A review on bi/polymetallic catalysts for steam methane reforming

(cid:1) Recent research on bi/polymetallic catalysts for (SE)SMR is reviewed. (cid:1) The catalysts are evaluated based on activity, stability, and physical properties. (cid:1) Metal loading and material structure greatly inﬂuence catalytic performance. (cid:1) Future research includes evaluation of sulphur-and sintering-resistant abilities.


a b s t r a c t
Blue hydrogen production by steam methane reforming (SMR) with carbon capture is by far the most commercialised production method, and with the addition of a simultaneous in-situ CO 2 adsorption process, sorption-enhanced steam methane reforming (SESMR) can further decrease the cost of H 2 production.Ni-based catalysts have been extensively used for SMR because of their excellent activity and relatively low price, but carbon deposition, sulphation, and sintering can lead to catalyst deactivation.One effective solution is to introduce additional metal element(s) to improve the overall performance.
This review summarizes recent developments on bi/polymetallic catalysts for SMR, including promoted nickel-based catalysts and other transition metal-based bi/polymetallic materials.The review mainly focuses on experimental studies, but also includes results from simulations to evaluate the synergistic effects of selected metals from an Abbreviations: at.%,Atomic percentage; CCUS, Carbon capture utilization and storage; CL-SMR, Chemical looping steam methane reforming; DFT, Density Functional Theory; SESMR, Sorption-enhanced steam methane reforming; SMR, Steam methane reforming; WGS, Water-gas shift reaction; wt.%, Weight percentage; YSZ, Yttria-stabilized zirconia.

Introduction
In recent years, hydrogen has gained increasing attention as an alternative to fossil fuels, enabling net zero targets to be realized.As stated in the 2021 UK Hydrogen Strategy [1], the UK is aiming for a total of 10 GW of low-carbon hydrogen production capacity by 2030 to decarbonize vital industries and provide clean energy across the heat, power, and transport sectors.This requires considerable effort in scaling up and optimizing carbon capture and storage systems as well as new hydrogen production processes, such as sorptionenhanced reforming.
Hydrogen can be produced from a variety of renewable and non-renewable sources, and can be divided into three categories depending on its production pathway.
Black/Grey/Brown hydrogen: from fossil fuel-based production (coal, natural gas, and lignite, respectively) with CO 2 released to the atmosphere.Blue hydrogen: from fossil fuel-based production with carbon capture, utilization, and storage (CCUS).Green hydrogen: from renewable sources, commonly electrolysis-based production.
Different hydrogen production pathways using renewable sources have been investigated, including water electrolysis [2], thermochemical (pyrolysis and gasification), or biological conversion (fermentation and photolysis) of biomass [3].However, various techno-economic studies have demonstrated that compared to these processes, fossil fuel reforming with CCUS remains the most cost-competitive option with the highest hydrogen production efficiency [4e6].Steam reforming, partial oxidation, and autothermal reforming are three main fossil fuel reforming technologies for hydrogen production, among which the steam reforming of methane (SMR) is by far the most deployed method.Although SMR is a mature technology, one of the most significant problems is its high CO 2 emission.It is estimated that without CCUS, hydrogen from SMR has an emission factor of 222e325 gCO 2eq per kWh of H 2 (10 tCO 2 /tH 2 ) [4,7].
Apart from the traditional approach of employing a downstream amine scrubbing process, an alternative option to mitigate carbon emissions is adding a simultaneous carbon capture step to the conventional SMR process.A novel hydrogen production technology, known as sorptionenhanced steam methane reforming (SESMR), combines the conventional SMR process with a simultaneous in-situ absorption of CO 2 using a solid sorbent (usually CaO).The main reactions involved in the SESMR process are as follows [8].
Steam reforming of methane Water-gas shift (WGS) reaction CO 2 sorption and sorbent regeneration Overall equation for SESMR In comparison to the traditional SMR process, SESMR enables the removal of CO 2 from the reaction zone, which shifts the equilibrium towards the product side, enhancing the production of hydrogen.The high-purity CO 2 stream released from the sorbent regeneration step can also be easily separated from the sorbent, and transported or stored for further use.The sorption enhanced steam reforming process has been applied to other feedstocks as well, including phenol [9], glycerol [10], ethanol [11], and biomass [12].In general, the CO 2 sorbent is combined with active catalytic metal(s) to form a bifunctional material, using alumina, perovskite or mayenite as the structural support.Since both SMR and SESMR require the use of catalysts to proceed, and the types of catalysts used for both processes are in general identical, they will be reviewed holistically in this paper.
The main reaction steps of SMR are listed in Table 1, including the dissociative adsorption of the reactants, dehydrogenation and bond reformation steps.It is generally agreed that the activation of the first CeH bond of the CH 4 decomposition step (step 1) is the rate-determining step of SMR [13e15]; but at lower temperatures (T < 500 C), the CO formation step (step 7) becomes dominant [14].The energy barrier for CeH bond activation over Ni surface is relatively low, and at the same time, the adsorption of C*, H* and O* is not so strong that the species cannot react off the surface easily.Nickel is also widely employed commercially due to its low price and high availability.However, Ni-based catalysts are prone to sintering and coke formation [16,17], hydrogen reduction of nickel-based catalysts before the reforming process is also necessary for activating the material.It is, therefore, of interest to investigate the anti-sintering ability, coke resistance, as well as the reducibility and self-activation ability of SMR catalysts.
