This technology is the easiest and most widespread hydrogen storage technique, but since hydrogen volumetric density is low, the compression and the volume of the storing tank needs to be high. For this technology, hydrogen is compressed in storage vessels. In practice it only requires three elements: (1) a hydrogen source, (2) a storing cylinder and (3) a compressor. There is currently five generation of compressed hydrogen vessel [10]:

• TYPE I: all metal cylinder;
• TYPE II: load-bearing metal liner hoop wrapped with resin-impregnated continuous filament;
• TYPE III: non-load-bearing metal liner axial and hoop wrapped with resin-impregnated continuous filament;
• TYPE IV: non-load-bearing non-metal liner axial and hoop wrapped with resin-impregnated continuous filament;

Type I cylinder is the first generation of compressed gas storage technology. It is made from either aluminum (M_Al = 26.98 g·mol⁻¹) or steel (M_Fe = 55.85 g·mol⁻¹) and can hold up to 200 bar of hydrogen gas. This technology remains cumbersome, heavy and has the lowest hydrogen energy content. It is mainly used for stationary hydrogen storage such as in industry for large scale use.

Type II cylinder is more sophisticated since it includes a glass fiber with aromatic polyamide or carbon fiber lining around the metal cylinder that can hold around 250 up to 300 bar. About 30-40% lighter than type I, it is about 50% more expensive than type I and still remains cumbersome, heavy and unsuitable for mobile applications [11].

Type III cylinder is made from composite material such as glass fiber with aromatic polyamide or carbon fiber lined with metal. They can hold 300 up to 440 bar. Its weight half less than type II but is twice as expensive to produce and fails to pass the ageing test at 70 MPa.

Type IV cylinder is made from composite materials such as carbon fiber with a polymer liner such as polyethylene or polyamide thermoplastic. Yet this technology is the lightest developed that can withstand the highest pressure. But it is also the most expensive solution [11].

Type V hydrogen vessel is from another technology. It is an experimental generation of cylinder composed of a full composite tanks liner-less. This technology requires further research to be safe enough to be used.

From a practical point of view, Hydrogen storage vessels are used for three purposes: (1) stationary storage for refueling station or industrial use, the target properties are large scale and low cost storage, (2) vehicular application for trains, cars, airship and boats, storage device needs to be light and energy dense and (3) bulk transportation from hydrogen producer to end user, to lower the cost of transportation the main requirements are light weight and large capacity [12].

The main challenge for compressed hydrogen storage is safety. Because a high energy capacity is nothing if it cannot be used properly without causing an explosion. The main phenomenon reducing hydrogen vessel longevity is hydrogen embrittlement. In metal, this phenomenon is due to the absorption of hydrogen that leads to the reduction of the metal ductility. Ductility is the ability to handle deformation without fracture for a metal.

The ISO 15869:2009 norm specifies the requirements for lightweight refillable fuel cylinders intended for the on-board storage of high-pressure compressed gaseous hydrogen or hydrogen blends on land vehicles.

Hydrogen vessels needs to pass various tests such as burst test, leak test, fatigue test bonfire test and the bullet test. The burst test consists of testing the bursting pressure of the vessel. It should be 2 times above the actual working pressure. The permeation test consists of measuring the molecular diffusion of hydrogen through the walls or interstices of the gas vessel over time. The lower the permeation, the better is the storage device. The fatigue test measures the mechanical damage created over cycling of charging/release of hydrogen in the vessel. And the bullet test consists of shooting bullets at the loaded vessel to determine the safety of the device in the case of a shooting by a crazy American.

Liquid hydrogen storage is performed at 20 K in cryo-compressed storage vessel [13]. The main advantage is the higher energy density compared to conventional compressed hydrogen storage technique. The amount of energy required to cool down to 20 K gaseous hydrogen represents 30% of the lower heating value of hydrogen (120.04 MJ·kg⁻¹), while for compression it represents only 15% (119.96 MJ·kg⁻¹).

For long term liquid hydrogen storage two detrimental phenomenon are to take into account: (1) boil off, due to heat transfer between cryo-compressed hydrogen and the environment and (2) evaporation due to hydrogen permeation through the vessel which can represent 2 to 3% of loss per day of storage.

