Catalytic Activity of Nickel and Ruthenium–Nickel Catalysts Supported on SiO 2 , ZrO 2 , Al 2 O 3 , and MgAl 2 O 4 in a Dry Reforming Process

: Dry reforming of methane (DRM) is an eco-friendly method of syngas production due to the utilization of two main greenhouse gases—methane and carbon dioxide. An industrial application of methane dry reforming requires the use of a catalyst with high activity, stability over a long time, and the ability to catalyze a reaction, leading to the needed a hydrogen / carbon monoxide ratio. Thus, the aim of the study was to investigate the e ﬀ ect of support and noble metal particles on catalytic activity, stability, and selectivity in the dry reforming process. Ni and Ni–Ru based catalysts were prepared via impregnation and precipitation methods on SiO 2 , ZrO 2 , Al 2 O 3 , and MgAl 2 O 4 supports. The obtained catalysts were characterized using X-ray di ﬀ ractometry (XRD), inductively coupled plasma optical emission spectrometry (ICP-OES), Brunauer–Emmett–Teller (BET) speciﬁc surface area, and elemental carbon-hydrogen-nitrogen-sulphur analysis (CHNS) techniques. The catalytic activity was investigated in the carbon dioxide reforming of a methane process at 800 ◦ C. Catalysts supported on commercial Al 2 O 3 and spinel MgAl 2 O 4 exhibited the highest activity and stability under DRM conditions. The obtained results clearly indicate that di ﬀ erences in catalytic activity result from the dispersion, size of an active metal (AM), and interactions of the AM with the support. It was also found that the addition of ruthenium particles enhanced the methane conversion and shifted the H 2 / CO ratio to lower values.


Introduction
Synthetic gas is one of the most important substrates in the chemical industry. It serves as a substrate in the synthesis of various chemicals, such as methanol, liquid synthetic motor fuels, naphtha, diesel, methane, and dimethyl ether [1,2]. Syngas is mainly produced via methane reforming and biomass gasification. Production technology and further applications of syngas strongly depend on the hydrogen to carbon monoxide molar ratio [1][2][3][4]. Among the reforming techniques, dry reforming (DRM) of methane has attracted much attention due to the utilization of two abundantly available greenhouse gases (methane and carbon dioxide) and a final H 2 /CO ratio that is close to unity [5,6]. Additionally, dry reforming requires only atmospheric pressure, however, the production of syngas via dry reforming at elevated pressures may be more practical because compression after the synthesis may lead to technical problems due to the high content of carbon monoxide [7,8].
Dry reforming is seen as an industrially immature process. A major limitation is the lack of an appropriate catalyst that exhibits high stability over a long time and high resistance to sintering and carbon deposition [1,9]. The most widely used and described catalysts are based on Ni particles. In comparison to noble metal catalysts, Ni catalysts are cheaper and exhibit similar or even higher The surface area determined using the Brunauer-Emmett-Teller method ranged from 391 to 6 m 2 g −1 for Ni/SiO 2 and both Ni/Al 2 O 3 and Ru-Ni/Al 2 O 3 , respectively. For silica supported catalysts, the BET surface area increased due to the high dispersion of nickel particles from 300 to 391 m 2 g −1 for SiO 2 and Ni/SiO 2 , respectively. The deposition of ruthenium oxide particles resulted in a two-fold reduction of S BET in comparison to Ni/SiO 2 . In the case of alumina-supported catalysts, the deposition of nickel oxide and ruthenium oxide particles did not influence the S BET -values for Al 2 O 3 , Ni/Al 2 O 3 , and Ru-Ni/Al 2 O 3 , which were almost the same. Values of the BET surface area for spinel-supported catalysts varied from 98 m 2 g −1 for Ni/MgAl 2 O 4 to 169 m 2 g −1 for Ru-Ni/MgAl 2 O 4 . Catalysts supported on zirconium(IV) oxide were characterized by a similar BET surface area and ranged from 23 to 30 m 2 g −1 for Ni/ZrO 2 and ZrO 2 , respectively.
Results of the XRD measurements are presented in Figure 1a-d and Table 1. The letter, S, refers to the reflections assigned to SiO 2 and N-NiO, R represents RuO 2 , A refers to Al 2 O 3 , M to MgAl 2 O 4 , and Z to ZrO 2 . Figure 1a represents the diffraction spectra of SiO 2 -supported catalysts. For SiO 2 and Ni/SiO 2 samples, the only reflection attributed silica were observed. A broad peak with a maximum at 23.5 • indicates the presence of amorphous silica [32,37]. In the case of the Ni-SiO 2 catalyst, the absence of reflections assigned to nickel oxide may indicate a high dispersion and small size of the NiO particles. As has already been reported in the literature, the lack of peaks attributed to particular components of a sample may be a result of high dispersion, low content, or small sizes of crystals [16,32,34]. The presence of nickel element was confirmed using inductively coupled plasma atomic emission spectroscopy nalysis. For Ru-Ni/SiO 2 , peaks attributed to NiO and RuO 2 were also observed. The observed reflections at 2θ equal to 28 [13,18,32]. The average crystalline size of RuO 2 and NiO equaled 8 and 19 nm. The content of RuO 2 and NiO crystals calculated using the Rietveld method equaled 19 ± 5 wt.% and 5 ± 5 wt.%, respectively.     It was previously reported that during the synthesis of nickel-spinel catalysts other compounds, such as MgO, Al 2 O 3 , and NiAl 2 O 4 , may be formed [39][40][41]. However, XRD analysis did not show the presence of other species, thus they were not formed during the preparation.
XRD patterns for the last group of catalysts-supported on ZrO 2 -are presented in Figure 1d. The average crystalline size of the bare support equaled 15 nm for Ni/ZrO 2 17 nm and for Ru-Ni/ZrO 2 , 15 nm. Peaks attributed to NiO were observed for both Ni/ZrO 2 and Ru-Ni Ni/ZrO 2 . Sizes of crystallites calculated based on the Scherrer equation were similar for the samples and equaled 22 Catalysts 2019, 9, 540 5 of 13 and 19 nm. No reflexions for the ruthenium species were observed, indicating small sizes and a high distribution [41]. The presence of the ruthenium element was confirmed using ICP analysis.

