Investigation of Catalyst Development from Mg2nih4 Hydride and Its Application for the CO2 Methanation Reaction

coatings

Article Investigation of Catalyst Development from Mg2NiH4 Hydride and Its Application for the CO2 Methanation Reaction

Martynas Lelis *, Sarunas Varnagiris, Marius Urbonavicius and Kestutis Zakarauskas Centre for Hydrogen Energy Technologies, Lithuanian Energy Institute, Breslaujos st. 3, 44403 Kaunas, Lithuania; [email protected] (S.V.); [email protected] (M.U.); [email protected] (K.Z.) * Correspondence: [email protected]; Tel.: +37-06-125-2924

Received: 11 November 2020; Accepted: 29 November 2020; Published: 1 December 2020

Abstract: 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 Mg2NiH4 hydride heating under CO2 and H2 gas flow initiates hydride decomposition, disproportionation and oxidation. These reactions empower catalytic properties of the material and promotes CO2 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 Mg2Ni alloy can be equally suitable for the hydride synthesis and its usage for the promotion of methanation reactions.

Keywords: Mg2NiH4; methanation; catalyst; CO2; Sabatier

1. 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 [4–6]. 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 CO2 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 + CO CH + 2H O ∆H = 165 kJ/mol (1) 2(g) 2(g) → 4(g) 2 (g) 298K − First stage: H + CO CO + H O ∆H = 41 kJ/mol (2) 2(g) 2(g) → (g) 2 (g) 298K

Coatings 2020, 10, 1178; doi:10.3390/coatings10121178 www.mdpi.com/journal/coatings Coatings 2020, 10, 1178 2 of 15

Second stage: 3H + CO CH + H O ∆H = 206 kJ/mol (3) 2(g) (g) → 4(g) 2 (g) 298K − Gao et al. [9] reported that based on free Gibbs energy calculations, the optimal temperature (in respect to CH4 selectivity and COx 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 CO2 methanation they also serve as catalytic drivers for the competing reaction (4) which produces elemental carbon and poisons the catalysts [12]:

Bosch reaction: 2H + CO C + 2H O (4) 2(g) 2(g) → (s) 2 (g) 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 CO2:H2 and CO:H2 gas ratios. The researchers determined that higher gas ratios have positive effects on lowering down carbon production and that preferable CO2:H2 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 CH4 production and do not support the Bosch reaction. In 1990, Selvam et al. [13] investigated the interaction between CO2, hydrogen storage alloys, and compounds including LaNi5, CaNi5, Mg2Ni, Mg2Cu, and FeTi. They reported that during exposure to air, all these compounds form surface oxides and hydroxides that, in turn, actively adsorb atmospheric CO2 and favor the formation of carbonate species on the top few layers of the surface. In a subsequent study of air-exposed Mg2NiH4 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 CO2 on an Mg2NiH4 surface could be beneficial for the Sabatier type methanation reaction. Their comprehensive study on the catalytic interactions of the hydride surface of Mg2NiH4 powder with CO2 provided valuable insights into material disproportionation and oxidation during cyclic hydrogen absorption and hydrogen desorption under CO2 (at temperatures up to 500 ◦C). Based on their findings, during the first few dehydriding cycles in a CO2-containing atmosphere, the simultaneous disproportionation of Mg2NiH4 and the selective oxidation of Mg take place. As a result, a layered structure consisting of Mg2NiH4/Mg2Ni/Ni/MgO was formed. According to the authors, when it was active, Ni helped to dissociate CO2 and CO molecules and promoted methanation [6]. Eventually, after approximately 20 cycles, Mg2NiH4 and Mg2Ni were no longer formed and only MgO and Ni phases were observed. In two more recent studies, Grasso et al. reported experimental data on CO2 methanation processes using as-sintered monoclinic [15] and as-milled cubic Mg2NiH4 [16] powders. In these studies, Mg2NiH4 powders served as the sole hydrogen sources and “providers” of catalytic sites for the promotion of the conversion of CO2 to CH4. 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 CO2 pressure and heated to 400 ◦C at a rate of 10 ◦C/min. The mass of the used Mg2NiH4 powder and the applied CO2 pressure (approximately 1.2 bars [16]) were proportioned in such a way that the calculated total molar quantity of H2 released by the complete decomposition of Mg2NiH4 would ensure an H2:CO2 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 Mg2NiH4‘s reaction with CO2 at 400 ◦C, the complete decomposition of Mg2NiH4 was not reached, although CO2 was totally consumed and Coatings 2020, 10, 1178 3 of 15 no carbon deposition was observed [15]. Considering the intermediate reaction products that were obtained after 1 and 5 h (namely, Mg2NiH4, MgH2, Mg2Ni, MgO, and CO), Grasso et al. concluded that the methanation of CO2 using Mg2NiH4 involves two simultaneous processes: (i) the catalytic conversion of CO2 through reactions (2) and (3) and (ii) the direct reduction of CO2 by the reducing effect of MgH2 (Equation (5)).

2MgH + CO 2MgO + CH (5) 2(s) 2(g) → (s) 4(g)

The chemical activity of the as-milled cubic polymorph form of Mg2NiH4 was found to be considerably higher, and its reaction with CO2 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 CO2 atmosphere, as–milled Mg2NiH4 rapidly decomposes directly to the Ni-Mg2Ni-MgNi2-/MgO (all of this phase remains after the reaction is complete) catalytic system Equation (6) which promotes the methanation reaction Equation (7).

