Content uploaded by Kaveh Edalati
Author content
All content in this area was uploaded by Kaveh Edalati on Nov 29, 2018
Content may be subject to copyright.
1
Acta Materialia 108 (2016) 293-303
Received 16 January 2016; Received in revised from 10 February 2016
Accepted 10 February 2012
New nanostructured phases with reversible hydrogen storage capability in
immiscible magnesium-zirconium system produced by high-pressure torsion
Kaveh Edalati1,2,*, Hoda Emami1, Yuji Ikeda3, Hideaki Iwaoka2, Isao Tanaka3,4,
Etsuo Akiba1,5, Zenji Horita1,2
1 WPI, International Institute for Carbon-Neutral Energy Research (WPI-I2CNER), Kyushu
University, Fukuoka 819-0395, Japan
2 Department of Materials Science and Engineering, Faculty of Engineering, Kyushu
University, Fukuoka 819-0395, Japan
3 Center for Elements Strategy Initiative for Structure Materials (ESISM), Kyoto University,
Sakyo, Kyoto 606-8501, Japan
4 Department of Materials Science and Engineering, Kyoto University, Kyoto 606-8501, Japan
5 Department of Mechanical Engineering, Faculty of Engineering, Kyushu University, Fukuoka
819-0395, Japan
Mg and Zr are immiscible in the solid and liquid states and do not form any binary phases. In
this study, Mg and Zr were significantly dissolved in each other by severe plastic deformation
(SPD) through the high-pressure torsion (HPT) method and several new metastable phases were
formed: nanostructured hcp, nano-twinned fcc, bcc or ordered bcc-based phases. These
supersaturated Mg-Zr phases, which did not decompose up to 773 K, exhibited reversible
hydrogen storage capability at room temperature. They absorbed ~1 wt.% of hydrogen under 9
MPa in ~20 s and fully desorbed the hydrogen in the air. First-principles phonon calculations
revealed that the disordered hcp and fcc solid solutions were dynamically stable in the whole
composition range of the Mg-Zr system. The bcc or bcc-based ordered phases, which were
formed only as intermediate phases during the phase transformation to the hcp solid solution
alloy, were energetically higher and were dynamically stable only under limited conditions in
the Mg-rich compositions.
Keywords: Mg-Zr phase diagram; phase transformations; metal hydrides; mechanical alloying;
density functional theory (DFT).
*Corresponding author:
Tel/Fax: +81-92-802-2992; E-mail: kaveh.edalati@zaiko6.zaiko.kyushu-u.ac.jp
2
1. Introduction
Phase transformations in the Mg-Zr system has not been investigated significantly
because Mg and Zr are immiscible in the solid state as well as in the liquid state [1]. Although
both Mg and Zr have the hcp structures with nearly similar lattice parameters at ambient
pressure and room temperature, earlier experimental and theoretical studies confirmed that no
binary phases can be formed in the Mg-Zr equilibrium phase diagram [1-3]. So far, there have
been few attempts to examine the formation of non-equilibrium (metastable) phases in the Mg-
Zr system.
Mg can store hydrogen in the form of MgH2 and Zr, which belongs to the group 4 of
the periodic table (Ti, Zr and Hf), can also store hydrogen in the form of ZrH2. Although the
hydrogen storage performance of the Mg-Ti-H system has been investigated significantly (e.g.,
see Refs. [4-11]), there have been limited attempts to examine the hydrogen storage
performance of the Mg-Zr-H system [8-18]. Goto et al. [12] and Okada et al. [13] employed a
cubic-anvil-type apparatus and synthesized hydrides MgZr2Hy or Mg2Zr3Hy with the
monoclinic structure by processing MgH2+ZrH2 mixture under a hydrogen pressure of 2-5 GPa
at 1073-1173 K for 2 h. Takasaki et al. [14], Kyoi et al. [15] and Moser et al. [16] employed a
high-pressure cell and synthesized a hydride Mg0.82Zr0.18H2 with the fcc structure by processing
a mixture of MgH2 and ZrH2 under a hydrogen pressure of 4-8 GPa at 873 K for 1 h. Bao et al.
[17] synthesized thin film of Mg0.82Zr0.18H2 with the fcc structure by co-sputtering of Mg and
Zr followed by hydrogenation. Guzik et al. [18] conducted reactive ball milling on Mg+Zr
powder mixtures under a deuterium pressure of 5.6-6.7 MPa, and synthesized a hydride
Mg0.40Zr0.60D1.78 with the fcc structure. First-principles calculations suggested that when the
fraction of Zr in Mg1-yZryH2 is higher than ~13 at.%, a ternary hydride with the fcc crystal
structure (fluorite-type structure) should be formed [19].
In this study, formation of new metastable phases (disordered hcp, partially ordered
bcc and disordered fcc) in the Mg-Zr system and their hydrogen storage performance are
examined by experiments using the high-pressure torsion (HPT) method [20-22] and by first-
principles calculations. The HPT method, which was first introduced by Bridgman in 1935 [23],
is employed for mixing Mg and Zr in the atomic scale. Although the HPT method is currently
used mainly as a severe plastic deformation (SPD) technique for grain refinement in metallic
materials [24-26], there have been long-time historical efforts from 1935 [23] to now [27-29]
to synthesize metastable phases by HPT processing (see a review in Ref. [30]). The HPT
method was employed in this study not only because of its capability to control phase
transformations [23-30], but also because of its effectiveness to activate hydrogen storage
materials [31-33], increase the hydrogenation kinetics [34-38] and synthesize Mg-based
hydrogen storage materials [39,40].
