Magnesium-based nanocomposites chemical hydrides

Article abstract of DOI:10.1016/S0925-8388(02)01306-3

The hydrolysis reaction of nanostructured MgH₂ and nanocomposites MgH₂-X (X = Ca, Li, LiAlH₄) prepared by ball-milling was studied as a function of milling time and component proportion. It was found that nanocrystallinity greatly enhances the hydrolysis kinetics. Moreover, in this new class of chemical hydrides, the reaction also proceeds to full completion, contrary to some conventional chemical hydrides where the reaction stops before total completion due to the formation of passivation layers. The effect of addition of acidic solutions was also investigated.

Full text of DOI:10.1016/S0925-8388(02)01306-3

Journal of Alloys and Compounds 353 (2003) L12–L15
L
www.elsevier.com/locate/jallcom
Letter
Magnesium-based nanocomposites chemical hydrides
*
J. Huot , G. Liang, R. Schulz
HERA, Hydrogen Storage Systems Inc., 577 Le Breton, Longueuil, Quebec, Canada J4G 1R9
Received 25 October 2002; accepted 11 November 2002
Abstract
The hydrolysis reaction of nanostructured MgH2 and nanocomposites MgH –X (X5Ca, Li, LiAlH ) prepared by ball-milling was
2
4
studied as a function of milling time and component proportion. It was found that nanocrystallinity greatly enhances the hydrolysis
kinetics. Moreover, in this new class of chemical hydrides, the reaction also proceeds to full completion, contrary to some conventional
chemical hydrides where the reaction stops before total completion due to the formation of passivation layers. The effect of addition of
acidic solutions was also investigated.

2002 Elsevier Science B .V . All rights reserved.
Keywords: Hydrogen absorbing materials; Nanostructured materials; Mechanical alloying; Gas–solid reactions
1
. Introduction
found that steam hydrolysis produces only hydrogen and
meet the purity requirement for fuel cells applications.
However, the price and availability of this type of chemical
hydrides could be problematic for future applications.
Therefore, there is a need for an inexpensive chemical
hydride with good hydrogen release properties, low-cost
and environmentally friendly.
Magnesium hydride could be considered as a candidate
for the hydrolysis reaction because it can be produced at a
relatively low cost. However, the hydrolysis reaction of
conventional Mg hydride is incomplete and very slow
because of the formation of a magnesium hydroxide layer.
The hydrolysis of magnesium hydride is expressed by the
following reaction.
An attractive source of hydrogen for hand-portable
Proton Exchange Membrane (PEM) fuel cells is the
aqueous hydrolysis of chemical hydrides. However, one
major problem associated with this method of hydrogen
production is the control of the hydrolysis reaction. The
reaction is either too fast or even explosive for some
chemical hydrides such as LiAlH4 or it is too slow for
practical applications in some other cases. The non-re-
versibility of the reaction can make this technology quite
expensive and therefore, efforts should be made on reduc-
ing the overall cost of ‘chemical hydride-storage tank’
systems. For example, it has been known for a long time
that hydrolysis of sodium borohydride (NaBH ) could be
used for generating hydrogen [1]. Recent developments of
4
MgH 1 2H O ⇒ Mg(OH) 1 2H
2
(1)
2
2
2
base-stabilised aqueous solutions of NaBH catalyzed by
4
ruthenium have shown some promise but problems with
material cost, shelf life and recycling of products are not
totally solved [2].
In a recent series of publications, Matthews et al. have
investigated the production of hydrogen from complex
hydrides [3,4] and from hydrolysis with steam [5]. They
The hydrogen yield of this reaction is 6.5 wt.% when the
weight of water is taken into account. In fuel cell applica-
tions, if the water produced by the fuel cell is redirected to
the chemical hydride, then the yield increases to 15.4
wt.%. The yield of this reaction compares well with the
hydrolysis of sodium borohydride which produces 7.3
wt.% if the weight of water is calculated and 21.2 wt.% if
the water weight is not taken into account. Moreover, the
reacted product (magnesium hydroxide) is environmentally
benign.
*
Corresponding author.
E-mail address: jh@herahydrogen.com (J. Huot).
In this paper, we present an investigation of the hy-
0
925-8388/02/$ – see front matter
 2002 Elsevier Science B .V . All rights reserved.
doi:10.1016/S0925-8388(02)01306-3

