H2-rich syngas production from biogas reforming: overcoming coking and sintering using bimetallic Ni-based catalysts

H 2 -rich syngas production from biogas reforming: Overcoming coking and sintering using bimetallic Ni-based catalysts S. Carrasco-Ruiz a , Q. Zhang b ,J.G  andara-Loe a , L. Pastor-P erez a , J.A. Odriozola a,b , T.R. Reina a,b,** , L.F. Bobadilla a,* a Inorganic Chemistry Department and Materials Science Institute, University of Seville-CSIC, 41092, Sevilla, Spain b Department of Chemical and Process Engineering, University of Surrey, Guildford, GU2 7XH, UK highlights graphical abstract H 2 -rich syngas can be obtained efficiently from biogas reforming. Ni-based catalysts are deactivated by metal sintering and coke deposition. Doping with low Rh loadings deactivation can be successfully prevented. article info Article history: Received 23 August 2022 Received in revised form 11 February 2023 Accepted 20 March 2023 Available online 19 April 2023 Keywords: Dry reforming H 2 -rich syngas NieRh catalyst Coking resistance abstract Dry reforming of methane is a very appealing catalytic route biogas (mainly composed by greenhouse gases: carbon dioxide and methane) conversion into added value syngas, which could be further upgraded to produce liquid fuels and added value chemicals. However, the major culprits of this reaction are coking and active phase sintering that result in catalysts deactivation. Herein we have developed a highly stable bimetallic NieRh catalyst supported on mixed CeO 2 eAl 2 O 3 oxide using low-noble metal loadings. The addition of small amounts of rhodium to nickel catalysts prevents coke formation and improves sintering resistance, achieving high conversions over extended reaction times hence resulting in promising catalysts for biogas upgrading. ©2023 The Author(s). Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). *Corresponding author. ** Corresponding author. Inorganic Chemistry Department and Materials Science Institute, University of Seville-CSIC, 41092, Sevilla, Spain. E-mail addresses: [email protected] (T.R. Reina), [email protected] (L.F. Bobadilla). Available online at www.sciencedirect.com ScienceDirect journal homepage: www.elsevier.com/locate/he international journal of hydrogen energy 48 (2023) 27907e27917 https://doi.org/10.1016/j.ijhydene.2023.03.301 0360-3199/©2023 The Author(s). Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. This is an open access article under the CC BY-NC- ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction Fighting global warming alert to mitigate the negative human impact on the environment is one of the urgent priorities of the scientific community raking top as Global Challenge priority for a sustainable development. The greenhouse effect caused by the emission of harmful gases into the atmosphere is provoking an increase in global temperature and therefore to changing climates causing droughts, deforestation, and desertification of the planet. The four main greenhouse gases (GHG) attracting serious global attention today are CO 2 ,CH 4, SO 2 , and N 2 O. Carbon dioxide is by far the most emitted gas into the atmosphere, accounting for 76% of total emissions in 2019 [1]. According to United in Science 2020 [2], a report compiled by the World Meteorological Organization (WMO) under the direction of the United Nations Secretary-General, lockdown-related fall in emissions caused by COVID-19 pandemic will not reduce sufficiently the CO 2 concentrations in the atmosphere. The growing trend of CO 2 and CH 4 emissions driven by the increase of global energy consumption makes mandatory the search of alternatives to achieve GHG valorisation [3]. The use of renewable energy sources represents “must do” action to favour the transition towards a low-carbon economy. Herein, biogas can play an important role in the development of the renewable energy market nowadays, as it has a wide variety of applications compared to other renewable energy sources [4]. Biogas is produced by anaerobic digestion of biodegradable wastes and is mainly compound by carbon dioxide and methane [5,6]. After its production, there are three main routes for biogas utilisation (Fig. 1b). Biomethane is a nearly pure methane stream that can be obtained by CO 2 separation of biogas. At present, biomethane has been widely used as an engine vehicle fuel in many countries and has broad development prospects [7,8]. Biogas is a raw material to produce syngas (H 2 /CO gas mixture) for industrial syntheses or energy purposes [7]. There are several ways to reform biogas into syngas such as Dry Reforming of Methane or