Usually, the vessel is surrounded by an insulation vest to lower heat transfer but increase the weight of the storing device.

SLUSH hydrogen is a mixture of solid and liquid hydrogen around triple point (between −259 °C and −253 °C). It is obtained through pressure decrease/cooling cycling around hydrogen boiling point. In practice, the fluid follows the mechanism: (1) Liquid hydrogen is heated up near its boiling point, (2) pressure is decreased with a vacuum pump, liquid hydrogen becomes gaseous, decreasing liquid hydrogen temperature, at the gas/liquid interface there is the formation of solid hydrogen due to temperature decrease (3) pressure is then increased and solid hydrogen partially melt and sink leaving the surface free. The process is repeated to increase the liquid solid ratio. Hydrogen slush main feature is its higher density (16–20% higher than liquid hydrogen).

The hydrogen atom electronegativity is 2.1 according to Pauling scale. In the science of atomic bonding, the electronegativity is a key parameter to determine the nature of the bonding between two atoms. Hydrogen atoms can be combined to various elements by chemical bonding, the nature and strength of the bonding is related to the electronegativity difference and number of electrons shared in the interaction/relationship (hydrogen atomic number is 1, meaning that it can share 1 electron). Thus, hydrogen atoms can be covalently bonded in chemicals such as ammonia (NH₃), formic acid (HCOOH), methanol (CH₃OH), methane (CH₄) or any Liquid Organic Hydrogen Carrier (LOHC) which correspond to organic compounds rich in hydrogen that can reversibly change their unsaturation degree. Covalent bonding corresponds to the sharing of valence electron between two atomic nuclei. Single bonding strength may vary between 150–600 kJ·mol⁻¹ while for multiple bonding, it may range between 400 and 1100 kJ·mol⁻¹. The higher the electronegativity difference between atoms the weaker is the bond. Reversely, the higher the bond multiplicity, the stronger the bond. The electronegativity difference between the atoms also triggers the polarity of the liaison quantified by the dipole moment. For extreme electronegativity difference, the electron involved in the bond is shared between the valence electron giver and the vacancy holder. In this case, instead of creating a partial charge in each atom (δ+/−), there is an actual charge (Cation+ or Anion−) leading to ionic bonds, whom strength is usually of some hundreds of kJ·mol⁻¹.

Complex metal hydrides are ionic compounds where the anion hold the hydride bonded to an electro-positive metal (usually an alkali metal or alkaline earth metal). They are complexes of atoms from Group I, II and III elements, like NaBH₄, LiNH₂, LiBH₄, Mg₂NiH₄ and Na₂AlH₅ [14,15]. The hydrogen atom is located in the corner of coordination polyhedron and hydrogen desorption is triggered by heat through cascades of decomposition of the compound [14]. These molecules, small molecules and complexes can reach quite high hydrogen density values. For instance, the theoretical value of NaBH₄ is 10.9 wt.-%, and for LiBH₄ it is 18.5 wt.-%, so two to three times higher than compressed hydrogen (5–6 wt.-%). Unfortunately, major technical and economical limitation remains against their broad production and use. For instance, for the use of formic acid, the development of a costly and safe infrastructure that can withstand the corrosive nature of this molecule is necessary [16,17,18]. For methanol, an additional chemical treatment is required to obtain a completely CO-free hydrogen [19]. Most of the hydride complexes, are unstable in water.

Hydrogen can be adsorbed on high specific surface area materials via physisorption. Van der Waals interaction happen via short range dipole-dipole interactions between molecular hydrogen and the host material. Many materials have the ability to adsorb hydrogen such as: carbon-based materials, zeolites, metal organic frameworks (MOF), covalent organic frameworks (COF) or highly porous polymers such as polymers of intrinsic micro-porosity (PIM) and hypercrossed-linked polymers (HCP). Hydrogen storage capacity of those materials is related to the high specific surface area materials due to high porosity, nano-scale or thin powders phase.

