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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Influence of the ECAP and HEBM processes and

the addition of Ni catalyst on the hydrogen storage
properties of AZ31-x Ni (x¼0,2,4) alloy

Song-Jeng Huang, Veeramanikandan Rajagopal*, Addisu Negash Ali

Department of Mechanical Engineering, National Taiwan University of Science and Technology, No. 43, Keelung Rd.,
Sec. 4, Da'an Dist., Taipei, Taiwan, ROC

article info abstract

Article history: In this work, the hydrogen sorption properties of AZ31 magnesium alloys with various
Received 6 September 2018 additions of nickel (Ni) (i.e., Ni (X ¼ 0, 2, 4) wt.%) were investigated. Cast ingots with
Received in revised form different AZ31/NiP compositions were fabricated using a gravitating mechanical stir cast-
22 October 2018 ing (GMSC) method. Two different processes, namely, equal channel angular pressing
Accepted 4 November 2018 (ECAP) and high energy ball milling (HEBM) were performed to improve its hydrogen
Available online 5 December 2018 storage properties. The particle size of the sample powders were measured by laser
diffraction analysis. The microstructures, powder morphologies, and phase transformation
Keywords: were characterized by optical microscopy (OM), scanning electron microscopy (SEM) and x-
Hydrogen storage ray diffraction (XRD) analysis. The hydrogen sorption kinetics were measured by a Sievert's
AZ31-Mg alloy type apparatus. The results showing the impact of the ECAP and HEBM processes on the
Ni catalyst hydrogen storage properties were compared. The AZ31-4 Ni ECAP processed sample
Equal channel angular pressing, showed a maximum hydrogen storage capacity of 7.0 wt.% at 2322 s with complete
(ECAP) desorption of all the hydrogen in less than 5 min at a temperature of 375  C. On the other
High energy ball milling (HEBM) hand, the pure AZ31 alloy which was treated with the HEBM process showed the maximum
hydrogen capacity of 6.5 wt.% at 2393 s with desorption of all the hydrogen within 6 min. In
addition, the activation energy, as illustrated by the Kissinger plot, revealed that the
activation energy of the ECAP processed AZ31-4 Ni was 104.73 (KJ/mol), obviously lower
than the material processed by HEBM and pure MgH2.
© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

the automobile industry as a source of green energy.

Introduction Furthermore, hydrogen is also considered as a potential
source of energy for future energy systems. However, the
Increasing demand for energy to meet the requirements of general use of hydrogen is presently obstructed by its storage
modern society drives efforts to discover and develop new technology, especially for applications such as automobiles
sources of sustainable energy. Hydrogen is a feasible ecolog- powered by hydrogen, fueled internal ignition engines. One
ically friendly substitute for fossil fuels that could be used in solution would be the deposition of hydrogen in metal

* Corresponding author.
E-mail address: [email protected] (V. Rajagopal).
https://doi.org/10.1016/j.ijhydene.2018.11.005
0360-3199/© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
1048 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

