Investigation of Catalyst Development from Mg 2 NiH 4 Hydride and Its Application for the CO 2 Methanation Reaction

: In current study various aspects of catalyst development for the Sabatier type methanation reaction were investigated. It was demonstrated that starting from 330–380 ◦ C Mg 2 NiH 4 hydride heating under CO 2 and H 2 gas ﬂow initiates hydride decomposition, disproportionation and oxidation. These reactions empower catalytic properties of the material and promotes CO 2 methanation reaction. Detailed structural, colorimetric and thermogravimetric analysis revealed that in order to have fast and full-scale development of the catalyst (formation of MgO decorated by nanocrystalline Ni) initial hydride has to be heated above 500 ◦ C. Another considerable ﬁnding of the study was conﬁrmation that potentially both high grade and low grade starting Mg 2 Ni alloy can be equally suitable for the hydride synthesis and its usage for the promotion of methanation reactions. pass energies, respectively. An Ar + ion gun (a 4 kV accelerating voltage and a 2 × 2 mm sputtering area) was used for sputtering. XPS spectra processing and analysis was done using the MultiPak software.


Introduction
In 2019, the European Commission declared an ambitious goal by 2050 to become the first climate-neutral continent in the world [1]. Such a transition will require comprehensive changes and improvements at every element of the energy system including its generation, storage, and utilization. Naturally, the current 20% share of the renewable energy sources in EU energy generation [2] will have to be substantially increased. In addition, a wide diversity of carbon-neutral technologies will have to be applied to balance the mismatch of energy generation and demand (both in time and site) [3]. Power-to-gas systems can be used to transform renewable electrical energy into chemical energy in the form of substitute natural gas (SNG) and is among the top candidate technologies to balance the mismatch [4][5][6]. In comparison to electricity, SNG has two important advantages. First, it can store energy for the long periods of time. Second, existing infrastructure makes it immediately available to be used as fuel for transport and a convenient measure to provide energy for remote energy consumers.
One of the most environmentally friendly power-to-gas solutions is to generate hydrogen (e.g., by the electrolysis of water or via a reversed PEM fuel cell) and then to react it with CO 2 through the Sabatier reaction [3]. Though there are still some disputes over the mechanisms of the individual steps [7] an overall Sabatier process Equation (1) can be described as a two stage (Equations (2) and (3)) reaction [8]: Overall reaction: 4H 2(g) + CO 2(g) → CH 4(g) + 2H 2 O (g) ∆H 298K = −165 kJ/mol (1) First stage: H 2(g) + CO 2(g) → CO (g) + H 2 O (g) ∆H 298K = 41 kJ/mol (2) 2 of 15 Second stage: 3H 2(g) + CO (g) → CH 4(g) + H 2 O (g) ∆H 298K = −206 kJ/mol (3) Gao et al. [9] reported that based on free Gibbs energy calculations, the optimal temperature (in respect to CH 4 selectivity and CO x conversion) for the Sabatier reaction (1) is 300-400 • C. However, due to the endothermic reaction (2), relatively high activation barrier [3], and limited kinetics, at this temperature range significant methanation reaction is observed only with the presence of suitable catalyst [10]. The need for the catalyst is lessened when temperature exceeds 500 • C and reverse water gas shift reaction (2) becomes exothermic. But at high temperatures the competing Bosch reaction takes place and the efficiency of methanation process (1) is starting to decrease [9]. The presence of conventional heterogeneous catalysts (for example Ni, Fe, Co or Ru [11]) do not help either because in addition to promoting CO 2 methanation they also serve as catalytic drivers for the competing reaction (4) which produces elemental carbon and poisons the catalysts [12]: Bosch reaction: 2H 2(g) + CO 2(g) → C (s) + 2H 2 O (g) (4) Looking for the solutions to avoid carbon build up at elevated temperatures Gao et al. [9] and Jurgensen et al. [12] investigated the role of CO 2 :H 2 and CO:H 2 gas ratios. The researchers determined that higher gas ratios have positive effects on lowering down carbon production and that preferable CO 2 :H 2 gas ratios should be at least 1:6 [9]. Additionally, it was demonstrated that by increasing gas mixture pressure from 1 to 11 bars, the starting temperature for carbon generation rises from 365 to 515 • C [12]. The other way to minimize carbon formation and the poisoning of the catalyst is to find new materials that have a higher selectivity for CH 4 production and do not support the Bosch reaction.
In 1990, Selvam et al. [13] investigated the interaction between CO 2 , hydrogen storage alloys, and compounds including LaNi 5 , CaNi 5 , Mg 2 Ni, Mg 2 Cu, and FeTi. They reported that during exposure to air, all these compounds form surface oxides and hydroxides that, in turn, actively adsorb atmospheric CO 2 and favor the formation of carbonate species on the top few layers of the surface. In a subsequent study of air-exposed Mg 2 NiH 4 hydride [14], Selvam et al. reached conclusion that later hydride also undergoes similar processes and forms surface carbonate species, especially in the presence of moisture.
Two decades later, Kato et al. [6] argued that the high absorptivity of CO 2 on an Mg 2 NiH 4 surface could be beneficial for the Sabatier type methanation reaction. Their comprehensive study on the catalytic interactions of the hydride surface of Mg 2 NiH 4 powder with CO 2 provided valuable insights into material disproportionation and oxidation during cyclic hydrogen absorption and hydrogen desorption under CO 2 (at temperatures up to 500 • C). Based on their findings, during the first few dehydriding cycles in a CO 2 -containing atmosphere, the simultaneous disproportionation of Mg 2 NiH 4 and the selective oxidation of Mg take place. As a result, a layered structure consisting of Mg 2 NiH 4 /Mg 2 Ni/Ni/MgO was formed. According to the authors, when it was active, Ni helped to dissociate CO 2 and CO molecules and promoted methanation [6]. Eventually, after approximately 20 cycles, Mg 2 NiH 4 and Mg 2 Ni were no longer formed and only MgO and Ni phases were observed.