2. Experimental Materials and Methods
The Mg and Zr elements in the powder form with high purity (99.5%) and particle
3
sizes smaller than ~44 m (325 mesh) were mechanically mixed. The HPT process was carried
out on the powder mixtures using a pair of Bridgman anvils [23], having flat-bottom holes of
14mm diameter and 0.4mm depth on the surface. The powder mixtures were processed by
torsional straining using HPT (
hrN /2
,
: shear strain, r: distance from disc center, N:
number of HPT turns, h: thickness of disc [20,21]) for N = 1, 10, 100 and 1000 turns under a
pressure of P = 1.5 GPa with a rotation speed of
= 1 rpm at room temperature. The HPT-
processed discs with 14mm diameter were examined using various experimental methods as
explained below.
In order to examine the microstructural homogeneity along the disc radii, the HPT-
processed discs were first mechanically polished and the Vickers microhardness was measured
using an applied load of 200 g for 15 s in three radial directions from the disc center to periphery.
Second, the microscopic distribution of elements was examined using scanning
electron microscopy (SEM) equipped with energy dispersive X-ray spectroscopy (EDS). The
SEM and EDS examinations were performed under an accelerating voltage of 15 kV at 4-6 mm
away from the disc center.
Third, the formation of new phases was examined using X-ray diffraction (XRD)
analyses using the Cu Kα radiation. Examinations were performed on the periphery of all discs
as well as on the center of discs processed for N = 100 and 1000 turns. The XRD profiles were
analyzed in details by Rietveld refinement using the FullProf Suite software [41,42].
Fourth, for microstructural examinations, transmission electron microscopy (TEM)
and Cs-corrected scanning transmission microscopy (STEM) were conducted using
accelerating voltages of 300 kV and 200 kV, respectively. Thin foils were prepared from 6 mm
away from the disc center using a focused ion beam system. Thin foils were examined by TEM
bright-field and dark-field imaging modes, selected-area electron diffraction (SAED) analysis,
STEM-EDS analysis using a silicon drift detector and high-resolution high-angle annular dark-
field (HAADF) imaging mode using a convergence angle of 46 mrad and detecting angles of
90-370 mrad.
Fifth, thermal stability of new phases was examined using differential scanning
calorimetry (DSC) by heating the samples to 773 K with a heating rate of 5 K/min under an
argon atmosphere. The samples after the DSC analysis were further examined by XRD analysis.
Sixth, the hydrogen pressure-composition isotherms (PCI) and the hydrogenation
kinetics were examined in a Sieverts-type gas absorption apparatus at 303 K in the pressure
range of 0.01-9 MPa for samples processed by HPT for N = 1000 turns. For PCI measurements,
two disc samples were crushed under an argon atmosphere and subjected to evacuation at room
temperature for 2 h before starting the measurements. It should be noted that the crushing was
performed to increase the total surface area of the samples in contact with the hydrogen gas.
The PCI measurements were performed on the crushed samples for three cycles, while the third
PCI cycle was terminated at the end of the hydrogenation step. Within 10-60 min after
hydrogenation in the third cycle, the sample was examined by XRD analysis to investigate the
4
crystal structure of the hydride. The hydrogenated sample was also examined within 20-120
min after hydrogenation using thermogravimetry analysis (TGA) and differential thermal
analysis (DTA) by heating the sample to 773 K with a heating rate of 5 K/min under an argon
pressure of 0.1 MPa.
3. Calculation Methods
Formation energies and dynamical stability of hcp, fcc, and bcc phases for the Mg-Zr
system were investigated using first-principles calculations. Phonon frequency calculations
were used to discuss the dynamical stability of different phases. Special quasi-random
structures (SQSs) [43] were constructed for chemical compositions of Mg1-xZrx (x = 0.25, 0.5,
and 0.75) to mimic atomic configurations. Ordered bcc-based structures, namely B2 and D03,
were also investigated to discuss the effect of the ordering on the dynamic stability of the bcc-
based phases.
Electronic-structure calculations were performed within the framework of density
functional theory (DFT) using the Vienna Ab initio simulation package (VASP) code [44-46]
employing the plane-wave basis projector augmented wave (PAW) method [47]. The
generalized gradient approximation in the Perdew-Burke-Ernzerhof (PBE) form was used [48].
The plane-wave energy cut-off was taken to be 250 eV. The Brillouin zones were sampled using
Γ-centered 18x18x12, 12x12x12, and 16x16x16 meshes (7776, 6912, and 8192 points per
atoms, respectively) for the conventional hcp, fcc and bcc unit cells, respectively. The
Methfessel-Paxton scheme [49] with a smearing width of 0.4 eV was employed together with
the k-point sampling meshes.
Phonon calculations were performed using the PHONOPY code [50,51]. Force
constants were obtained from the cells whose sizes correspond to the 3x3x2, 2x2x2, and 2x2x2
supercells of the conventional hcp, fcc, and bcc unit cells, respectively. Atomic displacements
of 1 pm were used to extract the force constants. Phonon frequencies were calculated at the
reciprocal-space points on the Γ-centered 72x72x48, 48x48x48, and 64x64x64 meshes for the
primitive hcp, fcc, and bcc unit cells, respectively. The tetrahedron method [52,53] was used to
obtain phonon density of states.
4. Results and Discussion
4.1. Formation of Metastable Phases
XRD profiles are shown in Fig. 1(a) for the Mg/Zr powder mixtures and for the
samples processed by HPT for N = 1, 10, 100 and 1000 turns. The XRD profiles were conducted
at the periphery of discs, where the average shear strain is
= 40, 400, 4000 and 40000 after N
= 1, 10, 100 and 1000 turns, respectively. The center of discs were also examined for samples
processed by HPT for N = 100 and 1000 turns, which correspond to average shear strains of
= 800 and 8000, respectively. An examination of Fig. 1(a) indicates that while the peaks for
hcp-Mg and hcp-Zr are visible after N = 1, 10 and 100 turns, they disappear after N = 1000 turns.
5
Figure 1. (a) XRD profiles from center and periphery of discs (indicted by C and P, respectively)
for samples processed by HPT for N = 1, 10, 100 and 1000 turns including powder mixtures;
(b) XRD profile from periphery of sample processed for N = 100 in enlarged scale to check
ordering in bcc phase; and (c) fraction of bcc phase against shear strain for samples processed
by HPT for various turns.
