Dehydrogenation behavior and microstructure evolution of hydrogenated magnesium–nickel–yttrium melt-spun ribbons

Abstract

This paper reports porous Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons which are prepared by a melt spinning method. The dehydrogenation behaviors, including the thermal decomposition process, dehydrogenation capacities, percentage and kinetics, are investigated in a temperature range of 523–623 K. The microstructure evolution during the dehydrogenation process of metal melt-spun ribbons is discussed with respect to the microstructures of hydrogenated and dehydrogenated ribbons. It is considered that the porous structure in the as-prepared ribbons accelerates the diffusion of hydrogen gas and atoms, and increases dehydrogenation nucleation locations and decreases the diffusion distance. Additionally, the addition of yttrium and melt-spinning decreases the activation energy of dehydrogenation. A low activation energy and a refined microstructure play dominant roles in decreasing the dehydrogenation temperature and increasing the kinetics of Mg-based metal melt-spun ribbons.

---

1. Introduction

Magnesium-based alloys are considered a promising medium for hydrogen storage for on-board vehicle and stationary applications to solve environment and energy issues, due to their high theoretical storage capacity (e.g. Mg 7.6 wt%), light weight and abundance. However, magnesium-based alloys have the high stability of the hydrides and a low hydrogen atom diffusion rate, which causes a high dehydrogenation temperature and low hydrogen desorption kinetics, and limits industrial applications.¹

A great many efforts have been devoted to improving the dehydrogenation properties of Mg-based alloys from two aspects, element doping and microstructure modification. Work has proved that the addition of transition metal elements,^2,3 rare metal elements^4,5 and catalysts^6,7 into Mg-based alloys decreases the stability of hydrides and increases the dehydrogenation kinetics.⁸ However, the excellent kinetics caused by the addition of transition metal elements are at the expense of low dehydrogenation capacities. It has been found that rare earth (RE) elements not only accelerate the dehydrogenation kinetics, but also enhance the dehydrogenation capacity compared with transition metal elements by substituting B side elements of A₂B-type Mg-based alloys, and decrease the hydrogen desorption temperature.⁹ By adding rare earth elements, the high activation of rare earth elements is able to catalyse the decomposition of hydrogen molecules on the surface of alloy particles.¹⁰ During the dehydrogenation process, the compounds and RE hydrides could be formed in the alloys and these products have been proved to provide positive catalysis on the formation and decomposition of other hydrides.¹¹ Yttrium (Y), especially, is lighter than other RE elements, and it favors an increase in gaseous hydrogen absorption/desorption capacities per unit mass. Additionally, strain and defects, which are induced by the formation and decomposition of YH₂/YH₃, have been proved to facilitate hydrogen absorption and desorption.¹⁰ The addition of elements and catalysts can significantly improve the hydrogenation/dehydrogenation rates,¹² however, these methods still cannot solve the problems of slow dehydrogenation kinetics well.¹³

The microstructure is closely related to the thermodynamics and kinetics of hydrogenation/dehydrogenation.⁸ The diffusion of hydrogen atoms in alloys and hydrides is an important factor in the hydrogen desorption kinetics. A nanoscale-refined grain size and amorphous structure can significantly enhance the hydrogen absorption/desorption kinetics and hydrogen storage capacity since the refined grain generates a large number of interfaces along which hydrogen atoms can diffuse more easily.^14,15 For microstructure modification, ball milling¹⁶ and melt spinning¹⁷ have been proved to be powerful techniques to produce amorphous and nanocrystalline Mg₂Ni alloys.⁸ However, ball milling is still limited by some unavoidable shortcomings: contamination of metal powders by milling equipment and oxygen even in argon atmosphere,¹⁸ small scale production and its relatively high cost for special safe handling of the powder such as heat-transfer and pipe blockage.³ In addition, the growth of ball milled particles occurs at a high reaction temperature and deteriorates the kinetic properties.¹⁹ On the contrary, the microstructure stability of melt spun ribbons is higher than that of ball milled particles during hydrogenation and dehydrogenation cycles.^8,20 Generally speaking, the amorphous and nanocrystalline Mg-based alloys prepared by melt-spinning have faster hydrogen absorption and desorption kinetics.²¹ Melt spinning has been proved an effective method to reduce grain size and possesses attractive hydrogenation/dehydrogenation properties.²² In addition to the enhanced diffusivity of hydrogen atoms, the diffusion distance is also decreased by melt spinning. The melt-spun ribbons can absorb and release hydrogen from two sides, therefore the ribbons can be hydrogenated completely with the thickness less than 50 μm.²³ In addition, increased porosity in metals significantly increases the diffusion of gas.²⁴ The dehydrogenation properties are closely related to the microstructure, and the microstructure evolution during dehydrogenation process is necessary to analyze the dehydrogenation mechanism of melt spun ribbons.