Carbon based materials are interesting hydrogen storage materials since carbon is a cheap and abundant resource. It is also relatively light (M_C = 12 g·mol⁻¹) compared to other alternatives such as metal hydrides or physical storage in metal tanks (the average weight of a full type I hydrogen vessel is 60 kg). Carbon can be used within various allotropes forms for hydrogen storage: nanotubes, graphite, fullerene, or carbon foam. Carbon nanotubes are smart folded graphite layers that have a 0.1 wt.-% gravimetric storagecapacity [21]. Hydrogen sorption properties can be tuned by varying the number of walls (single or multi-walls nanotubes) [22], or by crafting metals to dope the carbon. On those structures hydrogen can be adsorbed at the external surface of the tube or on the internal cavity. Hydrogen molecules condense at a high density inside of the nanometric tube [23] enhancing the overall hydrogen storage capacity. For multi walls doped carbon nanotubes, the combination of all the phenomenon leads to a higher hydrogen gravimetric capacity of about 3.7 wt.-% [24].

Generally speaking, carbon-based material storage capacity is mainly dependent on their preparation technique, their processing method, their purity, their structure and their stability. Hydrogen is linked by Van der Waals bonding (6 kJ·mol⁻¹) this means that the adsorption and desorption mechanisms require low activation energy and are thus faster and easier to perform. Ideally, those materials should be experimentally charged at temperature as low as possible, since temperature increase leading to hydrogen desorption.

Hydrogen can be also absorbed in metals forming metal hydrides in which hydrogen atom diffuse inside of the metal crystallographic lattice occupying interstitial sites. These metal hydrides are characterized by metal bonding between hydrogen and the metal atoms. The hydrogen atoms behave as interstitial elements in the metallic host matrix. By this way, the hydrogen storage proceeds under moderate temperature and pressure, giving to the related hydrides the important safety advantage over the gas and liquid storage methods.

Metal and hydrogen usually form two different kinds of hydride phases, α-phase at which only some hydrogen atoms are absorbed and β-phase at which hydride is fully formed. Hydrogen storage in metal hydrides depends on different parameters and consists of several mechanistic steps. Metals differ in the ability to dissociate hydrogen, this ability being dependent on surface structure, morphology and purity [25]. An optimum hydrogen-storage material is required to have the following properties:

• High hydrogen capacity per unit mass,
• Unit volume which determines the amount of available energy,
• Low dissociation temperature, moderate dissociation pressure,
• Low heat of formation in order to minimize the energy necessary for hydrogen release,
• Low heat dissipation during the exothermic hydride formation,
• Reversibility,
• Limited energy loss during charge and discharge of hydrogen,
• Fast kinetics,
• High stability against O₂ and moisture for long cycle life,
• Cyclability,
• Low cost of recycling and charging infrastructures and high safety.

Regarding hydrogen absorption in metal hydrides [26,27,28], in which atomic hydrogen is strongly bonded to the hosting crystal lattices, the exothermic character of the hydride formation reaction induces a heat release upon uptake of each mole of hydrogen. This heat must be removed during charging; otherwise, equilibrium temperature would be reached quickly and the reaction would stop. Conversely, to achieve a fast hydrogen release, it is necessary to supply the heat of the reaction, the desorption being an endothermic reaction. To achieve an efficient thermal management of a metal hydride, it is both necessary to improve the thermal conductivity of the metal hydride, e.g., with the introduction of expanded graphite [29] and to design a complex heat exchanger in the tank. Depending on the metal hydride nature, they can be ground to very fine particles with a high specific surface area [30,31] or directly produced as nanoparticles [30,32] to improve the sorption kinetics, but at the same time, inducing air and/or moisture sensitivity [33,34,35], requiring consequently a polymer coating to prevent the particle degradation [34,36,37,38,39]. Focusing on metal hydride nanoparticles, the achievement of super-stoichiometries may compensate the additional cost induced by coating, making these systems still economically and technically valuable.

These solids have good energy density by volume, although their energy density by weight is often worse than that of the leading hydrocarbon fuels.

