Syntheses of Novel Hydrides Containing Light Elements under High Pressure and High Temperature

Abstract

This paper describes high-pressure synthetic studies on novel hydrides. High-pressure hydrogenation experiments are carried out using a cubic-type multi-anvil apparatus. In-situ synchrotron radiation X-ray diffraction measurements are effectively used to explore synthetic conditions, to investigate the reaction mechanisms, and to characterize the thermodynamic stabilities of the obtained hydrides. Theoretical calculations based on first-principles calculations enable us to predict the thermal stability and crystal structure of the target hydrides before the high-pressure experiments, which leads to the rapid discovery of the novel hydrides. Lithium-containing hydrides, YLiFeH₆, LiNiH₃, Li₄FeH₆, and Li₃AlFeH₈ are synthesized. Syntheses of aluminum-based alloys hydrides, Al₂CuH and Al₃FeH₄ under high-pressure are also described. These results demonstrate that the high-pressure technique is useful for discovering novel hydrides.

1. Introduction

Synthesis of novel hydrides with functionalities such as hydrogen storage, high-temperature superconductivity,^1–4) fast ionic conduction,⁵⁾ and H⁻ conduction⁶⁾ is being investigated. The reason for such a wide variety of functionalities of hydrides is attributed to the fact that hydrogen can possess a wide variety of chemical states in materials. Recent advances in synthesis, measurement, and theoretical calculation techniques have made it possible to control the chemical state of hydrogen, and as a result, novel functional hydrides have been realized one after another.^7,8) High-pressure and high-temperature synthesis, which is the subject of this paper, is one of the most promising methods for syntheses of novel hydrides.

Applying high pressure to a material induces structural phase transitions, chemical reactions, and other phenomena, which do not occur around ambient pressure. The technique is widely applied in many branches of physics, chemistry, and biology.⁹⁾ High-pressure methods are also effective in the synthesis of novel materials. High-pressure methods are widely used for the synthesis of super-hard materials such as diamond, oxide superconductors, and dielectrics. It is also effective in the synthesis of new hydrides. This is mainly because highly pressurized hydrogen becomes reactive owing to a steep increase in the chemical potential of hydrogen¹⁰⁾ and it reacts with metals to form novel hydrides that can be synthesized only under high pressure. In the present paper, new hydrides recently synthesized by the author’s group under high pressure and high temperature are presented along with high-pressure hydrogenation techniques.

2. High-Pressure Hydrogenation Technique

Baranowski et al. have reported a pioneering high-pressure hydrogenation study using a piston-cylinder apparatus, in which the hydrogenation of pure nickel was demonstrated.¹¹⁾ Subsequently, Antonov et al. conducted a high-pressure hydrogenation study using a toroid-type apparatus. Fluid hydrogen evolved from an internal hydrogen source was sealed in a metal capsule and reacted with a metal sample. The high-pressure apparatus was designed so that the formed hydride could be quenched at the temperature of liquid nitrogen. Thermodynamically unstable hydrides were recovered at ambient pressure and precisely characterized.¹²⁾ They investigated hydrogenation reaction of transition metals intensively and characterized crystal structures of recovered hydrides by neutron diffraction.¹³⁾

Cubic-type multi-anvil presses, which are used by the authors group, are wildly used to synthesize novel hydrides under high pressure. Figure 1(a) shows a schematic diagram of a high-pressure generation using a cubic-type multi-anvil press. In this apparatus, a sample is set in a cubic pressure medium and compressed isotropically from six directions by pistons called anvils to generate high pressure at the sample position. Anvils are usually made of tungsten carbide. A pressure medium is made of clay mineral called pyrophyllite or a mixture of amorphous boron and epoxy resin. A truncation edge length of an anvil is about two-thirds a size of a cubic pressure medium. Therefore, when a pressure medium is compressed by anvils, a portion of the pressure medium is pushed out into gaps between anvils. The part extruded into the gaps functions as a gasket to support the pressure gradient.

Fig. 1

(a) Schematic of a cubic-type multi-anvil press. (b) Inner part of a cubic pressure medium for high-pressure hydrogenation experiments.

A sample is heated by a cylindrical graphite heater set together with a sample in a pressure medium. Figure 1(b) shows a schematic diagram of an inside of a cubic pressure medium. Electrodes are placed above and below a graphite heater. These electrodes contact upper and lower anvils for pressurization. Because anvils are metal, electric power can be supplied to a heater through anvils and electrodes. By supplying electric power to a heater, a sample is heated by Joule heat generated. A cubic-type multi-anvil press can generate temperatures of 2000°C or higher by appropriately combining these parts. However, due to the limitation of the melting points of the parts required for high-pressure and high-temperature hydrogenation as described below, the maximum temperature reached in high-pressure hydrogenation experiments is about 900°C at a pressure of 10 GPa.

A temperature can be measured using thermocouples under high pressure. Most of heat generated by a cylindrical graphite heater is released through anvils and press apparatus, thus, there is a proportional relationship between applied electric power and temperature at a sample position. A relationship between applied electric power and temperature was determined in advance and a temperature is estimated based on applied electric power and the relationship during synthesis experiments. The reproducibility of the temperature in this case is about ±20°C.

In the 1980s, Wakamori et al. used this apparatus to synthesize rare-earth hydrides,¹⁴⁾ in which a metal capsule similar to that used by Antonov et al. was used to confine high-pressure hydrogen. In the 1990s, Fukai et al. developed a new high-pressure cell design. A hydrogen-sealing capsule made of NaCl was used.¹⁵⁾ Fukai and co-workers have intensively studied metal-hydrogen systems under high pressure and high temperature.^10,16) They observed the formation of hydrogen-induced super abundant vacancies (SAV) in metal lattices under high pressure and high temperature.¹⁵⁾ SAV is thought to be a possible cause of hydrogen embrittlement.¹⁷⁾

The above-mentioned high-pressure hydrogenation technique developed by Fukai was used by the authors and is described in more detail here. Although hydrogen gas can be loaded into a compact high-pressure apparatus, such as a diamond-anvil cell,^18–20) it is difficult to load hydrogen gas into a high-pressure cell for a cubic-type multi-anvil apparatus. Instead of hydrogen gas, solid chemical hydrides are used as an internal hydrogen source. The internal hydrogen source evolves hydrogen when it is heated under high pressure. Powder mixtures of NaBH₄ and Ca(OH)₂ (molar ratio 1:2), AlH₃, and LiAlH₄ are typically used as the internal hydrogen source. Recently, BH₃NH₃ has also been used for internal hydrogen source.^21,22) The residue after the decomposition reaction of BH₃NH₃ at high pressure is a derivative of inert hexagonal BN.²³⁾ BH₃NH₃ with a high hydrogen density is a superior internal hydrogen source for high pressure hydrogenation.