Bi-Reforming of Methane. Green electricity can be produced from biogas by Combined Heat &Power system (CHP) [9]. Syngas is a combustible gas that can be used to produce electrical energy in turbines and fuel cells (Fig. 1c), but the current market demand uncovers the potential of syngas as a feedstock to produce fuels like diesel, naphtha, and gasoline, as well as for high value-added chemical inputs via Fischer- Tropsch synthesis (FTS) [10]. Focusing on biogas upgrading, there several thermal catalytic routes for syngas production from biogas namely: Oxy- Reforming of Methane (ORM), Bi-Reforming of Methane (BRM) and Dry Reforming of Methane (DRM) [11]: ORM :3CH4þCO2þO2#4CO þ6H2(Eq.1) DH0¼58 kJmol DG0¼1kJmol BRM :3CH4þCO2þ2H2O#4CO þ8H2(Eq.2) DH0¼220 kJmol DG0¼151 kJmol DRM :CH4þCO2#2CO þ2H2(Eq.3) DH0¼247 kJmol DG0¼170 kJmol Fig. 1 eBiogas life cycle: a) Biogas production; b) Biogas utilisation; c) Syngas utilisation. international journal of hydrogen energy 48 (2023) 27907e2791727908 Even though the ORM (Eq. (1)) is presumably the most attractive option given its autothermic nature which results in significant energy savings, safety concerns associated with oxygen use limit its interest to industrial and large-scale applications [12]. The most popular technologies are the BRM (Eq. (2)) and the DRM (Eq. (3)). The first one leads to a syngas H 2 /CO mixture ratio of a maximum of 2, meanwhile the last one can achieve a molar ratio of 1 [13]. However, in the context of a circular economy when a CO 2 utilisation route is pursued, CO 2 conversion may prevail over H 2 /CO ratio as selection criteria for the optimal route. Herein, DRM takes the edge over BRM as reported elsewhere [14]. In any case, the conditions used for reforming also favour other side reactions (Eqs. (4)e(7)), in which carbon deposits (C*) are formed as follows [15]: Methane decomposition :CH4#C*þ2H2(Eq.4) DH0¼75 kJmol Boudouard reaction :2CO #C*þCO2(Eq.5) DH0¼172 kJmol CO2hydrogenation :CO2þ2H2#C*þ2H2O(Eq.6) DH0¼90 kJmol CO hydrogenation :CO þH2#C*þH2O(Eq.7) DH0¼131 kJmol The formation of carbon deposits leads to catalysts deactivation, so it is important to find materials capable of minimizing carbon formation. Besides these reactions that can deactivate the catalyst, there is also a side reaction that would decrease the H 2 yield, the Reverse Water Gas Shift (RWGS): RWGS reaction :CO2þH2#CO þH2O(Eq.8) DH0¼46 kJmol When it comes to catalysts selection, Ni-based catalysts have shown a good activity and conversion of biogas into syngas [16e18], but they are textbook example for coking and sintering [19,20]. It has also been reported that noble metalbased catalysts have a greater activity and conversion, but in addition, they have more stability and higher coking resistance yet are expensive and less available [21]. For these very reasons bimetallic alloys comprised of Ni and low amounts of noble metals (i.e. Pd, Rh, Pt, Ru) are interesting in reforming processes. Irrespectively of the selected formulation, it is clear that catalysts stability (resistance to the different deactivation phenomena) is crucial for an optimal process design. Frequently in heterogeneous catalysis, deactivation issues are resolved by catalyst regeneration. For instance, thermal treatments to remove carbon deposits or get rid of potential sulfides [22]. The sequence of activity for the DRM using M/MgO (M≡Ir,Ni,Pd,Pt,Rh,Ru) catalysts is Ru >Rh >Ni »Ir,Pt,Pd considering the presence of sulfur compounds [23]. Comparing Ru and Rh, the regeneration of Ru catalysts is much lower than that of Rh, so it seems that the best option for our objective is a bimetallic NieRh catalyst [24]. Beyond the active phase, the support choice is also essential to ensure an adequate performance. For reforming reactions given the demanding process conditions (high temperatures) thermal and mechanical stability are key factors to consider. In this regard, alumina supports (Al 2 O 3 ) offers an excellent balance thermal/mechanical property while also displaying high specific area which enhances metallic dispersion. However, the acidic sites of this kind of supports can increase coke formation. That is why ceria (CeO 2 )isan interesting redox promoter. Indeed, ceria allows a flexible tunning of the acid/base properties of the support and provides excellent oxygen mobility, preventing carbon deposition via oxidation of the coke precursors [13]. Several authors have studied bimetallic NieRh catalysts for the DRM reaction [25,26], yet the specific role of Rh and how it benefits catalytic performance is not fully addressed. Particularly, short-term stability test and