Flexible NiRu Systems for CO2 Methanation: From Efficient Catalysts to Advanced Dual-Function Materials

CO2 emissions in the atmosphere have been increasing rapidly in recent years, causing global warming. CO2 methanation reaction is deemed to be a way to combat these emissions by converting CO2 into synthetic natural gas, i.e., CH4. NiRu/CeAl and NiRu/CeZr both demonstrated favourable activity for CO2 methanation, with NiRu/CeAl approaching equilibrium conversion at 350 °C with 100% CH4 selectivity. Its stability under high space velocity (400 L·g−1·h−1) was also commendable. By adding an adsorbent, potassium, the CO2 adsorption capability of NiRu/CeAl was boosted, allowing it to function as a dual-function material (DFM) for integrated CO2 capture and utilisation, producing 0.264 mol of CH4/kg of sample from captured CO2. Furthermore, time-resolved operando DRIFTS-MS measurements were performed to gain insights into the process mechanism. The obtained results demonstrate that CO2 was captured on basic sites and was also dissociated on metallic sites in such a way that during the reduction step, methane was produced by two different pathways. This study reveals that by adding an adsorbent to the formulation of an effective NiRu methanation catalyst, advanced dual-function materials can be designed.


Introduction
The reduction in carbon dioxide (CO 2 ) emissions is considered the way forward so to decrease the impact of greenhouse gases on climate and thus combat global warming. The amount of CO 2 in the atmosphere has reached 421 ppm, which is 50% higher than preindustrial levels [1]. The use of renewable energy sources to create carbon-free electricity is an effective method of dealing with CO 2 emissions from power generation. Electric power generated by renewable sources, such as wind and solar, is variable and fluctuating; therefore, long-term storage solutions are sought [2][3][4].
The use of power-to-gas (PtG) schemes can be a solution to this issue because PtG can make use of excess electricity from renewable sources to electrolyse water and produce hydrogen (H 2 ) [4][5][6][7]. Methane (CH 4 ) is generally considered a way to store the surplus H 2 , owing to the safety issues of H 2 during handling and transportation [8]. CH 4 can be produced when CO 2 and H 2 react during CO 2 methanation or Sabatier reaction (Equation (1)). The CO 2 used in this reaction can be obtained by industrial flue gases when carbon capture is implemented; therefore, the CH 4 or synthetic natural gas (SNG) resulting from the PtG scheme may be carbon-negative [4,[8][9][10][11][12][13].
SNG is a promising alternative fuel because it can be easily incorporated into existing industrial and transport infrastructure. The Sabatier reaction attracted a lot of attention A tubular fixed-bed quartz reactor (0.4 in ID) was used in all the experiments carried out in this study. The reactor was placed vertically in a furnace, and the catalysts were supported by quartz wool. All the experiments were conducted at atmospheric pressure, and the temperature was monitored by a thermocouple. The outlet gases, CO 2 , CH 4 , CO, and H 2 , were monitored by infrared and thermal conductivity detectors on an ABB AO2020 online gas analyser. A flow meter was employed so as to measure the outlet total volumetric flow rate of the gases. A water condenser was placed downstream of the reactor to collect all the water formed during the experiments. All the pipelines were well insulated to ensure that water did not condense before reaching the condenser.

Continuous CO 2 Methanation Experiment
In this work, the activity of NiRu/CeZr and NiRu/CeAl was tested in the CO 2 methanation reaction. In each test, 0.125 g of sample was reduced at 850 • C for 1 h under a 10% H 2 /N 2 flow at a rate of 50 mL min −1 . The catalytic activity was measured every 50 • C in the temperature range of 500 to 200 • C. The reactant inlet ratio of CO 2 /H 2 /N 2 was 1:4:5, with the weight hourly space velocity (WHSV) set to 24 L·g −1 ·h −1 . The total outlet flow rate was measured at every temperature of the activity experiment with a flow meter, which was located after the reactor in sequence. This flow meter was used to allow for accurate measurements of the total volumetric flow rate and to take into account the change in the volume of the reaction mixture. The CO 2 conversion error was ±5%, and each sample was tested twice. Equilibrium conversions were obtained via Chemstations ChemCad using the same inlet flow rates as those used in the activity experiment. Soave-Redlich-Kwong was selected as the equation of state, and a Gibbs reactor was used.
The equations used to calculate the CO 2 conversion, CH 4 selectivity, and CH 4 yield are shown below (2)(3)(4). The flow rates of the respective gases are symbolised by F, and the subscripts in and out correspond to the inlet and outlet streams, respectively. CO 2 conversion (%) = F CO 2 , in − F CO 2 ,out F CO 2 , in × 100 (2) CH 4 selectivity (%) = F CH 4 ,out F CO 2 , in − F CO 2 ,out × 100 CO selectivity (%) = F CO,out F CO 2 , in − F CO 2 ,out × 100 (4) Arrhenius plots were obtained to determine the apparent activation energy (Ea) of NiRu/CeAl and NiRu/CeZr. A temperature range of 200 • C to 250 • C was used, with CO 2 conversion ranging from 5% to 25% and the activation energy determined based on the slope of the reaction rate vs. 1/temperature graph because the slope was equal to the activation energy divided by the universal gas constant (R). In this work, the apparent reaction rate is defined as the moles of the product (CH 4 ) formed per active site (Ni and Ru) and per unit of time. It was assumed that all the Ni and Ru atoms of the sample mass were active to enable approximation of the activation energies. The Arrhenius expression is presented below (Equation (5)), where k is the rate constant, T is the temperature in K, A is the pre-exponential factor, and R is the universal gas constant (R = 8.314 J mol −1 K −1 ). The detailed mathematical derivation can be found in Appendix A.

