YMgGa as a hydrogen storage compound

Article abstract of DOI:10.1016/j.jssc.2009.04.029

The hydrogen absorption and desorption properties of the recently found ternary phase YMgGa have been studied. This compound absorbs 2.2 wt% hydrogen during the first cycle, but only 1.1 wt% can be stored reversibly for the following cycles under the appl

Full text of DOI:10.1016/j.jssc.2009.04.029

ARTICLE IN PRESS
Journal of Solid State Chemistry 182 (2009) 1833–1837
Contents lists available at ScienceDirect
Journal of Solid State Chemistry
journal homepage: www.elsevier.com/locate/jssc
YMgGa as a hydrogen storage compound
a,
Ã
b
b
a
Martin Sahlberg , Claudia Zlotea , Pietro Moretto , Yvonne Andersson
a
Department of Materials Chemistry, Uppsala University Box 538, SE-75121 Uppsala, Sweden
b
Institute for Energy, Joint Research Centre, European Commission, P.O. Box 2, NL-1755 Petten, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
The hydrogen absorption and desorption properties of the recently found ternary phase YMgGa have
been studied. This compound absorbs 2.2 wt% hydrogen during the first cycle, but only 1.1 wt% can be
stored reversibly for the following cycles under the applied pressure and temperature conditions.
Hydrogen absorption and desorption properties were investigated by measuring the thermal desorption
spectra and the pressure-composition isotherms while the crystal structure was determined using X-ray
diffraction (XRD). The compound absorbs hydrogen at pressures above 0.2 MPa and 250 1C by
Received 20 February 2009
Received in revised form
2
0 April 2009
Accepted 26 April 2009
Available online 3 May 2009
Keywords:
decomposing into YH
atmosphere; hydrogen is released and the YMgGa phase is partially recovered together with YGa
YH . The reformation of YMgGa occurs at temperatures below 450 1C on the expenses of hydrogen
desorption from YH . This is not expected under these temperature conditions as YH normally does not
desorb hydrogen below 800 1C.
3
and MgGa. This reaction is reversed when heating the hydride in a He
Magnesium intermetallics
Hydrides
2
and
2
Hydrogen storage
X-ray diffraction (XRD)
2
2
&
2009 Elsevier Inc. All rights reserved.
1
.
Introduction
The potential use of hydrogen as an energy carrier has been
mechanism. Here, the 1D growth is mediated by a host grain at
the top of the whisker [7].
Recently, the new ternary compound YMgGa was synthesized
and the crystal structure was determined to be the hexagonal
studied extensively in the last 20 years. Many different storage
systems have been suggested, from high-pressure tanks to
chemical storage in various compounds [1]. Many studies have
focused on metal hydrides as suitable compounds for hydrogen
storage [2].
˚
˚
ZrNiAl type-structure, a ¼ 7.2689 (10) A and c ¼ 4.4205 (9) A [5].
The atomic arrangement can be described by a 3D network of
Mg–Ga atoms with Y atoms situated inside the distorted
hexagonal channels.
Mg–Y and Mg–Ga based alloys have been studied in our group
recently [3–8]. This research has shown that the hydrogen
diffusion is improved with the amount of Y in Mg–Y thin films [9].
The present paper reports on the hydrogen absorption/
desorption properties and phase transformations of YMgGa.
The hydrogen absorption in Mg24
into YH and MgH . This reaction is irreversible as YH
stable and a temperature above 800 1C is needed to release the
hydrogen [10]. Mg Ga reacts with hydrogen gas and MgH and
Mg Ga are formed [6].
Hydrogen absorption in the Mg–Y and Mg–Ga systems has
been shown to be accompanied by microstructural modifications.
In Mg24 , hydrogen absorption leads to the formation of
nanosized 1D whiskers of MgH , growing from the surface of
Y
5
leads to decomposition
is very
2. Experimental details
3
2
2
2.1. Synthesis
5
2
2
2
5
The starting material was synthesized by heating appropriate
amounts of the element inside a sealed tantalum tube under
argon atmosphere inside an induction furnace up to ca. 1100 1C.
The elements were Mg (99.95% claimed purity, Alfa Aesar), Ga
(99.95% claimed purity, Johnson Matthey) and Y (99.9% claimed
purity, Highways International). No reaction between the sample
and the tantalum could be observed. The sample was crushed and
physically grounded to powder using an agate mortar.
Y
5
2
the host grains [7]. These whiskers are proved to be single
crystals. During hydrogen desorption the whiskers are trans-
formed to carved nanotubes of Mg [8]. The microstructural
modifications are different for hydrogen absorption in Mg
5
2
Ga .
MgH
2
whiskers were observed, but with a different growth
2.2. Hydrogenation
Ã
The sample was hydrogenated by solid–gas reactions at
Corresponding author.
E-mail address: martin.sahlberg@mkem.uu.se (M. Sahlberg).
different hydrogen pressures and temperatures for at least 10 h.
0
022-4596/$ - see front matter & 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2009.04.029

