Hydrogen storage properties of Mg[BH4]2

Article abstract of DOI:10.1016/j.jallcom.2007.05.054

Among the large variety of possible complex hydrides only few exhibit a large gravimetric hydrogen density and stability around 40 kJ mol^-1H₂. Mg[BH₄]₂ is based on theoretical approaches a complex hydride with an equilibrium hydrogen pressure of approximately 1 bar at room temperature and a hydrogen content of 14.9 mass%. The reaction of Li[BH₄] with MgCl₂ at elevated temperatures was investigated as a possible route to synthesize Mg[BH₄]₂. Li[BH₄] reacts with MgCl₂ at a temperature >523 K at a pressure of 10 MPa of hydrogen, where the product contains LiCl and Mg[BH₄]₂. The desorption pc-isotherm of the product obtained at 623 K shows two flat plateaus, which indicates that the decomposition of the product consists of a two-step reaction. The products of the first and the second decomposition reaction were analyzed by means of X-ray diffraction and found to be MgH₂ and Mg, respectively. The enthalpy for the first decomposition reaction was determined to be ΔH = -39.3 kJ mol^-1H₂ by the Van't Hoff plot of the equilibrium measurements between 563 K and 623 K, which is significantly lower than that for pure Li[BH₄] (ΔH = -74.9 kJ mol^-1H₂). However, only the second reaction step (MgH₂ → Mg) is reversible at the condition up to 623 K at 10 MPa of hydrogen.

Full text of DOI:10.1016/j.jallcom.2007.05.054

Journal of Alloys and Compounds 459 (2008) 583–588
Hydrogen storage properties of Mg[BH4]2
T. Matsunagaa,b,∗, F. Buchtera,b, P. Maurona,b, M. Bielmana,b, Y. Nakamoric,
c
d
d
d
a,b
¨
S. Orimo , N. Ohba , K. Miwa , S. Towata , A. Zuttel
a
¨
EMPA, Department Environment, Energy and Mobility, Abt. 138 “Hydrogen & Energy”, Uberlandstrasse 129, 8600 Du¨bendorf, Switzerland
b University of Fribourg, Physics Department, Pe´rolles, 1700 Fribourg, Switzerland
c Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan
d Toyota Central Research and Development Laboratories, Inc., Nagakute, Aichi 480-1192, Japan
Received 29 January 2007; received in revised form 13 May 2007; accepted 14 May 2007
Available online 18 May 2007
Abstract
Among the large variety of possible complex hydrides only few exhibit a large gravimetric hydrogen density and stability around 40 kJ mol−1H2.
Mg[BH4]2 is based on theoretical approaches a complex hydride with an equilibrium hydrogen pressure of approximately 1 bar at room temperature
andahydrogencontentof14.9mass%. ThereactionofLi[BH4]withMgCl2 atelevatedtemperatureswasinvestigatedasapossibleroutetosynthesize
Mg[BH4]2. Li[BH4] reacts with MgCl2 at a temperature >523 K at a pressure of 10 MPa of hydrogen, where the product contains LiCl and Mg[BH4]2.
The desorption pc-isotherm of the product obtained at 623 K shows two flat plateaus, which indicates that the decomposition of the product consists
of a two-step reaction. The products of the first and the second decomposition reaction were analyzed by means of X-ray diffraction and found to
be MgH2 and Mg, respectively. The enthalpy for the first decomposition reaction was determined to be ꢀH = −39.3 kJ mol−1H2 by the Van’t Hoff
plot of the equilibrium measurements between 563 K and 623 K, which is significantly lower than that for pure Li[BH4] (ꢀH = −74.9 kJ mol−1H2).
However, only the second reaction step (MgH2 → Mg) is reversible at the condition up to 623 K at 10 MPa of hydrogen.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Hydrogen absorbing materials; Thermodynamic properties; Thermal analysis
1. Introduction
the reversibility of the hydrogen sorption reaction of complex
hydrides.
