Hydrogen absorption study of high-energy reactive ball milled Mg composites with palladium additives

Article abstract of DOI:10.1016/j.jallcom.2013.01.086

Hydrogenation behaviour, structure, morphology and dehydrogenation/re- hydrogenation performances of Mg-Pd nanocomposites prepared by high-energy reactive ball milling in H₂ (HRBM) of Mg in the presence of amorphous and crystalline Pd black (0.

Full text of DOI:10.1016/j.jallcom.2013.01.086

Journal of Alloys and Compounds 580 (2013) S144–S148
Contents lists available at SciVerse ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jalcom
Hydrogen absorption study of high-energy reactive ball milled
Mg composites with palladium additives
⇑
M. Williams, J.M. Sibanyoni, M. Lototskyy , B.G. Pollet
South African Institute for Advanced Materials Chemistry, Faculty of Natural Sciences, University of the Western Cape, Private Bag X17, Bellville 7535, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 30 January 2013
Hydrogenation behaviour, structure, morphology and dehydrogenation/re-hydrogenation performances
of Mg–Pd nanocomposites prepared by high-energy reactive ball milling in H2 (HRBM) of Mg in the pres-
ence of amorphous and crystalline Pd black (0.1–5 wt.%) were studied. Improvements of hydrogenation
kinetics during HRBM were observed only for the materials prepared using crystalline Pd black. The
obtained nanocomposites were characterised by modest improvements in their dehydrogenation and
re-hydrogenation performances associated with the formation of Mg–Pd intermetallides.
Ó 2013 Elsevier B.V. All rights reserved.
Keywords:
Composite materials
Magnesium hydride
Mechanochemical processing
Kinetics
X-ray diffraction
2. Experimental
1. Introduction
Mg powder (ꢀ20 . . . +100 mesh, 99.8%; metal basis) was purchased from
Over the last two decades, numerous studies were undertaken in
an effort to enhance Mg hydrogenation rates [1]. One of the most
promising approaches was found to be mechanical milling (MM).
The products of MM often exhibit unusual physical and chemical
properties and enhanced reactivity, in particular with respect
to hydrogen. The enhancements are especially pronounced for
Mg-based materials where nanostructuring and surface modifica-
tion result in dramatic improvements in the hydrogenation
kinetics [2]. Additional improvement of the hydrogenation perfor-
mances of Mg was observed upon introduction of catalytic additives,
including transition metals, alloys and intermetallic compounds
[3–9]. High energy reactive ball milling (HRBM) of Mg with catalytic
additives in H2 has been proven to be the most efficient way to
further improve the re-hydrogenation process [6,7].
The catalytic effect of the metallic additives, most probably,
relates to facilitation of the reactions of hydrogen transfer, includ-
ing dissociative chemisorption and associative desorption of H2
molecules. One of the most efficient catalysts for these processes
is palladium. Indeed, Pd-coated magnesium thin films were shown
to be characterised by significantly improved kinetics of hydroge-
nation of Mg and dehydrogenation of the formed MgH2 [10–12].
At the same time, HRBM of Mg with Pd additive was shown to
inhibit both hydrogenation and re-hydrogenation of Mg [13].
In this work we report about investigation of HRBM of Mg in the
presence of a Pd black catalyst.
Alfa-Aesar.
Pd black was prepared as follows. PdCl2ꢁxH2O (1.0 g) and concentrated HCl
(1 mL) were dissolved in 100 mL of the de-ionised water. The solution was heated
to ꢂ70 °C under vigorous stirring, followed by drop-wise addition of 3–4 mL of
N2H4ꢁxH2O, or 10 g/L aqueous solution of NaH2PO2, which resulted in the formation
of black precipitates. The reduction of Pd black was driven to completion with an
excess of N2H4, or NaH2PO2. The deposit was collected by gravity filtration, washed
with copious quantities of water, and dried at 120 °C for 12 h in air. XRD studies
confirmed that Pd black obtained using N2H4ꢁxH2O as a reducing agent was crystal-
line, and that prepared by reduction with NaH2PO2 was amorphous, in accordance
with literature data and our earlier observations [14].