Apart from nickel, noble metals (Rh, Ru, Pd, Pt, and Ir) are also promising candidates for SMR because of their excellent catalytic ability and resistance to carbon formation.Currently, there is no definitive conclusion as to how the noble metals are ordered regarding their catalytic activity for SMR, however, several experimental [18,19] and numerical [20,21] studies reported that the catalytic activity of noble metals follows the order of Rh ~Ru > Ir > Pt ~Pd.Despite their advantages, noble metal-based catalysts are limited by their high prices.
One way to maintain the excellent performances of noble metals while maintaining a reasonable price is by combining two or more types of metals, using cheap transition metals (usually nickel or cobalt) as the base and noble metals as promoters.Bi/polymetallic catalysts have gained increasing attention in recent years, and the synergistic effect between commonly used metal elements has been investigated experimentally and numerically.Numerical studies focus on the reaction pathway, activation energies of certain reaction steps (in particular the CeH bond cleavage of the CH 4 decomposition step), as well as the adsorption energies of atomic or molecular species on the catalyst's surface, which are indicators of the material's catalytic activity and stability.Some materials were also tested experimentally, usually in lab-scale reactors, and evaluated based on their methane conversion ability, hydrogen yield, etc.To the best of the author's knowledge, there has been no literature that systematically summarizes the bi/polymetallic catalysts that have been employed in (SE)SMR.The aim of this review is therefore to provide an overview of the bi/polymetallic catalysts that have been tested for (SE)SMR, to summarize their advantages and limitations, and to identify the gap in current (SE)SMR catalyst development for future studies.

Nickel-based catalysts promoted by noble metals
Amongst the eight noble metals, platinum group metals (including Rh, Ru, Ir, Pt, and Pd) have the highest SMR activity and are most commonly used as a promoter for Ni-based catalysts.Their catalytic performance arises from the partially filled d-subshells e electrons can be easily added to or removed from these orbitals, resulting in an optimal interaction between the metal surface and the gas-phase adsorbate.A lower degree of d-band filling leads to the too-strong adsorption of both reactants and reaction products on the metal surface, which easily blocks the active catalytic surface.On the other hand, with a higher degree of filling, the metal surface does not interact strongly enough with the reactants, which is the case for both Ag and Au.This results in a relatively low catalytic activity but a more stable and clean metal surface, which is why Ag and Au are usually added for their coke-resistant property.There is currently no literature available regarding the use of Os as an SMR catalyst, possibly due to its tendency to form a volatile and toxic oxide e OsO 4 [18].
RueNi catalysts have been tested under lab-scale experimental SMR conditions by Jeong et al. [22] to study the effect of doping Ru over Ni/Al 2 O 3 and Ni/MgAl 2 O 4 .They concluded that adding a small amount of Ru (0.5 wt.%) greatly suppressed coke formation on the catalyst surface and facilitated NiO reduction.The coke-resisting ability of Ru was studied on an atomic scale using Density Functional Theory (DFT) study [23].It was demonstrated that when the noble metal was added, the activation energy of the CHO*-producing step (step 5 shown in Fig. 1) was significantly lower than that of the C* and H*-producing step (step 4), meaning CO production was favoured over carbon deposition.
Results also showed that Ru-promoted Ni catalysts were able to self-activate at a temperature of 700 C without any pre-reduction using hydrogen, which is beneficial from an economic and process operation point of view [24,25].Nibased catalysts doped with small (0.5 wt.%) or even trace amounts (0.05 wt.%) of Ru showed good self-activation properties without pre-reduction, and achieved a higher CH 4 conversion rate compared to monometallic Ru or Ni catalyst.Ru decreased the reduction temperature of Ni by inducing Table 1 e Main reaction pathway of SMR.

Reaction step
No.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 1 5 8 7 9 e1 5 8 9 3 hydrogen spill over on the Ni surface, a process in which H 2 molecules disassociate on the noble metal surface to H species and diffuse into Ni via the catalyst support.Similar to Ru, the addition of Rh was also found to facilitate Ni reduction and produce a synergistic effect [26,27].The bimetallic catalyst (with 0.2 wt.% Rh) showed higher activity compared to the linear combination of monometallic Rh or Ni catalyst.This was attributed to the enhanced textural properties of the bimetallic catalysts e the exposed metallic surface area and metal dispersion of the bimetallic catalysts, measured by CO chemisorption, were greatly promoted by the addition of Rh, and the increase was more significant with a higher Rh loading.The same synergistic effect was observed by Morales-Cano et al. as well [28], who studied the promoting effect of Ru, Rh, and Ir on Ni/a-Al 2 O 3 catalyst.The catalysts were also aged for 240 h at 800 C with an S/C ratio of 6 to induce the sintering of Ni particles.The activity of the aged NieRh and NieIr was found to be significantly higher than the aged monometallic catalysts, proving their ability to resist Ni sintering.This ability was attributed to the migration and diffusion of Ni into the Face Centred Cubic structure of Rh and Ir during the aging process, which enabled the formation of NieRh and NieIr alloys, and retained the high surface area of the materials.
The optimal loading of noble metals in Ni-based catalysts was also investigated.Katheria et al. [29] tested a series of Ni/ MgAl 2 O 4 catalysts with Rh concentration varying from 0.1 to 1 wt.%.Results showed that 0.1 wt.% of Rh was sufficient to increase CH 4 conversion by 20%, whereas a further increase in Rh concentration did not have a significant effect.A higher metal loading does not necessarily mean a better catalytic ability due to the less evenly distribution of active metal in the support.This observation was also verified by testing a series of Ni/MgAl 2 O 4 catalysts with Pt loading varying from 0.01 to 1 wt.% [30,31].Both studies reported that a Pt loading of 0.1 wt.% resulted in the highest catalytic activity.Further increase in Pt loading led to a decrease in both catalytic activity and stability.Results from the physical characterization of the materials showed that the highest surface area and maximum dispersion of active metal were achieved with 0.1 wt.% Pt loading, whereas higher Pt concentration, resulted in agglomeration on the material surface.
