Complex transition metal hydrides incorporating ionic hydrogen: thermal decomposition pathway of Na₂Mg₂FeH₈ and Na₂Mg₂RuH₈

Abstract

Complex transition metal hydrides have potential technological application as hydrogen storage materials, smart windows and sensors. Recent exploration of these materials has revealed that the incorporation of anionic hydrogen into these systems expands the potential number of viable complexes, while varying the countercation allows for optimisation of their thermodynamic stability. In this study, the optimised synthesis of Na₂Mg₂TH₈ (T = Fe, Ru) has been achieved and their thermal decomposition properties studied by ex situ Powder X-ray Diffraction, Gas Chromatography and Pressure-Composition Isotherm measurements. The temperature and pathway of decomposition of these isostructural compounds differs considerably, with Na₂Mg₂FeH₈ proceeding via NaMgH₃ in a three-step process, while Na₂Mg₂RuH₈ decomposes via Mg₂RuH₄ in a two-step process. The first desorption maxima of Na₂Mg₂FeH₈ occurs at ca. 400 °C, while Na₂Mg₂RuH₈ has its first maxima at 420 °C. The enthalpy and entropy of desorption for Na₂Mg₂TH₈ (T = Fe, Ru) has been established by PCI measurements, with the ΔH_des for Na₂Mg₂FeH₈ being 94.5 kJ mol⁻¹ H₂ and 125 kJ mol⁻¹ H₂ for Na₂Mg₂RuH₈.

---

Introduction

Transition metals are renowned for their diverse range of valence states and structural conformations.^1,2 As such, in the last five decades a swathe of homoleptic transition metal hydrides have been synthesised to determine their potential for technological applications. Mg₂NiH₄ was first realised for its reversible hydrogenation properties in 1968,³ and has since been investigated for a variety technological applications including smart windows and sensors.^4–7 Mg₂FeH₆, with a gravimetric hydrogen storage content of 5.5 wt% has since been developed,^8–14 along with a host of other transition metal hydride congeners and derivatives.^1,2,15–22

The transition metal hydrides of Group 8 often form octahedral [TH₆]⁴⁻ anions, of which are limited to four-fold coordination by counterions (M) in the form of M⁺M′⁺M′′⁺M′′′⁺, M²⁺M′⁺M′′⁺, M³⁺M′⁺, M²⁺M′²⁺M′′⁺. Expanding the diversity of coordination can be achieved by increasing the anionic charge of the system, for instance by the inclusion of H⁻. A recent DFT study by Takagi et al. established that the incorporation of anionic hydrogen into complex transition metal hydride compounds enables inclusion of a wider variety of cations, thereby allowing tuning of these materials in order to optimise their thermodynamic properties or hydrogen storage capacities.²³ To date, a variety of quaternary complex hydrides have been synthesised and their structural and physical properties explored, these include LaMg₂NiH₇ (La³⁺·2Mg²⁺·3H⁻·[NiH₄]⁴⁻),^18,24 Na₂Mg₂NiH₆ (2Na⁺·2Mg²⁺·2H⁻·[NiH₄]⁴⁻),^19,25 Na₂Mg₂TH₈ (2Na⁺·2Mg²⁺·2H⁻·[TH₆]⁴⁻) (T = Fe, Ru),²⁰ MMg₂FeH₈ (M²⁺·2Mg²⁺·2H⁻·[TH₆]⁴⁻) (M = Ba, Ca, Sr; T = Fe, Ru, Os)^15,16,26 and M₄Mg₄Fe₃H₂₂ (4Ca²⁺·4Mg²⁺·4H⁻·3[FeH₆]⁴⁻) (M = Ca, Yb).^27,28 Thermodynamic data for these materials are scarce, although some experimental^15,28 and DFT calculated²³ values have been determined. SrMg₂FeH₈ and BaMg₂FeH₈ decompose at ca. 440 and 450 °C under 0.1 MPa H₂,¹⁵ respectively, while Ca₄Mg₄Fe₃H₂₂ and Yb₄Mg₄Fe₃H₂₂ decompose at ca. 395 and 420 °C,^1,27,28 respectively. The enthalpy of desorption of Ca₄Mg₄Fe₃H₂₂ and Yb₄Mg₄Fe₃H₂₂ to their corresponding binary hydrides has been calculated to be 122 and 137 kJ mol⁻¹ H₂, respectively.²⁸ These values are significantly larger than those determined for the ternary Mg₂FeH₆ at 87 kJ mol⁻¹ H₂,²⁹ which decomposes at ca. 300 °C,¹³ and indicates the increased stabilisation offered by the incorporation of anionic hydrogen and varied cations into these quaternary compounds.