Hydrogen storage technology is one of the most impor-
tant subjects for the introduction of fuel cell vehicles in
transportation. The hydrogen storage density as well as the
moderate operating temperatures for desorption and absorp-
tion of hydrogen are key parameters for an onboard storage
system. Complex hydrides, consisting of light elements, are
very promising materials for hydrogen storage because of their
high gravimetric and volumetric hydrogen density [1]. Since
Bogdanovic and Schwickardi reported the reversibility of the
catalyzed hydrogen sorption reaction of NaAlH4 [2], many
efforts have been made to investigate complex hydrides as
hydrogen storage materials [3–7]. Besides the basic physical
properties, thechallengesaretotailorthestabilityandtoimprove
The hydrogen in the complex hydrides is located in the cor-
ners of a tetrahedron with B or Al in the center. The negative
charge of the anion, [BH4]− and [AlH4]−, is compensated by
a cation, e.g., Li or Na. The bonding character and the physical
properties of the complex hydrides are largely determined by the
difference in electronegativity between the cation and boron or
aluminum, respectively. From this point of view, first-principles
calculations [8,9] and experimental studies [10] on the stability
of complex hydrides have been performed. Recently, Nakamori
et al. have investigated the relation between stability of borohy-
drides and electronegativity of cations for various borohydrides
and have shown that there is a good correlation between them
[11].
Magnesium borohydride (Mg[BH4]2) is a very promising
material for hydrogen storage because of its high gravimetric
storage density (14.9 mass%). The Pauling electronegativity
of magnesium is 1.31, which is greater than that of lithium
(0.98), therefore, it is expected that Mg[BH4]2 is less stable than
∗
Corresponding author at: EMPA, Department Environment, Energy and
¨
Mobility, Abt. 138 “Hydrogen
Du¨bendorf, Switzerland. Tel.: +41 44 823 4038; fax: +41 44 823 4022.
& Energy”, Uberlandstrasse 129, 8600
E-mail address: tomoya@matsunaga.tec.toyota.co.jp (T. Matsunaga).
0925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2007.05.054

584
T. Matsunaga et al. / Journal of Alloys and Compounds 459 (2008) 583–588
Li[BH4]. However, properties of Mg[BH4]2 as a hydrogen stor-
age material, especially thermodynamic property, are not known
yet and the investigation requires sophisticated instruments and
a well-defined sample. Stasinevich and Egorenko investigated
the stability of various borohydrides by means of thermal anal-
ysis in hydrogen pressures up to 1.0 MPa [12]. They reported
that hydrogen desorption from Mg[BH4]2 occurs at 586–596
and 683–693 K, where the former corresponds to the desorp-
tion from Mg[BH4]2 to MgH2, the latter from MgH2 to Mg.
Konoplev and Bakulina also investigated the thermal analysis
of Mg[BH4]2, where they mentioned that Magnesium boride
(MgB2) may be formed after thermal decomposition [13]. Direct
synthesis of Mg[BH4]2 from Magnesium, boron and hydrogen
has been reported by Goerrig [14], where it has been claimed
that Mg[BH4]2 can be formed from these elements at 923 K and
15 MPa of hydrogen.
In this study, the reaction of Li[BH4] with MgCl2 at ele-
vated temperatures is investigated as a new route to synthesize
Mg[BH4]2. Lithium borohydride and magnesium chloride are
used as starting materials. The reactions of Li[BH4] with MgCl2
are examined by differential scanning calorimetry (DSC). The
synthesis products as well as the hydrogen desorption products
are investigated by X-ray diffraction and Raman spectroscopy.
Hydrogen desorption and absorption property of the synthesis
product is investigated by temperature programmed desorption
(TPD) and pc-isotherm measurement. Based on the equilib-
rium pressures of measured pc-isotherms the dehydrogenation
enthalpy and entropy were computed by means of the Van’t Hoff
equation.
air, each sample was filled into a sample holder in the argon glove box and cov-
ered with plastic wrap film during X-ray diffraction measurement and Raman
spectroscopy measurement.