Chaichi et al. [32] synthesized a novel supportless NiePdcarbon nanotube material and compared its performance with Ni and NiePd catalysts under SMR conditions.The addition of Pd facilitated the reduction of metallic oxides, whereas both Pd and carbon nanotube increased the specific surface area.The resultant CH 4 conversion of the NiePd-carbon nanotube material was 22% higher than the monometallic Ni/MgO catalyst.Reducibility enhancement by Pd was also reported by Batebi et al. [33] in a test of NiePd/Al 2 O 3 for combined steam and CO 2 reforming of methane.By adding Pd, the reduction degree was increased from 69% to 83%, leading to higher CH 4 conversion and H 2 yield while reducing coke deposition.Bimetallic NiePd materials have also been tested for the oxidative SMR process [34e38].Results from these studies demonstrated that the addition of Pd promoted the reduction of Ni, PdeNi alloy was also found to form preferentially on the material surface, contributing to its high activity and coke resistance.
Li and Miyata conducted a series of tests to study the doping effect of Ru [39,40], Rh [41,42], Pt [42e44], and Pd [42] on Ni/Mg(Al)O catalysts in a daily start-up and shut-down operation of SMR under steam purging.As was presented previously, the addition of all four types of promoters improved the reducibility of the catalyst by decreasing Ni reduction temperature and increasing the amount of hydrogen uptake on Ni.Ru, Rh and Pt were also capable of suppressing the deactivation of the catalyst due to Ni oxidation, which was attributed to the self-regeneration of Ni 0 from Ni 2þ assisted by hydrogen spill over on the noble metal surface and the reversible reduction-oxidation between Ni 0 and Ni 2þ in the Mg(Ni, Al)O periclase.Self-activation without any reduction treatment of Rh-, Pt-, and PdeNi bimetallic catalysts was also observed during the daily start-up and shut-down operation.Compared to the complete deactivation of the pure Ni catalyst after the first steam purging, the CH 4 conversion of the bimetallic catalysts was kept at the value of thermodynamic equilibrium even after 4 cycles of steam purging.The self-activation and self-regeneration properties of these bimetallic materials proved Ru, Rh, Pt, and Pd to be useful additives to the conventional Ni-based catalysts.
Although the reactivity of monometallic Ag and Au is relatively low, they have also been tested as promoters for Nibased SMR catalysts due to their excellent stability.DFT-based studies [45,46] showed that the Ag-doped Ni surface is less prone to carbon deposition e the threefold AgeNieNi adsorption site is unstable for carbon atoms.Carbon atoms initially positioned at these sites will therefore move to the adjacent NieNieNi site, which has lower adsorption energy.These negative interactions between the carbon atom and the AgeNi alloy surface indicate that Ag can be added as a cokeresistant promoter, which was also validated against experimental findings [47e49].Both research teams studied the promoting effect of Ag (0.03e1 wt.%) on Ni/g-Al 2 O 3 and concluded that even the minimum Ag loading of 0.3 wt.% could reduce carbon deposition significantly.However, this also compromised the catalytic activity of the material, which decreased by 25% compared to monometallic Ni catalysts.This is due to the fact that Ag atoms are energetically favoured to replace Ni atoms on the step edges, which are the most active sites for methane decomposition, compared to the terrace sites [50].Similar properties were found in Au-doped Ni catalysts.Both computational [51e53] and experimental [54,55] studies suggested that the overall catalytic activity of AueNi was affected by the higher energy barrier for CeH bond cleavage in the rate-determining CH 4 dissociation step; whereas the stability of the material was enhanced due to the suppression of carbon formation.
Ag and Au have also been employed as promoters to Ni electrodes for solid oxide fuel cells under internal SMR conditions.Ag [56] or Au [57] with a loading of 1e5 wt.% doped on Ni/ yttria-stabilized zirconia (YSZ), as well as Au with a loading of 1e4 at.% doped on Ni/CeO 2 -Gd 2 O 3 [58] anode were tested at temperatures ranging from 650 to 800 C. Both exhibited satisfying performance with enhanced tolerance to carbon formation.However, the performance of Ag-doped materials is largely influenced by the reaction temperature.At temperatures higher than 750 C, the AgeNi/YSZ cell degraded rapidly due to the low melting point of Ag [56].At temperatures lower than 600 C, the catalytic activity of the AgeNi/Al 2 O 3 catalyst decreased significantly, whereas the Au-doped catalyst still exhibited higher activity than monometallic Ni/Al 2 O 3 [59,60].Sapountzi et al. [61] reported that an Au amount of 2.3 wt.% promoted the reducibility of Ni catalyst, and the NieAu alloy formed on the catalyst surface was able to inhibit the formation of sulphuric compounds, including nickel sulphide.
