Production of Pure Hydrogen through Thermocatalytic Methane Decomposition Using NiO-MgO Catalysts Promoted by Chromium and Copper Prepared via Mechanochemical Method

,


Introduction
The burning of fossil fuels has enhanced the greenhouse gas emissions in the atmosphere and caused serious environmental problems, so it is necessary to find a replacement energy source [1][2][3]. Hydrogen is considered one of the candidates for clean and viable energy carriers with a high energy density (140 MJ/kg) compared to solid fuels (50 MJ/kg) [4][5][6]. Nevertheless, due to the high reactivity of H 2 with other elements such as carbon and oxygen, there is no pure source for it in nature, so it should be produced using other resources (such as natural gas) [7,8]. The main method for hydrogen production in the industry is methane reforming using water vapor which produces CO x as a by-product [9][10][11]. Hydrogen production from this method includes several stages: (1) catalytic reforming of methane (800-900°C) to produce synthesis gas, (2) water gas shift (WGS) reactions (173-257°C) to transform carbon monoxide to carbon dioxide, and (3) separation of the hydrogen and carbon dioxide gases with pressure-swing adsorption (PSA) unit [12,13].
Thermocatalytic methane decomposition allows the production of pure H 2 without any CO x impurities and solid carbon with high potentials. This reaction (CH 4 ⟶ C ðsÞ + 2H 2ðgÞ ; ΔH 298 = 74:8 kJ/mol) is endothermic, so the increase of reaction temperature leads to enhancement in the amount of produced hydrogen [14][15][16]. Although methane is the most inactive hydrocarbon and its decomposition requires temperatures higher than 1473 K, but with using suitable catalysts, this temperature can be significantly decreased [17,18]. A group of nickel-(Ni-), cobalt-(Co-), and iron-(Fe-) based catalysts was used for methane decomposition. Among these metals, Ni is known as the most active catalyst for the methane decomposition due to its exclusive 3D-orbital structure which facilitates methane dissociation through partially accepting electrons. However, Ni catalysts without support are not fully active for the decomposition of methane reaction due to the thermal sintering. The use of catalyst support can prevent the sintering of nickel particles by dispersing them on the catalyst support [19][20][21][22]. Rastegarpanah et al. [23] investigated methane decomposition of Ni-based catalysts over different supports including Al 2 O 3 , MgO, TiO 2 , and SiO 2 , and the result showed that Ni/MgO catalysts possessed higher activity of 57% methane conversion and stability than the other catalysts. Also, Liang et al. [24] investigated NiMgAl catalyst in thermocatalytic methane decomposition, and this catalyst showed 36% methane conversion. However, catalytic activity and stability are not only related to the type of support, but choosing a good promoter or using a suitable preparation method can affect the performance and lifetime of the catalyst.
The main challenge in this process is the deactivation of the catalyst due to the carbon deposition over catalyst active sites [25][26][27]. It is reported that copper addition can prevent the deposition of the carbon layers over active sites of the Ni-based catalyst and also increase the reducibility of Ni and methane adsorption on the catalyst surface. Because of the similar electron structure of copper and chromium, Cr can be also employed as a promoter to reach higher catalyst activity and lifetime [28][29][30].
In the present work, based on the literature, Ni and Mg were chosen as active phase and support of the catalyst, and also, Cr and Cu were chosen as promoters. In this study, comparison between Cr and Cu as promoters for 50NiO-MgO catalysts in the methane decomposition reaction was investigated. In this research, the loading of NiO was constant, and it was chosen as 50 wt.% based on the obtained results of our previous work [31].
The required catalysts were synthesized using the mechanochemical route. Between different synthesis routes for catalyst preparation, mechanochemical is one the most interesting synthesis method. The simplicity of the catalyst preparation process, the usage of conventional precursors (metal oxide or salts), and the lack of organic solvents for catalysts synthesis are some of the main advantages of the mechanochemical method.