Most metal bonds with hydrogen very strongly to form metal hydrides. As a result, high temperatures of around 120–200 °C are required to release their hydrogen content. But these temperatures are not so elevated for application issues. Hence, metal hydride storage is a safe, volume-efficient storage method for on-board vehicle applications.

Noble metals, and particularly Pd, exhibit a significant ability to dissociate hydrogen, making them valuable for metal hydride formation. At room temperature, palladium hydrides may contain two crystalline phases, α and β, the interstitial H in Pd metallic solid solution and the H in Pd defined compound, respectively. Pure α-phase exists at x < 0.017 whereas pure β-phase is achieved for x > 0.58; intermediate x values correspond to α − β mixtures [40]. Hydrogen absorption by palladium is reversible and therefore has been investigated for hydrogen storage. But the excessive Pd metal cost has definitely stopped all technological developments for any hydrogen storage industry implementation.

Light metals such as Li, Mg and Al, form another variety of elemental metal hydrides. These metals interact with hydrogen like Pd, forming first an unstable interstitial solid solution (α-phase) and then a hydride defined compound (β-phase). They are especially interesting due to their light weight and the number of hydrogen atoms per metal atom stabilized in their hydride phase, which is in many cases at the order of M/H = 2, reaching high volume storage density values, higher than that of that of gaseous hydrogen (0.99 H_atom.cm⁻³) or liquid hydrogen (4.2 H_atom.cm⁻³) [41]. It is for instance of 6.5 H_atom.cm⁻³ in the case of MgH₂ hydride [26,42,43].

Catalyzing and nanosizing are methods that can enhance hydrogenation and dehydrogenation in the case of metal/hydride system [42,43]. These processes consist in reducing the hydrogen release temperature and increasing the adsorption kinetic by reducing dehydrogenation enthalpy of hydride, modifying surface energy or/and facilitating hydrogen diffusion. In the case of magnesium hydride, catalyzing can be achieved by addition of metal, e.g., Pt, Mn, Fe in MgH₂ matrix or by mixing MgH₂ and precursors, e.g., TiF₄, HfCl₄, Ni₂P to make composites materials. Nanoengineering techniques enable to reach a few nanometers particles size. For example, MgH₂ nanoparticles (4–5 nm) can lead to a reversible hydrogen capacity of 6.7 wt% @ 30 °C and a dehydriding enthalpy decreased by 22% [42,43].

Intermetallic form another category of metal hydride for solid hydrogen storage. There are classified on the basis of their crystal structures, such as AB₂ type (Laves phase), AB₅ type phases (LaNi₅ as reference compound) and Ti-based body centered cubic (bcc) alloys are well known as hydrogen-storage materials. All these intermetallic are often obtained by combining an element forming a stable hydride with an element forming a non-stable hydride. As for the metallic hydrides, the dissociative chemisorption of hydrogen is followed by hydrogen diffusion into the interstitial sites.

AB₂ type compounds are derived from the Laves phases crystal structures. The potential AB₂ types are obtained with Ti and Zr on the A lattice site and the B sites being occupied by different combinations of 3d element like V, Cr, Mn and Fe. The hydrogen-storage capacity can reach up to 2 wt.-% in V–7.4%Zr–7.4%Ti–7.4%Ni Laves phase [47].

AB5 type alloys exhibit low working temperature and pressure compared to light metals but small hydrogen capacity values. The parent compound LaNi₅ absorbs about 1.0 H/LaNi₅, which corresponds to a hydrogen storage capacity of 1.4 wt.-%, lower than the 7.7 wt.-% of Mg [9,31,44]. Despite such storage performances, the cycling stability of such alloys still need to be improved to satisfy the growing demand of the society. Indeed, long-term hydrogen absorption/desorption cycles often evidence process degradation [48,49,50].

Bcc Ti-based alloys, and their FeTi leader compound, are well-known hydrogen-storage solids with a total hydrogen capacity of around 1.90 wt.-% with inexpensive elements [51]. However, the activation process of FeTi is troublesome due to the formation of titanium oxide layer. Both high-pressure and high temperature are required to achieve a reproducible absorption/desorption of the maximum amount of hydrogen in the compound [44]. New bcc alloys have been reported to absorb more hydrogen than the conventional intermetallic compounds. For instance, Ti–10Cr–18Mn–27V–5Fe and Ti–10Cr–18Mn–32V alloys have hydrogen-storage capacities of 3.01 and 3.36 wt%, respectively [48]. However, the high cost is one of the critical drawbacks limiting their successful practical applications.

