Hydrogenation Properties of Mg83.3Cu7.2Y9.5with Long Period Stacking Ordered Structure and Formation of Polymorphic γ-MgH2

Article abstract of DOI:10.1021/acs.inorgchem.0c02080

Nanosizing is known to affect the hydrogenation properties of magnesium. For this reason, the long period stacking ordered (LPSO) structures, made of the stacking of nanolayers of magnesium and nanolayers of Mg-A-B (with A = rare earth and B = transition metal), were herein considered. A Mg83.3Cu7.2Y9.5 LPSO compound with 18R structure was successfully synthesized. Its hydrogenation properties were investigated at temperatures between 150 and 400 °C. The X-ray diffraction (XRD) analysis indicates that the LPSO structure decomposes into magnesium hydride, yttrium hydride, and an intermetallic compound (Mg2Cu or MgCu2). The pressure composition (PC) isotherm for Mg83.3Cu7.2Y9.5 at 400 °C combined with XRD analysis allows one to understand the three-step hydrogenation pathway, detailed in this paper. At this hydrogenation temperature, the fully hydrogenated compound contains magnesium hydride exclusively crystallized in the most stable tetragonal structure (100% of α-MgH2 was formed). When the pristine LPSO was hydrogenated at lower temperature, the amount of α-MgH2 decreased, while its polymorphic structure, γ-MgH2, appeared. Finally, hydrogenation of Mg83.3Cu7.2Y9.5 at 150 °C led to the formation of γ-MgH2 with a high phase fraction (82% of γ-MgH2/MgH2). These results suggest that the crystallographic structure of the magnesium hydride can be controlled by the hydrogenation temperature of LPSO compounds.

Full text of DOI:10.1021/acs.inorgchem.0c02080

pubs.acs.org/IC
Article
Hydrogenation Properties of Mg83.3Cu7.2Y9.5 with Long Period
Stacking Ordered Structure and Formation of Polymorphic γ‑MgH2
́
Veronique Charbonnier,* Kohta Asano, Hyunjeong Kim, and Kouji Sakaki
Cite This: https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Nanosizing is known to affect the hydrogenation
properties of magnesium. For this reason, the long period stacking
ordered (LPSO) structures, made of the stacking of nanolayers of
magnesium and nanolayers of Mg-A-B (with A = rare earth and B
= transition metal), were herein considered. A Mg83.3Cu7.2Y9.5
LPSO compound with 18R structure was successfully synthesized.
Its hydrogenation properties were investigated at temperatures
between 150 and 400 °C. The X-ray diffraction (XRD) analysis
indicates that the LPSO structure decomposes into magnesium
hydride, yttrium hydride, and an intermetallic compound (Mg2Cu
or MgCu2). The pressure composition (PC) isotherm for
Mg83.3Cu7.2Y9.5 at 400 °C combined with XRD analysis allows
one to understand the three-step hydrogenation pathway, detailed
in this paper. At this hydrogenation temperature, the fully hydrogenated compound contains magnesium hydride exclusively
crystallized in the most stable tetragonal structure (100% of α-MgH2 was formed). When the pristine LPSO was hydrogenated at
lower temperature, the amount of α-MgH2 decreased, while its polymorphic structure, γ-MgH2, appeared. Finally, hydrogenation of
Mg83.3Cu7.2Y9.5 at 150 °C led to the formation of γ-MgH2 with a high phase fraction (82% of γ-MgH2/MgH2). These results suggest
that the crystallographic structure of the magnesium hydride can be controlled by the hydrogenation temperature of LPSO
compounds.
higher for β-MgH2 (−44.4 kJ/mol H214) and for γ-MgH2
1. INTRODUCTION
Hydrogen is a promising energy carrier. One of the bottlenecks
in hydrogen technologies is hydrogen storage. Magnesium is a
(−52.8 kJ/mol H214 or −44.66 kJ/mol H215) than for α-MgH2
(−53.2 kJ/mol H214). Note that these values must be taken
with care because ΔH(α-MgH2) = −53.2 kJ/mol H2 is
overestimated when compared to the experimental value of
−74.5 kJ/mol H2.16
Those two polymorphic structures (β- and γ-MgH2) are
interesting from the point of view of destabilization of
magnesium hydride, since they are thermodynamically
unstable compared to α-MgH2. On the other hand, this also
explains that they are difficult to synthesize. Indeed, the
calculated pressure−temperature phase diagram of MgH2
shows that ultrahigh pressure (on the order of the GPa) is
needed to form β- or γ-MgH2.17 This was confirmed
experimentally by the application of 2.5 GPa of pressurized
inert gas (Ar or He) on α-MgH2.12,13 Other techniques, such
as high pressure torsion (HPT)14 and ball-milling,1,15,18,19 also
light, inexpensive, and abundant metal extensively investigated
for this purpose. Upon hydrogenation, it forms tetragonal α-
MgH2 with a high hydrogen content of 7.6 wt %.1,2 However,
this hydride is thermodynamically very stable: its decom-
position temperature at 1 bar is 278 °C. This temperature is
too high for mobile application.3 In addition to its high
stability, magnesium hydride also suffers from low kinetics for
hydrogen absorption and desorption, so that even higher
temperatures are required to absorb and desorb hydrogen
(350−400 °C).4
In order to destabilize the hydride phase and improve its
reaction kinetics, several processes are under investigation
worldwide, including nanostructuration,1,5 addition of cata-
lysts,1,6−9 and alloying.10 It is worth noticing that, upon
hydrogenation, the formation of polymorphic MgH2 may
occur. MgH2 has several polymorphic forms: orthorhombic γ-
MgH2 with an α-PbO2-type structure (Pbcn)11−13 and cubic β-
MgH2 with a PdF2-type structure (Pa3).12 They are known for
being metastable and, thus, their formation should lead to
lower desorption temperature and faster kinetics. First
principle calculations for the polymorphic structures of
MgH2 revealed that the desorption enthalpy change ΔH is
Received: July 14, 2020
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Inorganic Chemistry
pubs.acs.org/IC
Article
highlighted the transformation (or partial transformation) of α-
MgH2 into γ-MgH2. Both techniques imply external mechan-
ical deformation, leading to the formation of defects, strain,
and reduction of crystallite size. The formation of γ-MgH2 was
also achieved by hydrogenation of Mg-thin films.20−22 It was
attributed to a smaller dimension of film and strain generated
by the clamping effect of the thin film to the substrate.20
Finally, to synthesize γ-MgH2, three key points emerge from
these methods (mechanical deformation and hydrogenation of
thin films): large strain, structural defects, and nanosizing.
Long period stacking ordered (LPSO) compounds (Figure
1) were originally studied for their outstanding mechanical
are made by stacking hcp-Mg layers (gray layers) and Mg-A-B
layers (red layers) along the c-axis. These structures resemble
multilayered thin films but with extremely thin layers (on the
order of nanometers). Therefore, there is a possibility that the
hcp-Mg layer forms MgH2 with different crystal structures
from most stable tetragonal α-MgH2, leading to a lower
desorption temperature of hydrogen.
In this study, the Mg−Cu−Y system was chosen because
Cu,30 Y, and YHx31−33 are known for their catalytic effect on
hydrogenation of Mg. Herein, we investigate the hydro-
genation properties of the Mg83.3Cu7.2Y9.5 LPSO compound.
This study shows a simple and new synthesis route for a large
amount of orthorhombic γ-MgH2.
2. EXPERIMENTAL SECTION
Chunks of Mg (99.9%) from Kojundo Chemical Laboratory Co., Ltd.,
Cu (99.99%) from Furuuchi Chemical Co., and Y (99.9%) from
Nippon Yttrium Co., Ltd. were used to synthesize Mg83.3Cu7.2Y9.5.
Due to its high vapor pressure, an excess of 7 wt % of magnesium was
added. The elements were melted together in an induction furnace
under a helium atmosphere. The obtained ingot was wrapped in
aluminum foil and subsequently annealed for 50 h at 400 °C under an
argon atmosphere.
X-ray diffraction (XRD) analysis was performed with a Rigaku
2500 V diffractometer with Cu Kα radiation in a 2θ range from 3° to
80° with a step size of 0.02°. The FULLPROF program34 based on
the Rietveld method35 was used to determine crystallographic
properties and phase amounts.
The compound was observed with a S-3400N scanning electron
microscope (SEM) from Hitachi. Back scattered electron (BSE)
mode and an energy dispersive X-ray (EDX) detector were used. A
piece of ingot was embedded in a resin and mechanically polished
using diamond paste. After polishing, conductive paste was used in
order to ensure the electronic conductivity.
Pressure composition (PC) isotherms were measured using
Sieverts’ method. The ingot was filed into powder in air.
