Introduction

ZrMn2 laves phase alloy attracts much attention due to its higher storage capacity (~1.7 wt%) with rapid H absorption kinetics [1]. However, hydrogen desorption occurs only at elevated temperatures which restricts its application potential for variety of devices such as heat pumps, Ni/MH batteries, stationary hydrogen storage cylinders, etc. Researchers have shown that the desorption characteristics of ZrMn2 system can be improved by doping/alloying with suitable metals [2–5]. The work of Matsumura et al. [6] revealed that the reason for the higher stability of the ZrMn2 hydride is due to the higher bond strength between the stored H with Mn atoms. They have also shown that the bond strengths between Zr–Mn, Mn–Mn and Mn–H all get significantly modified by doping by other suitable metals. Both the Zr and Mn sites can be substituted with hyper/hypo-stoichiometry of the laves phase which often makes the presence of substitute atoms even in excess of 10 per formula unit [7–13]. It may be pointed out that incorporation of several metal atoms results in unwanted side reaction [14]. As for example, dissolution of constituent metal atoms of MH electrode in the electrolyte of an electrochemical device costs its durability.

In an earliest study, Suzuki and Nishimiya [15] have investigated hydrogen storage characteristics of ZrMn1.6Ni0.4, ZrMnNi and ZrMn0.4Ni1.6 systems and shown that the dissociation pressure of the ZrMn2 system gets significantly enhanced with the incorporation of Ni content. They have also concluded that ZrMnNi is the most promising hydride with the hydrogen filling of ~3.5 hydrogen atoms per formula unit under near ambient conditions. Kim et al. [16] have studied several Zr–Mn–Ni–V alloys and observed that the equilibrium plateau pressure gets significantly enhanced with the concentration of Ni content. However, there was significant reduction in the hydrogen capacity and the rate of reaction. The authors attributed the effect to the amount of reactive surfaces, since the presence of more Ni limits the amount of reactive surfaces while offering excellent catalytic activity. We have studied a series of Zr based systems, namely ZrMn2−xNi x , where x = 0 to 2. XRD analysis revealed that the Mn rich compositions possess C14 hcp structure, whereas the Ni rich compositions exhibit C15 cubic structure. Hence, at a particular concentration of Zr–Mn–Ni the C14 structure transforms to C15 structure. We observed that except ZrNi2 all the other composition studied in the present investigation absorbs above 3 H/F.U. under room temperature. The dehydrogenation from ZrMn2 is found to be possible only above 250 °C, but Mn-Ni coexisting system exhibits very interesting desorption characteristics. Particularly, we observed that ZrMn0.5Ni1.5, ZrMn0.75Ni1.25 and ZrMn1.0Ni1.0 are the interesting compositions which can reversibly store ~2.2 H/F.U., ~3 H/F.U. and ~3.2 H/F.U., respectively. It is observed that x = 1.0–1.5 is a narrow range within which room temperature (and near STP) high capacity hydrides of this class can be tailored. The materials exhibit excellent thermodynamics, rapid absorption/desorption kinetics and good activation features.

Experimental techniques

All the required chemicals such as Zr, Mn and Ni were procured from Alfa Aesar Chemicals Ltd. The compositions of Zr/Mn/Ni were calculated as required for the formula ZrMn2−xNi x , (where x = 0–2). The powder mixer taken in right stoichiometry was palletized by 100 KN load (pellet thickness: 5 mm and diameter: 15 mm) and the pellets were melted under inert ambient using a high frequency induction furnace (20 kW, 12 kHz) supplied by Inductotherm India pvt ltd, Ahmedabad. For hydrogen storage measurements, 2 g of samples were loaded in a metal-hydrogen high pressure reaction chamber. PCT absorption/desorption measurements have been performed by a digitalized Sievert’s type apparatus. Heat of the reaction (∆H) was calculated from the van’t Hoff plots and the activation energy measurements were carried out using the standard Arrhenius relation. DSC measurements have been performed by PerkinElmer instrument and XRD analysis of all the samples were performed by Regaku mini Flex11 X-ray diffractometer employing CuKα radiation (λ = 1.541 Å). XRD measurements have been performed within the 2θ range of 30°–80° with a scanning speed of 2 s for the every 0.05° angles (step size). The lattice parameters were calculated using the unit cell refinement technique and the refined lattice constant values were compared with the standard JCPDF data source files. The crystallite size was calculated using Scherrer method and the strain available on the crystals was calculated using the Williamson-Hall (W-H) method. Further details regarding the detailed material synthesis and processing protocols can be seen in our earlier report [17].

