Intermediate products during the hydriding combustion synthesis of Mg 2NiH4

Article abstract of DOI:10.1016/j.jallcom.2004.04.092

Magnesium-based hydrogen storage alloy is one of the most attractive mediums, as a hydrogen storage tank for a fuel cell. However, the practical application is still limited because of its poor hydriding and dehydriding kinetics and relatively high reaction temperature. To overcome such problems, the hydriding combustion synthesis (HCS) was established in 1997, as one of the most effective production methods. The understanding of these microstructures is strongly related to the improvement of the reaction property. However, up to now, little efforts have been made as to the microstructures of final product and intermediate products. Therefore, the purpose of this paper is to prepare intermediate products at several stages during the HCS of Mg₂NiH ₄, and to study their microstructures by means of X-ray diffraction (XRD) and scanning electron microscopy (SEM) for the first time. Interestingly, the results revealed that: (1) the intermediate product generated at lower temperature than 800 K caused the unique particle morphology of final product of Mg₂NiH₄; (2) severe thermal corrosion, cotton-like material, was also observed on the surface of intermediate products at 705 K, which became enhanced with increasing temperature; (3) after the dehydriding reaction of MgH₂ at 750 K, a large amount of the Mg2Ni phase existed in the intermediate product; (4) hydrogen present during the heating period makes the product to have smaller size. Further investigation on the mechanism of the gas-solid reaction deserves attention, because lowering of the synthesis temperature by as much as 100' is very attractive and significant for industrializing this process.

Full text of DOI:10.1016/j.jallcom.2004.04.092

Journal of Alloys and Compounds 384 (2004) 157–164
Intermediate products during the hydriding combustion
synthesis of Mg2NiH4
Liquan Lia, Itoko Saitab, Tomohiro Akiyamab,∗
a
College of Materials Science and Engineering, Nanjing University of Technology, Nanjing 210009, PR China
b
Department of Chemical Engineering, Osaka Prefecture University, Sakai 599-8531, Japan
Received 12 February 2004; received in revised form 1 April 2004; accepted 8 April 2004
Abstract
Magnesium-based hydrogen storage alloy is one of the most attractive mediums, as a hydrogen storage tank for a fuel cell. However,
the practical application is still limited because of its poor hydriding and dehydriding kinetics and relatively high reaction temperature. To
overcome such problems, the hydriding combustion synthesis (HCS) was established in 1997, as one of the most effective production methods.
The understanding of these microstructures is strongly related to the improvement of the reaction property. However, up to now, little efforts
have been made as to the microstructures of final product and intermediate products. Therefore, the purpose of this paper is to prepare
intermediate products at several stages during the HCS of Mg2NiH4, and to study their microstructures by means of X-ray diffraction (XRD)
and scanning electron microscopy (SEM) for the first time. Interestingly, the results revealed that: (1) the intermediate product generated at
lower temperature than 800 K caused the unique particle morphology of final product of Mg2NiH4; (2) severe thermal corrosion, cotton-like
material, was also observed on the surface of intermediate products at 705 K, which became enhanced with increasing temperature; (3) after
the dehydriding reaction of MgH2 at 750 K, a large amount of the Mg2Ni phase existed in the intermediate product; (4) hydrogen present
during the heating period makes the product to have smaller size. Further investigation on the mechanism of the gas–solid reaction deserves
attention, because lowering of the synthesis temperature by as much as 100◦ is very attractive and significant for industrializing this process.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Hydrogen storage; Intermetallics; Miscellaneous; Reaction synthesis
1. Introduction
that, by means of mechanical alloying, the hydriding and
dehydriding properties at lower temperature than 573 K can
be much improved due to the physical control of the mi-
crostructures of the yield product, such as the generation of
nanocrystalline alloys, a large number of grain boundaries
and highly disordered interface phase [1–4]. Another effort
of a new production method of magnesium-based alloys is a
newly developed process called hydriding combustion syn-
thesis (HCS) [5].