2Mg NiH + CO 0.75Mg Ni + 2MgO + 0.25Ni + 0.5MgNi + C + 4H (6) 2 4(s) 2(g) → 2 (s) (s) (s) 2(s) (s) 2(g) C + 2H CH (7) (s) 2(g) → 4(g) Reports by Kato et al. [6] and Grasso et al. [15,16] provided clear evidence that during its interaction with CO2 gas at 400–500 ◦C, Mg2NiH4 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 Mg2NiH4 hydride [16]. Potentially, an Mg2NiH4 hydride-based CO2 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 Mg2NiH4 powders. Previous studies of Mg2NiH4 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 Mg2NiH4 hydride might be better suited for CO2 methanation catalysis than the other. Accordingly, in the current study, we investigated the structural transformations of Mg2NiH4 hydride films under a CO2 atmosphere and examined whether they correlated with the ones reported for Mg2NiH4 powders. In addition, our interest to investigate Mg2NiH4 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 Mg2NiH4 in powder and investigated its efficiency for the methanation of CO2 in order to compare processes in different forms of Mg2NiH4 hydride and to see how this material develops during the CO2 methanation reaction.

2. Materials and Methods

2.1. Material Synthesis

Mg2NiH4 hydride films were formed by using magnetron sputtering to deposit metallic Mg2Ni films and then hydriding them for 48 h under 20 bars of H2 pressure at a temperature of 250 ◦C. The deposition of Mg Ni films was realized under a 6 10 3 mbar Ar gas atmosphere inside a Kurt J. 2 × − 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 Coatings 2020, 10, 1178 4 of 15 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 Mg2NiH4 hydride [3,19] because it prevents the crystallization of localized MgNi2 phase nanoformations and makes full hydrogenation easier to attain. Mg Ni films were deposited on 20 20 mm fused silica 2 × substrates. The approximate thickness of the films was 500 nm. In previous studies [6,15,16], Mg2NiH4 catalysts were synthesized from high purity Mg2Ni 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 Mg2Ni material (Mg2Ni alloy granules of 99% purity obtained from American Elements) because it better reflected the potential conditions of Mg2NiH4 applications in commercial methanation reactors. At the same time, the usage of a less pure Mg2Ni 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 Mg2Ni 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 Mg2Ni powders [20,21] was achieved by applying 4 hydriding (16 h, 250 ◦C, and 20 bar H2) and dehydriding cycles (8 h, 250 ◦C, vacuum). Final hydriding was conducted for 24 h under 20 bars of H2 pressure at 250 ◦C. The specific surface areas of the unhydrided and hydrided Mg2Ni powders were estimated by the BET method and reached 1.59 and 2.69 m2/g, respectively.

2.2. In-Situ XRD

The in situ XRD characterization of Mg2NiH4 films during annealing under a CO2 atmosphere was performed with a Bruker D8 Discover (Hamburg, Germany) equipped with Mri TC-basic (Hamburg, Germany) chamber. For this type of experiment, Mg2NiH4 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 CO2 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 Mg2NiH4 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.

2.3. 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. Coatings 2020, 10, 1178 5 of 15

2.4. 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 Al2O3 TGA crucibles. For each measurement, 8.4–8.8 mg samplesCoatings 2020 were, 10, used.x FOR PEER REVIEW 5 of 15

2.5.2.5. Methanation CO methanation experiments were carried out with the custom build setup presented in Figure1. CO22 methanation experiments were carried out with the custom build setup presented in Figure For each CO methanation experiment, 5 g of hydride powder were mixed with 5 g of fine Al O 1. For each CO2 2 methanation experiment, 5 g of hydride powder were mixed with 5 g of fine Al22O33 powderpowder (Sigma-Aldrich,(Sigma‐Aldrich, 98%98% purity)purity) andand placedplaced insideinside aa cylindrical, high-temperature,high‐temperature ,stainless stainless steelsteel containercontainer (approximate(approximate lengthlength ofof 8080 mmmm andand internalinternal diameterdiameter ofof 1515 mm).mm). FromFrom bothboth ends,ends, powderpowder waswas closely packed by by high high purity purity quartz quartz wool wool plugs. plugs. The The container container was was heated heated up b up ay b vertical ay vertical high hightemperature temperature tubular tubular furnace, furnace, and and the thetemperature temperature of ofthe the powder powder was was measured measured by by an internalinternal thermocouplethermocouple probe probe that that was was inserted inserted into into the middlethe middle of the of powder. the powder. The gas The supply gas system supply consisted system of three mass flow controllers (MFCs) dedicated to CO ,H , and Ar gases. Flows from all MFC units consisted of three mass flow controllers (MFCs) dedicated2 2 to CO2, H2, and Ar gases. Flows from all wereMFC controlled units were by controlled one digital by gas one controller digital gas manufactured controller manufactured by Brooks (Seattle, by Brooks WA, USA). (Seattle, The WA, controller USA). was set up to maintain a total gas flow of 1 L/min while keeping the H :CO gas flow ratio at 6:1 The controller was set up to maintain a total gas flow of 1 L/min while keeping2 2 the H2:CO2 gas flow (followingratio at 6:1 the(following suggestion the suggestion by [9]). An by automatic [9]). An automatic constant pressure constant regulatorpressure regulator outside of outside the reaction of the vesselreaction was vessel used was to maintain used to maintain a constant a pressure constant inside pressure the inside reaction the zone. reaction A test zone. pressure A test of pressure 1 bar was of selected1 bar was to selected mimic the to mimic potential the reaction potential conditions reaction conditions of the methanation of the methanation reactor without reactor any without additional any technicaladditional measures, technical whereas measures, 10 whereas bars of pressure 10 bars of were pressure tested were in order tested to in limit order the to formation limit the offormation carbon, asof suggestedcarbon, as bysuggested [12]. by [12].

FigureFigure 1.1. ExperimentalExperimental schemescheme ofof COCO22 methanation set up.

TheThe supplementarysupplementary inertinert ArAr gasgas waswas usedused to maintain the recommended gas ﬂowflow requirements ofof thethe on-lineon‐line gasgas analyzeranalyzer (VISIT(VISIT 03H03H manufacturedmanufactured byby MesstechnikMesstechnik EHEIMEHEIM GmbH)GmbH) andand toto transfertransfer 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 on‐line gas analyzer.

3. Results and Discussions

As is typical for metallic films, the as‐deposited Mg2Ni 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 Mg2NiH4 [22–24]. The prevalence of the

Coatings 2020, 10, 1178 6 of 15

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 on-line gas analyzer.