Results and discussion

Structural studies

Figure 1 shows the XRDs of the ZrMn2−xNi x system, x correspond to = 0, 0.5, 1.0, 1.25, 1.5 and 2, respectively. The starting phase ZrMn2 corresponds to C14 structure and the end phase ZrNi2 corresponds to C15 structure. The C14 structure retains up to the Mn/Ni concentration of 1.50/0.50 and any further substitution of Mn by Ni leads to its transformation to the C15 phase. Particularly, the Mn/Ni concentration range of 1.00/1.00–0.75/1.25 appears to be quite sensitive for this structural transformation. Extensive discussions are being made by researchers regarding the correlation between the phase abundance of C15 and C14 and the corresponding hydrogen storage/electrochemical performances [18–22]. In the present case, when the C15 phase abundance is dominant, the precipitation of the non-laves phase Zr7Ni10 alloy does also occur. Since the Mn/Ni concentration region of 0.75/1.25–0.50/1.50 brings out the non-laves phase alloy in the system (see Fig. 1), we tried to understand the effect of minor variations (of concentration within the Mn/Ni concentration region of 0.75/1.25–0.50/1.50) in its phase abundance and hydrogen storage performances and the results were communicated elsewhere. As we decrease the amount of Mn in the composition, there is regular peak shifting toward the higher angle side. It shows that the lattice parameter of the system decreases with the increase of Ni content in the system. In a particular concentration (see Fig. 1), the contraction is followed by the precipitation of C15 phase also along with the C14 phase. Upon the further increase of Ni content the C14 phase gets eliminated and only C15 phase exists along with a non-laves phase of Zr7Ni10. The observed crystal structural details are summarized in the Table 1. The decrease in lattice parameters as we increase the Ni concentration can be attributed with the fact that both the covalent radius and atomic radius of Ni are smaller than that of Mn (Atomic radius of Mn and Ni are 1.79 and 1.62 Å, respectively; covalent radii of Mn and Ni are 1.17 and 1.15 Å, respectively).

Fig. 1
figure 1

XRDs of the ZrMn2−xNi x alloys (x = 0–2) prepared in the present study

Full size image
Table 1 The structural and microstructural details of the alloys prepared in the present study
Full size table

The XRDs of hydrogenated ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 are shown in Fig. 2a, b and c. For effective comparison, the corresponding XRDs of the unhydrided samples are also given in the figure. As can be seen, the XRDs of hydrogenated samples of ZrMnNi and ZrMn0.75Ni1.25 exhibit a considerable shifting toward the lower angle side. The shifting arises due to the fact that the insertion of hydrogen in the interstitial sites does lead to expansion of lattice (the upper limit of expansion is known to be one-fourth of the actual lattice parameter) [23]. The peak shifting gives us a clue that hydrides of both the materials could be stable under room temperature. However, our subsequent hydrogen liberation experiments over a range of temperatures proved that hydrogen liberation with slow kinetics is still active at room temperature itself. Thus, the presence of a small peak at the 2θ value around 42.5 (highlighted by a rectangular box) can be attributed to the existence of some portion of the material in dehydrogenated form. The other two interesting information explored from this comparison is that (a) the phase abundance of C14 appears to increase during hydrogenation of both the ZrMnNi, ZrMn0.75Ni1.25 samples. It can be understood from the appearance of some dominant peaks of C14 after hydrogenation, (b) the non-laves phase Zr7Ni10 which presents along with the ZrMn0.75Ni1.25 sample also gets hydrogenated and its usual orthorhombic phase itself undergoes phase transformation [24, 25]. Unlike the above two cases, in the case of ZrMn0.5Ni1.5 (Fig. 2c) the XRD of hydrogenated sample does not indicate any significant shifting from the usual XRD peak positions of its unhydrided sample. It clearly conveys us that there was no much hydrogen in the interstitial sites to cause any considerable lattice shifting. From this insight one may infer that the hydrogenated material is either not stable under room temperature or there is no hydrogenation at all. However, the subsequent experiments proved that in this case the former possibility exists (the low temperature experiments have proven that the equilibrium pressure of 1 atm occurs at temperatures lesser than room temperature).

Fig. 2
figure 2

Comparison of the XRDs of hydrogenated alloys, a ZrMnNi, b ZrMn0.75Ni1.25 and c ZrMn0.5Ni1.5. XRDs of the corresponding as-received materials are also given in the figure

Full size image

One another noticeable behavior is that the crystallite size of these materials got reduced to a greater extent after its interaction with hydrogen (see Table 1). The reduction in the crystallite size in the case of ZrMn0.5Ni1.5 is even closer to 50 %. The crystallite size vs strain diagram for all the compositions of the present study is shown in Fig. 3. As seen in the figure, the system with higher crystallite size exhibits smaller strain and vice versa. It is also seen that the hydrided materials are incorporated with higher strain as compared to the corresponding intermetallic system.

