- Email: [email protected]
Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg Subrata Panda a,b,c,*, Jean-Jacques Fundenberger a,b, Yajun Zhao a,b, Jianxin Zou a,c, Laszlo S. Toth a,b, Thierry Grosdidier a,b,** Universite de Lorraine, Laboratory of Excellence on Design of Alloy Metals for Low-mass Structures (DAMAS), Ile du Saulcy, Metz F-57045, France b Universite de Lorraine, Laboratoire d’Etude des Microstructures et de Mecanique des Materiaux (LEM3 UMR 7239), Ile du Saulcy, Metz F-57045, France c Shanghai Engineering Research Center of Mg Materials and Applications, National Engineering Research Center of Light Alloy Net Forming, Shanghai Jiao Tong University, Shanghai 200240, PR China a
article info
abstract
Article history:
While severe plastic deformation (SPD) on bulk samples has been widely applied for
Received 22 December 2016
modifying the H-sorption properties, there has been little attention towards the use of SPD
Received in revised form
on powder materials. In this context, the aim of the present work was to compare the H-
13 May 2017
storage properties of high-pressure torsion (HPT) consolidated products obtained from two
Accepted 14 May 2017
distinct Mg powder precursors: atomized micro-sized and condensed ultrafine powder
Available online xxx
particles. The results showed that the nature of the initial powder precursor had a pronounced effect on the H-sorption behavior. The HPT product obtained from the condensed
Keywords:
ultrafine powder showed faster absorption kinetics than the consolidated product obtained
HPT
from the atomized powder. However, the HPT product obtained from atomized powder
Powder consolidation
could absorb more hydrogen and showed faster desorption kinetics corresponding to a
Nanocomposites
lower activation energy. These results are discussed by taking into account the effective-
H-sorption
ness of the HPT process to refine the grain sizes and differences in the dispersion of fine
Hydrogenation/dehydrogenation
MgO oxide particles. © 2017 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
de Lorraine, Laboratory of Excellence on Design of Alloy Metals for low-mass Structures (DAMAS), Ile * Corresponding author. Universite du Saulcy, Metz F-57045, France. de Lorraine, Laboratory of Excellence on Design of Alloy Metals for low-mass Structures (DAMAS), Ile ** Corresponding author. Universite du Saulcy, Metz F-57045, France. E-mail addresses: [email protected] (S. Panda), [email protected] (J.-J. Fundenberger), [email protected] (Y. Zhao), [email protected] (J. Zou), [email protected] (L.S. Toth), [email protected] (T. Grosdidier). http://dx.doi.org/10.1016/j.ijhydene.2017.05.097 0360-3199/© 2017 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
2
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Introduction A considerable amount of research has been devoted to develop advanced H-storage media in solid-state. As a promising H carrier, magnesium has been considered as an excellent candidate for H-storage mainly due to its light weight, abundance in the earth's crust and high gravimetric storage capacity (7.6 wt% H) [1,2]. In addition, due to their high thermal energy density the Mg-based metal hydride systems are also considered as potential high temperature heat storage media [3,4]. Unfortunately, the sluggish hydriding/dehydriding kinetics, and stable thermodynamic properties [5] leading to higher processing temperatures are the main obstacles for metal-H systems for potential industrial applications. Significant improvements in the H-sorption kinetics of metal-H systems have been achieved by the use of highenergy-ball-milling (HEBM) [6e12], which were primarily considered to be the effect of nano-metric crystallite/grain size, introduction of lattice defects and increase of reactive surface area. The catalytic action of additions of transition metals or oxides was also considered to be beneficial. In particular, Barkhordarian et al. [8] showed that milling of MgH2 with 1 mol% of different oxides such as TiO2, V2O5 or Nb2O5 leads to an improvement of magnesium H-sorption kinetics. Aguey-Zinsou et al. [9] considered that the major effect of the MxOy oxides was related to the refinement of the size of MgH2 particles during milling. It was also reported that co-milling of Mg with its own MgO does not modify the structural and thermodynamic properties of MgH2 but leads to an efficient decrease of the particle size and therefore enhances the H-sorption kinetics [10]. Although the milling route results in considerable improvements in H-sorption properties, it suffers from two major drawbacks. Firstly, it is difficult to avoid agglomeration during milling of Mg or Mg alloys [11]. Secondly, impurities coming from the milling tools and gas adsorption are hard to control and can affect significantly the H-sorption kinetic properties [12]. To avoid these drawbacks, severe plastic deformation (SPD) techniques have been recently used such as equalchannel angular pressing (ECAP) and high pressure torsion (HPT) [13e18]. It was suggested from SPD on pure Mg that the refinement in grain size (i.e. large increase in amount of grain boundaries) played a more effective role than the amount of dislocations for modifying the H-storage properties [13]. Although the compaction of powders under high pressure plus torsional straining has been initiated already in the 90's [19,20], processing routes involving the consolidation of powders using SPD techniques for H-storage materials have received comparatively little attention [21e26]. Kusadome et al. [21] investigated the effect of high-pressure torsion (HPT) on MgNi2 powders and found that the intermetallic (normally accepted as non-absorber) could absorb 0.1 wt% H even at 100 C. Leiva et al. [22] have employed several SPD techniques (HPT, cold rolling and forging) on MgH2 and MgH2 e Fe powder mixtures and reported that SPD can also reduce crystallite size. Liang and Hout [24] showed that the application of cold rolling on MgH2 powders could be an efficient processing route to produce nanocrystalline structure for improving H-sorption properties.
A new method e applying HPT on arc plasma evaporated ultrafine grained powder e has been recently introduced to produce bulk Mg-MgO based nanocomposite for H storage [25,26]. The improved H-sorption properties of these bulk composites were mainly attributed to the refinement of the Mg matrix and the dispersion of fine MgO particles formed during intense shearing by HPT [25]. The effect of small amount of Fe and Ni addition was also examined [26]. This new processing route has potentially several advantages. Compared to HEBM, it is less likely to introduce contamination within the heavily deformed material. Also, in comparison with other SPD routes like, for example, ECAP, it is simpler to control and easier to apply on powder. In this context, in the present study the same processing route was employed with the aim of examining the effect of the nature of the initial powder precursors. To this end, two distinct Mg powder precursors consisting of micro-sized powder particles produced by inert gas atomization and ultrafine powder particles synthesized by an arc plasma evaporation method were employed. The idea of using atomized micro-sized powder instead of the evaporated ultrafine powder lies also in the fact that the production of atomized powder is much cheaper and can be done in more significant quantities.
Experimental Sample processing Two Mg powder precursors having different particle sizes were used in the present work for HPT processing. Commercial purity magnesium (99.8%) was gas atomized by SFM SA (Switzerland) to produce a micro-sized Mg powder with a particle size distribution in the range of 10e80 mm. Ultrafine magnesium powder was also produced by arc plasma evaporation/condensation method to generate particle sizes in the range of 50e600 nm. In the latter case, the apparatus was filled with a mixture of 0.70 atm Ar and 0.10 atm H2. The details of this procedure are available in Refs. [27,28]. In order to prevent these very fine Mg particles from burning in open air, a twostep passivation procedure was applied: maintaining the powder for 2 h in a mixture of 50 kPa Ar þ 10 kPa normal air followed by holding for 12 h in a mixture of 50 kPa Ar þ 50 kPa normal air, before opening the evaporation/condensation chamber [27,28]. Fig. 1 shows images illustrating the aspects of the different powders. The atomization process led to fairly rounded particles (see Fig. 1a) having a particle size distribution centered on ~15 mm in the micro-sized powder. Typical SEM image and TEM image of the ultrafine powder are shown in Fig. 1b. They demonstrate that the particle size distribution was essentially in the sub-micrometer range with an average particle size below 300 nm, and that each powder particle corresponded generally to a single grain. The hexagonal shape of some finer powder particles, visible in the TEM image, is due to the hexagonal structure of the Mg phase [25]. These two powders were separately processed by a twostep HPT facility. A schematic diagram of the processing route employed in this study is illustrated in Fig. 2. The
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
3
Fig. 1 e Morphological features of the initial powders: (a) SEM image of the atomized micro-sized Mg powder and (b) SEM image and a typical bright field TEM image (inset) of the condensed and passivated ultrafine Mg powder.
Fig. 2 e Schematic illustrations of a two-step HPT consolidation route employed in this study; step 1: uniaxial compression of the Mg powder into a precompacted disk (a) followed by step 2: torsional straining using HPT (b).
powders were first compacted by uniaxial compression under a hydrostatic pressure of 1.5 GPa (holding for 10 min) to form disks with a diameter of 20 mm and a thickness of 3 mm, as depicted in Fig. 2a. A subsequent torsional straining by HPT was conducted on the precompacted disks under a hydrostatic pressure of 1.2 GPa at room temperature, as portrayed in Fig. 2b. The final step was conducted under quasi-constrained conditions [21], where the material lateral flow was partially restricted. The torsional straining was imposed by rotating the lower anvil at 0.125 rpm up to 2 revolutions for a total maximum shear strain (g) of about 42 at the edge of the disks. In the following, the HPT-disks processed from the microatomized and condensed ultrafine Mg powders will be referred to as micro-HPT and nano-HPT products, respectively.
Characterization The crystallographic phases present in the initial powders, after HPT and after hydrogenation were identified by X-ray
diffraction (XRD) using a powder X-ray diffraction apparatus (D/max 2550VL/PCX) equipped with a monochromatic Cu-Ka radiation source. Microstructural characterization of the initial powder and the bulk HPT products was carried out by scanning electron microscopy (SEM) using a field emission gun scanning electron microscope (SEM e Zeiss Supra40). The H-storage properties of the HPT products were examined using a Sievert-type pressureecompositione temperature (PCT) volumetric apparatus (Type PCT-2, Shanghai Institute of Microsystem and Information Technology). Prior to the PCT and kinetic measurements, the bulk HPT products were broken into several smaller pieces (without using any file or attritor mill) to fit into the testing vessel. To characterize the effect of the HPT processing, the initial powders as well as their consolidated HPT products were activated at 400 C under a hydrogen pressure of 3.5 MPa. These one-cycled activated HPT products were subsequently cycled for the thermodynamic measurements of absorption/ desorption tests at 400 C, 375 C, 350 C and 325 C. Following this, the samples were used for the dynamic hydrogen
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
absorption measurements under a hydrogen pressure of 3.5 MPa at 200 C, 250 C, 300 C and 350 C. Then, after a final hydrogen absorption at 400 C for 6 h under a hydrogen pressure of 3.5 MPa, the dehydriding properties of the hydrogenated HPT products were investigated by simultaneous differential scanning calorimetry and thermogravimetry measurements (DSC-TG, Netzsch STA449F3 Jupiter) under 0.1 MPa of purified argon at a heating rate of 3, 5 and 10 C/min from 25 to 500 C. Except from the pressure of 3.5 MPa (instead of 4 MPa), this procedure and these cycles of hydrogen absorption/desorption are rather similar to the ones used in previous papers dealing with the hydrogen storage properties of ultrafine particles [29,30].