Nickel Catalysts
Results of the catalytic activity tests of nickel catalysts supported on SiO 2 , ZrO 2 , Al 2 O 3 , and MgAl 2 O 4 are presented in Figure 2a-c. In the graphs, the results for supports are not included due to their negligible catalytic activity. The highest conversion for methane and carbon dioxide was obtained for Ni/Al 2 O 3 and Ni/MgAl 2 O 4 catalysts. The conversion level of CH 4 for the Ni/Al 2 O 3 after two hours of the process remained stable at around 95%. In the case of the catalyst deposited on the magnesium-alumina spinel support, the efficiency of the methane conversion after 30 min of the process started decreasing to a final level of 80%. The consumption of carbon dioxide for both catalysts was observed at a steady level of 85%. Significant differences were observed in the generation of hydrogen and carbon monoxide. The highest molar ratio of hydrogen to carbon(II) oxide was close to what was also observed for the catalysts, Ni/Al 2 O 3 and Ni/MgAl 2 O 4 . Throughout the catalytic process, the n H 2 /n CO ratio for Ni/Al 2 O 3 ranged from 0.95 to 1.0, whereas for Ni/MgAl 2 O 4 , it was constant at 0.9.
The lowest activity was recorded for the catalyst obtained on a silica support. Throughout the process, the methane conversion reached 37%, while the carbon dioxide conversion level was about 50%. The Ni/ZrO 2 catalyst was characterized by an initial higher activity compared to Ni/SiO 2 ; however, after one hour, a 20% drop in methane conversion was observed. The decrease in the activity of the catalyst deposited on the support of zirconium(IV) oxide is probably a result of carbon deposits' formation on the catalyst surface, causing the blocking of active sites and throttling the flow of gases through the bed. The final molar ratio of n H 2 /n CO for Ni/ZrO 2 was dropped from 0.75 to 0.55.

