Hydrolysis of Mg-salt and MgH2-salt mixtures prepared by ball milling for hydrogen production

Article abstract of DOI:10.1016/j.jallcom.2005.09.008

It is well known that the reaction between Mg or MgH₂ and water produces hydrogen. However, this reaction stops rapidly, due to the formation of a passive Mg(OH)₂ layer onto the reactive material. The milling of Mg or MgH₂ with solid salts appears very effective to improve their hydrolysis reactivity. The milling time (0.5, 3 and 10 h), the amount of salt (1, 3 and 10 mol%) and the salt composition (KCl, NaCl, LiCl and MgCl₂) are all parameters that control the hydrolysis reaction in terms of conversion yield and kinetics. The salt additive acts as a process control agent upon milling that increases the specific surface area of the Mg powder (not valid for more brittle MgH₂ powder). In addition, the salt dissolution during the hydrolysis is assumed to break down the passive Mg(OH)₂ layer, favouring the reaction between Mg or MgH₂ and water. The driving force related to the exothermic dissolution of the salt additive such as MgCl₂ is also a key factor for accentuating the MgH₂ and Mg hydrolysis. The 0.5 h milled MgH₂-3 mol% MgCl₂ composite displays the best compromise in terms of milling duration and hydrogen production performance with 964 ml of hydrogen produced per gram of composite.

Full text of DOI:10.1016/j.jallcom.2005.09.008

Journal of Alloys and Compounds 416 (2006) 296–302
Hydrolysis of Mg–salt and MgH2–salt mixtures prepared
by ball milling for hydrogen production
∗
´ `
´
Marie-Helene Grosjean, Lionel Roue
´
INRS-Energie, Mate´riaux et Te´le´communications, 1650 boulevard Lionel Boulet, Varennes (QC), Canada J3X 1S2
Received 8 August 2005; accepted 5 September 2005
Available online 10 October 2005
Abstract
It is well known that the reaction between Mg or MgH2 and water produces hydrogen. However, this reaction stops rapidly, due to the formation
of a passive Mg(OH)2 layer onto the reactive material. The milling of Mg or MgH2 with solid salts appears very effective to improve their hydrolysis
reactivity. The milling time (0.5, 3 and 10 h), the amount of salt (1, 3 and 10 mol%) and the salt composition (KCl, NaCl, LiCl and MgCl2) are
all parameters that control the hydrolysis reaction in terms of conversion yield and kinetics. The salt additive acts as a process control agent upon
milling that increases the specific surface area of the Mg powder (not valid for more brittle MgH2 powder). In addition, the salt dissolution during
the hydrolysis is assumed to break down the passive Mg(OH)2 layer, favouring the reaction between Mg or MgH2 and water. The driving force
related to the exothermic dissolution of the salt additive such as MgCl2 is also a key factor for accentuating the MgH2 and Mg hydrolysis. The
0.5 h milled MgH2–3 mol% MgCl2 composite displays the best compromise in terms of milling duration and hydrogen production performance
with 964 ml of hydrogen produced per gram of composite.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Hydrogen production; Hydrolysis reaction; Chemical hydride; Mg; MgH2; Ball milling
1. Introduction
is irreversible and considering the high cost of NaBH4 (∼US$
50 kg−1), its use in high power systems such as propulsion appli-
The reaction of chemical hydrides (e.g. LiH [1,2], CaH2
[1–3], MgH2 [4–6], NaBH4 [1,7–17], LiBH4 [7,8,18], NaAlH4
[1,8], LiAlH4 [1,3,8]) with water appears an attractive means of
generating pure hydrogen for portable fuel cells without the need
and associated energy penalties for compression or liquefaction.
For example, the Millennium Cell company [9] has developed
a system, which generates hydrogen from an aqueous solution
of sodium borohydride (20% NaBH4–10% NaOH–70% H2O)
passing through a chamber containing a Ru-based catalyst. The
major advantage of this technology is its ability to store hydro-
gen in a safe and easily handling form and extract it only on
demand. In addition, the hydrogen is supplied with co-generated
moisture since the heat generated during the hydrolysis reaction
vaporizes a fraction of the water present. This is an added benefit
for PEM fuel cells, which require a constant humidification of
the proton exchange membrane. However, because the reaction
cations is not currently economically feasible. In contrast, its use
in small fuel cell systems (<1 kW) for portable applications such
as consumer electronics and military applications is attractive.
