Fe−Al Oxide Catalysts on Ceramic and

Aug 24, 2009 - Then an aqueous solution of 25 mL of 0.1 M HNO3 (Chem Lab) and 500 mL of distilled water were added. ... which works in the range 0−3...
2 downloads 0 Views 2MB Size
J. Phys. Chem. C 2009, 113, 16503–16516

16503

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts on Ceramic and Metallic Monoliths: Physicochemical Characterization, Effect of the Nature of the Slurry, and Comparison with LaMnO3 Catalysts E. Arendt,*,† A. Maione,† A. Klisinska,‡ O. Sanz,§ M. Montes,§ S. Suarez,| J. Blanco,| and P. Ruiz† Unite´ de catalyse et chimie des mate´riaux diVise´s, UniVersite´ catholique de LouVain, Croix du Sud 2/17, B-1348 LouVain-la-NeuVe, Belgium, Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences, ul. Niezapominajek 8, 30-239 Craco´w, Poland, Department of Applied Chemistry, UniVersity of the Basque Country, Paseo M. de Lardizabal 3, 20018 San Sebastia´n, Spain, and Instituto de Cata´lisis y Petroleoquı´mica, Consejo Superior de InVestigaciones Cientı´ficas, CSIC C/Marie Curie 2, 28049 Madrid, Spain ReceiVed: February 5, 2009; ReVised Manuscript ReceiVed: July 3, 2009

The main objective of this study has been twofold: (i) to investigate the effect of the nature of the slurry used to structurate a catalyst on monoliths; (ii) to study the influence of the nature of the monoliths on the physicochemical properties and on the catalytic activities in the combustion of methane. The slurry used in this work was made with an acidified solution of Pd(2 wt %)/Fe-Al oxide catalyst. This catalyst was prepared by the citrate method and deposited on ceramic and metallic monoliths according to a dip coating or an orbital stirring procedure. Both types of structured catalysts were studied using the following physicochemical techniques: specific surface area measurements (BET), scanning electron microscopy (SEM), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The Pd(2 wt %)/Fe-Al oxide catalyst used for monolith coating does not undergo any textural or crystallinity modification during the deposition procedure. Catalytic activities obtained for powder and monolithic catalysts are compared. The surface of the support plays an important role in the structuration of catalysts. The amount of deposited catalyst has an influence on the catalytic activity. The amount of deposited catalysts depends on the nature of the monolith, on the coating technique, and on the nature of the slurry. Comparisons between structured Pd(2 wt %)/Fe-Al oxide and structured LaMnO3 on ceramic and metallic monoliths (previous work) were made to clarify the influence of the nature of the slurry on the amount of deposited catalysts. The amount of deposited catalysts on ceramic monolith using Pd(2 wt %)/Fe-Al oxide catalyst is lower than that deposited using LaMnO3. In the case of metallic monoliths, the difference is less significant. The influence of a diffusion limitation of the reaction rate in structured catalysts with Pd(2 wt %)/Fe-Al oxide seems to be less important than with LaMnO3. Higher catalytic activities were observed for Pd(2 wt %)/Fe-Al oxide on the ceramic monoliths at high temperature. Coating on the metallic monoliths led to a better activity at low temperature compared to the corresponding powder catalyst. Introduction Catalytic combustion of methane has been widely investigated over the past decade as an alternative to conventional thermal combustion for energy production, in order to decrease the emission levels of noxious and/or greenhouse effect gases in the atmosphere.1-9 The methane oxidation reaction was studied over a wide variety of catalysts. The most studied systems are noble metal supported catalysts.3,4,7,10 Pd supported γ-alumina has been investigated due to its high combustion activity at low temperature.3-6,11-13 However, rapid deactivation is typically observed at high temperature. The deactivation of Pd/γ-Al2O3 catalyst is reported to be mainly due to a decrease of the specific surface area and to a phase transformation that can occur during reaction.3,4,14-16 The modification of γ-Al2O3 leads to a loss of the dispersion of Pd and to a loss of the specific surface area * Corresponding author. E-mail: [email protected]. † Universite´ catholique de Louvain. ‡ Polish Academy of Sciences. § University of the Basque Country. | Consejo Superior de Investigaciones Cientı´ficas.

during combustion.3,4,14-16 Both effects have as a consequence a loss in the catalytic performances. Development of heatresistant catalysts has received considerable attention for the study of high temperature combustion catalysts. A large amount of work has been undertaken to develop catalysts for generating power in high temperature natural gas turbines and for controlling methane emission in industrial plants.5 Considering the literature on performance of Pd catalysts supported on pure alumina, the potentialities of improving the thermal stability at high temperature and combustion activity of this system (Pd/ γ-Al2O3) are quite limited.17 Considerable attention has been paid to perovskites (general formula ABO3) as catalysts of various intermediate and high temperature processes (773-973 K) due to their high activity, thermal stability, and poison resistance.5,18-22 The catalytic activity of perovskites depends on the nature of A and B cations.23-26 Indeed, Mn- and Co-containing perovskites are the most active in methane combustion.25,27-30 Lanthanum is the cation commonly used in the A-site.25,27-30 Thus, LaMnO3 was found to be remarkably active toward methane oxidation.29,31 Hexaaluminates (BAl12O19) were also described as good can-

