Effect of graphite (GR) content on microstructure and hydrogen storage properties of nanocrystalline Mg24Y3–Ni–GR composites

Effect of graphite (GR) content on microstructure and hydrogen storage properties of nanocrystalline Mg24Y3–Ni–GR composites

Journal of Alloys and Compounds 726 (2017) 498e506 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: http:...

4MB Sizes 3 Downloads 27 Views

Journal of Alloys and Compounds 726 (2017) 498e506

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom

Effect of graphite (GR) content on microstructure and hydrogen storage properties of nanocrystalline Mg24Y3eNieGR composites Tai Yang a, b, c, Chunyong Liang a, b, *, Xinghua Wang a, b, Hongshui Wang a, b, Zeming Yuan c, Fuxing Yin a, b, Qiang Li a, b, **, Yanghuan Zhang c, *** a

Research Institute for Energy Equipment Materials, Hebei University of Technology, Tianjin 300130, China Tianjin Key Laboratory of Laminating Fabrication and Interface Control Technology for Advanced Materials, Hebei University of Technology, Tianjin 300130, China c Department of Functional Material Research, Central Iron and Steel Research Institute, Beijing 100081, China b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 June 2017 Received in revised form 19 July 2017 Accepted 3 August 2017 Available online 4 August 2017

Ternary Mg24Y3e5 wt.% Niex wt.% GR (x ¼ 0e10) composites were prepared by mechanical ball-milling. The phase composition, microstructure and hydrogen storage properties were characterized by X-ray powder diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and Sievert's testing method. Results show that adding an appropriate amount of GR has a beneficial effect on reducing particle size and promoting ball-milling efficiency. Sample with 3 wt.% GR content has a composite structure composed of amorphous and nanocrystalline, as a result, showing an excellent hydrogen absorption kinetics even at 100  C. Differential scanning calorimetry (DSC) tests were used to evaluate the dehydrogenation performance. The peak temperature was reduced from 325  C to 295  C with increasing GR content from 0 to 3 wt.%, which can be attributed to the grain refinement and contact improvement between catalytic Ni and Mg24Y3 alloy particles. However, larger flaky Mg24Y3 alloy particles were observed with a further increase of GR content, which is mainly due to the excessive GR addition reduces ball-milling efficiency. The benefit would be completely lost with the GR adding amount up to 10 wt.%. © 2017 Published by Elsevier B.V.

Keywords: Hydrogen storage Mg-based alloy Nanocrystalline Catalysis Kinetics

1. Introduction Hydrogen, a clean energy carrier, is of a great potential to be applied to fuel cell driven car and small stationary energy supply system [1,2]. However, hydrogen storage and transportation have been considered as one of the most challenging problem [3e5]. It is of particular importance to develop safe and efficient hydrogen storage system. Metal hydrides store atomic hydrogen in the bulk of the material, and attracted great attention for hydrogen storage material since they are capable of absorbing large quantities of hydrogen [6e8]. Among them, Mg-based alloys have a considerable potential as solid-state hydrogen storage material due to its high

* Corresponding author. Research Institute for Energy Equipment Materials, Hebei University of Technology, Tianjin 300130, China. ** Corresponding author. Research Institute for Energy Equipment Materials, Hebei University of Technology, Tianjin 300130, China. *** Corresponding author. E-mail addresses: [email protected] (C. Liang), [email protected] (Q. Li), [email protected] (Y. Zhang). http://dx.doi.org/10.1016/j.jallcom.2017.08.025 0925-8388/© 2017 Published by Elsevier B.V.

reversible capacity (7.6 wt.% H2 or 101 g H2/L for MgH2), high abundance and low cost [9,10]. However, the drawbacks are sluggish hydrogen absorption/desorption kinetics and high dehydrogenation temperature, which impedes its commercial applications [11e13]. Studies on Mg/MgH2 indicate that the hydrogen absorption and desorption kinetics greatly depend on the form of particles [14]. Bulk samples have limited surfaces, which results in reduced hydrogen diffusion paths and slow kinetics. Nanostructured materials with huge specific surface area can significantly promote the nucleation and grain growth in hydrogen absorption and desorption process [15,16], therefore, have better hydrogen storage properties. Ball-milling of Mg/MgH2, or with other additives, was shown to introduce a large density of crystal defects (involving cracks, dislocations and vacancies), form highly reactive surfaces, and reduce crystallite and particle size [17,18]. These features could improve hydrogen diffusion in particles, increase nucleation sites, and reduce activation energy for hydrogen absorption and desorption [19]. Ouyang et al. [20,21] prepared hydrogen storage