Fig. 3
figure 3

The crystallite size vs strain analysis of all the alloys prepared in the present study. The hydrided materials are also given in the figure

Full size image

Hydrogen storage characteristics

The hydrogen absorption kinetic curves of the as-prepared ZrMn2−xNi x alloys are shown in Fig. 4. Among all these alloys, the ZrMn2 alloy absorbs hydrogen with rapid kinetics when exposed to hydrogen without any activation treatments. A capacity of ~3 H/F.U. could be obtained within few seconds and the maximum capacity reaches 3.67 H/F.U. As we incorporate Ni in place of Mn, the absorption kinetics decreases and only little decrease in storage capacity is witnessed up to the Mn/Ni concentration of 0.75/1.25. But, when we further reduce the Mn concentration, both the absorption kinetics and storage capacity get reduced and only 0.85 H/F.U. can be stored when all the Mn is replaced by Ni. However, during hydrogen desorption, the kinetics of the Ni rich compositions are better than that of Mn rich compositions. The end materials ZrMn2, ZrNi2 and the intermediate composition ZrMn1.5Ni0.5 liberate hydrogen only marginally with very poor desorption kinetics. In fact, just traceable amount of hydrogen can be observed under 100 °C and no significant improvement is observed even at ~200 °C (see Fig. 5). It may be due to the fact that the Mn rich compositions quickly make very stable hydrides and Ni rich composition makes only less stable or unstable hydrides. Further experiments suggested that temperature in excess of 200 °C is required for hydrogen liberation from the Mn rich phases. However, better hydrogen storage characteristics could be witnessed from some intermediate compositions such as ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5. These hydrides absorb and liberate hydrogen under manageable temperatures. Particularly the alloy ZrMn0.5Ni1.5 liberates ~1.9 H/F.U. hydrogen at room temperature. The desorption activation energy values calculated for ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 hydrides are given in Table 2.

Fig. 4
figure 4

The hydrogen absorption kinetics curves for all the materials prepared in the present study. The curves represent the kinetics obtained during the first time hydrogen exposure to the as-prepared samples. Hydrogenation temperature: 30 °C

Full size image
Fig. 5
figure 5

The desorption kinetic curves of the hydrogenated samples. The calculated activation energy values are given in Table 2

Full size image
Table 2 Summary of the observed heat of desorption and activation energy values
Full size table

In the case of ZrMn2−xNi x we noticed that the x values in the range of 1–1.5 lead to promising Zr-based AB2 hydrides suitable for commercial applications. Our thorough investigation revealed that x in the range of 0 to <1 and 1.5 to 2 does not lead to any interesting Zr–Mn–Ni system for reversible hydrogen storage applications. The former case materials interact too strongly with hydrogen and the latter case interact poorly with hydrogen. The PCI (desorption isotherm) curves of three interesting systems within the x = 1–1.5 range, such as ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 are shown in Fig. 6a, b and c, respectively. From the nature of the PCT desorption curves obtained in these compositions it is clearly understood that any little changes made in the Mn concentration within this small x value range lead to big changes in the thermodynamics of the system. Our estimation reveals that ZrMnNi restores ~3.2 H/F.U. and works within the temperature range of 120–180 °C, ZrMn0.75Ni1.25 restores ~3 H/F.U. and works well within the temperature range of 60–140 °C and ZrMn0.5Ni1.5 restores ~2.1 H/F.U. and works well within the temperature range of 10–60 °C. One can understand from the PCT curves that, when we decrease the concentration of Mn = 1 to Mn = 0.5, the stability of the hydride phase gets dramatically reduced and the plateau pressure varies significantly with respect to temperature. The van’t Hoff plots for the said three compositions such as ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 are shown in Fig. 7 and the calculated thermodynamic data are given in Table 2. The equilibrium pressure of 1 atm in these three systems can be obtained at the temperatures of 120, 100 and −20 °C, respectively. The van’t Hoff plots also prove that thermodynamics of the system varies considerably within a narrow range of Mn concentration, i.e. 0.75–1.5. Such behavior can strike considerable applications for heat storage and electrochemical devices. Further, the van’t Hoff diagram clearly indicates that all these three system covers the working range of the commercial hydrogen fuel cells.