In two more recent studies, Grasso et al. reported experimental data on CO 2 methanation processes using as-sintered monoclinic [15] and as-milled cubic Mg 2 NiH 4 [16] powders. In these studies, Mg 2 NiH 4 powders served as the sole hydrogen sources and "providers" of catalytic sites for the promotion of the conversion of CO 2 to CH 4 . To proceed with the experiments, a certain amount of hydride powder was placed into a stainless steel reactor connected to Sieverts volumetric equipment. The reactor was provided a specific CO 2 pressure and heated to 400 • C at a rate of 10 • C/min. The mass of the used Mg 2 NiH 4 powder and the applied CO 2 pressure (approximately 1.2 bars [16]) were proportioned in such a way that the calculated total molar quantity of H 2 released by the complete decomposition of Mg 2 NiH 4 would ensure an H 2 :CO 2 molar ratio of 4:1. Interestingly, despite the nearly identical reaction conditions, the authors found evidence for two slightly different reaction mechanisms. After 10 h of the as-sintered monoclinic Mg 2 NiH 4 's reaction with CO 2 at 400 • C, the complete decomposition of Mg 2 NiH 4 was not reached, although CO 2 was totally consumed and no carbon deposition was observed [15]. Considering the intermediate reaction products that were obtained after 1 and 5 h (namely, Mg 2 NiH 4 , MgH 2 , Mg 2 Ni, MgO, and CO), Grasso et al. concluded that the methanation of CO 2 using Mg 2 NiH 4 involves two simultaneous processes: (i) the catalytic conversion of CO 2 through reactions (2) and (3) and (ii) the direct reduction of CO 2 by the reducing effect of MgH 2 (Equation (5)).
The chemical activity of the as-milled cubic polymorph form of Mg 2 NiH 4 was found to be considerably higher, and its reaction with CO 2 at 400 • C was completed in just 5 h [16]. The observed intermediate and final products of the methanation process were slightly different from the formerly described case. Therefore, the authors concluded that under a CO 2 atmosphere, as-milled Mg 2 NiH 4 rapidly decomposes directly to the Ni-Mg 2 Ni-MgNi 2 -/MgO (all of this phase remains after the reaction is complete) catalytic system Equation (6) which promotes the methanation reaction Equation (7).
Reports by Kato et al. [6] and Grasso et al. [15,16] provided clear evidence that during its interaction with CO 2 gas at 400-500 • C, Mg 2 NiH 4 powder disproportionates into catalytically active compounds that efficiently promote the generation of methane. However, in all three studies, different reaction products were observed and divergent mechanisms were proposed. Grasso et al. assumed that at least some of the observed variations could have related to the specific forms and preparation methods of the Mg 2 NiH 4 hydride [16].
Potentially, an Mg 2 NiH 4 hydride-based CO 2 methanation reaction catalyst can be used in powder, as well as in the supported film (coating) form, but up to now, all of the reported studies were only conducted with Mg 2 NiH 4 powders. Previous studies of Mg 2 NiH 4 hydride films [17,18] indicated that characteristic strains caused by the film-substrate interaction can introduce noticeable changes of properties in comparison to corresponding powders. Therefore, there is a possibility that one form of Mg 2 NiH 4 hydride might be better suited for CO 2 methanation catalysis than the other.
Accordingly, in the current study, we investigated the structural transformations of Mg 2 NiH 4 hydride films under a CO 2 atmosphere and examined whether they correlated with the ones reported for Mg 2 NiH 4 powders. In addition, our interest to investigate Mg 2 NiH 4 films had two more motives. Firstly, film condensation from physical vapor allows one to synthesize very uniform samples with strictly controlled component ratios that prevent the formation of undesirable phases. This makes them particularly suitable as the subject for in-situ structural analysis, and this is useful for resolving discrepancies between the proposed reaction pathways. Secondly, films are eligible to be approximated as 1D objects, which is a clear advantage for the observation and evaluation of surface changes, including the analysis of surface region depth profiles.
In addition, we synthesized Mg 2 NiH 4 in powder and investigated its efficiency for the methanation of CO 2 in order to compare processes in different forms of Mg 2 NiH 4 hydride and to see how this material develops during the CO 2 methanation reaction.

Material Synthesis
Mg 2 NiH 4 hydride films were formed by using magnetron sputtering to deposit metallic Mg 2 Ni films and then hydriding them for 48 h under 20 bars of H 2 pressure at a temperature of 250 • C. The deposition of Mg 2 Ni films was realized under a 6 × 10 −3 mbar Ar gas atmosphere inside a Kurt J. Lesker PVD-75 system (Jefferson Hills, PA, USA) by co-sputtering with two circular Torus 3 magnetrons. The nominal purities of Mg and Ni targets were 99.99% and 99.995%, respectively. By adjusting the output power for individual magnetrons (196 W for Mg and 120 W for Ni), the Mg:Ni ratio in the films was tuned up to approximately 68:32 (as measured by EDS). A fraction that is slightly higher than the stochiometric fraction of Mg has been proven to be beneficial for the synthesis of Mg 2 NiH 4 hydride [3,19] because it prevents the crystallization of localized MgNi 2 phase nanoformations and makes full hydrogenation easier to attain. Mg 2 Ni films were deposited on 20 × 20 mm fused silica substrates. The approximate thickness of the films was 500 nm.
In previous studies [6,15,16], Mg 2 NiH 4 catalysts were synthesized from high purity Mg 2 Ni powders that were prepared under controlled laboratory conditions. To extend the understanding of processes, in the current study, we chose to use a lower quality starting Mg 2 Ni material (Mg 2 Ni alloy granules of 99% purity obtained from American Elements) because it better reflected the potential conditions of Mg 2 NiH 4 applications in commercial methanation reactors. At the same time, the usage of a less pure Mg 2 Ni alloy (both in elemental composition and phase) was expected to help to identify any significant material purity-related shortcomings of the methanation process. The average size of the initial Mg 2 Ni alloy granules was 3 mm. Prior to hydriding, the granules were mechanically ground down to a grain size of 20-50 µm. The obtained powders were placed into a stainless steel container and pumped down to a vacuum for several hours. The initial activation of the Mg 2 Ni powders [20,21] was achieved by applying 4 hydriding (16 h, 250 • C, and 20 bar H 2 ) and dehydriding cycles (8 h, 250 • C, vacuum). Final hydriding was conducted for 24 h under 20 bars of H 2 pressure at 250 • C. The specific surface areas of the unhydrided and hydrided Mg 2 Ni powders were estimated by the BET method and reached 1.59 and 2.69 m 2 /g, respectively.

