MgH₂ RHC catalyzed by 5 mol% TiCl₃ was synthesized by Neves et al. by high energy ball milling (5g powder, planetary ball-mill, 20 h, BPR 10:1, 230 rpm, 20% volume filling of vial). Thermodynamic measurements under absorption conditions led to the following parameters for the 1D-interface-controlled Reaction (47): ∆H = 34 ± 2 kJ∙mol⁻¹ H₂, ∆S = 70 ± 3 J∙K⁻¹∙mol⁻¹ H₂, apparent activation energy E_a = 146 ± 3 kJ∙mol⁻¹ H₂ and A = (1.8 ± 1.0) 10⁸ s⁻¹ [144].

The nanocomposite 2LiBH₄-LiAlH₄/RFC (RFC = resorcinol formaldehyde carbon aerogel) was obtained by a two-step melt-infiltration process (0.0595 cm³/g, 45.3 m²/g BET data confirm nanoconfinement inside pores) and showed good H₂ storage properties with 5.7 reversible wt% [145]. AlB₂ formed during the second dehydrogenation step altered the decomposition pathway of LiBH₄, and incomplete rehydrogenation of LiBH₄ was confirmed by the identification of Li₂B₁₂H₁₂ in the FTIR spectrum (vibration peak at 2480 cm⁻¹) [145].

Plerdsranoy et al. prepared a pellet of RHC compressed under 976 MPa (LiBH₄-LiAlH₄ with 1:1 molar ratio), which exhibited good mechanical stability during cycling with 80% of theoretical H₂ capacity compared to the milled sample (65%) [146]. This improvement was also associated with an important reduction in activation energy ΔE_A = 69 kJ/mol H₂ compared to in the milled sample. The support used was a carbonaceous polyacrylonitrile (PAN)-based activated carbon nanofiber ACNF, that was prepared by electrospinning, carbonization and chemical activation ((KOH_aq). Solution impregnation (LiAlH₄ 1 M in Et₂O) and melt impregnation (LiBH₄, 310 °C, 110 bar H₂) took place with a 2:1 weight ratio ACNF: RHC (Reaction (48), 10.12 wt% theoretical H₂ content for milled RHC and 3.37 wt% for RHC-ACNF).

LiAlH₄ (l) + LiBH₄ (l) → 2LiH (s) + 1/2AlB₂(s) + 1/2Al (s) + 3H₂ (g)

(48)

Melt infiltration of 2LiBH₄-NaAlH₄ confined into CA (33% pore volume filling, 310 °C, 30 min, 110 bar H₂) produced a 2.4 wt% reversibility. In fact, 9.52 wt% was released during the first dehydrogenation (3.36%, 1st step; 6.16 wt%, 2nd step), but the formation of the eutectic LiBH₄-NaAlH₄ at 250 °C was not associated with any H₂ release [147]. Furthermore, it was shown that NaBH₄ is the compound from the original RHC that stores H₂ reversibly. The dehydrogenation reaction can be formulated according to (49).

2LiBH₄(s) + NaAlH₄(s) → [LiBH₄(s) + LiAlH₄(s) + NaBH4(s)] → 2LiH(s) + Na(s) + AlB₂(s) + 5H₂(g)

(49)

A LMBH eutectic comprising LiBH₄-Mg(BH₄)₂ in a 55:45 molar ratio was obtained by ball milling (BPR 40:1, 400 rpm, 2 h) [148]. High loading of LMBH in porous hollow carbon nanospheres (HCNS) by over-infiltration (33, 50 and 67 wt% at 190 °C, 60 bar H₂, 1 h) led to LMBH@HCNS nanocomposites. It was observed that an interfacial adhesion effect due to HCNSs avoids borohydride aggregation during de/rehydrogenation, thus having a beneficial effect on cyclability. The optimal composition of the composite was 50LMBH@HCNS, with a 4.5 wt% stable reversible H₂ capacity vs hydride content, or a 2.3 wt% actual cycling capacity (50 wt% LMBH loading in HCNS). The hydrogen release/uptake pathway is described by Reaction (50) for dehydrogenation, and by Reaction (51) for the rehydrogenation reaction [148].

LiBH₄ + Mg(BH₄)₂ → MgB₂ + LiH + B + 5.5H₂

(50)

2LiH + MgB₂ + 4H₂ → 2LiBH₄ + MgH₂

(51)

An interesting strategy for the inclusion of 2LiBH₄–MgH₂ in a porous Ni/C scaffold was proposed by Huang et al. [149]. The host was produced by the pyrolysis of the nickel-based metal organic framework, MOF-74-Ni. The hydrogen capacity of the nanocomposite 2LiBH₄–MgH₂–15%Ni/C was ~9 wt% and MgNi₃B₂ was identified as an active catalyst for the hydrogenation reaction [149].

Dansirima et al. nanoconfined 2LiBH₄-MgH₂ into biomass-derived, activated carbon (AC) to produce LiBH₄-MgH₂-AC composite [150]. LiBH₄-Mg (2:1 molar ratio) were milled (planetary ball mill, BPR 20:1, 10 h, 580 rpm), then milled in a 1:1 AC weight ratio to yield 2LiBH₄-Mg-AC which underwent further hydrogenation of Mg (400 °C, 5 °C/min, 40–50 bar H₂, 10 h), to yield the final composite: 2LiBH₄-MgH₂-AC. The H₂ storing composite “LiBH₄-MgH₂-AC” was used in a small hydrogen storage tank (21.7 mL packing volume), proving the feasibility of using nanoconfined RHC as a viable fuel (58% of the 5.7 wt% theoretical hydrogen achieved: 3.28 wt%, due to temperature gradient and poor hydrogen diffusion through hydride bed). However, the formation of thermally stable Li₂B₁₂H₁₂ limits the reversible H₂ storage capacity, which conforms to the Reaction (51) [150].

The RHC 2LiBH₄-MgH₂ was produced by planetary ball milling (BPR 10:1, 5 h), and infiltrated in. ZrCl₄–doped carbon aerogel scaffold CAS synthesized by the carbonization of cross-linked resorcinol-formaldehyde polymer [151]. The optimum loading was investigated by the melt infiltration technique, which produced composites RHC: ZrCl₄-CAS with weight ratios 1:1 (pore blocking), 1:2 (most suitable) and 1:3 (lower H₂ storage content). The optimum composition 1:2 featured 3.8 wt% theoretical H₂ capacity, and showed 3.7 wt% -1st dehydrogenation, and 3.54–3.45 wt%—second to fourth dehydrogenation cycles. Partial dehydrogenation and formation of [B₁₂H₁₂]²⁻ seemed to the lower overall H₂ storage capacity, which otherwise follows the pathway described by reactions (40) and (41) [151].

Dematteis et al. investigated mixed-cation borohydrides formed in ternary and quaternary systems: KCa(BH₄)₃, LiKMg(BH₄)₄, LiK(BH₄)₂ and also new eutectics. LiBH₄ promoted early H₂ release (~200 °C), while KCa(BH₄)₃ promoted a single-step reaction (higher temperature). The investigated ternary systems were LiBH₄-NaBH₄-Mg(BH₄)₂, LiBH₄-NaBH₄-Ca(BH₄)₂, LiBH₄-KBH₄-Mg(BH₄)₂, LiBH₄-KBH₄-Ca(BH₄)₂, LiBH₄-Mg(BH₄)₂-Ca(BH₄)₂, NaBH₄-KBH₄-Mg(BH₄)₂, NaBH₄-KBH₄-Ca(BH₄)₂, NaBH₄-Mg(BH₄)₂-Ca(BH₄)₂ and KBH₄-Mg(BH₄)₂-Ca(BH₄)₂. Whereas, the quaternary systems comprise LiBH₄-NaBH₄-KBH₄-Mg(BH₄)₂, LiBH₄-NaBH₄-KBH₄-Ca(BH₄)₂, LiBH₄-NaBH₄-Mg(BH₄)₂-Ca(BH₄)₂, LiBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂ and NaBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂ (Table 3) [152].

An investigation of new mutually destabilized reactive hydride system LiBH₄–Mg₂NiH₄. Mg₂NiH₄ and MgNi_2.5B₂ prepared in-house by Bergemann et al. revealed MgNi_2.5B₂ as an active intermediate, which was confirmed by ¹¹B MAS NMR (154 ppm) [153]. MgO (<5 wt%) was present after dehydrogenation in the samples, which could be a result of the high affinity of Mg for oxygen, and points to unavoidable side reactions during sample manipulation. The dehydrogenation of RHC follows Reaction (42) which could hypothetically consist of three individual steps [153].

2LiBH₄ + 2.5Mg₂NiH₄ → (2LiH + 2B +2.5Mg₂Ni +8H₂ → 2LiH + MgNi_2.5B₂ + 4Mg + 8H₂) → 2LiH + MgNi_2.5B₂ + 4MgH₂ + 4H₂

(42)

A TiCl₄-catalyzed CAS prepared by resorcinol-formaldehyde (RF) aerogels technique was used to yield composites 2LiBH₄–MgH₂–0.13TiCl₄. The RHC was confined by solution impregnation and melt infiltration in nanoporous catalyzed carbon aerogel host TiCl₄-CAS [154]. The superior behavior of TiCl₄-catalyzed RHC was due to the formation of Ti–MgH₂ alloys (Mg_0.25Ti_0.75H₂ and Mg₆TiH₂) during the first rehydrogenation. The dehydrogenation proceeded in three steps (140, 240 and 380 °C), with no detectable traces of B₂H₆ [154].

Other investigations started from neat RHC LiBH₄-NaBH₄ which was synthesized by ball milling (BPR 30:1, 120 min, 350 rpm) leading to formation of a solid solution. Notably, the orthorhombic-to-hexagonal LiBH₄ transition took place at 94 °C (15 °C lower than that of neat LiBH₄). A new eutectic with a composition between Li_0.65Na_0.35BH₄ and Li_0.70Na_0.30BH₄ was observed (mp = 216 °C), and ab initio and Calphad allowed the calculation of thermodynamic parameters and the phase diagram [155].

Liu et al. synthesized LiBH₄–Mg(BH₄)₂@NPC nanocomposites by melt infiltration (20 wt% hydride loading vs support; 200 °C, 60 bar H₂, 30 min.) [156]. The mixed borohydride Li/Mg(BH₄)₃ formed an eutectic with a structure similar to α-Mg(BH₄)₂, with varying diborane (B₂H₆) and triborane (B₃H₈) evolution—lower for bulk, higher for nanoconfined, confirming a suppressed reaction pathway and altered decomposition mechanism [156].

The 0.62LiBH₄-0.38NaBH₄ RHC catalyzed by nano-Ni was prepared by planetary ball milling (10 h, 1 bar Ar, 175 rpm) [157]. Nano Ni-addition destabilizes the dehydrogenation process (a three-step reaction, 20–25 °C reduction in peak temperatures compared to neat RHC). The cycling stability was unaffected by the nickel catalyst, which facilitates regeneration of LiBH₄. The dehydrogenation products were identified as Ni₄B₃ (first step) and Li_1.2Ni_2.5B₂ (third step). The possible reactions of LiBH₄ with Ni depend on the initial LiBH₄:Ni molar ratio conforming to Equations (52)–(54), and comprise various nickel boride formulations as catalyst-containing species [157].

LiBH₄ + 4/3Ni → 1/3Ni₄B₃ + LiH + 3/2H₂

(52)

LiBH₄ + 2Ni → Ni₂B + LiH + 3/2H₂

(53)

LiBH₄ + 3Ni → Ni₃B + LiH + 3/2H₂

(54)

Additives M in the RHC of type (2LiBH₄ + M) (M = LiAlH₄, Li₃AlH₆) were utilized to destabilize LiBH₄, lowering its mp to 445 °C and 435 °C, respectively [122]. The Li₃AlH₆ was synthesized from LiH and LiAlH₄ by ball milling (50 h, 300 rpm, BPR 25:1), and was part of the most promising sample, the 2LiBH₄-Li₃AlH₆ RHC, which featured a two-step release (3 and 6.1 wt% H₂) forming AlB₂ for a potentially reversible system [122].

The interfacial interaction of LiBH₄ in LiBH₄–Ca(BH₄)₂ RHCs (4:1, 2.1:1 -eutectic composition) with the surface of mesoporous silica (MCM-41 or SBA-15) revealed the liquid-like behavior starting at 95 °C when nanoconfined, which led to high diffusional mobility. Notably, utilizing (VT) NMR the authors revealed the co-infiltration of eutectic LiBH₄–Ca(BH₄)₂ (hand or ball milling) into mesopores far below the eutectic melting point [158].

When introduced via mechanical milling, 2D-MXene (Ti₃C₂) has served as a support for 4MgH₂-LiAlH₄ RHC [159]. Chen et al. obtained the composite 4MgH₂-LiAlH₄-Ti₃C₂ which showed improved de/re-hydrogenation kinetics (ΔT = 64 K lower than milled 4MgH₂-LiAlH₄ RHC). Additionally, thermodynamic parameters for the three-stage dehydrogenation were also deduced (E_A = 65.9 kJ mol⁻¹ H₂⁻¹, 70.6 kJ mol⁻¹ H₂⁻¹ and 74.3 kJ mol⁻¹ H₂⁻¹, respectively) [159].