Fig. 6
figure 6

a, b, c The PCT desorption isotherms of the alloys. a ZrMnNi, b ZrMn0.75Ni1.25 and c ZrMn0.5Ni1.5

Full size image
Fig. 7
figure 7

The van’t Hoff plots (dehydrogenation) for the compositions such as ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5

Full size image

The PCT (re)absorption isotherm of the composition ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 are given in Fig. 8a. The composition ZrMn0.5Ni1.5 exhibits much higher plateau pressure as compared to ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5. It clearly shows the large differences in their thermodynamic features between the x values of 1.25–1.5 in the ZrMn2−xNi x system. The reversible hydrogen absorption kinetic plots for the systems such as ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 are shown in Fig. 8b. As seen, ZrMnNi system restores hydrogen in excess of 3 H/F.U. and nearly half of this capacity is absorbed within 5 s. ZrMn0.75Ni1.25 also stores almost the similar capacity but ZrMn0.5Ni1.5 stores only 2.2 H/F.U. However, it should also be noted that unlike ZrMnNi and ZrMn0.75Ni1.25, ZrMn0.5Ni1.5 can absorb and release this capacity under room temperature. The DSC curves of all the ZrMn2−xNi x system are shown in Fig. 9. The DSC curves also confirm that the ZrMn0.5Ni1.5 sample behaves much differently as compared to all other compositions. This curious observation in the case of ZrMn0.5Ni1.5 may be related with the heat transfer associated with the evolution of residual hydrogen so as to transform as hydrogen free metal phase. Because, only in the case of ZrMn0.5Ni1.5 does the hydrogen liberation become effective at lower temperatures. Note that ZrNi2 do not possess considerable hydrogen above room temperature and any composition above Ni/Mn concentration of 0.5/1.5 release only very small amount of hydrogen at the temperatures less than 200 °C (also refer desorption PCTs). Thus, the DSC profile of all other samples except ZrMn0.5Ni1.5 looks as similar as each other. The hydrogen absorption/desorption behavior of ZrMn2−xNi x for x = 1–1.5 makes it clear that these materials are suitable for varieity of stationary applications (as for example, large stationary hydrogen storage cylinders in hydrogen production/distribution spots, H isotope stores etc.) and even mobile applications. Particularly, in the case of ZrMn0.75Ni1.25, the reversible hydrogen storage of ~2 H/F.U. within few seconds shows that these are candidates par excellence for device applications.

Fig. 8
figure 8

a The PCT isotherms for reversible hydrogen absorption for ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 and b the reversible hydrogen absorption kinetics of ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 alloys

Full size image
Fig. 9
figure 9

DSC curves of hydrided ZrMn2−xNi x alloys (x = 0–2)

Full size image

In the present Zr–Mn–Ni alloy system, the hydrogen capacity of C14 phase is, in general, higher than the capacity of C15 phase, which is contrary to the observation of some earlier reports [21, 26, 27]. However, from performance point of view, particularly for low temperature desorption, C15 phase co-existing with the non-laves phase of Zr7Ni10 is found to be better. Song et al. [28, 29] have earlier studied the ZrMn0.9−xV x Ni1.1 (x from 0.1 to 0.8) system and concluded that C15 phase with higher stacking faults performs better for electrochemical applications. In the present case, since Zr7Ni10 concurrently presents with the C15 phase, the higher sorption performances cannot be solely related to the stacking faults itself. Moreover, there are reports suggesting that the Zr7Ni10 offers catalytic activity for surface reactions and also acts as a medium for charge diffusion [30–33]. However, in spite of the presence of Zr7Ni10 phase since few C15 compositions did not store hydrogen appreciably, it appears to be the distinct electronic structure of some composition of Zr–Mn–Ni makes the difference. Particularly, any compositions within the ZrMn0.75Ni1.25 and ZrMn0.50Ni1.50 could be promising.

Conclusions

In the case of ZrMn2−xNi x laves phase system, thermodynamically promising hydrogen storage alloys can be made within the range of 1.5 ≤ x ≥ 1. ZrMn2 and ZrMn1.5Ni0.5 absorb ~3.6 H/F.U., but the absorbed hydrogen cannot be liberated under manageable conditions. ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 hydrides reversibly store hydrogen capacity such as 3.2 H/F.U., 3 H/F.U. and 2.2 H/F.U., respectively. Each of these hydrides stores at least half of its effective capacity within 1 min of charging time. The working range of ZrMnNi, ZrMn0.75Ni1.25 and ZrMn0.5Ni1.5 are 120–180 °C, 60–120 °C and 10–60 °C, respectively. Interesting thermodynamic behavior could be witnessed when we proceed from ZrMn0.75Ni1.25 to ZrMn0.5Ni1.5 system which is also confirmed by DSC analysis.