To confine hydrogen evolved from the internal hydrogen source, a hydrogen-sealing capsule made of NaCl is used as described above. The hydrogen-sealing capsule can confine hydrogen at 10 GPa and 800°C for more than 24 h.²⁴⁾ A sample is placed in a capsule made of BN, which allows hydrogen permeation but prevents contamination from residues of the decomposed internal hydrogen source. Instead of a BN capsule, kaolin wool has also been used for the same purpose.¹³⁾ The sample and the BN capsule are placed in the hydrogen-sealing capsule along with the pelletized internal hydrogen source. The hydrogen-sealing capsule is located in a cylindrical graphite heater. When electric current is supplied to a graphite heater, the inner parts of the heater (the sample, the BN capsule, the internal hydrogen source pellets, and the hydrogen-sealing capsule) are heated. Hydrogen is then released from the internal hydrogen source pellets and confined in the hydrogen-sealing capsule. Hydrogen can permeate in the BN capsule. The sample is immersed in hydrogen fluid and high-pressure and high-temperature hydrogenation reaction is realized. This technique has been used to synthesize novel hydrides under high pressure. Novel hydrides containing light-weight elements, such as lithium and magnesium, have been synthesized via the technique.^21,25–33) A number of reviews of high-pressure hydrogenation studies are also available.^10,13,34,35)

3. Recent Syntheses Studies of Novel Hydrides under High Pressure

3.1 Lithium containing complex hydrides

Complex hydrides were exhaustively investigated by around 2000,³⁶⁾ and it was thought that it would be difficult to synthesize new complex hydrides. In recent years, some new complex hydrides have been synthesized through a combination of theoretical predictions and high-pressure and high-temperature synthesis. In the present paper, a series of synthesis studies of Lithium-containing complex hydrides is explained.

3.1.1 LiYFeH₆

YMn₂ alloy with a MgCu₂-type structure is hydrogenated via a two-step process:^37,38)   

\begin{equation*} \text{YMn$_{2}$} + \text{3H$_{2}$} \to \text{YMn$_{2}$H$_{4.5}$} + \text{3/4H$_{2}$} \to \text{YMn$_{2}$H$_{6}$} \end{equation*}

The first hydrogenation reaction yields an interstitial hydride, YMn₂H_4.5, in which the YMn₂ metal lattice retains the MgCu₂-type structure. In contrast to the first hydrogenation reaction, the second hydrogenation reaction produces a complex hydride with a K₂PtCl₆-type structure in which half of the manganese atoms are surrounded by six hydrogen atoms to form complex anions, [MnH₆].^5,37–39) The rest of the manganese and yttrium atoms randomly occupy the same crystallographic sites in the ionic states Mn²⁺ and Y³⁺. Total charge neutrality is maintained as Y³⁺Mn²⁺[MnH₆]⁵⁻. Similar hydrogenation processes are expected for other alloys containing complex anion-forming atoms, such as iron, cobalt, and nickel.

We attempted to synthesize an iron-containing complex hydride from YFe₂ with the same MgCu₂-type structure as that of YMn₂. The hydrogenation reaction of YFe₂ yields YFe₂H_x (x < 5).⁴⁰⁾ A complex hydride containing the [FeH₆]⁴⁻ anion was not obtained via the hydrogenation reaction of YFe₂. Yttrium tends to be a trivalent cation, therefore, we inferred that the charge imbalance between Y³⁺ and [FeH₆]⁴⁻ prevented the formation of an ion-containing complex hydride. Here, the lightest metal element, lithium, was added to adjust the imbalanced charge. We hypothesized that the complex hydride YLiFeH₆ (Y³⁺Li⁺[FeH₆]⁴⁻) would be synthesized by the addition of lithium as an electron donor.⁴¹⁾

A powder mixture of YFe₂ and LiH with a molar ratio of 1:1 was compacted into a disk and was hydrogenated at 6 GPa and 800°C for 12 h. After the hydrogenation reaction, the sample was quenched at room temperature and depressurized. The recovered sample was characterized by synchrotron radiation X-ray diffraction (SR-XRD) measurement at the BL22XU beamline of SPring-8, Japan.

To investigate the crystal structure of YLiFeH₆, we performed a structural optimization by using first principles calculations. The calculations show that YLiFeH₆ has a K₂PtCl₆-type crystal structure with fcc symmetry. The simulated profile well reproduced the observed one, indicating that YLiFeH₆ is formed by hydrogenating the powder mixture of YFe₂ and LiH. The calculated crystal structure is shown in Fig. 2. The calculated electronic structure indicates that lithium acts as an electron donor to stabilize the hydride.

Fig. 2

Schematic of the crystal structure of YLiFeH₆.

The synthesis temperature was limited to 800°C; an unknown high-temperature phase appeared when the powder mixture was hydrogenated at 900°C, whereas full hydrogenation was not achieved below 700°C. When the sample was hydrogenated at 800°C, small Bragg peaks that were not from YLiFeH₆ were observed. These peaks were not indexed by known pure metals, alloys, or hydrides, suggesting that an unknown phase was partially formed. When the sample was hydrogenated at 900°C for 2 h, the unknown phase became dominant; therefore, the unknown phase was thought to be a high-temperature phase. The high-temperature phase was recovered under ambient conditions. Hydrogen evolution from the high-temperature phase was observed around 430°C at ambient pressure, which was comparable to the dehydrogenation temperature of YLiFeH₆ (460°C). When the sample was hydrogenated at 700°C for 12 h, YH₂ remained unreacted. High temperatures above 700°C are necessary to complete the hydrogenation reaction.

3.1.2 LiNiH₃

Complex anions [NiH₄]⁻ are reported to be stabilized with alkali metal counterions.^42,43) Considering charge neutrality, a lithium-containing complex hydride, LiNiH₄ could be synthesized. However, first-principles calculations have predicted that LiNiH₃ with a perovskite-type structure is thermodynamically stable⁴⁴⁾ and is synthesized through the reaction, LiH + Ni + H₂ → LiNiH₃. To confirm the theoretical prediction, we carried out the high-pressure synthesis of LiNiH₃. The reaction process was investigated by in situ SR-XRD⁴⁵⁾ at BL14B1, SPring-8.⁴⁶⁾

A powder mixture of LiH and nickel with a molar ratio of 2:1 was used as the starting material. The excess LiH was added because lithium diffuses from the sample container under high pressure and high temperature. This is in contrast to YLiFeH₆, in which excess LiH was not added. The diffusion of lithium may depend on the material being synthesized. The sample was pressurized to 3 GPa at room temperature, heated to 600°C for 3 min, and kept at 600°C for 250 min.

