Size-tailored Ru nanoparticles deposited over γ-Al 2 O 3 for the CO 2 methanation reaction

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published by Elsevier in Applied Surface Science on 31 July 2019, available at: https://doi.org/10.1016/j.apsusc.2019.03.248 ” Page 1 of 49 Size-tailored Ru nanoparticles deposited over -Al2O3 for the CO2 methanation reaction Authors Names: Sara Navarro-Jaén, Juan C. Navarro, Luis F. Bobadilla, Miguel A. Centeno, Oscar H. Laguna, José. A. Odriozola Authors Address: Departamento de Química Inorgánica, and Instituto de Ciencia de Materiales de Sevilla (Centro Mixto Universidad de Sevilla-CSIC). Av. Américo Vespucio 49, Seville 41092, Seville – Spain Keywords: Carbon capture and utilization (CCU); CO2 methanation; Sabatier reaction; Highpressure effect; Methane production; Ru nanoparticles; Ru/Al2O3 catalysts; CO2 methanation mechanism Corresponding author: Oscar H. Laguna e-mail: olag[email protected] Phone: +34 954489543 ORCID: 0000-0003-0747-069X Page 2 of 49 Abstract By means of the polyol method, a series of 5 wt.% Ru/Al2O3 catalysts was synthesized controlling the particle size of the ruthenium species. The physico-chemical characterization demonstrated the successful particle size control of the Ru species, in such a way that higher the Ru/PVP ratio, higher the Ru particle size. Moreover, there are evidences that suggest preferential growth of the RuO2 clusters depending on the Ru/PVP ratio. Regarding the catalytic activity during the CO2 methanation, the total conversion and the CH4 yield increased with the particle size of Ru. Nevertheless, a considerable enhancement of the catalytic performance of the most active system was evidenced at 4 bar, demonstrating the improvement of the thermodynamics (superior total conversion) and kinetics (superior reaction rate) of the CO2 methanation at pressures above the atmospheric one. Finally, the insitu DRIFTS study allowed to establish that CO2 was dissociated to CO* and O* species on the metallic Ru particles, followed by the consecutive hydrogenation of CO* towards CHO*, CH2O*, CH3O*, and finally CH4 molecules, which were further desorbed from the catalyst. Thus from the mechanistic point of view, a suitable particle size of the Ru nanoparticles along with the high-pressure effects results in the enhancement of the availability of hydrogen and consequently in the formation of CHxO species that enhance the cleavage of the C-O bond, which is the ratedetermining step of the overall CO2 methanation process. 1. Introduction The formulation of actions to curb the effects of climate change, beyond a scientific and technological problem, recently transcended the political scene after the Paris Agreement in 20151, 2. This is why, new policies around the world are being formulated such as the European Climate Policy3 that has proposed investment for a transition to low-carbon Page 3 of 49 economy, in order to reduce the emission of green-house gases in every possible scenario, within a plan with check points in 2030 and 2050. For instance, in the case of the domestic greenhouse emissions, by 2030 a reduction 40% of the highest levels, emitted in 1990, is expected. Furthermore, relevant sectors such as transport and industry also have to develop innovative technologies in order to fulfill with the requirements of using renewable energy sources and reducing emissions. For those purposes, strategies such as the deployment of electric mobility or Carbon Capture and Storage (CCS) have to be improved3-5. CCS technologies imply CO2 capturing instead of releasing this to the atmosphere, followed by a permanent storage underground in geological formations. Nevertheless, CO2 may be considered as a renewable raw material for the production of energy carriers, chemicals or materials, thus allowing the arise of a different approach, which is the Carbon Capture and Utilization (CCU)5. The potential of CCU as a climate mitigation alternative strongly depends on the balance between the benefits and the required energy input that also implies certain emission. Therefore, two main conditions have to be fulfill in order to ensure the suitability of a CCU technology: i) CO2 must come from fossil or atmospheric sources and ii) the energy input in the CO2 transformation must come from low-carbon or renewable sources5. However, for the accomplishment of these conditions, the processes for transforming CO2 into valuable products, including catalytic processes, have to be modernized and optimized in order to ensure their sustainability. Catalytically, there are relatively recent approaches such as the use of the Fischer-Tropsch Synthesis (FTS) as a suitable process for the valorization of CO2 coupled with the Reverse Water-Gas Shift reaction (R-WGS)6, 7 for obtaining long chain hydrocarbons. However, there are other widely used reactions such as the CO2 hydrogenation into oxygenates and/or hydrocarbons for obtaining fuels, which is the case of the CO2 hydrogenation Page 4 of 49 following the Sabatier reaction (Eq.1). This is especially attractive since it is the most advantageous reaction regarding thermodynamics, and because the transport and distribution of the produced methane may be carried out in the existing infrastructure of natural gas8, 9. CO2+ 4H2  CH4 + 2H2O ∆H25ºC = -165 kJ/mol Eq. 1 Although this exothermic reaction has been widely used in the production of ammonia for removing CO2 from the hydrogen employed as raw material, the new trends regarding the valorization of CO2 in order to reduce its emission to the atmosphere, has motivated new interest on this reaction. In fact, Jalama9, in a comprehensive review about kinetic and mechanistic aspects of Ni-, Ru- and Cu-based catalysts for the CO2 methanation, recently pointed out the need of further studies for acquiring fundamental understanding on this topic. For this purpose, not only the formulation of the catalysts may be