HCS has been proposed since 1997 and regarded recently
as an innovative processing and fabrication to produce hy-
drogen storage alloys, especially for magnesium-based al-
loys [6,7]. By means of the HCS, magnesium-based hydro-
gen storage alloys can be prepared at 873 K below the melt-
ing point of magnesium and the product was really active,
which was obtained by only one step combining the synthe-
sis of the alloys and the hydrogenation of the alloys without
any activation process [8–12]. It has been recognized widely
as a highly promising technology offering many benefits
Magnesium-based alloy is potentially most important hy-
drogen storage materials from the points of lightweight, high
specific storage capacity, abundant raw materials and low
environmental impact; for example, magnesium-based hy-
drogen storage alloy is one of the most attractive mediums,
as a hydrogen storage tank for fuel cells in transportations.
Nevertheless, the practical application of magnesium-based
alloys has been still limited unfortunately because of their
poor hydriding, dehydriding kinetics and high reaction tem-
perature (623–673 K). Many papers have been published to
overcome these problems, most of which addressing new
production methods. One of the well-known examples is
mechanical alloying or ball milling. The results reported
∗
Corresponding author. Tel.: +81-722-54-9303;
fax: +81-722-54-9911.
E-mail address: takiyama@chemeng.osakafu-u.ac.jp (T. Akiyama).
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2004.04.092

158
L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
such as short processing time, high product purity, low en-
ergy requirement and so on [13–15].
2. Experiment
The reaction mechanism of HCS of Mg2NiH4 from the
compact of a magnesium and nickel mixture at 2.0 MPa hy-
drogen was studied based on systematic experiments [16].
The interesting results were that as much as seven reac-
tions were detected from the peaks of heat flow in a dif-
ferential scanning calorimeter (DSC) curve and from X-ray
diffraction (XRD) analysis. That is, (1) Mg + H2 → MgH2,
with a wide temperature range from about 520–660 K; (2)
MgH2 → Mg + H2, between 675 and 700 K; (3) 2Mg + Ni
→ Mg2Ni (L), an eutectic reaction in the Mg–Ni system
at 790 K; (4) 2Mg + Ni → Mg2Ni, an exothermal reac-
tion of the combustion synthesis; (5) Mg2Ni + 0.15H2 →
Ng2NiH0.3, a solid solution reaction during the cooling pe-
riod; (6) Mg2Ni + H2 → Mg2NiH4 (HT), between 645 and
600 K, and (7) Mg2NiH4 (HT) → Mg2NiH4 (LT), initiating
at 510 K.
The most recent study revealed the reason why the prod-
uct catches hydrogen quickly, in which the kinetics of the
product of HCSed Mg2NiH4 was microscopically studied in
comparison to the commercially available product of Mg2Ni
based on ingot metallurgy (IM) [17]. The results showed
that the HCSed product was fully charged by hydrogen in
the form of Mg2NiH4 just after being synthesized and had
a very large reaction rate without any activation treatment;
only 5 min for full charge. One of the most noteworthy re-
sults is the high activity of the product obtained; that is, it
stored hydrogen even at room temperature. Moreover, obser-
vation by transmission electron microscopy (TEM) revealed
the mechanism of this improved kinetics of the HCSed prod-
uct. Many tree-like nano fissures emerged inside the HCSed
product just after the first dehydrogenation; in contrast, the
IM product has no such fissures even after three-times acti-
vation treatment.
Certainly, all attractive properties of the HCSed product
is related to the microstructures of the product, as the simi-
lar phenomena reported on the mechanical alloying or ball
milling of magnesium-based hydrogen storage alloys. There-
fore, further improvement of the HCSed product properties
depend clearly on the well understanding and control of
the microstructures of the product. Not only the microstruc-
ture of the final product of HCS but also the microstructure
of intermediate products in HCS process should be studied
because there are seven reactions during HCS process and
there must be a close correlation between the microstruc-
ture of intermediate products and final product. However,
no efforts, up to now, have been made to investigate the in-
termediate products of HCS although such research work is
very important to improve the properties of final product,
leading to the industrization of this process. Therefore, the
purpose of this paper is to prepare the intermediate products
in the hydriding combustion synthesis of Mg2NiH4 first and
then to study the microstructure of these intermediate prod-
ucts by means of XRD and scanning electron microscopy
(SEM).
