Elucidación de las especies activas en la reacción de dehidroaromatización de metano a aromáticos sobre sistemas Mo-M/Zeolita: efecto de la activación y de la adición de un segundo metal

I II Elucidación de las Especies Activas en la Reacción de Dehidroaromatización de Metano a Aromáticos sobre Sistemas Mo-M/Zeolita: Efecto de la Activación y de la Adición de un Segundo Metal Elucidation of Active Species in the Dehydroaromatization Reaction of Methane to Aromatics on Mo-M/Zeolite Systems: Effect of Activation and Second Metal Addition Memoria presentada para optar al grado de Doctor por la Universidad de Sevilla. Fdo. Ángeles López Martín DIRECTORES Fdo: Alfonso Caballero Martínez Fdo: Gerardo Colón Ibáñez Sevilla, febrero de 2022. III IV INDEX Chapter 1. Introduction ................................................................................................................. 1 1.1. Energy context and utilization of natural gas ............................................................... 3 1.2. Aromatic hydrocarbons production .............................................................................. 7 1.3. Zeolites in catalysis ............................................................................................................. 8 1.5. Methane Dehydroaromatization (MDA) .......................................................................... 13 1.5.1. Generalities ............................................................................................................... 13 1.5.2. Reaction mechanism ................................................................................................. 15 1.5.3 Role of acid sites and shape selectivity ...................................................................... 17 1.5.4. Species on Mo/H-ZSM-5 ............................................................................................ 19 1.5.5. Deactivation and regeneration ................................................................................. 24 1.5.6. Approaches to improve MDA .................................................................................... 28 1.6. Aim of this work ............................................................................................................... 33 1.7. References ........................................................................................................................ 34 Chapter 2. Catalyst synthesis and experimental techniques ...................................................... 45 2.1. Catalyst preparation ......................................................................................................... 47 2.1.1. ZSM-5 based catalyst................................................................................................. 47 2.1.2. MCM-22 based catalysts. .......................................................................................... 47 2.2. Characterization techniques. ........................................................................................... 48 2.2.1. Nitrogen physisorption.............................................................................................. 48 2.2.2. X-ray powder diffraction. .......................................................................................... 51 2.2.3. Thermogravimetric analysis ...................................................................................... 52 2.2.4. Transmission electron microscopy and scanning transmission electron microscopy ............................................................................................................................................. 53 2.2.5. Temperature-programmed reduction ...................................................................... 55 2.2.6. Diffuse reflectance UV-vis spectroscopy ................................................................... 57 2.2.7. Raman spectroscopy ................................................................................................. 58 2.2.8. Inductively coupled plasma atomic emission spectroscopy (ICP-OES) ..................... 60 2.2.9. X-ray photoelectron spectroscopy ............................................................................ 60 2.2.10. Ammonia adsorption microcalorimetry .................................................................. 63 2.3. Catalytic activity study ..................................................................................................... 64 2.4. References ........................................................................................................................ 65 Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction. ...... 69 3.1. Introduction ..................................................................................................................... 71 3.2. Catalysts characterization and performance ................................................................... 74 3.2.1. Catalytic activity ......................................................................................................... 74 V 3.2.2. Surface area measurements ...................................................................................... 76 3.2.3. X-ray diffraction ......................................................................................................... 78 3.2.4. Transmission electron microscopy ............................................................................ 79 3.2.5. Thermogravimetric analysis ....................................................................................... 79 3.2.6. Raman and UV-vis spectroscopy ................................................................................ 80 3.2.7. Temperature-programmed reduction ....................................................................... 82 3.2.8. X-ray photoelectron spectroscopy ............................................................................. 84 3.2.8.1. In-situ reduction .................................................................................................. 85 3.2.8.2. In-situ carburization ............................................................................................ 87 3.3. Conclusions ....................................................................................................................... 90 3.4. References ........................................................................................................................ 90 Chapter 4. Acid treatment of Mo/ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. ............................................................................................. 95 4.1. Introduction. .................................................................................................................... 97 4.2. Catalysts characterization and performance ................................................................... 98 4.3. Identifying and correlating Mo species with catalytic performances ............................ 104 4.4. Locating the position of Mo in/on the zeolite support .................................................. 104 4.5. Visualizing the Mo particles: HAADF-STEM studies ....................................................... 107 4.6. Raman characterization ................................................................................................. 112 4.7. The role of the Mo/Alumina phase: Why higher Mo loading yields worse catalyst? .... 114 4.8. Conclusions .................................................................................................................... 116 4.9. References ...................................................................................................................... 117 Chapter 5. Bimetallic Mo-Re supported ZSM-5 through sequential impregnation. ................. 123 5.1. Introduction ................................................................................................................... 125 5.2. Bimetallic catalysts synthesis ......................................................................................... 127 5.3. Catalysts characterization and performance ................................................................. 128 5.3.1. Catalytic activity ...................................................................................................... 128 5.3.2. Surface area measurements ................................................................................... 132 5.3.3. Ammonia adsorption microcalorimetry .................................................................. 136 5.3.4. Temperature-programmed reduction .................................................................... 138 5.3.5. X-ray photoelectron spectroscopy .......................................................................... 142 5.4. Conclusions .................................................................................................................... 150 5.5. References ...................................................................................................................... 151 Chapter 6. On the effect of the catalyst reduction pre-treatment on MDA reaction ................ 157 6.1. Introduction .................................................................................................................... 159 6.2. Catalysts characterization and performance ................................................................. 160 VI 6.2.1. Catalytic activity ....................................................................................................... 160 6.2.2. X-ray powder diffraction. ......................................................................................... 166 6.2.3. Surface area measurements. ................................................................................... 167 6.2.4. Transmission electron microscopy (TEM) and high angle annular dark field (HAADF). ........................................................................................................................................... 167 6.2.5. Temperature-programmed reduction ..................................................................... 171 6.3. Conclusions ..................................................................................................................... 176 6.4. References. ..................................................................................................................... 178 Chapter 7. Synthesis and characterization of Mo/MCM-22 systems for MDA reaction .......... 181 7.1. Introduction. .................................................................................................................. 183 7.2. Synthesis. ........................................................................................................................ 186 7.2.1. MCM-22. .................................................................................................................. 186 7.2.2. Metal loading. ......................................................................................................... 188 7.3. Results and discussion .................................................................................................... 188 7.3.1. Synthesis of MCM-22 zeolites ................................................................................. 188 7.3.2. Effects of different synthesis procedures for MCM-22 on the activity, selectivity, and stability of Mo/MCM-22 catalysts .............................................................................. 190 7.3.3. Effect of Mo loading and SiC dilution. ..................................................................... 195 7.3.4. Effect of the Re incorporation to the 4% wt. Mo/MCM-22 catalyst. ...................... 202 7.4. Conclusions .................................................................................................................... 208 7.5.References ....................................................................................................................... 210 Chapter 8. General conclusions, future work and outlook. ...................................................... 217 8.1. General conclusions ....................................................................................................... 219 8.2. Future work .................................................................................................................... 221 8.3. Outlook ........................................................................................................................... 222 1 Chapter 1. Introduction Chapter 1. Introduction. 8  Cracking of oil products from crude oil distillation, like naphtha (C5- C10) or gas oil (C14-C20) fractions, breaking the larger molecules into smaller ones.  Catalytic reforming of naphtha, obtaining molecules with the same number of carbon atoms but with different structures. It uses platinum/alumina catalysts to obtain benzene, and other aromatics.  Disproportionation of toluene, which is a product from the reforming of naphtha with a lower demand. However, it can be converted to dimethylbenzene and benzene using a catalyst. Ethylbenzene is the main derivate of benzene, being used in the synthetic rubber industry to manufacture resins and latex derived from styrene, which are mainly used in the automotive industry. Therefore, ethylbenzene can be considered as the main driver of the benzene market. Other benzene derivates are cumene, cyclohexane and nitrobenzene, used for the manufacture of protective coatings, nylon fibers and detergents. In the case of other aromatics, toluene is commonly used as organic solvent, for paint thinner. However, its largest industrial use is in the production of benzene and xylene. On the other hand, P-xylene is used to produce the polyester PET, for clothing, fibres, plastic bottles, film and the production of other plastic products. Hence, the demand in the market for p-xylene being highly correlated with the demand for PET.9 1.3. Zeolites in catalysis As described in the previous section, limited fossil fuels and global warming alerted the scientific community to move towards greener design processes. On this matter, catalysis offers green chemistry advantages.10 Catalysts are essential for the reduction of air and water pollution and thereby contribute to reducing the emissions of products Chapter 1. Introduction 9 that are harmful to human health and the environment. Also, it plays an important role in society and its well-being. Without catalysis the manufacture of many materials and foodstuffs would not be possible. The worldwide annual demand for catalysts was estimated to be 850,000 tons in 2007 and was projected to increase by 3.5-4 % annually through 2012.11 Some examples involving catalysts are:  Fuels and energy uses, as an example, the Fluid Catalytic Cracking (FCC), a process developed in 1942 which produces gasoline as well as heating oil, fuel oil, propane, butane, and chemical feedstocks. Over half the world’s gasoline is currently produced using this process.  Emissions control, for example CO, NOx and hydrocarbon emissions from automobiles, sulphur based emissions in the combustion of fuels or emissions from the world’s coal fired power plants.  Polymer production of adhesives, coatings, foams, and packaging materials, textile and industrial fibres, etc.  The pharmaceutical industry, for instance, William Knowles in the 1970s found that rhodium bonded to chiral phosphine ligands could perform asymmetric catalytic hydrogenation. The method was soon developed for the commercial production of the anti-Parkinson drug.12  The food industry uses enzymes in brewing, dairy processing, in oil and fat production, for the preparation of meat or fish products or for baking specialties. Summarizing, it is estimated that most of the chemical processes (85–90 %) involve at least one catalytic step. 10 In heterogeneous catalysis, the active sites and their coordination environment and electronic structure are the most important factors that affect the performance of the Chapter 1. Introduction. 10 catalysts. To obtain a high degree of metal dispersion, a solid with adequate surface area is used as a support for an active component that is dispersed as very small particles, with dimensions of 1 to 20 nm, often referred to as nanoparticles. Some typical supports are for example Al2O3, SiO2, TiO2, CeO2 and zeolite. But only a small fraction of these nanoparticles, located at apices, edges, steps, and corners are exposed to reactant.13,14 Advances in characterization and synthesis methods are allowing us to understand and control the variables that affects the performance of a catalysts in a reaction, considering that surface structure and electronic properties can change greatly in this size range. An example is the activity dependence of the size in Au nanoparticles, being active at ambient conditions for the oxidation of CO to CO2 only in the range of 2 to 3 nm.15 Support design is also critical, and advanced catalysts have been designed in the forms of nanocrystalline and nanoporous materials, tailoring the pore structures to provide for ultrahigh surface areas and product selectivity. This knowledge, the recent approaches to nanoparticle synthesis, and the increasing need for greener chemical manufacturing processes is contributing to the design and development of new catalysts. Also, due to the change in the structure and composition of the systems under reaction conditions, it is important to characterize catalysts in operando leading to develop these types of characterisation techniques. Zeolites are crystalline aluminosilicates consisting of SiO4 and AlO4- tetrahedra, which are inner linked through oxygen atoms to give a three-dimensional network through which long channels and cavities, also called cages, are developed with diameters of molecular dimensions (0.3 to 1.4 nm). Many zeolites are known, some of them of natural origin, but the majority synthesized. Petroleum refining and basic petrochemistry are based in the use of zeolites, some examples are the industrial use in fluid catalytic cracking (FCC) of heavy petroleum distillates producing more gasoline with a higheroctane rate. The synthesis of ethylbenzene, precursor of styrene and polystyrene, is also based on zeolites, the disproportionation of toluene into benzene and xylenes and also Chapter 1. Introduction 11 the isomerization of xylenes to produce paraxylene, the chemical precursor for terephthalic acid, precursor to the polyester PET.16 By April 2021, 253 zeolite framework types had been identified.17 Zeolites can be classified according to different criteria, the most used concerns the dimensions of the pore apertures. The diameter of the channels is determined by the number of tetrahedra that form it, which are a denominated number of member rings (MR). Based on that, zeolites can be classified as:  Small pore zeolites, like chabazite, with channels formed by 8 member rings and pore diameters of 3 to 4.5 Å.  