Formation of one-dimensional MgH2 nano-structures by hydrogen induced disproportionation

Article abstract of DOI:10.1016/j.jallcom.2006.02.024

Remarkable formation of one-dimensional single crystalline MgH₂ structures in the nano- and micro-meters ranges is reported. These structures have been tailored by hydrogen absorption and subsequent disproportionation of bulk Mg₂₄Y₅. The MgH₂ whiskers have been structurally and morphologically characterized by X-rays diffraction, scanning and transmission electron microcopies. A growth model is proposed for the early stage of the whiskers formation by combining surface chemical and morphological investigations. The formation of MgH₂ whiskers opens new engineering explorations and challenges for further experimental and theoretical studies.

Full text of DOI:10.1016/j.jallcom.2006.02.024

Journal of Alloys and Compounds 426 (2006) 357–362
Formation of one-dimensional MgH2 nano-structures
by hydrogen induced disproportionation
Claudia Zloteaa,∗, Jun Lub, Yvonne Anderssona
a Department of Materials Chemistry, Uppsala University, Box 538, Uppsala 751 21, Sweden
b Microstructure Laboratory, Uppsala University, Box 534, Uppsala 751 21, Sweden
Received 25 January 2006; received in revised form 16 February 2006; accepted 16 February 2006
Available online 29 March 2006
Abstract
Remarkable formation of one-dimensional single crystalline MgH2 structures in the nano- and micro-meters ranges is reported. These structures
have been tailored by hydrogen absorption and subsequent disproportionation of bulk Mg24Y5. The MgH2 whiskers have been structurally and
morphologically characterized by X-rays diffraction, scanning and transmission electron microcopies. A growth model is proposed for the early
stage of the whiskers formation by combining surface chemical and morphological investigations. The formation of MgH2 whiskers opens new
engineering explorations and challenges for further experimental and theoretical studies.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Magnesium hydride; One-dimensional nano-structures; Hydrogen absorbing materials
1. Introduction
and is considered to originate from the very slow hydrogen dif-
fusion through the ␤-MgH2 layer [2]. Alloying Mg is a possible
solution to overcome the inherent limitations of pure Mg [1,4].
For example, the kinetics of the hydrogen absorption in Mg2Ni
is found to be significantly faster than in pure Mg [5]. Mg2Ni
absorbs hydrogen but the uptake capacity is decreased as com-
pared to Mg [6]. So far, no valid solution for hydrogen storage
material for sustainable devices has been proposed. To alloy
Mg with Y might be a plausible route, since our recent study
on Mg1−xYx (x = 0–0.17) thin films has demonstrated that the
presence of yttrium improves the diffusion of hydrogen [7]. The
present study reports on the possibility of tailoring new one-
dimensional Mg-based structures in the nano- and micro-meters
ranges by hydrogen absorption and disproportionation of bulk
Mg24Y5. The formation of MgH2 whiskers by hydrogen absorp-
tion and subsequent disproportionation of bulk alloys has never
been reported so far. The proposed method has the unique advan-
tage of designing nano- and micro-structures directly from the
bulk material without contamination introduced by mechanical
alloying synthesis techniques. Comparable formation of metal-
lic Mg whiskers has been reported earlier [8].
The shape, size, surface composition and crystal structure of
materials are major factors that control the hydrogen absorption
properties for energy storage applications. To act as an efficient
energy carrier, hydrogen should be absorbed and desorbed in
materials easily and in high quantities. Magnesium is one of the
most promising candidates for future hydrogen storage materials
due to the formation of MgH2, where the hydrogen content cor-
responds to 7.6 wt.%. However, pure Mg presents major draw-
backs for use in robust, sustainable and long life storage device.
The Mg hydride formation is very slow and has poor thermody-
namic properties [1,2]. In order to overcome these drawbacks,
two promising ways have been extensively investigated. One
route is to reduce the Mg particle size to micro- or nano-ranges
and the other solution is to tailor Mg-rich alloys and compounds.