Approximately 1 g of powder was introduced in a stainless steel
vessel and heated at 400 °C for 5 h in an argon atmosphere. It was
evacuated, and the measurement was started. At the end of the
measurement, the temperature was kept and the sample holder was
evacuated under dynamic vacuum for several hours.
Figure 1. Crystal structures of LPSO compounds draw on the basis of
TEM results: 10H with space group Cmce,26 18R with space group
P3212,28 and 14H with space group P63/mcm.28 The polyhedron in
the Mg-A-B layer is a B6A8 cluster. The composition is indicated
below each crystal structure.
properties,23 but the high content of magnesium in these
LPSO structures makes them possible candidates for hydrogen
storage. They are Mg-based compounds containing a small
amount of rare earth (A) and transition metal (B). Up to now,
three different stacking sequences of LPSO were identified:
10H, 18R, and 14H.24 Crystal structures were proposed on the
basis of transmission electron microscopy (TEM) results.25−29
Table 1 summarizes the ideal compositions and proposed
space groups of the LPSO structures. As evidenced by their
crystallographic structures (Figure 1), the LPSO compounds
3. RESULTS
3.1. Phase Determination and Structural Properties.
Figure 2 shows the XRD pattern of Mg83.3Cu7.2Y9.5. This
composition is close to stoichiometric 14H phase determined
by TEM analysis.25,26,28 The low angle diffraction peak at 5.6°
corresponds to the interplanar distance dhkl of 1.58 nm. This
indicates that a long period crystallographic structure is
formed. This interplanar distance is consistent with the 18R
structure of LPSO36 and much shorter than that for the 14H
structure (1.81 nm) as shown in Table 1. To confirm the
a
Table 1. Tabulated Compositions and Proposed Space Groups for LPSO Compounds with 10H, 18R, and 14H Structures
Mg (at %)
76.7
B (at %)
10
A (at %)
13.3
B/A
0.75
hkl
dhkl (nm)
2θ (deg)
proposed space groups
ref.
26
27
29
26
28
28, 29
26
10H
18R
002
002
001
002
003
002
002
1.31
1.30
1.56
1.56
1.56
1.81
1.82
6.8
6.8
5.6
5.6
5.6
4.9
4.9
Cmce
P63/mcm
C2/m
C2/c
P3212
80.6
83.3
8.3
7.2
11.1
9.5
0.75
0.75
14H
P63/mcm
P6322
a
B is a transition metal, and A is a rare earth. The 2θ column indicates the position of the first diffraction peak for XRD measurements with Cu Kα
radiation.
B
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
3.2. Hydrogenation Properties and Phase Decom-
position upon Hydrogenation. Figure 4a shows the PC
isotherm of Mg83.3Cu7.2Y9.5 measured at 400 °C. 1.86 H/M
(atomic ratio of hydrogen and metal atoms) was absorbed and
1.62 H/M was desorbed, which corresponds to 4.7 wt % of
reversible capacity. When one takes a closer look at the details,
a three step-reaction for hydrogenation is observed: hydrogen
absorption in low hydrogen pressure (below 0.05 MPa) and
two plateaus at 1.2 and 4 MPa. In order to understand the
three step-reaction, XRD patterns were measured at different
states of charging (Figure 4a, points 1 to 4). The detailed
structure information obtained by Rietveld refinement is
shown in Table 2. Note that a small amount of Y2O3 is always
observed, indicating contamination from the air.
The sample hydrogenated at 400 °C under 0.5 MPa of
hydrogen pressure (point 1) shows that the LPSO structure
already decomposed and formed YH2, Mg, and Mg2Cu (Figure
4b). At this point, MgH2 was not observed. The lattice
constants of Mg and YH2 are almost the same as the literature
values. After further hydrogenation (point 2), the XRD analysis
shows that the main phases are α-MgH2, YH3, and Mg2Cu
(Figure 4c). Therefore, the plateau at 1.2 MPa corresponds
both to the further hydrogenation of YH2 into YH3 and to the
hydrogenation of Mg in MgH2. Then, pristine Mg83.3Cu7.2Y9.5
was fully hydrogenated at 400 °C under 4 MPa of hydrogen
pressure (point 3). The XRD pattern highlights the presence of
α-MgH2, YH3, and MgCu2 and traces of YH2 (2.5 wt %)
(Figure 4d). The plateau at 4 MPa can thus be attributed to
the decomposition of Mg2Cu into MgH2 and MgCu2. The
structure of MgH2, which was formed here, is tetragonal (P42/
mnm), and its cell parameters are very close to the literature
values, although hydrogenation of Mg-X systems (X = Sc,40
Ti41) can lead to the formation of (Mg,X)H2 with an Fm3m
space group (FCC structure). Similarly, the structure of YH3
was hexagonal even though Mg-doped Y can form FCC YH3.42
These structural features and the lattice constants indicate that
Mg and Y were not mixed with each other. Finally, the XRD
pattern recorded after PC isotherm measurement at 400 °C
(point 4) shows that dehydrogenation products were YH2, Mg,
and Mg2Cu (Figure 4e). In the whole process, the change in
phase fraction estimated by Rietveld refinements was
reasonable: the capacities derived from Rietveld refinement
(Table 2) are very close to the ones observed by PC isotherm
measurement (Figure 4a).
Figure 2. XRD pattern for Mg83.3Cu7.2Y9.5 after annealing treatment at
400 °C for 50 h. The C2/m space group determined by Saal and
Wolverton29 was used for the Le Bail refinement of Mg83.3Cu7.2Y9.5
with the 18R structure.
nature of the main phase, Le Bail refinement was performed.37
A small amount of silicon was added to the sample as a
reference in order to determine the displacement of the
sample. Then, it was fixed, and the Le Bail refinement of the
main phase was started. The three different space groups
proposed for the 18R structure and presented in Table 1 (C2/
m, C2/c, and P3212) were tried. The C2/m space group gave
the best goodness of fit, and the derived cell parameters are a =
11.157(1) Å, b = 19.478(2) Å, c = 16.620(3) Å, and β =
77.28(1)°. Most of the diffraction peaks were indexed by this
18R structure, while some of the tiny peaks, indicated in Figure
2, correspond to the Mg2Cu phase.
SEM−BSE micrographs are shown in Figure 3. The global
composition derived from SEM−EDX analysis at low
Figure 3. SEM images of Mg83.3Cu7.2Y9.5 annealed at 400 °C for 50 h
at low magnification (a) and high magnification (b).
After the first hydrogenation at 400 °C, the temperature was
kept and the sample holder was evacuated under dynamic
vacuum for several hours. Then, a second hydrogenation was
performed at 150 °C under 5 MPa. The maximal hydrogen
content after the second hydrogenation at 150 °C is 1.42 H/
M. It decreased by 0.44 H/M compared to the first
hydrogenation at 400 °C. No hydrogen desorption was
observed at this temperature. As shown in Figure 5, main
phases are α-MgH2, YH3, and Mg2Cu. Traces of unreacted
YH2 and Mg are also present. At this condition, the reaction
products were almost the same as those at 400 °C although
Mg2Cu was not hydrogenated into α-MgH2 and MgCu2.
In order to understand the temperature effect on hydro-
genation of Mg83.3Cu7.2Y9.5, a PC isotherm of pristine
Mg83.3Cu7.2Y9.5 was measured at 150 °C (Figure 6a). During
the first cycle, one single plateau is lying at 0.14 MPa.
Hydrogen was absorbed up to 1.43 H/M, but hydrogen
desorption was not observed. The sample cell was
subsequently evacuated under dynamic vacuum at 150 °C
magnification (Figure 3a) is Mg83.8 0.2Cu7.4 0.3Y8.8 0.1, which
is close to the target composition. High magnified SEM−BSE
micrographs and EDX analysis show different phases. The
main area is light gray and corresponds to
Mg83.31 0.07Cu7.39 0.05Y9.30 0.03. The dark gray area having a
composition of Mg85.9 0.7Cu6.3 0.2Y7.8
has a Mg content
0.5
slightly higher than the light gray area. As an impurity phase, a
white area, with Mg70.7 0.03Cu14.4 0.3Y14.9 composition, was
0.4
also observed. Light gray and dark gray areas have a
composition close to the stoichiometric 14H compound.
However, the XRD analysis does not show the 14H structure,
whereas the 18R structure is clearly present. Our results
suggest that our 18R phase contains more Mg than the
stoichiometric 18R phase (Table 1). This tendency is
consistent with refs 38 and 39. We will discuss this
compositional difference from the stoichiometric 18R phase
later.