In-Situ XRD
The in situ XRD characterization of Mg 2 NiH 4 films during annealing under a CO 2 atmosphere was performed with a Bruker D8 Discover (Hamburg, Germany) equipped with Mri TC-basic (Hamburg, Germany) chamber. For this type of experiment, Mg 2 NiH 4 hydride films were placed on a Pt:Rh heating foil strip, and the temperature was monitored with an S-type thermocouple that was laser-welded to the backside of the foil. Prior to the heating, the chamber was abundantly flushed and filled up with CO 2 up to an absolute pressure of approximately 1.6 bar. From earlier reports by Kato et al. [6] and our own experience working with the in-situ heating of Mg 2 NiH 4 films in air [22], it was known that at low temperatures (< 200 • C), the hydride remains satisfactorily stable. Therefore, in the beginning, XRD patterns were recorded with relatively large steps at temperatures of 30, 100, 150, and 200 • C. The heating rate between the steps was 1 • C/min, and at each temperature before pattern recording, samples were left for 60 min to reach thermal equilibrium with the heater. In temperature range from 200 to 500 • C, XRD patterns were recorded with 10 • C steps. The heating rate was kept to 1 • C/min, and the delay time was set to 30 min. The net measurement time for each pattern was approximately 90 min. Accordingly, below 200 • C, the total time samples were kept at each temperature step was 150 min, and an additional 50 min were used for heating. The corresponding values for the upper range measurements were 120 and 10 min, respectively.

Surface Characterisation Techniques
SEM measurements were performed using a Hitachi S-3400N (Tokyo, Japan) microscope. The surface depth profiles of the film samples was analyzed with an X-ray photoelectron spectrometer (PHI Versaprobe 5000, Boston, MA, USA) using monochromated 1486.6 eV Al radiation, 25 W of beam power, a 100 µm beam size, and a 45 • measurement angle. Survey and high resolution XPS spectra were acquired using 187.85 eV and 23.5 eV band pass energies, respectively. An Ar + ion gun (a 4 kV accelerating voltage and a 2 × 2 mm sputtering area) was used for sputtering. XPS spectra processing and analysis was done using the MultiPak software.

Mass Changes and Thermal Effects
Differential scanning calorimetry (DSC) and TGA were performed with a NETZSCH STA 449 F3 Jupiter (Hamburg, Germany) analyzer with an SiC furnace. The heating program started at 45 • C and continued up to 580 • C at a heating rate of 10 • C/min. Argon gas (with a flow rate of 60 mL/min) was used as the inert atmosphere. The oxidizing environment consisted of carbon dioxide (with a flow rate of 25 mL/min) and argon (with a flow rate 35 mL/min). The TGA sample carrier with an S-type thermocouple was calibrated for 200 mg Al 2 O 3 TGA crucibles. For each measurement, 8.4-8.8 mg samples were used.

Methanation
CO 2 methanation experiments were carried out with the custom build setup presented in Figure 1. For each CO 2 methanation experiment, 5 g of hydride powder were mixed with 5 g of fine Al 2 O 3 powder (Sigma-Aldrich, 98% purity) and placed inside a cylindrical, high-temperature, stainless steel container (approximate length of 80 mm and internal diameter of 15 mm). From both ends, powder was closely packed by high purity quartz wool plugs. The container was heated up b ay vertical high temperature tubular furnace, and the temperature of the powder was measured by an internal thermocouple probe that was inserted into the middle of the powder. The gas supply system consisted of three mass flow controllers (MFCs) dedicated to CO 2 , H 2 , and Ar gases. Flows from all MFC units were controlled by one digital gas controller manufactured by Brooks (Seattle, WA, USA). The controller was set up to maintain a total gas flow of 1 L/min while keeping the H 2 :CO 2 gas flow ratio at 6:1 (following the suggestion by [9]). An automatic constant pressure regulator outside of the reaction vessel was used to maintain a constant pressure inside the reaction zone. A test pressure of 1 bar was selected to mimic the potential reaction conditions of the methanation reactor without any additional technical measures, whereas 10 bars of pressure were tested in order to limit the formation of carbon, as suggested by [12].

Methanation
CO2 methanation experiments were carried out with the custom build setup presented in Figure  1. For each CO2 methanation experiment, 5 g of hydride powder were mixed with 5 g of fine Al2O3 powder (Sigma-Aldrich, 98% purity) and placed inside a cylindrical, high-temperature ,stainless steel container (approximate length of 80 mm and internal diameter of 15 mm). From both ends, powder was closely packed by high purity quartz wool plugs. The container was heated up b ay vertical high temperature tubular furnace, and the temperature of the powder was measured by an internal thermocouple probe that was inserted into the middle of the powder. The gas supply system consisted of three mass flow controllers (MFCs) dedicated to CO2, H2, and Ar gases. Flows from all MFC units were controlled by one digital gas controller manufactured by Brooks (Seattle, WA, USA). The controller was set up to maintain a total gas flow of 1 L/min while keeping the H2:CO2 gas flow ratio at 6:1 (following the suggestion by [9]). An automatic constant pressure regulator outside of the reaction vessel was used to maintain a constant pressure inside the reaction zone. A test pressure of 1 bar was selected to mimic the potential reaction conditions of the methanation reactor without any additional technical measures, whereas 10 bars of pressure were tested in order to limit the formation of carbon, as suggested by [12]. The supplementary inert Ar gas was used to maintain the recommended gas flow requirements of the on-line gas analyzer (VISIT 03H manufactured by Messtechnik EHEIM GmbH) and to transfer some heat away from the reaction zone. The optimal ratio between the reagent gas (H2 and CO2) and the inert gas (Ar) carrier was experimentally predetermined and kept at 1:1. During data processing, Ar gas contribution was numerically deduced. At certain points during the CO2 methanation process, gaseous reaction products were taken out for the verification by a stand-alone gas chromatographer (Agilent 7890A, Santa Clara, CA, USA). The observed results were consistent with the data from the The supplementary inert Ar gas was used to maintain the recommended gas flow requirements of the on-line gas analyzer (VISIT 03H manufactured by Messtechnik EHEIM GmbH) and to transfer Coatings 2020, 10, 1178 6 of 15 some heat away from the reaction zone. The optimal ratio between the reagent gas (H 2 and CO 2 ) and the inert gas (Ar) carrier was experimentally predetermined and kept at 1:1. During data processing, Ar gas contribution was numerically deduced. At certain points during the CO 2 methanation process, gaseous reaction products were taken out for the verification by a stand-alone gas chromatographer (Agilent 7890A, Santa Clara, CA, USA). The observed results were consistent with the data from the on-line gas analyzer.