The apparatus used in this study was the same as reported
in [7]. The reactor is an Inconel tube with inside length of
520 mm and inner diameter of 70 mm. Three CA-type ther-
mocouples were attached to three holes under the sample
container of stainless steel through the sample frame. These
thermocouples were used not only for measurement of the
reactor temperature distribution but also for fixing the sam-
ple container on the same place, the middle of the tube, every
time. The temperature distribution measured by the central
thermocouple was used for sample temperature control with
high accuracy. An accuracy within 1 K was confirmed by a
thermocouple inserted into the sample in a preliminary test.
As a result, a uniform temperature zone of 40 mm × 60 mm
was within 5 K. The atmosphere pressure inside of the re-
actor was also controlled precisely within 0.015 MPa. The
samples of the HCS product were prepared from commer-
cially available magnesium and nickel powders. The parti-
cle sizes of the magnesium, less than 180 ␮m, and nickel,
2–3 ␮m, were selected as the minimum sizes in commercial
available chemicals for maximizing the contacting area be-
tween the particles. The powders, with 99.9 mass% in purity,
were mixed well in 2:1 of Mg:Ni molar ratio by an ultra-
sonic homogenizer in acetone for 3.6 ks. After completely
dried, 10 g of the powder mixture was used for the HCS di-
rectly. Generally, in the combustion synthesis, the powder
mixture was usually compressed to increase the contacting
area between the particles. However, an uncompressed pow-
der mixture was used in this study for easy hydrogen pene-
tration into the sample.
In order to synthesize intermediate products in the HCS
of Mg2NiH4, three groups of samples were obtained at dif-
ferent stages according to Fig. 1, which is the DSC curve of
hydriding combustion synthesis of Mg2NiH4 at 4.0 MPa of
hydrogen pressure as reported in [9]. The samples of group
H (High temperature) were obtained by heating the pow-
der mixture up to 850 K after the combustion synthesis of
Mg2Ni. The samples of group L (Low temperature) were
heated up to 705 K or about 750 K after the hydriding reac-
tion of Mg and the dehydriding reaction of MgH2, respec-
tively, and the samples of group E (Eutectic temperature,
779 K) were obtained around 790 K near the eutectic tem-
perature of the Mg–Ni system. Table 1 gives the synthesis
conditions of all samples in this study. The obtained inter-
mediate products were identified by XRD analysis and were
carefully examined by SEM.
3. Results and discussion
Fig. 2 shows the XRD patterns of sample group H. Argon
was used in heating and cooling periods for sample 1902. In
contrast, hydrogen was used in heating and argon in cool-
ing for sample 1901 and hydrogen was used in heating and
cooling for sample 1701. Fig. 3 shows the SEM images of

L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
159
1.5
1
Combustion Synthesis of Mg2Ni
Hydriding of Mg2Ni
0.5
0
-0.5
-1
Hydriding of Mg
EutecticReaction
Dehydriding of MgH2
Heating
Cooling
-1.5
-2
460
560
660
760
860
Temperature (K)
Fig. 1. DSC curve of hydriding combustion synthesis of Mg2NiH4 at 4.0 MPa of hydrogen pressure.
Table 1
Experimental conditions of intermediate products synthesized during the hydriding combustion synthesis of Mg2NiH4
Group no.
Sample no.