C
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 4. (a) PC isotherm for Mg83.3Cu7.2Y9.5 measured at 400 °C. Points 1, 2, 3, and 4 indicate, respectively, the conditions for the measurements
of Rietveld refined XRD patterns of Mg83.3Cu7.2Y9.5: (b) hydrogenated at 400 °C under 0.5 MPa, (c) hydrogenated at 400 °C under 1.5 MPa, (d)
hydrogenated at 400 °C under 4 MPa, and (e) after PC isotherm measurement at 400 °C.
for several hours. Second and third cycles were measured, but
no absorption occurred. This suggests that evacuation under
dynamic vacuum does not allow hydrogen desorption. Indeed,
high temperatures are required to desorb hydrogen from MgH2
and YH2. At 150 °C, the desorption pressure for MgH2 is
about 0.0007 MPa and desorption kinetics is slow. As for YH2,
it was shown that its desorption starts at 650 °C under
dynamic vacuum.43 Figure 6b shows the XRD pattern recorded
after the first PC isotherm measurement and subsequent
evacuation. This pattern is completely different from those in
Figures 4d and 5.
The two diffraction peaks around 20° indicate that Mg2Cu
remained after hydrogenation. The diffraction peak at 5.6°,
characteristic of 18R LPSO, is still present, but its intensity
drastically decreased. In order to analyze the diffraction pattern
in detail, the powder was crushed in ethanol and only the
supernatant powder was used for Rietveld refinement. This fine
supernatant powder did not contain 18R LPSO (because
LPSO is a very coarse powder, it sank), and MgO and Y2O3
were introduced during this operation; however, the main
feature of the diffraction pattern mostly remained.
First, the Rietveld refinement was attempted by using a
similar strategy as in Figure 5. The quality of the refinement
was poor (Figures 6c and S1a): the position of the main peak
at 31° is close to that for YH3; however, the intensity at 31° is
underestimated, and those at 26.8 and 27.9° are overestimated.
Then, the orthorhombic γ-MgH2 was considered for Rietveld
refinement (Figures 6d and S1b). Note that, now, the
calculated intensities of the main diffraction peak at 31° and
other peaks (especially at 26.8° and 27.9°) match the observed
intensities. In addition, the peak at 25.3°, which was not
indexed using the first refinement models, is indexed by γ-
D
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Table 2. Results of Rietveld Refinement of Mg83.3Cu7.2Y9.5 after Hydrogenation at 400 °C under 0.5 MPa (Figure 4b), 400 °C
under 1.5 MPa (Figure 4c), 400 °C under 4 MPa (Figure 4d), PC Isotherm Measurement at 400 °C (Figure 4e), and
a
Rehydrogenation at 150 °C (Figure 5)
Mg P63/mmc
α-MgH2 P42/mnm
YH2 Fm3m
YH3 P3c1
Mg2Cu Fddd
MgCu2 Fd3m
Y2O3 Ia3
After Hydrogenation at 400 °C under 0.5 MPa
wt %
a (Å)
b (Å)
c (Å)
L (nm)
ε (%)
51.1(2)
3.20959(4)
a
5.2117(2)
b
b
20.7(1)
5.1848(2)
a
24.3(1)
3.9(2)
10.581(3)
a
a
c
5.2786(3)
9.0616(5)
18.338(1)
100(6)
a
90(10)
0.42
0.077
c
C (H/M)
χ2 (%)
0.15
7.91
After Hydrogenation at 400 °C under 1.5 MPa
wt %
a (Å)
b (Å)
c (Å)
L (nm)
ε (%)
2.5(1)
3.208(1)
a
5.217(3)
c
c
50.4(3)
4.5154(4)
a
3.0206(1)
230(70)
0.10
1.1(1)
5.194(1)
a
a
c
21.6(1)
6.3534(6)
a
6.6099(7)
75(5)
0.22
23.5(2)
5.2816(4)
9.068(1)
18.351(2)
71(6)
0.9(1)
10.57(1)
a
a
c
c
0.060
c
C (H/M)
χ2 (%)
1.57
2.03
After Hydrogenation at 400 °C under 4 MPa
wt %
a (Å)
b (Å)
c (Å)
L (nm)
ε (%)
60.2(3)
4.5190(2)
a
3.0235(2)
67(3)
2.4(1)
5.197(1)
a
a
19.1(1)
6.3598(4)
a
6.6123(5)
29(1)
0.16
16.3(1)
7.0597(3)
a
2.0(1)
10.60(1)
a
40(2)
0.19
c
c
0.080
C (H/M)
χ2 (%)
1.91
5.01
After PCT at 400 °C
wt %
a (Å)
b (Å)
c (Å)
L (nm)
ε (%)
53.3(4)
3.2112(2)
a
5.2155(4)
83(7)
23.3(2)
5.1992(4)
a
22.4(3)
5.2785(7)
9.075(1)
18.349(2)
61(6)
1.0(1)
10.59(1)
a
a
c
a
50(3)
0.39
0.095
0.078
c
C (H/M)
χ2 (%)
0.17
1.65
After PCT at 400 °C and Hydrogenation at 150 °C
wt %
a (Å)
b (Å)
c (Å)
L (nm)
ε (%)
C (H/M)
χ2 (%)
3.3(2)
3.209(2)
a
5.228(6)
c
c
55.2(6)
4.5187(6)
a
3.0228(4)
43(3)
13.7(2)
5.1992(7)
a
8.6(2)
6.361(2)
a
6.604(3)
19(2)
0.098
1.60
18.5(3)
5.282(1)
9.078(2)
18.360(5)
c
0.7(1)
10.603(8)
a
a
c
a
19(2)
0.38
0.13
c
c
4.06
a
L is the crystallite size (diameter); C indicates the hydrogen capacity (derived from weight percent of each phase); ε is the strain; χ2 is the
b
goodness of fit. The diffraction peaks are almost as sharp as the silicon reference used to determine instrumental parameters. The contributions of
the crystallite size and strain to the full width at half maximum is thus very low, and their value has no physical meaning. The crystallite size and
c
strain are not given because the phase fraction is too small and/or the diffraction peaks overlap with other phases, making it difficult to obtain
values with a physical meaning.
MgH2. Rietveld refinement (Figure 6d and Table 3) shows
that a high fraction of γ-MgH2 was formed (≈55 wt %). This
value corresponds to 82% of γ-MgH2/MgH2. One can notice
the large value of the a cell parameter and cell volume of γ-
MgH2 compared to the reported value of pure γ-MgH2 (Δa/a
= +2.6%, ΔV/V = +2.39%). This is likely to be triggered by
internal strain, resulting from the low temperature hydro-
genation process, and will be further discussed later. In
addition, it suggests that the magnesium is fully hydrogenated,
so that the hydrogen content of the obtained γ-MgH2 is close
to 2.0 H/M. These results clearly indicate that polymorphic γ-
MgH2 was formed upon hydrogenation of Mg83.3Cu7.2Y9.5 at
150 °C and under a low-pressure condition (0.14 MPa). Note
that the lattice parameters of α-MgH2 and YH3 are almost the
E
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
4. DISCUSSION
4.1. Structural Properties. Although the chemical
composition of our sample determined by SEM−EDX analysis
(Mg83.8 0.2Cu7.4 0.3Y8.8 0.1) is close to the ideal composition of
the 14H structure (Table 1), the XRD analysis shows that the
main phase of our Mg83.3Cu7.2Y9.5 is LPSO with a 18R
structure. Here, we would like to discuss such compositional
deviation from the stoichiometry of LPSO compounds.
Ternary phase diagrams for Mg−Cu−Y33,38 and Mg−Ni−Y
systems39 were reported. Investigations of the Mg−Cu−Y
system, with Mg content higher than 90 at %, have highlighted
the existence of the LPSO compound with the 14H structure
for Mg content between 90.3 and 91.7 at %.38 This magnesium
content is high compared to the ideal composition for the 14H
structure (Table 1). In the Mg−Ni−Y system,39 similar results
were reported and Mg content is from 90.3 to 92.3 at % for
14H, from 84.2 to 90 at % for 18R, and from 78.8 to 85.7 at %
for 10H. In all cases, LPSO contains more Mg than the
stoichiometric composition. This is consistent with our present
result.
Figure 5. Rietveld refined XRD pattern of Mg83.3Cu7.2Y9.5 after one
cycle at 400 °C followed by rehydrogenation at 150 °C.
same as those shown in Table 2, suggesting Mg and Y did not
substitute each other in this process. Therefore, the formation
of polymorphic γ-MgH2 is not related to a substitutional effect.
As far as we know, this is the first observation of formation of
polymorphic γ-MgH2 through the hydrogenation of LPSO
compounds, and its possible mechanism will be discussed later.