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

alloys by plasma milling, both of thermodynamics and kinetics were improved, which could be ascribed to the refined microstructure and reduced dehydrogenation activation energy. Alloying Mg with transition metal elements, in many cases, is beneficial for improving hydrogen storage performance [22e24]. This is because transition metals act as catalysts to break up the H2 molecule into absorbable atomic hydrogen [25]. Shahi et al. [26] reported that the dissociation energy for H2 molecule is 1.15 eV on the surface of magnesium, which required more than 400  C for dissociation. This energy barrier can be reduced to 0.56, 0.39 and 0.06 eV for the presence of Cu, Pd and Ni respectively [26,27]. Moreover, interfaces and crystal defects resulting from introduction of other elements also can accelerate the hydrogen diffusion rate [28]. Mg and Ni are able to form intermetallic Mg2Ni, which can absorb approximately 3.6 wt.% hydrogen [29,30]. The addition of Ni has been suggested to be the great choice for improving hydrogen storage properties of Mg-based materials [31,32]. Apart from transition metals, some carbonaceous additives also have been studied to improve hydrogen storage properties of Mg/ MgH2. These carbon materials include activated carbon [33,34], carbon nanotubes (CNTs) [35,36], graphene [37,38], carbon black [39,40], graphite [41,42], etc. Lototskyy et al. [43] prepared Mg/ MgH2ecarbon composites by high energy reactive ball-milling, the research results shown that the addition of carbon significantly changes the hydrogen absorption and desorption kinetics. Ranjbar et al. [44] reported that the particle size of MgH2 þ 10 wt.% Ti0.4Mn0.22Cr0.1V0.28 composites can be markedly reduced by adding 5 wt.% carbon nanotubes, the onset dehydrogenation temperature is only about 245  C. This is because these carbonaceous additives facilitate the emergence of hydrogen from the phase interfaces and grain boundaries during desorption [44]. Moreover, mechanical adhesion of particles would take place during ballmilling because of high ductility of magnesium, which would have a negative impact on hydrogenation kinetics [33]. This problem could be overcome by adding carbon as anti-sticking agent [45,46]. The hydriding/dehydriding process can be catalyzed by the combination of in situ formed extremely fine CeH2/CeH2.73 and Ni to Mg/MgH2, the formed nano-composite structure can effectively suppress Mg/MgH2 grain growth and enable the material to maintain its high performance for more than 500 hydrogenation dehydrogenation cycles [47]. Our previous paper reported that the as-cast MgeY binary alloy also exhibits good hydrogen storage kinetics and excellent cyclic characteristics [48]. However, the hydrogen desorption temperature is still too high (the dehydrogenation peak temperature is 385  C at the heating rate of 5  C/min) [48]. The hydrogen storage performance remains to be further improved. According to the above mentioned studies, one can conclude that both of transition metals and carbon allotropes can improve the hydrogen storage properties of Mg-based materials. In the present study, Mg24Y3e5 wt.% Niex wt.% GR (x ¼ 0e10) composites were synthesized by ball-milling. The structures and hydrogen storage properties of these composites were studied in detail. Aim of the present work is to investigate the effect of GR content on microstructure and hydrogen storage properties of ballmilled Mg-based composites. 2. Experimental The starting Mg24Y3 alloy powders were prepared by vacuum induction melting. Details on the preparation and characterization of the alloy were described in previous paper [48]. Other raw materials used in the present study included carbonyl nickel powders (99.99% purity) and GR powders (99.95% purity), all of which were purchased from Sinopharm Chemical Reagent Beijing Co., Ltd. The

499

ball-milling process was performed in a QM-3SP2 planetary ballmill, using stainless steel pots and balls. Each experiment involved 5.000 g of Mg24Y3 alloy powder and a weight fraction of 5 wt.% Ni powder. Then 0, 1, 3, 5, 10 wt.% of GR powders were added in the mixture, therefore, the chemical compositions were Mg24Y3e5 wt.% Niex wt.% GR (x ¼ 0e10). For the sake of easy expression, the composites with different GR quantity are designated as 0 wt.% GR, 1 wt.% GR, 3 wt.% GR, 5 wt.% GR and 10 wt.% GR, respectively. High purity argon was used as protective atmosphere and the pressure within the pots was kept about 0.15e0.2 MPa to prevent oxidation of the samples. The ball to powder weight ratio was 40:1, and the rotational speed was 350 rpm. After milling for 20 h, the milling pots were opened in a glove box and all the powder samples were handled in high purity argon. Structural properties and phase compositions of the samples were tested by XRD (Bruker D8 Advance powder diffractometer with Cu-Ka radiation). The tests were performed at a scanning rate of 2 /min in the 2q range from 10 to 90 . The morphologies of the sample particles were characterized by a SEM (FEI Quanta-400), and the high resolution images and electron diffraction (ED) patterns were performed by a TEM (FEI Tecnai G2 F20). For TEM observations, the sample powders were dispersed in ethanol with ultrasonic dispersion instrument and then dropped the suspension on a holey carbon film supported by a copper grid. Hydrogen absorption and desorption properties were tested by using a conventional Sievert's-type apparatus. About 0.500 g samples were placed in a stainless steel sample chamber. All the isothermal hydrogen absorption measurements were carried out under an initial hydrogen pressure of 3.0 MPa with 99.995 vol.% purity hydrogen. For the hydrogen desorption kinetics, the onset dehydrogenation pressure was about 0.01 MPa and ending pressure was below 0.03 MPa. In view of microstructural changes are expected to occur during the previous hydriding/dehydriding cycles, and leading to a variation of the kinetics [48], the samples have gone through 5 full activation cycles at 350  C and the kinetics almost do not vary with the increase of cycle numbers. Then the hydrogen absorption and desorption properties were performed at various temperatures. After each dehydrogenation test, the temperature was reset at 350  C, and the sample chamber was evacuated below 0.0001 MPa for 30 min to remove any remaining hydrogen in samples. Then the subsequent hydrogenation and dehydrogenation measurements were carried out at target temperatures. Hydrogen desorption behavior of the composites was also studied by using DSC (NETZSCH STA 449F3). Before the tests, the ball-mailed composites were also subjected to 5 activation cycles at 350  C, and the samples were fully hydrogenated and cooled to room temperature. Then the samples were heated from room temperature to 450  C at the heating rate of 5  C/min. The measurements were performed under high purity argon flow with a flow rate of 50 ml/min. 3. Results and discussion 3.1. Structural and morphological characterizations The XRD patterns of the ball-milled composites are presented in Fig. 1. Apparently, increasing GR content leads to apparent changes in diffraction peaks, which indicates that the crystal structure of the composites has been affected by the introduction of GR. Broad and short diffraction peaks can be observed when the GR content is below 3 wt.%, which is very different from the sharp and intense diffraction peaks of the as-cast Mg24Y3 alloy [48]. The broad peaks are considered to be associated with the formation of crystal defects, micro-stresses and the reduced crystallite and particle size

500

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

Fig. 1. XRD patterns of the ball-milled composites.