The Mg–xPd mixtures (x = 0.1,0.5,1,2 and 5 wt.%) containing crystalline
(reduced with N2H4) and amorphous (reduced with NaH2PO2) Pd black, were ball
milled in H2 using a Retsch PM 100 ball mill and 220 mL hardened steel vial with
pressure–temperature monitoring system (Evico Magnetics GmbH). The milling
was performed with a ball-to-powder-ratio of 40:1 (82 steel balls, 10 mm in diam-
eter) at 500 rpm, for 6 h in total. Prior to the milling, the vial was evacuated fol-
lowed by filling with H2 gas (P ꢂ 30 bar) supplied from a metal hydride hydrogen
storage and supply unit on the basis of AB5-type alloy. The hydrogen pressure in
the vial was kept above 20 bar by refilling each time (P ꢂ 30 bar) when the pressure
dropped below this value. When the vial temperature approached 70 °C, the milling
was stopped and resumed again after cooling the vial to room temperature. The
amount of hydrogen absorbed in the sample was calculated starting from actual
pressure–temperature values (with the correction to hydrogen compression factor)
and plotted versus milling time.
A High Resolution Transmission Electron Microscope (HRTEM), FEI Tecnai 30,
operating at 120 kV was used to analyze the morphology of the as-prepared and
re-hydrogenated samples. The sample holder was cooled to ꢂ90 K.
XRD studies were performed using Bruker AXS D8 Advance diffractometer with
Cu Ka radiation (k1 = 1.5406 Å, k2 = 1.5444 Å, k2/k1 = 0.5). The Bragg angle range was
2h = 20–90°, and scan rate was 1.2°/min with a step size of 0.02°. The collected XRD
data were refined by Rietveld whole-profile refinement using GSAS software [15],
where the peak shapes were described using the Thompson–Cox–Hastings pseu-
⇑
do-Voight-type function (CW profile function 2). A standard
a-Al2O3 sample was
Corresponding author. Tel.: +27 21 959 9314; fax: +27 21 959 9312.
E-mail address: mlototskyy@uwc.ac.za (M. Lototskyy).
used for the determination of the instrumental contribution into peak profile
0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jallcom.2013.01.086

M. Williams et al. / Journal of Alloys and Compounds 580 (2013) S144–S148
S145
Table 1
Characteristics of HRBM Mg and Mg–Pd composites prepared using crystalline Pd black.
Composite
Maximum H capacity (wt.% H)
Time of 90% hydrogenation (min)
TDS peakb (°C)
HRBM
Re-hydrogenationa
HRBM
Re-hydrogenationa
Mg
7.69
6.42
7.10
7.65
7.59
3.41
–
5.13
4.88
5.56
221
460
113
218
162
110
–
26
38
52
375
–
350
341
345
Mg–0.1Pd
Mg–0.5Pd
Mg–2Pd
Mg–5Pd
a
15 bar H2/230 °C, after TDS of the re-hydrogenated material.
For the re-hydrogenated sample.
b
8
7
6
5
4
3
2
1
0
MgH2
8
MgH2
A
B
Mg-5Pd
Mg
7
6
5
4
3
2
1
0
Mg-1Pd
Mg-0.1Pd
Mg-1Pd
Mg-0.5Pd
Mg
Mg-0.1Pd
Mg-0.5Pd
0
1
2
3
4
5
0
1
2
3
4
5
Milling time [hours]
Milling time [hours]
Fig. 1. H absorption during HRBM of Mg with amorphous (A) and crystalline (B) Pd black.
parameters. During refinement, Gaussian profile parameters (GU, GV and GW) were
fixed (kept the same as for the Al2O3 standard), and only two Lorentzian profile
parameters were refined, LX (size broadening) and LY (strain broadening).1 The
refinements yielded lattice periods of the constituent phases, as well as estimations
of their abundances and crystalline sizes. The latter were calculated using the formula
[15]:
threshold does not result in a noticeable increase of hydrogenation
and re-hydrogenation rates, as well as in a noticeable decrease of
TDS peak temperature.