Purehydrogencanbealsoobtainedbyreactinglessexpensive
MgH2 (∼US$ 25 kg−1) with water through the reaction:
MgH2 + 2H2O → Mg(OH)2 + 2H2
(1)
The theoretical hydrogen yield of this reaction is 6.4 wt%
when water is included in the calculation. The hydrogen yield
increases to 15.2 wt% if the water produced by the fuel cell
is recovered for the hydrolysis reaction. Moreover, this reac-
tion has the advantage of producing residual Mg(OH)2, which
is environmental friendly. It is also possible to produce hydro-
gen by using very cheap Mg material (∼US$ 2–5 kg−1) by the
following reaction:
Mg + 2H2O → Mg(OH)2 + H2
(2)
In this case, the hydrogen yield is 3.3 wt% (considering water
mass) and 8.2 wt% (no water included in the calculation). How-
ever, the hydrolysis reaction using conventional Mg or MgH2 is
∗
Corresponding author. Tel.: +1 450 929 8185; fax: +1 450 929 8102.
E-mail address: roue@emt.inrs.ca (L. Roue´).
0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2005.09.008

M.-H. Grosjean, L. Roue´ / Journal of Alloys and Compounds 416 (2006) 296–302
297
rapidly interrupted because of the formation of a passive mag-
nesium hydroxide layer on the reactive material. Acid can be
added to form soluble Mg2+ species but this is detrimental to
the equipment and provides a potential hazard for users.
The hydrolysis of MgH2 can be improved in terms of yield
and kinetics by using nanocomposites MgH2–X (X = Ca, Li,
LiAlH4, CaH2) materials prepared by ball milling [4]. The best
performance was obtained with MgH2 + 20 mol% CaH2 powder
mixture milled during 10 h, leading to a reaction yield close to
80% after 30 min of hydrolysis with excess water. It was stated
that the nanostructure formed upon extensive milling is essential
to obtain high yield and fast kinetics of the hydrolysis reaction.
The requirement of such an extended milling time (10 h) is, how-
ever, unfavourable to produce an economically viable hydrogen
source material.
The oxygen content in the powders was measured by inert gas fusion tech-
nique with a TC-600 detector from LECO. The iron content was determined by
EDX analysis.
The hydrolysis reactions were carried out at room temperature and atmo-
spheric pressure in a flask reactor of 1000 ml with two openings, one for water
addition (using funnel pressure equalization) and the other for hydrogen exhaust.
The gas produced was flowed through a condenser and drierite to remove all
water vapour before passing through a flowmeter (ADM 3000, Agilent Tech-
nologies). The flowmeter was connected to a computer for recording gas flow
and volume as a function of time. Ten millilitres of aqueous solution were added
to react with 250 mg of powder. A magnetic stirrer agitated the solution contin-
uously during the test. The background flow (0.15 ml/min) was subtracted from
the data. Each test was repeated at least two times and the precision of the mea-
surement was estimated at 5%. Hydrogen production is expressed as conversion
yield (%) defined as the volume of produced hydrogen over the theoretical vol-
ume of hydrogen that should be released assuming that all Mg or MgH2 material
is hydrolyzed.
In a recent work [5], we have shown that when hydrolysis is
performed in the presence of chloride ions, a significant increase
of the H2 production is observed with Mg powder milled for only
30 min. This performance was explained by the accentuation of
the pitting corrosion of Mg due to the creation of numerous
defects and fresh surfaces through the 30 min milling process.