An in situ SR-XRD measurement revealed that LiNiH₃ is formed through a three-step reaction. Figure 3 shows a series of X-ray diffraction profiles taken at 3 GPa and 600°C. Immediately after the sample was heated to 600°C, Bragg peaks from nickel with an fcc structure shifted toward the lower energy side. These peak shifts indicated the lattice expansion of nickel, which was caused by the formation of NiH_x (x < 1). Fifty minutes after the sample was heated to 600°C, new Bragg peaks appeared on the lower energy sides of the NiH_x peaks. These peaks were indexed by a larger fcc lattice than that of NiH_x, indicating that a Li_yNi_1−yH solid solution was formed. After the Li_yNi_1−yH solid solution became a single phase, drastic changes in the diffraction profile were observed. The new peaks were indexed with a simple cubic lattice, indicating the formation of LiNiH₃. The hydride was recovered at ambient conditions. The diffraction profile taken at ambient pressure agreed well with that calculated one for the theoretically predicted crystal structure.⁴⁴⁾ It was confirmed that LiNiH₃ is synthesized via the hydrogenation reaction of a powder mixture of LiH and Ni at 3 GPa and 600°C.

Fig. 3

Series of in-situ X-ray diffraction profile of a powder mixture of LiH and nickel hydrogenated at 3 GPa and 600°C. Open squares, closed circles, and open circles indicate diffraction peaks from NiH_x, Li_yNi_1−yH solid solution, and LiNiH₃, respectively.

The Li_yNi_1−yH solid solution is considered to be the precursor of the perovskite structure. The fcc lattice of metal atoms remained in the Li_yNi_1−yH solid solution. In the fcc structure, we can choose a body-centered tetragonal (bct) unit cell, which is similar to the perovskite lattice. The perovskite structure can be derived from the Li_yNi_1−yH solid solution by ordering the metal atoms and expanding the metal lattice through hydrogenation. The presence of the precursor would enable us to hydrogenate the powder mixture at a lower temperature than that of YLiFeH₆ and Li₄FeH₆ (see below).

3.1.3 Li₄FeH₆

After the successful synthesis of YLiFeH₆, we tried to replace yttrium with lithium to synthesize Li₄FeH₆, which contains an [FeH₆]⁴⁻ complex anion in the same manner as YLiFeH₆. First-principles calculations indicated that Li₄FeH₆ can be synthesized through the reaction, 4LiH + Fe + H₂ → Li₄FeH₆, and it has the K₄CdCl₆ type structure (Fig. 4), which is isostructural with Li₄RuH₆ and Li₄OsH₆.⁴⁷⁾ We demonstrated the formation of theoretically predicted Li₄FeH₆ by hydrogenating a powder mixture of LiH and iron at high pressures and high temperatures.^48,49)

Fig. 4

Schematic of the crystal structure of Li₄FeH₆.

The starting material was a powder mixture of LiH and iron with a molar ratio of 6:1. The excess LiH was added to compensate for lithium diffusing from the sample capsule in the same manner as in the case of LiNiH₃. The starting materials were pressurized to the target pressure (≤9.7 GPa) and then heated to the preset temperature (≤900°C). The sample was kept in hydrogen fluid at the constant preset temperature, quenched at room temperature, and depressurized.

In-situ SR-XRD measurement revealed the hydrogenation process of the powder mixture of LiH and iron. A body-centered cubic (bcc) to fcc structural phase transition of iron occurred during heating, which is known structural phase transition of iron under high pressure and high temperature. New Bragg peaks appeared immediately after the sample was heated to 900°C. After the formation of fcc iron, new Bragg peaks were observed. These Bragg peaks were indexed by the unit cell of the theoretically predicted Li₄FeH₆, thus confirming the formation of Li₄FeH₆. The Bragg peaks from Li₄FeH₆ became strong over time. Li₄FeH₆ became a single phase after a 30 min heat treatment. The formed Li₄FeH₆ was recovered at ambient conditions, although partial decomposition of the Li₄FeH₆ was observed during decompression.

The same experiments were conducted at different pressures and temperatures. The results are summarized in Fig. 5. Two important conclusions can be drawn. First, Li₄FeH₆ is thermodynamically stable near ambient conditions. We extrapolated the observed decomposition curve, and roughly estimated the decomposition temperature at ambient pressure. The decomposition temperature was estimated to be slightly lower than room temperature, which is consistent with the partial decomposition of Li₄FeH₆ during decompression. Second, full hydrogenation reactions were achieved above 900°C irrespective of the pressure. It is likely that Li₄FeH₆ is formed only at the boundary of iron and LiH particles in the powder mixture at low temperatures. The Li₄FeH₆ separates the unreacted iron and LiH particles and prevents the hydrogenation reaction from continuing.

Fig. 5

Reaction pressure-temperature diagram of a powder mixture of LiH and iron under high hydrogen pressure. Li₄FeH₆ is thermodynamically stable below the curve. Full hydrogenation was achieved only when a powder mixture was hydrogenated at a pressure-temperature conditions above the dotted line.

This is in contrast to the hydrogenation reaction of the powder mixture of LiH and nickel to form LiNiH₃ (see Section 3.1.2). In the case of LiNiH₃, NiH_x and LiH form a solid solution, Li_yNi_1−yH, which enables the lithium and nickel atoms to homogenize and acts as a precursor for the hydride formation. Formation of the solid solution allows the hydrogenation reaction to occur at a relatively low temperature of 600°C. When the powder mixture of iron and LiH is fully hydrogenated, iron and LiH are directly hydrogenated without forming any intermediate products, such as Li_yFe_1−yH, as mentioned above. When the powder mixture of iron and LiH is partially hydrogenated at low temperatures of less than 900°C, the remaining iron is hydrogenated to form FeH_x (x ≤ 1 below 10 GPa⁵⁰⁾). Despite of the formation of FeH_x, the Li_yFe_1−yH solid solution is not formed. The iron and LiH particles remain separated owing to the absence of a solid solution at low temperatures. This is one reason why high temperatures are necessary to hydrogenate the powder mixture fully. Consequently, it is necessary to stabilize Li₄FeH₆ at 900°C and a high pressure (above 6.1 GPa) is therefore needed to hydrogenate the powder mixture fully.