hydrides, a method which exhibits excellent characteristics been examined by monitoring the absorption and desorption
such as a large volumetric storage capacity and safety pro- properties. The desorption occurred at a faster rate at a tem-
tection [1,2]. perature of 300  C with a pressure of 2 atm. A 5.2 wt.%
MgH2 is one of the most attractive metal hydrides for hydrogen release was observed at 20 min in the materials
hydrogen storage owing to its higher theoretical gravimetric subjected to both ECAP and ball milling and 26 min in the
capacity of 7.6 wt.%, and the abundance and reuseability of Mg materials subjected to ECAP alone [19]. Thus, past studies
make it an economical [3,4]. Nevertheless, the practical show that ECAP and high energy ball milling (HEBM) can
implementation of MgH2 for hydrogen storage has not been enhance the hydrogen storage properties of Mg and Mg alloys.
attained because of its lower hydrogen absorption/desorption In this work, cast ingots composed of AZ31/NiP alloys were
kinetics, and larger thermodynamic stability. To achieve bet- produced by the gravitating stir casting method. Both ECAP
ter hydrogenation/dehydrogenation kinetics, the metal hy- plastic deformation and HEBM milling processes were per-
drides have to be heated up to a temperature of 673K with a formed to produce defects along the grain boundaries. The
hydrogen pressure of 3 MPa [3]. Furthermore, the addition of colliding of the particles in the powders is expected to
various catalysts to Mg metal can have a significant effect on enhance the kinetics of sorption. The hydrogenation proper-
stimulating the hydrogen sorption kinetics. In particular, the ties depend on the microstructures and the weight percent-
transition metals (Ni, Fe, Cu, Ti, V) can be used as a catalyst to ages of the Ni contents. Hence, the hydrogenation results
increase the sorption kinetics of hydrogen molecules on the were comparatively studied among the two processing effects
surface of Mg and its alloy materials. The hydrogen molecules with similar particle size distribution for different alloys by
diffuse quickly into the Mg crystal lattice leading to an constant volume method.
improved rate of hydrogenation and dehydrogenation.
Notably, the addition of the catalyst enhances the hydrogen
storage performance even at a moderate temperature [5e7]. Experimental procedure
Among the transition metals, Ni has been utilized to decrease
the thermodynamic stability of metal hydride [8e10]. Materials and methods
Currently, some researchers have focused on the adaption
of deformation methods, in particular the equal channel Material preparation
angular pressing (ECAP) technique to treat and deform bulk The Mg-based AZ31 alloy was obtained from Kuangyue Co.,
alloy materials for hydrogen storage [11,12]. In 2015, Asselli Ltd. and the 99.90% pure Ni-Particles (Nip) were purchased
et al. reported the hydrogen storage kinetics of a ZK60 alloy from First Chemical Works Co., Ltd. Nip with particles with
produced through the ECAP process and subsequent accu- ~25 mm in size was used as a catalyst with the AZ31 alloy
mulative roll-bonding (ARB) process [13]. In another study of a (AZ31/Nip). Samples with three different compositions, 0 wt.%
commercial ZK60 alloy, the grain size was decreased to of Nip (AZ31-0 Ni), 2 wt.% of Nip (AZ31-2 Ni) and 4 wt.% of Nip
250 nm by the ECAP process which is the lowest size value (AZ31-4 Ni), were fabricated through the gravitating me-
ever reached by this deformation method. More importantly, chanical stir casting (GMSC) method. During the casting pro-
this material showed a maximum hydrogen storage capacity cess, CO2þSF6 gases were supplied at 400  C to avoid burning
of 6.6 wt.%. An investigation of the durability of the materials AZ31 magnesium alloy and argon gas was also supplied at
showed no decline in storage capability or kinetics for up to 700  C to prevent oxidation. A stabilization time of 15 min for
1000 hydrogen-sorption cycles [14]. After subjecting Mg2Ni every 100  C incremental increase in temperature up to 700  C
monocrystalline powders to plastic strain by the ECAP and was maintained during the casting process. After 760  C the
cold rolling process it was concluded that the microstructure AZ31/Nip melt was stirred with two stirrers for 5 min at
and the nanostructure are directly correlated to the hydrogen 300 rpm to attain a uniform distribution of Nip throughout the
storage behaviors [15]. In order to attain sub-micron sized Mg alloy. Finally the melt alloy was poured into a mold located
micro grains in Mg2Ni alloys and Mg from the Mg-Ni, the alloy inside the furnace chamber to produce the AZ31/Nip cast in-
materials are first treated using the ECAP process, following by gots [20,21].
the cold rolling technique. The obtained material shows a Sample billets with dimensions of 11.5  11.5  75 mm
maximum hydrogen storage capacity of about 6 wt.% [16]. were sliced from the cast ingots and then kept in a furnace at
Others have focused on the effects of mechanical milling. a temperature of 410  C for 24 h for homogenization treat-
Composite alloys of various compositions were prepared with ment. The microhardness was measured using a Vickers
particle sizes lower than 400 nm by hydriding combustion microhardness tester (Akashi MVK-H1). After homogeniza-
synthesis followed by the mechanical milling process. In the tion, the sample billets were plastically deformed through the
Mg-Ni composites, the Mg2Ni compound dispersed uniformly equal channel angular pressing (ECAP) technique with a
results enhanced the hydrogen-sorption rate and capacity channel angle of 120 with 8 passes and route type Bc (rotation
even at low temperature. The quantity of Ni catalyst materials of 90 in the same direction for each pass) after 30 min of
also has a direct influence on the capacity and kinetics [17]. In preheating time. After each squeeze, samples were subse-
another study, alloys of different composition were synthe- quently immersed in water. The temperature of the ECAP
sized by mechanical alloying with varying milling times. The mold was sustained at 300  C for all passes. The temperature
Ni distribution over Mg after various ball milling times (15 and oscillation of the die was ±5  C. To remove the oxide layer
30 h) was evaluated. The results were based on the dispersion after ECAP process the surfaces of the sample billets were
of the Ni content and the milling time [18]. The combined ef- polished by silicon carbide polishing sheets. Then the solid
fect of ECAP and ball milling on a ZK60 magnesium alloy has billets were comminuted into powder by a filing process 1 h
 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8 1049

