T. Nakagawa et al. / Journal of Alloys and Compounds 446–447 (2007) 306–309
307
It is noteworthy that a finite hydrogen pressure is necessary
for the reaction of Mg with LiBH4 to yield MgB2. In contrast,
the Mg and B phases instead of the MgB2 phase are produced
after dehydrogenation under vacuum. This reaction is expressed
as follows:
gas flow, p-DSC under ∼0.3 MPa Ar gas flow and 0.5 MPa H2-
gas in the closed system. It can be seen that the decomposition
temperature of the TiCl3-doped MgH2 is ∼280 ◦C under the He-
gas flow (Fig. 1(a)), which is almost the same as that of under
the Ar flow condition (Fig. 1(b)), irrespective of sorts of inert
gases. This result indicates that an inert gas gives no influence
on the dehydrogenation properties of MgH2. On the other hand,
in the p-DSC profile under 0.5 MPa H2 gas pressure (Fig. 1(c)),
an endothermic peak corresponding to decomposition of MgH2
is observed at higher temperature (∼370 ◦C) than that under the
inert gas flow, where the pressure rises to ∼0.6 MPa at this tem-
perature. This indicates that the hydrogen dissociation pressure
is about 0.6 MPa at ∼370 ◦C, suppressing the dehydrogenation
of MgH2 by a reverse reaction.
2LiBH4 + MgH2 → 2LiH + Mg + 2B + 4H2.
(6)
At present, it is not yet clear which role the hydrogen atmo-
sphere does play in the reaction (5). Therefore, in this paper, we
report on the results of thermal analysis for the mixture of MgH2
and LiBH4 under different atmospheric conditions. And finally,
we try to clarify the role of hydrogen in this system.
2. Experimental
For LiBH4 doped with a small amount of TiCl3, three
endothermic reactions are observed in the DTA profile under
a He gas flow (Fig. 2(a)), which is almost the same as under
an Ar flow condition as shown in Fig. 2(b). Therefore, the
inert gases do not show any differences on the dehydrogena-
tion properties of LiBH4. Here, it should be noted that the
measuring temperature was limited up to 450 ◦C, because the
sample pan of the Al metal was used in this work. For this rea-
son, unfortunately, the third reaction peak could not be clearly
the endothermic reactions could be recognized above 400 ◦C
since the DSC signals are lower than the base lines. Comparing
In this work, MgH2 powder (95 mass% purity, Gelest Inc.), LiBH4 pow-
der (95 mass% purity, Sigma–Aldrich) and titanium chloride powder (TiCl3,
99.999 mass% purity, Sigma–Aldrich) were used as starting materials. Thermal
analyses were performed for MgH2, LiBH4 and a mixture of MgH2 and 2LiBH4,
in which 3 mol% TiCl3 was added as a catalyst. Prior to thermal analyses, all the
samples were ball-milled by a planetary ball-mill apparatus (Fritsch P7) for 2 h
at 400 rpm under 1.0 MPa highly pure hydrogen gas (7 N). Mixed powders of
∼300 mg and 20 pieces of steel balls with a diameter of 7 mm were loaded into
the milling container (Cr–steel pot with an internal volume of ∼30 ml). The ther-
mal analyses were examined by a pressurized differential scanning calorimetry
(p-DSC) (TA Instruments, DSC Q10P) and a thermogravimetry with differential
thermal analysis equipment (TG-DTA) (Rigku, TG8120). The measurement of
p-DSC was operated under ∼0.3 MPa Ar gas flow, or under 0.5 MPa H2 gas as
initial pressure (final pressure at 450 ◦C was ∼0.7 MPa) in a closed system, while
the TG-DTA measurement was operated under a He gas flow condition. The
heating rate was fixed at 5 ◦C/min for both measurements. The identification of
the products was carried out by X-ray powder diffraction (XRD) measurements
(Rigaku, RINT-2500, Cu K␣). All the processes from the sample preparation to
the thermal analyses were performed in Ar-filled gloveboxes with a recycling
purification system (MP-P60W, Miwa MFG Co., LTD) to avoid the sample
pollutions by water vapor and oxygen (the dew point of water was lower than
−80 ◦C and the oxygen concentration was below 1 ppm).
´
with the results reported by Fedneva et al. [3] and Soulie et al.
[13], three endothermic reactions in Fig. 2(a) and (b) correspond
to the phase transition (∼105 ◦C, orthorhombic to hexagonal),
the melting phenomenon (∼280 ◦C) and the decomposition into
LiH, B and H2 (above 400 ◦C). On the other hand, under 0.5 MPa
H2 gas pressure in Fig. 2(c), two endothermic peaks are observed
below 300 ◦C, but there is no other reaction above 300 ◦C. These
twopeaksarerecognizedatthesametemperaturesasthoseunder
the inert gas conditions, indicating that the hydrogen pressure
does not affect both the temperatures of phase transition and
melt of LiBH4. However, the third reaction above 300 ◦C disap-
pears in the DSC profile under 0.5 MPa H2 gas pressure, where
3. Result and discussion
The thermal analyses for the TiCl3-doped MgH2 and the
TiCl3-doped LiBH4 are shown in Figs. 1 and 2, respectively.
Fig. 1. TG-DTA and DSC profiles of TiCl3-doped MgH2. (a) TG-DTA profile
under a He flow, (b) DSC profile under an Ar flow and (c) DSC profile under
0.5 MPa H2 (initial pressure).
Fig. 2. TG-DTA and DSC profiles of TiCl3-doped LiBH4. (a) TG-DTA profile
under a He flow, (b) DSC profile under an Ar flow and (c) DSC profile under
0.5 MPa H2 (initial pressure).