Peaks for a new hcp phase with the lattice parameters of a = 0.321 nm and c = 0.516 nm clearly
appear after N = 1000 turns. The lattice parameters of the new hcp phase lies between the lattice
parameters of hcp-Mg (a = 0.321 nm and c = 0.521 nm) and hcp-Zr (a = 0.323 nm and c =
6
0.515 nm), suggesting that the two elements dissolve in each other with introducing severe
shear strain and finally a supersaturation of two elements in each other is achieved.
Close examination of Fig. 1(a) indicates that a bcc phase with a lattice parameter of a
= 0.3566 nm also forms after N = 100 turns. A magnified view of the XRD profiles for the bcc
phase, as shown in Fig. 1(b), finds a weak superlattice peak of the (100) reflection, indicating
the presence of an ordered bcc-based structure. The evolution of the bcc phase quantity with
the shear strain is shown more clearly in Fig. 1(c), where the fraction of bcc phase, calculated
by Rietveld analysis, is plotted against the shear strain. The bcc phase is formed when the shear
strain increases to
= 800 and its fraction reaches a maximum value at a shear strain of
=
4000. The fraction of the bcc phase decreases significantly (to almost zero) at very large shear
strains as
= 40000.
Fig. 2 shows SEM micrographs (left) and SEM-EDS mappings (right) for the samples
processed by HPT for N = 1, 10, 100 and 1000 turns, where SEM images and SEM-EDS
mappings were taken from r = 4-6 mm and r = 5.5 mm away from the disc centers, respectively.
It should be noted that the magnifications for the SEM images and the EDS mappings is
different. It is apparent that Mg and Zr phases (with dark and bright contrasts in the SEM
micrographs, and with red and green colors in the EDS mappings, respectively) are well
distinguishable after N = 1 and 10 turns. The two elements are partially dissolved in each other
after N = 100 and the particles are elongated in the shearing direction, especially at far distances
from the disc center. In good agreement with the XRD analyses, the two elements are
significantly mixed after N = 1000 turns, despite their thermodynamic immiscibility. Mixing of
the Mg and Zr atoms after N = 1000 can be seen even in higher magnifications, as shown in
STEM-EDS mappings of Figs. 3(a-c), although Mg-rich and Zr-rich regions are still visible.
TEM observations, as shown in Figs. 3(d-f), show that the microstructure after HPT
processing for N = 1000 turns consists of nanograins with sizes in the range of a few nanometers
to ~100 nm. The average grain size measured for 70 grains using several dark-field images
appears to be 40±30 nm. The SAED analysis with a ring-shaped pattern, which was taken from
a circular region with a diameter of ~350 nm, also indicates that microstructures contain
nanograins with random orientations. The reduction of grain size to the nanometer level in this
study is basically consistent with many earlier publications [20-22]; however, the reduction of
grain size in this study is more significant than those reported in single-phase metals and alloys
[24,25] including pure Mg [54,55] and Mg-based alloys [56,57], which may be ascribed to the
co-existence of different phases as well as to the occurrence of supersaturation [59,60]. The
large variations of grain sizes from a few nanometer to ~100 nm suggests that the shear strain
was not high enough even after N = 1000 turns to achieve a saturation of grain fragmentation
[26]. Microhardness measurements also confirm that the hardness and thus the microstructural
features and phase quantities do not reach the steady state even after N = 1000 turns (hardness
levels at r = 0.1, 3 and 6 mm away from the disc center were 80±8, 100±4 and 145±11 Hv,
respectively).
7
Figure 2. SEM micrographs taken from 4-6 mm away from disc center (left) and EDS mappings
taken from 5.5 mm away from disc center (right) for samples processed by HPT for (a) N = 1,
(b) N = 10, (c) N = 100 and (d) N = 1000 turns.
8
Figure 3. (a) STEM bright-field image and corresponding EDS mapping with (b) Mg and (c)
Zr; (d) TEM bright-field image and corresponding (e) dark-field image and (f) SAED pattern
for sample processed by HPT for N = 1000 turns and examined at 6 mm away from disc center.
Examination of the microstructure using high-resolution STEM (HAADF images) and
corresponding diffractograms obtained by the fast Fourier transform (FFT) analyses, as shown
in Fig. 4(a), confirm the presence of bcc structure after N = 100 turns, which is in good
agreement with the XRD analyses. In addition to the hcp and bcc crystal structures, an fcc phase
with a lattice parameter of a = 0.44-0.46 nm, as shown in Fig. 4(b), is detected in the
microstructure. The bcc phase, which could not be detected using the XRD analysis at very
large shear strains as
= 40000, is still visible in the microstructure, as shown in Fig. 4(c).
Moreover, some grains with nanotwins are visible in the microstructure, as shown in Fig. 4(d).
Examination of these grains shows that they correspond to the fcc structure with a lattice
parameter of a = 0.44-0.46 nm. It should be noted that, although nanotwins are rarely formed
in the nangrained hcp structures by plastic deformation, the nanotwins are formed frequently in
9
the nanograined fcc structures [58]. Therefore, the presence of nanotwins is another evidence
for the formation of the fcc phase in the Mg-Zr system. The current examinations suggest that
the material at very large shear strains contains mainly hcp phase and small amounts of bcc and
fcc phases.
Close examination of high-resolution lattice images in Fig. 4 shows that the
microstructure contains many Mg-based nanoclusters. Since the micrographs in Fig. 4 are
HAADF images, the dark regions correspond to Mg-rich regions. High-resolution EDS
mapping, as shown in Fig. 4(e) confirms that the dark regions in the lattice images correspond
to Mg-based nanoclusters. These nanoclusters can be appropriate sites for hydrogen storage in
the current material.
Figure 4. HAADF Lattice images of (a,c) bcc phase and (c,d) fcc phase including corresponding
FFT diffractograms; and (e) HAAF lattice image and corresponding EDS mappings to check
the formation of Mg-based nanoclusters for samples processed by HPT for (a,b) N = 100 and
(c-e) N = 1000 turns.