3. Results and Discussions

As is typical for metallic ﬁlms, the as-deposited Mg2Ni coatings had mirror like appearances. After hydriding, they became transparent and had a bright orange hue. Both of these features are Coatings 2020, 10, x FOR PEER REVIEW 6 of 15 characteristic for the low temperature (LT) phase of Mg2NiH4 [22–24]. The prevalence of the monoclinic

LTmonoclinic phase of LT Mg phase2NiH 4ofwas Mg subsequently2NiH4 was subsequently conﬁrmed confirmed by XRD (Figure by XRD2). In(Figure addition 2). In to addition the LT phase, to the smallLT phase, fractions small of fractions the so-called of the pseudo-cubic so‐called pseudo high temperature‐cubic high temperature (HT) phase of (HT) Mg 2phaseNiH4 [of17 ,Mg25]2NiH were4 observed[17,25] were at room observed temperature. at room temperature.

FigureFigure 2.2. In-situIn‐situ XRDXRD measuredmeasured duringduring MgMg22NiH4 filmﬁlm heating under the CO 2 atmosphere.atmosphere.

During the in-situin‐situ heatingheating ofof thethe MgMg22NiH4 filmfilm under the COCO22 atmosphere (Figure 22),), several several temperature ranges related toto substantial compositioncomposition changeschanges couldcould be identified.identified. Starting from the room temperaturetemperature up up till till 240 240◦ C,°C, the the LT LT and and HT HT phases phases of Mg of Mg2NiH2NiH4 co-existed,4 co‐existed, and and their their ratio ratio in the in film the didfilm not did change not change noticeably. noticeably. Above Above 240 ◦C, 240 the °C, fraction the fraction of LT phaseof LT graduallyphase gradually decreased, decreased, and at 300and◦ C,at only300 °C, the only HT phase the HT of Mgphase2NiH of4 persisted.Mg2NiH4 persisted. The initiation The of initiation Mg2NiH 4ofdecomposition Mg2NiH4 decomposition was observed was at 350–360observed◦ Cat when 350–360 weak °C peakswhen attributedweak peaks to attributed Mg2Ni arose. to Mg At2Ni 410 arose.◦C, the At transition 410 °C, the from transition Mg2NiH from4 to Mg2NiHNi was4 to complete,Mg2Ni was and complete, at 420 ◦C, and only at peaks 420 °C, of Mgonly2Ni peaks were of observed. Mg2Ni were A further observed. increase A further of the temperatureincrease of the was temperature followed by thewas gradual followed disproportionation by the gradual of disproportionation Mg2Ni and the crystallization of Mg2Ni and of MgO the oxide.crystallization The crystalline of MgO MgO oxide. phase The crystalline was at first MgO observed phase at was 450 at◦ Cfirst and observed remained at 450 as the °C mainand remained phase at theas the end main of the phase measurement at the end at of 500 the ◦measurementC. at 500 °C. Surface images made by SEM and the near surface region depth profileprofile recorded by XPS complemented thethe findings findings of of in-situ in‐situ XRD XRD and highlightedand highlighted structural structural changes changes of the film of the that film occurred that dueoccurred to the due heating to the in theheating CO2 atmosphere.in the CO2 atmosphere. More specifically, More the specifically, SEM images the (Figure SEM images3) show (Figure that after 3) heatingshow that in theafter CO heating2 atmosphere, in the CO the2 atmosphere, surface of the the films surface became of the considerably films became rougher, considerably which indicated rougher, significantwhich indicated mass transfersignificant processes mass transfer within theprocesses film. At within the same the time,film. At an the XPS same depth time, profile an (FigureXPS depth3d) specifiedprofile (Figure that the 3d) Ni specified concentration that the at Ni the concentration surface was less at the than surface 1/6 of was Mg. less This than was 1/6 clear of Mg. evidence This was that duringclear evidence the disproportionation that during the ofdisproportionation Mg2Ni, magnesium of Mg separated2Ni, magnesium from Ni, oxidized,separated and from had Ni, a oxidized, tendency toand segregate had a tendency at the near to segregate surface region. 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 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 Mg2NiH4 powders that were cyclically exposed to a CO2 atmosphere at a temperature of 500 °C. Furthermore, several tests with Mg2NiH4 powder that had an MgNi2 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 MgNi2 and Ni phases in in‐situ XRD data was the result of insufficient crystallization rather than different reaction pathways.

Coatings 2020, 10, 1178 7 of 15

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 substantial proof for that. For example, the observed quantitative concentration values, as well as their qualitative changes at the surface of the film (Figure3d) in great proximity, resembled the data presented by Kato et al. [6], who measured the XPS depth profile of Mg2NiH4 powders that were cyclically exposed to a CO2 atmosphere at a temperature of 500 ◦C. Furthermore, several tests with Mg2NiH4 powder that had an MgNi2 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 MgNi2 and Ni phases in in-situ XRD data wasCoatings the 2020 result, 10, x of FOR insu PEERfficient REVIEW crystallization rather than different reaction pathways. 7 of 15

Figure 3. SEM surface images of Mg–Ni ﬁlms:films: ( (aa)) as as-deposited,‐deposited, ( (bb)) hydrided, and ( c)) after it was

analyzed by in-situin‐situ XRD in the CO2 atmosphere. ( (dd)) XPS XPS near near surface surface region region depth profile proﬁle of the ﬁlmfilm presented in (cc).).