As illustrated by TEM images (Fig. 2), the as-milled powders con-
sist of agglomerates of polycrystalline MgH2 particles, 100–200 nm
in size. Twinning, whose appearance is typical for ball-milled MgH2
[16], was also observed.
18; 000K k
DV
¼
ð1Þ
p
X
Bright- (Fig. 2A and B) and dark-field (Fig. 2C and D) TEM micro-
graphs indicate the presence, in addition to MgH2, of less electron-
transparent phase(s) containing noticeable amounts of Pd. The
average Pd content in the sample determined by EDS was of
3.43 wt.% corresponding well to the starting sample composition.
Since neither Pd, nor PdH could be detected through XRD refine-
ment (see below), the appearance of the less electron-transparent
areas could be accredited to new phase(s) containing both Mg and
Pd (i.e. Mg–Pd intermetallides).
It is seen from the low magnification bright-field image (Fig. 2A)
that dark inclusions presenting Pd-rich particles are uniformly dis-
tributed in the MgH2 matrix. The inclusions vary in size: few rela-
tively big (50–100 nm; delimited by circle) and majority very small
(<10 nm; shown by arrows). In the dark field images (Fig. 2C and D)
it is visible that a separate phase exists in the Mg matrix, as
evidenced by the areas of bright contrast. This new phase exhibits
where DV is a volume-weighted crystallite size, K is the Scherrer constant (assumed
to be equal to 1), X is a Lorentzian profile parameter LX.
Dehydrogenation and re-hydrogenation performances of the materials after
HRBM were studied using a Sieverts-type volumetric setup. 200 mg of the sample
powder was loaded into the reactor which, together with the measurement system,
was further evacuated to <10ꢀ4 mbar. Thermal desorption measurements (TDS)
were carried out by heating the reactor at a heating rate of 5 °C/min from 25 to
460–470 °C under dynamic vacuum conditions; the vacuum sensor was calibrated
on the flow rate of H2 supplied into the measurement system. Further re-hydroge-
nation was carried out at ꢂ15 bar H2 and 230 °C for ꢂ4 h followed by cooling down
to room temperature. Two TDS – re-hydrogenation cycles were performed for each
sample.
3. Results and discussion
Table 1 summarises the data on hydrogenation (during HRBM),
re-hydrogenation and dehydrogenation performances of the stud-
ied composites. The selected data on H absorption during HRBM
are presented in Fig. 1.
For Mg–xPd minor improvements of the performances were ob-
served only when using crystalline Pd black introduced in the
amount x P 0:5 wt.%. The increase of the Pd content above this
a
basal plane alignment which has long-range arrangement
(Fig. 2D; the corresponding region is delimited by a rectangle).
Fig. 3 shows XRD patterns of the as-milled (A) and re-hydroge-
nated (B) Mg–5Pd (crystalline). Results of refinement of the typical
XRD patterns are summarised in Table 2.
