Effect of preparation routes on the performance of a multi-component AB2-type hydrogen storage alloy

This article presents experimental results on the preparation and characterisation of a multi-component AB2–type intermetallic hydrogen storage alloy (A = Ti0.85Zr0.15, B = Mn1.22Ni0.22Cr0.2V0.3Fe0.06). The alloy samples were prepared by induction melting using Y2O3-lined alumo-silica and graphite crucibles. The characterisation results were compared with the ones for the reference sample of the same composition prepared by arc melting. It has been shown that the induction-melted samples exhibit reduced hydrogen sorption capacities and sloping plateaux on the pressure composition isotherms (PCI’s). The origin of the observed effects has been shown to be in the inhomogeneity of the induction-melted alloys and their contamination due to crucible—melt interaction, particularly pronounced for the alloy melted in the alumo-silica crucible; this alloy was additionally characterised by the decrease of Zr/Ti ratio and, in turn, higher plateau pressures of the PCI’s.


Introduction
Metal hydrides (MHs) able to reversibly and selectively interact with hydrogen both from a gas and an electrolyte are characterised by unique properties, including extremely high-volume density of H atoms accommodated in the metal matrix, reversibility and fast kinetics of hydrogenation/dehydrogenation, selectivity of hydrogen uptake, significant changes of physical properties of the host metal (alloy) in the course of its hydrogenation, catalytic activity in hydrogen transfer reactions.The MH technologies utilising these features can provide very efficient solutions, particularly in energy storage and conversion technologies [1][2][3][4].
Hydride forming materials represented by a vast number of Laves-type AB 2±x intermetallics can reversibly interact with hydrogen in an extremely wide range of pressures.Their hydrogen sorption properties can be easily tuned in wide limits by a partial substitution of A-or B-component in a binary intermetallic compound or changing the B/A ratio.One example is C15-ZrFe 2 (plateau pressure above 700 bar at the ambient temperature) [5] and C14-ZrFe 2-x V x (x = 0.25-0.5)characterised by the plateau pressure at the same temperature far below the atmospheric one [6,7].Another example is C14-TiMn 2 .This intermetallic has a huge hysteresis in the system with H 2 gas (absorption/desorption plateau pressure at T = 25 • C of 415/34 bar [8]), but the under-stoichiometric C14-TiMn 1.5 absorbs more than 1.7 wt.% H at T = 20 • C and P < 45 bar and exhibits desorption plateau pressure about 7 bar at the same temperature.The substitution of Ti with Zr in the stoichiometric alloy noticeably reduces the hydride stability (plateau pressure below 0.1 bar at T = 20 • C for Ti 0.5 Zr 0.5 Mn 2 ) [9].The variation of composition and stoichiometry of the multi-component Laves-phase intermetallics allows to further extend the range of hydrogen equilibrium pressures for their hydrides covering more than ten orders of magnitude at operating temperatures below 300 • C.These materials have been used in several applications, including hydrogen storage and compression, heat management, getters and low-pressure sources of hydrogen, as well as rechargeable nickel-MH batteries [10].Wide tuneability of thermodynamic properties of hydrogen systems with the multi-component AB 2 -type alloys where A = Ti + Zr and B = Mn + Cr + V + Fe by the variation of the alloy composition makes these materials very promising for compact hydrogen storage at near-ambient conditions and thermally-driven hydrogen compression to several hundred bars in the temperature range 20 • C-150 • C [11,12].
We note that most of the multi-component hydrogen storage alloys belong to high-entropy alloys (HEAs) mainly represented by BCC solid solutions and/or C14-Laves phase intermetallics [13][14][15].Suitable relationships between the hydride-forming (e.g.A = Ti, Zr, Hf; B = V, Nb) and non-hydride-forming (e.g.B = Mn, Cr, Fe, Co, Ni) components in the latter materials, along with the possibility to vary the key HEA's properties including mixing entropy and enthalpy, valence electron concentration, size and electronegativity mismatch, etc., opens wide opportunities for the tuning the alloys' composition towards matching with the requirements of their target hydrogen storage applications.
Preparation routes of the Ti-and Zr-based AB 2 -type alloys include conventional metallurgical (smelting) and powder metallurgical (sintering, mechanical alloying) methods [10,16].Because of high inhomogeneity and uncontrolled contamination which result in poor reproducibility of hydrogen sorption performance of the final product prepared using the latter group of methods, the metallurgical methods based on melting the metallic components and further post-processing of the ingot (annealing, rapid solidification, etc) are preferred [10].These methods mainly include arc-melting and induction melting [17].The use of arc melting (with a non-consumable tungsten electrode and a water-cooled copper hearth) in the atmosphere of a purified inert gas allows to produce the highest purity ingots which are, however, highly inhomogeneous and require multiply re-melting and further annealing.This route is difficult to upscale and, as a rule, it is used in laboratories to prepare reference samples of the alloys.
Induction melting is more suitable for the upscaled manufacturing of the Ti-and Zr-based AB 2 -type alloys.At the same time, due to the high solidification temperatures of these alloys, along with the high reactivity of Ti-and Zr-containing melts, the preparation of large amounts of these materials requires special care.The alloys prepared in laboratory conditions (arc melting and annealing) and manufactured using industrial-scale facilities (induction melting) often show differences in their structure, morphology and hydrogen sorption performances [17].This effect has many origins, including non-uniformity of solidification, the concentration inhomogeneity, as well as the contamination of the ingot with the crucible material, also associated with the shortening of the crucible lifetime due to a crucible-melt interaction which may include capturing of oxygen from commonly used ceramic materials (e.g.Al 2 O 3 or Al 2 O 3 + SiO 2 ), or, when using crucibles made of graphite, formation of carbides or dissolution of carbon in the melt [10,16,18].
Thus, the induction melting of the AB 2 -type alloys (A = Ti + Zr) requires the melting crucible to be carefully selected and designed to avoid melt contamination at high temperatures; the crucible should also have a high thermal shock resistance and relatively low cost [16,19].One of the effective ways to solve this problem is in the use of conventional crucibles (e.g., made of Al 2 O 3 ) coated with a thin layer of yttrium oxide, which is one of the most resistant materials towards interaction with Ti-containing melts [20].Earlier, we applied this route for the induction melting of Ti 0.85 Zr 0.15 Mn 1.2 M 0.8 (M = V + Cr + Ni + Fe) AB 2 -type hydrogen storage alloy and showed that the Y 2 O 3 coating of alumo-silica and graphite crucibles prevents destruction of the crucible during the melting and allows to obtain suitable quality ingots [16].
In this work, we present further details on the preparation and results of characterisation of the multi-component AB 2 -type hydrogen storage alloy to demonstrate the influence of preparation route on performance.