A full completion of the hydrolysis reaction in the presence of
chloride ions was observed with Mg–10 at% Ni composite mate-
rial milled for 30 min. This behavior is related to the creation of
micro-galvanic cells between well-distributed Mg and Ni ele-
ments, which accentuates greatly the Mg corrosion. However,
these compounds have the inconvenience of requiring an aque-
ous media containing chlorine ions to be reactive. Thus, the pure
water produced by the fuel cell cannot be used for the hydroly-
sis reaction. Moreover, milled MgH2 and MgH2–Ni composites
display low reactivity even in KCl solution, probably because
the insulator character of MgH2 prevents its corrosion.
In this work, we present an investigation of the hydrogen
production from Mg–solid salt and MgH2–solid salt mixtures
treated by ball milling. The aim is to conceive highly reactive
materials to ensure a full completion of the hydrolysis reaction
using pure and neutral water solution. It is expected that the
concomitant salt dissolution upon hydrolysis could be beneficial
in breaking down the passive Mg(OH)2 layer on particles, thus
favouring the extensive reaction between freshly exposed Mg or
MgH2 surface and water.
3. Results and discussion
3.1. Characterization of milled Mg–KCl mixtures
Fig. 1 shows the evolution of the specific surface area of
Mg–x mol% KCl (x = 0, 1, 3 and 10) mixtures as a func-
tion of the milling time. For the Mg powder milled without
salt, its specific surface area decreases with increasing milling
time (from ∼0.7 m2 g−1 before milling to ∼0.1 m2 g−1 after
10 h of milling). In contrast, the specific surface area of the
Mg–x mol% KCl mixtures with x = 1 and 3 reaches a maxi-
mum of ∼1.2 m2 g−1 after 0.5 h of milling and then decreases to
∼0.2 m2 g−1 after 3 and 10 h of milling. With Mg–10 mol% KCl
mixture, no decrease in the specific surface area of the powder
is observed after prolonged milling and so, a maximum value of
3 m2 g−1 is obtained after 10 h of milling. This reflects the major
influenceofthesaltadditiveintherepeatedfracturing/coldweld-
ing processes upon ball milling. That is KCl acts as a process
control agent that impedes clean Mg-to-Mg contacts necessary
for cold welding and thus, the powder fracturing is promoted.
This effect was also observed by Ivanov et al. [19] with var-
2. Experimental
MgH2 (Th. Goldschmidt, 95 wt% MgH2, 5 wt% Mg, 20 ␮m), Mg and Ni
(Alfa Aesar, 99.8 wt%, −325 mesh) and KCl, NaCl, LiCl and MgCl2 salts (Alfa
Aesar, ACS grade, 99%) were used as starting materials. The salts were carefully
dried before using. Ball milling was performed under argon atmosphere with
a ball-to-powder mass ratio of 8:1 using a SPEX 8000 ball mixer. The milling
time was varied from 0.5 to 10 h.
The samples were characterized by X-ray diffraction (XRD) using a Bruker
AXSD8 Siemens diffractometer with Cu K␣ radiation. The crystallite size
and internal strain were determined from the peak broadening using the
Williamson–Hall plot.
Scanning electron microscopy (SEM) observations were made using a
Hitachi SE2300 SE/N microscope. The distribution of salts in composite pow-
ders was checked by energy dispersive X-ray (EDX) mapping.
The specific surface area of the powders was measured by N2 adsorption
(multipoint BET) using a Quantachrome Autosorb Automated Gas Sorption
system.
Fig. 1. Specific surface area of Mg and Mg–x mol% KCl (x = 1, 3 and 10) pow-
ders as a function of the milling time.

298
M.-H. Grosjean, L. Roue´ / Journal of Alloys and Compounds 416 (2006) 296–302
Fig. 2. SEM micrographs of (a) Mg powder and (b) Mg–10 mol% KCl mixture obtained after 30 min of milling.
ious ball-milled Mg–salt composite materials. However, when
the Mg–KCl composite is submitted to a prolonged ball milling
(>0.5 h) with a low amount of KCl (≤3 mol%), the embedding of
KCl into the soft Mg matrix may occur, decreasing the amount of
KCl at the surface of the Mg particles. Then, the creation of clean
Mg surfaces favouring the Mg powder coalescence becomes
again predominant, as observed with Mg material milled alone.