Liu et al. [62] reported that NieIr alloy supported on MgAl 2 O 4 was a durable catalyst for steam and CO 2 bi-reforming of methane under pressurized conditions.The bimetallic catalyst was composed of small metallic clusters (with a mean size of ~2 nm) and the cluster size was retained for a duration of 434 h, in contrast to the significant increase in cluster size of the monometallic Ni/MgAl 2 O 4 catalyst, showing the anti-sintering ability of Ir.The coke resistance of the bimetallic material was attributed to the combined effect of small ensemble sizes, increased surface oxophilicity, and higher activation barrier for CH 4 dissociation.The number of active sites was evaluated by H 2 chemisorption, it was found that an Ir loading of 0.1 wt.% was able to quadrupole the quantity of active sites of Ni/ MgAl 2 O 4 , and the promoting effect increased with a higher Ir loading.The bimetallic Ir 10 Ni 90 /MgAl 2 O 4 catalyst achieved CH 4 and CO 2 conversion of 95% and 98%, respectively, at 1 bar; and was able to maintain a relatively high CH 4 and CO 2 conversion of ~60% when pressurized to 20 bar.

Nickel-based catalysts promoted by non-noble metals
Despite multiple advantages, the use of noble metals is still constrained for economic reasons.Therefore, many researchers have turned to the application of non-noble metals as potential promoters of Ni-based catalysts.
NieFe-based catalysts/oxygen carriers have been tested for chemical looping steam methane reforming (CL-SMR).Hu et al. [63] used NieFe modified calcite as an oxygen carrier and concluded that a Fe/Ni ratio of 0.67 was optimal for the reaction, while higher Fe concentration led to sintering.The novel material exhibited good catalytic performance with the highest CH 4 conversion of 98.9%, and high stability during the long-term reaction process.Garai et al. [64] tested Ni-ferrite supported on ZrO 2 and CeO 2 , which showed high H 2 and CO selectivity, as well as high CH 4 conversion (93%, 98%, and 99%, respectively).No carbon deposition was observed on the used material, due to the ability of iron oxides (FeO x ) to remove carbon via a surface redox cycle to produce Fe and CO 2 [65].Djaidja et al. [66] reported that although the addition of Fe slightly decreased the CH 4 conversion of the (NieMg) 2 Al catalyst from 93% to 91%, both H 2 and CO yield were improved and carbon formation was significantly suppressed.
In addition, the Fe-promoted Ni catalyst was also proved to be sulphur-resistant.Tsodikov's team [67,68] tested NieFe/g-Al 2 O 3 catalysts prepared by epitaxial coating and a novel coreshell type catalyst containing Ni and Fe (Fig. 2) under SMR conditions in the presence of up to 30 ppm H 2 S. The materials showed good catalytic activity, unlike conventional Ni-based catalysts, which lose activity rapidly when the gas-vapour mixture contains H 2 S.This property was attributed to the core-shell structure, in which the core containing NieFe nanoparticles provided the catalytic ability while the g-Fe 2 O 3 shell provided vacancies for H 2 S to decompose to elemental sulphur following the reactions below: Cobalt is also considered a promising SMR catalyst additive because of its good activity for the WGS reaction, which assists in shifting the equilibrium towards higher H 2 production.However, one problem related to the usage of Co is its tendency to oxidize when the temperature and steam partial pressure are in the range used for SMR [69].Alloying it with Ni is a potential solution to this problem while preserving the advantages of both elements.A series of NieCo/ZrO 2 bimetallic catalysts with different Co loadings were tested and compared to monometallic Ni and Co catalysts by Harshini et al. [70].Their results suggested that a Ni/Co ratio of 1:1 was optimal for limiting both oxidation of Co and carbon formation caused by Ni.The material also exhibited long-term stability with a constant CH 4 conversion of 81.8% within 50 h with no surface carbon formation.You et al. [71] tested a series of NieCo/g-Al 2 O 3 catalysts and found that at a temperature of 800 C Co-modified catalysts exhibited the same reforming activity as unmodified ones with enhanced coke resistance.The performance of the bimetallic catalyst was not as good as Ni at low temperatures, possibly due to the formation of NieCo alloy, which increased the crystallite and particle size of the material, while decreasing metal dispersion and surface area, and blocking low-coordinate active Ni sites where the rate-determining CH 4 dissociation step takes place.
NieCu bimetallic catalysts have been used for the conventional SMR process [60] as well as low-temperature SMR [73e75].Results showed that by adding Cu as the promoter, a larger Ni crystallite size, surface area, and a better metal dispersion was obtained [74], and the overall carbon resistance of the material was enhanced [66,72].TGA results before and after a long-term SMR test (20 h) showed that carbon formation on NieCu/Al 2 O 3 was great suppressed (8.9%) compared to commercial Ni/Al 2 O 3 (28.3%).The addition of Cu has also been proven to enhance the activity of the WGS reaction [76], which explains the increase in CH 4 conversion when using CueNi as the catalyst.However, it should be noted that an upper limit exists in terms of the promoting effect of Cu.A Cu/Ni ratio equal to or higher than 5 in the material will result in reduced catalytic activity, as reported by Huang et al. [73].
The promoting effect of Zirconium was investigated by Lertwittayanon et al. [77] using Ni/a-Al 2 O 3 catalysts containing CaZrO 3 nanoparticles.CaZrO 3 loading between 10 and 15 wt.% showed the best catalytic performance with a CH 4 conversion of 67%.Results also showed that unlike conventional SMR catalysts requiring an S/C ratio of approximately 3, an S/C ratio of 1/3 or 1 was most appropriate for the CaZrO 3modified catalyst.This is because a high S/C ratio causes an excessive amount of steam to adsorb on the CaZrO 3 surface, which competes with the adsorption of CH 4 .