Several groups attempted to explain the deviation from the
ideal composition for LPSO compounds. First principle
calculations suggest that the L12 type B6A8 clusters (Figure
1) in 14H Mg83.3Zn7.2A9.5 and 18R Mg80.6Zn8.3A11.1 compounds
are distorted in such a way that A and Zn atoms are displaced
Figure 6. (a) PC isotherms for pristine Mg83.3Cu7.2Y9.5 measured at 150 °C. (b) XRD patterns of Mg83.3Cu7.2Y9.5 after one hydrogenation cycle at
150 °C; (c, d) after the third cycle at 150 °C, the powder was crushed in ethanol in air. The supernatant powder was used for Rietveld refinement.
(c) Rietveld refinement without the addition of the γ-MgH2 phase. The phase abundance of YH3 is overevaluated. (d) Rietveld refinement with the
addition of the γ-MgH2 phase. The fit has been greatly improved, and the results are gathered in Table 3.
F
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Table 3. Results of Rietveld Refinement for Mg83.3Cu7.2Y9.5 after Hydrogenation at 150 °C and Crushing in Ethanol (Figure
a
6d)
γ-MgH2 Pbcn
Δx/x (%)
Mg2Cu Fddd
YH3 P3c1
α-MgH2 P42/mnm
Y2O3 Ia3
MgO Fm3m
wt %
53.8(6)
4.624(2)
5.394(2)
4.964(2)
123.8(1)
23(2)
4.3(1)
5.283(3)
9.074(6)
18.34(2)
879.1(5)
b
1.6(1)
6.36(1)
a
6.634(1)
232.3(4)
b
b
11.7(3)
4.526(5)
a
3.022(7)
61.96(8)
b
b
3.7(1)
10.59 (1)
a
a
1186(1)
b
b
24.9(6)
4.223(6)
a
a
75.29(9)
b
b
a (Å)
b (Å)
c (Å)
V (Å3)
L (nm)
ε (%)
χ2 (%)
+2.44
−0.82
+0.48
+2.08
0.61
b
2.18
a
The column Δx/x compares the cell parameters of γ-MgH2 of the sample to the tabulated cell parameters of γ-MgH2. L is the crystallite size
b
(diameter); ε is the strain; χ2 is the goodness of fit. The crystallite size and strain are not given because the phase fraction is too small and/or the
diffraction peaks overlap with other phases, making it difficult to obtain values with a physical meaning.
toward and outward the center of the cluster, respectively. This
leads to the formation of a large interstitial site at the center of
the B6A8 clusters.28 These calculations also show that Mg, as
interstitial atom, is the most preferred element to occupy the
center of the B6A8 cluster.29 If we consider that each Cu6Y8
cluster contains only Mg atom as interstitial atom, the Mg
contents will increase to 83.7, 80.8, and 77.0 at % for 14H,
18R, and 10H structures, respectively. This interstitial model
does not make a substantial contribution to the composition.
Therefore, such a large deviation of Mg content in LPSO
cannot be solely explained by occupation at the center of
Cu6Y8 clusters.
Next, we simply consider the possibility that some of Cu6Y8
clusters are replaced by Mg14 clusters. In the 18R structure
model, there are four Cu6Y8 clusters in one unit cell. If one
(two) of the eight Cu6Y8 clusters in two unit cells is replaced,
the chemical composition becomes Mg83.0Cu7.29Y9.71
(Mg85.42Cu6.25Y8.33). Our SEM−EDX results are close to
these cases. As described above, Xu et al. showed that the
composition range of Mg for 14H and 18R in the Mg−Ni−Y
system are from 90.3 to 92.3 at % and from 84.2 to 90 at %,
respectively.39 In both cases, the maximum Mg content can be
explained by the model where a half of Cu6Y8 clusters in the
unit cell are replaced by Mg14 clusters. To further examine the
phase-composition relationship of Mg−Cu−Y LPSO, we
synthesized a Mg−Cu−Y compound with a higher Mg
content. The SEM−EDX analysis showed a chemical
composition of Mg89.0 0.6Cu5.1 0.2Y5.9 0.5. As we expected, the
crystal structure of this main phase is the 14H structure. This
suggests that two of the Cu6Y8 clusters are replaced in three
unit cells of the 14H structure and not only 18R but also 14H
has more Mg in Mg−Cu−Y system. This result indicates that
the LPSO phases can tolerate a large difference in their
stoichiometric compositions. The replacement of B6A8 clusters
by Mg clusters may explain it.
suggests that extra Cu atoms might be inserted in the Cu6Y8
cluster as an interstitial atom in the Mg−Cu−Y system. If we
consider two 18R cells and insert Cu at the center of 3 of the 8
clusters (in two 18R cells), the Cu/Y ratio becomes 0.80. It
suggests that Cu atoms might be inserted in the B6A8 cluster.
However, the first principle calculation for the Mg−Y−Zn
system has shown that Zn is the most unstable element to
occupy the center of B6A8 clusters, even though Zn has the
smallest atomic radius among them.26
Another possibility is the substitution. Egusa and Abe
proposed that a substitution of A and B by Mg is very likely to
happen.28 When one considers the atomic radius of Y (1.776
Å), Cu (1.278 Å), and Mg (1.599 Å), Mg is more likely to
substitute Y rather than Cu (Hume−Rothery rules). Finally,
we considered that one (two) of the eight Cu6Y8 clusters in
two unit cells of the 18R structure is replaced by one (two)
Mg14 cluster and that four (two) Y atoms are replaced by Mg
atoms. In this case, the composition becomes Mg85.76Cu6.25Y7.83
(Mg83.68Cu7.29Y9.03), which is extremely close to our SEM−
EDX data and explains the high Mg content as well as the high
Cu/Y ratio. Our results clearly show that the high Mg content
is mainly due to the substitution of Cu and Y by Mg. It would
be interesting to confirm the stability of this simple LPSO
structure model with the density functional theory or by
confronting experimental annular bright-field scanning trans-
mission electron micrographs to calculated ones.
4.2. Hydrogenation Reaction and Change in Crystal
Structure. As shown in Figures 4 and 6, 18R Mg83.3Cu7.2Y9.5
decomposes upon hydrogenation and the crystal structure of
the resulting Mg hydride depends on the hydrogenation
temperature. Here, we would like to discuss the reaction
pathway and the dependence on hydrogenation temperature of
the crystal structure of Mg hydride.
4.2.1. High Hydrogenation Temperature: 400 °C. The
XRD pattern of the fully hydrogenated sample (Figure 4d)
shows that the Mg83.3Cu7.2Y9.5 LPSO compound decomposed
into YH3, α-MgH2, and MgCu2 upon hydrogenation with a
maximal capacity of 1.86 H/M (4.9 wt %). This is consistent
with previous studies investigating the hydrogenation of 18R-
type Mg−Ni−Y LPSO compounds at 300−350 °C.44,45
The enthalpies of formation of YH2, YH3, and α-MgH2 are
−228, −178, and −74.5 kJ·mol−1 H2, respectively.46 Con-
sequently, the formation of YH2 is the most favorable and
should take place prior to the formation of α-MgH2. Such
behavior was observed by Zlotea et al. for Mg24Y5,46 Ishikawa
et al. for Mg85Zn6Y9,47 and Li et al., who studied hydrogenation
of the Mg−Ni−Y system by means of in situ synchrotron
Now, we would like to focus on the B/A ratio. According to
SEM−EDX analysis, our sample shows a B/A ratio close to 0.8
for both Mg-rich and highly Mg-rich 18R phases, which is
slightly larger than the ideal value (0.75 = 6/8). It means that
Cu content is slightly larger or Y content is slightly lower. This
is in line with the results of Jiang et al.,38 who measured a Cu/
Y ratio of 0.82 0.07 for the 14H structure in the Mg−Cu−Y
system after annealing at 400 °C for 1500 h. In fact, ABF-
STEM images of 18R and 10H Mg−Zn−Y showed that Mg, Y,
Zn, or a vacancy occupies the interstitial site at the center of
the B6A8 clusters.26 The atomic radius of Cu is the smallest
among our constituent elements. Therefore, the Cu/Y ratio
G
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Table 4. Hydrogenation and Dehydrogenation Pathways for Mg83.3Cu7.2Y9.5 at 400 °C and the Amount of Hydrogen Associated
with Each Step
hydrogen content (H/
M)
by step
cumulative
hydrogenation pathway
Mg83.8 0.2Cu7.4 0.3Y8.8
→ 69Mg + 8.8YH2 + 7.4Mg2Cu
→ 69α-MgH2 + 8.8YH3 + 7.4Mg2Cu
→ 80.1α-MgH2 + 8.8YH3 + 3.7MgCu2
0.176
1.468
0.222
0.176
1.644
1.866
0.1
dehydrogenation pathway
80.1α-MgH2 + 8.8YH3 + 3.7MgCu2
→ 69α-MgH2 + 8.8YH3 + 7.4Mg2Cu
→ 69Mg + 8.8YH2 + 7.4Mg2Cu
−0.222
−1.468
−0.222
−1.690
XRD.48 Similar results were obtained for LaMg12, where first
disproportionation of the alloy into lanthanum hydride and
metallic magnesium takes place, followed by hydrogenation of
Mg,49 matching with the enthalpy of formation of each
hydride. The PC isotherm of Mg83.3Cu7.2Y9.5 recorded at 400
°C also follows the same steps. When one takes the average
LPSO composition obtained from the SEM−EDX study, i.e.,
Mg83.8 0.2Cu7.4 0.3Y8.8 0.1, its decomposition should lead to the
formation of 8.8 YH2 as the first step, which is equivalent to
0.176 H/M (Table 4). This is very close to the measured
hydrogen content in the low pressure range, 0.18 H/M. Thus,
Y is consumed to form YH2 at the first step of hydrogenation.