caused by the ball-milling [49]. Diffraction peaks corresponding to GR are invisible when the GR content is below 3 wt.%, which is most likely due to the crystalline structure of the GR has been destroyed during ball-milling, leading to amorphization [50]. As to the case of 5e10 wt.% GR samples, the diffraction peaks at 2q ¼ 26.5 belongs to GR are clearly detected due to higher amounts of GR. Moreover, diffraction peaks corresponding to Ni can be clearly observed in the ball-milled composites, implying that the added Ni exists as a simple substance. No new carbide or Ni-based crystalline phases were observed, showing that the ball-milled composites are a mechanical mixture. Furthermore, the broad and short peaks corresponding to Mg24Y5 phase tend to be much sharper and more intense when the GR content reaching to 10 wt.%, indicating that too much GR content reduces ball-milling efficiency. SEM images of the ball-milled composites are presented in Fig. 2. Particles with distinctly different sizes and morphologies of the different composites can be clearly observed. As is shown in Fig. 2(a), the 0 wt.% GR sample contains irregularly shaped particles with the size range from 20 to 50 mm. Most of them are adhered and agglomerated into large cluster under the effect of ball-milling. For the 3 wt.% GR composite, the particle size significantly reduced and also more uniform. The size of the particles are about 5 mm, which is much smaller than that of 0 wt.% GR sample. The results clearly suggest the beneficial effect of GR in reducing the particle size of

the composites during ball-milling, and more importantly, promoting milling efficiency. It means that doping a proper content of GR can make the Ni and GR additives dispersed more homogeneously in Mg24Y3 master alloy. This phenomenon can be ascribed to the lubricant effect of the GR [44,51]. However, the benefit will be lost with the further increase of GR content. It can be seen from Fig. 2(c) that 10 wt.% GR composite is characterized by the large flaky particle morphology. These flakes exhibit smooth and uniform surfaces, with an average thickness of about 5 mm. The formation of such flaky particles caused by ball-milling in the presence of GR can be attributed to the “forging” of the individual Mg24Y3 alloy particle [43]. It is generally known that the GR is a soft solid material, and can be used as an excellent solid lubricant. Because of the existence of excessive lubricating graphite, it is more likely to occur slipping rather than mechanical grinding and alloying between raw material particles during ball-milling. The excessive graphite addition also reduces the cold welding and fracturing during ball-milling, which helps to preserve the crystal structure of Mg24Y3 alloy particles. Above results are in agreement well with the XRD measurements (Fig. 1). The large flaky particles of the ballmilled 10 wt.% GR composite doubtlessly have an inhomogeneous microstructure due to the fact that the Ni and GR additives can only attach themselves to the surfaces of these flakes. TEM has been used to further clarify the details on the morphological studies. Fig. 3 shows the TEM images and ED patterns of the ball-milled composites. It is clear that the 0 wt.% GR composite has a nanocrystalline structure, with the average grain size about 10 nm. As for the 3 wt.% GR composite, it shows a complex structure with nano-sized particles embedded in amorphous matrix. Therefore, it can be inferred that 3 wt.% of GR addition can increase the amorphous tendency of the Mg24Y3 alloy during ball-milling. Nano-graphite also can be observed in this sample, indicating that ball-milling results in a pulverization and even a partial amorphization of GR powders. This result is in conformity with the disappearance of the diffraction peak corresponding to GR in 3 wt.% GR sample described in Fig. 1. For the case of the 10 wt.% GR sample, the GR particles was also decomposed into thicker nano-sheets under the impact of ball-milling. Decomposing and slipping of the GR powders significantly reduces the milling efficiency, leading to the formation of flaky alloy particles shown in Fig. 2(c). Bright and sharp ED patterns indicate that the 10 wt.% GR sample is still a crystal structure. XRD patterns of the hydrogenated and dehydrogenated composites are presented in Fig. 4. It can be seen from Fig. 4(a) that the diffraction peaks corresponding to MgH2, YH2, YH3 and Mg2NiH4 can be observed, indicating that the Mg24Y5 and Mg phases in the alloy have completely changed into metal hydrides. The broad and

Fig. 2. SEM images of the ball-milled composites: (a) 0 wt.% GR; (b) 3 wt.% GR; (c) 10 wt.% GR.

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

501

Fig. 3. TEM images and ED patterns of the as-prepared composites: (a) 0 wt.% GR; (b) 3 wt.% GR; (c) 10 wt.% GR.

Fig. 4. XRD patterns of the full hydrogenated composites.

short XRD peaks tend to be much sharper and more intense, which is probably attributable to the elimination of microstrains and growth of crystals during hydrogenation [40,52]. Moreover, the

existence state of GR has not been changed after hydrogenation, meaning that the GR plays the role of chemically inert substance during hydrogen absorption and desorption reaction. The hydrogenation process of Mg24Y5 and Mg has been reported in previous work [48], now we will discuss the phase evolution of Ni during hydrogen absorption. It can be seen from Fig. 4 that the Mg2NiH4 was formed for all the hydrogenated samples, implying that the Ni reacted with Mg during hydrogenation and produced Mg2NiH4. This reaction can be described as: Ni þ 2 Mg þ 2H2 / Mg2NiH4. However, a very weak peak corresponding to Ni still exist in XRD patterns except the 3 wt.% GR sample, which means that this phase transformation is an incomplete reaction. We believe that this phenomenon associates with the distribution of Ni in the alloy particles. It can be seen from Fig. 2 that the ball-milled 3 wt.% GR composites have much less particle size and more homogeneous mix. No doubt that the decreased particle size and uniformly distributed Ni should facilitate the diffusion of Ni and Mg atoms [53], thus, making the Ni more likely to react with Mg and forming Mg2NiH4. For the 10 wt.% GR composite, it has an extremely limited contact area of Ni and Mg. Diffusion between atomic Ni and Mg is relatively more difficult, making Ni has not completely formed into Mg2NiH4. Therefore, the diffraction peaks of unreacted Ni can be clearly observed in the XRD patterns of full hydrogenated composites. The XRD patterns of the dehydrogenated composites are shown in Fig. 4(b), from which it can be seen that the dehydrogenated composites consist of Mg, YH2 and Mg2Ni phase. The disappearances of MgH2, YH3 and Mg2NiH4 indicate completely decomposition of these phases. It is reported that Mg2NiH4 can play a catalytic role for the dehydrogenation of MgH2 due to the fact that Mg2NiH4 has less thermal stability and lower dehydrogenation temperature than MgH2 [54,55]. Moreover, the hydrogen absorption and desorption kinetics of Mg-based alloy can be improved by Y addition, which is mainly attributed to the catalytic role of YH2/ YH3 phase and changed microstructures [48,56]. The lattice strain caused by YH2 4 YH3 transformation provides the driving force to trigger the hydrogen absorption/desorption of MgH2 [57], therefore promoting the sorption kinetics of the alloy. Fig. 5 shows TEM images and ED patterns of the full hydrogenated composites. Clearly, the hydrides are typical nano-crystalline structure. High resolution images show that the in situ formed YH2 nanoparticles embedded in the MgH2 matrix. Bright ED patterns show that all the samples are crystalline structure. The grain size of 3 wt.% GR sample is smaller than that of 0 and 10 wt.% GR samples, which is about 5e10 nm for YHx and 10e20 nm for MgH2. As for the 0 and 10 wt.% GR samples, the crystal size is from 20 to 50 nm. Clearly, the coexisted nanoparticles and amorphous structure of the

