Bimetallic Ni-Based Catalysts for CO2 Methanation: A Review

CO2 methanation has recently emerged as a process that targets the reduction in anthropogenic CO2 emissions, via the conversion of CO2 captured from point and mobile sources, as well as H2 produced from renewables into CH4. Ni, among the early transition metals, as well as Ru and Rh, among the noble metals, have been known to be among the most active methanation catalysts, with Ni being favoured due to its low cost and high natural abundance. However, insufficient low-temperature activity, low dispersion and reducibility, as well as nanoparticle sintering are some of the main drawbacks when using Ni-based catalysts. Such problems can be partly overcome via the introduction of a second transition metal (e.g., Fe, Co) or a noble metal (e.g., Ru, Rh, Pt, Pd and Re) in Ni-based catalysts. Through Ni-M alloy formation, or the intricate synergy between two adjacent metallic phases, new high-performing and low-cost methanation catalysts can be obtained. This review summarizes and critically discusses recent progress made in the field of bimetallic Ni-M (M = Fe, Co, Cu, Ru, Rh, Pt, Pd, Re)-based catalyst development for the CO2 methanation reaction.


Introduction
During the last hundred years, rapid industrialization and the high energy demands of our society have disrupted the carbon cycle through ever increasing greenhouse gas emissions, and the ramp-up of renewable energy production has yet to offset the negative effects on our planet's climate and ecosystems [1,2]. However, progress made in hydrogen production technologies through water electrolysis has raised hopes for the utilization of this green fuel that produces no CO 2 emissions upon its combustion [3], despite the fact that its storage and transportation remain challenging when compared to other traditional energy carriers, such as natural gas [4]. In the last decade, research efforts have been focused on the development of catalysts that can utilize this excess renewable hydrogen in order to hydrogenate CO 2 released from industrial flue gases. This way, H 2 can be transformed into a reliable energy carrier, that is, CH 4 or synthetic natural gas (SNG), with a significantly higher energy density, all the while creating a closed carbon cycle [5]. The complete hydrogenation of CO 2 into CH 4 , or CO 2 methanation, is also known as the Sabatier reaction and is an exothermic reaction with the following equation: CO 2 + 4H 2 → CH 4 + 2H 2 O ∆H 298 K = −165 kJ/mol (1) Ni has become a favourite active metal for this reaction, since its high methanation activity, low cost and natural abundance render it attractive for industrial-scale applications [6]. Since CH 4 yield peaks at a relatively low temperature (300-400 • C, depending on the reaction conditions) [7], structural degradation of Ni-based catalysts, though not completely avoided, plays a minor role compared to other reactions (e.g., methane dry Figure 1. Scheme of the CO2 + H2 reaction mechanisms over Ni/CeO2 and Ni/Al2O3 catalysts. Reproduced with permission from [12]. Copyright: Elsevier, Amsterdam, The Netherlands, 2020. The activity of Ni-based catalysts can be further improved via modification of the metal oxide supports. For example, alkali and alkaline earth metals [13], transition metals [6] and rare-earth metals [14] can be used as promoters that modify the physicochemical properties of metal oxide supports. In some cases, these ions can enter the lattice of the metal oxide supports (e.g., Ca 2+ ions in CeO2 and ZrO2 lattices) [15], or form segregated metal oxide phases supported on the support surface (e.g., La2O3, CeO2 and MnOx in Al2O3) [16]. Such modifications can lead to an increase in support basicity, so that the initial step of CO2 chemisorption step is accelerated, or to an increase in the active metal dispersion [17]. In most cases, the low-temperature activity and stability of Ni-based catalysts is enhanced following modification of the metal oxide supports [13,14,16].
Besides Ni, Ru and Rh noble metals have been extensively studied as active metallic phases in CO2 methanation and they usually achieve a much higher activity at low temperatures [18,19]. Since CH4 is thermodynamically favoured over other CO2 hydrogenation products such as CO, at low temperatures, CH4 selectivity can be significantly higher when using noble metal catalysts [7]. Among the two noble metals, Ru can achieve higher activity and its price is considerably lower compared to Rh, while it can also provide significant methanation activity when supported on cheap supports (e.g., Al2O3 or TiO2) at a metal loading as low as 1% or even 0.5% [20]. Ru is also preferable to Ni for application in the combined capture and methanation of CO2 derived from industrial flue-gases since the high reducibility of RuOx oxides allows for isothermal operation at low temperatures [21,22].
A popular method to counter some of the drawbacks of Ni-based catalysts is to use a second metal (e.g., Fe, Co or Ru) as a dopant, in order to create appropriate bimetallic Figure 1. Scheme of the CO 2 + H 2 reaction mechanisms over Ni/CeO 2 and Ni/Al 2 O 3 catalysts. Reproduced with permission from [12]. Copyright: Elsevier, Amsterdam, The Netherlands, 2020.
The activity of Ni-based catalysts can be further improved via modification of the metal oxide supports. For example, alkali and alkaline earth metals [13], transition metals [6] and rare-earth metals [14] can be used as promoters that modify the physicochemical properties of metal oxide supports. In some cases, these ions can enter the lattice of the metal oxide supports (e.g., Ca 2+ ions in CeO 2 and ZrO 2 lattices) [15], or form segregated metal oxide phases supported on the support surface (e.g., La 2 O 3 , CeO 2 and MnO x in Al 2 O 3 ) [16]. Such modifications can lead to an increase in support basicity, so that the initial step of CO 2 chemisorption step is accelerated, or to an increase in the active metal dispersion [17]. In most cases, the low-temperature activity and stability of Ni-based catalysts is enhanced following modification of the metal oxide supports [13,14,16].
Besides Ni, Ru and Rh noble metals have been extensively studied as active metallic phases in CO 2 methanation and they usually achieve a much higher activity at low temperatures [18,19]. Since CH 4 is thermodynamically favoured over other CO 2 hydrogenation products such as CO, at low temperatures, CH 4 selectivity can be significantly higher when using noble metal catalysts [7]. Among the two noble metals, Ru can achieve higher activity and its price is considerably lower compared to Rh, while it can also provide significant methanation activity when supported on cheap supports (e.g., Al 2 O 3 or TiO 2 ) at a metal loading as low as 1% or even 0.5% [20]. Ru is also preferable to Ni for application in the combined capture and methanation of CO 2 derived from industrial flue-gases since the high reducibility of RuO x oxides allows for isothermal operation at low temperatures [21,22].
A popular method to counter some of the drawbacks of Ni-based catalysts is to use a second metal (e.g., Fe, Co or Ru) as a dopant, in order to create appropriate bimetallic CO 2 methanation catalysts [6]. Such an approach has been successfully employed in other reactions. For example, NiFe alloys are active and stable catalysts for dry reforming of methane, since Fe can promote carbon gasification and significantly reduce coking through an intricate dealloying and realloying mechanism [23]. The combination of Ni with other metals can either lead to the formation of Ni-M alloys, or monometallic heterostructures

Promotion with Fe
Among all metals, Fe is by far the most studied element in bimetallic Ni-based catalysts for CO 2 methanation, since it is quite cheap and highly available and it exhibits a high solubility into the Ni lattice, favouring the formation of NiFe alloys [25]. It has been suggested, based on computational screening and catalytic experiments, that alumina supported and Ni-rich NiFe catalysts can improve the rate of CO 2 conversion, with the optimal Ni/(Ni + Fe) ratio being around 0.7-0.9 [40,41]. Many more works have focused on NiFe alloys prepared with different methods and supported on various metal oxides [42].
Generally, the Ni/Fe ratio in the alloy and the reducibility of the metal-oxide support appear to be the most critical parameters that determine whether Fe will promote or suppress the catalyst's CO 2 methanation activity.
Pandey et al. [43] were among the first to perform a systematic study about the Fe promotion in Ni-based methanation catalysts. The optimal active metal content of the catalysts consisted of 75% Ni and 25% Fe. When compared to their monometallic counterparts (Ni and Fe), the bimetallic NiFe alloy catalysts supported on alumina and silica were shown to exhibit higher CH 4 yields and this enhancement was more apparent in the alumina supported catalysts. This was attributed to the creation of a suitable alloy phase and to the increased CO 2 chemisorption at unreduced Fe 3 O 4 sites. The authors then compared the activity of Ni 3 Fe catalysts supported on different metal oxide supports, such as Al 2 O 3 , SiO 2 , ZrO 2 and TiO 2 , and noted that Al 2 O 3 supported catalysts yielded the best results and provided the largest promotion due to Fe alloying [42]. Regarding the optimal alloy composition, a statistical technique, namely response surface methodology (RSM), was also used. The model equation predicted that a 32.78% Ni loading and 7.67% Fe loading would lead to the optimal methane yield, and the obtained experimental results confirmed this model prediction [44].
Ray et al. [38] from the same group managed to develop a descriptor for the methanation of CO 2 over Ni 3 M bimetallic catalysts, with the second metal (M) being Fe and Cu. It was shown that the turnover frequency for methane production (TOF CH4 ) could be linearly correlated with the number of d-density of states (d-DOS) at the Fermi level (N EF ) based on density functional theory (DFT) calculations, as shown in Figure 2. The N EF descriptor successfully predicted the enhancement of CO 2 methanation performance over the Ni 3 Fe catalyst, due to a favourable change in the electronic properties of the active Ni phase, while Ni 3 Cu alloy formation was detrimental for the production of CH 4 . In a follow-up study, Ray et al. [45] also included Ni 3 Co catalysts in their calculations and concluded that Co alloying could also enhance the CO 2 methanation performance but by a smaller degree compared to Fe. They also showed that Co alloying led to the highest promotion regarding the dry methane reforming (DMR) reaction, with the location of the Ni d-band centre (ε d ) being the most appropriate descriptor to assess the catalytic activity for this reaction.