The proportions of remaining elements are 91 at % of Mg and
9 at % of Cu. According to the phase diagram of Cu−Mg,50 in
this composition range and temperature (400 °C), there is
formation of Mg and Mg2Cu, which is in good agreement with
our XRD results showing YH2, Mg, and Mg2Cu as
decomposition products (Figure 4b).
When one takes into account the enthalpy, the next reaction
should be the hydrogenation of YH2 into YH3. However, the
plateau was not observed clearly in our PC isotherms, probably
because of the slow kinetics. Then, the hydrogenation of Mg
into α-MgH2 should occur at about 1.86 MPa at 400 °C (since
ΔH = −74.5 kJ·mol−1 H2 and ΔS = 135 J·K−1·mol−1 H216).
This is close to the plateau at 1.2 MPa and consistent with the
results of Couillaud et al.51 If YH2 and metallic Mg are
hydrogenated, the increase in hydrogen content should be
1.468 H/M. This is close to the width of this plateau in the PC
isotherm (Figure 4a). The XRD analysis of Mg83.3Cu7.2Y9.5
hydrogenated at 400 °C under 1.5 MPa shows that the sample
mainly consists of α-MgH2, Mg2Cu, and YH3 (Figure 4c and
Table 2). The detection of small quantities of both Mg and
YH2 suggests that their hydrogenation into α-MgH2 and YH3
happens simultaneously. Therefore, this intermediate plateau is
attributed to the formation of α-MgH2 and YH3.
high (82%). Hydrogenation properties of LPSO phases have
been investigated,44,45,47,48 but the formation of γ-MgH2 has
not been reported. This is the first time that γ-MgH2 is directly
synthesized by a low pressure (about 0.14 MPa) and low
temperature (150 °C) route with such a high content. Here,
therefore, we would like to discuss the possible formation
mechanism of γ-MgH2 from the viewpoints of nanosizing,
strain, and structural defects.
γ-MgH2 is known to form through the phase transition of α-
MgH2 under very high pressure (2.5 GPa) at 250 °C as a
thermally equilibrium phase.13,17 On the other hand, the
synthesis condition in the present study is about 0.14 MPa and
150 °C. According to the pressure−temperature phase diagram
of MgH2, this is far from its equilibrium condition.17
Therefore, γ-MgH2 formed by hydrogenation of LPSO
compounds must be in a metastable state. In order to confirm
it, the effect of hydrogenation temperature on the crystal
structure of MgH2 was investigated. As described in the
experimental results, a high fraction of γ-MgH2 (82%) was
formed at 150 °C, while only α-MgH2 was observed at 400 °C.
When it was hydrogenated at 200 °C, clear Bragg peaks from γ-
MgH2 still appeared with those from α-MgH2 as shown in
Figure 7 but peak intensities from γ-MgH2 decreased. The
At 400 °C, disproportionation of Mg2Cu into α-MgH2 and
MgCu2 occurs under 5.4 MPa of hydrogen.30,51 This is
equivalent to the absorption of 0.222 H/M (Table 4). The
expected plateau pressure and capacity match well with the last
plateau at 4 MPa. In fact, the XRD pattern after hydrogenation
at 400 °C under 4 MPa of hydrogen pressure showed the
formation of MgCu2 instead of Mg2Cu. Therefore, the last
reaction corresponds to the hydrogenation of Mg2Cu. Finally,
the hydrogenation/dehydrogenation process of Mg83.3Cu7.2Y9.5
at 400 °C is summarized in Table 4.
Figure 7. XRD patterns of Mg83.3Cu7.2Y9.5 after hydrogenation at 150,
200, and 400 °C.
results from the Rietveld refinement are shown in Figure S2
and Table S1. The relative content of γ-MgH2, γ-MgH2/MgH2,
was about 50%. It is clear that the fraction of γ-MgH2/MgH2
depends on the temperature and that the low hydrogenation
temperature promotes the formation of γ-MgH2. These results
indicate that, when hydrogenated at high temperature, LPSO
forms thermodynamically stable α-MgH2 and, when hydro-
genated at low temperature, LPSO preferably forms metastable
4.2.2. Formation of Polymorph of MgH2. Surprisingly, the
formation of less stable γ-MgH2 is clearly observed after
hydrogenation of Mg83.3Cu7.2Y9.5 at 150 °C (Figure 6 and
Table 3). In addition, the relative content of γ-MgH2/MgH2 is
H
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 8. Crystal structures of (1) hcp Mg with P63/mmc space group,63 (2) tetragonal α-MgH2 with P42/mnm space group,64 and (3)
orthorhombic γ-MgH2 with Pbcn space group65 and their coordination system. The large colored balls are Mg atoms; the small black balls are H
atoms. The Mg atoms with light colors (light green, light blue, and light brown) belong to the front face of the cell with black edges. The dark Mg
atoms are located on the back face of the cell, and the Mg atoms with intermediate color are in the intermediate plane. The cells with colored faces
are the conventional cells. The vertical axis on the left represents the direction of c for the LPSO. It is parallel to the c-axis of hcp-Mg (Figure 1).
The evolution of the lattice dimensions of the hydrides are compared to hcp-Mg. The three red arrows indicate the slight displacement of one
atomic layer, leading to the formation of α-MgH2.
γ-MgH2. Therefore, an important parameter to form γ-MgH2
via hydrogenation of LPSO compounds is the hydrogenation
temperature.
MgH2 resulting from hydrogenation at 400 °C contains a
negligible amount of strain. Therefore, our results suggest that
the key parameter to form γ-MgH2 is lattice strain and that it
affects the phase stability.
Formation of metastable MgH2 was reported by the
following methods: HPT of α-MgH2,14 ball-milling of α-
MgH2,1,18,19 and hydrogenation of Mg-thin films.20−22,52−56
Edalati et al. synthesized γ-MgH2 by HPT of α-MgH2 at 27
°C.14 They demonstrated that the formation of γ-MgH2 is
promoted with an increase in the applied pressure and the
number of turns (deformation cycle). They explained that the
formation of γ-MgH2 was triggered by the formation of defects
induced by strain. In addition, TEM analysis evidenced that
HPT leads to smaller grain size (down to 70 35 nm). Ball-
milling also implies external mechanical deformation, and it
was reported that it induced the formation of γ-MgH2 even
though the fraction is only 18 wt %.1 Similarly to HPT, it was
also concluded that the formation of γ-MgH2 is due to the
creation of strains and structural defects. Upon these
mechanical deformation processes implying external stress,
strain (structural defects) and/or small grain size must affect
the formation of γ-MgH2.
From here, we will discuss crystallite size and lattice strain
estimated by our Rietveld refinement. When LPSO is
hydrogenated at 400 °C or rehydrogenated at 150 °C after
hydrogenation at 400 °C, the crystallite size and lattice strain
for α-MgH2 are 67 nm and 0.080% and 43 nm and 0.13%,
respectively. Those for γ-MgH2 hydrogenated at 150 °C are 23
nm and 0.61% (the lattice strain and crystallite size for α-
MgH2 cannot be evaluated because of the small phase
fraction). Those for α-MgH2 and γ-MgH2 hydrogenated at
200 °C are 10 nm and 0.18% and 25 nm and 0.69%,
respectively. In all cases, crystallite size for both α-MgH2 and γ-
MgH2 are smaller than that for HPT samples.14 Therefore, it is
difficult to conclude whether there is a limitation for crystallite
size to form γ-MgH2. On the other hand, the lattice strain for
γ-MgH2 is always much higher than for α-MgH2. Especially, α-
The formation of metastable orthorhombic MgH2 (either γ-
MgH2 or CaCl2-type MgH2) was also observed upon direct
hydrogenation of Mg-based thin films without external
stress.20,21,52−56 The trend that comes out is that α-MgH2 is
more likely to form when the Mg-thin film is porous or exhibits
columns perpendicular to the surface, whereas the content of
orthorhombic MgH2 is higher when the film is dense. It is
thought that the low porosity of the dense thin films restricts
the free expansion of magnesium during its hydrogenation into
MgH2, which favors the formation of metastable MgH2. In fact,
Siviero et al., who hydrogenated 200 nm thick Mg-thin films at
100 °C, proposed that it was the result of compressive stress
generated by clamping of the film to the substrate.20 Nagengast
et al. estimated that a stress of about 3.2 GPa was induced in
their Mg-based thin films upon hydrogenation, resulting in the
formation of γ-MgH2.22 Therefore, the spatial constraint or
compressive stress is also an important factor in the formation
of γ-MgH2.