Results HPT-processing and sintered products Fig. 3 compares the XRD patterns of the initial powders and their HPT-processed products. The peaks of the Mg hexagonal structure are the main contributors in these XRD patterns. These patterns are, however, different for two reasons. Firstly, as already reported in Refs. [25,26], since the ultrafine powder was passivated after the evaporation/condensation process, a thin layer of MgO formed around the Mg particles. Therefore, a minor peak of MgO phase is detected for both ultrafine powder and its nano-HPT product (see Fig. 3b). In addition, also in the XRD traces of the ultrafine and nano-HPT consolidated product, some Mg(OH)2 is identified through its (101) peak. This phase e which was not previously reported for these condensed ultrafine Mg powder [27,28] e might have formed together with the MgO during the passivation process in normal air. In contrast, the atomized micro-sized powder and its consolidated HPT product did not show such kind of additional peaks (see Fig. 3a). Secondly, the difference in peak ratios indicate different kinds of texture developments in these two distinct powder precursors subjected to severe plastic deformation through HPT.
The deformed microstructures in the consolidated products have been characterized by SEM observations at the periphery and at the middle-thickness of the HPT-disks. Since a similar kind of localized deformation at the middle-section of the HPT-processed disks for bulk materials has been recently reported by Panda et al. [31] e where the materials had experienced a massive shear deformation as compared to the top and bottom surfaces, the microstructural developments at the middle-section of the HPT-disks were observed in the present study. Typical SEM images obtained under backscattered electron (BSE) imaging conditions are displayed in Fig. 4 together with a sketch showing the examined location. These images illustrate that the grains are clearly inclined at about 45 from the shear plane after HPT treatments. In the HPT product consolidated from the condensed and passivated ultrafine powder (see Fig. 4b), the fine MgO particles (as also reported by Zou et al. [25]) were introduced within the grains during the massive shear straining, and appear as white spots randomly distributed throughout the Mg matrix. It is interesting to notice that some oxide particles also formed alignments inclined at about 45 . These oxide particles have generated significant Zener pinning that has restricted the grain size evolution within fine elongated domains with an average thickness of about ~200e300 nm. Comparatively, in the oxide free micro-HPT product (see Fig. 4a), and consistently with the results reported by Edalati et al. [13], severe plastic deformation has led to the development of a dynamically recrystallized microstructure consisting of rather equiaxed grains. In our case, these grains had an average grain size of ~0.5e1 mm.
Activation kinetics of the powders and their HPT products In order to assess the effects of the SPD processing route on the H-sorption behaviors of the two distinct powder particles, the initial powders as well as the consolidated bulk products were tested for activation kinetics. Fig. 5 compares the first H absorption kinetics measured at 400 C under a hydrogen pressure of 3.5 MPa for all the samples. Note that, because of
Fig. 3 e Normalized X-ray diffraction patterns for the initial powders and their products after HPT consolidation: (a) atomized micro-sized Mg powder and its micro-HPT product and (b) condensed ultrafine Mg powder and its nano-HPT product. Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
5
Fig. 4 e SEM micrographs taken under back scattered imaging conditions of the (a) micro-HPT and (b) nano-HPT products, examined at the outer edge of the disks and at their middle thicknesses as shown in the inset.
Fig. 5 e First hydrogen absorption kinetics measured from the initial loose powders as well as their HPT consolidated products at 400 C under a hydrogen pressure of 3.5 MPa.
their more sluggish behavior, 10 h were used instead of 8 h for the runs carried out on the atomized and micro-HPT products to try to reach a more steady state plateau. It is readily apparent from the comparison between the two powders that, while the atomized powder could reach almost the theoretical value of 7.6 wt% H for pure Mg, the condensed powder showed comparatively a reduced H uptake capacity with a plateau located at about ~7.1 wt% H. It is also clear from Fig. 5 that the kinetics of H storage is much faster for the condensed nanosized powder than for the atomized micro-sized one. The increase in surface area due to the reduced scale of the condensed powder and the presence of the MgO phase formed during their passivation process that acts as a catalyst, can explain the fastest kinetics. In return, as a fraction of these condensed powder particles had reacted with oxygen, less Mg was available to react with hydrogen; which explains a lower H-storage capacity of the condensed and passivated powder. Comparing now in Fig. 5 the behavior of the powders with their HPT consolidated products, it is clear that the effect of the HPT process is, in both cases, to increase the storage kinetics while reducing the storage capacity. For example, the
capacity of the atomized powder after 10 h of storage is reduced from about 7.6% down to 6.1% after HPT processing. The structural defects (vacancies, dislocations) as well as the reduction in grain size induced by the severe plastic deformation are likely to create diffusion path and increase the reactivity of the material surface, thus improving the storage kinetics. Comparatively, the reduction of the storage capacity is a little more surprising. As will be shown by the XRD analyses of the hydrogenated HPT products (Fig. 8), this lower capacity is due to the fact that some of the magnesium did not react. In addition, it may be considered, on this first cycle, that the fully consolidated and sintered HPT samples produced in the present study did not allow H to fully penetrate into the interior of the bulk HPT products. Alternatively, it is possible that, while the increased amount of structural defects and grain boundaries that act as faster diffusion path, some of these crystallographic perturbations may alter the amount of H that can be stored in the magnesium lattice. Further analysis, both theoretical and experimental, should be carried out to elucidate the effect of the nature and quantity of structural defects induced by severe plastic deformation on the storage ability.
H-sorption properties of the HPT products Thermodynamics of absorption/desorption Thermodynamics of H uptake and release were assessed on the one-cycled activated HPT samples by measuring the pressureecompositionetemperature (PCT) isotherms in the temperature range of 325e400 C. The PCT diagrams for both HPT products are depicted in Fig. 6a and c, and the data obtained from these cycles are given in Table 1. In Table 1 are also given experimental results from the literature [29,30] obtained from condensed ultrafine Mg powders under the same type of pressureecompositionetemperature isotherm cycles. In these cycles, as seen from the PCT diagrams and the data at 400 C, the maximum H uptake capacity in the microHPT product was determined to be 5.5 wt% H higher than the storage capacity of the nano-HPT product ~5.2 wt% H. The fact that these capacities are comparatively lower than those obtained for the activation process given in Fig. 5 (micro-HPT: 6.1 wt% H and nano-HPT: 5.7 wt% H), is obviously related to the different measurement conditions; while one is at near
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
6
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Fig. 6 e Pressure-composition-temperature (PCT) isotherms obtained at different temperatures and their absorption van't Hoff plots for the micro-HPT (a, b) and nano-HPT (c, d) products.
Fig. 7 e Dynamic hydrogen absorption profiles recorded at different temperatures on the samples already activated, and recycled during the thermodynamic measurements for the (a) micro-HPT and (b) nano-HPT products.
equilibrium, where the pressure is adjusted by the thermodynamics, the other is obtained under a fixed pressure of 3.5 MPa which produces an external driving force and leads to higher H storage quantities. Consistently with the fact that the equilibrium pressure depends on temperature [32], the equilibrium pressure and H uptake capacities decrease when the
operating temperature is lowered. This is true for the HPT products as well as for the powders tested in the literature. However, it can also be noticed that the decrease in storage capacity with the temperature is faster for the micro-HPT product than for the nano-HPT one. Indeed, the maximum storage decreases from 5.5 wt% at 400 C down to 4.26 wt% at
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
7
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
desorption followed by a faster desorption kinetics for the micro-HPT than the nano-HPT products. Based on the PCT curves, the van't Hoff plots (ln P vs. 1000/ T) for H absorption of both HPT consolidated products are displayed in Fig. 6b and d. The hydrogenation enthalpy (DHabs) was calculated from the slope of a linear fitting of the van't Hoff equations (ln P ¼ 9.33/T þ 14.52 for micro-HPT and ln P ¼ 9.07/T þ 13.85 for nano-HPT products), and are reported as 77.5 and 75.4 kJ/mol H2 for the micro-HPT and nano-HPT products, respectively (Table 2). These values are rather close to the one recorded on the initial ultrafine Mg powders (78.6 kJ/mol H2) [29,30].
Kinetics of absorption
Fig. 8 e Normalized X-ray diffraction patterns of the hydrogenated products obtained by dynamic hydrogen absorption at 400 C for 6 h under a hydrogen pressure of 3.5 MPa from the samples already activated, and recycled during the thermodynamic and kinetic measurements.