3.1.4 Incorporation of H⁻ anions

Takagi et al. reported a guideline for syntheses of new complex hydrides, where H⁻ anions were incorporated to enhance turnabilities of complex hydrides.^7,51) In order to satisfy charge neutrality, a number of combinations cations is limited as illustrated in Fig. 6(a). Here, H⁻ was incorporated to increase the number of combinations of cations (Fig. 6(b)). As a result, a wide variety of hydrides can be synthesized. They evaluated the stability of complex hydrides with H⁻ by first-principles calculations and showed that they can be explained in terms of the electronegativity of cations. Based on these results, we have tried to synthesize theoretically predicted Li₃AlFeH₈.

Fig. 6

Number of combinations of cations in iron containing complex hydride (a) without incorporation of H⁻ anions and (b) with incorporation of H⁻ anions.

A powder mixture of LiH, AlH₃, and pure iron at a molar ratio of 3:1:1 was hydrogenated at 5 GPa and 600°C, and the reaction process was monitored by in-situ SR-XRD measurement system. As time progressed, new Bragg peaks began to appear. These peaks can be indexed by the theoretically predicted unit cell of Li₃AlFeH₈, and together with the analysis results of the recovered sample, it was confirmed that the target Li₃AlFeH₈ was obtained by hydrogenation at 5 GPa and 600°C. A schematic of the crystal structure of Li₃AlFeH₈ is shown in Fig. 7(a). The time variation of the peak intensities indicated that the hydrogenation reaction was completed in approximately 5 hours. Since peaks from unknown phases other than Li₃AlFeH₈ were also observed after the completion of the reaction, it was also clear that a single phase of Li₃AlFeH₈ could not be obtained at the temperature and pressure conditions.

Fig. 7

Schematics of the crystal structures of (a) Li₃AlFeH₈ (b) LiAlFeH₆.

Unfortunately, the single phase of Li₃AlFeH₈ was not obtained at temperatures and pressures up to 9 GPa and 900°C. It was, however, found that several new hydrides were formed by changing the temperature and pressure conditions. Although not all the phases could be identified, one of them was found to be H⁻ free LiAlFeH₆ (Fig. 7(b)). For LiAlFeH₆, after determining the unit cell of the crystal structure, the chemical formula was estimated from a trend of volume filling ratios of ions in unit cells of hydrides,⁵²⁾ and the composition and crystal structure were determined by first-principles calculations based on the estimated chemical formula.

3.2 Aluminum-based hydrides

Compounds of metal and hydrogen are called metal hydrides, and among them, alloys that can easily absorb and desorb hydrogen under moderate temperature and pressure conditions are called hydrogen storage alloys. There are very well-known guidelines for obtaining hydrogen storage alloys and hydrides. The guideline is to alloy metals with high and low hydrogen affinities.⁵³⁾ Here, a metal with high hydrogen affinity refers to an element that can easily form metal hydride with a hydrogen-to-metal atomic ratio (H/M ratio) of 0.5 or higher near ambient pressure. In contrast, a metal with low hydrogen affinity refers to an element that cannot form a metal hydride such that H/M ≥ 0.5 near ambient pressure. It is also known that hydrogen affinity can be well classified by the periodic table. Elements in groups 1 through 5 have high hydrogen affinity, while elements in groups 6 through 13 have low hydrogen affinity with the exception of Pd. It has been reported that the hydrogen affinity of a metal is related to the softness of the metal.⁵⁴⁾ Let us now review the guideline for the search of hydrogen storage alloys and hydrides. Hydrogen storage alloys must contain at least an element with high hydrogen affinity. However, all metals with high hydrogen affinity are expensive with the exception of Mg, making it difficult to reduce costs. In addition, the choice of lightweight elements is also limited, making it difficult to increase the hydrogen density per weight. Mg and its alloys form very stable hydrides. It is difficult to release hydrogen from Mg and its alloy hydrides, and research and developments are underway to make it easier to release hydrogen. Thus, under the conventional search guidelines for hydrogen storage alloys, it is considered difficult to obtain a material that can solve the problems of hydrogen storage alloys, such as gravimetric hydrogen density and cost. To solve the problems, we are conducting exploratory research on hydrides of alloys composed solely of metals with low hydrogen affinity, contrary to the conventional guidelines.

The main target of the exploratory research is aluminum-transition metal alloys, where both aluminum and transition metals are with low-hydrogen affinities. Aluminum seems to be a promising material for hydrogen storage applications because of its light weight, abundance, and safety. Complex hydrides called alanate compounds (NaAlH₄, LiAlH₄, etc.), which are composed of complex ions formed by aluminum and hydrogen, have been studied. However, no practical hydrogen storage material has been realized. If we can realize aluminum-based alloy hydrides other than alanates, the scope of exploratory research on hydrogen storage materials using aluminum is expected to increase dramatically. However, as aluminum alloys are used as materials resistant to hydrogen embrittlement, the reactivity of hydrogen with aluminum alloys is extremely low, and the realization of aluminum-based alloy hydrides has been considered difficult.

3.2.1 Al₂CuH

We have clarified that Al₂Cu alloy can be hydrogenated above 10 GPa and forms interstitial hydride, Al₂CuH.⁵⁵⁾ Figure 8 shows in situ SR-XRD profiles of Al₂Cu alloy hydrogenated at 10 GPa. The second profile from the bottom was measured immediately after the Al₂Cu alloy was heated to 600°C in a hydrogen fluid at 10 GPa. At this stage, the sample retained its Al₂Cu-type structure, which is an ambient pressure phase. No hydrogenation reaction was observed when the alloy was kept in the hydrogen fluid for 24 hours or when the temperature was raised to 850°C in steps of 50°C. When the temperature increased to 900°C, the Al₂Cu alloy decomposed into AlCu_1.5 and aluminum. When the decomposed sample was again lowered to 800°C, Al₂Cu was regenerated by the reverse reaction of the decomposition reaction. The Al₂Cu regenerated in the reverse reaction was hydrogenated in the order of minutes, and a single-phase Al₂CuH was formed after approximately 40 minutes.

Fig. 8

Series of in-situ X-ray diffraction profiles of Al₂Cu under high hydrogen pressure. Squares, open and closed triangles, and circles indicate Bragg peaks from Al₂Cu, Al_1.5Cu, aluminum, and Al₂CuH, respectively.