Long-Term Stability Test
The stability of the NiRu/CeAl catalyst in the CO 2 methanation reaction was tested in a continuous-flow quartz tube reactor, as described earlier, for 20 h. First, the sample was Nanomaterials 2023, 13, 506 5 of 18 reduced at 850 • C for 1 h under a 10% H 2 /N 2 flow; then, the temperature was set to 350 • C. The total flow was 200 mL min −1 , with 10% CO 2 , 40% H 2 and 50% N 2 . The WHSV was 400 L·g −1 ·h −1 , and the gas hourly space velocity (GHSV) was 12,000 h −1 . The aforementioned equations were used to measure the performance of the catalyst, and those high values were chosen to test its durability at a low reactor volume. The total outlet flow rate was measured every 30 min with a bubble meter, with the exception of night-time hours. A trend line was created based on these measurements, allowing for calculation of various parameters.

CO 2 Capture and Reduction Cycle
A cycle of CO 2 capture and methanation was performed for the NiRuK/CeAl in a similar manner, as described elsewhere [51] using 0.1 g of DFM. The initial reduction was carried out at 800 • C for 1 h under a 10% H 2 /N 2 flow at a rate of 50 mL min −1 . A slightly lower reduction temperature was used for the DFM compared to the two catalysts to optimise the reduction temperature and avoid sintering of the adsorbent, as it was demonstrated [51] that no reduction took place after 800 • C for the NiRuK/CeAl DFM. Subsequently, the reactor was cooled down to 350 • C under N 2 flow, and a cycle of CO 2 capture-N 2 purge-CO 2 methanation was carried out. The capture and methanation steps lasted 20 min each, and the purge step lasted 5 min. During the CO 2 capture step, 10% CO 2 in N 2 with a total flow rate of 50 mL min −1 was used, and during the CO 2 methanation step, 10% H 2 in N 2 with a total flow rate of 50 mL min −1 was used. The flow of N 2 was set to 45 mL min −1 throughout the cycle in order to be used as an internal standard to calculate the flow rates of the remaining gases according to Equation (6). The percentages of CO 2 , CH 4 , CO, and H 2 were recorded every 5 s by an ABB analyser. The amounts of CH 4 produced and CO 2 desorbed were calculated based on the area under the curve of the flow rate (mL min −1 ) vs. time (min) graph.

Time-Resolved Operando DRIFTS MS Experiment
Time-resolved operando DRIFTS-MS measurements were carried out in a hightemperature reaction cell supported in a Praying Mantis (Harrick) optical system with ZnSe windows. A Thermo Nicolet iS50 FTIR spectrometer equipped with a liquid-N 2 -cooled MCT detector was used to record the spectra at 4 cm −1 resolution with an average of 128 scans. The outlet gases passed through a mass spectrometer (Prisma plus from Pfeiffer Vacuum), and the data were logged with Quadera software. Only the NiRuK/CeAl sample was tested. In this experiment, approximately 50 mg of sample was loaded to the cell and reduced at 600 • C for 1 h (10 • C min −1 ) with a 10% H 2 /Ar flow at a rate of 50 mL min −1 . This reduction temperature was chosen to allow for better signal quality because higher temperatures lead to troubles with background noise. Then, the temperature was decreased to 250 • C with the same gas mixture, and 5 successive cycles of capture and reduction were performed, maintaining each step for 10 min. CO 2 capture was performed with 10% CO 2 /Ar and reduction with 10% H 2 /Ar. The total flow rate was 50 mL min −1 in both cases. That temperature was chosen to better observe the intermediates.