The two important factors affecting hydrogen desorption kinetics are the dehydrogenation energy barrier of hydrides and the diffusion of hydrogen atoms. The addition of RE elements facilitates the glass formation of Mg₂Ni-based alloys.⁸ The melt spinning technique combined with the addition of Y into Mg₂Ni alloys is employed in this work to decrease the dehydrogenation energy barrier of hydrides and facilitate the diffusion of hydrogen atoms. Special attention was paid to the microstructure evolution and phase transformation of Y-doped Mg-based melt-spun ribbons during the dehydrogenation process. The activation energy of the dehydrogenation reaction is calculated by the Kissinger method. In addition, the effect of temperature on dehydrogenation kinetics is also emphasized and the corresponding influencing factors are discussed. The corresponding dehydrogenation mechanism of Mg–Ni–Y melt-spun ribbons is proposed in this work.

2. Experimental

Mg-based hydrogen storage alloys with the composition Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6 at.%) were melted under the protection of covering flux in a resistance furnace. The alloys were re-melted and spun on a single copper roller with linear velocity 40 ms⁻¹ in an argon atmosphere of 500 mbar.²⁵ The melt-spun ribbons were about 50 μm in thickness, 5 mm in width and 10–30 mm in length. The ribbons, which were prepared by the melt-spinning technique, were directly hydrogenated in an automatic Sieverts apparatus (PCT Pro2000) under an initial hydrogen pressure of 30 bar at two temperatures of 523 and 623 K for 4 h. The dehydrogenation kinetics were measured under initial vacuum at the corresponding hydrogenation temperature for 1 h. Before hydrogenation/dehydrogenation measurements, all the alloy ribbons were activated by three continuous hydrogenation/dehydrogenation cycles maintaining an initial hydrogen pressure of 30 bar at 10 h for hydrogenation and vacuum for 1 h for dehydrogenation. The metallographic samples were prepared without any etching reagent even water. The phase composition and microstructure of the hydrogenated ribbons were characterized by X-Ray Diffraction (XRD), Scanning Electron Microscopy (SEM, TESCAN, MIR A3 XMU) and Transition Electron Microscopy (TEM, TECNAI, F30 G²). The phase transformations during dehydrogenation of the hydrogenated ribbons were analyzed by Differential Scanning Calorimetry (DSC, NETZSCH, STA449C) at various heating rates of 5, 10, 15 and 20 K min⁻¹ from ambient temperature to 723 K in an argon atmosphere.

3. Results and discussion

3.1 Phase analysis of hydrogenated ribbons

The XRD patterns of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons are shown in Fig. 1a. The dominant phase of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons is monoclinic Mg₂NiH₄. Bragg peaks of Mg₂NiH_0.3 and MgH₂ are also observed in the XRD patterns. In addition, several Bragg peaks of MgYNi₄, YH₂, YH₃ and cubic structure Mg₂NiH₄ phases are observed in the XRD pattern of the hydrogenated ribbon with x = 6. The intensity (counts) of the Bragg peaks of the MgH₂ phase (∼27.9°) increases from 599 to 2375 with the increasing Y content in the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons. It indicates that the phase content of MgH₂ increases with the increase of Y content, which coincides with the content of reactant α-Mg phase in melt-spun ribbons.²⁵ On the contrary, the phase content of Mg₂NiH_0.3 decreases with the increase of Y content, which can be seen from the decreasing intensities of the Bragg peaks from Mg₂NiH_0.3 (about 21, 37 and 47°). Mg₂NiH_0.3 is a transitional phase of the Mg₂Ni–Mg₂NiH₄ system. The addition of Y is beneficial to decrease the content of the transitional phase Mg₂NiH_0.3 by facilitating the transformation of Mg₂NiH_0.3 to Mg₂NiH₄. The hydride of MgYNi₄ is not observed in the XRD patterns of the hydrogenated melt-spun ribbons because the MgYNi₄ absorbs hydrogen reversibly at room temperature,²⁶ and almost all of the hydrogen is released under 500 K.²⁷ The formation of hydride YH₂, as the catalyst, facilitates the formation of Mg₂NiH_0.3 and MgH₂ during the hydrogenation process.¹¹ Then Mg₂NiH_0.3 and YH₂ are completely hydrogenated to Mg₂NiH₄ and YH₃, respectively. The strain and defects induced by YH₂/YH₃ are believed to facilitate the hydrogenation/dehydrogenation process.¹¹

Fig. 1 The XRD patterns of the (a) hydrogenated and (b) dehydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons.