Metal hydrides can be obtained by two main techniques: (1) by electrochemistry, the selected metal is immersed in an electrolyte solution, and hydrogen penetrates at the metal-liquid interface or (2) by physical method where the selected metal is immersed in hydrogen gas under high pressure and hydrogen penetrates at the metal-gas interface.

Electrochemistry is the study of chemical reactions triggered by a controlled potential difference. It is a powerful tool to prepare metal hydrides. The best example is that of Ni−MH batteries with the NiOOH/Ni(OH)₂ positive electrode and M/MH negative electrode. The overall charge-discharge reaction in a Ni−MH battery (in alkaline solution) is as follow:

During charging, the metal hydride, MHx, is formed at the anode material M, while the nickel hydroxide Ni(OH)₂ at the cathode is transformed into nickel oxyhydroxide, NiOOH. The stored hydrogen is oxidized and the reduction of NiOOH occurs during the discharge. The M/MH electrode should be capable of reversible hydrogen storage with an insignificant self-discharge. To fulfil the reversibility condition, and thus enable efficient charge-discharge cycles, the bonding between the hydrogen and the metal host should be moderately stable. Very stable and very unstable metal-hydride bonding are unsuitable parameters for the reversibility of the charging-discharging process. Since very stable bonding correspond to high activation energy to trigger the desorption process and very unstable bonding correspond to low stability of the hydride and thus spontaneous desorption of hydrogen from the matrix. Because of this, detailed information about the energetic and kinetic of hydrogen transfer processes occurring on/in electrode materials, as well as the specific energy, charge-discharge efficiency and cyclic lifetime of metal-hydride electrodes, are of a great importance for their future use in commercial batteries.

• Hydrogen electrosorption at the solid-liquid interface:

2.

Hydridation of the solid phase:

where M stands for the metal or the multi-metallic alloy forming the hydride, MH_ads the metal with hydrogen atoms adsorbed on the electrode surface, α−MH the interstitial metal solid solution and β−MH the metal hydride defined compound.

During charging, the hydrogen adsorbed atoms are generated at the electrode surface by the electroreduction of water molecules. The adsorbed hydrogen atom then penetrates and diffuses into the host material, forming first a solid solution (α−MH phase). And as the hydrogen content increase, the interstitial sites of the matrix are fully filled which leads to the formation of the stable metal hydride (β−MH). Hydrogen penetration and diffusion kinetic is far slower than the initial charge-transfer step.

The hydrogen adsorbed atoms participate in chemical and/or electrochemical recombination, resulting in evolution of molecular hydrogen:

The electrochemical method for hydrogen preparation involves the reversal of the above reactions during electrode discharge. Hydrogen atoms are released from the hydride phase or formed at the electrode surface through dissociative chemisorption of molecular hydrogen, and then undergo electrooxidative desorption. A reduction in the surface concentration of hydrogen adsorbed atoms leads to the diffusion of hydrogen from the bulk of the hydride phase towards the electrode-solution interface. It should be noted that the efficiency of hydriding-dehydriding cycles of MH electrodes can be reduced by hydrogen evolution. Hence, it is essential to develop alloy materials that facilitate faster Volmer reaction and hydrogen diffusion in the bulk of the electrode material, compared to Tafel and/or Heyrovsky reactions. For instance, LaNi₅ alloy electrode has been found to achieve maximum capacity (360 mAh·g⁻¹) in the first cycle, but its discharge capacity rapidly declines in the subsequent cycles. However, this deterioration rate can be overcome by partially replacing nickel with cobalt or aluminium, which increases the cycle life of the electrode [54]. Moreover, substitution of Ce, Nd, and Pr at the La site has been shown to result in improved capacity retention during cycling [55].