10
Since Mg and Zr are thermodynamically immiscible in each other, the main
mechanism of formation of supersaturated Mg-Zr alloys after processing by HPT should be due
to the mechanical alloying. For example, Mg/Z particles with initial particle sizes of 10-44 m
which are located next to each other at a radius of 5 mm from the disc center are significantly
elongated and their thicknesses are nominally reduced to 0.2-1 nm after N = 1000 turns (
=
40,000). Therefore, the two particles are mechanically mixed at the sub-nanometer level with
the application of such ultrahigh strains. Earlier studies also showed that supersaturated alloys
can be formed after severe plastic straining in different metallic systems [59,60] including the
Mg-Ti system [39].
The formation of Mg-Zr solid solution with the hcp structure in this study is due to
strain-induced mixing of two elements and formation of supersaturated Mg-rich or Zr-rich
alloys. The formation of bcc phase in intermediate levels of strain and its disappearance at very
large strains suggest that the bcc phase is an intermediate phase during the phase transformation
from the mixture of hcp-Mg and hcp-Zr to the hcp-Mg-Zr solid solution. Formation of fcc phase
may be a consequence of formation of partial dislocations and accumulation of stacking faults
in nanograins of supersaturated hcp phase with low stacking fault energy. An earlier study using
the first-principles calculations showed that the accumulation of stacking faults in nanograins
of hcp metals such as Co can finally lead to the formation of an fcc phase [61]. It should be
noted that in the Mg-Ti system, in which the Mg and Ti atoms are immiscible in the solid and
liquid states similar to the Mg-Zr system, the formation of hcp, bcc and fcc phases was also
reported using different processing techniques [4-11] including the HPT method [39].
The variation of heat flow against temperature after DSC analysis with a heating rate
of 5 K/min is shown in Fig. 5(a) for the sample processed by HPT for N = 1000 turns. No
exothermic peaks corresponding to the decomposition of metastable phases, dislocation
recovery and grain growth appear by increasing the temperature to 773 K (this temperature is
the upper limit of our DCS facility). The XRD profile for the sample before and after the DSC
analysis is shown in Fig. 5(b). It is apparent that the XRD profiles are exactly the same before
and after the DSC analysis (the same lattice parameters and the same peak broadenings),
suggesting that the metastable phases in the Mg-Zr are stable at temperatures up to 773 K and
no significant dislocation recovery and grain growth occur during heating. The stability of these
phases is not due to their thermodynamic stability because no binary phases exist in the
equilibrium phase diagram of the Mg-Zr system [1-3]. The main reason for the stability of these
phases should be due to high kinetic barrier for the elemental decomposition in these phases.
4.2. Hydrogen Storage in Mg-Zr Phases
Results of PCI measurements at room temperature are shown in Fig. 6(a) for the
sample processed by HPT for N = 1000 turns. It should be noted that the hydrogenation during
the PCI measurements was continued until the final pressure reached a pressure limit of 9 MPa
in the whole volume of the apparatus (9 MPa is the maximum achievable pressure in our
11
Sieverts-type gas absorption apparatus). It is apparent that the sample absorbs ~1 wt.% of
hydrogen in three hydrogenation cycles and desorbs the hydrogen during evacuation at room
temperature. It should be noted that the pressure of 9 MPa, in which a plateau-like region
appears in Fig. 6(a), is higher than the equilibrium hydrogenation pressure of MgH2 [62-64]
and Mg-Zr-based hydrides [12-18], and thus it may not correspond to the equilibrium plateau
pressure. The hydrogen absorbed during each step of PCI measurements are plotted in Fig. 6(b)
against the time required to reach the steady state (the time after that no increase in hydrogen
absorption occurs). Examination of Fig. 6(b) indicates that the hydrogenation occurs very fast
and the material absorbs ~1 wt.% of hydrogen in ~20 s.
Figure 5. (a) DSC curve obtained using heating rate of 5 K/min for sample processed by HPT
for N = 1000 turns, and (b) corresponding XRD profiles before and after heat treatment in DSC.
12
Figure 6. (a) PCI results and (b) corresponding hydrogenation kinetic plots for sample processed
by HPT for N = 1000 turns.
In order to investigate the formation of hydrides, the sample after hydrogenation in the
third cycle was quickly examined by TGA/DTA and XRD analysis. The TGA/DTA and XRD
analyses were conducted within 20-120 min and 10-60 min after exposing the samples to an
argon atmosphere, respectively. The TGA/DTA results, as shown in Fig. 7(a), do not show any
indication for hydrogen desorption. The XRD profile of sample after hydrogenation, as shown
in Fig. 7(b), is quite similar to the one taken before the hydrogenation (the hcp structure with
the same lattice parameters) and no peaks for the hydride phase can be detected. The current
TGA/DTA and XRD results indicate that the hydrogen is desorbed from the Mg-Zr phases
quickly under a partial hydrogen pressure of zero (either under vacuum, air or argon
atmospheres).
Hydrogen absorption under a pressure of 9 MPa and hydrogen desorption under a
partial hydrogen pressure of zero are thermodynamically reasonable [62-64]. However, because
of kinetic barriers for hydrogenation and dehydrogenation, bulk Mg-based materials usually
cannot show this kind of reversibility at room temperature [62]. Since the current Mg-Zr phases
absorb and desorb hydrogen very fast even when compared to Mg-based alloys after
13
deformation by HPT [34-38] or equal-channel angular pressing [65-68], it is not easy to find
out where the hydrogen atoms are (in situ analyses using X-ray and neutron scattering under
high hydrogen pressure should be conducted). However, such fast hydrogenation and
dehydrogenation suggest that the hydrogen atoms may be absorbed in the Mg-based nanocluster,
as shown in Fig. 4(e), under a high hydrogen pressure of 9 MPa. If 2/3 of Mg atoms stay in the
form of nanoclusters in the material, the material can absorb 1 wt.% of hydrogen. Earlier studies
showed that Mg nanoclusters can absorb and desorb hydrogen very fast [69]. Moreover, as
suggested in earlier studies [32,33], a large fraction of grain boundaries introduced by HPT
processing can also act as pathways for hydrogen transport between the sample surface and the
nanoclusters. An earlier study using the first-principles calculations also suggested that the Mg-
rich phases in the Mg-Zr system can produce hydrides with lower thermodynamic stability than
MgH2 [19].