The patterns of the as-milled samples (see Fig. 3A as an
1
example) exhibit broad peaks of tetragonal
a- and orthorhombic
In the ball milled samples, the LX and LY for
constrained to be equal.
a- and c-modifications of MgH2 were

S146
M. Williams et al. / Journal of Alloys and Compounds 580 (2013) S144–S148
Fig. 2. Bright (A and B) and dark (C and D) field TEM micrographs of as-milled sample Mg–5 wt.% of crystalline Pd black. Circle delimits a typical large particle of Pd-
containing phase, arrows show examples of small Pd-containing particles, and rectangle delimits region of Pd-containing phase exhibiting long-range basal plane alignment.
c
-modifications of MgH2 with lattice periods corresponding well to
better refinement was obtained assuming additional presence of
poorly-crystallized MgPd (Fig. 3A, Table 2), but in lesser amounts
than in the as-milled sample. The fitted lattice period of MgPd cor-
responded well to the literature data [19]. Finally, the XRD patterns
of the re-hydrogenated samples exhibited the presence of MgO
formed due to oxidation of the sample.
the literature data [17,18], and crystallite size of 10–12 nm, with
the trend of a slight increase with increasing Pd content in the
composite. Palladium and palladium hydride were not observed
in the patterns. However, for the samples containing P2 wt.% of
Pd black, the better refinement was obtained assuming presence
of a poorly-crystallized MgPd intermetallide (CsCl-type; space
group # 221; a = 3.12 Å [19]), crystallite size below 10 nm. For
Mg–5 wt.% Pd, the refinement was further improved assuming
Our observations allow us to conclude that mechanical alloying
of magnesium with palladium takes place already during HRBM.
Most probably, this process starts from the formation of Mg0.9Pd1.1
,
the presence of Mg0.9Pd1.1 (AuCu-type; space group
#
123;
the most stable intermetallide in Mg–Pd system [22], on the sur-
face of Pd particles. During the milling followed by dehydrogena-
tion/re-hydrogenation, due to diffusion of Mg, this phase is
enriched with Mg to form MgPd and, further, Mg6Pd. Similar re-
sults were obtained by Roquefere et al. [23] who prepared Mg6Pd
by MM of stoichiometric mixture of Mg and Pd, and Pasquini
et al. [24] who observed the formation of MgPd, Mg5Pd2 and Mg3Pd
in the course of PVD of Pd onto Mg nanoparticles.
In summary, the Mg–xPd nanocomposites are characterised by
modest improvements of their dehydrogenation and re-hydroge-
nation behaviour as compared to HRBM Mg (Table 1). The reason
for that seems to be in the high ‘‘metallurgical affinity’’ between
Mg and Pd forming a wide spectrum of intermetallic compounds
a = 3.03 Å, c = 3.42 Å [20]) with quite large crystallites (75 nm).
The estimated abundances and crystallite sizes of Pd-containing
phases in Mg–5 wt.% Pd (Table 2) are in good correspondence with
TEM observations (Fig. 2A).
The re-hydrogenated samples (see example in Fig. 3B)
contained major phase of
a-MgH2; abundance >75 wt.%, and
crystallite size about 90 nm. A significant amount (10–17 wt.%) of
non-hydrogenated Mg, crystallite size above 100 nm, was observed
as well. The patterns of the re-hydrogenated samples containing
P2 wt.% Pd exhibit well-resolved peaks belonging to the impurity
of Mg6Pd (space group # 216; a = 20.108 Å [21]). The most inten-
sive peaks of Mg6Pd are marked by arrows in Fig. 3B. Furthermore,

M. Williams et al. / Journal of Alloys and Compounds 580 (2013) S144–S148
S147
Table 2
Characteristics of constituent phases in Mg–Pd composites prepared using crystalline Pd black.
