Monitoring the Reaction Mechanism in Model Biogas Reforming by In Situ Transient and Steady-State DRIFTS Measurements

FULL PAPER Monitoring the reaction mechanism in model biogas reforming by in situ transient and steady-state DRIFTS measurements Luis F. Bobadilla,[a]* Victoria Garcilaso, [a] Miguel A. Centeno, [a] and José A. Odriozola[a]* Abstract: In this work, the reforming of model biogas reaction has been investigated on a Rh/MgAl2O4 catalyst. In situ transients and steady-state DRIFTS measurements were used to gain insights about the reaction mechanism involved in the activation of CH4 and CO2. It was found that the reaction proceeds through of an initial pathway in which methane and CO2 are both dissociated on Rh metallic sites, and additionally a bifunctional mechanism in which methane is activated on Rh sites and CO2 is activated on the basic sites of the support surface via formate intermediate by H-assisted CO2 decomposition. Moreover, this plausible mechanism is able to explain why the observed apparent activation energy of CO2 was much lower than that of CH4. Our results suggest that the CO2 dissociation facilitates the CH4 activation because the oxygen adsorbed species formed in the decomposition of CO2 are capable to react with the CHx species derived from the methane decomposition. Introduction A key element to achieve a sustainable development is introducing renewable energy in the chemical production chain to reduce the impact on environment and greenhouse gas (GHG) emissions. The use of biomass (particularly waste) and reuse of CO2 are two key elements enabling a new scenario for the chemical production [1]. Biogas generated by anaerobic digestion processes from the organic wastes is considered as one of the important bio-renewable resources for the next generation. The composition of biogas varies depending on the fermentation process and the feedstock, although the major constituents are methane (60%) and carbon dioxide (40%) with trace amounts of other gases [2]. Biogas reforming have received a great attention due to its environmental benefits from utilizing these two greenhouse gases and producing highly valuable synthesis gas (syngas, H2 and CO) as a feedstock [3]. Syngas can be used as a raw material for the production of synthetic fuels with low environmental impact such as hydrogen, methanol, dimethylether (DME) and higher hydrocarbons [4]. Since model biogas is composed essentially by methane and carbon dioxide, the dry reforming of methane (DRM) is one of the more interesting reaction than can be applied to produce syngas: DRM: CH4 + CO2 → 2CO + 2H2 ΔH = 247 kJ mol-1 (Eqn. 1) Several studies summarized in different reviews have showed the high efficiency of supported transition metals in this reaction [5]. Noble metals such as Rh, Ru, Pt, Pd and Ir are generally very active for DRM, without significant coke formation, due to small equilibrium constants for methane decomposition and low dissolution of carbon into their lattices [6]. Rh appears to be one of the most active and stable metals to catalyse CO2 reforming of CH4 [7]. It is generally accepted that the dissociative adsorption of both CH4 and CO2 on the metallic sites of the catalysts depend on both electronic and geometric factors. Evidence has been provided to indicate that differences in activity for DRM over a given metal dispersed on different supports may be due to metal–support interactions and/or the participation of O or OH species from the support in the metal– support interfacial region [5b]. As the adsorptive dissociation of methane, occurring on the metal surface, is one of the ratelimiting steps, the support must offer a high metal dispersion. Moreover, the support must provide the active sites for the adsorption and dissociation of CO2. Wang and Ruckenstein [7b] reported that reforming rates on Rh-supported catalysts depend on the reducibility of the supports. Methane conversion on Rh supported non-reducible oxides (Al2O3, SiO2, MgO, Y2O3, La2O3) is higher than on reducible ones (CeO2, TiO2). They concluded that MgO and Al2O3 are the most promising supports for CO2 reforming of methane. On the other hand, it is well known that increased Lewis basicity of the support leads to increased adsorption of CO2 which produces surface species able to react with carbon to form CO. For instance, it has been reported that the addition of MgO enhances the catalytic performance of alumina based catalysts as an effective promoter [8]. During the last decades, numerous studies have been devoted to understand the reaction mechanism and elementary steps at the atomic level of CO2 dry reforming over noble metal supported catalysts [5a]. However, the nature of the adsorbed