A New Approach to the Preparation of Stable Oxide-Composite Cobalt-Samarium Catalysts for the Production of Hydrogen by Dry Re-Forming of Methane

1. Introduction

The increasing threat of global climate change has prompted the governments of 195 countries to adopt the “Paris Agreement” that aims to reduce greenhouse gas emissions, primarily by decarbonizing the economy [1,2]. In this regard, hydrogen is considered the most environmentally friendly alternative fuel, since its use as a fuel is accompanied only by the release of water vapor into the air. Hydrogen has the highest energy intensity per unit mass but low density. In addition to being a promising fuel, hydrogen is used in increasing volumes in the processes of ammonia synthesis and oil refining and as a synthesis gas component for the production of methanol and a number of petrochemicals [1,2,3,4].

The main industrial process of hydrogen production is the steam reforming of methane (1), which a high-energy endothermic process requiring the production of considerable volumes of water vapor [2,5,6].

СН₄+Н₂O⇌СО+3Н₂    ΔH⁰₂₉₈= +206 kJ

(1)

Promising processes for hydrogen production from methane are partial oxidation of methane, POM, (2) and dry reforming of methane, DRM, (3).

2СН₄ + O₂⇌ 2СО + 4Н₂    ΔH⁰₂₉₈= –36 kJ

(2)

СН₄ + СО₂ ⇌ 2СО + 2Н₂    ΔH⁰₂₉₈= +247 kJ

(3)

Since reaction (2) is exothermic, the POM process in combination with the steam reforming of methane is implemented in the autothermal mode. In addition, POM allows the production of synthesis gas (H₂: CO = 2:1) suitable for direct processing into petrochemical products [5,6,7,8]. The practical application of POM is largely limited by the need to use pure oxygen, the related explosion hazard, and the possible sintering of metal active centers in the event of "hot spots" in the catalyst layer.

The process of hydrogen production by the DRM reaction has been most extensively studied in recent decades. For example, the number of publications on this topic in 1990-2022 exceeded 1000, including 863 scientific articles with the citation number above 6, and 34 reviews [9]. The explosive interest in this topic is due to a number of factors. Firstly, these are the above-mentioned efforts of most countries to reduce the greenhouse effect [1,2,3]. The DRM process allows the utilization of methane and carbon dioxide, which are the main greenhouse gases. Secondly, the DRM process is a way to produce hydrogen and other valuable products by processing of renewable raw materials, biogas, one of the promising decarbonization strategies [10]. Thirdly, the industrial process of hydrogen production by the steam reforming of methane includes an additional stage of the steam reforming of CO (4):

CO + H₂O ⇌ CO₂ + H₂

(4)

CO₂ generated by this process must also be disposed of.

In addition, the joint methane steam reforming and DRM can also be used for the utilization of CO₂ captured from the flue gases of methane steam reforming plants [2]. Finally, the DRM process renders it possible to produce synthesis gas, which is a suitable intermediate for the synthesis of dimethyl ether, Fischer-Tropsch synthesis of hydrocarbons etc. [1,2,3,4,6,9,11,12,13,14,15,16]. Furthermore, the synthesis gas of the DRM process can be used in fuel cells [13].

As for the use of the DRM process is practice, this process is realized mostly on a pilot scale and in combination with the steam reforming of methane in order to control the composition of the resulting synthesis gas [15,16,17,18]. The CALCOR process is known [15,16,17,19]; however, it is aimed at obtaining primarily carbon monoxide (H₂:CO = 0.42:1) and, therefore, is accomplished at a large excess of CO₂. The industrial implementation of DRM is constrained by a number of features of this process, such as high endothermicity associated with the stability of CO₂ and CH₄ molecules and a significant formation of coke deposits and "sintering" of active catalyst centers associated with a high temperature of the process. A high temperature of the process also requires complex hardware design. In addition, it was shown that the DRM process conducted at elevated pressures is accompanied by an enhanced carbonization of catalysts. Therefore, the synthesis gas produced by DRM and intended for subsequent processing will require a costly stage of its compression [1,13,15,17,20,21,22].

According to thermodynamic calculations, at fairly low temperatures DRM does not allow a high yield of synthesis gas. To achieve high yields of synthesis gas (H₂ : CO = 1 : 1) by reaction (3) with the reagents conversion close to 100% and reduced carbon formation by the methane pyrolysis reaction (5), a ratio of CO₂ : CH₄ = 1 : 1, a temperature above 850°C, and atmospheric pressure are required [12,13,16,17,22,23,24,25].