The LiBH₄/KBH₄ mixed borohydride system was hand-ground and used as a eutectic electrolyte (0.725LiBH₄/0.275KBH₄; ~50% of whole RHC system was the hydride content) to facilitate dehydrogenation/rehydrogenation of theSn-catalyzed MgH₂ and 0.5Sn system (ball milled; 2.3 wt% theoretical H₂). Sn provides destabilization by the formation of Mg₂Sn, thus lowering ΔH_{dehydrogenation} (Reaction (55)).

2MgH₂ + Sn ↔ Mg₂Sn + 2H₂

(55)

Kinetic improvements account for a 10-fold increase in reaction rate of both de-and rehydrogenation. E_A was also decreased from ~150 kJ/mol to ~100 kJ/mol. Notably, the working regime imposed a restricted temperature profile, below mp (Sn) = 232 °C [160].

Using a Ti-based catalyst (TiF₃, Ti and TiO₂), Ma et al. showed the strong influence of titanium over dehydrogenation temperature and kinetics of the 4LiAlH₄–Mg₂NiH₄ system [161]. A low activation energy of E_{A, desorption}= 81.56 kJ mol⁻¹ was deduced based on the Johnson-Mehl-Avrami model, and the most effective form of the catalyst was TiF₃. Mg₂NiH₄ catalyzes the decomposition of LiAlH₄, and it further accelerates dehydrogenation by reaction with in-situ formed Al (reactions (21), (22) and (56)).

Al + Mg₂NiH₄ → Mg₁₇Al₁₂ + Al₃Ni + Al_1.1Ni_0.9 + H₂ (220~330 °C)

(56)

Mesoporous carbon (C-replica of SBA-15, namely CMK-3) was used as a host for the x LiBH₄– (1 − x) Ca(BH₄)₂ system (x = 0.50, 0.60, 0.65, 0.68, 0.70, 0.75, and 0.8) [162]. The mixed borohydride was introduced in CMK-3 via melt infiltration (230 °C, 30 min, 3 bar H₂) producing RHC@CMK-3 when mixed in a 1:1 weight ratio. Half of the initial dehydrogenation content was recovered after rehydrogenation, pointing to a synergy nanoconfinement-mutually-destabilized system. Analysis by XRD data of reaction products revealed Ca₃(BH₄)₃(BO₃) and LiCa₃(BH₄)(BO₃)₂ as a result of [BH₄]⁻ oxidation, presumed to form outside the mesopores, due to unit cell size mismatch with a mesoporous channel diameter. The rehydrogenated material comprised LiCa₃(BH₄)(BO₃)₂ (ex situ XRD) [162].

Nanoconfinement of 0.55LiBH₄–0.45Mg(BH₄)₂ system in activated carbon aerogel via melt infiltration was shown to facilitate both hydrogen release and uptake, while also maintaining a high H₂ release content of 8.3 wt% in RHC@CA after the fourth hydrogen release [163]. A clear role in the high reversible H₂ content was attributed to the high porosity and pore volume (689–2660 m²/g, 1.21–3.13 cm³/g) of the activated scaffold [163].

MWCNTs were utilized by Meethom et al. to confine LiBH₄–LiAlH₄ by ball milling [164]. The MWCNTs (5 wt%) provided enhanced thermal conductivity and a high surface area for reactive contact between Al and LiBH₄/LiH, leading to a reversible in H₂ storage with faster kinetics (three times faster) and reduced hydrogen release onset by ΔT = −120 °C. The quenching of the first desorption step (220 °C) was introduced for the first time, and among the reaction products, AlB₂ was identified to form during reaction of LiAlH₄ with Al (Equation (18)).

Other techniques like vacuum vapor deposition produced Mg@NaBH₄/MgB₂ core-shell structures affording ~6 wt% hydrogen storage [165]. Mg@NaBH₄ displayed better hydrogenation properties to pristine Mg, while nanosized MgB₂ produced NPs dispersed on Mg enhance sorption kinetics (E_A = 60.1 kJ/mol H₂, ΔH_ab = −73.8 kJ/mol H₂, ΔH_des. = 89.7 kJ/mol H₂, dehydrogenation onset of 245 °C was lower than for MgH₂). The reaction mechanism confirms the intermediacy of MgH₂-NaBH₄ RHC, and conforms to Equation (57):

2MgH₂ + 2NaBH₄ → Mg + MgB₂ + 6H₂ + 2Na

(57)

By ball milling, 4MgH₂-LiAlH₄-10 wt%TiO₂ composite was obtained (1 h, 400 rpm, Ar) and showed that the TiO₂ destabilized RHC-TiO₂ compared to the neat RHC (lowered onset of the first step from 100 °C to 70 °C, and lowered the second step from 270 °C to 200 °C). The activation energy was also reduced from 133.3 (4MgH₂-LiAlH₄) to 102.5 kJ/mol (4MgH₂-LiAlH₄-TiO₂) [166].

Composites LiBH₄- x AlH₃ (x = 0.5, 1.0, 2.0) introduced by Liu et al. [167] showed enhanced destabilization as [AlH₃] increased in RHCs prepared by ball milling (proposed kinetic model: JMA) with E_{A, RHC desorption}~122.0 kJ/mol, while E_{A, LIBH4} = 169.8 kJ/mol. Dehydrogenation seems to be controlled by AlB₂ intermediate and its increased nucleation rate. The simple binary hydride AlH₃ decomposes in the first step into H₂ and Al, which is hypothesized to partake in the second dehydrogenation step (Reaction (58)). The nature of the “Li-Al-B” product was unclear [167].

LiBH₄ + Al →’Li-Al-B’ + AlB₂ + H₂

(58)

The reactive mixture Mg(NH₂)₂–2LiH–0.07KOH was prepared by Chen et al. and yielded a new RHC containing Li₃K(NH₂)₄–MgNH–LiNH₂, with a decomposition dependent on H₂ pressure [168]. The KOH role was highlighted in the Mg(NH₂)₂–2LiH RHC (90 °C theoretical desorption temperature from thermodynamic data, but ~220 °C in practice due to slow kinetics), which follows a reversible pathway described by (59):

$${\mathrm{Mg}\left({\mathrm{NH}}_2\right)}_2 +2\mathrm{LiH} \leftrightarrow 1/2{\mathrm{Li}}_2{\mathrm{Mg}}_2 {\left(\mathrm{NH}\right)}_3 +1/2{\mathrm{LiNH}}_2 +1/2\mathrm{LiH}+3/2{\mathrm{H}}_2 \leftrightarrow {\mathrm{Li}}_2{\mathrm{Mg(NH)}}_2+2{\mathrm{H}}_2$$

(59)

Reversibility and enhanced kinetics were reported for the RHC 2LiBH₄–NaAlH₄ melt infiltrated into CAS at 185 °C (mp_NaBH4) and 310 °C(mp_LiBH4), using weight ratios CAS:RHC of 2:1, 2:1.5 and 1:1 [169]. The composite 2:1 showed the best results storing ~2.4 wt% hydrogen up to 400 °C. Nanoconfinement by melt impregnation in CAS (resorcinol–formaldehyde aerogels) inhibits Na evaporation and promotes NaBH₄ regeneration, which was impossible in as milled RHC due to the loss of Na source (Reaction (60)).

LiH + Al + 2NaH + 3/2H₂ ↔ LiNa₂AlH₆

(60)

In a report by Peru et al. [170], a 5 nm pore-size CMK-3 carbon was utilized to infiltrate a eutectic mixture 0.725 LiBH₄–0.275 KBH₄ resulting in 2.5–3 wt% reversible hydrogen storage due to a favorable synergy C_surface—nano-sized MBH₄ (M = Li, K), which helped circumventing or reducing irreversible side-reactions [170].

More exotic borohydrides of lanthanides were also investigated. For instance, the mixture 3LiBH₄ and Er(BH₄)₃ and 3LiH was shown to desorb up to 3.5 wt% after 3 a/d cycles [171]. LiBH₄ was generated in-situ via metathesis, and its high temperature form (h-LiBH₄) was stabilized in RHC at 40–60 K (61).

Er(BH₄)3(s) + 3LiH(s) → ErH₂(s) + 3LiBH₄(s) + 0.5H₂(g)

(61)

The desorption (62)—resorption (63) are similar transformations, which also imply a redox behavior of the Er(II/III) center [171].

4LiBH₄(s) + ErH₂(s) → ErB₄(s) + 4LiH(s) + 7H₂(g)

(62)

ErB₄(s) + 4LiH(s) + 7.5H₂(g) → 4LiBH₄(s) + ErH₃(s)

(63)

When confined in activated carbon nanofiber ACNFs, the 2LiBH₄-MgH₂ RHC produced 2LiBH₄-MgH₂ -30 wt% ACNFs composites storing up to 4.5 wt% hydrogen [172]. The dehydrogenation onset was reduced by 50 °C (from 350 °C to 300 °C) due to doping with ACNFs and subsequent compaction to pellet (under 600 MPa). The activation energy E_A was reduced by 67 kJ/mol, and H₂ permeability was increased in compacted RHC@ACNFs (10 times), while the heat transfer was also increased (1.5 times). ACNFs alter the dehydrogenation pathway, from a one-step process (neat RHC: 2LiBH₄-MgH₂) to a two-step process (RHC@ACNFs). H₂ desorbs from both individual hydrides (yielding Mg and LiH), rather than reacting according to the neat 2LiBH₄-MgH₂ (Equations (40) and (41)) [172].

The RHC K₂Mn(NH₂)₄–8LiH was recently investigated [173]. K₂Mn(NH₂)₄ was synthesized by ball milling metallic 1 Mn: 2 K under seven bar NH₃, and activated by ball milling (12 h, 200 rpm, BPR 40:1, 10 bar H₂). Final dehydrogenation products contain Li₂NH, Mn₃N₂ and MnN, while rehydrogenation mixture contains LiH, LiNH₂ and other “K-Mn-species”. Notably, the activation energy E_A = 65 kJ/mol (K₂Mn(NH₂)₄–8LiH) is comparable with E_A = 66 kJ/mol for the more mainstream RHC: LiNH₂-2LiH [173].

Javadian et al. utilized the eutectic mixture 0.62LiBH₄–0.38NaBH₄produced by melt infiltration composites RHC@CA (CA = CO₂-activated resorcinol-formaldehyde carbon aerogel host with 37–38 nm pores and S_BET = 690–2358 m²/g) [174]. The authors recorded important thermodynamic enhancement (ΔT = −100 °C of nanoconfined vs. neat RHC), while cycling capacity was three times higher. The activation energy E_A was decreased from 139 kJ/mol (bulk) to 116–118 kJ/mol (C aerogel nanoconfined), with 4.3 wt% hydrogen storage (0.7 of initial) after four a/d cycles. By contrast, the overall reversibility of neat 0.62LiBH₄–0.38NaBH₄ was poor with 1.6 wt% (0.22 of theoretical) after four a/d cycles [174]. Another example of an eutectic mixture of borohydride comes from the same group, when 0.7LiBH₄–0.3Ca(BH₄)₂ was infiltrated into activated carbon aerogel (pristine: 689 m²/g, 1.21 cm³/g and CO₂-activated: 2660 m²/g, 3.13 cm³/g) filling the pores up to 60 vol% [175]. Thermodynamic improvements recorded are a clear reduction of dehydrogenation temperature ΔT = −83 °C (CA-RHC; 156 kJ/mol) and ΔT = −95 °C (CA_CO2-RHC, E_A = 130 kJ/mol) compared to neat RHC (E_A = 204 kJ/mol). CO₂-activated CA produces less borates and oxides, highlighting the narrow pore’s role in maintaining stability and altering the thermodynamic pathway in RHCs. Reversible behavior is reasonably well described by Reaction (64).

4LiBH₄ + Ca(BH₄)₂ ↔ 4LiH + CaB₆ + 10H₂

(64)

Taking advantage of the investigation possibilities offered when labelled borohydrides are used, in 2012 Sartori et al. described the melt infiltration of Li¹¹BD₄–Mg(¹¹BD₄)₂ into carbon nanoscaffolds (IRH33, 1.17 cm³g⁻¹, 2587 m² g⁻¹_, 0.5 to 4.5 nm pores) yielding Li¹¹BD₄–Mg(¹¹BD₄)₂/IRH33 composites (190 °C, 40 bar D₂, 1 h) [176]. By-products (like the toxic B₂D₆) and poorly regenerative dodecaborane and [B₁₂D₁₂]²⁻ were avoided by nanoconfinement (no signals in the expected region of ¹¹B NMR), while also reducing the desorption temperature by ΔT = −60 °C.

A proof of thermodynamic destabilization of NaBH₄ by Li₃AlH₆ was shown in NaBH₄–Li₃AlH₆ composite [132]. Na, Al and AlB₂ species formed during dehydrogenation were key to the observed improvement. The activation energies recorded were reduced to E_A,1 = 162.1 kJ/mol and E_A,2 = 68.1 kJ/mol for NaBH₄ decomposition [132].

The predicted hydrogen release in RHC do not usually match actual experimental data. For instance, although predicted to release 4.2 wt% H₂ at 64 °C and 1 bar H₂, the 6Mg(NH₂)₂–9LiH–LiBH₄ RHC faces kinetic barriers, and YCl₃/Li₃N additives were used to alter the kinetics [177].

Other TM salt additives were used in more widespread RHC. The 2LiH and MgB₂/2LiBH₄ + MgH₂ system was catalyzed with 0.05 TiCl₃ and used as an additive [144]. The following thermodynamic parameters were determined: ΔH = −34 ± 2 kJ∙mol H₂⁻¹ and entropy ΔS = −70 ± 3 J∙K⁻¹∙mol H₂⁻¹_; while the apparent E_A= 146 ± 3 kJ∙mol H₂⁻¹ and the Arrhenius pre–exponential factor was A= (1.8 ± 1.0) 10⁸ s⁻¹ [144].

