Repositorio Institucional de Documentos

Proyecto Fin de Carrera Hidrogenación Selectiva de Nitroestireno catalizada por Platino: Optimización del catalizador y de la reacción Autor Ana Olasolo Alonso Director y ponente Prof. Dr. Lioubov Kiwi Prof. Dr. Javier Herguido Escuela de Ingeniería y Arquitectura, Universidad de Zaragoza 2013 Anexos Anexos Anexos NOTA: Estos anexos se han obtenido llevando a cabo una selección de información de la memoria completa (más extensa) realizada como informe final del trabajo desarrollado en el GGRC-EPFL. Anexos Anexos Anexo I-Concepts 1. Concepts of Heterogeneous Catalysis Catalysis is a technology of modifying the rate at which a desired chemical reaction takes place. By using the catalyst, the yield to a specific product can be improved; the selectivity to the target product can be modified, and the generation of undesired compounds reduced or suppressed1. Catalytic materials are required to decrease the activation energies of the reactions without altering the thermodynamic equilibrium of the process2. Although theoretically they are not consumed during the reaction, their properties can change significantly with use3. Therefore, they must be active, selective, stable and/or easy to regenerate. Heterogeneous catalysis refers to the type of catalysis where the catalyst phase differs from that of the reactants. This technology represents a promising solution for the production of chemicals, petrochemicals and fine chemicals4 where new challenges and unceasing improvement in the area of the preparation of catalysts tailored for specific reaction and/or processes appear continuously3. Some examples of new type of heterogeneous catalysts include catalytic fibers, membranes, catalyst operating in supercritical conditions and heterogeneous enantioselective catalysts3. In Heterogeneous catalysis, the reaction takes place on the active sites of the catalyst suface where molecules are adsorbed and activated. In general, during a reaction over a solid porous catalyst, the following steps take place (Figure-appendix 1): Firstly, there is a mass transfer of the reactants from the bulk phase to the external surface of the catalyst (external diffusion) and after, to the catalyst pores surface (Internal diffusion). The chemical reaction only takes place if the reactants are adsorbed and located and orientated correctly. Finally when the products are formed, they desorb and diffuse to the external catalyst surface (internal diffusion) and then to the bulk (external diffusion). Figure-appendix 1.The catalytic cycle. Anexos 2. Selective Hydrogenation of Nitroarenes 2.1. Applications and Conventional (Non-Catalytic) Synthesis Concerning the hydrogenation of nitroarenes, the production of aminoaromatics is the main application. The reason is that functionalized anilines are essential intermediates in the production of agrochemicals, dyes, pharmaceuticals, polymers and pigments5 and therefore, nowadays they are of a great importance in the chemical industry. Nevertheless, apart from the target product (aminoarenes), by-products are also formed. Therefore, the main challenges in the reduction of functionalized nitroarenes are the reduction of only the specific functional group, keeping untouched the other reducible groups presents in the corresponding Nitroarenes6 (Figure-appendix 2) and the avoidance of decomposition of the molecule, which can lose its functional groups (Figure-appendix 3). Figure-appendix 2. Hydrogenation of functional groups in nitroarenes Figure-appendix 3. Decomposition of Nitroamines To overcome these troubles non-catalytic processes are carry out industrially like Béchamp process with Fe/HCl or sulfide reduction with H2S or NaSH as reducing agent7. On the one hand, the Béchamp process uses stoichiometric amounts of Fe-HCl as reducing system and includes the following steps8: Fe + 2HCl + 2H2O  FeCl2 + 2H3O + 2H3O+ FeCl2 + 2HCl + 2H2O  FeCl42- + 2H3O+ RNO2 + 3Fe + 5H2O  RNH2 + Fe(OH)2 + Fe(OH)3 + FeO + H3O+ Fe(OH)2 + 2Fe(OH)3 Fe3O4 + 4H2O Anexos However, the large amount of Fe-FeO sludge produced as by-product together with the difficulties in the separation of the desired product and the use of corrosive acids make this process inconvenient for the environment. On the other hand, the reduction of nitroarenes by negative divalent sulphur (sulfide, sulfhydrate and polysulfides) is called Zinin reduction9. The overall reaction is as follows: 4ArNO2 + 6S2-+ 7H2O  4ArNH2 +3S2O32- + 6OHThe mechanism of reduction nitro-aromatics in the Liquid-Liquid interphase is described in the following figure: Figure-appendix 4. Mechanism of Sulfide reduction for Nitroaromatics compounds9 Both methods have as a main disadvantage the production of large amounts of waste and therefore it is important to develop economically feasible alternatives to non-catalytic processes7. Anexos 2.2. Alternative Nitroarene Hydrogenation It has been discovered that solid catalysts combined with H2 as reducing agent are very efficient in nitroarene hydrogenation10 and as a consequence, research is focused on finding new and better