XRD patterns of commercialcommercial MgMg22Ni powders that contained Mg andand MgNiMgNi22 components are provided in Figure Figure 44a.a. After After hydriding hydriding these these powders, powders, the the crystal crystal phases phases of ofMg Mg2NiH2NiH4, MgH4, MgH2, and2, andMgNi MgNi2 were2 were obtained obtained (Figure (Figure 4b).4 b).The The XRD XRD patterns patterns of of the the hydrided hydrided powders powders had had no peakspeaks ofof unreacted MgMg and and Mg Mg2Ni2Ni phases, phases, so weso assumewe assume that the that loss the of hydrogenloss of hydrogen concentration concentration (see DSC/ TGA(see analysesDSC/TGA below) analyses was below) caused was by thecaused aggregative by the aggregative effect of hydrogen effect of non-adsorbing hydrogen non phases‐adsorbing like MgNiphases2. MgOlike MgNi was2 not. MgO observed was not by observed XRD, but by we XRD, nevertheless but we nevertheless assume that assume it also that contributed it also contributed to the lower to hydrogenthe lower concentration hydrogen concentration because commercial because Mg–Ni commercial grains wereMg–Ni delivered grains inwere an air-containing delivered in canisteran air‐ andcontaining had successive canister and handling had successive in air. handling in air. The DSC analysis of hydrided powders under Ar flow indicated the presence of two distinct thermal events (Figure5a). 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 Mg NiH . Proceeding further, most of the sample mass ( 2.6 wt.%) was lost during 2 4 ≈ 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.

Figure 4. XRD patterns of commercial Mg2Ni alloy powders: (a) unhydrided and (b) hydrided.

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.

Coatings 2020, 10, x FOR PEER REVIEW 7 of 15

Figure 3. SEM surface images of Mg–Ni films: (a) as‐deposited, (b) hydrided, and (c) after it was

analyzed by in‐situ XRD in the CO2 atmosphere. (d) XPS near surface region depth profile of the film presented in (c).

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 Coatingsthe lower2020 , hydrogen10, 1178 concentration because commercial Mg–Ni grains were delivered in an8 ofair 15‐ containing canister and had successive handling in air.

FigureFigure 4.4. XRDXRD patternspatterns ofof commercialcommercial MgMg22Ni alloy powders: (a) unhydrided and (b) hydrided. Coatings 2020, 10, x FOR PEER REVIEW 8 of 15 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.

Figure 5.5. DiDifferentialﬀerential scanningscanning calorimetrycalorimetry (DSC)(DSC) andand TGATGA chartscharts ofof hydridedhydrided MgMg22Ni alloy powders under (aa)) Ar andand ((bb)) ArAr andand COCO22 gas ﬂows.flows.

The firstfirst thermalthermal eventevent duringduring hydridedhydrided powderpowder heatingheating underunder thethe ArAr andand COCO22 gas flowflow started withwith thethe LT–HT LT–HT phase phase transition transition at 235at 235◦C (Figure°C (Figure5b). A5b). nearly A nearly identical identical onset temperatureonset temperature and width and ofwidth the Mg of 2theNiH Mg4 phase2NiH4 transition phase transition range (235–300 range (235–300◦C) were °C) also were observed also observed for the Mg for2NiH the4 Mgfilm2NiH (Figure4 film2) and(Figure powder 2) and samples powder (Figure samples5a) that(Figure were 5a) tested that underwere tested CO 2 andunder Ar CO gas2 atmospheres,and Ar gas atmospheres, respectively. Thisrespectively. indicated This that belowindicated the phasethat below transition the phase temperature, transition the temperature, activation barrier the foractivation Mg2NiH barrier4 hydride for decompositionMg2NiH4 hydride and decomposition/or oxidation was and/or relatively oxidation high, was and relatively at near-atmosphere high, and pressures, at near‐atmosphere Mg2NiH4 hydridepressures, demonstrated Mg2NiH4 hydride enduring demonstrated stability regardless enduring of itsstability form and regardless surrounding of its form gas phase and surrounding composition. gas phaseDisparities composition. between Mg2NiH4 hydride behavior under the inert and reactive gases arose soon afterDisparities start of the between LT–HT phase Mg2NiH transition.4 hydride Namely, behavior under under the the CO inert2 and and Ar reactive gas flow, gases there arose were soon no massafter start change of the events LT–HT up tophase approximately transition. Namely, 330 ◦C. An under analysis the CO of2 the and results Ar gas of flow, the in-situ there were XRD no of Mgmass2NiH change4 film events heating up under to approximately CO2 gas (Figure 3302) suggested°C. An analysis that at of later the temperatures, results of the the in dehydriding‐situ XRD of ofMg Mg2NiH2NiH4 film4 phase heating should under have CO taken2 gas (Figure place. The 2) suggested decomposition that at reaction later temperatures, of Mg2NiH4 thewas dehydriding expected to resultof Mg2 inNiH approximately4 phase should 2.7 have wt.% taken mass place. loss and The a decomposition strong endothermic reaction minimum of Mg2NiH in the4 was TGA expected and DSC to curves, respectively (Figure5a). Instead, under the CO and Ar gas flow, a prolonged small ( 0.1 wt.%) result in approximately 2.7 wt.% mass loss and a strong2 endothermic minimum in the TGA≈ and DSC declinecurves, ofrespectively mass and a(Figure low intensity 5a). Instead, exothermic under upswing the CO2 was and observed. Ar gas flow, Therefore, a prolonged it is clear small that (≈ the0.1 endothermalwt.%) decline reaction of mass (8) and [28 a] waslow simultaneouslyintensity exothermic accompanied upswing by was some observed. additional Therefore, reactions. it is clear that the endothermal reaction (8) [28] was simultaneously accompanied by some additional reactions. Mg2NiH4(s) Mg2Ni(s) + 2H2(g) ∆H298K = 63.5 kJ/mol (8) Mg2NiH4(s) →→ Mg2Ni(s) + 2H2(g) ΔH298K =− −63.5 kJ/mol (8)

COCO(g) ++ H2(g) → CC(s) ++ HH2OO(g) Δ∆HH298K= = −131.3 kJkJ/mol/mol (9) (9) (g) 2(g) → (s) 2 (g) 298K − 2CO(g) → C(s) + CO2(g) ΔH298K = −172.4 kJ/mol (10)