analyzed, but also the possible effects of the reaction conditions such as the composition of the feedstream (H2/CO2 molar ratio, the presence of H2O), the effect of the space velocity and the occurrence of possible side reactions such as CO methanation, inversed methane CO2 reforming, water-gas shift (WGS), reverse-water-gas shift (R-WGS), and methane cracking, among others10. Within the most studied catalysts for CO2 methanation, the Ni- and Ru-based ones can be highlighted as the systems that almost produce exclusively CH4, while other active phases such as Pd, Pt, Rh, Mo, Re and Au, may result more easily in the co-production of CO (by means of R-WGS)11. Moreover, oxides such as Al2O3, TiO2, ZrO2, and CeO2 are the most common materials used as supports for depositing the cited metals7, 11. Particularly Ni-based catalysts are attractive, because of their good activity and low price. In fact, the principal catalysts manufacturers (e.g. Johnson Matthey, Topsøe, and Cariant-Süd Page 5 of 49 Chemie) offer methanation catalysts with Ni within their formulation. However, Ni catalysts present sintering and considerable carbon deposition at high temperatures 12, so Ru arises as an alternative since this noble metal exhibits a superior specific activity due to its strong ability to activate the CO2 molecules, thus resulting in a considerable decreasing of the reaction temperature, while the carbon deposition and the sintering are both inhibited9, 13. Nevertheless, Ru is able not only of activating CO2 molecules but also to dissociate H2. Thus, the coexistence of both functionalities in this active phase, that are required for the CO2 methanation, makes the Ru-based catalysts performance dependent on the size of the metal clusters, as pointed out by Kwak et al.14 in their study of Ru/Al2O3 systems. In this case, these authors have proposed a superior selectivity to CH4 rather than CO with larger Ru nanoparticles because, apart to their ability for the rapid formation of CO species from CO2, these appeared to be large enough to adsorb H2 and then supply the required amounts of atomic H for the production of CH4. In general, these results are in agreement with different reports that also observed a superior selectivity towards CH4 with larger metal nanoparticles in different series of catalyst with size controlled Ru clusters15-17. However, a controversy arises when it is intended to propose a reaction mechanism, not only for Ru-based catalysts but also for the other studied systems, even though there is a more or less general agreement about the effects of the metal particle size. Up to now, two main possible pathways can be identified11, 18. For one hand, a pathway where CO2 is converted to CO and subsequently transformed into CH419, 20, and on the other hand, a direct methanation of CO2 without the formation of a CO intermediate21, 22. In both cases the explanations of the pathways seem to agree with the effect of the size of Ru nanoparticles. Even some authors such as Panagiotopoulou and Verykios23 have studied the selective methanation of CO including CO2 in the mixture of reaction over Ru/TiO2 catalysts and Page 6 of 49 they proposed that the mechanism of reduction of both molecules are not influenced by the Ru particle size. Nevertheless, which is clear is that further studies are required in order to clarify mechanistic aspect of the cited reaction in Ru-based catalysts. Other aspect about the CO2 methanation that has to be pointed out is that, according to the Le Chatelier’s principle, this reaction is favored with pressure due to the volume contraction in the direction of the methane production, and this has been demonstrated in different thermodynamic analysis such as that presented by Gao et al.10. However, despite the theoretical benefits in the yield of CH4 if the Sabatier reaction is carried out at pressures above the atmospheric one, systematic experimental studies on the effect of operating at high pressures are currently scarce, especially for the Ru-based catalysts. To our knowledge, one of the few reports about this regard is that reported by Weatherbee et al. 24 where the rates for the CO2 conversion and CH4 formation increased with the pressure using Ru/SiO2 catalysts. However, more efforts have to be focused in order to establish the possible effects of pressure over the mechanism of reaction of Ru-based catalysts during the CO2 methanation, especially because the thermodynamic studies have demonstrated that pressure may alter the product distribution by means of the occurrence of side reactions such as the R-WGS10. The advantages of high-pressure during the reduction of CO2 for obtaining methanol instead of methane has been pointed out by Gaikwad et al.25. These authors have demonstrated that under high-pressure, a dense phase formation by products condensation occurs, limiting the overall reaction rate by internal mass transfer. However, with the suitable size of the catalysts, such problem can be effectively overcoming, resulting in the promotion of kinetics (high reaction rate) and thermodynamics (superior conversions) for enhancing the methanol yield. All about this approach could be very adaptable for the Page 7 of 49 production of CH4 from CO2, thus the development of catalysts able to maximize the benefits of high-pressure should be carried out. According to the scenario described above, the present work proposes the synthesis of Ru/Al2O3 catalysts by means of adapting the polyol method in order to establish effects of the particle size over some physicochemical properties and, more importantly, over the catalytic activity during the CO2 methanation. The most active system will be also studied under high-pressure conditions for evaluating possible modifications in the catalytic performance. Finally, this study will be complemented with in-situ DRIFTS experiments in order to establish possible clues about the reaction pathway of the CO2 methanation at high pressure. In this sense, the present manuscript combines the application of a methodology for controlling the size of the ruthenium species for obtaining catalysts for the Sabatier reaction, with measures of catalytic activity at high pressure. Therefore, this will be one of the few works up to now presenting experimental data of the pressurized CO2 methanation reaction. 