Fig. 1b shows the XRD patterns of the dehydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons. The dominant phase of dehydrogenated Mg₆₇Ni_33−xY_x ribbons is Mg₂Ni. The MgNi₂, which is eliminated by the melt-spinning technique, is separated out from the dehydrogenated ribbons after 4 hydrogenation/dehydrogenation cycles. The Bragg peak intensity of MgNi₂ decreases with the increase of Y in the dehydrogenated ribbons. The Bragg peaks of Mg (∼43.3°) and MgYNi₄ (∼42.7°) are in the dehydrogenated Mg₆₇Ni_33−xY_x (x = 1, 3, 6) ribbons, their intensities increase with increasing Y in the dehydrogenated ribbons. The YH₂, which is an incomplete dehydrogenation product of YH₃, also appears in the Bragg peaks of dehydrogenated Mg₆₇Ni_33−xY_x ribbons with x = 6. The phases of the dehydrogenated ribbons are similar to those of the melt-spun ribbons²⁵ except for the addition of MgNi₂ and YH₂ after 4 hydrogenation/dehydrogenation cycles.

From comparing the phase structures of the hydrogenated and dehydrogenated samples, the phase transformations contain Mg₂NiH₄ to Mg₂Ni, Mg₂NiH_0.3 to Mg₂Ni, MgH₂ to Mg, and YH₃ to YH₂. The MgYNi₄ phase appears both in hydrogenated and dehydrogenated samples, but the hydride of MgYNi₄ is not observed. The hydride of MgYNi₄ is probably dehydrogenated in this experimental condition.²⁷ In addition, the trace MgNi₂ phase precipitates out in 4 hydrogenation/dehydrogenation cycles. The broadening of the XRD patterns indicates that both amorphous and nanocrystalline regions are included in the hydrogenated and dehydrogenated ribbons.

3.2 Microstructures of hydrogenated ribbons

The crystallite sizes of the ribbons increase from 5 nm to 150 nm with the addition of Y content in the melt-spun Mg₆₇Ni_33−xY_x (x = 1, 3, 6) ribbons. Fig. 2 shows a high resolution image of the hydrogenated ribbons, the corresponding Selected Area Electron Diffraction (SAED) pattern and Electron Energy-Loss Spectroscopy (EELS) hydrogen mapping. The ribbons contain amorphous and nanocrystalline regions as shown in Fig. 2a, which is consistent with the results of the XRD. The circled area is amorphous and the straight-line portion is nanocrystalline. In addition, the continuous rings with scattered diffraction points in the SAED pattern also indicate that the microstructures of ribbons are amorphous and nanocrystalline. Nanocrystalline hydrides enhance the hydrogenation kinetics.²⁸ The SAED pattern shows that the ribbons are polycrystalline within an electron beam diameter. The major phase Mg₂NiH₄ is observed in the hydrogenated ribbons in the polycrystalline SAED rings. It has been proven that the amorphous and nanocrystalline regions can provide passageways for the diffusion of hydrogen atoms.²⁹ The grains of the hydrogenated ribbons are refined for the effect of melt-spinning. The amorphous and nanocrystalline regions formed by melt-spinning can shorten the diffusion distance and enhance the diffusivity of hydrogen.

Fig. 2 HRTEM image and SAED pattern of the Mg₆₇Ni_33−xY_x (x = 1, 3, 6) ribbons and EELS hydrogen mapping. (a) x = 1, (b) x = 3, (c) x = 6, (d) EELS hydrogen mapping of x = 3 ribbons.