Promoter Effects in the Continuous Methanation of CO 2 over NiRu Catalysts
The NiRu/CeAl and NiRu/CeZr samples were tested under CO 2 methanation conditions after they had been reduced in situ. The results of the activity experiment are presented in Figure 1A. As CO 2 methanation is a highly exothermic reaction, the increase in the reaction temperature decreases CO 2 conversion [29,52]. It was observed that at low temperatures, i.e., at 200 • C, the conversion was less than 5%. CO 2 conversion reached a maximum value of 85% at 350 • C for the NiRu/CeAl and 68% at 400 • C for the NiRu/CeZr.

Long-Term Stability Test
Due to its higher activity, NiRu/CeAl was chosen to perform a long-term stability experiment. The reaction conditions were selected to be far from equilibrium conditions in order to better understand the catalytic behaviour. Moreover, good catalytic performance at high space velocities is favoured in industry because it is associated with a reduction in reactor volume and therefore lower capital cost.
The results of the stability experiment are shown in Figure 2. It was observed that the CO2 conversion remained stable at 60% during the 20 h experiment, indicating a great stability at such a high space velocity. The CH4 selectivity was 94% throughout the experiment without experiencing any drop. Accordingly, the CO selectivity was stable at 6%, and the carbon balance remained closed (100 ± 4%) during the 20 h experiment. It was therefore shown that the NiRu/CeAl was active and highly durable in the CO2 methanation reaction. Moreover, it was observed that the NiRu/CeAl exhibited better CH 4 selectivity of 100% up to 350 • C. On the other hand, the NiRu/CeZr had lower selectivity than NiRu/CeAl, reaching 98% in the 250-350 • C temperature range. As temperature increased, selectivity towards CO through the reverse water-gas shift reaction (RWGS) (Equation (7)) was enhanced because RWGS was favoured at a higher temperature due to its endothermic nature. CO 2 methanation and RWGS are competitive reactions, and careful catalyst design is needed for the desired reaction, in this case, the CO 2 methanation reaction [22].
Because the conversion of CO 2 into CH 4 is currently the subject of considerable investigation, Table 1 summarises the performance results of Ni-and Ru-based catalysts reported in the literature. In this study, the NiRu/CeAl catalyst achieved excellent performance in the CO 2 methanation reaction compared to similar catalysts, proving the advancement offered by this work for this reaction.
As previously mentioned, CO 2 methanation has a kinetic regime at low temperatures (T < 300 • C), and the CO 2 conversions are well below the equilibrium. Therefore, the apparent activation energy calculation based on product formation can be made in a differential reactor from the Arrhenius plots. The Arrhenius plots of the two catalysts in the 200-250 • C temperature range are presented in Figure 1B, with the apparent activation energies of NiRu/CeAl and NiRu/CeZr calculated to be 80.2 and 135.7 kJ mol −1 , respectively. It should be mentioned that these results were estimated based on two points because differential conditions were not achieved at higher temperatures. However, these estimations are in accordance with other literature data indicating that they are within the expected range. The Ea of NiRu/CeAl was comparable to previous literature findings of Ru-and NiRu-based catalysts estimated to be approximately 70 kJ mol −1 [8,20,29,53]. However, the Ea of NiRu/CeZr was more than 70% higher than that of NiRu/CeAl, similar to Ni-based catalysts reported in the literature [22,[54][55][56]. Therefore, it was demonstrated that there was a higher kinetic barrier for the NiRu/CeZr sample compared to that of the NiRu/CeAl sample. Nevertheless, the results showcase the superiority of the latter in the CO 2 methanation reaction.

Long-Term Stability Test
Due to its higher activity, NiRu/CeAl was chosen to perform a long-term stability experiment. The reaction conditions were selected to be far from equilibrium conditions in order to better understand the catalytic behaviour. Moreover, good catalytic performance at high space velocities is favoured in industry because it is associated with a reduction in reactor volume and therefore lower capital cost.
The results of the stability experiment are shown in Figure 2. It was observed that the CO 2 conversion remained stable at 60% during the 20 h experiment, indicating a great stability at such a high space velocity. The CH 4 selectivity was 94% throughout the experiment without experiencing any drop. Accordingly, the CO selectivity was stable at 6%, and the carbon balance remained closed (100 ± 4%) during the 20 h experiment. It was therefore shown that the NiRu/CeAl was active and highly durable in the CO 2 methanation reaction.