The LPSO compounds consist of infinite layers of hcp-Mg
sandwiched by Mg-A-B layers and can be compared to dense
Mg-thin films. It suggests the possibility of formation of
metastable MgH2 upon hydrogenation of LPSO compounds,
as observed in the present work. In addition, the plane of the
hcp-Mg layers, which is perpendicular to the cLPSO-axis, is
(002) (Figure 1), like what is usually observed for Mg-thin
films. From this basis and similarly to the images shown in refs
52 and 53, we examined the relationship between hcp-Mg (1)
and the magnesium hydrides, α-MgH2 (2) and γ-MgH2 (3)
(Figure 8). The cells with black edges can easily be compared
to each other, and for a better understanding, the conventional
cells (with colored faces) are also represented. It turns out that,
in the case of the transformation of hcp-Mg to α-MgH2, the
I
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
cell expansion is mainly parallel to chcp‑Mg (+22.7%) and the
atoms located in the middle plane shifted following the red
arrows, in good agreement with refs 52−54. As for the
transformation to orthorhombic MgH2, in the case of certain
thin films, a shear deformation of the hcp-Mg cell leads to
MgH2 with a CaCl2-type structure.52,53 On the other hand, in
the present study, the hcp-Mg cell expands along the [210]
direction and slightly shrinks along the c-axis to form γ-MgH2
with an α-PbO2-type structure.
hydrogen diffusion.62 To determine the cycling properties of
the metastable γ-MgH2, differential scanning calorimetry
(DSC) was carried out under hydrogen pressure (Figure
S3). The XRD analysis performed after one dehydrogenation/
hydrogenation cycle showed that the γ-MgH2 phase
completely disappeared and is replaced by α-MgH2. However,
the peaks for absorption and desorption are not shifted upon
cycling. A detailed thermodynamic study of γ-MgH2 is still on
going.
The structural relationship between hcp-Mg and γ-MgH2 as
well as the large strain observed in our γ-MgH2 (Table 3) allow
us to propose a possible mechanism for the formation of γ-
MgH2 in LPSO compounds. The hydrogenation starts in the
Mg-A-B layers with the formation of YH2. Because of the low
temperature, YH2 nucleates randomly and aggregates in each
Mg-A-B layer, which results in the collapse of Cu6Y8 clusters.
The nucleation of MgH2 can easily occur at lattice defects or
contaminating phases.57 Therefore, the formation of MgH2
easily takes place near the YH2 clusters. Local expansion of the
Mg-A-B layers, triggered by the formation of YH2 clusters,
generates compressive stress perpendicular to the Mg layers
and tensile stress parallel to the Mg layers. The compressive
stress prevents Mg from expanding along the c-axis upon
hydrogenation, whereas the tensile stress favors its expansion
in the (ab)-plane. As already shown in Figure 8 (3), a slight
contraction of the cell occurs in the chcp‑Mg direction (−5.2%).
This suggests that γ-MgH2 is preferably formed when it is
difficult for Mg to expand along the c-axis ([0001] direction),
just like our LPSO hydrogenated at 150 °C, where the Mg
layers undergo compressive stress parallel to c. The most
significant change of the Mg cell upon formation of γ-MgH2
occurs along the Mg [2−10] direction. This expansion is
extremely large for γ-MgH2 formed in the present study
(+44%).
Note that layered-phase transformation is also reported in
the similar stacking structure materials, ABy compounds (2 < y
< 5). The structure of ABy compounds consists of the piling of
[AB2] and [AB5] units along the c-axis. These compounds
showed stepped hydrogenation. The [AB2] units hydrogenate
to form [AB2Hx] units in the first step, and then, [AB5] units
hydrogenate to form the full hydride. In the first hydrogenation
step, the expansion of the [AB2] units along the c-axis can
increase by more than 50%,58,59 while a shrinkage of the outer
part of the [AB5] units (close to the [AB2] units) was observed
(down to −8% along the c-axis60). The hydrogen-induced
structural change in our LPSO compounds at 150 °C is
probably similar to the structural change in ABy compounds.
The low hydrogenation temperature inhibits the release of
strains, while high temperature processes can release it.
Therefore, strains highlighted by X-ray diffraction analysis
explain the formation of metastable γ-MgH2 at low hydro-
genation temperature and the increase in α-MgH2 content at
higher hydrogenation temperature (Figure 7). Vajeeston et al.
calculated the total energy of several MgH2 polymorphs.61 It
turns out that the difference between α- and γ-MgH2 is very
tiny (a few meV); thus, the formation of metastable γ-MgH2
may become possible by the introduction of constraints or
strains.
5. CONCLUSIONS
In order to understand the effect of nanosizing of magnesium
on hydrogenation properties, Mg83.3Cu7.2Y9.5 was synthesized
and its hydrogenation properties were investigated at different
temperatures. The Mg83.3Cu7.2Y9.5 LPSO compound with a
18R structure was successfully synthesized. At each temper-
ature, the LPSO structure decomposed upon hydrogenation.
At 400 °C, Mg83.3Cu7.2Y9.5 undergoes a three-step dispropor-
tionation into α-MgH2, YH3, and MgCu2. Interestingly, with a
decrease in hydrogenation temperature, the polymorphic γ-
MgH2 phase was formed together with α-MgH2 and the
fraction of γ-MgH2/MgH2 reached 82% at 150 °C. We
propose that the formation of γ-MgH2 may result from a
combination of several factors: first, the small difference in
total energy between α-MgH2 and γ-MgH2; second, the
structure of LPSO compounds providing regular and nano-
sized Mg layers; third, the inner stress rising during the
hydrogenation process induced by the peculiar LPSO structure
and promoted by the low hydrogenation temperature, which
inhibits atomic diffusion.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.0c02080.
■
sı
XRD analysis for Mg83.3Cu7.2Y9.5 after hydrogenation at
150 and 200 °C; results of Rietveld refinement; DSC
analysis for Mg83.3Cu7.2Y9.5 after hydrogenation at 150
°C (PDF)
AUTHOR INFORMATION
Corresponding Author
■
́
Veronique Charbonnier − Energy Process Research Institute,
National Institute of Advanced Industrial Science and
Technology (AIST), Tsukuba West, Tsukuba, Ibaraki 305-
8569, Japan; orcid.org/0000-0002-9939-2213;
Email: v.charbonnier@aist.go.jp
Authors
Kohta Asano − Energy Process Research Institute, National
Institute of Advanced Industrial Science and Technology
(AIST), Tsukuba West, Tsukuba, Ibaraki 305-8569, Japan;
orcid.org/0000-0003-4208-7303
Hyunjeong Kim − Energy Process Research Institute, National
Institute of Advanced Industrial Science and Technology
(AIST), Tsukuba West, Tsukuba, Ibaraki 305-8569, Japan
Kouji Sakaki − Energy Process Research Institute, National
Institute of Advanced Industrial Science and Technology
(AIST), Tsukuba West, Tsukuba, Ibaraki 305-8569, Japan;
orcid.org/0000-0003-4781-1073
The presence of γ-MgH2 is interesting because it was
observed that the compounds containing γ-MgH2 exhibit
higher desorption pressure and improved kinetics compared to
pure α-MgH2.1,15,18 This behavior is due to the low stability of
γ-MgH2 and its distorted structure, leading to enhanced
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.0c02080
J
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
(16) Stampfer, J. F.; Holley, C. E.; Suttle, J. F. The Magnesium-
Hydrogen System. J. Am. Chem. Soc. 1960, 82, 3504−3508.
(17) Moser, D.; Baldissin, G.; Bull, D. J.; Riley, D. J.; Morrison, I.;
Ross, D. K.; Oates, W. A.; Noreus, D. The pressure-temperature phase
diagram of MgH2 and isotopic substitution. J. Phys.: Condens. Matter
2011, 23, 305403.