The Al₂Cu alloy immediately after pressurization did not hydrogenate even after holding in hydrogen at 800°C, whereas the Al₂Cu alloy regenerated through the decomposition reaction hydrogenated immediately. This may be due to the chemically stable oxide layer on the surface of the Al₂Cu alloy. The pristine Al₂Cu alloy does not hydrogenate because the surface oxide inhibits hydrogenation reaction inside the alloy. When the decomposition reaction progresses by heating up to 900°C, atomic diffusion occur and induce modifications of the surface oxide layer. Since the sample is surrounded by a sufficient amount of hydrogen, reformation of the oxide layer does not proceed. If thermodynamically stable hydrides existed at 900°C, hydrogenation could proceed at this stage, but no such reaction was observed. Hydrides form when the temperature is lowered to reach the hydride stable region. In the case of Al₂CuH, the hydride stable temperature is lower that the reverse transition temperature, and thus, the hydrogenation reaction of the regenerated Al₂Cu was observed at 800°C.

Al₂CuH formed under high temperature and high pressure did not decompose during the depressurization process and could be recovered at ambient conditions. The recovered sample released hydrogen at approximately 150°C by heating at ambient pressure. However, this hydrogen release temperature does not reflect thermodynamically stable conditions because the decomposition reaction is suppressed by the surface oxide layer of the hydride. In other words, it does not indicate that Al₂CuH is stable at ambient conditions. Other analysis indicates that Al₂CuH is thermodynamically unstable at ambient conditions, and that a pressure of about 2 GPa is required for hydrogenation of the Al₂Cu alloy. Evaluation of the crystal structure and first-principles calculations indicate that hydrogen exists in a near-neutral state in interstitial sites of the Al₂Cu alloy.

In this way, aluminum-based alloys hydride was obtained. The hydrogenation reaction of Al₂Cu requires high pressure of 2 GPa or higher, so it cannot be used as a practical hydrogen storage material as it is. In the next section, another aluminum-based hydride that is thermodynamically more stable, is described.

3.2.2 Al₃FeH₄

We investigated the possibility of replacing Cu with other transition metals to synthesize novel hydrides and clarified that novel aluminum-iron hydride can be synthesized under high pressure.⁵⁶⁾ A small piece of Al₁₃Fe₄ alloy prepared in an arc melting furnace was used as the starting material. After pressurizing the starting material to 9 GPa at room temperature, the sample was heated to 750°C while maintaining the pressure at 9 GPa and hydrogenated the sample. After hydrogenation, the sample was cooled to room temperature, then depressurized to ambient pressure, and recovered at ambient conditions. The hydrogenation reaction process was monitored by in-situ SR-XRD measurement system.

Al₁₃Fe₄ alloy kept its crystal structure after pressurization to 9 GPa at room temperature. When the sample was heated, the intensities of the Bragg peaks from the Al₁₃Fe₄ alloy decreased around 700°C, and new peaks began to appear and grow. These profile changes were not observed even when the sample was treated at high pressure and high temperature in the absence of hydrogen. It was also confirmed that the recovered sample contained hydrogen, as described below, indicating the change in the crystal structure was induced by a formation of novel hydride, Al₃FeH₄. The results of in-situ SR-XRD observations indicate that the hydrogenation of the Al₁₃Fe₄ alloy completed within approximately 5 minutes after reaching 750°C, and that no further hydrogen absorption or other reactions proceed thereafter. It was also confirmed that the generated hydride did not change its crystal structure or decompose during cooling and depressurization. The obtained hydride can be recovered at ambient conditions.

The sample recovered at ambient conditions was heated and the hydrogen release profile was measured as is shown in Fig. 9. When the sample was heated, hydrogen was released from approximately 150°C. The weight hydrogen density was calculated to be 2.9 mass% based on the weight change during the hydrogen release. The hydrogen density of Al₃FeH₄ is comparable to those of typical hydrogen storage alloys such as LaNi₅H₆ (1.4 mass%) and TiFeH₂ (1.9 mass%).

Fig. 9

Hydrogen desorption profile of Al₃FeH₄ measured at ambient pressure.

Using in-situ SR-XRD measurement technique, it is possible not only to quickly determine pressure and temperature conditions where the target hydride is formed, but also to determine pressure and temperature conditions where the formed hydride decomposes. The thermodynamic stability of a synthesized hydride can be evaluated from the decomposition pressure-temperature conditions. The obtained results for Al₃FeH₄ (Fig. 10) indicate that Al₃FeH₄ is likely to be thermodynamically stable even near ambient pressure. In other words, Al₁₃Fe₄ alloy is thermodynamically capable of hydrogenation near ambient pressure. However, at present, hydrogenation of Al₁₃Fe₄ alloys requires high pressure of more than 7 GPa. This is probably due to the influence of the chemically stable surface oxide layer, which is present on the surface of aluminum alloys. If the influence of the surface oxide can be suppressed in some way, hydrogen absorption at near ambient pressure can be realized using this material, and research for this purpose is currently in progress. Al₃FeH₄ has relatively high thermodynamic stability despite the fact that it does not contain any metal with high-hydrogen affinity. Therefore, we investigated how Al₃FeH₄ is stabilized from the viewpoint of crystal structure.

Fig. 10

Decomposition pressure-temperature diagram of Al₃FeH₄.

The schematic of the crystal structure of Al₃FeH₄ refined by Rietveld analysis of SR-XRD and neutron diffraction data is shown in Fig. 11(a). Judging from the crystallographic symmetry and a size of the unit cell, the crystal structure is relatively simple. However, there are some partially occupied sites. The arrangement of the hydrogen around the metal allows us to consider in more detail what kind of bonding forms between the metal and the hydrogen.

Fig. 11

Schematics of the crystal structure of (a) Al₃FeH₄, (b) FeH₂ structural unit in Al₃FeH₄, and (c) [FeH₆]⁴⁻ complex anion in Li₄FeH₆.

Figure 11(b) shows a magnified view of the arrangement of hydrogen around an iron atom (Fe–H structural unit) in Al₃FeH₄. Iron atom is surrounded by six hydrogen sites to form an octahedron. This shape and the Fe–H distance are similar to those of [FeH₆]⁴⁻ complex anion in iron-containing complex hydrides (Fig. 11(c)). The site occupancy of the hydrogen site in Al₃FeH₄ is 1/3 and therefore, the chemical formula of the Fe–H structural unit in Al₃FeH₄ can be written as FeH₂, which is clearly different from [FeH₆]⁴⁻ in iron-containing complex hydrides such as Li₄FeH₆. This structural unit, FeH₂ may be a precursor that appears during the formation of complex anions.^57,58) We believe that the existence of such a unique structural unit may have enabled stable hydrides to be obtained from metal combinations that were thought to be unstable according to conventional guidelines.