Results and Discussions
As is typical for metallic films, the as-deposited Mg 2 Ni coatings had mirror like appearances. After hydriding, they became transparent and had a bright orange hue. Both of these features are characteristic for the low temperature (LT) phase of Mg 2 NiH 4 [22][23][24]. The prevalence of the monoclinic LT phase of Mg 2 NiH 4 was subsequently confirmed by XRD ( Figure 2). In addition to the LT phase, small fractions of the so-called pseudo-cubic high temperature (HT) phase of Mg 2 NiH 4 [17,25] were observed at room temperature.
Coatings 2020, 10, x FOR PEER REVIEW 6 of 15 monoclinic LT phase of Mg2NiH4 was subsequently confirmed by XRD ( Figure 2). In addition to the LT phase, small fractions of the so-called pseudo-cubic high temperature (HT) phase of Mg2NiH4 [17,25] were observed at room temperature. During the in-situ heating of the Mg2NiH4 film under the CO2 atmosphere ( Figure 2), several temperature ranges related to substantial composition changes could be identified. Starting from the room temperature up till 240 °C, the LT and HT phases of Mg2NiH4 co-existed, and their ratio in the film did not change noticeably. Above 240 °C, the fraction of LT phase gradually decreased, and at 300 °C, only the HT phase of Mg2NiH4 persisted. The initiation of Mg2NiH4 decomposition was observed at 350-360 °C when weak peaks attributed to Mg2Ni arose. At 410 °C, the transition from Mg2NiH4 to Mg2Ni was complete, and at 420 °C, only peaks of Mg2Ni were observed. A further increase of the temperature was followed by the gradual disproportionation of Mg2Ni and the crystallization of MgO oxide. The crystalline MgO phase was at first observed at 450 °C and remained as the main phase at the end of the measurement at 500 °C.
Surface images made by SEM and the near surface region depth profile recorded by XPS complemented the findings of in-situ XRD and highlighted structural changes of the film that occurred due to the heating in the CO2 atmosphere. More specifically, the SEM images ( Figure 3) show that after heating in the CO2 atmosphere, the surface of the films became considerably rougher, which indicated significant mass transfer processes within the film. At the same time, an XPS depth profile (Figure 3d) specified that the Ni concentration at the surface was less than 1/6 of Mg. This was clear evidence that during the disproportionation of Mg2Ni, magnesium separated from Ni, oxidized, and had a tendency to segregate at the near surface region.
By comparing earlier reports by Kato et al. [6] and Grasso et al. [15,16], who worked with powders, with current in-situ XRD data that were observed for film, an outstanding feature was found: in the later case, there were no signs of the crystalline MgNi2 and Ni phases. On one hand, this could be considered a sign of a slightly different reaction pathway for the film samples in comparison to the powder equivalent. On the other hand, it seems that a full-scale comparison does not provide During the in-situ heating of the Mg 2 NiH 4 film under the CO 2 atmosphere (Figure 2), several temperature ranges related to substantial composition changes could be identified. Starting from the room temperature up till 240 • C, the LT and HT phases of Mg 2 NiH 4 co-existed, and their ratio in the film did not change noticeably. Above 240 • C, the fraction of LT phase gradually decreased, and at 300 • C, only the HT phase of Mg 2 NiH 4 persisted. The initiation of Mg 2 NiH 4 decomposition was observed at 350-360 • C when weak peaks attributed to Mg 2 Ni arose. At 410 • C, the transition from Mg 2 NiH 4 to Mg 2 Ni was complete, and at 420 • C, only peaks of Mg 2 Ni were observed. A further increase of the temperature was followed by the gradual disproportionation of Mg 2 Ni and the crystallization of MgO oxide. The crystalline MgO phase was at first observed at 450 • C and remained as the main phase at the end of the measurement at 500 • C.