The High Resolution Transmission Electron Microscopy (HRTEM) image and the corresponding Fast Fourier Transform (FFT) pattern of the hydrogenated Mg₆₇Ni_33−xY_x (x = 3) ribbon are shown in Fig. 2b. The amorphous structure and nanocrystallinity are observed in the HRTEM image. The FFT pattern is an amorphous ring with slight scattered diffraction points. Fig. 2c shows the HRTEM image of the hydrogenated Mg₆₇Ni_33−xY_x (x = 6) ribbons. The interplanar spacing and Miller indices of Mg₂NiH₄ are marked in the HRTEM image and SAED pattern. A lot of stacking faults are observed in the HRTEM image. Crystal defects are induced by hydrogen absorption/desorption in Mg₂Ni and other hydrogen storage alloys.³⁰ It has been reported that the stacking faults are beneficial to the nucleation of hydrides and lead to lower hydrogen desorption temperatures and better kinetics.³¹ Comparing with the ribbons with x = 1 and 3, the hydrogenated ribbons with x = 6 contain more crystalline structures as shown in Fig. 2c. Plenty of amorphous structures are observed in the melt-spun ribbons of x = 6 before hydrogenation. The crystallization of the amorphous structure occurs in the hydrogenation process.²¹ In addition, stacking faults are formed in the crystallization process. It has been reported that changing the structure of the Mg₂Ni alloy from polycrystalline to nanocrystalline results in a drop of about 100 K in absorption/desorption temperature.³² The amorphous alloy exhibits faster hydrogenation kinetics than the crystallized alloy with the same composition due to the faster hydrogen diffusion in the amorphous phase.⁴ Amorphous and nanocrystalline regions are helpful to the hydrogenation/dehydrogenation.

In order to analyze the distribution of hydrides, the EELS hydrogen mapping is characterized as shown in Fig. 2d. The white portion is the area including hydrogen atoms. The dark particle surface provides evidence of partial hydrogen atom-release from the particle surface under the high vacuum of the TEM and the effect of the electron beam. The hydrogen desorption process is clearly seen in the EELS maps at ∼400 °C,³³ but the Mg-based hydrides are rather stable at room temperature. The SAED pattern shows that the ribbons have been hydrogenated and the dominant hydride can be ascribed to Mg₂NiH₄.

3.3 Microstructure evolution during dehydrogenation

The dehydrogenation process of ball milled particles has been described by Tanniru et al.³⁴ However, the dehydrogenation process of melt-spun ribbon has not been described. In order to analyze the dehydrogenation process, the microstructures of hydrogenated and dehydrogenated melt-spun ribbons are analyzed in this work as shown in Fig. 3. Plenty of connected pores, which have been observed in the microstructure of melt-spun ribbons,²⁵ are also observed in the microstructure of hydrogenated and dehydrogenated ribbons after 4 hydrogenation/dehydrogenation cycles. Owing to the insulativity and lower z-contrast (atomic number) of the hydride, the hydride phase appears darker than the metal phase in the BSE mode.^35,36 The areas around the pores are preferentially hydrogenated as shown in Fig. 3a. The process of hydrogen desorption is the inverse process of absorption.³⁷ The areas around the pores in melt spun ribbons are similar to the surface of ball milled particles. The areas around the pores are the preferential positions for dehydrogenation. Cracks can be observed in the dehydrogenated ribbons as shown in Fig. 3b. The hydrogen absorption and desorption of alloys results in the expansion and contraction of the crystal lattice during the hydrogenation/dehydrogenation cycles, inducing stress in the grain. Hydrogenation/dehydrogenation cycles increase the brittleness of the ribbons, but serious damage did not occur. The melt spun ribbons are distorted during hydrogenation/dehydrogenation cycles and the distortion counteracts the partial distortion energy. Plenty of amorphous and nanocrystalline regions in the melt-spun ribbons provide more diffusion paths for hydrogen atoms. The grain size reduction to the nanoscale decreases the diffusion distance. The porous surface structure increases the hydrogen absorption rate.²⁸ Hydrogen diffusion through the bulk metal (hydride) phase is usually quite slow and often rate-limiting.³⁸ Hydrogen diffusion in the gaseous phase is fast, greater than that in the bulk metal phase. It has been reported that fast hydrogen diffusion through or along the porous carbon phase benefits the kinetics of hydrogen release and uptake.³⁸ The pores in the melt-spun ribbons provide plenty of surfaces, increasing dehydrogenation nucleation locations and decreasing the diffusion distances. In addition, the released hydrogen gas can be exhausted from the ribbons by the connected pores, further improving the hydrogen desorption kinetics. The areas around the pores are the preferential positions for hydrogenation/dehydrogenation.

Fig. 3 The SEM/BSE images of (a) hydrogenated and (b) dehydrogenated melt-spun ribbons.