catalysts tailored for specific reactions and/or processes or improving the performance of the existing catalysts. The generally accepted reaction pathway for the catalytic reduction of aromatic nitro compounds is based in the model presented by Haber11 . There are two different routes in the proposed mechanism: direct and condensation route, which are shown in the Figure-appendix 512. Figure-appendix 5. Possible reaction paths for functional Nitroarenes hydrogenation12 The nitro compound is reduced into a nitroso intermediate followed by its hydrogenation which produces the corresponding hydroxylamine specie (HA). These steps take place very fast and generate an intermediate, capable to react by two different ways; either direct or condensation route12. In the direct route the HA is directly hydrogenated into the desired amine while in condensation route, HA together with the nitroso compound are condensate into an azoxy group which is reduced rapidly into the corresponding azo, hydrazo (condensation products) and finally aniline compounds. Anexos hand, when another type of metal is added to the catalyst, an electronic shift takes place thanks to the interactions of atoms with different electronegativities36, which induces changes in the catalytic properties. In addition, in some cases, depending on the relative position of the active sites, the catalytic response could be different; this could be the case of the addition of a second metal which might change the relative position of the active sites. Regarding the hydrogenation of 3-NS with bimetallic catalysts, Pt is the most active metal reported in the literature, which gives close selectivity to the ones given by gold, the most selective metal. For this reason, it is logic to believe that designing a catalyst of Pt modified by a second metal; the selectivity could be improved, keeping relatively high activity. In a first attempting of using bimetallic catalyst, Corma et al25 studied Au-Pd and Au-Pt catalyst over TiO2 without finding good results in terms of selectivity to 3-VA. However, Cárdenas-Lizana et al.37 studied the hydrogenation of other nitroaromatics like pChoronitrobenzene over Pd/ZnO, and concluded that Pd/ZnO promote the reduction of the nitro group due to the formation of Pd-Zn alloy, achieving 100 % selectivity to p-choroaniline. Crespo-Quesada et al.38 studied the catalyst Pd/ZnO, where Pd-Zn alloy is also formed, focus on the selective hydrogenation of the 2-methyl-3-butyn-2-ol to 2-methyl-3-buten-2-ol and showing an excellent catalytic performance for the reduction of the triple bond with a complete hindrance of the over-hydrogenation reaction. , In addition, in platinum-based bimetallic catalysts, some electropositive metals, such as Fe, Sn, Zn, Ni and Co have prominent promotion effects39, for instance, Silvestre-Albero et al40 suggested that PtZn alloy formation can change the electronic structure of Pt, which may hinder the adsorption of olefinic bond (C=C) during the hydrogenation of crotonaldehyde. Therefore, as the combination of Pd-Zn promotes the selective reduction of nitro compounds and both Pd-Zn and Pt-Zn block the hydrogenation of the double bond, it is reasonable to expect that the combination of Pt-Zn could show promising results in the hydrogenation of 3-NS to 3-VA. Anexos 5. Role of the Support The catalysts support is a material where active sites are fixed.Usually, the most suitable supports for the preparation of heterogeneous catalysts are those which have high porosity and specific surface area, since more amount of active area can be dispersed. There are different types of support but the most common ones are oxides of transition metals (alumina, silica, zirconia, titania) and activated carbon41. Supports play very important role in the catalytic properties. Throught modification of the surface area and pore distribution of the catalyst, the internal mass transfer can be controlled. Supports also affect the metal distribution, structure and morphology of metal particles and their acid/base properties might influence the catalyst performance42. In addition, by the metal-support interactions not only the metal particles are stabilized, reducing sintering, but also the activity and selectivity of the catalyst can be tuned43. There are different types of supports; metallic supports such as carbon nanofibers, zeolites, SiO2, polymers or metal oxide supports, which can be divided in turn in reducible and non-reducible. Examples like ZnO, TiO2, CeO2, Fe2O3, and SnO2are reducible while others such as MgO or Al2O3 are not. The redox characteristics of the oxide support are crucial where partial reduction of carrier can result in partial/total blockage of the active site via migration of suboxide species and encapsulation of the metal particles, and/or the formation of new bimetallic phase with modified electronic and geometric properties affecting catalytic performance41. In this way, reducible supports permit control the activity and selectivity through these metal-support interactions. For example, chemical interactions between noble metal