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 minimum in the DSC curve. However, the CO2 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 Mg2Ni 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 Mg2NiH4. To reaffirm this view, we note that the TGA curve reached its minimum slightly above 400 °C, which was consistent with the disappearance of Mg2NiH4 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

Coatings 2020, 10, 1178 9 of 15

2CO C + CO ∆H = 172.4 kJ/mol (10) (g) → (s) 2(g) 298K − 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 minimum in the DSC curve. However, the CO2 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 Mg2Ni 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 Mg2NiH4. To reaffirm this view, we note that the TGA curve reached its minimum slightly above 400 ◦C, which was consistent with the disappearance of Mg2NiH4 XRD peaks at 410–420 ◦C (Figure2). 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 (Figure5a). 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 Figure6a,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 (Figure5a) and was attributed to the plateau pressure levels of Mg 2NiH4 [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 Figure5b and comments above). The second thing to notice from Figure6 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 CO2 or intermediate products from reaction (2), namely CO. A nearly 100% CO2 conversion was maintained up to approximately 530 ◦C, with an estimated CH4 yield of approximately 75%. On the other hand, methanation at 1 bar of pressure was able to reach a 100% CO2 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 Coatings 2020, 10, x FOR PEER REVIEW 9 of 15

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 Coatingsthe hydrogen2020, 10, 1178desorption in the last section of the sample heating under Ar (Figure 5a). Accordingly,10 of 15 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 90%decomposition at 500 ◦C) is and usually maintained kinetically a downtrend limited [31] as and temperature the heating rose rate to during 550 ◦C. DSC/TGA At both analysis investigated was pressures,relatively high CO2 (10methanation °C/min); (ii) by the the dehydriding reaction product temperature of Mg2 NiHis known4 powder to be had postponed lower CH by4 surfaceyields than were achieved with standard commercial Ni-based catalysts at 10 bar (H2:CO2 ratio of 4:1) and oxides [15,32]. After MgH2 decomposition, the powder reaction with CO2 was finalized by some more 30thermal bar (H events,2:CO2 ratio thus ofrepresenting 6:1) [9]. When rapid operating oxidation at 10of barsMg that of pressure, was produced we did by not the detect dehydriding CO or other of carbon-containing species in gaseous reaction products; therefore, we assumed that relative difference MgH2 and continued the disproportionation of Mg2Ni and, eventually, MgNi2. between initial CO2 and CH4 product fluxes qualitatively reflected the actual solid C yields. This was The CO2 methanation reaction dynamics at 1 and 10 bars of gas pressure are provided in Figure quite6a,b, respectively. surprising because Both charts theoretical have temperature and experimental peaks that study correspond by Gao et to al. the [9] fast estimated decomposition that at the of specifically used gas pressure and H2:CO2 ratios, carbonization should have been prohibited. The cause hydrides (Mg2NiH4 and MgH2) and the oxidation of Mg2Ni disproportionation products. After the ofheat such burst, a high we carbonfound a yield new hascatalyst not yet with been substantially identified anddifferent will be properties. subject of our future studies.

Figure 6. CO2 methanationmethanation dynamics: ( a) at 1 bar and ( b) at 10 bars of gas pressure. Hydrided MgMg22Ni alloy powders were usedused asas thethe startingstarting catalystcatalyst material.material.

ABy study comparing by Kato methanation et al. [6] demonstrated dynamics at that different the properties pressures, of powder it could catalysts be noticed can that be significantly at 1 bar in a improvedlow temperature after they region are cycled (prior throughto the temperature H2 (25 bar and burst), 350 ◦thereC), CO was2 (1 a bar slight and increase 500 ◦C), in and H vacuum2 (and a (500decrease◦C). in They CO reported2) concentration, that after whereas 18 cycles, at 10 the bars, methanation the H2:CO process2 ratio remained was initiated stable. at 250–260At 1 bar◦ C,of whereaspressure, an such uncycled behavior high was purity not surprising Mg2NiH4 catalyst because initiated a similar a hydrogen methanation release reaction was at also approximately observed in 330–340the TGA ◦patternC (in comparison (Figure 5a) toand the was current attributed study to with the plateau a relatively pressure low gradelevels Mgof Mg2NiH2NiH4 catalyst4 [33,34]. that On initiatedthe other methanation hand, Mg2NiH at 3304 stability◦C for 1at bar 10 ofbars pressure of pressure and 380 demonstrated◦C for 10 bars that of pressure).high partial hydrogen pressureAfter was considering a stronger the factor potential for the benefits stabilization of catalyst of Mg cycling2NiH4 than after the the oxidative first methanation potential of run-up CO2, ateven 10 barsthough of pressureat low pressure was finished, Mg2NiH we4 decomposition gradually decreased and partial furnace oxidation temperature happened down at to the 250 same◦C. Duringtime (see the Figure cool-down 5b and periodcomments and above). 30 min after the target temperature was reached, the container was constantlyThe second flushed thing to with notice hydrogen. from Figure Subsequently, 6 is that at 10 the bars container of pressure, was the filled methanation up with hydrogen reaction tostarted 10 bars at approximately of pressure and 50 confined °C lower for(330 16 °C h versus at 250 380◦C. °C), The and next as day, soon we as repeated temperature the methanationreached 420– process (Figure7) with same parameters as were used before (10 bars of pressure and H :CO ratio of 2 2 6:1). This time, the methanation reaction started immediately at 250 ◦C (close to the lowest reported values [35,36]), reached a 100% CO2 conversion at 320–330 ◦C, and continued without the significant deterioration of CO2 conversion up to the maximum tested temperature of 600 ◦C. The CH4 yield at Coatings 2020, 10, 1178 11 of 15

500 ◦C was slightly higher and reached just over 80%. Looking at the methanation dynamics (Figure7a), one can notice that there was no temperature burst or hydrogen concentration increase that could be attributed to the decomposition of Mg2NiH4 or MgH2 hydride. This suggested that during the ﬁrst methanation experiment, Mg2NiH4 and MgH2 hydrides completely disproportionated, and only the MgO and Ni phases were formed. Accordingly, such a powder catalyst did not adsorb any additional hydrogenCoatings 2020 and, 10, x was FOR able PEER to REVIEW provide an enhanced performance after one just methanation cycle.11 of 15

FigureFigure 7.7. Methanation withwith thethe developeddeveloped powderpowder catalyst:catalyst: ((aa)) productproduct fractionfraction ofof COCO22 methanationmethanation atat didifferentﬀerent temperaturestemperatures and and ( (bb)) methanation methanation dynamics. dynamics.