The isostructural compounds of Na₂Mg₂FeH₈ (5.1 wt% H) and Na₂Mg₂RuH₈ (4.0 wt% H) hold potential as hydrogen storage materials.²⁰ To make a fair assumption of their prospective application, a true understanding of the physical properties of these novel quaternary complex transition metal hydrides must be determined, unto which the data is extremely sparse. As a consequence, the influence of H⁻ on the thermal stability and decomposition process of these materials is generally unknown and must be understood. As such, ex situ powder X-ray diffraction (PXD) and Pressure-Composition Isotherm Measurements (PCI) have been conducted. Their temperatures and pathways of decomposition have been established and the associated enthalpies and entropies of H₂ desorption have been calculated and compared to literature values.

Experimental

All preparation and manipulation was performed in a Miwa glove box filled with purified argon (<1 ppm O₂ and the dew point of H₂O below 190 K) to avoid contamination.

The synthesis of Na₂Mg₂FeH₈ was carried out by two methods: S1 followed a four step process, which first required the synthesis of Mg₂FeH₆. This was achieved by mechanically milling (Fritsch Pulverisette 7) MgH₂ (hydrogen storage grade, Sigma Aldrich) and Fe (99.99%, Mitsuwa) powders at a molar ratio of 2 : 1 for 2 h at 400 rpm (ball-to-powder ratio 40 : 1), under argon with subsequent heat treatment of the pelletised powder at 400 °C for 20 h under 3 MPa H₂. The resultant olive green powder was then mechanically milled with NaH (95%, Sigma Aldrich) at a molar ratio of 1 : 2 for 20 h under argon (S1-BM) with subsequent heat treatment of the pelletised powder at 400 °C for 20 h under 30 MPa H₂. The product was yielded as an olive green powder.

The synthesis of S2 (Na₂Mg₂FeH₈) followed a two-step process. NaH, MgH₂ and Fe powders at a molar ratio of 2 : 2 : 1 were mechanically milled for 20 h at 400 rpm (ball-to-powder ratio 40 : 1), under argon (S2-BM) with subsequent heat treatment of the pelletised powder at 400 °C for 60 h under 30 MPa H₂. The product was yielded as an olive green powder.

The synthesis of Na₂Mg₂RuH₈ followed a two-step process. MgH₂, NaH and Ru (99.9%, Kojundo Chemical Laboratory) were mechanically milled (identical parameters as employed with Na₂Mg₂FeH₈) at a molar ratio of 2 : 2 : 1 for 20 h under argon (S3-BM), before subsequent heat treatment of the pelletised powder at 500 °C for 20 h under 30 MPa H₂. The product was yielded as a light grey powder.

Powder X-ray diffraction (PXD) measurements were conducted using a conventional X-ray diffractometer (Lab-PXD, PANalytical X'Pert-Pro, CuKα radiation) in flat plate mode. Data were collected using a X'Celerator X linear position sensitive detector within a 2θ range of 10–90° using 0.02° steps at 0.04°/s with X-ray generator operating conditions of 45 kV and 40 mA. The PXD samples were loaded in an Ar glovebox and the sample holder covered by Mylar film to prevent oxygen/moisture contamination during data collection. PANalytical HighScore Plus v. 3.0, DICVOL,³⁰ CHEKCELL³¹ and GSAS^32,33 were used for phase identification, indexing, space group identification and Rietveld refinement, respectively.