CH₄ ⇌ C + 2H₂    ΔG⁰₂₉₈ ≤ 0 at T≥ 658 C.

(5)

These conditions also contribute to the reverse water gas shift reaction (6) that affects the ratio of DRM products and the value of CO₂ conversion:

CO₂ + H₂ ⇌ CO + H₂O    ΔG⁰₂₉₈ ≤ 0 at T ≥ 700 C.

(6)

Carbon formation in the DRM process is possible not only by reaction (5) but also by the disproportionation of CO (7) and the hydrogenation of CO (8) and CO₂ (9) [12,26].

2СО ⇌ С+ CO₂   ΔH⁰₂₉₈= -172 kJ

(7)

СO+Н₂⇌С+Н₂О   ΔH⁰₂₉₈= -131 kJ

(8)

СO₂+2Н₂⇌С+2Н₂О   ΔH⁰₂₉₈= -90 kJ

(9)

However, these reactions are exothermic, thermodynamically unfavorable at high temperatures, and probably proceed in the opposite direction.

One of the key approaches to the practical implementation of the DRM process considered in most relevant publications is to design selective, stable, and carbonization resistant catalysts. Objectives are to prevent sintering of the active centers of catalysts and formation of stable forms of surface carbon (graphite, carbon fibers, and nanotubes), which deactivate active centers, destroy the catalyst, and block the passage of gases through a reactor. Complex oxides of the perovskite structure are in most common use as promising DRM catalysts [1,2,3,5,6,10,11,12,13,14,15]. As a result of pre-reduction or contact with reagents perovskite structure catalysts are transformed into composites containing a metal phase highly dispersed in an oxide matrix, which in some cases enables one to achieve high activity, selectivity, and stability in DRM. Among these compounds, there are catalysts based on nickelates and cobalates of rare earth elements [5,11,27,28,29,30,31,32,33]. According to these studies, perovskite precursors of catalysts can be synthesized by various methods, such as solid-phase synthesis, self-propagating high-temperature synthesis, sol-gel synthesis, decomposition of specially synthesized complex compounds, etc. The choice of the synthesis method largely determines the properties of the DRM catalysts.

Previously, we synthesized and tested in DRM lanthanum nickelate [30] and samarium cobaltate [29] perovskite structures obtained by the thermal decomposition of specially synthesized heterometallic complex compounds (M¹(phen)x][M²(NO₃)_y(H₂O)]·zMeCN, where M¹ is Ni or Co, M² is La or Sm, phen is o-phenanthroline, and MeCN is methyl acetate). These compounds served as precursors of efficient DRM catalysts, composites containing metallic nickel or cobalt dispersed in a matrix of lanthanum or samarium oxides. However, during the DRM process, despite the thermodynamically favorable conditions of the process, their surface was subjected to significant carbonization which resulted in blocking of the gas flow in the reactor. Using a complex procedure for the supercritical antisolvent deposition of a complex compound--the precursor of perovskite SmCoO₃, the catalyst particle size was reduced and catalyst carbonization was avoided [29].

However, we have shown that the production of efficient DRM catalysts does not require the mandatory synthesis of a completely single-phase initial perovskite [34,35]. Composites containing, along with the perovskite phase, the phases of nickel, cobalt, and rare earth elements oxides can be synthesized using a simple evaporation of aqueous solutions of nickel, cobalt and rare earth elements (REE) salts. In the DRM process, these composites form catalysts consisting of metallic nickel or cobalt dispersed in an REE oxide. These catalysts make it possible to obtain synthesis gas with a yield above 90% but undergo significant carbonization.

In order to create a DRM catalyst less susceptible to carburization, we synthesized an oxide composite containing 2 wt. % cobalt. This material was prepared by the simple evaporation of an aqueous solution of cobalt and samarium nitrates, followed by calcination of the resulting material at 700^oC [36]. It was assumed that due to a decrease in the cobalt content in this material compared to the SmCoO₃ perovskite containing 23 wt. % cobalt metallic cobalt particles more resistant to sintering and subsequent carbonization would be obtained. It is known that the deposition of perovskite systems on various substrates often increases their efficiency in DRM catalysis [11,14,16,20]. It was supposed that the synthesized material would contain samarium cobaltate SmCoO₃ dispersed in a matrix of samarium oxide. However, the formed composite consisted of samarium oxide and samarium cobaltite, Sm₂CoO₄, which apparently resulted from the interaction of SmCoO₃ with an excess of samarium oxide.