325 C for the micro-HPT products while it moves in the nanoHPT products from 5.2 wt% to 4.78 wt% for the same temperature range. Also, it is evident from the comparison of Fig. 6a and c with Table 1 that both HPT products cannot desorb H at the lowest temperature of 325 C while this was possible in the initial condensed ultrafine powders tested in the literature [29,30]. Another drawback of the HPT processing, clearly visible when comparing the behavior of the consolidated products with the powders (Table 1), is that it lowers substantially the H uptake capacity of the material. Comparatively, one of the positive effects of the HPT processing, clearly visible in Table 1 when comparing the nanoHPT products with its ultrafine condensed precursor, is that it reduces the hysteresis between the absorption and desorption plateau pressures. Finally, the comparison of the curves in Fig. 6a and c reveals a stronger incubation for
The dynamic absorption profiles recorded at different temperatures under a hydrogen pressure of 3.5 MPa for the two HPT products are depicted in Fig. 7. These measurements were carried out on the samples already activated, and recycled during the thermodynamic measurements. It is apparent from the figures that the nano-HPT product can absorb 3.5 wt % H within 5 min and 5.2 wt% after 2 h at 300 C. Comparatively, the micro-HPT product can absorb ~2 wt% and 4.5 wt% H, respectively, for the same experimental conditions. Moreover, it is interesting to note that, at 250 C the nano-HPT product can still absorb more than 4.5 wt% H within 2 h, while, comparatively, the absorption in the micro-HPT sample is restricted to only ~0.7 wt% at this temperature. Furthermore, due to its nanostructural features of the Mg matrix and the catalytic effects of the dispersed MgO oxides, the nanoHPT composite can able to absorb 3.1 wt% H within 3 h even at 200 C (see Fig. 7b). The kinetic curves displayed in Fig. 7 were fitted by the JohnsoneMehleAvramieKolmogorov (JMAK) model [33,34]. This model describes a process of nucleation and growth using the relation ln [ln (1 a)] ¼ n ln k þ n ln t, where a is the fraction of Mg transformed into MgH2 at time t, k is an effective kinetic parameter, and n is the Avrami exponent or the order of the reaction, which provides information about the dimensionality of the transformation. The initial part of the absorption curves was used to linearly fit with the JMAK equation and from the ln [ln (1 a)] vs. ln t plot, the effective
Table 1 e Comparative thermodynamic properties obtained from the PCT analyses of the HPT products tested here and the ultrafine condensed powder from the literature [29,30]. Samples
Micro-HPT (this work)
Nano-HPT (this work)
Ultrafine condensed powder [29,30]
Temperature ( C) 400 375 350 325 400 375 350 325 400 375 350 325
[29,30] [29,30] [26,30] [30]
Habs content (wt%)
Hdes content (wt%)
H retained (wt%)
Absorption plateau pressure (MPa)
Desorption plateau pressure (MPa)
Pressure hysteresis (MPa)
5.50 5.21 4.47 4.26 5.20 4.99 4.84 4.78 6.14 5.88 5.69 5.48
5.10 5.01 4.32 e 5.10 4.85 4.64 e 5.94 5.63 5.29 4.67
0.40 0.20 0.15 e 0.10 0.14 0.20 e 0.20 0.25 0.40 0.81
1.247 0.928 0.915 0.287 1.440 0.855 0.483 0.266 1.467 0.792 0.443 0.251
1.154 0.826 0.795 e 1.371 0.799 0.434 e 1.04 0.682 0.368 0.183
0.093 0.102 0.120 e 0.069 0.056 0.049 e 0.427 0.110 0.075 0.068
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
8
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Table 2 e Summary of the hydrogenation and dehydrogenation properties obtained for the micro-HPT and nano-HPT products. HPT products/ properties
Micro-HPT Nano-HPT
Hydrogenation properties
Dehydrogenation properties
DHabs (kJ/mol H2)
Activation energy (Ea) (kJ/mol H2)
H uptake (wt%) at 400 C
H release (wt%) from TG
Onset temperature ( C)
Activation energy (Ea) (kJ/mol H2)
77.5 75.4
70.2 43.5
6.6 5.5
5.1 4.9
402.8 380.7
170.3 209.1
kinetic parameter (k) was calculated. Based on the experimental data obtained at different temperatures, the Arrhenius equation was employed to determine the average apparent activation energies; the results are provided in Table 2. It was found that the HPT processed product obtained from the condensed ultrafine powder shows a lower value of apparent activation energy (Ea ¼ 43.5 kJ/mol H2) compared to that of the micro-HPT one (Ea ¼ 70.2 kJ/mol H2). However, both HPT products have much lower activation energy than those obtained for the initial ultrafine Mg powders (92.9 kJ/mol H2) [30] or pure Mg nanocrystals prepared from solution (115e122 kJ/ mol H2) [35].
Hydrogenated products After the PCT and kinetic measurements, the samples were hydrogenated again at 400 C for 6 h under a hydrogen pressure of 3.5 MPa. After these different cycling processes, the H uptake capacities in these hydrogenated samples were determined to be 6.6 and 5.5 wt% H for the micro-HPT and nano-HPT products, respectively (see Table 2). Part of these hydrided samples was used for characterization of the hydrogenated products by XRD analyses (this section) while other parts were used for the characterization desorption properties (next section). The XRD patterns of the hydrogenated products are shown in Fig. 8. It is apparent from these patterns that most of the hcp-Mg phase has transformed into b-MgH2 upon hydrogenation, in both HPT products. However, the presence of hcpMg peaks and some MgO are also visible in the XRD patterns. The presence of hcp-Mg indicates an incomplete hydrogenation reaction even at 400 C and for a fairly long duration which explains the inherent lower H uptake capacities revealed previously for the HPT products compared to the initial powder precursors (Fig. 5 and Table 1). It can also be noticed from the relative intensity of the peaks that the quantity of remnant Mg is higher in the micro-HPT product than in the nano-HPT one. After the different cycles that the materials have previously sustained, this is consistent with the evolutions shown in Fig. 6 and the data given in Table 1 indicating a faster decrease in storage capacity after cycling for the micro-HPT product than the nano-HPT one. Concerning the presence of MgO in the XRD traces e in particular for the micro-HPT product for which it was not detected after the HPT processing (see Fig. 3a) e it might have formed at the surface during handling of the hydrogenated products out of the glove box before the XRD analyses. However, it is also interesting to notice that, for the nano-HPT product, that the Mg(OH)2 peaks observed in the XRD patterns of ultrafine powder and nano-HPT product before
H retained (wt%)
1.5 0.6
hydrogenation (see Fig. 3b) have disappeared after hydrogenation (see Fig. 8). This is likely due to the endothermic decomposition of the Mg(OH)2 phase at high temperature (above 330 C) [36] into MgO during the H-sorption cycles, which thereby increased the MgO content in the nano-HPT product.
Kinetics of desorption The desorption behavior of the hydrogenated samples were examined using differential scanning calorimetryethermogravimetry (DSC-TG) technique operated under 0.1 MPa of purified argon after a storage cycle of 6 h at 400 C under a hydrogen pressure of 3.5 MPa. The heating rates applied in these measurements were 3, 5 and 10 C/min in the temperature range of 25e500 C. The DSC-TG curves for the heating rate of 10 C/min are displayed in Fig. 9. The most characteristic findings from the DSC-TG analysis are given in Table 2. The amount of desorbed H was calculated from the TG curves. While the micro-HPT product can desorb 5.1 wt% H, the nano-HPT product can desorb ~4.9 wt% H for the same experimental conditions. Thus, despite a lower storage capacity for the nano-HPT product, the differences between the absorbed and desorbed amounts indicates that a lower amount of H was retained after desorption in the nano-HPT (0.6 wt% H) than in the micro-HPT one (1.5 wt% H). In addition, it can also be noticed from the DSC curves that the onset temperature for desorption is much lower (380.7 C) for the nano-HPT product than for the micro-HPT one (402.8 C). These values for the HPT products are lower than the one of the initial condensed ultrafine Mg powders (423 C) [30], indicating that one of the positive effect of the HPT process is to reduce the dehydriding temperature of the Mg hydride. Besides, it is interesting to note that the slope of the TG curve (see Fig. 9a and c) after the onset of desorption is much steeper in case of micro-HPT product (slopes: 1.93 wt% H/min versus 1.65 wt% H/min). This indicates that, after a slightly longer incubation period, the hydrogenated micro-HPT product has a faster desorption rate than its hydrogenated nano-HPT counterpart. These results are consistent with the aspect of the PCT curves given in Fig. 6a which clearly reveal the necessity of higher depression (incubation) followed by fastest desorption for the micro-HPT product than in Fig. 6c for the nano-product. The Kissinger method [37], based on DSC analysis, and the related peak temperature depending on heating rate were employed to determine the activation energy of the desorption kinetics; the corresponding Kissinger plots (ln (b/T2) vs. 1000/T) for both products are shown in Fig. 9b and d. As can be
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
9
Fig. 9 e DSC and TG curves (a, c) recorded at a heating rate of 10 C/min from the hydrogenated products, and the corresponding Kissinger plots (b, d) for the micro-HPT (a, b) and nano-HPT (c, d) products.
seen from these figures, the apparent activation energy for the micro-HPT product is 170.3 kJ/mol H2, lower than the value obtained for the nano-HPT composite (209.1 kJ/mol H2), and fairly close to the results recorded for Mg nanocrystals (126e160 kJ/mol H2) [35].
Discussion The current investigation revealed that, despite a substantial amount of severe plastic deformation imparted by the HPT processing to the two types of Mg powders, the exact nature of the starting powder precursor had a pronounced effect on the microstructure development in the HPT products and in their H-sorption behaviors. These aspects are recalled and discussed hereafter in three different sections.
Microstructure of the HPT products Consistently with the recent results obtained on pure bulk Mg subjected to HPT [13,31], the occurrence of dynamic recrystallization during the HPT processing led to the formation of an equiaxed microstructure in the micro-HPT product obtained by consolidation of the micro-sized atomized powder. While the HPT consolidation of the micro-sized atomized Mg powder resulted in a recrystallized equiaxed microstructure,
the consolidation of the ultrafine condensed and passivated powder has led to the formation of an Mg-MgO based nanocomposite, characterized by a much finer grain size. The MgO particles were essentially aligned along the shear direction and their interaction with the migrating grain boundaries has prevented the occurrence of significant dynamic recrystallization during the HPT processing. Thus, the microstructure consisted essentially of fine elongated grains e about 200 nm/ 300 nm in diameter e surrounded by fine oxide dispersion in the nano-HPT product. These differences in the microstructure have generated strong modifications in the H-sorption behaviors.
Effect of the HPT processing The effect of the HPT processing has been analyzed by activating the HPT products at 400 C and their initial precursor powders under similar conditions (Fig. 5) as well as by cycling the HPT products for absorption/desorption tests carried out at different temperatures (Fig. 6), and comparing these data with those obtained in the literature on evaporated ultrafine powders using similar conditions (Table 1). For both types of powder precursors, through the introduction of structural defects, the effect of the HPT processing was to increase substantially the kinetics of H-sorption while reducing the storage capacity. The reduction in the H uptake capacities of
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
10
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
the HPT products was the consequence of incomplete hydrogenation reactions that was evidenced from the presence of un-reacted Mg phase in the XRD patterns of the hydrogenated products (see Fig. 8). On cycling, for all types of samples, the equilibrium pressure and H uptake capacities obtained from the PCT experiments decreased when the operating temperature was lowered. However, another advantage of the HPT processing is that it reduces the hysteresis between the absorption and desorption plateau pressures.