A GC323 (Gas Chromatography) (GL sciences Inc.) was used to detect the desorbed H₂ by means of a TCD detector, with a column temperature of 200 °C. Samples were heated at a rate of 5 °C min⁻¹ under an Ar flow of 40 ml min⁻¹.

Typical Pressure-Composition Isotherm Measurements (PCI) were conducted inside custom-built manometric apparatus, where the sample cell was placed in a furnace and heated to the desired temperature at a hydrogen pressure of 6 MPa or 30 MPa. The experiment was controlled by software developed by Suzuki Shokan Co., Ltd.

Results and discussion

Synthesis of Na₂Mg₂TH₈ (T = Fe, Ru)

The synthesis of Na₂Mg₂FeH₈ was first reported to utilise Mg₂FeH₆ as a starting material, which was consequently milled for 20 h with NaH (S1-BM), followed by hydrogenation at 30 MPa H₂ at 400 °C (eqn (1) and (2)).²⁰ Overall, this method is time intensive (in excess of 100 h due to the requirement of two milling and two hydrogenation procedures), and the final product also contains Fe, NaH, and Mg₂FeH₆ impurities. In order to reduce the time requirements and levels of impurities, alternative synthetic routes were investigated. It was determined that successful synthesis is achievable by ball milling 2NaH + MgH₂ + Fe (S2-BM), followed by hydrogenation at 30 MPa H₂ at 400 °C (eqn (3)), with an associated time requirement of 80 h. The products after milling of each material differ considerably (Fig. 1), the constituents of S1-BM, identified by PXD are not altered during milling (Mg₂FeH₆ + NaH) despite a broadening of the peaks due to a decrease of crystallite size (or reduced crystallinity).³⁴ The products of S2-BM were NaMgH₃ + Fe.³⁵

Fig. 1 PXD of ball milled samples S1-BM (Mg₂FeH₆ + 2NaH), S2-BM (2NaH + 2MgH₂ + Fe) and S3-BM (2MgH₂ + 2NaH + Ru) and hydrogenated samples of S1-3. λ = CuKα. * = Na₂Mg₂TH₈ (T = Fe, Ru); # = Mg₂FeH₆; @ = NaH; x = T (Fe, Ru); $ = NaMgH₃; ^ = NaOH; + = Mg; ∼ MgH₂; ! = unknown phase.

Hydrogenation of the ball milled samples S1-BM and S2-BM under standard conditions (30 MPa H₂ and 400 °C for 60 h), resulted in the formation of Na₂Mg₂FeH₈ as the major phase (Fig. 1). Analysis of hydrogenated S1 by Reitveld refinement indicates that the sample is 80.3(9)% pure and as such contains residual Mg₂FeH₆ (8.9(4)%), Fe (7.9(2)%), NaMgH₃ (2.5(3)%) and NaH (2.7(2)%) starting materials. Analysis of hydrogenated S2 indicates that the sample contains 90.6(4)% Na₂Mg₂FeH₈ and 9.4(3)% Fe, without residual Mg₂FeH₆ or NaH. This indicates that the optimal method of synthesis is via the S2 method due to the elimination of the prerequisite Mg₂FeH₆ synthesis (eqn (1)) and overall decrease in impurities compared to S1, primarily due to the initial (complete) formation of NaMgH₃ during the milling reaction. Milling initiates the breaking of the strong Na–H bonds, which is required to ensue during the annealing phase when synthesised via Mg₂FeH₆ (S1). This ultimately leads to the observation of unreacted NaH and Mg₂FeH₆ starting materials.

