WO2025251102A1 - Catalyst for selectively converting carbon dioxide and hydrogen to carbon monoxide and water - Google Patents

Abstract

Catalyst for selectively converting CO₂ + H₂ to CO and H₂O, the catalyst comprising a porous support, the support carrying particles, wherein the porous support comprises an oxide of Ce or a mixed oxide of Ce and Zr and wherein the particles comprise at least one ferromagnetic compound, wherein said ferromagnetic compound comprises (i) at least one oxide of Fe, Co, and Ni, or (ii) at least one metal of Fe, Co, and Ni or (iii) at least one oxide of Fe, Co, and Ni and at least one metal of Fe, Co, and Ni.

Description

CATALYST The present invention relates to a catalyst for selectively converting CO₂ + H₂ to CO and H₂O, the catalyst comprising a porous support, the support carrying particles. Further, this invention relates to a method of manufacturing a catalyst for selectively converting CO2 + H2 to CO and H₂O. Finally, the invention relates to a method for selectively converting CO₂ + H₂ to CO and H₂O by reacting gaseous CO₂ with gaseous H₂ in the presence of a catalyst. BACKGROUND OF THE INVENTION Many important industrial chemical reactions are catalyzed by a solid and require temperatures above 200 °C. One such reaction is the reverse water gas shift (RWGS), a newly emerging route for reducing the carbon footprint of platform chemicals, because it can convert captured CO2 streams into syngas (namely CO + H₂) (Ref.1) CO2 + H2 → CO + H2O ΔH298K = + 41 kJ/mol RWGS is an endothermic reaction and requires a catalyst to activate the poorly reactive CO₂. This means that RWGS reactors typically have to be operated at temperatures well above 400 °C to achieve substantial conversions. One important competitive reaction to the RWGS is the exothermic methanation of CO₂ (ΔH_298K = - 165 kJ/mol) which needs to be avoided for selective formation of syngas. RWGS is catalyzed by noble metals, but also by abundant metals such as Cu, Fe, Co, Mo and Ni, and bimetallic combinations of them (Ref. 2). While noble metals catalyze the reaction at 200 to 450 °C, group VIII metals need about 100 °C higher temperatures to achieve comparable conversions. Recent studies of this reaction have advanced the understanding on the role of the metal particle size, morphology and composition, as well as effect of the electron transfer from the support (Refs. 2 – 5). A very recent report shows the advantage of performing RWGS at high temperatures using electrically heated walls (Ref.6). The use of self-heating catalysts via magnetic induction can bring catalytic systems to high energetic efficiencies (Ref.7). Some metal and metal oxide nanoparticles have ferromagnetic properties and can release heat under an alternating magnetic field. The ability to inductively heat and the maximum temperature achievable is related to the chemical composition and Curie temperature (upper limit of ferromagnetic properties) of the nanoparticles. Feng-man Sun et al: “Ni/Ce–Zr–O catalyst for high CO₂ conversion during reverse water gas shift reaction(RWGS)", INTERNATIONAL JOURNAL OF HYDROGEN ENERGY, 40 (46), 2015; 15985-15993, DOI: 10.1016/J.IJHYDENE.2015.10.004 discloses an Ni/Ce-Zr-O catalyst. Natalia M. Martin et al: “Catalytic hydrogenation of CO2 to methane over supported Pd, Rh and Ni catalysts”, CATALYSIS SCIENCE & TECHNOLOGY, 7 (5); 2017; 1086-1094, DOI: 10.1039/C6CY02536F discloses Ni/CeO₂ catalysts for CO₂ hydration. Lea R. Winter: “Tuning Ni-catalyzed CO₂ hydrogenation selectivity via Ni-ceria support interactions and Ni-Fe bimetallic formation”, APPLIED CATALYSIS B. ENVIRONMENTAL, 224 (16), 2017; 442-450, DOI: 10.1016/j.apcatb.2017.10.036 discloses Ni/CeO2 and Fe/CeO2 catalysts for CO2 hydration. M. V. Konishcheva et al: “Selective CO methanation in H2-rich stream over Ni–, Co– and Fe/CeO2: Effect of metal and precursor nature”, INTERNATIONAL JOURNAL OF HYDROGEN ENERGY, 40 (40), 2015; 14058-14063, DOI: 10.1016/J.IJHYDENE.2015.07.071 discloses Ni/CeO₂, Fe/CeO₂ and Co/CeO₂ catalysts. BRIEF DESCRIPTION OF THE INVENTION RWGS is endothermic and therefore requires catalysts that can work at temperatures above 500 °C to reach high CO2 conversions. In addition, the involved gas mixtures may trigger oxidations, reductions or phase changes in the solid material (e.g. from oxide to metallic state). At the targeted reaction temperatures, the surface of magnetic nanoparticles is subject to dynamic changes and can also chemically react with the environment, changing in composition and structure and, thus, affecting its magnetic properties. Therefore, implementing both catalytic and magnetic functions in the same material for self-heating catalysis of the RWGS reaction is a complex task. One main problem is the stability of nanoparticle surfaces under reaction conditions. Having regard to the state of the art, the technical problem underlying the invention consisted in developing a catalyst that is selective for the conversion reaction of CO2 + H2 to CO and H₂O, that is self-heating but still shows no undesired reactions such as methanation of CO₂. The technical problem outlined above is solved by a catalyst according to claim 1. The catalyst for selectively converting CO₂ + H₂ to CO and H₂O comprises a porous support, the support carrying particles, wherein i_{. the porous support comprises an oxide of Ce or a mixed oxide of Ce and Zr and ii. wherein the particles comprise at least one ferromagnetic compound, wherein said} ferromagnetic compound comprises (_{1) at least one oxide of Fe, Co, and Ni, or (2) at least one metal of Fe, Co, and Ni or (3) at least one oxide of Fe, Co, and Ni and at least one metal of Fe, Co, and Ni.