Boudjeloud et al. [78] tested a series of La-promoted Ni/a-Al 2 O 3 catalysts.The highest CH 4 conversion (97%) and H 2 yield (94%) were obtained with a Ni/La ratio of 7:3.The improved activity compared to monometallic Ni was credited to the decrease in Ni particle size and enhanced metal dispersion, which prevented the agglomeration and sintering of the bimetallic material.The addition of La also facilitated the reduction of Ni, however, its effect on coke resistance was not significant [27].
The effect of doping Molybdenum was studied by Maluf and Assaf [79] using MoeNi/Al 2 O 3 catalysts with different Mo concentrations.The addition of Mo decreased the surface area of the catalyst, possibly due to the blockage of active Ni sites by MoO x .However, the specific activity of each active site was increased, which was attributed to the transfer of electrons from MoO x to Ni particles resulting in an increase in electron density in Ni.Molybdenum carbide has also been employed in the methane reforming process and is known to have good catalytic activity and stability at high pressures [80].However, its stability quickly drops at atmospheric pressure due to the surface oxidation of Mo 2 C to MoO x by CO 2 .This problem can be mitigated by combining Mo 2 C with nickel.Despite being a major reason for the deactivation of traditional Ni catalysts, carbon deposited on the bimetallic NieMo 2 C surface promotes the carburization of NiMoO x back to its carbide form [81,82]. NieMo 2 C catalysts have been tested for dry methane reforming [81e83], steam reforming of methanol [84], as well as steam-CO 2 bi-reforming of methane [85], and have shown more promising results than unpromoted Mo 2 C catalyst or NieMo catalyst in their reduced form.
The promoting effect of the rare earth element, rhenium, was reported by Xu et al. [86].They concluded that by coating a NieRe bimetallic layer on the surface of a high cell density Ni monolith catalyst, the reducibility and catalytic performance of the material were enhanced.The low hydrogen adsorption energy of Re atoms also facilitates the adsorption of hydrogen atoms on Re and adjacent Ni atoms, which suppressed the oxidation of Ni and led to enhanced catalyst stability.

Nickel-based catalysts promoted by metalloids
Silicon is one of the most studied metalloids as it is often employed as the catalyst support for the SMR process, usually in its oxide or carbide form, because of its thermal stability and potentially high surface area.Silica is generally considered an inert material, as it has weak metal-support interaction with the active metal (Ni in most cases) due to its low reducibility [87].The lack of metal-support interaction in Ni/SiO 2 is also a Fig. 2 e Core-shell configuration with the NieFe alloy core surrounded by a superparamagnetic g-Fe 2 O 3 shell.Reprinted from Ref. [67] with permission from Elsevier.source of filamentous whisker carbon formation [88].To improve the interaction between Ni and the silica support matrix, Majewski et al. [89] synthesized a core-shell typed Ni/ SiO 2 catalyst using the St€ ober-deposition-precipitation method, and tested it under different SMR conditions.Results showed that the core-shell structure increased the coke resistance of the catalyst, as deposited carbon was only detected at a low s/c ratio (1:1) and temperature (550 C).Other characteristics of the support material, including crystallite size and metal dispersion, are affected by the acidity/basicity of the support and are also known to influence the rate of carbon growth on the catalyst surface.The acidity of the silica support facilitates the decomposition of methane, but at the same time promotes cracking and polymerization leading to catalyst deactivation because of carbon formation [88].To achieve a better acid-basic balance, basic metal oxides, including ceria [90,91] and magnesia [92,93] are often added to Ni/SiO 2 to tune the surface acidity of the support, and it was found that a homogeneous distribution of the basic sites on the acidic silica framework improved the long-term stability of the catalyst.
Elements with a similar electronic structure of carbon include tetra-and penta-valent p such as Sn, Sb, As, Ge, Pb, Ag, etc.The addition of these metals was predicted to have a cokeresisting effect, because, similar to the formation of nickel carbide (interaction between 2p electrons of carbon and 3d electrons of Ni), the reaction between these metals with Ni could potentially reduce the chance of NieC interaction.A few of the above candidates were tested by D.L.Trimm [94], and their coke-resistant ability followed the order of As > Ag > Sb > Sn > Pb.
The addition of Sn was also investigated by Nikolla et al.
[95e97] experimentally and numerically.The bimetallic NieSn catalysts showed lower CH 4 conversion during the first 30 min of the reaction, however, its long-term stability was greatly enhanced.The carbon resistance of the NieSn/YSZ catalyst was explained by DFT calculated reaction energy barriers, which showed that the NieSn alloy surface preferentially oxidizes C* rather than forming CeC bonds.The presence of Sn also lowers the binding of C* to low-coordinated sites, which is the position for carbon nucleation.The decrease in the catalytic activity of NieSn alloy is possibly due to the blockage of low-coordinate Ni sites by Sn, as these sites are the most active for the rate-determining CeH bond activation step.
Similar to Sn, boron-promoted Ni catalyst has also demonstrated good stability due to the reduction in carbon nucleation centres.A boron loading of 1 wt.% was sufficient to enhance the overall stability of the material without compromising its catalytic activity [98].Apart from this, Ligthart et al. [27] also reported the structural-promoting ability of boron for obtaining small Ni particles.However, one limitation of the bimetallic material is that the addition of boron strongly impeded the reduction of Ni.