ARTICLE IN PRESS
1834
M. Sahlberg et al. / Journal of Solid State Chemistry 182 (2009) 1833–1837
Prior to hydrogenation the reactor was evacuated to ꢀ1 Pa and
flushed with argon several times. Oxide formation due to the
short exposure to air during loading of the reactor is considered to
be negligible.
2.3. Absorption/desorption isotherms and thermal desorption
spectra (TDS)
The absorption/desorption isotherms and the thermal deso-
rption spectra were recorded by using a volumetric thermal
desorption analyzer (Hiden Isochema). The system was used
either in Sieverts mode or in thermal desorption mode by
applying a constant temperature increase under a flow of inert
gas. Prior to any experiment the reactor was evacuated to a base
ꢁ
4
pressure of 10 Pa.
Absorption/desorption isotherms were recorded volumetri-
cally (Sieverts method) by a step-wise increase/decrease of the
pressure. The volumes of dosing and reactor chambers are
Fig. 1. XRD pattern of the YMgGa compound, recorded at 23 1C. The total amount
of YGa and Y MgGa phases was estimated to be less than 10 wt%.
3
approximately 5 cm . The maximum pressure and temperature
2
2
were 8 MPa and 375 1C, respectively, and the thermal stability
during the isothermal measurements was 70.3 1C. The instru-
ment was leak-tested prior to all measurements at high pressure
and room temperature over a period of several hours. The
isothermal measurements were repeated and found to be in good
agreement when performed using a similar volumetric instru-
ment which gave comparable results. Long equilibrium times
were used for the isothermal measurements allowing more than
5
h per aliquot.
For the TDS measurements, the reactor was connected to a
high sensitivity dynamic sampling mass spectrometer by a heated
flexible quartz capillary. The absorbed hydrogen was thermally
desorbed in a constant helium flow by heating the sample with a
linear temperature increase (5 1C/min). The amount of desorbed
hydrogen was determined by measuring the hydrogen partial
pressure of the mass spectrometer as a function of time.
2.4. Phase analysis with X-ray powder diffraction
Diffraction data were recorded on a X-ray powder diffract-
ometer, Bruker D8, equipped with a V a˚ ntec PSD (41 opening) using
CuK radiation. The reaction pathways were investigated by in
Fig. 2. XRD pattern from YMgGa hydrogenated at 6 MPa and 375 1C. YH
3
and MgGa
a
1
were the majority phases with small amounts of MgH , Mg Ga and YGa .
2
2
2
situ measurements in a hydrogen flow at different temperatures.
Pressures up to 600 kPa and temperatures up to 350 1C were used,
with a temperature step of 50 1C. Each pattern was recorded for
program. The obtained structural parameters agreed with the
previously reported crystal structure determinations [3,5,13].
Several different samples and thermal treatments were
3
6
0 min at a constant temperature. LaB was used as an internal
calibration standard. Phase analysis were performed using the
Rietveld method [11] implemented in the program FULLPROF [12].
The previously determined crystal structure parameters of the
observed phases were used as input parameters in the phase
analysis. The peak shape was described by the pseudo-Voigt
function and the background was modelled by interpolation
performed but it was not possible to gain
monophased sample.
a completely
3.2. Hydrogen absorption
between fixed points. The varied profile parameters were: 2
y
zero
At 6 MPa hydrogen pressure and 375 1C, YMgGa reacted with
point (1); scale factor (1); profile shape parameter (1); half-width
parameters (3); asymmetry parameters (2). Refinements of the
structural parameters showed no significant deviation from the
previously reported crystal structure determinations.
hydrogen and formed YH and MgGa and small amounts of MgH ,
3
2
Mg Ga and YGa (estimated to be less than 6 wt%), see Fig. 2.
2
2
In situ powder X-ray diffraction (XRD) measurements, at
600 kPa hydrogen pressure, showed that YMgGa absorbs hydrogen
3 2
through decomposition into YH and YGa and some small
amounts of other phases, as shown in Fig. 3. The compound
3. Results
started to absorb hydrogen at temperatures above 200 1C, and
2 3 2
YGa , YH and Mg Ga were formed.
3
.1. YMgGa
Phase analysis after hydrogenation at different hydrogen
pressures and temperatures are presented in Table 1; no
significant deviation from previously reported crystal structure
determinations were observed [13–18]. At low hydrogen pressures
The YMgGa sample contained small impurities of YGa and
Y
2
2
MgGa (see Fig. 1); the total amount of these phases is
estimated to be less than 10 wt%, as calculated by the FULLPROF
2
the yttrium containing compounds were mainly YGa and yttrium