3)
Effect of the initial powder precursors 4) The overall absorption kinetics was faster for the Mg-MgO based nano-HPT composite processed from the condensed ultrafine Mg powder. Comparatively, the micro-HPT product consolidated from the atomized micro-sized powder had the advantage of a higher H-sorption capacity. This behavior can be attributed to the presence of the additional MgO phase in the nano-HPT composite. While the presence of an MgO layer is often regarded as an obstacle for H-sorption e as it prevents the H atoms from entering into the Mg substrate [12] e it is claimed that the faster H-uptake rate can be related to the presence of oxides that can serve as nucleation sites for hydride growth. In addition, it has also been reported in Ref. [9] that if MgH2 is well intermixed with MgO, an improved Hsorption kinetics is obtained due to the reduced particle size of MgH2 and some catalytic effects of MgO particles. In terms of storage, the only disadvantage with MgO noticed in the present study is that it has impaired the storage capacity in the nano-HPT composite in comparison with the micro-HPT product. In terms of desorption, after some incubation, the microHPT product revealed a faster desorption rate, accompanied with much lower activation energy. It is also interesting to note, however, that the hydrogenated nano-HPT material showed a lower onset temperature of desorption and a relatively lower peak maximum in the DSC analyses. This suggests that some trapping sites were less effective in this type of consolidated product. Further analysis is under way to clarify if differences exist in the character of the grain boundaries between the two consolidated products.
Conclusions A comparative study of H-sorption behaviors was conducted on two types of HPT products obtained from the consolidation of two kinds of Mg powder precursors: atomized micro-sized powder, and condensed and passivated ultrafine powder particles. The deformed microstructures were characterized, and the H-sorption properties were investigated. The main findings of the present study can be summarized as follows. 1) HPT consolidation applied on the atomized micro-sized Mg powder developed a dynamically recrystallized microstructure with fairly equiaxed 0.5e1 mm grains in the micro-HPT product. 2) The HPT consolidation applied on the condensed and passivated ultrafine powder led to the formation of an Mg-
5)
6)
MgO based nanocomposite, the so-called nano-HPT product, within which slightly elongated 200 nm/300 nm thick grains were pinned by the nano-sized MgO oxide particles. For both types of powder precursors, through the introduction of structural defects and microstructural refinement, the HPT processing has provided with significant improvements of the hydrogenation kinetics for the HPT products compared to their initial powder precursors. Another advantage of the HPT processing is to reduce the hysteresis between the absorption and desorption plateau pressure during the PCT experiments. The major drawback of the HPT processing was, for both types of powder precursors, that it reduced the maximum H uptake capacities. The nano-HPT product showed faster absorption kinetics accompanied with a much lower activation energy (43.5 kJ/ mol H2), which was attributed to its nanostructural features and to the catalytic effect of the MgO oxide particles. Comparatively, the micro-HPT product had the advantage of having higher H-sorption capacity. Despite a slightly lower onset temperature of desorption, the nano-HPT product exhibited slower desorption kinetics and a higher activation energy than the micro-HPT product.
Acknowledgements This work was supported by the French State through the program “Investment in the future” operated by the National Research Agency (ANR) and referenced by ANR-11-LABX-000801 (Labex DAMAS). This work was also partly supported by the National Key Research and Development Program of China (2016YBF0701203), Science and Technology Commission of Shanghai Municipality (14JC1491600) and “111” project from China's Ministry of Education (B16032).
references
[1] Sakintuna B, Lamari-Darkrim F, Hirscher M. Metal hydride materials for solid hydrogen storage: a review. Int J Hydrogen Energy 2007;32:1121e40. [2] Jain IP, Lal C, Jain A. Hydrogen storage in Mg: a most promising material. Int J Hydrogen Energy 2010;35:5133e44. [3] Reiser A, Bogdanovic B, Schlichte K. The application of Mgbased metal-hydrides as heat energy storage systems. Int J Hydrogen Energy 2000;25:425e30. [4] Corgnale C, Hardy B, Motyka T, Zidan R. Metal hydride based thermal energy storage system requirements for high performance concentrating solar power plants. Int J Hydrogen Energy 2016;41:20217e30. [5] Jain IP. Hydrogen the fuel for 21st century. Int J Hydrogen Energy 2009;30:7368e78. [6] Liang G, Huot J, Boily S, Van Neste A, Schulz R. Hydrogen storage properties of the mechanically milled MgH2eV nanocomposite. J Alloys Compd 1999;291:295e9. [7] Oelerich W, Klassen T, Bormann R. Metal oxides as catalysts for improved hydrogen sorption in nanocrystalline Mg-based materials. J Alloys Compd 2001;315:237e42.
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
[8] Barkhordarian G, Klassen T, Bormann R. Fast hydrogen sorption kinetics of nanocrystalline Mg using Nb2O5 as catalyst. Scr Mater 2003;49:213e7. [9] Aguey-Zinsou KF, Nicolaisen T, Fernandez JA, Klassen T, Bormann R. Effect of nanosized oxides on MgH2 (de) hydriding kinetics. J Alloys Compd 2007;434:738e42. [10] Aguey-Zinsou KF, Fernandez JA, Klassen T, Bormann R. Using MgO to improve the (de) hydriding properties of magnesium. Mater Res Bull 2006;41:1118e26. [11] Cakmak G, Ozturk T. ECAP processing and mechanical milling of Mg and MgeTi powders: a comparative study. J Mater Sci 2011;46:5559e67. [12] Ares JR, Aguey-Zinsou KF, Klassen T, Bormann R. Influence of impurities on the milling process of MgH2. J Alloys Compd 2007;434:729e33. [13] Edalati K, Yamamoto A, Horita Z, Ishihara T. High-pressure torsion of pure magnesium: evolution of mechanical properties, microstructures and hydrogen storage capacity with equivalent strain. Scr Mater 2011;64:880e3. [14] Skripnyuk VM, Rabkin E, Estrin Y, Lapovok R. The effect of ball milling and equal channel angular pressing on the hydrogen absorption/desorption properties of Mge4.95 wt% Zne0.71 wt% Zr (ZK60) alloy. Acta Mater 2004;52:405e14. [15] Leiva DR, Fruchart D, Bacia M, Girard G, Skryabina N, Villela AC, et al. Mg alloy for hydrogen storage processed by SPD. Int J Mater Res 2009;100:1739e46. [16] Botta WJ, Jorge Jr AM, Veron M, Rauch EF, Ferrie E, Yavari AR, et al. H-sorption properties and structural evolution of Mg processed by severe plastic deformation. J Alloys Compd 2013;580:S187e91. [17] Lima GF, Triques MR, Kiminami CS, Botta WJ, Jorge AM. Hydrogen storage properties of pure Mg after the combined processes of ECAP and cold-rolling. J Alloys Compd 2014;586:S405e8. [18] Grill A, Horky J, Panigrahi A, Krexner G, Zehetbauer M. Longterm hydrogen storage in Mg and ZK60 after severe plastic deformation. Int J Hydrogen Energy 2015;40:17144e52. [19] Korznikov AV, Safarov IM, Laptionok DV, Valiev RZ. Structure and properties of superfine-grained iron compacted out of ultradisperse powder. Acta Mater 1991;39:3193e7. [20] Valiev RZ, Mishral RS, Grozal J, Mukherjee AK. Processing of nanostructured nickel by severe plastic deformation consolidation of ball-milled powder. Scr Mater 1996;34:1443e8. [21] Kusadome Y, Ikeda K, Nakamori Y, Orimo S, Horita Z. Hydrogen storage capability of MgNi2 processed by high pressure torsion. Scr Mater 2007;57:751e3. [22] Leiva DR, Jorge AM, Ishikawa TT, Huot J, Fruchart D, Miraglia S, et al. Nanoscale grain refinement and H-Sorption properties of MgH2 processed by high-pressure torsion and other mechanical routes. Adv Eng Mater 2010;12:786e92.
11
nya Z, Verebelyi T, Szabo PJ, Zhilyaev AP, [23] Revesz A, Ka Spassov T. The effect of high-pressure torsion on the microstructure and hydrogen absorption kinetics of ballmilled Mg70Ni30. J Alloys Compd 2010;504:83e8. [24] Liang J, Hout J. A new approach to the processing of metal hydrides. J Alloys Compd 2011;509:L18e22. [25] Zou JX, Perez-Brokate CF, Arruffat R, Bolle B, Fundenberger JJ, Zeng XQ, et al. Nanostructured bulk MgþMgO composite synthesized through arc plasma evaporation and high pressure torsion for H-storage application. Mater Sci Eng B 2014;183:1e5. [26] Grosdidier T, Fundenberger JJ, Zou JX, Pan YC, Zeng XQ. Nanostructured Mg based hydrogen storage bulk materials prepared by high pressure torsion consolidation of arc plasma evaporated ultrafine powders. Int J Hydrogen Energy 2015;40:16985e91. [27] Shao H, Wang Y, Xu H, Li X. Hydrogen storage properties of magnesium ultrafine particles prepared by hydrogen plasma-metal reaction. Mater Sci Eng B 2004;110:221e6. [28] Zou JX, Sun HQ, Zeng XQ, Ji G, Ding WJ. Preparation and hydrogen storage properties of Mg-rich Mg-Ni ultrafine particles. J Nanomater 2012;2012:2. [29] Stephane P, Zou JX, Zeng XQ, Sun HQ, Ding WJ. Preparation and hydrogen storage properties of ultrafine pure Mg and Mg-Ti particles. Trans Nonfer Met Soc China 2012;22:1849e54. [30] Zou JX, Zeng XQ, Ying Y, Chen X, Guo H, Zhou S, et al. Study on the hydrogen storage properties of coreeshell structured MgeRE (RE¼Nd, Gd, Er) nano-composites synthesized through arc plasma method. Int J Hydrogen Energy 2013;38:2337e46. [31] Panda S, Toth LS, Fundenberger JJ, Perroud O, Guyon J, Zou JX, et al. Analysis of heterogeneities in strain and microstructure in aluminum alloy and magnesium processed by high-pressure torsion. Mater Charact 2017;123:159e65. [32] Bogdanovic B. Magnesium hydride: a homogeneouscatalysed synthesis and its use in hydrogen storage. Int J Hydrogen Energy 1984;9:937e41. [33] Mintz MH, Zeiri Y. Hydriding kinetics of powders. J Alloys Compd 1995;216:159e75. [34] Kelton KF. Analysis of crystallization kinetics. Mater Sci Eng A 1997;226:142e50. [35] Norberg NS, Arthur TS, Fredrick SJ, Prieto AL. Size-dependent hydrogen storage properties of Mg nanocrystals prepared from solution. J Am Chem Soc 2011;133:10679e81. [36] Burke PJ, Bayindir Z, Kipouros GJ. X-ray photoelectron spectroscopy (XPS) investigation of the surface film on magnesium powders. Appl Spectrosc 2012;66:510e8. [37] Kissinger HE. Reaction kinetics in differential thermal analysis. Anal Chem 1957;29:1702e6.