10.1021/jp901056z CCC: $40.75  2009 American Chemical Society Published on Web 08/24/2009

16504

J. Phys. Chem. C, Vol. 113, No. 37, 2009

didates at very high temperatures reactions conditions of 1100-1400 K.5,14,32-35 It has been revealed that some additives (alkali, alkaline earth, and rare earth metal oxides) to alumina lead to the formation of a hexaaluminate phase, which possesses a high heat resistance against sintering at very high temperatures (>973 K).14 The unique layered crystal structure retards the sintering of the material and hence makes them good candidates for use at extreme temperatures.14,32-35 As for the perovskites, various cations can be substituted into the hexaaluminate structure.9,32,33 However, replacing one Al atom with Cr, Ni, or Co in BAl12O19 did not improve the CH4 oxidation activity, although a high surface area was maintained.32,36 On the other hand, replacing one Al by Fe or Mn greatly increased the methane oxidation activity.32,36 Doping hexaaluminate samples by a transition metal, or by noble metal (substituted inside the hexaaluminate network or supported on it), also resulted in an increase in activity for the CH4 combustion.37,38 Therefore, metal-substituted hexaaluminates are a promising alternative to supported-Pd catalysts for the high temperature catalytic combustion of methane.5,8,39-43 Hexaaluminate having high thermal stability and sufficient specific surface area32,33 can be used to support Pd and simultaneously contribute to the reactivity of the whole catalytic system.17 Indeed, a strong synergetic effect in the catalytic activity in methane oxidation after introduction of 0.5-2 wt % Pd to manganese-alumina catalysts stabilized in the form of Mn3O4 spinel doped with Al3+ cations or as hexaaluminate (Mn, Mg)LaAl11O19 was observed.17 Moreover, excellent stabilities in the methane reaction were obtained on palladium-doped hexaaluminate samples.38 This clearly showed the beneficial effect of the support for the stabilization of the PdO active phase at high reaction temperature. The application of catalytic combustion in high power energy devices requires catalysts with high attrition resistance and low pressure drop, due to the high space velocities necessary for this process.42 Some constraints are involved when the catalyst is used for the catalytic abatement of pollutants in industrial emissions. These problems can be overcome by the use of structured catalysts on monoliths. Both ceramic and metallic monoliths could be used.4,42-46 Presently, metallic monoliths are more interesting because they exhibit higher thermal and electric conductivities even if the adherence of the catalyst to the metallic substrate presents more difficulties than in the case of the ceramic ones.42-46 In addition, the possibility of thinner walls allowing higher cell density and lower pressure drop is a strong advantage of metallic monoliths.44 An additional advantage of metallic substrates is the easy way to produce different and complicated forms adapted to a wide variety of problems and uses.44 Our work concerns catalytic combustion in power energy devices, but the results and insights can be useful for the abatement of methane in industrial plants. In a recent paper47 we carried out a detailed study of the preparation of structured LaMnO3 perovskite catalysts in order to determine the principal parameters influencing the catalytic performances in methane combustion. The parameters influencing the amount of deposited catalyst, such as the nature of the monolith and the coating technique (dip coating or orbital stirring), were investigated. When depositing an active phase on a monolith by wash coating, the major challenge is to achieve a homogeneous and well-adhered catalyst layer on the monolith walls. A lower adherence of the coating on the metallic monolith was observed. The nature of the monolith itself strongly affects the catalytic activity of the monolithic catalysts. A better behavior of structured LaMnO3 on metallic monoliths was observed at a lower temperature, and higher catalytic activities