The corrosion and degradation of the MH electrode represent a significant and serious challenge for Ni-MH batteries. Hydrogen-absorbing alloys are made up of diverse metallic constituents, each of which has a distinctive oxidation potential in an alkaline electrolyte. Nonetheless, there are several approaches to improve the corrosion resistance of the negative electrode. These include adding small amounts of Al, Cr, or Si to shield the underlying metal from further oxidation, enhancing the crystallinity of the alloy through rapid cooling and heating sequences, and slightly reducing the specific surface area of the electrode material [55].

The Sievert method is a thermodynamic process where hydrogen is forced to penetrate the material and the mechanism of absorption happen in 3 steps: (1) under pressure, hydrogen molecules are adsorbed at the surface of the material, (2) hydrogen molecules dissociate into atomic hydrogen, (3) hydrogen atoms penetrate the metal matrix and diffuse in the tetrahedral and octahedral metal lattice sites. Then the metal goes through two states. During hydrogenation, the metal undergoes two phases. Firstly, the formation of the α-phase occurs, where hydrogen atoms (H) and metal atoms (M) combine to create an interstitial solid solution. Here, some of the interstitial sites of the metal crystallographic lattice are occupied by hydrogen atoms, while others remain empty. This solid solution is metastable, indicating that hydrogen can spontaneously desorb from the matrix. Increasing pressure stabilizes a mixture of α and β-phases, which eventually evolves into a unique phase, the β-phase. The β-phase corresponds to the fully saturated metal hydride, which is thermodynamically stable and has its own crystallographic structure that may differ from the starting metal.

The probability of collision between gaseous hydrogen molecules and the metal are increased by increasing the pressure. The dissociation step is triggered by the catalytic property of the material and is thus the kinetic limitation step of the mechanism. This mechanism is shown by the Pressure-Composition-Temperature (PCT) diagram. The solid solution (α-phase) appears first at low hydration ratio until reaching a plateau pressure where the metastable α-phase turns into thermodynamically stable β-phase. The structural transition between the metal crystallographic structure (α-phase) to the hydride one (β-phase) happens at constant pressure. It means that both phases coexist in the plateau pressure until hydrogen saturation in the material, thus the formation of the thermodynamically stable hydride.

In real materials, hydrogen intake and extake P-C-T curves may differ from ideal systems. Experimentally, the α to β phase transition plateau pressure has a slope increasing with temperature. This phenomenon can happen due to temperature heterogeneity in the sample leading to kinetic disparity in the material. It can also be due to the intrinsic heterogeneity of the material (due to alloying or doping of the metal matrix) or the physical properties change (such as deformation capacity) in the transition from the solid solution to the hydride. Another phenomenon is the hysteresis between hydrogen intake and extake P-C-T curves. In thermodynamic, hysteresis is a proof of a non-reversible phenomenon. There are three main reasons for such a hysteresis: (i) the pressure needed for hydride formation is greater than that of hydride decomposition, (ii) the terminal solubility of hydrogen in the host alloy is greater for hydride formation than for hydride decomposition and (iii) the hydride formation temperature is lower than the hydride decomposition temperature [56].

The reached hydrogen storage capacity depends on the nature of the β phase. Some metals or intermetallics exhibit only one metal hydride phase, others may exhibit several, each existing in a specific T and P domain. For instance, in the FeTi–H system there are several hydride crystalline phases corresponding to different formation enthalpy values:−28.10 kJ/mol_H2 for FeTiH and −33.72 kJ/mol_H2 for FeTiH₂, respectively [57], the latter being formed in a T and P domain closer to the ambient one.

In the past three decades, an innovative solid-gas hydriding process has been explored as a means of preparation of solid hydrogen storage material. This method involves the implantation of hydrogen ions into a metal matrix under a potential difference between a plasma source and the sample.

The plasma ions energy in the sheath is assumed to be the difference between the plasma potential V_p and the substrate potential V applied:

V_p being the plasma potential and V the bias voltage.