Several studies have been reported on nano-structured Mg mate-
rials by mechanical synthesis [3]. This approach addresses the
kinetic problem of hydrogen diffusion. The uptake tends to stop
once the hydride layer formed at the surface exceeds 30–50 ␮m
∗
2. Experimental methods
Corresponding author. Present address: European Commission, Directorate-
General, Joint Research Center, Institute for Energy, P.O. Box 2, NL-1755 ZG
Petten, The Netherlands.
Mg24Y5 bulk compounds were prepared from appropriate amount of the ele-
ments: Y 99.9% purity (Highways International) and Mg 99.95% purity (Alfa
E-mail address: claudia.zlotea@gmail.com (C. Zlotea).
0925-8388/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2006.02.024

358
C. Zlotea et al. / Journal of Alloys and Compounds 426 (2006) 357–362
Table 1
Aesar). The elements were placed in a Ta tube, which is sealed by welding in an
Ar atmosphere and subsequently heated up to 1073 K in a high frequency induc-
tion furnace under 300 mbar of Ar. In order to avoid long time air exposure, the
samples were manipulated and stored inside a glove box under Ar atmosphere.
Prior to any characterization, the powder samples were ground and filtered to
particle sizes less then 60 ␮m.
Hydrogen absorptions were performed through the gas–solid interaction
using an autoclave system at different hydrogen pressures and temperature pro-
files. Prior to hydrogen absorption, the system was evacuated to a base pressure
of 10−2 mbar and flushed several times with hydrogen gas. The hydrides were
prepared by repeating several heat-treatments in hydrogen atmosphere in order
to crack the oxide layer and to ensure a direct access of hydrogen to the pure
compound. The temperature profile was 1 h heating to 643 K, 5–20 h dwelling
and 8 h cooling to 300 K, 2–5 times at the applied hydrogen pressure.
The structural properties were determined by X-rays diffraction (XRD) mea-
surements using a Guinier-Ha¨gg camera with Cu K␣1 radiation and Si as internal
standard. The experimental crystallographic constants were refined using the
CELREF program based on the least-squares method. Mg24Y5 crystallizes in
Phases formed by hydrogen disproportionation of Mg24Y5 at different applied
pressures (bar) at 643 K
P(H2) (bar)
Disproportionation phases
1.5–5
10
20–30
YH2 + Mg
YH2 + Mg + MgH2 + YH3
YH3 + MgH2
The results are obtained for 3 heat-treatment cycles in hydrogen atmosphere
with a temperature profile of 1 h heating, 5–20 h dwelling and 8 h cooling to
300 K.
from 300 to 643 K several times. This heat-treatment in hydro-
gen atmosphere causes disproportionations of the compound for
pressures between 1.5 and 30 bar. The obtained phases at differ-
ent pressures, as determined by XRD, are shown in Table 1 and
Fig. 1.
˚
the cubic ␣-Mn type structure with the refined lattice parameter a = 11.257(2) A,
which is in good agreement with earlier results [9–11]. In the XRD pattern of
Mg24Y5, extra diffraction peaks with very weak intensities were observed and
were identified as Mg2Y.
At hydrogen pressures below 5 bar, Mg24Y5 undergoes a
complete disproportionation with the formation of YH2 and
metallic Mg. At hydrogen pressures above 10 bar, Mg24Y5
decomposes into YH2, YH3, MgH2 and metallic Mg. The reac-
tion rate was slow and thermodynamic equilibrium conditions
were not obtained despite several heat-treatment cycles. At the
hydrogen pressure of 20 bar, a complete reaction of Mg and Y
with hydrogen occurs. Thermodynamic equilibrium is reached
and only YH3 and MgH2 are seen in the XRD patterns. Y hydride
might have a catalytic effect in the formation of MgH2, as pre-
viously suggested for rare-earth hydrides in the Ce–Mg–H [16]
and La–Mg–H [17] systems. Although, the role of the Y hydride
in the formation of MgH2 remains an open question and is still to
address. The diffraction peaks of the Y hydrides are very broad
indicating either small particles sizes or strain in the crystallites.