Arendt et al. were observed for structured LaMnO3 on ceramic monoliths at a higher temperature. Comparisons between structured and powder catalysts showed a higher catalytic activity of the former. The fraction of coverage of the surface of ceramic monoliths by LaMnO3 slurry seems to be the same. We showed that the orbital technique led to a smaller thickness and that the higher the amount of catalyst, the higher the thickness of the coating. We also noticed that the smaller the thickness of the coating, the higher the specific catalytic activity (moles of converted methane per gram per hour). This can suggest that, over ceramic monoliths, a limitation due to diffusional limitation of reactant cannot be excluded. The present work follows the same strategy as for the structured LaMnO3 perovskite catalysts, but using a different catalyst, namely palladium supported on a mixed Fe-Al oxide support. This support has the same composition as FeAl12O19 hexaaluminate. However, in this work, the Fe-Al oxides have not been calcined at a temperature high enough to produce the hexaaluminate phase transformation (1700 K). Our objective is to synthesize structured catalysts to be used at low (as Pd/ Al2O3 does) and intermediate temperatures (about 800 K), namely the same range of application of LaMnO3 perovskite studied previously, in order to compare both results. Another important aspect to underline is that the mixed Fe-Al oxide has been prepared via the citrate method. This method allows preparing a homogeneous amorphous gel which after calcination at an adequate temperature allows the nucleation of the oxide phases. As the selected calcination temperature is not high enough to produce the hexaaluminate phase in all the bulk, it can be expected that the Fe and the Al will form on the surface a layer of a mixed Fe-Al oxide phase. This layer could have properties of some hexaaluminate-type structure and could give some new thermal stability to the alumina support, avoiding the loss of the dispersion of Pd and deactivation. In order to facilitate comparison, the same ceramic and metallic monoliths used in previous work were used in this study. The main aim of the paper is to investigate the physicochemical properties of the powder and structured catalysts. Similar conditions used to measure catalytic activity in the previous work47 were selected, and the same methane concentration was used. The main goal of this work is to compare the structured Pd(2 wt %)/Fe-Al oxide catalysts with the LaMnO3 perovskite catalysts presented in the previous work, in order to clarify the influence of the nature of the slurry used to structure the catalysts over ceramic and metallic monoliths. Characterization of the resulting monoliths was carried out using specific surface area measurements (BET), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and scanning electron microscopy (SEM), and their catalytic activities were tested in methane combustion. Data obtained by further characterization of the structured catalysts could allow complete determination of their principal characteristics (for instance, mechanical properties, viscosity, density, and particle size of the slurry) in order to study the parameters involved in the manufacture of the structured monolith catalyst for practical applications. It is expected that information obtained using the above-mentioned physicochemical techniques can contribute to understanding some of the fundamental aspects related to the structuration of palladium supported on mixed Fe-Al oxide catalysts as a new nonconventional support. Another aspect concerned the monolith. Our objective was to use a metallic monolith and a ceramic one in order to get insight into the role of the nature of the surface. For practical reasons it was not possible to get both monoliths

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts with the same geometric characteristics. Commercial monoliths are proposed with predesigned geometries and for precise applications. This is why, in this work, monoliths present different geometric properties [different cells per square inch (cpsi)]. In addition, it is well-known that ceramic monoliths are porous with well-characterized textures and high surface areas, contrary to metallic monoliths. Metallic monoliths are soft, are nonporous, and have a low surface area. It is expected that using these considerably different monoliths the fundamental information obtained will also be useful for future practical applications. Experimental Section Preparation of Catalysts. Powder Catalysts. Preparation of Fe-Al Oxide Support. Fe-Al oxide support was prepared by the citrate method. Fe(NO3)3 · 9H2O (Aldrich, 98+%, 3.59 g), Al(NO3)3 · 9H2O (Aldrich, 98+%, 40 g), and citric acid monohydrate (Merck, 99.5%) were used as starting materials. Aqueous solution with a cation ratio Fe:Al of 1:12 was prepared. The corresponding nitrates were dissolved in 0.250 L of distilled water. The citric acid was added in 10 wt % excess over the stoichiometric quantity (Merck, 99.5%, 26.67 g of C6H8O7 · H2O) to ensure complete complexation of the metal ions. The resulting solution was stirred for 1 h and then evaporated in a rotary evaporator (438 K) under reduced pressure until a transparent and a viscous gel was obtained. This gel was then placed overnight in a vacuum stove at 343 K. During this treatment an intense production of nitrous vapors occurred. The resulting spongy material was finely crushed and calcined for 5 h at 1073 K. Preparation of Pd(2 wt %)/Fe-Al Oxide Catalyst. A 2 g sample of Fe-Al oxide (25.2 m2 g-1) and 0.10 g of Pd(NH3)4Cl2 · H2O (Aldrich 99.99%) were both put into 0.15 L of distilled water, and the pH of the impregnation solution was fixed up to 10.6 by addition of a 25 vol % solution of ammonia (Merck). After 1 h stirring, water was evaporated under reduced pressure in a rotary evaporator (318 K) and dried overnight in an oven at 383 K. Then the sample was heated to 673 K under O2 (Indugas 99.995%, 60 mL min-1) and kept for 1 h at this temperature. The same treatment as in Pd/γ-Al2O3 catalyst was followed. O2 was evacuated under a flow of N2 (Indugas 99.996%, 60 mL min-1) for 30 min, and palladium was reduced by means of a gaseous treatment under 5 vol % H2 in N2 (Indugas 60 mL min-1) for 3 h at 673 K. Finally, the reduced catalyst was calcined under air for 1 h at 773 K and 1 h at 1073 K. Preparation of Catalyst LaMnO3. This catalyst was prepared previously.47 LaMnO3 catalyst was prepared by the citrate method. La(NO3)3 · 6H2O (Fluka, >99%), Mn(NO3)2 · 4H2O (Merck, 98.5%), and citric acid monohydrate (Merck, 99.5%) were used as starting materials. Aqueous solution with cation ratio La:Mn of 1:1 was prepared. In the preparation procedure, the corresponding nitrates (La(NO3)3 · 6H2O, Mn(NO3)2 · 4H2O) in appropriate quantities (53.72 and 31.14 g, respectively) were dissolved in 2.48 L of distilled water. The citric acid was added in 10 wt % excess over the stoichiometric quantity (47.80 g of C6H8O7 · H2O) to ensure complete complexation of the metal ions. The resulting solution was stirred for 1 h and then evaporated in a rotary evaporator (438 K) under reduced pressure until a viscous gel was obtained. The gel was then placed overnight in a vacuum stove at 343 K. During this treatment, an intense production of nitrous vapors occurred. The resulting spongy material was finely crushed and calcined for 5 h at 973 K.