and the support, TiO2, are referred as strong metal support interaction (SMSI)44 and they are associated with the formation of new chemical bonds and drastic changes in the chemisorption properties of the noble metal45. Reducible supports, such as TiO2, CeO2 and SnO2 were used to promote the hydrogenation of carbonyl bond by creating sites at the interface between the metal particle and suboxide species from the support, which decorate the particle surface46,47. Alloy formation was observed in Pt-ZnO and Pt-SnO2 systems48. Conventionally, supported catalysts are used in the form of powders (ZnO, TiO2, Fe2O3, MgO, CNFs), but they present some disadvantages like mass transfer limitations, difficulties to scale-up due to separation costs. To overcome these drawbacks structured support are gaining importance. They have regular structures free of randomness at a reactor level 49 . In spite of all their advantages, structured supports possess a much lower surface area, which can limit the dispersion of the active phase and therefore, low active metal loadings are normally used. To resolve this problem, either higher quantities of structured catalyst are required or non-structured supports are coated by a porous metal oxide layer with the aim of increasing the specific surface area. Anexos Three types of structured catalysts can be distinguished50: 3. Monolithic catalysts: Continuous unitary structures containing many narrow parallel straight or zigzag passages which are made of ceramics or metals. The active material is dispersed uniformly over the whole porous monolithic structure. Using this supports, the pressure drop is low, the flow distribution is uniform, and the catalyst has high specific surface49. Nevertheless this kind of supports is expensive and shortly robust, requiring laminar flow profile and adiabatic conditions. 4. Membrane catalysts: Monoliths with communication between passages because the walls are permeable. Their main advantage is that they can combine catalytic reaction with reactant/product separation. 5. Arranged catalysts: Catalyst arranged in arrays or any other non-particulate catalyst such as packings covered with catalytically active material. In general lines, all of them allow a relatively fast mass transport over reaction zone in the direction perpendicular to the flow.For example, filamentous supports have easy recuperation of the catalyst, no catalyst agglomeration, low pressure drop and low resistance to external and internal mass transfer49. In particular, sintered metal fibers (SMF) are mechanically and chemically stable 3D thin structure of metal microfilament, so they are a kind of filamentous support. They can be made of different special alloys such as stainless steel, Inconel or Fecralloy. For being metal wires, they possess high mechanical strength and a high thermal conductivity which allows isothermal conditions in the reactor, an especially useful feature for highly exo/endothermic reactions. However, to achieve better performance, non-porous supports should be coated by a layer with high surface area, such as ZnO51. Concerning the hydrogenation of NS, it was found high activities and selectivities to VA using TiO224,27 , Al2O327,28 as supports for the following catalysts: Pt/TiO2, Au/TiO2, Au/ Al2O3, Ag/ Al2O3.In addition, Beier et al.52reported the effect of the support on the catalytic response of the hydrogenation of 3-NS, using Pt as active metal. Considerable differences in terms of selectivity were not found (12 % of maximum difference) by using supports such as SiO2, Al2O3, TiO2, carbon nanotubes (CNTs) or unsupported Pt, being TiO2 the most selective. In addition CNTs showed the highest activity (~250 mol mol-1 h-1) while TiO2 provided the lowest one (~70 mol mol-1 h-1). The reason could be that the catalyst Pt/TiO2 is generally expected to have strong interaction between the support and the metal52 and TiOx species can be generated by partial reduction of TiO2 which could block the active surface of Platinum. Anexos Boronat et al.53also studied the influence of the support on the hydrogenation of 3-NS, but using gold as active metal. Table-appendix 3. Effect of the support in 3-NS hydrogenation over Au catalyst53 Support SVA (X(%)) (%) r (mol molM-1 h-1)* TiO2a 96 (99) 173 Fe2O3b 95(23) 23 SiO2c 30(10) 10 Cb 41(6) 6 X: NS conversion, SVA: Selectivity to 3-Vinilaniline, r: total reaction rate Reaction conditions: 1ml feeding; 5.8mol% 3-NS, 90.5 mol% toluene, 1 mol% o-xylene (internal standar), 120°C, 9 bar of H2. a1.5 wt%Au loading, b 4.5 wt% Au loading, c1.6 wt% Au loading The authors have concluded that reducible supports such as TiO2 and Fe2O3 lead to selective processes, whereas non-reducible ones such as SiO2 or C produce both reduction of the double bond and the nitro functions53, being the TiO2 the one which shows highest activity. The study of different supports was also carried out by Shimizu et al.27,28 using 4-NS and Au and Ag as active metals: Table-appendix 4. Effect of the support in 4-NS hydrogenation over Au catalyst27 Support SVA (X(%)) (%) r (mol molM-1 h-1)* Al2O3a 