Experimental Section
2.1. Catalyst Preparation. All of the materials were purchased from Sigma-Aldrich with 99.99% purity.
The NiO-MgO catalysts promoted by Cr 2 O 3 and CuO and different loadings of these metals were synthesized with the mechanochemical method. The initial materials used in this method were Ni(NO 3 ) 2 ·6H 2 O, Mg(NO 3 ) 2 ·H 2 O, Cu(NO 3 ) 2 ·3H 2 O, and Cr(NO 3 ) 3 ·9H 2 O as the metal precursor, and also, ammonium carbonate ((NH 4 ) 2 CO 3 ) was used as the precipitation agent. For preparing desired catalysts with the mechanochemical route, the calculated extents of metal precursors and precipitation agent were milled using a mortar for 20 min at ambient temperature. In this route, the released hydrate water of the metal precursors plays the role of solvent, and the ammonium carbonate acts as a precipitating agent [32]. After reaching a pasty state, the mixture was dried for 24 h in an oven with a temperature of 100°C, and then, it was calcinated at the temperature of 600°C for 3 h with a heating ramp of 3 (°C)/min.

Characterization of Catalysts.
The N 2 adsorption/desorption analysis was conducted at 77 K to study the BET surface area and pore characteristics of the prepared catalysts. This analysis was performed using a BELSORP-mini II analyzer. It should be noted that before the measurement, the catalysts were pretreated under vacuum at 250°C for 2 h to remove moisture. The crystal structure of the catalysts was analyzed with X-ray diffraction (XRD) analysis using a PANalytical X'Pert-Pro instrument. The reducibility and coke formation of the catalysts were evaluated by temperature-programmed reduction and oxidation (TPH and TPO) analysis using a similar instrument (Micromeritics ChemiSorb 2750) with different inlet gas flow. For temperature-programmed oxidation test, a gas stream containing a mixture of oxygen and helium (95 : 5) (20 ml/min) was used, and also for temperatureprogrammed reduction test, a gas stream containing a mixture of hydrogen and argon (95 : 5) (20 ml/min) was used. The morphological structure of the fresh catalysts and also the deposited carbon on the surface of the spent catalysts were determined with a scanning electron microscopy (SEM) test using a MIRA3 TESCAN apparatus.

Catalyst Activity Evaluation.
For evaluating catalyst activity in the thermocatalytic methane decomposition reaction, the exact amount of the catalyst (25 mg, 0.25-0.5 mm) with 75 mg of inert quartz sands was replaced in a vertical quartz microreactor. After that, the catalyst was reduced in an H 2 gas stream with a flow of 25 ml/min at a temperature of 700°C for 3 h in atmospheric pressure. After this step, the temperature was set to 575°C, and reactant gas (20 ml/min) containing CH 4 and N 2 with the proportion of 15 to 85 was introduced to the microreactor, and the activity of the catalyst was specified at various temperatures from 575 to 700°C with a heating step of 25°C in atmospheric pressure. The outlet gas from the microreactor was specified using an online gas chromatograph instrument (Varian 3400) equipped with a packed column and thermal conductivity detector. H 2 yield and CH 4 conversion were obtained with the following equations:  [30,36,37]. As can be seen in Figure 1, enhancement of Cu content intensified diffraction peaks, and this was due to the increase in agglomeration degree of particles on the catalyst surface with the rise of Cu loading. The enhancement of Cr loading increased the dispersion of the NiO particles and had a negative effect on the intensity of diffraction peaks. In other words, the increasing of Cu and Cr contents resulted in larger and smaller crystallite sizes, respectively. Also, the crystallite size of the calcined catalyst was calculated using Scherrer's correlation (D = 0:9λ/β cos θ, where D stands for the crystallite size, θ stands for the Braggs angle, λ stands for the wavelength of X-ray, and β stands for the full width at half the maximum of the peak), and the obtained results are gathered in Table 1. The obtained results for Cu-doped and Cr-doped catalysts were between 21-28.5 and 17.8-29.9, respectively, which were compatible with results obtained from XRD patterns.