These results allow us to propose a two-step mechanism for the
hydrogen absorption of Mg24Y5. During the first step, YH2 and
pure metallic Mg are formed. The second step promotes the for-
mation of YH3 and MgH2. The disproportionation mechanism
can be understood in terms of the thermodynamics of elementary
hydride formation. The heats of formation ꢀHf of YH2, YH3
and MgH2 are −114, −89 and −37 kJ/mol H, respectively [18].
The formation of YH2 is the most favorable, which explains
Surface analyses of the initial compound were performed at room temper-
ature by X-rays photoelectron spectroscopy (XPS) measurements using a Phi
Quantum 2000 instrument with a monochromatic Al K␣ source (1486.6 eV).
In order to determine the surface and bulk chemical modulation with depth,
selected areas were bombarded by Ar+ and sequentially analyzed. The total
sputtering time for depth profiling was approximately 7 min. The investigated
electronic levels were Mg 2p, Y 3d and O 1s. Both metallic and oxidized Mg and
Y atoms are present at the surface. The oxidized Mg and Y peaks completely
vanish after approximately 1 min of sputtering. The binding energy of Mg 2p
peak (49.8 eV) after 1 min of sputtering is in agreement with the value of pure
Mg or other Mg-rich alloys [1]. The binding energy of the metallic Y 3d5/2
peak (156.4 eV) is shifted compared to the value of pure Y (155.8 eV) [12].
The chemical shift is understood in terms of changes in the atomic environment
of Y occurring upon the compound formation. A surface segregation with Mg
enrichment is evidenced since the Mg/Y ratio is approximately twice as large in
the oxidized layer of the surface (after 40 s of Ar+ sputtering) compared to the
bulk (after 7 min of Ar+ sputtering), which is in agreement with earlier studies
[13]. Previously, it was established that the driving force towards surface seg-
regation is the difference in surface energy of the alloying elements in vacuum
conditions [14,15]. Within this model, Mg atoms will preferentially segregate
to the surface, even before air contact.
Microstructural analyses were performed by scanning electron microscopy
(SEM), energy dispersive spectroscopy (EDS) and transmission electron
microscopy (TEM) techniques before and after heat-treatments in hydrogen
atmosphere. The SEM measurements were carried out using a high resolution
LEO 1550 microscope equipped with an in-lens detector. The TEM specimen
was characterized by a Jeol2000FXII microscope operating at 200 kV. The EDS
analysis was performed by an INCA instrument. The preparations of the spec-
imens for SEM and TEM investigations were made by dispersing the powder
materials on a carbon tape and a copper grid with a thin carbon film, respectively.
Prior to the SEM and EDS measurements, the specimens were exposed to air.
3. Results and discussion
Mg24Y5, with a small homogeneity range on the Mg-rich
side, crystallizes in a related cubic ␣-Mn type structure, space
group I-43m [9–11]. According to the XPS measurements, the
surface of the initial sample is enriched with Mg and partially
covered with both Mg and Y oxides. The Mg24Y5 sample was
exposed to different hydrogen pressures in the temperature range
550–650 K. Mg24Y5 starts to react with hydrogen at 573 K, but
to crack the oxide layer and to ensure a direct access of hydro-
gen to the pure compound, the temperature has to be increased
Fig. 1. XRD patterns of hydrogenated Mg24Y5 at 643 K and 1.5, 10 and 20 bar
hydrogen pressures.

C. Zlotea et al. / Journal of Alloys and Compounds 426 (2006) 357–362
359
the presence of YH2 already at moderate pressure. At higher
pressures both MgH2 and YH3 are always formed, which is in
agreement with the heats of formation.