89(100) 3014 SiO2b 76(79) 134 MgOc 59(90) 414 Cc 6(28) 20 TiO2d 95(100) 500 X: Conversion, SVA: Selectivity to 3-Vinilaniline, r: total reaction rate. Reaction conditions: 160°C, 3 bar of H2,THF(15ml),2mmol of 4-NS, 0.04mol% catalyst, Au loading (a 1wt%, b 2.5wt%, c 0.2wt%,d 3wt%) *Initial reaction rate (X up to 40%) It can be seen that SiO2, MgO and C show lower selectivity to 4-VA than Al2O3 and TiO2. The intrinsic activity of the gold NPs having similar mean size in all the catalysts(1.9-3.6 nm) depends strongly on the acid-base features of the support material27. The fact is that the catalyst with a strong basic character support (MgO) and the one with an acidic character support (SiO2) give low initial reaction rate. This, in conjunction with the fact that the Au supported on carbon (neutral) shows the lowest activity, suggests that both acidic and basic surface sites are necessary. That could be why the Au supported on alumina, being the acidbased bifunctional support, shows the highest activity, almost 10 times more than in the other cases. Anexos For the study with Ag, Shimizu et al, stated the following results: Table-appendix 5. Effect of the support in 4-NS hydrogenation over Ag catalyst25 Catalyst SVA (X(%)) (%) r (mol molM-1 h-1)b Al2O3a 96 (100) 183 TiO2 a 92(100) 65 Sepiolite a 89(100) 19 WO3 98(11) 13 ZrO2a 76(61) 19 CeO2 74(59) 19 SnO2 79(3) 2 MgO 21(3) 0.32 X: Conversion, SVA: Selectivity to 3-Vinilaniline, r: total reaction rate Reaction conditions: 160°C, 3 bar of H2,THF(15ml),2mmol of 4-NS, 2mol% catalyst, Au loading 5wt). a 0.04 mol % catalalyst, b Initial reaction rate (X up to 40%) The selectivity did not markedly depend on the support material except for MgO28. The result shows that the support with strong basic character (CeO2 and MgO) and that with acidic character (SnO2 and WO3) result in low activity, being the acid-based bifunctional support (Al2O3) the one which gives the highest activity and also selectivity. Anexos 6. Role of the Solvent Solvents play an important role in heterogeneous catalysis to dissolve or dilute reactants and products, to control high reaction rates, to dissipate any reaction exothermicity; indeed, the solvent may have to be considered as a potential participant in the overall reaction.54 In the production of fine chemicals over heterogeneous catalysts, the choice of the solvent has a crucial impact since they could strongly influence the selectivity and activity of the catalyst. On the one hand, solvents can interact with the reactant and products positively: helping the mass transfer processes and freeing the catalyst surface of active site blockers. On the other hand, they can be adsorbed on the catalyst surface, generating competitive adsorption of the solvent and the reactant molecules on the active sites, and modifying catalytic properties. Therefore, the choice of a solvent and the understanding of solvent effects require a careful investigation55. For these reasons, solvent properties such as polarity, dielectric constant, acidity and factors like solvation of reactants and products, gas solubility and other mass transfer effects are needed to be considered54. Fujita et al.56studied the influence of the solvent in the hydrogenation of 3-NS to 3-VA. During all the study the catalyst employed was Pt/TiO2 and using the same reaction conditions. Table-appendix 6. Effect of the solvent in 3-NS hydrogenation over Pt/TiO2 56 Solvent X(%) SVA (%) Neat* 41 70 scCO2** 64 75 tol b 43 55 E tOH a 99 48 X: Conversion, SVA: Selectivity to 3-VA. Reaction conditions: 3.6mmol of NS, 20mg of 0.5wt% Pt/TiO2, 40 bar of H2, 10 cm3 of organic solvent, 100 bar of CO2, 50°C, 1h of reaction. *No solvent.**Supercritical carbon dioxide (scCO2). On the one hand, it can be deduced from the data that the highest activity was achieved by using ethanol (EtOH) as a solvent, although the selectivity to 3-VA kept low. On the other hand, the best selectivity to 3-VA was achieved by employing supercritical carbon dioxide, being its activity a middle point between the best and the worst activity. Therefore, the solvent affects both activity and selectivity. Anexos 7. Catalyst Preparation Transition metals, especially precious metals NPs are more and more employed in catalysis; they possess a high surface-area-to-volume ratio compared to bulk materials, being able to improve catalytic performances with respect conventional catalystsfor many organic reactions34.In particular, they play a decisive role in the 3-NS to 3-VA hydrogenation. NPs are defined as a small object, sized between 1 and 100 nanometers that behave as a whole unit in terms of its properties. Their main advantage is that they may or may not exhibit properties that differ significantly from those observed in fine particles or bulk materials57, for instance, Gold NPs melt at much lower temperatures (~300 °C for 2.5 nm size) than the gold slabs (1064 °C)58. NPs can be amorphous or crystalline; formed by several shapes and forms, composed of single or multi-chemical elements and they can exist individually or incorporated in a matrix59. The