Surface images made by SEM and the near surface region depth profile recorded by XPS complemented the findings of in-situ XRD and highlighted structural changes of the film that occurred due to the heating in the CO 2 atmosphere. More specifically, the SEM images ( Figure 3) show that after heating in the CO 2 atmosphere, the surface of the films became considerably rougher, which indicated significant mass transfer processes within the film. At the same time, an XPS depth profile (Figure 3d) specified that the Ni concentration at the surface was less than 1/6 of Mg. This was clear evidence that during the disproportionation of Mg 2 Ni, magnesium separated from Ni, oxidized, and had a tendency to segregate at the near surface region. By comparing earlier reports by Kato et al. [6] and Grasso et al. [15,16], who worked with powders, with current in-situ XRD data that were observed for film, an outstanding feature was found: in the later case, there were no signs of the crystalline MgNi 2 and Ni phases. On one hand, this could be considered a sign of a slightly different reaction pathway for the film samples in comparison to the powder equivalent. On the other hand, it seems that a full-scale comparison does not provide substantial proof for that. For example, the observed quantitative concentration values, as well as their qualitative changes at the surface of the film (Figure 3d) in great proximity, resembled the data presented by Kato et al. [6], who measured the XPS depth profile of Mg 2 NiH 4 powders that were cyclically exposed to a CO 2 atmosphere at a temperature of 500 • C. Furthermore, several tests with Mg 2 NiH 4 powder that had an MgNi 2 phase (more details are provided in the following paragraphs) showed that there was a very close agreement between on-set process temperatures for powder and film samples. Altogether, we presume that the absence of MgNi 2 and Ni phases in in-situ XRD data was the result of insufficient crystallization rather than different reaction pathways.  XRD patterns of commercial Mg2Ni powders that contained Mg and MgNi2 components are provided in Figure 4a. After hydriding these powders, the crystal phases of Mg2NiH4, MgH2, and MgNi2 were obtained (Figure 4b). The XRD patterns of the hydrided powders had no peaks of unreacted Mg and Mg2Ni phases, so we assume that the loss of hydrogen concentration (see DSC/TGA analyses below) was caused by the aggregative effect of hydrogen non-adsorbing phases like MgNi2. MgO was not observed by XRD, but we nevertheless assume that it also contributed to the lower hydrogen concentration because commercial Mg-Ni grains were delivered in an aircontaining canister and had successive handling in air.  XRD patterns of commercial Mg 2 Ni powders that contained Mg and MgNi 2 components are provided in Figure 4a. After hydriding these powders, the crystal phases of Mg 2 NiH 4 , MgH 2 , and MgNi 2 were obtained (Figure 4b). The XRD patterns of the hydrided powders had no peaks of unreacted Mg and Mg 2 Ni phases, so we assume that the loss of hydrogen concentration (see DSC/TGA analyses below) was caused by the aggregative effect of hydrogen non-adsorbing phases like MgNi 2 . MgO was not observed by XRD, but we nevertheless assume that it also contributed to the lower hydrogen concentration because commercial Mg-Ni grains were delivered in an air-containing canister and had successive handling in air.
The DSC analysis of hydrided powders under Ar flow indicated the presence of two distinct thermal events (Figure 5a). The first endothermal event started at approximately 235 • C and was attributed to the LT-HT phase transition of Mg 2 NiH 4 [26,27]. Close to the beginning of the phase transition, the sample started to lose some mass (up to 0.2 wt.%), and this indicated the partial decomposition of Mg 2 NiH 4 . Proceeding further, most of the sample mass (≈2.6 wt.%) was lost during the second endothermal event starting at approximately 315-320 • C. This event represented the rapid hydrogen release during the supposedly full dehydriding of Mg 2 NiH 4 . Lastly, there was a lengthy mass loss of approximately 0.1 wt.% that led to a total mass loss of nearly 2.8 wt.%. Accordingly, we attributed the last phase of mass loss to the dehydriding of MgH 2 .
MgNi2 were obtained (Figure 4b). The XRD patterns of the hydrided powders had no peaks of unreacted Mg and Mg2Ni phases, so we assume that the loss of hydrogen concentration (see DSC/TGA analyses below) was caused by the aggregative effect of hydrogen non-adsorbing phases like MgNi2. MgO was not observed by XRD, but we nevertheless assume that it also contributed to the lower hydrogen concentration because commercial Mg-Ni grains were delivered in an aircontaining canister and had successive handling in air. The DSC analysis of hydrided powders under Ar flow indicated the presence of two distinct thermal events (Figure 5a). The first endothermal event started at approximately 235 °C and was attributed to the LT-HT phase transition of Mg2NiH4 [26,27]. Close to the beginning of the phase transition, the sample started to lose some mass (up to 0.2 wt.%), and this indicated the partial decomposition of Mg2NiH4. Proceeding further, most of the sample mass (≈2.6 wt.%) was lost during the second endothermal event starting at approximately 315-320 °C. This event represented the rapid hydrogen release during the supposedly full dehydriding of Mg2NiH4. Lastly, there was a lengthy mass loss of approximately 0.1 wt.% that led to a total mass loss of nearly 2.8 wt.%. Accordingly, we attributed the last phase of mass loss to the dehydriding of MgH2.  The first thermal event during hydrided powder heating under the Ar and CO2 gas flow started with the LT-HT phase transition at 235 °C (Figure 5b). A nearly identical onset temperature and width of the Mg2NiH4 phase transition range (235-300 °C) were also observed for the Mg2NiH4 film ( Figure 2) and powder samples (Figure 5a) that were tested under CO2 and Ar gas atmospheres, respectively. This indicated that below the phase transition temperature, the activation barrier for Mg2NiH4 hydride decomposition and/or oxidation was relatively high, and at near-atmosphere pressures, Mg2NiH4 hydride demonstrated enduring stability regardless of its form and surrounding gas phase composition.
Disparities between Mg2NiH4 hydride behavior under the inert and reactive gases arose soon after start of the LT-HT phase transition. Namely, under the CO2 and Ar gas flow, there were no mass change events up to approximately 330 °C. An analysis of the results of the in-situ XRD of Mg2NiH4 film heating under CO2 gas ( Figure 2) suggested that at later temperatures, the dehydriding of Mg2NiH4 phase should have taken place. The decomposition reaction of Mg2NiH4 was expected to result in approximately 2.7 wt.% mass loss and a strong endothermic minimum in the TGA and DSC curves, respectively (Figure 5a). Instead, under the CO2 and Ar gas flow, a prolonged small (≈0.1 wt.%) decline of mass and a low intensity exothermic upswing was observed. Therefore, it is clear that the endothermal reaction (8) [28] was simultaneously accompanied by some additional reactions. Kato et al. [6] reported that under a CO2 and H2 gas flow over Mg2NiD4 covered with surface oxide layers, the conversion of CO2 to CO (Equation (2)) and methane (Equation (3)) commenced at 330 °C. The sum of these reactions was exothermic, and, considering that there was some hydrogen released from the Mg2NiH4 methanation of CO/CO2, this could explain the lack of an endothermic The first thermal event during hydrided powder heating under the Ar and CO 2 gas flow started with the LT-HT phase transition at 235 • C (Figure 5b). A nearly identical onset temperature and width of the Mg 2 NiH 4 phase transition range (235-300 • C) were also observed for the Mg 2 NiH 4 film ( Figure 2) and powder samples (Figure 5a) that were tested under CO 2 and Ar gas atmospheres, respectively. This indicated that below the phase transition temperature, the activation barrier for Mg 2 NiH 4 hydride decomposition and/or oxidation was relatively high, and at near-atmosphere pressures, Mg 2 NiH 4 hydride demonstrated enduring stability regardless of its form and surrounding gas phase composition.