Medium pore zeolites, like MCM-22, with channels consisting of 10 member rings and pore diameters of 4.5 to 6 Å.  Large pore zeolites, like mordenite, with channels consisting of 12 member rings and pore diameters of 6 to 8 Å.  Extra-large pore zeolites, like UTD-1, with channels formed by 14 member rings and pore diameters up to 10 Å. 18 The inner pore structure depends on the composition, the zeolite type, and the cations. The latter is another aspect of zeolites to describe, zeolites have extraframework cations to compensate for the excess negative charge from the AlO4–. They are exchangeable by other cations or by a proton. In the case of the latter, the zeolite is exchanged with an ammonium ion, usually NH4NO3, and performing a thermal dissociation in which it decomposes, obtaining the acid-zeolite form. This allows the production of protonic zeolites, which are very strong solid Brönsted acids. The Brönsted acid strength is related to the Si/Al ratio which cannot be lower than one. However, each acid site becomes a somewhat weaker Brönsted site as the population of protons increases, in this way the acid strength of Brönsted acid sites can be modulated through isomorphic substitution. 19 20 Lewis acidity in a zeolite generally results from aluminum which is not tetrahedrally bound in the framework. It is usually much weaker in zeolites than Brönsted acidity. Extra-framework aluminum is often an artifact of the synthesis process but, also, can be Chapter 1. Introduction. 12 created by purpose steaming a zeolite to remove Al from the framework. This process is used to increase the Si/Al ratio of zeolite-Y for cracking catalysts. Besides the acidity of the Brönsted sites mentioned before, this 3D network gives the zeolite some interesting properties, like a high surface area and therefore a high adsorption capacity, the possibility of being used as a molecular sieve, or as a shapeselective molecule. Zeolites, with their precise pore dimensions, can discriminate reagents and products by size and shape if they present significant differences, which will influence diffusivity through channels. 21 Currently, the techniques for the preparation of zeolites involve the use of templates or organic structure directing agents (SDA), which have allowed the preparation of several new zeolites with different structures by guiding the formation and controlling the size of zeolite channels, cavities, etc. Zeolites are prepared by hydrothermal crystallization at 100-200°C in basic medium in order to facilitate the dissolution of silicon precursors, and using cationic templates that are decomposed and eliminated by washing or burning.22 In recent years, techniques to adapt the nature and properties of zeolites have been studied and developed. To do this, two strategies can be tried: one is to try to obtain the zeolite with the desired properties by modifying its synthesis procedure, and the other is to try to modify the properties with post-synthesis treatments. Some examples are the incorporation of different heteroelements into zeolites, the reduction of the zeolite crystal size or the generation of intracrystalline mesoporosity. The replacement of framework atoms in a crystalline compound such as a zeolite is also called isomorphic substitution. In that, the type of crystal does not change but other properties of the zeolite can be tuned such as the nature and catalytic function of the formed Brönsted and Lewis acid sites. For this, the critical difference for isomorphic substitution is assumed to be 15% for the radii of the ions, and the difference in electronegativity values is 0.4 a.u. on the Pauling scale. Following these criteria, zeolites Chapter 1. Introduction 13 can be isomorphically substituted by Be +, Zn2+, B3+, Al3+, Ga3+ , Fe3+ , Ge4+ , Ti4+ ,Sn4+ among others.23 Another example of the variation of the zeolites and their properties is the reduction of the crystal size at nanometric scale or the design of layered zeolites. For the former, the objective is decreasing the use of expensive SDA and seed-assisted crystallization of zeolite nanocrystals seems to be a good solution. For the latter the design of layered zeolites, has been successfully applied for MFI, BEA, and SAPO-34.24,25,26 The way to introduce mesoporosity to the crystal structure has also been studied in recent years. One option is to use hard templates, such as carbon particles, that are trapped in the synthesis but are subsequently removed by burning, thus forming pores.27 Some examples of mesoporous zeolites prepared by this method are MFI, MWT, BEA. Another option to achieve mesoporous zeolites is using different post-synthesis modifications such as dealumination, performed by steaming or mineral acid treatment, or desilication in alkaline medium.28,29 Some of these processes could help us overcome some of the drawbacks found when carrying out our work. As an example, the introduction of porosity helps reduce the limitations of mass transport, shortening the access of reactive molecules to sites assets and the output of products from zeolite channels. The possibility of adjusting the different properties of the zeolite would allow its optimization for specific uses, having a key role in the green and sustainable chemistry of the future.30. 1.5. Methane Dehydroaromatization (MDA) 1.5.1. Generalities MDA involves the selective conversion of methane to a mixture of aromatics, predominantly benzene but also naphthalene and toluene, and hydrogen. This process is as follows: Chapter 1. Introduction. 14 6𝐶𝐻4(𝑔)→ 𝐶6𝐻6 (𝑔)+9𝐻2 (𝑔) ; ∆𝐺𝑟 °=+433𝐾𝐽 𝑚𝑜𝑙 ⁄; ∆𝐻𝑟 °= +532𝐾𝐽 𝑚𝑜𝑙 ⁄ Along with this reaction, these others occur: 2𝐶𝐻4↔ 𝐶2𝐻6+ 𝐻2 2𝐶𝐻4↔ 𝐶2𝐻4+ 2𝐻2 Methane dehydroaromatization is an extremely unfavorable reaction thermodynamically,3,35,31 leading to very low equilibrium conversions and benzene yields. For that reason, high temperature is required to obtain a significant benzene production.32,33 Moreover, within this high temperature the formation of coke is thermodynamically favoured. This limitation makes necessary to develop catalysts showing increasing methane reactivity by activating the carbon-hydrogen bonds and, at the same time decreasing the carbon formation. In 1993, Wang et al.34 reported that methane could be converted to benzene and hydrogen using Mo/ZSM-5 as catalyst. Since then, many different catalytic systems have been studied and applied to MDA. Regarding the metallic phase, a better performance has been observed using Mo, W, Fe, V, Cr and Re. On the other side, concerning the supports, a lot of variety of inorganic oxides like SiO2, Al2O3 or TiO2, or also zeolites like ZSM-5, ZSM-8, ZSM-11, SAPO, MCM-41 or MCM-22 have been considered. However, Mo/H-ZSM-5 system is still the catalyst that has showed the best results in terms of conversion and selectivity for MDA reaction and, therefore, the most studied. The large number of studies carried out in these years have allowed us to obtain a wide knowledge of both this catalyst and the aromatization process itself. This fact has let reaching a certain consensus in the scientific community regarding some important aspects: 35  Mo-based systems are usually significantly more active and selective as compared to other transition metals. Chapter 1. Introduction 15  Mo/H-ZSM-5 catalysts contain highly dispersed Mo(VI)-oxo centers, which are reduced or carburized under the reaction conditions generating the active Mosites.  The Brönsted acidity of the zeolite plays a critical role in the catalytic activity, l  eading to catalyst deactivation when there is an excess of Brönsted acid sites.  Zeolite shape selectivity is critical for MDA performance. Only zeolite topologies with pore sizes of about 5.5 Å demonstrate high benzene selectivity.  There are different stages during the MDA reaction: induction, quasi-steady- state production of benzene, and deactivation. All these aspects and processes will be discussed in more depth over the following sections. 1.5.2. Reaction mechanism After many studies on MDA reaction, a consensus on the mechanism of reaction that takes place has not yet been reached. Different mechanisms have been proposed for this reaction: the mechanism via CH3• was one of the early hypotheses 36 In this mechanism the free radicals dimerize and dehydrogenate to form ethane and ethylene. Subsequently, ethylene aromatized to benzene with the aid of the available protons of H-ZSM-5 zeolite. It can be seen in the following equations: 𝐶𝐻4→𝐶𝐻3 •+𝐻• 2 𝐶𝐻3 •→𝐶2𝐻4+𝐻2 3 𝐶2𝐻4→𝐶6𝐻6+3 𝐻2 Later, bifunctional mechanism was suggested for the mechanism of methane dehydroaromatization over Mo/zeolite. This mechanism has been widely accepted in the recent literature studies.37,38 According to this, methane activation occurs at the Mo active site to form ethane, ethylene or acetylene intermediates. This is the so-called induction period in which the maximum conversion of the reaction occurs. After that, Chapter 1. Introduction. 16 conversion decreases due to carbon deposition that is taking place. It is during the induction period when Mo6+ ions are reduced, forming a metal carbide, the structure of which is the subject of much discussion (Mo2C or MoOxCy, among others). Based on this mechanism, some studies suggest that metal carbide is the active site where methane reacts to ethylene. Thus, by surface migration or by diffusion in the gas phase, ethylene reaches the Brönsted acid sites of the zeolite, where the reaction intermediates are oligomerized, cyclized and aromatized, forming benzene and other aromatic compounds. A scheme of this mechanism is shown in the next figure: Figure 3. Scheme of the bifunctional mechanism by two steps Recently, a new mechanism for the MDA reaction was proposed showing certain similarities with the reaction mechanism proposed for the methanol to hydrocarbons reaction. In this way, methane is activated on the Mo active sites, producing reactive radicals or C2 species that react with linear polyaromatic species trapped in the pores of ZSM-5 zeolite. These compounds are the called “hydrocarbon pool” and it is of great importance for the MDA reaction. Thus, according to this mechanism, benzene would be produced from secondary reactions of these compounds with the initial products of CH4 activation.39,40 However, many details about this hypothesis are still unknown, such as the condensed aromatics species that could be formed in the micropores of ZSM-5 zeolite. Chapter 1. Introduction 17 On this basis, it is clear that the aromatization process of methane is very complex, and there are still many doubts about the reaction mechanism. Furthermore, the rapid deactivation and high reaction temperatures make it difficult to determine the species formed in each of the stages. Also, depending on the catalyst used, the species can be very diverse, suggesting different mechanisms. The possibility of several mechanisms in the process cannot be ruled out either, all of which makes the determination of the reaction mechanism and the active species a challenge for the field of catalysis, as well as of great importance. 1.5.3 Role of acid sites and shape selectivity As Wang et al. demonstrated, by loading Mo species on zeolite supports MDA activity improves markedly. The zeolite Brönsted acid sites (BAS) have a very important role in the reaction, they serve as anchor points for molybdenum species inside or outside the pores. Some of the formed Mo species are also believed to be the active centers for the reaction. Therefore, the quantity of BAS, as well as its location and distribution are important factors to consider and that may be modulated if needed.41 Xu et al. characterized the interaction between ammonium heptamolybdate and NH4- ZSM-5 zeolite at different stages of catalyst preparation, concluding that by calcining at a suitable temperature, molybdenum species interact with Brönsted acid sites, migrating to channels of the zeolite. Furthermore, they showed that there is a strong interaction between the Mo species and the Al frame, leading to the extraction of the Al frame.42, 43 Using 'H MAS NMR in-situ spectroscopy to study proton species on Mo/H-ZSM-5, Bao et al. observed that the Mo species are kept in contact with the Al framework through two oxygen bridges. In addition, they observed that Brönsted acid sites attracted Mo species, that migrated to the channels anchoring themselves in the aluminum sites by an oxygen bridge. Hence, their studies supported that bifunctional catalytic mechanism for MDA Chapter 1. Introduction. 24 crucial in the development of efficient catalysts.72,73 They also studied the active site using 13C NMR after precarburizing Mo/H-ZSM-5 with carbon monoxide and observed two types of Mo (oxy-)carbidic molecular species in addition to Mo2C nanoparticles. Furthermore, they found that the carbon from the preformed carbide was incorporated into the main products of the MDA reaction, pointing to a dynamic active site.74 In agreement to Beale and others, they proposed an oxidation state between 4+ and 6+ for Mo species and explained that Mo (oxy-)carbidic species no longer transform during the induction period. This would imply that the build-up of a hydrocarbon pool is necessary for the catalyst to achieve maximum activity.75 To obtain more information about the viability of this mechanism, they performed a computational study showing that the exposure of Mo/H-ZSM-5 to the MDA conditions yields a range of reduced sites including mono- and binuclear Mo-oxo and Mo carbide complexes. These sites can catalyse the reaction via two alternative reaction channels, namely, the C−C coupling (ethylene) and the hydrocarbon-pool propagation mechanisms. Their results showed that the binuclear Mo carbide species that operate through the hydrocarbon pool mechanism are the most effective species.76 Results shown by Gascón were corroborated by Hensen et al. They combined a pulse reaction technique with isotope labelling of methane and found MDA is shown to be governed by a hydrocarbon pool mechanism in which benzene is derived from secondary reactions of confined polyaromatic carbon species with the initial products of methane activation.39 1.5.5. Deactivation and regeneration To date, rapid deactivation of catalysts in the MDA reaction remains the main drawback that hinders the final industrial application of this reaction.3,32,77,78 There are several factors to consider and that influence deactivation of the catalyst.79 Active sites are known to be easily covered with carbon deposits during the reaction, deactivating the zeolite, which can even lead to channel blockage. This may be due to the presence of an excess of Brönsted acid sites, causing condensation reactions between aromatic rings to Chapter 1. Introduction 25 continue beyond naphthalene, causing coke deposition. There is no consensus on the nature of these deactivation coke deposits, but according to the literature there are different types of carbons formed during the reaction that will affect the reaction: coke could be associated with active Mo sites or with Brönsted acid sites, it could be a carbide or pre-graphite carbon. Another factor to consider that can also lead to deactivation of the catalyst is the dealumination of the zeolite produced during the reaction, which would lead to the loss of the zeolite structure. Finally, it is indicated that given the high temperature at which the reaction is carried out, the sintering of the active phase is another factor to be considered. 60,80 Despite not reaching a consensus on this issue, modulation of certain parameters has been proposed to avoid, or at least slow down, the deactivation of the systems. On the one hand, modulation of the amount, distribution and strength of the accessible Brönsted acid centers in the zeolite is of interest, so as to avoid the formation of coke that covers the active sites or blocks the channels. On the other hand, the improvement in the diffusion in the zeolite can also benefit this task. Both approaches are usually related and the modification of one can lead to the variation of the other, so you have to be cautious and study the different properties well. The study and understanding of systems as well as the understanding of reaction is the only way to overcome this barrier, and that is why many research groups focus their work on this task. Below are some of the studies and results obtained related to deactivation, mainly of the Mo/H-ZSM-5 catalyst, as well as their regeneration, or their proposals to slow down the deactivation. Lunsford et al. distinguished three types of carbon by using X-ray photoelectron spectroscopy (XPS). These include carbide carbon in Mo2C (C 1s 282.7 eV), carbon in pregraphite carbonaceous deposits (sp-type, C 1s 283.2 eV) and carbon in carbonaceous deposits with graphite structure (C1s 284.6 eV). They explained that pre-graphite deposits are formed on the external surface of the zeolite, whereas graphite is formed Chapter 1. Introduction. 26 in the zeolite channels. 