Y is chemically related to the rare-earth metals and it is of
interest to compare our results to previous studies of hydrogena-
tion of rare-earth Mg alloys and compounds, with similar hydro-
gen disproportionation reactions [4,16,19]. RMg12, R2Mg17 and
R5Mg41 (R = La, Ce and mischmetal) form RH3 and MgH2 at
hydrogen pressures between 5 and 50 bar in the temperature
range 600–653 K [4]. An in situ X-rays diffraction study of
LaMg12 and La2Mg17 has clearly shown direct decompositions
into LaH3 and MgH2 at 10, 30 and 50 bar for the temperature
range 553–603 K [19]. On the contrary, a calorimetric study
demonstrates a two-step mechanism for hydrogen interactions
in the La–Mg system [20], which is similar to our proposed
model.
Microstructural studies have been performed by high resolu-
tion scanning electron microscopy (SEM) and energy dispersive
spectroscopy (EDS) before and after hydrogen absorption at dif-
ferent pressures. The initial powder with particle sizes less than
60 ␮m showed a common particle morphology and size distri-
bution. The SEM investigation performed after the first step of
the hydrogen induced disproportionation demonstrated no sig-
nificant changes in the morphology of the particles. A more
compact particle size distribution was noticed and might be
understood in terms of agglomeration due to the heat-treatments.
However, a large number of cracks on the surface of the parti-
cles were formed after the first step of disproportionation. A
nest-like microstructure with particles in the range of hundreds
of nanometers can be seen between the cracks. This might be
attributed to the disproportionation and formation of YH2 and
pure Mg.
Significant morphology changes were proved to be induced
by the second step of the hydrogen induced disproportionation.
Remarkable microstructures are shown in Fig. 2. A large number
of one-dimensional structures with thicknesses in the range of
nanometers (further called nano-whiskers), occasionally up to
micrometers, were observed on the surface of the host particles
of the parent compound.
The crystal symmetry and the approximate lattice parameters
of the nano-whiskers were obtained from transmission electron
microscopy measurements by selected area electron diffraction
(SAED) (see Fig. 3). SAED patterns from two whiskers were
acquired, which are illustrated in the insets of the Fig. 3.
The nano-whiskers are single crystals with tetragonal sym-
˚
metry and the lattice parameters a = b = 4.55(5) A and c = 3.00(5)
˚
A. This result is consistent with the crystal structure of MgH2,
Fig. 2. SEM images obtained after the hydrogen induced disproportionation of
bulk Mg24Y5 at 20 bar and 643 K: one-dimensional nano-structures grown on
the host particles of the parent compound (a), net of one-dimensional nano-
structures (b) and polyhedral grains on the surface of the host material (c).
which crystallizes in space group P42/mnm with the general
reflection conditions 0 k l: k + l = 2n, 0 0 l: l = 2n and h 0 0: h = 2n
[21]. It should be noted that the extinct reflection {1 0 0} is
weakly present in one of the SAED pattern. The thickness and
width of the nano-whisker is about 120 nm, which is too thick
for the kinematical condition to be valid and thus extinction
reflections are present. No conclusion could be made regarding
the crystallographic growth direction of the nano-whiskers. The
identification of the nano-whiskers as MgH2 was also supported
by the EDS analysis, which showed that Mg is the main con-
stituent element. Small amount of oxygen was detected in con-
trast with hydrogen, which cannot be observed by this method.
A quantitative result cannot be proposed because of the smaller
size of the nano-whiskers as compared to the interaction volume
of the electron beam with the probed material.

360
C. Zlotea et al. / Journal of Alloys and Compounds 426 (2006) 357–362
Fig. 3. TEM images of nano-whiskers obtained after hydrogen induced dispro-
portionation of Mg24Y5 at 20 bar hydrogen pressure and 643 K. The insets show
the selected area electron diffraction patterns.
Fig. 4. SEM image of the surface of the host particle showing a microstruc-
ture of small grains (bright color) with thickness in the nanometer range. The
agglomerates of polyhedral grains (dark color) correspond to the MgH2 phase.