2.2. Hydrogen desorption and absorption
Hydrogen desorption property of the product synthesized by the heat treat-
ment of Li[BH4] with MgCl2 (for 2:1 mole ratio) at 593 K under 10 MPa of
hydrogen was investigated as follows.
Temperature programmed desorption (TPD) measurements were carried out
in vacuum after the heat treatment in a stainless steel cylinder with a heating
rate of 0.2–5.0 K min−1. After the heat treatment, the samples were cooled down
to room temperature under 10 MPa of hydrogen. Subsequently, hydrogen gas
was extracted from the cylinder to vacuum at room temperature. The samples
were heated up under vacuum from room temperature to 773 K. During the
measurements, desorbed gas volume was measured using a mass flow controller
(Brooks instruments, 5850E).
Desorption isotherms were measured at 563, 593 and 623 K using the same
mass flow controller mentioned above. A total of 1.6 g of the sample were
used for each measurement. Hydrogen pressure was decreased from 10 MPa
to vacuum with a flow rate of 1.0 cm3 min−1. It took approximately 70 h to per-
form each desorption measurement with this gas flow rate. Absorption isotherm
measurement was performed at 623 K continuously after the desorption mea-
surement at the same temperature. Hydrogen was filled into the cylinder with a
flow rate of 1.0 cm3 min−1 until the pressure was increased up to 10 MPa.
After desorption or absorption measurement, the cylinder was cooled down
to room temperature. The samples were taken out in an argon glove box. Crys-
tal structures of the samples were investigated by powder X-ray diffraction
measurement as mentioned above at room temperature.
3. Results and discussion
3.1. Reaction of Li[BH4] with MgCl2
2. Experimental
The DSC profiles of Li[BH4] and MgCl2 mixture are shown
in Fig. 1. Profiles of pure Li[BH4] and pure MgCl2 were also
investigated as references. For Li[BH4], there is an endother-
mic peak (at T = 386 K, ꢀQ = −206.0 J g−1) during heating
from 313 to 513 K and an exothermic peak during cooling
both of them corresponding to the phase transition of Li[BH4],
whereas no peak was observed for MgCl2 in this temperature
range. For the Li[BH4] and MgCl2 mixture sample, there is
an endothermic peak (at T = 385 K, ꢀQ = −56.3 J g−1) during
the 1st heating procedure, corresponding to the phase tran-
sition of Li[BH4]. However, this peak disappears after the
cooling period of 1st cycle. This result is an evidence that
there is no pure Li[BH4] remains after the 1st heating. In addi-
tion, for the Li[BH4] and MgCl2 mixture, other exothermic
and endothermic peaks are observed after the 1st cooling at
T = 440 K (ꢀQ = 23.6 J g−1). These peaks are not observed for
pure Li[BH4] or pure MgCl2. This result implies that some reac-
tion took place during the 1st heating. The endothermic peak at
440 K observed after the 1st cooling is in agreement with the
DSC measurement result reported for Mg[BH4]2 [12]. There-
fore, the following reaction during the 1st heating procedure is
assumed:
The starting materials used in this study were purchased from Aldrich Co.
Ltd.: The purities are >95% for Li[BH4] and >99.9% for MgCl2, respectively.
The samples were handled in pure argon atmosphere (glove box), in high vacuum
and under pure hydrogen gas in order to avoid any possible contamination by
moisture or oxygen from air.
2.1. Reaction of Li[BH4] with MgCl2
The reaction of Li[BH4] with MgCl2 was investigated by differential scan-
ning calorimetry (DSC) (Mettler Toledo Inc. HP DSC827e). Five milligrams of
the sample (for 2:1 mole ratio) was mixed in an argon glove box and filled in
a DSC sample cell made of aluminum. The sample cell was sealed up in argon
atmosphere. Therefore, it has never been exposed to any gases except pure argon
during the measurement. The DSC measurement of Li[BH4] and MgCl2 mixture
was carried out in the temperature range from 313 to 513 K at the heating (or
cooling) rate of 5 K min−1 in the sealed sample cell for three heating and cooling
cycles. Measurements with pure Li[BH4] and MgCl2 were also performed at the
same heating condition.