} Particles comprising at least one ferromagnetic compound on the carrier have the advantage that they are self-heating if an alternating magnetic field is applied. Hence, the catalyst can be brought to the ideal reaction temperature for the reaction CO2 + H2 to CO and H2O without having to apply an external heat source. In a preferred embodiment the particles comprising the at least one ferromagnetic compound are nanoparticles. A nanoparticle is a particle of any shape with an equivalent diameter of approximately 1 to 100 nm (nanoparticle as defined in IUPAC Compendium of Chemical Terminology, 5th ed. International Union of Pure and Applied Chemistry; 2025. Online version 5.0.0, 2025. https://doi.org/10.1351/goldbook.09538). Preferably, the nanoparticles have a particle size of 10 to 80 nm, preferably 20 to 55 nm. The particle size can be determined by applying the Scherrer equation to X-ray diffraction reflections. By using bifunctional (ferromagnetic and catalytic) nanoparticles there is the advantage of a high surface area for the catalysis, and an excellent heat transfer and fast response to temperature changes. Nanoparticle size is a critical parameter in both catalysis and inductive heating, with distinctive effects for each of them (see Fig. 1). Catalysis occurs on the surface and therefore its activity per mass of catalyst generally decreases with particle size. In addition, the structure-sensitivity of many reactions makes size effects even more pronounced (Ref.8). The specific absorption rate (SAR), and hence the heating power of the ferromagnetic materials, is proportional to the area within the magnetization hysteresis curve for a given temperature. The heat loss in ferromagnetic nanoparticles reaches a maximum at a critical volume at the transition from single to multiple magnetic domains. The catalyst according to the invention targets a particle size distribution that falls in the range were the catalytic and magnetic function are adapted to each other (Fig.1). In addition, it should be noted that the working conditions of energy-intensive reactions entails high-pressure reactive gas feeds and temperatures that will trigger changes in the morphology, surface structure and composition of nanoparticles (Ref. 9). Therefore, self-heating catalysts must be designed in such a way that they generate stable nanostructures under the harsh conditions of the application. RWGS is endothermic and therefore requires catalysts that can work at temperatures above 500 °C to reach high CO₂ conversions. High temperatures have the additional advantage of avoiding CO2 methanation, which is exothermic and causes unnecessary consumption of costly H2. Given the temperature and catalysis requirements of the energy-intensive reactions, the inventors found that ferromagnetic nanoparticles of naturally abundant group VIIIb elements of the iron group (Fe, Co, Ni) forming bi- or multimetallic combinations with other transition metals, and as their mixed complex oxides fulfil these requirements. These ferromagnetic materials have adequate heating rates and elevated Curie temperatures, which is the upper temperature limit for induction heating. A catalyst able to achieve chemical conversions under inductive heating must have: i) adequate magnetic properties at the reaction T range that allow heat release under easily applicable magnetic field strengths (i.e. up to 100 mT), and ii) good and stable catalytic performance at the targeted reaction temperatures and pressure gas feeds required in the application (often in the range of 1 to 40 bar). Using supported ferromagnetic nanoparticles according to the invention has the benefit of a high surface area for the catalysis, stable particle size, an excellent heat transfer and fast response to temperature changes. Therefore, supporting these ferromagnetic materials as nanoparticles on a mesoporous oxide ensures their stabilization against thermal degradation (sintering). In addition, supports alter the redox properties of metal particles, therefore they can stabilize the particles in the targeted oxidation state against changes in the redox potential of the reacting gas phase. In particular, Fe3O4 and metallic Fe phases are found both