2. Materials and methods 2.1. Synthesis of the catalysts A series of 5 wt.% Ru/Al2O3 catalysts was prepared by means of the polyol method26, using the Ru/PVP molar ratio reported by Yin et al.27 during the synthesis of Ru nanoplates with irregular forms by means of a hydrothermal process. For this purpose, the following commercial reagents were used: -Al2O3 [alumina - Puralox Scca 30/200] as support, Ru(NO)(NO3)3 [ruthenium (III) nitrosyl nitrate - Johnson Matthey® - 14.32 wt.% metal content] as precursor of the active phase, polyvynilpyrrolidone [PVP - Sigma- Aldrich® - Mw ~10000] as surfactant, ethylene glycol [Alfa Aesar® – 99+%] as solvent, and NaBH4 [sodium borohydride Sigma-Aldrich® - 99%] as reducing agent. Page 8 of 49 Firstly, the alumina (3.98 g) was dispersed in ethylene glycol (41 mL) within a round flask. Subsequently, the ruthenium (III) nitrosyl nitrate (1.98 mmol) and the suitable amount of PVP (according to the desired Ru/PVP ratio - Table 1) were dissolved in the ethylene glycol under continuous stirring. The liquid turned brown and the mixture was heated at 50 ºC. Once at this temperature, the NaBH4 (5.20 mmol) was added and the system was heated at 100 ºC for 1h and subsequently cooled down until room temperature. In all cases, a black colloidal suspension was obtained, suggesting the reduction of the Ru species. Afterwards, the ethylene glycol was separated using a cold trap vacuum system. The obtained solid was washed four times with a 50/50 acetone/methanol mixture, and then it was dried at 100 ºC overnight and finally calcined at 400 ºC for 4h. The catalysts obtained with variable Ru/PVP ratio (A-F) were labelled as RuA-F, as presented in Table 1. 2.2. Physico-chemical characterization The XRD patterns of the powder catalysts were recorded in a Philips X’pert Pro diffractometer, equipped with a high-temperature chamber Anton Paar HTK 1200, using CuKα radiation (40 mA, 45 kV), with 0.05º step size and 300 seconds of step time, over a 2θ range from 10º to 90º. Firstly, for every catalyst, an XRD pattern was measured at 30 ºC, using the temperature control of the chamber. Afterwards, the samples were reduced with a feed-stream of H2 10 vol.% in N2, heating from 30 º to 400 ºC (10 ºC/min). Once at 400 ºC, a XRD pattern of the reduced catalyst was measured. The temperature of the reduction process was selected from the H2-TPR experiments in order to ensure that all the reducible species are reduced. The software X’Pert HighScore Plus was used to identify the crystallographic phases present in each material and for the treatment of the Page 15 of 49 production of Ni0. In a similar way, it has been demonstrated that for gold catalysts deposited over Fe-doped ceria supports, the coexistence of gold, cerium and iron cationic species resulted in the enhancement of the reduction of the materials at lower temperatures35. Therefore, it has to be noticed that in our case, the reducibility is being determined by the parameters of the synthesis. Apart from this, it is clear that the reduction was complete at 400 ºC in all cases, and further remaining PVP were likely gasified, so this temperature is selected for carrying out the reduction process in the XRD equipment. Regarding the textural properties, all the prepared materials present type IV isotherms (Figure S1) associated to mesoporous materials, and type H1 hysteresis suggesting aggregates or agglomerates of spheroidal particles36. The obtained BET surface area, average pore size and pore volume values are included in Table S1. Firstly, the BET surface area values are quite similar among theme since these are between 143 and 152 m2/g and this variation is within the expected error of the measurement (10 m2/g). Therefore, no alterations of the surface area (that has to be principally due to support) are observed due to the variation of the Ru/PVP molar ratio during the synthesis. Regarding the pore volume, although slight modifications are evidenced, there is not a clear trend of the values. Firstly, there is a progressive increment from RuA to RuD but then, RuE and RuF present lower values. Therefore, probably the modifications of the pore volume are within the error of the measurement. In line with this, the pore size values also reveal a small modification with the Ru/PVP molar ratio, but in this case, the sizes follow an increasing trend with such molar ratio. There are some works that have reported the enlargement of the porous structure of materials such as silica37, ceria/alumina38 or ceria39, due to the introduction of deposited gold nanoparticles. Page 16 of 49 Hence, probably a similar behavior is being noticed for the Ru/Al2O3 catalysts, revealing a different interaction between Ru and the support, depending on the Ru/PVP ratio. The XRD patterns of the calcined catalysts, as well as those of the reduced systems at 400 ºC with a mixture H2 10 vol.