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg Subrata Panda a,b,c,*, Jean-Jacques Fundenberger a,b, Yajun Zhao a,b, Jianxin Zou a,c, Laszlo S. Toth a,b, Thierry Grosdidier a,b,** Universite de Lorraine, Laboratory of Excellence on Design of Alloy Metals for Low-mass Structures (DAMAS), Ile du Saulcy, Metz F-57045, France b Universite de Lorraine, Laboratoire d’Etude des Microstructures et de Mecanique des Materiaux (LEM3 UMR 7239), Ile du Saulcy, Metz F-57045, France c Shanghai Engineering Research Center of Mg Materials and Applications, National Engineering Research Center of Light Alloy Net Forming, Shanghai Jiao Tong University, Shanghai 200240, PR China a
article info
abstract
Article history:
While severe plastic deformation (SPD) on bulk samples has been widely applied for
Received 22 December 2016
modifying the H-sorption properties, there has been little attention towards the use of SPD
Received in revised form
on powder materials. In this context, the aim of the present work was to compare the H-
13 May 2017
storage properties of high-pressure torsion (HPT) consolidated products obtained from two
Accepted 14 May 2017
distinct Mg powder precursors: atomized micro-sized and condensed ultrafine powder
Available online xxx
particles. The results showed that the nature of the initial powder precursor had a pronounced effect on the H-sorption behavior. The HPT product obtained from the condensed
Keywords:
ultrafine powder showed faster absorption kinetics than the consolidated product obtained
HPT
from the atomized powder. However, the HPT product obtained from atomized powder
Powder consolidation
could absorb more hydrogen and showed faster desorption kinetics corresponding to a
Nanocomposites
lower activation energy. These results are discussed by taking into account the effective-
H-sorption
ness of the HPT process to refine the grain sizes and differences in the dispersion of fine
Hydrogenation/dehydrogenation
MgO oxide particles. © 2017 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
de Lorraine, Laboratory of Excellence on Design of Alloy Metals for low-mass Structures (DAMAS), Ile * Corresponding author. Universite du Saulcy, Metz F-57045, France. de Lorraine, Laboratory of Excellence on Design of Alloy Metals for low-mass Structures (DAMAS), Ile ** Corresponding author. Universite du Saulcy, Metz F-57045, France. E-mail addresses: [email protected] (S. Panda), [email protected] (J.-J. Fundenberger), [email protected] (Y. Zhao), [email protected] (J. Zou), [email protected] (L.S. Toth), [email protected] (T. Grosdidier). http://dx.doi.org/10.1016/j.ijhydene.2017.05.097 0360-3199/© 2017 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
2
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Introduction A considerable amount of research has been devoted to develop advanced H-storage media in solid-state. As a promising H carrier, magnesium has been considered as an excellent candidate for H-storage mainly due to its light weight, abundance in the earth's crust and high gravimetric storage capacity (7.6 wt% H) [1,2]. In addition, due to their high thermal energy density the Mg-based metal hydride systems are also considered as potential high temperature heat storage media [3,4]. Unfortunately, the sluggish hydriding/dehydriding kinetics, and stable thermodynamic properties [5] leading to higher processing temperatures are the main obstacles for metal-H systems for potential industrial applications. Significant improvements in the H-sorption kinetics of metal-H systems have been achieved by the use of highenergy-ball-milling (HEBM) [6e12], which were primarily considered to be the effect of nano-metric crystallite/grain size, introduction of lattice defects and increase of reactive surface area. The catalytic action of additions of transition metals or oxides was also considered to be beneficial. In particular, Barkhordarian et al. [8] showed that milling of MgH2 with 1 mol% of different oxides such as TiO2, V2O5 or Nb2O5 leads to an improvement of magnesium H-sorption kinetics. Aguey-Zinsou et al. [9] considered that the major effect of the MxOy oxides was related to the refinement of the size of MgH2 particles during milling. It was also reported that co-milling of Mg with its own MgO does not modify the structural and thermodynamic properties of MgH2 but leads to an efficient decrease of the particle size and therefore enhances the H-sorption kinetics [10]. Although the milling route results in considerable improvements in H-sorption properties, it suffers from two major drawbacks. Firstly, it is difficult to avoid agglomeration during milling of Mg or Mg alloys [11]. Secondly, impurities coming from the milling tools and gas adsorption are hard to control and can affect significantly the H-sorption kinetic properties [12]. To avoid these drawbacks, severe plastic deformation (SPD) techniques have been recently used such as equalchannel angular pressing (ECAP) and high pressure torsion (HPT) [13e18]. It was suggested from SPD on pure Mg that the refinement in grain size (i.e. large increase in amount of grain boundaries) played a more effective role than the amount of dislocations for modifying the H-storage properties [13]. Although the compaction of powders under high pressure plus torsional straining has been initiated already in the 90's [19,20], processing routes involving the consolidation of powders using SPD techniques for H-storage materials have received comparatively little attention [21e26]. Kusadome et al. [21] investigated the effect of high-pressure torsion (HPT) on MgNi2 powders and found that the intermetallic (normally accepted as non-absorber) could absorb 0.1 wt% H even at 100 C. Leiva et al. [22] have employed several SPD techniques (HPT, cold rolling and forging) on MgH2 and MgH2 e Fe powder mixtures and reported that SPD can also reduce crystallite size. Liang and Hout [24] showed that the application of cold rolling on MgH2 powders could be an efficient processing route to produce nanocrystalline structure for improving H-sorption properties.
A new method e applying HPT on arc plasma evaporated ultrafine grained powder e has been recently introduced to produce bulk Mg-MgO based nanocomposite for H storage [25,26]. The improved H-sorption properties of these bulk composites were mainly attributed to the refinement of the Mg matrix and the dispersion of fine MgO particles formed during intense shearing by HPT [25]. The effect of small amount of Fe and Ni addition was also examined [26]. This new processing route has potentially several advantages. Compared to HEBM, it is less likely to introduce contamination within the heavily deformed material. Also, in comparison with other SPD routes like, for example, ECAP, it is simpler to control and easier to apply on powder. In this context, in the present study the same processing route was employed with the aim of examining the effect of the nature of the initial powder precursors. To this end, two distinct Mg powder precursors consisting of micro-sized powder particles produced by inert gas atomization and ultrafine powder particles synthesized by an arc plasma evaporation method were employed. The idea of using atomized micro-sized powder instead of the evaporated ultrafine powder lies also in the fact that the production of atomized powder is much cheaper and can be done in more significant quantities.
Experimental Sample processing Two Mg powder precursors having different particle sizes were used in the present work for HPT processing. Commercial purity magnesium (99.8%) was gas atomized by SFM SA (Switzerland) to produce a micro-sized Mg powder with a particle size distribution in the range of 10e80 mm. Ultrafine magnesium powder was also produced by arc plasma evaporation/condensation method to generate particle sizes in the range of 50e600 nm. In the latter case, the apparatus was filled with a mixture of 0.70 atm Ar and 0.10 atm H2. The details of this procedure are available in Refs. [27,28]. In order to prevent these very fine Mg particles from burning in open air, a twostep passivation procedure was applied: maintaining the powder for 2 h in a mixture of 50 kPa Ar þ 10 kPa normal air followed by holding for 12 h in a mixture of 50 kPa Ar þ 50 kPa normal air, before opening the evaporation/condensation chamber [27,28]. Fig. 1 shows images illustrating the aspects of the different powders. The atomization process led to fairly rounded particles (see Fig. 1a) having a particle size distribution centered on ~15 mm in the micro-sized powder. Typical SEM image and TEM image of the ultrafine powder are shown in Fig. 1b. They demonstrate that the particle size distribution was essentially in the sub-micrometer range with an average particle size below 300 nm, and that each powder particle corresponded generally to a single grain. The hexagonal shape of some finer powder particles, visible in the TEM image, is due to the hexagonal structure of the Mg phase [25]. These two powders were separately processed by a twostep HPT facility. A schematic diagram of the processing route employed in this study is illustrated in Fig. 2. The
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
3
Fig. 1 e Morphological features of the initial powders: (a) SEM image of the atomized micro-sized Mg powder and (b) SEM image and a typical bright field TEM image (inset) of the condensed and passivated ultrafine Mg powder.
Fig. 2 e Schematic illustrations of a two-step HPT consolidation route employed in this study; step 1: uniaxial compression of the Mg powder into a precompacted disk (a) followed by step 2: torsional straining using HPT (b).
powders were first compacted by uniaxial compression under a hydrostatic pressure of 1.5 GPa (holding for 10 min) to form disks with a diameter of 20 mm and a thickness of 3 mm, as depicted in Fig. 2a. A subsequent torsional straining by HPT was conducted on the precompacted disks under a hydrostatic pressure of 1.2 GPa at room temperature, as portrayed in Fig. 2b. The final step was conducted under quasi-constrained conditions [21], where the material lateral flow was partially restricted. The torsional straining was imposed by rotating the lower anvil at 0.125 rpm up to 2 revolutions for a total maximum shear strain (g) of about 42 at the edge of the disks. In the following, the HPT-disks processed from the microatomized and condensed ultrafine Mg powders will be referred to as micro-HPT and nano-HPT products, respectively.
Characterization The crystallographic phases present in the initial powders, after HPT and after hydrogenation were identified by X-ray
diffraction (XRD) using a powder X-ray diffraction apparatus (D/max 2550VL/PCX) equipped with a monochromatic Cu-Ka radiation source. Microstructural characterization of the initial powder and the bulk HPT products was carried out by scanning electron microscopy (SEM) using a field emission gun scanning electron microscope (SEM e Zeiss Supra40). The H-storage properties of the HPT products were examined using a Sievert-type pressureecompositione temperature (PCT) volumetric apparatus (Type PCT-2, Shanghai Institute of Microsystem and Information Technology). Prior to the PCT and kinetic measurements, the bulk HPT products were broken into several smaller pieces (without using any file or attritor mill) to fit into the testing vessel. To characterize the effect of the HPT processing, the initial powders as well as their consolidated HPT products were activated at 400 C under a hydrogen pressure of 3.5 MPa. These one-cycled activated HPT products were subsequently cycled for the thermodynamic measurements of absorption/ desorption tests at 400 C, 375 C, 350 C and 325 C. Following this, the samples were used for the dynamic hydrogen
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
absorption measurements under a hydrogen pressure of 3.5 MPa at 200 C, 250 C, 300 C and 350 C. Then, after a final hydrogen absorption at 400 C for 6 h under a hydrogen pressure of 3.5 MPa, the dehydriding properties of the hydrogenated HPT products were investigated by simultaneous differential scanning calorimetry and thermogravimetry measurements (DSC-TG, Netzsch STA449F3 Jupiter) under 0.1 MPa of purified argon at a heating rate of 3, 5 and 10 C/min from 25 to 500 C. Except from the pressure of 3.5 MPa (instead of 4 MPa), this procedure and these cycles of hydrogen absorption/desorption are rather similar to the ones used in previous papers dealing with the hydrogen storage properties of ultrafine particles [29,30].