Synthesis temperature
(K)
Atmosphere in heating
(MPa)
Atmosphere in cooling
(MPa)
Comments
H (high)
1902
1901
1701
850
853
870
Ar, 0.1
H2, 4.0
H2, 4.0
Ar, 0.1
Ar, 0.1
H2, 4.0
After combustion synthesis of Mg2Ni
After combustion synthesis of Mg2Ni
Final product of Mg2NiH4
L (Low)
1702
3001
1803
705
750
757
H2, 4.0
Ar, 0.1
H2, 4.0
H2, 4.0
Ar, 0.1
Ar, 0.1
After hydriding of Mg
After dehydriding of MgH2
After dehydriding of MgH2
E (Eutectic)
1801
2202
794
795
H2, 4.0
Ar, 0.1
Ar, 0.1
Ar, 0.1
Near eutectic point
Near eutectic point
Reference
1700
None
None
None
Raw material
these three samples. Comparing the XRD patterns of sam-
ples 1901 with that of 1902 as shown in Fig. 2, almost no
differences existed and all showed a single phase of Mg2Ni,
except that the diffraction peaks of sample 1901 shifted to
the left side slightly. However, if comparing the SEM images
of the samples 1901 and 1902 as shown in Fig. 3, apprecia-
ble differences existed in their product morphology and the
size of the product on the right hand side of the figure when
Mg2Ni
Mg2NiH4
Ar 850K Ar, after CS of Mg2Ni
H2 853K Ar, after CS of Mg2Ni
1902
1901
1701
H2 870K H2, final of Mg2NiH4
10
20
30
40
50
60
70
80
2 Theta
Fig. 2. XRD pattern of mixture powder 2Mg + Ni heated up to 850 and 870 K under different atmospheres by the process of hydriding combustion
synthesis.

160
L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
Fig. 3. SEM images (1500×) of mixture powder 2Mg + Ni heated up to 850 and 870 K under different atmospheres by the process of hydriding
combustion synthesis.
hydrogen was used in the heating period is obviously smaller
than that on the left hand side when argon was used in the
same period. The reason why a hydrogen atmosphere during
heating reduced the particle size is not clear at present. The
point here is that a hydrogen atmosphere during the heating
period generates a small particle size, increasing the spe-
cific surface area as a result, although the phase identified
by XRD of the product was the same (Mg2Ni).
period does not affect the XRD pattern of the intermediate
product of Mg2Ni, but reduces the particle size. It should be
beneficial for improving the kinetics of the hydrogen stor-
age alloys due to an expansion of the surface area. (3) On
the final product of HCS, some cracks in the particles of
Mg2NiH4 were generated. The cracks are also useful to ac-
celerate the hydriding and dehydriding of the hydrogen stor-
age alloys at the initial stage.
The XRD pattern of sample 1701 in Fig. 2 showed a very
pure phase as the final product of the hydriding combus-
tion synthesis of Mg2NiH4. From the SEM image of sam-
ple 1701 in Fig. 3, one can recognize that the size became
slightly larger in comparison to that in 1901 and a large
number of cracks appeared on the particle surfaces and/or
across the particles. The cracks must be from the expansion
of the intermediate product of the combustion synthesis of
Mg2Ni during the hydriding reaction during the cooling pe-
riod under hydrogen atmosphere, because there are no cracks
in those particles when argon was used during cooling as
shown in the SEM images of samples 1901 and 1902.
Some of the conclusions, which can be derived from
Figs. 2 and 3, are: (1) Pure Mg2Ni, as an intermediate prod-
uct during the HCS of Mg2NiH4, is obtainable from the
powder mixture of magnesium and nickel by heating up to
850 K under argon or hydrogen. Moreover, pure Mg2NiH4,
as the final product of HCS, is generated at 4.0 MPa of hy-
drogen pressure during the cooling period as reported pre-
viously [18], (2) A hydrogen atmosphere during the heating
Fig. 4 gives the XRD patterns of samples of group L.