3.4 Thermal decomposition behavior of hydrogenated ribbons

Fig. 4a shows the dehydrogenation DSC curves and the corresponding first derivative curves (DDSC) of hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons at a heating rate of 10 K min⁻¹. The obvious endothermic peak around 520–560 K observed in each DSC curve can be ascribed to the dehydrogenation peak of the dominant Mg₂NiH₄ phase. The dehydrogenation phases are Mg₂NiH₄, Mg₂NiH_0.3, MgH₂ and YH₃ from the phase analysis. The dehydrogenation peak of the intermediate hydride Mg₂NiH_0.3 phase overlaps Mg₂NiH₄. The dehydrogenation temperature of MgH₂ is higher than that of Mg₂NiH₄, the endothermic peak of MgH₂ powder is about 664 K, but it is still over 593 K in the Mg₈₀Ni₁₀Y₁₀ alloy with the addition of Ni and Y.⁹ The trace endothermic peaks of MgH₂ in Fig. 4a are about 600–660 K for the minor content of MgH₂. It has been reported that the desorption temperature of YH₃ into YH₂ in vacuum starts at 443 K and the peak temperature is 533 K.³⁹ The endothermic peak of YH₃ to YH₂ overlaps that of Mg₂NiH₄ in the Mg₆₇Ni_33−xY_x ribbon with x = 6. The main endothermic peak of Mg₂NiH₄ is not symmetric, peaks overlap the main endothermic peak as can be seen in the DDSC curve. The decomposition of Mg₂NiH₄ is divided into two steps, which are dehydrogenation of Mg₂NiH₄ to Mg₂NiH_0.3 and dehydrogenation of Mg₂NiH_0.3 to Mg₂Ni.⁴⁰ Therefore, the main dehydrogenation peaks contain overlapping peaks. The area under the peaks reveals the heat of the absorption of ribbons per unit mass. The heat of the absorption of the ribbon with x = 0 exceeds 1700 J g⁻¹, and it is below 1300 J g⁻¹ with x = 1, 3 and 6. The addition of Y decreases the heat of the absorption of ribbons per unit mass in the dehydrogenation process.

Fig. 4 Dehydrogenation DSC and corresponding first derivative curves of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons. (a) Heating rate of 10 K min⁻¹ in Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons, (b) heating rates of 5, 10, 15 and 20 K min⁻¹ in x = 1 ribbons.

It is obvious that the peak dehydrogenation temperature (∼525 K) of the melt-spun ribbon with x = 1 is lower compared with that of the melt-spun ribbons with x = 0, 3 and 6 (∼542, ∼539 and ∼554 K, respectively) in Fig. 4a. In order to analyze the activation energy of the dehydrogenation reaction of Mg₂NiH₄, the dehydrogenation DSC curves of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons at heating rates of 5, 10, 15 and 20 K min⁻¹ are measured and the DSC curves of x = 1 ribbons are shown in Fig. 4b. The corresponding dominant dehydrogenation peak temperatures of Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons are summarized in Table 1. The Arrhenius equation,⁴¹, indicates that the lower activation energy E_a is related to faster reaction rate k and lower temperature T. The apparent activation energy for dehydrogenation of the dominant phase Mg₂NiH₄ in the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons is calculated based on the dehydrogenation peak temperatures in Table 1 using the Kissinger equation (eqn (1)).⁴²

where β is the heating rate, T_p is the peak temperature, A_k is the preexponential factor, E_k is the dehydrogenation activation energy, R is the gas constant (8.3145 J (K mol)⁻¹). ln(β/T_p²) and 1/T_p are linear correlation coefficients. Fitting curves of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons using the Kissinger method present a good linear relation between ln(β/T_p²) and 1/T_p with a slope of –E_k/R, as shown in Fig. 5a. The activation energy E_k of the Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons in the dehydrogenation can be summarized as 102, 73, 98 and 80 kJ mol⁻¹ by the Kissinger equation,⁴² respectively. The E_k of melt-spun ribbons with x = 0 is 102 kJ mol⁻¹, which is close to that of Mg₂NiH₄.⁴³ The trend line of activation energy of the Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons is shown in Fig. 5b. The dehydrogenation activation energy of the ribbons with x = 1, 3, 6 is lower than that of the hydrogenated ribbon with x = 0 with the addition of Y. The hydrogenated melt-spun ribbon with x = 1 possesses the lowest activation energy of the dehydrogenation reaction among the experimental alloys. The small quantity addition of Y and refinement of the particles and grains decreases the apparent activation energy for hydrogen desorption by shortening the diffusion distance and enhancing the diffusivity with the increase of the micro strain and lattice distortion.⁴⁴ YH₂ and YH₃ are generated in the Mg-based alloys in hydrogenation and dehydrogenation processes. Fine YH₂/YH₃ particles provide a large number of active sites and paths to enhance the H atom diffusion along abundant phase boundaries and hydride phases.⁴⁵ The refined grain size of alloys, the multiphase structure and YH₂/YH₃ decrease the activation energy during dehydrogenation.