Promotion with Fe
Among all metals, Fe is by far the most studied element in bimetallic Ni-based catalysts for CO2 methanation, since it is quite cheap and highly available and it exhibits a high solubility into the Ni lattice, favouring the formation of NiFe alloys [25]. It has been suggested, based on computational screening and catalytic experiments, that alumina supported and Ni-rich NiFe catalysts can improve the rate of CO2 conversion, with the optimal Ni/(Ni + Fe) ratio being around 0.7-0.9 [40,41]. Many more works have focused on NiFe alloys prepared with different methods and supported on various metal oxides [42]. Generally, the Ni/Fe ratio in the alloy and the reducibility of the metal-oxide support appear to be the most critical parameters that determine whether Fe will promote or suppress the catalyst's CO2 methanation activity.
Pandey et al. [43] were among the first to perform a systematic study about the Fe promotion in Ni-based methanation catalysts. The optimal active metal content of the catalysts consisted of 75% Ni and 25% Fe. When compared to their monometallic counterparts (Ni and Fe), the bimetallic NiFe alloy catalysts supported on alumina and silica were shown to exhibit higher CH4 yields and this enhancement was more apparent in the alumina supported catalysts. This was attributed to the creation of a suitable alloy phase and to the increased CO2 chemisorption at unreduced Fe3O4 sites. The authors then compared the activity of Ni3Fe catalysts supported on different metal oxide supports, such as Al2O3, SiO2, ZrO2 and TiO2, and noted that Al2O3 supported catalysts yielded the best results and provided the largest promotion due to Fe alloying [42]. Regarding the optimal alloy composition, a statistical technique, namely response surface methodology (RSM), was also used. The model equation predicted that a 32.78% Ni loading and 7.67% Fe loading would lead to the optimal methane yield, and the obtained experimental results confirmed this model prediction [44].
Ray et al. [38] from the same group managed to develop a descriptor for the methanation of CO2 over Ni3M bimetallic catalysts, with the second metal (M) being Fe and Cu. It was shown that the turnover frequency for methane production (TOFCH4) could be linearly correlated with the number of d-density of states (d-DOS) at the Fermi level (NEF) based on density functional theory (DFT) calculations, as shown in Figure 2. The NEF descriptor successfully predicted the enhancement of CO2 methanation performance over the Ni3Fe catalyst, due to a favourable change in the electronic properties of the active Ni phase, while Ni3Cu alloy formation was detrimental for the production of CH4. In a follow-up study, Ray et al. [45] also included Ni3Co catalysts in their calculations and concluded that Co alloying could also enhance the CO2 methanation performance but by a smaller degree compared to Fe. They also showed that Co alloying led to the highest promotion regarding the dry methane reforming (DMR) reaction, with the location of the Ni d-band centre (εd) being the most appropriate descriptor to assess the catalytic activity for this reaction. Linear correlation between the turnover frequency for CH4 production (TOFCH4) and the number of d-density of states (d-DOS) at Fermi level (NEF) for Ni, Ni3Fe and Ni3Cu catalysts. Reproduced with permission from [38]. Copyright: Elsevier, Amsterdam, The Netherlands, 2017. Linear correlation between the turnover frequency for CH 4 production (TOF CH4 ) and the number of d-density of states (d-DOS) at Fermi level (N EF ) for Ni, Ni 3 Fe and Ni 3 Cu catalysts. Reproduced with permission from [38]. Copyright: Elsevier, Amsterdam, The Netherlands, 2017.
The Grunwaldt group have been amongst the pioneers in the development of NiFebased methanation catalysts and the elucidation of the role of Fe in the overall catalytic performance. Mutz et al. [26] prepared Ni 3 Fe catalysts supported on Al 2 O 3 via depositionprecipitation. The formed alloy nanoparticles exhibited a small size and high dispersion, and the NiFe-based catalyst proved to be more active and stable at lower temperatures compared to the monometallic Ni-based catalyst. The alloy catalyst was also proven to have a quite stable and robust performance after a 45 h time-on-stream operation under industrially oriented conditions [26]. No carbon deposition could be observed under various gas feeds for such catalysts using operando Raman spectroscopy [46]. Farsi et al. [47] investigated the CO 2 methanation kinetics on such Ni 3 Fe methanation catalysts under technical operation conditions. CO selectivity over CH 4 was found to increase over shorter residence times and higher temperatures, while water concentration was indicated as the main inhibiting factor. Furthermore, Serrer et al. [48] focussed specifically on the role of Fe during CO 2 methanation, employing advanced operando spectroscopic methods. Fe was shown to act as a protective or "sacrificial" dopant upon cases of H 2 -dropout during CO 2 methanation. While Ni-based catalysts are oxidized under such events and then fail to regain their initial activity upon the restoration of H 2 flow, NiFe-based catalysts retain the Ni 0 sites in their reduced state, due to the preferential oxidation of their Fe sites into FeO. Under normal operation, Fe was shown to increase the reducibility of Ni and result in the formation of small FeO x clusters at the surface of the alloy nanoparticles, with Fe being found in various oxidation states [49]. It has been suggested that increased methanation performance of NiFe alloy catalysts could be due to these small FeO x clusters on top of alloy nanoparticles that can potentially favour CO 2 chemisorption and activation, as shown in Figure 3.
based methanation catalysts and the elucidation of the role of Fe in the overall catalytic performance. Mutz et al. [26] prepared Ni3Fe catalysts supported on Al2O3 via depositionprecipitation. The formed alloy nanoparticles exhibited a small size and high dispersion, and the NiFe-based catalyst proved to be more active and stable at lower temperatures compared to the monometallic Ni-based catalyst. The alloy catalyst was also proven to have a quite stable and robust performance after a 45 h time-on-stream operation under industrially oriented conditions [26]. No carbon deposition could be observed under various gas feeds for such catalysts using operando Raman spectroscopy [46]. Farsi et al. [47] investigated the CO2 methanation kinetics on such Ni3Fe methanation catalysts under technical operation conditions. CO selectivity over CH4 was found to increase over shorter residence times and higher temperatures, while water concentration was indicated as the main inhibiting factor. Furthermore, Serrer et al. [48] focussed specifically on the role of Fe during CO2 methanation, employing advanced operando spectroscopic methods. Fe was shown to act as a protective or "sacrificial" dopant upon cases of H2-dropout during CO2 methanation. While Ni-based catalysts are oxidized under such events and then fail to regain their initial activity upon the restoration of H2 flow, NiFe-based catalysts retain the Ni 0 sites in their reduced state, due to the preferential oxidation of their Fe sites into FeO. Under normal operation, Fe was shown to increase the reducibility of Ni and result in the formation of small FeOx clusters at the surface of the alloy nanoparticles, with Fe being found in various oxidation states [49]. It has been suggested that increased methanation performance of NiFe alloy catalysts could be due to these small FeOx clusters on top of alloy nanoparticles that can potentially favour CO2 chemisorption and activation, as shown in Figure 3. . Scheme of the CO2 activation mechanism on Ni-Fe alloy-based catalysts during methanation reaction under realistic conditions. Reproduced with permission from [49]. Copyright: Royal Society of Chemistry, London, UK, 2020.
Mebrahtu et al. [39] used hydrotalcite-precursors with a tailored Fe/Ni ratio in order to prepare NiFe/(Mg,Al)Ox catalysts with high levels of metal intermixing and dispersion. The Fe/Ni ratio played a crucial role in the physicochemical properties and the catalytic performance of the prepared catalysts. An Fe/(Ni + Fe) ratio of 0.1 was shown to provide high metal dispersion, small nanoparticle sizes and an optimum amount of surface basic sites. Consequently, the catalyst with this specific ratio offered the best low-temperature catalytic performance in terms of CO2 conversion and CH4 selectivity, whereas catalysts with higher Fe contents experienced a significant drop for these values (Figure 4). Mebrahtu et al. [50] also indicated a possible deactivation pathway for monometallic Ni Figure 3. Scheme of the CO 2 activation mechanism on Ni-Fe alloy-based catalysts during methanation reaction under realistic conditions. Reproduced with permission from [49]. Copyright: Royal Society of Chemistry, London, UK, 2020.
Mebrahtu et al. [39] used hydrotalcite-precursors with a tailored Fe/Ni ratio in order to prepare NiFe/(Mg,Al)O x catalysts with high levels of metal intermixing and dispersion. The Fe/Ni ratio played a crucial role in the physicochemical properties and the catalytic performance of the prepared catalysts. An Fe/(Ni + Fe) ratio of 0.1 was shown to provide high metal dispersion, small nanoparticle sizes and an optimum amount of surface basic sites. Consequently, the catalyst with this specific ratio offered the best lowtemperature catalytic performance in terms of CO 2 conversion and CH 4 selectivity, whereas catalysts with higher Fe contents experienced a significant drop for these values (Figure 4). Mebrahtu et al. [50] also indicated a possible deactivation pathway for monometallic Ni catalysts via the formation of Ni hydroxides caused by the water produced in situ upon methanation. It was found that the introduction of Fe prevented the formation of Ni-OH species, thus increasing the catalytic activity of such systems. It was also argued that Fe formed spinel phases on the alumina nanosheets and did not form alloys with Ni.
catalysts via the formation of Ni hydroxides caused by the water produced in situ upon methanation. It was found that the introduction of Fe prevented the formation of Ni-OH species, thus increasing the catalytic activity of such systems. It was also argued that Fe formed spinel phases on the alumina nanosheets and did not form alloys with Ni. Figure 4. Effect of Fe content on CH4 (○) and CO (△) selectivity in comparison to sole metals. The highest methane selectivity, with 90%, was achieved for the catalyst containing 10% Fe. Reproduced with permission from [39]. Copyright: Royal Society of Chemistry, London, UK, 2018.
Giorgianni et al. [51] attributed the beneficial role of Fe into such hydrotalcite-derived catalysts to the presence of Fe(II) species. These species could activate CO2 molecules and adsorb H2O produced in situ during the reaction, thus preventing the hydroxylation of nearby Ni 0 active sites. However, these Fe(II) species inevitably undergo oxidation into Fe(III) over time, reducing the catalytic performance.
In a similar fashion, Huynh et al. [52] observed high CO2 methanation activity and stability for cheap hydrotalcite-derived NiFe/(Mg,Al)Ox catalysts, which could achieve around 75% CO2 conversion and 95% CH4 selectivity at just 300 °C under high gas space velocities. They also showed that an Fe/(Ni + Fe) ratio of 0.2 (Ni4Fe) could lead to the highest promotion in CH4 yield at low temperatures, due to a considerable decrease in the activation energy for CH4 formation [53]. The formate pathway was observed to be favoured over the direct dissociation of CO2 into CO via in-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and density functional theory (DFT) [53]. Furthermore, NiFe-containing layered double hydroxides (LDHs) could be in situ grown over alumina and silica washcoated cordierite monoliths via an urea hydrolysis preparation method [54]. Using different washcoat materials and metal concentrations, a structured NiFe bimetallic catalyst with a thin catalytic layer on the cordierite substrate was prepared. This structured NiFe monolithic catalyst achieved high CH4 yields under high gas flow rates, thus making it attractive for industrial-scale applications.
Burger et al. [55] prepared NiAlOx coprecipitated catalysts modified with Fe and Mn. Both metals improved the performance of the NiAlOx catalysts. Mn formed separate MnOx phases that enhanced the CO2 adsorption capacity and metal dispersion, while Fe formed NiFe alloys, promoting the electronic properties of Ni and enhancing the catalyst stability. The optimal Ni to promoter molar ratio was five. Such catalysts were also proven to be resistant upon sulphur poisoning, since H2S was preferentially adsorbed at the metal promoter phases [56], thus preserving the Ni active sites [57].
Burger et al. [58,59] also studied the mechanism of Fe promotion in coprecipitated Ni/Al2O3 catalysts. In one study [58], Fe was deposited adjacent to Ni nanoparticles via a surface redox mechanism creating alloyed surface phases. Fe alloying also increased the activity of Ni/Al2O3 catalysts prepared via deposition-precipitation and the thermal stability of co-precipitated NiAlOx catalysts. (γFe,Ni) alloy nanoparticles were observed at Giorgianni et al. [51] attributed the beneficial role of Fe into such hydrotalcite-derived catalysts to the presence of Fe(II) species. These species could activate CO 2 molecules and adsorb H 2 O produced in situ during the reaction, thus preventing the hydroxylation of nearby Ni 0 active sites. However, these Fe(II) species inevitably undergo oxidation into Fe(III) over time, reducing the catalytic performance.
In a similar fashion, Huynh et al. [52] observed high CO 2 methanation activity and stability for cheap hydrotalcite-derived NiFe/(Mg,Al)O x catalysts, which could achieve around 75% CO 2 conversion and 95% CH 4 selectivity at just 300 • C under high gas space velocities. They also showed that an Fe/(Ni + Fe) ratio of 0.2 (Ni 4 Fe) could lead to the highest promotion in CH 4 yield at low temperatures, due to a considerable decrease in the activation energy for CH 4 formation [53]. The formate pathway was observed to be favoured over the direct dissociation of CO 2 into CO via in-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and density functional theory (DFT) [53]. Furthermore, NiFe-containing layered double hydroxides (LDHs) could be in situ grown over alumina and silica washcoated cordierite monoliths via an urea hydrolysis preparation method [54]. Using different washcoat materials and metal concentrations, a structured NiFe bimetallic catalyst with a thin catalytic layer on the cordierite substrate was prepared. This structured NiFe monolithic catalyst achieved high CH 4 yields under high gas flow rates, thus making it attractive for industrial-scale applications.
Burger et al. [55] prepared NiAlO x coprecipitated catalysts modified with Fe and Mn. Both metals improved the performance of the NiAlO x catalysts. Mn formed separate MnO x phases that enhanced the CO 2 adsorption capacity and metal dispersion, while Fe formed NiFe alloys, promoting the electronic properties of Ni and enhancing the catalyst stability. The optimal Ni to promoter molar ratio was five. Such catalysts were also proven to be resistant upon sulphur poisoning, since H 2 S was preferentially adsorbed at the metal promoter phases [56], thus preserving the Ni active sites [57].
Burger et al. [58,59] also studied the mechanism of Fe promotion in coprecipitated Ni/Al 2 O 3 catalysts. In one study [58], Fe was deposited adjacent to Ni nanoparticles via a surface redox mechanism creating alloyed surface phases. Fe alloying also increased the activity of Ni/Al 2 O 3 catalysts prepared via deposition-precipitation and the thermal stability of co-precipitated NiAlO x catalysts. (γFe,Ni) alloy nanoparticles were observed at co-precipitated NiFeAlO x catalysts and the presence of Fe(II) species upon aging provided additional CO 2 activation sites [59]. The generation of active Fe(II) species appeared to partially offset the negative effects caused by active metal sintering upon aging.
Other reports about Fe promotion in Ni/Al 2 O 3 catalysts include the work of Li et al. [60]. It was found that adding 3% Fe on a 12% Ni/Al 2 O 3 catalyst led to a mild improvement for CO 2 conversion and CH 4 selectivity, whereas the increase in the Fe content to 12% (Fe/Ni ≈ 1) caused a worse methanation performance. Liang et al. [61] studied the effect of various additives on Ni/Al 2 O 3 catalysts and found the presence of an increased number of oxygen vacancies over the Fe-modified catalyst, as evidenced by electron paramagnetic resonance (EPR), that caused favourable changes to the reaction pathway. Finally, in contrast to other works, Daroughegi et al. [62] found that a 25% Ni/Al 2 O 3 catalyst modified with 5% Fe exhibited a much worse methanation performance in terms of CO 2 conversion compared to the corresponding monometallic Ni catalyst.
Up until now, the NiFe-based catalysts studied were supported on Al 2 O 3 or Al-based "inert" supports. The oxide support can, however, also play a major role in the methanation performance of alloy catalysts [63]. Ren et al. [64] showed that modification of a 30% Ni/ZrO 2 catalyst with 3% Fe enhanced its low temperature methanation performance. Modification with 5% Fe also yielded better results compared to the monometallic catalyst, but higher Fe contents led to a decline in methanation activity. The authors suggested that the majority of Fe in the catalysts was not fully reduced following pretreatment under H 2 flow, but remained at an Fe(II) oxidation state. These species could potentially improve the dispersion and reducibility of the Ni phase and also promote the reduction of ZrO 2 , thus facilitating the presence of oxygen vacancies, that together with Fe(II) sites typically enhance CO 2 chemisorption and dissociation.
Yan et al. [65] employed low metal loadings of Ni and Fe (1.5% Ni and 0.5-4.5% Fe) supported ZrO 2 and assigned the various interfacial sites on the catalysts as selective for different CO 2 hydrogenation products. The Ni-ZrO 2 interface in the monometallic catalyst was characterized as active for the methanation reaction. Fe addition up to an equimolar amount to that of Ni led to a small improvement in CO 2 conversion and CH 4 selectivity and probably preserved the methane selective active metal-ZrO 2 interface. Only with the addition of a large amount of Fe (Fe/Ni ratio at 3) does the CO selectivity increase, via the creation of Ni-FeO x interface that binds intermediate CO weakly and favours the reverse water gas shift (RWGS) reaction. Their experimental results are summarized in Figure 5. Furthermore, other studies showed that Fe promotion provided an enhancement of the methanation performance of Ni-based catalysts supported on Al 2 O 3 and mesoporous clay modified with ZrO 2 [66,67].
It has been previously mentioned that CeO 2 support can offer significant advantages regarding the CO 2 methanation performance of Ni-based catalysts, when compared with other metal oxide supports [12]. However, Fe modification of Ni catalysts supported on CeO 2 -based supports appears to exert a negative influence on CO 2 methanation. Winter et al. [68] studied CeO 2 -supported NiFe catalysts with low metal loadings of Ni and Fe (1.5% Ni and 0.5-1.5% Fe). In contrast to the ZrO 2 supported catalysts reported by Yan et al. [65], Fe modification, even in small amounts, led to a significant drop in CO 2 conversion and rise in CO selectivity. The majority of Fe remained oxidized according to X-ray absorption near edge structure (XANES) analysis and these oxidized Fe species probably led to a weakened interaction between metal and intermediate CO, thus facilitating CO desorption and higher CO selectivity [68].
Pastor-Pérez et al. [69] prepared Fe and Co modified Ni catalysts supported on CeO 2 -ZrO 2 and observed a negative effect of Fe addition, since the 3% Fe, 15% Ni/CeO 2 -ZrO 2 catalyst exhibited lower values for CO 2 conversion and CH 4 selectivity compared to the monometallic Ni catalyst. Likewise, le Saché et al. [70] observed that Fe addition on a Ni/CeO 2 -ZrO 2 catalyst led to a drop in CO 2 conversion and CH 4 selectivity. Furthermore, Frontera et al. [71] also reported a drop in CO 2 methanation activity upon Fe alloying over Ni catalysts supported on Gd-doped CeO 2 (GDC). In contrast to NiFe catalysts supported on inert supports such as Al 2 O 3 and SiO 2 [26,43], Fe addition appeared to suppress the population of surface oxygen vacancies on the already defect-rich GDC support. amount to that of Ni led to a small improvement in CO2 conversion and CH4 selectivity and probably preserved the methane selective active metal-ZrO2 interface. Only with the addition of a large amount of Fe (Fe/Ni ratio at 3) does the CO selectivity increase, via the creation of Ni-FeOx interface that binds intermediate CO weakly and favours the reverse water gas shift (RWGS) reaction. Their experimental results are summarized in Figure 5. Furthermore, other studies showed that Fe promotion provided an enhancement of the methanation performance of Ni-based catalysts supported on Al2O3 and mesoporous clay modified with ZrO2 [66,67]. It has been previously mentioned that CeO2 support can offer significant advantages regarding the CO2 methanation performance of Ni-based catalysts, when compared with other metal oxide supports [12]. However, Fe modification of Ni catalysts supported on CeO2-based supports appears to exert a negative influence on CO2 methanation. Winter et al. [68] studied CeO2-supported NiFe catalysts with low metal loadings of Ni and Fe (1.5% Ni and 0.5-1.5% Fe). In contrast to the ZrO2 supported catalysts reported by Yan et al. [65], Fe modification, even in small amounts, led to a significant drop in CO2 conversion and rise in CO selectivity. The majority of Fe remained oxidized according to X-ray absorption near edge structure (XANES) analysis and these oxidized Fe species probably led to a weakened interaction between metal and intermediate CO, thus facilitating CO desorption and higher CO selectivity [68].
Pastor-Pérez et al. [69] prepared Fe and Co modified Ni catalysts supported on CeO2-ZrO2 and observed a negative effect of Fe addition, since the 3% Fe, 15% Ni/CeO2-ZrO2 catalyst exhibited lower values for CO2 conversion and CH4 selectivity compared to the monometallic Ni catalyst. Likewise, le Saché et al. [70] observed that Fe addition on a Ni/CeO2-ZrO2 catalyst led to a drop in CO2 conversion and CH4 selectivity. Furthermore, Frontera et al. [71] also reported a drop in CO2 methanation activity upon Fe alloying over Ni catalysts supported on Gd-doped CeO2 (GDC). In contrast to NiFe catalysts supported on inert supports such as Al2O3 and SiO2 [26,43], Fe addition appeared to suppress the population of surface oxygen vacancies on the already defect-rich GDC support.
Carbon is another type of support in heterogeneous catalysis, which has been deemed as inactive regarding CO2 methanation when using Ni-based catalysts [72]. However, Gonçalves et al. [73] managed to improve the performance of activated carbon (AC) supported Ni catalysts by modifying the surface chemistry of carbon and by introducing Fe as a second metal. They reported that 5% Fe addition on a 15% Ni catalyst supported on activated carbon with increased basicity (AC-R) improved the catalytic performance Carbon is another type of support in heterogeneous catalysis, which has been deemed as inactive regarding CO 2 methanation when using Ni-based catalysts [72]. However, Gonçalves et al. [73] managed to improve the performance of activated carbon (AC) supported Ni catalysts by modifying the surface chemistry of carbon and by introducing Fe as a second metal. They reported that 5% Fe addition on a 15% Ni catalyst supported on activated carbon with increased basicity (AC-R) improved the catalytic performance for CO 2 methanation. More specifically, NiFe alloy formation improved the low-temperature activity of the supported catalysts, increased the CH 4 selectivity and favoured catalyst stability, as shown in Figure 6. for CO2 methanation. More specifically, NiFe alloy formation improved the low-temperature activity of the supported catalysts, increased the CH4 selectivity and favoured catalyst stability, as shown in Figure 6. The formation of NiFe alloy phase can sometimes take place via the migration of Fe atoms that are incorporated in the support lattice, towards the surface Ni particles, upon exposure to a reducing atmosphere. Wang et al. [74] impregnated Ni on an olivine support and observed the formation of NiFe alloyed phase from the Fe contained in the support. The good CO2 methanation performance was ascribed to NiFe alloy formation and its fa- The formation of NiFe alloy phase can sometimes take place via the migration of Fe atoms that are incorporated in the support lattice, towards the surface Ni particles, upon exposure to a reducing atmosphere. Wang et al. [74] impregnated Ni on an olivine support and observed the formation of NiFe alloyed phase from the Fe contained in the support. The good CO 2 methanation performance was ascribed to NiFe alloy formation and its favourable interaction with unreduced FeO x segregated from the olivine support upon calcination and reduction treatments. Thalinger et al. [75], however, reported that NiFe alloy formation through Fe exsolution from a reducible La 0.6 Sr 0.4 FeO 3-δ perovskite with supported Ni nanoparticles could spoil the methanation chemistry of Ni through an unfavourable change in nickel's electronic properties. The suppression of CO 2 methanation was less pronounced on a Ni/SrTi 0.7 Fe 0.3 O 3-δ catalyst, due to the better structural stability of this perovskite support [76] and, thus, the lesser extent of Fe exsolution and alloying with Ni.
Lastly, Pandey et al. [77] performed a study in which they used unsupported NiO as a catalyst precursor and introduced various Fe amounts. Similarly to many of their supported counterparts [26,39], Ni catalysts modified with Fe in a stoichiometry of Ni/Fe > 1 exhibited an increase in catalytic activity for CO 2 methanation compared to metallic Ni catalyst. The catalysts that consisted of Ni-rich NiFe alloys with 10% Fe and 25% Fe yielded the best results. The one with 10% Fe was shown to be the most active and the higher activity was ascribed to the absence of unalloyed Fe in this catalyst [77].
By taking into consideration the large number of literature reports regarding Fe modified Ni catalysts [26,39,65], it can be concluded that, under specific circumstances, Fe can significantly promote the CO 2 methanation performance. The molar ratio between Ni and Fe appears to be the most important factor that determines whether Fe will promote or demote the methanation performance. An Fe/Ni ratio between 0.1 and 0.25 mostly yields the best results, while Fe-rich alloys significantly degrade the catalytic activity [39]. Furthermore, the type of support also influences the degree of promotion. Al 2 O 3 , ZrO 2 , SiO 2 and carbon supported Ni catalysts are typically promoted upon Fe addition at a specific Fe/Ni ratio, but Ni catalysts supported on defect-rich CeO 2 -based supports do not exhibit a similar behaviour [42,68]. Through spectroscopic techniques, it has been reported that partly reduced Fe(II) species play a major role in the promotion mechanism [49,50]. In general, Fe constitutes an ideal promoter for Ni-based catalysts, since it is also a cheap and earth-abundant metal and, through the favourable formation of NiFe alloy, it can be used to explicitly tailor the electronic properties of the catalytically active phase. Some of the most representative studies that include Ni-M (M = transition metal) bimetallic catalysts for CO 2 methanation are comparatively presented in Table 1.