502

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

Fig. 5. TEM images and ED patterns of the full hydrogenated composites: (a) 0 wt.% GR; (b) 3 wt.% GR; (c) 10 wt.% GR.

ball-milled 3 wt.% GR composite transformed into a complete nanocrystalline during hydrogenation, which can be attributed to the nucleation and growth of hydride grains [58]. The result is consistent well with the sharper and intense XRD peaks shown in Fig. 4. These formed crystal defects and phase interfaces not only play the role of passage to enhance hydrogen diffusivity in the particles but also reduce the activation energy for hydrogenation and dehydrogenation process [59]. Nano-GR sheets attached to the particle surface can be clearly observed in 10 wt.% GR samples. By comparison, Mg2NiH4 cannot be readily distinguished in the hydrogenated composites perhaps due to their low content. 3.2. Isothermal absorption and desorption kinetics Isothermal hydrogen absorption and desorption curves were measured to evaluate the kinetic performance of the composites. Fig. 6 shows the hydrogenation kinetic curves of the composites at different temperatures. Hydriding rate of the samples were found to be similar at a higher temperature (350  C), almost reaching saturation within 20 min. Their maximum hydrogen absorption capacity decreases from 5.26 to 4.53 wt.% with the increase of GR content. It is commonly thought that GR does not absorb hydrogen to form hydrides under the current experimental condition, so the overall hydrogen absorption capacity declines with increasing GR content. However, as can be noted from Fig. 6(b and c) that the hydrogen absorption capacity and kinetics of the composites are very different from each other. When absorption temperature goes down to 100  C, fastest absorption kinetics were observed for the 3 wt.% GR samples, taking only 30 s to nearly complete the hydriding process. Compared with Mg3RE, Mg3RENi0.1 and MgeNieRE alloys, the hydrogen storage capacity and absorption kinetics were significantly improved [60e62]. The fast kinetics should be related to the large amount of defects, smaller crystallite size as well as the catalytic effect of Ni produced by mechanical milling. However, further increasing GR content would lead to a quite negative influence on absorption rate. The hydrogenation kinetics can be characterized by its hydrogen absorption saturation ratio (R1min), which is defined as R1min ¼ C1min/Cmax,350  C, where C1min is the hydrogen absorption capacity at 1 min, and Cmax,350  C is the hydrogen absorption capacity at 350  C within 20 min. The R1min values of the composites

at 350, 260 and 100  C are shown in Fig. 6(d). Apparently, the adding amount of GR has a significant impact on the hydrogenation kinetics. At a higher temperature (350  C), the R1min value declines from 78.6% to 68.6% with the GR content increases from 0 to 3 wt.%, and then rises to 87.5% when the GR amount further increases to 10 wt.%. However, the absorption kinetics at lower temperature (260 and 100  C) reveals a remarkably different behavior. The R1min values first increase and then decrease with the increase of GR content. As we all known that the hydriding process is an exothermic reaction, higher temperature is harm for hydrogen absorption process, so the decrease of temperature facilitates the hydrogen absorption process. However, the hydrogen diffusion rate in particles decreases with the increase of temperature, which shows an unfavorable effect [63,64]. That is to say, the hydrogen absorption kinetics is controlled by the above two opposite factors. It can be seen from Figs. 2 and 3 that the 3 wt.% GR sample has the smallest particle size and amorphous/nanocrystalline structure. The interfaces and defects provide a preferential hydrogen diffusion path that permits the combination of atomic H and Mg to become the rate limiting process, that is, hydrogen diffusion is not the main influencing factors of hydrogenation kinetics [65]. Therefore, the R1min values of the 3 wt.% GR sample increases from 68.6% to 97.3% when hydrogenation temperature decreases from 350  C to 100  C. As for the case of 10 wt.% GR sample, the microstructure were not essentially changed during ball-milling, and still have larger particle size. The kinetic performance is mainly controlled by hydrogen diffusion. Moreover, Ni particles distributed on the large flaky Mg24Y3 alloy surfaces have little catalytic effects on hydrogenation. For these reasons, the hydrogenation rate of 10 wt.% GR sample becomes slow with the temperature decreasing. Hydrogen desorption kinetics curves of the composites were also investigated, as shown in Fig. 7. Similarly, the content of GR has a significant impact on hydrogen desorption capacity and kinetics. Sample with 3 wt.% amount of GR still have the best dehydrogenation kinetics, insufficient or excessive GR addition all reduces the dehydrogenation rate. The hydrogen desorption kinetics strongly depends on the temperature. As shown in Fig. 7, the 3 wt.% GR composite could release about 5 wt.% hydrogen within 4 min at 350  C. When the temperature drops to 260  C, the composite still could release the same amount of hydrogen within 80 min. The 10 wt.% GR sample, by contrast, has a really slow dehydrogenation

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

503

Fig. 7. Isothermal hydrogen desorption kinetic curves of the composites at different temperatures: (a) 350  C; (b) 260  C.

kinetics. Note that the desorption rate of 10 wt.% GR composite seems faster before the hydrogen content reaches around 1 wt.%, but it becomes very sluggish afterwards. This could be because Ni particles distributed in Mg24Y3 alloy particle surfaces only have catalytic effect on those areas in superficial layer. Dehydrogenation process in inside of alloy particles is still very slow. 3.3. Dehydrogenation behavior To further analyse the effect of GR content on dehydrogenation for the hydrogenated composites, the thermal properties were characterized by using DSC, and the traces are presented in Fig. 8. Obviously, two endothermic peaks can be observed. According to the phase composition of the hydrogenated composites described in Fig. 4, we can confirm that these endothermic peaks mainly attributed to the decomposition of MgH2 because of high thermal stability for YH2 and small amount of Mg2NiH4 and YH3. It is undeniable that the huge difference of dehydrogenation peak temperature between these composites should be related to the

Fig. 6. Isothermal hydrogen absorption kinetic curves of the composites at different temperatures: (a) 350  C; (b) 260  C; (c) 100  C; (d) hydrogenation saturation ratio after 1 min.