Sample characteristics
Mg–1Pd HRBM
Constituent phases
-MgH2
a
c
-MgH2
Abundance (wt.%)
87.3(–)
12.7(4)
Unit cell parameters (Å)
a = 4.517(1)
c = 3.016(1)
a = 4.479(5)
b = 5.544(5)
c = 4.931(4)
Crystallite size, DV (nm)
10
Mg–2Pd HRBM
a-MgH2
c
-MgH2
MgPd
Abundance (wt.%)
89.0(–)
10.4(3)
0.6(1)
Unit cell parameters (Å)
a = 4.5197(9)
c = 3.0186(7)
a = 4.483(5)
b = 5.542(7)
c = 4.941(4)
a = 3.000(3)
Crystallite size, DV (nm)
11
8
Mg–5Pd HRBM
a-MgH2
c
-MgH2
MgPd
Mg0.9Pd1.1
Abundance (wt.%)
78.5(–)
14.5(3)
5.0(1)
2.0(1)
Unit cell parameters (Å)
a = 4.5188(6)
c = 3.0235(6)
a = 4.527(5)
b = 5.435(5)
c = 4.957(4)
a = 3.023(3)
a = 3.107(4)
c = 3.284(7)
Crystallite size, DV (nm)
12
7
74
Mg–2Pd re-hydrogenated
a
-MgH2
Mg
Mg6Pd
1.4(1)
MgPd
MgO
Abundance (wt.%)
82.0(–)
10.6(3)
0.3(1)
5.7(3)
Unit cell parameters (Å)
a = 4.5130(3)
c = 3.0182(2)
89
a = 3.2090(4)
c = 5.2098(8)
117
a = 20.198(5)
a = 3.151(7)
a = 4.228(2)
Crystallite size, DV (nm)
123
10
16
Mg–5Pd re-hydrogenated
a-MgH2
Mg
Mg6Pd
MgPd
MgO
Abundance (wt.%)
76.1(–)
16.7(5)
4.5(2)
0.6(1)
2.1(3)
Unit cell parameters (Å)
a = 4.5121(3)
c = 3.0177(3)
90
a = 3.2077(5)
c = 5.213(1)
118
a = 20.202(4)
a = 3.146(2)
a = 4.214(2)
Crystallite size, DV (nm)
138
29
33
B
A
I(obs)
I(obs)
I(calc)
I(calc)
(obs-calc)
(obs-calc)
α-MgH2
Mg6Pd
α-MgH2
γ-MgH2
Mg
Pd Mg0.9
PdMg
PdMg
1.1
MgO
20
30
40
50 60
2θ [o]
70
80
90
20
30
40
50 60
2θ [o]
70
80
90
Fig. 3. XRD patterns of as-milled (A) and re-hydrogenated (B) samples Mg–5 wt.% of crystalline Pd black. The most intensive well-resolved peaks of Mg6Pd (B) are labelled by
arrows.
[22]. Hydrogenation of Mg–Pd intermetallides is quite difficult: so
far it was reported about formation of -solid solution Pd3MgH0.7
4. Conclusions
a
[25] and ternary hydride PdMg2H [26]; reaction of Mg6Pd with
H2 yielding Mg3.65Pd and MgH2 proceeds very slowly [23]. We sup-
pose that the formation of Mg–Pd intermetallides observed in the
course of the XRD studies results in the lowering of the catalytic
activity of initial Pd particles thus exhibiting the minor (if any)
effect of improvement of hydrogen sorption performances of the
Mg-based nanocomposites.
Improvements of hydrogenation kinetics during HRBM of Mg–
Pd, as compared to HRBM Mg, were observed only for the materials
prepared using crystalline Pd black (P0.5 wt.%). The obtained
nanocomposites were characterised by modest improvements of
their dehydrogenation and re-hydrogenation performances associ-
ated with the formation of Mg–Pd intermetallides that results in
the lowering of the catalytic activity of initial Pd particles.

S148
M. Williams et al. / Journal of Alloys and Compounds 580 (2013) S144–S148
[12] J. Qu, B. Sun, J. Zheng, R. Yang, Y. Wang, X. Li, J. Power Sources 195 (2010)
Acknowledgements
1190–1194.
[13] A.Ye. Yermakov, N.V. Mushnikov, M.A. Uimin, V.S. Gaviko, A.P. Tankeev, A.V.
Skripov, A.V. Soloninin, A.L. Buzlukov, J. Alloys Comp. 425 (2006) 367–372.