Disparities between Mg 2 NiH 4 hydride behavior under the inert and reactive gases arose soon after start of the LT-HT phase transition. Namely, under the CO 2 and Ar gas flow, there were no mass change events up to approximately 330 • C. An analysis of the results of the in-situ XRD of Mg 2 NiH 4 film heating under CO 2 gas (Figure 2) suggested that at later temperatures, the dehydriding of Mg 2 NiH 4 phase should have taken place. The decomposition reaction of Mg 2 NiH 4 was expected to result in approximately 2.7 wt.% mass loss and a strong endothermic minimum in the TGA and DSC curves, respectively (Figure 5a). Instead, under the CO 2 and Ar gas flow, a prolonged small (≈0.1 wt.%) decline of mass and a low intensity exothermic upswing was observed. Therefore, it is clear that the endothermal reaction (8) [28] was simultaneously accompanied by some additional reactions. Kato et al. [6] reported that under a CO 2 and H 2 gas flow over Mg 2 NiD 4 covered with surface oxide layers, the conversion of CO 2 to CO (Equation (2)) and methane (Equation (3)) commenced at 330 • C. The sum of these reactions was exothermic, and, considering that there was some hydrogen released from the Mg 2 NiH 4 methanation of CO/CO 2 , this could explain the lack of an endothermic minimum in the DSC curve. However, the CO 2 methanation through Reactions (2) and (3) does not explain the lack of the mass loss. The Bosch reaction (reaction (4)), carbon monoxide reduction (reaction (9)), and Boudouard Reaction (10) could potentially lead to solid carbon build up [29,30]. However, most of these reactions become depressed as temperature is increased [9]. Therefore, without the complete denial of carbonization, we suggest that the disproportionation of Mg 2 Ni powder surface and its oxidation were the most dominant processes in our experiments. A similar conclusion was drawn up by other researchers [6,15,16] and was supported by the eventual observation of an MgO phase by in-situ XRD. In this reaction pathway, finite oxygen adsorption compensated the mass of the desorbed hydrogen; meanwhile, the strongly exothermic formation of MgO overcame the endothermic decomposition of Mg 2 NiH 4 . To reaffirm this view, we note that the TGA curve reached its minimum slightly above 400 • C, which was consistent with the disappearance of Mg 2 NiH 4 XRD peaks at 410-420 • C ( Figure 2). The exhaustion of hydrogen sources should have stopped all of the above-mentioned carbon production reactions, stabilized sample mass, and translated into severe changes of the DSC curve. Instead, both curves maintained small increases, thus indicating the superior role of other processes.
A strong endothermal event at 475 • C was accompanied by a 0.1% mass loss that was equal to the hydrogen desorption in the last section of the sample heating under Ar (Figure 5a). Accordingly, following the same reasoning as above, we attribute this event to the decomposition of MgH 2 . The higher than usual decomposition temperature can be explained by two factors: (i) MgH 2 decomposition is usually kinetically limited [31] and the heating rate during DSC/TGA analysis was relatively high (10 • C/min); (ii) the dehydriding temperature is known to be postponed by surface oxides [15,32]. After MgH 2 decomposition, the powder reaction with CO 2 was finalized by some more thermal events, thus representing rapid oxidation of Mg that was produced by the dehydriding of MgH 2 and continued the disproportionation of Mg 2 Ni and, eventually, MgNi 2 .
The CO 2 methanation reaction dynamics at 1 and 10 bars of gas pressure are provided in Figure 6a,b, respectively. Both charts have temperature peaks that correspond to the fast decomposition of hydrides (Mg 2 NiH 4 and MgH 2 ) and the oxidation of Mg 2 Ni disproportionation products. After the heat burst, we found a new catalyst with substantially different properties. By comparing methanation dynamics at different pressures, it could be noticed that at 1 bar in a low temperature region (prior to the temperature burst), there was a slight increase in H 2 (and a decrease in CO 2 ) concentration, whereas at 10 bars, the H 2 :CO 2 ratio remained stable. At 1 bar of pressure, such behavior was not surprising because a similar hydrogen release was also observed in the TGA pattern ( Figure 5a) and was attributed to the plateau pressure levels of Mg 2 NiH 4 [33,34]. On the other hand, Mg 2 NiH 4 stability at 10 bars of pressure demonstrated that high partial hydrogen pressure was a stronger factor for the stabilization of Mg 2 NiH 4 than the oxidative potential of CO 2 , even though at low pressure Mg 2 NiH 4 decomposition and partial oxidation happened at the same time (see Figure 5b and comments above).