78 These results were corroborated by Bao et al. who also showed how the amount of the sp-type carbon increases during the reaction, until it gradually covers both the zeolite surface and Mo2C phase. Thus, it is considered responsible for the deactivation of Mo catalysts.3,50 Ismagilov et al. determined that during the reaction carbonaceous deposits were formed as graphite layers on the surface of Mo2C nanoparticles that were >2 nm in size, and as friable layers and disordered structure on the external surface of the zeolite. According to the energy dispersive X-Ray (EDX), X-ray diffraction analysis (XRD), and differential thermal analysis (DTA) results they obtained, they confirmed that the content of the carbonaceous deposits and the extent of their condensation increases with the time on stream, methane concentration in the feed, temperature and feed flow rate.68 Liu et al. also investigated the carbonaceous deposits on Mo/H-MCM-22 finding two types of coke: the first type located near the Brönsted acid sites, inert and what they pointed as the cause of catalyst deactivation, and the second type associated with Mo species which are reactive towards H2, also forming ethylene and benzene.81 However, they repeated the same analysis on Mo/H-ZSM-5 catalysts obtaining different results. They characterised the nature of deposited carbon on this last catalyst using several techniques including XPS, thermo gravimetric analysis (TGA), differential thermal analysis (DTA) and high-resolution transmission electron microscopy (HRTEM). Their results confirmed that the sp2/sp3 bonding ratio of coke species increases during the reaction, suggesting that the deposition of heavier aromatic-type carbon is the main reason for catalyst deactivation.82 Hensen et al. studied the deactivation of Mo/H-ZSM-5 catalysts using X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, thermo gravimetric analysis (TGA), and TEM characterisation techniques. They found two types of carbon deposition which denominated as “hard” and “soft”. This distinction was related to the location of coke species inside the pores and on the external surface, respectively. They also found Chapter 1. Introduction 27 during this study that MoO3 species act as an active oxidation catalyst, reducing the combustion temperature of a certain fraction of coke.40 At present, Beale et al. carried out different studies in which some of the factors mentioned in the literature as possible causes of the deactivation of the catalysts during the reaction were discarded. In their initial work, they suggested dealumination during reaction was minimal and that deactivation was due to carbon deposition which occurs in the zeolite outer surface.70 Subsequently, they studied Mo/H-ZSM-5 catalysts after reacting for different periods of time. They used for their study the quasielastic neutron scattering technique (QENS) and the results of this work indicated that the diffusion of methane in the samples did not appear strongly affected by the presence of the deposited coke, suggesting the access to the pore network is not significantly held back by the reaction coke.83 In another later study, they observed that Mo1.6C3 clusters grow on the outer zeolite surface, which appears to be the primary cause of catalyst deactivation. They finally concluded that the deactivation was not only due to a decrease in the amount of active Mo available on the surface but also due to a loss in shape-selectivity which leads to an increased carbon deposition at the outer shell of the zeolite crystals and eventually to pore blockage.71 Despite the studies carried out by the different research groups on the factors that influence the deactivation of the catalysts during the MDA reaction, and considering the speed with which it takes place, the importance of the regeneration of the catalysts becomes evident. Recent studies in this field include, for example, the use of O2 and H2, the use of cyclic regenerators that allow obtaining a stable and continuous benzene production, etc. Some of these examples are shown below: Zhang et al. compared the O2 and the H2 regeneration of Mo/H-ZSM-5 systems, concluding that under oxidative and high temperature conditions Mo is found as MoO3 which is easily sublimable. Also, they found that this specie easily reacted with the framework aluminium of the zeolite to form inactive Al2(MoO4)3. They concluded that the regeneration of coked Mo/H-ZSM-5 has to be performed in a reductive Chapter 1. Introduction. 28 environment, demonstrating a full recovery of the benzene formation activity of Mo/H- ZSM-5 at its working temperatures using H2 for regeneration.84 Hensen et al. investigated the structural and textural stability of H-ZSM-5 as a function of the Mo loading in air at high temperature (550–700 °C). They found differences in the behaviour of catalysts with low Mo load (1–2 wt% Mo), and catalysts with higher Mo load. For the first case, the catalysts presented a high stability making them suitable candidates for isothermal (700 °C) reaction - air regeneration protocol of methane dehydroaromatization. However, at higher loading, catalysts rapidly lost their initial activity due to the irreversible damage to the zeolite framework, as a result of the aluminium molybdate formation after reaction of mobile MoO3 species with framework Al.85 Kim et al. also studied a series of Mo/H-ZSM-5 catalysts with different Mo loadings and applied reaction-regeneration cycles under both oxidative and reductive regeneration conditions, also using different temperatures for it. Their results revealed that, for 5%Mo/H-ZSM-5 the most active of their catalyst, reductive regeneration required a temperature higher than 700 °C for coke removal, which resulted in thermal degradation of the catalyst. However, oxidative regeneration at 450 °C is the most efficient for maintaining catalytic activity and stability, while at higher temperatures of 550–850 °C, irreversible deactivation was observed. They observed that the selective recovery of Brönsted acid sites near Mo sites, as opposed to the isolated acid sites were sufficient to restore the catalytic activity in terms of benzene formation. 86 1.5.6. Approaches to improve MDA Many studies have been devoted to the optimization of the MDA reaction system. Some of the strategies followed by the different research groups have been the creation of mesopores in zeolite, the reduction of the size of its crystals, the co-feeding of other gases with methane to reduce the formation of coke, as well as the creation of new catalysts, the “chemical looping” or the use of different types of reactors. Chapter 1. Introduction 29 However, carbon formation seems inevitable, but the experimental findings show this can be kinetically limited using H2, oxidants, CO2 or steam in the CH4 feed. 31 On the ZSM- 5 catalysts, various approaches have been attempted: Iglesia et al. showed that the stability of MoCx/H-ZSM-5 catalysts can be improved by adding CO2 to CH4 feeds at low concentrations. They explained the improvement as a result of the H2 formed by CO2–CH4 reactions.87 Liu and co-workers also found that H2 can be added to the feed, finding that it minimizes carbon formation, particularly on the Bronsted acid sites of the zeolite.88 The addition of O2 or NO also has also been studied during these years. Au el al. used these gases as oxidant and their results showed there was an optimum, below which presumably coke is not oxidized sufficiently rapidly, and above which oxidation of the active Mo carbide decreases the activity.89 Another method used to try to avoid deactivation of the catalyst was used by Lin et al. They showed that by hydrothermal post-synthesis treatment in NaOH solution, a commercial zeolite could be restructured, modifying the microporous structure and the acidity. The process stabilized the framework with high crystallinity and prevented the creation of substantial extra pores. Consequently, the restructured Mo/H-ZSM-5 catalyst showed significant improvement in both catalytic durability and selectivity to aromatics in the methane dehydroaromatization reaction.90 Over the last few years, many researchers have developed new catalysts with the aim of finding a more robust one, which allows its use in industry for the MDA reaction. Some of them are based on zeolites and some are not: Yung et al. used Ag−Ga/H-ZSM-5 catalyst for co-aromatization of methane and propylene at 400 °C. Their results showed an excellent catalytic performance, which they Chapter 1. Introduction. 30 explained was due to a better dispersion of Ag and Ga on the zeolite surface and moderate amount of strong Brönsted and Lewis surface acid sites. They also suggested the existence of a synergetic effect where methane was activated and converted into aromatics with the help of a coexisting olefin. However, benzene was not the major product.91 Veser et al. studied the nature of the active species in Fe/H-ZSM-5 catalysts prepared by different methods. They showed that coking is reduced with increasing Fe dispersion and that atomically dispersed Fe2+ inside the zeolite micropores constitutes the selective species for MDA reaction. However, the presence of Brönsted acid sites in the zeolite results in continued coke formation.92 Xiaoguang Guo et al. studied another type of catalyst, like the Fe©SiO2 which showed excellent stability, methane conversion, and aromatic selectively in MDA reaction. Neither coking nor deactivation happened in a test of 60 h.93 Li et al. studied GaN nanowires, which were first studied for direct conversion of methane into benzene and molecular hydrogen under ultraviolet illumination. During this work, they observed that GaN exhibited high activity toward methane C−H bond activation. In a second study they analyzed the thermal catalytic activity of GaN. This was a commercial catalyst consisted of GaN powders with a wurtzite crystal structure. Their results showed a superior stability and reactivity for converting light alkanes, including methane into benzene with high selectivity. The catalyst was also highly robust and was used repeatedly without noticeable deactivation.94,95 The fixed bed reactor has disadvantages in terms of industrial use; however, it remains the most widely used reactor for the study of the methane aromatization reaction. Adopting a type of reactor that enables an instantaneous regeneration of coked Mo/H- ZSM-5 catalyst or selectively removing hydrogen for shifting the chemical equilibrium to the products side are some of the strategies studied. Like this, one of the strategies followed to improve thermodynamic limitation is through the method that combines MDA with chemical loop. Xu et al. developed this strategy by coupling MDA reaction Chapter 1. Introduction 31 with chemical looping to achieve reactive separation of H2 from the MDA products by using a Fe3O4/FeO redox pair. Their results indicated that the yield of aromatics was significantly increased.96 Regarding the use of different reactors, the application of membrane reactor, fluidized bed reactors, or even plasma reactor are some examples: Kjølseth et al. integrated an electrochemical BaZrO3-based membrane in the reactor which exhibited proton and oxide ions conductivity. Their results demonstrated high aromatic yields and a better stability of the catalyst reducing the coke production rate.97 Xu et al. proposed a two beds type of circulating fluidized bed reactor system for industrial application of the MDA reaction. In this system, one catalyst bed serves as CH4 convertor and the other as catalyst regenerator, and under operation catalyst particles are continuously circulated between the two beds to realize continuous regeneration of coked catalyst and hence ensure the simultaneous realization of industrially required continuous operation and high catalyst activity.84 Kim et al. decreased the reaction temperature by using a dielectric barrier discharge (DBD) plasma-catalyst hybrid system. When the plasma was applied to the Mo/H-ZSM- 5 catalytic reactor, methane was converted at much lower temperatures than the conventional process. They showed as the benzene selectivity could be improved by enhancing the interaction between plasma and catalyst, and the catalyst deactivation was effectively prevented when H2 gas was mixed with the reaction feed.98 Some research demonstrated that by incorporating an additional metal ion into the Mo/H-ZSM-5 system, catalytic activity could be enhanced. These are some examples: Xu et al. doped several metal species, like Fe, Cu, Co, Mn, on Mo/H-ZSM-5 catalyst finding that catalytic performances and stability were promoted when the catalyst was doped with Fe due to the generation of carbon nanotubes induced by iron species. They explained these carbon nanotubes could suppress the formation of surface coke.99 Chapter 1. Introduction. 32 Scurrell et al. doped Pt and Sn on Mo/H-ZSM-5 catalyst for methane aromatization. Their catalytic results of Pt−Mo/HZSM-5 catalyst showed an increase of methane conversion and a reduction of coke deposition. This last phenomenon happened as well when they doped the catalysts with a low load of tin. Furthermore, they co-impregnated the Mo/H- ZSM-5 catalyst with Pt and Sn finding that these catalysts exhibited higher selectivity to aromatic and lower selectivity to carbon deposition. They observed that the order of impregnation affected to the catalytic results and the structural and electronic properties of the catalyst.100 Martinez et al. exchanged H+ in H-ZSM-5 with alkali (Na+, Cs+) and alkaline-earth (Ca2+, Mg2+) cations in order to modulate both strength and density of Brönsted acid sites. They found that partially exchanging H+ with Na+ reduced the density of Brönsted acid sites showing lower coke formation and higher aromatics selectivity.101 Currently, Hensen and co-workers have studied other types of strategies to improve the MDA efficiency, such as introducing O2 pulses into methane feeding or performing the reaction at high pressure. In the first strategy they observed that a periodic supply of short pulses of oxygen into the methane feed led to doubling of the benzene yield, decreasing the rate of coke formation four times. Their results showed how the rapid cycling between the oxidic and carbidic forms of molybdenum did not affect the catalyst framework, and neither is molybdenum lost by sublimation. They explained these facts were due to the stabilization of molybdenum-oxo complexes on cation-exchange sites of the zeolite. 102 Using the second strategy, Hensen et al. showed that catalyst productivity can be improved by nearly an order of magnitude by raising the reaction pressure to 15 bar, demonstrating that performing the MDA reaction at elevated pressure leads to an increased rate of surface species hydrogenation, lower overall coke selectivity, and higher methane conversion capacity. Their results showed that the improvement of the catalytic performance at increased pressure is independent of the Mo loading, reaction temperature and methane space velocity. Operando XAS results evidenced that, although the Mo-oxide precursor is reduced easier under elevated pressure, the structure of the active sites during the actual reaction is independent of pressure. They explained this improvement in the performance as the result of a faster Chapter 1. Introduction 33 coke hydrogenation at high pressure, which resulted in a lower coke selectivity and better utilization of the zeolite micropore space. 103 These have been some of the studies carried out to improve the results obtained by performing the MDA reaction. As we have seen, different strategies have been developed; however, despite all the studies carried out, the low efficiency and the deactivation of carbon deposits continue to prevent this reaction from being commercialized. 1.6. Aim of this work The main objective of this thesis is to understand the guiding principles of Mo/zeolite catalyst systems in the aromatization reaction of methane. Through the catalytic study and the characterization of these, it is intended to expand the knowledge of the active species in the reaction, the processes involved in the rapid deactivation of the catalyst, etc. This study would facilitate the development of new, more efficient catalysts that would allow the reaction to be commercialized. To achieve this, several partial objectives have been met: 1. Investigate the effect of preparing Mo/ZSM-5 catalyst with different loadings on the structure and catalytic performance for methane dehydroaromatization in order to optimize the system and obtain the best performance for the studied reaction (Chapter 3). 2. Study the nature of the active species in the Mo/zeolite catalyst as well as their influence on the deactivation mechanism (Chapter 3, 4 and 7). 3. A deep understanding of the effect of the addition of rhenium as a promoter of the classic Mo/zeolite system on the structure and activity performance for MDA reaction is also a subject of this work. (Chapter 5 and 7). Chapter 1. Introduction. 40 studies. Kinet. Catal. 47, 389–394 (2006). 68. Matus, E. V. et al. Study of methane dehydroaromatization on impregnated Mo/ZSM-5 catalysts and characterization of nanostructured molybdenum phases and carbonaceous deposits. Ind. Eng. Chem. Res. 46, 4063–4074 (2007). 69. Jie Gao, Yiteng Zheng, Jih-Mirn Jehng, Y. T. & Israel E. Wachs, S. G. P. Identification of molybdenum oxide nanostructures on zeolites for natural gas conversion. Science. 348, (2015). 70. Lezcano-González, I. et al. Molybdenum Speciation and its Impact on Catalytic Activity during Methane Dehydroaromatization in Zeolite ZSM-5 as Revealed by Operando X-Ray Methods. Angew. Chemie - Int. Ed. 55, 5215–5219 (2016). 71. Agote-Arán, M. et al. Determination of Molybdenum Species Evolution during Non-Oxidative Dehydroaromatization of Methane and its Implications for Catalytic Performance. ChemCatChem 11, 473–480 (2019). 72. Vollmer, I. et al. Relevance of the Mo-precursor state in H-ZSM-5 for methane dehydroaromatization. Catal. Sci. Technol. 8, 916–922 (2018). 73. Vollmer, I. et al. Applied Catalysis A , General Quantifying the impact of dispersion , acidity and porosity of Mo / HZSM-5 on the performance in methane dehydroaromatization. Appl. Catal. A, Gen. 574, 144–150 (2019). 74. Vollmer, I. et al. On the dynamic nature of Mo sites for methane dehydroaromatization. Chem. Sci. 9, 4801–4807 (2018). 75. Vollmer, I. et al. A site-sensitive quasi-in situ strategy to characterize Mo/HZSM- 5 during activation. J. Catal. 370, 321–331 (2019). 76. Li, G., Vollmer, I., Liu, C., Gascon, J. & Pidko, E. A. Structure and Reactivity of the Mo/ZSM ‑ 5 Dehydroaromatization Catalyst: An Operando Computational Study. ACS Catal. 9, 8731–8737 (2019). 77. Tang, P., Zhu, Q., Wu, Z. & Ma, D. Methane activation: The past and future. Energy Environ. Sci. 7, 2580–2591 (2014). 78. Weckhuysen, B. M., Rosynek, M. P. & Lunsford, J. H. Characterization of surface Chapter 1. Introduction 41 carbon formed during the conversion of methane to benzene over Mo/H-ZSM-5 catalysts. Catal. Letters 52, 31–36 (1998). 79. Zheng, H. et al. Methane dehydroaromatization over Mo/HZSM-5: A study of catalytic process. Catal. Letters 111, 111–114 (2006). 80. Tempelman, C. H. L. & Hensen, E. J. M. On the deactivation of Mo/HZSM-5 in the methane dehydroaromatization reaction. Appl. Catal. B Environ. 176–177, 731– 739 (2015). 81. Liu, H. et al. The chemical nature of carbonaceous deposits and their role in methane dehydro-aromatization on Mo/MCM-22 catalysts. Appl. Catal. A Gen. 236, 263–280 (2002). 82. Liu, B. S., Jiang, L., Sun, H. & Au, C. T. XPS, XAES, and TG/DTA characterization of deposited carbon in methane dehydroaromatization over Ga-Mo/ZSM-5 catalyst. Appl. Surf. Sci. 253, 5092–5100 (2007). 