catalytically and magnetically suitable for the RWGS application. In this embodiment, the ferromagnetic compound comprises an oxide of Fe, or metallic Fe or an oxide of Fe and metallic Fe. In one embodiment the particles comprise Fe3O4. It could be shown that in all cases where the ferromagnetic compound is Fe and/or oxides of Fe, the self-heating properties of the catalyst under an alternating magnetic field yield up to 700 °C which is a good temperature for the CO₂ + H2 to CO and H2O conversion. The doping of these Fe phases with different amounts of Co serves to adjust particle properties such as reducibility, average particle diameter, specific absorption rate, and Curie temperature. Hence, in the preferred embodiment, the ferromagnetic compound comprises ^ _{oxides of Fe and Co, or ^ metallic Fe and Co, or ^ oxides of Fe and Co and metallic Fe and Co.} The molar ratio between Co and Fe of the ferromagnetic metal and metal oxide phases is 1:∞ preferably between 1 : 2 and 1 : 4. It could be shown that in all cases where the ferromagnetic compound is Fe and Co (either as metals and / or oxides thereof), the self-heating properties of the catalyst under an alternating magnetic field yield up to around 900 °C which is even a better temperature for the CO₂ + H₂ to CO and H2O conversion than the pure Fe-embodiments. In an alternative embodiment, ferromagnetic compound comprises ^ _{oxides of Fe and Ni, or ^ metallic Fe and Ni, or ^ oxides of Fe and Ni and at metallic Fe and Ni.} Though nickel has the effect that the self-heating properties of the catalyst under an alternating magnetic field is lower than 700 °C there may be certain applications where lower temperatures for the reaction of CO2 with H2 are advantageous. The molar ratio between Ni and Fe of the ferromagnetic metal and metal oxide phases is 1:∞ preferably between 1 : 2 and 1 : 4. The bifunctional performance of magnetic nanoparticles and their structural reliability upon usage depend not only on particle size, but also on atomistic structural details related to their surface and substrate interface, the interface with supports and/or self-developed outer shells. The combination of both catalytic and heating functions on Fe, Co, or Ni bimetallic nanoparticles will overcome the energy challenges of the reaction, allowing the catalytic sites to efficiently reach the necessary high temperatures for converting CO₂ into syngas. The porous support comprises either a pure oxide of Ce such as CeO2 or a mixed oxide comprising another metal ion such as an oxide with Zr. The mixed oxide with Zr could be Ce_xZr_yO₂ with 1 > x > 0,5 and y = 1-x. In a preferred embodiment the porous support is CeO₂ and/or Ce0.5Zr0.5O2, most preferably the support is CeO2. The object of the present invention is also solved by a method of manufacturing a catalyst as defined above, comprising the steps of a1) providing a porous support with an oxide of Ce or a mixed oxide of Ce and Zr, and impregnating that support with solution comprising at least one of Fe, Co, and/or Ni-ions a2) removing the solvent to obtain dried, impregnated support, a3) calcinating the dried, impregnated support, thereby generating anti-ferromagnetic metal oxides on the support, b) reducing the calcined material from step a3) at temperatures ranging from 400 to 700 °C, preferably about 550 °C, to obtain ferromagnetic metal particles and stabilizing the particles against re-oxidation by contact with air, c1) at least partially oxidizing Fe, Co, and/or Ni in the alloy with CO₂, or c2) partially oxidizing Fe, Co, and/or Ni in the alloy, preferably with diluted O2 at room temperature (approximately 20 °C to 30 °C). Step a1) comprises the step of providing a porous support an oxide of Ce and impregnating that support with solution comprising Fe, Co, and/or Ni-ions. The metal ions were incorporated via wet impregnation on different mesoporous supports (SiO₂, γ-Al₂O₃, CeO₂ and mixed oxide Ce_0.5Zr_0.5O₂) with iron and cobalt nitrate solutions in amounts necessary to reach 20 wt.% metal loading in all materials. Of these supports those comprising Ce turned out to be useful. Porous supports of oxides of Ce typically have a specific surface area of 20 to 125 m²/g. The particle sizes of the support is typically in the range of 1 nm to 1 mm. Table 1 shows surface areas of some of the tested supports. Table 1: Surface area of mesoporous supports measured by N2 physisorption Support Surface area (m²/g) SiO₂ 254 γ-Al₂O₃ 247 CeO2 45 Ce_0.5Zr_0.5O₂ 102 Depending on the desired catalyst, the solution comprises the necessary amounts of Fe, Co, and/or Ni-ions. If the catalyst – as preferred – is a mixed Fe/Co catalyst, then Co and Fe ions need to be present. In the preferred embodiment, the Co : Fe molar ratios are between 1 : ∞, preferably between 1 : 2 and 1 : 4. In step a2) the solvent to obtain dried, impregnated support is removed. This can be done e.g. by heating and/or reducing the pressure. Step