AA comparisoncomparison of of SEM SEM images images of theof the as-ground as‐ground Mg2 NiMg alloy2Ni alloy grains grains (Figure (Figure8b) and 8b) the and final the catalyst final powdercatalyst (Figurepowder8c) (Figure provided 8c) some provided more evidencesome more regarding evidence how regarding specific features how specific of hydrides features could of becomehydrides beneficial could become for the beneficial formation for of the effi formationcient catalysts of efficient for the catalysts methanation for the reaction.methanation The reaction. initially groundThe initially Mg2 Niground alloy Mg grains2Ni werealloy relativelygrains were course, relatively and theircourse, size and ranged their from size ranged 10 to 50 fromµm. 10 During to 50 hydrideμm. During formation hydride (hydriding formation/dehydriding (hydriding/dehydriding cycling), the catalystcycling), material the catalyst underwent material huge underwent volume changeshuge volume (up to changes 30% [ 37(up]) thatto 30% introduced [37]) that largeintroduced amount large of structuralamount of defects,structural widened defects, upwidened grain boundaries,up grain boundaries, and ultimately and ultimately fractured initialfractured grains initial into grains sub-micrometer-size into sub‐micrometer fine powders‐size fine (Figure powders8c). The(Figure first 8c). temperature The first temperature run-up under run CO‐up2 underand H 2COgas2 and flow H promoted2 gas flow promoted CO2 methanation CO2 methanation and initiated and hydrideinitiated decomposition hydride decomposition and oxidation. and These oxidation. highly energetic These highly chemical energetic reactions chemical empowered reactions Mg2Ni disproportionationempowered Mg2Ni anddisproportionation MgO–Ni phase and separation MgO–Ni (Figure phase9). separation As the result, (Figure MgO 9). particles As the result, decorated MgO byparticles nanocrystalline decorated Ni by were nanocrystalline formed. Such Ni a catalystwere formed. is able toSuch promote a catalyst CH4 isformation able to atpromote a relatively CH4 lowformation temperature at a relatively (250 ◦C) low and temperature an do this with (250 a °C) significantly and an do higher this with CH a4 significantlyoutput than higher that not CH oxidized4 output hydridethan that powder. not oxidized The hydride superior powder. catalytic The properties superior catalytic of disproportionated properties of disproportionated MgO–Ni systems MgO–Ni can be attributedsystems can to be the attributed role of MgO to the because role of MgOMgO doesbecause not MgO just workdoes not as an just inert work support as an inert for catalytic support Nifor nanoparticlescatalytic Ni nanoparticles but also actively but also changes actively the methanationchanges the methanation pathway for Nipathway [38]. For for example, Ni [38]. For studies example, have shownstudies thathave MgO shown enhances that MgO the enhances adsorption the and adsorption activation and of activation CO2 [39]. of On CO the2 [39]. other On hand, the other MgO hand, was alsoMgO found was also to promote found to H promote2O formation H2O formation and its desorption and its desorption from a catalyst from a catalyst [15]. [15].

Coatings 2020, 10, 1178 12 of 15 Coatings 2020, 10, x FOR PEER REVIEW 12 of 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 CO2-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 Mg2NiH4 powder kept below 500 ◦C might take more thanCoatings 20 2020 hydriding, 10, x FOR/dehydriding PEER REVIEW cycles [6] or up to 10-fold more time for the reaction to complete12 of [15 15].

Figure 8. Optical and SEM images: (a) as received Mg2Ni alloy grains, (b) Mg2Ni grains after grinding,

(c) catalyst–Al2O3 mixture after methanation test, and (d) EDS elemental mapping of catalyst–Al2O3 mixture after methanation test.

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 CO2‐ 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),Figure 8. OpticaltheOptical full and development SEM images: of ( ( aathe)) as as catalystreceived received can Mg2 NibeNi alloyobtained alloy grains, grains, in ( (onebb)) Mg Mg cycle22NiNi grains grainswithin after after 1 h. grinding, grinding, Meanwhile, the catalyst(c) catalyst–Alcatalyst–Al development22O3 mixturemixture from after a methanation Mg2NiH4 powder test, and kept(d)) EDSEDS below elementalelemental 500 mapping°Cmapping might ofof take catalyst–Al catalyst–Al more 22thanOO33 20 hydriding/dehydridingmixture after methanation cycles test. [6] or up to 10‐fold more time for the reaction to complete [15].

FigureFigure 9.9. XRDXRD patternpattern ofof catalyst–Alcatalyst–Al22O3 mixture after methanation test.

4. Conclusions In the current study, various aspects of Mg NiH =based catalyst development for a Sabatier In the current study, various aspects of Mg2NiH2 4=based4 catalyst development for a Sabatier type typemethanation methanation reaction reaction were investigated. were investigated. Based on Based the congruous on the congruous data of reaction data on of set reaction temperatures on set temperatures and other evidence, we conclude that despite some differences in the crystallization and other evidence, we conclude that despite some differences in the crystallization of MgNi2 and Ni of MgNi and Ni phases, both Mg NiH hydride films and powders react with CO equally. It was phases, both2 Mg2NiH4 hydride films2 and4 powders react with CO2 equally. It was demonstrated2 that demonstrated that Mg NiH hydride heating above 330–380 C under a CO and H gas flow initiates Mg2NiH4 hydride heating2 4 above 330–380 °C under a CO◦ 2 and H2 gas2 flow 2 initiates hydride Figure 9. XRD pattern of catalyst–Al2O3 mixture after methanation test. hydridedecomposition, decomposition, disproportionation, disproportionation, and oxidation. and oxidation. These These reactions reactions empowered empowered the the catalytic properties of the material and promoted the CO2 methanation reaction. However, structural and 4.properties Conclusions 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 catalystIn the current (the formation study, various of MgO aspects particles of Mg decorated2NiH4=based by nanocrystallinecatalyst development Ni), it for has a Sabatierto be heated 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

Coatings 2020, 10, 1178 13 of 15 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 Mg2Ni 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.