J. Phys. Chem. C, Vol. 113, No. 37, 2009 16505 Structured Catalysts. Preparation of the Slurry of Powder Catalyst. A 25 g sample of powder catalyst (either Pd(2 wt %)/ Fe-Al oxide (Pd(2 wt %)/FeAl12O19) or LaMnO3) was added to 225 mL of distilled water and kept under stirring during 24 h. Then an aqueous solution of 25 mL of 0.1 M HNO3 (Chem Lab) and 500 mL of distilled water were added. Finally, 50 g of powder catalyst and 125 mL of water were added and kept under stirring at room temperature. The final pH was 5.6 and 3.5 for LaMnO3 slurry and Pd(2 wt %)/Fe-Al oxide, respectively. Monolith Preparation. The same monolith used in the structuration of LaMnO3 powder catalysts were used.47 Ceramic monoliths were prepared by the Instituto de Cata´lisis y Petroleoquı´mica (CSIC, Spain) with the following composition: 40% sepiolite; 40% boehmite. Monolithic supports were made by extrusion of a dough prepared by kneading a mixture of boehmite (Pural, Condea Chemie) and sepiolite (Tolsa S.A.) powders with water. The boehmite presented the following impurities: C, 0.3 wt %; SiO2, 0.01 wt %; Fe2O3, 0.01 wt %; Na2O, 0.01 wt %; K2O, 0.02 wt %. The sepiolite presented the following impurities (expressed as oxides): CaO, 2.36 wt %; Fe2O3, 1.84 wt %; Na2O, 0.53 wt %; K2O, 1.73 wt %. The monolithic shaped materials were dried at room temperature for 48 h, then dried at 383 K for 24 h, and finally heat-treated at 973 K for 5 h. The selected Al2O3/sepiolite ratio was 2/3 by weight according to previous results where the composition of the support was optimized.48 The resulting monoliths presented the following geometric characteristics: cylinder (length × diameter), 3 cm × 17.5 mm; square cell size, 2.5 mm; wall thickness, 0.9 mm; geometric surface, 865 m2/m3; cell density, 8.5 cells/cm2. Monoliths (without catalyst) were calcined at 973 K for 5 h, before coating. The aluminum-based monoliths (metallic monoliths) were prepared by the anodization process according to a procedure reported elsewhere42 and briefly summarized here. The aluminum sheets used for the anodization were obtained from INASA (Industrial Navarre del Aluminio S.A. Spain). Their weight average composition is the following: 0.34% Fe, 0.10% Si, 0.002% Mg, 0.005% Mn, 0.002% Cu, 0.0009% Cr, 0.001% Pb, 0.011% Ti, 0.005% Zn, and 99.53% Al. The aluminum foils (24 cm × 3 cm) were cleaned with detergent and water and rinsed with water. Acetone was then used to remove possible organic impurities and water, and finally the foils were dried. The anodization was carried out in a polypropylene buret immersed in a thermostatic water bath. The aluminum foil, to be anodized, was connected to the anode of the power supply, and placed between two aluminum foils connected to the cathode. These three foils were placed inside the electrolytic bath, and subsequently separated at a constant distance of 1.5 cm from each other. Aluminum foils were separately anodized for 30 min, in a 2.5 M H2SO4 electrolyte with a current density of 2 A dm-2. The power supply used was Gold Source dc power supply DF 1731SB, which works in the range 0-30 V and 0-3 A. Once the anodization process ended, the anodized foil was taken out of the electrolytic bath and washed thoroughly with tap water in order to eliminate the remaining electrolyte. It was then washed with distilled water and dried by air flow. The foils obtained in this way were further dried a 393 K for 30 min and calcined at 873 K for 2 h. The monoliths were prepared with a rolled-up flat and corrugated anodized sheet.42 The resulting monoliths presented the following geometric characteristics: cylinder (length × diameter), 3 cm × 16.0 mm; wall thickness, 0.1 mm; cell density, 355 cells/ in2.42 Coating Techniques: Orbital Stirring and Dip Coating. The powder catalyst was deposited on ceramic and metallic mono-

16506

J. Phys. Chem. C, Vol. 113, No. 37, 2009

Arendt et al.