Figure 12 summarizes the synthesis pressure–temperature conditions for the hydrides explained in the present paper. The Al-based hydrides require high pressure of more than 7 GPa for hydrogenation because such conditions are necessary to suppress the effects of chemically stable surface oxide layers. The Li-containing complex hydrides have a relatively wide range of synthesis temperature–pressure conditions. All of these material systems have relatively high thermodynamic stabilities among hydrides synthesized under high pressure, however, the temperature required for hydrogen differs depending on the reactivity of the mixed powder used as the starting material. For a material with a slow synthesis reaction, a high pressure is needed to stabilize the hydride to a high temperatures to suppress its slow kinetics. Therefore, it can be seen that synthesis temperatures monotonously increase with pressures.

Fig. 12

Pressure–temperature conditions of synthesized hydrides. Closed and open symbols show synthesis conditions of Li-containing complex hydrides and Al-based hydrides, respectively.

Finally, a comparison of the thermodynamic stability of hydrides synthesized using multi-anvil apparatus and those synthesized near ambient pressure is discussed. Considering the reaction M + x/2 H₂ = MH_x, in which the metal M is hydrogenated to form a hydride MH_x, the equilibrium pressure–temperature conditions are expressed by a van’t Hoff equation.   

\begin{equation} \ln(p_{\text{H2}}) = 2\varDelta H_{0}/x\text{R}T - 2\Delta S_{0}/x\text{R}, \end{equation} (1)

where p_H2 is the hydrogen pressure, T is the absolute temperature, R is the gas constant, ΔH₀ and ΔS₀ are the standard heat of formation and standard entropy change of the hydrogenation reaction, respectively. The schematic of a van’t Hoff relationship is shown in Fig. 13. It is clear from the equation that there is a linear relationship between ln(p_H2) and 1/T, the slope corresponds to ΔH₀.

Fig. 13

Schematic diagram of pressure–temperature dependence of equilibrium conditions of hydrogenation reaction of metals. The solid line shows a van’t Hoff relation. The dashed curve shows another one considering fugacity at a high pressure region.

At pressures of a few tens of MPa, the behavior of hydrogen gas begins to dissociate from that of an ideal gas. In the high pressure region where hydrogen does not behave as an ideal gas, the pressure in the van’t Hoff equation is replaced by a fugacity. In such high pressure region, the interaction between hydrogen molecules becomes non-negligible, and a fugacity becomes larger than a pressure. Therefore, as shown in Fig. 13, the stability region of hydride increases. In the case of Al₃FeH₄, which can be synthesized at 7 GPa and 750°C, ΔH₀ is estimated to be −10 to −20 kJ/mol H₂.

Hydrogen storage alloys obtained around ambient pressure have also hydrogenated at high pressures of several GPa using a multi-anvil apparatus. TiFe alloy was hydrogenated at 5 GPa to form its hydride with disordered metal sublattice.⁵⁹⁾ LaNi₅ was also hydrogenated to form hydrogen rich hydride phase, LaNi₅H_x (x < 9).⁶⁰⁾ However, in general, metals with high-hydrogen affinities form too stable metal hydrides at several GPa region resulting in phase disproportionations of original alloys. Thus, it is not easy to obtain novel hydride by hydrogenating hydrogen storage alloys at several GPa region.

4. Conclusions

We have presented an advance in the synthesis of novel hydrides under high pressure and high temperature. Lithium-containing hydrides were predicted by theoretical calculations, synthesized by a high-pressure technique, and characterized by in-situ SR-XRD. YLiFeH₆ was synthesized, in which lithium was added as an electron donor to compensate for imbalanced charge. Theoretically predicted LiNiH₃ with a perovskite structure was synthesized. In situ SR-XRD revealed that the perovskite hydride was formed through a three-step reaction and that the Li_yNi_1−yH solid solution was the precursor to the perovskite formation. The theoretically predicted Li₄FeH₆ complex hydride was also synthesized, which was thermodynamically stable near ambient conditions, although it was synthesized only at high pressures. The incorporation of H⁻ ions allows us to increase the number of cation combinations while maintaining charge neutrality, and we have successfully synthesized theoretically predicted hydride, Li₃AlFeH₈. According to conventional guidelines for synthesizing hydrogen storage alloys and hydrides, alloys composed solely of metals with low-hydrogen affinities are not considered to form hydrides. We have hydrogenated Al₂Cu and Al₁₃Fe₄ alloys to form novel hydrides Al₂CuH and Al₃FeH₄, where all the elements aluminum, iron, and copper are metals with low-hydrogen affinities. Especially, aluminum has very low-hydrogen affinity and though to be difficult to hydrogenate, though many attempts have been done to obtain light-weight and low-cost hydrogen storage materials. These results demonstrate that the high-pressure technique is useful for discovering novel hydrides.

Acknowledgments

This work was supported by JSPS KAKENHI Grant-in-Aid for Scientific Research on Innovative Areas “Hydrogenomics” (Nos. JP18H05513 and JP18H05518) as well as JSPS KAKENHI (No. JP22H01821) and grants from the Inter-University Cooperative Research Program of the Institute for Materials Research, Tohoku University (Proposal Nos. 20K0022, 202012-RDKGE-0066, and 202112-RDKGE-0025). Crystal structures shown in the present paper were drown using the VESTA program.⁶¹⁾