Heat treatments of Li[BH4] and MgCl2 mixture were carried out in a stain-
less steel cylinder under 10 MPa of hydrogen at 453, 523 or 593 K. In each
heat treatment, approximately 600 mg of Li[BH4] and MgCl2 mixture (for 2:1
mole ratio) was pressed in order to make a pellet and filled into a cylinder in
an argon glove box. After evacuating by rotary vacuum pump for 1 h at room
temperature, 10 MPa of hydrogen was introduced into the cylinder and heated
up. It was kept at the final temperature mentioned above for 3 h and then slowly
cooledtoroomtemperature. Crystalstructuresofthesamplesafterheattreatment
wereinvestigatedatroomtemperaturebypowderX-raydiffractionmeasurement
(SIEMENS, D-500, Cu K␣ radiation). Raman spectra of the samples were mea-
sured at room temperature using a Dilor Labram Raman spectrometer equipped
with a HeNe laser (632.8 nm) and an optical microscope. To avoid exposure to
2Li[BH4] + MgCl2 → Mg[BH4]2 + 2LiCl
XRD measurement results after heat treatment also indicate
that Li[BH4] react with MgCl2 during heat treatment. Fig. 2
shows the XRD measurement results after heat treatment at

T. Matsunaga et al. / Journal of Alloys and Compounds 459 (2008) 583–588
585
heat treatment at the temperature >523 K. The X-ray diffrac-
tion peaks around 2Θ = 18–19◦ after heat treatment at 453 K
may be due to an intermediate compound of the reaction,
which are not observed after heat treatment at higher temper-
atures.
The Raman spectra of pure Li[BH4] are shown in Fig. 3(a
and b) and Li[BH4] with MgCl2 after heat treatment at 593 K
under 10 MPa of hydrogen in Fig. 3(c and d). The Raman shifts
observedforpureLi[BH4](Fig. 3(aandb))areinagreementwith
the bending mode and the stretching mode reported for [BH4]−
in Li[BH4], respectively [15]. After heat treatment, two of new
Raman shifts occurred at 1200 and 1380 cm−1 in the bending
mode region as is shown in Fig. 3(c). The Raman spectra after
heat treatment is consistent with the Raman spectra [11] and
the IR spectra [13] reported for Mg[BH4]2. This result shows
that Mg[BH4]2 is included in the product after heat treatment. In
the stretching mode region, the splitting at 2300 cm−1 has disap-
peared after heat treatment as is shown in Fig. 3(d). Probably this
spectrum shows overlap of stretching mode in Mg[BH4]2 and
Li[BH4] [11,15]. The product after the heat treatment at 593 K
was also examined by DSC measurement. The result shows that
there is an endothermic peak around 440 K, whereas no peak
from LiBH4 was observed. This result indicates that most of the
product is supposed to be Mg[BH4]2.
Fig. 1. Differential scanning calorimetry profiles of Li[BH4] and MgCl2 mixture
with a sealed sample cell. The heating rate is 5 K min−1. Measurements with
pure Li[BH4] and MgCl2 were also performed at the same heating condition.
Solid lines and dotted lines show heating and cooling, respectively. A total of
5 mg of the sample was used for each measurement.