species involved in the process as well as the key intermediates remain still under debate. In order to know what exactly is occurring on the catalyst surface during the reaction it is necessary to combine the direct observation by spectroscopic techniques with the computational modelling methods. Recently, Polo-Garzón et al. have reported a complete mechanistic study employing ab initio methods [9] and steady-state isotopic transient kinetic analysis (SSITKA) [10] to gain insights into the dry reforming of methane reaction on Rh-substituted lanthanum zirconate pyrochlore catalyst. Combining both experimental and theoretical studies, they found that the main reaction pathway consists in the dehydrogenation of adsorbed methane to [a] Dr. L.F.Bobadilla, V. Garcilaso, Dr. M.A. Centeno, Prof. J.A. Odriozola Instituto de Ciencia de Materiales de Sevilla Centro Mixto Universidad de Sevilla-CSIC 49 Av. Américo Vespucio, 41092 Sevilla (Spain) E-mail: [email protected] (Luis F. Bobadilla) [email protected] (José A. Odriozola) Supporting information for this article is given via a link at the end of the document FULL PAPER generate CH2 species, which then are oxygenated to CH2O species by atomic oxygen from CO2 dissociation. CH2O dehydrogenates to create CHO species, which undergoes dissociation to release a hydrogen atom and an adsorbed CO that later desorbs to yield product gas. These adsorbed hydrogens can also interact with surface bound species derived from CO2 to yield additional CO product and water via the reverse water gas shift (RWGS) reaction. Erdöhelyi et al. [7c] investigated the CH4 and CO2 dissociation as well as the reforming reaction of CH4 with CO2 over Rh/Al2O3 by means of FTIR and kinetic measurements. They suggested that the CO2 dissociation is promoted by the presence of methane. The products distribution depended strongly on the surface coverage by oxygen and carbon deposits. In-situ transient diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) has been widely employed as a useful tool to elucidate the mechanism and uncover the role of the involved species during the reaction. However, the kinetic responses of the surface species analysed under transient conditions could be characteristically distinct from those measured from steady-state experiments. On the other hand, steady-state rate measurements cannot provide direct information about the nature and coverage of surface intermediates. Therefore, it would be desirable to integrate both transient and steady-state in situ approaches to gain a more comprehensive and definitive vision of the investigated catalytic reaction. A good understanding of the reaction mechanism will provide useful guidance to achieve an optimal catalyst in terms of activity, selectivity and durability. The present contribution deals with a comprehensive study to clarify some aspects of the activity of a Mg-promoted Rh/Al2O3 catalyst toward biogas reforming reaction and understand the mechanism using transient and steady-state DRIFTS experiments. Results and Discussion The activity of Rh supported on MgAl2O4 catalyst was measured under continuous flow-conditions at atmospheric pressure using a CH4/CO2 molar ratio equal to 1.5. Figure 1 depicts the influence of reaction temperature on CH4 and CO2 conversion, and on the H2/CO molar ratio. It can be observed that both CH4 and CO2 conversion values increase with the temperature, in accordance with the strong endothermic character of the reaction. As CO2 is the limiting reactant, it is expected that CH4 conversion exhibit lower values for the whole temperature range compared to CO2. The high conversion of CH4 observed at higher temperatures could be ascribed to the methane decomposition reaction, as the predominant reaction to form hydrogen and carbon: CH4 decomposition: CH4 ↔ C + H2 ΔH = 75 kJ mol-1 (Eqn. 2) Furthermore, this reaction could be favored by the secondary reaction of reverse water gas shift (RWGS), where the hydrogen produced is consumed in the reaction with CO2 in agreement with results obtained by Parkhomenko et al. [11]. RWGS: CO2 + H2 ↔ CO + H2O ΔH = 41 kJ mol-1 (Eqn. 3) On the other hand, it is noticeable that the H2/CO molar ratio values obtained increase with the increment of temperature and it becomes slightly greater than the stoichiometric value of 1 at 750 ºC. This fact could be explained by the dominance of the cracking reaction of CH4 to form H2 and carbon at higher temperatures. Moreover, according to thermodynamic equilibrium, for CH4/CO2 molar ratios superior to 1, the amount of H2 produced increases with the temperature since CO2 is the limiting reactant and the RWGS reaction cannot simultaneously improve along with the dry reforming reaction [12]. Figure 1. Effect of reaction temperature on the conversions of CO2 and CH4 and the molar ratio of H2/CO over Rh/MgAl2O4 catalyst. Reaction conditions: WHSV = 150 L g-1 h-1, CH4/CO2 = 1.5 and 1 bar The results of catalytic activity presented in this study are in accordance with the thermodynamic model predicted by Nikoos et al. [13]. They reported that for CH4/CO2 molar ratio equal to 1.5, the methane decomposition reaction is considered to be the main reaction forming H2 and carbon at higher temperatures. In situ transient experiments were performed using two different combinations of reactants in order to observe the gas phase reactants and products and to identify the surface species present on the Rh/MgAl2O4 catalyst. We have examined the formation of surface species from gas streams containing CO2 or H2 + CO2 at 600ºC, and subsequently, we have monitored the evolution of the formed species by switching a gas stream containing CH4 every 560 s. Noteworthy that in this study we do not aim to identify which surface species are involved in H2 production but rather to investigate the mechanism of CO2 and CH4 dissociation using DRIFTS as one of the most sensitive techniques for following the interaction of CO2 and/or CH4 with the catalyst. This type of experiment enables direct correlation of specific surface species with the production of gaseous products. A series of IR spectra recorded during alternate switching between CO2 and CH4 feeds are presented in Fig.2A. The IR features obtained during CO2 exposure show that adsorption of CO2 at 600 °C leads to the formation of metallic carbonyls and CO adsorbed species as well as to CO gas-phase. This fact FULL PAPER reveals that CO2 adsorption dissociative takes place on the surface catalyst and more concretely on the metallic active sites. In addition, the rhodium-based catalyst presents several bands in the range 1350-1650 cm-1 that can be ascribed to carbonate species. It is generally accepted that CO2 chemisorption and dissociation on a transition metal surface is dominated by electron transfer and requires the formation of a carbonate precursor [14]. When CO2 stream was switched by CH4 stream, the IR bands attributed to carbonates species almost disappeared completely while that the intensities of the bands attributed to CO adsorbed and metallic carbonyls were only partially decreased. Figure 2. DRIFT spectra collected at 650ºC when the feed gas was switched from CO2/N2 to CH4/N2 over Rh/MgAl2O4 catalyst. The temporal evolution of the bands attributed to the different species can be seen more clearly in Fig. 2B and Fig. 2C. The bands appearing at 1450-1680, 1376-1545 and 1350-1637 cm-1 can be assigned to three types of carbonates surface adsorbed species on the Rh supported catalyst: bicarbonate, monodentate and bidentate carbonates, respectively [15]. The different CO2 adsorption modes depend on the chemical nature of the surface oxygen atoms. The chemisorption of CO2 reveals acidic sites where the CO2 molecule preserves its linear geometry upon adsorption, and basic sites as well as acid-base pairs when carbonate-likes species are formed [15a]. Monodentate carbonates formation involves low-coordination or isolated surface O2- anions such as those present in corners or edges, bidentate carbonate forms on Lewis-acid–base pairs (M-O2- pair site, where M is the metal cation Mg or Al), and bicarbonate formation involves surface hydroxyl groups [15d, 16]. CO adsorbed on Rh-based catalysts has been widely investigated and different CO species can be present. The IR bands at 2010 and 1860 cm-1 can be attributed to linear and bridged adsorbed CO species on Rh metallic sites, respectively [17]. There is also a weaker feature at 2084 cm-1 that can be assigned to a germinal (gem-) dicarbonyl species [18]. The presence of gem-dicarbonyl species is associated to de the oxidation of Rh0 to Rh+ cations due to the presence of hydroxyl groups on the support surface according to [19]: 2Rh + 4CO + 2Al(Mg)OH → H2 + 2Al(Mg)-O-Rh(CO)2 (Eqn. 4) This oxidation disrupts the Rh-Rh bonds generating the gemdicarbonyl species and provoking a redispersion of the Rh particles [20]. Thus, the detection of gem-dicarbonyl complexes is considered significant for the existence of highly dispersed Rh metal particles on the catalyst surface. Apart from this, it is noticeable a broad feature in the 2100-2250 cm-1 region which correspond to CO gas (inside the cell) and two weak peaks at 2170 and 2223 cm-1 that can be assigned to CO adsorbed to low coordinated Al3+ ions on the MgAl2O4 spinel structure and alumina support, respectively [21]. Figure 3. DRIFT spectra collected at 650ºC when the feed gas was switched from (CO2+H2)/N2 to