% in N2, are presented in Figure 2. Firstly, for the calcined catalysts (Figure 2A), the more intense reflection lines agree with those of RuO2 (ICDD 00-040-1290) and -Al2O3 (ICDD 00-010-0425), discarding phase transitions from the starting alumina. Additionally, weak reflections ascribable to the presence of metallic Ru (ICDD 00.006-0663) are observable, even though the solid were calcined in air. Apart from those mentioned nor other crystalline phases, nor the formation of solid solution between Ru and Al2O3 is evident in the diffraction patterns. Regarding the reduced catalysts (Figure 2B), the XRD patterns preserve the reflections of the -Al2O3, and those of the metallic Ru become more intense, while the reflections of RuO2 disappear. Apart from this, no evidences of new crystalline phases or solid solution formation is noticed. A close analysis of the reflections due to RuO2 and Ru in the calcined and the reduced catalysts respectively, reveals an increment of their intensity as well as the decreasing of their full-width at half maximum (FWHM) with the enhancement of the Ru/PVP molar ratio. Therefore, using the more intense reflections of the RuO2 (28º 2 - (110) crystallographic plane – Figure 2A) and Ru (44º 2 - (101) crystallographic plane Figure 2B) phases, the Scherrer equation was applied for obtaining the crystallite sizes (Table S1). Comparing these values with the Ru/PVP molar ratios, it is observed that the RuO2 crystallite size increases while the PVP loading decreases during the synthesis (Figure 3A), thus resulting in the decreasing of protective molecules that inhibit the agglomeration of Ru species. Furthermore, comparing the sizes of the RuO2 crystallites Page 17 of 49 and those of the corresponding Ru crystals generated after the reduction treatment into the XRD chamber (Figure 3B), the main result is that higher the size of the RuO2, higher the size of the Ru crystallites. However, the increments of the RuO2 and Ru species with the Ru/PVP molar ratio (Figures 3A and 3B), present a linear trend that changes the slope, which is in agreement with the slope change observed for the reducibility percentage (Figure 1B). According to this, the reducibility and now the crystallite size of RuO2 species may be suggesting not only their enlargement depending on the Ru/PVP ratio but also a probable morphology alteration. Nevertheless, this has to be noticeable through preferential orientation of crystallographic planes. This is why a comparison of different area ratios of the more intense crystallographic planes of RuO2 ((110), (101) and (211)) was performed, as a function of the Ru/PVP ratio (Figure S2). The results suggest that a preferential orientation may be occurring with the involvement of the (211) and (110) (Figure S2A) crystallographic planes rather than with that of the (101) and (110) (Figure S2B) since in the first case, a progressive increment of the area ratio is observed while in the last case, a trend more similar to a constant is noticed. Regarding the coexistence of RuO2 and Ru in the calcined catalysts that was highlighted above (Figure 2A), a comparison of the area ratios between the most intense crystallographic plane of Ru (101) and those of RuO2 ((110) and (211)) was carried out as a function of the Ru/PVP ratio (Figure S3). In both cases, the area ratios increase with the Ru/PVP ratio, pointing out that as the size of the deposits of increases, greater the population of metallic Ru species. Comparing the theoretical cell volumes of RuO2 and Ru (62.90x106 pm3 and 27.15x106 pm3 respectively), the reduction of RuO2 implies a volume loss percentage (V. Loss (%)) Page 18 of 49 of 56.8%, calculated according to the Eq. 3. For the prepared materials, the crystallite volumes (Figure 3B inset) are obtained assuming spherical geometry of the RuO2 and Ru crystals, and using the crystallite sizes (Table S1) as the corresponding diameters. V.Loss(%) =Volume(RuO2) - Volume (Ru) Volume(RuO2)X 100 Eq.3 The RuA-RuD series of materials presents V. Loss (%) values close to the theoretical one as a result of the diminishing of the structural volume due to the releasing of oxygen atoms after the reduction. This behavior agrees with that of other systems where the volume loss produced by the reduction of the oxide to the metallic state have allowed the formulation of empirical approaches for the estimation of the particle size of metal clusters from that of the corresponding oxide. For instance, that proposed by Schanke et al.40 for the estimation of crystallite size of Co nanoparticles using the size and the molar volume molar volume of the corresponding Co3O4. Despite this, for the last systems (RuE and RuF), the V. Loss(%) are below 56.8% suggesting that as PVP is removed, not only the size of the deposits of RuO2 increases but also there is a greater tendency to sinter during the reduction of RuO2 to Ru. Furthermore, the deviation of a lineal trend of the evolution of the crystallite size of RuO2 and Ru (Figure 3B) may be also related with the coexistence of both phases in the calcined solids, since the metallic Ru that does not have to be the same in all the cases, altering the sintering of the metal clusters during the reduction. Other aspect that has to be noted is that while the Ru/PVP molar ratio influences the size of the deposited species of Ru, regardless oxide or metal species, the availability of Ru atoms is also being altered, thus it is expected that the Ru metal dispersion decreases with the increase of the particle size. In this sense, there are models that have established the Page 19 of 49 relationship between the metallic crystallite size and the fraction of exposed metal in supported metal catalysts such as that proposed by Borodziński et al.41, which we have applied using the crystallite size of the metal clusters of Ru previously obtained by means of the Scherrer equation. In our case we used the model for estimating dispersions below 20%41, and the results included in Table S1, as expected, demonstrate that the Ru dispersion decreases with the increment of the crystallite size. Up to now, the modification of the reflections of RuO2 or Ru phases has been associated to alterations in the crystallite sizes mainly. However, for such assumption, the content of Ru must be the same in all the studied systems, or at least comparable. An estimation of the real ruthenium content has been carried out by means of the EDX analysis of different SEM micrographs for every material (such as those included in Figure S4 for the RuA catalyst). The Ru wt.