TABLE 1: Summary of Experimental Conditions Used To Prepare Structured Monolithic Catalysts: Coating Technique, Iteration Number, Weight of Fresh Monoliths, and Notation of the Sample deposited catalyst LaMnO3

monolith

coating technique

iteration no.

wt of fresh monolith (g)

notation

ceramic

orbitalstirring dip coating

metallic

orbitalstirring

3 3 1 11 3 9 3 3 3 3 3

4.2028 4.2717 4.2442 4.6727 4.6770 4.6726 4.5619 4.3091 4.3973 4.7112 4.7004

LaCMOS3 LaCMDC3 LaCMDC1 LaMMOS11 LaMMOS3 LaMMDC9 LaMMDC3 PdCMOS3 PdCMDC3 PdMMOS3 PdMMDC3

dip Pd(2 wt %)/Fe-Al oxide

ceramic metallic

coating

orbitalstirring dip coating orbitalstirring dip coating

liths by two different methods: orbital stirring and dip coating. The same coating techniques were used in our previous work to prepare LaMnO3 perovskite catalysts.47 The orbital stirring technique consists in an orbital stirring at constant speed (46 rpm) of the monolith in 20 mL of the catalyst made slurry during 30 min. The dip coating technique consists in dipping and withdrawing at constant speed (12 cycles/ min) the monolith in 50 mL of slurry during 30 min. The monolith was hung and was kept in the channels’ direction perpendicular to the surface of the slurry during the whole dipping and withdrawing processes. Each dipping (either by orbital stirring or dip coating) was followed by the removal of the excess of material by dry air blowing and a subsequent heating at 473 K for 30 min. The coating and heating procedures were repeated several times, until the desired loading was obtained. Finally, the ceramic monoliths (with catalyst) were calcined at 973 K (the same temperature as that used for the calcination of the powder catalyst) for 5 h and the metallic ones (with catalyst) were calcined at 873 K for 2 h. The calcination temperature for the metallic monoliths is limited by the melting of aluminum. The experimental conditions used in coating technique are indicated in Table 1. Figure 1 indicates the number of iteration cycles necessary for coating the monoliths and the amount of deposited Pd/Fe-Al oxide catalyst on the monoliths. Two ceramic monoliths and two metallic monoliths were impregnated by the two methods (orbital stirring and dip coating) and by the same number of cycles of impregnation (three cycles). In order to identify the prepared monoliths, a specific notation was used: (i) the first two letters indicate the structured catalyst (“La” for LaMnO3 catalyst and “Pd” for Pd/Fe-Al oxide); (ii) the next two letters indicate the nature of the monolith (“CM” meaning ceramic monolith and “MM” meaning metallic monolith); (iii) the next two letters indicate the technique used to coat the monoliths (“OS” for orbital stirring and “DC” for dip

Figure 1. Amount of deposited catalyst per iteration cycle in the coating of ceramic and metallic monoliths.

coating); (iv) the final number indicates the number of cycles used to prepared the sample. Physicochemical Characterization. The BET specific surface areas of the monoliths were measured on a Micromeritics ASAP 2000/2010 using N2 (a 70/30 vol % mixture of helium and nitrogen Praxair) adsorption/desorption at liquid N2 temperature. The BET specific surface area and pore size distribution of the powder catalyst were measured through N2 adsorption at liquid nitrogen temperature by using a Micromeritics TriStar 3000 instrument. XRD patterns of the fresh catalysts were recorded on a Siemens D5000 diffractometer using the KR radiation of Cu (λ ) 1.5418 Å). The 2θ range was scanned between 5° and 80° at a rate of 0.02 deg s-1. The identification of the phases was achieved by using the ICDD-JCPDS database. SEM analysis was performed with a LEO 983 GEMINI microscope equipped with a gun with field emission, which works with a voltage of acceleration of 1 kV. XPS analyses were performed with a photoelectron spectrometer SSI X-probe (SSX-100/206) from Surface Science Instruments using Al KR radiation (E ) 1486.6 eV) and a hemispherical analyzer. The powder samples pressed in small stainless steel troughs of 4 mm diameter were placed on a ceramic carousel in order to avoid differential charging effect. Moreover, an electron flood gun adjusted at 8 eV and a nickel grid placed 2 mm above the sample surface were used for charge stabilization. The samples were outgassed overnight under vacuum (10-5 Pa) and then introduced into the analysis chamber where the pressure was around 10-6 Pa. The spot size was approximately 1 mm in diameter. The resolution for the general spectra and element spectra was fixed at 4. Binding energy (BE) values were referred to the C-(C,H) component of the C 1s peak fixed at 284.8 eV. Data treatment was performed with the CasaXPS program; the spectra were decomposed using 85% Gaussian and 15% Lorentzian types and a Shirley nonlinear sigmoid-type baseline. For each sample, a survey spectrum was recorded, followed by detailed scans of C 1s, O 1s, specific sample elements and finally C 1s again in order to control the charge compensation stability throughout the analysis time. The analysis of the XPS spectra of the catalysts was based on the following photopeaks: Pd 3d, Fe 2p, and Al 2p. The intensity ratio I(Pd 3d5/2)/I(Pd 3d3/2) was fixed at 1.50 with an energy difference of 5.26 eV.49 As far as ceramic and metallic monoliths are concerned, their XPS analyses were based on Si 2s and Mg 2p and Al 2p photopeaks. Catalytic Testing. The catalytic performances of the powder catalyst in methane combustion were investigated by using a fixed-bed microreactor operating at atmospheric pressure. A 320 mg sample of powder catalyst (100-315 µm) was introduced