CH4/N2 over Rh/MgAl2O4 catalyst. A series of spectra reported in Fig. 3A shows the evolution of the surface species formed on the Rh/MgAl2O4 catalyst during alternate switching between CO2 + H2 (1:4 molar ratio) and CH4- only feeds. During the CO2 + H2 introduction, compared with the CO2 exposure (Fig. 2A), the appearance of new IR features at 1580, 1398 and 1378 cm-1 as well as a low intense band at 2915 cm-1 (not shown) indicates the presence of adsorbed formate species on the catalyst surface. Moreover, the formation of gasphase CO reveals that RWGS reaction takes place. As has been claimed in the literature, it is assumed that the formation of CO proceeds via decomposition of formate species as intermediates in the RWGS reaction. On the other hand, it can be also observed the presence of CO-bridged adsorbed species. The FULL PAPER CO produced by RWGS reaction can interact with the Rh particles forming CO linear, CO bridged and/or gem-dicarbonyl species. The only presence of CO-bridged species suggests their inactivity and its possible role as spectator during the reaction. Figure 4 illustrates the concentration profiles of CO2 and CH4 during the modulation of both combination of reactants, CO2 vs CH4 and CO2 + H2 vs CH4, for a period of 3 modulation cycles. As can be noticed, when only CO2 was exposure the concentration of CO2 was gradually increased until the CO2 saturation level was reached. According to the DRIFTS results discussed above (Figure 2), it is evident that CO2 is dissociated to CO and oxygen active species during an initial transient or induction period where the metallic surface particles become entirely covered of CO-adsorbed species and at the same time CO2 is trapped over the surface support on the basic adsorption sites expectedly in the form of surface carbonates. Subsequently, by switching the gas feed from CO2 to CH4, the oxygen active species on metallic sites favored the CH4 decomposition to H2 and CHx species and the concentration of methane delays a few of seconds to achieve the saturation level. On the other hand, during the CO2 + H2 mixture feed, one may notice that the level of CO2 in the effluent was always below the saturation level reveling a continuous formation of CO via RWGS reaction. As the gas stream was switched to methane, the level of methane directly reached the saturation without the ongoing increase during the initial several seconds. This reveals the absence of active oxygen species capable to activate the CH4 dissociation since in presence of hydrogen predominates the H-assisted CO2 dissociation to produce CO via formate intermediates. Figure 4. CH4 and CO2 concentration profiles as a function of time when the feed gas was switched from H2/N2 or (CO2+H2)/N2 to CH4/N2 over Rh/MgAl2O4 catalyst at 650ºC. On the basis of the above understanding, two possibilities could be considered for the production of CO: i) direct dissociation of CO2 on Rh metallic particles in absence of hydrogen, and ii) H-assisted CO2 reduction by dehydroxylation of intermediate formate species formed on the support. Concerning to the former possibility, we have observed that carbon dioxide dissociation is achieved in absence of hydrogen and a large amount of CO adsorbed species on Rh is produced. By contrast, the H-assisted carbon monoxide formation can be assumed to occur through an activated formate-type intermediates, which accounts for the amount of formate species on the catalyst surface. In absence of hydrogen, our data show the activation of CO2 on Rh catalyst occur via direct dissociation to CO adsorbed and O. When methane is introduced, the adsorbed oxygen facilitates the dissociation of methane by reaction with the CHx formed on the catalyst surface. In opposite, the H-assisted activation of CO2 produces formate species adsorbed that decompose to CO and H2O. The absence of active oxygen species on the metal surface in this case enables the activation of methane and only RWGS reaction is possible. Thus, one can conclude that the CO2 dissociation results in the covering of the metallic sites with adsorbed active oxygen species which favors the methane dissociation. Figure 5. (A) DRIFT spectra obtained during CO2 reforming of methane in the temperature range from 450 to 750ºC. (B) Evolution of the peak height attributed to CO-linear, CO-bridged and formate surface species as a function of the reaction temperature. Reaction conditions: 1 bar, total flow rate of 50 mL min-1 and CH4/CO2 = 1.5 (molar ratio). In situ DRIFTS spectroscopy was used to study the chemical nature of intermediate species formed during the biogas reforming reaction in the temperature range of 450–750 ºC on the Rh/MgAl2O4 catalyst. Figure 5A shows the DRIFTS