83. Silverwood, I. P., Arán, M. A., González, I. L., Kroner, A. & Beale, A. M. QENS study of methane diffusion in Mo/H-ZSM-5 used for the methane dehydroaromatisation reaction. AIP Conf. Proc. 1969, (2018). 84. Xu, Y., Song, Y. & Zhang, Z. G. A binder-free fluidizable Mo/HZSM-5 catalyst for non-oxidative methane dehydroaromatization in a dual circulating fluidized bed reactor system. Catal. Today 279, 115–123 (2017). 85. Kosinov, N. et al. Stable Mo/HZSM-5 methane dehydroaromatization catalysts optimized for high-temperature calcination-regeneration. J. Catal. 346, 125–133 (2017). 86. Han, S. J. et al. Non-oxidative dehydroaromatization of methane over Mo/H-ZSM- 5 catalysts: A detailed analysis of the reaction-regeneration cycle. Appl. Catal. B Environ. 241, 305–318 (2019). 87. Lacheen, H. S. & Iglesia, E. Stability, structure, and oxidation state of Mo/H-ZSM- 5 catalysts during reactions of CH4 and CH4-CO2 mixtures. J. Catal. 230, 173–185 (2005). Chapter 1. Introduction. 42 88. Zheng Liu, Michael A. Nutt, and E. I. The effects of CO2, CO and H2 co-reactants on methane reactions catalyzed by Mo/H-ZSM-5. Catal. Letters 81, (2002). 89. Tan, P. L., Leung, Y. L., Lai, S. Y. & Au, C. T. Methane aromatization over 2 wt% Mo/HZSM-5 in the presence of O2 and NO. Catal. Letters 78, 251–258 (2002). 90. Song, Y., Sun, C., Shen, W. & Lin, L. Hydrothermal post-synthesis of HZSM-5 zeolite to enhance the coke-resistance of Mo/HZSM-5 catalyst for methane dehydroaromatization. Catal. Letters 109, 21–24 (2006). 91. He, P., Gatip, R., Yung, M., Zeng, H. & Song, H. Co-aromatization of olefin and methane over Ag-Ga/ZSM-5 catalyst at low temperature. Appl. Catal. B Environ. 211, 275–288 (2017). 92. Lai, Y. & Veser, G. The nature of the selective species in Fe-HZSM-5 for nonoxidative methane dehydroaromatization. Catal. Sci. Technol. 6, 5440–5452 (2016). 93. Guo, X. et al. Direct, nonoxidative conversion of methane to ethylene, aromatics, and hydrogen. Science. 344, 616–619 (2014). 94. Li, L., Fan, S., Mu, X., Mi, Z. & Li, C. J. Photoinduced conversion of methane into benzene over GaN nanowires. J. Am. Chem. Soc. 136, 7793–7796 (2014). 95. Li, L. et al. Thermal non-oxidative aromatization of light alkanes catalyzed by gallium nitride. Angew. Chemie - Int. Ed. 53, 14106–14109 (2014). 96. Brady, C., Murphy, B. & Xu, B. Enhanced Methane Dehydroaromatization via Coupling with Chemical Looping. ACS Catal. 7, 3924–3928 (2017). 97. Morejudo, S. H. et al. Direct conversion of methane to aromatics in a catalytic coionic membrane reactor. Science. 353, 563–566 (2016). 98. Park, S. et al. Plasma-Assisted Non-Oxidative Conversion of Methane over Mo/HZSM-5 Catalyst in DBD Reactor. Top. Catal. 60, 735–742 (2017). 99. Xu, Y., Wang, J., Suzuki, Y. & Zhang, Z. G. Improving effect of Fe additive on the catalytic stability of Mo/HZSM-5 in the methane dehydroaromatization. Catal. Today 185, 41–46 (2012). Chapter 1. Introduction 43 100. Tshabalala, T. E., Coville, N. J., Anderson, J. . & Scurrell, M. S. Dehydroaromatization of methane over Sn–Pt modified Mo/H-ZSM-5 zeolite catalysts: Effect of preparation method. Appl. Catal. A Gen. 485503, 218–226 (2015). 101. Martínez, A. & Peris, E. Non-oxidative methane dehydroaromatization on Mo/HZSM-5 catalysts: Tuning the acidic and catalytic properties through partial exchange of zeolite protons with alkali and alkaline-earth cations. Appl. Catal. A Gen. 515, 32–44 (2016). 102. Kosinov, N., Coumans, F. J. A. G., Uslamin, E., Kapteijn, F. & Hensen, E. J. M. Selective Coke Combustion by Oxygen Pulsing During Mo/ZSM-5-Catalyzed Methane Dehydroaromatization. Angew. Chemie - Int. Ed. 55, 15086–15090 (2016). 103. Kosinov, N. et al. Reversible Nature of Coke Formation on Mo/ZSM-5 Methane Dehydroaromatization Catalysts. Angew. Chemie - Int. Ed. 58, 7068–7072 (2019). Chapter 2. Catalyst synthesis and experimental techniques In this chapter the experimental techniques used during the development of this thesis have been defined. The equipment used for the characterization of the samples, as well as the experimental setups used to determine the catalytic activity, are also briefly detailed. The choice of these techniques is due, on the one hand, to the intention of providing valuable information on the physical-chemical, structural, textural and morphological properties of the samples, and on the other hand, to the availability of them in our facilities or in other accessible centers. Chapter 2. Catalyst synthesis and experimental techniques 47 2.1. Catalyst preparation 2.1.1. ZSM-5 based catalyst Commercial ZSM‐5 (SiO2/Al2O3=23, Alfa Aesar) was also used as support for the catalysts. The commercial ZSM‐5 was first calcined at 550 °C for 3 h in order to have its protonated form (HZSM-5). For methane dehydroaromatization, molybdenum was loaded onto the HZSM-5: the dried zeolite was impregnated with a solution containing the stoichiometric amounts ammonium heptamolybdate tetrahydrate (Aldrich) in water and kept under stirring for 20 hours. After that, water was eliminated in a rotary evaporator. After impregnation, the samples were dried for 4 h, and calcined in air at 550 °C for 3 h, at a rate of 5°C /min. The impregnation procedure is equivalent for other metals we use for MDA reaction, like rhenium and also for different types of zeolites. 2.1.2. MCM-22 based catalysts. MCM-22 was obtained by hydrothermal sythesis, the classic method or microwave assisted. The classic method consists of PTFE-lined stainless-steel reactors of 170 mL that can be introduced inside a heater at the required temperature. The second method consists of a set of hydrothermal reactors of smaller size which rotate and are heated using microwave radiation whith internal agitation in each reactor (Milestone EthosEZ Microwave). For the preparation, the following compounds were used: hexamethylenimine (HMI, 99 %, Aldrich), silicic acid (SiO2·H2O, 100-200 mesh, Sigma- Aldrich), sodium aluminate (NaAlO2·2H2O, Al2O3: 50-56 % Na2O: 37-45 %, Sigma-Aldrich), sodium hydroxide (NaOH, Merck), and deionized water. Typically, the relation for the synthesis mixtures were SiO2/Al2O3= 30, Na/SiO2= 0.18, HMI/SiO2=0.35 and H2O/SiO2=19.5. The general procedure consists on dissolving the sodium hydroxide and the sodium aluminate in deionized water. To this solution, the amount of silicic acid and hexamethyleneimine were added, and the resulting mixture was stirred until it was Chapter 2. Catalyst synthesis and experimental techniques 48 visibly homogeneous. After this, the resulting gels were introduced in the hydrothermal reactor and crystallization of the dry gel was carried out at 155 °C, for different periods of time depending on the version of the method used and under the autogenous pressure. After quenching the autoclaves in cold water to room temperature, the samples were filtered, washed, dried at 110 °C overnight and calcined in air at 550 °C for 8 h, at a rate of 2 °C/min. The calcined samples were ion-exchanged in a 1 M NH4NO3 solution for 12 h, and after washed and filtered. This procedure was repeated three times. After drying overnight at 110 °C, the solids were calcined at 550 °C for 3 h in air to finally obtain the H-form zeolite. Molybdenum was supported on the zeolite by means of the impregnation method described before, obtaining catalysts of different metal loads supported on MCM-22. 2.2. Characterization techniques. 2.2.1. Nitrogen physisorption. Surface area and pore structure are properties of key importance for any catalyst or support material. Gas adsorption measures are widely used for the characterization of a wide variety of porous solids, such as oxides, carbons, zeolites or organic polymers. At physisorption, a gas interacts with a solid surface by a van der Waals bond, weaker than a typical chemical bond. The formation of many layers of adsorbed molecules which is achieved due to the van der Waals interactions between adsorbed molecules is not much different from the van der Waals interaction between the molecules and the surface. The amount of gas adsorbed at a given temperature for a relative gas pressure depends on different factors such as specific surface, porosity, etc, and it can be represented as a graph called adsorption isotherm. Chapter 2. Catalyst synthesis and experimental techniques 49 The IUPAC classification of adsorption isotherms is illustrated in Figure 1. The six types of isotherm (IUPAC classification) are characteristic of adsorbents that are microporous (type I), nonporous or macroporous (types II, III, and VI), or mesoporous (types IV and V). Figure 1. The IUPAC classification of adsorption isotherms for gas. Several mathematical models have been developed to describe the equilibrium adsorption capacity of adsorbents. Among the most common is the Brunauer, Emmet y Teller (BET), it is based on the Langmuir model and extended to multilayers adsorption allowing the determination of the specific surface of a solid.1,2 𝑃 𝑉𝑎𝑑𝑠(𝑃0−𝑃)=1 𝑉𝑚𝐶+𝐶−1 𝑉𝑚𝐶·𝑃 𝑃0 (Eq. 1) Where P is the partial pressure of the adsorbed gas, P0 the vapor pressure of the gas, Vm the adsorbed gas volume when a monolayer is created, C is a constant, and Vads the adsorbed gas volume at a pressure P. If it is represented 𝑃 𝑉𝑎𝑑𝑠(𝑃0−𝑃) vs 𝑃 𝑃0 a straight line will be obtained, with a slope and an intercept that will allow Vm and C to be obtained. Using the value of Vm the specific surface BET can be obtained using this equation: Chapter 2. Catalyst synthesis and experimental techniques 56 A reducing gas mixture (typically hydrogen diluted in argon, but in our case also methane diluted in helium) flows over the sample, which is placed in a fixed-bed reactor while the temperature is increased according to a linear temperature programme. The difference between the inlet and outlet concentration of the gas stream is measured using a thermal conductivity detector and/or a mass spectrometer. The resulting TPR profile contains qualitative information on the oxidation state of the reducible species present, but the technique is intrinsically quantitative and the information obtained is of a kinetic nature and can be correlated with catalytic behaviour. Like this, it is possible to determine the energetic nature of catalysts and to identify the temperature required to activate the metallic phase.9 Commercial patterns are used for calibration, converting the TCD signal into active gas concentration and integrating the area under the concentration vs time graph. Knowing the stoichiometry of the reduction reaction, it is easy to determine the consumed hydrogen. Figure 3. Simplified representation of the TPR setup. In our case, TPR analysis were carried out using a Quantachrome Chemstar instrument equipped with a thermal conductivity detector and mass spectrometer. About 35 mg of catalyst was first heated to 150 °C in an inert flow of argon at 30 mL/min for 60 min Chapter 2. Catalyst synthesis and experimental techniques 57 previous to each experiment for water desorption. The experimental conditions were chosen to fulfil the resolution conditions. 2.2.6. Diffuse reflectance UV-vis spectroscopy UV-visible spectroscopy is based on the electronic absorption of electromagnetic radiation when it interacts with matter in the wavelength range between 190 nm and 800 nm. For powder samples, the diffuse reflectance measure is used because in powders of randomly oriented particles part of the incident light goes back at all angles into the hemisphere of provenance of the light. The phenomenon resulting from the reflection, refraction, diffraction, and absorption by particles oriented in all directions is called diffuse reflection, in contrast with regular reflection from a plane phase boundary. Figure 4. Projection of the spectrometer beam into the sample where it is reflected, scattered and transmitted through the sample material. The ultraviolet-visible spectroscopy by diffuse reflectance (UV-Vis DRS) provides information on the environment of the metallic and organic species present in the material, as long as they present transitions between molecular levels separated by energies of the order of the region UV and / or visible from electromagnetic radiation.8 In a UV-Vis DRS spectrum, the ratio of the light scattered from an infinitely thick layer and the scattered light from an ideal non-absorbing reference sample is measured as a function of the wavelength. The resulting spectrum is usually obtained as a percentage Chapter 2. Catalyst synthesis and experimental techniques 58 of reflectance against the wavelength, establishing as 100 % of reflectance that was obtained for the non-absorbing reference sample. The studies of diffuse reflectance spectra in this work were obtained on a UV–vis scanning spectrophotometer Agilent Cary 300, equipped with an integrating sphere, using Spectralon® as reference material and ZSM-5 as background. UV-vis spectra were performed in the diffuse reflectance mode (R) and transformed to a magnitude proportional to the extinction coefficient (K) through the Kubelka-Munk function, F(R∞). For the sake of comparison, all spectra were arbitrary normalized in intensity to 1. 2.2.7. Raman spectroscopy Raman spectroscopy is a technique based on inelastic monochromatic light scattering in the visible, near infrared, or near ultraviolet range, and it is a powerful tool for determining chemical species. It exploits the existence of Stokes and Anti-Stokes scattering to examine molecular structure: when a laser falls upon the sample, most of it is scattered elastically (Rayleigh scattering), but a small amount of light is scattered inelastically, and therefore has a change in its frequency. That is, the electronic cloud of the molecule is excited and goes into a virtual excited state. To return to the ground state it emits a photon, however, since the dispersion is inelastic, energy has been lost or gained, and the molecule does not return to the same state, but goes into a rotational or vibrational other than the initial one. The difference in energy between these states is what modifies the frequency of the emitted photon and is characteristic of the chemical nature and the physical state of the sample. This is typically measured as the change in the wavenumber from the incident light source. Chapter 2. Catalyst synthesis and experimental techniques 59 Figure 5. Vibrational energy-level diagram showing the states involved in Raman scattering phenomenon. Raman spectroscopy requires a change in polarizability and allows obtaining complementary spectral information on homonuclear molecules. This is a measure of the deformability of a bond in an electric field and depends on how easy it is for the electrons in the bond to be displaced, inducing a temporary dipole. 10 Figure 6. Scheme of operation of a Raman equipment. Micro-Raman measurements were performed using a LabRAM Jobin Yvon spectrometer equipped with a microscope. Laser radiation (λ = 532 nm) was used as excitation source Chapter 2. Catalyst synthesis and experimental techniques 60 at 5 mW. All measurements were recorded under the same conditions (2s of integration time and 30 accumulations) using a 100x magnification objective and a 125 mm pinhole. 2.2.8. Inductively coupled plasma atomic emission spectroscopy (ICP-OES) Inductively coupled plasma atomic emission spectroscopy is a spectral method used to quantitatively determine the elemental composition of samples at trace and ultra-trace levels. ICP-AES works by the emission of photons from analytes that are brought to an excited state by the use of high-energy plasma. The sample, in liquid form, it is transported to the nebulizer system using a peristaltic pump where it is transformed into an aerosol thanks to the action of argon gas. This aerosol is led to the ionization zone consisting of a plasma in which temperatures of up to 7000 °C can be reached. The plasma source is induced when passing argon gas through an alternating electric field that is created by an inductively couple coil. The atoms are ionized or excited and emit radiation of a wavelength that is characteristic of each element in order to return to their ground state. The radiation is separated according to its wavelength and a detector measures the intensity of each radiation, obtaining like this the concentration of each element in the sample.11 The elemental determination of the samples was performed using a SPECTRO SectroBLUE spectrometer equipped with a CCD detector. Previously, the samples were digested in an EthosOne Microwave, firstly using a mixture of acids and heating at 240 °C, and then in boric acid at 220 °C to end the digestion. 2.2.9. X-ray photoelectron spectroscopy XPS is based on the photoelectric effect and it provides information on the elemental composition, oxidation state and, in favourable cases, on the dispersion or chemical Chapter 2. Catalyst synthesis and experimental techniques 61 environment. When a sample is irradiated with X-ray, an electron can be ejected. This happens if the energy of the X-ray is bigger than the binding energy of the electron. The kinetic energy of the ejected electron will follow this equation: Ek = hν - Eb – ϕ (Eq. 6.) Where Ek is the kinetic energy of the photoelectron, h is Planck’s constant, ν is the frequency of the exciting radiation, Eb is the binding energy of the photoelectron with respect to the Fermi level of the sample and ϕ is the work function of the spectrometer. Figure7. Scheme of an XPS spectrometer. By measuring the number of ejected electrons (intensity) and the kinetic energy, a xps spectrum is obtained, and it provides information about elemental composition of the sample, oxidation state, dispersion of the element and chemical environment. The change of this in the sample produces a chemical shift, which consists of a change in binding energy of a core electron of an element. To remove or to add electronic charge as a result of changes in bonding will alter the shielding: the withdrawal of valence electron charge produces an increase in binding energy (oxidation), and the addition a decrease in binding energy. Photoelectrons do not travel far in solids due to interactions with atoms resulting in inelastic scattering. The number of photoelectrons that reach a surface are described by the Beer-Lambert law: Chapter 2. Catalyst synthesis and experimental techniques 62 N= N0 e-x/ λ sinϴ (Eq. 7.) Where λ is the inelastic mean free path (IMPF, average distance that an electron with a given energy travels between successive inelastic collisions), N0 and N are the number of electrons before and after passage through the sample, x is the thickness of sample and ϴ is the angle between the surface and the plane of acceptance of the electron optics. With a path length of one λ, 63 % of all electrons are scattered and no longer contribute to the peak in the XPS spectrum. Sampling Depth is defined as the depth from which 95% of all photoelectrons are scattered by the time they reach the surface (3λ). Most λs are in the range of 1 – 3.5 nm for Al Kα radiation, so the sampling depth (3λ) for XPS under these conditions is 3-10 nm Figure 8. Probability of emitted electron reaches the surface and being analysed. XPS data was recorded on pellets which were outgassed in the prechamber of the instrument at room temperature up to a pressure below 2·10−8 torr to remove chemisorbed water from their surfaces. Spectra were recorded using a Leybold-Heraeus LHS-10 spectrometer, working with constant pass energy of 50 eV. The spectrometer main chamber, working at a pressure <2·10−9 Torr, is equipped with an EA-200 MCD hemispherical electron analyser with a dual X-ray source working with Al Kα (hν= 1486.6 Chapter 2. Catalyst synthesis and experimental techniques 63 eV) at 120 W and 30 mA. Si 2p and C 1s signal (103.0 and 284.6 eV respectively) were used as internal energy reference depending on the experiments. 