Results HPT-processing and sintered products Fig. 3 compares the XRD patterns of the initial powders and their HPT-processed products. The peaks of the Mg hexagonal structure are the main contributors in these XRD patterns. These patterns are, however, different for two reasons. Firstly, as already reported in Refs. [25,26], since the ultrafine powder was passivated after the evaporation/condensation process, a thin layer of MgO formed around the Mg particles. Therefore, a minor peak of MgO phase is detected for both ultrafine powder and its nano-HPT product (see Fig. 3b). In addition, also in the XRD traces of the ultrafine and nano-HPT consolidated product, some Mg(OH)2 is identified through its (101) peak. This phase e which was not previously reported for these condensed ultrafine Mg powder [27,28] e might have formed together with the MgO during the passivation process in normal air. In contrast, the atomized micro-sized powder and its consolidated HPT product did not show such kind of additional peaks (see Fig. 3a). Secondly, the difference in peak ratios indicate different kinds of texture developments in these two distinct powder precursors subjected to severe plastic deformation through HPT.
The deformed microstructures in the consolidated products have been characterized by SEM observations at the periphery and at the middle-thickness of the HPT-disks. Since a similar kind of localized deformation at the middle-section of the HPT-processed disks for bulk materials has been recently reported by Panda et al. [31] e where the materials had experienced a massive shear deformation as compared to the top and bottom surfaces, the microstructural developments at the middle-section of the HPT-disks were observed in the present study. Typical SEM images obtained under backscattered electron (BSE) imaging conditions are displayed in Fig. 4 together with a sketch showing the examined location. These images illustrate that the grains are clearly inclined at about 45 from the shear plane after HPT treatments. In the HPT product consolidated from the condensed and passivated ultrafine powder (see Fig. 4b), the fine MgO particles (as also reported by Zou et al. [25]) were introduced within the grains during the massive shear straining, and appear as white spots randomly distributed throughout the Mg matrix. It is interesting to notice that some oxide particles also formed alignments inclined at about 45 . These oxide particles have generated significant Zener pinning that has restricted the grain size evolution within fine elongated domains with an average thickness of about ~200e300 nm. Comparatively, in the oxide free micro-HPT product (see Fig. 4a), and consistently with the results reported by Edalati et al. [13], severe plastic deformation has led to the development of a dynamically recrystallized microstructure consisting of rather equiaxed grains. In our case, these grains had an average grain size of ~0.5e1 mm.
Activation kinetics of the powders and their HPT products In order to assess the effects of the SPD processing route on the H-sorption behaviors of the two distinct powder particles, the initial powders as well as the consolidated bulk products were tested for activation kinetics. Fig. 5 compares the first H absorption kinetics measured at 400 C under a hydrogen pressure of 3.5 MPa for all the samples. Note that, because of
Fig. 3 e Normalized X-ray diffraction patterns for the initial powders and their products after HPT consolidation: (a) atomized micro-sized Mg powder and its micro-HPT product and (b) condensed ultrafine Mg powder and its nano-HPT product. Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
5
Fig. 4 e SEM micrographs taken under back scattered imaging conditions of the (a) micro-HPT and (b) nano-HPT products, examined at the outer edge of the disks and at their middle thicknesses as shown in the inset.
Fig. 5 e First hydrogen absorption kinetics measured from the initial loose powders as well as their HPT consolidated products at 400 C under a hydrogen pressure of 3.5 MPa.
their more sluggish behavior, 10 h were used instead of 8 h for the runs carried out on the atomized and micro-HPT products to try to reach a more steady state plateau. It is readily apparent from the comparison between the two powders that, while the atomized powder could reach almost the theoretical value of 7.6 wt% H for pure Mg, the condensed powder showed comparatively a reduced H uptake capacity with a plateau located at about ~7.1 wt% H. It is also clear from Fig. 5 that the kinetics of H storage is much faster for the condensed nanosized powder than for the atomized micro-sized one. The increase in surface area due to the reduced scale of the condensed powder and the presence of the MgO phase formed during their passivation process that acts as a catalyst, can explain the fastest kinetics. In return, as a fraction of these condensed powder particles had reacted with oxygen, less Mg was available to react with hydrogen; which explains a lower H-storage capacity of the condensed and passivated powder. Comparing now in Fig. 5 the behavior of the powders with their HPT consolidated products, it is clear that the effect of the HPT process is, in both cases, to increase the storage kinetics while reducing the storage capacity. For example, the
capacity of the atomized powder after 10 h of storage is reduced from about 7.6% down to 6.1% after HPT processing. The structural defects (vacancies, dislocations) as well as the reduction in grain size induced by the severe plastic deformation are likely to create diffusion path and increase the reactivity of the material surface, thus improving the storage kinetics. Comparatively, the reduction of the storage capacity is a little more surprising. As will be shown by the XRD analyses of the hydrogenated HPT products (Fig. 8), this lower capacity is due to the fact that some of the magnesium did not react. In addition, it may be considered, on this first cycle, that the fully consolidated and sintered HPT samples produced in the present study did not allow H to fully penetrate into the interior of the bulk HPT products. Alternatively, it is possible that, while the increased amount of structural defects and grain boundaries that act as faster diffusion path, some of these crystallographic perturbations may alter the amount of H that can be stored in the magnesium lattice. Further analysis, both theoretical and experimental, should be carried out to elucidate the effect of the nature and quantity of structural defects induced by severe plastic deformation on the storage ability.
H-sorption properties of the HPT products Thermodynamics of absorption/desorption Thermodynamics of H uptake and release were assessed on the one-cycled activated HPT samples by measuring the pressureecompositionetemperature (PCT) isotherms in the temperature range of 325e400 C. The PCT diagrams for both HPT products are depicted in Fig. 6a and c, and the data obtained from these cycles are given in Table 1. In Table 1 are also given experimental results from the literature [29,30] obtained from condensed ultrafine Mg powders under the same type of pressureecompositionetemperature isotherm cycles. In these cycles, as seen from the PCT diagrams and the data at 400 C, the maximum H uptake capacity in the microHPT product was determined to be 5.5 wt% H higher than the storage capacity of the nano-HPT product ~5.2 wt% H. The fact that these capacities are comparatively lower than those obtained for the activation process given in Fig. 5 (micro-HPT: 6.1 wt% H and nano-HPT: 5.7 wt% H), is obviously related to the different measurement conditions; while one is at near
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
6
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Fig. 6 e Pressure-composition-temperature (PCT) isotherms obtained at different temperatures and their absorption van't Hoff plots for the micro-HPT (a, b) and nano-HPT (c, d) products.
Fig. 7 e Dynamic hydrogen absorption profiles recorded at different temperatures on the samples already activated, and recycled during the thermodynamic measurements for the (a) micro-HPT and (b) nano-HPT products.
equilibrium, where the pressure is adjusted by the thermodynamics, the other is obtained under a fixed pressure of 3.5 MPa which produces an external driving force and leads to higher H storage quantities. Consistently with the fact that the equilibrium pressure depends on temperature [32], the equilibrium pressure and H uptake capacities decrease when the
operating temperature is lowered. This is true for the HPT products as well as for the powders tested in the literature. However, it can also be noticed that the decrease in storage capacity with the temperature is faster for the micro-HPT product than for the nano-HPT one. Indeed, the maximum storage decreases from 5.5 wt% at 400 C down to 4.26 wt% at
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
7
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
desorption followed by a faster desorption kinetics for the micro-HPT than the nano-HPT products. Based on the PCT curves, the van't Hoff plots (ln P vs. 1000/ T) for H absorption of both HPT consolidated products are displayed in Fig. 6b and d. The hydrogenation enthalpy (DHabs) was calculated from the slope of a linear fitting of the van't Hoff equations (ln P ¼ 9.33/T þ 14.52 for micro-HPT and ln P ¼ 9.07/T þ 13.85 for nano-HPT products), and are reported as 77.5 and 75.4 kJ/mol H2 for the micro-HPT and nano-HPT products, respectively (Table 2). These values are rather close to the one recorded on the initial ultrafine Mg powders (78.6 kJ/mol H2) [29,30].
Kinetics of absorption
Fig. 8 e Normalized X-ray diffraction patterns of the hydrogenated products obtained by dynamic hydrogen absorption at 400 C for 6 h under a hydrogen pressure of 3.5 MPa from the samples already activated, and recycled during the thermodynamic and kinetic measurements.