prior to hydrogenation to reduce the possibility of oxide layer experimental and the hydrogenation mechanisms are sum-
formation. marized in Scheme 1.
In the HEBM process, homogenized samples were rasped
manually then milled in a planetary type machine (Retsch
PM100) with a ball to power ratio of 30:1 in an inert atmo- Results and discussion
sphere to prevent oxidation. The rotation speed milling was
maintained at 300 rpm for a milling time of 4 h for the AZ31- Microhardness
0 Ni sample, and 3 h for the AZ31-2 Ni and AZ31-4 Ni (AZ31-
Ni) samples. The powder transmission process was carried The effects of adding the catalyst on the microhardness re-
out inside a glove box in an inert atmosphere. sponses of AZ31/Nip alloys were investigated. The micro
hardness values (HV) of the pure AZ31 Mg alloy, 2 and 4 wt.%
Characterization Ni magnesium alloys in as-cast conditions were 46.08 ± 3.70,
Particle size distributions were measured by a Malvern mas- 57.50 ± 4.41 and 60.82 ± 4.04 HV and for homogenized samples
tersizer model 2000. The microstructure of the solid samples the micro hardness values were 43.14 ± 3.62, 47.04 ± 2.39 and
was observed by optical digital microscopy and the powder 50.62 ± 3.32, respectively. In general, the microhardness of the
morphologies were obtained through scanning electron mi- samples showed higher values for AZ31-Ni as compared to
croscopy (SEM) (model JSM-6390LV) equipped with an EDS pure AZ31. Increasing the weight percentage of Nip had a
(Energy dispersive spectroscope) before and after hydrogena- significant influence on the microhardness showing an
tion. The phases of the prepared alloys were observed through increasing trend. After homogenization the hardness of the
XRD analysis (model Bruker D2 phaser with Cu Ka radiation). materials was apparently lower than the as-cast condition
which could reduce the phenomenon of fracturing during
Hydrogen storage kinetics plastic deformation.
The hydrogenation kinetic properties were investigated using
a Sivert's type volumetric apparatus with a fixed volume. The Microstructural analysis
volume of the system 0.0000217 m3 was calculated while the
0.1 g of powder samples inside the chamber. During the acti- To inspect the microstructures, optical digital microscopic
vation, temperature was elevated to 375  C from room tem- (OM) analysis was carried out for the samples of AZ31-0 Ni,
perature while the powder samples were placed inside the AZ31-2 Ni, and AZ31-4 Ni under three conditions, namely cast,
chamber. The heating process was conducted at the rate of homogenized, and ECAP processed samples, as displayed in
5  C/min for 80 min with a hydrogen atmosphere. The ab- Fig. 1. The As-cast AZ-Mg alloys generally exhibited Al12Mg17
sorption rate during hydrogenation was calculated at 375  C as a secondary phase which is known as the brittle phase
under 3.5 MPa hydrogen pressure while the desorption con- which caused a higher micro hardness. Here, the homogeni-
ducted at the pressure of 0.8 MPa. The average of five stabi- zation process was performed at 410  C for 24 h to increase the
lized cycle values was taken in the calculation to get precise ductility during ECAP deformation. Fig. 1(a) and (b) show the
results. The properties of thermal decomposition were char- microstructure of the as-cast-AZ31 alloy before and after ho-
acterized by DSC (Differential scanning calorimetry) with N2 mogenization. The alloy contains cluster-like Al12Mg17 as
atmosphere. The activation energy of AZ31-4 Ni was calcu- shown in Fig. 1(a) while Fig. 1(b) shows that the secondary
lated by the Kissinger equation method. Furthermore, the phase has dissolved. During the homogenization process, the

Scheme 1 e Diagram outlining preparation and hydrogenation mechanisms of the alloy materials.
1050 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

Fig. 1 e OM microstructure of: (a) as-cast AZ31, (b) homogenized AZ31, (c) ECAPed AZ31 (8pass), (d) as-cast AZ31-2 Ni (e)
homogenized AZ31-2 Ni, (f) ECAPed AZ31-2 Ni (8 pass), (g) as-cast AZ31-4 Ni and (h) homogenized AZ31-4 Ni, (i) ECAPed AZ31-
4 Ni.

elevated temperature increased the Al solubility which caused boundaries and higher defect density assists in the improve-
the Al12Mg17 to dissolve into the matrix Mg phase [22]. ment of the rate of hydrogenation.
Fig. 1(d) and (g) clearly show that the addition of Nip during The energy dispersive spectrometry (EDS) elemental anal-
the casting process produced the AlNi phase which agglom- ysis of surface compositions for the different cast materials
erated on the surface of the a-Mg phase. The addition of nickel are shown in Fig. 2. Fig. 2(a and b) shows highly concentrated
contents to the liquefied Mg-Al in the temperature range from secondary phases containing the Mg,Al,Zn, Ni elements which
650 C to 900  C with different wt.% of (110) Al can produce confirms the compositions of pure AZ31. The detection of Mg
phase precipitation of the AlNi. This confirms the dissolution and Al with larger amount exhibit the presence of Al12Mg17
reaction between the Al and Ni elements that were present in phase. The spectrum for the AZ31-Ni alloys are presented in
the casting crucible [23]. In addition, the presence of the Fig. 2(cef) from observations for the alloy AZ31-Ni (2,4 wt.%.),
Al12Mg17 phase also confirms the reaction between the Mg-Al detection of Al, Ni elements as a major peak for a particular
elements. As it can be seen from Fig. 1(e) and (h), Al12Mg17 phase indicates that the possibilities of AlNi phase on the
became immersed into the Mg phase, with some precipitation surface. Hence the addition of Ni element produces the AlNi
of AlNi on the surface of the magnesium. As shown in Fig. 1(c), phase, evident that the reaction between Al and Ni in the cast
(f) and (i), the grain sizes were refined through the plastic crucible.
deformation process and this created more grain boundaries. Scanning electron microscopy was performed to analyze
After 8 passes of ECAP, the uniformity of grain refinement was the features of the surface morphology of the prepared AZ31/
higher for the alloy with the Nip addition compared to the pure Nip alloy powders; see the images presented in Fig. 3. Fig. 3(a)
one. Furthermore, observations showed a large number of shows the AZ31/Nip alloy after being treated through the ECAP
coarse grains surrounded by refined grains with more defects process combined with manual filing, Observe the defined
induced through deformation. Similar defects were detected elongated nature of the structures in the powder, which are
in the microstructure of the AZ31 that was processed by the less than a hundred micrometers. In general, the hydrogen
ECAP process [24]. The existence of large numbers of grain storage performance of the alloys showed higher efficiency
 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8 1051