The second thing to notice from Figure 6 is that at 10 bars of pressure, the methanation reaction started at approximately 50 • C lower (330 • C versus 380 • C), and as soon as temperature reached 420-430 • C, there were almost no unreacted CO 2 or intermediate products from reaction (2), namely CO. A nearly 100% CO 2 conversion was maintained up to approximately 530 • C, with an estimated CH 4 yield of approximately 75%. On the other hand, methanation at 1 bar of pressure was able to reach a 100% CO 2 conversion within a narrow time frame when the hydrogen desorption from powders significantly increased the hydrogen concentration in the gas flow. When the hydrogen source was depleted, the methanation reaction lost its temporarily boosted efficiency (approximately 90% at 500 • C) and maintained a downtrend as temperature rose to 550 • C. At both investigated pressures, CO 2 methanation by the reaction product of Mg 2 NiH 4 powder had lower CH 4 yields than were achieved with standard commercial Ni-based catalysts at 10 bar (H 2 :CO 2 ratio of 4:1) and 30 bar (H 2 :CO 2 ratio of 6:1) [9]. When operating at 10 bars of pressure, we did not detect CO or other carbon-containing species in gaseous reaction products; therefore, we assumed that relative difference between initial CO 2 and CH 4 product fluxes qualitatively reflected the actual solid C yields. This was quite surprising because theoretical and experimental study by Gao et al. [9] estimated that at the specifically used gas pressure and H 2 :CO 2 ratios, carbonization should have been prohibited. The cause of such a high carbon yield has not yet been identified and will be subject of our future studies. translated into severe changes of the DSC curve. Instead, both curves maintained small increases, thus indicating the superior role of other processes.
A strong endothermal event at 475 °C was accompanied by a 0.1% mass loss that was equal to the hydrogen desorption in the last section of the sample heating under Ar (Figure 5a). Accordingly, following the same reasoning as above, we attribute this event to the decomposition of MgH2. The higher than usual decomposition temperature can be explained by two factors: (i) MgH2 decomposition is usually kinetically limited [31] and the heating rate during DSC/TGA analysis was relatively high (10 °C/min); (ii) the dehydriding temperature is known to be postponed by surface oxides [15,32]. After MgH2 decomposition, the powder reaction with CO2 was finalized by some more thermal events, thus representing rapid oxidation of Mg that was produced by the dehydriding of MgH2 and continued the disproportionation of Mg2Ni and, eventually, MgNi2.
The CO2 methanation reaction dynamics at 1 and 10 bars of gas pressure are provided in Figure  6a,b, respectively. Both charts have temperature peaks that correspond to the fast decomposition of hydrides (Mg2NiH4 and MgH2) and the oxidation of Mg2Ni disproportionation products. After the heat burst, we found a new catalyst with substantially different properties. By comparing methanation dynamics at different pressures, it could be noticed that at 1 bar in a low temperature region (prior to the temperature burst), there was a slight increase in H2 (and a decrease in CO2) concentration, whereas at 10 bars, the H2:CO2 ratio remained stable. At 1 bar of pressure, such behavior was not surprising because a similar hydrogen release was also observed in the TGA pattern ( Figure 5a) and was attributed to the plateau pressure levels of Mg2NiH4 [33,34]. On the other hand, Mg2NiH4 stability at 10 bars of pressure demonstrated that high partial hydrogen pressure was a stronger factor for the stabilization of Mg2NiH4 than the oxidative potential of CO2, even though at low pressure Mg2NiH4 decomposition and partial oxidation happened at the same time (see Figure 5b and comments above).
The second thing to notice from Figure 6 is that at 10 bars of pressure, the methanation reaction started at approximately 50 °C lower (330 °C versus 380 °C), and as soon as temperature reached 420- A study by Kato et al. [6] demonstrated that the properties of powder catalysts can be significantly improved after they are cycled through H 2 (25 bar and 350 • C), CO 2 (1 bar and 500 • C), and vacuum (500 • C). They reported that after 18 cycles, the methanation process was initiated at 250-260 • C, whereas an uncycled high purity Mg 2 NiH 4 catalyst initiated a methanation reaction at approximately 330-340 • C (in comparison to the current study with a relatively low grade Mg 2 NiH 4 catalyst that initiated methanation at 330 • C for 1 bar of pressure and 380 • C for 10 bars of pressure).
After considering the potential benefits of catalyst cycling after the first methanation run-up at 10 bars of pressure was finished, we gradually decreased furnace temperature down to 250 • C. During the cool-down period and 30 min after the target temperature was reached, the container was constantly flushed with hydrogen. Subsequently, the container was filled up with hydrogen to 10 bars of pressure and confined for 16 h at 250 • C. The next day, we repeated the methanation process ( Figure 7) with same parameters as were used before (10 bars of pressure and H 2 :CO 2 ratio of 6:1). This time, the methanation reaction started immediately at 250 • C (close to the lowest reported values [35,36]), reached a 100% CO 2 conversion at 320-330 • C, and continued without the significant deterioration of CO 2 conversion up to the maximum tested temperature of 600 • C. The CH 4 yield at Coatings 2020, 10, 1178 11 of 15 500 • C was slightly higher and reached just over 80%. Looking at the methanation dynamics (Figure 7a), one can notice that there was no temperature burst or hydrogen concentration increase that could be attributed to the decomposition of Mg 2 NiH 4 or MgH 2 hydride. This suggested that during the first methanation experiment, Mg 2 NiH 4 and MgH 2 hydrides completely disproportionated, and only the MgO and Ni phases were formed. Accordingly, such a powder catalyst did not adsorb any additional hydrogen and was able to provide an enhanced performance after one just methanation cycle. A comparison of SEM images of the as-ground Mg2Ni alloy grains ( Figure 8b) and the final catalyst powder (Figure 8c) provided some more evidence regarding how specific features of hydrides could become beneficial for the formation of efficient catalysts for the methanation reaction. The initially ground Mg2Ni alloy grains were relatively course, and their size ranged from 10 to 50 μm. During hydride formation (hydriding/dehydriding cycling), the catalyst material underwent huge volume changes (up to 30% [37]) that introduced large amount of structural defects, widened up grain boundaries, and ultimately fractured initial grains into sub-micrometer-size fine powders (Figure 8c). The first temperature run-up under CO2 and H2 gas flow promoted CO2 methanation and initiated hydride decomposition and oxidation. These highly energetic chemical reactions empowered Mg2Ni disproportionation