504

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

temperature is reduced from 325  C to 295  C with increasing GR content from 0 wt.% to 3 wt.%, suggesting that the addition of GR has beneficial effect on dehydrogenation. This can be attributed to the refinement of particle size and contact improvement between Mg24Y3 alloy particles and catalytic Ni [17,67]. Moreover, nanocrystalline structure produced by ball-milling provides quite a lot of defects and interfaces, promoting hydrogen diffusion and accelerating nucleation in dehydrogenation process, due to the extra interfacial free energy stored in the boundary [68,69]. As a result, the 3 wt.% GR sample shows the best hydrogen desorption performance. Similarly, further increase of the GR content does not produce further benefit to dehydrogenation, leading to the intensity degradation of lower temperature peaks and intensity increasing of higher temperature peaks. When the mass fraction of GR increased to 10 wt.%, the endothermic peak at lower temperature (305  C) almost disappeared, instead the enhancement of the higher temperature peak (382  C). That is, the desorption properties get worse with increasing GR content, this phenomenon is also consistent well with the hydrogen desorption kinetics shown in Fig. 7. In regard to the reduction of the hydrogen desorption kinetics and increase of dehydrogenation temperature caused by excessive GR addition, there are two potential explanations: (1) excessive GR addition reduces ball-milling efficiency, the number of interfaces and defects in alloy particles has not grown significantly, making nucleation step and hydrogen diffusion more difficult, and (2) the catalytic Ni and GR additives only distribute in the surfaces of Mg24Y3 alloy particles and are difficult to enter into inside of the alloy particles, the adhesion between Ni and Mg24Y3 alloy particles is quite poor, reducing catalytic efficiency of Ni. Above reasons make the hydrogen desorption properties of ball-milled 10 wt.% GR composite are no better than the as-cast Mg24Y3 alloy [48]. Fig. 8. DSC desorption curves for the hydrogenated composites.

4. Conclusions microstructure and distribution of Ni. During hydrogen desorption process, Mg2NiH4 first desorbs hydrogen and undergoes a volume contraction, causing a contraction strain on MgH2 around it, hence, facilitating dehydrogenation of MgH2 [30,66]. That is, the formed Mg2Ni/Mg2NiH4 can act as catalytic sites for the composites, reducing dehydrogenation temperature. Moreover, previous study results shown that the hydrogen desorption peak temperature of the as-cast Mg24Y3 alloy was about 385  C in the same heating rate [48], which is consistent with the higher peak temperatures in the present ball-milled composites. Therefore, we can conclude that the lower endothermic peak temperatures, from 295 to 325  C, are attributed to the dehydrogenation of MgH2 under the catalytic effect of Mg2Ni/Mg2NiH4 (catalytic active area shown in Fig. 9), and the higher temperature ones, around 390  C, are regarded as the dehydrogenation of MgH2 which far away from Mg2Ni/Mg2NiH4 catalytic sites (ordinary inert area shown in Fig. 9). The GR content also has an obvious effect on lower peak temperatures. It can be seen from Fig. 8 that the dehydrogenation peak

The effect of GR content on structural and hydrogen storage properties of ball-milled Mg24Y3e5 wt.% Niex wt.% GR (x ¼ 0e10) composites were investigated. Appropriate amount of GR addition can enhance ball-milling efficiency, reduce particle size, and promote forming ability of amorphous and nanocrystalline. The optimal hydrogen storage properties could be achieved by ballmilling of Mg24Y3e5 wt.% Ni with 3 wt.% GR, which can absorb about 5 wt% hydrogen within 30 s at 100  C and desorb the same amount of hydrogen within 80 min at 260  C, respectively. The dehydrogenation peak temperature also reduces from 325  C to 295  C by adding 3 wt.% of GR. The explanation for the improved hydrogen storage behavior can be ascribed to the refinement of particle and grain size as well as better contact between the catalytic Ni and Mg24Y3 alloy particles. However, further increasing adding amount of GR leads to the reduction of ball-milling efficiency and a serious degradation of the hydrogen absorption and desorption properties. When the mass fraction of GR increases up to 10 wt.%, there is only one endothermic peak at 382  C, and the

Fig. 9. Schematic diagram of the catalytic active area and ordinary inert area caused by different GR addition.

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506

hydrogen absorption and desorption performance is nearly as bad as as-cast Mg24Y3 alloy. Acknowledgements This work was financially supported by the National Natural Science Foundations of China (Nos. 51371094 and 51471054) and Natural Science Foundation of Inner Mongolia, China (2015MS0558).