Another report of using TiCl₃ for the 2LiBH₄–MgH₂ RHC’s inclusion in the porosity of resorcinol–formaldehyde carbon aerogel scaffold (RF–CAS) comes from Gosalawit-Utke et al. [178]. The carbon scaffold RF-CAS was decorated with TiCl₃ (1.6 wt%) by solution impregnation, and after melt impregnation of RHC produced composites 2LiBH₄–MgH₂–TiCl₃@RF-CAS capable to release 3.6 wt% hydrogen during the fourth a/d cycle [178]. The kinetics were twice as fast relative to undoped RHC, due to TiCl₃ doping. The 2LiBH₄–MgH₂–TiCl₃ system proved to be reversible, and regeneration of LiBH₄ and MgH₂ was assessed by FTIR and SR-PXD data, overall conforming to the Equations (40) and (41).

The reversible behavior of the 2LiBH₄-LiAlH₄ system was assessed by nanoconfinement in mesoporous carbon (MC) scaffolds, with ~8.5 wt% reversibility confirmed by seven a/d cycles (Figure 5) [179].

The dehydrogenation events in 2LiBH₄-LiAlH₄@MC were identified at 80 °C (ΔT = −40 °C) and 230 °C (ΔT = −145 °C), considerably lower than the milled 2LiBH₄-LiAlH₄. The favorable synergy nanoconfinement–thermodynamic destabilization suppressed B₂H₆ emissions, forming catalytically active AlB₂ instead of Li₂B₁₂H₁₂. The decomposition conforms to Equation (65) (Figure 6) [179].

2LiBH₄ + LiAlH₄ ↔ 3LiH + AlB₂ + 9/2H₂

(65)

As oftentimes seen in XRD diffractogram, metal borides arise as potential intermediates in dehydrogenation studies. Hence, their effect was checked by independently preparing Fe₃B–catalyzed LiBH₄-MgH₂ RHCs soring ~2.9 wt% hydrogen after the seventh a/d cycle, with the main release step occurring below 265 °C [180]. A new processing method, Ball Milling with Aerosol Spraying (BMAS) was introduced by Ding et al., producing Mg(BH₄)₂ at RT from (nano-) MgH₂ and LiBH₄ [180].

Other investigations expanded on the multi-borohydride systems with an example of quinary RHC, namely LiBH₄-NaBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂ [181]. The equimolar composition of complex borohydrides was produced by a planetary ball mill (10 bar H₂, 10 mm diameter stainless steel balls and BPR 30:1, 1–50 h, 350 rpm), yet even after 50 h of milling no miscibility was obtained. The only phase formed upon milling was KCa(BH₄)₃, while dehydrogenation occurred from the liquid phase as a complex multi-step process (in-situ SR PXRD, Synchrotron Radiation Powder X-ray Diffraction). This was the first report of five-component liquid borohydride obtained via the eutectic approach [181].

Another eutectic mixture 0.68LiBH₄–0.32Ca(BH₄)₂ (“LiCa” eutectic) was nanoconfined in the porosity of a carbon aerogel scaffold CAS (S_BET = 2421 ± 189 m²/g, V_tot = 2.46 ± 0.46 mL/g, 13 nm pore size) [182]. The as-synthesized nanocomposite had a maximum theoretical H₂ capacity of 14.34 wt%, when Reaction (64)) occurred via melt infiltration. It seemed that the LiBH₄ component was affected only during the first a/d cycle by CAS confinement (ΔT = −40 °C), with diminishing returns after four cycles (ΔT = −10 °C). LiBH₄@CAS continuously loses H₂ capacity after the second cycle, but “LiCa”@CAS and neat “LiCa” retain stability during cycling up to the seventh cycle. Ca(BH₄)₂ decomposition products (CaO, CaH₂ and/or CaB₆) are believed to facilitate full LiBH₄ reversibility [182].

Due to its high gravimetric capacity, Ca(BH₄)₂ was investigated in other RHCs as well, for instance in 2NaAlH₄ + Ca(BH₄)₂ with 5 wt% TiF₃ as additive [183]. Mustafa et al. described TiF₃-doped RHC (ball milling, 6 h, 400 rpm, BPR 40:1), converting the initial RHC_i: NaAlH₄–Ca(BH₄)₂ system into RHC_f: Ca(AlH₄)₂–NaBH₄ (Reaction (44)). The intermediate phases in (2NaAlH₄-Ca(BH₄)₂)-to-(Ca(AlH₄)₂-NaBH₄) transition were identified as CaAlH₅ (Reaction (66)) and CaH₂ (Reaction (67)).

Ca(AlH₄)₂ → CaAlH₅ + Al + 3/2H₂

(66)

3Ca(AlH₄)₂ + 2TiF₃ → 3CaF₂ + 2Al₃Ti + 12H₂

(67)

The a/d of RHC-TiF₃ was significantly improved by TiF₃, with a significantly decreased desorption onset (125 °C to 60 °C for 1st stage), and corresponding similar reduction of activation energies of CaAlH₅ (79.3 kJ/mol; ΔE_A = −63.6 kJ/mol) and NaBH₄ (124.6 kJ/mol; ΔE_A = −21.9 kJ/mol). The complex hydride CaAlH₅ produced in (66) would decompose to generate CaH₂ and Al (Reaction (68)).

CaAlH₅ → CaH₂ + Al + 3/2H₂

(68)

Notably, CaH₂ plays a critical role because it will generate Al-Ca alloys by reaction with Al (Reaction (69)).

CaH₂ + 3Al → 1/2Al₄Ca + 1/2Al₂Ca + H₂

(69)

The two formulations of Al-CA alloys further react with the more thermodynamically stable NaBH₄ (Reactions (70) and (71)), again producing active boride species (CaB₆, AlB₂).

14NaBH₄ + Al₄Ca → 14Na + CaB₆ + 4AlB₂ + 28H₂

(70)

10NaBH₄ + Al₂Ca → 10Na + CaB₆ + 2AlB₂ + 20H₂

(71)

When 2D materials of the MXene type were introduced by ball milling as nanoadditives (1, 3, 5 and 7 wt%) in the 2LiH and MgB₂ system, RHCs 2LiBH₄ and MgH₂ were produced [184]. The source of MgH₂ regeneration was discussed by the authors in the context of TiB₂ formed during dehydrogenation, which functions as heterogeneous nucleation nuclei for MgB₂, allowing the control over nanosizing of the system during cycling. Another interesting feature of the system was the observed reduction of MXene Ti₃C₂ to generate Ti(0)) active metal sites, a crucial event for generating the catalytic TiB₂ species (Reaction (72)) (Figure 7) [184].

Ti + 2LiBH₄ → TiB₂ + 2LiH + 3H₂

(72)

A ternary RHC system LiBH₄-MgH₂-NaAlH₄ was introduced by Plerdsranoy in 2016 and transformed in another RHC LiAlH₄–MgH₂–NaBH₄ when nanoconfined by melt infiltration in CAS (resorcinol-formaldehyde synthesis; CAS-RHC 1:1 weight ratio) [185]. This led to thermodynamic improvement (ΔT = −70 °C) and reversible behavior after four cycles (65% and 55% H₂ released by nanocomposites). Notably, thermodynamic alteration by nanoconfinement reduced the multi-step release for neat RHC to only one major H₂ release event. NaBH₄ starts to decompose at 360 °C (@CAS) vs. 475 °C (milled RHC) or 540 °C (bulk NaBH₄). Re-hydrogenation proceeds also at 360 °C, under 50 bar H₂ for 12 h, making NaBH₄, MgH₂ and Li₃AlH₆ reversible, but Li₃AlH₆ and NaBH₄ were partially reversible when nanoconfined (p_{rehydrogenation} > 80 bar H₂). A downside when dealing with fully reduced light metals, is that the production of Na(l) is followed by its evaporation and the subsequently reduced H₂ wt% capacity (Reactions (21), (22) and (73)) [185].

NaAlH₄ + LiBH₄ → LiAlH₄ + NaBH₄

(73)

Intermediate phases form as a result of the reaction of MgH₂-Al (46) and MgH₂—LiH (74).

7MgH₂ + 3LiH → Li₃Mg₇ + 8.5H₂

(74)

There are some reports of less-common complex hydrides as components of innovative RHCs, like the (1 − x)LiBH₄ − x Mg₂FeH₆ system (x = 0.25, 0.5, 0.75), releasing 6.0 wt% hydrogen up to 630K [186]. Pressure-composition-isothermal (PCT) data revealed the same reaction occurring within all investigated samples (x = 0.5). The reversibility was demonstrated for equimolar RHC (LiBH₄ − Mg₂FeH₆) after 4 a/d cycles, with no H₂ wt% loss. The dehydrogenation reaction conforms to Reaction (75) (ΔH = 64 kJ/mol H₂; ΔS = 125 J/(K molH₂)), and confirms the crucial role of FeB [186].

LiBH₄ + Mg₂FeH₆ → LiH + 2MgH₂ + (Fe,FeB) + 5/2H₂

(75)

Another experimental data was the equilibrium pressure of 1.85 MPa, very close to the theoretical one (2.1 MPa, 643 K—from thermodynamic considerations). Mg resulted from MgH₂ dehydrogenation could react with LiBH₄ (Reaction (15)), and the final H₂ release step was assigned to Mg₂FeH₆ dehydrogenation (Reaction (76)).

Mg₂FeH₆ → 2Mg + Fe + 3H₂

(76)

Lastly, reversible capacity of 5.0 wt% was recorded in NaAlH₄-7 wt%NP-TiH₂@G composites prepared by using reactive titanium (II) hydride–catalyzed graphene nanosheets [187]. The reactive TiH₂ was prepared by ultrasonication (4 h, 40 kHz) of TiCl₄ with LiH in THF (handled in an Ar-filled glovebox; Reaction (77)).

TiCl₄ + 4LiH⟶TiH₂ + 4LiCl + H₂↑

(77)

The nanocomposite NaAlH₄-7 wt% NP-TiH₂@G resulted after incorporation of NaAlH₄ into NP-TiH₂@G scaffolds, could be fully dehydrogenated very close to RT, at 30 °C. The dehydrogenation occurred in two steps with thermodynamically accessible energy barriers (E_A,1 = 80 ± 3:3 kJ/mol and E_A,2 = 70 ± 2:8 kJ/mol), probably lowered by formation of reactive, catalytic Al-Ti species (Figure 8) [187].

A summary of the main aspects discussed on nanosized RHCs is given in

Table 5. Li-RHC-based nanostructured systems and specific details regarding hydrogenation storage capacity.