Promotion with Co
Co is another transition metal commonly used to prepare bimetallic NiCo catalysts for CO 2 methanation. Co is right adjacent to Ni in the periodic table, it can dissolve into the lattice of metallic Ni similarly to Fe and its easy transition between the oxidation states of Co(III), Co(II) and Co 0 can induce modifications to the electronic properties of Nibased catalysts [24,78]. Among the first works, Guo et al. [79] prepared several bimetallic NiCo catalysts supported on SiO 2 with different Co/Ni ratios. They found that higher Co loadings improved the activity of the catalysts and that a Co/Ni ratio of 0.4 led to the best performing catalyst. The formation of a homogeneous NiCo alloy supported on SiO 2 has also been shown to promote CO dissociation and hydrogen spillover during CO methanation, leading to higher activity [78].
Further works focussed on Al 2 O 3 supported NiCo catalysts, with Xu et al. [80] and Liu et al. [37] introducing ordered mesoporosity into the catalyst structure via a one-pot evaporation induced self-assembly (EISA) synthesis method. Xu et al. [80] reported that a suitable Co/(Ni + Co) ratio of 20% could increase the catalytic activity and stability. The authors suggested that Ni and Co were supported as adjacent monometallic phases on the mesoporous Al 2 O 3 structure and that they served as active sites for the chemisorption of H 2 and CO 2 , respectively. The synergy between these two metals acted to decrease the activation energy for CO 2 methanation. Liu et al. [37] also prepared similar ordered mesoporous NiCo/Al 2 O 3 composites. The authors claimed that NiCo alloy formation and the confinement effect of the mesoporous structure were responsible for the improved low-temperature activity and stability of the bimetallic catalysts. Their best performing 10N3COMA catalyst (3% Co 3 O 4 and 10% NiO on ordered mesoporous Al 2 O 3 ) exhibited 78% CO 2 conversion and 99% CH 4 selectivity at 400 • C, as well as great stability upon a 60 h time on stream test, as shown in Figure 7.
Further works focussed on Al2O3 supported NiCo catalysts, with Xu et al. [80] and Liu et al. [37] introducing ordered mesoporosity into the catalyst structure via a one-pot evaporation induced self-assembly (EISA) synthesis method. Xu et al. [80] reported that a suitable Co/(Ni + Co) ratio of 20% could increase the catalytic activity and stability. The authors suggested that Ni and Co were supported as adjacent monometallic phases on the mesoporous Al2O3 structure and that they served as active sites for the chemisorption of H2 and CO2, respectively. The synergy between these two metals acted to decrease the activation energy for CO2 methanation. Liu et al. [37] also prepared similar ordered mesoporous NiCo/Al2O3 composites. The authors claimed that NiCo alloy formation and the confinement effect of the mesoporous structure were responsible for the improved lowtemperature activity and stability of the bimetallic catalysts. Their best performing 10N3COMA catalyst (3% Co3O4 and 10% NiO on ordered mesoporous Al2O3) exhibited 78% CO2 conversion and 99% CH4 selectivity at 400 °C, as well as great stability upon a 60 h time on stream test, as shown in Figure 7. Alrafei et al. [81] prepared Ni and NiCo catalysts with various metal loadings supported on Al 2 O 3 extrudates. According to their findings, 10% Ni and 10% Co were the most suitable metal loadings for the bimetallic catalysts. The bimetallic catalyst with these metal loadings outperformed the monometallic 10% Ni catalyst, as Co helped to increase the reducibility and dispersion of the Ni phase. However, the 20% monometallic Ni catalyst exhibited better performance in terms of CO 2 conversion and CH 4 selectivity when compared to the bimetallic catalyst, with a total Ni and Co metal loading of 20%. All catalysts presented a remarkable stability upon 200 h of operation. In contrast to other works, Fatah et al. [82] reported that 5% Co addition on a 5% Ni/Al 2 O 3 catalyst caused roughly a 30% drop in CO 2 conversion and lower CH 4 selectivity compared to the Ni monometallic catalyst. The deactivation was attributed to the occurrence of larger particles in the NiCo bimetallic catalyst, while an increased formation of formate intermediate species was also identified via in situ FTIR.
Besides SiO 2 and Al 2 O 3 , a lot of works focus on NiCo catalysts supported on ZrO 2 , CeO 2 -ZrO 2 and other CeO 2 -based supports. As discussed before, Ren et al. [64] studied 30% Ni/ZrO 2 catalysts modified with Fe, Co and Cu. The Co-modified catalyst exhibited higher CO 2 conversion, but slightly lower CH 4 selectivity, and Fe was shown to be more suitable as a promoter metal. Razzaq et al. [83], on the other hand, concluded that a Co promoted Ni/CeO 2 -ZrO 2 catalyst was more suitable compared to other catalysts (modified with Mo and Fe) for the co-methanation of CO and CO 2 in CH 4 -rich coke oven gas. Following that work, Zhu et al. [84] confirmed that 5% Co addition on a 15% Ni/CeO 2 -ZrO 2 catalyst could promote catalytic activity and stability for CO 2 methanation. The optimum CeO 2 /ZrO 2 molar ratio of 1/3 (Ce 0.25 Zr 0.75 O 2 ) was shown to also play a major role by providing a maximum amount of oxygen vacancies close to active metal sites that favour CO 2 activation and dissociation.
As previously discussed, Pastor-Pérez et al. [69] prepared Fe-and Co-modified Ni catalysts supported on CeO 2 -ZrO 2 . While Fe modification impeded the CO 2 methanation activity, the addition of 3% Co on a 15% Ni/CeO 2 -ZrO 2 catalyst enhanced the CO 2 conversion, CH 4 selectivity and the long-term stability and it allowed for the use of higher space velocities. Following that work, Pastor-Pérez et al. [85] also showed that such Co modified Ni catalyst could prevent coke deposition during the reaction. The catalyst was quite robust upon long-term operation tests, while the presence of methane in the feed gas did not appear to impede the catalytic activity, but instead promoted it. In contrast to other works, the addition of Co in Ni catalysts supported on defect-rich GDC support did not offer any advantages to the catalytic performance for CO 2 methanation, as reported by Frontera et al. [86].
Apart from the typical catalyst supports, Jia et al. [87] supported the two metals of Ni and Co on TiO 2 coated silica spheres. The electronic effect of the reducible TiO 2 layer acted to promote metal dispersion and the adsorption of H 2 and CO 2 , while the addition of Co second metal enhanced the intrinsic activity and long-term stability of the NiCo/TiO 2 @SiO 2 catalyst in a fluidized bed reactor. Varun et al. [88] prepared NiO-MgO nanocomposites, via a sonochemical method, that were active for the CO 2 methanation reaction. The authors further modified the composites by impregnating 2% Co, Fe and Cu. Among the three dopants, only Co led to significant activity improvement by lowering the activation energy for the methanation pathway.  Co improved the catalytic performance for CO 2 methanation above 250 • C. NiCo modified catalyst had great stability after many hours and could retain high activity under increased gas space velocities. Co increased Ni dispersion and could decrease coke deposition due to its redox properties. Cu addition improved Ni dispersion and reducibility and NiCu alloys were formed. Cu introduction also decreased CO 2 conversion and CH 4 selectivity, favouring instead the production of CO via the reverse water-gas shift (RWGS) reaction.
Lastly, Zhang et al. [89] attempted to increase the metal intermixing between Ni and Co by incorporating Co into the lattice of LaNiO 3 perovskite and supporting the perovskite on a mesostructured cellular foam (MCF) silica support. Following reduction, NiCo alloy nanoparticles and adjacent La 2 O 3 phase, highly dispersed on the MCF support, were generated. The promoting effect of Co alloying, the increased basicity and dispersion induced by the La 2 O 3 phase and the confinement effect of the mesoporous support acted to yield a catalyst with increased activity and great stability. A low Co/(Ni + Co) ratio of 5% at a perovskite with nominal composition LaNi 0.95 Co 0.05 O 3 was found to be optimal by leading to NiCo alloy particles with the smallest size ( Figure 8). yield a catalyst with increased activity and great stability. A low Co/(Ni + Co) ratio of 5% at a perovskite with nominal composition LaNi0.95Co0.05O3 was found to be optimal by leading to NiCo alloy particles with the smallest size ( Figure 8). Overall, it can be concluded that Co mostly acts to promote the performance of CO2 methanation catalysts. However, based on the reported literature, it appears that the type of support, either inert or reducible, does not affect the promotion mechanism [37,69]. The effect of the Co/(Ni + Co) ratio is also not as apparent when compared to similar NiFe bimetallic catalysts. However, it appears that Ni-rich NiCo catalysts lead to a more favourable activity [89].