Figure 7. (a) TGA/DTA results after hydrogenation in third PCI cycle and (b) XRD profiles
before and after hydrogenation for sample processed by HPT for N = 1000 turns.
4.3. First-Principles Calculations
In order to examine the dynamical stability of the disordered hcp, fcc and bcc structures
as well as that of the ordered bcc-based structures, first-principles phonon calculations were
systematically made for different compositions in the Mg-Zr system (Mg, Mg0.75Zr0.25,
14
Mg0.5Zr0.5, Mg0.25Zr0.75 and Zr). Plots of phonon density of states are shown in Fig. 8, in which
the imaginary phonon frequencies are shown by negative values. Several important features can
be seen in Fig. 8. First, for all selected compositions, disordered hcp and fcc crystal structures
do not exhibit imaginary phonon frequencies, which implies that these two phases are
dynamically stable. Second, ordered D03-Mg0.75Zr0.25 and B2-Mg0.5Zr0.5 phases and disordered
bcc-Mg0.75Zr0.25 are also dynamically stable. Third, the bcc crystal structure is dynamically
unstable (neither stable nor metastable) in the other selected compositions. The last two features
suggest that the bcc phase cannot be formed in the Zr-rich side of the Mg-Zr system.
Figure 8. Phonon density of states curves obtained by first-principles calculations for hcp, fcc,
disordered bcc and ordered bcc phases with different compositions (Mg, Mg0.75Zr0.25, Mg0.5Zr0.5,
Mg0.25Zr0.75, Zr). Imaginary phonon frequencies are shown by negative values.
15
Figure 9. Dependence on Zr content for formation energy with reference to mixture of hcp-Mg
and hcp-Zr obtained by first-principles calculations. Open marks indicate phases that exhibit
imaginary phonon frequencies and are dynamically unstable (neither stable nor metastable) and
close marks indicate phases that are dynamically stable.
The calculations of formation energy, as shown in Fig. 9, indicate several important
points. First, the formation energies for all phases are positive with reference to the mixture of
hcp-Mg and hcp-Zr. The formation energy of the disordered hcp-Mg0.5Zr0.5 phase is 60
meV/atom (5790 J/mol), which corresponds to T = 5790 / (RLn2) = 1000 K of the phase
separation temperature, assuming ideal solid solution at x = 0.5 with an entropy of -R[xlnx+(1-
x)ln(1-x)] = RLn2 (R: gas constant). Since Mg metal is known to melt at 923 K, this result
suggests that all solid-state Mg-Zr phases should decompose to the mixture of hcp-Mg and hcp-
Zr under thermal equilibrium conditions. However, metastable Mg-Zr phases (shown with close
marks in Fig. 9) can remain if the kinetic barrier for the decomposition is not able to be
overcome. Second, the energy of formation for the hcp and fcc phases are much smaller than
those for the bcc phases or the bcc-based ordered phases, indicating that the bcc or the bcc-
based ordered phases should subsequently transform to the hcp or fcc phases. Third, the energy
of formation for the fcc-Mg0.5Zr0.5 phase is slightly lower than that for the hcp-Mg0.5Zr0.5 phase,
suggesting that the hcp phase can subsequently transform to the fcc phase. Although trace
amounts of fcc phase could be detected in this study using high-resolution STEM, it is expected
that the amount of the fcc phase can be increased, if the number of HPT turns increases to N >
1000 (increasing N is technically difficult at the present time).
In summary, both experiments using the HPT method and first-principles calculations
show that several metastable phases can be formed in the immiscible Mg-Zr system. Unlike the
immiscible Mg-Ti system, in which the metastable phases decompose at 643 K [39], the
metastable phases in the Mg-Zr system remain unchanged at least up to 773 K. The metastable
phases in the Mg-Zr system exhibit much better hydrogenation/dehydrogenation kinetics at
room temperature when compared to the Mg-Ti metastable phases [4-11].
16
5. Conclusions
Powder mixtures of Mg - 50 at.% Zr, which are thermodynamically immiscible in the
solid and liquid states, were subjected to intense shear strain through the high-pressure torsion
(HPT) method at room temperature and the following conclusions were drawn.
1. A supersaturation of two elements in each other occurred and several metastable phases
with average grain sizes of 40±30 nm were formed: (i) the hcp phase with a = 0.3213 nm
and c = 0.5160 nm; (ii) the trace amount of the fcc phase with a = 0.44-0.46 nm; (iii) the
bcc or bcc-based ordered phase with a = 0.3566 which was formed as an intermediate phase
during the phase transformation from the mixture of hcp-Mg and hcp-Zr to the hcp-Mg-Zr
solid solution.
2. The metastable phases remained unchanged up to 773 K during heating with a rate of 5
K/min. This may be ascribed to the presence of the kinetics barriers for the phase separation.
3. The HPT-processed Mg-Zr phases absorbed ~1 wt.% of hydrogen very fast (in ~20 s) under
a hydrogen pressure of 9 MPa at room temperature. The absorbed hydrogen was released
quickly at room temperature under air or argon atmospheres. It was suggested that the
hydrogen is mainly absorbed in the Mg-based nanoclusters.