The synthesis of Na₂Mg₂RuH₈ follows a two-step reaction, where stoichiometric quantities of Ru, NaH and MgH₂ are milled for 5 h (S3-BM) before hydrogenation under 30 MPa H₂ at 500 °C for 20 h (S3) (eqn (4)).²⁰ The composition of the milled material is unchanged from the starting materials (Fig. 1), whereas after hydrogenation the sample appears to be mostly Na₂Mg₂RuH₈, but also comprises of some residual Ru. An unknown material is also identifiable within the Na₂Mg₂RuH₈ powder (S3), of which only a few reflections are discernible. The occurrence of these additional Bragg peaks were also noted previously.²⁰

Thermal decomposition of Na₂Mg₂TH₈ (T = Fe, Ru)

In order to determine the thermal decomposition temperatures of Na₂Mg₂TH₈ (T = Fe, Ru), aliquots of each material were heated at a rate of 5 °C min⁻¹ with GC detecting the corresponding desorbed H₂. The chromatograms obtained (Fig. 2a) are similar, with two stages of H₂ desorption being observed for both materials (S1 and S3). The onset of decomposition for Na₂Mg₂FeH₈ occurs at ca. 280 °C, with the first maxima being observed at ca. 400 °C, and the second at ca. 430 °C. Decomposition concludes at ca. 475 °C. Conversely, the onset of H₂ desorption for Na₂Mg₂RuH₈ occurs at ca. 325 °C, with the first maxima being observed at ca. 421 °C, and the second at ca. 460 °C. H₂ is no longer detected after ca. 500 °C.

Fig. 2 (a) GC analysis of Mg₂FeH₆, Na₂Mg₂FeH₈ (S1) and Na₂Mg₂RuH₈ (S3) (ΔT = 5 °C min⁻¹). Ex situ PXD analysis of (b) Na₂Mg₂FeH₈ (S1) and (c) Na₂Mg₂RuH₈ (S3) at selected temperatures (λ = CuKα). Numbers on GC plots correlate to temperatures at which samples were heated prior to PXD analysis. * = Na₂Mg₂TH₈ (T = Fe, Ru); # = Mg₂FeH₆; @ = NaH; x = T (Fe, Ru); $ = NaMgH₃; ^ = NaOH; + = Mg; α = Mg₃Ru₂; ! = unknown phase; θ = unknown phase.

To ascertain the pathway of decomposition, ex situ PXD was conducted on samples heated to selected temperatures in vacuo (Fig. 2b and c). Analysis of Na₂Mg₂FeH₈ after heating at 360 °C indicates a miniscule quantity of Na₂Mg₂FeH₈ resides, although the majority has decomposed into NaH, Mg and Fe, while NaMgH₃ is also detected. By 400 °C NaH has decomposed, while only Mg and Fe are observable by PXD. Na is not observed due to the low vapor pressure of Na at elevated temperatures. No further changes to the material are observed at higher temperatures. Therefore the decomposition of Na₂Mg₂FeH₈ is determined to occur according to eqn (5)–(7).

Na₂Mg₂FeH₈ → 2NaMgH₃ + Fe + H₂ (1.3 wt%) (5)

2NaMgH₃ + Fe → 2NaH + 2Mg + 2H₂ + Fe (2.5 wt%) (6)

2NaH + 2Mg + Fe → 2Na + 2Mg + Fe + H₂ (1.3 wt%) (7)

Na₂Mg₂RuH₈ was first heated to 320 °C, where PXD determined that decomposition has not yet started. By 420 °C full decomposition appears to be complete, with Mg₃Ru₂ and Ru being the products. At this temperature, NaH instantly decomposes to Na and evaporates from the sample. As a result, the decomposition of Na₂Mg₂RuH₈ is determined to occur according to eqn (8) and (9). Presumably the excess Ru required to form Mg₃Ru₂ (without leaving excess Mg) comes from the excess Ru that remained in the starting material.