REFERENCES

• 1)   A.P.  Drozdov,  M.I.  Eremets,  I.A.  Troyan,  V.  Ksenofontov and  S.I.  Shylin: Nature 525 (2015) 73–76. doi:10.1038/nature14964
• 2)   A.P.  Drozdov,  P.P.  Kong,  V.S.  Minkov,  S.P.  Besedin,  M.A.  Kuzovnikov,  S.  Mozaffari,  L.  Balicas,  F.F.  Balakirev,  D.E.  Graf,  V.B.  Prakapenka,  E.  Greenberg,  D.A.  Knyazev,  M.  Tkacz and  M.I.  Eremets: Nature 569 (2019) 528–531. doi:10.1038/s41586-019-1201-8
• 3)   M.  Somayazulu,  M.  Ahart,  A.K.  Mishra,  Z.M.  Geballe,  M.  Baldini,  Y.  Meng,  V.V.  Struzhkin and  R.J.  Hemley: Phys. Rev. Lett. 122 (2019) 027001. doi:10.1103/PhysRevLett.122.027001
• 4)   E.  Snider,  N.  Dasenbrock-Gammon,  R.  McBride,  M.  Debessai,  H.  Vindana,  K.  Vencatasamy,  K.V.  Lawler,  A.  Salamat and  R.P.  Dias: Nature 586 (2020) 373–377. doi:10.1038/s41586-020-2801-z
• 5)   M.  Matsuo,  Y.  Nakamori,  S.I.  Orimo,  H.  Maekawa and  H.  Takamura: Appl. Phys. Lett. 91 (2007) 224103. doi:10.1063/1.2817934
• 6)   G.  Kobayashi,  Y.  Hinuma,  S.  Matsuoka,  A.  Watanabe,  M.  Iqbal,  M.  Hirayama,  M.  Yonemura,  T.  Kamiyama,  I.  Tanaka and  R.  Kanno: Science 351 (2016) 1314–1317. doi:10.1126/science.aac9185
• 7)   S.  Takagi and  S.  Orimo: Scr. Mater. 109 (2015) 1–5. doi:10.1016/j.scriptamat.2015.07.024
• 8)   R.  Mohtadi and  S.  Orimo: Nat. Rev. Mater. 2 (2016) 16091. doi:10.1038/natrevmats.2016.91
• 9)  M. Eremets: High Pressure Experimental Methods, (Oxford University Press, New York, 1996).
• 10)  Y. Fukai: The Metal-Hydrogen System, Second, (Springer, Berlin, 2005).
• 11)   B.  Baranowski and  S.M.  Filipek: J. Alloy. Compd. 404–406 (2005) 2–6. doi:10.1016/j.jallcom.2005.02.102
• 12)   V.E.  Antonov,  K.  Cornell,  V.K.  Fedotov,  A.I.  Kolesnikov,  E.G.  Ponyatovsky,  V.I.  Shiryaev and  H.  Wipf: J. Alloy. Compd. 264 (1998) 214–222. doi:10.1016/S0925-8388(97)00298-3
• 13)   V.E.  Antonov: J. Alloy. Compd. 330–332 (2002) 110–116. doi:10.1016/S0925-8388(01)01532-8
• 14)   K.  Wakamori,  S.M.S.M.  Filipek and  A.  Sawaoka: Rev. Sci. Instrum. 54 (1983) 1410. doi:10.1063/1.1137257
• 15)   Y.  Fukai and  N.  Okuma: Jpn. J. Appl. Phys. 32 (1993) L1256–L1259. doi:10.1143/JJAP.32.L1256
• 16)   Y.  Fukai: J. Alloy. Compd. 404–406 (2005) 7–15. doi:10.1016/j.jallcom.2005.02.075
• 17)   Y.  Fukai: J. Alloy. Compd. 356–357 (2003) 263–269. doi:10.1016/S0925-8388(02)01269-0
• 18)   K.  Takemura,  P.C.  Sahu,  Y.  Kunii and  Y.  Toma: Rev. Sci. Instrum. 72 (2001) 3873–3876. doi:10.1063/1.1396667
• 19)   A.  Machida,  A.  Ohmura,  T.  Watanuki,  T.  Ikeda,  K.  Aoki,  S.  Nakano and  K.  Takemura: Solid State Commun. 138 (2006) 436–440. doi:10.1016/j.ssc.2006.04.011
• 20)   A.  Ohmura,  A.  Machida,  T.  Watanuki,  K.  Aoki,  S.  Nakano and  K.  Takemura: Phys. Rev. B 73 (2006) 104105. doi:10.1103/PhysRevB.73.104105
• 21)   K.  Puhakainen,  E.  Stoyanov,  M.J.  Evans,  K.  Leinenweber and  U.  Hussermann: J. Solid State Chem. 183 (2010) 1785–1789. doi:10.1016/j.jssc.2010.05.033
• 22)   V.E.  Antonov,  B.M.  Bulychev,  V.K.  Fedotov,  D.I.  Kapustin,  V.I.  Kulakov and  I.A.  Sholin: Int. J. Hydrogen Energ. 42 (2017) 22454–22459. doi:10.1016/j.ijhydene.2017.03.121
• 23)   J.  Nylén,  T.  Sato,  E.  Soignard,  J.L.  Yarger,  E.  Stoyanov and  U.  Häussermann: J. Chem. Phys. 131 (2009) 104506. doi:10.1063/1.3230973
• 24)   H.  Saitoh,  A.  Machida,  Y.  Katayama and  K.  Aoki: Appl. Phys. Lett. 94 (2009) 151915. doi:10.1063/1.3120559
• 25)   D.  Kyoi,  E.  Rönnebro,  N.  Kitamura,  A.  Ueda,  M.  Ito,  S.  Katsuyama and  T.  Sakai: J. Alloy. Compd. 361 (2003) 252–256. doi:10.1016/S0925-8388(03)00431-6
• 26)   R.  Kataoka,  T.  Sato,  T.  Kuriiwa,  A.  Kamegawa and  M.  Okada: Mater. Trans. 52 (2011) 49–53. doi:10.2320/matertrans.M2010208
• 27)   A.  Kamegawa,  R.  Kataoka and  M.  Okada: Energy Procedia 29 (2012) 276–282. doi:10.1016/j.egypro.2012.09.033
• 28)   D.  Moser,  D.J.  Bull,  T.  Sato,  D.  Noréus,  D.  Kyoi,  T.  Sakai,  N.  Kitamura,  H.  Yusa,  T.  Taniguchi,  W.P.  Kalisvaart and  P.  Notten: J. Mater. Chem. 19 (2009) 8150–8161. doi:10.1039/b911263d
• 29)   E.  Rönnebro: J. Phys. Chem. Solids 71 (2010) 1154–1158. doi:10.1016/j.jpcs.2010.03.025
• 30)   K.  Spektor,  W.A.  Crichton,  S.  Konar,  S.  Filippov,  J.  Klarbring,  S.I.  Simak and  U.  Häussermann: Inorg. Chem. 57 (2018) 1614–1622. doi:10.1021/acs.inorgchem.7b02968
• 31)   K.  Spektor,  W.A.  Crichton,  S.  Filippov,  S.I.  Simak,  A.  Fischer and  U.  Haüssermann: Inorg. Chem. 59 (2020) 16467–16473. doi:10.1021/acs.inorgchem.0c02294