(18) Varin, R. A.; Czujko, T.; Wronski, Z. Particle size, grain size and
-MgH2 effects on the desorption properties of nanocrystalline
commercial magnesium hydride processed by controlled mechanical
milling. Nanotechnology 2006, 17, 3856−3865.
(19) Rizo-Acosta, P.; Cuevas, F.; Latroche, M. Hydrides of early
transition metals as catalysts and grain growth inhibitors for enhanced
reversible hydrogen storage in nanostructured magnesium. J. Mater.
Chem. A 2019, 7, 23064−23075.
(20) Siviero, G.; Bello, V.; Mattei, G.; Mazzoldia, P.; Battaglin, G.;
Bazzanella, N.; Checchetto, R.; Miotello, A. Structural evolution of
Pd-capped Mg thin films under H2 absorption and desorption cycles.
Int. J. Hydrogen Energy 2009, 34, 4817−4826.
(21) Gautam, Y. K.; Chawla, A. K.; Walia, R.; Agrawal, R. D.;
Chandra, R. Hydrogenation of Pd-capped Mg thin films prepared by
DC magnetron sputtering. Appl. Surf. Sci. 2011, 257, 6291−6295.
(22) Nagengast, D. G.; van Gogh, A. T. M.; Kooij, E. S.; Dam, B.;
Griessen, R. Contrast enhancement of rare-earth switchable mirrors
through microscopic shutter effect. Appl. Phys. Lett. 1999, 75, 2050.
(23) Lu, F.; Ma, A.; Jiang, J. Review on long-period stacking-ordered
structures in Mg-Zn-RE alloys. Rare Met. 2012, 31, 303−310.
(24) Abe, E.; Ono, A.; Itoi, T.; Yamasaki, M.; Kawamura, Y.
Polytypes of long-period stacking structures synchronized with
chemical order in a dilute Mg-Zn-Y alloy. Philos. Mag. Lett. 2011,
91, 690−696.
(25) Yokobayashi, H.; Kishida, K.; Inui, H.; Yamasaki, M.;
Kawamura, Y. Enrichment of Gd and Al atoms in the quadruple
close packed planes and their in-plane long-range ordering in the long
period stacking-ordered phase in the Mg-Al-Gd system. Acta Mater.
2011, 59, 7287−7299.
(26) Kishida, K.; Nagai, K.; Matsumoto, A.; Yasuhara, A.; Inui, H.
Crystal structures of highly-ordered long-period stacking-ordered
phases with 18R, 14H and 10H-type stacking sequences in the Mg-
Zn-Y system. Acta Mater. 2015, 99, 228−239.
(27) Yamasaki, M.; Matsushita, M.; Hagihara, K.; Izuno, H.; Abe, E.;
Kawamura, Y. Highly ordered 10H-type long-period stacking order
phase in a Mg-Zn-Y ternary alloy. Scr. Mater. 2014, 78−79, 13−16.
(28) Egusa, D.; Abe, E. The structure of long period stacking/order
Mg-Zn-RE phases with extended non-stoichiometry ranges. Acta
Mater. 2012, 60, 166−178.
(29) Saal, J. E.; Wolverton, C. Thermodynamic stability of Mg-based
ternary long-period stacking ordered structures. Acta Mater. 2014, 68,
325−338.
(30) Karty, A.; Grunzweig-Genossar, J.; Rudman, P. S. Hydriding
and dehydriding kinetics of Mg in a Mg/Mg2Cu eutectic alloy:
Pressure sweep method. J. Appl. Phys. 1979, 50, 7200−7209.
(31) Si, T. Z.; Liu, Y. F.; Zhang, Q. A. Hydrogen storage properties
of the supersaturated Mg12YNi solid solution. J. Alloys Compd. 2010,
507, 489−493.
(32) Kalinichenka, S.; Rontzsch, L.; Riedl, T.; Gemming, T.;
Weißgarber, T.; Kieback, B. Microstructure and hydrogen storage
properties of melt-spun Mg-Cu-Ni-Y alloys. Int. J. Hydrogen Energy
2011, 36, 1592−1600.
(33) Chen, R.; Ding, X.; Chen, X.; Li, X.; Su, Y.; Guo, J.; Ding, H.;
Fu, H. In-situ hydrogen-induced evolution and de/hydrogenation
behaviors of the Mg93Cu7‑xYx alloys with equalized LPSO and eutectic
structure. Int. J. Hydrogen Energy 2019, 44, 21999−22010.
(34) Rodriguez-Carvajal, J. Recent advances in magnetic structure
determination by neutron powder diffraction. Phys. B 1993, 192, 55−
69.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by Japan Society for the Promotion
of Science under a short-term program (PE18009) and
International Joint Research Program for Innovative Energy
Technology by the Ministry of Economy, Trade and Industry
(METI).
REFERENCES
■
(1) Huot, J.; Liang, G.; Boily, S.; Van Neste, A.; Schulz, R. Structural
study and hydrogen sorption kinetics of ball-milled magnesium
hydride. J. Alloys Compd. 1999, 293−295, 495−500.
(2) Crivello, J.-C.; Denys, R. V.; Dornheim, M.; Felderhoff, M.;
Grant, D. M.; Huot, J.; Jensen, T. R.; de Jongh, P.; Latroche, M.;
Walker, G. S.; Webb, C. J.; Yartys, V. A. Mg-based compounds for
hydrogen and energy storage. Appl. Phys. A: Mater. Sci. Process. 2016,
122, 85.
(3) Larsson, P.; Araujo, C. M.; Larsson, J. A.; Jena, P.; Ahuja, R. Role
of catalysts in dehydrogenation of MgH2 nanoclusters. Proc. Natl.
Acad. Sci. U. S. A. 2008, 105, 8227−8231.
(4) Zaluska, A.; Zaluski, L.; Strom-Olsen, J. O. Nanocrystalline
magnesium for hydrogen storage. J. Alloys Compd. 1999, 288, 217−
225.
(5) Zlotea, C.; Latroche, M. Role of nanoconfinement on hydrogen
sorption properties of metal nanoparticles hybrids. Colloids Surf., A
2013, 439, 117−130.
(6) Oelerich, W.; Klassen, T.; Bormann, R. Metal oxides as catalysts
for improved hydrogen sorption in nanocrystalline Mg-based
materials. J. Alloys Compd. 2001, 315, 237−242.
(7) Cheng, Y.; Zhang, W.; Liu, J.; Cheng, K.; Zhao, Z. Effect of the
nanometric LiFeO4 on the hydrogen storage properties of MgH2. Int.
J. Hydrogen Energy 2017, 42, 356−365.
(8) Liang, G.; Huot, J.; Boily, S.; Van Neste, A.; Schulz, R. Catalytic
effect of transition metals on hydrogen sorption in nanocrystalline ball
milled MgH2-Tm (Tm = Ti, V, Mn, Fe and Ni) systems. J. Alloys
Compd. 1999, 292, 247−252.
(9) Aguey-Zinsou, K.-F.; Ares Fernandez, J. R.; Klassen, T.;
Bormann, R. Using MgO to improve the (de)hydriding properties
of magnesium. Mater. Res. Bull. 2006, 41, 1118−1126.
(10) Manivasagam, T. G.; Kiraz, K.; Notten, P. H. L. Electro-
chemical and Optical Properties of Magnesium-Alloy Hydrides
Reviewed. Crystals 2012, 2, 1410−1433.
(11) Bortz, M.; Bertheville, B.; Bottger, G.; Yvon, K. Structure of the
high pressure phase 1-MgH2 by neutron powder diffraction. J. Alloys
Compd. 1999, 287, L4−L6.
(12) Vajeeston, P.; Ravindran, P.; Hauback, B. C.; Fjellvag, H.;
Kjekshus, A.; Furuseth, S.; Hanfland, M. Structural stability and
pressure-induced phase transitions in MgH2. Phys. Rev. B: Condens.
Matter Mater. Phys. 2006, 73, 224102.
(13) Bastide, J.-P.; Bonnetot, B.; Letoffe, J.-M.; Claudy, P.
Polymorphisme de l’hydrure de magnesium sous haute pression.
Mater. Res. Bull. 1980, 15, 1215−1224.
(14) Edalati, K.; Kitabayashi, K.; Ikeda, Y.; Matsuda, J.; Lia, H.-W.;
Tanaka, I.; Akiba, E.; Horita, Z. Bulk nanocrystalline gamma
magnesium hydride with low dehydrogenation temperature stabilized
by plastic straining via high-pressure torsion. Scr. Mater. 2018, 157,
54−57.
́
(35) Rietveld, H. M. A Profile Refinement Method for Nuclear and
Magnetic Structures. J. Appl. Crystallogr. 1969, 2, 65−71.