spectra recorded under steady-state conditions for the Rh/MgAl2O4 catalyst after stabilization under the reaction mixture at the indicated temperature. The spectra recorded at every temperature present intense bands ascribed to absorbed carbonyls with peaks around 1860 and in the range 2000-2050 cm-1 range corresponding to linear and bridged carbonyls on metallic rhodium, respectively, while the IR bands in the 1300- 1650 cm-1 range signify carbonate species [15a]. Furthermore, it can be appreciated the presence of intense bands at 1590-1375 cm-1 corresponding to formates [15b]. Fig. 5B reports a plot of the height peaks associated with surface formates (1595 cm-1), CO- bridged (1860 cm-1) and CO-linear (2020 cm-1) as a function of the reaction temperature. As can be observed, by increasing the reaction temperature the amount of formate species is notably FULL PAPER decreased while that of linearly adsorbed CO species are slightly decreased and the content of bridged carbonyls remains unaltered. This suggest that the bridged adsorbed CO species are spectator during the reaction. It can be understood considering that the Rh-CO bonding in bridged carbonyls is much more stable and stronger than in linear carbonyls and gem-dicarbonyls. The turnover frequency (TOF) can be defined as the number of product molecules produced or the number of reactant molecules transformed per catalyst site and per unit time. The most common manner to determine the TOF is from the relationship between the reaction rate and the surface metal exposed. The calculation of TOF is complicated since it is impossible to know exactly the real number of accessible active sites under reaction conditions. Moreover, it depends on whether the reaction is structure-sensitive or structureinsensitive. Thus, the quantification of the exact number of surface sites exposed requires an approximation [22]. Although it is well known that both the methane activation and the dissociation of CO2 are structure sensitive and promoted at defect sites such as edges and corners [5b], we have assumed that all the surface atoms of Rh exposed are active sites in the main reaction pathway. We suppose that if they are not all equally active, the turnover rate will have an average value. Accordingly, the TOF was calculated by using the following equation [23]: 𝑇𝑂𝐹 = 𝑟 · 𝑀𝑅ℎ 𝐷𝑅ℎ [𝑠−1] where r is the reaction rate expressed in molreactant converted molRh-1 s-1, MRh is the atomic weight of rhodium (102.9 g mol-1) and DRh is the metal dispersion. In all cases, rhodium dispersion was estimated to be 85 % on the basis of the cuboctahedral particles model proposed by Polisset [24] and assuming an average Rh particle size of 1 nm obtained by TEM measurements (Fig. S2 in Supporting Information). The reaction rates of CO2 and CH4 converted were estimated from the data analysis of gas-phase in the DRIFTS cell. Figure 6. Turnover frequencies (TOFs) of CO2 and CH4 consumption as a function of reciprocal temperature expressed in Kelvin units on Rh/MgAl2O4 obtained in the DRIFTS cell. Reaction conditions: 1 bar, total flow rate of 50 mL min-1 and CH4/CO2 = 1.5 (molar ratio). Figure 6 shows the CH4 and CO2 turnover rates (TOFs) plotted against reciprocal temperature expressed in Kelvin units in order to estimate the apparent activation energy. As can be observed, the apparent activation energy of CO2 consumption (30 kJ mol-1) was much lower than that of CH4 one (64 kJ mol-1). Similar values of activation energy for CH4 and CO2 were obtained by Zhang et al. [25] using a bimetallic Ni-Co/Mg(Al)O catalyst. This points to the important influence of the support in the reaction and likely this notably difference between activation energies of CO2 and CH4 was due to the presence of strong Lewis basic sites on MgAl2O4 spinel, which facilitates the CO2 activation. On the other hand, the presence of formate species observed by in situ DRIFTS indicates that RWGS reaction plays a major role in the CO2 activation and the support has an important influence in the reaction mechanism. This observation is in agreement with the results reported by Nagaoka et al. [26] for Pt-based catalysts, where the activation energy for the rate of CH4 and CO2 disappearance varies for different supports. The majority of values reported in the literature using different noble metals as active phase shows that the activation energy for methane is greater than the one estimated for CO2. This is consistent with other reactions involving methane and it is mainly due to the high energy required to activate the stable C-H bond on methane. Most of the reaction mechanism reported in the literature for the CO2 reforming of methane reaction are based on the Langmuir-Hinshelwood