Promotion with Cu
In all of the works discussed until now, whenever Cu is added in Ni-based catalysts, the CO2 methanation reaction is hindered and the antagonistic reverse water-gas shift (RWGS) reaction is instead promoted. Thus, the Cu-containing bimetallic catalysts turned out to be inferior to monometallic Ni ones for CO2 methanation [38,45,64,88]. Dias et al. [90] also showed that even 1% Cu addition on 10% Ni/SiO2 could drastically decrease CO2 conversion and increase the selectivity for CO. Although Cu enhanced the dispersion and reducibility of Ni, similarly to other transition metals (Fe and Co), it deactivated the catalyst for the CO2 methanation reaction. This was probably a result of NiCu alloy formation and the inability of pure Cu and NiCu alloy to adsorb hydrogen. Despite that, the Cucontaining catalysts could be suitable for CO production, due to their high stability and CO selectivity. Table 1 below summarises typical bimetallic Ni-based catalysts promoted with transition metals (Fe, Co and Cu) for the methanation of CO2. Overall, it can be concluded that Co mostly acts to promote the performance of CO 2 methanation catalysts. However, based on the reported literature, it appears that the type of support, either inert or reducible, does not affect the promotion mechanism [37,69]. The effect of the Co/(Ni + Co) ratio is also not as apparent when compared to similar NiFe bimetallic catalysts. However, it appears that Ni-rich NiCo catalysts lead to a more favourable activity [89].