The MgH2 nano-whiskers form complex three-dimensional
nets interconnecting the particles of the initial compound (see
Fig. 2a and b). A large number of nano-whiskers have straight
and uniform widths along the growth direction. Kinked whiskers
with complex shapes growing in irregular branches were also
observed. Moreover, the nano-whiskers were interacting and
deforming each other during the growth. The most extraordi-
nary feature of these one-dimensional MgH2 structures is the
length, which occasionally exceeds 150 ␮m. As a general ten-
dency, the density and numbers of nano-whiskers are increasing
with the applied hydrogen pressure. They are very brittle and
a great number is broken during the SEM or TEM specimen
preparations and measurements.
Together with these unexpected one-dimensional structures,
agglomerates of micro-particles were observed along cracks
and on the surface of the particles (see Fig. 2c). These micro-
particles have polyhedral regular shapes and sizes in the nano
and micrometer ranges. A large number of MgH2 nano-whiskers
are originating from the polyhedral particles. One micro-particle
can grow several nano-whiskers with different thicknesses and
directions. We conclude that the polyhedral micro-particles are
MgH2 with different morphologies. This is also supported by
EDSanalysis, whichrevealedaMg-richcompositionandasmall
oxygen content (less then 3 at.%).
The microstructure of the parent particles revealed also inter-
esting features. The surface of the particles contains a large
number of cracks and flakes of the material. The fine struc-
ture of the parent particles was found to be very porous and
formed by a large number of nano-particles (see Fig. 4). The
average chemical composition of the porous particles, as deter-
mined by EDS, revealed a richer Y concentration as compared
to the starting material. This result can be understood by long-
range diffusion of Mg atoms towards the surface and formation
of different MgH2 morphologies.
The formation of nano-whiskers is most favorable after at
least 3 heat-treatment cycles in hydrogen atmosphere at 643 K
and pressures above 10 bar. Samples exposed to hydrogen pres-
sure of 12 bar for 1 or 2 heat-treatment cycles in hydrogen
atmosphere show a complete decomposition of the initial com-
pound into YH2, metallic Mg and small traces of both YH3 and
MgH2. SEM investigations of these samples demonstrate that
the early stage of the one-dimensional MgH2 growth consists
of very few whiskers without any polyhedral micro-particles.
The microstructure revealed that several whiskers were growing
compactly and the thickness of the single structures was between
100 and 200 nm (see Fig. 5a). The first whiskers extrude from the
host particle in straight or kinked shapes by bursting the surface
layer (see Fig. 5b). Further heat-treatment cycles in hydrogen
atmosphere favor the formation of MgH2 in both complex nets
of nano-whisker and polyhedral micro-particles.
Several questions about the origin and growth mecha-
nism of the MgH2 nano-whiskers have been aroused. Several
types of mechanism are currently accepted to well explain
the growth of one-dimensional nano-structures [22–26]. The
vapor–liquid–solid (VLS) process is the most used mechanism
to understand the growth of semiconductor nano-whiskers [26].
A catalyst metal seed, which is located at the end of the whisker,
always assists this process and its size influences the diameter of
the whisker. Within this model, metallic Mg whiskers have been
reported to grow from hydrogenated Mg powder [8]. The for-
mation of Mg whiskers was assured by the presence of sublimed
Mg vapor and hydrogen as carrier gas. Catalyst Mg-Al eutectic
seeds were located at the tips of the whiskers. However, in our
case, SEM or TEM investigations cannot detect any metal seeds
at the ends of the MgH2 whiskers. The vapor–solid (VS) process
is widely used to describe the growth of nano-belts and nano-
rods of oxides [23,24]. Within the VS mechanism, the oxide
vapor directly deposits on a substrate at lower temperatures and

C. Zlotea et al. / Journal of Alloys and Compounds 426 (2006) 357–362
361
must be thin enough to crack and to allow the extrusion of the
compressed metal [27,28]. Other authors proposed recently an
oxidation-based process, which assumes that oxygen diffuses
and forms oxide at interfaces providing the driving force for
whisker extrusion [29]. However, the whiskering phenomenon
is not completely understood.