The formation of smaller Mg particles when milled with KCl
is confirmed by SEM observations (Fig. 2). For Mg powder
milled alone during 0.5 h, large Mg platelets with a mean size
larger than 150 ␮m were observed in comparison to smaller and
irregular Mg flakes with a mean diameter lower than 50 ␮m for
0.5 h milled Mg–10 mol% KCl composite. Note that the KCl
particles are not clearly discernible but the EDX mapping (not
shown) confirms that after 30 min of milling, KCl is homoge-
neously distributed in the composite.
as the milling time increases, reflecting the reduction of the
crystallite size and the accumulation of microstrains. The Mg
crystallite size is reduced from 139 (as-received state) to 94 nm
after 0.5 h of milling and reaches a steady-state value of 66 nm
after 10 h of milling. The Mg lattice strain is 0.16% after 10 h of
milling. Close values were obtained for Mg milled without KCl
(d = 54 nm, ε = 0.25%after10 hmilling)[20]. Therelativelyhigh
limit value of the Mg crystallite size and low strain level indi-
cate that the recovery process is facilitated during the milling
process, which is usually observed with metals having a low
melting temperature such as Mg [21]. A structural refinement
is also observed for KCl powder: its crystallite size is reduced
from 110 nm before milling to 77 and 23 nm after 0.5 and 10 h
of milling, respectively.
3.2. Hydrolysis reactivity of milled Mg–salt mixtures
The iron contamination of the Mg and Mg–KCl powders
remains very low (<0.3 wt% after 10 h of milling), indicating
that the erosion of the vial and the balls upon milling is very
limited even in presence of KCl. Oxygen contamination appears
more important since ca. 2.5 wt% of oxygen is measured after
10 h of milling for Mg and Mg–KCl powders compared to ca.
0.5 wt% for unmilled powders. However, no significant Mg oxi-
dation occurs in the first 30 min of milling. The source of oxygen
during milling is residual/leaking air present in the container and
oxides present at the surface of the milling tools [20].
Fig. 4 displays the hydrogen production profiles for reaction
with pure water of Mg and Mg–x mol% KCl (x = 1, 3 and 10)
Fig. 3 displays the X-ray diffraction patterns of the
Mg–10 mol% KCl powder as a function of the milling time.
Except for an extra peak attributed to MgO, just discernable after
10 h of milling, no new peak appears and no peak displacement
occurs with milling, indicating that no reaction occurs between
MgandKCl. AnincreaseoftherelativeintensityoftheMg(0 0 2)
peak at 34.4◦ is observed for 0.5 h milled Mg powder, indicat-
ing a preferential orientation along the c-axis. It may result from
the weaker energy required to carry out a deformation by slip
perpendicular to the c-axis than in another direction. During pro-
longed milling (t ≥ 3 h), considerable energy is accumulated into
the powder and dislocations occur in all directions, i.e. deforma-
tions are random and the texture disappears. An accentuation of
the broadening of the Mg and KCl diffraction peaks is observed
Fig. 3. X-ray diffraction patterns of Mg–10 mol% KCl mixture as a function of
the milling time.

M.-H. Grosjean, L. Roue´ / Journal of Alloys and Compounds 416 (2006) 296–302
299
Fig. 4. Hydrogen production profiles for reaction with water of (a) Mg, (b) Mg–1 mol% KCl, (c) Mg–3 mol% KCl and (d) Mg–10 mol% KCl powders milled for
different times.
powdersmilledfordifferenttimes. AsseeninFig. 4a, thereactiv-
ity of the Mg powder milled without KCl is not improved by the
ball milling treatment since all curves display a conversion yield
lower than 20% after 1 h of reaction. This is due to the forma-
tion of a Mg(OH)2 layer onto the Mg particles, which prevents
further contact between water and unreacted Mg material [20].
In contrast, the ball milling affects the reactivity of the Mg–KCl
composites in pure water. For x = 1 and 3 mol% KCl (Fig. 4b and
c), the optimum milling time is 0.5 h with, respectively, a conver-
sion yield of 35 and 45%. Further milling leads to a substantial
decrease of their hydrolysis reactivity. For x = 10 mol% KCl, no
decay of the conversion yield is observed with extended milling
since a conversion yield of ∼50% is measured with composites
milled 0.5, 3 and 10 h (Fig. 4d).