Preparation of the samples
The high-entropy AB 2 type hydrogen storage alloy (A = Ti 0.85 Zr 0.15 , B = Mn 1.22 Ni 0.22 Cr 0.2 V 0.3 Fe 0.06 ; mixing entropy of the components ∆S mix = 12.92 J/(g-at.M K) = 1.55 R, i.e. above the threshold of 1.5 R for the HEAs [13,14]; the mixing enthalpy ∆H mix = −11.41kJ/g-at M) was intended for the use for hydrogen storage on board fuel cell vehicle at near-ambient temperatures [21].The alloy was prepared by induction melting of the metallic components under Ar atmosphere.The starting components were all purchased from South African suppliers with a purity of >98%; these included electrolytic Mn (low oxygen flake), Ti (grade 2, cut-offs), Zr (sponge), Cr (pieces), Ni (powder), and Ferrovanadium (powder; 80 wt.% V). 1 wt.%La was added to the charge to assist with the activation of the alloy according to the recommendation found in [22].
The samples (m = 100 g) were prepared in SM 10 K VHT vacuum induction melting furnace with a crucible capacity of 200 g.Two types of crucibles were used: ceramic alumo-silica (Al 2 O 3 + SiO 2 ) and graphite, ∼150 g in weight each.Each crucible was placed into alumo-silica crucible of the bigger size (see top images in figure 1).
To minimise interaction between the melt and the crucible material, both crucibles were coated inside by the deposition of a protective coating whose surface contacting the melt was yttria (Y 2 O 3 ).In both cases, the total coating thickness was of 0.05 inches (see figure 1, bottom).
The alumo-silica crucible was directly spray-painted with Y 2 O 3 slurry-pure yttria powder suspended in 5% aqueous solution of caboxylmethylcellulose (CMC) stabiliser.The coated crucible was dried in an oven at The coating of the graphite crucible had three layers, which were shown to provide good adhesion of the external Y 2 O 3 refractory layer to the graphite substrate.The inner molybdenum metallic layer was applied as a diffusion barrier to carbon, and the intermediate Mo + Y 2 O 3 layer was used to absorb thermal shocks.Spray painting was used to first coat Mo slurry (powder with particle size below 5 µ suspended in the CMC solution), then Mo + Y 2 O 3 slurry, and finally Y 2 O 3 slurry, with drying in an oven after each stage at T = 80 • C. The CMC was vaporized during pre-heat of the crucible before the melting.
The charge was loaded into the Y 2 O 3 lined alumo-silica-or graphite-based crucible, which was then placed into the vacuum furnace chamber, followed by flushing the chamber three times by evacuation to 10 -3 mbar and Ar refilling (0.3-1 bar).After flushing, the vacuum furnace chamber was filled with Ar (1 bar), and the electric power supplied to the inductor was gradually increased until the complete melting of the charge was visually observed.The charge was maintained in the molten state at T = 1600 • C-1800 • C for 1 min.After solidification in the crucible and cooling to ambient temperature, the ingot could be easily removed [16].
The reference sample of the same composition (m = 10 g) was prepared by arc-melting performed on a water-cooled copper hearth in Ar atmosphere (99.999%; pressure 1 bar).The alloy was turned over and re-melted three times to provide the ingot homogeneity.