2.2.10. Ammonia adsorption microcalorimetry The study of adsorption calorimetry allows the determination of the number, strength, and force distribution of acid / base sites in solids and is useful in predicting the effect of these parameters on catalytic properties. This technique measures the heats of adsorption of a base probe molecule. In order to measure the acid sites of our samples we used ammonia as our probe molecule. It is one of the most used in calorimetry since it is extremely basic and also small, so its diffusion is hardly affected by the porous structure. The disadvantage of using ammonia is that it adsorbs very strongly even in the weakest acid sites, so its interactions with the acid sites are relatively nonspecific. In the adsorption microcalorimetry technique, the sample is kept at a constant temperature while a known volume of ammonia is adsorbed on its surface, and a heat flux detector emits a signal proportional to the amount of heat transferred per unit of time that it is collected and integrated. Adsorption takes place by repeatedly sending small aliquots of gas over the sample, which has been previously degassed, until saturation ranges are reached. The calorimetric data is generally displayed as a plot of enthalpy of adsorption versus coverage.12,13 Ammonia adsorption microcalorimetry measurements were performed with a Tian- Calvet heat flow microcalorimeter (Setaram), equipped with a volumetric vacuum line. Each sample (ca. 0.1 g, 40-80 mesh) was thermally pretreated at 80 °C for 12 h under vacuum (5 mPa). Ammonia adsorption was carried out by admitting successive doses of the probe gas and measuring the equilibrium pressure relative to each adsorbed amount by means of a differential pressure gauge (Datameritics) up to a final equilibrium pressure of about 133 Pa. The thermal effect corresponding to the adsorption of each Chapter 2. Catalyst synthesis and experimental techniques 64 dose was simultaneously recorded. The adsorption temperature was kept at 80 °C, in order to limit physisorption. After overnight outgassing at the same temperature, a readsorption run was carried out, in order to distinguish the acid sites weak enough to be made free through the outgassing step. The adsorption and calorimetric isotherms were obtained from each adsorption (and re-adsorption) experiment. Combining the two sets of data, a plot of the differential heat of adsorption, Qdiff, as a function of the adsorbed amount was drawn, which gives information on the strength distribution of the adsorbing sites. 2.3. Catalytic activity study The catalysts were tested in a fixed‐bed stainless steel tube reactor with 0.1g of catalyst diluted in the same quantity of SiC (Sigma Aldrich) and located between two quartz wool plugs. Methane dehydroaromatization reaction was performed under CH4/N2 (85:15) with a flow of 5 mL/min, at 700 °C with a 10 °C temperature ramp. Obtained products were analyzed by using online GC (Agilent, 7890B) equipped with 3 channels for separate analyses of light gases (HAYESEP Q precolumn, MolSieve 5Å column, thermal conductivity detector, TCD), light hydrocarbons (GS-GASPRO column, flame ionization detector, FID) and aromatics (HP-88, flame ionization detector, FID). The methane conversion (𝑋𝐶𝐻4), selectivity (𝑆𝐶6𝐻6, etc) and yield (𝑌𝐶6𝐻6, etc) for the different products obtained were calculated using Equations 10 to 12. 𝑋𝐶𝐻4(%)=[𝐶𝐻4]𝑖−[𝐶𝐻4]𝑡 [𝐶𝐻4]𝑖×100 (Eq. 10.), where [𝐶𝐻4]𝑖 are the methane mol at the reactor inlet, and [𝐶𝐻4]𝑡 the methane mol at the reactor outlet at a time t. 𝑆𝐶6𝐻6(%)=[𝐶6𝐻6]𝑡×6 ∑([𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖]𝑡×𝑛𝑖)×100 (Eq. 11), Chapter 2. Catalyst synthesis and experimental techniques 65 where [𝐶6𝐻6]𝑡 are the benzene mol at a time t, 𝑛𝑖 the stoichiometric factor of the product i and [𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖]𝑡 the products mol at a time t. 𝑌𝐶6𝐻6(mmol/𝑔𝑐𝑎𝑡)=∑([𝐶6𝐻6]𝑡×𝐶) (Eq. 12), Where C is a constant where it is included different factors as the loop volume, time of each GC analysis and methane flow. 2.4. References 1. Brunauer, B. S. & Emmett, P. H. in Multimolecular. 60, 309–319 (1938). 2. Harkins, W. D. & Jura, G. Surfaces of Solids. XIII. A Vapor Adsorption Method for the Determination of the Area of a Solid without the Assumption of a Molecular Area, and the Areas Occupied by Nitrogen and Other Molecules on the Surface of a Solid. J. Am. Chem. Soc. 66, 1366–1373 (1944). 3. Union, I., Pure, O. F. & Chemistry, A. International Union of Pure Commission on Colloid and Surface Chemistry Including Catalysis * Reporting Physisorption Data for Gas / Solid Systems with Special Reference to the Determination of Surface Area and Porosity. 57, 603–619 (1985). 4. Barrett, E. P., Joyner, L. G. & Halenda, P. P. The Determination of Pore Volume and Area Distributions in Porous Substances. I. Computations from Nitrogen Isotherms. J. Am. Chem. Soc. 73, 373–380 (1951). 5. Methods, X. D. X-ray Diffraction Methods. (2008). 6. Vyazovkin, S. Thermogravimetric Analysis: Characterisation of Materials. Charact. Mater. 344–362 (2012). 7. Loganathan, S., Valapa, R. B., Mishra, R. K., Pugazhenthi, G. & Thomas, S. Thermogravimetric Analysis for Characterization of Nanomaterials. Thermal and Rheological Measurement Techniques for Nanomaterials Characterization 3, (Elsevier Inc., 2017). 8. Niemantsverdriet, J. W. (Eindhoven U. of T., Catalysis, S. I. of, 2, D. D., Eindhoven, Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction If the kinetic diameter of the molecules involved in MDA is observed, the largest would be that of the aromatics, approximately 5.85 Å. It is a diameter greater than the size of the ZSM-5 channels, however we know that these products diffuse through this zeolite. This is because the pore diameters indicated are only approximations obtained for empty zeolites. Instead, zeolite structures are flexible and if there are small molecules inside, the channel system can expand or deform depending on the shape of the occluded molecule. The Mo/H-ZSM-5 catalysts discussed in this chapter are generally prepared by impregnation with an aqueous solution containing (NH4)6Mo7O24, as in our case. Molybdenum is therefore in the form of an ion, which, being large, remains on the surface of the zeolite. It is by increasing the temperature that the molybdenum species migrate to the pore channels of the zeolite and become anchored.8 Thus, during calcination process, which is carried out at temperatures between 500 °C and 700 °C with low heating ramps, it is possible to achieve a good dispersion of molybdenum species. If, on the other hand, the temperature is raised rapidly, the structure of the zeolite can be damaged and Al can be extracted, forming Al2MoO4 and hampering the catalytic activity of these systems.9 The reported results during these last years on the existing molybdenum species in the Mo/H-ZSM-5 catalytic systems remain a subject of intense debate. In this sense, the exact nature of the Mo-oxo precursor in the catalysts will depend on several parameters to take into account at the time of synthesis, such as the preparation method, the Si/Al ratio and the Mo loading. According to the bibliography, after catalyst calcination Mo is found in the oxidation states of 6+ and 5+, the latter forming aluminium molybdate. Some of the Mo-oxo intermediates observed are (MoO2)2+ and MoO2(OH)+ monomers and (Mo2O5)2+ dimers (Figure 2). The species can stabilize on two proximate BAS but also on silanols. On the external surface larger Mo(VI)-oxo oligomers have also been detected. According to previous publications, if the Si/Al ratio or the Mo content is increased, the formation of dimers is favoured. If, on the other hand, the Si/Al ratio or the Mo content is reduced, the formation of (MoO2)2+ monomers is favored.10–12 Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Figure 2. Schematic of proposed molecular Mo-oxo species adsorbed in ZSM-5 structure. Moreover, these precursor molybdenum species can be found in different positions in the ZSM-5 structure, both at the surface or inside the porous structure. In the case that Mo species were inside the channels, they could also be found in straight channels, sinusoidal channels or at the intersection of both. Thus, we have to highlight the great complexity of the catalyst systems frequently used in the MDA reaction. Therefore, the first stage of our work will be to characterize and understand, as far as possible, the systems from which we start. A detailed knowledge of the catalytic system will require the analysis of its physicochemical, textural and structural properties, combining different characterization techniques. By correlating this information with the performance of the catalyst, a better understanding of the processes that are carried out in this reaction and a more efficient future design of catalysts will be possible. In this chapter, the physico-chemical characterization of the Mo/H-ZSM-5 systems has been carried out using different techniques such as XRD, TPR and TEM, in addition to the in-situ XPS during the catalyst reduction and carburization process. For this, different molybdenum systems supported in commercial ZSM-5 have been synthesized following the protocol detailed in Chapter 2. Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction 3.2. Catalysts characterization and performance 3.2.1. Catalytic activity Figure 3 shows the time-on-stream reaction for a series of Mo/H-ZSM-5 systems, formed by 1, 2, 4, 8 and 10 wt% of Mo supported on the zeolite. Figure 3. Conversion plots for different Mo/H- ZSM-5 catalysts during methane aromatization reaction. As a general trend, methane conversion progressively decays with reaction time, indicating a clear catalyst deactivation with it. In fact, the obtained values of methane conversion at 700°C lies between 15% for 4%Mo and 5% for 1%Mo, showing a rapid and progressive decay. This feature is in agreement with most of the reported values in the literature.4 0100 200 300 400 500 600 0 1 2 3 4 5 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 Conversion to aromatics (%) Time (min) 0100 200 300 400 500 600 0 5 10 15 20 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 Conversion (%) Time (min) 0100 200 300 400 500 600 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 Conversion to C2-C3 (%) Tiempo (min) 0100 200 300 400 500 600 0 1 2 3 4 5 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 Conversion to C +H2 (%) Time (min) Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Among the studied systems, 4 wt% molybdenum loading attained the maximum CH4 conversion rate. When the content of supported Mo exceeds that value, the catalytic performance quickly decreases. The main hydrocarbon product observed was benzene, being naphthalene the main aromatic side product. Thus, achieved selectivities for aromatic compounds are ca 80 % in all cases. It can be stated that the lower aromatic conversion in the other systems can be related to a higher occurrence of intermediate products such as ethylene, ethane or propene (Figure 3). At the same time to aromatization reaction, important CH4 cracking is taking place with continuous conversion to H2 (Figure 3). Thus, the continuous coke formation during the reaction would lead to the progressive catalyst deactivation. If we consider the yield for desired compounds the 4%Mo/H-ZSM-5 gave a significantly higher amount of benzene after ca. 12 hours of reaction (Figure 4). Thus, the selectivity to aromatics was 80 % for 4%Mo/H-ZSM-5 sample. Contrary to benzene and aromatics formation, ethylene is continuously formed along the whole reaction time (Figure 4). Thus, it can be said that the deactivation of catalysts affects specifically to the aromatic production. This fact was already described and was explained by considering the progressive blocking of the porous structure and the accessibility to Brönsted acid sites by coke deposition during the reaction13 In accordance with the bifunctional mechanism, at this point the formed C2 and C3 products (ethane/ethylene, propane/propene) cannot proceed its polymerization to bigger C-compounds. Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Figure 4. Product yields for different Mo/H- ZSM-5 catalysts during methane aromatization reaction. 3.2.2. Surface area measurements The BET surface area of Mo/H-ZSM-5 systems was measured to determine the changes in the surface area of former ZSM-5 upon Mo deposition and after reaction. All catalysts showed type IV isotherms with a clear hysteresis loop, indicating the presence of micro- and mesoporosity (Figure 5. (left)): 0100 200 300 400 500 600 0 1 2 3 4 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 C6H6 yield (mmol / g cat) Time (min) 0100 200 300 400 500 600 0 1 2 3 4 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 Aromatics yield (mmol / g cat) Time (min) 0100 200 300 400 500 600 0 20 40 60 80 100 120 140 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 H2 yield (mmol / g cat) Time (min) 0100 200 300 400 500 600 0 1 2 3 4 5 1% Mo/H-ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 8% Mo/H-ZSM-5 10% Mo/H-ZSM-5 C2-C3 yield (mmol / g cat) Time (min) Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Figure 5. Adsorption-desorption isotherms for different Mo/H-ZSM-5 catalysts (left); Comparison of adsorption-desorption isotherms for 4%Mo/H-ZSM-5 fresh and spent catalysts (right). In Table 1 we summarize the BET surface areas of selected Mo/H-ZSM-5 systems. After Mo deposition, it can be evidenced that the hysteresis loop progressively disappeared as Mo loading increases, denoting that mesopores are highly affected by metal deposition while microporosity remains almost unaltered for low Mo content and slightly diminished for Mo loading higher than 2% (Table 1). Moreover, after reaction the coke deposition strongly affects to the surface area and porosity of the catalyst. Thus, for spent 4%Mo sample ca. 20 % of the total BET surface area is lost after reaction, being this diminution more drastic for mesopores. In Figure 5 (right) it can be seen the noticeable modification of the adsorption-desorption isotherms as well as hysteresis feature. Thus, for spent 4 % Mo sample ca. 20 % of the total BET surface area is lost after reaction, being this diminution more drastic for mesopores which decreases at 50 % (Table 1). We may say that, in principle such textural collapse would be in part the responsible of the catalytic activity deactivation. 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 120 140 Quantity Adsorbed (cm³/g STP) Relative Pressure (P/Po) ZSM-5 2% Mo/H-ZSM-5 4% Mo/H-ZSM-5 10% Mo/H-ZSM-5 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 120 140 Quantity Adsorbed (cm³/g STP) Relative Pressure (P/Po) 4% Mo/H-ZSM-5 fresh 4% Mo/H-ZSM-5 spent Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Catalyst Surface Area (m2/g) Pore Volume (cm3/g) Total Micro* Total Micro* H-ZSM-5 361 267 0.20 0.12 2%Mo 336 265 0.18 0.12 4%Mo 313 236 0.17 0.11 4%Mo spent 245 207 0.14 0.10 10%Mo 272 209 0.15 0.10 *S micro and V micro were calculated using t-plot method. Table 1. Textural properties of the different samples 3.2.3. X-ray diffraction The XRD of fresh samples are shown in Figure 6. Figure 6. XRD patterns for Mo/H-ZSM-5 catalysts. All catalysts exhibit diffraction peaks that can be ascribed to ZSM-5. Although the former diffraction patterns are preserved, the gradual reduction in the intensity of typical diffraction peaks of H-ZSM-5 with Mo loading would confirm that metal incorporation and diffusion through the zeolite pores causes indeed some losses in the crystallinity.14 Moreover, since no evidence of MoOx can be found it can be inferred that Mo clusters 510 15 20 25 30 35 40 45 50 ZSM-5 1%Mo/H-ZSM-5 2%Mo/H-ZSM-5 8%Mo/H-ZSM-5 4%Mo/H-ZSM-5 10%Mo/H-ZSM-5 Intensity (a.u) 2q Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction are well dispersed on the zeolite channels even at higher loadings.15 After reaction, catalysts suffers a great change in the surface features due to the formation of metal carbide and coke deposition. In this sense, it has been reported that remarkable deposition of coke as well as Mo species sintering and detaching from zeolite structure would be the responsible of the catalyst deactivation.16 3.2.4. Transmission electron microscopy In the images obtained by TEM of fresh samples, it is difficult to distinguish the presence of Mo species, especially in 4%Mo/H-ZSM-5 (Figure 7). After reaction, all samples show noticeable clusters that could be ascribed to metal carbide formed during reaction. FRESH SPENT Figure 7. TEM images for 4%Mo/H-ZSM-5 (top) and 10%Mo/H-ZSM-5 (bottom) fresh and spent catalysts. 3.2.5. Thermogravimetric analysis As pointed out before, the formation of coke that could be in principle associated to the catalyst deactivation can be studied by thermogravimetric analysis of spent carburized samples at different times. As reported in the literature, different weight loss regions can be identified in the range from 100 °C to 800 °C. By observing the DTG plots (Figure 8 (left)), three weight loss steps can be clearly identified. The first weight loss below 220 Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction °C would correspond to the physi- and chemisorbed water evolution, as corroborated from TPD experiment (Figure 8 (right)). Figure 8. Thermogravimetric analysis of different 4 and 10%Mo/H-ZSM-5 catalysts after 30, 60 and 120 min reaction (left), TPD experiment of 4%Mo/H-ZSM-5 sample (right). Above this temperature, we may expect the decomposition of metal carbide and coke. This second process occurs at temperature above 300 °C and involves two clear stages at around 450 °C and 550 °C that can be ascribed indistinctly to the elimination of coke and the decomposition of metal carbide. If we follow the evolution of the different contributions to the weight loss with the reaction time, it is evident that the process a higher temperature (ca. 530 °C) become more noticeable as reaction time increases (Figure 8 (left)). Therefore, we may infer that carbide decomposition would occur at lower temperature after which coke combustion would take place. 