Typically, positively charged hydrogen ions generated in the gas plasma are accelerated towards the negatively polarized sample, where they penetrate at the extreme surface of the metal. They then recombine with an electron from the sample according to the reactions H⁺ + e⁻ → H and H₂⁺ + e⁻ → H₂, before diffusing into the metal matrix. One major advantage of this technique is that it eliminates the need for dissociation of hydrogen molecules into atomic hydrogen, a feature that is directly related to the catalytic properties of the host matrix. However, the main disadvantages of this method are the high cost of energy required for plasma generation and the extremely thin thickness of hydrogen implantation in the sample. In other words, the formation of metal hydride can occur only in the outermost atomic layers of the metal sample.

Very few works exist in the literature even if all the authors claimed that plasma-based ion implantation could be used as an effective route for metal hydrogenation. Historically, plasma-based ion implantation was first studied by Myers et al. to tentatively achieve deuterium D superstoichiometries in high purity annealed Palladium foil of 0.25 mm in thickness [58]. Experimentally, deuterium with an energy of 10 keV were obtained by accelerating D₃⁺ through a potential drop of 30 kV. The implantation beam was magnetically separated to exclude D⁺, D₂⁺, and impurity species and was swept for uniformity. Pd foils were deposited on continuously cryogenically cooled copper block. Within these conditions, they succeeded to introduce D atoms into Pd at atomic ratios greater than one by ion implantation at cryogenic temperatures. They observed that at implantation temperatures of 41 and 81 K, a saturation concentration ratio [D]/[Pd] of 1.6 was reached, substantially above the limit of 1:0 observed in gas-phase charging. But as the temperature was ramped upward, [D]/[Pd] abruptly decreased to approximately 1.0 near 120 K, reflecting a process of accelerated transport unique to the superstoichiometric state. They suggested that the diffusion mechanism was driven by D hopping among octahedral and tetrahedral interstitial sites [58]. Of course all these phenomena proceeded at the Pd surface within a maximal depth of 100–150 nm, depending on deuterium ions energy.

Replacing deuterium gas by hydrogen gas, Tavares et al. evidenced that hydrogenation of high purity annealed palladium foil of 0.14 mm in thickness by plasma ion implantation is possible. Within their operating conditions they observed a rearrangement of the fcc metal lattice with the formation of a vacancy-ordered structure. They showed thereby that very high pressure hydrogenation is not mandatory to induce superabundant vacancy phase generation, which might further trap hydrogen atoms [59].

Experimentally, they performed their implantation of hydrogen experiences using a distributed electron cyclotron resonance (DECR) plasma reactor. Their plasma was generated in an argon 10%—hydrogen 90% gas mixture at a total pressure of 0.33 Pa for a 1.8 kW microwave input power (2.45 GHz). They insisted on the role of argon. It is assumed to help sustaining the hydrogen-based discharge and cleaning the substrate surface (carbon and oxygen desorption). During the processing, the substrate-holder was continuously cooled with a circulation of insulating oil, to make the substrate temperature never higher than 200 °C, under the operating conditions.

The formed Ar⁺ (major), Ar₂⁺ and Ar₂⁺ (minor), H⁺ (major), H₂⁺ and H₃⁺ (minor) ion species were accelerated onto the substrate under a pulsed voltage difference. The pulse voltage (V), duration (τ) and frequency (f) were fixed to −40 kV, 38 ms and 50 Hz, respectively. Therefore, the total dose (cm⁻²) implanted in palladium was proportional to the total time of high voltage pulses, also named the effective time of implantation (T) which was fixed to 12 s (leading to a processing time t of 106 min). It was estimated to be of the order of 5 × 10¹⁶ cm⁻². As previously the implantation in Pd was proceeded into to the extreme surface. The implantation depth was calculated and found to be about 201 nm, 115 nm and 24 nm at 40 keV for H⁺, H₂⁺ and Ar⁺ plasma ions respectively [59].

Some years later, 20 nm thick Pd coatings deposited on Si substrates with 800 nm SiO₂ and 1 nm Cr buffer layers were treated by Wulff et al. in a 2.45 GHz microwave plasma source at 700 W plasma power and 40 Pa working pressure without substrate cooling or heating. They tuned the sample voltage from 0 to −150 V at constant gas flow and demonstrated hydrogen ion implantation and α-PdH and/or β-PdH phases formation, crystallizing all within the fcc cubic structure [60].