Apart from the experimental work mentioned above, the SMR activity of metalloid-promoted nickel catalysts was also studied using numerical methods.In the work by Xu et al. [21], the catalytic activity of a series of bimetallic alloys was predicted based on a microkinetic model, and DFT-calculated atomic adsorption energies on the bimetallic M 1 M 2 (211) surface.By setting the conditions as 793 C, 12.2 bar, and with gas composition as 50% to equilibrium, alloys including Ni 3 M (M ¼ Sn, Sb, Ge, and As) and Co 3 Ge were predicted to have the highest activity (Fig. 3).Although these elements showed promising results, experimental validation of the in-silico study has not been found.Further study can be carried out on these metalloids (Sb, As, Ge)-based catalysts in search of an optimal balance between activity and stability.

Cobalt-based and other transition metal-based bimetallic catalysts
Apart from nickel, some other transition metals, such as Co, Cu, and Fe, have been used as catalysts for reforming processes (dry and steam reforming of hydrocarbons, glycerol, or bio-derived material).Shen et al. [99] tested a series of monometallic Co/CeO 2 and bimetallic Co-M/CeO 2 catalysts (M ¼ Ni, Al, and Cu) under conventional SMR conditions to Fig. 3 e Turnover frequencies for CO production under industrial (a) inlet conditions: T ¼ 638 K, P ¼ 14.3 bar, with a gas composition of 14.5% CH 4 , 83.1% H 2 O, 0.1% CO, 0.4% N 2 and 1.9% H 2 ; and (b) outlet conditions: T ¼ 1066 K, P ¼ 12.2 bar, with a gas composition of 2.4% CH 4 , 65.8% H 2 O, 6.5% CO, 0.3% N 2 and 2.5% H 2 .Reprinted from Ref. [21] with the permission from IOP Publishing.
study the effect of Co loading and different promoters.As mentioned previously, higher active metal loading does not necessarily mean better performance because of the uneven distribution of active compounds in the support.A Co loading of 12% was found to be optimal in terms of CH 4 conversion and H 2 yield.The addition of both Ni and Al increased CH 4 conversion, whereas Cu slightly reduced the overall catalytic activity, possibly due to the sintering of Cu.The combination of NieCo was chosen over AleCo because it exhibited higher CH 4 conversion (76.1%),H 2 selectivity (58.5%), and H 2 yield (44.5%).
The cobalt-based catalyst prepared from hydrotalcite precursors using the anion-exchange method was tested by Lucredio and Assaf [100] with low H 2 O/CH 4 feed ratios of 2 and 0.5 to test the stability of the material under extreme conditions.For an H 2 O/CH 4 ratio of 2, CH 4 conversion was maintained at 80% during 6 h of reaction; the carbon amount on the used catalyst was found to be only 2.7 wt.% after 30 h of reaction.The catalyst began to show a deactivation tendency due to the deposition of excess carbon when H 2 O/CH 4 ratio is further decreased to 0.5, and CH 4 conversion decreased from ~60% to 40% during 6 h of reaction.
Although Co is less prone to coke formation compared to Ni, the interaction between Co and the metal oxide support is strong, leading to the formation of cobalt oxides with limited reducibility [101].As presented in section 2.1, by adding a small amount of noble metal the reducibility of the transition metal-based catalysts can be largely improved because of the hydrogen spill over effect.Profeti et al. [102] explored the effect of noble-metal promoters (0.3 wt.% Pt, Pd, Ru, and Ir) on Co/Al 2 O 3 catalysts.Results showed that the addition of the noble metals significantly decreased the reduction temperature of cobalt species, with their promoting effect following the order of Pd > Pt > Ru > Ir.In terms of their catalytic activity, the Co-based bimetallic catalysts did not show satisfying results.Average CH 4 conversion of 50e60% was obtained for Pd-, Pt-and IreCo, 30% for RueCo, and only 7% for Co/Al 2 O 3 , which was possibly due to the partial oxidation of cobalt active sites in the presence of water molecules.
Akbari-Emadabadi et al. tested a CaeCo bi-functional catalyst/sorbent (with a mass ratio of Ca/Co ¼ 9) in the CL-SMR process, and investigated the promoting effect of yttrium [103] and zirconium [104], with a mass ratio of Ca/Y ¼ Ca/ Zr ¼ 4.5.Both promoted samples remained stable at 700 C for up to 16 redox cycles, whereas the unpromoted one was deactivated after 10 cycles.The catalytic performance of the materials is summarized in the table below (Table 2).Both Y and Zr showed promoting effect regarding catalytic activity for SMR and H 2 selectivity, and the usage of Zr was more advantageous in comparison to Y. Based on the results from catalyst characterization, the addition of Co reduced the overall surface area of the material by more than 30%, whereas the addition of Zr compensated this negative effect to some extent.Results also showed that Zr prevented the formation of Ca 2 Co 2 O 5 spinel in the structure of the bimetallic material, which lowered the risk of losing active sites of Co.The study proved Y and Zr to be promising textural promoters of the bi-functional catalyst/sorbent materials employed in CL-SMR.
Apart from Co-based bimetallic catalysts, catalysts combining two types of noble metal have also been studied.The research by Roy et al. [105] focused on a novel PteRh (1.2 wt.%) catalyst supported on metal foam.The sample was tested in a multichannel heat exchanger platform reactor to evaluate its potential in solid oxide fuel cell application.The combination of Pt and Rh was proven to enhance the production of hydrogen by SMR, with CH 4 conversion, H 2 yield, and H 2 /CO ratio of 97.2%, 3.16 mol per mol of CH 4 input and 6.03, respectively, all of which were higher than commercial monometallic Ni and Ru catalysts.The novel catalyst also showed excellent stability, negligible coke deposition was found after 200 h of SMR reaction at 800 C. Further research from an economic point of view should be carried out to evaluate the potential of these materials in large-scale applications.