Based on the growth mechanism of compressive stress and
subsequently relief of the stress, we propose a three-step model
for the early stage formation of the MgH2 nano-whiskers. At this
stage, theyaredistinctandagrowthmodelcanthenbesuggested.
During the first and second steps, the nucleation and growth of
MgH2 take place underneath the oxidized surface layer. Due to
the hydrogen absorption an important expansion of the volume
occurs (33% of volume expansion from Mg to MgH2). During
the third step, a compressive stress is induced and subsequently
relieved by fracturing the oxidized surface and extrusion of the
whiskers. For further heat-treatment cycles in hydrogen atmo-
sphere, the growth mechanism of MgH2 in polyhedral particles
and intricate nets of whiskers is far from clear. Furthermore, the
role of the Y hydrides in the formation of the whiskers is not yet
elucidated. Improvements for understanding the formation of
different morphologies of MgH2 and the role of the Y hydrides
in the growth of the whiskers are currently under investigation.
It is worth mentioning that previous hydrogenation studies on
other rare-earth Mg materials did not show any whisker forma-
tion. Few studies have been reported regarding the microstruc-
ture of hydrogen decomposition phases of Mg-rich materials
[16,17]. Theses reports have investigated the morphology of
CeMg12 [16] and LaMg12 [17] after decomposition and showed
a refinement of the grains together with an increased roughness
of the surface.
Fig. 5. SEM images of the early stage of nano-whiskers growth induced by
hydrogen disproprotionation of bulk Mg24Y5 at 12 bar, 643 K and 1–2 heat-
treatment cycles in hydrogen atmosphere. A compact structure of several nano-
whiskers with thickness between 100 and 200 nm can be observed (a). A whisker
extrusion and the zoom of the cracked surface around the whisker base (b).
4. Conclusions
We have found unique formation of a large number of single
crystalline MgH2 nano-whiskers by hydrogen induced dispro-
portionation of bulk Mg24Y5. Occasionally, the whiskers have
lengths and thicknesses in the micrometer range. The result
opens new roads of designing Mg-based materials for different
technological applications. The desorption of hydrogen and/or
separation of these whiskers have not been investigated yet
but might appear to be a promising approach for tailoring Mg
assemblies with high active surface for hydrogen storage. Fur-
thermore, it offers the unique possibility to study the physical
properties of large MgH2 single crystals. These features have
never been investigated in bulk materials due to embrittlement
by hydrogen absorption. Our results are a challenge for further
experimental and theoretical work in order to understand the
growth mechanism of MgH2 whiskers and to study their physical
properties.
grows into different morphologies. MgO has been reported to
grow into nano-rods or nano-belts structures by heating a halide
source in a flow of argon–oxygen gas mixture [24,25]. The VLS
and VS growth models are based on the fact that the reactants
are supplied in the vapor phase. In our case, the growth process
of the MgH2 whiskers is far from clear. Either a vapor phase
is formed during the hydrogen absorption or a long-range solid
diffusion mechanism has to be taken into account. The starting
hypothesis of gaseous reactants is believed to be hardly fulfilled
in our hydrogenation conditions. The working temperature is
643 K and well below the melting point of metallic Mg (923 K).
No evidence for loss of material or evaporation on the walls of
the crucibles and autoclave were noticed, regardless the tem-
peratures and pressures. Therefore, the VLS and VS models
whiskers growth seem to be unlikely.
Many models are currently proposed in order to explain the
spontaneous growth of the low melting point metal whiskers at
room temperature, which pose serious problems in the semi-
conducting industry [27–29]. The compressive stress is mostly
accepted but insufficient condition for the growth of Sn, Cd,
Zn, Al whiskers. It has been proposed that an oxidized surface
Acknowledgments
This work was financially supported by the Swedish Energy
Administration(STEM)andtheNordicEnergyResearchProject
(NERP). The authors would like to thank to Ola Wilhelmsson
for the help with the XPS measurements.