Promotion with Cu
In all of the works discussed until now, whenever Cu is added in Ni-based catalysts, the CO 2 methanation reaction is hindered and the antagonistic reverse water-gas shift (RWGS) reaction is instead promoted. Thus, the Cu-containing bimetallic catalysts turned out to be inferior to monometallic Ni ones for CO 2 methanation [38,45,64,88]. Dias et al. [90] also showed that even 1% Cu addition on 10% Ni/SiO 2 could drastically decrease CO 2 conversion and increase the selectivity for CO. Although Cu enhanced the dispersion and reducibility of Ni, similarly to other transition metals (Fe and Co), it deactivated the catalyst for the CO 2 methanation reaction. This was probably a result of NiCu alloy formation and the inability of pure Cu and NiCu alloy to adsorb hydrogen. Despite that, the Cucontaining catalysts could be suitable for CO production, due to their high stability and CO selectivity. Table 1 below summarises typical bimetallic Ni-based catalysts promoted with transition metals (Fe, Co and Cu) for the methanation of CO 2 .

Promotion with Noble Metals
Although noble metals are significantly pricier than transition metals, they can provide significant advantages regarding the CO 2 methanation reaction, due to their excellent low-temperature activity, as well as their high reducibility and stability once oxidized [18,19,[91][92][93]. The combination of Ni with a noble metal aims to transfer some of these properties into Ni-based catalysts, without the need for the use of a high noble metal loading [33]. Combined Ni and Ru catalysts exist mostly in the form of heterostructures, rather than alloys, and Ru can provide additional methanation sites, as well as spillover hydrogen into nearby Ni sites [94]. On the other hand, Pt and Pd mostly modify the electronic properties of Ni via alloy formation [31,32].

Promotion with Ru
Ru is the cheapest among the discussed noble metals and, at the same time, the most active metallic phase in CO 2 methanation. The combination of the metals Ru and Ni has been described as "privileged" due to the great benefits that can arise from the NiRu bimetallic synergy [94]. Ru can drastically improve the reducibility of Ni catalysts, as well as improve the Ni metallic dispersion and provide additional methanation sites.
In an early work, Zhen et al. [30] impregnated Ni and Ru on γ-Al 2 O 3 . Ru was found to be segregated at the outer surface and not forming an alloy with the dominant Ni metallic phase. The catalyst with 1% Ru and 10% Ni presented high activity, CH 4 selectivity and stable performance upon 100 h of operation. It was suggested that CO 2 molecules dissociated on Ru particles and H 2 molecules on Ni ones, followed by hydrogen spillover to hydrogenate the adsorbed carbon species (Figure 9). Lange et al. [95] also found that part of Ru noble metal could be substituted by Ni in Ru/ZrO 2 catalysts without compromising the catalytic activity. They claimed that low metal loadings led to the formation of alloys, but higher loadings led to the separation of the two metallic phases.
Although noble metals are significantly pricier than transition metals, they can provide significant advantages regarding the CO2 methanation reaction, due to their excellent low-temperature activity, as well as their high reducibility and stability once oxidized [18,19,[91][92][93]. The combination of Ni with a noble metal aims to transfer some of these properties into Ni-based catalysts, without the need for the use of a high noble metal loading [33]. Combined Ni and Ru catalysts exist mostly in the form of heterostructures, rather than alloys, and Ru can provide additional methanation sites, as well as spillover hydrogen into nearby Ni sites [94]. On the other hand, Pt and Pd mostly modify the electronic properties of Ni via alloy formation [31,32].

Promotion with Ru
Ru is the cheapest among the discussed noble metals and, at the same time, the most active metallic phase in CO2 methanation. The combination of the metals Ru and Ni has been described as "privileged" due to the great benefits that can arise from the NiRu bimetallic synergy [94]. Ru can drastically improve the reducibility of Ni catalysts, as well as improve the Ni metallic dispersion and provide additional methanation sites.
In an early work, Zhen et al. [30] impregnated Ni and Ru on γ-Al2O3. Ru was found to be segregated at the outer surface and not forming an alloy with the dominant Ni metallic phase. The catalyst with 1% Ru and 10% Ni presented high activity, CH4 selectivity and stable performance upon 100 h of operation. It was suggested that CO2 molecules dissociated on Ru particles and H2 molecules on Ni ones, followed by hydrogen spillover to hydrogenate the adsorbed carbon species (Figure 9). Lange et al. [95] also found that part of Ru noble metal could be substituted by Ni in Ru/ZrO2 catalysts without compromising the catalytic activity. They claimed that low metal loadings led to the formation of alloys, but higher loadings led to the separation of the two metallic phases. Hwang et al. [96] modified the Ru content on a 35% Ni and 5% Fe/Al2O3 xerogel. A volcano-shaped trend was observed regarding the Ru content, with the small loading of 0.6% yielding the best results. Liu et al. [97] synthesized Ni and Ru-doped ordered mesoporous CaO-Al2O3 nanocomposites. The confinement of the active nanoparticles due to the ordered mesoporous structure of the catalyst prevented their sintering, while the CaO component increased the basicity and favoured CO2 chemisorption. The addition of a small quantity of Ru as a second metal enhanced the activation of H2 and CO2, due to its synergy with the Ni primary phase, leading to a final nanocomposite catalyst with superior catalytic activity and stability. In a more recent study, Chein et al. [98] also reported Hwang et al. [96] modified the Ru content on a 35% Ni and 5% Fe/Al 2 O 3 xerogel. A volcano-shaped trend was observed regarding the Ru content, with the small loading of 0.6% yielding the best results. Liu et al. [97] synthesized Ni and Ru-doped ordered mesoporous CaO-Al 2 O 3 nanocomposites. The confinement of the active nanoparticles due to the ordered mesoporous structure of the catalyst prevented their sintering, while the CaO component increased the basicity and favoured CO 2 chemisorption. The addition of a small quantity of Ru as a second metal enhanced the activation of H 2 and CO 2 , due to its synergy with the Ni primary phase, leading to a final nanocomposite catalyst with superior catalytic activity and stability. In a more recent study, Chein et al. [98] also reported that a 1% Ru and 10% Ni/Al 2 O 3 catalyst had a better low-temperature activity compared to the respective monometallic catalysts, also agreeing with the fact that a small amount of added Ru can induce significant changes to the catalytic activity of Ni-based catalysts.
NiRu bimetallic catalysts are also viable candidates to be studied under more industriallike conditions. Bustinza et al. [99] prepared bimetallic NiRu structured monolithic catalysts with low Ru contents by homogeneously dispersing Ni and Ru precursors over alumina washcoated monoliths. Through an appropriate preparation method, Ni was supported in the form of small nanoparticles (2-4 nm), while Ru was atomically dispersed over the structured support. The bimetallic catalyst with these characteristics provided stable CO 2 methanation performance for many hours under high space velocities. Navarro et al. [100] structured a catalyst consisting of MgAl 2 O 4 supported 0.5% Ru and 15% Ni washcoated on metal micromonoliths. The structured catalyst was stable upon 100 h of continuous operation. The effect of CH 4 presence in the initial gas stream (simulated biogas) was also addressed by Stangeland et al. [101]. CO 2 conversion was increased upon promotion of a 20% Ni/Al 2 O 3 catalyst with 0.5% Ru and the catalyst was not greatly affected by the CH 4 presence in the gas stream, exhibiting a stable methanation activity.
Another practical implementation of the NiRu bimetallic combination is in dualfunction materials (DFMs), used for the integrated capture and methanation of CO 2 [102]. Having Ni active methanation phase in DFMs presents a major problem, since Ni is deactivated during the CO 2 capture step from flue gases and cannot be reactivated under H 2 flow at the usual operation temperature of 320 • C [21,103]. Adding just 1% Ru in a DFM consisting of 10% Ni increased the reducibility of the Ni phase by 70% and thus enabled CH 4 production under isothermal cyclic operation [104]. Employing in-situ DRIFTS, it was shown that both Ru and Ni are active sites for the methanation reaction, with bicarbonate and bidentate carbonate intermediate species spilling over to both metals before being hydrogenated into CH 4 [105].
Besides Al 2 O 3 -based supports, ZrO 2 , CeO 2 -ZrO 2 and other CeO 2 -based defect-rich supports are also commonly used in NiRu-based catalysts and will be discussed thereafter. Ocampo et al. [106] modified 5% Ni catalysts supported on CeO 2 -ZrO 2 by changing the CeO 2 /ZrO 2 ratio and by adding a noble metal phase (Ru and Rh). A mass ratio between CeO 2 and ZrO 2 of 1.5 in the support led to the best results and 0.5% noble metal addition increased the catalytic activity and stability by enhancing Ni dispersion. In a later study, Shang et al. [107] prepared Ru-doped 30% Ni catalysts supported on CeO 2 -ZrO 2 via an one-pot hydrolysis method. They found that Ru addition could enhance Ni dispersion and promote the basicity of the catalysts, thereby achieving a quite increased low-temperature activity. They reported that their best performing catalyst 3% Ru, 30% Ni/Ce 0.9 Zr 0.1 O 2 could achieve 98.2% CO 2 conversion and 100% CH 4 selectivity under the reaction conditions tested.
Renda et al. [33] performed an extensive study on the effect of noble metal loading on the catalytic activity of 10% Ni catalysts supported on CeO 2 . At first, CeO 2 was identified as the most suitable catalyst support over CeZrO 4 and CeO 2 /SiO 2 , which is in agreement with other studies [12]. In the case of Ru, a volcano-shaped trend could be observed between the Ru loading and the yield for methane at low-temperatures, with 1% Ru leading to the best performance, as shown in Figure 10. The yield for CH 4 at 300 o C was almost double for the 1% Ru promoted 10% Ni/CeO 2 catalyst since Ru offered additional active sites for CO 2 methanation in synergy with Ni. The worse performance at higher Ru loadings was attributed to the worsening of active metal dispersion. Renda et al. [108] also showed that using ruthenium acetylacetonate instead of ruthenium chloride as the Ru precursor salt could improve the dispersion of the active metals, due to the templating effect of the precursor salt, leading to an enhanced methanation performance. In another interesting study, le Saché et al. [70] reported that Ru addition on a Ni/CeO2-ZrO2 catalyst could improve the dispersion of Ni and increase the overall intrinsic activity for the reduction of CO2. The bimetallic NiRu catalyst exhibited enhanced CO2 methanation performance in terms of CO2 conversion and CH4 selectivity at lower temperatures compared to the monometallic Ni catalyst. Furthermore, the NiRu-based catalyst was also active for CO production via the reverse water-gas shift (RWGS) reaction upon temperature increase. It was shown that the catalyst could be stable and high-per- Figure 10. CO 2 conversion and CH 4 yield trends for the samples doped with 0.5%, 1% and 3% ruthenium (H 2 :Ar:CO 2 = 4:5:1; WHSV = 60 l/(h·g cat )). Reproduced with permission from [33]. Copyright: Elsevier, Amsterdam, The Netherlands, 2020.
In another interesting study, le Saché et al. [70] reported that Ru addition on a Ni/CeO 2 -ZrO 2 catalyst could improve the dispersion of Ni and increase the overall intrinsic activity for the reduction of CO 2 . The bimetallic NiRu catalyst exhibited enhanced CO 2 methanation performance in terms of CO 2 conversion and CH 4 selectivity at lower temperatures compared to the monometallic Ni catalyst. Furthermore, the NiRu-based catalyst was also active for CO production via the reverse water-gas shift (RWGS) reaction upon temperature increase. It was shown that the catalyst could be stable and high-performing over long time for both reactions (CO 2 methanation and RWGS) at the respective reaction temperatures (350 and 700 • C). Their experimental results are summarized in Figure 11. It should also be noted that Fe addition on the Ni/CeO 2 -ZrO 2 catalyst led to inferior CO 2 methanation performance, as discussed previously.