Promoter Effects on NiRu Catalysts for CO2 Methanation: The Effect of Surface Basicity
In general, CO2-TPD profiles are employed to assess the basicity of the materials and the CO2 adsorption sites. Depending on the temperature at which CO2 is desorbed, basic sites can be categorised as weak, medium, or strong. CO2 is desorbed from weak basic sites up to 250 °C, from medium basic sites between 250 °C and 700 °C, and from strong basic sites over 700 °C [51,56,59]. Because the typical methanation temperature is below 400 °C, the basic sites that are aimed for in this reaction are the weak and intermediate

Promoter Effects on NiRu Catalysts for CO 2 Methanation: The Effect of Surface Basicity
In general, CO 2 -TPD profiles are employed to assess the basicity of the materials and the CO 2 adsorption sites. Depending on the temperature at which CO 2 is desorbed, basic sites can be categorised as weak, medium, or strong. CO 2 is desorbed from weak basic sites up to 250 • C, from medium basic sites between 250 • C and 700 • C, and from strong basic sites over 700 • C [51,56,59]. Because the typical methanation temperature is below 400 • C, the basic sites that are aimed for in this reaction are the weak and intermediate ones because the strong sites are not expected to have high reactivity. Normally, the weak peaks correspond to the CO 2 bonded onto surface OH− groups, the medium peaks represent bidentate carbonates, and the strong peaks correspond to unidentate carbonates and strong basic surface O 2 − anions [55,60,61]. Figure 3 shows the CO 2 -TPD profiles of NiRu/CeAl and NiRu/CeZr. Both samples presented weak-intermediate basic sites. The first peak of both samples at ca. 80 • C and the second of peak the NiRu/CeZr at 120 • C correspond to CO 2 desorption from the weak Bronsted OH− groups. The peaks located at ca. 200 • C were assigned to the decomposition of bidentate carbonates formed on oxygen-metal pairs [55,60,61]. In addition, Figure 3 implies that there were more weak basic sites on the NiRu/CeAl sample based on the signal intensity. Because the basic sites that participated in this reaction were the weakintermediate sites, it was concluded that the NiRu/CeAl sample was able to provide more sites for CO 2 adsorption, leading to better catalytic activity. Therefore, its increased performance was attributed to its enhanced basicity, as reflected by the CO 2 -TPD results, in addition to its lower activation energy barrier.   Because NiRu/CeAl demonstrated good activity and stability in the CO2 methanation reaction, it was selected to be upgraded to a dual-function material by incorporating an adsorbent into its catalytic formulation. In this work, the chosen adsorbent was potassium. In addition to the CO 2 -TPD results presented herein, Table 2 summarises some important characterisation information about the two catalysts and the DFM obtained in our previous studies [26,40]. It can be concluded that besides the enhanced basicity and lower kinetic barrier of the NiRu/CeAl catalyst compared to the NiRu/CeZr catalyst, its higher surface area, better Ni dispersion, and smaller particle size were also crucial aspects that led to better catalytic activity.

Adding CO 2 Capture Functionality to Synthesise a Methanation DFM
Because NiRu/CeAl demonstrated good activity and stability in the CO 2 methanation reaction, it was selected to be upgraded to a dual-function material by incorporating an adsorbent into its catalytic formulation. In this work, the chosen adsorbent was potassium. The aim of the TGA experiment was to observe the capturing ability of the reduced Kbased DFM and, consequently, its best-performing temperature by simply ramping the temperature with CO 2 and Ar. An experiment with only Ar was also carried out on the reduced DFM to be used as reference for any material degradation and any weakly desorbed species. The results of this study are displayed in Figure 4. It was observed that the DFM was able to adsorb CO 2 immediately once it came into contact with the CO 2 -containing stream at 150 • C, with a maximum weight uptake slightly above that temperature, i.e., at 170 • C. At that temperature, the DFM adsorbed 0.013 mg of CO 2 /mg of sample or, alternatively, 0.3 mmol of CO 2 /g of sample. Moreover, it appeared that the DFM was able to keep that amount of CO 2 adsorbed onto its surface up to 400-500 • C. However, when the temperature was increased beyond that temperature range, a decline in the weight percentage was observed, indicating that the formed carbonates were not strong enough and that their decomposition started to take place, yielding K 2 O and CO 2 [62]. At the end of the experiment (950 • C), only a very small amount of CO 2 remained on the surface: 0.0003 mg of CO 2 /mg of sample or 2.4% of the initial CO 2 uptake, which may also be attributed to experimental error. decomposition started to take place, yielding K2O and CO2 [62]. At the end of the experiment (950 °C), only a very small amount of CO2 remained on the surface: 0.0003 mg of CO2/mg of sample or 2.4% of the initial CO2 uptake, which may also be attributed to experimental error. These results are in accordance with previous literature findings, which showed that K-based adsorbents are appropriate for low-intermediate temperature applications when T < 400 °C [42,51,[63][64][65][66][67]. Therefore, the use of K-based DFMs in the CO2 methanation reaction, which also occurs at low-intermediate temperatures, is desirable. It is worth noting that the current adsorption results should be compared with the results for other DFMs only, despite the potential existence of better CO2 adsorbents, because it is important to understand their adsorption behaviour under dynamic DFM operation even if the experiments are conducted under ideal conditions.