Polymetallic catalysts
Compared to the relatively simple mono and bimetallic system, the application of catalysts containing three or more types of active metals in SMR has not been investigated in detail.Existing literature mainly examined Ni-based material with the addition of two or three commonly used elements, such as Co, Cu, Ru, Pt, etc.
The effect of the simultaneous presence of copper and zinc in Ni/Al 2 O 3 catalyst was investigated by Nazari and Alavi [106].They reported that Cu and Zn affect the Ni-based catalyst in different ways e Cu enables a better resistance to coke formation while Zn improves the catalyst's activity, stability, and H 2 selectivity.The optimal combination of the three metals for SMR was found to be 15%Nie1%Cue5%Zn, which achieved a CH 4 conversion of 94% and an H 2 yield of 3.12.
Jeon et al. [107] investigated the performance of a selection of bi-and trimetallic Ni-based catalysts under steam-deficient conditions.A series of bimetallic catalysts containing 5 wt.% of alkaline earth metal (Mg, Ca, Sr, Ba) or 0.5 wt.% noble metal (Ru, Rh, Pt, Pd), and trimetallic catalysts containing both alkaline earth metal and Ru were synthesized.Results from the tests demonstrated that adding Mg or Ca enhanced the coke resistance of Ni-based catalysts, whereas the effect of Sr and Ba was not significant.Among the noble metals, Ru was the best candidate for suppressing coke deposition.Based on these conclusions, a catalyst with optimized composition e 0.5%Rue5%Mge10%Ni/g-Al 2 O 3 e was selected and tested.A CH 4 conversion of 96% was maintained for 250 h, proving its excellent long-term stability.
Bi-functional polymetallic materials combining the catalytic ability of transition metals and CO 2 sorbents are commonly employed for the SESMR process [108].Chen et al. [109,110] found that a simple physical mixture of Ni/Al 2 O 3 (20 wt.%) and CaO was able to improve H 2 purity to above 95%, compared to 72% without in-situ CO 2 removal.Based on an elemental mapping analysis of the samples, each Ni elemental point was surrounded by several Ca points, which allowed the efficient capture of CO 2 produced during the reaction.Dewoolkar and Vaidya [111] synthesized hybrid NieCaO/ Al 2 O 3 and Ni-hydrotalcite materials by coprecipitation and incipient wet impregnation, respectively.The materials were tested at T ¼ 500 C with an S/C ratio of 9 for 20 cycles for stability evaluation.Results showed that CH 4 conversion and H 2 yield of the hybrid materials were higher than those obtained using a physical mixture of catalyst and sorbent, due to the more efficient mass and heat transfer.The hybrid materials also exhibited better stability, NieCaO/Al 2 O 3 and Nihydrotalcite maintained high H 2 purity of 90% for up to 11 and 16 cycles, respectively, compared to the sintering and deactivation of the mixed material after only 2 cycles.
Di Giuliano et al. [112,113] reported the use of mayenite as for bi-functional NieCaO material.Results from multicycle sorption/regeneration TGA demonstrated the stable sorption capacity of the material after 20 cycles.This enhanced stability compared to commercial CaO was attributed to the presence of mayenite as an inert binding, preventing CaO from sintering.The same phenomenon was also observed by Dang et al. [9], confirming the role of mayenite as a structural stabilizer.Di Giuliano et al. also concluded that the nickel precursor used for material synthesis may affect the texture and reducibility properties, and nickel nitrate hexahydrate was found to the most suitable precursor.
Kim et al. used ruthenium as the reforming catalyst, and tested the performance of RueCaO/Ca 3 Al 2 O 6 under SESMR conditions [114].As Ru is a highly active catalyst for SESMR, the mass fraction of CaO was able to be increased significantly compared to conventional NieCaO-based materials.Ca 3 Al 2 O 6 acted as the structural stabilizer against sintering, and maintained the surface area of RueCaO/Ca 3 Al 2 O 6 at 18 m 2 /g after 10 cycles of SESMR, compared to 5 and 6 m 2 /g for Ru/lime and Ru/CaO.
Hafizi et al. [115] modified conventional calcium-based CO 2 sorbent with CeO 2 , and tested it for CL-SESMR together with Co 3 O 4 /SiO 2 .The addition of CeO 2 significantly improved the morphology of CaO by increasing its surface area and uniformly distributed pores in the sorbent structure.Combined with the Co-based catalyst, the material was able to produce high purity H 2 (93e96%) for 8 redox cycles, and maintain the same CO 2 removal efficiency for 3 carbonation/calcination cycles.
Ghungrud et al. [116] reported a novel trimetallic bifunctional material for SESMR, consisting of Ni and Co (with concentrations varying from 0 to 40%) supported on hydrotalcite and promoted by 2.5 wt.% cerium.The hybrid material was evaluated in terms of its H 2 production ability, sorption capacity, and cyclic stability.Results revealed that CH 4 conversion increases with the Co content in the material, which is possibly due to the enhancement of WGS reaction by Co.Under optimal reaction conditions, CH 4 conversion of 95.7% and 90.7% were obtained by two Ce-promoted NieCo samples (with Ni/Co ratios of 1/3 and 1, respectively).The maximum sorption capacity was found to be 1.74 and 1.51 mol CO 2 /kg, respectively.The material was also tested at optimal conditions for 25 cycles, samples showed good stability by maintaining an H 2 concentration higher than 90%, and remained stable for 21 and 16 cycles.This property was attributed to the effective metal-support interaction and higher active metal dispersion within the promoted material.The author concluded that the hydrotalcite-supported CeeNieCo (2.5, 10, 30 wt.%) trimetallic catalyst/sorbent shows good performance and could be a promising candidate for large-scale SESMR application.