[14] M.V. Lototsky, M. Williams, V.A. Yartys, Ye.V. Klochko, V.M. Linkov, J. Alloys
Comp. 509 (2011) S555–S561.
[15] A.C. Larson, R.B. Von Dreele, General Structure Analysis System (GSAS), Los
Alamos National Laboratory Report LAUR, 86–748, 2004.
[16] M. Danaie, D. Mitlin, J. Alloys Comp. 476 (2009) 590–598.
[17] J.M. Haschke, M.R. Clark, High Temp. Sci. 7 (1975) 152–158.
[18] M. Bortz, B. Bertheville, G. Böttger, K. Yvon, J. Alloys Comp. 287 (1999) L4–L6.
[19] H.H. Stadelmaier, W.K. Hardy, Z. Metallk. 52 (1961) 391–396.
[20] P.I. Kripyakevich, E.I. Gladyshevskii, Sov. Phys. Crystallogr. 5 (1960) 552–554.
translated from Kristallografiya.
This development was funded by Department of Science and
Technology of South Africa via HySA Program, Project KP3-S04.
ML also thanks National Research Foundation of South Africa
(NRF) for the support via Incentive Funding Grant.
References
[1] I.P. Jain, C. Lal, A. Jain, Int. J. Hydrogen Energy 35 (2010) 5133–5144.
[2] A. Zaluska, L. Zaluski, J.O. Ström-Olsen, J. Alloys Comp. 288 (1999) 217–225.
[3] G. Liang, J. Huot, S. Boily, A. Van Neste, R. Schulz, J. Alloys Comp. 292 (1999)
247–252.
[4] J. Charbonnier, P. de Rando, D. Fruchart, S. Miraglia, L. Pontonnier, S. Rivoirard,
N. Skryabina, P. Vulliet, J. Alloys Comp. 383 (2004) 205–208.
[5] M.Y. Song, S.N. Kwon, H.R. Park, J.L. Bobet, Int. J. Hydrogen Energy 36 (2011)
12932–12938.
[21] S. Samson, Acta Crystallogr., B 28 (1972) 936–945.
[22] A.A. Nayeb Hashemi, J.B. Clark, Mg–Pd (Magnesium–Palladium), in: T.B.
Massalski (Ed.), Binary Alloy Phase Diagrams, vol. 3, second ed., Springer,
Berlin, 1990, pp. 2535–2536.
[23] J.-G. Roquefere, J. Lang, A. Yonkeu, J. Dufour, J. Huot, Int. J. Hydrogen Energy 36
(2011) 2165–2169.
[24] L. Pasquini, E. Callini, M. Brighi, F. Boscherini, A. Montone, T.R. Jensen, C.
Maurizio, M.V. Antisari, E. Bonetti, J. Nanopart. Res. 13 (2011) 5727–5737.
[25] H. Kohlmann, G. Renaudin, K. Yvon, C. Wannek, B. Harbrecht, J. Solid State
Chem. 178 (2005) 1292–1300.
[26] Y. Goto, H. Kakuta, A. Kamegawa, H. Takamura, M. Okada, J. Alloys Comp. 404–
406 (2005) 448–452.
[6] J.L. Bobet, B. Chevalier, M.Y. Song, B. Darriet, J. Etourneau, J. Alloys Comp. 336
(2002) 292–296.
[7] M.V. Lototsky, R.V. Denys, V.A. Yartys, Int. J. Energy Res. 33 (2009) 1114–1125.
[8] Y.B. Pan, Y.F. Wu, Q. Li, Int. J. Hydrogen Energy 36 (2011) 12892–12901.
[9] N. Mahmoudi, A. Kaflou, A. Simchi, J. Power Sources 196 (2011) 4604–4608.
[10] A. Krozer, B. Kasemo, J. Phys.: Condens. Matter l (1989) 1533–1538.
[11] J. Dufour, J. Huot, J. Alloys Comp. 439 (2007) L5–L7.