% values calculated from the SEM analysis are presented in Table S1 and in all cases the loading of noble metal is comparable to the intended one (5 wt.%), thus making the main differences between the obtained materials related to the modification of the metal crystallite size and morphology due to the different Ru/PVP ratio used during the synthesis. Regarding the morphology of the samples, no defined shapes are observed in the SEM micrographs. Only amorphous agglomerates of alumina, which is the majority phase, are appreciated. A closer exploration of the Ru-containing species in the surface of the catalysts was carried out by means of TEM. In Figure 4, two representative micrographs of the RuF catalysts are presented, including the corresponding compositional analysis represented through the mapping of the relevant components of the formulation of the materials, where the possible presence of remaining elements of the precursors such as Na and B species was also taken into account. Page 20 of 49 The mapping of the different elements in both TEM micrographs confirms the presence of clusters highly rich in Ru and O over a matrix combining Al and O. However, the inclusion of B and Na in the mapping analyses reveals that despite the washing processes during the synthesis of the materials, there is a considerable amount of remnant Na species (much more than B ones) highly dispersed on the surface of the catalysts and in close contact with clusters of RuO2. Other aspect that has to be remarked from the TEM micrographs is that although the average size of the RuO2 deposits are of the order of those calculated using XRD (Table S1) and follows a similar tendency as a function of the Ru/PVP (see Figure 3A), the distribution is heterogeneous and no patterns of shape are observed depending on the Ru/PVP ratio. Therefore, the applied synthesis method does not allow generating ruthenium species with appreciable regular sizes or shapes. The Raman spectra of the calcined catalysts are presented in Figure 5 and in all cases, the profiles are very similar. Firstly, there is a wide region of overlapped small signals (281, 308, 349, 381 and 420 cm-1), whose may correspond to vibrations of the Al-O bond. In agreement with this, some authors that have studied -Al2O3 through Raman spectroscopy, observed similar signals in such region of the spectra and assigned them to vibrations of Al-O bonds in tetrahedral coordination AlO433, 42. Despite this, in our case, a precise allocation of these signals common to all materials that allows the identification of the possible vibrational modes is hardly achievable because these signals are not only displaced but also the signal-to-noise ratio is too poor to perform reliable deconvolution. Therefore, the presence of AlO4 structure is inferable in the prepared catalysts, but some structural and morphological parameters may be altering this symmetry, since the Page 21 of 49 influence on shape, intensity and position of the Raman bands of many materials as been thoroughly demonstrated because of the phonon confinement behavior43, 44. The more intense bands at 522, 635 and 701 cm–1 are due to the vibrational modes of RuO245, that confirms the presence of the oxide of this noble metal, as previously stablished by means of the XRD analysis (Figure 2A). The three signals agree with active Raman vibrational modes described for the TiO2 rutile structure44, which is the same of the RuO2. In this sense, the Raman active modes A1g (~635 cm-1) and B2g (701 cm-1) are produced by motions of anions with respect to the central cation (O-Ru-O) perpendicular to the C axis, while the Eg mode (~522 cm-1) is produced along the C axis. It is important to remark that the position, as well as the FWHM of the Eg, A1g and B2g Raman bands appear modified in the different catalysts. Even the relative intensity between them looks different. Therefore, the Eg/A1g and A1g/B2g intensities ratios have been obtained and depicted against the crystallite size of the RuO2 (Figure 6), previously stablished through the Scherrer equation, in order to observe if any tendency on these intensity ratios is induced by the crystallite size, since the phonon confinement may be altered by the crystallite size. Firstly, the two trends presented in Figure 6 demonstrate that the increment of the crystallite size of the RuO2 clusters produces a growing trend of both intensities ratios (Eg/A1g and A1g/B2g). This effect of the crystallite size in the alteration of the intensity of the vibrational modes may be related with the fact that the growth generated with the modification of the Ru / PVP ratio is preferentially oriented towards some crystallographic planes. This may be resulting in the alteration of the staking of the planes generated by the O-Ru-O bonds that seems to be modifying their motions perpendicular or along the C axis described above. Page 22 of 49 In line whit this, other authors have also observed the alteration of the intensities of the Raman vibrational modes in RuO2. For instance Mar et al.45, through the characterization of RuO2 thin films, observed a close relationship between Raman shift of the signals and finite-size effects as well as stress induced by lattice mismatch and different thermal expansion coefficients. In addition, they observed that the deposition temperature, which is related with the final size of the RuO2 films, determines the linewidth of the Raman signals. Nevertheless, in our samples, apart from the effect of the size of the particles, the morphology alteration that was suggested in the TPR and XRD results should be also considered. Especially because a reasonable agreement arises between the observed change of trends in the reducibility percentages or the RuO2 crystallite sizes depicted against the Ru/PVP ratios (Figures 1B and 3B respectively), and the variation of the slope of the Raman intensity ratios suggested, especially in the case of the A1g/B2g ratio (Figure 6B).. In this sense, considering that the three compared vibrational modes occur in different crystallographic planes of the RuO2 structure, including the most intense one (110) that may be involved in preferential orientations as demonstrated through XRD analysis (Figure S2), it is inferable that the possible morphology alteration may also result in the distortion of the vibrations of the O-Ru-O bonds. 