ARTICLE IN PRESS
M. Sahlberg et al. / Journal of Solid State Chemistry 182 (2009) 1833–1837
1835
Fig. 4. The PCI curves at 375 1C for the first and second cycle. The second isotherm
starts with a hydrogen content of ꢀ1.0 wt%.
Fig. 3. In situ powder X-ray diffraction, measured at 600 kPa hydrogen pressure
6
and a stepwise increase of temperature from 50 to 350 1C. LaB has been used as
internal calibration standard.
Table 1
Phase analysis after hydrogenation.
Hydrogen-pressure (MPa) Temperature (1C) Phase content
Unreacted
23
400
375
375
375
YMgGa+(YGa+Y
YMgGa+YH +YGa
+YGa +(Mg
+Mg Ga
+MgGa+(YGa +Mg
2
MgGa
+(Y
Ga+unknown)
2
)
0
0
2
6
.2
2
2
2 2
MgGa )
.5
YH
YH
YH
2
3
3
+YH
+YGa
3
2
2
2
2
2
2
Ga+MgH
2
)
The minority phases are in brackets (estimated to be less than 10 wt%).
hydrides (YH
hydride increases with increasing hydrogen pressures and at
MPa the amount of YGa is negligible. At these high hydrogen
pressures MgH and MgGa were also formed. Any influence on the
2 3
and YH ). The amount of the formed yttrium
6
2
2
hydrogen absorbing properties due to the small amounts of
impurity phases is considered to be negligible.
3.2.1. Pressure-composition-isothermal measurements
All pressure-composition isotherms (PCI) were recorded at
375 1C at pressures up to 8 MPa. The first (circle) and second
(triangle) isotherms are shown in Fig. 4. At 8 MPa the first
hydrogen absorption cycle showed a maximum uptake of 2.2 wt%,
which is not completely desorbed during dehydrogenation (see
Fig. 4). The second cycle stores hydrogen reversibly with a
capacity of 1.1 wt%. This behaviour is stable, as discussed later,
for further hydrogenation/dehydrogenation up to 13 cycles. A
plateau pressure was noticed below 10 kPa. This plateau was
followed by an increase of the hydrogen content with pressure.
The total storage capacity of 2.2 wt% is explained by 1.6 wt%
Fig. 5. (a) Thermal desorption spectrum from YMgGa after hydrogenation at 6 MPa
and 375 1C. (b) XRD pattern of the desorbed sample.
hydrogen in YH
3
and 0.6 wt% originated from the formation of
. The second absorption of hydrogen
small amounts of MgH
2
showed that the first hydrogen uptake is not completely reversible
at 375 1C. During desorption ꢀ1.1 wt% hydrogen was recovered.
The shape of the second isotherm was similar to the part of the
first isotherm after the plateau. Very long equilibrium times were
needed for the isothermal measurements. The small apparent
increase in uptake during desorption between 8 and 6 MPa
indicated
a small additional absorption with long thermo-
dynamic equilibration.