3.2.6. Raman and UV-vis spectroscopy It has been reported that “bulk” MoO3 shows Raman bands located at 822 and 998 cm-1 and weaker bands around 300 cm-1, corresponding to the antisymmetric stretching mode of Mo–O– Mo, the stretching mode of Mo=O, and the bending mode of Mo=O bonds in bulk MoO3 crystallites, respectively.17 In Figure 9 (left) we show the Raman spectra for different Mo/H-ZSM- 5 samples. 100 200 300 400 500 600 Intensity (a.u) Temperature (ºC) MASS 18 TCD signal 100 200 300 400 500 600 520 ºC CH4, 120 min CH4, 60 min CH4, 30 min 10% Mo/H-ZSM-5 4%Mo/H-ZSM-5 DTG(%/ºC) Temperatura (ºC) 420 ºC 470 ºC 540 ºC Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Figure 9. Raman (left) and diffuse reflectance UV-vis spectra (right) for different Mo/H-ZSM-5 catalysts. In our case, 10%Mo/H-ZSM-5 sample exhibits the characteristics bands of MoO3, showing intense bands at 822, 956 and a small shoulder at 998 cm-1. The relative intensities of 822 and 998 cm-1 bands would indicate that the number of bridging Mo– O–Mo bonds are relative in higher extent than terminal Mo=O bonds. Moreover, the occurrence of 956 cm-1 would also denote the presence of polymeric molybdate species.18,19 This later band is predominantly present in 2%Mo and 4%Mo/H-ZSM-5 samples, for which Mo-O-Mo stretching mode and terminal Mo=O are almost disappeared. From Raman results, it can be assessed the existence of small clusters or oligomers of MoO3 at lower loading levels (below 10 wt%), while observable “bulk” crystal phase of MoO3 would only emerge at significantly higher loading levels (10 wt% or above). Complementary, we have also performed the diffuse reflectance spectra of the studied samples (Figure 9 (right)). For low Mo-loading catalysts, only a single ligand-to-metal charge transfer (LMCT) transition of the O 2p to Mo 3d orbital is observed at 240-300 nm, becoming more pronounced at higher molybdenum loading. Moreover, for 4%Mo and 10%Mo/H-ZSM-5 samples a LMCT tiny transition at around 400 nm arises, 200 300 400 500 600 700 800 Wavelength (nm) 1% Mo 2% Mo 4% Mo 8% Mo 10% Mo 250 MoO3 Absorbance (a.u.) 305 350 200 400 600 800 1000 H-ZSM-5 2% Mo 4% Mo 10% Mo 292 670 998 822 Intensity (a.u) Raman shift (cm-1) 956 880 Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction Figure 15. XPS evolution with time C 1s under in-situ reaction with CH4 at 700 °C. From the observation of the Mo 3d signals, we could envisage the formation of molybdenum carbide that would be accomplished at temperature higher than 600 °C. 295 290 285 280 275 270 CH4, 500ºC 30min CH4, 700ºC 60min CH4, 700ºC 30min CH4, 700ºC 90min CH4, 700ºC 10min Binding Energy (eV) Intensity (cps) 2%Mo/H-ZSM-5 O2, 500ºC 30min 284.2 295 290 285 280 275 270 CH4, 700ºC 90min CH4, 700ºC 60min CH4, 700ºC 30min CH4, 700ºC 10min CH4, 600ºC 30min CH4, 500ºC 30min Intensity (cps) Binding Energy (eV) 1%Mo/H-ZSM-5 O2, 500ºC 30min 278.6 284.6 295 290 285 280 275 270 air exposed CH4, 700ºC 90min CH4, 700ºC 40min CH4, 700ºC 30min CH4, 700ºC 20min CH4, 600ºC 30min CH4, 700ºC 10min CH4, 500ºC 30min Intensity (cps) Binding energy (eV) 4%Mo/H-ZSM-5 O2, 500ºC 30min 284.3 Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction This is in accordance with CH4-TPR, in which the CH4 consumption is taking place at ca 700 °C. New XPS band at ca. 228 eV is observed for 2%Mo and 4%Mo samples. This binding energy has been assigned in the literature to Mo2+ in the Mo2C.22 It is worthy to note that for lower Mo content systems, the appearance of Mo-carbide band seems to be delayed at higher temperature. Thus, for Mo loadings higher than 4% the band assigned to carbide species located at 228 eV is present after CH4 treatment at 600 °C. However, the formation of Mo-carbide cannot be clearly detected from the C 1s band. It has been reported that the carbidic carbon shows a XPS band at around 282 eV.23 Since CH4 cracking process and coke formation would happen close to carburization, C 1s from carbide species would strongly overlap with the graphitic carbon peak at 284.6 eV. Nevertheless, the shift of the big C 1s peak towards lower BE region could serve as an indication for significant Mo2C formation. The amount of carbon after carburization treatment significantly increases, denoting the high degree of carbon deposition (Table 4): Table 4. Surface chemical composition from XPS (wt%). The presence of an intense C 1s peak is observed even after few minutes of carburization. The formation of carbon during this process strongly affects to the surface Mo/H-ZSM-5 O2 @ 500°C, 30 min Mo C Si Al O 1wt% 1.39 0,06 46,81 4.03 47.71 2wt% 4.53 0.28 43.99 4.84 46.35 4wt% 9.95 1.30 41.77 5.23 41.75 Mo/H-ZSM-5 CH4 @ 700°C, 90 min Mo C Si Al O 1wt% 0.47 3.08 47.34 3.59 45.51 2wt% 1.61 25.16 37.58 3.62 32.03 4wt% 9.01 22.35 35.78 2.82 30.04 Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction features of the catalysts. As previously stated, coke deposition produces an important diminution of zeolite mesoporosity. The decrease in accessibility to the Brönsted acid sites of micropores has been reported as the responsible of the rapid deactivation of the systems.13 Furthermore, from the quantitative analysis of carburized samples, it can be noted that the amount of surface Mo drastically decays specially for 1%Mo and 2%Mo/H-ZSM-5. This fact would be probably associated with the formation of graphitic carbon that could cover Mo sites. 3.3. Conclusions We have prepared a series of Mo/H-ZSM-5 systems by impregnation method with different metal loadings. In all cases, Mo species present high dispersion and were not detected by XRD even at high loading. The TPR study clearly reveals the great complexity of Mo species present, especially as loading increases. Indeed, in-situ reduction and carburization analysis by XPS also denotes the different behaviour of catalysts by metal loading. Thus, the formation of Mo-carbide would take place at different temperatures conditioned by the metal content. This clearly points out the different nature and reactivity of metal species, denoting the great complexity of the Mo/H-ZSM-5 systems indicated in the introduction to this chapter. The optimum performance has been attained for 4% metal loading, yielding to ca. 2 mmolbenzene/gcat at the end of the reaction. We have correlated the important deactivation observed to the coke formation by means of CH4 cracking being more pronounced for most active systems. 3.4. References 1. Spivey, J. J., Hutchings, G. Catalytic aromatization of methane. Chem. Soc. Rev. 43, 792–803 (2014). 2. Jang, W. J., Shim, J. O., Kim, H. M., Yoo, S. Y., Roh, H. S. A review on dry reforming of methane in aspect of catalytic properties. Catal. Today 324, 15–26 (2019). 3. Fujita, K. I. Development and application of new iridium catalysts for efficient dehydrogenative reactions of organic molecules. Bull. Chem. Soc. Jpn. 92, 344– Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction 351 (2019). 4. Galadima, A., Muraza, O. Advances in Catalyst Design for the Conversion of Methane to Aromatics: A Critical Review. Catal. Surv. from Asia 23, 149–170 (2019). 5. Olson, D. H., Kokotailo, G. T., Lawton, S. L. Crystal Structure and Structure-Related Properties of ZSM-5. J. Phys. Chem. 85, 2238–2243 (1981). 6. Meier, W. M., Olson, D. H., Kokotailo, G. T., Lawton, S. L. Structure of synthetic zeolite ZSM-5. Nature 272, 437–438 (1978). 7. Hoff, T. C., Thilakaratne, R., Gardner, D. W., Brown, R. C., Tessonnier, J. P. Thermal Stability of Aluminum-Rich ZSM-5 Zeolites and Consequences on Aromatization Reactions. J. Phys. Chem. C 120, 20103–20113 (2016). 8. Ma, D. et al. MAS NMR, ESR and TPD studies of Mo/HZSM-5 catalysts: Evidence for the migration of molybdenum species into the zeolitic channels. Catal. Letters 66, 155–160 (2000). 9. Tan, P. L., Leung, Y. L., Lai, S. Y., Au, C. T. The effect of calcination temperature on the catalytic performance of 2 wt.% Mo/HZSM-5 in methane aromatization. Appl. Catal. A Gen. 228, 115–125 (2002). 10. Jie Gao, Yiteng Zheng, Jih-Mirn Jehng, Y. T., Israel E. Wachs, S. G. P. Identification of molybdenum oxide nanostructures on zeolites for natural gas conversion. Science. 348, (2015). 11. Vollmer, I. et al. Relevance of the Mo-precursor state in H-ZSM-5 for methane dehydroaromatization. Catal. Sci. Technol. 8, 916–922 (2018). 12. Tessonnier, J. P., Louis, B., Rigolet, S., Ledoux, M. J., Pham-Huu, C. Methane dehydro-aromatization on Mo/ZSM-5: About the hidden role of Brønsted acid sites. Appl. Catal. A Gen. 336, 79–88 (2008). 13. Tempelman, C. H. L., Hensen, E. J. M. On the deactivation of Mo/HZSM-5 in the methane dehydroaromatization reaction. Appl. Catal. B Environ. 176–177, 731– 739 (2015). Chapter 3. Physico-chemical characterization of Mo/H-ZSM-5 systems for MDA reaction 14. Hu, J. et al. Effect of the particle size of MoO3on the catalytic activity of Mo/ZSM- 5 in methane non-oxidative aromatization. New J. Chem. 39, 5459–5469 (2015). 15. S. Rajagopal, H.J. Marini, J. A. Ma., R. M. Silica-Alumina-Supported Acidic Molybdenum Catalysts- TPR and XRD Characterization. J. Catal. 147, 417–428 (1994). 16. Lezcano-González, I. et al. Molybdenum Speciation and its Impact on Catalytic Activity during Methane Dehydroaromatization in Zeolite ZSM-5 as Revealed by Operando X-Ray Methods. Angew. Chemie - Int. Ed. 55, 5215–5219 (2016). 17. Windom, B. C., Sawyer, W. G., Hahn, D. W. A raman spectroscopic study of MoS2 and MoO3: Applications to tribological systems. Tribol. Lett. 42, 301–310 (2011). 18. Liu, B. et al. Methanol-to-hydrocarbons conversion over MoO3 /H-ZSM-5 catalysts prepared via lower temperature calcination: a route to tailor the distribution and evolution of promoter Mo species, and their corresponding catalytic properties. Chem. Sci. 6, 5152–5163 (2015). 19. Wei Li, George D. Meitzner, Richard W. Borry III, E. I. Raman and X-Ray Absorption Studies of Mo Species in Mo/H-ZSM5 Catalysts for Non-Oxidative CH4 Reactions. J. Catal. 191, 373–383 (2000). 20. Choi, J. G., Thompson, L. T. XPS study of as-prepared and reduced molybdenum oxides. Appl. Surf. Sci. 93, 143–149 (1996). 21. Xia, T., Li, Q., Liu, X., Meng, J., Cao, X. Morphology-controllable synthesis and characterization of single-crystal molybdenum trioxide. J. Phys. Chem. B 110, 2006–2012 (2006). 22. Oshikawa, K., Nagai, M., Omi, S. Characterization of molybdenum carbides for methane reforming by TPR, XRD, and XPS. J. Phys. Chem. B 105, 9124–9131 (2001). 23. Mir, R. A., Sharma, P., Pandey, O. P. Thermal and structural studies of carbon coated Mo2C synthesized via in-situ single step reduction-carburization. Sci. Rep. 7, 1–12 (2017). Chapter 4. Acid treatment of Mo/ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. Á. López-Martín, F. Platero, G. Colón, A. Caballero “Elucidating the nature of Mo species on ZSM-5 and its role in the methane aromatization reaction” Reaction Chemistry & Engineering, 6, 1265-1276, 2021 In this chapter some of the catalysts that have been previously described above have been subjected to a postsynthesis treatment with sulfuric acid. With the combination of chemical (TPR), spectroscopic (XPS), HAADF and other techniques have allowed us to identify the different Mo precursors stabilized in the calcined ZSM-5 support, their nature (monomers, dimers and bulk Mo oxides), location in the zeolite framework (external surface or micropores), and the partial segregation of aluminum during the preparation of catalysts. The role of each Mo phase promoting or hindering the transformation of methane in aromatics has been also clarified. Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 97 4.1. Introduction. As above anticipating, many of reported previous studies have been devoted to the mechanism of the MDA reaction. All of them discussed about the role of different factors such as the preparation method, the catalyst pretreatment, the loading and dispersion of metal active site or the acidity of the zeolite support.1–7 These variables seem to determine important features such as the dispersion of molybdenum precursors, its location in the internal microporous structure of zeolite or the external surface, and/or the geometry of these precursors.8–12 There is a general consensus that under reaction conditions, typically pure methane at a temperature around 700 °C, the Mo(VI)-oxo centers are reduced or carburized generating a metal carbide phase, which is accepted as the active phase.13,14 Thus, all these issues are essential to establish the role played by each specific phase in the MDA reaction, from highly active phases to merely inactive or spectator phases, or even phases that could hinder the catalytic performance of the system. In this sense, the role of these aspects concerning the nature and amount of coke, that would block the zeolite micropores and the access to the active phase, turns crucial and determines the MDA reaction mechanism.11,12,15–21 As mentioned above, there is a general agreement that the more important molybdenum phase is a highly dispersed one, located inside the micropores of the zeolite, which under reaction conditions generates small molybdenum (oxy)-carbide entities. However, active phases for methane aromatization as well as the reaction mechanism have not yet been completely understood. Several of them have been proposed in the literature and was previously described in Chapter 1.22 This chapter delves deeper into the understanding of the catalytic series described in the previous section. We already discussed in Chapter 3 on how the catalytic performance decreased when the Mo load was higher than 4%. In addition, mesopores were affected and the presence of MoO3 bulk phases were observed at metallic loading up to 10 wt%. On this basis, it is not strange to think that removing this excess of molybdenum could have some effect on the MDA reaction. To eliminate these Mo phases that hinder the aromatization we submit the catalyst to acid treatment in order Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 104 4.3. Identifying and correlating Mo species with catalytic performances From these results, it is evident that several Mo species coexist in the catalysts, playing different roles in the catalytic performance. But first, although we must consider the possibility that some of the reduction processes identified in Figure 1 could correspond to the reduction of molybdenum to intermediate species (V, IV and/or II), a first set of interesting conclusions can be yielded from these results. Thus, it can be assumed that the reduction processes A and E, the two appearing in the 2%Mo/H-ZSM-5 catalyst, would correspond to different Mo species, given that only A is eliminated from the zeolite support by the soft acid treatment. Along with A, the D reduction process, which only appears in the 10%Mo sample, is also eliminated by this soft acid treatment. Therefore, it can also be assigned to a different Mo phase. Regarding the two remaining processes, B and C, they are also selectively extracted by respectively, the soft (B) and hard (C) acid treatment, also indicating that they should correspond to different Mo phases. Indeed, the absence of C species is one of the significant differences when compared the acid treated catalysts with the former 4%Mo sample, negatively affecting the catalytic performance of this acid treated 10%Mo. In summary, these results even without being conclusive, seem to indicate the coexistence of up to 5 different Mo phases in the calcined Mo/H-ZSM5 catalytic system as a function of the metal loading. Moreover, these results suggest that the so named phase C clearly improves the catalytic performance of the Mo/H-ZSM-5 catalysts, while phase D seems to hinder the transformation process of methane in aromatics. 4.4. Locating the position of Mo in/on the zeolite support A fundamental point in the characterization of these systems is to clarify where the metal entities are located, specifically on the external surface and/or the internal mesopores of the zeolite. For this, we have paid attention to the in-situ XPS study of the Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 105 reduction process of the 2, 4 and 10%Mo/H-ZSM-5 catalytic systems. Specifically, we have focused on the results obtained after hydrogen treatments at 650 °C. The Mo 3d regions spectra obtained after calcination and reduction treatments, depicted in Figure 6, show Mo6+ as the only oxidation state present after the calcination treatment. In the same way the reduction treatment at 650 °C almost completely reduced the metal to Mo0. Figure 6. Mo 3d region XPS spectra of Mo/H-ZSM-5catalysts calcined and after in-situ reduction. In particular, the spectrum of the 2%Mo/H-ZSM-5 shows two contributions after the reduction treatment. A main peak centered at 227.2 eV, accounting for 79 % of total intensity and corresponding to Mo0 species, and a minor peak (17 % of total intensity) at 233.6 eV, the same position of the original Mo6+ species. Also, the remaining 4 % of intensity can be fitted with a small contribution Mo(IV). These results are quite interesting when compared and contrasted to the TPR in Figure 1. Taken in account that the reduction is kinetically driven,24 from this profile it can be deduced that A species must be completely reduced at 650 °C, while the E species should remain oxidized. Even more, from the TPR the relative amount of A/E seems to be around 50/50, while the XPS 245 240 235 230 225 220 215 H2/Ar, 650ºC, 30min Binding Energy (eV) Intensity (cps) 2%Mo/H- ZSM-5 O2, 500ºC, 30min 227.2 245 240 235 230 225 220 215 H2/Ar, 650ºC 30min Intensity (cps) Binding Energy (eV) 10%Mo/H-ZSM-5 O2, 500ºC 30min 227.7 245 240 235 230 225 220 215 Intensity (cps) Binding Energy (eV) 4%Mo/H-ZSM-5 O2, 500ºC, 30min 227.2 H2/Ar, 650ºC, 30min Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 106 result shows a very different proportion, with the peak of the metallic molybdenum being more much intense. Thus, two important conclusions can be extracted. First of all, A and E are effectively two different Mo species, with A completely reduced at 650 °C while E remains oxidized. Secondly, taking into account that XPS is a surface sensitive technique, the very low intensity of E can be explained accepting that it is mainly located inside the microporous structure of zeolite and consequently, mostly invisible by this technique. This result is especially relevant as is generally assumed that at low metal loadings, all the molybdenum species are located in the microporous structure of zeolite.6 The possibility that E is agglomerated, thus giving a low intensity peak in XPS can be discarded, as it will be shown later.25 Some additional conclusions can be also extracted from the XPS spectra of 4%Mo and 10%Mo/H-ZSM-5 catalysts. In both cases, the Mo appears completely reduced after the treatment in hydrogen at 650 °C. However, from the TPR profiles of Figure 1 only species A and B must be effectively reduced, while C, D and E species should remain oxidized at this reduction temperature. Once again, these results seem to prove that also the TPR peak assigned to B species should correspond to a different Mo phase that is reduced to Mo0 at 650 °C, a phase that must be located on the external surface of the zeolite and detectable by XPS. Accordingly, the rest of species (C and D) which remain oxidized and undetected by XPS, should be located inside the micropores of the zeolite and/or agglomerated forming large particles. For this reason, it was decided to carry out the HAADF-STEM measurements shown below. Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 107 4.5. Visualizing the Mo particles: HAADF-STEM studies To further characterize these Mo species, a HAADF-STEM and elemental analysis study has been accomplished. According to the images of 2%Mo/H-ZSM-5 presented in Figure 7, no individual metallic particles can be visualized, suggesting that Mo species are extremely well dispersed and distributed throughout the zeolite particles. The images also show that Mo species have the same distribution pattern as Al and Si. Thus, it can be concluded that both Mo phases A/E observed in the 2%Mo/H-ZSM-5 TPR profile of Figure 1 are homogeneously distributed throughout the zeolite support and indistinguishable by TEM. Additionally, this result allows us to discard that the low intensity of XPS peak corresponding to the oxidized phase E is not due to the agglomeration of this phase in large particles (Figure 6), confirming that it must be located inside the micropores of the zeolite. Figure 7. HAADF-STEM images of the 2%Mo/H-ZSM-5catalyst. Composition profile of the signaled line is also included. 050 100 150 200 250 300 350 Intensity (a.u) Position (nm) Al Si Mo Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 108 In contrast, the images obtained for the 4%Mo/H-ZSM-5 sample (Figure 8) allow us to visually identify at least two different Mo phases. Thus, in addition to the homogeneously distributed Mo species (Figure 8, upper panel), it is detected the presence of larger Mo particles associated with Al-enriched areas of the zeolite support. (Figure 8, bottom panel) According to the dark field image it seems that such particles would be located on the external surface of the zeolite particles. This is especially relevant since neither the 2%Mo/H-ZSM-5 catalyst nor the original zeolite submitted to the same preparation treatment but without the addition of Molybdenum present any inhomogeneity in the Si/Al distribution (Figures 8 and 9). Figure 8. HAADF-STEM images of the 4%Mo/H-ZSM-5catalyst. Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 109 This result indicates that some Al have been partially segregated from the ZSM-5 network during the preparation method (impregnation and calcination). Thus, as previously discussed, this phase located on the external surface of zeolite can be associated with the B phase. Figure 9. HAADF-STEM images of the ZSM-5 support. Finally, the images in Figure 10 corresponding to the 10%Mo/H-ZSM-5 sample show, together with the previous ones, the presence of a new phase consisting in large molybdenum particles (around 100 nm in diameter) clearly associated to aluminum. The composition profiles, also included in Figure 10, unambiguously confirms the formation of a Mo/Al2O3 phase, where Si cannot be detected. As it has been also incipiently observed in the 4%Mo/H-ZSM-5 catalyst, its formation seems to be due to some type of chemical etching of aluminum during the calcination treatment of the catalyst, maybe through the formation of a surface aluminum molybdate phase. In fact, the segregation of Al from zeolites induced by Mo has been previously observed by other authors, and identified as a Al2(MoO4)3 phase which is assumed to be inactive for this reaction.26,27 This etching process seems to be favored as the amount of Mo increases. Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 110 Figure 10. HAADF-STEM images of the 10% Mo/H-ZSM-5 catalyst. Composition profiles of the signaled round particle are also included. In order to confirm the formation of a Mo/Alumina phase, we have used the same methodology to prepare a model 4%Mo/Al2O3 catalyst. The TPR profile of this sample, included as Figure 11 shows a main peak centered between 800 and 900 °C, the same range as the phase D. Also, this catalyst does show no activity in the MDA reaction. Both features support the assumption that a Mo/Alumina phase is effectively formed in the 10%Mo/H-ZSM-5 catalyst. 050 100 150 200 250 300 350 Intensity (a.u) Position (nm) Al Si Mo Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 111 020 40 60 80 100 120 140 TCD Signal (a.u) Time (Min) 4Mo/Al2O3 TPR-H2 1000ºC 0 200 400 600 800 1000 T (ºC) Figure 11. TPR profile of a 4%Mo/Al2O3 catalyst. Therefore, all these findings strongly support that the new Mo/Alumina phase can be associated to the D peak detected in the TPR profile of the 10%Mo/H-ZSM-5 catalyst. As a conclusion, by STEM/HAADF analysis we have identified different molybdenum phases, which seems to correspond to the previous A/E, B/C and D phases. Between them, phases A, C and E are undistinguishable by TEM, while phases B and D appear associated to aluminum, relatively dispersed in the case of B and as large particles in the case of D. As shown in a previous work, it is important to point out that no molybdenum phase can be detected by XRD in any of the catalysts, showing only peaks coming from the zeolite.28 This can be clearly seen in Figure 12, where the three catalytic systems present similar profiles, showing basically the same XRD pattern, with no major changes from the zeolite support. Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 112 Figure 12. XRD patterns for ZSM-5 systems 4.6. Raman characterization A further approximation on the nature of these molybdenum phases. For this, the results obtained by Raman spectroscopy (Figure 13) were analysed in greater depth with the aim of ascribing the species if possible. Figure 13. Raman spectra of calcined 2, 4 and 10%Mo/H-ZSM-5 catalysts. As previously reported, ZSM-5 shows bands at 292 and 375 cm-1, coming from the bending mode of six and five-membered rings of the zeolite, respectively.29 The well- 200 400 600 800 1000 H-ZSM-5 2% Mo 4% Mo 10% Mo 292 670 998 822 Intensity (a.u) Raman shift (cm-1) 956 880 510 15 20 25 30 35 40 45 50 ZSM-5 2%Mo/H-ZSM-5 4%Mo/H-ZSM-5 10%Mo/H-ZSM-5 Intensity (a.u) 2q Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 113 known bands at 820 and 996 cm-1 ascribed respectively to Mo-O-Mo and Mo=O asymmetric and symmetric stretching of crystalline MoO3 are also detected in the 10 %Mo/H-ZSM-5 catalyst,30 indicating the presence of large MoO3 entities in this catalyst.31,32 The presence of these big particles must correspond to those previously detected by HAADF and associated to aluminum (Figure 10). The assignment of well dispersed molybdenum species supported on oxide supports by Raman spectroscopy is still debatable, particularly between the spectra obtained under different hydrated conditions.30 Nevertheless, although it is not possible to obtain analogous bulk species to compare with, some of these isolated Mo species, with different degrees of polymerization have been reported in the literature and could serve to elucidate the species present in our systems. Molybdate is usually considered to be in different forms from isolated monomers to dimers/olygomeric and polymeric species. Thus, the band at around 960 cm-1 could be attributed to terminal Mo=O groups of isolated tetrahedrally coordinated monomers, MoO42-, while the weak and broad band at 870 cm-1 has been associated to bridge Mo-O-Mo entities of dimeric/oligomeric species.30 As can be observed in the Figure 13, these bands are also present in the 10%Mo/H-ZSM-5 system, together with the MoO3 bulk ones. In agreement with our previous findings, the presence of such different bands denotes the complexity of Mo speciation over zeolite support.33 In regards to 4%Mo/H-ZSM-5 catalyst, it exhibits two main bands centered at 870 and 960 cm-1, previously described as typical for small 2D clusters of molybdenum oxide species forming both dimers and monomers, respectively.30,34,35 Finally, the 2%Mo/H- ZSM-5 sample only present a well-defined band at ca. 960 cm-1, coming from monomeric Mo species can be found. The band associated to dimeric species at around 870 cm-1 appears but as a very wide peak, almost undistinguishable from the background. Therefore, mainly monomeric and dimeric molybdenum species seem to be present at these low Mo loadings.33 Thus, according to these results, and assuming the presence of the previous characterized molybdenum phases (A-E), it seems plausible to assign the A and E species Chapter 4. Acid treatment of Mo/H-ZSM-5 catalysts: Study of the Mo phases and the performance in the MDA reaction. 120 applications of MoO 3-ZrO2 composite oxides towards one pot synthesis of amidoalkyl naphthols. Catal. Commun. 12, 1255–1259 (2011). 33. Hu, H. & Wachs, I. E. Surface structures of supported molybdenum oxide catalysts. Characterization by raman and Mo L3-edge XANES. J. Phys. Chem. 99, 10897–10910 (1995). 34. Wei Li, George D. Meitzner, Richard W. Borry III, and E. I. Raman and X-Ray Absorption Studies of Mo Species in Mo/H-ZSM5 Catalysts for Non-Oxidative CH4 Reactions. J. Catal. 191, 373–383 (2000). 35. Chen, K., Xie, S., Bell, A. T. & Iglesia, E. Structure and properties of oxidative dehydrogenation catalysts based on MoO3/Al2O3. J. Catal. 198, 232–242 (2001). 36. Agote-Arán, M. et al. Determination of Molybdenum Species Evolution during Non- Oxidative Dehydroaromatization of Methane and its Implications for Catalytic Performance. ChemCatChem 11, 473–480 (2019). Chapter 5. Bimetallic Mo-Re supported ZSM-5 through sequential impregnation. A. López-Martín, M.F. Sini, M.G. Cutrufello, A. Caballero, G. Colón “Characterization of Re-Mo/ZSM-5 catalysts: How Re improves the performance of Mo in the methane dehydroaromatization reaction” Applied Catalysis B: Environmental 304 (2022) 120960 In this chapter the promoting effect of rhenium addition on Mo/H-ZSM-5 catalysts system has been studied. Hence, bimetallic catalysts have been synthesized using a sequential impregnation methodology. The catalytic performance for direct aromatization of methane reaction has been determined and correlated with their physical and chemical state, combining chemical (TPR), spectroscopic (XPS), electron microscopy (TEM, HAADF) and other characterization techniques. An important synergy between Mo and Re, which it is strongly affected by the sequential impregnation, has been observed. Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 125 5.1. Introduction As has been shown throughout this thesis, the challenges with the MDA reaction are two-fold: the reaction is thermodynamically limited with low one-pass methane conversion and even the best catalytic systems, Mo/zeolites, suffer rapid deactivation from coking.1 Within this framework, several characteristics are yet unknown and how to prevent its deactivation caused by coke formation still remains an unsolved problem that hinders its industrial application.2–4 In order to improve the catalytic performance and to decrease the deactivation rate, multiple studies have focused on the addition of different metals as a co-dopant to the Mo catalysts for the MDA reaction. However, though reported results showed interesting effects, attained improvements appear always temporary and the stability of the catalysts was always a challenge.5 For this sake, a catalyst design strategy is crucial to improve stability. So, the use of multifunctional Mo-X/zeolite systems where X is a codopant capable of modulating the stability has been widely reported. In this sense, the addition of metal dopants to would improve not only the catalytic performance but also the stability of Mo/H-ZSM-5 catalysts by influencing the catalyst deactivation step. Thus, Scurrell and co-workers added platinum and/or tin on Mo/H-ZSM-5 zeolite catalyst for methane aromatization reaction.6 They concluded that catalysts with additional tin led to lower methane conversion but higher aromatic products selectivity. The incorporation of Fe modifying conventional Mo/H-ZSM-5 catalysts induces high stability and selectivity to aromatics 7 Zhang et al. studied the effect of indium on Mo/H-ZSM-5 in MDA concluding that although methane conversion decreases, the coke selectivity is reduced to 50 % that of Mo/H-ZSM-5.8 Such stabilization seems to be related to the close proximity of In and Mo that would suppress coke formation. Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 126 More recently, Sridhar et al. reported the addition of different loadings of Co and Ni on Mo/H-ZSM-5 catalyst in order to evaluate the promoting effect of these metals on reactivity and stability of the Mo catalysts.9 They observed a synergistic effect between Mo and the promoters, resulting in benzene yield and catalytic stability improvements. Since it is proposed that Brönsted acid sites would play an important role in the reaction pathways, the modification of the zeolite acidity would be achieved through metal codoping.10 Thus, Cr, Ag and Ga were employed as additive that increased the catalyst acidity and resulted in improved methane conversion and benzene selectivity.5 Rhenium was discovered in 1925 and, in spite of the sparse distribution of rhenium and its low content in the earth's crust, the annual world mine production of rhenium from mining is about 50 tons.11 During the past 30 years, the two most important uses of rhenium have been in high-temperature superalloys and platinum-rhenium catalysts for producing gasoline, where it can successfully replace the more expensive platinum group metals.12,13 Rhenium has many stable oxidation states: +3, +4, +6, +7, and its metallic form Re˚, which allows its compounds to show a rich chemistry and a multifunctionality of Re catalysts. The most stable oxidation state is +7 state, unlike other group VII metals such as Mn+7, which is a highly reactive and oxidizing species. It can also form numerous oxocations and oxyanions.14 This metal has certain similarities with the metals of group VIII in the periodic table in terms of its catalytic properties, which gives it the ability to catalyse both hydrogenation and dehydrogenation reactions, among others. In this way, we could say that rhenium has a combination of properties that makes it different from any other transition element with catalytic activity. Overall, among the catalytic properties of rhenium we could highlight its high selectivity and its high resistance to poisoning and deactivation by nitrogen, sulfur, etc, and thus, improves the yield of the process.15 Therefore, rhenium is a very promising metal for its application in different catalytic processes, in Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 127 fact, rhenium-based catalysts have been extensively used in reforming, alkene metathesis and hydrodesulfurization of crude oil fractions. 15–17 Some studies have used Re/H-ZSM-5 catalysts for the aromatization of ethane. Krogh et al. compared Re/H-ZSM-5 and Zn/H-ZSM-5 and found out that although Re exhibited lower activity initially, it was deactivated more slowly showing higher and stable ethane conversion values at long reaction times. The difference between both catalysts was attributed to the sublimation of Zn species that did not occur in the more stable case of Re/ZSM-5.18 Solymosi et al. reported that Re in the Re/ZSM-5 catalysts was promoting the activation of the C-H bond that resulted in more ethane conversion and higher aromatics selectivity. Also, it was observed that the increase in the rhenium load in the catalysts caused less metal dispersion and made the catalyst less active. 19 In this work we study the effect of rhenium addition as co-dopant to the classical Mo/H- ZSM-5 catalyst, and how it affects, on the one hand to the physicochemical properties, and on the other hand to the catalytic activity and stability during MDA reaction. We will study the important effect of the sequential addition of metals, preparing two different series of catalyst in which Mo has been loaded first in a case (Mo1st), and Re in the other (Re1st). Both catalyst series have been widely characterized by several techniques, in particular XPS and TPR, and also tested in order to evaluate its catalytic performance in the MDA reaction. 5.2. Bimetallic catalysts synthesis The bimetallic catalysts were prepared using a commercial NH4-ZSM-5 zeolite (Alfa Aesar, SiO2/Al2O3=23/1) previously calcined at 550 °C for 3 hours to obtain the acid form. Mo and Re were sequentially added by impregnation procedure. Thus, for each series, one of the metal was firstly impregnated followed by calcination at 550 °C, 3h. After which, the second metal was subsequently introduced in a second step and further Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 128 calcined again at the same temperature. The obtained series were named as Mo1st and Re1st, respectively. Molybdenum was loaded using ammonium heptamolybdate tetrahydrate of appropriate concentration ((NH4)6Mo7O24·4H2O, Sigma Aldrich) to obtain 4 wt% Mo, while rhenium was loaded using rhenium (VII) oxide (Re2O7, Sigma Aldrich) at the corresponding stoichiometric amount leading to 1 wt% and 4 wt%. As a result, the following bimetallic systems from both series were obtained: 1%Re-4%Mo and 4%Re- 4%Mo. 5.3. Catalysts characterization and performance 5.3.1. Catalytic activity Before reaction, catalysts were pretreated under O2/N2 (20% in O2) flow of 15 mL/min at 250 °C for 1 h. Then the temperature was lowered to 150 °C under inert gas. Then, the methane dehydroaromatization reaction was performed as usual, with a CH4/N2 (85% in CH4) flow of 5mL/min. Meanwhile, the temperature was ramped up with a rate of 10 °C/min from 150 °C until 700 °C, and held at this temperature for 18 h. The first important issue to point out is that Re incorporation significantly improves the benzene/aromatic yield in both series with respect to Mo monometallic system. Moreover, a second point to be highlighted is that the catalytic activity for methane aromatization to benzene reveals that the addition sequence followed would have a marked effect (Figure 1). Thus, we may state that Re1st series shows better catalytic performance with respect to Mo1st one. Moreover, while for Re1st series the amount of Re seems not to affect to the aromatic yields, for Mo1st series the amount of Re appears to have certain influence. Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 129 Figure 1. Accumulated benzene and aromatics yields for 4%Mo/H-ZSM-5, 4%Re/H-ZSM-5 and bimetallic catalysts for Mo1st and Re1st series. The maximum activity is attained by 1%Re-4%Mo/H-ZSM-5 catalyst in the Re1st series, for which higher production of benzene and aromatics has been found (Figure 1). It is also worth of noting that the increasing amount of Re does not affect to the yield of benzene in the Re1st series. On the contrary, for Mo1st series increasing amount of Re seem to be detrimental. Moreover, in the case of aromatics production, irrespective of the sequence addition, higher Re content leads to a slightly lower yield. Thus, from both results it can be argued that 1%Re-4%Mo/H-ZSM-5 from Re1st series would favour the 0200 400 600 800 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0200 400 600 800 1% Re-4% Mo/H-ZSM-5 4% Mo/H-ZSM-5 4% Re-4% Mo/H-ZSM-5 4% Re/H-ZSM-5 Aromatics yield (mmol/gcat) Time (min) Time (min) Re 1st Mo 1st 0200 400 600 800 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0200 400 600 800 1% Re-4% Mo/H-ZSM-5 4% Mo/H-ZSM-5 4% Re-4% Mo/H-ZSM-5 4% Re/H-ZSM-5 Benzene yield (mmol/gcat) Time (min) Time (min) Re 1st Mo1st Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 136 Furthermore, when Re is incorporated in second place, it could be preferentially deposited close to Mo in close interaction as will be further discussed. This could explain that bimetallic samples do not show important diminution in the surface area. In spite of the different impregnation procedure, we have evidenced from HADDF a good dispersion of Re ions on the zeolite support (Figure 6). This fact denotes that Re would easily diffuse through zeolite pore structure upon different calcination treatments. 5.3.3. Ammonia adsorption microcalorimetry Ammonia adsorption microcalorimetry experiments were carried out to determine the total number, strength and strength distribution of surface acid sites. The differential heats of ammonia adsorption at 80 °C on selected catalysts are presented in Figure 7 (left panel) with increasing surface coverage. The populations of the surface acid sites at various heat intervals of adsorption are shown in Figure 7 (right panel). Figure 7. Differential heats of ammonia adsorption (left) and acid sites strength distribution obtained after ammonia adsorption for different catalysts (right). 0200 400 600 800 1000120014001600 0 25 50 75 100 125 150 175 200 H-ZSM5 4% Mo 1% Re-4% Mo (Mo1st) 1% Re-4% Mo(Re1st) Qdiff (KJ/mol) Amount of NH3 adsorbed (mmol/g) 60-100 100-120 >120 0 1000 2000 3000 4000 5000 6000 7000 8000 nNH3 (mmol/g) Qdiff (KJ/mol) H-ZSM-5 4% Mo 1% Re-4% Mo (Re1st) 1% Re-4% Mo (Mo1st) Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 137 For all samples, after outgassing at 80 °C, NH3 readsorption occurred only to a limited extent and with low values of differential heat of adsorption (Qdiff from 100 kJ/mol down to 60 kJ/mol). This indicates that the sites characterized by Qdiff ≤ 100 kJ/mol are weak enough to readily desorb ammonia molecules adsorbed during the first adsorption run. Therefore, they can be classified as weak. By observing the different distribution of sites for weak, medium and strong acid sites it can noted that the incorporation of metals largely affects the weak acid sites. Moreover, it can be pointed out that Mo incorporation only affects these weak sites (60-100 kJ/mol). When Re is present, it is worthy to note that strongest sites (> 100 kJ/mol) are specifically affected. In addition, it can be said that the sequence of metal incorporation in bimetallic systems would not have a clear effect on the number of these stronger acid sites. From these results, it can be argued that Re naturally trends to occupy the strongest sites while Mo the weakest ones. Such differential behaviour could be related with the different acid character of the Mo and Re ions. In a recent work, Gao et al. described that the strong interaction between acidic proton and Mo sites correlates with the products formation and catalyst lifetime.26 These authors argued that this issue would have an important implication in modulating the catalytic MDA performance by taking advantage of the proximity of the acidic proton- Mo sites at atomic-level. Furthermore, other authors stated that tuning the surface acidity of charged catalyst would balance the aromatization performance and coke resistance for stability enhancement.27 Thus, the strong Lewis and Brönsted acidity of Ag/H-ZSM-5 would be closely related to its good catalytic performance towards aromatization In our case, as stated from NH3 microcalorimetry, such stronger acid sites would be occupied by Re ions that would condition Mo location and therefore favour the Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 138 interaction with these acid sites. This fact would be associated to the particular ethane evolution observed (Figure 2). 5.3.4. Temperature-programmed reduction Figure 8 depicts the TPR profiles obtained for the samples. As we have extensively discussed above, 4%Mo/H-ZSM-5 sample showed a complex TPR profile.28,29 Figure 8. TPR profiles of monometallic 4%Mo/H-ZSM-5, and bimetallic Re-Mo/H-ZSM-5 catalysts from both series. Such complexity that is denoted from TPR profile renders quite difficult the identification of the active site for this reaction.30 In this sense, Kosinov et al. concluded that MoCx species inside zeolite channels are the active sites, whereas MoCx particles on the zeolite outer surface are “spectators” that only produce coke.1 However, later they reassigned the active sites from MoCx to single Mo atoms, which shows the intricacy of this question.31 Thus, we proposed the occurrence of different molybdenum oxides species, that changed as the metal loading increases. Essentially, we reported four main reduction TCD signal (a.u.) Time (min) 1% Re-4% Mo 4% Re-4% Mo 4% Mo Re1st 20 40 60 80 100 120 Mo1st TCD signal (a.u.) Time (Min) 20 40 60 80 100 120 0 100 200 300 400 500 600 700 800 900 1000 Temperature (ºC) Temperature (ºC) 0 100 200 300 400 500 600 700 800 900 1000 Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 139 processes for this monometallic Mo sample (whose reducing processes peaked are at 510 °C, 615 °C, 755 °C and 1000 °C), showing the high heterogeneity of the chemical states of molybdenum supported on zeolite. However, Re reduction is much simpler and only one defined reduction peak is observed at ca. 400 °C (Figure 9). Figure 9. TPR profiles of monometallic 1%Re/H-ZSM-5 and 4%Re/H-ZSM-5 reference catalysts. In the case of bimetallic catalysts, the reduction profile is quite different from monometallic references and cannot be represented as the sum of both profiles. This fact would point out that certain Mo-Re interaction is present. Such assumption would also be in agreement with the particular evolution of surface feature discussed above (Figure 5). Moreover, particularly for 1%Re-4%Mo systems, it appears that metal introduction sequence also affects to the reduction profile. Thus, this catalyst from Re1st series (Figure 8), the sharp peak associated to Re reduction appears shifted to a slightly higher temperature with lower intensity and showing a new notable shoulder. At the same time, most of the reduction peaks associated to Mo appeared in certain extent diminished. This change in the TPR profile would imply a different reorganization of Mo ions in the structure with respect to the monometallic system. As we have previously stated,28 the highest reduction temperature (ca. 1000 °C) has been associated to small clusters of well-dispersed Mo monomers located at the inner microporous channels of the zeolite. As we have discussed, the role of these species would be similar Time (min) 1% Re/H-ZSM-5 4% Re/H-ZSM-5 20 40 60 80 100 120 0 100 200 300 400 500 600 700 800 900 1000 Intensity (a.u) Temperature (ºC) Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 140 to those small clusters of well-dispersed Mo monomers located on the external surface of the zeolite and would show certain activity. Therefore, the diminution of this particular reduction peak would be associated to the presence of Re ions that could avoid in part the insertion and formation of Mo small clusters at these inner sites. However, the observed diminution of these Mo-species seems not to affect notably the catalytic performance for bimetallic systems. Moreover, the new important reduction process at ca 450 °C together with the partial disappearance of Mo reduction peaks would indicate that Mo is reduced at lower temperatures in the presence of Re. It is worthy to mention that 1%Re-4%Mo/H-ZSM-5 sample from Re1st series showed a more marked different profile with respect to monometallic 4%Mo/H-ZSM-5. Thus, in addition to the identified Mo active sites, new active sites probably associated to Re-Mo entities are expected and would be responsible of the improved performance. We could infer that main diminution in the TPR profile is attained in the region 800° - 920 °C (Figure 8). If we consider the different species described in a previous work, this reduction temperature range would be associated to external located bulk MoO3. From HAADF images we unequivocally showed that these species would be supported on aluminium oxide segregated from the zeolite network and would be responsible of the heavy coke formation.28 The particular modifications observed in this sample were not so marked for the same catalyst in the Mo1st series (Figure 8); so the different sequential deposition clearly induces a different structural situation. For this series, it seems that the TPR peak that appeared more affected was that at higher temperature. Thus, reduction peak has been associated to the well dispersed Mo-monomers located at the inner micropores with strong interaction with the support. The diminution of this reduction peak in this series would point out that upon Re incorporation these Mo species would diffuse to other positions, becoming easily reducible. Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 141 In the case of 4%Re-4%Mo catalysts, the TPR profiles do not show important differences between both series. It is worthy to note that in spite of the higher Re loading with respect to 1%Re-4%Mo catalysts the low temperature peak in principle associated to Re reduction appears with similar intensity as in the case of 1%Re-4%Mo catalysts. This fact would denote that the reduction would not be complete. If we consider the metal content for both series (Table 1), it is worthy to note that while Mo content appears in all cases close to the nominal one, in the case of Re it can be noticed that for Re1st series the metal content is lower. This fact could be related with the high mobility of Re previously mentioned. Indeed, it has been reported that Re2O7 sublimation would occur at 250 °C.32 This fact has been used for achieving stable Re-oxo species grafted on zeolite precursors via Re2O7 reactions with OH groups.25 That issue has been pointed out from surface acidity discussion. Catalysts Re wt% * Mo wt% * µmol H2/mgcat Average reduction % 4%Mo --- 4.2 1.16 88 1%Re 1.3 --- 0.23 93 4%Re 3.6 --- 0.66 97 1%Re-4%Mo (Re1st) 0.8 4.2 0.98 67 4%Re-4%Mo (Re1st) 2.3 3.8 1.43 91 1%Re-4%Mo (Mo1st) 1.2 3.8 1.25 99 4%Re-4%Mo (Mo1st) 3.0 3.8 1.36 81 * Metal content from ICP analysis. Table 1. Metal content, H2 consumption from TPR and oxidation state variation for Re-Mo/H- ZSM-5 from both series. Indeed, for Re1st series the second calcination after Mo impregnation would provoke certain metal sublimation that lead to the slight metal loading diminution. Moreover, this process would explain that inner micropores that initially were blocked became free Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 142 after the second calcination due to the migration toward the external surface. Such high mobility of Re ions would play an important role in the explanation of TPR profiles above discussed. From H2 consumption and considering the metal content observed from ICP, we have also calculated the variation in the oxidation state for supported metals (Table 1). In the case of monometallic samples, the obtained reduction degree for each catalyst highlight that after TPR treatment, metal ions were fully reduced from 6+ and 7+ oxidation state for Mo and Re monometallic catalysts respectively. Additionally, by observing the reduction degree for the bimetallic systems it is clear that 1%Re-4%Mo/H-ZSM-5 from Re1st series would show a particular behaviour. For this catalyst, a fraction of the metal content would remain partially oxidized (67 % of ions are reduced). This would clearly imply the coexistence of Re-Mo species that are stabilized in a partial oxidized state. Such behaviour is not observed in the similar catalyst from Mo1st series or even for 4%Re-4%Mo catalysts. Therefore, the order in the sequential incorporation notably condition the final structural and chemical situation of metals. 5.3.5. X-ray photoelectron spectroscopy In Figure 9 we show the XPS Mo 3d and Re 4f spectra for all samples after soft oxidation at 250°C. The binding energy for Mo 3d5/2 in 4%Mo/H-ZSM-5 is located at 233.3 eV, which has been reported for Mo6+ in MoO3.33,34 Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 143 245 240 235 230 225 245 240 235 230 225 233.2 eV 233.4 eV Intensity (counts) Binding Energy (eV) 4% Mo 1% Re-4% Mo (Re 1st) 1% Re-4% Mo (Mo 1st) 233.0 eV 233.4 eV 233.0 eV 233.0 eV Intensity (counts) Binding Energy (eV) 4% Mo 4% Re-4% Mo (Re 1st) 4% Re-4% Mo (Mo 1st) Figure 9. XPS Mo 3d and Re 4f core-level spectra for monometallic and bimetallic Mo-Re/H- ZSM-5 catalysts after soft oxidation at 250 °C. 60 55 50 45 40 35 30 45,8 eV 40,0 eV 4% Re 1% Re-4% Mo (Re 1st) 1% Re-4% Mo (Mo 1st) Intensity (counts) Binding energy (eV) 46,1 eV 60 55 50 45 40 35 30 25 46,1 eV Intensity (counts) Binding energy (eV) 4% Re 4% Re-4% Mo (Re 1st) 4% Re-4% Mo (Mo 1st) 40,0 eV Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 144 For the 4%Re/H-ZSM-5 catalyst, the observed binding energy for Re 4f7/2 level was 46.1 eV, which is in accordance with values associated to Re7+.35 In the case of bimetallic 1%Re-4%Mo/H-ZSM-5catalysts, Mo 3d and Re 4f bands show a slight shift toward lower binding energy values. Such small shift would point out a strong interaction with Re. Indeed, as previously discussed above from TPR experiments we have denoted a different reduction profile and a lower reduction degree in the case of Re1st series. Additionally, from BET surface area and acidity measurements we also pointed out such possibility. Thus, when rhenium is present it can be assumed that Mo and Re form a new entity in close interaction. * calculated from Mo 3d ** calculated from Re 4f Table 2. XPS binding energies for Mo 3d5/2 and Re 4f7/2 levels and surface atomic ratio for monometallic and bimetallic Re-Mo/H-ZSM-5 catalysts. If we consider the surface composition (Table 2), it can be noticed that in the case of Mo similar Mo/Si ratios have obtained for all catalysts, being in all cases close to Mo content in the monometallic one. However, for rhenium the surface content for 4%Re-4%Mo in the Re1st series lies markedly below the reference value of monometallic one (0.013 vs 0.007 for 4%Re/H-ZSM-5 and 4%Re-4%Mo/H-ZSM-5 Re1st, respectively). The observed differences could be related in principle to a different surface content probably due to a certain loss of Re during second calcination but more importantly to a different distribution of rhenium ions due to the second calcination treatment. Catalyst Mo 3d5/2 (eV) Re 4f7/2 (eV) Mo/Si* ratio Re/Si** ratio Re/Mo ratio 4%Mo 233.3 --- 0.061 --- --- 4%Re --- 46.0 --- 0.013 --- 1%Re-4%Mo (Re1st) 233.0 46.0 0.050 0.004 0.08 4%Re-4%Mo (Re1st) 233.2 46.0 0.059 0.007 0.12 1%Re-4%Mo (Mo1st) 233.0 46.0 0.044 0.005 0.11 4%Re-4%Mo (Mo1st) 233.0 46.0 0.055 0.012 0.22 Chapter 5. Bimetallic Re-Mo supported ZSM-5 through sequential impregnation 145 In order to confirm the above behaviour, we have studied the evolution of the surface during reduction treatment. The results obtained by in-situ XPS study of the reduction process of the monometallic (4%Mo/H-ZSM-5 and 4%Re/H-ZSM-5) and bimetallic catalysts from both series are shown in Figure 10. Figure 10. Mo 3d and Re 4f regions XPS spectra for the 4% Mo/H-ZSM-5 and 4% Re/H-ZSM-5 monometallic catalyst calcined and after in-situ reduction treatment. The Mo 3d and Re 4f regions spectra obtained along different reduction treatments of the monometallic 4%Mo/H-ZSM-5and 4%Re/H-ZSM5 systems have been represented. As previously indicated, for Mo system the only oxidation state present after evacuation pretreatment in O2 is Mo6+. Upon reduction, the Mo 3d peak doublet becomes less defined pointing out the co-existence of other species as it was stated from the complex TPR profile. Just after reduction at 650 °C, the Mo 3d peak becomes wider and clearly shifted to lower binding energy (being Mo 3d5/2 at ca. 228.0 eV) denoting that surface Mo is almost reduced to Mo0 (Table 2). Therefore, it can be said that at this temperature all Mo species at the surface got reduced. On the other hand, regarding to 4%Re/H-ZSM- 5, after calcination the wide peak located at ca 46.1 eV would denote that the main oxidation state is 7+ (Figure 9). As observed from TPR, total reduction is achieved at a 244 240 236 232 228 224 O2/Ar RT H2/Ar 780ºC H2/Ar 650ºC H2/Ar 550ºC H2/Ar 400ºC Fresh sample H2/Ar 350ºC O2/Ar 250ºC Intensity (a.u) Binding energy (eV) Mo 3d 60 55 50 45 40 35 Binding energy (eV) Intensity (a.u) Re 4f O2/Ar RT H2/Ar 780ºC H2/Ar 650ºC H2/Ar 550ºC H2/Ar 400ºC Fresh sample H2/Ar 350ºC O2/Ar 250ºC