Characterisation of the samples
The composition, morphology and the degree of crystallinity of the AB 2 alloy samples were studied by SEM/EDS (Zeiss Auriga scanning electron microscope) and TEM/SAED (JEOL JEM-2100 transmission electron microscope operating at 200 kV).For the SEM/EDS analyses, the sample powders were mounted onto fresh carbon discs.The samples for TEM/SAED studies were prepared by the deposition of the sample powder, using ultrasonic dispersant, onto a substrate (copper mesh covered with electrically conductive amorphous film).
The studies of the elemental composition of the alloys were supplemented by atomic emission spectroscopy (AES) for the detection of impurities using a laser analyser LEA-S500 (SOL Instrument).Improvement of the resolution of the impurity spectral lines overlapping with the ones for the metallic components was achieved by optimising the energy of laser irradiation impulse in the range between 18 and 21.5 J.
The alloys phase composition was determined by XRD (Rigaku Minflex diffractometer) using Cu-K α radiation (λ 1 = 1.5406Å, λ 2 = 1.5444Å, λ 2 /λ 1 = 0.5, 2θ = 15-90 • ).The XRD data was further processed by Rietveld full-profile analysis using GSAS software [23] and the following reference data: A standard α-Al 2 O 3 sample was used to determine the instrumental contribution into peak profile parameters.During refinements, the Gaussian profile parameters (GU, GV and GW) were fixed (kept the same as for the Al 2 O 3 standard), and only two Lorentzian profile parameters were refined, LX (size broadening) and LY (strain broadening).
Further details about the XRD studies are presented in supplementary information, section S1.The pressure-composition isotherms (PCI) of the samples were measured using a commercial automated Sievert-type apparatus (PCT Pro-2000).Approximately 1 g of the sample was loaded into the reactor; the sample was then activated by heating to 300 • C under dynamic vacuum for 1 h.The sample was then allowed to cool down to ambient temperature, followed by charging with hydrogen at the starting H 2 pressure above 30 bar.The hydrogen absorption and desorption isotherms were recorded at the temperatures of 20, 40 and 60 • C and hydrogen pressures from 0.1 to 100 bar.

SEM and EDS
The SEM images of the AB 2 type alloy samples prepared by the different routes (figure 2) show that there are no visual differences between the alloys prepared by arc and induction melting: all the alloys exhibit a smooth, continuous surface.The EDS results (table 1) show that the measured amounts of the metallic components are in satisfactory correspondence with the nominal composition of the alloy.However, for the induction melted samples several deviations were observed.Mainly, it relates to the underestimation of Zr content in the alloy induction-melted in alumo-silica crucible.
No impurities were detected in the arc-melted alloy, while both alloys prepared by induction melting contained oxygen (O) impurities.The presence of oxygen can be attributed to the contamination of the sample with Y 2 O 3 particles from the crucible coating (most probable reason for the sample melted in the graphite crucible where oxygen was detected only in a few separate points), as well as to the interaction of the melt with the residual atmosphere of the induction furnace chamber and with the inner surface of the crucible.The latter also relates to the source of the silicon (Si) impurity, which was found only in the sample prepared in the alumo-silica crucible.
The EDS results allow us to conclude that the applied protecting coating, which eliminates severe contamination of the ingot and destruction of the crucible due to its interaction with the melt [16] does not result in the complete elimination of the latter process accompanied by the leaching of some metallic components and contamination of the ingot with non-metallic impurities.However, this effect is much less pronounced for the graphite crucible.
Due to interference with the background signal from carbon support, the reliable determination of carbon in the studied samples by EDS was not possible.