325 C for the micro-HPT products while it moves in the nanoHPT products from 5.2 wt% to 4.78 wt% for the same temperature range. Also, it is evident from the comparison of Fig. 6a and c with Table 1 that both HPT products cannot desorb H at the lowest temperature of 325 C while this was possible in the initial condensed ultrafine powders tested in the literature [29,30]. Another drawback of the HPT processing, clearly visible when comparing the behavior of the consolidated products with the powders (Table 1), is that it lowers substantially the H uptake capacity of the material. Comparatively, one of the positive effects of the HPT processing, clearly visible in Table 1 when comparing the nanoHPT products with its ultrafine condensed precursor, is that it reduces the hysteresis between the absorption and desorption plateau pressures. Finally, the comparison of the curves in Fig. 6a and c reveals a stronger incubation for
The dynamic absorption profiles recorded at different temperatures under a hydrogen pressure of 3.5 MPa for the two HPT products are depicted in Fig. 7. These measurements were carried out on the samples already activated, and recycled during the thermodynamic measurements. It is apparent from the figures that the nano-HPT product can absorb 3.5 wt % H within 5 min and 5.2 wt% after 2 h at 300 C. Comparatively, the micro-HPT product can absorb ~2 wt% and 4.5 wt% H, respectively, for the same experimental conditions. Moreover, it is interesting to note that, at 250 C the nano-HPT product can still absorb more than 4.5 wt% H within 2 h, while, comparatively, the absorption in the micro-HPT sample is restricted to only ~0.7 wt% at this temperature. Furthermore, due to its nanostructural features of the Mg matrix and the catalytic effects of the dispersed MgO oxides, the nanoHPT composite can able to absorb 3.1 wt% H within 3 h even at 200 C (see Fig. 7b). The kinetic curves displayed in Fig. 7 were fitted by the JohnsoneMehleAvramieKolmogorov (JMAK) model [33,34]. This model describes a process of nucleation and growth using the relation ln [ln (1 a)] ¼ n ln k þ n ln t, where a is the fraction of Mg transformed into MgH2 at time t, k is an effective kinetic parameter, and n is the Avrami exponent or the order of the reaction, which provides information about the dimensionality of the transformation. The initial part of the absorption curves was used to linearly fit with the JMAK equation and from the ln [ln (1 a)] vs. ln t plot, the effective
Table 1 e Comparative thermodynamic properties obtained from the PCT analyses of the HPT products tested here and the ultrafine condensed powder from the literature [29,30]. Samples
Micro-HPT (this work)
Nano-HPT (this work)
Ultrafine condensed powder [29,30]
Temperature ( C) 400 375 350 325 400 375 350 325 400 375 350 325
[29,30] [29,30] [26,30] [30]
Habs content (wt%)
Hdes content (wt%)
H retained (wt%)
Absorption plateau pressure (MPa)
Desorption plateau pressure (MPa)
Pressure hysteresis (MPa)
5.50 5.21 4.47 4.26 5.20 4.99 4.84 4.78 6.14 5.88 5.69 5.48
5.10 5.01 4.32 e 5.10 4.85 4.64 e 5.94 5.63 5.29 4.67
0.40 0.20 0.15 e 0.10 0.14 0.20 e 0.20 0.25 0.40 0.81
1.247 0.928 0.915 0.287 1.440 0.855 0.483 0.266 1.467 0.792 0.443 0.251
1.154 0.826 0.795 e 1.371 0.799 0.434 e 1.04 0.682 0.368 0.183
0.093 0.102 0.120 e 0.069 0.056 0.049 e 0.427 0.110 0.075 0.068
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
8
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
Table 2 e Summary of the hydrogenation and dehydrogenation properties obtained for the micro-HPT and nano-HPT products. HPT products/ properties
Micro-HPT Nano-HPT
Hydrogenation properties
Dehydrogenation properties
DHabs (kJ/mol H2)
Activation energy (Ea) (kJ/mol H2)
H uptake (wt%) at 400 C
H release (wt%) from TG
Onset temperature ( C)
Activation energy (Ea) (kJ/mol H2)
77.5 75.4
70.2 43.5
6.6 5.5
5.1 4.9
402.8 380.7
170.3 209.1
kinetic parameter (k) was calculated. Based on the experimental data obtained at different temperatures, the Arrhenius equation was employed to determine the average apparent activation energies; the results are provided in Table 2. It was found that the HPT processed product obtained from the condensed ultrafine powder shows a lower value of apparent activation energy (Ea ¼ 43.5 kJ/mol H2) compared to that of the micro-HPT one (Ea ¼ 70.2 kJ/mol H2). However, both HPT products have much lower activation energy than those obtained for the initial ultrafine Mg powders (92.9 kJ/mol H2) [30] or pure Mg nanocrystals prepared from solution (115e122 kJ/ mol H2) [35].
Hydrogenated products After the PCT and kinetic measurements, the samples were hydrogenated again at 400 C for 6 h under a hydrogen pressure of 3.5 MPa. After these different cycling processes, the H uptake capacities in these hydrogenated samples were determined to be 6.6 and 5.5 wt% H for the micro-HPT and nano-HPT products, respectively (see Table 2). Part of these hydrided samples was used for characterization of the hydrogenated products by XRD analyses (this section) while other parts were used for the characterization desorption properties (next section). The XRD patterns of the hydrogenated products are shown in Fig. 8. It is apparent from these patterns that most of the hcp-Mg phase has transformed into b-MgH2 upon hydrogenation, in both HPT products. However, the presence of hcpMg peaks and some MgO are also visible in the XRD patterns. The presence of hcp-Mg indicates an incomplete hydrogenation reaction even at 400 C and for a fairly long duration which explains the inherent lower H uptake capacities revealed previously for the HPT products compared to the initial powder precursors (Fig. 5 and Table 1). It can also be noticed from the relative intensity of the peaks that the quantity of remnant Mg is higher in the micro-HPT product than in the nano-HPT one. After the different cycles that the materials have previously sustained, this is consistent with the evolutions shown in Fig. 6 and the data given in Table 1 indicating a faster decrease in storage capacity after cycling for the micro-HPT product than the nano-HPT one. Concerning the presence of MgO in the XRD traces e in particular for the micro-HPT product for which it was not detected after the HPT processing (see Fig. 3a) e it might have formed at the surface during handling of the hydrogenated products out of the glove box before the XRD analyses. However, it is also interesting to notice that, for the nano-HPT product, that the Mg(OH)2 peaks observed in the XRD patterns of ultrafine powder and nano-HPT product before
H retained (wt%)
1.5 0.6
hydrogenation (see Fig. 3b) have disappeared after hydrogenation (see Fig. 8). This is likely due to the endothermic decomposition of the Mg(OH)2 phase at high temperature (above 330 C) [36] into MgO during the H-sorption cycles, which thereby increased the MgO content in the nano-HPT product.
Kinetics of desorption The desorption behavior of the hydrogenated samples were examined using differential scanning calorimetryethermogravimetry (DSC-TG) technique operated under 0.1 MPa of purified argon after a storage cycle of 6 h at 400 C under a hydrogen pressure of 3.5 MPa. The heating rates applied in these measurements were 3, 5 and 10 C/min in the temperature range of 25e500 C. The DSC-TG curves for the heating rate of 10 C/min are displayed in Fig. 9. The most characteristic findings from the DSC-TG analysis are given in Table 2. The amount of desorbed H was calculated from the TG curves. While the micro-HPT product can desorb 5.1 wt% H, the nano-HPT product can desorb ~4.9 wt% H for the same experimental conditions. Thus, despite a lower storage capacity for the nano-HPT product, the differences between the absorbed and desorbed amounts indicates that a lower amount of H was retained after desorption in the nano-HPT (0.6 wt% H) than in the micro-HPT one (1.5 wt% H). In addition, it can also be noticed from the DSC curves that the onset temperature for desorption is much lower (380.7 C) for the nano-HPT product than for the micro-HPT one (402.8 C). These values for the HPT products are lower than the one of the initial condensed ultrafine Mg powders (423 C) [30], indicating that one of the positive effect of the HPT process is to reduce the dehydriding temperature of the Mg hydride. Besides, it is interesting to note that the slope of the TG curve (see Fig. 9a and c) after the onset of desorption is much steeper in case of micro-HPT product (slopes: 1.93 wt% H/min versus 1.65 wt% H/min). This indicates that, after a slightly longer incubation period, the hydrogenated micro-HPT product has a faster desorption rate than its hydrogenated nano-HPT counterpart. These results are consistent with the aspect of the PCT curves given in Fig. 6a which clearly reveal the necessity of higher depression (incubation) followed by fastest desorption for the micro-HPT product than in Fig. 6c for the nano-product. The Kissinger method [37], based on DSC analysis, and the related peak temperature depending on heating rate were employed to determine the activation energy of the desorption kinetics; the corresponding Kissinger plots (ln (b/T2) vs. 1000/T) for both products are shown in Fig. 9b and d. As can be
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
9
Fig. 9 e DSC and TG curves (a, c) recorded at a heating rate of 10 C/min from the hydrogenated products, and the corresponding Kissinger plots (b, d) for the micro-HPT (a, b) and nano-HPT (c, d) products.
seen from these figures, the apparent activation energy for the micro-HPT product is 170.3 kJ/mol H2, lower than the value obtained for the nano-HPT composite (209.1 kJ/mol H2), and fairly close to the results recorded for Mg nanocrystals (126e160 kJ/mol H2) [35].
Discussion The current investigation revealed that, despite a substantial amount of severe plastic deformation imparted by the HPT processing to the two types of Mg powders, the exact nature of the starting powder precursor had a pronounced effect on the microstructure development in the HPT products and in their H-sorption behaviors. These aspects are recalled and discussed hereafter in three different sections.
Microstructure of the HPT products Consistently with the recent results obtained on pure bulk Mg subjected to HPT [13,31], the occurrence of dynamic recrystallization during the HPT processing led to the formation of an equiaxed microstructure in the micro-HPT product obtained by consolidation of the micro-sized atomized powder. While the HPT consolidation of the micro-sized atomized Mg powder resulted in a recrystallized equiaxed microstructure,
the consolidation of the ultrafine condensed and passivated powder has led to the formation of an Mg-MgO based nanocomposite, characterized by a much finer grain size. The MgO particles were essentially aligned along the shear direction and their interaction with the migrating grain boundaries has prevented the occurrence of significant dynamic recrystallization during the HPT processing. Thus, the microstructure consisted essentially of fine elongated grains e about 200 nm/ 300 nm in diameter e surrounded by fine oxide dispersion in the nano-HPT product. These differences in the microstructure have generated strong modifications in the H-sorption behaviors.
Effect of the HPT processing The effect of the HPT processing has been analyzed by activating the HPT products at 400 C and their initial precursor powders under similar conditions (Fig. 5) as well as by cycling the HPT products for absorption/desorption tests carried out at different temperatures (Fig. 6), and comparing these data with those obtained in the literature on evaporated ultrafine powders using similar conditions (Table 1). For both types of powder precursors, through the introduction of structural defects, the effect of the HPT processing was to increase substantially the kinetics of H-sorption while reducing the storage capacity. The reduction in the H uptake capacities of
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
10
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
the HPT products was the consequence of incomplete hydrogenation reactions that was evidenced from the presence of un-reacted Mg phase in the XRD patterns of the hydrogenated products (see Fig. 8). On cycling, for all types of samples, the equilibrium pressure and H uptake capacities obtained from the PCT experiments decreased when the operating temperature was lowered. However, another advantage of the HPT processing is that it reduces the hysteresis between the absorption and desorption plateau pressures.
3)
Effect of the initial powder precursors 4) The overall absorption kinetics was faster for the Mg-MgO based nano-HPT composite processed from the condensed ultrafine Mg powder. Comparatively, the micro-HPT product consolidated from the atomized micro-sized powder had the advantage of a higher H-sorption capacity. This behavior can be attributed to the presence of the additional MgO phase in the nano-HPT composite. While the presence of an MgO layer is often regarded as an obstacle for H-sorption e as it prevents the H atoms from entering into the Mg substrate [12] e it is claimed that the faster H-uptake rate can be related to the presence of oxides that can serve as nucleation sites for hydride growth. In addition, it has also been reported in Ref. [9] that if MgH2 is well intermixed with MgO, an improved Hsorption kinetics is obtained due to the reduced particle size of MgH2 and some catalytic effects of MgO particles. In terms of storage, the only disadvantage with MgO noticed in the present study is that it has impaired the storage capacity in the nano-HPT composite in comparison with the micro-HPT product. In terms of desorption, after some incubation, the microHPT product revealed a faster desorption rate, accompanied with much lower activation energy. It is also interesting to note, however, that the hydrogenated nano-HPT material showed a lower onset temperature of desorption and a relatively lower peak maximum in the DSC analyses. This suggests that some trapping sites were less effective in this type of consolidated product. Further analysis is under way to clarify if differences exist in the character of the grain boundaries between the two consolidated products.