model. It involves a dissociative adsorption of the reactants (CH4 and CO2) followed by ratedetermining surface reaction of the adsorbed species to the final products [27]. Based on the results obtained in this work and the reaction mechanisms postulated in the literature, the following elementary reaction steps seem appropriated to describe the mechanism involved in the biogas reforming reaction: 1) CO2 (g) ↔ CO* + O* CO2 direct dissociation 2) CH4 (g) ↔ CH3* + H* CH4 dissociation on metal 3) CH3* ↔ CH2* + H* 4) CH2* ↔ CH* + H* 5) CH* ↔ C* + H* 6) H* + H* ↔ H2(g) H2 desorption 7) CO2 (g) ↔ CO2* CO2 adsorption as carbonates 8) CO2* + H* ↔ COOH* H-assisted CO2 activation 9) COOH* ↔ CO* + OH* 10) CH* + O* ↔ COH* Oxidation of CHx fragments FULL PAPER 11) CHO* ↔ CO* + H* 12) CO* ↔ CO (g) Desorption of CO 13) OH* + H* ↔ H2O* 14) H2O* ↔ H2O (g) Desorption of H2O Here, (*) correspond to an adsorption site on the catalyst surface. Among these steps, methane dissociation could be considered the rate-determining steps in the reforming reaction since it is necessary to activate the stable C-H bond [28]. Figure 7 shows an illustrative representation of the steps than can be tentatively involved in the reaction mechanism of the model biogas reforming. In a first step where the support plays a minor role, the dissociative adsorption of methane on metal Rh sites produces hydrogen and CHx species whereas CO2 is also adsorbed dissociatively on the metallic particles to form CO and O. Then, this oxygen adsorbed species reacts with CHx and leads to formation of CHxO species. Finally, the CHxO species then dissociates to form CO and H2 in a fast step. We can assume that the CO2 dissociation facilitates the CH4 activation by oxygen adsorbed formed in the decomposition of CO2. Besides we believe that the formation of oxygen species provokes a partial oxidation of the Rh metallic particles. As we mentioned above, this oxidation disrupts the Rh-Rh bonds generating the gem-dicarbonyl species and provoking a redispersion of the Rh particles. Likely this gem-dicarbonyl species are unstable and rapidly are transformed to bridged CO species. On the other hand, the hydrogen generated in the first step initiates a new bifunctional mechanism in which two types of active centers, basic sites of support and metallic particles, are involved in the reaction pathway where the role of the support is crucial. The CO2 is adsorbed on the basic sites of the support to form carbonates/bicarbonates species that subsequently are transformed to CO and H2O via formation of formate-type intermediates by H-assisted CO2 decomposition. The later process participates in the reaction mechanism through the RWGS reaction and it is obvious that the formation of formate species requires a close vicinity of the carbonate species and the metal adsorbed hydrogen. Therefore, the formation of formate species proceeds mainly at the metalsupport interface although they can migrate to the surface support where they are stably adsorbed and detected by DRIFTS [29]. Figure 7. Schematic representation of the plausible reaction mechanism for model biogas reforming over Rh/MgAl2O4 catalyst. The majority of the reaction mechanism postulated in the literature for different supported catalysts, except those developed for inert supports, are based on bifunctional pathway. This bifunctional pathway is widely accepted, with the reaction taking place on catalysts with active metal and support having either electrophilic or nucleophilic character. For active metals on inert supports, the metal catalyzes all the surface reactions and is more prone to deactivation due to carbon deposition or sintering [5a]. Conclusions On the basis of in situ transient and steady-state DRIFTS studies, a reaction mechanism focused on the activation of CO2 and CH4 on MgAl2O4-supported Rh catalyst was proposed for the model biogas reforming reaction. It was found that dissociative adsorption of CO2 takes place initially and the active oxygen species formed on the metallic sites facilitates the CH4 activation. The hydrogen generated in this first step initiates a new bifunctional mechanism in which CO2 is activated on the basic sites of the support surface via formate intermediate by H- assisted CO2 decomposition. Therefore, CO2 can be transformed into CO through two pathways: CO2 direct dissociation on metallic sites and H-assisted dissociation on metal-support interface. Experimental Section Rh (1 wt.% nominal) supported over MgAl2O4 catalyst was prepared by impregnation using the incipient wetness method. For the support synthesis, γ-alumina (Spheralite SCS505) sieved between 100-200 μm was impregnated with an alcoholic solution of Mg(NO3)2·6H2O (28 wt.