Promotion with Rh
Rhodium (Rh) is one of the most expensive noble metals and it is known to be highly active for CO2 methanation at low temperatures, even when used under very low loadings [19,113]. However, its presence in bimetallic Ni-Rh catalysts does not always lead to a desirable activity increase, especially when taking into consideration the increased cost of the catalytically active material. One of the earlier works that does report a promoting effect of Rh in Ni-based catalysts was conducted by Ocampo et al. [106]. The addition of just 0.5% Rh on a 5% Ni/CeO2-ZrO2 catalyst greatly increased the catalytic activity of the catalyst, and also by a larger degree compared to the modification with 0.5% Ru. Higher stability for the noble metal modified catalysts was also observed and the better performance was attributed to the enhancement of the dispersion of nickel nanoparticles. Swalus et al. [114] followed a different path by mechanically mixing 1% Rh/Al2O3 and 1% Ni/activated carbon catalysts. The combined catalyst presented increased methane production through a synergistic collaboration between Ni and Rh metallic sites. More specifically, Ni sites were active for H2 adsorption and Rh sites for CO2 adsorption, respec- Through a novel preparation method, Polanski et al. [109] prepared Ru nanoparticles supported on metallic Ni grains via first preparing Ru nanoparticles supported on an intermediate silica carrier. Ru and Ni were then deposited on the carrier and the silica was etched upon suspension in a strong basic solution to yield a 1.5% Ru/Ni catalyst with an oxide passivation layer. The prepared catalyst exhibited a greatly enhanced lowtemperature activity with about 100% CO 2 conversion being achieved at a temperature as low as 200 • C. Carbon deposition was shown to deactivate the catalyst over time, with the initial activity being regained upon treatment under H 2 atmosphere. In a followup work, Siudyga et al. [110] synthesized 1.5% Ru catalysts supported on Ni nanowires via a similar preparation method. The higher surface area of the 1D nanostructure of the Ni support provided an additional advantage, with around 100% CO 2 conversion being reached at 179 • C and the onset temperature for the reaction being observed at just 130 • C. Siudyga et al. [111] also observed that Ru/Ni catalysts are highly active for CO methanation from syngas, with 100% CO conversion being reached at 178 • C and observable CO conversion being detected even at −7 • C.
The choice of Ru as a promoter for the CO 2 methanation reaction in Ni-based catalysts is a sensible option, since the combination of these two metals appears to yield favourable results, as reported by numerous works [30,33,70]. The synergistic effect through the presence of Ru and Ni adjacent phases can be observed via the increased metal dispersion, the higher reducibility, the improved low-temperature activity and the superior stability. Ni and Ru are both very active CO 2 methanation sites on their own and their combination can lead to an enhancement of the intrinsic catalytic activity, and to a decrease in the activation energy, through the favourable activation of H 2 and CO 2 reactant molecules [33,94]. The only drawback for the use of Ru is its higher price compared to transition metals like Ni, Fe and Co, even though it is significantly cheaper when compared to other noble metals such as Rh or Pt. On rare occasions, Ru addition does lead to a drop in CO 2 methanation activity, like in the work of Wei et al. [112] regarding Ni catalysts supported on zeolites. Some of the most representative studies that include Ni-M (M = noble metal) bimetallic catalysts for CO 2 methanation are comparatively presented in Table 2.

Promotion with Rh
Rhodium (Rh) is one of the most expensive noble metals and it is known to be highly active for CO 2 methanation at low temperatures, even when used under very low loadings [19,113]. However, its presence in bimetallic Ni-Rh catalysts does not always lead to a desirable activity increase, especially when taking into consideration the increased cost of the catalytically active material. One of the earlier works that does report a promoting effect of Rh in Ni-based catalysts was conducted by Ocampo et al. [106]. The addition of just 0.5% Rh on a 5% Ni/CeO 2 -ZrO 2 catalyst greatly increased the catalytic activity of the catalyst, and also by a larger degree compared to the modification with 0.5% Ru. Higher stability for the noble metal modified catalysts was also observed and the better performance was attributed to the enhancement of the dispersion of nickel nanoparticles. Swalus et al. [114] followed a different path by mechanically mixing 1% Rh/Al 2 O 3 and 1% Ni/activated carbon catalysts. The combined catalyst presented increased methane production through a synergistic collaboration between Ni and Rh metallic sites. More specifically, Ni sites were active for H 2 adsorption and Rh sites for CO 2 adsorption, respectively. Active hydrogen species could then spill over to Rh sites to facilitate the hydrogenation of chemisorbed CO 2 molecules.
Via a complex material architecture, Arandiyan et al. [115] prepared highly active NiRh CO 2 methanation catalysts. They first synthesized a 3-dimensionally ordered macroporous (3DOM) perovskite with nominal stoichiometry LaAl 0.92 Ni 0.08 O 3 via a poly(methyl methacrylate) (PMMA) colloidal crystal-templating preparation method. Rh was then introduced via wet impregnation of a Rh precursor salt. Upon a high-temperature reduction treatment, exsolution of Ni nanoparticles occurred, followed by NiRh alloy formation with the surface Rh atoms. The Ni exsolution procedure facilitated a high dispersion of Ni on the entire perovskite surface and a strong metal-support interaction that could otherwise not be achieved by traditional impregnation techniques [116]. The Rh-Ni/3DOM LAO catalyst with highly dispersed NiRh alloy nanoparticles on a macroporous perovskite support exhibited a significantly enhanced CO 2 methanation activity, when compared to similar monometallic Ni and Rh catalysts [115].
The architecture of bimetallic NiRh catalysts supported on Al 2 O 3 was also modified by Wang et al. [117]. On the one hand, the method of galvanic replacement (GR) led to the formation of a Ni@Rh core-shell catalyst, where Ni nanoparticles were encapsulated by an atomically thin RhO x shell. On the other hand, the method of chemical reduction (CR) led to a higher degree of intermixing between Ni and Rh. The catalyst prepared by galvanic replacement presented a superior CO 2 methanation performance, leading to the conclusion that Rh atoms highly dispersed as a thin shell over Ni nanoparticles were more active compared to large Rh nanoparticles or NiRh alloys ( Figure 12). Dissociative adsorption of CO 2 over Rh was observed as the key intermediate step. Reducing the loading of Rh could also further enhance the specific activity per noble metal atom. Thus, the galvanic replacement method, which rests on the reduction potential difference between the various redox metal pairs, was shown to be an effective strategy for the preparation of bimetallic catalysts [118].
The architecture of bimetallic NiRh catalysts supported on Al2O3 was also modified by Wang et al. [117]. On the one hand, the method of galvanic replacement (GR) led to the formation of a Ni@Rh core-shell catalyst, where Ni nanoparticles were encapsulated by an atomically thin RhOx shell. On the other hand, the method of chemical reduction (CR) led to a higher degree of intermixing between Ni and Rh. The catalyst prepared by galvanic replacement presented a superior CO2 methanation performance, leading to the conclusion that Rh atoms highly dispersed as a thin shell over Ni nanoparticles were more active compared to large Rh nanoparticles or NiRh alloys ( Figure 12). Dissociative adsorption of CO2 over Rh was observed as the key intermediate step. Reducing the loading of Rh could also further enhance the specific activity per noble metal atom. Thus, the galvanic replacement method, which rests on the reduction potential difference between the various redox metal pairs, was shown to be an effective strategy for the preparation of bimetallic catalysts [118]. Contrasting the results discussed until now [115,117], there are a number of works that observe a deactivating effect of Rh presence in Ni-based catalysts. Mutz et al. [46] reported that a NiRh0.1 catalyst supported on Al2O3 showed inferior CO2 conversion to a Contrasting the results discussed until now [115,117], there are a number of works that observe a deactivating effect of Rh presence in Ni-based catalysts. Mutz et al. [46] reported that a NiRh 0.1 catalyst supported on Al 2 O 3 showed inferior CO 2 conversion to a similar monometallic Ni catalyst over the entire temperature range tested. Moreover, Mihet et al. [119] and Renda et al. [33] both showed that 0.5% Rh addition on a 10% Ni catalyst supported on Al 2 O 3 and CeO 2 , respectively, led to a drop in the CO 2 conversion and the overall CH 4 yield. A NiRh combined catalyst supported on Al 2 O 3 and prepared Heyl et al. [120] was also less active for CO 2 methanation compared to a monometallic Rh catalyst.
Therefore, Rh insertion in Ni-based catalysts does not always lead to an activity enhancement and it appears that certain requirements need to be followed, such as the existence of highly dispersed NiRh alloys with strong metal-support interaction, or the fine dispersion of Rh on top of Ni particles [115,117]. In any case, the very high cost of Rh renders it unattractive for use as promoter in Ni-based catalysts, especially when comparable activity promotion can be achieved via the use of cheaper metals, like Ru.