Structural Characteristics of As-Prepared Catalysts.
The adsorption/desorption isotherms and pore size distribution of 50NiO-MgO and Cu-and Cr-doped samples are exhibited in Figures 2(a) and 2(b). These figures show that Cr and Cu additives did not alter the isotherm type, and according to the IUPAC classification, each catalyst presented a type II isotherm with an H3-shaped hysteresis loop. Type II isotherm is related to macroporous powders or nonporous  3 International Journal of Energy Research solids, and the H3 hysteresis loop is a feature of solid materials which have aggregated or agglomerated particles with gap-shaped pores having nonuniform shape and size [38]. As mentioned before, the Cr additive did not change the isotherm type, but it caused the enhancement in the slope of the isotherm and width of the hysteresis loop as compared with the undoped and Cu-doped catalysts, suggesting the higher surface area of the Cr-doped catalysts. All the catalysts showed pore size distribution in the range of 10-60 nm, and this extent of pore size showed that all calcined samples possessed meso-and macropores in their structure.
The physicochemical properties of 50NiO-MgO, Cudoped, and Cr-doped catalysts are reported in Table 1. The specific surface area and pore volume for Cu-doped samples were between 10.3-126 m 2 /g and 0.032-0.041 cm 3 /g, respec-tively. The BET surface area and pore volume were diminished with the addition of CuO to the catalysts, which can be due to the plugging of the catalyst pores by the CuO and the collapse of the porous structure. Additionally, the particle size was also enhanced with the doping of Cu due to the agglomeration of particles, while with the doping of Cr resulted in an increase in the BET area and pore volume.

H 2 -TPR Analysis of the Catalysts. H 2 -TPR technique was
performed to investigate the reducibility of the synthesized samples, and the results are presented in Figure 3. The 50NiO-MgO catalyst exhibited two main reduction peaks. The low-temperature peak is attributed to the reduction of nickel oxide to nickel in the catalyst with weak interaction among nickel oxide and magnesium oxide, and the high-  International Journal of Energy Research temperature peak is recognized to the reduction of NiO-MgO solid solution [23]. With doping of Cu as the promoter, a shift in the first reduction peak was observed and resulted in lowering the reduction temperature. This may be a result of a higher activated H 2 production over Cu, which results in a higher rate of Ni nucleation and consequently a higher NiO reducibility. Also, both reduction peaks were sharper as an effect of Cu addition [39]. For the chromium-doped catalysts, two main peaks at nearly 400 and 500°C and a broad peak between 600 and 800°C were observed. The first peak at 400°C was majorly attributed to the reduction of Cr 6+ to Cr 4+ , which became sharper with the enhancement of Cr additive, and the weak peak at near 500°C was attributed to the reduction of NiO with mild interaction (i.e., NiCr 2 O 4 or MgCr 2 O 4 ), and the higher temperature reduction peak was belonged to the reduction of the nickel oxide particles which were in solid solution of nickel and magnesium oxide [40]. As shown in this figure, the center of the broad peak moved to the lower temperature as a result of the addition of Cr which caused better reducibility of the sample.