• 32)   K.  Spektor,  W.A.  Crichton,  S.  Filippov,  S.I.  Simak and  U.  Häussermann: Inorg. Chem. 58 (2019) 11043–11050. doi:10.1021/acs.inorgchem.9b01569
• 33)   S.  Burapornpong,  S.  Tsuchiya,  T.  Kimura and  A.  Kamegawa: J. Alloy. Compd. 850 (2021) 156754. doi:10.1016/j.jallcom.2020.156754
• 34)   S.M.  Filipek: J. Adv. Sci. 19 (2007) 1–10. doi:10.2978/jsas.19.1
• 35)   S.M.  Filipek,  V.  Paul-Boncour,  N.  Kuriyama,  N.  Takeichi,  H.  Tanaka,  R.-S.  Liu,  R.  Wierzbicki,  R.  Sato and  H.-T.  Kuo: Solid State Ion. 181 (2010) 306–310. doi:10.1016/j.ssi.2010.01.009
• 36)   K.  Yvon: Chim. Int. J. Chem. 52 (1998) 613–619. doi:10.2533/chimia.1998.613
• 37)   V.  Paul-Boncour,  S.M.  Filipek,  M.  Dorogova,  F.  Bourée,  G.  André,  I.  Marchuk,  A.  Percheron-Guégan and  R.S.  Liu: J. Solid State Chem. 178 (2005) 356–362. doi:10.1016/j.jssc.2004.05.066
• 38)   M.  Matsuo,  K.  Miwa,  S.  Semboshi,  H.-W.  Li,  M.  Kano and  S.  Orimo: Appl. Phys. Lett. 98 (2011) 221908. doi:10.1063/1.3597363
• 39)   M.  Matsuo,  D.  Matsumura,  Y.  Nishihata,  G.  Li,  N.  Hiyama,  S.  Semboshi and  S.  Orimo: Appl. Phys. Lett. 100 (2012) 044101. doi:10.1063/1.3678032
• 40)   V.  Paul-Boncour,  M.  Escorne,  A.  Mauger,  M.  Latroche and  A.  Percheron-Guegan: J. Appl. Phys. 79 (1996) 4253. doi:10.1063/1.361793
• 41)   M.  Matsuo,  H.  Saitoh,  A.  Machida,  R.  Sato,  S.  Takagi,  K.  Miwa,  T.  Watanuki,  Y.  Katayama,  K.  Aoki and  S.  Orimo: RSC Adv. 3 (2013) 1013–1016. doi:10.1039/C2RA22497F
• 42)   K.  Kadir and  D.  Noréus: Inorg. Chem. 46 (2007) 2220–2223. doi:10.1021/ic0620586
• 43)   M.  Orlova,  J.-P.  Rapin and  K.  Yvon: Inorg. Chem. 48 (2009) 5052–5054. doi:10.1021/ic802417a
• 44)   S.  Takagi,  H.  Saitoh,  N.  Endo,  R.  Sato,  T.  Ikeshoji,  M.  Matsuo,  K.  Miwa,  K.  Aoki and  S.  Orimo: Phys. Rev. B 87 (2013) 125134. doi:10.1103/PhysRevB.87.125134
• 45)   R.  Sato,  H.  Saitoh,  N.  Endo,  S.  Takagi,  M.  Matsuo,  K.  Aoki and  S.  Orimo: Appl. Phys. Lett. 102 (2013) 091901. doi:10.1063/1.4794067
• 46)   H.  Saitoh,  A.  Machida and  K.  Aoki: Chin. Sci. Bull. 59 (2014) 5290–5301. doi:10.1007/s11434-014-0543-8
• 47)   M.  Kritikos,  D.  Nore’us,  A.F.  Andresen and  P.  Fischer: J. Solid State Chem. 92 (1991) 514–519. doi:10.1016/0022-4596(91)90357-N
• 48)   H.  Saitoh,  S.  Takagi,  M.  Matsuo,  Y.  Iijima,  N.  Endo,  K.  Aoki and  S.  Orimo: APL Mater. 2 (2014) 076103. doi:10.1063/1.4886219
• 49)   T.  Ogata,  T.  Sato,  S.  Takagi,  H.  Saitoh,  Y.  Iijima,  B.  Paik and  S.  Orimo: Mater. Trans. 58 (2017) 157–159. doi:10.2320/matertrans.M2016345
• 50)   K.  Sakamaki,  E.  Takahashi,  Y.  Nakajima,  Y.  Nishihara,  K.  Funakoshi,  T.  Suzuki and  Y.  Fukai: Phys. Earth Planet. Inter. 174 (2009) 192–201. doi:10.1016/j.pepi.2008.05.017
• 51)   S.  Takagi,  T.D.  Humphries,  K.  Miwa and  S.  Orimo: Appl. Phys. Lett. 104 (2014) 203901. doi:10.1063/1.4878775
• 52)   T.  Sato,  S.  Takagi,  S.  Deledda,  B.C.  Hauback and  S.  Orimo: Sci. Rep. 6 (2016) 23592. doi:10.1038/srep23592
• 53)   A.  Züttel: Mater. Today 6 (2003) 24–33. doi:10.1016/S1369-7021(03)00922-2
• 54)   H.  Mizoguchi,  S.W.  Park and  H.  Hosono: J. Am. Chem. Soc. 143 (2021) 11345–11348. doi:10.1021/jacs.1c06307
• 55)   H.  Saitoh,  S.  Takagi,  N.  Endo,  A.  Machida,  K.  Aoki,  S.  Orimo and  Y.  Katayama: APL Mater. 1 (2013) 032113. doi:10.1063/1.4821632
• 56)   H.  Saitoh,  T.  Sato,  M.  Tanikami,  K.  Ikeda,  A.  Machida,  T.  Watanuki,  T.  Taguchi,  S.  Yamamoto,  T.  Yamaki,  S.  Takagi,  T.  Otomo and  S.  Orimo: Mater. Des. 208 (2021) 109953. doi:10.1016/j.matdes.2021.109953
• 57)   K.  Miwa,  T.  Sato,  M.  Matsuo,  K.  Ikeda,  T.  Otomo,  S.  Deledda,  B.C.  Hauback,  G.  Li,  S.  Takagi and  S.  Orimo: J. Phys. Chem. C 120 (2016) 5926–5931. doi:10.1021/acs.jpcc.6b00341
• 58)   T.  Sato,  K.  Ikeda,  M.  Matsuo,  K.  Miwa,  T.  Otomo,  S.  Deledda,  B.C.  Hauback,  G.  Li,  S.  Takagi and  S.  Orimo: Int. J. Hydrogen Energ. 42 (2017) 22449–22453. doi:10.1016/j.ijhydene.2017.02.196
• 59)   N.  Endo,  H.  Saitoh,  A.  Machida and  Y.  Katayama: Int. J. Hydrogen Energ. 38 (2013) 6726–6729. doi:10.1016/j.ijhydene.2013.03.120
• 60)   T.  Sato,  H.  Saitoh,  R.  Utsumi,  J.  Ito,  Y.  Nakahira,  K.  Obana,  S.  Takagi and  S.  Orimo: Molecules 28 (2023) 1–12. doi:10.3390/molecules28031256
• 61)   K.  Momma and  F.  Izumi: J. Appl. Cryst. 41 (2008) 653–658. doi:10.1107/S0021889808012016