Promotion with Pt
Monometallic Pt catalysts used in the reduction of CO 2 favour the formation of CO rather than CH 4 and are attractive catalysts for the reverse water-gas shift reaction (RWGS) [121]. However, NiPt alloys can have a different performance compared to monometallic Ni and Pt. Mihet et al. [119] observed a promoting effect for CO 2 methanation by 0.5% Pt addition on 10% Ni/Al 2 O 3 , through the enhancement of Ni dispersion and NiO reducibility. Renda et al. [33] observed a similar behaviour when 0.5% Pt was added on a 10% Ni/CeO 2 catalyst. Interestingly, as shown in Figure 13, the increase in Pt loading above 0.5% decreased the quantity of produced CH 4 . This behaviour was attributed to the fact that Pt sites could easily dissociate CO 2 into CO, but the further conversion of intermediate CO species took place at the nearby Ni sites. When Pt was finely dispersed, CH 4 production was high, whereas higher Pt loadings could lead to increased CO production that blocked the active Ni methanation sites. rable activity promotion can be achieved via the use of cheaper metals, like Ru.

Promotion with Pt
Monometallic Pt catalysts used in the reduction of CO2 favour the formation of CO rather than CH4 and are attractive catalysts for the reverse water-gas shift reaction (RWGS) [121]. However, NiPt alloys can have a different performance compared to monometallic Ni and Pt. Mihet et al. [119] observed a promoting effect for CO2 methanation by 0.5% Pt addition on 10% Ni/Al2O3, through the enhancement of Ni dispersion and NiO reducibility. Renda et al. [33] observed a similar behaviour when 0.5% Pt was added on a 10% Ni/CeO2 catalyst. Interestingly, as shown in Figure 13, the increase in Pt loading above 0.5% decreased the quantity of produced CH4. This behaviour was attributed to the fact that Pt sites could easily dissociate CO2 into CO, but the further conversion of intermediate CO species took place at the nearby Ni sites. When Pt was finely dispersed, CH4 production was high, whereas higher Pt loadings could lead to increased CO production that blocked the active Ni methanation sites. Figure 13. CO2 conversion and CH4 yield trends for the samples doped with 0.5%, 1% and 3% of platinum (H2:Ar:CO2 = 4:5:1; WHSV = 60 L/(h·gcat)). Reproduced with permission from [33]. Copyright: Elsevier, Amsterdam, The Netherlands, 2020.
Kikkawa et al. [31] elucidated the promoting role of isolated Pt atoms that are finely dispersed in Ni-rich NiPt alloy nanoparticles supported on Al2O3 for the CO2 methanation reaction. Through Ni and Pt precursor salt impregnation and reduction, Pt atoms could dissolve into the lattice of Ni nanoparticles and were slightly negatively charged as Pt δ-, surrounded by Ni δ+ atoms. It was shown that isolated Pt δ-enhanced the adsorption of intermediate CO species, while also weakening the C-O bond energy. These chemisorbed CO species on Pt δ-could then be easily hydrogenated into CH4. As a result, Ni95Pt5 supported on Al2O3 with 5% Pt atoms in the NiPt alloy led to a higher CH4 formation rate and CH4 yield compared to the monometallic Ni catalyst and other NiPt alloys with a higher Figure 13. CO 2 conversion and CH 4 yield trends for the samples doped with 0.5%, 1% and 3% of platinum (H 2 :Ar:CO 2 = 4:5:1; WHSV = 60 L/(h·g cat )). Reproduced with permission from [33]. Copyright: Elsevier, Amsterdam, The Netherlands, 2020.
Kikkawa et al. [31] elucidated the promoting role of isolated Pt atoms that are finely dispersed in Ni-rich NiPt alloy nanoparticles supported on Al 2 O 3 for the CO 2 methanation reaction. Through Ni and Pt precursor salt impregnation and reduction, Pt atoms could dissolve into the lattice of Ni nanoparticles and were slightly negatively charged as Pt δ-, surrounded by Ni δ+ atoms. It was shown that isolated Pt δenhanced the adsorption of intermediate CO species, while also weakening the C-O bond energy. These chemisorbed CO species on Pt δcould then be easily hydrogenated into CH 4 . As a result, Ni 95 Pt 5 supported on Al 2 O 3 with 5% Pt atoms in the NiPt alloy led to a higher CH 4 formation rate and CH 4 yield compared to the monometallic Ni catalyst and other NiPt alloys with a higher Pt content. Overall, the promoting role of isolated Pt atoms on Ni 95 Pt 5 /Al 2 O 3 rested on them acting both as CO adsorption sites, as well as H 2 dissociation sites.
In a later work [122], it was shown, based on in-situ FTIR observations, that, on such NiPt alloy catalysts (iso-Pt), the adsorbed intermediate CO species that form CH 4 were rather bridging CO between isolated Pt atoms and nearby Ni atoms. On the monometallic Pt catalyst, strongly anchored CO species were mainly found as on-top CO (terminal mode, Pt-C≡O), which were difficult to be hydrogenated into CH 4 . These on-top CO species were then preferentially desorbed as CO rather than being converted into CH 4 and, hence, the monometallic Pt catalyst exhibited a high CO selectivity and reverse water-gas shift (RWGS) reactivity. However, the initially formed on-top CO species over isolated Pt atoms in NiPt alloys were quickly transformed into bridging CO between these isolated Pt atoms and their neighbouring Ni atoms (µ 2 -bridging mode, Pt-(C=O)-Ni). These bridging intermediate CO species were also similar to those found in the monometallic Ni catalyst between Ni-Ni neighbours and were easily further converted into CH 4 . Therefore, it was concluded that CH 4 (and CO) selectivity could be determined by the type of adsorbed intermediate CO species. High CH 4 selectivity for the monometallic Ni catalyst and the NiPt alloy catalyst (iso-Pt) that contained isolated Pt atoms was due to the occurrence of bridging CO, while CO selectivity was drastically increased in the monometallic Pt catalyst where CO was adsorbed as on-top CO.
Furthermore, the rate determining step (RDS) for CO 2 methanation over such NiPt alloy catalysts with isolated Pt atoms was suggested to be the chemisorption of H 2 , based on kinetic studies [122]. That could be explained by the fact that although Pt sites favoured the dissociation of H 2 at low temperatures, isolated surface Pt atoms were preferentially covered by CO rather than H 2 , due to the enhanced Pt-CO bonding. As a result, this limited availability of H 2 dissociation sites on isolated Pt atoms became the factor that restricted the formation of CH 4 over NiPt alloys. Therefore, the RDS for CO 2 methanation over NiPt alloys was suggested to be H 2 dissociation rather than H-assisted CO dissociation or hydrogenation of surface carbon species, which are commonly indicated as the RDS over monometallic Ni methanation catalysts [7,122]. The experimental results and reaction mechanism for these two works are summarized in Figure 14 [31,122].

Promotion with Pd
Pd is another noble metal with a high number of applications in heterogeneous catalysis [123,124]. Stand-alone Pd catalysts are not commonly used for CO2 methanation, but NiPd combinations can sometimes have a superior activity for this reaction. An example of such promotion is given by the work of Mihet et al. [119]. The addition of 0.5% Pd on 10% Ni/Al2O3 catalysts remarkably increased the low-temperature activity and the NiPd bimetallic catalyst was quite stable, outperforming other bimetallic Ni-based catalysts promoted with Pt and Rh. The addition of Pd, as well as that of other noble metals, increased the reducibility of NiO and favoured higher metal dispersion and H2 chemisorption. Arellano-Treviño et al. [104] also showed that Pd addition could increase NiO reducibility in DFMs, though its effectiveness for CH4 production is not so high compared to Ru and Pt addition.
Li et al. [32] synthesized an effective CO2 methanation catalyst composed of NiPd alloys supported on an SBA-15 mesoporous siliceous support. There was a strong synergetic effect between the two metals, as evidenced by H2-TPR, and this synergy led to the bimetallic NiPd catalysts exhibiting enhanced CO2 conversion compared to the monometallic Pd and Ni catalysts. The content of Ni and Pd in the alloy was also varied and it was shown that the Ni0.75Pd0.25 alloy with a Ni/Pd molar ratio of three led to the best results.
Furthermore, Yan et al. [125] prepared a composite hierarchically structured bimetallic nanocatalyst that consisted of metallic Pd nanoclusters adjacent to NiO with local As discussed previously with regard to Ru doping, adding Pt in Ni-based DFMs could also increase the low-temperature reducibility of NiO. The addition of 1% Pt increased the NiO reducibility by 50%, compared to the 70% increase caused by the addition of 1% Ru [104]. Employing in-situ DRIFTS, it was shown that Pt acted to further promote CO 2 chemisorption and dissociation by forming Pt-CO species [105]. However, since Pt is not active for CH 4 formation, these chemisorbed carbonyl intermediates were found to spill over towards Ni sites upon H 2 inflow, where they were further hydrogenated into CH 4 . Overall, the NiRu bimetallic DFMs yielded more CH 4 compared to the NiPt ones, since Ru sites could also participate in the CO 2 methanation reaction [104].
In short, despite monometallic Pt catalysts being very active for the RWGS reaction, Ni-rich NiPt alloys appear to promote CH 4 formation [31,33]. The dilution of Pt atoms over Ni metallic nanoparticles changes the CO 2 methanation pathway, since Pt sites accommodate adsorbed CO intermediates due to their high affinity with carbonyl [31]. In general, it seems that a very low amount of Pt added in Ni catalysts, even compared to other noble metals, can lead to a significant promotion for CO 2 methanation [33]. It should, however, be argued whether the extent of such promotion can be achieved via less expensive metals, like Ru.