Host Type	Catalyst/Active Intermediate; H2 Storing Composite RHC	wt% H2	Ref.
Adaptive poly(4-methyl-1-pentene), TPX™ Polymer	2LiH + MgB2+ 7.5(3TiCl3·AlCl3) and 2LiH + MgB2+ 7.5(3TiCl3·AlCl3)@TPX	~9.1 wt% (non-confined); 7.3 wt% (TPX-confined)	[140]
Carbon aerogel scaffold with pore size Dmax ∼21 nm	Al12Mg17 and Mg1−xAlxB2 proposed as intermediates; 2 LiBH4: MgH2	4.3 wt%; ~4 wt% reversible	[142]
-	Li-RHC (2LiH + MgB2/2LiBH4 + MgH2) composite doped with TiO2 (Li-RHC-Ti).	10.1 wt% (stable after cycles 11–15)	[143]
-	2 LiH + MgB2/2 LiBH4- MgH2. Additive used: 0.05 mol TiCl3/mol MgB2	- (kinetic study by PCI isotherms in the 350–400 °C temperature interval).	[144]
Resorcinol formaldehyde carbon aerogel (RFC): 1.1452 cm3/g, 687.1 m2/g.	2LiBH4-LiAlH4/RFC; AlB2 as catalyst in-situ. Li2B12H12 was also identified as potential intermediate.	5.7 wt% (reversible) (tonset, step 1 = 100 °C; tonset,step 2 = 280 °C); rehydrogenation at 350 °C, 5 MPa H2 (5 h)	[145]
Activated carbon nanofibers (ACNF)	LiBH4-LiAlH4	6.6 wt% (milled LiBH4-LiAlH4); 2.7 wt% (nano RHC-ACNF)	[146]
Mesoporous carbon aerogel CA (Dmax = 30 nm, SBET = 689 m2/g and Vtot = 1.21 cm3/g)	2LiBH4-NaAlH4; AlB2 and Al detected as final dehydrogenation products (can act as catalysts, also Li3AlH6 and LiAl3 was observed); C surface can also exhibit catalytic properties.	2.4 wt% (2.6 wt% theoretical; 33% pore filling); 7.8 wt% (estimated based on 100% pore filling of CA); tonset = 100 °C; 83% reversible of theoretical wt% H2 (RHC-CA); 47% reversible (RHC). Rehydrogenation at 126 bar H2, 400 °C, up to 10 h (NaBH4 recovered)	[147]
Porous hollow carbon nanospheres (HCNS)	LMBH: LiBH4-Mg(BH4)2 in 55:45 mole ratio, eutectic formation reported. (LMBH@HCNS final composite). Stable [B12H12]2− anion was detected by FTIR (dehydrogenated products)	4.3 wt% (33LMBH@HCNS); 6.1 wt% (50LMBH@HCNS); 8.1 wt% (67LMBH@HCNS). Rehydrogenation (100 bar H2, 6 h, 280 °C)	[148]
Nano–Ni(doped)in Ni/C scaffold	2LiBH4–MgH2; final composite: 2LiBH4–MgH2–15%Ni/C and reactive intermediate (nanosized) Ni4B3 as a pre-catalyst for MgNi3B2 species.	~9 wt%	[149]
Activated carbon	2LiBH4-MgH2; “LiBH4-MgH2-AC” composite; Li2B12H12—intermediate (incomplete rehydrogenation)	3.56–4.55 wt% (dehydrogenation); 2.03–3.28 wt% H2 (rehydrogenation) for the first 5 cycles	[150]
Carbon aerogel scaffold (CAS) doped with ZrCl4	2LiBH4–MgH2; ZrCl4 catalyst/dopant (Zr-P63mmc and ZrB2 detected by XRD in dehydrogenated samples).	2.5–5.4 wt% (nano-RHC:AC, 1:3–1:1); 8.7 wt% (milled RHC)	[151]
-	ternary and quaternary mixtures in the system LiBH4-NaBH4-KBH4-Mg(BH4)2-Ca(BH4)2:	various (structural investigation study)	[152]
-	2 LiBH4–2.5 Mg2NiH4; reversible H2 storage via MgNi2.5B2 (intermediate)	~4.8 wt% (under H2 pressure of 1 or 5 bar)	[153]
Carbon aerogel scaffold (CAS)	2LiBH4–MgH2–0.13TiCl4; TiCl4 catalyst ([B12H12]2− stable intermediate identified at 2480 cm−1).	3.6 wt% (1.5 h, tonset= 140 °C, release in range 140–380 °C)	[154]
-	LiBH4–NaBH4	N/A	[155]
Ordered Nanoporous Carbon (NPCs) 1012 m2/g and 0.65 cm3/g	LiBH4–Mg(BH4)2	~7 wt% (1st cycle), ~ 1 wt% (2nd cycle)	[156]
-	0.62LiBH4-0.38NaBH4; nano-Ni as catalyst, producing 0.91 (0.62LiBH4-0.38NaBH4)-0.09Ni catalyzed RHC.	8.1 wt% H2 (tpeak = 350 °C, up to 650 °C, Ar). Decreasing performace in subsequent cycles (5.1 wt%, 1.1 wt% and 0.6 wt%).	[157]
-	2LiBH4 + M (M = LiAlH4, Li3AlH6)	8 wt% (2LiBH4-LiAlH4), 9.1 wt% (2LiBH4-Li3AlH6, 150 min)	[122]
Nanoporous silica (fumed silica; mesoporous MCM-41 and SBA-15)	LiBH4–Ca(BH4)2 (4:1, 2.1:1 -eutectic composition)	N/A	[158]
2D-MXene (Ti3C2)	4MgH2-LiAlH4-Ti3C2 (Ti3C2 MXene acts as pre-catalyst of Ti(0) species in 4MgH2-LiAlH4 RHC). TiH1.942 intermediate identified in situ hints at reaction Ti(0) with RHC.	~ 6.5 wt% (onset at 336K, peak at 594 K)	[159]
-	LiBH4/KBH4 as eutectic catalyst for MgH2/Sn destabilized hydride, and for rehydrogenation of spent MgB2 to Mg(BH4)2. 0.025 MgI2 was found to bring further kinetic enhancement	~1.7 wt% (tonset = 150 °C; 40 h); ~2 wt% rehydrogenated (920–1000 bar. at 215–175 °C); 1.9 wt% (2nd dehydrogenation).	[160]
-	4LiAlH4–Mg2NiH4; Ti-based catalyst used (3 mol% TiF3); potential intermediacy of Al3Ti (Ti0) phase (in-situ XPS data). [F-] effect possibly due to replacement of H in Al-species, forming [AlF4]− and [AlF6]3− which facilitate dehydrogenation.	5.62 wt% (TiF3-cat.); 5.02–5.09 (undoped, Ti-or TiO2-cat.), with tonset = 50 °C for 1st step (21); tonset = 128 °C for 2nd step (22); 1.38 wt% (reversible rehydrog., 3 MPa H2, 300 °C, 150 s).	[161]
Mesoporous carbon CMK-3 (C-replica of SBA-15 silica)	x LiBH4–(1 − x) Ca(BH4)2 eutectic with x = 0.50, 0.60, 0.65, 0.68, 0.70, 0.75, and 0.80 (mp~200 °C). CMK–3 porosity: 1229 m2/g, 3.5 nm pore size, 1.63 cm3/g–0.59 cm3/g micro & 1.04 cm3/g meso)	~11 wt% (eutectic RHC); ~ 5 wt% (RHC@CMK-3, up to 400 °C, tpeak~300 °C, 2 h, under 3 bar H2). Rehydrogenation (400 °C, 24 h, 110 bar H2)	[162]
Activated carbon aerogel CA (689–2660 m2/g, 1.21–3.13 cm3/g)	0.55LiBH4–0.45Mg(BH4)2	13.3 wt% H2 (RHC@CA); 8.4 wt% (RHC). Reversible: 8.3 wt% (RHC@CA,4th cycle); 3.1 wt% (RHC, 4th cycle).	[163]
MWCNTs	LiBH4–LiAlH4; active phases/catalysts formed in-situ: AlB2 and LiAl.	2.0–3.0 wt% (5.5 h, 400 °C for RHC@MWCNTs), 2.3–2.8 wt% (RHC) decreasing after 3 cycles.	[164]
-	Mg@NaBH4/MgB2 core-shell structures	5.98 wt%	[165]
-	4MgH2 + LiAlH4, TiO2 catalyst. Active catalyst phases are considered Al3Ti and TiH2.	4.7 wt% (450 °C); 2.5 wt% (13 min, 320 °C); rehydrogenation confirmed (2.9 wt%, 10 min)	[166]
-	LiBH4-x AlH3 (x = 0.5, 1.0, 2.0)	11.0 wt% (release, 450 °C, 6 h) for LiBH4 − 0.5 AlH3 (best result); 3.2 wt% reversible capacity (2nd dehydrogenation, 400 °C, 5 MPa H2, 5 h).	[167]
-	Mg(NH2)2–2LiH–0.07KOH	4.5–4.9 wt% under 0–5 bar H2,	[168]
Carbon aerogel scaffold (CAS)	2LiBH4–NaAlH4; LiNa2AlH6 was identified as short-lived intermediate (60)	2.39 wt% (400 °C, for 2:1 weight ratio CAS-RHC); ~2.0 wt% reversible (4th cycle, 400 °C, 80 bar H2, 12 h)	[169]
CMK-3 Type Carbon (5 nm pore size)	LiBH4-KBH4 Eutectic (0.725 LiBH4–0.275 KBH4)	reversible 2.5–3 wt% (5th cycle)	[170]
	3LiBH4 + Er(BH4)3 + 3LiH	4.2 (1st cycle); 3.7 (2nd cycle); 3.5 (3rd cycle) out of 6 wt% (theoretically accessible for given RHC)	[171]
Activated carbon nanofibers (ACNFs)	2LiBH4-MgH2; ACNFs as dopant (30 wt%)	1.8 to 4.5 wt% H2 (reversible)	[172]
-	K2Mn(NH2)4–8LiH; K3MnH5 and K–Mn-species (probably hydrides)	6 wt% (open system); rehydrogenation almost complete (230 °C, 50 bar H2, very fast at 1 wt%/min); 3 wt% reversible (375 °C release, 300 °C uptake)	[173]
CO2-activated carbon aerogel (37–38 nm; 690–2358 m2/g)	0.62LiBH4–0.38NaBH4 (eutectic)	1.6 wt% (4th cycle, 22% of full theoretical capacity)	[174]
Activated carbon aerogel (pristine: 689 m2/g, 1.21 cm3/g and CO2-activated: 2660 m2/g, 3.13 cm3/g)	0.7LiBH4–0.3Ca(BH4)2	12.08 wt% (RHC), 7.71 wt% (RHC-CACO2), 3.36 (RHC-CA)	[175]
Carbon nanoscaffold (IRH33, 1.17 cm3 g−1, 2587 m2 g−1, 0.5 to 4.5 nm pores)	Li11BD4–Mg(11BD4)2	4.5 wt% (460 °C, 6.7 wt% maximum theoretical based on 46% filling, 14.6 wt% maximum of RHC)	[176]
-	NaBH4-Li3AlH6	4.2 wt% (330 °C); 6.1 wt% (420 °C, 30 atm, 60 min)	[132]
-	6Mg(NH2)2–9LiH–LiBH4 (additives: YCl3 and Li3N)	4.2 wt% uptake (180 °C, 85 bar H2, 8 min or 90 °C, 185 bar H2)	[177]
-	2LiH + MgB2/2LiBH4 + MgH2; 0.05 TiCl3 was used as additive	N/A	[144]
Resorcinol–formaldehyde carbon aerogel scaffold (RF–CAS)	2LiBH4–MgH2–TiCl3	3.6–3.75 wt% (1st release, 95–98.6% of maximum theoretical; 425 °C, 3.4 bar, 4.5 h); 3.25 wt% (2nd release, 8 h); 3.6–3.75 (3rd, 4th release, 15 h); uptake (425 °C, 130–145 bar)	[178]
Mesoporous carbon (MC) scaffolds	2LiBH4-LiAlH4	10 wt% (300 °C, 300 min)	[179]
-	LiBH4-MgH2; Fe3B was used as additive.	4.11 wt% (1st cycle, des.); 2.35 wt% (1st abs); 2.89 wt% (7th cycle, des.); 3.32 wt% (7th cycle, abs.).	[180]
-	LiBH4-NaBH4-KBH4-Mg(BH4)2-Ca(BH4)2	N/A (up tp 500 °C, incomplete desorption)	[181]
Carbon aerogel scaffold CAS (SBET = 2421 ± 189 m2/g, Vtot = 2.46 ± 0.46 mL/g, 13 nm pore size)	0.68LiBH4–0.32Ca(BH4)2 (“LiCa” eutectic)	5.4 wt% (bulk RHC), 3.7 wt% (RHC@CAS), 1.1 wt% (LiBH4@CAS), at 5th cycle.	[182]
-	2NaAlH4 + Ca(BH4)2; 5 wt% TiF3 as additive and precursor of active catalytic species Al3Ti and CaF2 (synergy)	~8 wt% (des, RHC-TiF3); ~7.5 wt% (des., RHC); 4.1 wt% (abs, RHC-TiF3); 3.5 wt% (abs, RHC), after 60 min, 420 °C, 30 atm H2.	[183]
2D MXene—nanoadditive (x wt%, x = 1, 3, 5, 7 wt%)	2LiH + MgB2/2LiBH4 + MgH2; 2D Ti3C2 MXene as additive	9.0 wt% H2 (400 °C, 20 min). Regeneration by hydrogenated at 350 °C/10 MPa, <5 min.	[184]
Manoporous carbon host (CAS; 458 m2/g; 0.56cm3/g; 4.86 nm pore size)	LiBH4-MgH2-NaAlH4 transforming into LiAlH4–MgH2–NaBH4, Intermediate phases detected as Li3Mg7, Mg17Al12 (PXD).	3.0 wt% (tpeak = 329 °C, up to 450 °C, ~72% of maximum theoretical capacity of RHC@CAS)	[185]
–	(1 − x) LiBH4– x Mg2FeH6, x = 0.25, 0.5, 0.75.	6.0 wt% (580–630K; rehydrogenation with no H2 wt% loss after 4 cycles)	[186]
NP-TiH2@G (TiH2 nanoplates 50 nm lateral; 15 nm thick; additive)	NaAlH4-7 wt%NP-TiH2@G	5 wt% (reversible, full dehydrogenation at 80 °C; rehydrogenation 30 °C, 100 atm H2)	[187]

Compared to other reports [147], changing the reagent molar ratio (from 2:1 to 1:1) [185], in the case of LiBH₄-NaAlH₄ RHC, can yield different products—some of which with proven catalytic activity in hydrogenation cycling (AlB₂ for instance [147]). It seems feasible that tuning the molar ratio of reagents and the consideration of possible chemical reactions occurring during de/rehydrogenation can shape the future of RHC, including avenues poorly explored thus far.

.

2.1.6. LiBH₄-Adducts—Ammoniates

Ammine metal borohydrides of the general formula M(BH₄)_n · xNH₃ are particular cases of metal borohydride adducts M(BH₄)_n · x L (L = dative ligand, typically containing an electron-donor atom/group). While providing good H₂ storage capacity, the main drawback of borohydride adducts is the lack of a generally applicable regeneration treatment.

In case of ammonia-adduct LiBH₄·NH₃, a three-step process was proposed, involving digestion (H⁺ addition to Li-B-N polymer formed by dehydrogenation; CH₃OH treatment producing LiB(OCH₃)₄ depicted by Reaction (78)), reduction (H⁻ addition; LiAlH₄ converts LiB(OCH₃)₄ into LiAl (OCH₃)₄ and regenerated LiBH₄) and NH₃-complexation (exposure of LiBH₄ to ammonia NH₃ atmosphere) [188].