CO2 Capture and Reduction Study over NiRuK/CeAl DFM
As demonstrated by the capture experiment in TGA, the NiRuK/CeAl sample was able to capture CO2 in the typical methanation temperature range. In addition, the NiRu/CeAl catalyst showed impressive activity, selectivity, and stability in the methanation reaction, which motivated the development of a DFM using this catalyst. A cycle of CO2 capture, N2 purge, and CO2 methanation was performed in order to demonstrate the ability to 'transform' a typical methanation catalyst into a highly effective DFM. The temperature chosen for this experiment was 350 °C because at this temperature, adsorption was feasible according to the CO2 capture experiment in TGA, and the NiRu/CeAl catalyst These results are in accordance with previous literature findings, which showed that Kbased adsorbents are appropriate for low-intermediate temperature applications when T < 400 • C [42,51,[63][64][65][66][67]. Therefore, the use of K-based DFMs in the CO 2 methanation reaction, which also occurs at low-intermediate temperatures, is desirable. It is worth noting that the current adsorption results should be compared with the results for other DFMs only, despite the potential existence of better CO 2 adsorbents, because it is important to understand their adsorption behaviour under dynamic DFM operation even if the experiments are conducted under ideal conditions.

CO 2 Capture and Reduction Study over NiRuK/CeAl DFM
As demonstrated by the capture experiment in TGA, the NiRuK/CeAl sample was able to capture CO 2 in the typical methanation temperature range. In addition, the NiRu/CeAl catalyst showed impressive activity, selectivity, and stability in the methanation reaction, which motivated the development of a DFM using this catalyst. A cycle of CO 2 capture, N 2 purge, and CO 2 methanation was performed in order to demonstrate the ability to 'transform' a typical methanation catalyst into a highly effective DFM. The temperature chosen for this experiment was 350 • C because at this temperature, adsorption was feasible according to the CO 2 capture experiment in TGA, and the NiRu/CeAl catalyst exhibited the highest conversion according to the activity experiment. Moreover, at this temperature, the DFM could take advantage of the reversibility of the captured CO 2 so as to transform it into CH 4 more easily. The DFM mode of operation differs from typical adsorption processes, which require a change in temperature or pressure for the desorption step to occur. In particular, the fact that DFMs can operate under atmospheric pressure, thus avoiding pressure swing adsorption (PSA), is one of their main assets, leading to lower costs. It is worth noting that in comparison with the continuous CO 2 methanation experiments reported in previous sections, these cyclic experiments operate under dynamic conditions. Therefore, it is not obvious how the DFM will behave in terms of methanation activity compared to steady-state operation of the parent catalyst [40]. Figure 5 shows the results of the CO 2 capture and methanation experiment. It was noticed that CO 2 breakthrough occurred within the first 3 min of exposing the DFM to the CO 2 -containing stream. Breakthrough was estimated to have taken place within the first 10-15 s, but there was an anticipated time delay due to the setup of the pipelines and reactor. This material displayed the typical CO 2 adsorption curve, as adsorption includes a fast step followed by a slower step [51,68]. During methanation, only CH 4 and CO 2 were detected. NiRuK/CeAl displayed 100% CH 4 selectivity despite the fact that K is typically used as a promoter in the RWGS reaction [69][70][71][72][73]. The amounts of CH 4 produced and CO 2 desorbed were calculated based on the area under the curve in Figure 5. Hence, 0.264 mol of CH 4 /kg of sample was produced, and 0.139 mol of CO 2 /kg of sample was desorbed. The sum of the CH 4 moles formed and the CO 2 moles desorbed was higher than the moles of CO 2 adsorbed in the TGA experiment. However, this was expected because there was still some CO 2 in the gas phase before the methanation step started, as shown in Figure 5; moreover, the CO 2 adsorption temperature was constant at 350 • C in this experiment but not in the TGA experiment. Nevertheless, the sum of the CH 4 production and CO 2 desorption capacities were consistent with similar K-based DFMs found in the literature [40,74]. In addition, a delay in CO 2 desorption was detected, proving that the heat released from the methanation reaction supplied the required heat for CO 2 desorption to take place. Furthermore, it should be noted that CH 4 was detected in the first 3 min of the methanation step, indicating that the CO 2 capture and methanation steps can have the same duration in future experiments, as opposed to the fact that material regeneration has typically been a slower step in other DFM studies [74,75].
The methanation capacity obtained in this work is comparable to previous K-based DFM results reported in the literature or slightly lower [40,51,67,74]. However, a very important parameter that needs to be taken into consideration is the N 2 purge step. In the typical methanation temperature range, CO 2 is weakly adsorbed onto the surface of the DFM. This means that when N 2 purge takes place after the capture step, weakly adsorbed CO 2 is desorbed, affecting the DFM methanation capacity and demonstrating that the desorption capacity of DFMs affects their performance, in agreement with previous studies [51]. Consequently, in the future, it is advisable to perform experiments by tracking the outlet flow rates and shortening the times of capture and methanation steps so as to continuously clean the effluents from CO 2 and produce valuable products [62]. However, when performing O 2 -containing CO 2 capture followed by a hydrogenation reaction, a N 2 purge is required for safety reasons. Other uncommon impurities, such as NOx and SOx, which are present in real-world applications, also need to be taken into consideration, as they affect the DFM performance [76][77][78]. Furthermore, material recyclability allowing for the sustainable recovery of the main metals from the spent catalyst is another important aspect when designing such materials.