Promotion with Pd
Pd is another noble metal with a high number of applications in heterogeneous catalysis [123,124]. Stand-alone Pd catalysts are not commonly used for CO 2 methanation, but NiPd combinations can sometimes have a superior activity for this reaction. An example of such promotion is given by the work of Mihet et al. [119]. The addition of 0.5% Pd on 10% Ni/Al 2 O 3 catalysts remarkably increased the low-temperature activity and the NiPd bimetallic catalyst was quite stable, outperforming other bimetallic Ni-based catalysts promoted with Pt and Rh. The addition of Pd, as well as that of other noble metals, increased the reducibility of NiO and favoured higher metal dispersion and H 2 chemisorption. Arellano-Treviño et al. [104] also showed that Pd addition could increase NiO reducibility in DFMs, though its effectiveness for CH 4 production is not so high compared to Ru and Pt addition.
Li et al. [32] synthesized an effective CO 2 methanation catalyst composed of NiPd alloys supported on an SBA-15 mesoporous siliceous support. There was a strong synergetic effect between the two metals, as evidenced by H 2 -TPR, and this synergy led to the bimetallic NiPd catalysts exhibiting enhanced CO 2 conversion compared to the monometallic Pd and Ni catalysts. The content of Ni and Pd in the alloy was also varied and it was shown that the Ni 0.75 Pd 0.25 alloy with a Ni/Pd molar ratio of three led to the best results.
Furthermore, Yan et al. [125] prepared a composite hierarchically structured bimetallic nanocatalyst that consisted of metallic Pd nanoclusters adjacent to NiO with local tetrahedral symmetry. The support consisted of acid-treated carbon nanotubes (CNTs) and the catalysts were further decorated with a thin tetramethyl orthosilicate (TMOS) layer. This composite NiPd bimetallic catalyst produced a maximum amount of CH 4 and CO at 300 • C, compared to the monometallic Pd catalyst and other active catalysts with similar metal loadings that are reported in the literature. As shown in Figure 15, the higher performance of NiO T Pd-T catalyst was attributed to the local synergy at the interface between the Ni phase and the adjacent Pd phase, with H and CO being, respectively, chemisorbed over Ni and Pd interfacial sites. Adsorbed H atoms then helped reduce nearby NiO sites and thus increase the number of metallic Ni sites that are available for the methanation reaction.

Promotion with Re
Rhenium (Re) is not a very common element in heterogeneous catalysis, and it is only sometimes considered to be a noble metal. An example of its application in heterogeneous catalysis is the deoxydehydration (DODH) of biomass-derived polyols to produce adipic acid over ReOx supported on ZrO2 [126]. There has, however, recently been a handful of reports where Re is studied as a promoter in Ni-based catalysts for methanation.

Promotion with Re
Rhenium (Re) is not a very common element in heterogeneous catalysis, and it is only sometimes considered to be a noble metal. An example of its application in heterogeneous catalysis is the deoxydehydration (DODH) of biomass-derived polyols to produce adipic acid over ReO x supported on ZrO 2 [126]. There has, however, recently been a handful of reports where Re is studied as a promoter in Ni-based catalysts for methanation.
Initially, Han et al. [127] screened 63 different elements as potential promoters in Ni/Al 2 O 3 catalysts for the methanation of CO, with the help of data mining technology. Their model predicted that promoting a Ni/Al 2 O 3 catalyst with Re could potentially lead to the largest drop in the temperature for 50% conversion (T 50 ) by 78 • C, compared to the unpromoted Ni/Al 2 O 3 . In another computational study, Yuan et al. [128] showed that the high oxygen affinity of Re can render it a favourable element to facilitate C-O bond scission during CO 2 methanation. Their calculations indicated that isolated Re atoms on top of Ni(111) facets can accommodate O adatoms cleaved from COOH*/HCOO* and CHO*/CO* intermediates, whereas CO* and H* can be found adsorbed on the Ni (111) surface. The synergistic collaboration between the Ni surface and Re dopant atoms could promote the scission of C-O bonds, increasing CH 4 selectivity and favouring the overall catalytic activity for CO 2 methanation.
Dong et al. [129] experimentally tested the promoting effect of Re on Ni catalysts supported on H 2 O 2 -modified manganese sand for CO methanation coupled with watergas shift. Re was reported to decrease the activation energy for CO methanation and lead to a faster CH 4 formation rate. Subsequently, Dong et al. [130] studied Re-promoted Ni catalysts supported on coal combustion fly ash (CCFA) derived from industrial wastes for the CO 2 methanation reaction. The bimetallic Ni-Re/CCFA catalyst exhibited higher CO 2 conversion and CH 4 selectivity (99.5% and 70.3%, respectively), compared to the monometallic Ni/CCFA catalyst ( Figure 16). The promoting effect of Re was attributed to the increase in Ni dispersion, as well as to the resistance of the bimetallic catalyst towards sintering and coke formation. Bicarbonate, bidentate formate and methoxyl species were detected as the reaction intermediates via in-situ DRIFTS. Table 2, below, summarises typical bimetallic Ni-based catalysts promoted with noble metals (Ru, Rh, Pt, Pd and Re) for the methanation of CO 2 . The introduction of 1% Ru in 15% Ni/CeO 2 -ZrO 2 improved the dispersion of Ni and the intrinsic activity for CO 2 reduction. The catalyst was active for both CO 2 methanation at 350 • C and reverse water-gas shift (RWGS) at 700 • C.  Pd and Pt addition improved the catalytic activity of 10% Ni/Al 2 O 3 , while Rh addition led to worse performance. Pd and Pt increased the dispersion and reducibility of NiO and provided active sites for H 2 chemisorption and activation.
The Pd-promoted catalyst was slightly better than the Pt-promoted one. The synergy between Ni and Pd metals led to active NiPd alloy catalysts for CO 2 methanation supported on mesoporous SBA-15 support. All bimetallic catalysts were better than the monometallic ones. A Ni/(Ni + Pd) ratio of 3 (Ni 0.75 Pd 0.25 ) led to the best performing catalyst. [32] Pd 30% NiO T Pd-TMOS/acid-treated CNTs (Pd/Ni molar ratio = 1.5) Ni wet impregnation on acid-treated CNTs, Pd addition with NaBH 4 and TMOS decoration WHSV = 100,000 mL g −1 h −1 H 2 /CO 2 = 3/1 Methane yield: Y CH4 = 1905.1 µmol CH 4 g cat −1 , at 300 • C (NiO T Pd-TMOS/acid-treated CNTs) The catalyst consisted of Pd nanoclusters adjacent to NiO with tetrahedral symmetry, supported on acid-treated CNTs and decorated with a layer of tetramethyl orthosilicate (TMOS).
Re addition improved Ni dispersion and the catalyst resistance towards sintering and coking, leading to better performance for CO 2 methanation.
conversion and CH4 selectivity (99.5% and 70.3%, respectively), compared to the monometallic Ni/CCFA catalyst ( Figure 16). The promoting effect of Re was attributed to the increase in Ni dispersion, as well as to the resistance of the bimetallic catalyst towards sintering and coke formation. Bicarbonate, bidentate formate and methoxyl species were detected as the reaction intermediates via in-situ DRIFTS. Table 2, below, summarises typical bimetallic Ni-based catalysts promoted with noble metals (Ru, Rh, Pt, Pd and Re) for the methanation of CO2.

Conclusions
The race for the development of low-cost and high-performing CO 2 methanation catalysts stems from the need to efficiently convert excess electricity and H 2 generated from renewables, as well as CO 2 captured from flue gases, into a reliable energy carrier. Ni is the standard option to be used in CO 2 methanation catalysts, due to its high activity and low cost. However, insufficient low-temperature activity and the degradation of Ni catalysts over time due to oxidation and sintering creates the need for the employment of specific metal additives to counter such drawbacks. These additives can fall in two generalized categories: other transition metals (including Fe and Co) and noble metals (including Ru, Rh, Pt, Pd and Re).
The transition metals Fe and Co offer the obvious advantage of being cheap like Ni and their similar size and electronic properties allow for their intricate interaction with the Ni primary phase and their easy dissolution into the Ni lattice, forming NiFe and NiCo alloys, respectively. The composition of the formed alloy is of great importance, since only specific bimetallic combinations can lead to an optimal CO 2 methanation performance, especially in the case of NiFe alloys. The combined bimetallic catalysts can also offer additional advantages, such as higher stability, as well as resistance towards oxidation and sulphur poisoning.
Noble metals generally increase the reducibility and dispersion of the Ni primary phase and they can also participate in the reaction as active CO 2 methanation phases.
Stand-alone Ru catalysts are highly active for low-temperature CO 2 methanation and the presence of Ru in bimetallic Ni catalysts as a separate monometallic phase also boosts catalytic activity. Additionally, the cost-effectiveness of Ru compared to other noble metals renders the bimetallic NiRu combinations quite popular in the field of heterogeneous catalysis. Rh and Pt can also greatly enhance the catalytic activity for CO 2 methanation when dissolved or deposited upon Ni in small quantities. Lastly, Pd and Re have been also tested as potential promoters in Ni-based catalysts.
The assumed trade-off between cost and catalytic activity for CO 2 methanation catalysts can be potentially overcome via the development of bimetallic Ni-containing catalysts with an optimised Ni-dopant metal synergy. Recent advances in operando spectroscopic techniques can shed light on how the reaction mechanism differs between Ni-based alloys or Ni-dopant metal interfaces and monometallic Ni, allowing for the development of catalysts with the lowest possible cost and highest possible performance. Funding: This research has been co-financed by the European Union and Greek national funds through the operational program "Regional Excellence" and the operational program Competitiveness, Entrepreneurship and Innovation, under the call Research-Create-Innovate (Project code: T1EDK-00782).

Informed Consent Statement: Not applicable.
Data Availability Statement: Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Conflicts of Interest:
The authors declare no conflict of interest.