LiN_xBH_y + CH₃OH ⟶ LiB(OCH₃)₄ + xNH₃ + yH₂

(78)

In the process, a CoCl₂ catalyst was also used, yielding LiBH₄ · NH₃-2 mol% CoCl₂ composite which was released upon heating to 200° (2 °C/min) ~13.6 wt% H₂ (3 equiv.), thus producing LiN_xBH_y. Regeneration of ammonia adduct of LiBH₄ from spent fuel LiN_xBH_y was achieved through the three-step process (MeOH-LiAlH₄-NH₃ method) and was confirmed by ¹¹B NMR and XRD data (Figure 9) [188].

The same ammoniate complex of LiBH₄ was also investigated by nanoconfinement in nanoporous SiO₂ (LiBH₄·NH₃@SiO₂, 1:2 wt/wt)) [189]. This strategy led to an onset dehydrogenation reaction at 60 °C, and an enhanced conversion of NH₃ to H₂ (85% of total gas evolved). The H₂ release was measured to be 1.26, 2.09 and 2.35 equivalent of hydrogen at 150 °C, 200 °C, and 250 °C, respectively. It was proposed that the new “ammonia-deliquescence” method—which avoids the usage of additional solvents—affords a low dehydrogenation temperature due to the NH₃ interaction with siloxanic support, which leads to the stabilization of ammonia by nanoconfinement into nanopores, and enhanced interaction of LiBH₄ and NH₃ in the nanoscaffold (Figure 10) [189].

2.1.7. LiNH₂BH₃ (Lithium Amidoborane)

Lithium amidoborane (LiNH₂BH₃) is a promising material for hydrogen storage due to its high H₂ content (10.9 wt%), which can be desorbed below 100 °C without formation of side-products like borazine. It can be synthesized by reaction of ammonia borane (NH₃BH₃) with either LiH or Li₂NH (79).

$$\mathrm{LiH}+{\mathrm{NH}}_3{\mathrm{BH}}_3 \stackrel{-{\mathrm{H}}_2}{\to } {\mathrm{LiNH}}_2{\mathrm{BH}}_3 \stackrel{-\frac{1}{2}{\mathrm{NH}}_3}{\leftarrow }½{\mathrm{Li}}_2\mathrm{NH}+{\mathrm{NH}}_3{\mathrm{BH}}_3$$

(79)

Bearing similarities to NH₃BH₃, lithium amidoborane will release H₂ in two consecutive steps (90 °C, and 150 °C), yielding LiNHBH₂ and finally, LiNBH (80).

$${\mathrm{LiNH}}_2{\mathrm{BH}}_3 \stackrel{-{\mathrm{H}}_2}{\to } {\mathrm{LiNHBH}}_2 \stackrel{-{\mathrm{H}}_2}{\to } \mathrm{LiNBH}+{\mathrm{H}}_2$$

(80)

Ammoniates of LiNH₂BH₃ are also known, and LiNH₂BH₃.NH₃ can generate H₂ at very reasonable temperature ranges ~40–70 °C. Regeneration of lithium amidoborane can be achieved by reaction of solid residue LiNBH with methanol with the formation of LiB(OCH₃)₄, and distillation to form B(OCH₃)₃ which can be reduced by LiAlH₄ and NH₄Cl to NH₃BH₃ (AB). As a last step, reaction of AB with LiH reforms LiNH₂BH₃ [190]. Even under a nanoconfined state, LiNH₂BH₃ has an exothermic decomposition and cannot be regenerated under 10 MPa H₂ pressure [168].

2.1.8. Li-N-H System; Li₃BN₃H₁₀

The Li-N-H system has received attention due to important thermodynamic features which allured the community into the promise of a material that would release H₂ at temperatures near ambient condition. However, as is often the case with many hydrogen storage systems, its reversibility is not trivial [136]. As previously mentioned, RHC typically contain an amide source and a hydride partner, oftentimes with the aid of a third component (complex hydride—borohydride or catalyst), as is the case in the 6Mg(NH₂)₂-9LiH-LiBH₄/(YCl/Li₃N) system discussed before [177].

An interesting report of an Li-N-H system involved nanoconfinement of Li₃BN₃H₁₀ in the pores (4.4 nm) of highly ordered nanoporous carbon NPC (S_BET = 1012 m²/g, V_{pore, BJH} = 0.65 cm³/g). Thermodynamic destabilization was confirmed by an onset dehydrogenation temperature of 110 °C (160 °C lower than bulk). More importantly, the exothermic reaction of bulk Li₄BN₃H₁₀ was altered to a two-step, endothermic process which released lower toxic gases—NH₃ and B₂H₆—upon confinement (Figure 11) [191].

The sheer size of the mesopores (4.4 nm) appeared to play an important role in driving decomposition and suppressing B₂H₆ emission, while other reports utilizing mesoporous carbon of larger sizes (13 nm) did not bypass this issue [191].

2.2. Na-Based Complex Hydrides

2.2.1. NaBH₄

NaBH₄ bears similarities with its lighter counterpart (LiBH₄); however, its thermodynamic stability is too high for use in hydrogen storage materials in the bulk form (580 °C, 1 atm H₂, E_A = 275 kJ/mol). Catalysts and nanostructuring have shown very promising results in the recent past, achieving system destabilization and implicitly lower dehydrogenation temperatures and, in some case, partial reversibility [192], by controlling particle size growth during cycling [193].

Sodium borohydride remains an important pillar of the hydrogen storage puzzle, whether used in RHC with other borohydrides [152,156,174,181], or with other complex hydride systems (alanates [132] or Mg₂NiH₄ [194]), nanostructured [192], on development of new synthesis methods to be obtained in NP form [60,195], pursuing features like restrained growth [193], fast Li-ion conductors [66], catalyzed systems (GdF₃ [196,197], ScF₃-YF₃ [198], V-based catalysts [199], SiS₂ [103], MgFe₂O₄ [200]), inclusion in core-shell-structures (with Mg/MgB₂ as partners) [165], or investigation accounts based on physical methods investigation of {BH₄]⁻ anion mobility (neutron scattering [71]) or interphase evolution (LiI [201]). Other approaches have employed the use of FeCl₃ catalyst for NaBH₄-based proton exchange membrane fuel cells [202].

2.2.2. NaAlH₄

Sodium aluminate (NaAlH₄) is another promising complex hydride which promises a theoretical H₂ storage capacity of 5.5 wt% at 250 °C, which—at least on paper—complies to US DOE’s upcoming target for the year 2025 [78,113,203]. However, regeneration is still an issue and can be aided by catalyst/dopant addition, or by nanostructuring techniques [110,111,204] which oftentimes alter the hydrogen dynamics [205], ionic conductivity [206,207] and thermodynamic behavior of alanate systems [114,208].

Recent reports focus on confinement in carbonaceous hosts: in ordered mesoporous carbon [209,210], ScOCl-functionalized carbon aerogel [211], MOFs (Ti-functionalized MOF(Mg) [212], new additives or scaffolds (Ni Raney with pore size 3 nm –affording onset of desorption at ~85° [213], graphene G [214], Ti-doped CO₂-activated carbon aerogel [215], polymer nanocomposites based on g of polyaniline or sulfonated polyetherimide as polymer matrices, affording 1.1 wt% storage after 12 h at 120 °C and 32 bar H₂ [216], N-doped nanoporous carbon NPC that lowered desorption E_A by 70 kJ/mol [217], graphene oxide GO [218], Al [219], exploring plasmonic heating effect on Au–hydride interface for local, light-activated heating [220], Co-catalyzed porous carbon hosts–affording a 3.3 wt% reversible gravimetric capacity after five cycles a/d [221], carbon nanotubes CNTs [222], CeO₂ hollow nanotubes HNTs [223], ordered mesoporous carbon OMC–producing a highly stable material with 80% H₂ capacity retention after 15 cycles [224], CeF₃/Ti₃C₂ MXene [225], TiH₂/G [187], C@TiO₂/Ti₃C₂ [226]). In addition, novel investigations on RHCs (NaAlH₄-Ca(BH₄)₂ [183]), undoped– [227] or TiCl₃-catalyzed micro-mesoporous carbon produced by resorcinol-formaldehyde method [228], multi-wall carbon nanotubes MWCNTs [229], Ni-nanoporous sheets of carbon [125], Ti-functionalized MOF–74(Mg) [212] or improved synthetic methods that generate alanates as NPs [59] were focused on. Destabilization of NaBH₄ was reviewed recently by correlation to TM fluorides used as dopants [230]. Destabilization can also be achieved using an ionic liquid–vinylbenzyl trimethylammonium chloride [231], alkali metals addition –leading to enhanced reversibility [232] or carefully-chosen TM ferrites like NiFe₂O₄ NPs [233]. Using RHC has proved advantageous, especially when combined with light metal borohydrides (2LiBH₄-NaAlH₄ [169], quinary equimolar mixture of light borohydrides [181] or various eutectic compositions in Al nano-framework [234]).

2.3. Mg-Based Complex Hydrides

Magnesium based hydrides have puzzled scientists because they offered remarkable H₂ storage properties largely unattainable by other complex or simple hydrides, including the real prospect of reversibility during cycling [235,236,237]. Results stemmed from the high energy ball milling process [238], magnesium borohydride [239], or advances in the Li-Mg-Al systems [240].

2.3.1. Mg(BH₄)₂

Perhaps one of the most fascinating borohydrides, Mg(BH₄)₂, has the largest number of polymorphs (at least six of them characterized), complex crystal structures and true permanent porosity (γ-polymorph, ~1100 m²/g surface area), all the while exhibiting one of the highest H₂ wt% among characterized borohydrides. The influence of H₂ pressure (up to 1000 bar H₂) on the thermodynamics of Mg(BH₄)₂ impregnation into carbonaceaous supports was studied recently by White el al. [241]. The field of solid-state electrolytes for batteries was enriched with new examples of Li/Na/Mg nanoconfined borohydrides [66].

Improved hydrogen storage capabilities were recorded through nanoconfinement [54] in C-based materials ([80], high surface area graphite [242], ordered mesoporous carbons like CMK-3 and CMK-8 [241] and mesoporous carbon [243]. Also recorderd were IRH33 carbon [176]) or graphene (G) and its derivatives (G [244], rGO [245] and G-aerogels [241]) as part of RHC (LiBH₄-Mg(BH₄)₂ [148,163,242], the isotopycally-labeled RHC: Li¹¹BD₄-Mg(¹¹BD₄)₂ [176], or produced in-situ by cycling from the prior-RHC: LiBH₄-MgH₂ [180]). In addition, LiBH₄-NaBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂ [152,181], hybrid dual-component Mg(BH₄)₂-Metal–Organic Borohydride: tetramethylammonium borohydride (TMAB) [246] or ternary systems Mg₂NiH₄-LiBH₄-Mg(BH₄)₂ [194]). Catalysts (Al₂O₃ produced by atomic layer deposition in γ-Mg(BH₄)₂@Al₂O₃ composite, Ni-Pt core-shell NPs [243]), additives (MgCl₂ [247], or others [248], Ti₃C₂ MXenes—a 2D layered material [249]) or exhibiting the influence of the nanoscale modification of dehydrogenation product—MgB₂—on the overall improved hydrogenation behavior [250,251]. Given the high affinity towards catalysts, addition of Ni NPs dispersed in mesoporous carbon yielded clear improvements in hydrogenation behavior of Mg(BH₄)₂ [252,253]. Bipyridine-functionalized MOF was shown to improve reversibility of Mg(BH₄)₂, potentially through the B-N synergy [254].

2.3.2. Mg(B₃H₈)₂

While probably not enough explored, the speciation of borohydrides is rich and other complex hydrides could be nerated as a result of this. For instance, Mg(B₃H₈)₂ was synthesized and the conversion to tetrahydridoborate [BH₄]⁻ ion was studied in the RHC: MgH₂-Mg(B₃H₈)₂ [255]. Mg(B₃H₈)₂ was synthesized by a metathesis reaction between 2NaB₃H₈ and MgBr₂ under Ar in a ball milling process (Reaction (81)).

2NaB₃H₈+ MgBr₂ → Mg(B₃H₈)₂ + 2NaBr

(81)

With about 22 wt% conversion [B₃H₈]²⁻-to–[BH₄]⁻, and a decomposition onset as low as ~100 °C, the RHC proposed by Gigante et al. Mg(B₃H₈)₂-4MgH₂ shows real promise for MgH₂-Mg(B₃H₈)₂ as a hydrogen carrier, managing tetrahydridoborate conversion (85–88 wt%) below 200 °C within 1 h, without B-losses as toxic boranes (B₂H₆ and higher homologues) (Reaction (82)).

Mg(B₃H₈)₂ + 4MgH₂ → 3Mg(BH₄)₂ + 2Mg

(82)

In this process, the leading role was that of activated MgH₂ addition, since neat octahydrotriborate does not undergo this transformation under such mild conditions [255]. Moreover, the use of unsolvated Mg(B₃H₈)₂ affords minimal boranes by-products, in stark contrast to its diglyme adduct that released mainly B₅H₉ [255].