Fig. 2 e EDS analysis results: (a,b) as-cast AZ31-0 Ni, (c,d) as-cast AZ31-2 Ni, (e,f) as-cast AZ31-2 Ni.

with the ECAP and filing process than without the filing pro- morphology of the alloy powders upon hydrogenation treat-
cess due to its higher surface to volume ratio produced by ment is presented in Fig. 3(c) and (d), for the ECAP-processed
filing [14,25]. Such a process produces a synergistic effect for and ball milled samples, respectively. The hydrogenated
proficient diffusion of the hydrogen molecules which leads to filed sample appears to be fragmented with a significant
favorable hydrogenation properties. It is suspected that ball reduction in particle size. The ball milled and hydrogenated
milling also helps to produce an effective surface to volume samples had rough surfaces due to the occurrence of nucle-
ratio, as indicated by the morphology of the ball milled AZ31/ ation throughout the surface of the alloys. The rough surface
Nip alloys displayed in Fig. 3(b). This material contains parti- is clearly visible in Fig. 3(e). There are two reasons for this: (1)
cles quite similar in size with smoother surfaces than the filed the metal matrix changes into a tetragonal MgH2 lattice
powders. structure from a hexagonal Mg structure, that produces
The SEM results of the filed alloy powders revealed saw- changes in the lattice dimensions; (2) the unequal pressure
toothed surfaces and the ball milled powders had smooth concentration engendered by the pressurized hydrogen gas
surfaces decorated with tiny particles. The evolution of the nucleation, leads to ruptures on the surface.
1052 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

Fig. 3 e SEM morphology of the AZ-Alloy (a) ECAP: before hydrogenation, (b) HEBM: before hydrogenation, (c) ECAP: after
hydrogenation, (d) HEBM: after hydrogenation, (e) magnified image after hydrogenation.

Particle size distribution analysis influence of the surface area. Fig. 4 presents the particle size
distribution curves in the same region. The observations show
The average particle size values and particle size distributions the average particle values to be in the range of 45e60 mm.
were evaluated using a laser diffraction method before hy- From the analysis, it can be seen that the increment in the Ni
drogenation. Table 1 shows the average particle sizes for the particle volume influences the hardness and the particle size
obtained AZ31/Nip alloys that were filed after the ECAP pro- values in terms of the milling time.
cess and HEBM, after a milling time of 4 h for pure AZ31 and
3 h for the AZ31-Ni alloys. Different milling times were XRD pattern analysis
required in order to obtain similar particle sizes to reduce the
X-Ray diffraction analysis was carried out for different con-
ditions such as-cast, homogenized, before and after hydro-
Table 1 e Average particle size of the AZ-alloy powder. genation on samples with three different compositions to
S. no Sample Process Average particle determine the different phases. The XRD patterns of the AZ31-
name size (mm) 0 Ni (wt. %) alloys are presented in Fig. 5. From the observa-
1 AZ31-0 Ni ECAP 51.709
tions, it can be seen that the AZ31 alloys exhibited sharp Mg
2 AZ31-0 Ni HEBM 53.927 peaks as a major phase, ascribed to the good crystallization
3 AZ31-2 Ni ECAP 50.861 phenomenon attained during the casting process. The exis-
4 AZ31-2 Ni HEBM 47.898 tence of the Mg peaks after hydrogenation confirmed the
5 AZ31-4 Ni ECAP 51.512 incompleteness of the reaction between Mg and hydrogen to
6 AZ31-4 Ni HEBM 58.923
produce MgH2.
 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8 1053