It was found that this material after prereduction in CH₄ and CO₂ mixture or in hydrogen flow formed catalyst which demonstrated low efficiency in DRM. And only after a long-term exposure in the CH₄/CO₂ flow showed synthesis gas yields of 88-90 % at 900°C.

The aim of the present study is to find optimal conditions for the preparation of highly efficient stable and carbonization resistant Co/Sm₂O₃ catalysts with various cobalt content for hydrogen production by DRM.

2. Materials and Methods

2.1. Materials

We used the following commercial Sigma-Aldrich reagents: Co(NO₃)₂·6H₂O CAS 10026-22-9, and Sm(NO₃)₃·6H₂O CAS 13759-83-6.

2.2. Preparation of Catalyst

The designations of the synthesized materials--catalysts precursors--and the amounts of reagents used for their synthesis are given in Table 1. The amounts of reagents corresponded to the cobalt content in the resulting catalyst (Table 1). The Co(NO₃)₂•6H₂O and Sm(NO₃)₃•6H₂O samples were dissolved in 30 ml of distilled water under stirring in a glass beaker. The obtained solutions were heated under stirring until water evaporation. The resulting mass was transferred into an alundum crucible and heated in a Nabertherm muffle furnace for 3 h at 300^oC. The solid product was crushed, heated in the muffle furnace for 2.5 h to 800^oC, and kept for 2 h at this temperature.

2.3. Characterization

The specific surface area S_BET of the samples was measured by the method of low-temperature nitrogen adsorption on an ATKh-06 analyzer (Katakon, Russia). The samples were degassed in a nitrogen flow (1 atm) at 200^oC for an hour prior to analysis. Based on the data obtained, the specific surface area of the samples was calculated using the Brunauer–Emmett–Teller (BET) model and the five-points method in the partial pressure range of 0.05–0.25 P/P₀.

The powder X-ray diffraction study (XRD) of freshly prepared and spent catalysts was carried out on a Rigaku MiniFlex 600 diffractometer (CuKα radiation, detector with graphite monochromator, and Cu anticathode). The XRD data were processed using the database of the International Center for Diffraction Data (ICDD).

Thermogravimetric analysis (TGA) was performed on a TGA/DSC 3+ simultaneous thermal analyzer (Mettler Toledo) in an air flow in the range of 30–1000°C at a heating rate of 10°C/min. The TGA data were processed using the STARe Excellence software.

The micromorphology of all samples was studied by scanning electron microscopy (SEM) on a Carl Zeiss NVision 40 high-resolution microscope equipped with an Oxford Instruments X-MAX detector (80 mm²) and operating at an accelerating voltage of 1-20 kV. SEM images were taken in InLens (SE2) and ESB modes with an Everhart-Thornley detector at accelerating voltages of 1 and 7 kV.

The microstructure and elemental mapping of some spent catalysts were investigated on a TJEOL JEM-2100 UHR transmission electron microscope (TEM) operating at an accelerating voltage of 200 kV. Sample powders were dispersed in ethanol and dropped onto a TEM cooper grid (Ted Pella, Inc.). TEM micrographs were collected in the bright-field mode using an Olympus Quemesa 11-megapixel CCD camera. The dark-field sample images and elemental mapping were performed in the STEM mode.

The temperature-programmed reduction (H₂-TPR) was performed in a flow quartz reactor with an inner diameter of 2 mm at a heating rate of 7.5°C/min. Temperature was measured with the aid of a chromel-alumel thermocouple (K-type). The Н₂/Ar flow (5 vol. % H₂) velocity was 50 ml/min. The H₂ content in the outlet gas flow was assessed on a Krystallyuks-4000M chromatograph equipped a thermal conductivity detector.

In addition, the H₂-TPR of the catalysts was carried out in the same mode on an USGA-101/M3 chemisorption analyzer (LLC UNISIT, Russia).