The reactivity decay with prolonged milling for Mg–x mol%
KCl (x = 1 and 3) powders can be explained by a decrease of
their specific surface area (see Fig. 1). The Mg oxidation upon
prolonged milling can also affect the hydrolysis reactivity as
reported previously for Mg powders [5]. Moreover, the embed-
ding of KCl into the Mg particles during a prolonged milling
(decreasing the amount of KCl in direct contact with water) may
limit the rupture process of the passive Mg(OH)2 layer through
the salt dissolution, which is assumed to favour the hydrolysis
reaction.
For comparison, it appears that the addition of KCl in the
aqueous media [5] is more effective in increasing the conver-
sion yield of the Mg powder than its incorporation in solid
state to form a leaching Mg–KCl composite. Indeed, a conver-
sion yield of 89% was reached with 0.5 h milled Mg immersed
in 1 M KCl solution [5]. The increase of the Mg conversion
yield in KCl solution was explained by the destabilization of
the Mg(OH)2 passive layer due to the substitution of the OH−
ions by the Cl− ions to form soluble MgCl2. The hydroly-
sis reaction is, however, more rapid with Mg–KCl composites
since no induction period for hydrolysis initiation is required
(in contrast to that observed with Mg in KCl solution). This
indicates that the mechanism favouring the hydrolysis reac-
tion is not the same: lixiviation process for Mg–salt mix-
tures versus pitting corrosion phenomenon for Mg immersed
in KCl solution. The acceleration of the hydrolysis reaction
by in situ KCl leaching process was also observed with milled
Mg–Ni–KCl composite materials (not shown) but their reactiv-
ity was so high that a fire took place immediately after water
addition.

300
M.-H. Grosjean, L. Roue´ / Journal of Alloys and Compounds 416 (2006) 296–302
Fig. 5. Hydrogen production profiles for reaction with water of 0.5 h milled
Mg–3 mol% salt (KCl, NaCl, KCl, MgCl2) powders.
Fig. 6. Conversion yield in water of 0.5 h milled Mg–3 mol% salt (KCl, NaCl,
KCl, MgCl2) powders as a function of the enthalpy of dissolution of the salt
[22].
Fig. 5 shows the influence of the salt nature (KCl, NaCl,
LiCl and MgCl2) on the hydrolysis reactivity in pure water of
Mg–3 mol% salt mixtures milled for 0.5 h. MgCl2 appears as
the best salt additive, leading to a conversion yield of 77% after
1 h of hydrolysis in comparison to 57, 49 and 45% with LiCl,
NaCl and KCl, respectively. A conversion yield close to 100% is
even obtained with the composite Mg–10 mol% MgCl2 powder
milled 0.5, 3 and 10 h. This order (MgCl2 > LiCl > NaCl > KCl)
in terms of hydrolysis yield and kinetics is observed for all com-
posites (x = 1, 3 and 10) and is maintained whatever the unit
used for expressing the amount of salt (mol%, wt%, vol% or Cl−
concentration). Such a behaviour cannot be simply explained by
a dependence of the salt composition on the composite surface
areasince, forinstance, Mg–3 mol%KCl, Mg–3 mol%NaCland
Mg–3 mol% LiCl mixtures milled for 0.5 h display a similar spe-
cific surface area (1.1 m2/g). However, it must be noted that the
milling of Mg with 3 mol% MgCl2 for 0.5 h leads to a composite
material with a higher specific surface area (SBET = 3.0 m2/g).