shows that the AlNi phase dissolved into the matrix during
the ECAP process; see Fig. 6(a, c). The phenomenon of dynamic
recrystallization (DRX) and the thermal energy produced
during ECAP could have caused the dissolution of AlNi into the
Mg matrix. After squeezing, the samples were subsequently
dipped into water which could stop the phase evolution. In
contrast, the XRD results show the presence of the AlNi phase
for ball milling. It is evident that the energy produced during
high energy ball milling does not affect the AlNi thus it is
present in the matrix before hydrogenation. The AlNi phase
can be identified even after hydrogenation; see Fig. 6(b, d). The
presence of the AlNi phase after hydrogenation indicates that
AlNi does not react with the hydrogen molecules. The ob-
tained XRD patterns reveal that the a-Mg phase was trans-
formed into the b-MgH2 phase during the hydrogenation
process, and the Mg2NiH4 phase was also found to have a very
Fig. 4 e Particle size distribution of AZ-alloy powders. small intensity in the AZ31-4 Ni alloy. The alloy powders
processed by ball milling contained a higher intensity of a-Mg
after hydrogenation as compared to the alloys that were
processed by the ECAP technique.
The XRD results for the AZ31-2 Ni, AZ31-4 Ni alloys before
and after hydrogenation are exhibited in Fig. 6. AZ31 alloyed Hydrogenation properties
with Ni, can be seen to have compounds of pure Mg, Al12Mg17,
AlNi, Mg2Ni as major phases in the as-cast condition which The hydrogen storage and kinetic properties of pure AZ31 and
confirms the presence of Mg-Al-Ni elements while the AZ31-Ni alloys which were treated through different methods
Al12Mg17 phase has vanished during homogenization. From such as ECAP and ball milling were measured. The effects of
past studies, it is known that the addition of Ni to a Mg-Al alloy the different processing procedures and the addition of Ni
with 3.3% Al content will have a solubility of 0.65% Ni in the particles were compared. The kinetic properties were calcu-
Mg matrix. The residual Ni element reacts with Al to produces lated at the same temperature of 375  C and a similar particle
the AlNi phase in the liquid state [23]. Here, in the AZ31 with size distribution. Fig. 7 shows the absorption and desorption
Ni alloys, it is noted that the AlNi phase identified by XRD characteristics of the AZ31/Nip alloy samples. From the ob-
verifies the reaction between the Al and the Ni and affects the servations, it can be noted that the samples treated thorough
solubility of Ni in the Mg matrix. It is also found that the Mg2Ni ECAP show more stable absorption than the powdered sam-
phase has a tiny intensity which is evidence of the reaction of ples milled by the HEBM process; see Fig. 7(a, c, e). The milled
excess Ni with the major Mg. After the homogenization pro- alloy materials show a tendency for hydrogenation to prog-
cess, the AlNi phase is still present in the AZ31-Ni alloys, with ress, indicating that the reaction is still not finished. Hence,
small changes in the intensity. The Al12Mg17 phase is found the ECAP process significantly influences the rate of the re-
after plastic deformation (ECAP) and mechanical ball milling action. Materials treated with the ECAP process (with the Bc
in pure AZ31 and in the alloy with the addition of Ni. The route) were deformed in three planes which produces the
addition of Ni into the Mg-Al alloy allowed this to form, and more grain boundaries causing the grain structure to be
enhanced the quantity of the Al12Mg17 phase in the matrix refined uniformly [27]. The hydrogenation consists of three
[26]. Detection of the AlNi phase with its negligible intensity main steps: nucleation, diffusion of hydrogen atoms into the

Fig. 5 e XRD patterns of pure AZ31: (a) ECAP and (b) HEBM.
1054 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

Fig. 6 e XRD patterns of AZ31 with Ni: (a) AZ31-2 Ni: ECAP, (b) AZ31-2 Ni: HEBM, (c) AZ31-4 Ni: ECAP, (d) AZ31-4 Ni: HEBM.

Mg matrix and reaction between the magnesium and of Mg2NiH4 which enhances the hydrogenation kinetics.
hydrogen. During deformation, the dislocation density in- However, the attainment of saturated hydrogenation for
creases with an increasing number of passes to produce a Mg2NiH4 is extremely difficult even after 450 min at high
larger quantity of dislocations [28,29]. The produced disloca- temperature, as observed through in-situ XRD [32]. In the case
tions act as trapping sites [30], while the grain boundaries act of the ball milled sample, the AlNi phase does not dissolve
as the nucleation paths for the hydrogen diffusion which during this milling time. A non-reacting AlNi phase influences
further enhances nucleation [13]. Thus, processing with the the hydrogenation capacity. It can be seen that the addition of
ECAP method encourages the kinetics of hydrogenation and Ni could increase the AlNi phase in the matrix causing a
increases the volume of hydrogen reaction to magnesium for reduction in capacity. Therefore, the maximum absorption
the same reaction time. capacity of 6.5 wt.% was observed for the pure AZ31 alloy in
The hydrogen storage capacity and reaction time for which the compound AlNi does not exist in the matrix. For
different compositions of AZ31/Nip alloys are shown in AZ31-2 Ni, the maximum capacity reached was 6.4 wt.%.
Table 2. From the XRD patterns, it can be seen that the AlNi Similarly, AZ31-4 Ni which had a higher Ni content, had the
phase dissolves into the matrix during the ECAP process lowest hydrogen storage capacity of 6.2 wt.% after 2359 s. It
which is considerably produces the Mg-solid solution. It is can be seen that the material subjected to the ECAP process
presumed that the dissolved solid solution phase influences demonstrated comparatively better results than that the
the quantity of magnesium involved in the reaction. It can be material processed by high energy ball milling.
observed from the absorption capacity of the different mate-
rials that an increase in the Ni content in the alloy increases Activation energy
the magnesium solid solution during the deformation process
which could enhance the quantity of Mg available to react Additionally, the thermal decomposition properties were
with the hydrogen thereby increasing the kinetics. Hence, the calculated for the AZ31-4 Ni alloy after deformation by ECAP
volume of reactive magnesium (R-Mg) is comparatively larger and high energy ball milling. In order to determine the influ-
for the AZ31-4Ni alloy than other materials for a specific re- ence of different processing and catalytic effects on the acti-
action time. It reaches the maximum capacity of 7.0 wt.% in vation energy, the alloys (AZ31-4 Ni -ECAP and HEBM) were
2322 s which is faster than in a previous study [31] and the hydrogenated for preparation of hydrides. Then the powdered
entire amount of hydrogen is desorbed in less than 5 min. The materials were subjected to differential scanning calorimetric
reaction between the Mg- Ni elements produces a tiny amount analysis in the temperature range of 30e500  C. Three types of
 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8 1055

Fig. 7 e Kinetic curve of AZ31-alloys: ECAP VS HEBM: (a) AZ31-0 Ni absorption, (b) AZ31-0 Ni desorption, (c) AZ31-2 Ni
absorption, (d) AZ31-2 Ni desorption, (e) AZ31-4 Ni absorption, (f) AZ31-4 Ni desorption.