[23]

[24]

[25]

[26]

References [27] [1] M.V. Lototskyy, M.W. Davids, I. Tolj, Y.V. Klochko, B.S. Sekhar, S. Chidziva, F. Smith, D. Swanepoel, B.G. Pollet, Metal hydride systems for hydrogen storage and supply for stationary and automotive low temperature PEM fuel cell power modules, Int. J. Hydrogen Energy 40 (2015) 11491e11497. [2] H. Wang, H.J. Lin, W.T. Cai, L.Z. Ouyang, M. Zhu, Tuning kinetics and thermodynamics of hydrogen storage in light metal element based systems e a review of recent progress, J. Alloys Compd. 658 (2016) 280e300. [3] X.B. Yu, Z.W. Tang, D.L. Sun, L.Z. Ouyang, M. Zhu, Recent advances and remaining challenges of nanostructured materials for hydrogen storage applications, Prog. Mater. Sci. 88 (2017) 1e48. [4] M.L. Ma, R.M. Duan, L.Z. Ouyang, X.K. Zhu, Z.L. Chen, C.H. Peng, M. Zhu, Hydrogen storage and hydrogen generation properties of CaMg2-based alloys, J. Alloys Compd. 691 (2017) 929e935. [5] L.Z. Ouyang, J.J. Tang, Y.J. Zhao, H. Wang, X.D. Yao, J.W. Liu, J. Zou, M. Zhu, Express penetration of hydrogen on Mg(10T3) along the close-packed-planes, Sci. Rep. 5 (2015) 10776. [6] N.A.A. Rusman, M. Dahari, A review on the current progress of metal hydrides material for solid-state hydrogen storage applications, Int. J. Hydrogen Energy 41 (2016) 12108e12126. [7] L.Z. Ouyang, W. Chen, J.W. Liu, M. Felderhoff, H. Wang, M. Zhu, Enhancing the regeneration process of consumed NaBH4 for hydrogen storage, Adv. Energy Mater. 7 (2017) 1700299. [8] M. Anika, A.B. Aybara, N. Küçükdevecia, H. Erkena, B. Baksana, H. Gas¸anb, €kçü, Synthesis of La2Ni7 hydrogen storage alloy by the N.B. Hatirnaz, E. Lo electro-deoxidation technique, Int. J. Hydrogen Energy 40 (2015) 2248e2254. [9] T. Sadhasivam, H.T. Kim, S. Jung, S.H. Roh, J.H. Park, H.Y. Jung, Dimensional effects of nanostructured Mg/MgH2 for hydrogen storage applications: a review, Renew. Sustain. Energy Rev. 72 (2017) 523e534. [10] G.L. Xia, Y.B. Tan, X.W. Chen, D.L. Sun, Z.P. Guo, H.K. Liu, L.Z. Ouyang, M. Zhu, X.B. Yu, Monodisperse magnesium hydride nanoparticles uniformly selfassembled on graphene, Adv. Mater. 27 (2015) 5981e5988. [11] L.Z. Ouyang, Z.J. Cao, H. Wang, R.Z. Hu, M. Zhu, Application of dielectric barrier discharge plasma-assisted milling in energy storage materials e a review, J. Alloys Compd. 691 (2017) 422e435. [12] M. Anik, F. Karanfil, N. Küçükdeveci, Development of the high performance magnesium based hydrogen storage alloy, Int. J. Hydrogen Energy 37 (2012) 299e308. [13] H. Zhong, H. Wang, J.W. Liu, D.L. Sun, F. Fang, Q.A. Zhang, L.Z. Ouyang, M. Zhu, Enhanced hydrolysis properties and energy efficiency of MgH2-base hydrides, J. Alloys Compd. 680 (2016) 419e426. [14] H.J. Lin, C. Zhang, H. Wang, L.Z. Ouyang, Y.F. Zhu, L.Q. Li, W.H. Wang, M. Zhu, Controlling nanocrystallization and hydrogen storage property of Mg-based amorphous alloy via a gas-solid reaction, J. Alloys Compd. 685 (2016) 272e277. [15] B. Zhang, Y.J. Lv, J.G. Yuan, Y. Wu, Effects of microstructure on the hydrogen storage properties of the melt-spun Mg-5Ni-3La (at.%) alloys, J. Alloys Compd. 702 (2017) 126e131. [16] S. Kumar, A. Singh, G.P. Tiwari, Y. Kojima, V. Kain, Thermodynamics and kinetics of nano-engineered Mg-MgH2 system for reversible hydrogen storage application, Thermochim. Acta 652 (2017) 103e108. [17] L. Popilevsky, V.M. Skripnyuk, M. Beregovsky, M. Sezen, Y. Amouyala, E. Rabkin, Hydrogen storage and thermal transport properties of pelletized porous Mg-2 wt.% multiwall carbon nanotubes and Mg-2 wt.% graphite composites, Int. J. Hydrogen Energy 41 (2016) 14461e14474. [18] L.S. Xie, J.S. Li, T.B. Zhang, L. Song, H.C. Kou, Microstructure and hydrogen storage properties of Mg-Ni-Ce alloys with a long-period stacking ordered phase, J. Power Sources 338 (2017) 91e102. [19] M. Lototskyy, M.W. Davids, J.M. Sibanyoni, J. Goh, B.G. Pollet, Magnesiumbased hydrogen storage nanomaterials prepared by high energy reactive ball milling in hydrogen at the presence of mixed titaniumeiron oxide, J. Alloys Compd. 645 (2015) S454eS459. [20] L.Z. Ouyang, Z.J. Cao, H. Wang, J.W. Liu, D.L. Sun, Q.A. Zhang, M. Zhu, Enhanced dehydriding thermodynamics and kinetics in Mg(In)eMgF2 composite directly synthesized by plasma milling, J. Alloys Compd. 586 (2014) 113e117. [21] L.Z. Ouyang, Z.J. Cao, L.L. Li, H. Wang, J.W. Liu, D. Min, Y.W. Chen, F.M. Xiao, R.H. Tang, M. Zhu, Enhanced high-rate discharge properties of La11.3Mg6.0Sm7.4Ni61.0Co7.2Al7.1 with added graphene synthesized by plasma milling, Int. J. Hydrogen Energy 39 (2014) 12765e12772. [22] C. Lu, J.X. Zou, X.Y. Shi, X.Q. Zeng, W.J. Ding, Synthesis and hydrogen storage

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45] [46]

[47]

[48]

[49]