Actually, the Mg–salt composite hydrolysis reactivity can
be related to the enthalpy of dissolution of the salt in water
as shown in Fig. 6. In other words, the driving force related
to the largely negative value of the enthalpy of dissolution of
MgCl2 in water (−155 kJ mol−1) greatly favours the hydrolysis
reaction. The extra heat released through the exothermic salt dis-
solution would also help to increase the yield and kinetics of the
hydrolysis reaction. However, from the enthalpy of dissolution
of MgCl2 in water, the calculated global temperature increase
of the aqueous media associated with the MgCl2 dissolution
appears very limited (∼1 ◦C for 250 mg of Mg–3 mol% MgCl2
immersed in 10 ml of water) in comparison with the global tem-
perature increase of the aqueous solution (∼77 ◦C for 250 mg of
Mg–3 mol% MgCl2 in 10 ml of water) due to the Mg hydrolysis
reaction (ꢀHr = −354 kJ mol−1). Thus, to be effective, the peak
temperature associated with the salt dissolution must be local-
ized in micrometric zones where the salt dissolution from the
Mg matrix occurs.
(Fig. 7). This could reflect the major influence of the global tem-
perature of the media on the Mg hydrolysis reactivity.
3.3. Hydrolysis reactivity of milled MgH2–MgCl2 mixtures
As observed for Mg-based materials, the milling of MgH2
with MgCl2 accentuates greatly the MgH2 reactivity (Fig. 8).
After 60 min in pure water, a conversion yield of 61% is obtained
with the 0.5 h milled MgH2–3 mol% MgCl2 composite in com-
parison to only 25 and 11% for 0.5 h milled and unmilled
MgH2 powders, respectively. In addition, as observed with
milled Mg–salt composites, the MgH2 powder conversion yield
increases with the amount of salt and reaches its maximum for a
milling duration of 0.5 h (Fig. 9). Note that a conversion yield as
high as 81% is obtained with the MgH2–10 mol% MgCl2 mate-
rial milled for 10 h. However, the 0.5 h milled MgH2–3 mol%
Note that the conversion yield decreasesdrasticallyasthevol-
ume of water added to react with the composite powder increases
Fig. 7. Conversion yield of 0.5 h milled Mg–3 mol% MgCl2 powder (250 mg)
as a function of the added volume of water.

M.-H. Grosjean, L. Roue´ / Journal of Alloys and Compounds 416 (2006) 296–302
301
Fig. 10. Conversion yield in water of various MgH2 and MgH2–salt powders as
a function of their effective surface area.
Fig. 8. Hydrogen production profiles for reaction with water of as-received
MgH2, 0.5 h milled MgH2 and 0.5 h milled MgH2–3 mol% MgCl2 powders.
4. Conclusion
MgCl2 composite displays the best compromise in terms of
milling duration and hydrogen production performance with
964 ml of hydrogen produced per gram of composite, which
corresponds to a hydrogen yield of 8.6 wt% (the water weight is
not included in the calculation).
Fig. 10 displays the conversion yield of various milled MgH2
and MgH2–MgCl2 composite materials as a function of their
effective surface area. For the MgH2 powders milled without
salt, their conversion yield depends upon their specific surface
area (see trend dotted line in Fig. 10). The conversion yields of
the milled MgH2–MgCl2 composites are all positioned above
this trend line. This confirms that the accentuation of the hydrol-
ysis reactivity of the MgH2–MgCl2 materials is not associated
with an increase of their specific surface area but is rather related
to the MgCl2 dissolution upon hydrolysis that favours the exten-
sive reaction between MgH2 and water.
Mg and MgH2 powders were ball milled with different solid
salts (KCl, NaCl, LiCl, MgCl2). Depending of the synthesis
parameters (milling duration, salt amount and composition), a
great improvement of the Mg and MgH2 hydrolysis reactivity
was observed. This was explained by: (i) an increase of the effec-
tive surface area of the Mg powder upon milling (not valid for
MgH2 powder), (ii) the salt leaching from the composite upon
hydrolysis creating fresh Mg and MgH2 surfaces and (iii) the
driving force related to the exothermic dissolution of the salt
additivesuchasMgCl2. Thebestperformancewasobtainedwith
the MgH2–3 mol% MgCl2 composite milled for 0.5 h, namely
964 ml of hydrogen produced per gram of composite that corre-
sponds to a hydrogen yield of 8.6 wt% when the water weight is
not included in the calculation.