Na₂Mg₂RuH₈ → 2NaH + 2/3Mg₃Ru₂ + 3H₂ (3.0 wt%) (8)

2NaH + Mg₃Ru₂ → 2Na + Mg₃Ru₂ + H₂ (1.0 wt%) (9)

PCI analysis of Na₂Mg₂TH₈ (T = Fe, Ru) (Fig. 3a and b) enables the intricacies of the decomposition process to be truly understood. The experiments on Na₂Mg₂FeH₈ were conducted at initial pressures of 30 MPa and 400 °C, mimicking conditions used for synthesis. Consequently, it was ascertained that the first thermal reaction according to eqn (5), occurs at an equilibrium pressure of 15.5 MPa, releasing ca. 0.8 wt% H₂ at 400 °C (Fig. 4a). PXD analysis of the products recovered at 6 MPa characterised the products to be NaMgH₃ and Fe (Fig. 3e). ΔH_dec for this process was determined by means of a van't Hoff plot of H₂ desorption equilibrium pressures and the linear fit (R² = 0.975) to the data to be 93 kJ mol⁻¹ H₂ (Fig. 3c), while the corresponding ΔS_dec was calculated as 180 J mol⁻¹ H₂/K. However, at lower temperatures this step is kinetically hindered and as a result ΔH and ΔS may be obscured. The values reported above are an average of the plateau pressures.

Fig. 3 PCI analysis of Na₂Mg₂FeH₈ (a) and Na₂Mg₂RuH₈ (b) at selected temperatures. van't Hoff plot of H₂ desorption equilibrium pressures and the linear fit to the data for Na₂Mg₂FeH₈ (c) and Na₂Mg₂RuH₈ (d). Ex situ PXD analysis of Na₂Mg₂FeH₈ (S1) (e) and Na₂Mg₂RuH₈ (S3) (f) for samples collected after PCI analysis and after rehydrogenation (λ = CuKα). * = Na₂Mg₂TH₈ (T = Fe, Ru); # = Mg₂FeH₆; @ = NaH; x = T (Fe, Ru); $ = NaMgH₃; ^ = NaOH; + = Mg; Δ = Na; α = Mg₃Ru₂; δ = Mg₂RuH₄; ! = unknown phase; θ = unknown phase; γ = unknown phase; ε = unknown phase.

Fig. 4 Energy diagram illustrating the experimentally determined enthalpies of formation (ΔH_form kJ mol⁻¹) of Na₂Mg₂FeH₈, Mg₂FeH₆,²⁹ and Na₂Mg₂RuH₈. The excess Ru required to form Mg₃Ru₂ is acquired from the impurity Ru remaining in the Na₂Mg₂RuH₈ starting material.

The further two equilibrium plateaus below 1 MPa H₂ correspond to the decomposition of NaMgH₃, exhibiting mass losses of 2.5 and 1.5 wt% for eqn (6) and (7), respectively. The overall hydrogen content released was therefore determined to be 4.7 wt% at 400 °C (theoretical maximum of 5.1 wt%). ΔH_dec was calculated to be 87 and 111 kJ mol⁻¹ H₂ for the latter two processes, in accord with the literature values.³⁶ The corresponding ΔS_dec also agreed with literature values with 132 and 158 J mol⁻¹ H₂/K for eqn (6) and (7), respectively. Therefore ΔH_des for the entire system is surmised to be 378 kJ mol⁻¹ (94.5 kJ mol⁻¹ H₂). The identity of the species at each decomposition stage was determined by PXD by ending selected PCI experiments at specified pressures. Fig. 3e illustrates the final products after the PCI experiments conducted at 360 and 400 °C and also those observed after the first and during the second equilibrium step (eqn (6)). During the second equilibrium step, NaMgH₃, Fe and NaH, and Mg are observed, indicating that NaMgH₃ is decomposing. After the third equilibrium (final products), Na, Mg and Fe are the main constituents, although residual NaH is also observed. Therefore the decomposition process can be described according to eqn (5)–(7) and Fig. 4.