3.2. Hydrogen desorption and absorption
3.2.1. Temperature programmed desorption
Fig. 4 shows the temperature programmed desorption spec-
tra at a heating rate of 0.2 K min−1. Although the product
after heat treatment contains LiCl, LiCl does not decompose
below 773 K. Therefore, all of the desorbed gas is supposed
to originate from Mg(BH4)2. In addition, more than 99.7%
of the desorbed gas from the sample was confirmed to be
hydrogen by mass spectroscopy. The hydrogen desorption tem-
perature of Li[BH4] and MgCl2 mixture after the heat treatment
is approximately 100 K lower as compared to that of pure
Li[BH4]. The sample after heat treatment has two desorption
peaks. It starts to decompose at 500 K and the first desorp-
tion peak appears around 563 K. This peak is dominant and
more than half of hydrogen is evolved. The second desorption
peak starts at 600 K and it is very sharp as compared to the
first desorption peak. The shape of two independent desorp-
tion peaks implies that decomposition reaction consists of two
steps.
various temperatures. Mixing Li[BH4] and MgCl2 in a mor-
tar in an inert atmosphere (Ar) at room temperature does not
lead to any reaction. However, after heat treatment, the peaks
corresponding to Li[BH4] and MgCl2 disappear. New peaks
were observed which correspond to LiCl. Although there is
still a small amount of Li[BH4] and MgCl2 after heat treat-
ment at 453 K, neither Li[BH4] nor MgCl2 was observed after
Desorption spectra of Li[BH4] and MgCl2 mixture after heat
treatment with different heating rates are shown in Fig. 5(a).
The peak temperatures of the spectra were used to determine
the activation energy of the first and the second desorption.
The logarithm of the ratio of the heating rate divided by the
square of the peak temperature is a linear function of the inverse
temperature if the hydrogen desorption is a first order reac-
tion (Fig. 5(b)). The slope of the fitted line corresponds to
the activation energy for the reaction (Kissinger plot) [16].
The activation energy of the first and the second desorption
obtained by Fig. 5(b) were 122 and 166 kJ mol−1, respec-
tively.
Fig. 2. Powder X-ray diffraction profiles of Li[BH4] and MgCl2 mixture: (a)
before heat treatment; (b) after heat treatment at 453 K; (c) after heat treatment
at 523 K; (d) after heat treatment at 593 K, respectively. Each heat treatment was
carried out under 10 MPa of hydrogen.

586
T. Matsunaga et al. / Journal of Alloys and Compounds 459 (2008) 583–588
Fig. 3. Raman spectra of pure Li[BH4] (a and b) and Li[BH4] with MgCl2 after heat treatment at 593 K under 10 MPa of hydrogen (c and d).
3.2.2. PC desorption and absorption isotherms
Fig. 6shows the pressure composition desorption isotherms
at 563, 593 and 623 K. Two plateaus were observed at 623 K
at more than 0.1 MPa, whereas only one plateau was observed
at 563 and 593 K within this pressure range. Equilibrium pres-
sures of the first desorption varied from 1.4 MPa at 563 K to
3.1 MPa at 623 K. The equilibrium pressure (Peq) as a func-
tion of temperature is related to the changes ꢀH and ꢀS of
enthalpy and entropy, respectively, by the Van’t Hoff equa-
tion:
ꢀ
ꢁ
Peq
Pe0q
ꢀH 1
R T
ꢀS
R
ln
=
−
(1)
Based on the Van’t Hoff plot shown in Fig. 7,
the dehydrogenation enthalpy and entropy were found to
be ꢀH = −39.3 kJ mol−1H2, ꢀS = −91.3 J mol−1 K−1, respec-
tively. Whereas the dehydrogenation enthalpy of pure Li[BH4]
is obtained to be ꢀH = −74.9 kJ mol−1H2 from the equi-
librium pressure of Li[BH4] reported in the literature
[17]. Therefore, the product of Li[BH4] with MgCl2 after
heat treatment is thermodynamically less stable than pure
Li[BH4].
Absorption pc-isotherm was measured at 623 K subsequently
after the desorption measurement at the same temperature. As
is shown in Fig. 6, only one absorption plateau was observed
around 1.3 MPa, where the absorbed hydrogen amount was
almost the same as the desorbed hydrogen amount in the sec-
ond reaction step of the desorption measurement. This result
Fig. 4. Temperature programmed desorption spectra of: (a) Li[BH4] and MgCl2
mixture after heat treatment at 593 K under 10 MPa of hydrogen; (b) pure
Li[BH4]. The samples were heated after evacuation at room temperature with
a heating rate of 0.2 K min−1. The upper graph and the lower graph show the
desorbed gas volume and the gas flow as a function of temperature, respectively.