insufficient of postreaction analysis in the current reports pose some questions regarding the fundamental role of Rh in the bimetallic formulation. So, under these premises, this work focuses on the development of an advanced multicomponent catalysts NieRh/CeO 2 eAl 2 O 3 whose performance is compared to a bare monometallic Ni/CeO 2 eAl2O 3 system. Catalytic activity and stability results as well as pre- and post-reaction characterisation will be discussed to understand the beneficial effect of noble metal addition. Experimental Catalysts preparation The support used was a commercial ceria-alumina oxidic support with 20 wt% of CeO 2 (Puralox, SASOL). The monometallic catalyst was prepared by wet impregnation, where the support was first impregnated with Ni(NO 3 ) 2 ·6H 2 O (Sigma-Aldrich) diluted in distilled water, evaporated at reduced pressure in a rotavapor, dried overnight at 100 C and calcined at 550 C for 4 h. In a similar procedure, the bimetallic catalyst was prepared by wet co-impregnation, where the support was impregnated with Ni(NO 3 ) 2 ·6H 2 O (Sigma-Aldrich) and RhCl 3 (Sigma-Aldrich) diluted in distilled water, evaporated at reduced pressure in a rotavapor, dried overnight at 100 C and calcined at 550 C for 4h. In all cases the NiO content is calculated to be 10 wt% and 0.5 wt% Rh for the bimetallic sample. These ratios were chosen based on previous works [27e29]. Characterisation techniques The textural properties of the samples were evaluated from nitrogen adsorption-desorption isotherms at liquid nitrogen temperature in a Micromeritics Tristar II apparatus. Before international journal of hydrogen energy 48 (2023) 27907e27917 27909 analysis, the samples were degassed at 150 C for 8 h in vacuum. On one hand, the specific surface area (S BET ) was determined by the Brunauer-Emmet-Teller (BET) method [30] and corresponds to the sum of the inner surface of the pore plus the outer surface of the grains. Additionally, the Barrett- Joyner-Halenda (BJH) method was used for determining the pore size distribution using the desorption isotherm. Pore volume represents the inner and outer granular volume, and the average pore size was calculated as the ratio of the pore volume and the specific surface area and normalized using a coefficient that depends on the pores shape [31]. Scanning electron microscopy (SEM) analysis was performed on the calcined samples in a vacuum, using a JEOL 5400 microscope equipped with an EDS analyser (Oxford Link). X-ray diffraction measurements were carried out in a X'Pert Pro PANalytic instrument. The diffraction patterns were recorded at 40 mA and 45 kV using Cu-Karadiation (l¼0.154 nm). The 2qangle was increased using a step size of 0.05and a step time of 300 s in a range of 10 to 90. Temperature programmed reduction (TPR) measurements were conducted in a conventional U-shaped quartz reactor connected with a thermal conductivity detector (TCD), passing a flow of 50 mL/min of 5% H 2 diluted in Ar. TPR experiments were performed using approximately 50 mg of each catalyst at a heating rate of 10 C/min from room temperature (RT ¼25 C) to 900 C. A mixture containing acetone and dry ice was used as a cold trap to remove the water formed throughout the procedure. Temperature programmed oxidation (TPO) experiments were carried out after the long-term stability tests to investigate the carbon species deposited on the catalyst surface. TPO analysis were conducted in a U-shaped quartz reactor coupled to a PFEIFFER Vacuum PrismaPlus mass spectrometer. 25 mg of both samples were heated up to 900 C at a rate of 10 C/min in a calibrated flow of 50 mL/min (5% O 2 , 95% He). Catalytic activity and stability Catalytic performance was evaluated in a fixed-bed continuous-flow reactor described elsewhere [32] in which 100 mg of undiluted catalyst were loaded for each run. Prior to the reaction, the samples were in situ reduced in a flow of 10% H 2 /N 2 at 850 C for 1 h. The reaction was performed at atmospheric pressure passing a flow of 100 mL/min of CO 2 /CH 4 /N 2 (molar ratio of 1:1:6) and decreasing successively the temperature from 850 C down to 550 C until achieving the steady state each 50 C. The WHSV (Weight Hourly Space Velocity) was fixed at 60 L/g cat $h. All the gases involved in the experiments were monitored by utilising an on-line gas analyser (ABBAO2020) which was equipped with both IR and TCD detectors. The catalytic stability of both mono and bimetallic materials was also studied. For this purpose, 100 mg of each catalyst were reduced under the same conditions as for the catalytic activity tests, and then the reaction conditions