The thermal stability of Na₂Mg₂FeH₈ is enhanced compared to that of Mg₂FeH₆, which exhibits a H₂ desorption maxima at ca. 360 °C (Fig. 2) with an associated ΔH_des of 261 kJ mol⁻¹.²⁹ The additional stability achieved by the incorporation of Na⁺ and H⁻ into the compound, induces a significant increase in desorption temperature maxima to 400 °C and a total ΔH_des of 378 kJ mol⁻¹ (Fig. 4). This value correlates very well with the previous DFT calculations conducted on this compound, which determined ΔH_f to be −328 kJ mol⁻¹.²³

In contrast to Na₂Mg₂FeH₈, Na₂Mg₂RuH₈ is stable above pressures of 0.19 MPa H₂ and T >500 °C (Fig. 3a). PXD analysis of material annealed at 6 MPa H₂ and 450 °C (Fig. 3f) indicates that the only modification is the disappearance of the unknown phase (observed after initial synthesis (Fig. 1)) which is replaced by another unknown phase. This material can be indexed to an orthorhombic unit cell of a = 14.5331, b = 7.9841 and c = 6.2429 and crystallises in a possible space group of Pmmn, although structural identification is inhibited by the low concentration and weak intensity of the Bragg peaks. As was observed from the GC results (Fig. 1), decomposition is also noted to follow a two-step decomposition route by PCI. At 500 °C, the first plateau is observed at an equilibrium pressure of 0.19 MPa, while the second occurs at ca. 0.07 MPa. Each step was determined to have an associated mass loss of ca. 1.9 wt%, with a total of 3.8 wt% H₂ being desorbed out of a maximum theoretical capacity of 4.0 wt%. This process was also carried out at 475 and 450 °C. This allowed for ΔH_des and ΔS to be determined to be 131 kJ mol⁻¹ H₂ and 176 ΔS (J mol⁻¹ H₂/K), respectively for step 1 (R² = 0.984) and ΔH_des = 119 kJ mol⁻¹ H₂ and ΔS = 151 ΔS (J mol⁻¹ H₂/K) for step 2 (R² = 0.999) (Fig. 3d). Therefore ΔH_des for the entire system is surmised to be 500 kJ mol⁻¹ (125 kJ mol⁻¹ H₂). PXD of the products at each stage allows a greater insight into those determined by ex situ heating in vacuo. After the first plateau, a substantial level of Mg₂RuH₄ is identifiable in the powder, along with Ru, NaOH and a small quantity of Mg₃Ru₂. The highly oxidisable Na (residual after evaporation) is the source of NaOH (occurring during PXD analysis), while the thermally unstable Mg₂RuH₄ is the source of Mg₃Ru₂. An unknown phase is also observed at this temperature and pressure, which due to the low intensity of the Bragg peaks associated with this material, indexing and as such, structural refinement was not possible. After the second plateau, the remaining powder consists of Mg₃Ru₂ and Ru. Presumably the excess Ru required to form Mg₃Ru₂ (without leaving excess Mg) comes from the excess Ru that remains in the starting material (Fig. 1). Therefore the decomposition process can be described according to eqn (10) and (11) and Fig. 4.

Na₂Mg₂RuH₈ → 2Na + Mg₂RuH₄ + 2H₂ (2.0 wt% H₂) (10)

Mg₂RuH₄ + 2Na → Mg₃Ru₂ + 2Na + 2H₂ (2.0 wt% H₂) (11)

The decomposition pathway of these materials differ significantly in that NaMgH₃ is the intermediate for Na₂Mg₂FeH₈, while Na₂Mg₂RuH₈ disassembles via Mg₂RuH₄ (Fig. 4). The versatile 4d Ru metal center is known to form a variety of complex anions including [Ru₂H₆]¹²⁻, [RuH₄]_n⁴ⁿ⁻, [RuH₅]_av⁵⁻, [RuH₆]⁴⁻ and [RuH₇]³⁻, while the 3d Fe metal center is only known to form [FeH₆]⁴⁻ anions.¹ As a consequence, Na₂Mg₂RuH₈ forms the [RuH₄]_n⁴ⁿ⁻ polyanionic intermediate upon decomposition,³⁷ whereas Na₂Mg₂FeH₈ preferentially decomposes to the thermally stable NaMgH₃³⁶ rather than Mg₂FeH₆ (Fig. 4). Mg₂RuH₄ was not observed during the ex situ heating experiments of Na₂Mg₂RuH₈ (Fig. 2e) as it is not thermodynamically stable at the temperatures at which Na₂Mg₂RuH₈ decomposes (420 °C, in vacuo).³⁷ Although, the synthesis of Mg₂RuH₄ is accomplished at 450 °C under 0.2 MPa H₂, it would presumably decompose at the temperatures imposed here, especially in vacuo. Stabilisation of this species is therefore viable under H₂ pressures of 0.05–0.02 MPa.