Table 2 e Hydrogen storage absorption capacity and reaction time for AZ31/Ni alloys with different compositions.
Sample Temperature  C Absorption Desorption Absorption Desorption
capacity (wt.%) capacity (wt.%) time (s) time (s)
AZ31-0 Ni ECAP 375 6.5 ± 0.04 6.5 ± 0.05 2256 ± 350.65 355 ± 47.97
AZ31-2 Ni ECAP 375 6.7 ± 0.04 6.7 ± 0.04 2152 ± 371.9 260 ± 15.84
AZ31-4 Ni ECAP 375 7.0 ± 0.04 7.0 ± 0.04 2322 ± 196.01 294 ± 34.51
AZ31-0 Ni HEBM 375 6.5 ± 0.05 6.5 ± 0.04 2393 ± 174.34 332 ± 17.27
AZ31-2 Ni HEBM 375 6.4 ± 0.07 6.4 ± 0.00 2390 ± 113.96 287 ± 19.32
AZ31-4 Ni HEBM 375 6.2 ± 0.04 6.2 ± 0.04 2359 ± 187.05 253 ± 31.10
1056 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

Fig. 8 e DSC curves: (a) AZ31-4 Ni-ECAP, (b) AZ31-4 Ni-HEBM.

Fig. 9 e Kissinger plots: (a) AZ31-4 Ni- ECAP, (b) AZ31-4 Ni-HEBM.

changes in temperature rate were followed in (DSC), namely, alloy showed a maximum hydrogen storage capacity of
15  C, 20  C, 25  C/min, as shown in Fig. 8. The activation en- 7.0 wt.% within 2322 s and all the hydrogen was desorbed in
ergies of the prepared alloys were calculated by Kissinger less than 5 min at a temperature of 375  C. The reason for the
plots. The Kissinger plots of the AZ31-4 Ni alloy are shown in high hydrogen storage and rapid release of AZ31-4 Ni was
Fig. 9. The dehydrogenation behavior of the material is rele- presumed that the dissolving of AlNi phase into the matrix
vant to the activation energy barrier for the release of H2 gas during the ECAP process. It is presumed that this influenced
from the MgH2. It can be seen that the alloy AZ31-4 Ni and enhanced the quantity of reactive magnesium (R-Mg)
deformed by ECAP has activation energy of 104.73 (KJ/mol), which was involved in the reaction, causing the increasing
and for the same alloy (AZ31-4 Ni) which was milled by HEBM tendency upon increase of the Ni contents. Furthermore,
process shows the activation energy of 111.45 (KJ/mol). The some of the Ni content may also react with Mg to form a tiny
materials processed by ECAP exhibited the lower activation amount Mg2NiH4 which also the reason for the hydrogenation
energy means that the applying processing route could in- rate enhancement.
fluence the activation energy. Furthermore, the addition of The AlNi phase was not seen to dissolve during the ball
different Ni composites to MgH2 affect the activation energy milling process and it was also found that AlNi formed after
by decreases the onset temperature level for hydrogen the hydrogenation process which confirms that AlNi does not
desorption. Hence, the Alloy AZ31-4 Ni treated through two participate in the reaction with the hydrogen molecules. Thus,
process shows the activation apparently lower than the pure the increment of the Ni contents in the Mg matrix caused a
MgH2 which has a value of 161.9 (KJ/mol) [33]. reduction in the hydrogen storage capacity. As a result, in the
HEBM process, the pure AZ31 alloy absorbed the maximum
capacity of 6.5 wt.%; for AZ31-2 Ni and AZ31-4 Ni the absorbed
Conclusions capacity was 6.4 and 6.2 wt. % respectively. The activation
energy for the AZ31-4 Ni alloy was estimated through a Kis-
In this study, AZ31-0 Ni, AZ31-2 Ni, and AZ31-4 Ni alloys were singer plot. It was revealed that the addition of Ni and the
fabricated through the gravitating mechanical stir casting deformation processes reduced the activation energy by
(GMSC) process. Significant grain size refinements and 104.73 (KJ/mol) for the samples deformed by ECAP which is
increased defect density were observed after deformation obviously lower than the material processed by HEBM and
through the ECAP process. The ECAP-processed AZ31-4 Ni pure MgH2. Overall, the observations that the AZ31-4 Ni
 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8 1057