2.4. Catalytic Experiments

The catalytic performance of 2%Co/Sm₂O₃, 5%Co/Sm₂O₃, 10%Co/Sm₂O₃, 23%Co/Sm₂O₃ materials in the DRM process was tested at atmospheric pressure in a fixed bed flow quartz reactor (inner diameter 18 mm) with an axial pocket for thermocouple (outer diameter 8 mm). The tip of the thermocouple (chromel-alumel (K)) was positioned in the center of the catalyst layer. Powdered catalysts precursors were pressed into pellets, grinded, and a fraction with 0.5-1 mm grains was used in further experiments. The catalyst precursor (0.2 g, layer height 1 mm) was placed on a quartz fiber substrate. The catalyst precursor was heated to 900 ^oC in a nitrogen stream (JSC NIIKM, Russia, nitrogen content 99.999%). When the catalyst was heated in the nitrogen flow, after its supply was stopped, a gas mixture of CH₄ and CO₂ (JSC "Moscow Gas Processing Plant", Russia, purity at least 99.9%) was fed into the reactor. The ratio of CH₄ : CO₂ was 1, and the feed rate of the nitrogen or gas mixture was 15 L × g⁻¹ × h⁻¹. The gas velocity at the reactor inlet and outlet was measured with a foam flow meter. The temperature change in the reactor was carried out using a programmable temperature controller. The products were analyzed at a fixed temperature, and the temperature was adjusted to other set values without stopping the supply of reagents.

The composition of gaseous products was analyzed online by GLC on GALS 311 chromatographs equipped with thermal conductivity detectors using He as a carrier gas. The concentrations of H₂, CH₄, and CO were estimated at 30°C in a 2 m × 3 mm steel column packed with NaX zeolite. The H₂ concentration was calculated using a special calibration graph. To detect CH₄, CO₂, ethylene, and ethane a similar column filled by a Porapak Q at 70°С was used. The admixtures of C₂₊ hydrocarbons were determined at 70°C using a similar column filled by 5% Na₂CO₃ on alumina. The chromatographic data were analyzed using the EKOCHROM software (SKB of the Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences). All chromatograms were calculated by internal normalization with correction made for the molecular weights of the components.

2.4.1. Calculations

Methane conversion, X(CH₄), was calculated as

$${\rm X}(C{H}_4)=\frac{W_{in}(C{H}_4)-W_{out}(C{H}_4)}{W_{in}(C{H}_4)}\times 100\%,$$

where W_in(CH₄) is the quantity (mol) of CH₄ injected in the reactor and W_out(CH₄) is the quantity (mol) of CH₄ at the reactor outlet.

CO₂ conversion X(CO₂) were calculated similarly.

Hydrogen yield, Y(H₂), was calculated as

$$Y(H_2)=\frac{W_{out}(H_2)}{2W_{in}(C{H}_4)}\times 100\%,$$

where W_out(H₂) is quantity (mol) of H₂ at the reactor outlet and W_in(CH₄) is the quantity (mol) of CH₄ injected in the reactor.

CO yield, Y(CO), was calculated as:

$$Y(CO)=\frac{W_{out}(CO)}{W_{in}(C{H}_4)+W_{in}(C{O}_2)}\times 100\%$$

where W_out(CO) is the quantity (mol) of CO at the reactor outlet and W_in(CH₄) and W_in(CO₂) are, accordingly, the quantities (mol) of CH₄ and CO₂ injected in the reactor.

3. Results and Discussion

In specially blank experiments conducted in the reactor without a catalyst, a slight carbonization of reactor walls was detected mainly after a layer of quartz nozzle and quartz fiber. At the same time, the trace amounts of CO and hydrogen were recorded in exhaust gases along with unreacted reagents.

3.1. Characterization of freshly prepared materials

According to the XRD data, the synthesized sample of 2%Co/Sm₂O₃ contains phases Sm₂O₃ (ICDD 96-101-0341, 88 wt. %) and Sm₂CoO₄ (ICDD 96-200-2267, 12 wt. %) (Figure 1a). The phase content was calculated by the Rietveld method [37].

Sample 5%Co/ Sm₂O₃ (Figure 1b), along with Sm₂O₃ (ICDD 96-101-0341, 86 wt. %), contains the SmCoO₃ phase (ICDD 96-412-4856, 14 wt. %).

Sample 10%Co/ Sm₂O₃ (Figure 1c), along with Sm₂O₃ (ICDD 99-208-1967, 47 wt. %), contains SmCoO₃ (ICDD 99-204-5928, 53 wt. %).

Sample 23%Co/ Sm₂O₃ (Figure 1d), in which the amount of cobalt corresponds to its content in the SmCoO₃ perovskite, according to XRD data, contains only 35 wt. % of SmCoO₃ (ICDD 96-152-1745), as well as Sm₂O₃ (ICDD 96-101-0590, 42 wt. %) and Co₃O₄ (ICDD 96-153-8532, 23 wt. %). Thus, the samarium/cobalt ratio in the synthesized complex oxides significantly affects their phase composition.

The SEM images of the synthesized materials (Figure 2 a-d) show that the 2%Co/Sm₂O₃ sample (Figure 2a) is formed by flat particles with an undeveloped porous structure, which corresponds to a low specific surface area of the sample (Table 1).