Conclusions

The optimised syntheses of Na₂Mg₂TH₈ (T = Fe, Ru) have been reported. Ball milling of stoichiometric quantities of NaH, MgH₂ and Fe followed by hydrogenation allows for a yield of >90% purity of Na₂Mg₂FeH₈via the formation of a NaMgH₃ intermediate. On the contrary, no intermediate is observed during the synthesis of Na₂Mg₂RuH₈ using an identical procedure.

The thermal decomposition of both Na₂Mg₂TH₈ materials have been studied by ex situ PXD, GC and PCI measurements. The first desorption maxima of Na₂Mg₂FeH₈ has been established to occur at ca. 400 °C, while Na₂Mg₂RuH₈ has its first maxima at 420 °C. The decomposition pathways of these isostructural compounds differs considerably, with Na₂Mg₂FeH₈ proceeding via NaMgH₃ in a three-step process, while Na₂Mg₂RuH₈ decomposes via Mg₂RuH₄ in a two-step process. The dissimilarity between the pathways originates from the capability of the 4d Ru metal centre to exist in a variety of [RuH_x]ⁿ⁻ complexes compared to Fe, which only exists as [FeH₆]⁴⁻.

The enthalpy and entropy of desorption for Na₂Mg₂TH₈ (T = Fe, Ru) for each stage of decomposition has been established to be by PCI measurements. The total enthalpy of desorption for Na₂Mg₂FeH₈ is 95 kJ mol⁻¹ H₂ and 125 kJ mol⁻¹ H₂ for Na₂Mg₂RuH₈.

Acknowledgements

We would like to acknowledge Ms. Warifune for the synthesis of the Mg₂FeH₆ starting material. We also appreciate the financial support from JSPS KAKENHI Grant Number 25220911.