material which was processed by ECAP showed better kinetics [13] Asselli AAC, Leiva DR, Huot J, Kawasaki M, Langdon TG,
with higher sorption capacity. Botta WJ. Effects of equal-channel angular pressing and
accumulative roll-bonding on hydrogen storage properties of
a commercial ZK60 magnesium alloy. Int J Hydrogen Energy
2015;40:16971e6. https://doi.org/10.1016/
Acknowledgment j.ijhydene.2015.05.149.
[14] Krystian M, Zehetbauer MJ, Kropik H, Mingler B, Krexner G.
The authors would like to acknowledge the financial support Hydrogen storage properties of bulk nanostructured ZK60 Mg
alloy processed by equal channel angular pressing. J Alloy
received from the Ministry of Science and Technology of
Comp 2011;509:S449e55. https://doi.org/10.1016/
Taiwan for this research under the grant number MOST 106- j.jallcom.2011.01.029.
2221-E-011 -008 -MY2. We also wish to express our gratitude to [15] Re
ve
Gajdics M, Varga LK, Kra

sz A,
llics G, Pe
ter L, Spassov T.
Prof. Chun Chiu (NTUST-Taiwan) for the assistance with the Hydrogen storage of nanocrystalline Mg-Ni alloy processed
Sievert's apparatus and also sincerely thank Prof. Eugen Rab- by equal-channel angular pressing and cold rolling. Int J
kin (Technion-Israel) for his support and guidance. Hydrogen Energy 2014;39:9911e7. https://doi.org/10.1016/
j.ijhydene.2014.01.059.
[16] Skripnyuk V, Buchman E, Rabkin E, Estrin Y, Popov M,
references Jorgensen S. The effect of equal channel angular pressing on
hydrogen storage properties of a eutectic Mg-Ni alloy. J Alloy
Comp 2007;436:99e106. https://doi.org/10.1016/
j.jallcom.2006.07.030.
[1] Jia Y, Sun C, Shen S, Zou J, Mao SS, Yao X. Combination of [17] Tan Y, Mao Q, Su W, Zhu Y, Li L. Remarkable hydrogen
nanosizing and interfacial effect: future perspective for storage properties at low temperature of Mg-Ni composites
designing Mg-based nanomaterials for hydrogen storage. prepared by hydriding combustion synthesis and
Renew Sustain Energy Rev 2015;44:289e303. https://doi.org/ mechanical milling. RSC Adv 2015;5:63202e8. https://doi.org/
10.1016/j.rser.2014.12.032. 10.1039/c5ra09754a.
[2] Krishna R, Titus E, Salimian M, Okhay O, Rajendran S, [18] Zaldı́var-Cadena AA, Dı́az-Pen~ a I, Caban ~ as-Moreno JG.
Rajkumar A. Hydrogen storage for energy application. Dispersion of niquel on the microstructure in magnesium
Hydrog Storage 2012. https://doi.org/10.5772/51238. based alloys for hydrogen storage. J Magnes Alloy
[3] Eberle U, Arnold G, Von Helmolt R. Hydrogen storage in 2013;1:292e6. https://doi.org/10.1016/j.jma.2013.11.002.
metal-hydrogen systems and their derivatives. J Power [19] Skripnyuk VM, Rabkin E, Estrin Y, Lapovok R. The effect of
Sources 2006;154:456e60. https://doi.org/10.1016/ ball milling and equal channel angular pressing on the
j.jpowsour.2005.10.050. hydrogen absorption/desorption properties of Mg-4.95 wt%
[4] Schlapbach L, Schlapbach L, Züttel A. Hydrogen-storage Zn-0.71 wt% Zr (ZK60) alloy. Acta Mater 2004;52:405e14.
materials for mobile applications. Nature 2001;414:353e8. https://doi.org/10.1016/j.actamat.2003.09.025.
https://doi.org/10.1038/35104634. [20] Huang SJ, Chiu C, Chou TY, Rabkin E. Effect of equal channel
[5] Hanada N, Ichikawa T, Fujii H. Catalytic effect of angular pressing (ECAP) on hydrogen storage properties of
nanoparticle 3d-transition metals on hydrogen storage commercial magnesium alloy AZ61. Int J Hydrogen Energy
properties in magnesium hydride MgH2 prepared by 2018;43:4371e80. https://doi.org/10.1016/
mechanical milling. J Phys Chem B 2005;109:7188e94. https:// j.ijhydene.2018.01.044.
doi.org/10.1021/jp044576c. [21] Huang S, Ali AN. Materials Science & Engineering A
[6] Lu J, Choi YJ, Fang ZZ, Sohn HY, Boule V. Prepared by Effects of heat treatment on the microstructure and
ultrahigh-energy - high-pressure milling. 2009. p. 15843e52. microplastic deformation behavior of SiC particles
[7] Varin RA, Zaranski Z, Czujko T, Polanski M, Wronski ZS. The reinforced AZ61 magnesium metal matrix composite.
composites of magnesium hydride and iron-titanium Mater Sci Eng A 2018;711:670e82. https://doi.org/10.1016/
intermetallic. Int J Hydrogen Energy 2011;36:1177e83. https:// j.msea.2017.11.020.
doi.org/10.1016/j.ijhydene.2010.06.092. [22] Zheng L, Nie H, Liang W, Wang H, Wang Y. Effect of pre-
[8] Takahashi Y, Yukawa H, Morinaga M. Alloying effects on the homogenizing treatment on microstructure and mechanical
electronic structure of Mg2Ni intermetallic hydride. J Alloy properties of hot-rolled AZ91 magnesium alloys. J Magnes
Comp 1996;242:98e107. https://doi.org/10.1016/0925-8388(96) Alloy 2016;4:115e22. https://doi.org/10.1016/
02268-2. j.jma.2016.04.002.
[9] Si T, Ma Y, Li Y, Liu D. Solid solution of Cu in Mg2NiH4 and its [23] TathgarvH S, Bakke P, Ovrelid E, Fenstand J, Frisvold F,
destabilized effect on hydrogen desorption. Mater Chem Engh TA. Solubility of nickel in molten magnesium-
Phys 2017;193:1e6. https://doi.org/10.1016/ aluminum alloys above 650 C. Magnesium Technology
j.matchemphys.2017.02.004. 2000:181e9. https://doi.org/10.1002/9781118808962.ch27.
[10] Pozzo M, Alfe D. Hydrogen dissociation and diffusion on 2000.
transition metal ( Ti , Zr , V , Fe , Ru , Co , Rh , Ni , Pd , Cu , Ag ) [24] Muralidhar A, Narendranath S, Shivananda Nayaka H. Effect
-doped Mg ( 0001 ) surfaces. Int J Hydrogen Energy of equal channel angular pressing on AZ31 wrought
2009;34:1922e30. https://doi.org/10.1016/ magnesium alloys. J Magnes Alloy 2013;1:336e40. https://
j.ijhydene.2008.11.109. doi.org/10.1016/j.jma.2013.11.007.
[11] Soyama J, de Araujo Silva R, Jose JM, Leiva DR, Floriano R, [25] Soyama J, Martins Triques MR, Leiva DR, Jorge AM, Da
Ishikawa TT, et al. Processing of MgH 2 by extensive cold Silva EP, Pinto HC, et al. Hydrogen storage in heavily
rolling under protective atmosphere. Int J Hydrogen Energy deformed ZK60 alloy modified with 2.5 wt.% Mm addition. Int
2016;2. https://doi.org/10.1016/j.ijhydene.2016.10.056. J Hydrogen Energy 2016;41:4177e84. https://doi.org/10.1016/
[12] Li J, Xie L, Zhang T, Song L. Microstructure and absorption/ j.ijhydene.2016.01.036.
desorption kinetics evolutions of Mg e Ni e Ce alloys during [26] Bououdina M, Guo ZX. Comparative study of mechanical
hydrogenation and dehydrogenation cycles. Int J Hydrogen alloying of (MgþAl) and (MgþAlþNi) mixtures for hydrogen
Energy 2018;43:8404e14. https://doi.org/10.1016/ storage. J Alloy Comp 2002;336:222e31. https://doi.org/
j.ijhydene.2018.03.113. 10.1016/S0925-8388(01)01856-4.
1058 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 4 4 ( 2 0 1 9 ) 1 0 4 7 e1 0 5 8