were evaluated at atmospheric pressure passing the same flow conditions as above at 650 C for 48 h. A second stability study was carried out at 850 C for 48 h. The catalytic results obtained have been expressed in terms of conversions of both reactant gases and the H 2 /CO molar ratio. The equations used to estimate these parameters are as follows: CH4conversion ð%Þ¼nCH4in nCH4out nCH4in $100 (Eq.9) CO2conversion ð%Þ¼nCO2in nCO2out nCO2in $100 (Eq.10) H2=CO ¼nH2 nCO $100 (Eq.11) Being nthe molar flow of CH 4 ,CO 2 ,H 2 , and CO respectively and the subscripts in or out correspond to the inlet or the outlet reactor flow. Results and discussion Characterisation of solid catalysts The textural properties (S BET , and pore size and volume) of the catalysts and the commercial support are summarized in Table 1. As can be observed, the specific surface areas of the catalysts are quite similar. This indicates a good dispersion of the metals on the surface of the support, which will have a positive effect on the activity of the materials. Likewise, both pore volume (V pore ) and pore width (D pore ) are almost identically, showing that no agglomeration of metal particles has occurred on the catalyst and that the pores support have not been blocked. Besides that, the textural properties obtained for the ceriapromoted alumina support suggest that both Ni and Rh are introduced into the pores of the support, so that the specific surface area, pore volume and pore width decrease. In addition, the nitrogen adsorption/desorption isotherms at 77 K have been obtained for both catalysts as shown in Figure S1. According to IUPAC classification [33], four isotherm types are usually found in catalyst characterisation, and each isotherm shape depends directly on the solid porous texture. In our case, both isotherms correspond to a type IV isotherm, which is related to a mesoporous material (pore size between 2 nm and 50 nm). Besides, there is also a classification of the type of hysteresis, being the type H1 found in our isotherms. This kind of hysteresis is characteristic of solids consisting of particles traversed by quasi-cylindrical channels or consisting of aggregates (consolidated) or agglomerates (unconsolidated) of spheroidal particles, and the pores may be uniform in size and shape. Figure S2 shows the pore size distribution obtained by BJH method for both catalytic materials. As can be noted, both pore distribution curves are very similar confirming that there Table 1 eTextural properties of the support and prepared catalysts. Sample S BET (m 2 /g) V pore (cm 3 /g) D pore ( A) CeO 2 eAl 2 O 3 159 0.39 71.9 Ni/CeO 2 eAl 2 O 3 132 0.31 67.7 NieRh/CeO 2 eAl 2 O 3 138 0.32 66.6 international journal of hydrogen energy 48 (2023) 27907e2791727910 are not agglomeration of metal particles and the material cavities have not been blocked. The morphology of the calcined catalysts is shown in the SEM micrographs provided in Fig. 2. Analysing the distribution of elements in the different mappings, we can confirm that both Ni and Rh are well dispersed and homogeneously distributed on the surface of the catalysts, corroborating a successful synthesis for both samples. The structure of both prepared catalysts was analysed by means of X-Ray Diffraction (XRD). Both samples were firstly analysed after calcination at 550 C(Fig. 3a). From this figure, it can be observed that all samples contain the typical diffraction lines corresponding to the (111), (200), (220) and (311) crystalline planes expected for the CeO 2 fluorite-like phase while the peaks corresponding to the (311) and (440) crystalline planes are ascribed to the g-Al 2 O 3 phase [34]. Fig. 2 eSEM micrographs of both calcined catalysts: (I)Ni/CeO 2 eAl 2 O 3 ; (II)NieRh/CeO 2 eAl 2 O 3 ; where (a) general view; (b) Ce mapping; (c) Ni mapping; (d) Al mapping; (e) Rh mapping. Fig. 3 eXRD of: a) calcined samples (550 C); b) reduced samples (850 C). international journal of hydrogen energy 48 (2023) 27907e27917 27911 Fig. 3b shows the XRD patterns for the samples after reduction in a flow of 10% H 2 /N 2 at 850 C for 1 h. Hence, the diffraction lines related to metallic Ni can be clearly seen, although the metallic Rh peaks were hardly observed due to the high dispersion and/or low loading of the noble metal. An overall idea of the redox behaviour and the metalsupport interactions was gathered by Temperature Programmed Reduction (TPR) analysis. The TPR profiles obtained for both catalysts are presented in Fig. 4. Notably three different regions are distinguished. First, aRegion (150e350 C) shows the reduction of surface ceria [35e37], which