4. First-principles calculations showed that the disordered hcp and fcc solid solutions were
metastable (i.e., dynamically stable but not thermodynamically most stable), in the whole
composition range of the Mg-Zr binary system. On the other hand, the disordered bcc or the
bcc-based ordered phases were energetically higher than the hcp and fcc phases. They were
dynamically stable only under limited conditions in the Mg-rich compositions. This is
consistent with the fact that the bcc or the bcc-based phase was formed only as an
intermediate phase before the formation of the hcp solid solution.
Acknowledgments
One of the authors (KE) acknowledges a grant from Kyushu University
Interdisciplinary Programs in Education and Projects in Research Development (P&P) (No.
27513) and a grant from WPI-I2CNER for Interdisciplinary Researches. This work was
supported in part by the Light Metals Educational Foundation of Japan, and in part by the
Grant-in-Aids from the MEXT, Japan (No. 26220909 and No. 15K14183). The HPT process
was carried out in the International Research Center on Giant Straining for Advanced Materials
(IRC-GSAM) at Kyushu University.
References
[1] H. Okamoto, Mg-Zr (magnesium-zirconium), J. Phase Equilib. Diff. 28 (2007) 305-306.
[2] M. Hamalainen, K. Zeng, Thermodynamic evaluation of the Mg-Zr system, Calphad 22 (1998)
375-380.
[3] R. Arroyave, D. Shin, Z.K. Liu, Modification of the thermodynamic model for the Mg-Zr system,
Calphad 29 (2005) 230-238.
[4] Y. Tsushio, S. J. Choi, H. Enoki, E. Akiba, Mg-Ti binary alloys with the BCC structure,
Proceedings of the Spring Meeting of the Japan Institute of Metals 130 (2002) 499.
[5] G. Liang, R. Schulz, Synthesis of Mg-Ti alloy by mechanical alloying, J. Mater. Sci. 38 (2003)
17
1179-1184.
[6] D.M. Borsa, R. Gremaud, A. Baldi, H. Schreuders, J.H. Rector, B. Kooi, P. Vermeulen, P.H.L.
Notten, B. Dam, R. Griessen, Structural, optical, and electrical properties of MgyTi1−yHx thin films,
Phys. Rev. B. 75 (2007) 205408.
[7] K. Asano, H. Enoki, E. Akiba, Synthesis of HCP, FCC and BCC structure alloys in the Mg-Ti
binary system by means of ball milling, J. Alloys Compd. 480 (2009) 558-563.
[8] K. Asano, H. Enoki, E. Akiba, Synthesis of Mg-Ti FCC hydrides from Mg-Ti BCC alloys, J.
Alloys. Compd. 4478 (2009) 117-120.
[9] S. Rousselot, M.P. Bichat, D. Guay, L. Rouez, Structural and electrochemical hydriding
characteristics of Mg-Ti-based alloys prepared by high energy ball milling, J. Electrochem. Soc.
156 (2009) A967-A973.
[10] G. Calmak, Z. Karoly, I. Mohai, T. Ozturk, J. Szepvolgyi, The processing of Mg-Ti for hydrogen
storage; mechanical milling and plasma synthesis, Int. J. Hydrogen Energy 35 (2010) 10412-
10418.
[11] K. Maveja, M. Phasha, N. van der Berg, Microstructure and crystal structure of an equimolar Mg-
Ti alloy processed by Simoloyer high-energy ball mill, Power Technol. 199 (2010) 256-263.
[12] Y. Goto, T. Hayashi, R. Kataoka, H. Kakuta, A. Kamegawa, H. Takamura, M. Okada, High-
pressure synthesis of novel hydrides in Mg-TM systems (TM = Zr, Nb and Mo), Mater. Trans. 46
(2005) 1798-1801.
[13] M. Okada, Y. Goto, R. Kataoka, Y. Yambe, A. Kamegawa, H. Takamura, Novel hydrides in Mg-
TM systems synthesized by high pressure (TM = Zr, Nb, Hf and Ta), J. Alloys Compd. 446-447
(2007) 6-10.
[14] T. Takasaki, D. Kyoi, N. Kitamura, S. Tanase, T. Sakai, Reversible hydrogen storage property and
structural analysis for face-centered cubic hydride Mg0.82Zr0.18H2 prepared by gigapascal hydrogen
pressure method, J. Phys. Chem. B 111 (2007) 14102-14106.
[15] D. Kyoi, T. Sakai, N. Kitamura, A. Ueda, S. Tanase, Synthesis of FCC Mg-Zr and Mg-Hf hydrides
using GPa hydrogen pressure method and their hydrogen-desorption properties, J. Alloys Compd
463 (2008) 311-316.
[16] D. Moser, D.J. Bull, T. Sato, D. Noreus, D. Kyoi, T. Sakai, N. Kitamura, H. Yusa, T. Taniguchi,
W.P. Kalisvaart, P. Notten, Structure and stability of high pressure synthesized Mg-TM hydrides
(TM = Ti, Zr, Hf, V, Nb and Ta) as possible new hydrogen rich hydrides for hydrogen storage, J.
Mater. Chem. 19 (2009) 8150-8161.
[17] S. Bao, Y. Yamada, K. Tajima, P. Jin, M. Okada, K. Yoshimura, Switchable mirror based on Mg-
Zr-H thin films, J. Alloys Compd. 513 (2012) 495-498.
[18] M.N. Guzik, S. Deledda, M.H. Sørby, V.A. Yartys, B.C. Hauback, New FCC Mg-Zr and Mg-Zr-
Ti deuterides obtained by reactive milling, J. Sol. State Chem. 226 (2015) 237-242.
[19] B.R. Pauw, W.P. Kalisvaart, S.X. Tao, M.T.M. Koper, A.P.J. Jansen, P.H.L. Notten, Cubic MgH2
stabilized by alloying with transition metals: a density functional theory study, Acta Mater. 56
(2008) 2948-2954.
[20] R.Z. Valiev, R.K. Islamgaliev, I.V. Alexandrov, Bulk nanostructured materials from severe plastic
deformation, Prog. Mater. Sci. 45 (2000) 103-189.