and MgO-Ni phase separation ( Figure 9). As the result, MgO particles decorated by nanocrystalline Ni were formed. Such a catalyst is able to promote CH4 formation at a relatively low temperature (250 °C) and an do this with a significantly higher CH4 output than that not oxidized hydride powder. The superior catalytic properties of disproportionated MgO-Ni systems can be attributed to the role of MgO because MgO does not just work as an inert support for catalytic Ni nanoparticles but also actively changes the methanation pathway for Ni [38]. For example, studies have shown that MgO enhances the adsorption and activation of CO2 [39]. On the other hand, MgO was also found to promote H2O formation and its desorption from a catalyst [15]. A comparison of SEM images of the as-ground Mg 2 Ni alloy grains (Figure 8b) and the final catalyst powder (Figure 8c) provided some more evidence regarding how specific features of hydrides could become beneficial for the formation of efficient catalysts for the methanation reaction. The initially ground Mg 2 Ni alloy grains were relatively course, and their size ranged from 10 to 50 µm. During hydride formation (hydriding/dehydriding cycling), the catalyst material underwent huge volume changes (up to 30% [37]) that introduced large amount of structural defects, widened up grain boundaries, and ultimately fractured initial grains into sub-micrometer-size fine powders (Figure 8c). The first temperature run-up under CO 2 and H 2 gas flow promoted CO 2 methanation and initiated hydride decomposition and oxidation. These highly energetic chemical reactions empowered Mg 2 Ni disproportionation and MgO-Ni phase separation ( Figure 9). As the result, MgO particles decorated by nanocrystalline Ni were formed. Such a catalyst is able to promote CH 4 formation at a relatively low temperature (250 • C) and an do this with a significantly higher CH 4 output than that not oxidized hydride powder. The superior catalytic properties of disproportionated MgO-Ni systems can be attributed to the role of MgO because MgO does not just work as an inert support for catalytic Ni nanoparticles but also actively changes the methanation pathway for Ni [38]. For example, studies have shown that MgO enhances the adsorption and activation of CO 2 [39]. On the other hand, MgO was also found to promote H 2 O formation and its desorption from a catalyst [15].
In-situ XRD, TGA/DSC, and methanation reaction products analyses showed that in order for all these reactions to complete, it is recommended to heat a hydride powder above 500 • C (in a CO 2 -containing atmosphere) or, for a short period of time, as high as 600 • C. When such high temperature conditioning is induced (or in the current study case, a spontaneous temperature burst is not suppressed), the full development of the catalyst can be obtained in one cycle within 1 h. Meanwhile, the catalyst development from a Mg 2 NiH 4 powder kept below 500 • C might take more than 20 hydriding/dehydriding cycles [6] or up to 10-fold more time for the reaction to complete [15]. In-situ XRD, TGA/DSC, and methanation reaction products analyses showed that in order for all these reactions to complete, it is recommended to heat a hydride powder above 500 °C (in a CO2containing atmosphere) or, for a short period of time, as high as 600 °C. When such high temperature conditioning is induced (or in the current study case, a spontaneous temperature burst is not suppressed), the full development of the catalyst can be obtained in one cycle within 1 h. Meanwhile, the catalyst development from a Mg2NiH4 powder kept below 500 °C might take more than 20 hydriding/dehydriding cycles [6] or up to 10-fold more time for the reaction to complete [15].

Conclusions
In the current study, various aspects of Mg2NiH4=based catalyst development for a Sabatier type methanation reaction were investigated. Based on the congruous data of reaction on set temperatures and other evidence, we conclude that despite some differences in the crystallization of MgNi2 and Ni phases, both Mg2NiH4 hydride films and powders react with CO2 equally. It was demonstrated that In-situ XRD, TGA/DSC, and methanation reaction products analyses showed that in order for all these reactions to complete, it is recommended to heat a hydride powder above 500 °C (in a CO2containing atmosphere) or, for a short period of time, as high as 600 °C. When such high temperature conditioning is induced (or in the current study case, a spontaneous temperature burst is not suppressed), the full development of the catalyst can be obtained in one cycle within 1 h. Meanwhile, the catalyst development from a Mg2NiH4 powder kept below 500 °C might take more than 20 hydriding/dehydriding cycles [6] or up to 10-fold more time for the reaction to complete [15].

Conclusions
In the current study, various aspects of Mg2NiH4=based catalyst development for a Sabatier type methanation reaction were investigated. Based on the congruous data of reaction on set temperatures and other evidence, we conclude that despite some differences in the crystallization of MgNi2 and Ni phases, both Mg2NiH4 hydride films and powders react with CO2 equally. It was demonstrated that Mg2NiH4 hydride heating above 330-380 °C under a CO2 and H2 gas flow initiates hydride decomposition, disproportionation, and oxidation. These reactions empowered the catalytic properties of the material and promoted the CO2 methanation reaction. However, structural and thermogravimetric analyses revealed that in order to provide the fast and full-scale development of the catalyst (the formation of MgO particles decorated by nanocrystalline Ni), it has to be heated

Conclusions
In the current study, various aspects of Mg 2 NiH 4 =based catalyst development for a Sabatier type methanation reaction were investigated. Based on the congruous data of reaction on set temperatures and other evidence, we conclude that despite some differences in the crystallization of MgNi 2 and Ni phases, both Mg 2 NiH 4 hydride films and powders react with CO 2 equally. It was demonstrated that Mg 2 NiH 4 hydride heating above 330-380 • C under a CO 2 and H 2 gas flow initiates hydride decomposition, disproportionation, and oxidation. These reactions empowered the catalytic properties of the material and promoted the CO 2 methanation reaction. However, structural and thermogravimetric analyses revealed that in order to provide the fast and full-scale development of the catalyst (the formation of MgO particles decorated by nanocrystalline Ni), it has to be heated above 500 • C. Another considerable finding of the study was the confirmation that both high grade and low grade starting Mg 2 Ni alloys could be equally suitable for hydride synthesis and in methanation reactors. Nevertheless, it was also observed that during methanation, such catalysts produced considerable amounts of solid carbon. The cause of the carbonization, as well as its effects on the long time efficiency of the catalysts, will have to be key issues of future research.