2.3.3. Mg(BH₄)₂-Adducts/Ammoniates: Case of Mg(BH₄)_2.6NH₃

The practical use of NH₃-adducts of metal borohydrides is plagued by the exothermic effects. However, upon nanoconfinement into a microporous activated carbon (AC, S_BET = 2051 m²/g, V_micro = 0.835 cm³/g) scaffold, the dehydrogenation of the hexaammoniate Mg(BH₄)₂·6NH₃ changes the thermodynamics of the process, which now becomes endothermic with an improvement in H₂ release temperature of ~40 °C [256]. The microporosity of the AC was found to be a critical factor (mean pore size 4 nm). Notably, the ammoniation was the result of a solvent exchange occurring within AC porosity (Figure 12, Reaction (83)).

$$\mathrm{Mg}({\mathrm{BH}}_4{)}_2\stackrel{2{ \mathrm{Et}}_2\mathrm{O}; \mathrm{AC}}{\to } \mathrm{Mg}({\mathrm{BH}}_4{)}_2\cdots 2{\mathrm{Et}}_2\mathrm{O}@\mathrm{AC}\stackrel{6{ \mathrm{NH}}_3;-2{\mathrm{Et}}_2\mathrm{O}}{\to } \mathrm{Mg}({\mathrm{BH}}_4{)}_2\cdots 6{\mathrm{NH}}_2@\mathrm{AC}$$

(83)

Among the five investigated samples xMg(BH₄)₂·6NH₃ @ yAC (1:1, 0.8:1, 0.6:1, 0.4:1 and 0.2:1), the 1:1 Mg(BH₄)₂·6NH₃ @ AC sample released H₂ at a low temperature of ~40 °C, 85 °C lower than its bulk, non-nanoconfined counterpart, with the main dehydrogenation event occurring in the range 150–350 °C (Figure 13).

Increasing the AC content lowered the dehydrogenation peak even further. The 0.6Mg(BH₄)₂·6NH₃ @ 1AC had a peak dehydrogenation temperature of 148 °C, with the H₂ release ending at 240 °C [256]. The intermediacy of a novel Mg-B-N compound was identified; however, rehydrogenation attempts were not successful.

2.3.4. Mg(NH₂)₂

Recent reports on magnesium amide are concerned with RHC systems: Mg(NH₂)₂-2LiH-0.07KOH [168] and 6Mg(NH₂)₂-9LiH-LiBH₄ co-catalyzed by YCl₃/Li₃N [177].

Chen et al. started from an RHC: Li₃K(NH₂)₄–MgNH–LiNH₂ showed that the dehydrogenation is dependent on p(H₂). Up to 4.5–4.9 wt% H₂ was released under pressures up to 5 bar H₂ (Figure 14).

The role of KOH was important, since it enhances the reversibility of the Mg(NH₂)₂–2LiH system (Reaction (59)) [168].

Cao et al. have used chloride (YCl₃) and nitride (Li₃N) catalysts to improve the hydrogenation in 6Mg(NH₂)₂–9LiH–LiBH₄ system, and achieved 4.2 wt% H₂ capacity which can be regenerated applying 85 bar H₂ (180 °C, 8 min) or 185 bar H₂ (90 °C) [177]. The nanocrystaline catalysts (2–10 nm) and catalyst-derived species (YH₃, YB_x, all nanocrystaline 2–10 nm) provide kinetic alteration to the de/rehydrogenation Reaction (84).

6Mg(NH₂)₂ + 9LiH + LiBH₄ ↔ 3Li₂Mg₂(NH)₃ + Li₄(BH₄)(NH₂)₃ + 9H₂

(84)

The origin of the Li₄(BH₄)(NH₂)₃ product can be traced back to the successive reactions occurring within the proposed RHC system (Reactions (85)–(87)) (Figure 15 and Figure 16) [177].

3LiH + YCl₃ → YH₃ + 3LiCl

(85)

Li₃N + 2H₂ → LiNH₂ + 2LiH

(86)

3LiNH₂ + LiBH₄ → Li₄(BH₄)(NH₂)₃

(87)

2.3.5. Mg(AlH₄)₂

Only a few reports of Mg(AlH₄)₂ have been published so far [125,257]. When confined into nickel-containing porous carbon sheets (Ni-PCSs, S_BET = 2572 m²/g, pore size 2.8 nm), LiAlH₄, NaAlH₄ and Mg(AlH₄)₂ were shown to lower the desorption temperature by as much as 29 °C (Figure 17) [125].

The Mg(AlH₄)₂-Ni-PCS sample started to desorb H₂ at 125 °C (146 °C in ball-milled version), with a total hydrogen content of 3.34 wt%—a lower amount compared to the theoretical 9.3 wt% due to the inclusion in a porous support, but also due to added weight of LiCl from metathetical synthesis of Mg(AlH₄)₂ (LiAlH₄ and MgCl₂) [125].

Xiao et al. synthesized Mg(AlH₄)₂ nanoparticles (2–7 nm) using a solvent-free, mechanochemical strategy and studied their hydrogenation properties [257]. These NPs start to desorb H₂ at 80 °C (completing the first dehydrogenation step at 120 °C, 30 min), recording a 65 °C improvement compared to microparticles of Mg(AlH₄)₂ [257]. Similar improvement was deduced regarding the activation energy of dehydrogenation: a reduced value of 105.3 kJ/mol vs. 123.6 kJ/mol in micro-Mg(AlH₄)₂. Interestingly, the presence of LiCl from the metathesis reaction was found to keep the Mg/MgH₂ products in the nanorange (<10 nm), hence improving the cycling behavior of the system [257].

2.4. Other First Group-Derived Borohydrides

Since the general trend regarding stability of alkali metal borohydrides is to increase down the group 1 elements of the periodic table, the thermodynamic destabilization should be even greater in order to bring their operation regime into a fuel cell achievable domain (100–200 °C). Hence, the reports on heavier alkali metal borohydrides, coupled with their lower theoretical H₂ content, are rather scarce. For instance, only a few reports of RbBH₄ and CsBH₄ exist—and which concern reorientation mobility of [BH₄]⁻ anions [71], a report on complex hydride K[Al(NH₂BH₃)₄] [258] and slightly more research performed on lighter KBH₄ [60,71,152,160,170,181,194]. In one of these reports, a mixture of complex hydrides (LiBH₄/KBH₄) was used as an electrolyte to perform full rehydrogenation in the MgH₂/Sn coupled system [160].

Bearing the familiar amidoborane ligand, K[Al(NH₂BH₃)₄] was synthesized by Møller et al. and featured a triclinic unit cell crystal structure (P-1) [258]. Following a mechanochemical approach, KAlH₄ and AB (NH₃BH₃) were milled and afforded the complex hydride K[Al(NH₂BH₃)₄] which featured two-step, exothermic decomposition peaks in DSC analysis (94 °C and 138 °C), releasing a total of ~6.0 wt% H₂ along with amorphous KBH₄ (¹¹B-and ²⁷Al NMR) [258]. Still, the system was not reversible under employed conditions (260 °C, 110 bar H₂).

Nanostructuring KBH₄-LiBH₄ RHC into CMK-3 type ordered mesoporous carbon has shown improved dehydrogenation behavior [170]. Exploiting an eutectic composition of 0.725 LiBH₄–0.275 KBH₄ featuring a low melting point (105 °C), melt-infiltration into mesoporous carbon CMK-3 afforded a reversible H₂ uptake-release of 2.5–3.0 wt% during five hydrogenation cycles [170]. The quinary equimolar mixture LiBH₄-NaBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂ was investigated by Dematteis et al. [181], and showed that the pure borohydrides only form KCa(BH₄)₃ as a new phase, yielding a liquid phase (after up to 50 h of ball milling) from which dehydrogenation was shown to be a complicated, multi-step process. This approach of combining five borohydrides relied on the concept of high energy alloys [181]. Dematteis et al. have also explored the opportunity of Mg₂NiH₄—confirmed to act as a hydrogen pump in the Ni-doped Mg/MgH₂ system—to form novel RHC-like systems Mg₂NiH₄-LiBH₄-M(BH₄)_x (M = Na, K, Mg, Ca) [194]. In particular, the Mg₂NiH₄-LiBH₄-KBH₄ was prepared by ball milling in the eutectic borohydride composition (reversible, mp = 110 °C), yielding 0.56 Mg₂NiH₄, 0.32 LiBH₄, 0.12 KBH₄ [194]. Even so, a modest temperature enhancement relative to the bulk was recorded (ΔT = −10 °C) and the PXD spectra confirmed formation of Mg₂NiH_0.3, MgH₂ and MgNi_2.5B₂ only from reaction of LiBH₄ with Mg₂NiH₄, with KBH₄ remaining largely unaffected [194].

2.5. Ca-Based Complex Hydrides; Ca(BH₄)₂

Magnesium and calcium are two representative examples of group 2 complex hydrides. Ca(BH₄)₂ in particular has some attractive features that set it apart from its Mg-counterpart.

Ca(BH₄)₂ was also investigated in the context of RHC when mixed with other borohydrides in the ternary/quaternary [152] or quinary [181] systems based on mixtures LiBH₄-NaBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂, while DFT computations shed light on the decomposition of Ca(BH₄)₂ altered by nanoconfinement effects [259]. In the quantum mechanical computations, thin films of β-Ca(BH₄)₂ were shown to afford the decrease of dehydrogenation enthalpy by ΔH = −5 kJ/mol H₂, through the formation of the active intermediates CaH₂ (path (88a)) or CaB₂ (pathway (88b), Reaction (88)) [259].

$${\mathrm{CaH}}_2+2\mathrm{B}+3{\mathrm{H}}_2 \stackrel{\mathrm{a}}{\leftarrow } \mathrm{Ca}({\mathrm{BH}}_4{)}_2\stackrel{\mathrm{b}}{\to }2/3{\mathrm{CaH}}_2+1/3{\mathrm{CaB}}_6+10/3{\mathrm{H}}_2$$

(88)

Nanoconfinement effects in actual physical systems were confirmed by Comanescu et al., who designed a micro-mesoporous carbon scaffold and used the activated scaffold (MC 650-a, S_BET = 1780 m²/g, V_pore = 1 cm³/g) of high porosity to confine Ca(BH₄)₂ by the incipient wetness method of an MTBE solution of calcium borohydride. This approach led to ~ 2.4 wt% reversible and roughly stable H₂ capacity over 18 cycles of hydrogen release/uptake (Figure 18). Remarkably, the onset of H₂ desorption was reduced to ~100 °C and rehydrogenation conditions were significantly improved, compared to bulk Ca(BH₄)₂ (20–45 atm H₂, 6.5 h). Additionally, ~68.7% of Ca(BH₄)₂ was proved to behave reversibly after the evaluation of support weight contribution to the Ca(BH₄)₂@MC-a nanocomposite [260].

Other supports for Ca(BH₄)₂ confinement were also investigated: SBA-15 and MCM-41 (RHC: LiBH₄-Ca(BH₄)₂) [158], mesoporous carbon (RHC, eutectic composition LiBH₄-Ca(BH₄)₂) [162], activated carbon aerogel (RHC: LiBH₄−Ca(BH₄)₂) [175], ordered mesoporous carbon CMK-3 (1320 m²/g, 1.48 cm³/g) catalyzed by TiCl₃ additive [261]. Other RHCs containing Mg₂NiH₄ (Mg₂NiH₄-LiBH₄-Ca(BH₄)_x) highlighted the role of complex hydride Mg₂NiH₄ in lowering the desorption temperature of the eutectic mixture of borohydrides [194]. Even in the absence of a Ni-based hydride, reversibility of LiBH₄ was enhanced by using the eutectic composition of RHC: LiBH₄-Ca(BH₄)₂, namely 0.68LiBH₄–0.32Ca(BH₄)₂, when nanoconfined into high porosity carbon aerogel (S_BET = 2421 ± 189 m²/g, V_tot = 2.46 ± 0.46 mL/g, 13 nm pore size) [182]. The catalyst effect strategy was explored in the system comprising 2NaAlH₄ and Ca(BH₄)₂, when enhancements were recorded upon TiF₃ addition, presumably due to the formation of intermediates of the form [AlF₆]³⁻ [183]. General catalytic principles and implications of using d-block metal derivatives were reviewed in correlation to metal hydride gravimetric capacity, and their potential for kinetic improvements. Among them, Ni, Co, V, Ti, Fe and Nb appear to have offered the most promising results [262].

2.6. Al-Based Complex Hydrides

While Al(BH₄)₃ may have a very high theoretical storage capacity, its volatile and extremely reactive nature have precluded its use as a hydrogen storage material. However, the corresponding ammoniate Al(BH₄)₃(NH₃)₆ could be stabilized by confinement in a PSDB (poly(styrene-co-divinylbenzene) polymeric matrix [263] where it also showed partially-reversible behavior using a hydrazine/ammonia treatment step. An interesting detail regarding the report from Tang et al. is the understanding that the diffusion of Al(BH₄)₃ (produced by mixing AlCl₃ and LiBH₄ in a 1:3 stoichiometric ratio) into the pores of PSDB was enhanced by complexation to the phenyl rings of PSDB to produce Al(BH₄)₃/PSDB nanocomposites. Subsequent exposure to ammonia afforded final Al(BH₄)₃·6NH₃/PSDB nanocomposites. The host could also be tuned for inorganic component, such as nanostructured porous carbon [264]. Inclusion of Al(BH₄)₃(NH₃)₆ in pore-expanded mesoporous carbon with high textural characteristics (S_BET = 980 m²/g, V_pore = 1.629 cm³/g and d_pore = 12.7 nm) had a beneficial effect in reducing borane emissions from the composite ammoniate@nanoporous carbon, yielding an H₂ purity of up to 93.5%.

2.7. TM- and RE-Based Complex Hydrides

2.7.1. Adducts/Ammoniates: Zr(BH₄)_4.8NH₃

Featuring a low desorption temperature, Zr(BH₄)₄.8NH₃ is considered a promising material for hydrogen storage and was recently synthesized by Wu et al. [265] using a physical vapor deposition approach, terming the method “heating-ball milling vial”. Additionally, 10 wt% NaBH₄ inclusion in Zr(BH₄)₄.8NH₃-10 wt%NaBH₄ RHC showed a thermodynamioc improvement by lowering the desorption temperature from 130 °C (bulk ammoniate) to 75 °C (composite), while also supressing alternative reaction pathways leading to B₂H₆/NH₃ (Reaction (89)) [265].