[27] Iwahashi Y, Horita Z, Nemoto M, Langdon TG. An boundaries. Acta Mater 2016;107:168e77. https://doi.org/
investigation of microstructural evolution during equal- 10.1016/j.actamat.2016.01.069.
channel angular pressing. Acta Mater 1997;45:4733e41. [31] Huang S-J, Fang C, Rajagopal V, Chen Y-L. Effect of different
https://doi.org/10.1016/S1359-6454(97)00100-6. processes on hydrogen storage properties of AZ magnesium
[28] Chiu C, Huang SJ, Chou TY, Rabkin E. Improving hydrogen alloy. Mater Sci Forum 2018;920:230e5. https://doi.org/
storage performance of AZ31 Mg alloy by equal channel 10.4028/www.scientific.net/MSF.920.230. MSF.
angular pressing and additives. J Alloy Comp [32] Denys RV, Poletaev AA, Petter J, Ketil J, Tarasov BP,
2018;743:437e47. https://doi.org/10.1016/ Yartys VA. Nanostructured rapidly solidified LaMg 11 Ni alloy
j.jallcom.2018.01.412. . II . In situ synchrotron X-ray diffraction studies of hydrogen
[29] Yang T, Li Q, Liu N, Liang C, Yin F, Zhang Y. Improved absorption e desorption behaviours. Int J Hydrogen Energy
hydrogen absorption and desorption kinetics of magnesium- 2011;37:5710e22. https://doi.org/10.1016/
based alloy via addition of yttrium. J Power Sources j.ijhydene.2011.12.133.
2018;378:636e45. https://doi.org/10.1016/ [33] Zhang Q, Zang L, Huang Y, Gao P, Jiao L, Yuan H, et al.
j.jpowsour.2018.01.003. Improved hydrogen storage properties of MgH2with Ni-
[30] Iwaoka H, Arita M, Horita Z. Hydrogen diffusion in ultrafine- based compounds. Int J Hydrogen Energy 2017;42:24247e55.
grained palladium: roles of dislocations and grain https://doi.org/10.1016/j.ijhydene.2017.07.220.