In the images of 5%Co/Sm₂O₃ (Figure 2b) and 23%Co/Sm₂O₃ (Figure 2d) samples taken at a higher resolution, particles of a similar shape are seen; their lateral faces contain mesopores with a diameter of about 20 nm.

The 10%Co/Sm₂O_3` (Figure 2b) sample also contains flat particles with the undeveloped porous structure.

The H₂-TPR profile of 2%Co/Sm2O3 (Figure 3a) reveals the presence of three regions of hydrogen absorption. Less intense peaks with maxima at 310 and 400^oC are probably related to reduction of a small amount of Co₃O₄ contained in the catalyst [38], which was not recorded by XRD. An intense peak with a maximum at 540^oC, according to [11,31,39], can be assigned to the reduction of samarium cobaltite by reaction (10):

Sm₂CoO₄ + H₂ = Sm₂O₃ + Co + H₂O.

(10)

The H₂-TPR profile of 5%Co/Sm₂O₃ (Figure 3b) exhibits a low-intensity peak with maxima at 420^oC and an intense peak with a maximum at 530 ^oC, which, according to [11,31,39], can be attributed to the staged reduction of samarium cobaltate by reactions (11) and (12):

2SmCoO₃ + H₂ = 2SmCoO_2,5 + H₂O

(11)

2SmCoO_2,5 + 2H₂ = Sm₂O₃ + 2Co + 2H₂O

(12)

The H₂-TPR profiles of 10%Co/Sm₂O₃ (Figure 3c) and 23%Co/Sm₂O₃ (Figure 3d) are recorded by the USGA analyzer. The H₂-TPR profile of 10%Co/Sm₂O₃ shows two intense peaks with maxima at 390 and 532^oC, which, according to [11,31,39], can also be attributed to the staged reduction of samarium cobaltate by reactions (11) and (12).

The H₂-TPR graph of 23%Co/ Sm₂O₃ exhibits a peak with a maximum at 318^oC corresponding to the reduction of cobalt oxides and an intense peak with a maximum at 419^oC, which can be assigned to the reduction of both cobalt oxides and samarium cobaltate by reaction (11). A less intense peak with a maximum at 517^oC probably corresponds to reaction (12). In general, the H₂-TPR data obtained for the synthesized samples are consistent with the XRD analysis (Figure 1 and comments).

3.2. Results of Catalytic Experiments

As indicated in our previous paper [36], it takes a long time for the 2%Co/Sm₂O₃ catalyst preheated to 900^oC in the flow of CH₄/CO₂ (1 : 1) or hydrogen flow to reach the optimal operating mode. In the present work, 2%Co/Sm₂O₃ was preheated to 900 ^oC for an hour in a high-purity nitrogen flow. The results are shown in Figure 4.

It can be seen that in contrast to [36] results, immediately after the mixture of CH₄ and CO₂ was fed to the reactor, the yield of the synthesis gas (CO : H₂ = 1 : 1) equal to 96 was achieved. The conversion of CH₄ was 97%, and the conversion of CO₂ was 99%. In a long-term experiment with intermediate cooling and reheating to 900^oC in the high-purity nitrogen, the stable performance of the catalyst for 50 h was observed. The values of CO₂ conversion and CO yield higher than the values of methane conversion and hydrogen yield are apparently due to an insignificant reverse water gas shift reaction (6) along with the DRM reaction (3). Since, in accordance with [36], in the case of the 2%Co/Sm₂O₃ catalyst, a decrease in temperature leads to a decrease in the conversion of reagents and the yields of products, this catalyst was not tested at temperatures less than 900^oC. Our results demonstrate that a highly efficient and stable DRM catalyst is formed in situ immediately after the contact of 2%Co/Sm₂O₃ preheated to 900 ^oC in N₂ flow with the CH₄-CO₂. mixture.

The 5%Co/Sm₂O₃ and 10%Co/Sm₂O₃ samples, which were also preheated to 900^oC for an hour in the high-purity nitrogen flow, as well as the 2%Co/Sm₂O₃ sample, also immediately formed catalysts which showed high yields of CO and H₂ (Figure 5 and Figure 6).