J. Huot et al. / Journal of Alloys and Compounds 353 (2003) L12–L15
L13
drolysis of nanocomposite magnesium hydrides. In this
new class of chemical hydrides, reaction kinetics are
improved and often proceed to full completion, contrary to
conventional chemical hydrides where reactions usually
stop before total completion due to the formation of
passivation layers. Tests on additions of acidic solutions
were also performed.
2
. Experimental details
The nanocomposites were synthesized by first mixing
the raw materials inside an argon filled glove box. The
mixtures were then milled for up to 20 h in a Spex 8000
model shaker mill using a vial and balls of stainless steel
with a ball to powder weight ratio of 10:1. Magnesium
Fig. 1. Hydrogen release in hydrolysis of polycrystalline MgH2 and
MgH2 milled for 20 h. The reacted fraction (F) is the ratio of the volume
of released hydrogen over the theoretical volume of hydrogen that should
be released assuming that all material is hydrolyzed.
hydride from Th. Goldschmidt (95 wt.% MgH , 5 wt.%
2
Mg), lithium (Alfa, 99%), calcium (Alfa, 98.8%), and
LiAlH powder (Aldrich, 95%) were used.
4
The hydrolysis reactions were carried out in an Erlen-
meyer flask provided with two openings, one for water
addition and the other for hydrogen exhaust. Between 80
and 100 mg of nanocomposite was loaded into the flask
and allowed to react with 10 ml of deionized water.
Hydrogen release was quantified by water displacement in
an inverted graduated cylinder positioned over a water-
filled tank. Measurements with acidic solutions were
performed by first injecting 10 ml of deionized water and
afterward, by periodic additions of 5 ml of a diluted HCl
solution. The diluted HCl was prepared by mixing 1 part
HCl in 100 parts of deionized water. In all measurements,
the reaction mixture was continuously agitated with a
magnetic stirrer.
hydride. Therefore, the high rate of reaction during the first
10 h of hydrolysis could be explained by the high specific
surface area and the nanostructure. Fig. 1 shows that after
10 h of reaction the hydrolysis rate for the nanostructured
magnesium hydride is similar or lower to that of the
polycrystalline case, indicating the formation of a passiva-
tion layer.
To achieve a faster and more complete reaction, we
added a second element which can be leached away during
the electrolysis, thus exposing fresh magnesium hydride
surface that react readily with water. It is known that ionic
metal hydrides react readily with water and produce
soluble hydroxides. Therefore, a nanocomposite of mag-
nesium hydride with an ionic metal hydride leads to
synergetic effects between the two components and faster
and more complete hydrolysis reactions. Selected ionic
hydrides were tested and the results of the hydrolysis
experiments are presented in Fig. 2. It can be seen that for
the lithium addition, the initial reaction rate is faster than
in the case of pure magnesium hydride. However, the
reaction quickly slows down to a smaller rate than ball-
milled magnesium hydride afterwards. In the case of the
LiAlH4 addition, the general behavior of the reaction is
3
. Results and discussion
Fig. 1 shows the hydrogen release during hydrolysis of
polycrystalline and nanocrystalline magnesium hydride.
For ease of comparison, the reacted fraction as a function
of time is plotted where the reacted fraction F is defined as
the ratio of the volume of hydrogen released over the
theoretical volume of hydrogen that could be released
assuming that all material is hydrolyzed. Hydrogen dis-
charge of polycrystalline MgH2 is seen to be linear and
quite slow. After 20 h, only 54% of the theoretical
hydrogen capacity has been released. This is in striking
contrast with the 20 h ball-milled magnesium hydride that
has released 74% of its theoretical hydrogen capacity after
similar to that of milled MgH . The most interesting case
2
is the nanocomposite MgH –Ca 5 at.% where the initial
2
reaction is fast and gives a higher yield than the other
reactions. After the initial period, the system decomposes
at a rate similar to that of the nanocrystalline MgH . The
2
2
0 h of hydrolysis. Moreover, the hydrogen release of the
results shown in Fig. 2 indicate that MgH –Ca is the best
2
nanocrystalline sample is much faster at the onset of the
reaction. As reported previously [6], high energy milling of
magnesium hydride for a period of 20 h produces a
nanocrystalline structure with crystallite size of 11.960.1
nanocomposite tested in the present study.
The MgH –Ca system was further investigated to see
2
the effect of milling time and calcium proportion on the
hydrolysis reaction. The results are presented in Fig. 3. The
addition of 5 at.% of calcium to magnesium hydride and
milling for 1 h produced a powder that reacted very
2
21
nm and with a specific surface area of 9.9 m
g
2
21
compared to 1.2 m
g
for the unmilled magnesium