a3) requires the calcinating of the dried, impregnated support, in the presence of O₂ thereby generating anti-ferromagnetic metal oxides on the support. This can be done e.g. at temperatures of 400 °C to 600 °C for 1 to 10 hours. At this point, metal oxide particles in the range of 10 to 30 nm are formed on the surface of the support (Table 2). Table 2: Particle size and crystalline composition of the different Fe species determined by in- situ XRD. All materials contain 20 wt. % total metal unless specified otherwise. One decisive step is step b) which requires the reducing the calcined material from step a3) at temperatures ranging from 400 to 700 °C, preferably about 550 °C. In this step ferromagnetic metal particles in the form of nanoparticles are created. The nanoparticles stabilized on the surface of the support, wherewith these nanoparticles are comprised of an alloy of Fe and at least one of Fe, Co, and/or Ni. These nanoparticles are partially resting on the support and partially integrated in the support. In an example procedure, 130 mg of the catalyst were placed inside a quartz tube of 9 mm inner diameter, fixed with quartz wool, and was heated to 550 °C (with a conventional lab-scale oven) under 100 ml/min of 10% hydrogen in Argon for 1h. This step serves to reduce the anti- ferromagnetic metal oxides (mostly α-Fe₂O₃) to the targeted metallic phases (see Fig. 9 and Table 2), namely α-Fe with a Curie temperature of 780°C and alloys of α-Fe with Co which are expected to have higher Curie temperatures. These particles are then stabilized against re-oxidation by contact with air. The support with the alloy is stabilized against oxidation with air by partial oxidation (passivation). Since Fe phases are particularly susceptible to re-oxidation under ambient conditions, a stabilization treatment, prior to removal of the material from the reactor used in b, is necessary to hinder the oxidation of magnetic nanoparticles by air. This way, the material can be used for induction heating directly after inserting it in the inductively-heated reactor (no external heating is necessary for re-activation). Two successful pretreatments have been identified: In step c1) the metal particles at least partially oxidized with CO₂, thereby maintaining the ferromagnetic properties. After purging with inert gas such as Ar to remove any hydrogen left from activation, the gas flow is switched to pure CO₂ at 550 °C. This leads to the partial oxidation of the metallic nanoparticles and in the case of Fe and Co to form Co-doped magnetite (Fe3O4) particles with a size of 10-45 nm (Table 2). Magnetite is also a ferromagnetic phase with a curie temperature of 585°C and it can be reversibly reduced to metal Fe upon contact with H₂-containing streams (such as RWGS reacting mixture). Alternatively, in step c2) the metal particles are partially oxidized with diluted O₂ at room temperatures. After H2-treatment, the material is cooled down to room temperature under Ar flow, and then the atmosphere is switched to 1% O₂ diluted in inert, similar to the procedure described by Vinum et al. (Ref.11). This generates a thin oxide layer around the metal particles that protects them from further oxidation upon contact with air. Notably, the formation of this oxide layer can be monitored by a slight increase in temperature of the solid (as a result of the exothermic oxidation reaction), in the order of magnitude of a few degrees Celsius. The catalyst was removed from the reactor and exposed to air only once it has cooled down to room temperature. After steps c1) or c2) usually metallic Fe, Ni, or Co and oxides of Fe, Ni, or Co, are present. However, during the catalytic reaction CO2 + H2 to CO and H2O conversion the catalyst may change as to comprise: ^ _{at least one oxide of Fe, Co, and Ni, or ^ at least one metal of Fe, Co, and Ni or ^ at least one oxide of Fe, Co, and Ni and at least one metal of Fe, Co, and Ni.} The invention also relates to a method for selectively converting CO₂ + H₂ to CO and H₂O by reacting gaseous CO₂ with gaseous H₂ in the presence of the catalyst in a vessel, wherein an alternating magnetic field is applied to the vessel so as to heat the catalyst, wherein the alternating magnetic field is applied until CO2 and H2 have been converted to CO and H2O. DETAILED DESCRIPTION OF THE INVENTION Further details and advantages of the invention are described with the attached figures and examples. Fig.1 schematically shows the influence of particle size in magnetic heating and catalysis. Fig.2 shows a comparison of the mean CO formation rate normalized per mass of catalyst (left) and per mass of metal (right) against the average temperature on stream for all the tested catalysts, prepared by steps a1, a2, a3, b and c.1. Hollow inverse triangle stands for 20%Fe₄Co₁/CeO₂ treated by method c.2 (1% O₂ at room temperature). Reaction conditions: contact time 5.2 g_cat ml_CO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < dp < 420 µm, RWGS mixture: 25% CO₂, 25% H₂ balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.3 shows the catalytic activity and temperature on stream for 20%Fe4Co1/γ-Al2O3 upper left), 20%Fe4Co1/CeO2 (upper right), 20%Fe4Co1/Ce0.5Zr0.5O2 (bottom left) and 20%Fe₄Co₁/SiO₂ (bottom right). Reaction conditions: contact time 5.2 g_cat mlCO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < dp < 420 µm, RWGS mixture: 25% CO2, 25% H2 balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.4 shows the catalytic activity and temperature on stream for 20% (left) and 50% (right) metal content on γ-Al2O3. Fe:Co ratio = 4:1, Reaction conditions: contact time 5.2 g_cat ml h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < d_p < 420 µm, RWGS mixture: 25% CO₂, 25% H₂ balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.5 shows the catalytic activity and temperature on stream for Fe-containing materials supported on CeO₂ (left) and Ce_0.5Zr_0.5O₂ (right). Metal content 20% wt. Reaction conditions: contact time 5.2 gcat mlCO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < dp < 420 µm, RWGS mixture: 25% CO2, 25% H₂ balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.6 shows the catalytic activity and temperature on stream of 20%Fe4Co1/CeO2 after treatment with CO2 at 550°C (left) and after dilute O2 treatment at room temperature (right). Reaction conditions: contact time 5.2 g_cat ml_CO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < d_p < 420 µm, RWGS mixture: 25% CO₂, 25% H₂ balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.7 shows the catalytic activity and temperature on stream for an unsupported spine) CoFe₂O₄. Reaction conditions: contact time 5.2 g_cat ml_CO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < dp < 420 µm, RWGS mixture: 25% CO2, 25% H2 balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.8 Shows H2-TPR of Fe- (left) and FeCo- (right) containing materials. Fe:Co ratio of 4:1. Conditions: 30 ml/min total flow, 5% H2 in Ar, temperature ramping rate: 5°C/min. Fig.9 Shows the X-ray diffractograms of the materials of Table 2 obtained after calcination, after in-situ treatment in H2 at 550°C and after in-situ treatment in CO2 at 550°C. Fig.10 Shows the catalytic activity and temperature on stream for 20%Fe₄Co₁/CeO₂ (upper left), 20%Fe3Co1/CeO2 (upper right), 20%Fe2Co1/CeO2 (bottom left) and 20%Fe1Co1/CeO2 (bottom right). Reaction conditions: activation as in c2, contact time 5.2 g_cat ml_CO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < dp < 420 µm, RWGS mixture: 25% CO2, 25% H2 balanced with Ar. Applied magnetic field strength 47 mT at 323 kHz. Fig.11 Shows the catalytic activity and temperature on stream for Fe4Co1/CeO2 with in- situ activation (under induction) according to protocol b. The temperature is monitored with an ultra-thin thermocouple and connected to a control loop. Reaction conditions: contact time 5.2 g_cat ml_CO2 h^-1, total pressure 1 bar, total flow 100 ml/min, material grain size 215 µm < d_p < 420 µm, RWGS mixture: 25% CO2, 25% H2 balanced with Ar. Applied magnetic field strength during reduction 44 mT at 323 kHz, field strength during reaction 46 mT at 323 kHz. Catalytic performance of materials under induction heating: To test the materials under reaction conditions in magnetic induction environment, the pretreated catalyst was pressed, ground and sieved to a material grain size 215 µm < d_p < 420 µm and then ca. 100 mg were placed in a quartz 9 mm internal diameter quartz reactor surrounded by a copper coil of 40 mm length 50 mm diameter with 6 windings. The catalyst was subsequently heated via induction under 100 ml/min Ar flow until the temperature reached a steady state value. The alternating magnetic field was generated by passing an alternating current through the coil of variable intensity and voltage. This setup can generate a 47 mT field at 323 kHz frequency at 2 kW. The temperature in the induction reactor was monitored with an infra-red sensor equipped with a camera, with a measurement spot of 0.7 mm diameter focused on the centre of the catalyst bed. In all cases the heating to a stable temperature of the material in Ar was completed within less than 100 seconds from switching on the magnetic field. The gas flow was then changed to the reaction flow consisting of 100 ml/min total flow containing 25% CO2, 25% H2 and balanced by Ar. The materials reach a new steady temperature under reaction conditions that is related to the catalytic, magnetic and redox properties of the material. The products were analysed with a gas chromatographer equipped with a TCD detector. The activity in RWGS of every catalyst was evaluated in terms of the molar formation rate of CO normalised per mass of catalyst or per mass of metal contained in the catalyst. As a reference, the catalytic performance of the materials was also tested in the same reactor but via heating at comparable temperatures in a conventional