3.2. Catalytic activity measurements Measurements at atmospheric pressure: As for the catalytic activity of the studied materials, the CO2 conversion percentages at different temperatures are presented in Figure 7. Firstly, in all cases evaluated at 1 bar, the conversion increases with the temperature, although no systems achieve the equilibrium CO2 conversion. Despite this, Page 23 of 49 which is evident is that the CO2 conversion increases with the increment of the particle size of the Ru nanoparticles, especially at higher temperatures. Other relevant aspect during the methanation reaction is the probable occurrence of collateral reactions that also may result in the production of CO46, thus making the selectivity towards methane production an additional aspect that has to be accounted for designing methanation catalysts. In this sense, the CO and CH4 Vol.% in dry basis (discarding water since it was condensed prior the chromatographic analysis) were established at every temperature in the outlet of the reactor for all the catalytic tests and the results are presented in Figure 8. The production of CO increases with the temperature, but in no case the CO vol.% is superior to the CH4 vol.%. In addition, the CO formation diminishes with the increment of the particle size of the Ru nanoparticles. Regarding the CH4 vol.%, the general trend exhibits the enhancement of the CH4 vol.% with the temperature, but for the RuF system, with the largest Ru nanoparticles, a tendency to reach a maximum CH4 vol.% that later decreases is observed. This behavior agrees with that described by the CH4 vol.% obtained from the equilibrium compositions as well as with the theoretical behavior described by Gao et al.10, who associated the decreasing of the methane production and the increment of carbon monoxide one to the occurring of the R-WGS reaction that is endothermic and consequently favored with temperature. Moreover, the results presented in Figure 8 show that the CH4 vol.% presents an opposite trend respect to that of the CO vol.%. Therefore, it can be inferred that as the size of the Ru clusters increases, the catalysts become not only more active but also more selective towards the production of CH4. Page 24 of 49 The effect of the size of the active phase of methanation catalysts on their catalytic activity as well as on the selectivity has been widely studied not only for Ru-based catalyst but also for others such as Pt, Cu, or Ni-based systems9, 46. Particularly, regarding Ru/Al2O3 catalysts, different authors have observed tendencies in good agreement with our catalytic results and their dependence on particle size of the active phase. For instance, Lange et al.17 reported that the increasing of the loading of Ru, from 1 to 3% in Ru/ZrO2 catalysts, resulted in the enhancement of the CO2 methanation. Hu et al.47 also observed a similar behavior using Ru/TiO2 with the Ru loading up to ca. 3% before declining, while the selectivity to CO or CH4 remained unchanged. However, Kwak et al.14 established that at low Ru loadings (≤0.5%), where the active metal phase is highly dispersed on the alumina support, the systems tend to produce CO, while through the increment of the metal loading, the selectivity toward CH4 increases, diminishing the CO production. In this last work, the authors suggested that the mechanism of the reaction is different on small particles with high dispersion and on large Ru nanoparticles because, while the small Ru atoms favor the CO formation, the bigger Ru clusters are able to supply atomic hydrogen, promoting the production of CH4. Moreover, Kowalczyk et al.16 established the relationship between the size of the Ru nanoparticles of different Ru-supported catalysts and their catalytic performance during the CO2 methanation. Firstly, they observed for all the studied materials a general decreasing of the catalytic activity with the particle size but when the same comparison is made for different catalysts prepared with the same support, a slight increment of the conversion is observed as the size of the Ru particles increases. Therefore, based on these results they suggested that other aspects such as the morphology of the nanoparticles promoted by the supports could be also related with the catalytic performance, and not only their size. Page 31 of 49 The number of studies published analysing the effect of pressure on CO2 methanation reaction over Ru-based catalysts is limited. As we mentioned above, the reaction proceeds with gas volume contraction and thus the equilibrium conversion is increased with the pressure. From the mechanistic point of view, the methanation rate depends on the amount of adsorbed H* and CHO* species on the metal particles, whose concentration depends primarily on adsorption equilibrium with gas-phase hydrogen and CO2. CHO* species are stabilized on the metal sites through the formation of a double bond as a result of back-donation of electrons from metal and donation of electrons from carbine/carbine carbon, respectively 62. As a result the of stabilization, CHO* species became less reactive and more stable on the metal surface. However, an increase of the pressure on CO2 methanation leads to an increase of the surface coverage of H* species that facilitates the hydrogenation of CHxO species to form methane. 4. Conclusions The synthesis of a series of 5 wt.