[21] R.Z. Valiev, Y. Estrin, Z. Horita, T.G. Langdon, M.J. Zehetbauer, Y.T. Zhu, Producing bulk
ultrafine-grained materials by severe plastic deformation, JOM 58(4) (2006) 33-39.
[22] A.P. Zhilyaev, T.G. Langdon, Using high-pressure torsion for metal processing: fundamentals and
applications, Prog. Mater. Sci. 53 (2008) 893-979.
[23] P.W. Bridgman, Effects of high shearing stress combined with high hydrostatic pressure, Phys.
Rev. 48 (1935) 825-847.
[24] M. Kawasaki, R.B. Figueiredo, T.G. Langdon, An investigation of hardness homogeneity
throughout disks processed by high-pressure torsion, Acta Mater. 59 (2011) 308-316.
[25] D.J. Lee, E.Y. Yoon, D.H. Ahn, B.H. Park, H.W. Park, L.J. Park, Y. Estrin, H.S. Kim, Dislocation
density-based finite element analysis of large strain deformation behavior of copper under high-
pressure torsion, Acta Mater. 76 (2014) 281-293.
[26] R. Pippan, S. Scheriau, A. Taylor, M. Hafok, A. Hohenwarter, A. Bachmaier, Saturation of
fragmentation during severe plastic deformation, Annu. Rev. Mater. Res. 40 (2010) 319-343.
[27] Y. Ivanisenko, I. MacLaren, X. Sauvage, R.Z. Valiev, H.J. Fecht, Shear-induced →
18
transformation in nanoscale Fe-C composite, Acta Mater. 54 (2006) 1659-1669.
[28] A.P. Zhilyaev, I. Sabirov, G. Gonzales-Doncel, J. Molina-Aldarequia, B. Srinivasarao, M.T. Perez-
Prado, Effect of Nb additions on the microstructure, thermal stability and mechanical behavior of
high pressure Zr phases under ambient conditions, Mater. Sci. Eng. A 528 (2011) 3496-3505.
[29] B.B. Straumal, A.S. Gornakova, O.B. Fabrichnaya, M.J. Kiregel, A.A. Mazilkin, B. Baretzky, A.M.
Gusak, S.V. Dobatkin, Effective temperature of high pressure torsion in Zr-Nb alloys, High Temp.
Mater. Proc. 31 (2012) 339-350.
[30] K. Edalati, Z. Horita, A review on high-pressure torsion (HPT) from 1935 to 1988, Mater. Sci.
Eng. A 652 (2016) 325-352.
[31] K. Edalati, J. Matsuda, H. Iwaoka, S. Toh, E. Akiba, Z. Horita, High-pressure torsion of TiFe
intermetallics for activation of hydrogen storage at room temperature with heterogeneous
nanostructure, Int. J. Hydrogen Energy 38 (2013) 4622-4627.
[32] K. Edalati, J. Matsuda, M. Arita, T. Daio, E. Akiba, Z. Horita, Mechanism of activation of TiFe
intermetallics for hydrogen storage by severe plastic deformation using high-pressure torsion,
Appl. Phys. Lett. 103 (2013) 143902.
[33] K. Edalati, J. Matsuda, A. Yanagida, E. Akiba, Z. Horita, Activation of TiFe for hydrogen storage
by plastic deformation using groove rolling and high-pressure torsion: similarities and differences,
Int. J. Hydrogen Energy 39 (2014) 15589-15594.
[34] D.R. Leiva, A.M. Jorge, T.T. Ishikawa, J. Huot, D. Fruchart, S. Miraglia, C.S. Kiminami, W.J.
Botta, Nanoscale grain refinement and H-sorption properties of MgH2 processed by high-pressure
torsion and other mechanical routes, Adv. Eng. Mater. 12 (2010) 786-792.
[35] A. Revesz, Z. Kanya, T. Verebelyi, P.J. Szabo, A.P. Zhilyaev, T. Spassov, The effect of high-
pressure torsion on the microstructure and hydrogen absorption kinetics of ball-milled Mg70Ni30,
J. Alloys Compd. 204 (2010) 83-88.
[36] M. Zehetbauer, R. Grossinger, H. Krenn, M. Krystian, R. Pippan, P. Rogl, T. Waitz, R. Wurschum,
Bulk nanostructured functional materials by severe plastic deformation, Adv. Eng. Mater. 12
(2010) 692-700.
[37] J.X. Zou, C.F. Perez-Brokate, R. Arruffat, B. Bolle, J.J. Fundenberger, X.Q Zeng, T. Grosdidier,
W.J. Ding, Nanostructured bulk Mg plus MgO composite synthesized through arc plasma
evaporation and high-pressure torsion for H-storage application, Mater. Sci. Eng. B 183 (2014) 1-
5.
[38] T. Hongo, K. Edalati, M. Arita, J. Matsuda, E. Akiba, Z. Horita, Significance of grain boundaries
and stacking faults on hydrogen storage properties of Mg2Ni intermetallics processed by high-
pressure torsion, Acta Mater. 92 (2015) 46-54.
[39] K. Edalati, H. Emami, A. Staykov, D.J. Smith, E. Akiba, Z. Horita, Formation of metastable
phases in magnesium-titanium system by high-pressure torsion and their hydrogen storage
performance, Acta Mater. 50 (2015) 150-156.
[40] H. Emami, K. Edalati, A. Staykov, T. Hongo, H. Iwaoka, Z. Horita, E. Akiba, Solid-state reactions
and hydrogen storage in magnesium mixed with various elements by high-pressure torsion:
experiments and first-principles calculations, RCS Adv. 6 (2016) 11665-11674.
[41] P. Thompson, D.E. Cox, J.B. Hastings, Rietveld refinement of Debye-Scherrer synchrotron X-ray
data from Al2O3, J. Appl. Crystallogr. 20 (1987) 79-83.
[42] J. Rodriguez-Carvajal, Recent advances in magnetic structure determination by neutron powder
diffraction, Physica B 192 (1993) 55-69.