Table 1 The dehydrogenation peak temperatures of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) melt-spun ribbons at heating rates of 5, 10, 15 and 20 K min⁻¹

Composition Mg67Ni33−xYx	Dehydrogenation peak temperature
	5 K min^−1	10 K min^−1	15 K min^−1	20 K min^−1
x = 0	527	542	553	556
x = 1	507	525	534	547
x = 3	526	539	549	557
x = 6	538	554	568	577

---

Fig. 5 Fitted Kissinger curves and corresponding dehydrogenation activation energy curves of the hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons. (a) Fitted Kissinger curves, (b) the activation energy curves of the dehydrogenation reaction.

3.5 Dehydrogenation kinetics

The activation energy is related to the reaction temperature as well as the reaction kinetics. The hydrogen desorption capacities of the Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons detected at 523 and 623 K are shown in Fig. 6. The hydrogen desorption capacities of all the ribbons with added Y are higher than Mg₆₇Ni_33−xY_x (x = 0) ribbons at 523, 573 (ref. 25) and 623 K. The addition of Y decreases the activation energy of the dehydrogenation reaction and increases the hydrogen desorption capacity corresponding with Mg₆₇Ni_33−xY_x (x = 0) ribbons. The Mg₆₇Ni_33−xY_x ribbons are not dehydrogenated completely at the same hydrogenation temperature from the dehydrogenation percent shown in Fig. 6. It can be explained that partial hydrides (such as YH₂) do not release hydrogen at this temperature. In addition, hydrogen desorption is related to hydrogen pressure as well as the temperature.⁴⁶ Hydrogen pressure increases with the hydrogen desorption during the dehydrogenation process, and the partially dissolved hydrogen in the alloys is not released under the increased hydrogen pressure. Further decreasing the hydrogen pressure and raising the temperature can facilitate the dehydrogenation of hydrides. The content of MgYNi₄ increases with the increasing Y content in the ribbons with x = 1, 3 and 6.²⁵ Mg is consumed by forming MgYNi₄ but the hydride of MgYNi₄ is not observed in the hydrogenated ribbons. In addition, partial hydrides (such as YH₂) do not release hydrogen at the experimental temperature. The maximum dehydrogenation capacity of ribbons with x = 6 is lower than that of x = 3. The low activation energy of the dehydrogenation reaction makes it easy to release hydrogen. The ribbon with x = 1 has the maximum dehydrogenation capacity of the Mg₆₇Ni_33−xY_x ribbons for the lowest activation energy of the dehydrogenation reaction. But the ribbon with x = 1 has a relatively high dehydrogenation percentage even at 523 K. The low activation energy of the dehydrogenation reaction of the Mg₆₇Ni_33−xY_x (x = 1) ribbon not only increases the hydrogen desorption capacity and kinetics but decreases the dehydrogenation temperature.

Fig. 6 Hydrogen desorption content of the Mg₆₇Ni_33−xY_x ribbons.

Fig. 7 shows the dehydrogenation kinetics curves of the Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons at 523 and 623 K in the initial 1 and 0.5 h. As can be seen from the kinetics curves, the dehydrogenation rates of ribbons at 523 K are lower than those at 623 K. The increase of temperature improves the dehydrogenation kinetics of ribbons. After the addition of Y, the dehydrogenation kinetics of the Mg₆₇Ni_33−xY_x (x = 1, 3, 6) ribbons are better than that of ribbons with x = 0 at 523 K. Considering the values of the activation energy of dehydrogenation, the values of the Mg₆₇Ni_33−xY_x (x = 1, 3, 6) ribbons are lower than that of the ribbon with x = 0. The properties of dehydrogenation kinetics are related to the activation energy of dehydrogenation. In addition, the dehydrogenation kinetics are also related to the microstructure of the ribbons. The grain size of ribbons increases with the addition of Y elements. The dehydrogenation kinetics are the results of the activation energy of the dehydrogenation reaction combined with the microstructure. The ribbon with x = 1 has the highest kinetics in the Mg₆₇Ni_33−xY_x (x = 1, 3, 6) ribbons at 523 and 623 K. The high dehydrogenation kinetics of Mg₆₇Ni_33−xY_x ribbons with x = 1 are also related to the high pore percentage.²⁵

Fig. 7 The dehydrogenation kinetics curves of the Mg₆₇Ni_33−xY_x (x = 0, 1) ribbons at 523 and 623 K in the initial 1 and 0.5 h. (a) 523 K in the initial 1 h, (b) 623 K in the initial 0.5 h.