362
C. Zlotea et al. / Journal of Alloys and Compounds 426 (2006) 357–362
[15] A.R. Miedema, Z. Metallkde. 69 (1978) 455.
References
[16] J.M. Boulet, N. Gerard, J. Less Common Met. 89 (1983) 151.
[17] M. Stoui, A. Grayevski, A. Resnik, D. Shaltiel, N. Kaplan, J. Less Common
Met. 123 (1986) 9.
[18] The NBS tables of chemical thermodynamic properties, J. Phys. Chem.
Ref. Data 11 (Suppl. 2) (1982).
[1] P. Selvam, B. Viswanathan, C.S. Swamy, V. Srinivasan, Int. J. Hydrogen
Energy 11 (3) (1986) 169.
[2] A.S. Pedersen. Hydrogen Metal System I, Solid State Phenomena 35, 1996,
pp. 49–50.
[3] A. Zaluska, L. Zaluska, J.O. Stro¨m-Olsen, Appl. Phys. A 72 (2001) 157.
[4] M. Pezat, B. Darriet, P. Hagenmuller, J. Less-Common Met. 74 (1980) 427.
[5] M.Y. Song, M. Pezat, B. Darriet, P. Hagenmuller, J. Less-Common Met.
103 (1984) 145.
[6] J.J. Reilly, R.H. Wiswall, Inorg. Chem. 7 (1968) 2254.
[7] C. Chacon, E. Johansson, C. Zlotea, Y. Andersson, B. Hjo¨rvarsson, J. Appl.
Phys. 97 (2005) 104903.
[19] D. Sun, F. Gingl, Y. Nakamura, H. Enoki, M. Bououdina, E. Akiba, J.
Alloys Compd. 333 (2002) 103.
[20] B. Bernhardt, K. Bohmhammel, Thermochim. Acta 382 (2002) 249.
[21] F.H. Ellinger, C.E. Holley Jr., B.B. McInteer, D. Pavone, R.M. Potter, E.
Staritzky, W.H. Zachariasen, J. Am. Chem. Soc. 77 (1955) 2647.
[22] Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim, H.
Yan, Adv. Mater. 15 (2003) 353.
[23] Z.W. Pan, Z.R. Dai, Z.L. Wang, Science 291 (2001) 1947.
[24] J. Zhang, L. Zhang, X. Peng, X. Wang, Appl. Phys. A 73 (2001) 773.
[25] Z. Cui, G.W. Meng, W.D. Huang, G.Z. Wang, L.D. Zhang, Mat. Res. Bull.
35 (2000) 1653.
[26] K.A. Dick, K. Deppert, M.W. Larsson, T. Ma˚rtensson, W. Seifert, L.R.
Wallenberg, L. Samuelson, Nat. Mater. 3 (2004) 380.
[27] E. Iwamura, K. Takagi, T. Ohnishi, Thin Solid Films 349 (1999) 191.
[28] K. Zeng, K.N. Tu, Mater. Sci. Eng. R 38 (2002) 55.
[29] M.W. Barsoum, E.N. Hoffman, R.D. Doherty, S. Gupta, A. Zavaliangos,
Phys. Rev. Lett. 93 (2004) 206104.
[8] P.J. Herley, W. Jones, B. Vigeholm, J. Appl. Phys. 58 (1985) 292.
[9] AMS International, Binary Alloys Phase Diagram, Second Edition plus
updates, CD Version, 1996.
[10] J.F. Smith, D.M. Bailey, D.B. Novotny, J.E. Davison, ActaMetall. 13(1965)
889.
[11] F. Bonhomme, K. Yvon, J. Alloys Compd. 232 (1996) 271.
[12] A. Fujimori, L. Schlapbach, J. Phys. C: Solid State Phys. 17 (1984) 341.
[13] P. Selvam, B. Viswanathan, C.S. Swamy, V. Srinivasan, J. Less Common
Met. 163 (1990) 89.
[14] A.R. Miedema, Z. Metallkde. 69 (1978) 287.