Similarly, Dewoolkar and Vaidya investigated the promoting effect of Ce and Zr on a Ni/hydrotalcite bi-functional material [117].Results showed that both Ce and Zr were able to increase the surface area and surface basicity, which inhibited coke formation.Both Ce and Zr promoted materials remained stable for 13 and 17 cycles, respectively, whereas the unpromoted material became unstable after 9 cycles.The addition of Ce was found to be particularly beneficial, as the promoted bi-functional material reached a high CH 4 conversion of 96.4% and an adsorption capacity of 1.41 mol CO 2 /kg sorbent.
Apart from the type of promoter added to Ni-based catalysts, the structure of the promoted catalyst also influences its overall performance.Cho et al. [118]  i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 1 5 8 7 9 e1 5 8 9 3 monometallic 1 wt.%Ru/Al 2 O 3 synthesized by the conventional wet impregnation method.The comparison between (b) and (c) showed that the novel structure was able to improve methane conversion and maintain it at a higher level when gas hourly space velocity was largely increased.This was because the novel structure enables the active metal, ruthenium, to be mainly deposited on the outer region of the alumina pellets and can therefore be utilized more efficiently.With the addition of nickel, (a) achieved an even higher methane conversion than (b) at higher gas hourly space velocity.This proved that the novel structure enables the efficient utilization of active noble metals, reducing the metal loading necessary and, therefore, the overall cost.
Obradovic et al. [119,120] proposed a novel plate-type catalyst for SESMR, as demonstrated in Fig. 4.This material was synthesized by depositing Pt and Al 2 O 3 on a static mixer element made of Ni alloy.During the SESMR process, it acts simultaneously as the catalyst, the distributor for solid sorbent, as well as the gas-phase radial mixer.The CH 4 conversion obtained by this novel catalyst was 15 times higher than Table 3 e A summary of bi/polymetallic catalysts for SMR (Green/Red: promoter has a positive/negative effect on the overall performance, compared to the monometallic base material.Blank: no data available.*: result based on numerical modelling.).monometallic Ni under the same conditions.However, its catalytic activity rapidly decreased after 10 h of reaction, temperatures higher than 590 C also led to activity loss due to carbon accumulation on the material surface.Further investigation on modifying relevant parameters (such as the promoter type, metal loading, and reaction conditions) to increase its stability may be of interest.
Based on the above review, it can be concluded that an increase in catalytic activity is usually achieved either by modifying the textural properties of the material (e.g.increasing surface area and metal dispersion) using a second active metal; or by modifying the energetics of the surface reactions, in particular the rate-determining step (e.g.decreasing activation energy of the first CeH bond in CH 4 dissociation).Although a variety of metal combinations have been studied for their performance in the SMR reaction, there exist many other combinations of elements that have been predicted to be active or have not been evaluated at all regarding their reforming activity.

Conclusions and outlook
SMR is currently the most dominant hydrogen production technology, and extensive research on the catalytic aspect of this process has been carried out.Carbon emissions from SMR can be reduced by adding a CO 2 -sorption step.This review provides insights on recent developments in the use of bi/ polymetallic catalysts for (SE)SMR.The performance of the bi/ polymetallic catalysts presented in this review is briefly summarized in Table 3 and is evaluated based on three main factors: stability (resistance to carbon, sulphur, sintering, and oxidation), catalytic activity and selectivity, as well as their physical/chemical properties (reducibility and self-activation ability).
The most widely used SMR catalyst to date is ceramicsupported nickel because of its relatively good performance and inexpensive price, but problems such as sintering and coke formation still exist.In search of catalysts with better performance, various elements have been added as promoters to conventional Ni-based catalysts.Noble metal-promoted catalysts generally have superior reactivity, coke resistance, and enhanced reducibility, but their applications are often limited by their prices.Researchers have therefore turned to non-noble metals and metalloids.The addition of iron was found to be coke and sulphur-resistant due to the surface redox reaction between Fe 0 and FeO x .Both cobalt and copper can enhance the activity of the WGS reaction, thus shifting the reaction towards more hydrogen production.Elements including zirconium, yttrium, and lanthanum were found to be good textural promoters due to their ability to increase the surface area and metal dispersion of the material.Silicon is often used as the catalyst support in its oxide or carbide form, and the addition of ceria or magnesia was able to tune the surface acidity of silica for better long-term stability.The addition of other metalloids (tin, boron) led to enhanced cokeresisting ability, but often at the expense of losing catalytic activity.A DFT-based study has also predicted germanium, arsenic, and antimony to be effective promoters of nickel-based catalysts, however, further experimental verification is still necessary.
Apart from the type of element selected as the promoter, the loading of each component is also a critical parameter that influences the overall performance of alloy material.A higher active metal loading does not necessarily indicate a higher activity due to the restricted distribution of active metal in the material.The influence of material structure on catalytic activity was also investigated.Although novel core-shell type, plate type, and metal foam-support structures were found to be beneficial to the overall catalytic performance, this largely complicates the synthesis process and limits the wide application of the materials.It is, therefore, necessary to find a balance between improvements to the material properties and a viable and efficient material preparation process.

Table 2 e
Comparison between the catalytic activity of unpromoted and promoted CaeCo samples.