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts

J. Phys. Chem. C, Vol. 113, No. 37, 2009 16507

into a U-shape quartz reactor (internal diameter 8 mm) and then covered by carborundum beads (Prolabo 99%, diameter >800 µm), whose inactivity was previously checked under reaction conditions. The reactor was inserted into a furnace. The temperature of the catalyst was monitored using a thermocouple, in contact with the bed. The thermocouple was fixed during experiments. No measure of the axial or radial temperature profile was performed. Classical conditions to measure the catalytic combustion of methane were followed. A gaseous mixture of 1 vol % CH4 (Indugas, 99.95%), 10 vol % O2 (Indugas, 99.995%), and 89 vol % He (Indugas, 99.996%) was fed to the reactor. The total flow rate was set at 100 mL · min-1. This flow rate corresponds to a gas hourly space velocity equal to 333.33 min-1. The reaction was run between 573 and 973 K. Analysis of reactants and products was performed by online gas chromatography (Varian CP 4900) during the whole catalytic test. The detection and quantification of compounds were performed using a thermal conductivity detector. Methane conversion was measured stepwise every 50 K with a staying time of 90 min at each step. In order to compare the experiments, the results will be expressed in terms of T10, T30, T50, T70, and T90 values, which are the temperatures to be reached to obtain 10, 30, 50, 70, and 90% of methane conversion, respectively. The catalytic activities of the structured catalysts were evaluated, without any other treatment, by using the same apparatus described above, except that the reactor was modified. The monoliths were introduced into a quartz reactor (length, 50 cm; internal diameter, 2.2 cm) and covered with quartz wool and carborundum beads. The reaction was run between 473 and 873 K (stepwise every 50 K). The temperature was measured by using a thermocouple placed in contact with the monolith. Since the catalyst loading can differ in monolithic catalysts, the specific catalytic performances will be expressed in terms of catalytic activity based on the catalytic mass deposited on the monolith (moles of converted methane per gram per hour) according to the following equation:

XCH4F mol ) g·h mcatal(100)

(1)

where XCH4 represents the conversion rate of methane at a given temperature (%), F is the methane flow (mol of CH4 h-1), and mcatal is the catalyst mass deposited on the monolith (g). Both pure ceramic and metallic monoliths have been tested in order to check their activity toward methane combustion. No conversion of methane was observed.

TABLE 2: Specific Surface Area (SBET), Average Pore Volume (Vp), and Pore Diameter (Dp) of Fe-Al Oxide, Pd(2 wt %)/Fe-Al Oxide Powdered Catalysts (before and after Catalytic Test), Ceramic and Metallic Monoliths, and Structured Monolithic Catalysts after Catalytic Test SBET (m2 g-1) Fe-Al oxide fresh Fe-Al oxide tested Pd(2 wt %)/Fe-Al oxide fresh Pd(2 wt %)/Fe-Al oxide tested CM PdCMOS3 PdCMDC3 MM PdMMOS3 PdMMDC3

Vp (cm3 g-1)

Dp (nm)

25.2 24.2 27.3

0.05 0.05 0.06

5.2 5.5 5.6

27.7

0.06

5.6

140.0 116.7 119.6 1.5 1.8 2.2

-

-

respectively. The most intense peak of PdO at 2θ ) 33.836° can be attributed to reflection plane (101). No peaks indicating an Fe oxide phase were observed. A peak at 57.5° is observable. This peak could be attributed to the presence of some Fe (or/ and Pd) compounds which originate a specific layered structure. This peak is not attributed to iron oxide or palladium oxide. The position and intensity of the peak are the same after the catalytic reaction. Table 3 summarizes the XPS atomic intensity ratios of Pd/ (Fe + Al) and Fe/Al for Fe-Al oxide and Pd(2 wt %)/Fe-Al oxide catalyst. It is observed that the Fe/Al atomic ratio of Fe-Al oxide decreases after the test from 0.075 to 0.056 (the theoretical value is 0.083). An increase of carbon percentage after the reaction (14.92 to 18.68%) was observed. As far as Pd(2 wt %)/Fe-Al oxide is concerned, the fresh XPS Fe/Al atomic ratio (0.055) is significantly lower than that of Fe-Al oxide (0.075). After the test, a decrease to 0.041 is observed. If the Pd/(Fe + Al) atomic ratio is considered, the ratio remains nearly unchanged (0.017 before and 0.019 after the reaction) after the test. The theoretical value for Pd/(Fe + Al) is 0.010. The binding energy (BE) values of the Pd 3d5/2 peak indicate that the surface of the palladium particles is in a fully oxidized state corresponding to PdO, which is observed before and after the catalytic test (336.9 and 337.1 eV, respectively). The assignment of BE values observed to the overall surface oxidation state of palladium is based on literature data.50,51 The