Notes and references

1. K. Yvon and G. Renaudin, Encyclopedia of Inorganic Chemistry, John Wiley & Sons, Ltd, 2006.
2. K. Yvon, Chimia, 1998, 52, 613–619.
3. J. J. Reilly and R. H. Wiswall, Inorg. Chem., 1968, 7, 2254–2256.
4. H. Blomqvist and D. Noréus, J. Appl. Phys., 2002, 91, 5141–5148.
5. M. Lelis, D. Milcius, E. Wirth, U. Halenius, L. Eriksson, K. Jansson, K. Kadir, J. Ruan, T. Sato, T. Yokosawa and D. Noréus, J. Alloys Compd., 2010, 496, 81–86.
6. R. J. Westerwaal, M. Slaman, C. P. Broedersz, D. M. Borsa, B. Dam, R. Griessen, A. Borgschulte, W. Lohstroh, B. Kooi, G. ten Brink, K. G. Tschersich and H. P. Fleischhauer, J. Appl. Phys., 2006, 100, 063518.
7. T. J. Richardson, J. L. Slack, R. D. Armitage, R. Kostecki, B. Farangis and M. D. Rubin, Appl. Phys. Lett., 2001, 78, 3047–3049.
8. M. Polanski, T. K. Nielsen, Y. Cerenius, J. Bystrzycki and T. R. Jensen, Int. J. Hydrogen Energy, 2010, 35, 3578–3582.
9. M. Polanski, K. Witek, T. K. Nielsen, L. Jaroszewicz and J. Bystrzycki, Int. J. Hydrogen Energy, 2013, 38, 2785–2789.
10. J. J. Didisheim, P. Zolliker, K. Yvon, P. Fischer, J. Schefer, M. Gubelmann and A. F. Williams, Inorg. Chem., 1984, 23, 1953–1957.
11. G. Li, M. Matsuo, S. Deledda, R. Sato, B. C. Hauback and S. Orimo, Mater. Trans., 2013, 54, 1532–1534.
12. T. J. Richardson, J. L. Slack, B. Farangis and M. D. Rubin, Appl. Phys. Lett., 2002, 80, 1349–1351.
13. B. Bogdanović, A. Reiser, K. Schlichte, B. Spliethoff and B. Tesche, J. Alloys Compd., 2002, 345, 77–89.
14. S. F. Parker, K. P. J. Williams, M. Bortz and K. Yvon, Inorg. Chem., 1997, 36, 5218–5221.
15. B. Huang, K. Yvon and P. Fischer, J. Alloys Compd., 1995, 227, 121–124.
16. B. Huang, F. Gingl, F. Fauth, A. Hewat and K. Yvon, J. Alloys Compd., 1997, 248, 13–17.
17. B. Huang, K. Yvon and P. Fischer, J. Alloys Compd., 1994, 204, 5–8.
18. G. Renaudin, L. Guenee and K. Yvon, J. Alloys Compd., 2003, 350, 145–150.
19. M. Orlova, J.-P. Rapin and K. Yvon, Inorg. Chem., 2009, 48, 5052–5054.
20. T. D. Humphries, S. Takagi, G. Li, M. Matsuo, T. Sato, M. H. Sørby, S. Deledda, B. C. Hauback and S. Orimo, J. Alloys Compd., 2015 DOI:10.1016/j.jallcom.2014.12.113.
21. M. Matsuo, H. Saitoh, A. Machida, R. Sato, S. Takagi, K. Miwa, T. Watanuki, Y. Katayama, K. Aoki and S. Orimo, RSC Adv., 2013, 3, 1013–1016.
22. H. Saitoh, S. Takagi, M. Matsuo, Y. Iijima, N. Endo, K. Aoki and S. Orimo, APL Mater., 2014, 2, 076103.
23. S. Takagi, T. D. Humphries, K. Miwa and S. Orimo, Appl. Phys. Lett., 2014, 104, 203901.
24. M. Di Chio, L. Schiffini, S. Enzo, G. Cocco and M. Baricco, J. Alloys Compd., 2007, 434, 734–737.
25. K. Kadir and D. Noréus, Inorg. Chem., 2007, 46, 3288–3289.
26. B. Huang, K. Yvon and P. Fischer, J. Alloys Compd., 1992, 187, 227–232.
27. B. Huang, K. Yvon and P. Fischer, J. Alloys Compd., 1992, 190, 65–68.
28. B. Huang, K. Yvon and P. Fischer, J. Alloys Compd., 1993, 197, 65–68.
29. J. A. Puszkiel, P. Arneodo Larochette and F. C. Gennari, J. Alloys Compd., 2008, 463, 134–142.
30. A. Boultif and D. Louer, J. Appl. Crystallogr., 2004, 37, 724–731.
31. J. Laugier and B. Bochu, LMGP-Suite Suite Programs Interpret. X-ray Exp. by Jean laugier Bernard Bochu, ENSP/Laboratoire des Matériaux du Génie Phys. BP 46. 38042 Saint Martin d'Hères, Fr.
32. B. H. Toby, J. Appl. Crystallogr., 2001, 34, 210–213.
33. A. C. Larson and R. B. Von Dreele, Los Alamos National Laboratory Report, 2000, Laur 86-748.
34. B. E. Warren, X-ray Diffraction, Courier Dover Publications, 1969.
35. H. Reardon, N. Mazur and D. H. Gregory, Prog. Nat. Sci., 2013, 23, 343–350.
36. D. A. Sheppard, M. Paskevicius and C. E. Buckley, Chem. Mater., 2011, 23, 4298–4300.
37. F. Bonhomme, K. Yvon, G. Triscone, K. Jansen, G. Auffermann, P. Müller, W. Bronger and P. Fischer, J. Alloys Compd., 1992, 178, 161–166.

---