Sample 1702 was heated up to 705 K then kept there for
half an hour and a hydrogen atmosphere was used during
both heating and cooling. Our idea was to obtain the inter-
mediate product of the metal hydride of MgH2 based on the
DSC curves in Fig. 1. Sample 1803 was heated up to 757 K,
in which hydrogen and argon were used during heating and
cooling, respectively. The heating temperature of 757 K was
selected and argon atmosphere was used during cooling, be-
cause 757 K is just after the dehydriding reaction of MgH2
as shown in Fig. 1. Therefore, the intermediate product was
produced shortly after the dehydriding reaction. As a refer-
ence, sample 3001 was heated up to 750 K but argon was
used during both heating and cooling. Fig. 5 shows the SEM
images of the samples of group L. Here, the image of sam-
ple 1700 is from the original mixture of metal powder of
magnesium and nickel.
The XRD pattern of sample 1702 in Fig. 4 shows that it
was a mixture of three phases; these were MgH2, unreacted
Mg and Ni. This result indicated that the hydriding of Mg

L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
161
Mg2Ni
Ni
Mg
H2 705K H2 after hyd. of Mg
1702
3001
1803
MgH2
Ar 750K Ar after dehyd. of MgH2
H2 757K Ar after dehyd. of MgH2
10
20
30
40
50
60
70
80
2 Theta
Fig. 4. XRD pattern of mixture powder 2Mg + Ni heated up to 705 and 750 K under different atmosphere by the process of hydriding combustion synthesis.
was not completed even at 4.0 MPa pressure of hydrogen
and at 705 K for about 30 min. Comparing the XRD patterns
of sample 3001 with sample 1803, all consist of a mixture
of three phases. These were Mg2Ni, unreacted Mg and Ni,
except the peak height of Mg and Ni was a little high in
sample 1803 when hydrogen was used during heating. It
seems that there is an effect of hydrogen on the synthesis
of Mg2Ni during heating. On other hand, the existence of
Mg2Ni just after the dehydriding reaction of MgH2 in the
samples 3001 and 1803 is not predicted from the knowledge
of the reaction mechanism of the process reported in Ref.
[16]. Actually, the amount of Mg2Ni was not small but large
from the relative XRD peak heights of these phases.
From the SEM images as shown in Fig. 5, two very in-
teresting features of the morphologies of these intermediate
products can be observed. One is the corrosion of the sur-
Fig. 5. SEM images (500×) of mixture powder 2Mg + Ni heated up to 705 and 750 K under different atmospheres by the process of hydriding combustion
synthesis.

162
L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
faces of the magnesium particles. This kind of corrosion can
be called thermal corrosion because it was developed dur-
ing increasing temperature. It is not difficult to observe the
development of corrosion if comparing the SEM image of
sample 1702, heated up to 705 K with that of sample 1700,
the raw material of mixture powder and with that of sample
3001 or sample 1803 heated up to 750 K. Another feature of
the morphology of these intermediate products is the growth
of the particle size of the nickel. Again, it was developed
with increasing temperature obviously and it occurred even
at only 705 K if comparing the particle sizes, as indicated
by the arrows, of sample 1702 with sample 1700 in Fig. 5.
Taking into account of the XRD results, that is a large
amount of Mg2Ni existed, as shown in Fig. 4 and the ther-
mal corrosion of the particles of Mg mentioned above, it
seems to be possible to understand that the growth of nickel
particles was from the synthesis of Mg2Ni, that is the mag-
nesium was evaporated by thermal corrosion and transferred
to the particles of nickel in gas phase, then the synthesis of
Mg2Ni by diffusion reaction, even though the temperature
of 705 K and 750 K was much lower than 790 K, the eutectic
temperature of Mg–Ni system as shown in Fig. 2. Because
there was not much interface contact between the nickel par-
ticles and magnesium particles and the amount of Mg2Ni
was not small, as well as the temperatures of 705 and 750 K
are very low, the synthesis mechanism such as gas–solid re-
action must have played an important role in the process of
hydriding combustion synthesis. A detailed investigation of
this point is planned and the result will be in the near future.