505

properties of coreeshell structured binary [email protected] and ternary [email protected]@Ni composites, Int. J. Hydrogen Energy 42 (2017) 2239e2247. D.F. Wu, L.Z. Ouyang, C. Wu, Q.F. Gu, H. Wang, J.W. Liu, M. Zhu, Phase transition and hydrogen storage properties of Mg17Ba2 compound, J. Alloys Compd. 690 (2017) 519e522. X.J. Yang, L.L. Li, W.L. Sang, J.L. Zhao, X.X. Wang, C. Yu, X.H. Zhang, C.C. Tang, Boron nitride supported Ni nanoparticles as catalysts for hydrogen generation from hydrolysis of ammonia borane, J. Alloys Compd. 693 (2017) 642e649. D. Pukazhselvan, G. Capurso, A. Maddalena, S.L. Russo, D.P. Fagg, Hydrogen storage characteristics of magnesium impregnated on the porous channels of activated charcoal scaffold, Int. J. Hydrogen Energy 39 (2014) 20045e20053. R.R. Shahi, A. Bhatanagar, S.K. Pandey, V. Shukla, T.P. Yadav, M.A. Shaz, O.N. Srivastava, MgH2eZrFe2Hx nanocomposites for improved hydrogen storage characteristics of MgH2, Int. J. Hydrogen Energy 40 (2015) 11506e11513. , Hydrogen dissociation and diffusion on transition metal (¼ M. Pozzo, D. Alfe Ti, Zr, V, Fe, Ru, Co, Rh, Ni, Pd, Cu, Ag)-doped Mg(0001) surfaces, Int. J. Hydrogen Energy 34 (2009) 1922e1930. A.R. Yavari, J.F.R. de Castro, G. Vaughan, G. Heunen, Structural evolution and metastable phase detection in MgH2e5%NbH nanocomposite during in-situ H-desorption in a synchrotron beam, J. Alloys Compd. 353 (2003) 246e251. S. Aminorroaya, A. Ranjbar, Y.H. Cho, H.K. Liu, A.K. Dahle, Hydrogen storage properties of Mg-10 wt% Ni alloy co-catalysed with niobium and multi-walled carbon nanotubes, Int. J. Hydrogen Energy 36 (2011) 571e579. €ntzsch, C. Baehtz, B. Kieback, Hydrogen desorption kiS. Kalinichenka, L. Ro netics of melt-spun and hydrogenated Mg90Ni10 and Mg80Ni10Y10 using in situ synchrotron, X-ray diffraction and thermogravimetry, J. Alloys Compd. 496 (2010) 608e613. T. Fujimoto, S. Ogawa, T. Kanai, N. Uchiyama, T. Yoshida, S. Yagi, Hydrogen storage property of materials composed of Mg nanoparticles and Ni nanoparticles fabricated by gas evaporation method, Int. J. Hydrogen Energy 40 (2015) 11890e11894. L.Z. Ouyang, Z.J. Cao, H. Wang, J.W. Liu, D.L. Sun, Q.A. Zhang, M. Zhu, Dualtuning effect of in on the thermodynamic and kinetic properties of Mg2Ni dehydrogenation, Int. J. Hydrogen Energy 38 (2013) 8881e8887. A.D. Rud, A.M. Lakhnik, Effect of carbon allotropes on the structure and hydrogen sorption during reactive ball-milling of MgeC powder mixtures, Int. J. Hydrogen Energy 37 (2012) 4179e4187. C.Z. Wu, P. Wang, X. Yao, C. Liu, D.M. Chen, G.Q. Lu, H.M. Cheng, Effect of carbon/noncarbon addition on hydrogen storage behaviors of magnesium hydride, J. Alloys Compd. 414 (2006) 259e264. E. Ruse, S. Pevzner, I.P. Bar, R. Nadiv, V.M. Skripnyuk, E. Rabkin, O. Regev, Hydrogen storage and spillover kinetics in carbon nanotube-Mg composites, Int. J. Hydrogen Energy 41 (2016) 2814e2819. S.J. Hwang, Y.S. Chuang, Enhanced hydrogen storage properties of MgH2 cocatalyzed with zirconium oxide and single-walled carbon nanotubes, J. Alloys Compd. 664 (2016) 284e290. M.K. Singh, A. Bhatnagar, S.K. Pandey, P.C. Mishra, O.N. Srivastava, Experimental and first principle studies on hydrogen desorption behavior of graphene nanofibre catalyzed MgH2, Int. J. Hydrogen Energy 42 (2017) 960e968. J. Zhang, H. Qu, G. Wu, L.B. Song, X.F. Yu, D.W. Zhou, Remarkably enhanced dehydrogenation properties and mechanisms of MgH2 by sequential-doping of nickel and graphene, Int. J. Hydrogen Energy 41 (2016) 17433e17441. T. Spassov, Z. Zlatanova, M. Spassova, S. Todorova, Hydrogen sorption properties of ball-milled MgeC nanocomposites, Int. J. Hydrogen Energy 35 (2010) 10396e10403. Z.G. Huang, Z.P. Guo, A. Calka, D. Wexler, H.K. Liu, Effects of carbon black, graphite and carbon nanotube additives on hydrogen storage properties of magnesium, J. Alloys Compd. 427 (2007) 94e100. A. Kubota, H. Miyaoka, M. Tsubota, K. Shimoda, T. Ichikawa, Y. Kojima, Synthesis and characterization of magnesiumecarbon compounds for hydrogen storage, Carbon 56 (2013) 50e55. H. Imamura, M. Kusuhara, S. Minami, M. Matsumoto, K. Masanari, Y. Sakata, K. Itoh, T. Fukunag, Carbon nanocomposites synthesized by high-energy mechanical milling of graphite and magnesium for hydrogen storage, Acta Mater. 51 (2003) 6407e6414. M. Lototskyy, J.M. Sibanyoni, R.V. Denys, M. Williams, B.G. Pollet, V.A. Yartys, Magnesiumecarbon hydrogen storage hybrid materials produced by reactive ball milling in hydrogen, Carbon 57 (2013) 146e160. A. Ranjbar, M. Ismail, Z.P. Guo, X.B. Yu, H.K. Liu, Effects of CNTs on the hydrogen storage properties of MgH2 and MgH2-BCC composite, Int. J. Hydrogen Energy 35 (2010) 7821e7826. C.Z. Wu, H.M. Cheng, Effects of carbon on hydrogen storage performances of hydrides, J. Mater. Chem. 20 (2010) 5390e5400. A.D. Rud, A.M. Lakhnik, V.G. Ivanchenko, V.N. Uvarov, A.A. Shkola, V.A. Dekhtyarenko, L.I. Ivaschuk, N.I. Kuskov, Hydrogen storage of the MgeC composites, Int. J. Hydrogen Energy 33 (2008) 1310e1316. L.Z. Ouyang, X.S. Yang, M. Zhu, J.W. Liu, H.W. Dong, D.L. Sun, J. Zou, X.D. Yao, Enhanced hydrogen storage kinetics and stability by synergistic effects of in situ formed CeH2.73 and Ni in CeH2.73-MgH2-Ni nanocomposites, J. Phys. Chem. C 118 (2014) 7808e7820. T. Yang, Z.M. Yuan, W.G. Bu, Z.C. Jia, Y. Qi, Y.H. Zhang, Evolution of the phase structure and hydrogen storage thermodynamics and kinetics of Mg88Y12 binary alloy, Int. J. Hydrogen Energy 41 (2016) 2689e2699. Y.H. Cho, S. Aminorroaya, H.K. Liu, A.K. Dahle, The effect of transition metals