2.7.2. NaMgH₃ and NaZn(BH₄)₃

The NaMgH₃ perovskite type complex ternary hydride synthesized from MgH₂ and NaH, and activated by a ball milling approach (under Ar, 2–15 h) showed a two-step decomposition behavior, releasing 5.8 wt% H₂ from 287 °C to 408 °C within 2 h [266]. Notably, rehydrogenation of decomposed products were rehydrogenated under 10 bar H₂, at ~200 °C [266].

NaMgH₃ → NaH + Mg + H₂ → Na + Mg + H₂

(89)

The Rietveld refinement of rehydrogenated samples XRD data showed a regeneration of NaMgH₃ (27 wt%), NaH (6 wt%) and oxides (67 wt%) [266].

Nanoconfinement strategy was applied in the case of NaZn(BH₄)₃ by the inclusion of a complex hydride into the mesoporosity of SBA-15 silica—S_BET = 274.6 m²/g; V_pore = 0.4596 cm³/g—(hydride:support weight ratio 1:3, ball milling, 1 h, and infiltration of 25 drops NaZn(BH₄)₃–THF solution over 0.2 g SBA-15). Pure hydrogen was released in the range 50–150 °C, and a reduction of 5.3 kJ/mol in the activation energy was computed based on the Arrhenius plot (E_A = 38.9 kJ/mol for nanoconfined NaZn(BH₄)₃@SBA-15) (Figure 19) [267].

The decomposition reaction can be formulated according to Reaction (90), and is supported by observance of NaBH₄, amorphous B and Zn among decomposition products, along with the release of pure H₂ (6 wt%); however, rehydrogenation attempts in the solid state (400 °C, 10 MPa H₂) were unsuccessful [267].

NaZn(BH₄)₃ (nano) → NaBH₄ + Zn + B + 2H₂

(90)

2.7.3. XTiH₃ (CaTiH₃, MgTiH₃)

CaTiH₃ and MgTiH₃ perovskite hydrides were studied by Selgin using DFT and showed the predicted 4.01 wt% hydrogen storage for MgTiH₃ based on electronic band structures and energy of states computation (ductile material) vs. CaTiH₃ (brittle nature) [268].

2.7.4. Mg₂NiH₄

While not yet a commercial material, Mg₂NiH₄ can be synthesized from MgH₂ and Ni, and was utilized in a series of RHC: in LiBH₄–Mg₂NiH₄ where it showed mutually-destabilized system [153]; in the formation of nanocomposites Mg₂NiH₄@G nanosheets, where the surface MgO layer actually protected the complex hydride from further oxidation; in affording a low E_A = 31.2 kJ/mol [269] and in LiAlH₄-Mg₂NiH₄ doped with TiF₃ affording a dehydrogenation onset of 50 °C and E_A = 81.56 kJ/mol [161]. In addition, in the formation in-situ of Mg₂NiH₄-Mg₂Ni@MOF from a MgH₂@Ni-MOF starting nanocomposite (ΔH_des = 69.7 ± 2.7 kJ/mol; E_A,des = 144.7 ± 7.8 kJ/mol H₂; ΔH_abs = −65.7 ± 2.1 kJ/mol; E_A,abs = 41.5 ± 3.7 kJ/mol) [270], or in Mg₂NiH₄-LiBH₄-M(BH₄)_x (M = Na, K, Mg, Ca) composites where it afforded improvement in thermodynamic behavior by reduction with up to 40 °C the desorption temperature [194].

2.7.5. Mg₂FeH₆

The ternary hydride Mg₂FeH₆ was synthesized (2.1 MgH₂: 1 Fe, ball milling for 15 h, 200 rpm, 20 bar H₂) and used in composite systems Mg₂FeH₆ + 4LiNH₂ (Reaction (36)), allowing further reactivity enhancement that eliminated ~4.8 wt% H₂ at 225 °C and absorbed 3.7 wt% H₂ at 200 °C and 50 bar H₂ [137]. The report from Zhang et al. also highlights that the ternary hydride Li₄FeH₆ could be synthesized for the first time in a metathetical reaction, rather than applying GPa H₂ pressure, as previously reported in the literature [137].

When used in a (1 − x)LiBH₄ + xMg₂FeH₆ nanocomposite, the system showed for composition x = 0.5 a reversible behavior at 400 °C and 20 MPa H₂ for more than four cycles [186].

2.7.6. K₂Mn(NH₂)₄

K₂Mn(NH₂)₄ was synthesized by ball milling Mn and K under seven bar NH₃, and was utilized in the RHC K₂Mn(NH₂)₄–8LiH where it released 6.0 wt% H₂ in an open system. Moreover, very fast reabsorption kinetics (>1 wt%/min) allowed RHC recharging within minutes, at 230 °C and 50 bar H₂ [173]. The final identifiable products for dehydrogenation were Li₂NH, Mn₃N₂ and MnN, while after rehydrogenation an unknown “K-Mn-species2” was produced, alongside LiH and LiNH₂ [173].

2.7.7. Ti(BH₄)₄

Ti(BH₄)₃ could be a promising candidate for H₂ storage application (13.0 wt% theoretical storage capacity), if it were not for its high instability—it spontaneously decomposes at RT. Although volatile in pristine form, Ti(BH₄)₃ was synthesized and stabilized by the nanoconfinement strategy (trapping at 200 K) in the MOF structure of UiO-66 ((Zr₆O₄(BDC)₆, BDC = 1,4-benzenedicarboxylate, S_BET = 1200 m²/g). LiBH₄ and TiCl₃ were milled in a 3:1 molar ratio using a planetary mill (10 min) (Reaction (91)).

3LiBH₄ + TiCl₃ → Ti(BH₄)₃ + 3LiCl

(91)

While it would immediately decompose in bulk form into B₂H₆ and TiH₂, the nanoconfined composite Ti(BH₄)₃@MOF (S_BET = 770 m²/g, confirming confinement of borohydride species, but also that the pores have not been completely filled) was stable up to 350 K under vacuum, which is a great improvement in stabilization of such a reactive species [271]. The narrow cage size of UiO-66 (1.6–1.7 nm diameter) was demonstrably essential in achieving this stabilization effect.

2.7.8. (RE)(BH₄)_x

Frommen et al. studied the thermal decomposition of rare earth (RE) borohydrides—either obtainable starting from RECl₃-LiBH₄ mixtures (molar ratios 1:3 and 1:4), yielding RE(BH₄)₃/LiRE(BH₄)₃Cl and LiRE(BH₄)₄, respectively, (Reactions (92) and (93)), but also by exploring wet chemical sysnthesis which allow for a LiCl-free borohydride upon filtering and subsequent vacuum drying [272].

LiRE(BH₄)3Cl + 2LiCl ← RECl₃ + 3LiBH₄ → RE(BH₄)₃ + 3LiCl

(92)

RECl3 + 4LiBH₄ → LiRE(BH₄)₄ + 3LiCl

(93)

Y(BH₄)₃ is probably among the most promising RE borohydride due to its high gravimetric hydrogen capacity (9.1 wt%), but also due to reasonable dehydrogenation temperature (onset at 187 °C, peak at 250 °C).

Heere et al. investigated the composite 3LiBH₄ + Er(BH₄)₃ + 3LiH (9 wt% theoretical hydrogen capacity) and revealed a practical hydrogen release of 4.2, 3.7 and 3.5 wt% during the first three a/d cycles [171].

2.8. Ammonia Borane (AB) and Related Compounds

Ammonia-borane remains a highly investigated material, due essentially to its high hydrogen capacity. However, the system is not reversible without nanoconfinement into mesoporous materials [273] or other further tuning [274,275,276,277,278,279]. Yet, mechanistic investigations have not completely elucidated the dehydrogenation pathways and as such, rehydrogenation mechanisms are still under investigation to date.

2.8.1. NH₃BH₃ (AB)

In general, nanoconfinement of AB (NH₃BH₃, 19.6 wt% theoretical gravimetric capacity) into mesoporosity of various supports was shown to promote H₂ desorption starting at a lower temperature onset. Various supports were investigated, including mesoporous silica, GaO₃ and Co-substituted AlPO₄-5 (when polyiminoborane was not detected among thermolysis of AB@support using operando Raman-Mass Spectrometry, suggesting AB-SBA-15 interaction modifying typical decomposition pathway) [273]. Nanosized (average size 110 nm) AB synthesis using cetyltrimethylammonium bromide (CTAB) was also investigated, in addition to dodecane C₁₂H₂₆ as a counter-solvent in aqueous media (overall weight loss between 80–200 °C of 24–57.3 wt%) [280]. Mechanistic investigations of energy release from AB included using different oxidants (NH₄ClO₄ rerouting the decomposition pathway of AB through [NH₃BH₂NH₃]⁺[ClO₄]⁻ salt which inhibits BNH_x species and affords AB complete oxidation) [281], the effect of polyacrylamide-grafted organically modified mesoporous silica (PAM-COOH-MSNs and PAM-Ph-MSNs) as nanocarriers for AB confinement [282].Further investigations include silica aerogel scaffold (up to 60 wt% AB loading, thermolysis onset at 80 °C due to SiOH and SiOSi groups involed in the interaction AB-SiO₂ silica) [283], ZIF-67-derived fcc-Co@porous carbon nano/microparticles used as catalysts to enhance H₂ release from AB [284], carbon nanotubes CNTs array (CMK-5 with 1650 m² g⁻¹ and 1.69 cm³ g⁻¹, bimodal porosity; 9.4 wt% H₂ released from (30AB:30AlH₃)@CMK-5 nanocomposite at 95 °C in 10 min) [285]. Moreover, investigation subjects include nanoporous carbon (termed MDC –S_BET = 2222 m²/g, 2.49 cm³/g, 0.40 nm pore size–, obtained from MOF-5 calcination in N₂ at 1000 °C, yielding after confinement AB@MDC composites using a solution infiltration technique; 4.7 wt% H₂ was released at 80 °C with the lowest t_onset = 72 °C) [286]. GO/rGO derivatives (AB@GO and AB@rGO integrated AB into graphene derivative without solvent or melt infiltration, in a one-step “ice-templating” process, releasing no harmful gases such as B₂H₆, B₃H₆N₃ or NH₃) [287]. TiO₂(B)-catalyzed C-scaffold (synergistic role of C-TiO₂(B) in the enhancement of H₂ desorption from C-TiO₂(B)/NH₃BH₃ nanocomposites probably by added H+ ions from TiO₂ hydrolysis reaction) [288]. Microporous carbon with narrow PSD of 1.05 nm pore size (t_onset = 50 °C, main event at 86 °C and ~12 wt% H₂ release at 90 °C in 30 min) [289]. The study of MOF nature and effect on AB dehydrogenation behavior (MOFs: IRMOF-1, IRMOF-10, UiO-66, UiO-67 and MIL-53(Al)) [290], of the effect of nanoconfinement into MOF with MIL-53 topology: Al-MIL-53 (86% H₂ release), Al-MIL-OH (38% H₂ release) and Al-MIL-NH₂ (67 wt% H₂ release within 60–110 °C, suggesting interaction AB–functional groups on MOFs) [291]. Mesoporous monolithic BN as AB–scaffolds (S_BET = 584–728 m² g⁻¹, high V_pore = 0.75–0.93 cm³/g. H₂ release at 100 °C up to 8.1 wt%) [292]. Ni-matrix affording ~50 nm size reduction of AB (suppression of toxic gases like diborane, and rehydrogenation occurring at 200 °C under 6 MPa H₂) [293]. Investigation of an AB/PEO (polyethylene oxide) cocrystal for the potential energy storage applications [294]. Incorporation of AB into Pd/halloysite nanotubes (small Pd catalyst sizes of ~1.4 nm; H₂ release was recorded at 60 °C and E_A = 46 kJ/mol vs. 183 kJ/mol for neat AB) [295]. Evaluation of the feasible usage of 40:60 wt% AB:AC nanococomposite in a portable power tank (solution impregnation; ~6.0 wt% H₂ with t_onset = 96 °C) [279]. Alternatively, using less explored hosts such as hypercrosslinked porous poly(styrene-co-divinylbenzene) resin (PSDB) for AB nanoconfinement [296]. Shen et al. have used Cu NPs and showed that 6.8 nm Cu NPs are active catalysts in the decomposition of NH₃BH₃ to release H₂, as well as affording pure polybenzoxazole (PBO) in a one-pot reaction of NH₃BH₃, diisopropoxy-dinitrobenzene and terephthalaldehyde. [297]. Highly pure and chemically-resistant PBO (M_W = 19 kDa) was also reported by the same group using 8–18 nm Cu₂O, as a catalyst also able to enhance ammonia borane dehydrogenation [298].

2.8.2. Tetraalkyl Ammonium Borohydrides [NR₄][BH₄]

The parent ammonium borohydride, NH₄BH₄ (ABH₂), has a high gravimetric capacity of 18 wt% H₂ that can be released below 160 °C. However, decomposition occurs even at room temperature into [(NH₃)₂BH₂][BH₄] (DADB) and hence, a stabilization method needs to be developed before adopting its wider use as a H₂ storing media. To this end, MCM-41 mesoporous silica was used for nanoconfinement of ABH₂, allowing its storage at temperatures below −30 °C [299].