TMD Test.
The influence of copper content on the performance of the 50NiO-MgO catalysts in the temperature range of 575-700°C in the TMD reaction was investigated, and the acquired results are illustrated in Figure 4(a). For the 50NiO-MgO catalyst, methane conversion decreased significantly after the first temperature as an outcome of the agglomeration of catalyst particles and carbon deposition over the sample active sites. For the Cu-doped catalysts, the result showed that the initial conversion of methane at low temperatures diminished with the enhancement of Cu content, which could be due to the inactivity of copper at these temperatures in the TMD reaction. With the enhancement of temperature, the methane conversion increased which was an outcome of the endothermic nature of methane decomposition reaction. Also, an increase in the copper content led to better catalytic activity and stability at temperatures higher than 625°C which was due to the effect of copper in the protection of the catalyst surface.  shows the activity of Cr-doped catalysts in the temperature range of 575-700°C. A comparison between the catalytic activity of Cr-doped catalysts and 50NiO-MgO indicates that adding Cr improved the catalytic performance and stability. According to the TPR analysis of the catalysts, the higher performance of the catalysts was attributed to the better reducibility of Cr-doped catalysts. Also, addition of Cr caused higher dispersion of active particles. With the rise of reaction temperature, the catalytic activity enhanced; as mentioned before, this was a result of the endothermic nature of methane decomposition reaction. Figure 4(b) also displays that the catalytic activity and lifetime enhanced with the rise of Cr loading due to better reducibility of catalysts.
3.5. Catalytic Stability. The comparison between catalytic activity and lifetime of the 50NiO-MgO, 50NiO-MgO-15CuO, and 50NiO-MgO-15Cr 2 O 3 at the temperature of 575°C within 300 min of methane decomposition is presented in Figure 5. A comparison between catalytic activity of 50NiO-MgO and 50NiO-MgO-15CuO shows that, even  International Journal of Energy Research though at first, the activity of Cu-doped catalyst was less than 50NiO-MgO due to the inactivity of Cu at the lower temperatures, but the Cu-doped catalyst showed a better catalytic lifetime as compared to the 50NiO-MgO. The results confirmed that the Cr-doped catalyst presented much higher activity and stability compared to other catalysts since Cr is much more reactive than Cu at lower temperatures.  7 International Journal of Energy Research was produced over the surface of the catalyst. Moreover, the random growing direction of the carbon filaments might be due to the space competition during carbon growth [30,41].

TPO Analysis and Carbon
Accumulation. The TPO results of used 50NiO-MgO, 50NiO-MgO-15CuO, and 50NiO-MgO-15Cr 2 O 3 catalysts after 300 min at 575°C are presented in Figure 7(a). As shown in TPO profiles of 50NiO-MgO and Cu-doped catalysts, just one oxidation peak was noticed, and it was assigned to the carbon formation on the surface of the sample. The oxidation peak of the chromium-doped sample had a higher area because of its higher carbon formation during the decomposition of the methane reaction. However, for the chromium-doped sample, two oxidation peaks were detected, and the sharp oxidation peak was related to the fast burning of carbon with a crystalline structure. The other oxidation peak just like other samples was attributed to the formation of carbon on the surface of the catalyst.
The accumulated carbon measured by carbon balance is reported in Figure 7(b). The carbon content over the chromium-doped sample was much more than the other catalysts which was compatible with the result obtained from TPO results of catalysts.
3.8. Effect of GHSV. The effect of GHSV on the performance of the 50NiO-MgO-15Cr 2 O 3 catalyst in decomposition of the methane reaction at 525°C is presented in Figure 8. As can be seen, the GHSV enhanced the methane conversion diminished, which was due to the less contact time between methane and the surface of the catalyst, and also, the enhancement of GHSV diminished the amount of adsorbed methane on the active sites of the catalysts [42,43].

Conclusion
In this article, the 50NiO-MgO, Cu-doped, and Cr-doped samples were prepared using the mechanochemical method. The effect of Cr and Cu loadings on the activity and lifetime of the catalysts in the TMD reaction was studied. It was observed that the addition of Cu (due to its effect in the protection of the surface of the catalyst) and Cr (due to its effect on better reducibility of catalyst) enhanced the catalyst activity and lifetime. Overall, the Cr-doped catalyst showed higher performance (with methane conversion of 65%) and lifetime compared to other catalysts during 300 min at 575°C since Cr is much more reactive than Cu at lower temperatures and also the higher dispersion of active particles with the addition of Cr. Also, the effect of GHSV on the methane conversion of Cr-doped catalyst was studied, and the obtained results showed that the enhancement of GHSV diminished the conversion of methane from almost 44% to 34% due to the less contact time between methane and the surface of the catalyst and also decrease in the ratio of accessible active sites per methane entrance molecules.

Data Availability
No underlying data was collected or produced in this study.

Conflicts of Interest
The authors declare that they have no conflicts of interest.