Author Contributions: Conceptualization, M.L., M.U., and S.V.; methodology, M.U., S.V., and K.Z.; formal analysis, M.L.; investigation, M.U., S.V., and K.Z.; writing—original draft preparation, M.L.; writing—review and editing, M.U. and S.V.; visualization, M.U. and S.V. All authors have read and agreed to the published version of the manuscript. Funding: This research has received funding from European Regional Development Fund (project No. 01.2.2-LMT- K-718-01-0005) under grant agreement with the Research Council of Lithuania (LMTLT). Acknowledgments: Special thanks goes to the rest project team members: project manager Nerijus Striugas;¯ investigators: Andrius Tamošiunas,¯ Rolandas Paulauskas, Vilma Snapkauskiene.˙ Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. European Commission. The European Green Deal. Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee and the Committee of the Regions. COM/2019/640 Final; Secretariat-General (European Commission): Brussels, Belgium, 2019. 2. Eurostat. Renewable Energy Statistics. Share of Energy from Renewable Sources in the EU Mamber States. 2018. Available online: https://ec.europa.eu/eurostat/statistics-explained/index.php/Renewable_energy_statistics (accessed on 9 October 2020). 3. Gallandat, N.; Bérard, J.; Abbet, F.; Züttel, A. Small-scale demonstration of the conversion of renewable energy to synthetic hydrocarbons. Sustain. Energy Fuels 2017, 1, 1748–1758. [CrossRef] 4. Züttel, A.; Mauron, P.; Kato, S.; Callini, E.; Holzer, M.; Huang, J. Storage of Renewable Energy by Reduction

of CO2 with Hydrogen. Chimia 2015, 69, 264–268. [CrossRef][PubMed] 5. Gahleitner, G. Hydrogen from renewable electricity: An international review of power-to-gas pilot plants for stationary applications. Int. J. Hydrog. Energy 2013, 38, 2039–2061. [CrossRef] 6. Kato, S.; Borgschulte, A.; Ferri, D.; Bielmann, M.; Crivello, J.C.; Wiedenmann, D.; Parlinska-Wojtan, M.; Rossbach, P.;

Lu, Y. CO2 hydrogenation on a metal hydride surface. Phys. Chem. Chem. Phys. 2012, 14, 5518–5526. [CrossRef] [PubMed]

7. Miao, B.; Ma, S.S.K.; Wang, X.; Su, H.; Chan, S.H. Catalysis mechanisms of CO2 and CO methanation. Catal. Sci. Technol. 2016, 6, 4048–4058. [CrossRef]

8. Stangeland, K.; Kalai, D.; Li, H.; Yu, Z. CO2 Methanation: The Eﬀect of Catalysts and Reaction Conditions. Energy Procedia 2017, 105, 2022–2027. [CrossRef] 9. Gao, J.; Wang, Y.; Ping, Y.; Hu, D.; Xu, G.; Gu, F.; Su, F. A thermodynamic analysis of methanation reactions of carbon oxides for the production of synthetic natural gas. RSC Adv. 2012, 2, 2358–2368. [CrossRef] 10. Brooks, K.P.; Hu, J.; Zhu, H.; Kee, R.J. Methanation of carbon dioxide by hydrogen reduction using the Sabatier process in microchannel reactors. Chem. Eng. Sci. 2007, 62, 1161–1170. [CrossRef] 11. Jessop, P.G. Greenhouse Gas Carbon Dioxide Mitigation: Science and Technology By Martin M. Halmann (Weizmann Institute of Science, Israel) and Meyer Steinberg (Brookhaven National Laboratory). Lewis Publishers: Boca Raton, FL. 1999. xix + 568 pp. $99.95. ISBN 1-5667. J. Am. Chem. Soc. 2001, 123, 7197. [CrossRef] 12. Jürgensen, L.; Ehimen, E.A.; Born, J.; Holm-Nielsen, J.B. Dynamic biogas upgrading based on the Sabatier process: Thermodynamic and dynamic process simulation. Bioresour. Technol. 2015, 178, 323–329. [CrossRef] 13. Selvam, P.; Viswanathan, B.; Srinivasan, V.Evidence for the formation of surface carbonates on some hydrogen storage intermetallic compounds: An XPS study. Int. J. Hydrog. Energy 1990, 15, 133–137. [CrossRef]

14. Selvam, P.; Viswanathan, B.; Srinivasan, V. The inﬂuence of atmospheric CO2 on the surface properties of Mg2NiH4 and a comparison with some hydrogen storage alloys. J. Less Common Met. 1990, 158, L1–L7. [CrossRef] Coatings 2020, 10, 1178 14 of 15

15. Grasso, M.L.; Puszkiel, J.; Fernández Albanesi, L.; Dornheim, M.; Pistidda, C.; Gennari, F.C. CO2 reutilization for methane production via a catalytic process promoted by hydrides. Phys. Chem. Chem. Phys. 2019, 21, 19825–19834. [CrossRef][PubMed]

16. Grasso, M.L.; Puszkiel, J.; Gennari, F.C.; Santoru, A.; Dornheim, M.; Pistidda, C. CO2 reactivity with Mg2NiH4 synthesized by in situ monitoring of mechanical milling. Phys. Chem. Chem. Phys. 2020, 22, 1944–1952. [CrossRef][PubMed] 17. Lelis, M.; Milcius, D.; Wirth, E.; Hålenius, U.; Eriksson, L.; Jansson, K.; Kadir, K.; Ruan, J.; Sato, T.;