is enhanced on the bimetallic catalyst due to the spill-over phenomenon [38]. The bRegion (450e650 C) exhibits the reduction of NiO with low to medium strength of interaction with the support. Finally, the reduction of NiO with high strength of interaction with the support occurs at temperatures above 700 C(gRegion)[39]. Comparing both reduction profiles, we can see that the addition of rhodium improves notably the reducibility of Ni, facilitating its reduction at lower temperatures. This higher reducibility implies better redox properties, and these redox characteristics may enhance the catalytic performance during the DRM reaction. DRM catalytic activity The reduced catalysts were tested in the DRM reaction at a temperature range of 550e850 C and at atmospheric pressure with a CO 2 /CH 4 molar ratio 1:1. As we can see in Fig. 5.a, CH 4 conversions were lower than CO 2 conversions, which may suggest that the CH 4 activation is more difficult and requires higher temperatures in good agreement with DFT results [40]. The activation and cleavage of CeH bonds is more energetic than CO 2 dissociation. On the other hand, comparing the results between both catalysts, no significant differences are observed in the carbon dioxide and methane conversions, with similar conversion levels for the monometallic catalyst and the bimetallic one. In terms of the H 2 /CO molar ratio (Fig. 5b), both catalysts also showed similar results. However, this molar ratio stays over 1 throughout the reactions, so we can conclude that H 2 concentration is higher than CO concentration. This is an interesting for biogas reforming via DRM since typical H 2 /CO are slightly lower (typically close to 1). The main reason of this observation is the presence of multiple parallel reactions described above in the introduction section either consuming CO or producing H 2 ,i.e. the RWGS/WGS process which is effectively catalysed by NieCeO 2 systems [41]. Fig. 6 shows the XRD of the spent catalysts. Apart from the diffraction peaks of the metallic and support phases, we can observe a higher intensity in the 20-402qzone for the monometallic catalyst than for the bimetallic one. This testifies the formation of carbon deposits as will be well discussed further below. DRM catalytic stability Stability tests show how efficient catalysts are over long reaction runs and commonly stability is more important than activity in commercial catalysts design since it determines the frequency of start-up/shutdowns operations and the catalysts regeneration cost. Fig. 4 eH 2 -TPR profiles of both samples. Fig. 5 eDRM catalytic activity: a) CO 2 &CH 4 conversions; b) H 2 /CO molar ratio. international journal of hydrogen energy 48 (2023) 27907e2791727912 Hence, we have performed stability studies at two different temperatures. This way, we will be able to study the carbon deposition tolerance and metal sintering resistance of both catalysts in more in-depth manner. On the hand we will assess the coking tolerance at two different operation regimes and on the other hand we will evaluate the sintering resistance when increasing the thermal stress on the catalyst's particles. The first stability test was performed at 650 C. As shown in Fig. 7, both methane and carbon dioxide conversions for the Ni/CeO 2 eAl 2 O 3 sample decrease with time-on-stream, while conversions for the NieRh/CeO 2 eAl 2 O 3 sample remain practically constant over the 48 h of the test. This result clearly evidences the superiority of the bimetallic formulation showcasing the positive impact of small amounts of Rh. In order to study the carbonaceous deposits over both catalysts, TPO and XRD analyses were performed on the spent samples after the stability test. The peaks observed in the TPO diagram (Fig. 8a) represent the generation of CO 2 by the oxidation of the carbon deposits formed, so it can be affirmed that the carbon deposited in the NieRh catalyst is minimal compared to that of the reference sample. The temperature at which the CO 2 maximums appear provides information on the nature of the carbonaceous species deposited on the catalyst surface. Generally speaking, carbonaceous deposits have different structure orders and morphologies, and, in our case, we can distinguish between three types [42e50]: The first peak appears at 300e400 C and corresponds to a very labile amorphous carbon (C a ) which can be removed at low temperatures. The second peak appears at ~500 C and corresponds to graphitic carbon (C b ) which is very thermally stable. The C b is hardly removed by gasification so it leads to severe catalytic deactivation. Finally, the C g species