Time-Resolved Operando DRIFTS-MS Experimental Analysis
In order to achieve improved understanding of the surface intermediates involved in CO2 capture and reduction processes, we further performed an operando DRIFTS-MS study on the most optimal material, NiRuK/CeAl. For this purpose, the DFM was reduced in situ at 600 °C for 1 h in a flow of 50 mL min −1 of 10% H2/Ar. The repeated CO2 capture and reduction cycles were carried out by alternating a feed flow of 50 mL min −1 of 10% CO2 in Ar versus 10% H2 in Ar under isothermal conditions at 250 °C. Time-resolved DRIFTS spectra were successively recorded, and the dynamics evolution of the bands related to surface species was displayed in a 2D map, as shown in Figure 6A. It should be mentioned that all DRIFTS spectra were subtracted with respect to the spectrum of the surface after activation for clarification. As shown in Figure 6A, the development of bands ascribed to gaseous CO2 (2360 cm −1 ) and the symmetric/asymmetric stretching OCO vibrations characteristic of carbonate species were clearly visible under the CO2 capture feed [79]. In addition, it is worth observing the appearance of bands associated with CO adsorbed on metal sites, revealing that during the capture step, CO2 may be dissociated into CO* and O* adsorbed species. It is well known that ruthenium decreases the energy barrier for CO2 dissociation and facilitates the subsequent hydrogenation of CO* adsorbed, yielding CHx fragments [80]. Likewise, it is also possible to reduce CO2 to CO on the oxygen-deficient sites of ceria promoter and CO to migrate to metal sites [81]. After switching from the CO2 capture to H2 reduction stream, all the bands gradually decreased, and a new band emerged at 3014 cm −1 , which was attributed to gaseous methane.
A closer inspection of the DRIFTS spectra collected during the first cycle of capture/reduction is displayed in Figure 6B. As can be observed, the spectra recorded during the CO2 capture step (0-10 min) were dominated by the development of two intense bands in the 1800-1200 cm −1 region. These bands were ascribed to the symmetric and asymmetric