T. Matsunaga et al. / Journal of Alloys and Compounds 459 (2008) 583–588
587
Fig. 7. Van’t Hoff plots for (a) Li[BH4] with MgCl2 after heat treatment at 593 K
and (b) pure Li[BH4]. Equilibrium pressures of (a) and (b) are obtained from
PC desorption isotherms in Fig. 5, from ref. [17], respectively.
after desorption in vacuum to 593 K, (c) after desorption in vac-
uum to 623 K and (d) after absorption up to 10 MPa of hydrogen
at 623 K, respectively. As it is shown in Fig. 8(b), after desorp-
tion in vacuum to 593 K, the peaks which correspond to MgH2
are observed, whereas the diffraction pattern of metallic Mg is
not observed. On the other hand, as is shown in Fig. 8(c), after
desorption in vacuum to 623 K, the peaks corresponding to Mg
are clearly present, where the peaks from MgH2 are no longer
observed. These results indicate that MgH2 desorbs hydrogen
to be Mg during the second desorption step. After absorption
up to 10 MPa of hydrogen at 623 K, as is shown in Fig. 8(d),
MgH2 forms again, which implies the hydrogenation of Mg
during the absorption procedure. LiCl remains during desorp-
tion/absorption procedure as is shown in Fig. 8. Consequently,
the overall results of X-ray diffraction analysis can be explained
by the following two-step reaction:
Fig. 5. (a) Temperature programmed desorption spectra of Li[BH4] and MgCl2
mixtureafterheattreatmentat593 Kunder10 MPaofhydrogen. Numbersshown
in the graph indicate the heating rate in K min−1. A total of 20 mg of the sample
was used for the measurements. (b) Kissinger plot of thermal desorption spectra
at a heating rate of 0.2−5.0 K min−1, where Tp is a peak temperature of the
spectra and β is a heating rate.
is a clear evidence that only the second step of the reaction is
reversible under this measuring condition.
3.2.3. X-ray diffraction analysis
Mg[BH4]2 → MgH2 + 2B + 3H2(first step)
Fig. 8 shows the XRD measurement results of the Li[BH4]
and MgCl2 mixture (a) after heat treatment at 593 K for 3 h, (b)
Fig. 6. Absorption/desorption isotherms for Li[BH4] and MgCl2 mixture after
heat treatment at 593 K: (a) desorption at 563 K; (b) desorption at 593 K; (c)
desorption at 623 K; (d) absorption at 623 K. Curve (d) was measured after the
desorption measurement at the same temperature.
Fig. 8. Powder X-ray diffraction profiles of Li[BH4] and MgCl2 mixture: (a)
after heat treatment at 593 K for 3 h, (b) after desorption in vacuum to 593 K,
(c) after desorption in vacuum to 623 K and (d) after absorption up to 10 MPa
of hydrogen at 623 K, respectively.

588
T. Matsunaga et al. / Journal of Alloys and Compounds 459 (2008) 583–588
International Joint Research under the “Development for Safe
MgH2 → Mg + H2(second step)
UtilizationandInfrastructureofhydrogen” Project (2005-2006),
the Swiss Federal Office of Energy (BFE) project “Hydrogen
storage in metal- and complex hydrides”, No. 100324 (2003-
2006), EU FP6 IP NESSHY “Novel Efficient Solid Storage for
Hydrogen, Contract no.:518271 (SES6).
Taking this result into consideration, the first and the second
plateau of the desorption pc-isotherm at 623 K correspond to the
desorption from Mg[BH4]2 to MgH2 and from MgH2 to Mg,
respectively. This assumption is consistent with the desorbed
hydrogen amount ratio (3:1) in the first and the second plateau.