Tetraalkyl ammonium borohydrides [NR₄][BH₄] are a novel class of organic borohydrides that were investigated for CO₂ capture via triformatoborohydride ([HB(OCHO)₃]⁻) and converted to more useful chemicals [300].

3. Improvement Strategies and Most Encouraging Results—An Overview

Porous materials have emerged as a class of materials featuring high versatility regarding both structure and envisioned applications. They can be tailored to accommodate smaller or larger reagents, offering shorter or longer reaction times depending on the application field [113,208,240]. In the recent past, these materials have been explored in the energy storage field: MOFs [204,212,270,301,302,303], carbon nanomaterials [100,101,108,126,127,172,175,176,178,210,227,228,242,286,288,304,305], graphene derivatives [269,287], MXenes [159,184,225,306,307], TM oxides (CeO₂ [223]), siloxanic materials [93,107,189,267], ordered mesoporous carbon OMC [102,106,170,179,185,191,209,224,261], metal scaffolds for microencapsulation (Al [234]), polymers [296] and nitrides [105]. Machine learning was recently employed for predicting the hydrogen release of systems based on lithium borohydride [58,62].

Improvement attempts have tackled the hydrogen storage issue from various angles, and today many approaches have shown their effectiveness (Figure 20). The current strategies that were summarized in Figure 20 are detailed as follows.

3.1. Nanoconfinement

Nanoconfinement was a preferred strategy to improve hydrogenation behavior by incorporating simple or complex hydrides in MOFs [204,212,270,290,291,301,302,303,306]; metal-doped carbon originating from thermally-collapsed MOFs (Co-[284]) and(Zn-[286]); carbon materials [232,305]; microporous carbon [256]; mesoporous carbon [72,162,170,179,185,210,253,261]; micro-mesoporous carbon (0.5–4.5 nm, IRH33 [176]); nanoporous carbon aerogels [163,169,227]; CO₂–activated carbon aerogel [174,175]; TiCl₃-catalyzed carbon aerogels [178,228]; ScOCl-catalyzed carbon aerogel [211]; carbon nanofibers CNFs [95,172]; carbon nanotubes CNTs [101,285]; activated carbon catalyzed by metal salts (CeF₃ [100]); graphene G; graphene oxide GO and reduced graphene oxide rGO (GO, rGO–[287]), Fe₃O₄–G [98] and G-MC [99]; highly microporous carbon (PSD~ 1 nm, [289]); ordered mesoporous carbon OMC [106,191,209,224]; SWCNTs [108]; Zr-CMK-3 [308]; MWCNTs [164,229]; high surface area graphite [242]; fluorographite [130]; hollow carbon nanospheres HCNs [127]; various other mesoporous materials (BN–[292]), Ni-matrix [293] and Pd/halloysite NTs [295]); 1D nanoporous materials [304]; CeO₂ hollow nanotubes HNTs [223]; silica-based scaffolds [74,75,76,93,107,158,189,267]; MXenes Ti₃C₂ [96,159,184,307]; CeF₃-Ti₃C₂ MXene [225]; C@TiO₂ 2D scaffold [226]; nanoporous Ni-based alloy [97]; Ni–Pt core-shell nanoparticles [243]; Ni-catalyzed C nanosheets [125]; TiO₂/porous C [126]; NbF₅–MC [102]; porous Al scaffold [234]; h-BN [131]; polymeric matrix (hypercrosslinked porous poly(styrene-co-divinylbenzene) resin [296]) or PcB (poly (methyl methacrylate)–co–butyl methacrylate) [229].

3.2. Destabilization

Two apparently divergent strategies were pursued. Firstly, the stabilization of complex hydride systems that are too unstable for practical use under normal conditions and above (like Al(BH₄)₃ in the form of ammoniate complex Al(BH₄)₃.6NH₃, or Ti(BH₄)₃ for instance). Secondly, the destabilization of those compounds that present too high thermodynamic stability such that they lower their desorption/resorption into the achievable realm (100–150 °C ideally). However, due to the high thermal stability of complex hydrides in general, the destabilization methods prevail [309]. More specifically, nanostructured γ-Mg(BH₄)₂ was destabilized by the Al₂O₃ atomic layer [310], mutual destabilization is typically observed in RHC comprising metal borohydrides (LiBH₄–NaBH₄ [155]), RHCs based on borohydride-light hydride (LiBH₄-AlH₃ [167]), borohydride-alanate systems (NaBH₄-Li₃AlH₆ [132]), SiS₂-destabilized light borohydrides (6LiBH₄–SiS₂, 8.2 wt%H₂ released with t_onset = 92 °C [103]), TM fluorides destabilize alkali borohydrides (NaBH₄ [196,230]) and ionic liquids (ILs) were also reported to destabilize NaBH₄ (vinylbenzyl trimethylammonium chloride IL [231]). Others have reported confinement strategies to stabilize/destabilize NH₄BH₄ [299], or the synergistic effect of ternary/binary Mg-hydrides (Mg(AlH₄)₂/MgH₂ [257]).

3.3. Cation/Anion Substitution

Utilizing a catalyzed RHC system is a proven strategy to improve hydrogenation behavior. For instance, the nanoconfined 2LiBH₄–MgH₂–0.13TiCl₄ system showed good reversibility behavior possibly due to the formation of Ti-MgH₂ alloys (Mg_0.25Ti_0.75H₂ and Mg₆TiH₂, which can be regarded as cation substitutions of Mg²⁺ in MgH₂ by Ti³⁺ [154]). By forming eutectic compositions, mixed alkali borohydrides could be formulated (Li_0.65Na_0.35BH₄– Li_0.70Na_0.30BH₄ [155]) as well as anion substitutions (I⁻/NH₂⁻ substitutions of BH₄⁻, in LiBH₄–LiI and LiBH₄-LiNH₂ composites producing Li₂(BH₄)_xI_1−x and Li₂(BH₄)_x(NH₂)_1−x [139]). The presence of F⁻ anion in the 3NaBH₄-GdF₃ system produced dehydrogenation in the LT stage of NaBH₄, due to F⁻ substitution of H⁻ [196]. The stability of mixed-cation, mixed-anion borohydrides was also evaluated for solid electrolyte application in batteries (LiM(BH₄)₃Cl, M = La, Ce, Gd [70]).

3.4. Eutectic Formation Approach

A strategy to perform a more effective infiltration of RHC based on mixtures of borohydrides has been utilized recently, with reports of ternary and quaternary mixtures in the LiBH₄-NaBH₄-KBH₄-Mg(BH₄)₂-Ca(BH₄)₂ system [152], LiBH₄-NaBH₄ [66,155], 0.62LiBH₄-0.38NaBH₄ [157], LiBH₄-Ca(BH₄)₂ [158,162,182], LiBH₄–KBH₄ [160,170], LiBH₄-Mg(BH₄)₂ [163] and various combinations of alkali- and alkali-earth metal borohydrides (0.725LiBH₄-0.275KBH_4, 0.68NaBH₄-0.32KBH_4, 0.4NaBH₄-0.6 Mg(BH₄)₂) [234], LiBH₄-M(BH₄)_x (M = Na, K, Mg, Ca) [194].

3.5. Doping Strategy

Various dopants have been used so far to enhance the resorption parameters in hydrogen storage systems: TiO₂(B) nanoparticles embedded in ordered carbon (for AB [288]), TiO (for LiBH₄ [91]), TiCl₄ confined in nanoporous carbon aerogel (for RHC: 2LiBH₄–MgH₂ [154]), nano-Ni (for a 0.62LiBH₄-0.38NaBH₄ [157]), ZrC (for LiAlH₄ [121]), TM fluorides (for NaBH₄, [230]), V₂O₅ or VO₂ (for NaBH₄ [199]), MWCNTs (for RHC: LiBH₄–LiAlH₄ [164]), various dopants introducing strain (Na, K, Al, F, or Cl for LiBH₄ [61]), NiFe₂O₄ (for NaAlH₄ [233]), YCl₃ and Li₃N (for RHC: 6Mg(NH₂)₂-9LiH-LiBH₄ [177]) and ScOCl–functionalized carbon aerogel (for NaAlH₄ [211]), C@TiO₂/Ti₃C₂ (for NaAlH₄ [226]).

In the context of dopants used, the presence of many forms of titanium (TiO, TiO₂, TiO₂(B), Ti₃C₂) typically supported by nanoscaffolds is noteworthy; this important role of Ti in particular may originate from the relative ease with which it can access multiple oxidation states, including metallic state (0, +2, +3, +4).

3.6. Electrolyte-Assisted Dehydrogenation

While nanoconfined borohydrides are usually studied at lower temperatures and confined in siliceous supports for solid-state electrolyte application [67], recent reports employ electrolyte systems as a means to accelerate de-/rehydrogenation of MgH₂/Sn systems (LiBH₄/KBH₄ electrolyte [160]).

3.7. Additives

Many classes of compounds were used as additives [262]: (TM)H_x (TiH₂ [187]); (TM)F_x, chlorides (TiCl₃ [228] [178], TiCl₄ [154], FeCl₃, YCl₃ [177]); fluorides (of transition metals TM [196,230], TiF₃ [161,183], CeF₃ [100,225], K₂NbF₇ [124], NbF₅ [102], YF₃ [197,198], GdF₃ [197] and ScF₃ [198]); iodide (LiI [201]); tetrafluoroborates (LiBF₄ [68]); main group hydrides (LiH, MgH₂ and AlH₃ [167]); sulfide (Ce₂S₃ [94] and SiS₂ [103]); nitride (Li₃N [177], BN and TiO₂(B) [288]); borides and composites (MgB₂/Mg [165]); ferrites (NiFe₂O₄ [233] and MgFe₂O₄ [200]) and hexaferrites (SrFe₁₂O₁₉ [120] and BaFe₁₂O₁₉ [123]); oxides (TiO [91], TiO₂ [126,143,166,226], Fe₃O₄ [98], Al₂O₃ [310], V₂O₅ or VO₂ [199], V₂O₅ supported by MWCNTs [311,312] and various oxides [112]); oxochlorides (ScOCl [211]); hydroxide (KOH [168] and LiB(OH)₄ [108]); metals (nano-Ni [97,125,157,253,293], alkali metals [232], Zr [308], Al [234], Mg [104] and Ti(Al) [129]); intermetallics and derivatives (CoNiB [92] and Ni–Pt [243]); MXenes (Ti₃C₂ [159,225,226,249]); activated carbon [248] and SWCNTs [108].

3.8. Electron-Tuning of the Scaffold

Electron modification of the host component can oftentimes yield positive results. Three-dimensional mesoporous boron nitride BN, prepared as a C-replica of monolithic activated carbon, showed exceptional H₂ storage capacity when nanoconfining AB (8.1 wt% at 100 °C, [292]), while h-BN confirmed a strong catalytic effect in dehydrogenation of LiAlH₄ and Li₃AlH₆ intermediate generated from a LiAlH₄@h-BN nanocomposite [131]. On a similar note, surface-modified AlN (with O-H and C-H groups providing electronic interactions of hydridic protons of LiBH₄ and surface H^δ+) was used as a scaffold for LiBH₄ with increased reversible gravimetric capacity (6.1 wt% when cycling, [105]). Nitrogen-annealing of SWCNTs was shown to facilitate acidic and hydridic hydrogen interactions in a LiBH₄@SWCNT-N nanocomposite, affording dehydrogenation onset as low as 108 °C [108].

3.9. Host Modification

Finally, various modifications of the scaffold allow tuning the reactivity towards improved cycling behavior of complex hydrides, and this approach oftentimes alters the desorption pathway. Various types of scaffolds were investigated: carbonaceous supports– SWCNTs, MWCNTs, carbon nanofibers CNFs, carbon nanospheres CNSs, carbon aerogel CA and ordered mesoporous carbon OMC. The latter includes CMK-3, graphene oxide GO, reduced graphene oxide rGO, modified C-scaffold by incorporation of metals (Al, Zr, Ti, Ni, Pd), siloxanic materials (SBA-15, MCM-41 etc.), metal salts (chlorides, fluorides, iodides, nitrates), bases (KOH), borides, sulfides, ferrites or MXenes. A survey of the nanostructuring procedure of complex hydrides reveals that most of the known classes of substances were used either in neat form (carbon, metals –Al, Ni etc.) or incorporating active catalysts/dopants to alter thermodynamics and accelerate de-/rehydrogenation reactions. These aspects were discussed in relation to the complex hydride utilized, in the previous sections.

4. Outlook and Future Directions

The current state-of-the-art in the field of nanosized complex hydrides was presented, with their most recent advances in hydrogen storage technology. Various tactics for achieving improved thermodynamics and for overcoming sluggish kinetics have been overviewed. These include using thermodynamic destabilization, cation/anion substitution, the eutectic formation approach, doping/additives, choosing the right scaffold and its electron-tuning, along with the—by now proverbial—nanoconfinement. The main advantages and the most resilient drawbacks of using these methods were discussed in relation to practical examples reported in the literature. Additionally, mechanistic insights are discussed and supported by the successful de-/rehydrogenation behavior of reported nanocomposites. Additionally, it is the author’s belief that among the improvement pillars presented before, the electronic synergy direction is not yet fully explored, and is an area where significant advances can be foreseen based on recent developments. Theoretical and experimental research efforts will join hands in order to better predict the composition, hydrogenation and recycling behavior of future nanocomposites for reaching an energy storage goal: the sustainable hydrogen economy.