[43] A. Zunger, S.H. Wei, L.G. Ferreira, J.E. Bernard, Special quasirandom structures, Phys. Rev. Lett.
65 (1990) 353-356.
[44] G. Kresse, Ab initio molecular dynamics for liquid metals, J. Non-Cryst. Solids 192 (1995) 222-
229.
[45] G. Kresse, J. Furthmuller, Efficiency of ab-initio total energy calculations for metals and
semiconductors using a plane-wave basis set, Comput. Mater. Sci. 6 (1996) 15-50.
[46] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method,
Phys. Rev. B 59 (1999) 1758-1775.
[47] P.E. Blochl, Projector augmented-wave method, Phys. Rev. B 50 (1994) 17953-17979.
[48] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made simple, Phys.
Rev. Lett. 77 (1996) 3865-3868.
19
[49] M. Methfessel, A.T. Paxton, High-precision sampling for Brillouin-zone integration in metals,
Phys. Rev. B 40 (1989) 3616-3621.
[50] A. Togo, F. Oba, I. Tanaka, First-principles calculations of the ferroelastic transition between
rutile-type and CaCl2-type SiO2 at high pressures, Phys. Rev. B 78 (2008) 134106.
[41] A. Togo, I. Tanaka, First principles phonon calculations in materials science, Scripta Mater. 108
(2015) 1-5.
[52] A.H. MacDonald, S.H. Vosko, P.T. Coleridge, Extensions of the tetrahedron method for
evaluating spectral properties of solids, J. Phys. C: Solid State Phys. 12 (1979) 2991-3002.
[53] P.E. Blochl, O. Jepsen, O.K. Andersen, Improved tetrahedron method for Brillouin-zone
integrations, Phys. Rev. B 49 (1994) 16223-16233.
[54] K. Edalati, A. Yamamoto, Z. Horita, T. Ishihara, High-pressure torsion of pure magnesium:
Evolution of mechanical properties, microstructures and hydrogen storage capacity with
equivalent strain, Scripta Mater. 64 (2011) 880-883.
[55] X.G. Qiao, Y.W. Zhao, W.M. Gan, Y. Chen, M.Y. Zheng, K. Wu, N. Gao, M.J. Starink, Hardening
mechanism of commercially pure Mg processed by high pressure torsion at room temperature,
Mater. Sci. Eng. A 619 (2014) 95-106.
[56] F. Meng, J.M. Rosalie, A. Singh, H. Somekawa, K. Tsuchiya, Ultrafine grain formation in Mg-Zn
alloy by in situ precipitation during high-pressure torsion, Scripta Mater. 78 (2014) 57-60.
[57] S.A. Torbati-Sarraf, S. Sabbaghianrad, T.G. Langdon, Microstructural properties, thermal
stability and superplasticity of a ZK60 Mg alloy processed by high-pressure torsion, Lett. Mater.
53 (2015) 287-293.
[58] Y.T. Zhu, X.Z. Liao, X.L. Wu, Deformation twinning in nanocrystalline materials, Prog. Mater.
Sci. 57 (2012) 1-62.
[59] X. Sauvage, F. Wetscher, P. Pareige, Mechanical alloying of Cu and Fe induced by severe plastic
deformation of a Cu-Fe composite, Acta Mater. 53 (2005) 2127-2135.
[60] A. Bachmaier, M. Kerber, D. Setman, R. Pippan, The formation of supersaturated solid solutions
in Fe-Cu alloys deformed by high-pressure torsion, Acta Mater. 60 (2012) 860-871.
[61] G.P. Zheng, Y.M. Wang, M. Li, Atomistic simulation studies on deformation mechanism of
nanocrystalline cobalt, Acta Mater. 53 (2005) 3893-3901.
[62] B. Sakintuna, F. Lamari-Darkrim, M. Hirscher, Metal hydride materials for solid hydrogen
storage: a review, Int. J. Hydrogen Energy 32 (2007) 1121-1140.
[63] M. Dornheim, S. Doppiu, G. Barkhordarian, U. Boesenberg, T. Klassen, O. Gutfleisch, R.
Bormann, Hydrogen storage in Mg-based hydrides and hydride composites,
Scripta Mater. 56
(2007) 841-846.
[64] I.P. Jain, C. Lal, A. Jain, Hydrogen storage in Mg: a most promising material, Int. J. Hydrogen
Energy 35 (2010) 5133-5144.
[65] V.M. Skripnyuk, E. Rabkin, Y. Estrin, R. Lapovok, The effect of ball milling and equal channel
angular pressing on the hydrogen absorption/desorption properties of Mg-4.95 wt% Zn-0.71 wt%
Zr (ZK60) alloy, Acta Mater. 52 (2004) 405-414.
[66] M. Krystian, M.J. Zehetbauer, H. Kropik, B. Mingler, G. Krexner, Hydrogen storage properties
of bulk nanostructured ZK60 Mg alloy processed by equal channel angular pressing, J. Alloys
Compd. 509S (2011) S449-S455.
[67] A.M. Jorge Jr., E. Prokofiev, G.F. de Lima, E. Rauch, M. Veron, W.J. Botta, M. Kawasaki, T.G.
Langdon, An investigation of hydrogen storage in a magnesium-based alloy processed by equal-
channel angular pressing, Int. J. Hydrogen Energy 38 (2013) 8306-8312.
[68] A.M. Jorge Jr., G.F. de Lima, M.R.M. Triques, W.J. Botta, C.S. Kiminami, R.P. Nogueira, A.R.
Yavari, T.G. Langdon, Correlation between hydrogen storage properties and textures induced in
magnesium through ECAP and cold rolling, Int. J. Hydrogen Energy 39 (2014) 3810-3821.
[69] K.J. Jeon, H.R. Moon, A.M. Ruminski, B. Jiang, C. Kisielowski, R. Bardhan, J.J. Urban, Nature
Mater. 10 (2011) 286-290.