Since 750 K is lower than the usual temperature of HCS of
Mg2NiH4, that is 850 K by 100◦ it must be very attractive
and significant to an industrialization of this process.
sample 2202 considering the low diffraction sensitivity of a
light metal and the fact that the peak of Ni appeared in the
pattern. This result is similar to the observation that more
magnesium remained when hydrogen was used than when
argon was used during heating, as follows from the results
of sample 3001 and sample 1803 in Fig. 4. Why the appear-
ance of hydrogen retarded the synthesis reaction between
magnesium and nickel needs to be further studied from the
point of view of reaction kinetics and thermodynamics of
our system.
Some shifts of the Mg2Ni diffraction peaks to lower an-
gles appeared again when hydrogen was used during heat-
ing. This can be seen by comparing the XRD patterns of
samples 1801 and 2202 in Fig. 6. This kind of diffraction
peak shifts comes from the expansion of the crystal lattice
as mentioned previously. The expansion of the crystal lat-
tice is due to the solid solution of hydrogen into the crystal
lattice of Mg2Ni because the hydrogen used during heating
period is not easily completely removed from the furnace
tube at high temperature, and remained in the argon atmo-
sphere used during cooling.
Fig. 7 shows the SEM images of samples 1801 and 2202.
The most obvious difference is the particle sizes. Here a
particle is actually an aggregation of particles. We find again
that the particle sizes are smaller when hydrogen used than
when argon used during heating as discussed previously for
samples 1901 and 1902 in Fig. 3. The result indicates that
hydrogen strongly interfered in the particle size growth, and
that such interference happened already below 800 K during
heating. Since the particle sizes of sample 1801 was almost
the same as that of sample 1901 and the particle sizes of
sample 2202 was almost the same as that of sample 1902
as shown in Fig. 3 and Fig. 7, this result indicates that the
intermediate products of the eutectic below 800 K already
possessed the morphology characteristic of particles of the
final product.
Fig. 6 shows the XRD patterns of intermediate products
of samples 1801 and 2202 synthesized at 793 K near the
eutectic point of the Mg–Ni system but before the combus-
tion synthesis reaction of Mg2Ni. Hydrogen during heating
and argon during cooling were used for sample 1801. How-
ever, argon was used during heating and cooling for sample
2202. The XRD patterns showed that three phases, Mg2Ni,
Mg and Ni, existed in samples 1801 and 2202 and the main
phase was Mg2Ni. Although the diffraction peaks of Mg did
not appear in the pattern of sample 2202, Mg did exist in
Based on the reaction mechanism of HCS of Mg2NiH4,
the purpose to use hydrogen in the process was only for
the formation of the final product Mg2NiH4 during cooling,
though it was also involved in the hydriding and dehydriding
reactions of Mg and the hydriding reaction of Mg2Ni. The
results here, however, revealed that hydrogen also played
Mg2Ni
Mg
Ni
1801
H2 794K Ar, near eutectic point
Ar 795K Ar, near eutectic point
2202
10
20
30
40
50
60
70
80
2 Theta
Fig. 6. XRD pattern of mixture powder 2Mg + Ni heated up to 795 K under different atmosphere by the process of hydriding combustion synthesis.

L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
163
Fig. 7. SEM images (1500×) of mixture powder 2Mg + Ni heated up to 795 K under different atmospheres by the process of hydriding combustion
synthesis.
an important role in the morphology of the particles of the
intermediate product during heating, which determined sub-
stantially the morphology of the particles of final product.
This is possibly related to the hydriding and dehydriding
reactions of magnesium before the eutectic reaction during
heating. This role played by hydrogen in the morphology of
particles remains to be further investigated.
Further investigation of this mechanism is deserved be-
cause the lowering of synthesis temperature by 100◦ is at-
tractive and significant.
Acknowledgements
This research was supported by National Natural Sci-
ence Foundation of China (Project 50371037), Fellowship
Program of Japan Society for the Promotion of Science
(S-03117) and the Ministry of Education, Science, and
Culture, Japan (Grant No. 14350402). The authors would
like to express their great thanks to TAIATSU TECHNO
CORPORATION for the co-operation in the development
of the new version of the synthesis reactor used in this
study.