At 0 V, solid solution and hydride (PdH_0.14 and PdH_0.57) were formed. At −50 V, only hydride phases were obtained. Typically, PdH_0.57 was formed directly. At −100 V and −150 V, shrinking of the unit cell were observed assumed to be due to the formation of two fcc vacancy palladium hydride clusters PdH_vac(I) and PdHvac(II). Under longtime plasma exposure the fcc PdH_vac(II) phase led to the cubic PdH_1.33 hydride [60].

In parallel to these studies on Pd thin films, plasma hydrogen ion implantation was also successfully carried out on Ni films. Typically, 20 nm thick nickel films were exposed to Ar (10%)—H₂ (90%) microwave plasma using different negative bias voltages to study the hydride formation [61,62]. Plasma power and gas pressure were kept constant at 700 W and 50 Pa. It was established that without external bias voltage no chemical reaction occurred. At negative substrate voltages (−10 V, −25 V, −50 V, −75 V) a hexagonal Ni₂H phase was formed in a first quick reaction step.

In subsequent plasma chemical reactions Ni₂H was transformed into cubic NiH, the reaction transformation rate increasing with increasing the negative bias voltage. These results were very interesting, since the applied hydrogen gas pressure did not exceed 50 Pa while Ni metal usually requires ultrahigh hydrogen gas pressure to form hydrides by classical hydrogenation route. Indeed, according the Ni−H phase diagram established by Shizuku et al., stable hydrides NiH_x can be formed up to x = 0:8 at temperatures T lower to 800 °C, and quite high hydrogen pressures (1.1 and 5.4 GPa) [63].

All these preliminary results confirmed that plasma-based ion implantation may offer an alternative route for solid hydrogen storage and open real perspectives replacing ultra-thin films by finely divided metal particles with average sizes comparable to the tested film thicknesses. In other words, metal hydrides would be generated at low-temperature and low-pressure using microwave assisted Ar/H₂ (10/90) plasma. The gas mixture appeared to be mandatory [59]. Argon ions facilitate the ignition of the plasma, its maintenance and its stability as they pretreat the metal surface, eliminating adsorbed carbon and/or oxygen-based residues. And since Argon ions penetrate much less deeply than the ions of interest hydrogen ion species [59] they should not affect the metal interaction with hydrogen species.

This issue was also successfully explored on compacted (porosity of ~50%) polyol-made Nickel particles of ~25 nm diameter to Ar (10 %)—H₂ (90%) using a microwave plasma (2.45 GHz) for 2 h of exposition [64] leading to the production of similarly Ni₂H particles. Plasma power and gas pressure were kept constant at 180 W and 1 mbar, respectively. Plasma treatment was carried out for an incident ion energy range of 10 to 20 eV. In these conditions, argon and hydrogen ions do not allow defect formation in Nickel and the sputtering yield is negligible, they can be only implanted [65]. The ionic current recovered at the sample holder was stabilized a couple of minutes after the ignition of the plasma. For a current density crossing the Ni pellet surface of 5 mA.cm⁻² an incident ionic flux in the order of 10¹⁶ ions.cm⁻².s⁻¹ was expected. To avoid any uncontrolled heating of the exposed samples, and consequently implanted hydrogen release, the as-prepared Ni pellets were deposited on a copper sample holder continuously water cooled during the whole plasma experiments.

Despite all the highlighted added value of the so-called cold plasma hydrogen ion implantation metal hydride processing remains still confidential. Indeed, usually, plasma hydrogen ion interaction with metallic nanostructures, reside in the specific use of high energy density processes, which may generate sever damages (hydrogen bubbles, blistering...) in the exposed metals [66]. These defects usually act as hydrogen trapping sites, which are detrimental to hydrogen storage, as already observed in ball-milled metals, that are known to usually form hydride easily like LaNi₅ [67,68] and ZrCo [69,70]. Besides, plasma generation is electrical energy costly and this must be considered in any energy yield calculation.