Yokosawa, T.; et al. A mechanically switchable metal–insulator transition in Mg2NiH4 discovers a strain sensitive, nanoscale modulated resistivity connected to a stacking fault. J. Alloys Compd. 2010, 496.[CrossRef] 18. Pranevicius, L.; Wirth, E.; Milcius, D.; Lelis, M.; Pranevicius, L.L.; Bacianskas, A. Structure transformations and hydrogen storage properties of co-sputtered MgNi films. Appl. Surf. Sci. 2009, 255.[CrossRef] 19. Lelis, M.; Milcius, D.; Noréus, D. Substrate effects on formation and hydrogenation of Mg-Ni films. Appl. Surf. Sci. 2012, 263.[CrossRef] 20. Li, L.; Akiyama, T.; Kabutomori, T.; Terao, K.; Yagi, J. Hydriding and dehydriding behavior of the product in

hydriding combustion synthesis of Mg2NiH4. J. Alloys Compd. 1999, 287, 98–103. [CrossRef] 21. Li, L.; Akiyama, T.; Yagi, J. Activation behaviors of Mg2NiH4 at different hydrogen pressures in hydriding combustion synthesis. Int. J. Hydrog. Energy 2001, 26, 1035–1040. [CrossRef] 22. Rašković-Lovre, Ž.; Mongstad, T.T.; Karazhanov, S.; You, C.C.; Lindberg, S.; Lelis, M.; Milcius, D.; Deledda, S. Annealing-induced structural rearrangement and optical band gap change in Mg-Ni-H thin films. Mater. Res. Express 2017, 4.[CrossRef]

23. Blomqvist, H.; Noréus, D. Mechanically reversible conductor–insulator transition in Mg2NiH4. J. Appl. Phys. 2002, 91, 5141–5148. [CrossRef] 24. Gremaud, R.; van Mechelen, J.L.M.; Schreuders, H.; Slaman, M.; Dam, B.; Griessen, R. Structural and optical properties of MgyNi1–yHx gradient thin ﬁlms in relation to the as-deposited metallic state. Int. J. Hydrog. Energy 2009, 34, 8951–8957. [CrossRef]

25. Yamamoto, S.; Fukai, Y.; Rönnebro, E.; Chen, J.; Sakai, T. Structural changes of Mg2NiH4 under high hydrogen pressures. J. Alloys Compd. 2003, 356–357, 697–700. [CrossRef]

26. Noréus, D.; Werner, P.-E. The structure of the low temperature phase Mg2NiH4(LT). Mater. Res. Bull. 1981, 16, 199–206. [CrossRef]

27. Zolliker, P.; Yvon, K.; Baerlocher, C.H. Low-temperature structure of Mg2NiH4: Evidence for microtwinning. J. Less Common Met. 1986, 115, 65–78. [CrossRef]

28. Pivak, Y.; Palmisano, V.; Schreuders, H.; Dam, B. The clamping effect in the complex hydride Mg2NiH4 thin films. J. Mater. Chem. A 2013, 1, 10972–10978. [CrossRef] 29. Mills, G.A.; Steffgen, F.W. Catalytic Methanation. Catal. Rev. 1974, 8, 159–210. [CrossRef] 30. Nahar, G.A.; Madhani, S.S. Thermodynamics of hydrogen production by the steam reforming of butanol: Analysis of inorganic gases and light hydrocarbons. Int. J. Hydrog. Energy 2010, 35, 98–109. [CrossRef]

31. Evard, E.; Gabis, I.; Yartys, V.A.Kinetics of hydrogen evolution from MgH2: Experimental studies, mechanism and modelling. Int. J. Hydrog. Energy 2010, 35, 9060–9069. [CrossRef] 32. Borgschulte, A.; Gremaud, R.; Griessen, R. Interplay of diﬀusion and dissociation mechanisms during hydrogen absorption in metals. Phys. Rev. B 2008, 78, 94106. [CrossRef]

33. Li, L.; Akiyama, T.; Yagi, J. Activity and capacity of hydrogen storage alloy Mg2NiH4 produced by hydriding combustion synthesis. J. Alloys Compd. 2001, 316, 118–123. [CrossRef] 34. Saita, I.; Li, L.; Saito, K.; Akiyama, T. Pressure-Composition-Temperature Properties of Hydriding

Combustion-Synthesized Mg2NiH4. Mater. Trans. 2002, 43, 1100–1104. [CrossRef] 35. Rönsch, S.; Schneider, J.; Matthischke, S.; Schlüter, M.; Götz, M.; Lefebvre, J.; Prabhakaran, P.; Bajohr, S. Review on methanation—From fundamentals to current projects. Fuel 2016, 166, 276–296. [CrossRef]

36. Schaaf, T.; Grünig, J.; Schuster, M.R.; Rothenﬂuh, T.; Orth, A. Methanation of CO2—storage of renewable energy in a gas distribution system. Energy Sustain. Soc. 2014, 4, 2. [CrossRef] 37. Ono, S.; Ishido, Y.; Imanari, K.; Tabata, T.; Cho, Y.K.; Yamamoto, R.; Doyama, M. Phase transformation and thermal expansion of Mg Ni alloys in a hydrogen atmosphere. J. Less Common Met. 1982, 88, 57–61. [CrossRef] Coatings 2020, 10, 1178 15 of 15

38. Huang, J.; Li, X.; Wang, X.; Fang, X.; Wang, H.; Xu, X. New insights into CO2 methanation mechanisms on Ni/MgO catalysts by DFT calculations: Elucidating Ni and MgO roles and support eﬀects. J. CO2 Util. 2019, 33, 55–63. [CrossRef]

39. Guo, M.; Lu, G. The eﬀect of impregnation strategy on structural characters and CO2 methanation properties over MgO modiﬁed Ni/SiO2 catalysts. Catal. Commun. 2014, 54, 55–60. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional aﬃliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).