were ascribed to carbon nanotubes, they appear at ~650 C. As we can see in Fig. 8.a, post-stability (650 C) TPO analysis show more C b deposits on the monometallic catalyst than on the bimetallic one. This justifies the continuous decrease of CH 4 and CO 2 conversions for the Ni/CeO 2 eAl 2 O 3 which is related to carbon poisoning and evidences the poorer performance of the reference material compared to the advanced bimetallic formulation. In addition, the XRD analyses of both spent samples (Fig. 8b) also show a clear difference since there is a peak at 26approximately in the monometallic sample that it is not seen in the bimetallic sample. This peak corresponds to graphitic carbon deposits because it has a crystalline structure determined [51], thus confirming that there are more graphitic carbon deposits on the monometallic catalyst in fair agreement with the TPO analyses. The second stability test was performed at 850 C(Fig. 9). In this instance, the CO 2 conversions decrease as the same way for both catalysts, going from almost 100%e95% at the end of the experiment. Nonetheless, the methane conversion for the Ni/CeO 2 eAl 2 O 3 sample decreases from 97% down to 75% while the CH 4 conversion for the NieRh/CeO 2 eAl 2 O 3 sample declines less, going from almost 95%e80%. Even so, these conversions are higher than those achieved during the stability test at 650 C given the endothermic nature of the reaction. As in the previous case, the TPO analyses of the poststability samples at 850 C also represent three very distinct peaks corresponding to the three types of carbon deposits mentioned above (Fig. 10a). In this case, the area under the curve of NieRh/CeO 2 eAl 2 O 3 sample is significantly lower than the area under the curve of Ni/CeO 2 eAl 2 O 3 sample, which reconfirms that the addition of Rh to the Ni catalyst improves coking resistance. In fact, with a view to facilitate its visualisation, the signal of the former has been multiplied by a factor of 4, as it was difficult to separate the two peaks on a normal scale. It must be emphasised the carbon deposition is thermodynamically more favoured at lower temperatures between 600 and 750 C. In any case our bimetallic sample outperformed the reference catalysts in both temperature regimes. However, graphitic carbon formation was lower on the monometallic catalyst in this test, as coke formation is more favoured at lower temperatures. A particularity of this stability test is that it is possible to observe how the sintering phenomenon affects the monometallic catalyst with the XRD Fig. 6 eXRD of spent samples. Fig. 7 e650 ºC-Stability test results. international journal of hydrogen energy 48 (2023) 27907e27917 27913 of these samples (Fig. 10b). The peak of Ni located at 51.8was used to estimate the crystallite size by means of Scherrer equation. It was found that the crystallite size is higher in the monometallic catalyst. To make a comparison with the Ni crystallite size before reaction, the same procedure was performed with the deconvoluted Ni peak of both reduced catalysts. In the case of the monometallic sample, the Ni crystallite size increased from 12 nm to 17.5 nm, while in the bimetallic catalyst this value increased from 11.7 nm to 13 nm. This is evidence of how the addition of a bimetallic active phase prevents the sintering deactivation of the catalysts since the segregation of Ni species from stable bimetallic formulations is somewhat less energetically favourable in comparison to pure Ni particles [52,53]. Fig. 8 ePost characterisation after stability test at 650 C: a) TPO; b) XRD. Fig. 9 e850 ºC-Stability test results. Fig. 10 ePost characterisation after stability test at 850 C: a) TPO; b) XRD. international journal of hydrogen energy 48 (2023) 27907e2791727914 Comparing the results obtained in the stability tests at 650 C and 850 C, we can state that the deactivation at 650 C is mainly caused by the formation of coke, while at 850 C the deactivation by sintering prevails, as less carbon formation is observed. In both situations, the bimetallic catalyst stands out as a promising alternative. At 650 C the predominant secondary reaction is methane decomposition (Eq. (4)) where carbon deposits and hydrogen are formed from methane. This explains the low methane yields and that the H 2 /CO ratio remains above 1 at this temperature. At 850 C, however, the reverse reaction of the Boudouard reaction (Eq. (5)) starts to occur, where the carbon formed with CO 2 is converted to CO. This could decrease the H 2 / CO ratio, but C b isverythermallystable,andtheCO formationis not large enough to counteract the excessive