XRD
The refined XRD patterns of the alloy samples (background subtracted) are presented in figure 3. Results of the Rietveld refinement of the as-measured patterns are shown in supplementary information (figures S1-S3) and summarised in table 2 where calculated abundancies and lattice periods of the constituent phases, along with the unit cell volumes for the major C14-AB 2 phase, are presented.The last column in  ).It has to be noted that the presented values are only rough estimations because, for the major C14-AB 2 phase in the arc melted and induction-melted in graphite crucible samples, the calculated crystallite size exceeds or approaches the value of 100-200 nm above which the precision of the size calculations from XRD data drops dramatically [27,28].Nevertheless, this observation testifies about high degree of crystallinity of the major phase in these samples further confirmed by the TEM/SAED studies (section 3.3) while the impurity phases have a nanocrystalline nature.It is seen from table 2 that the sample prepared by arc melting is single-phase, which is in line with the EDS results (table 1).In La-containing induction-melted samples, La 2 TiO 5 formed during interaction of La deoxidiser with molten Ti and oxygen was detected.In the sample induction-melted in alumo-silica crucible main impurity phase was found to be the (Ti,Zr)O 2 mixed oxide characterised by broad peaks testifying about nanocrystalline nature of this impurity.
The lattice periods and unit cell volumes of the major C14-AB 2 phase in the studied samples are higher as compared to the reference data [25]; the increase of the unit cell volume for the arc melted alloy was of 3.81% at the similar c/a ratio (1.637 against 1.641 for the studied and reference samples, respectively).The increase in the lattice periods is caused by the higher content of zirconium (atomic radius 6 1.60 Å) in the studied alloy as compared to the reference intermetallic mainly containing Ti (atomic radius 1.46 Å).
When comparing the unit cell volume of the major phase in the arc melted and induction melted alloys, it can be seen that for the alloy melted in alumo-silica crucible, it is reduced by 0.91% while increases by 0.71% for the alloy melted in the graphite crucible.The smaller lattice periods and unit cell volume for the sample induction melted in alumo-silica crucible can be explained by the reduced content of Zr in this alloy (see section 3.1).The reason for the increase of the lattice periods of the C14-AB 2 phase in the sample induction melted in the graphite crucible is less clear.Presumably, it could be explained by the presence of interstitial carbon in this sample.

TEM, SAED and AES
The TEM images and selected area electron diffraction (SAED) patterns taken from the areas containing several alloy particles are shown in figure 4. The patterns exhibit the presence of both rings and spots, signifying that the alloys are mainly polycrystalline in nature.However, all the samples exhibit the presence of separate, almost completely monocrystalline particles (figure 5), which for the sample melted in the graphite crucible (E, F) have the highest degree of crystallinity.This sample also does not contain completely polycrystalline separate particles, which were found in other samples.The reason for these observations appears to be due to the slower cooling rate of this sample since both the electrically conductive graphite crucible and the sample were heated during induction melting, therefore the hot mass (100 g of the sample and 150 g of graphite) will cool down slower than in the case of alumo-silica crucible (100 g of sample only) at the similar cooling conditions.Conversely, very fast cooling of the sample during arc melting may result in a slightly poorer crystallinity; see figures 5(A) and (B).For the sample induction melted in alumo-silica crucible and characterised by the poorest crystallinity (see figures 5(C) and (D)), the origins of the observed effect may also be in the presence of noticeable amounts of impurity phases, including nanocrystalline (Ti,Zr)O 2 (see XRD data in table 2) which was not detected in the SAED patterns, as well as introducing silicone into metallic matrix (see table 1) associated with appearance of additional lattice defects.
The SAED patterns for all the samples exhibit the presence of the major phase C14-AB 2 in good correspondence with XRD results (table 2), as illustrated by the example in figures 6(A) and (B) for the sample melted in alumo-silica crucible.
Of the impurity phases, only traces of La 2 TiO 5 poorly crystallised monocrystals were found in the induction-melted samples (figure 6(C)).The low probability of occurrence of the corresponding reflections in SAED patterns taken from an area of ∼1 µ in diameter (figure 6(A)) is caused by the small content of this phase in the sample (⩽2 wt.% according to the XRD data).
Further details about TEM and SAED studies are presented in section S2 of supplementary information.
The AES studies gave a semi-quantitative validation of the EDS (table 1) and XRD (table 2) results, according to which the main impurity in the induction-melted samples was oxygen.It is illustrated by figure 7. The maximum amount of the oxygen impurity was determined for the sample induction melted in alumo-silica crucible, followed by the sample induction melted in graphite crucible and the arc melted sample.Additionally, the sample induction melted in alumo-silica crucible showed the presence of aluminium impurity that, along with the presence of silicon in the ingot, confirms our assumption about incomplete elimination of the crucible-melt interaction by the protective Y 2 O 3 coating, at least, for the alumo-silica crucible (section 3.1).
Due to the overlapping of the most intensive carbon line (247.8561nm) with the lines of Fe, Cr and Ti, it could not be resolved during AES studies.We assume that all three samples do not contain noticeable amounts of carbon, but presence of its traces in the sample melted in graphite crucible is possible.