Acknowledgment
This work has been financially supported by the Natural Sci-
ences and Engineering Research Council (NSERC) of Canada.
References
[1] V.C.Y. Kong, F.R. Foulkes, D.W. Kirk, J.T. Hinatsu, Int. J. Hydrogen
Energy 24 (1999) 665.
[2] V.C.Y. Kong, F.R. Foulkes, D.W. Kirk, J.T. Hinatsu, Int. J. Hydrogen
Energy 28 (2003) 205.
[3] C.A. Ward, D. Stanga, L. Pataki, R.D. Venter, J. Power Sources 41
(1993) 335.
[4] J.-P. Tessier, P. Palau, J. Huot, R. Schulz, D. Guay, J. Alloys Compd.
376 (2004) 180.
[5] M.-H. Grosjean, M. Zidoune, L. Roue´, J.-Y. Huot, Int. J. Hydrogen
Energy 31 (2006) 109–119.
[6] Y. Kojima, K.I. Suzuki, Y. Kawai, J. Mater. Sci. 39 (2004) 2227.
[7] J.E. Stearns, M.A. Matthews, D.L. Reger, J.E. Collins, Int. J. Hydrogen
Energy 23 (1998) 469.
Fig. 9. Conversion yield in water of MgH2–x mol% MgCl2 composites (x = 0,
1, 3 and 10) as a function of the milling time.
[8] R. Aiello, J.H. Sharp, M.A. Matthews, Int. J. Hydrogen Energy 24
(1999) 1123.

302
M.-H. Grosjean, L. Roue´ / Journal of Alloys and Compounds 416 (2006) 296–302
[9] S.C. Amendola, S.L. Sharp-Golman, M. Daleem Janjua, M.T. Kelly, P.J.
[16] C. Wu, H. Zhang, B. Yi, Catal. Today 93–95 (2004) 477.
[17] J.-H. Kim, K.-T. Kim, Y.-M. Kang, H.-S. Kim, M.-S. Song, Y.-J. Lee,
P.S. Lee, J.-Y. Lee, J. Alloys Compd. 379 (2004) 222.
[18] Y. Kojima, Y. Kawai, M. Kimbara, H. Nakanishi, S. Matsumoto, Int. J.
Hydrogen Energy 29 (2004) 1213.
Petillo, M. Binder, J. Power Sources 85 (2000) 186.
[10] S. Suda, Y.-M. Sun, B.-H. Liu, Y. Zhou, S. Morimitsu, K. Arai, N.
Tsukamoto, M. Uchida, Y. Candra, Z.-P. Li, Appl. Phys. A 72 (2001)
209.
[11] Y. Kojima, K.-I. Suzuki, K. Kukumoto, M. Sasaki, T. Yamamoto, Y.
Kawai, H. Hayashi, Int. J. Hydrogen Energy 27 (2002) 1029.
[12] R.D. Hua, Y. Hanxi, A. Xinping, C. Chuansin, Int. J. Hydrogen Energy
28 (2003) 1095.
[19] E. Ivanov, I. Konstanchuk, B. Bokhonov, V. Boldyrev, J. Alloys Compd.
359 (2003) 320.
[20] M.-H. Grosjean, M. Zidoune, L. Roue´, J. Huot, R. Schulz, Electrochim.
Acta 49 (2004) 2461.
[13] Y. Kojima, Y. Kawai, H. Nakanishi, S. Matsumoto, J. Power Sources
135 (2004) 36.
[21] S. Hwang, C. Nishimura, P.G. McCormick, Mater. Sci. Eng. A 318
(2001) 22.
[14] Y. Kojima, K.-I. Suzuki, K. Kukumoto, Y. Kawai, M. Kimbara, H.
Nakanishi, S. Matsumoto, J. Power Sources 125 (2004) 22.
[15] J.-H. Kim, H. Lee, S.-C. Han, H.-S. Kim, M.-S. Song, J.-Y. Lee, Int. J.
Hydrogen Energy 29 (2004) 263.
[22] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 73rd ed.,
CRC Press, London, 1992–1993.