% Ru/Al2O3 catalysts by means of the polyol method, modifying the Ru/PVP molar ratio resulted in the successful obtaining of materials with controlled particle size of the deposits of phases containing Ru. The physico-chemical characterization revealed that, as expected, the particle size of such deposits increased with the Ru/PVP molar ratio, although with a heterogeneous distribution according to the TEM micrographs. However, the growth of the Ru-containing clusters exhibits probable alterations in the crystallinity that suggest preferential orientations, at least through the comparison of the (110) and (211) crystallographic planes. These crystallinity alterations agree with changes in the reducibility of the materials, as well as with the vibrational modes in the Raman spectra due to the vibrations of the O-Ru-O bonds located along the cited crystallographic planes. Despite this, further studies would be required in order to explore probable shape control in the obtained materials with the applied synthesis Page 32 of 49 procedure, and correlate this with the catalytic performance of the materials during the Sabatier reaction. Regarding the catalytic activity measurements, the CO2 conversion increased with the size of the Ru nanoparticles in the reduced catalysts. Additionally the CH4 yield was also enhanced with larger Ru clusters, since the CO production was deplete, which demonstrates that Ru particle sizes determines the contribution of the side reaction R- WGS. Finally, as for the advantages of CO2 methanation carried out at high-pressure, these were noticed in the enhancement of the thermodynamics of the process through the increasing of the methane equilibrium concentration. In fact, the RuF catalyst achieved this experimentally. Additionally, the kinetics of the process was also favored since the maximum conversion occurs at low temperatures. Finally, from the mechanistic point of view, the pressurized reaction resulted in the enhancement of available hydrogen species required for the weakening and cleavage of the C-O bond (the rate determinin-step) through the production of CHxO species that finally resulted in CH4 molecules. Consequently, it is inferable that larger Ru nanoparticles are the most suitable scenarios to take advantage of carrying out the Sabatier reaction at pressures above the atmospheric one, since multiple Ru active sites are required in the same particle for the adsorption and activation of CO2 and H2 molecules. 5. Associated content Supporting information: Tables with some physico-chemical data from different characterizations techniques. Furthermore, Figures with N2 – isotherms, crystallographic planes areas ratio comparisons, SEM micrographs and Arrhenius plots will be also included. All of these Tables and Figures are described in the text. Page 33 of 49 6. Acknowledgements The authors thank the financial support provided by the Spanish Ministry of Economy and Competitivity through the project ENE2015-66975-C03-2-R. 7. References 1. Paris Agreement - United Naitons 2015, 2015. 2. J. C. Navarro, M. A. Centeno, O. H. Laguna and J. A. Odriozola, Catalysts, 2018, 8, 578. 3. J. Delbeke and P. Vis, EU Climate Policy Explained, European Union, 2016. 4. Intergovermental Panel on Climate Change (IPCC). Climate Change 2007: Mitigation of Climate Change, IPCC 4th Assessement Report, 2007. 5. K. Onarheim, A. Arasto, I. hannula, J. Kärki, J. Lehtonen and P. Vainikka, VTT (Discussion Paper), 1-21. 6. L. Pastor-Pérez, F. Baibars, E. Le Sache, H. Arellano-García, S. Gu and T. R. Reina, J. CO2 Util., 2017, 21, 423-428. 7. G. Centi, E. A. Quadrelli and S. Perathoner, Energy Environ. Sci., 2013, 6, 1711-1731. 8. I. Graça, L. V. González, M. C. Bacariza, A. Fernandes, C. Henriques, J. M. Lopes and M. F. Ribeiro, Appl. Catal., B, 2014, 147, 101-110. 9. K. Jalama, Catal. Rev., 2017, 59, 95-164. 10. J. Gao, Y. Wang, Y. Ping, D. Hu, G. Xu, F. Gu and F. Su, RSC Adv., 2012, 2, 2358-2368. 11. K. Stangeland, D. Kalai, H. Li and Z. Yu, Energy Procedia, 2017, 105, 2022- 2027. 12. S. Ma, Y. Tan and Y. Han, J. Nat. Gas Chem., 2011, 20, 435-440. Page 34 of 49 13. C. Janke, M. S. Duyar, M. Hoskins and R. Farrauto, Appl. Catal., B, 2014, 152- 153, 184-191. 14. J. H. Kwak, L. Kovarik and J. Szanyi, ACS Catal., 2013, 3, 2449-2455. 15. M. Kuśmierz, Catal. Today, 2008, 137, 429-432. 16. Z. Kowalczyk, K. Stołecki, W. Raróg-Pilecka, E. Miśkiewicz, E. Wilczkowska and Z. Karpiński, Appl. Catal., A, 2008, 342, 35-39. 17. F. Lange, U. Armbruster and A. Martin, Energy Technol., 2014, 3, 55-62. 18. D. C. Upham, A. R. Derk, S. Sharma, H. Metiu and E. W. McFarland, Catal. Sci. Technol., 2015, 5, 1783-1791. 19. M. Marwood, R. Doepper and A. Renken, Appl. Catal., A, 1997, 151, 223-246. 20. A. L. Lapidus, N. A. Gaidai, N. V. Nekrasov, L. A. Tishkova, Y. A. Agafonov and T. N. Myshenkova, Pet. Chem., 2007, 47, 75-82. 21. S. Fujita, H. Terunuma, H. Kobayashi and N. Takezawa, React. Kinet. Catal. Lett., 1987, 33, 179-184. 22. C. Schild, A. Wokaun, R. A. Koeppel and A. Baiker, J. Phys. Chem., 1991, 95, 6341-6346. 23. P. Panagiotopoulou and X. E. Verykios, J. Phys. Chem. C, 2017, 121, 5058- 5068. 24. G. D. Weatherbee and C. H. Bartholomew, J. Catal., 1984, 87, 352-362. 25. R. Gaikwad, A. Bansode and A. Urakawa, J. Catal., 2016, 343, 127-132. 26. L. F. Bobadilla, C. García, J. J. Delgado, O. Sanz, F. Romero-Sarria, M. A. Centeno and J. A. Odriozola, J. Magn. Magn. Mater., 2012, 324, 4011-4018. 27. A.-X. Yin, W.-C. Liu, J. Ke, W. Zhu, J. Gu, Y.-W. Zhang and C.-H. Yan, J. Am. Chem. Soc., 2012, 134, 20479-20489. Page 35 of 49 28. P. Betancourt, A. Rives, R. Hubaut, C. E. Scott and J. Goldwasser, Appl. Catal., A, 1998, 170, 307-314. 29. V. Mazzieri, F. Coloma-Pascual, A. Arcoya, P. C. L’Argentière and N. S. Fígoli, Appl. Surf. Sci., 2003, 210, 222-230. 30. K. R. Kahsar, D. K. Schwartz and J. W. Medlin, J. Am. Chem. Soc., 2014, 136, 520-526. 31. R.-Y. Zhong, K.-Q. Sun, Y.-C. Hong and B.-Q. Xu, ACS Catal., 2014, 4, 3982- 3993. 32. F. Romero-Sarria, S. Garcia-Dali, S. Palma, E. M. Jimenez-Barrera, L. Oliviero, P. Bazin and J. A. Odriozola, Surf. Sci., 2016, 648, 84-91. 