The data in Figure 5 indicate that for the catalyst derived from 5%Co/Sm₂O₃ the CH₄ conversion is 95-98%, the CO₂ conversion is 100%, the CO yield is 97-98%, and the H₂ yield is 95-97%. When temperature was reduced to 800°C, the CH₄ conversion decreased to 83-84%, the CO₂ conversion decreased to 90-91%, the CO yield decreased to 81-83%, and the H₂ yield decreased to 82%. At 700^oC, the CH₄ conversion was 46-47%, the CO₂ conversion was 58-59%, the CO yield was 47-49%, and the H₂ yield was 39-41%, while at 600^oC almost no DRM was observed. A subsequent increase in temperature to 900^oC restored the activity of the catalyst. When the experiment was continued up to 50 h, high values of CH₄ conversion (94-96%), CO₂ conversion (99-100%), CO yield (94-97%), and H₂ yield (94-96%) were maintained. The observed values of CO₂ conversion and CO yield higher than those of methane conversion and hydrogen yield also indicate that, along with the DRM reaction (3), an insignificant reverse water gas shift reaction (6) occurs and its contribution increases with decreasing temperature.

According to Figure 6, the catalyst derived from 10%Co/Sm₂O₃ demonstrates similar results in the DRM reaction. After initial heating in nitrogen flow to 900^oC, and the subsequent supply of a mixture of CH₄ and CO₂ to the reactor, the CH₄ conversion achieves 97-98% and the CO₂ conversion is as high as 100%. The CO and H₂ yields are 94-95 and 95%, respectively. When temperature was reduced to 800 C, the CH₄ conversion decreased to 84%, the CO₂ conversion decreased to 89-92%, the CO yield decreased to 82-85%, and the H₂ yield decreased to 81-82%. At 700^oC a CH₄ conversion of 47%, a CO₂ conversion of 57-61%, a CO yield of 46-47%, and a H₂ yield of 38-40% were observed, while at 600oC almost no DRM occurs. The subsequent increase in temperature to 900°C also restores the activity of the catalyst. When the experiment was continued to 50 h, the high values of CH₄ conversion (94-98%), CO₂ conversion (97-100%), CO yield (95-98%), and H₂ yield (94-98%) were maintained. The higher values of CO₂ conversion and CO yield are also consistent with an insignificant occurrence, along with the DRM reaction (3), of the reverse water gas shift reaction (6), the contribution of which increases with decreasing temperature.

For comparison, we synthesized and tested a 23%Co/Sm₂O₃ material in the DRM reaction (Figure 7). The same material was previously tested in the DRM reaction after preheating in the reagent stream [35]. According to [35], at 900^oC the formed catalyst shows a synthesis gas yield close to quantitative but contains 44.5 wt. % of carbon deposits. Taking into account the data from [35] and the tests of catalysts based on 2%Co/Sm₂O₃, 5%Co/Sm₂O₃, and 10%Co/Sm₂O₃, in the present work, this material was tested in the DRM reaction after preheating to 900^oC for an hour in the high-purity nitrogen stream. Figure 7 demonstrates that for 55 h the catalyst formed from 23%Co/Sm₂O₃ retained 95-98% CH₄ conversion, 98-100% CO₂ conversion, 95-98% CO yield, and 94-97% H₂ yield. At the same time, the data in Figure 7 confirm that, along with the DRM reaction (3), an insignificant reverse the water gas shift reaction (6) proceeds.

The catalysts formed during the DRM process and discharged from the reactor were investigated by XRD, TGA and TEM methods.

The XRD data of the spent catalysts are shown in Figure 8.

The XRD pattern of the spent catalyst derived from 2%Co/Sm₂O₃ (Figure 8a) shows only reflections due to cubic Sm₂O₃ (ICDD 96-101-0590) and rhombic Sm₂O₃ (ICDD 96-153-0725). The absence of reflections due to cobalt and its compounds may be explained by a small number of particles and their small size.

The XRD pattern of the spent catalyst derived from 5%Co/Sm₂O₃ (Figure 8b) exhibits, along with reflections due to cubic Sm₂O₃ (ICDD 96-901-5549) and rhombic Sm₂O₃ (ICDD 96-153-0725), reflections corresponding to metallic cobalt (ICDD 96-901-0969). In this case, the size of cobalt particles could not be correctly estimated according to the Debye-Scherrer formula.

The XRD pattern of the spent catalyst derived from 10%Co/Sm₂O₃ (Figure 8c) contains reflections due to cubic Sm₂O₃ (ICDD 96-154-2110), rhombic Sm₂O₃ (ICDD 96-153-0725), and metallic cobalt (ICDD 96-900-8467). In accordance with the Debye-Scherrer formula, the size of cobalt particles is 52 and 56 nm.