electric oven (1.8 kW). It was found that for same temperature, the same reaction rates were achieved, regardless of the type of heating. Fig. 2 shows the CO rates and temperatures achieved by Fe(Co) nanoparticles on different mesoporous supports and compared to an unsupported cobalt ferrite (CoFe₂O₄). In this case, the materials tested were subjected to the stabilization pretreatment c.1 unless else is stated. The inventors have found that Ce-based supports show a superior performance, and this is attributed to the ability of such supports to stabilize Fe(Co) particles in adequate size range for both catalysis and inductive heating. Conversely, Al₂O₃ has a negative effect due to the stabilization of Fe(Co) species as small particles with difficult reducibility. This leads to a lower fraction of particles with ferromagnetism, and therefore a smaller fraction of metal that effectively heats the catalyst. While Al₂O₃ stabilizes Fe(Co) particles with a diameter of 20 nm or smaller, CeO2 stabilizes Fe(Co) particles of 44 nm (Table 2). Fig. 2 shows an overview of all catalysts at isofield conditions (maximum field achievable at 2 kW of the power source). Increasing the overall metal loading in the material up to 50% on Al2O3 achieved to increase the temperature somewhat (Fig.4) but the reaction rates remain low, especially considering the higher metal loading per gram of catalyst for this sample. In addition, the 50%(Fe₄Co₁)/ Al₂O₃ catalyst shows poor stability under reaction conditions, probably due to loss of active surface over time. Fig.5 shows how for two materials with similar induction heating capability and metal phase composition, the catalytic activity is different depending on the nature of the support. Fe particles are more active in RWGS when supported on Ce0.5Zr0.5O2 than when supported on pure CeO₂. Effect of stabilization method on catalytic-magnetic performance: The materials shown in Fig. 2 were subjected to a CO₂ treatment after reduction as described in c.1. This treatment route convert metallic Fe(Co) nanoparticles into Co-doped Fe₃O₄ phases. Fig.6 shows that, when the materials are pre-treated according to procedure c.2, higher temperatures are generally achieved both in Ar and under RWGS conditions for a given magnetic field. Under RWGS reaction conditions the steady temperature achieved with c.2. procedure is ca.80-100 °C higher than if the material is treated with procedure c.1. After losing 10-15 % of the original activity in the first 1 h on stream, the catalyst reaches at steady state performance and remains stable for at least 2 h. Comparison of supported Fe(Co) nanoparticles with unsupported Fe(Co) materials: The inventors have examined the self-heating and catalytic performance of a ferromagnetic cobalt ferrite (CoFe₂O₄) as an unsupported alternative to the catalyst invention. The CoFe₂O₄ was synthesized following a well-known sol-gel procedure (Ref 14). This is a ferromagnetic material that can self-heat up to ca.520°C (Table 3). The material as-prepared, is placed directly in the inductively heated reactor without undergoing steps b) or c). Table 3: Curie temperature of the different Fe and Co phases (Ref.12): As evident in Fig. 7, the ferrite reached a temperature equal to its Tc of 520°C, then upon switching to the RWGS gas flow, which contains H₂, partial to total reduction of the ferrite takes place (see also Fig.8). This is detected in the gas outlet as excess consumption of H₂ in the feed mixture. As a consequence of generation of large fraction of Fe(Co) particles with particle size above the ferromagnetic threshold, the temperature increase under induction conditions. The high temperature and the reducing environment in the reactor combined with the lack of stabilization via mesoporous, non-reducible supports (like Al2O3 or CeO2) cause severe sintering of the metal particles generated from the ferrite reduction, leading to significant activity loss within a few minutes under 785°C. In any case, considering that the metal loading in this material is 100% in comparison with the supported nanoparticles (20%), the activity of the ferrite-derived Fe(Co) particles in RWGS is significantly lower (Fig. 2). Despite the superior self-heating capability of the ferrite, its suitability as a catalyst under induction heating conditions seems to be limited owing to the rapid deactivation and loss of surface area. Characterization of the supported Fe(Co) nanoparticles: A PANalytical X’Pert Pro diffractometer with CuKa radiation equipped with an X’Celerator linear detector in Bragg-Brentano geometry was used to carry out X-ray diffraction (XRD) measurements on powderised samples. In-situ XRD experiments were performed in an XRK900 chamber in a controlled gas atmosphere set by calibrated mass flow controllers (Bronkhorst). The samples were exposed to a H2 atmosphere at 550^oC for 20 minutes, followed by a change to pure CO2 at 550^oC for 20 minutes. The 2θ range of 