4. Conclusions

The dehydrogenation phase transformations of Mg₂NiH₄ to Mg₂Ni, Mg₂NiH_0.3 to Mg₂Ni, MgH₂ to Mg and YH₃ to YH₂ of hydrogenated Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons were investigated. Amorphous and nanocrystalline structures and stacking faults are contained in the Mg₆₇Ni_33−xY_x (x = 0, 1, 3, 6) ribbons. H is evenly distributed in the hydrogenated ribbons as shown by EELS analysis. The pores in the melt-spun ribbons are the preferential areas for hydrogenation/dehydrogenation. The pores in the metal ribbons generated by melt-spinning to facilitate the diffusion of hydrogen, increase dehydrogenation nucleation locations and decrease the diffusion distances. The high dehydrogenation kinetics are the result of low activation energy combined with amorphous and nanocrystalline regions. The low activation energy of the dehydrogenation reaction, caused by melt-spinning and the addition of Y, decreases the dehydrogenation temperatures and increases the dehydrogenation kinetics of Mg-based melt-spun ribbons.

Acknowledgements

This work is financially supported by the Doctorate Foundation of Northwestern Polytechnical University (CX201107) and Project supported by the Research Fund of the State Key Laboratory of Solidification Processing (NWPU) (95-QZ-2014). The Program of Introducing Talents of Discipline to Universities (B08040) is also acknowledged. Prof. Zhong-min Wang from Guilin University of Electronic Technology is gratefully appreciated for his help in hydrogen absorption/desorption measurements.