Hydrogen sorption properties
Figure 8(A) shows the as-measured hydrogen absorption and desorption isotherms (T = 20 • C) for the studied alloy samples.The corresponding van't Hoff plots at H/M = 0.35 are shown in figure 8(B), and the main hydrogen sorption parameters taken from the pressure-composition isotherms are summarised in table 3.
As it can be seen, the induction melted samples exhibit significant reduction in both maximum and reversible hydrogen storage capacity, especially pronounced for the sample induction melted in alumo-silica crucible.This sample (# 2 in figure 8) is also characterised by the lowest thermal stability of hydride, resulting in the higher hydrogen equilibrium pressures.The latter effect can be explained by the lower content of the zirconium in this sample (section 3.1), which also results in the decrease of the lattice periods and the unit cell volume of the C14-AB 2 major phase (section 3.2).The decrease of the maximum hydrogen storage capacity (∼82% as compared to the one for the arc melted alloy; table 3) also correlates well with the minimum abundance of this phase as shown by XRD (∼88%; table 2) where the impurity phases do not absorb hydrogen.Thus, we can conclude that during induction melting in the alumo-silica crucible, even protected by Y 2 O 3 coating, the crucible-melt interaction, along with oxygen contamination, results in the predominant removal from the melt of Zr characterised by higher than Ti affinity to oxygen [16].
The induction melted samples are also characterised by the increased plateau slope, which is especially pronounced for the sample melted in the graphite crucible.This effect can be associated with the inhomogeneity of these samples.An additional reason may be in the incorporation of the traces of non-metallic impurities into interstitials of the C14-AB 2 major phase.Such effects for boron and carbon  were observed for TiFe accompanied by a drastic decrease in hydrogen sorption capacity and increase in the plateau slope [18].
We note that despite significantly more incline plateau for the sample induction melted in the graphite crucible, the thermal stability of its hydride is very close to the one for the arc melted sample (figure 8(B), table 3).All the alloys are characterised by a moderate hydrogen absorption desorption hysteresis, Ln(P A /P D ) = 0.3-0.5, or hysteresis energy loss, RT Ln(P A /P D ) = 0.75-1.25 kJ/mol H 2 .Hysteresis slightly decreases in the series: arc melted → induction melted in alumo-silica crucible → induction melted in graphite crucible.
All the studied samples exhibit excellent hydrogen absorption and desorption kinetics.Typical example for the hydrogen absorption in the arc melted sample is shown in figure 9.As it can be seen from figure 9(A), even the first H 2 absorption after vacuum heating is very fast: the equilibrium corresponding to point a in figure 8(A) is achieved in 1 h while about 90% of hydrogen is absorbed during the first 2 min accompanied by significant self-heating the sample due to exothermic nature of the hydrogenation process.Similar behaviour was observed during the measurements of the pressure-composition isotherms characterised by the low pressure driving force.It is illustrated by figure 9(B) which shows hydrogen absorption kinetics for the same alloy when passing from point b to point c on the absorption isotherm (figure 8(A)).The equilibrium is achieved in less than 10 min, and 95% of hydrogen is absorbed in 2 min.