ARTICLE IN PRESS
1836
M. Sahlberg et al. / Journal of Solid State Chemistry 182 (2009) 1833–1837
was lowered from 320 to 260 1C and the temperature of the
maximum desorption rate (maximum peak) was decreased from
79 to 363 1C after cycling.
3
4
.
Discussion
According to in situ and ex situ XRD phase analysis, YMgGa
decomposed during hydrogen absorption, see Fig. 3 and Table 1.
The first hydrogen absorption cycle showed a maximum uptake of
2
.2 wt% and a plateau pressure was noticed below 10 kPa, which
indicated the formation of YH , in agreement with the XRD phase
analysis and previous results by Vajda and Daou [19]. By
increasing the hydrogen pressure YH and MgH are formed
together with a few percent of magnesium and yttrium gallides.
After desorption, YH or MgH were not observed in the
sample. The amount of YGa was increasing significantly and
the initial YMgGa phase was also reformed to some extent. The
2
3
2
3
2
2
Fig. 6. PCI curve at 375 1C after 13 absorption/desorption cycles.
amount of YH
amount of YH
2
was considerably less than the corresponding
in the fully hydrogenated sample. This suggests
3
that hydrogen desorption from yttrium hydride occurred below
50 1C which is unexpected since pure YH desorbed hydrogen at
temperatures above 800 1C [10]. The destabilization of YH is
believed to be due to the presence of gallium in the sample. YH
was destabilized due to the favourable formation of YGa
4
2
2
2
2
,
resulting in hydrogen desorption. There were no signs of metallic
yttrium in any of the samples, hydrogenated or desorbed.
PCI measurements after cycling proved that the material was
stable after the first hydrogenation and had a maximum reversible
hydrogen storage capacity of 1.1 wt% under the applied tempera-
ture and pressure conditions. The cycling was not affecting the
thermodynamic properties; however, it changed the kinetics of
hydrogen desorption. The temperature stability of hydrogen
desorption was lowered after cycling.
5
.
Conclusions
Combining the results from hydrogen absorption/desorption
measurements with in situ and ex situ X-ray diffraction gave us a
better understanding of the different reactions and intermediate
phases taking place in the YMgGa system.
Fig. 7. Thermal desorption spectrum after cycling. For a comparison, the first
thermal desorption spectrum is also displayed.
The hydrogen absorption and desorption properties of the new
ternary compound YMgGa have been investigated; the compound
can store ꢀ2 wt% hydrogen but only 1.1 wt% reversibly. Improved
thermodynamic and kinetic properties of desorption are found as
3
.3. Thermal desorption of hydrogen
The desorption spectrum for the fully hydrogenated sample
3
compared to pure YH . The temperature of hydrogen desorption
(6 MPa) was recorded up to 450 1C, as shown in Fig. 5. The TDS
from YH has been lowed by almost 400 1C by alloying with Mg
2
spectrum shows four desorption peaks and one small shoulder
and the temperature of the maximum desorption rate was 379 1C.
The corresponding XRD pattern of the desorbed sample (Fig. 5b)
and Ga. This opens up the road of tailoring the desorption
temperature of alloy materials for hydrogen storage applications.
2 2
showed formation of YGa , YH , Mg and YMgGa.
Acknowledgments
3.4. Cycling effect
Partial funding from the Swedish Research Council and the
Royal Swedish Academy of Sciences is gratefully acknowledged.
C.Z and P.M. gratefully acknowledge partial financial support by
the European Commission DG Research, under the contract SES6-
Thirteen hydrogen absorption/desorption cycles were per-
formed by keeping the temperature constant at 375 1C and
varying the pressure from vacuum to 8 MPa of hydrogen. One
isotherm was recorded after the cycling (Fig. 6).
2006-518271/NESSHY.
The shape of the absorption and desorption isotherm recorded
after cycling was very similar to the second isotherm (Fig. 4). The
maximum hydrogen uptake was close to 1.1 wt% at 8 MPa, in
agreement with the value from the second isotherm.
The hydrogen thermal desorption spectrum of the cycled
material is shown in Fig. 7. For comparison, the first spectrum of
desorption is also displayed. The onset temperature of desorption
References
[1] L. Schlapbach, A. Zuttel, Nature 414 (2001) 353.
[
2] B. Sakintuna, F. Lamari-Darkrim, M. Hirscher, Int. J. Hydrogen Energy 32
2007) 1121.
3] M. Sahlberg, Y. Andersson, Acta Crystallogr. C 65 (2009) i7.
(
[
[4] M. Sahlberg, Y. Andersson, J. Alloys Compd. 446–447 (2007) 134.

ARTICLE IN PRESS
M. Sahlberg et al. / Journal of Solid State Chemistry 182 (2009) 1833–1837
1837
[
[
[
[
5] M. Sahlberg, T. Gustafsson, Y. Andersson, Acta Crystallogr. E 63 (2007) i195.
6] C. Zlotea, Y. Andersson, Acta Mater. 54 (2006) 5559.
7] C. Zlotea, J. Lu, Y. Andersson, J. Alloys Compd. 426 (2006) 357.
8] C. Zlotea, M. Sahlberg, S. Oezbilen, P. Moretto, Y. Andersson, Acta Mater. 56
(
[12] J. Rodr ı´ guez-Carvajal, FullProf2k computer program, version 290, 2004.
[13] O. Schob, E. Parthe, Acta Crystallogr. 19 (1965) 214.
[14] K. Frank, K. Schubert, J. Less-Common Met. 20 (1970) 215.
[15] K. Schubert, F. Gauzzi, K. Frank, Z. Metallk. 54 (1963) 422.
[16] Y. Wang, M.Y. Chou, Phys. Rev. B 51 (1995) 7500.
2008) 2421.
9] C. Chacon, E. Johansson, B. Hjorvarsson, C. Zlotea, Y. Andersson, J. Appl. Phys.
7 (2005) 104903/1.
10] V.A. Yartys, O. Gutfleisch, V.V. Panasyuk, I.R. Harris, J. Alloys Compd. 253–254
1997) 128.
11] H.M. Rietveld, J. Appl. Cryst. 2 (1969) 65.
[
[17] A. Pebler, W.E. Wallace, J. Phys. Chem. 66 (1962) 148.
9
[18] F.H. Ellinger, C.E.J. Holley, B.B. McInteer, D. Pavone, R.M. Potter, E. Staritzky,
W.H. Zachariasen, J. Am. Chem. Soc. 77 (1955) 2647.
[19] P. Vajda, J.N. Daou, Hydrogen Met. Syst. I Solid State Phenom. 49–50 (1996)
71–157.
[
[
(