H 2 formation. Conclusions Biogas conversion to H 2 -rich syngas represents a straightforward approach for greenhouses valorisation in the context of a circular economy. Herein, a multicomponent catalyst based on NieRh bimetallic active phase supported on a ceriapromoted alumina (NieRh/CeO 2 eAl 2 O 3 ) has been prepared and tested in biogas upgrading via DRM. The performance of this advanced system has been compared to that of a reference monometallic nickel homologue (Ni/CeO 2 eAl 2 O 3 ). Different pre-reaction and post-reaction characterisation tests have been carried out, as well as a reaction monitoring based on the CH 4 and CO 2 conversions and the H 2 /CO molar ratio obtained. The starting hypothesis suggested better performances for the bimetallic catalyst than for the monometallic one. TPR experiments showed a higher reducibility for the bimetallic sample compared to the monometallic one, which implies better redox properties, enhancing the catalytic performance during the DRM reaction. In the catalytic activity test, the differences observed are negligible however key discrepancies are observed in the catalysts’ stability. The results obtained in the different stability tests show that there is indeed a synergic NieRh effect that improves catalytic stability, thus the conversions on the bimetallic catalyst remain almost constant over long-term runs while the monometallic catalyst suffers a significant activity drop showcasing its deactivation. Post-stability TPO experiments showed the formation of two different carbonaceous species, an amorphous carbon and a crystalline carbon, the latter being the cause of deactivation as it is very thermally stable. Our results demonstrate that the monometallic systems nucleate greater concentration of carbon than the advanced bimetallic sample being carbon deposition favoured in the low-temperature range (i.e. 650 C). With the stability test at 850 C the sintering of nickel particles can be also observed, with a remarkable increase on the Ni clusters size on the monometallic catalyst compared to the bimetallic which demonstrates a greater tolerance towards active phase agglomeration. Overall, this paper showcases the key role played by heterogeneous catalysis within CO 2 conversion technologies and the research still needed to keep refining catalytic formulations to avoid deactivation seeking for an optimal operation. In particular, this work has demonstrated that the addition of low amounts of Rh significantly improves the catalytic performance, reducing coke formation and sintering. A novel angle of this work is the proven coking resistance of our multicomponent catalysts in the low-temperature DRM window. This is a remarkable result beyond the catalysts optimization opening a new research avenue for the design of low-temperature DRM units resulting in a potential tangible reduction in energy consumption and overall process operation costs. In any case, the design of advanced multicomponent catalysts as the one presented herein will contribute to facilitate the transition towards greener industrial processes to produce low-carbon fuels and added value chemicals. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements Financial support for this work was gather from Spanish Ministry of Science and Spanish Ministry of Science and Innovation through the projects PLEC2021-008086, RYC2018- 024387-I as well as the project PID2019-108502RJ-I00. This work was also partially funded by the University of Seville via the VI PPIT grant scheme for talented researchers. Sergio Carrasco would also like to acknowledge the Sociedad Espa- ~ nola de Cat alisis (SECAT) for his fellowship. Funding from the European Commission through the BIOALL project (Grant Agreement: 101008058) SASOL is kindly acknowledged for providing the catalysts'supports. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijhydene.2023.03.301. references [1] Yoro KO, Daramola MO. CO2 emission sources, greenhouse gases, and the global warming effect. In: Advances in carbon capture. Elsevier; 2020. p. 3e28. https://doi.org/10.1016/b978- 0-12-819657-1.00001-3. [2] Kappelle M. United in science 2020 - UN climate report. 2020. https://doi.org/10.13140/RG.2.2.12801.28004. [3] Boucher O, Friedlingstein P, Collins B, Shine KP. The indirect global warming potential and global temperature change potential due to methane oxidation. Environ Res Lett 2009;4. https://doi.org/10.1088/1748-9326/4/4/044007. [4] Ullah Khan I, Hafiz Dzarfan Othman M, Hashim H, Matsuura T, Ismail AF, Rezaei-DashtArzhandi M, Wan international journal of hydrogen energy 48 (2023) 27907e27917 27915