catalytic performance of NPs depends highly on their surface morphology.The number of planes, corners, and edges and the ratio between them depends ultimately on the NP size48. For these reasons, achieving controlled nanocatalyst morphology is the key to improve activity and selectivity60. However the main problem is that NPs are not thermodynamically stable due to their small size and high surface area. They possess a huge driving force to interact quickly with the surrounding atoms, trying to minimize their surface energy and increasing in this way their particle sizes. Small NPs tend to get dissolved while the larger particles grow up, generating a broad size distribution. Therefore, the main challenge is the generation of stable monodispersed NPs of different sizes. 7.1. Size-Controlled Preparation Colloidal methods In order to have good crystallographic control, colloidal methods have been widely implemented. They are advantageous because no specialized equipment is required, solutionbased processing and assembly is easily implemented and large amounts of NPs can be obtained48. In most cases, a metal salt precursor is reduced in solution in the presence of a stabilizing and a reducing agent. Stabilizer adsorbs on the surface of the colloidal particles, thus guarding them against coagulation or improving their chemical stability, but at the same time modifying their catalytic properties61. These methods are based on two steps: nucleation and growth. Nucleation During nucleation, seed particles (nuclei) are formed. They act as nucleation sites for subsequent metal reduction, allowing the growth. Therefore, the seed particles formation is Anexos the key step because they determine the crystallographic growth48. The nucleation can be homogeneous or heterogeneous: In homogeneous nucleation the seed particles are formed in situ and according to the LaMer model. Nucleation only takes place when the metal concentration achieves a saturation point48 and all subsequent growth occurs on the pre-existing nuclei by atom addition. When concentration drops below the critical level, nucleation stops and the particles continue to grow until the equilibrium concentration of the precipitated species is reached62. In order to achieve highly monodisperse nanocrystals, nucleation must occur instantaneously, which is obtained by fast reduction of the metal precursor above the critical concentration. In heterogeneous nucleation the seed particles are pre-formed ex-situ. They are later added to a growth medium to facilitate the reduction of the metal ions48. In this way, size control can be considered as overgrowth process and it allows wider range of growth conditions: milder reducing agents, lower temperatures or aqueous solutions48. In addition, it allows the introduction of seed particles of one metal into the growth medium of another metal. Growth The growth of the particles can be carried out through atom addition where reduced metal species are deposited on the solid surface of the nuclei or by aggregation between particles (secondary addition). In order to stabilize NPs, either some surface protecting reagents such as organic ligands or inorganic capping materials must be added to the reaction mixture or NPs must be placed into an inert environment such as inorganic or polymers62. PVP is a linear polymer widely used as stabilizer due to its versatility and good results in size control of many metals42. For example monodisperse Pt particles of 1.7-3.6 nm were synthesized with alcohol as a reducing agent, PVP as stabilizer and methanol (MeOH), EtOH and ethylene glycol as solvents63. In general colloidal methods allow obtaining well size-controlled NPs with sharp size distribution. However, the main problem arrives when the formed NPs must be deposited on the support followed by stabilizer removal. Normally high temperature treatments are required and during these processes NPs agglomerations can take place resulting in broader size distributions. Non-colloidal methods Using supports is an alternative way to stabilize metal NPs. The principle of NPs formation is the same; reduction of metal precursors, but in this case, the reduction takes place on the support, allowing the formation of NPs directly on the surface of the support where they are more or less firmly anchored and effectively separated from each other. In this Anexos way they can be prepared quite easily, guaranteeing a high level of stability43. Indeed, this type of immobilization to prepare heterogeneous catalysts has been well investigated34. The average distance between particles will depend on the metal content, the particle size and the surface area of the support. To achieve adequate separation, metal loadings usually lie between 0.3-3 wt% and surface areas for supports are typically in the range from 50 to 500 m2 g-1 43. In addition, the metal-support preparation systems possess useful features: the location of the metal NPs within the support is in principle controllable, the particle size can be tuned by metal loading or temperature, further components like modifiers can