Ruthenium-Nickel Catalysts
In Figure 3a-c, the results of tests on the activity of ruthenium-nickel catalysts are presented. The highest activity was observed for the Ru-Ni/Al2O3 catalyst. In comparison to Ni/Al2O3, an increase in activity due to the deposition of small ruthenium particles was observed. The conversion of methane remained stable-100% during the entire process-while the conversion of carbon dioxide ranged from 90% at the beginning to 71% after two hours. The molar ratios of hydrogen to

Ruthenium-Nickel Catalysts
In Figure 3a-c, the results of tests on the activity of ruthenium-nickel catalysts are presented. The highest activity was observed for the Ru-Ni/Al 2 O 3 catalyst. In comparison to Ni/Al 2 O 3 , an increase in activity due to the deposition of small ruthenium particles was observed. The conversion of methane remained stable-100% during the entire process-while the conversion of carbon dioxide ranged from 90% at the beginning to 71% after two hours. The molar ratios of hydrogen to carbon monoxide were lower in comparison to the catalysts containing only nickel particles. At the beginning of the process, it was 0.75 and it decreased during the process to 0.6. In the case of the Ru-Ni/MgAl 2 O 4 catalyst, the complete conversion of methane was also observed. The conversion of carbon(IV) oxide ranged from 85% to 75% at the beginning and after 2 h of the process, respectively. The molar ratio of hydrogen/carbon monoxide was also slightly reduced from 0.67 to 0.61. The results obtained for the generation of CO and H 2 are similar to the results obtained for Ni/MgAl 2 O 4 ; however, the molar ratio of H 2 :CO was lower. The lack of catalytic activity enhancement is probably assigned to the deposition of large RuO 2 particles. The lowest catalytic activity among the tested catalysts was observed for the Ru-Ni/SiO 2 catalyst. The deposition of small ruthenium(IV) oxide particles on the silica surface resulted in an increase in the catalytic activity compared to Ni-SiO 2 . The conversion of methane increased from 37% to 60% to 64%. The degree of conversion of carbon(IV) oxide, however, remained at a similar level, around 50%. The H 2 :CO molar ratio was noted to be in the range of 0.70 to 0.53.  Among the tested ruthenium-nickel catalysts, Ru-Ni/Al2O3 was characterized by one of the highest activity and stability during 2 h test, thus its stability was examined during the 6 h test. Figure  4a,b show the results of the analysis. The methane conversion rate remained constant at 98% to 100% for 6 h of the process. The degree of CO2 conversion during the process was also constant; after six hours the conversion rate reached 75%. The molar ratio of H2:CO decreased from 0.75 to 0.55 at the beginning and after 6 h of the process, respectively. Among the tested ruthenium-nickel catalysts, Ru-Ni/Al 2 O 3 was characterized by one of the highest activity and stability during 2 h test, thus its stability was examined during the 6 h test. Figure 4a,b show the results of the analysis. The methane conversion rate remained constant at 98% to 100% for 6 h of the process. The degree of CO 2 conversion during the process was also constant; after six hours the conversion rate reached 75%. The molar ratio of H 2 :CO decreased from 0.75 to 0.55 at the beginning and after 6 h of the process, respectively. Among the tested ruthenium-nickel catalysts, Ru-Ni/Al2O3 was characterized by one of the highest activity and stability during 2 h test, thus its stability was examined during the 6 h test. Figure  4a,b show the results of the analysis. The methane conversion rate remained constant at 98% to 100% for 6 h of the process. The degree of CO2 conversion during the process was also constant; after six hours the conversion rate reached 75%. The molar ratio of H2:CO decreased from 0.75 to 0.55 at the beginning and after 6 h of the process, respectively.

Carbon Deposits
Carbon deposits on the catalyst surface after the dry reforming process were quantified using CHNS elemental analysis. Table 2 presents the results of the measurements for selected catalysts. The carbon content is expressed in % by weight of the reactor bed. The lowest contents were observed for catalysts showing the highest catalytic activity-Ni/MgAl 2 O 4 , Ru-Ni/MgAl 2 O 4 , Ni/Al 2 O 3 , and Ru-Ni/Al 2 O 3 . The carbon content was less than 0.1 wt.%. Carbon deposits were observed only for catalysts supported on silica-the carbon content was comparable to 0.38 wt.% and 0.42 wt.% for Ni/SiO 2 and Ru-Ni/SiO 2 , respectively. The total carbon contents for Ni/SiO 2 and Ru-Ni/SiO 2 equaled 0.19 and 0.21 ug of carbon per 1 g of catalyst.