Results Powder Catalysts. Physicochemical Characterization. The BET specific surface areas, the pore volumes, and the pore diameters of Pd(2 wt %)/Fe-Al oxide powder catalyst measured before and after reaction are listed in Table 2. The BET specific surface area of Pd(2 wt %)/Fe-Al oxide powder is a little higher than that of Fe-Al support (27.3 and 25.2 m2 g-1, respectively). No significant variation of the BET specific surface area, the pore volume, and the pore diameter were observed after the catalytic tests for both catalysts. The diffractogram of Fe-Al oxide shows the presence of γ-Al2O3 phase (File No. 10-0425) (data not shown). The Pd(2 wt %)/Fe-Al oxide also shows the presence of γ-Al2O3 and of PdO phases (File No. 41-1107) (Figure 2). The most intense peaks of γ-Al2O3 phase at 2θ ) 37.603°, 45.862°, and 67.032° can be attributed to reflection planes (311), (400), and (440),

Figure 2. Diffractograms of Pd(2 wt %)/Fe-Al oxide before and after the catalytic test. (*) γ-Al2O3 phase; (×) PdO phase.

16508

J. Phys. Chem. C, Vol. 113, No. 37, 2009

Arendt et al.

TABLE 3: XPS Analysis Carbon Percentage (%), XPS Intensity Ratios (Pd/(Fe + Al) and Fe/Al), and Binding Energies of Pd 3d5/2, Fe 2p3/2, and Al 2p of Fe-Al Oxide and Pd(2 wt %)/Fe-Al Oxide Powdered Catalyst before and after the Catalytic Test and of Monoliths after the Catalytic Testsa intensity ratio Fe-Al oxide fresh Fe-Al oxide tested Pd(2 wt %)/Fe-Al oxide fresh Pd(2 wt %)/Fe-Al oxide tested CM PdCMOS3 PdCMDC3 MM PdMMOS3 PdMMDC3 a

BE (eV)

Ctot (%)

Pd/(Fe + Al)

Fe/Al

14.9 18.7 13.1 13.6 8 6 9.7 15.5 11.4 15.00

0.017 (0.010) 0.019 (0.010) 0.010(0.010) 0.008 (0.010) 0.005 (0.010) 0.004 (0.010)

0.075 (0.083) 0.056 (0.083) 0.055 (0.083) 0.041 (0.083) 0.086 (0.083) 0.075 (0.083) 0.042 (0.083) 0.040 (0.083)

Pd 3d5/2 336.9 337.1 336.9 337.3 337.3 337.6

-

Fe 2p3/2

Al 2p

711.7 711.9 711.8 711.8

74.1 74.1 74.3 74.3 74.6 74.1 74.4 74.7 74.5 74.7

-

712.0 711.8

-

711.8 712.0

Theoretical values are reported in parentheses in the table. CM means ceramic monolith and MM means metallic monolith.

Figure 3. Conversion of CH4 (%) as a function of reaction temperature for Pd(2 wt %)/Fe-Al oxide and Fe-Al oxide powder catalysts.

binding energy values of the Al 2p peak for Fe-Al oxide and Pd(2 wt %)/Fe-Al oxide catalyst are the following: 74.1 and 74.3 eV. These BE values are characteristic of γ-Al2O3.49 The binding energies of used catalyst are characterized by values very close to those of the fresh catalysts. The BE values of the Fe 2p3/2 peak for Fe-Al oxide and Pd(2 wt %)/Fe-Al oxide are the following: 711.7 and 711.9 eV. These values are not characteristic of iron oxide species because the BE values of the Fe 2p3/2 peak for FeO, Fe2O3, or Fe2O4 are lower than those observed in this work.49 Catalytic ActiWity. The evolution of the conversion of CH4 as a function of the temperature for Pd(2 wt %)/Fe-Al oxide catalysts is shown in Figure 3. At 723 K, the catalyst shows 10% methane conversion, and at 900 K, the conversion reaches 90%. These temperatures are largely lower than those found for Fe-Al oxide, which displayed a 10% methane conversion at 798 K and reached 90% at 968 K. Structured Catalysts on Monoliths. Physicochemical Characterization. BET results of structured catalysts on monoliths are presented in Table 2. Catalyst coating on ceramic monoliths leads to lower specific surface areas (116.7 m2 g-1 for PdCMOS3 or 119.6 m2g-1 for PdCMDC3, whereas the SBET of pure ceramic monolith is 140.0 m2 g-1). Coating a catalyst on metallic monoliths leads to a specific surface area increase when compared to the metallic monolith itself (1.8 and 2.2 m2 g-1 for PdMMOS3 and PdMMDC3, respectively, compared to 1.5 m2 g-1 for the pure metallic monolith). In Figure 4, the peaks at 2θ ) 21.36°, 26.74°, and 53.64° (indicated by #) can be attributed to the ceramic monolith. It is