The XRD pattern of the spent catalyst derived from 23%Co/Sm₂O₃ (Figure 8d) shows reflections related to cubic Sm₂O₃ (ICDD 99-201-7616), rhombic Sm₂O₃ (ICDD 99-202-8230), and metallic cobalt (ICDD 99-101-0059). In accordance with the Debye-Scherrer formula, the size of cobalt particles is 43 and 48 nm.

Thus, the XRD data (Figure 8) indicate that the true DRM catalysts are Co/Sm₂O₃ composites of metallic cobalt dispersed in samarium oxide, which are derived from 2%Co/Sm₂O₃, 5%Co/Sm₂O₃, 10%Co/Sm₂O₃, and 23%Co/Sm₂O₃. It should be noted that all the formed composites maintain high activity in DRM for at least 50 h.

The TGA data obtained for the spent catalysts are shown in Figures 9a-d.

All the samples are characterized by a slight initial weight loss associated with the removal of adsorbed water and gases upon heating to 300^oC. At higher temperatures, marked differences in TGA profiles are observed.

The TGA profile of the spent catalyst derived from 2%Co/Sm₂O₃ (Figure 9a) shows that upon heating from 350 to 450 ^oC, the weight slightly increases apparently due to the oxidation of metallic cobalt. At 450-650^oC the weight decreases by 0.26%, which correlates with the combustion of carbonaceous deposits and the decomposition of carbonates. A slight weight loss at 650-750^oC may be attributed to the combustion of carbon, and a subsequent increase in weight can be explained by the resynthesis of cobalt-samarium complex oxides.

The TGA profile of the spent catalyst derived from 5%Co/Sm₂O₃ (Figure 9b) indicates that upon heating from 320 to 440^oC weight increases by 0.42% evidently due to the oxidation of metallic cobalt. At 440-730^oC the weight decreases by 4%, in correlation with the combustion of carbonaceous deposits and the decomposition of carbonates. A slight weight loss at 650-750^oC may also be associated with the combustion of carbon. A subsequent slight increase in weight is presumably caused by the resynthesis of cobalt-samarium complex oxides.

The TGA profile of the spent catalyst derived from 10%Co/Sm₂O₃ (Figure 9c) shows that upon heating from 260 to 460^oC weight increases by 2%, which is apparently related to the oxidation of metallic cobalt. At 460-720^oC the weight loss is 2.2%. This is apparently associated with the combustion of carbonaceous deposits and the decomposition of carbonates. Above 720^oC, weight increases by 0.31% which is evidently due to the resynthesis of cobalt-samarium complex oxides.

The TGA data on the spent catalyst derived from 23%Co/Sm₂O₃, (Figure 9d) demonstrate that upon heating from 260 to 460^oC an increase in weight is as low as 0.54%. This can apparently be explained by oxidation of metallic cobalt. At 460-620^oC weight decreases by 20.72%. This is apparently due to the combustion of amorphous carbonaceous deposits and the decomposition of carbonates. Heating from 620 to 720^oC resulted in a weight loss of 23.87% probably due to the combustion of graphite-like carbon. Above 720^oC, there is a 3.5% increase in weight, which is apparently related to the resynthesis of cobalt-samarium complex oxides.

Thus, the TGA data of all the studied samples indicate the presence of metallic cobalt in the formed composites. The TGA data suggest that the catalyst formed from 2%Co/Sm₂O₃ is almost not subject to carburization. Catalysts based on 5%Co/Sm₂O₃ and 10%Co/Sm₂O₃ are carbonized to a small extent, whereas the catalyst based on 23%Co/Sm₂O₃ undergoes strong carbonization although it does not lose activity in DRM for more than 50 h.

The TGA results are consistent with the SEM data (Figure 10 a-d).

The spent catalyst derived from 2%Co/Sm₂O₃ (Figure 10a) contains almost no carbon deposits. The spent catalysts based on 5%Co/Sm₂O₃ (Figure 10b) and 10%Co/Sm₂O₃ (Figure 10c), underwent noticeable carbonization, and the surface of the spent catalyst derived from 23%Co/Sm₂O₃ (Figure 10d) is almost completely covered with carbon deposits.

Spent catalysts with the lowest cobalt content, which were derived from 2%Co/Sm₂O₃ and 5%Co/Sm₂O₃, were additionally investigated by the TEM method. The TEM micrograph of the spent catalyst based on 2%Co/Sm₂O₃ (Figure 11a) demonstrates formation of an insignificant amount of carbon nanotubes. Figure 11b shows that the catalyst contains cobalt particles with a size of about 20 nm.