20°-80° was scanned. The crystallite size was estimated by applying the Scherrer equation (Eq.1) for the highest intensity peak of each phase. In case of peak interference, the second highest intense peak was chosen. ^ ^^ In eq. (1), the term ^ stands for a dimensionless shape factor, ^ denotes the X-ray wavelength. ^ is the line broadening at half the maximum intensity and ^ the Bragg angle. Temperature Programmed Reduction in H₂ Temperature programmed reduction of the materials in diluted H2 is shown here. The support affects the reducibility of Fe(Co) phases and therefore the preparation method, generating metal particles under reducing conditions at 550°C (dashed line) is decisive in the proportion of metal and metal oxide present in the material. The nature of the support has an effect on the average particle size of segregated Fe₂O₃ oxides after calcination (Table 1). In addition, the enhanced reducibility of Fe(Co) species when supported on CeO₂ and Ce_0.5Zr_0.5O₂ leads to the formation of larger concentration of metal particles after reduction and, likely, this is associated to a larger particle size on average. Conversely, the lower reducibility of Fe(Co) phases on Al₂O₃ leads to less metallic phase content after H₂ treatment at 550°C and this is presumably related also with stabilizing a small average particle size in the calcined materials. Overall, this means that on CeO2-based supports, a larger fraction of the metal content is able to evolve into Fe(Co)°/Fe3O4 particles with a diameter above the threshold to generate heat via magnetic inductions. Consequently, such heaters lead to higher temperatures across the solid catalyst under magnetic induction. Conversely, strong metal support interactions are observed in γ-Al2O3 materials which stabilizes nanoparticles in smaller sizes that leads to inferior self-heating performance. Fig.10 shows samples with higher Co loading. The activation according to step c2) is described above. The same procedure as explained in section “Catalytic performance of materials under induction heating“ was applied. Stability test of FeCo materials: In a further example, it is shown that the fresh material (prior activation) can be heated under induction directly, thanks to its content in CoFe2O4 (ferromagnetic mixed oxide). In Fig. 11, the activation of Fe4Co1/CeO2 material under induction is shown now with control loop via an ultrathin thermocouple. Activation at 550 °C in H₂ is immediately followed by RWGS reaction. T set point is fixed at 500 °C (requiring X mT applied field at 323 kHz for the Fe₄Co₁/CeO₂ material) and the reaction was tested for 4 h. After initial slight deactivation, CO rate is constant and about 91 % of the rate obtained in conventional heating in an electric oven at 500 C. This additional advantage of the FeCo materials allows for reaching higher T if needed. References: (1) Kondratenko, E. V.; Mul, G.; Baltrusaitis, J.; Larrazábal, G. O.; Pérez-Ramírez, J. Energy Environ. Sci.2013, 6 (11), 3112. https://doi.org/10.1039/c3ee41272e. (2) Wang, Y.; Winter, L. R.; Chen, J. G.; Yan, B. Green Chem. 2021, 23 (1), 249-267. https://doi.org/10.1039/DOGC03506H. (3) Lin, L.; Yao, S.; Liu, Z.; Zhang, F.; Li, N.; Vovchok, D.; Martínez-Arias, A.; Castaňeda, R.; Lin, J.; Senanayake, S. D.; Su, D.; Ma, D.; Rodriguez, J. A. J. Phys. Chem. C 2018, 122 (24), 12934-12943. https://doi.org/10.1021/acs.jpcc.8b03596. (4) Beierlein, D.; Häussermann, D.; Pfeifer, M.; Schwarz, T.; Stöwe, K.; Traa, Y.; Klemm, E. Applied Catalysis B: Environmental 2019, 247, 200-219. https://doi.org/10.1016/j.apcatb.2018.12.064. (5) Peng, L.; Jurca, B.; Primo, A.; Gordillo, A.; Parvulescu, V. I.; García, H. ACS Sustainable Chem. Eng. 2021, 9 (28), 9264-9272. https://doi.org/10.1021/acssuschemeng.1c01401. (6) Thor Wismann, S.; Larsen, K.; Molgaard Mortensen, P. Angew Chem Int Ed 2022, 61 (8), e202109696. https://doi.org/10.1002/anie.202109696. (7) Ambrosetti, M. Chemical Engineering and Processing - Process Intensification 2022, 182, 109187. https://doi.org/10.1016/j.cep.2022.109187. (8) Van Santen, R. A. Acc. Chem. Res. 2009, 42 (1), 57-66. https://doi.org/10.1021/ar800022m. (9) Behrens, M. Angew Chem Int Ed 2016, 55 (48), 14906-14908. https://doi.org/10.1002/anie.201607600. (10) Almind, M. R.; Vinum, M. G.; Wismann, S. T.; Hansen, M. F.; Vendelbo, S. B.; Engbwk, J. S.; Mortensen, P. M.; Chorkendorff, I.; Frandsen, C. ACS Appl. Nano Mater.2021, 4 (11), 11537¬11544. https://doi.org/10.1021/acsanm.1c01941. (11) Vinum, M. G.; Almind, M. R.; Engbwk, J. S.; Vendelbo, S. B.; Hansen, M. F.; Frandsen, C.; Bendix, J.; Mortensen, P. M. Angew Chem Int Ed 2018, 57 (33), 10569-10573. https://doi.org/10.1002/anie.201804832. (12) Coey, J. M. D.1st ed.; Cambridge University Press, 2001. https://doi.org/10.1017/CBO9780511845000. (13) Lu, Z.; Gao, P.; Ma, R.; Xu, J.; Wang, Z.; Rebrov, E. V. Journal of Alloys and Compounds 2016, 665, 428-434. https://doi.org/10.1016/j.jallcom.2015.12.262.