33. M. Baronskiy, A. Rastorguev, A. Zhuzhgov, A. Kostyukov, O. Krivoruchko and V. Snytnikov, Opt. Mater., 2016, 53, 87-93. 34. H. Schmalzried, Berichte der Bunsengesellschaft für physikalische Chemie, 1979, 83, 964-964. 35. O. H. Laguna, M. A. Centeno, M. Boutonnet and J. A. Odriozola, Catal. Today, 2016, 278, 140-149. 36. G. Leofanti, M. Padovan, G. Tozzola and B. Venturelli, Catal. Today, 1998, 41, 207-219. 37. Z. Konya, V. F. Puntes, I. Kiricsi, J. Zhu, J. W. Ager, M. K. Ko, H. Frei, P. Alivisatos and G. A. Somorjai, Chem. Mater., 2003, 15, 1242-1248. 38. M. I. Domínguez, M. Sánchez, M. A. Centeno, M. Montes and J. A. Odriozola, J. Mol. Catal. A: Chem., 2007, 277, 145-154. 39. O. H. Laguna, F. Romero Sarria, M. A. Centeno and J. A. Odriozola, J. Catal., 2010, 276, 360-370. 40. D. Schanke, S. Vada, E. A. Blekkan, A. M. Hilmen, A. Hoff and A. Holmen, J. Catal., 1995, 156, 85-95. 41. A. Borodzinski and M. Bonarowska, Langmuir, 1997, 13, 5613-5620. Page 36 of 49 42. P. V. Thomas, V. Ramakrishnan and V. K. Vaidyan, Thin Solid Films, 1989, 170, 35-40. 43. A. K. Arora, M. Rajalakshmi, T. R. Ravindran and V. Sivasubramanian, J. Raman Spectrosc., 2007, 38, 604-617. 44. Y. Zhang, C. X. Harris, P. Wallenmeyer, J. Murowchick and X. Chen, J. Phys. Chem. C, 2013, 117, 24015-24022. 45. S. Y. Mar, C. S. Chen, Y. S. Huang and K. K. Tiong, Appl. Surf. Sci., 1995, 90, 497-504. 46. J. Gao, Q. Liu, F. Gu, B. Liu, Z. Zhong and F. Su, RSC Adv., 2015, 5, 22759- 22776. 47. J. Hu, K. P. Brooks, J. D. Holladay, D. T. Howe and T. M. Simon, Catal. Today, 2007, 125, 103-110. 48. R. L. Narayan and T. S. King, Thermochim. Acta, 1998, 312, 105-114. 49. G. C. Bond, Catal. Today, 1999, 49, 41-48. 50. P. J. Lunde and F. L. Kester, J. Catal., 1973, 30, 423-429. 51. G. Du, S. Lim, Y. Yang, C. Wang, L. Pfefferle and G. L. Haller, J. Catal., 2007, 249, 370-379. 52. G. D. Weatherbee and C. H. Bartholomew, J. Catal., 1981, 68, 67-76. 53. N. M. Gupta, V. S. Kamble, R. M. Iyer, K. R. Thampi and M. Gratzel, J. Catal., 1992, 137, 473-486. 54. S. M. Kozlov, G. F. Cabeza and K. M. Neyman, Chem. Phys. Lett., 2011, 506, 92-97. 55. L. Árnadóttir, E. M. Stuve and H. Jónsson, Chem. Phys. Lett., 2012, 541, 32-38. 56. G. E. Mitchell, P. L. Radloff, C. M. Greenlief, M. A. Henderson and J. M. White, Surf. Sci., 1987, 183, 403-426. Page 37 of 49 57. S. Sharma, K. B. Sravan Kumar, Y. M. Chandnani, V. S. Phani Kumar, B. P. Gangwar, A. Singhal and P. A. Deshpande, J. Phys. Chem. C, 2016, 120, 14101- 14112. 58. B. Miao, S. S. K. Ma, X. Wang, H. Su and S. H. Chan, Catal. Sci. Technol., 2016, 6, 4048-4058. 59. Y. Traa and J. Weitkamp, Chem. Eng. Technol., 1999, 22, 291-293. 60. A. Karelovic and P. Ruiz, J. Catal., 2013, 301, 141-153. 61. I. M. Ciobica and R. A. van Santen, J. Phys. Chem. B, 2003, 107, 3808-3812. 62. Y. Mori, T. Mori, T. Hattori and Y. Murakami, J. Phys. Chem., 1990, 94, 4575- 4579. Page 38 of 49 TABLES Table 1. Nomenclature of the catalysts and the corresponding Ru/PVP molar ratio Catalyst Ru/PVP molar ratio RuA 17.4 RuB 47.8 RuC 65.2 RuD 151.2 RuE 254.8 RuF 358.4 Page 39 of 49 FIGURE CAPTIONS Figure 1. H2-TPR profiles of the prepared catalysts: A) TPR profiles of the catalysts; B) Reducibility percentage as a function of the Ru/PVP molar ratio Figure 2. XRD pattern of the prepared catalysts: A) catalysts calcined in air; B) catalysts reduced in H2/N2 at 400 ºC Figure 3. Comparison of the crystallite sizes of the materials: A) RuO2 crystallite size vs Ru/PVP molar ratio; B) Ru crystallite size vs RuO2 crystallite size; C) Volume loss percentage during the reduction of RuO2 species to Ru at 400 ºC Figure 4. TEM micrographs of the RuF catalyst including mapping of Ru, Al, Na, O and B: A) selected zone 1; B) selected zone 2 Figure 5. Raman spectra of the prepared catalysts Figure 6. Correlation between the intensities ratio of the signals associated to vibrational modes of RuO2 and the RuO2 crystallite sizes obtained through the Scherrer equation: A) Eg/A1g; B) A1g/B2g Figure 7. Catalytic activity of the prepared catalysts Figure 8. CO and CH4 Vol.% in the outlet of the reactor for the different catalytic activity measurements Figure 9. Apparent activation energy Ea as a function of Ru particle size Figure 10. Evolution of DRIFTS spectra obtained during CO2 methanation when the RuF catalyst was exposed to N2 (12.5 mL/min), CO2 (7.5 mL/min), and H2 (30 mL/min) at 250 °C. Catalyst was pre-treated under a H2/N2 (50/50) mixture at 400 °C for 1 h: a) experiments at 1 bar; b) experiments at 4 bar Page 40 of 49 FIGURES Figure 1. H2-TPR profiles of the prepared catalysts: A) TPR profiles of the catalysts; B) Reducibility percentage as a function of the Ru/PVP molar ratio 100 200 300 400 500 307 ºC 263 ºC 234 ºC 218 ºC 243 ºC 305 ºC A) RuF RuE RuD RuC RuB Ru/PVP molar ratio TCD signal (mV) Temperature (ºC) RuA B) (0.5 mV) 311 ºC 302 ºC 278 ºC 228 ºC 223 ºC 281 ºC 242 ºC 200 ºC 0150300 150 200 250 300 350 400 RuF RuE RuD RuC Reducibility Percent. (%) RuA RuB Page 47 of 49 Figure 8. CO and CH4 Vol.% in the outlet of the reactor for the different catalytic activity measurements 200 300 400 0.0 0.5 1.0 1.5 2.0 2.5 3.0 RuA RuB RuC RuD RuE RuF RuF_4bar Vol.% Temperature (ºC) A) Vol.% CO Eq. 1bar Eq. 4bar 200 300 400 0 5 10 15 20 25 30 35 40 Vol.% CH4 B) Eq. 1bar Eq. 4bar Page 48 of 49 Figure 9. Apparent activation energy Ea as a function of Ru particle size 16 18 20 22 24 26 28 40 50 60 70 80 90 100 110 120 RuF_4bar 71kJ/mol RuF 103kJ/mol RuE 104kJ/mol RuD 117kJ/mol RuC 69kJ/mol RuB 61kJ/mol -Ea (1bar) -Ea (4bar) Ea (kJ/mol) Ru particle size (nm) RuA 42kJ/mol Page 49 of 49 Figure 10. Evolution of DRIFTS spectra obtained during CO2 methanation when the RuF catalyst was exposed to N2 (12.5 mL/min), CO2 (7.5 mL/min), and H2 (30 mL/min) at 250 °C. Catalyst was pre-treated under a H2/N2 (50/50) mixture at 400 °C for 1 h: a) experiments at 1 bar; b) experiments at 4 bar