L14
J. Huot et al. / Journal of Alloys and Compounds 353 (2003) L12–L15
Fig. 2. Hydrogen release in hydrolysis of MgH2 milled for 20 h, MgH2 –
Fig. 4. Hydrogen release in reaction of polycrystalline and nanocrystal-
line MgH2 under the injection of 5 ml of diluted HCl solution. The gray
bars indicate the reacted fraction if all HCl contained in the 5 ml injection
reacts to produce hydrogen.
Ca 5 at.% milled for 1 h, MgH –Li 5 at.% milled for 1 h, and
2
MgH –LiAlH 5 at.% milled for 10 h. For definition of F, see Fig. 1.
2
4
quickly upon addition of water. On the other hand, after
less than 1 h, the reaction rate was almost identical to that
of pure milled magnesium hydride. This is in striking
creation of complex hydride. Further experiments are
needed in order to understand fully the hydrolysis reaction
in this kind of nanocomposite. Fig. 3 also shows the effect
of the calcium proportion. As expected, a higher propor-
tion of calcium resulted in a faster hydrolysis reaction and
completion is reached after 4 h.
contrast to the MgH –Ca 5 at.% composite milled for 10
2
h. In this case, the hydrolysis reaction could be divided in
three distinct regimes. First, a fast initial reaction which is
almost instantaneous and is of the same magnitude (reacted
fraction50.1) as that of the 20 h milled MgH . This
2
In acid solution, the following reaction takes place.
proportion is smaller than in the system MgH –Ca 5 at.%
2
milled for 1 h. Afterward, the 10 h milled system enters
into a second regime that last up to 4 h and where the
reaction is much faster than that of the powder milled for 1
h. After 4 h of hydrolysis, the reaction rate is similar to
those of the milled MgH and the 1 h milled MgH –Ca 5
at.%. At this moment, we do not have a definitive
explanation for the observed improvement of hydrolysis
with milling time. Possible explanations are: structural
effects due to a better distribution of calcium in the
magnesium hydride matrix; creation of calcium hydride;
MgH 1 2HCl ⇒ MgCl 1 2H
2
(2)
2
2
One of the advantages of this reaction over pure hydrolysis
is that the end product is soluble thus making the recovery
and treatment of the end product much simpler. The
reaction of nanocrystalline and polycrystalline magnesium
hydride to acidic solutions was tested and the results are
presented in Fig. 4. In the case of the polycrystalline
material, it is clear that only reaction (2) is taking place
upon each addition of HCl solution. Moreover, the reaction
is complete and very fast. Between HCl additions, no
significant hydrolysis reaction (reaction (1)) is taking
place. This is in striking difference with the nanocrystal-
line case where for the first 15 min both reactions (1) and
2
2
(
2) are taking place. Close inspection of the curves with
and without acidic injection shows that there is no
synergetic effect and that both reactions happen indepen-
dently.
4
. Conclusion
The hydrolysis reaction of MgH2 and MgH –Ca
2
nanocomposites was studied. It was found that
nanocrystalline magnesium hydride shows enhanced hy-
drolysis reaction compared to the conventional material.
Further improvement of reaction kinetics was achieved by
Fig. 3. Hydrogen release in hydrolysis of MgH2 milled for 20 h, MgH2 –
Ca 5 at.% milled for 1 h, MgH –Ca 5 at.% milled for 10 h, and
2
MgH –Ca 20 at.% milled for 10 h. For definition of F, see Fig. 1.
the nanocomposite MgH –Ca 5 at.%. Investigations
2
2

J. Huot et al. / Journal of Alloys and Compounds 353 (2003) L12–L15
L15
[
2] S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, N.C. Spencer,
M.T. Kelly, P.J. Petillo, M. Binder, Int. J. Hydrogen Energy 25
showed that the milling of the nanocomposite should be
performed for a relatively long time in order to get a fine
nanostructure which results in a fast and complete hy-
drolysis. Increasing the calcium content to 20 at.% results
in faster kinetics and a complete hydrolysis reaction was
achieved in only 4 h.
(
2000) 969.
[3] J.E. Stearns, M.A. Matthews, D.L. Reger, J.E. Collins, Int. J.
Hydrogen Energy 23 (1998) 469.
[4] R. Aiello, M.A. Matthews, D.L. Reger, J.E. Collins, Int. J. Hydrogen
Energy 23 (1998) 1103.
[
5] R. Aiello, J.H. Sharp, M.A. Matthews, Int. J. Hydrogen Energy 24
1999) 1123.
6] J. Huot, G. Liang, S. Boily, A .V . Neste, R. Schulz, J. Alloys Comp.
93–295 (1999) 495.
(
[
References
2
[1] H.I. Schlesinger, H.C. Brown, A.E. Finholt, J.C. Gilbreath, H.R.
Hoekstra, E.K. Hyde, J. Am. Chem. Soc. 75 (1953) 215.