% of MgO). It should be mentioned that the theoretical amount of MgO required to completely transform the alumina into the “ideal” spinel phase is ca. 28 wt.% [30]. Further, the solid was dried overnight at 100 ºC and then calcined in air 24h at 900 ºC. Regarding the active phase impregnation, an aqueous solution of Rh(NO3)3·2H2O was impregnated over the support and it was dried and calcined in air at 500 °C for 8 h (2 ºC min-1). The elemental composition of the catalyst was determined by ICP analysis. Textural properties, phase structure and morphology of the catalyst were investigated by nitrogen adsorption/desorption isotherms at liquid nitrogen temperature, powder X-ray diffraction and transmission electron microscopy, respectively. Further details on the catalyst characterization are described in the Supporting Information. The dry reforming activity of the catalyst was measured in a fixed bed continuous-flow reactor working at atmospheric pressure using a commercial Microactivity Reference® equipment. The amount of catalyst used was 50 mg diluted in crushed quartz both sieved in the 100–200 μm range. Prior to reaction, the catalyst was reduced at 750 °C for 3 h in a FULL PAPER 100 mL min−1 H2 (50 vol.% in N2) stream. The CH4/CO2 mixture, with molar ratio maintained at 1.5 simulating a model biogas, was passed through the reactor at a total flow rate of 167 ml min-1 (WHSV = 150 L g-1 h-1) and the reaction temperatures were maintained at 450, 550, 650 and 750 °C in a tubular furnace. A thermocouple was placed into the center of the catalyst bed to record the reaction temperature and to control the furnace. Gas products were analyzed on line using a microGC (Varian 4900) equipped with Porapak Q and MS-5A columns. The conversions of CH4 and CO2 were calculated based on the difference between the input and output of CH4 and CO2, respectively. In-situ DRIFTS measurements were carried out using a high temperature environmental reaction chamber supported in a Praying Mantis (Harrick) DRIFTS optical system with ZnSe windows. The spectra were collected using a Thermo Nicolet Nexus FT-IR spectrometer equipped with a liquid nitrogen cooled MCT detector at 4 cm-1 resolution and average of 64 scans. The whole optical path was purged with CO2- and H2O-free nitrogen. About 50 mg of catalyst finely ground was loaded in the cell for each measurement. The gases pass through the catalyst packed-bed, mimicking the plug-flow condition of the reaction. Different mixtures of gases could be introduced into the cell, and the gas flows were controlled by mass-flow controllers from AALBORG appropriately calibrated. All the valves, connexions and pipelines of the reaction system were heated at 100 ºC to prevent the water condensation produced during the reaction. Prior to the measurements, the catalyst was activated in situ at 750 ºC in 50% H2/N2 in a total flow rate of 50 mL min-1. For the transient experiments, the different mixtures at a total flow rate of 50 mL min-1 (10% CO2 diluted in N2, 15% CH4 diluted in N2, 40% H2 + 10% CO2 diluted in N2) were switched and introduced into the IR cell under isothermal conditions at 600ºC every 560 s. A four-way valve was installed to allow us to switch between two gas mixtures. The experiment of steady-state was performed feeding into the IR cell a mixture of 15% CH4 and 10% CO2 in nitrogen balance for a total flow rate of 50 mL min-1. The temperature range used varies from 450 to 750 °C with stepwise at 50 °C intervals and remaining at each temperature for 30 min in order to obtain steady-state conditions. The concentration of CO2 was monitored by time-resolved analysis using a VAISALA CO2 measurement meter. The concentration of CH4 was estimated by integration of the band area of νas (C-H) attributed to gas phase methane (3015 cm-1). A previous calibration by flowing CH4/N2 mixtures over α-Al2O3 at different temperatures was performed to obtain the relationship between the band area of νas (C-H) and the partial pressure of methane inside the cell. In the temperature range used, no reaction between methane and α-Al2O3 was observed, although the thermal effect must be considered. As temperature increases, the intensity of gaseous compounds DRIFTS signal decrease. With this kind of calibration, absorption coefficients can be estimated, and product quantitative analysis can be made in catalytic tests. This one, for instance, has been used in CO and CO2 methanation reactions over Rh-supported catalysts [31]. Acknowledgements Financial support for this work has been obtained from the Spanish Ministerio de Economía y Competitividad (MINECO) (ENE2013-47880-C3-2-R and ENE2015-66975-C3-2-R) and from Junta de Andalucía (TEP-8196). Victoria Garcilaso ackowledges MINECO by their FPI fellowship (BES-2013- 062806). Keywords: Biogas reforming • Reaction mechanism • In situ transient DRIFTS [1] S. Perathoner, G. Centi, ChemSusChem 2014, 7, 1274-1282. [2] S. Zinoviev, F. Müller-Langer, P. Das , N. Bertero , P. Fornasiero , M. Kaltschmitt , G. Centi , S. Miertus, ChemSusChem 2010, 3, 1106-1133. [3] aV. P. Rathod, J. Shete, P. V. Bhale, Int. J. Hydrogen Energy 2016, 41, 132-138; bP. Djinović, I. G. O. Črnivec, A. Pintar, Catal. Today 2015, 253, 155-162; cV. Chiodo, A. Galvagno, A. Lanzini, D. Papurello, F. Urbani, M. Santarelli, S. Freni, Energy Conv. Manag. 