Figure 12 presents the TEM data on the elemental composition and distribution of cobalt and samarium in the spent catalyst derived from 2%Co/Sm₂O₃. It can be seen that cobalt atoms are evenly distributed in samarium oxide.

For comparison, the spent catalyst derived from 5%Co/Sm₂O₃, which is more prone to carbonization, was also investigated by TEM (Figure 13).

The TEM micrographs shows that the sample contains non-surface-bound carbon nanotubes and cobalt-containing particles up to 50 nm in size. A higher tendency of this sample to coking may be explained by an increase in the size of cobalt-containing particles.

The study of the elemental composition of this material (Figure 14) also showed that the distribution regions of cobalt and samarium atoms in the sample are coincident.

Thus, the simple method of synthesizing catalyst precursors, namely, evaporation of aqueous solutions of cobalt and samarium nitrates with subsequent calcination of the resulting material enables one to obtain materials that are precursors of selective and stable catalysts of the DRM reaction. This method is much simpler than most of the known methods used to prepare highly efficient and stable catalysts of this reaction on the basis of cobalt-samarium precursors [7,11,29,30,31], including single-phase perovskite SmCoO₃ [29,30,31].

An effective way is developed to achieve high synthesis gas yields using the synthesized precatalysts which consists in heating samples in a nitrogen flow to an optimal temperature of 900^oC. This DRM procure does not require catalyst pre-reduction by hydrogen or a CH₄/CO₂ mixture. The stable operation of the catalysts lasts for 50 h, and the yields of H₂ and CO close to the thermodynamically predicted limits are achieved. The TGA and SEM data indicate that a high cobalt content in the precatalyst is not required to create a selective, stable, and carbonization resistant DRM catalyst. An increase in the cobalt content in the samples does not affect their stability for 50 h but significantly increases carbonization, which may hinder a longer stable operation of DRM catalysts.

4. Conclusions

The new simple method has been developed for the synthesis of materials that are effective precursors of catalysts for the production of hydrogen and carbon monoxide by the DRM reaction. It is shown that the amount of cobalt introduced into the precatalyst affects its phase composition. For example, according to XRD data, the 2%Co/Sm₂O₃ sample contains Sm₂O₃ and Sm₂CoO₄ phases. Materials based on 5%Co/Sm₂O₃ and 10%Co/Sm₂O₃ contain SmCoO₃ along with Sm₂O₃. The 23%Co/Sm₂O₃ sample containing the same amount of cobalt as SmCoO₃ perovskite consists of SmCoO₃, Sm₂O₃, and Co₃O₄.

It has been demonstrated that the formation of efficient and stable DRM catalysts based on the synthesized materials does not require their pre-reduction with hydrogen or the initial mixture of reagents. The heating of the synthesized materials up to 900^oC in a nitrogen flow and the subsequent supply of the equimolar CH₄/CO₂ mixture allow one to in situ obtain a selective and stable DRM catalyst operating for at least 50 h.

It has been shown that the cobalt content in the pre-catalyst significantly affects the carbonization resistance of the formed DRM catalyst. For example, the catalyst derived from 2%Co/Sm₂O₃ is not subject to carbonization when tested in DRM for 50 h. The catalysts based on 5%Co/Sm₂O₃ and 10%Co/Sm₂O₃ undergo noticeable carbonization, which, however, does not worsen the results of DRM for 50 h. The 23%Co/Sm₂O₃ derived catalyst containing the same amount of cobalt as perovskite SmCoO₃ maintains stability in the DRM reaction for 50 h but forms a significant amount of carbon, which likely affects its stability during long-term operation. This implies that a cobalt content of 2 wt. % is optimal for obtaining the stable cobalt-samarium DRM catalyst, provided its preheating to 900^oC in a nitrogen flow followed by the supply of reagents. The stability of this catalyst is associated with the formation of sintering-resistant cobalt metal particles with a size of about 20 nm. Similar observations were made in the study of DRM using supported nickel and nickel-cobalt catalysts containing these metals in an amount of no more than 2 wt. % [5,40].

In general, the catalyst formed from the 2%Co/Sm₂O₃ material is promising for practical use, since it is almost not subject to carbonization and sintering of cobalt particles and demonstrates consistently high yields of H₂ and CO. Its application in the DRM reaction makes it possible to efficiently utilize two greenhouse gases, methane and carbon dioxide, which is of great ecological importance. In addition, it can potentially be used to produce "green" hydrogen by the conversion of renewable raw material--biogas.

6. Patents

Russian Patent Application RU 2023 113 594 A (25.05.2023).