be incorporated and two or more metal can be used to make supported alloys or bimetallic clusters43. Conventionally, oxides and zeolites are the dominant catalyst supports, because many of them are inexpensive and can be prepared with wide ranges of pore structures64. There are three main methods of supported NPs preparation:  Impregnation method: The support is microporous, it soaks up by capillary action a solution of a metal salt. In this way, the metal solution is deposited in the pore, without exceed their volume. Then, the system is dried and finally reduced to generate small particle metal43 (normally by heating treatment with hydrogen).Corma et al26 achieved good results in terms of selectivity in 3-NS3-VA hydrogenation by preparation of 0.2 wt% Pt/ TiO2, 1.5wt% Ni/TiO2, 5wt% Ru/TiO2using incipient wetness technique and the following precursors: H2PtCl6.6H2O, Ni(NO3)2 and RuCl3.  Deposition-precipitation method mainly consists of a controlled precipitation of a hydroxide or other insoluble compound of the metal onto the support suspended in a liquid medium, followed by filtration, washing, drying and reduction. It is effective in generating well-dispersed metal NPs. Serna et al10 designed 1wt% Au/TiO2 catalyst by deposition-precipitation, using HAuCl4 as precursor and obtained a particle of a 3.5nm size when the solution pH of 7 was employed, which gave good selectivity to 3-VA in the hydrogenation of 3-NS.  Ion exchange method: the metal precursor reacts with the functional groups on a support (e.g., OH groups and oxygen atoms on oxides and zeolites), bonding chemically to the support64. This ion exchange process is followed by washing, drying and reduction and it also gives small NPs of metal. Hydroxyl groups at the surface of oxide supports may undergo either anion or cation exchange with a suitable metal ion complex, depending on the point of zero charge (PZC) of the support43: S+ + OH- S – OH  S-O- + H+ An important characteristic of ion exchange methods is that as the precursor ions are chemically linked to the support before reduction, the possibility of movement of the Anexos formed metal particles is reduced and as a consequence, the aggregation is diminished with respect to the impregnation and deposition-precipitation methods, where there is not such as initial chemical interaction. Another type of supports that is becoming more and more popular are the polymers where metal precursor is introduced into the highly porous polymer framework and reduced to formed metal NPs, which remain stable by confinement in the small pores of the matrix. The main disadvantage is that since the NPs size corresponds to the polymer pores sizes, changing the NPs size implies designing a new polymeric matrix with the desired pores size. This approach has been successfully employed with complexes of several metals such as ruthenium 65 or palladium66 resulting in increasing activity/selectivity/stability as compared to conventional supported metal catalysts. While applications to date have been mainly directed at selective oxidation65,67,68 69, recent use in hydrogenations shows promise66 70. 7.2. Synthesis of Bimetallic Catalysts From the different existing methods of synthesis, the reduction of the metal precursors is broadly used. As occurs with monometallic catalyst, the bimetallic NPs can be formed in colloidal solution or by reduction of metal precursors on a support. However, controlling the size of the NPs is challenging since it depends on the metal composition34. Among the existing colloidal and non-colloidal methods, there are two mainly used techniques34:  Co-reduction where the two metal precursors are chemically reduced together and in presence of stabilizers.  Successive reduction metal precursors, where one metal is deposited on pre-formed monometallic NPs of the other metal. This is especially efficient for the formation of core-shell bimetallic structures. 7.3. Catalyst Activation In order to finish the catalyst preparation, the catalyst must be activated prior to use. This step is crucial since it normally requires temperature treatments which can lead to many transformations on the surface of the catalyst; NPs can sinter, increasing their particle size and broaden the size distribution; while applying reducible support, they can interact with the active phase, generating alloy or decorating the metal NPs with suboxide species; or when bimetallic NPs are used different phases can appear. In this way, selectivity and activity to a certain reaction is strongly influenced by this temperature treatment. Some authors have investigated the effect of the catalyst reduction temperature on the 3-NS3-VA hydrogenation, using different catalysts reduction temperatures. Anexos 0.0 0.1 0.2 0.3 0.4 0.0 0.2 0.4 0.6 CNS/Cdod ANS/Adod K=1.5 Figure-appendix 12. Calibration of NS with an internal standard 0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.05 0.10 0.15 0.20 K=1.7 CVA/Cdod AVA/Adod Figure-appendix 13. Calibration of VA with an internal standard 0.0 0.1 0.2 0.3 0.4 0.0 0.2 0.4 0.6 K=1.7 CEA/Cdod AEA/Adod Figure-appendix 