2015, 98, 252-258. [4] A. Vita, L. Pino, F. Cipitì, M. Laganà, V. Recupero, Fuel Process. Technol. 2014, 127, 47-58. [5] aD. Pakhare, J. Spivey, Chem. Soc. Rev. 2014, 43, 7813-7837; bM. C. J. Bradford, M. A. Vannice, Catal. Rev. Sci. Eng. 1999, 41, 1-42; cC. Papadopoulou, H. Matralis, X. Verykios, in Catalysis for Alternative Energy Generation (Eds.: L. Guczi, A. Erdôhelyi), Springer New York, New York, NY, 2012, pp. 57-127; dB. C. Enger, R. Lødeng, A. Holmen, Int. J. Hydrogen Energy 2012, 37, 10418-10424. [6] aY. H. Hu, E. Ruckenstein, in Adv. Catal., Vol. Volume 48, Academic Press, 2004, pp. 297-345; bJ. R. Rostrup-Nielsen, J. H. B. Hansen, J. Catal. 1993, 144, 38-49; cJ. Wei, E. Iglesia, J. Phys. Chem. B 2004, 108, 4094-4103. [7] aJ. Wei, E. Iglesia, J. Catal. 2004, 225, 116-127; bH. Y. Wang, E. Ruckenstein, Appl. Catal. A Gen. 2000, 204, 143-152; cA. Erdöhelyi, J. Cserenyi, F. Solymosi, J. Catal. 1993, 141, 287-299. [8] L. Zhang, Q. Zhang, Y. Liu, Y. Zhang, Appl. Surf. Sci. 2016, 389, 25-33. [9] F. Polo-Garzón, M. He, D. A. Bruce, J. Catal. 2016, 333, 59-70. [10] F. Polo-Garzón, D. Pakhare, J. J. Spivey, D. A. Bruce, ACS Catal. 2016, 6, 3826-3833. [11] K. Parkhomenko, A. Tyunyaev, L. M. Martinez Tejada, D. Komissarenko, A. Dedov, A. Loktev, I. Moiseev, A.-C. Roger, Catal. Today 2012, 189, 129-135. [12] N. D. Charisiou, G. Siakavelas, K. N. Papageridis, A. Baklavaridis, L. Tzounis, D. G. Avraam, M. A. Goula, J. Natural Gas Sci. Eng. 2016, 31, 164-183. [13] M. K. Nikoo, N. A. S. Amin, Fuel Process. Technol. 2011, 92, 678-691. [14] F. Solymosi, J. Mol. Catal. 1991, 65, 337-358. [15] aC. Morterra, G. Ghiotti, F. Boccuzi, S. Coluccia, J. Catal. 1978, 51, 299-313; bF. Solymosi, A. Erdöhelyi, M. Kocsis, J. Catal. 1980, 65, 428-436; cY.-G. Chen, K. Tomishige, K. Yokoyama, K. Fujimoto, J. Catal. 1999, 184, 479-490; dD. Li, R. Li, M. Lu, X. Lin, Y. Zhan, L. Jiang, Appl. Catal. B Environ. 2017, 200, 566-577; eJ. I. Di Cosimo, V. K. Dıez, M. Xu, E. Iglesia, C. R. Apesteguıa, J. Catal. 1998, 178, 499-510. [16] M. A. Al-Daous, A. A. Manda, H. Hattori, J. Mol. Catal. A Chem. 2012, 363–364, 512-520. [17] aC. M. Carbonell, C. O. Areán, Vib. Spectrosc. 1996, 12, 103-107; bM. Cavers, J. M. Davidson, I. R. Harkness, L. V. C. Rees, G. S. McDougall, J. Catal. 1999, 188, 426-430. [18] S. Trautmann, M. Baerns, J. Catal. 1994, 150, 335-344. [19] F. Solymosi, M. Pasztor, J Phys. Chem. 1985, 89, 4789-4793. [20] P. Malet, J. J. Benitez, M. J. Capitan, M. A. Centeno, I. Carrizosa, J. A. Odriozola, Catal. Lett. 1993, 18, 81-97. [21] aC. M. Carbonell, C. O. Areán, Vibr. Spectrosc. 1996, 12, 103-107; bE. Finocchio, G. Busca, P. Forzatti, G. Groppi, A. Beretta, Langmuir 2007, 23, 10419-10428. [22] M. Boudart, Chem. Rev. 1995, 95, 661-666. [23] T. R. Reina, S. Ivanova, J. J. Delgado, I. Ivanov, V. Idakiev, T. Tabakova, M. A. Centeno, J. A. Odriozola, ChemCatChem 2014, 6, 1401-1409. [24] M. Polisset, University of Paris VI 1990. [25] J. Zhang, H. Wang, A. K. Dalai, Ind. Eng. Chem. Res. 2009, 48, 677- 684. [26] K. Nagaoka, K. Seshan, J. A. Lercher, K.-i. Aika, Catal. Lett. 2000, 70, 109-116. FULL PAPER [27] Y. Kathiraser, U. Oemar, E. T. Saw, Z. Li, S. Kawi, Chem. Eng. J. 2015, 278, 62-78. [28] A. C. Luntz, J. Harris, Surf. Sci. 1991, 258, 397-426. [29] D. Heyl, U. Rodemerck, U. Bentrup, ACS Catal. 2016, 6, 6275-6284. [30] A. Penkova, L. Bobadilla, S. Ivanova, M. I. Domínguez, F. Romero- Sarria, A. C. Roger, M. A. Centeno, J. A. Odriozola, Appl. Catal. A Gen. 2011, 392, 184-191. [31] J. J. Benítez, I. Carrizosa, J. A. Odriozola, Appl. Spectrosc. 1993, 47, 1760-1766. FULL PAPER FULL PAPER Understanding the mechanism: The production of syngas from biogas represents an attractive approach for reducing greenhouse emissions. We contribute to gain a better understanding of the reaction mechanism in the biogas reforming for developing more economic and efficient catalysts. Luis F. Bobadilla,* Victoria Garcilaso, Miguel A. Centeno and José A. Odriozola* Page No. – Page No. Monitoring the reaction mechanism in model biogas reforming by in situ transient and steady-state DRIFTS measurements