Time-Resolved Operando DRIFTS-MS Experimental Analysis
In order to achieve improved understanding of the surface intermediates involved in CO 2 capture and reduction processes, we further performed an operando DRIFTS-MS study on the most optimal material, NiRuK/CeAl. For this purpose, the DFM was reduced in situ at 600 • C for 1 h in a flow of 50 mL min −1 of 10% H 2 /Ar. The repeated CO 2 capture and reduction cycles were carried out by alternating a feed flow of 50 mL min −1 of 10% CO 2 in Ar versus 10% H 2 in Ar under isothermal conditions at 250 • C. Time-resolved DRIFTS spectra were successively recorded, and the dynamics evolution of the bands related to surface species was displayed in a 2D map, as shown in Figure 6A. It should be mentioned that all DRIFTS spectra were subtracted with respect to the spectrum of the surface after activation for clarification. As shown in Figure 6A, the development of bands ascribed to gaseous CO 2 (2360 cm −1 ) and the symmetric/asymmetric stretching OCO vibrations characteristic of carbonate species were clearly visible under the CO 2 capture feed [79]. In addition, it is worth observing the appearance of bands associated with CO adsorbed on metal sites, revealing that during the capture step, CO 2 may be dissociated into CO* and O* adsorbed species. It is well known that ruthenium decreases the energy barrier for CO 2 dissociation and facilitates the subsequent hydrogenation of CO* adsorbed, yielding CH x fragments [80]. Likewise, it is also possible to reduce CO 2 to CO on the oxygen-deficient sites of ceria promoter and CO to migrate to metal sites [81]. After switching from the CO 2 capture to H 2 reduction stream, all the bands gradually decreased, and a new band emerged at 3014 cm −1 , which was attributed to gaseous methane. electron cloud density on Ni metallic sites. Moreover, the absence of Ru carbonyl ban also indicated the interaction between Ni and Ru, meaning that a synergy between two active metals took place [24,30]. Alloy formation could be possible according to o previous work [51], but more studies are needed to confirm that. Nevertheless, the pr ence of carbonyl-adsorbed species reveals that CO2 suffered partial dissociation on me sites or reduction on oxygen vacancies. Next, the series of DRIFTS spectra collected during the following 10 min under 1 H2/Ar flow ( Figure 6B) showed that the features related to metallic carbonyl species r idly vanished with the concurrent production of gaseous methane (Figure 6B inset). It w also evident that the intensity of carbonate bands (1600-1200 cm −1 ) decreased with A closer inspection of the DRIFTS spectra collected during the first cycle of capture/reduction is displayed in Figure 6B. As can be observed, the spectra recorded during the CO 2 capture step (0-10 min) were dominated by the development of two intense bands in the 1800-1200 cm −1 region. These bands were ascribed to the symmetric and asymmetric stretching vibrations related to carbonate-like species with different adsorption geometry on the Al-O-K and Ce-O-K basic sites of the support. In more detail, the pair of bands at 1602-1323 cm −1 corresponded to bidentate carbonates (∆ν 3 splitting = 321), whereas the bands at 1560-1347 cm −1 (∆ν 3 splitting = 213) were associated with bridged carbonates [79,82]. It should be stressed that bands related to bicarbonate species were scarcely observed. The absence of bicarbonates can be explained by the fact that alkali promotion neutralises the hydroxyl groups on the alumina surface, which are responsible for the formation of bicarbonate species by reacting with CO 2 [83,84]. On the other hand, another band around 1753 cm −1 was also notable, which was ascribed to the formation of K 2 CO 3 over agglomerated potassium oxide particles when exposed to a CO 2 stream [85]. With respect to the bands developed at higher frequencies in the 2100-1800 cm −1 region, two bands at 2012 and 1890 cm −1 were clearly appreciable. These features were ascribed to linear and bridged carbonyl-adsorbed species, respectively, on nickel metallic sites [86]. Presumably, the observed vibrational frequencies were shifted to relatively lower values than those reported for nickel monometallic catalysts. This fact could be attributed to the presence of ruthenium, which induced electron transfer from Ru to Ni sites, resulting in higher electron cloud density on Ni metallic sites. Moreover, the absence of Ru carbonyl bands also indicated the interaction between Ni and Ru, meaning that a synergy between the two active metals took place [24,30]. Alloy formation could be possible according to our previous work [51], but more studies are needed to confirm that. Nevertheless, the presence of carbonyl-adsorbed species reveals that CO 2 suffered partial dissociation on metal sites or reduction on oxygen vacancies.
Next, the series of DRIFTS spectra collected during the following 10 min under 10% H 2 /Ar flow ( Figure 6B) showed that the features related to metallic carbonyl species rapidly vanished with the concurrent production of gaseous methane (Figure 6B inset). It was also evident that the intensity of carbonate bands (1600-1200 cm −1 ) decreased with the exposure time, but those species still remained after 10 min of reduction. This observation clearly showed that the reduction rate of basic carbonates was much slower than that of carbonyl species. Notably, the band at 1730 cm −1 exhibited a shift to low frequency with the simultaneous emergence of two weak bands at 2765 and 2680 cm −1 in the C-H stretching region ( Figure 6B inset). After 7 min of 10% H 2 /Ar exposure, all these bands decreased notably in intensity. Solymosi and Knözinger [87] observed similar features and reported that they were related to the formation of potassium formates from the reduction of potassium carbonates. It is worth mentioning that the associative mechanism of CO 2 methanation proposes that carbon dioxide reacts with surface hydroxyls on the support, producing bicarbonates that are progressively hydrogenated to methane passing through the formation of formates at the metal-support interface [88,89]. Our findings suggest that formates may also be formed on the potassium sites and subsequently be hydrogenated toward methane, following an alternative pathway in the absence of bicarbonates.
Moreover, Figure 6C shows the evolution of the IR band intensities related to carbonates (1323 cm −1 ), bridged carbonyls (1897 cm −1 ), linear carbonyls (2012 cm −1 ), and gaseous methane (3016 cm −1 ) as a function of the time on stream during the capture/reduction cycles. Understanding the evolution trend of adsorbed *CO species and carbonates is critical to determine the reduction mechanism of the captured CO 2 . It was evident that the intensity of both types of carbonyls decayed much faster than that of carbonates such that they can be distinguished between fast and slow reduction steps, as shown in Figure 6D. The linearly bounded *CO species decreased sharply and diminished within 4-5 min after switching the stream to 10% H 2 in an Ar feed. The bridged carbonyls were somewhat less reactive than the linear carbonyls, but they also decreased moderately fast. These observations are in line with the results reported by Vogt et al. [90], who found that linearly adsorbed CO is more weakly adsorbed than carbonyls in bridge conformation; thus, the rate-determining step in CO 2 methanation is governed by the concentration of linearly bounded CO species and their availability to be hydrogenated into methane. Meanwhile, the intensity of the band of carbonates depleted more slowly, remaining after 10 min of reduction. Simultaneously, it was evident that methane was produced rapidly during the first minutes, and its production decayed progressively. Even if some carbonates species remained on the surface after 10 min, most were eventually converted into CH 4 , as shown in Figure 6C. Although we could not rule out the presence of inactive carbonate species, the initial rapid production of methane, along with a slight decay in production, was consistent with the coexistence of fast (carbonyl) and slow (carbonate) reduction pathways, as depicted in Figure 6D.
In summary, we can rationalise that during the capture step, CO 2 is dissociated on metal sites or reduced in oxygen-deficient sites, producing linear and bridged carbonyladsorbed species. On the other hand, CO 2 is also adsorbed on the support basic sites, forming carbonate-like species. In the reduction step, carbonyl species are rapidly hydrogenated into CH 4 , whereas carbonates are more stable and require a longer time to be reduced. This study provides a primary approach to understanding the capture/reduction process at the fundamental level so as to design more efficient dual-function materials. In addition, the operando DRIFTS-MS results reported herein demonstrate that our DFM achieved promising results in terms of material stability, and more studies are currently underway to better comprehend its behaviour over time, especially under non-realistic conditions.