Claims

CLAIMS 1. Catalyst for selectively converting CO₂ + H₂ to CO and H₂O, the catalyst comprising a porous support, the support carrying particles, characterized in that the porous support comprises an oxide of Ce or a mixed oxide of Ce and Zr and wherein the particles comprise at least one ferromagnetic compound, wherein said ferromagnetic compound comprises ^ _{at least one oxide of Fe, Co, and Ni, or ^ at least one metal of Fe, Co, and Ni or ^ at least one oxide of Fe, Co, and Ni and at least one metal of Fe, Co, and Ni.}

2. Catalyst according to claim 1, wherein said ferromagnetic compound comprises ^ _{an oxide of Fe, or ^ metallic Fe, or ^ an oxide of Fe and metallic Fe.}

3. Catalyst according to claim 1 or claim 2, wherein the particles are nanoparticles.

4. Catalyst according to any one of claims 1 to 3, wherein the porous support comprises CeO₂ and/or Ce_xZr_yO₂ with 1 > x > 0,5 and y = 1-x.

5. Catalyst according to any one of claims 1 to 4, wherein the particles comprise Fe3O4, preferably are comprised of Fe₃O₄.

6. Catalyst according to any one of claims 1 to 5, wherein the ferromagnetic compound comprises ^ _{oxides of Fe and Co, or ^ metallic Fe and Co, or ^ oxides of Fe and Co and metallic Fe and Co.}

7. Catalyst according to claim 6, wherein the molar ratio between Co and Fe of the ferromagnetic compound is between 1 : 2 and 1 : ∞.

8. Catalyst according to one of claims 1 to 5, wherein the ferromagnetic compound comprises ^ _{oxides of Fe and Ni, or ^ metallic Fe and Ni, or ^ oxides of Fe and Ni and at metallic Fe and Ni,} preferably wherein the molar ratio between Ni and Fe of the ferromagnetic compound is between 1 : 2 and 1 : ∞.

9. Method of manufacturing a catalyst according to one of claims 1 to 8, comprising the steps of a1) providing a porous support of an oxide of Ce or a mixed oxide of Ce and Zr, and impregnating said support with a solution comprising at least one of Fe, Co, and/or Ni-ions, a2) removing the solvent to obtain dried, impregnated support, a3) calcinating the dried, impregnated support, thereby generating anti-ferromagnetic metal oxides on the support, b) reducing the calcined material from step a3) at temperatures ranging from 400 °C to 700 °C, preferably about 550 °C, to obtain ferromagnetic metal particles and stabilizing the particles against re-oxidation by contact with air, c1) at least partially oxidizing Fe, Co, and/or Ni in the alloy with CO₂, or c2) partially oxidizing Fe, Co, and/or Ni in the alloy, preferably with diluted O2.

10. Method for selectively converting CO₂ + H₂ to CO and H₂O by reacting gaseous CO₂ with gaseous H2 in the presence of a catalyst according to one of claims 1 to 8 in a vessel, wherein an alternating magnetic field is applied to the vessel so as to heat the catalyst, wherein the alternating magnetic field is applied until CO₂ and H₂ have been converted to CO and H₂O.