Discussion
The catalytic activity of nickel and ruthenium-nickel catalyst deposited on ZrO 2 , Al 2 O 3 , MgAl 2 O 4 , and SiO 2 supports was investigated. The highest activity was observed for catalysts for both nickel and ruthenium-nickel catalysts supported on Al 2 O 3 and MgAl 2 O 4 . The level of methane and carbon dioxide conversions for Ni/Al 2 O 3 was constant through all catalytic process and equaled 95% and 85%, respectively. A similar effect in CO 2 conversion was observed for Ni/MgAl 2 O 4 , however, a lower level of methane conversion occurred. The hydrogen to carbon monoxide ratio for the most active samples was also the highest, close to unity, as is expected for dry reforming processes. Moreover, for most active samples, no carbon deposits were detected. The Ni/SiO 2 catalyst revealed the lowest activity and carbon deposition. The total carbon content equaled 0.19 ug C g cat −1 . No correlation between the BET surface area and catalytic activity was observed. The highest activity was observed for supports that  [37,42], whereas with Al 2 O 3 and MgAl 2 O 4 , nickel reduction temperatures are significantly higher, indicating strong interactions [9,42]. The high activity of catalysts supported on MgAl 2 O 4 may also be a result of the basic features. Supports of the basic properties facilitate dissociative adsorption of CO 2 on the catalyst [1,9,43]. Park et al. [41] investigated the effect of supports on the catalytic activity of Co-Al catalysts during the dry reforming process. The catalytic activity decreased as follows: CoAl/MgAl 2 O 4 > CoAl/Al 2 O 3 > CoAl/ZrO 2 > CoAl/CeO 2 > CoAl/SiO 2 . The differences in catalytic activity were attributed to the balance between the rate of carbon decomposition and surface oxidation, which are strictly dependent on the nature of the support. Han et al. [42] also investigated the effect of the support of catalysts during DRM reactions. They found that a high catalytic activity was a combined result of the nickel nanoparticle size, process temperature, and strong interactions of Ni with the support, especially with basic properties. Further modification with ruthenium particles resulted in an increase in the methane conversion rate and product generation yield for catalysts obtained on silica supports and commercial aluminium (III) oxide. For Ru-Ni/Al 2 O 3 , almost all CH 4 was converted; even for the less active sample, Ru-Ni/SiO 2 , the methane conversion increased to a level of around 65%. A decrease of the H 2 :CO ratio for all ruthenium-nickel catalysts in comparison to catalysts containing only nickel particles was noted. For example, for the Ni/Al 2 O 3 and Ru-Ni/Al 2 O 3 sample, the H 2 :CO ratio decreased from nearly 1.0 to 0.7, respectively, whereas for Ni/MgAl 2 O 4 , it decreased from about 0.9 to 0.65, respectively. According to the literature, noble metal particles, such as platinum, ruthenium, and rubidium, increase catalyst reducibility and resistance to carbon deposition by catalyzing the gasification reaction of carbon atoms, which results from carbon dioxide disproportionation, methane decomposition, and the carbon dioxide reaction with hydrogen (RWGS) [15]. The enhanced hydrogen generation as a result lowers the H 2 :CO ratio, which may be caused by the additional promotion of methane dissociation observed in the presence of noble metal particles [44]. On the basis of X-ray diffraction crystallite size analysis (see Table 1), it was found that the average size of ruthenium(IV) oxide Ru-Ni/SiO 2 crystals was 8 nm, while for Ru-Ni-MgAl 2 O 4 , this was 32 nm. For Ru-Ni/Al 2 O 3 , no reflections were observed for RuO 2 , which may indicate the presence of very small RuO 2 crystallites, resulting in an enhanced catalytic activity. According to the literature and the free energy values of Gibbs for dry reforming at 800 • C, the methane conversion rate can reach 95%, while carbon dioxide can reach 85% [7]. For ruthenium-nickel catalysts supported on MgAl 2 O 4 and Al 2 O 3 , nearly maximum possible conversions were reached. Fluctuations in carbon dioxide and hydrogen can result from the occurrence of side reactions, including the oxide disproportionation reaction (called the Boudouard reaction) and the conversion of carbon(II) oxide with water vapor (RWGS) [10]. CHNS elemental analysis confirmed the effective gasification of carbon deposits for the most active catalysts. Carbon deposits were only detected only for Ru-Ni/SiO 2 at the amount of 0.21 ug c g cat −1 . For the Ru/Ni-Al 2 O 3 catalyst, a 6 h test for stability during DRM conditions was performed. The catalyst exhibited high activity and stability. The methane and carbon dioxide conversions were constant throughout the whole process and equaled 98% to 100% and 75%, respectively. The H 2 :CO ratio also remained at a constant level of 0.7.