506

[50] [51]

[52]

[53]

[54]

[55]

[56] [57]

[58]

T. Yang et al. / Journal of Alloys and Compounds 726 (2017) 498e506 on hydrogen migration and catalysis in cast MgeNi alloys, Int. J. Hydrogen Energy 36 (2011) 4984e4992. C.X. Shang, Z.X. Guo, Effect of carbon on hydrogen desorption and absorption of mechanically milled MgH2, J. Power Sources 129 (2004) 73e80. S. Aminorroaya, H. Liu, Y. Cho, A. Dahle, Microstructure and activation characteristics of MgeNi alloy modified by multi-walled carbon nanotubes, Int. J. Hydrogen Energy 35 (2010) 4144e4153. denas, Z.X. Guo, K.F. Aguey-Zinsou, D. Cazorla-Amoro  s, M.A. Lillo-Ro A. Linares-Solano, Effects of different carbon materials on MgH2 decomposition, Carbon 46 (2008) 126e137. Y. Jia, Y.N. Guo, J. Zou, X.D. Yao, Hydrogenation/dehydrogenation in MgH2activated carbon composites prepared by ball milling, Int. J. Hydrogen Energy 37 (2012) 7579e7585. J.W. Liu, C.C. Zou, H. Wang, L.Z. Ouyang, M. Zhu, Facilitating de/hydrogenation by long-period stacking ordered structure in Mg based alloys, Int. J. Hydrogen Energy 38 (2013) 10438e10445. J.X. Zou, S. Long, X. Chen, X.Q. Zeng, W.J. Ding, Preparation and hydrogen sorption properties of a Ni decorated Mg based [email protected] nano-composite, Int. J. Hydrogen Energy 40 (2015) 1820e1828. L.Z. Ouyang, H.W. Dong, M. Zhu, Mg3Mm compound based hydrogen storage materials, J. Alloys Compd. 446e447 (2007) 124e128. F.P. Luo, H. Wang, L.Z. Ouyang, M.Q. Zeng, J.W. Liu, M. Zhu, Enhanced reversible hydrogen storage properties of a MgeIneY ternary solid solution, J. Hydrogen Energy 38 (2013) 10912e10918. M. Pourabdoli, S. Raygan, H. Abdizadeh, D. Uner, A comparative study for synthesis methods of nano-structured (9Nie2MgeY) alloy catalysts and effect of the produced alloy on hydrogen desorption properties of MgH2, Int. J. Hydrogen Energy 38 (2013) 16090e16097.

[59] M. Konarova, A. Tanksale, J.N. Beltramini, G.Q. Lu, Effects of nano-confinement on the hydrogen desorption properties of MgH2, Nano Energy 2 (2013) 98e104. [60] M. Zhu, H. Wang, L.Z. Ouyang, M.Q. Zeng, Composite structure and hydrogen storage properties in Mg-base alloys, Int. J. Hydrogen Energy 31 (2006) 251e257. [61] L.Z. Ouyang, F.X. Qin, M. Zhu, The hydrogen storage behavior of Mg3La and Mg3LaNi0.1, Scr. Mater. 55 (2006) 1075e1078. [62] L.Z. Ouyang, X.S. Yang, H.W. Dong, M. Zhu, Structure and hydrogen storage properties of Mg3Pr and Mg3PrNi0.1 alloys, Scr. Mater. 61 (2009) 339e342. [63] N. Xing, Y. Wu, W. Han, S.X. Zhou, Improved hydrogenation-dehydrogenation characteristics of nanostructured melt-spun Mg-10Ni-2Mm alloy processed by rapid solidification, Prog. Nat. Sci. Mater. 20 (2010) 49e53. [64] P. Adelhelm, P.E. de Jongh, The impact of carbon materials on the hydrogen storage properties of light metal hydrides, J. Mater. Chem. 21 (2011) 2417e2427. [65] F.J. Castro, V. Fuster, G. Urretavizcaya, Hydrogen sorption properties of a MgH2e10 wt.% graphite mixture, J. Alloys Compd. 509S (2011) S595eS598. [66] A. Zaluska, L. Zaluski, J.O. StromeOlsen, Nanocrystalline magnesium for hydrogen storage, J. Alloys Compd. 288 (1999) 217e225. [67] H. Imamura, K. Masanari, M. Kusuhara, H. Katsumoto, T. Sumi, Y. Sakata, High hydrogen storage capacity of nanosized magnesium synthesized by high energy ball-milling, J. Alloys Compd. 386 (2005) 211e216. [68] T.B. Zhang, W.J. Song, H.C. Kou, J.S. Li, Surface valence transformation during thermal activation and hydrogenation thermodynamics of MgeNieY meltspun ribbons, Appl. Surf. Sci. 371 (2016) 35e43. [69] L.Z. Ouyang, S.Y. Ye, H.W. Dong, M. Zhu, Effect of interfacial free energy on hydriding reaction of MgeNi thin films, Appl. Phys. Lett. 90 (2007) 021917.