Conclusions A comparative study of H-sorption behaviors was conducted on two types of HPT products obtained from the consolidation of two kinds of Mg powder precursors: atomized micro-sized powder, and condensed and passivated ultrafine powder particles. The deformed microstructures were characterized, and the H-sorption properties were investigated. The main findings of the present study can be summarized as follows. 1) HPT consolidation applied on the atomized micro-sized Mg powder developed a dynamically recrystallized microstructure with fairly equiaxed 0.5e1 mm grains in the micro-HPT product. 2) The HPT consolidation applied on the condensed and passivated ultrafine powder led to the formation of an Mg-
5)
6)
MgO based nanocomposite, the so-called nano-HPT product, within which slightly elongated 200 nm/300 nm thick grains were pinned by the nano-sized MgO oxide particles. For both types of powder precursors, through the introduction of structural defects and microstructural refinement, the HPT processing has provided with significant improvements of the hydrogenation kinetics for the HPT products compared to their initial powder precursors. Another advantage of the HPT processing is to reduce the hysteresis between the absorption and desorption plateau pressure during the PCT experiments. The major drawback of the HPT processing was, for both types of powder precursors, that it reduced the maximum H uptake capacities. The nano-HPT product showed faster absorption kinetics accompanied with a much lower activation energy (43.5 kJ/ mol H2), which was attributed to its nanostructural features and to the catalytic effect of the MgO oxide particles. Comparatively, the micro-HPT product had the advantage of having higher H-sorption capacity. Despite a slightly lower onset temperature of desorption, the nano-HPT product exhibited slower desorption kinetics and a higher activation energy than the micro-HPT product.
Acknowledgements This work was supported by the French State through the program “Investment in the future” operated by the National Research Agency (ANR) and referenced by ANR-11-LABX-000801 (Labex DAMAS). This work was also partly supported by the National Key Research and Development Program of China (2016YBF0701203), Science and Technology Commission of Shanghai Municipality (14JC1491600) and “111” project from China's Ministry of Education (B16032).
references
[1] Sakintuna B, Lamari-Darkrim F, Hirscher M. Metal hydride materials for solid hydrogen storage: a review. Int J Hydrogen Energy 2007;32:1121e40. [2] Jain IP, Lal C, Jain A. Hydrogen storage in Mg: a most promising material. Int J Hydrogen Energy 2010;35:5133e44. [3] Reiser A, Bogdanovic B, Schlichte K. The application of Mgbased metal-hydrides as heat energy storage systems. Int J Hydrogen Energy 2000;25:425e30. [4] Corgnale C, Hardy B, Motyka T, Zidan R. Metal hydride based thermal energy storage system requirements for high performance concentrating solar power plants. Int J Hydrogen Energy 2016;41:20217e30. [5] Jain IP. Hydrogen the fuel for 21st century. Int J Hydrogen Energy 2009;30:7368e78. [6] Liang G, Huot J, Boily S, Van Neste A, Schulz R. Hydrogen storage properties of the mechanically milled MgH2eV nanocomposite. J Alloys Compd 1999;291:295e9. [7] Oelerich W, Klassen T, Bormann R. Metal oxides as catalysts for improved hydrogen sorption in nanocrystalline Mg-based materials. J Alloys Compd 2001;315:237e42.
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 7 ) 1 e1 1
[8] Barkhordarian G, Klassen T, Bormann R. Fast hydrogen sorption kinetics of nanocrystalline Mg using Nb2O5 as catalyst. Scr Mater 2003;49:213e7. [9] Aguey-Zinsou KF, Nicolaisen T, Fernandez JA, Klassen T, Bormann R. Effect of nanosized oxides on MgH2 (de) hydriding kinetics. J Alloys Compd 2007;434:738e42. [10] Aguey-Zinsou KF, Fernandez JA, Klassen T, Bormann R. Using MgO to improve the (de) hydriding properties of magnesium. Mater Res Bull 2006;41:1118e26. [11] Cakmak G, Ozturk T. ECAP processing and mechanical milling of Mg and MgeTi powders: a comparative study. J Mater Sci 2011;46:5559e67. [12] Ares JR, Aguey-Zinsou KF, Klassen T, Bormann R. Influence of impurities on the milling process of MgH2. J Alloys Compd 2007;434:729e33. [13] Edalati K, Yamamoto A, Horita Z, Ishihara T. High-pressure torsion of pure magnesium: evolution of mechanical properties, microstructures and hydrogen storage capacity with equivalent strain. Scr Mater 2011;64:880e3. [14] Skripnyuk VM, Rabkin E, Estrin Y, Lapovok R. The effect of ball milling and equal channel angular pressing on the hydrogen absorption/desorption properties of Mge4.95 wt% Zne0.71 wt% Zr (ZK60) alloy. Acta Mater 2004;52:405e14. [15] Leiva DR, Fruchart D, Bacia M, Girard G, Skryabina N, Villela AC, et al. Mg alloy for hydrogen storage processed by SPD. Int J Mater Res 2009;100:1739e46. [16] Botta WJ, Jorge Jr AM, Veron M, Rauch EF, Ferrie E, Yavari AR, et al. H-sorption properties and structural evolution of Mg processed by severe plastic deformation. J Alloys Compd 2013;580:S187e91. [17] Lima GF, Triques MR, Kiminami CS, Botta WJ, Jorge AM. Hydrogen storage properties of pure Mg after the combined processes of ECAP and cold-rolling. J Alloys Compd 2014;586:S405e8. [18] Grill A, Horky J, Panigrahi A, Krexner G, Zehetbauer M. Longterm hydrogen storage in Mg and ZK60 after severe plastic deformation. Int J Hydrogen Energy 2015;40:17144e52. [19] Korznikov AV, Safarov IM, Laptionok DV, Valiev RZ. Structure and properties of superfine-grained iron compacted out of ultradisperse powder. Acta Mater 1991;39:3193e7. [20] Valiev RZ, Mishral RS, Grozal J, Mukherjee AK. Processing of nanostructured nickel by severe plastic deformation consolidation of ball-milled powder. Scr Mater 1996;34:1443e8. [21] Kusadome Y, Ikeda K, Nakamori Y, Orimo S, Horita Z. Hydrogen storage capability of MgNi2 processed by high pressure torsion. Scr Mater 2007;57:751e3. [22] Leiva DR, Jorge AM, Ishikawa TT, Huot J, Fruchart D, Miraglia S, et al. Nanoscale grain refinement and H-Sorption properties of MgH2 processed by high-pressure torsion and other mechanical routes. Adv Eng Mater 2010;12:786e92.
11
nya Z, Verebelyi T, Szabo PJ, Zhilyaev AP, [23] Revesz A, Ka Spassov T. The effect of high-pressure torsion on the microstructure and hydrogen absorption kinetics of ballmilled Mg70Ni30. J Alloys Compd 2010;504:83e8. [24] Liang J, Hout J. A new approach to the processing of metal hydrides. J Alloys Compd 2011;509:L18e22. [25] Zou JX, Perez-Brokate CF, Arruffat R, Bolle B, Fundenberger JJ, Zeng XQ, et al. Nanostructured bulk MgþMgO composite synthesized through arc plasma evaporation and high pressure torsion for H-storage application. Mater Sci Eng B 2014;183:1e5. [26] Grosdidier T, Fundenberger JJ, Zou JX, Pan YC, Zeng XQ. Nanostructured Mg based hydrogen storage bulk materials prepared by high pressure torsion consolidation of arc plasma evaporated ultrafine powders. Int J Hydrogen Energy 2015;40:16985e91. [27] Shao H, Wang Y, Xu H, Li X. Hydrogen storage properties of magnesium ultrafine particles prepared by hydrogen plasma-metal reaction. Mater Sci Eng B 2004;110:221e6. [28] Zou JX, Sun HQ, Zeng XQ, Ji G, Ding WJ. Preparation and hydrogen storage properties of Mg-rich Mg-Ni ultrafine particles. J Nanomater 2012;2012:2. [29] Stephane P, Zou JX, Zeng XQ, Sun HQ, Ding WJ. Preparation and hydrogen storage properties of ultrafine pure Mg and Mg-Ti particles. Trans Nonfer Met Soc China 2012;22:1849e54. [30] Zou JX, Zeng XQ, Ying Y, Chen X, Guo H, Zhou S, et al. Study on the hydrogen storage properties of coreeshell structured MgeRE (RE¼Nd, Gd, Er) nano-composites synthesized through arc plasma method. Int J Hydrogen Energy 2013;38:2337e46. [31] Panda S, Toth LS, Fundenberger JJ, Perroud O, Guyon J, Zou JX, et al. Analysis of heterogeneities in strain and microstructure in aluminum alloy and magnesium processed by high-pressure torsion. Mater Charact 2017;123:159e65. [32] Bogdanovic B. Magnesium hydride: a homogeneouscatalysed synthesis and its use in hydrogen storage. Int J Hydrogen Energy 1984;9:937e41. [33] Mintz MH, Zeiri Y. Hydriding kinetics of powders. J Alloys Compd 1995;216:159e75. [34] Kelton KF. Analysis of crystallization kinetics. Mater Sci Eng A 1997;226:142e50. [35] Norberg NS, Arthur TS, Fredrick SJ, Prieto AL. Size-dependent hydrogen storage properties of Mg nanocrystals prepared from solution. J Am Chem Soc 2011;133:10679e81. [36] Burke PJ, Bayindir Z, Kipouros GJ. X-ray photoelectron spectroscopy (XPS) investigation of the surface film on magnesium powders. Appl Spectrosc 2012;66:510e8. [37] Kissinger HE. Reaction kinetics in differential thermal analysis. Anal Chem 1957;29:1702e6.
Please cite this article in press as: Panda S, et al., Effect of initial powder type on the hydrogen storage properties of high-pressure torsion consolidated Mg, International Journal of Hydrogen Energy (2017), http://dx.doi.org/10.1016/j.ijhydene.2017.05.097