Conclusions
This work has confirmed the high sensitivity of hydrogen sorption properties of AB 2 -type hydrogen storage alloys (A = Ti + Zr) to their preparation routes, particularly when moving from arc melting to induction melting.The changes observed for the induction-melted alloys include lowering the hydrogen storage capacity and increasing the plateau slope mainly associated with the crucible-melt interaction accompanied by contamination of the ingot.Another reason for the observed effects is the higher inhomogeneity of the induction-melted alloys.
The crucible-melt interaction, which cannot be completely suppressed by the protective 0.05 inches-thick coating of the alumo-silica crucible deposited by spray-painting of suspension of Y 2 O 3 powder, results in the contamination of the ingot with the components of the crucible material (oxygen, silicon, aluminium) and simultaneous leaching of zirconium (having the strongest affinity to oxygen) from the melt.As a result of the latter process, thermal stability of the formed hydride significantly decreases, leading to further decrease of the hydrogen sorption capacity of the alloy at the applied temperatures and hydrogen pressures.
The alloy induction melted in the graphite crucible which was protected by a three-layer coating of the same thickness where the outer layer was Y 2 O 3 exhibited the lesser reduction of the hydrogen sorption capacity, and the hydride stability was very close to the one for the arc melted alloy.However, this alloy was characterised by the highest plateau slope, presumably originating from the incorporation of trace amounts of carbon into interstitials of the matrix of the major C14-AB 2 Laves phase.
This study has shown that selecting crucibles and looking for ways to suppress the crucible-melt interaction is essential in the preparation of MH alloys containing Ti and Zr by induction melting.Further studies should be aimed at the optimisation of the protective coating of graphite crucibles due to the lesser deterioration of the hydrogen sorption properties of the alloy prepared by induction melting in this crucible as compared to the arc melted one.An important stage of these activities would be in further clarifying the mechanisms of the ingot contamination and related deterioration of its hydrogen sorption performance.

Figure 1 .
Figure 1.Images and schematic cross-sections of the Y2O3-coated crucibles.

Figure 2 .
Figure 2. SEM images of the AB2-type alloy prepared by (A)-arc-melting and annealing, (B)-induction melting in alumo-silica crucible, and (C)-induction melting in graphite crucible.Each scale bar (thick yellow line) corresponds to 100 µm.

Figure 3 .
Figure 3. Observed (points) and calculated (red lines) XRD patterns of the AB2-type alloy prepared by (A)-arc-melting, (B)-induction melting in alumo-silica crucible, and (C)-induction melting in graphite crucible.Bottom ticks show the calculated peak positions of the constituent phases.

Figure 4 .
Figure 4. TEM images (A), (C), (E) and the corresponding SAED patterns (B,D,F) for the arc-melted sample (A), (B), the sample induction melted in the alumo-silica crucible (C), (D) and the sample induction melted in the graphite crucible (E), (F).

Figure 6 .
Figure 6.Selected area of the TEM image (A) taken for the sample induction melted in alumo-silica crucible and the corresponding SAED patterns supplemented by Miller indexes of rings and spots for the major C14-AB2 phase (B) and La2TiO5 impurity (C).

Figure 7 .
Figure 7. Fragments of atomic emission spectra for the arc-melted sample (1) and samples induction melted in alumo-silica (2) and graphite (3) crucible.The most intensive oxygen spectral lines are shown by the vertical arrows.

Figure 8 .
Figure 8. Hydrogen absorption and desorption isotherms at T = 20 • C (A) and van't Hoff plots for the AB2-type alloy samples (B): (1)-arc melted, (2)-induction-melted in alumo-silica crucible and (3)-induction-melted in graphite crucible.Points a, b and c on the absorption isotherm for the arc melted sample (1 in A) correspond to hydrogen pressures and concentrations for which kinetic data (figures 9(A) and (B)) are presented.

Figure 9 .
Figure 9. Hydrogen absorption kinetics and changes of the sample temperature at T0 = 21 • C for the arc melted sample: (A)-first absorption at the starting H2 pressure of 33 bar and final H2 pressure of 27 bar (point a in figure 8(A)).(B)-hydrogen absorption during the measurement of the pressure-composition isotherm (points b and c in figure 8(A)).

Table 1 .
Summary of compositional properties of the alloy samples according to EDS data collected from typical points (1-3 in figure2).
table 2 also presents the estimations of the crystallite size of the constituent phases calculated from the LX profile parameter (see section 2.2

Table 2 .
Summary of Rietveld refinement of the XRD patterns of the studied alloy samples (figure3).Errors of the values calculated during refinement (shown in brackets) relate to the last decimal digit.