(36) Okuda, H.; Horiuchi, T.; Tsukamoto, T.; Ochiai, S.; Yamasaki,
M.; Kawamura, Y. Evolution of long-period stacking ordered
(15) Zhou, S.; Zhang, Q.; Chen, H.; Zang, X.; Zhou, X.; Wang, R.;
Jiang, X.; Yang, B.; Jiang, R. Crystalline structure, energy calculation
and dehydriding thermodynamics of magnesium hydride from
reactive milling. Int. J. Hydrogen Energy 2015, 40, 11484−11490.
K
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
structures on annealing as-cast Mg85Y9Zn6 alloy ingot observed by
synchrotron radiation small-angle scattering. Scr. Mater. 2013, 68,
575−578.
three-layered Pd/Mg/Pd thin films. J. Alloys Compd. 2002, 330−332,
526−530.
(56) Le-Quoc, H.; Lacoste, A.; Miraglia, S.; Bechu, S.; Bes, A.;
Laversenne, L. MgH2 thin films deposited by one-step reactive plasma
sputtering. Int. J. Hydrogen Energy 2014, 39, 17718−17725.
(57) Mooij, L.; Dam, B. Hysteresis and the role of nucleation and
growth in the hydrogenation of Mg nanolayers. Phys. Chem. Chem.
Phys. 2013, 15, 2782−2792.
(37) Le Bail, A. Whole powder pattern decomposition methods and
applications: A retrospection. Powder Diffr. 2005, 20, 316−326.
(38) Jiang, M.; Su, X.; Li, H.; Ren, Y.; Qin, G. The phase equilibria
and thermal stability of the long-period stacking ordered phase in the
Mg-Cu-Y system. J. Alloys Compd. 2014, 593, 141−147.
(39) Xu, K.; Liu, S.; Huang, D.; Du, Y. Experimental investigation of
the isothermal section of the Mg-Ni-Y system with LPSO phases at
400°C. J. Mater. Sci. 2018, 53, 9243−9257.
(58) Yartys, V. A.; Riabov, A. B.; Denys, R. V.; Sato, M.; Delaplane,
R. G. Novel intermetallic hydrides. J. Alloys Compd. 2006, 408−412,
273−279.
(59) Iwase, K.; Sakaki, K.; Nakamura, Y.; Akiba, E. Phase
Transformation and Crystal Structure of La2Ni7Hx Studied by in
situ X-ray Diffraction. Inorg. Chem. 2010, 49, 8763−8768.
(60) Charbonnier, V.; Zhang, J.; Monnier, J.; Goubault, L.; Bernard,
P.; Magen, C.; Serin, V.; Latroche, M. Structural and hydrogen
storage properties of Y2Ni7 deuterides studied by neutron powder
diffraction. J. Phys. Chem. C 2015, 119, 12218−12225.
(61) Vajeeston, P.; Ravindran, P.; Kjekshus, A.; Fjellvag, H.
Pressure-Induced Structural Transitions in MgH2. Phys. Rev. Lett.
2002, 89, 175506.
(40) Latroche, M.; Kalisvaart, P.; Notten, P. H. L. Crystal structure
of Mg0.65Sc0.35Dx deuterides studied by X-ray. J. Solid State Chem.
2006, 179, 3024−3032.
(41) Asano, K.; Kim, H.; Sakaki, K.; Jimura, K.; Hayashi, S.;
Nakamura, Y.; Ikeda, K.; Otomo, T.; Machida, A.; Watanuki, T.
Structural Variation of Self-Organized Mg Hydride Nanoclusters in
Immiscible Ti Matrix by Hydrogenation. Inorg. Chem. 2018, 57,
11831−11383.
(42) Kataoka, R.; Kimura, T.; Takeichi, N.; Kamegawa, A.
Stabilization of Face-Centered Cubic High-Pressure Phase of REH3
(RE = Y, Gd, Dy) at Ambient Pressure by Alkali or Alkaline-Earth
Substitution. Inorg. Chem. 2018, 57, 4686−4692.
(43) Yartys, V. A.; Gutfleisch, O.; Panasyuk, V. V.; Harris, I. R.
Desorption characteristics of rare earth (R) hydrides (R = Y, Ce, Pr,
Nd, Sm, Gd and Tb) in relation to the HDDR behaviour of R-Fe-
based compounds. J. Alloys Compd. 1997, 253−254, 128−133.
(44) Zhang, Q. A.; Liu, D. D.; Wang, Q. Q.; Fang, F.; Sun, D. L.;
Ouyang, L. Z.; Zhu, M. Superior hydrogen storage kinetics of
Mg12YNi alloy with a long-period stacking ordered phase. Scr. Mater.
2011, 65, 233−236.
(62) Shen, C.; Aguey-Zinsou, K.-F. Can 6-MgH2 improve the
hydrogen storage properties of magnesium? J. Mater. Chem. A 2017, 5,
8644−8652.
(63) Busk, R. S. Effect of Temperature on the Lattice Parameters of
Magnesium Alloys. JOM 1952, 4, 207−209.
(64) Haschke, J. M.; Clark, M. R. Phase equilibria and crystal growth
of alkaline earth and lanthanide dihydrides. High Temp. Sci. 1975, 7,
152−158.
(65) Moriwaki, T.; Akahama, Y.; Kawamura, H.; Nakano, S.;
Takemura, K. Structural phase transition of rutile-type MgH2 at high
pressures. J. Phys. Soc. Jpn. 2006, 75, 074603-1−074603-6.
(45) Liu, J. W.; Zou, C. C.; Wang, H.; Ouyang, L. Z.; Zhu, M.
Facilitating de/hydrogenation by long-period stacking ordered
structure in Mg based alloys. Int. J. Hydrogen Energy 2013, 38,
10438−10445.
(46) Zlotea, C.; Lu, J.; Andersson, Y. Formation of one-dimensional
MgH2 nano-structures. J. Alloys Compd. 2006, 426, 357−362.
(47) Ishikawa, K.; Kawasaki, T.; Yamada, Y. Hydrogenation behavior
of Mg85Zn6Y9 crystalline alloy with long period stacking ordered
structure. Int. J. Hydrogen Energy 2015, 40, 13014−13021.
(48) Li, Y.; Gu, Q.; Li, Q.; Zhang, T. In-situ synchrotron X-ray
diffraction investigation on hydrogen-induced decomposition of long
period stacking ordered structure in Mg-Ni-Y system. Scr. Mater.
2017, 127, 102−107.
(49) Poletaev, A. A.; Denys, R. V.; Solberg, J. K.; Tarasov, B. P.;
Yartys, V. A. Microstructural optimization of LaMg12 alloy for
hydrogen storage. J. Alloys Compd. 2011, 509S, S633−S639.
(50) Mezbahul-Islam, M.; Kevorkov, D.; Medraj, M. The
equilibrium phase diagram of the magnesium-copper-yttrium system.
J. Chem. Thermodyn. 2008, 40, 1064−1076.
(51) Couillaud, S.; Gaudin, E.; Andrieux, J.; Gorsse, S.; Gayot, M.;
Bobet, J. L. Study of the hydrogenation mechanism of LaCuMg8
ternary phase: The decomposition induces kinetics improvement. Int.
J. Hydrogen Energy 2012, 37, 11824−11834.
(52) Ham, B.; Junkaew, A.; Arroyave, R.; Chen, J.; Wang, H.; Wang,
P.; Majewski, J.; Park, J.; Zhou, H.-C.; Arvapally, R. K.; Kaipa, U.;
Omary, M. A.; Zhang, X. Y.; Ren, Y.; Zhang, X. Hydrogen sorption in
orthorhombic Mg hydride at ultra-low temperature. Int. J. Hydrogen
Energy 2013, 38, 8328−8341.
(53) Ham, B.; Junkaew, A.; Arroyave, R.; Park, J.; Zhou, H.-C.;
Foley, D.; Rios, S.; Wang, H.; Zhang, X. Size and stress dependent
hydrogen desorption in metastable Mg hydride films. Int. J. Hydrogen
Energy 2014, 39, 2597−2607.
(54) Tan, Z.; Chiu, C.; Heilweil, E. J.; Bendersky, L. A.
Thermodynamics, kinetics and microstructural evolution during
hydrogenation of iron-doped magnesium thin films. Int. J. Hydrogen
Energy 2011, 36, 9702−9713.
(55) Higuchi, K.; Yamamoto, K.; Kajioka, H.; Toiyama, K.; Honda,
M.; Orimo, S.; Fujii, H. Remarkable hydrogen storage properties in
L
https://dx.doi.org/10.1021/acs.inorgchem.0c02080
Inorg. Chem. XXXX, XXX, XXX−XXX