Conclusions
Two catalysts, NiRu/CeAl and NiRu/CeZr, were tested in the CO 2 methanation reaction. It was observed that the NiRu/CeAl sample was more active and selective in the CO 2 methanation reaction compared to the NiRu/CeZr sample, as it almost reached equilibrium conversion at 350 • C with 100% CH 4 selectivity. Furthermore, during the stability experiment of the NiRu/CeAl at a high space velocity, i.e., 400 L g −1 h −1 , which lasted 20 h, no drop in CO 2 conversion was detected.
The NiRu/CeAl catalyst was upgraded by incorporating an adsorbent, i.e., potassium, so as to obtain a dual-function material. The sample was tested in a cycle of CO 2 capture-N 2 purge-CO 2 methanation, exhibiting a methanation capacity of 0.264 mol of CH 4 /kg of sample and CO 2 desorption capacity of 0.139 mol of CO 2 /kg of sample. By carrying out time-resolved operando DRIFTS, it was demonstrated that CO 2 was dissociated on the metallic sites and reduced on the oxygen-deficient sites of Ce. Moreover, K 2 CO 3 and carbonate-like species were observed on the Al-O-K and Ce-O-K basic sites of the support. During CO 2 methanation, it was demonstrated that the carbonyl and carbonate species were converted into CH 4 at different rates.
Overall, in this work, we have investigated the effectiveness of Ni-Ru catalysts for the CO 2 methanation reaction and performed kinetic and mechanistic analyses. Notably, the comprehension of the CO 2 adsorption and subsequent methanation mechanisms of our dual-function material was accomplished. Further fundamental understanding is nevertheless needed to fine-tune the formulation of these materials. In any case our dualfunction materials are able to both capture and convert CO 2 into synthetic natural gas showcasing an innovative approach to address current environmental issues. Funding: This research was funded by the Department of Chemical and Process Engineering/School of Chemistry and Chemical Engineering and the Doctoral College of the University of Surrey. This work was partially sponsored by the Spanish Ministry of Science and Innovation through the project PLEC2021-008086 sponsored by MCIN/AEI/10.13039/501100011033 Next Generation Europe and Junta de Andalucía PAIDI programme through the projects P20_00667 and P20_00594. L.P. Merkouri acknowledges the Erasmus + programme, which allowed her to conduct part of the experiments in Sevilla, Spain.
Data Availability Statement: Original data will be available upon request.