14. Calibration of EA with an internal standard Anexos 0.0 0.1 0.2 0.3 0.0 0.1 0.2 K=0.7 CDimer/Cdod ANS/Adod Figure-appendix 15. Calibration of Condensate products with an internal standard 2. Calculation Equations of the Reaction Parameters It was already mention that the reactant molar concentration is calculated as: dod dod i iC A A iKC  )( The conversion of 3-NSwas calculated from the difference between the reaction inlet and outlet concentrations: inNS outNSinNS HC C CC X , ,, 22   The selectivity to a certain compound (i) was given by: outiNS outi iCC C S ,int, ,   For example, For 3-VA, outVANS outVA VA CC C S ,int, ,   For 3-ENB, outENBNS outENB ENB CC C S ,int, ,   For 3-EA, outEANS outEA EA CC C S ,int, ,   Anexos The yield to a certain compound (i) was defined as: 100 i i SX Y  In order to estimate the activity of the catalyst, the initial reaction rate was measured. This one was defined as a number of consumed moles of reactant at maximum conversion of 20% per time and number of moles of active metal of catalyst: tn r Pt  NS 0 n Anexos Anexo III-Catalyst Characterization 1. Atomic Absorption Spectroscopy (AAS) Table-appendix 9. AAS analysis for the three designed catalysts Catalysts Theoretical Pt loading (wt%) AAS analysis (wt%) Pt/ZnO 0.4 0.4 Pt-Zn/HPS 2 1.6 Pt/ZnO 2 1.2 2. Temperature-Programmed Reduction (TPR) The TPR activation profile is showed in figure-appendix 16 for Pt/ZnO catalyst. In this technique the Pt precursor, after being linked to the ZnO support is reduced. 50 100 150 200 250 300 350 Signal (a.u.) Temperature (°C) Figure-appendix-16. TPR profiles generated for reduction of Pt/ZnO at the heating rates of 5 K min-1. Anexos 3. Transmission Electron Microscopy (TEM) Pt-Zn/HPS: In this case, the Pt and Zn NPs are formed directly in the support by treatment in H2 of the corresponding precursors. <1 1.01-2 2.01-3 3.01-4 4.01-5 5.01-6 6.01-7 7.01-8 8.01-9 9.01-10 10.01-11 11.01-12 12.01-13 13.01-14 14.01-15 >15 0 10 20 30 40 50 60 Percentage in Range (%) Particle size (nm) d=4.6nm Figure-appendix 17. TEM image of Pt-Zn/HPS and metal particle size distribution Pt/ZnO: In this case, Pt NPs are also formed directly on the support by reduction of the precursor after being linked chemically to the surface of the support. <1 1.01-2 2.01-3 3.01-4 4.01-5 5.01-6 6.01-7 7.01-8 8.01-9 9.01-10 10.01-11 11.01-12 12.01-13 13.01-14 14.01-15 >15 0 10 20 30 40 50 60 Percentage in Range (%) Particle size (nm) d=2.3nm Figure-appendix 18. TEM image of Pt/ZnO and metal particle size distribution. Anexos Anexo IVCalculation of Mass Transfer Limitations To evaluate the influence of internal mass transfer on the observed reaction rate, the Weisz modulus ψ for the Pt/ZnO catalyst was calculated according to the equation: surfeff v p p CD obsr n A V      )( 2 1  To implement the calculation, the active phase is supported on ZnO powder (>99%, Aldrich) with a particle diameter of 60 µm. In this way, 6 2 4 2 3 4 2 3 p p p p pd d d A V                      Concerning the diffusion of NS, the effective diffusion coefficient, Deff, in the catalyst layer was estimated based on a molecular diffusion coefficient of liquid in liquid at 75°C of 10-9 m2/s. Assuming middle values of porosity and tortuosity ( 5.0 p  , 4 p  )80, Deff was calculates as sm /1025.1 210  . The reaction order n was assumed as that of nitrostyerene as 1, calculated in appendix number IX. In order to ensure the absence of mass transfer limitation during all the experiment, the observed reaction rate must be obtained at harsh reaction conditions (110°C, 10 bar). The value was smol mol Pt  5366.1 which correspond to sl mol  4 1063.2 with a surface concentration of l mol 0223.0 . Based on these data the value for ψ is calculated as 0.009 and since it is lower than 0.1, it indicates an absence of internal mass transfer limitations in the catalytic surface layer. The molecular diffusion coefficient (Dm) is much smaller in liquid-liquid diffusion than in gas-liquid diffusion, since the gas molecules are generally much smaller and move easier. For this reason it is logic to think that since there is not internal mass transfer limitations for NS, H2 will neither has such as limitations. Anexos Anexo VMass Balance - Accumulation of Hydroxylamines The figure-appendix 19 shows the evolution of the concentrations and mass disbalance with time for the hydrogenation of 3-NS using 0.4wt% Pt/ZnO. Figure-appendix 19. Evolution of the concentrations and disbalance (cyan blue line) with time for the hydrogenation of 3-NS using 0.4wt%Pt/ZnO. Reaction conditions: 75°C, 10 bar, EtOH, 2000 r.p.m., nNS : nPt = 135 : 1. In this figure a disbalance (up to 20%)in the middle of the reactions was observed. The mass balance was calculated as the sum of moles of all the measured compounds at the time the sample was taken and compared with the initial amount of 3-NS. This disbalance is attributed to the accumulation of hydroxylamines (HA), which cannot be quantified by GC due to their degradations at high temperature during the injection. Many authors have study this phenomenom17,18; Studer et al17 