Catalyst Characterization
XRD analysis was performed using a Rigaku Intelligent X-ray diffraction system SmartLab equipped with a sealed tube X-ray generator (a copper target; operated at 40 kV and 30 mA), a D/teX high-speed position sensitive detector system, and an ASC-10 automatic sample changer. Data acquisition conditions were as follows: 2θ range: 5-90 • , scan speed: 1 • min −1 and scan step 0.008. Analysis of the crystal structure and planes, including Rietveld analysis, was performed using Rigaku PDXL software (Version 2.0, Rigaku Corporation, Neu-Isenburg, Germany, 2007).
The Brunauer, Emmett, and Teller (BET) surface analysis method was performed using the Micromeritics Gemini V apparatus (model 2365) (Norcross, GA, USA). Helium was used as the gas filling free spaces. The range of partial pressures, p/p 0 , ranged from 0.05 to 0.3. Before the adsorption measurement, the samples were degassed at 473 K for two hours.
The final metal loading (wt.%) of the final catalysts was carried out using atomic emission spectroscopy with excitation in induced plasma (ICP-OES). The ICP-OES Agilent 5100 VDV (Santa Clara, CA, USA) apparatus with a Mars CEM sample mineralizer was used for the tests. Analysis of the elemental composition was preceded by the stage of mineralization of catalysts with nitric acid at 200 • C and pressure not exceeding 24 bar.
Carbon deposits on spent catalysts were evaluated using elemental analysis CHNS. The sample of the spent catalyst was encapsulated in tin foil and combusted in an excess of oxygen at 1060 • C. The amount of generated carbon dioxide was measured chromatographically using a Thermo Fisher Scientific Flash 2000 analyzer (Walthman, MA, USA) equipped with a thermal-conductivity detector. Nitrogen was used as the carrier gas. The limit od detection equaled 0.1 wt.%.

Catalyst Preparation
Monometallic and bimetallic nickel and ruthenium catalysts were obtained using impregnation and precipitation methods. Detailed descriptions of the preparation are presented in the subsections below.

Support Preparation
The magnesium-alumina spinel was obtained by chemical precipitation in a basic environment. Magnesium nitrate and aluminium nitrate were dissolved in deionized water and mixed to obtain a clear solution. The molar ratio of Mg:Al equaled 1:2. Subsequently, the pH was adjusted to pH 10 with ammonia. The resulting precipitate was subjected to 24 h of ageing at 50 • C. Then, the precipitated MgAl 2 O 4 was separated, washed twice with deionized water, dried overnight at 100 • C, and calcinated at 400 • C for 6 h.
The preparation of silica was based on the Stöber method. First, SiO 2 was obtained via the hydrolysis of tetraethylorthosilicate (TEOS) in a basic environment. The hydrolysis was carried out by dropping a solution of tetraethylorthocoate in anhydrous ethyl alcohol into an aqueous solution of ammonia water. The amounts of reagents used were due to the stoichiometry of the reaction and our own preliminary research. The weight ratio of the reagents used for preparation equaled TEOS:H 2 O:EtOH:NH 3 = 11:1.8:30:1. The sol obtained from the hydrolysis was aged at room temperature for 24 h. The resulting precipitate was then separated, washed with ethyl alcohol, water, dried overnight at 80 • C, and calcinated at 500 • C for 6 h. The temperature increase was maintained at 2 • C·min −1 .
Zirconium(IV) oxide was obtained by chemical precipitation in a basic environment. Zirconium(IV) nitrate was dissolved in deionized water at 40 • C. The resulting solution was then cooled to room temperature and alkalized with ammonia until the pH reached pH = 9. The obtained precipitate was separated, washed twice with water, and dried at 80 • C. The last stage was calcination at 600 • C for 5 h.