4. Conclusion
Intermediate products at several stages during the hydrid-
ing combustion synthesis (HCS) of Mg2NiH4 were prepared
and their microstructures were studied by means of XRD
and SEM.
(1) The intermediate product of eutectic composition at tem-
perature lower than 800 K affected strongly the particle
morphology of the final product of Mg2NiH4 obtained
at temperature higher than 850 K by HCS.
References
(2) Hydrogen during heating leads to the product smaller
size. The hydrogen atmosphere would be quite effective
for improving the kinetics due to large surface area in
hydrogen storage alloys.
[1] S.S. Han, N.H. Goo, W.T. Jeong, K.S. Lee, J. Power Sources 92
(2001) 157–162.
[2] T. Aizawa, Mater. Sci. Forum 350–351 (2000) 299–310.
[3] S. Bouaricha, J.P. Dodelet, D. Gulay, J. Huot, S. Boily, R. Schulz,
J. Alloys Compd. 307 (2000) 226–233.
(3) Severe thermal corrosion; cotton-like feature, was ob-
served on the surface of the intermediate products of
Mg obtained near 750 K. It was observed even on the
intermediate product of MgH2 obtained at 705 K.
(4) A large amount of Mg2Ni appeared in the intermedi-
ate product at 750 K just after the dehydriding reaction
of MgH2. From the observation of the particle size of
nickel, the synthesis of Mg2Ni happened even at only
705 K, which is much lower than 790 K, the eutectic
temperature of Mg–Ni system.
[4] Y. Hui, C. Lixin, L. Yongquan, Rare Met. Mater. Eng. (China) 29
(2000) 193–196.
[5] T. Akiyama, H. Isogai, J.-i. Yagi, J. Alloys Compd. 252 (1997) L1–
L4.
[6] J.-i. Yagi, T. Akiyama, L. Li, In: Proceedings of the Processing
and Fabrication of Advanced Materials X, 5–8 November, 2001,
Indianapolis, Indiana, ASM international, pp. 223–237
[7] I. Saita, L. Li, K. Saito, T. Akiyama, Mater Trans. 43 (2002) 1100–
1104.
[8] L. Li, I. Saita, K. Saito, T. Akiyama, J. Alloys Compd. 345 (2002)
189–195.
[9] L. Li, T. Akiyama, J.-i. Yagi, Intermetallics 7 (1999) 201–205.
[10] L. Li, T. Akiyama, J.-i. Yagi, J. Alloys Compd. 316 (2001) 118–123.
[11] L. Li, T. Akiyama, J.-i. Yagi, Int. J. Hydrogen Energy 26 (2001)
1035–1040.
(5) Since no effective interface contact tightly between the
particles of nickel and magnesium, gas–solid reaction
must have played an important role before the temper-
ature reached to 790 K in the HCS.
[12] L. Li, T. Akiyama, J.-i. Yagi, J. Alloys Compd. 308 (2000) 98–103.

164
L. Li et al. / Journal of Alloys and Compounds 384 (2004) 157–164
[13] L. Li, T. Akiyama, T. Kabutomori, J.-i. Yagi, J. Mater. Trans. JIM
40 (1999) 1079–1083.
[16] L. Li, T. Akiyama, J.-i. Yagi, Intermetallics
677.
7 (1999) 671–
[14] L. Li, T. Akiyama, T. Kabutomori, K. Terao, J.-i. Yagi, J. Alloys
Comp. 281 (1998) 175–180.
[17] T. Akiyama, K. Saito, I. Saita, J. Electrochem Soc. 150/9 (2003)
450–454.
[15] L. Li, T. Akiyama, T. Kabutomori, K. Terao, K. Terao, J.-i. Yagi, J.
Alloys Compd. 287 (1999) 98–103.
[18] L. Li, T. Akiyama, J.-i. Yagi, J. Mater. Syn. Process 8 (2000) 7–
14.