reported that the accumulation of HA during the hydrogenation of nitroaromatic compounds can be very significant, for example, high accumulation of 70-80% was observed with unmodified Raney nickel catalyst for the hydrogenation of aromatic nitro compounds, or even worse, the hydrogenation could stop in some special cases. Since there was not disbalance at the end of the reaction over the investigated Pt-Zn catalysts, one can assume that the HA was converted into the final product, 3-VA, resulting in a high selectivity towards desired product. Figure-appendix 20 shows the evolution of selectivity to the desired product versus conversion of 3-NS. 0100 200 0.00 0.01 0.02 Concentration (mol/l) Time (min) Anexos 025 50 75 100 0 20 40 60 80 100 S (%) X (%) Figure-appendix 20. Evolution of selectivity (  3-VA) over conversion of 3-NS for the hydrogenation of 3-NS using 0.4wt%Pt/ZnO. Reaction conditions: 75°C, 10 bar, EtOH, 2000 r.p.m., 135 mol NS mol Pt-1 The selectivity is reduced in the middle of the reactions, caused by the accumulation of HA, which slowly transform into VA. Then, it is observed that the selectivity increases drastically after full conversion thanks to the transformation of these HA into the final product. Anexos Anexo VICalculations and Tables of Activities 1. Reaction Rate at Different Temperatures In order to be able to compare the results in terms of activity, the initial reaction rates over Pt/TiO2 and Au/TiO2 were recalculated for our experimental conditions (75°C). It was found in the literature that the activation energy for hydrogenation of different disubstituted nitroarenes82, over Pt catalysts is in the range of 24-40 KJ mol-1. So, considering an activation energy of 35 KJ mol-1, the initial reaction rate of 3-NS hydrogenation using Pt/TiO2 would be: 232 mol molPt-1 h-1 at 75°C and 2 bar and using Au/TiO2 the maximum initial rate would be 43 mol molAu-1 h-1 at 75°C and 8 bar. The required calculations were: Arrhenius’ law: TR Ea ekk    0 , Initial reaction rate equation: TR E b H a NS a ePCkr    0,0,00 2 Considering two different temperatures:         212,0 1,0 11 ln TTR Ea r r 2. TOF (Turnover Frecuency) Calculation of the activity per surface atom (TOF, turnover frequency) cannot be implemented rigorously since the amount of formed Pt-Zn alloyed is not known and as a consequence, the surface area of Pt covered by Zn cannot be measured. However, a roughly approximation of the activity per active site can be done, considering spherical Pt NPs free of Zn. Van Hardeveld et al71 have calculated the dispersion of the atoms of spherical NPs for different sizes, hence the percentage of atoms on the surface of the NPs (active sites) can be calculated and TOF estimated (dividing the reaction rate value by the dispersion). Applied to this study and working at 75°C, 10 bar, Pt/ZnO (PVP) with an average diameter of 6.8nm, possess a dispersion of 15%, with a TOF of 2186 h-1, Pt-Zn/HPS is characterized with a NPs size of 4.6nm which lead to a dispersion of 22% and an TOF value of 5609 h-1 and finally, Pt/ZnO with a particle size of 2.3nm in average, gives a dispersion of 41% which is translated into 3804 h-1 of TOF. Calculating as well, activity per active site for the literature results, it can be observed that Pt/TiO2 with a NPs size of 5.5 nm achieved a TOF value of 1450 h-1 at 75°C, 2 bar while Au/TiO2 characterized with a NPs size of 4 nm has a maximum TOF value of 215 h-1 at 75°C, 8 bar. Table-appendix 10. Summary of activities in the hydrogenation of 3-NS for the catalysts Catalyst Particle size (nm) Dispersion (%) r0 (mol molM-1 h-1) TOF (h-1) Pt/ZnO (PVP) 6.8 15% 328 2186 Pt-Zn/HPS 4.6 22% 1206 5609 Pt/ZnO 2.3 41% 1560 3804 Pt/TiO2 5.5 17% 232 1450 Au/TiO2 4 20 43 215 Anexos Anexo VIIEffect of the Solvent In the production of fine chemicals over heterogeneous catalysts, the choice of the solvent has a crucial impact; since it can strongly influence the selectivity and activity of the catalyst55. In particular, some solvent properties such as polarity, dielectric constant and acidity and factors like solvation of reactants and products, gas solubility, competitive adsorption on the catalyst surface and other mass transfer effects may be responsible of this effect54. For these reasons, the influence of the solvent on the catalyst performance has been studied. In order to classify the solvents, the polarity was chosen as parameter. Polarity in organic chemistry refers to a separation of charge in a bond or an entire molecule and depends on the difference in electronegativity of atoms and the asymmetry of the molecular structure. Polarity underlies a number of physical properties such as solubility of the reactants in the solvent. Dielectric constant is the most common measure of solvent polarity and characterizes the capability of the solvent to solvate and stabilize charges81. Table-appendix 11 shows the dielectric constants of the solvents used in this study: Table-appendix 11. Dielectric constants of the different solvents Solvent Dielectric constant Toluene 2.4 THF 7.5 IPrOH 18 EtOH 24 MeOH 33 Acetonitrile 37