Preparation of Ni and Ni-Ru Catalysts
Nickel-based catalysts were obtained via impregnation of the support with metal ions followed by chemical precipitation. Previously prepared support, as described in Section 4.3.1, MgAl 2 O 4 , SiO 2 , ZrO 2 , or commercial Al 2 O 3 , after the drying step, was dispersed in deionized water. Then, solutions of metal salts (nickel(II) nitrate and/or ruthenium(III) chloride) were added dropwise and the pH was adjusted to 9. The precipitate was separated, washed twice with water, and dried at 80 • C. Obtained powders were calcinated at 400, 500, and 600 • C for Al 2 O 3 , MgAl 2 O 4 , SiO 2 , and ZrO 2 supports [30,41,42], respectively. The metal content used for preparation was 7 wt.% and 1 wt.% for nickel and ruthenium, respectively.

Catalytic Activity
The catalytic activity of the obtained catalysts was investigated in a model reaction of the synthesis of carbon(II) oxide and hydrogen (synthesis gas). The tests were carried out in a tubular reactor with a length of 600 mm and an internal diameter of 4.5 mm. The catalytic bed consisted of catalyst in an amount of 5 wt.% diluted with quartz sand with a grain size in the range of 0.1 mm to 0.5 mm. A model mixture of substrates of carbon dioxide and methane were introduced into the reactor with a total flow rate of 400 cm 3 min −1 in a 1:1 volume ratio. The process temperature was maintained at 800 • C. The temperature inside the reactor was controlled throughout the process using a K120764 type thermocouple. The catalytic activity was expressed through the substrate conversion rate (CO 2 and CH 4 ) and product generation yield (H 2 and CO). The content of individual compounds in the mixture leaving the reactor was determined chromatographically using a Perkin Elmer Clarus 500 (Shelton, CT, USA) gas chromatograph equipped with a ShinCarbon ST100/120 (Restek Corporation, Bellefonte, PA, USA) packed column (length 1 m, internal diameter 1 mm) and a thermal-conductivity detector (TCD). Argon was used as the carrier gas.
Conversion levels and the molar ratio of hydrogen to carbon monoxide (II) were determined using the formulas shown below: X CH 4 (%) = n CH 4,in − n CH 4,out n CH 4,in × 100, X CO 2 (%) = n CO 2,in − n CO 2,out n CO 2,in × 100, Molar ratio = n H 2,out n CO,out , where n i,in and n i,out are the moles of each component in the feed or effluent, respectively.

Conclusions
In this work, we reported the effect of supports on the catalytic activity of nickel and ruthenium-nickel catalyst deposited on ZrO 2 , Al 2 O 3 , MgAl 2 O 4 , and SiO 2 . Unmodified supports did not reveal catalytic activity, which indicates the key role of nickel and ruthenium particles in the dry reforming process. The catalytic activity for nickel catalysts decreased in the following order: Ni/Al 2 O 3 > Ni/MgAl 2 O 4 > Ni/ZrO 2 > Ni/SiO 2 , while for ruthenium-nickel catalysts, the same order was observed: Ru-Ni/Al 2 O 3 > Ru-Ni/MgAl 2 O 4 > Ru-Ni/ZrO 2 > Ru-Ni/SiO 2 . The highest activity was observed for supports that strongly interact with nickel particles, which indicates the key role of support features on activity under DRM conditions. A comparison of substrate conversions and the H 2 :CO product ratios for Ni and Ni-Ru catalysts revealed that the introduction of ruthenium particles to the process system results in the switching of the hydrogen to carbon monoxide ratio to lower values as a consequence of enhanced methane dissociation. Differences in the catalytic activities were probably a result of differences in the dispersion and particle size of active metals, as well as interactions between support particles and active metals. Study on the effect of the particle size of nickel and ruthenium-nickel catalysts will be continued during further research.