Similarly, the absorption plateau in Fig. 6(d) corresponds to the
reaction from Mg to MgH2. The plateau pressure of 1.3 MPa
in Fig. 6(d) is almost the same as the equilibrium pressure of
Mg/MgH2 system at the same temperature [18].
References
[1] A. Zu¨ttel, Materialstoday Sept. (2003) 24.
[2] B. Bogdanovic, M. Schwickardi, J. Alloys Comp. 253–254 (1997) 1.
[3] A. Zu¨ttel, S. Rentsch, P. Fischer, P. Wenger, P. Sudan, Ph. Mauron, Ch.
Emmenegger, J. Alloys Comp. 356 (2003) 515.
[4] P. Chen, Z. Xiong, J. Luo, J. Lin, K.L. Tan, Nature 420 (2002) 302.
[5] M. Fichtner, O. Fuhr, O. Kircher, J. Alloys Comp. 356–357 (2003) 418.
[6] J.J. Vajo, S.L. Skeith, F. Mertens, J. Phys. Chem. B 109 (9) (2005) 3719.
[7] M. Aoki, K. Miwa, T. Noritake, G. Kitahara, Y. Nakamori, S. Orimo, S.
Towata, Appl. Phys. A 80 (2005) 1409.
[8] K. Miwa, N. Ohba, S. Towata, Y. Nakamori, S. Orimo, J. Alloys Comp.
404–406 (2005) 140.
[9] M. Yoshino, K. Komiya, Y. Takahashi, Y. Shinzato, H. Yukawa, M. Mori-
naga, J. Alloys Comp. 404–406 (2005) 185.
4. Conclusions
Li[BH4] reacts with MgCl2 by heat treatment at the tem-
perature >523 K at 10 MPa of hydrogen, where the product
contains LiCl and Mg[BH4]2. Desorption isotherm of the prod-
uct obtained at 623 K shows two flat plateaus, which indicates
that the decomposition of the product after heat treatment con-
sists of a two-step reaction. Using X-ray diffraction analysis, the
products of the first and the second decomposition reaction are
found to be MgH2 and Mg, respectively. From the Van’t Hoff
plot obtained by equilibrium pressures of the first decomposition
measured at 563–23 K, the dehydrogenation enthalpy of the first
decomposition is found to be ꢀH = −39.3 kJ mol−1H2, which
is lower than that of pure Li[BH4] (ꢀH = −74.9 kJ mol−1H2).
However, only the second reaction step (MgH2 → Mg) is
reversible at the condition up to 623 K at 10 MPa of hydrogen.
[10] S. Orimo, Y. Nakamori, G. Kitahara, K. Miwa, N. Ohba, S. Towata, A.
Zu¨ttel, J. Alloys Comp. 404–406 (2005) 427.
[11] Y. Nakamori, K. Miwa, A. Ninomiya, H. Li, N. Ohba, S. Towata, A. Zuttel,
S. Orimo, Phys. Rev. B 74 (2006) (045126(1)–045126(9)).
[12] D.S. Stasinevich, G.A. Egorenko, Russian J. Inorg. Chem. 13 (3) (1968)
341.
¨
[13] V.N. Konoplev, V.M. Bakulina, Russian Chem. Bull. 20 (1) (1971) 136.
[14] D. Goerrig, German Patent 1,077,644 (December 27, 1958).
[15] S. Gomes, H. Hagemann, K. Yvon, J. Alloys Comp. 346 (2002) 206.
[16] H.E. Kissinger, Anal. Chem. 29 (1957) 1702.
[17] NEDO international joint research under the “Development for Safe
Utilization and Infrastructure of Hydrogen” project progress report;
http://www.nedo.go.jp/nenryo/seika/suisoh16-17/seikah h17 5.pdf.
[18] L.B. Pankratz, Thermodynamic Properties of Carbides, Nitrides, and other
Selected Substances, U.S. Dept. of the Interior, Washington, DC, 1995, p.
957.
Acknowledgements
This study was partially supported by the New Energy
and Industrial Technology Development Organization (NEDO),