References

1. P. Vajeeston, P. Ravindran, B. C. Hauback, H. Fjellvåg, A. Kjekshus, S. Furuseth and M. Hanfland, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 73, 224102.
2. D. Vyas, P. Jain, J. Khan, V. Kulshrestha, A. Jain and I. P. Jain, Int. J. Hydrogen Energy, 2012, 37, 3755–3760.
3. Y. H. Cho, S. Aminorroaya, H. K. Liu and A. K. Dahle, Int. J. Hydrogen Energy, 2011, 36, 4984–4992.
4. Á. Révész, Á. Kis-Tóth, L. K. Varga, E. Schafler, I. Bakonyi and T. Spassov, Int. J. Hydrogen Energy, 2012, 37, 5769–5776.
5. J. X. Zou, X. Q. Zeng, Y. J. Ying, X. Chen, H. Guo, S. Zhou and W. J. Ding, Int. J. Hydrogen Energy, 2013, 38, 2337–2346.
6. R. K. Singh, H. Raghubanshi, S. K. Pandey and O. N. Srivastava, Int. J. Hydrogen Energy, 2010, 35, 4131–4137.
7. X. J. Hou, R. Hu, T. B. Zhang, H. C. Kou, J. S. Li and X. Y. Xue, Int. J. Hydrogen Energy, 2013, 38, 12904–12911.
8. Y. Zhang, T. Zhai, B. Li, H. Ren, W. Bu and D. Zhao, J. Mater. Sci. Technol., 2014, 30, 1020–1026.
9. S. Kalinichenka, L. Röntzsch, C. Baehtz and B. Kieback, J. Alloys Compd., 2010, 496, 608–613.
10. Z. N. Li, X. P. Liu, L. J. Jiang and S. M. Wang, Int. J. Hydrogen Energy, 2007, 32, 1869–1874.
11. J. W. Liu, C. C. Zou, H. Wang, L. Z. Ouyang and M. Zhu, Int. J. Hydrogen Energy, 2013, 38, 10438–10445.
12. S. Kalinichenka, L. Röntzsch and B. Kieback, Int. J. Hydrogen Energy, 2009, 34, 7749–7755.
13. S. T. Sabitu, G. Gallo and A. J. Goudy, J. Alloys Compd., 2010, 499, 35–38.
14. X. J. Hou, R. Hu, T. B. Zhang, H. C. Kou and J. S. Li, Int. J. Hydrogen Energy, 2013, 38, 16168–16176.
15. J. Cermák and L. Král, Acta Mater., 2008, 56, 2677–2686.
16. Á. Révész and D. Fátay, J. Power Sources, 2010, 195, 6997–7002.
17. C. Pohlmann, B. Kieback and L. Röntzsch, Int. J. Hydrogen Energy, 2014, 39, 8331–8339.
18. C. Suryanarayana, Prog. Mater. Sci., 2001, 46, 1–184.
19. Y. S. Au, M. K. Obbink, S. Srinivasan, P. C. M. M. Magusin, K. P. de Jong and P. E. de Jongh, Adv. Funct. Mater., 2014, 24, 3604–3611.
20. G. Y. Liang, D. C. Wu, L. Li and L. J. Huang, J. Power Sources, 2009, 186, 528–531.
21. Q. A. Zhang, L. X. Zhang and Q. Q. Wang, J. Alloys Compd., 2013, 551, 376–381.
22. E. A. Lass, Int. J. Hydrogen Energy, 2012, 37, 9716–9721.
23. I. P. Jain, C. Lal and A. Jain, Int. J. Hydrogen Energy, 2010, 35, 5133–5144.
24. V. Bérubé, G. Radtke, M. Dresselhaus and G. Chen, Int. J. Energy Res., 2007, 31, 637–663.
25. W. J. Song, J. S. Li, T. B. Zhang, H. C. Kou and X. Y. Xue, J. Power Sources, 2014, 245, 808–815.
26. K. Aono, S. Orimo and H. Fujii, J. Alloys Compd., 2000, 309, L1–L4.
27. N. Hanada, S.-I. Orimo and H. Fujii, J. Alloys Compd., 2003, 356–357, 429–432.
28. B. Sakintuna, F. Lamari-Darkrim and M. Hirscher, Int. J. Hydrogen Energy, 2007, 32, 1121–1140.
29. Y. H. Zhang, D. L. Zhao, S. H. Guo, Y. Qi, Z. W. Wu and X. L. Wang, J. Alloys Compd., 2009, 476, 457–461.
30. D. Liu, Y. Zhu and L. Li, Int. J. Hydrogen Energy, 2007, 32, 2417–2421.
31. N. Takeichi, K. Tanaka, H. Tanaka, T. T. Ueda, Y. Kamiya, M. Tsukahara, H. Miyamura and S. Kikuchi, J. Alloys Compd., 2007, 446–447, 543–548.
32. L. H. Kumar, B. Viswanathan and S. S. Murthy, J. Alloys Compd., 2008, 461, 72–76.
33. K. Tanaka, T. Miwa, K. Sasaki and K. Kuroda, J. Alloys Compd., 2009, 478, 308–316.
34. M. Tanniru, H. Y. Tien and F. Ebrahimi, Scr. Mater., 2010, 63, 58–60.
35. M. Vittori Antisari, A. Aurora, D. Mirabile Gattia and A. Montone, Scr. Mater., 2009, 61, 1064–1067.
36. H.-Y. Tien, M. Tanniru, C.-Y. Wu and F. Ebrahimi, Int. J. Hydrogen Energy, 2009, 34, 6343–6349.
37. K. C. Chou and K. Xu, Intermetallics, 2007, 15, 767–777.
38. P. Adelhelm and P. E. de Jongh, J. Mater. Chem., 2011, 21, 2417–2427.
39. V. A. Yartys, O. Gutfleisch, V. V. Panasyuk and I. R. Harris, J. Alloys Compd., 1997, 253–254, 128–133.
40. S. Kalinichenka, L. Röntzsch, C. Baehtz, T. Weißgärber and B. Kieback, J. Alloys Compd., 2011, 509(suppl. 2), S629–S632.
41. S. Arrhenius, Z. Phys. Chem., 1889, 4, 226.
42. H. E. Kissinger, Anal. Chem., 1957, 29, 1702–1706.
43. I. C. Atias-Adrian, F. A. Deorsola, G. A. Ortigoza-Villalba, B. De Benedetti and M. Baricco, Int. J. Hydrogen Energy, 2011, 36, 7897–7901.
44. Y. P. Pang, Y. F. Liu, X. Zhang, M. X. Gao and H. G. Pan, Int. J. Hydrogen Energy, 2013, 38, 1460–1468.
45. F. P. Luo, H. Wang, L. Z. Ouyang, M. Q. Zeng, J. W. Liu and M. Zhu, Int. J. Hydrogen Energy, 2013, 38, 10912–10918.
46. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353–358.

---