worth noticing that there are no significant differences between the diffractograms of the fresh and tested ceramic monoliths. In the case of structured catalysts on ceramic monoliths, PdCMOS3 and PdCMDC3, the peaks attributed to the powder catalyst (γ-Al2O3 and PdO phase) are less intense than the peaks attributed to the ceramic monolith. The peaks characterized by 2θ ) 21.36°, 38.56°, 44.86°, 65.86°, and 78.18° (indicated by #) can be attributed to the metallic support (Figure 5). As far as the structured catalysts on metallic monoliths (PdMMOS3 and PdMMDC3) are concerned, the peaks attributed to the metallic monolith are more intense than the peaks of the phases attributed to the powder catalyst (γ-Al2O3 phase and PdO phase). As was already observed in the case of pure ceramic monolith, there is no significant difference between the diffractograms of the fresh and the tested pure metallic monoliths. Thus it is safe to assume that the monolith does not undergo any change during the catalytic reaction. For practical reasons, ceramic and metallic monoliths were crushed and were analyzed only after the catalytic test; the XPS results are summarized in Table 3. In the case of structured catalysts on ceramic monoliths, the Fe/Al ratios are quite close to the theoretical value observed for the powder (0.083). For instance, the Fe/Al ratio for PdCMOS3 is 0.086. However, when the values obtained for structured catalysts are compared to the value of the powder after reaction (namely 0.041), the Fe/Al atomic ratio is higher in the case of the structured catalyst (0.086 and 0.075). On the other side, for the metallic monoliths, it could be noticed that the values of Fe/Al ratios (0.042 and 0.040 for

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts

J. Phys. Chem. C, Vol. 113, No. 37, 2009 16509

Figure 4. Diffractograms of a ceramic monolith before and after the catalytic test (CM fresh and CM tested, respectively), tested Pd(2 wt %)/ Fe-Al oxide powder catalyst (Pd(2 wt %)/Fe-Al oxide tested), and Pd(2 wt %)/Fe-Al oxide coated on ceramic monoliths (PdCMDC3, PdCMOS3). (#) monoliths; (*) γ-Al2O3 phase; (×) PdO phase.

Figure 5. Diffractograms of a metallic monolith before and after the catalytic test (MM fresh and MM tested, respectively), tested Pd(2 wt %)/ Fe-Al oxide powder catalyst (Pd(2 wt %)/Fe-Al oxide tested), and Pd(2 wt %)/Fe-Al oxide coated on metallic monoliths (PdMMDC3, PdMMOS3). (#) monoliths; (*) γ-Al2O3 phase; (×) PdO phase.

PdMMOS3 and PdMMDC3, respectively) are smaller than the theoretical value observed for the powder (0.083) but are very close to the value observed for the Pd(2 wt %)/Fe-Al oxide powder catalyst after the reaction (namely 0.041). Values obtained for structured catalysts could be compared to the value of the powder Pd(2 wt %)/Fe-Al oxide catalyst after the reaction. When the experimental Pd/(Fe + Al) atomic ratio is considered, its values are lower than the value of the tested powder Pd(2 wt %)/Fe-Al oxide catalyst (0.019) for whatever monolith (0.010, 0.008, 0.005, and 0.004 for PdCMOS3, PdCMDC3, PdMMOS3, and PdMMDC3, respectively). The binding energy (BE) values of the Pd 3d5/2 peak indicate that the surfaces of the palladium particles are in a fully oxidized state and thus were not modified by the deposition of the catalyst on the monoliths. In fact, the BE of Pd 3d5/2 for the powder after the test is 337.1 eV; for the structured catalysts it is 336.9, 337.3, and 337.3 eV for PdCMOS3, PdCMDC3, and PdMMOS3, respectively. Only for PdMMDC3 was a weak increase of the binding energy value observed (BE of Pd 3d5/2 is 337.6 eV). Thus, the catalysts structured on metallic or ceramic monoliths and powder Pd(2 wt %)/Fe-Al oxide catalysts present

the same surface environment after the reaction; in all cases palladium is present as PdO. The binding energy (BE) values of the Al 2p peak indicate that the surface corresponds to γ-Al2O3.49 Binding energy values of the Fe 2p3/2 peak for the structured catalysts are the same as that for the powder. In the case of the structured monoliths, it can be observed by SEM that the catalyst covers the surface of the monolith. Figure 6a shows that the surface of the ceramic monoliths is not smooth; rather, a certain rugosity is visible. The surface of the ceramic monoliths does not undergo any significant morphological change during the catalytic test (Figure 6b). SEM micrographs of ceramic monolith and Pd(2 wt %)/Fe-Al oxide structured on ceramic monolith by orbital stirring (PdCMOS3) and by dip coating (PdCMDC3), after the catalytic test, are presented in Figure 6. The catalyst covers rather well the surface of the monolith. It is worth noticing that there are no significant differences between the SEM micrographs of PdCMOS3 and PdCMDC3 at low magnification (500×; Figure 6c,e). The micrographs at higher magnification (3000× and 5000×, respectively) underline the presence of some regions where the

16510

J. Phys. Chem. C, Vol. 113, No. 37, 2009

Arendt et al.

Figure 6. SEM micrographs of a ceramic monolith without catalyst [(a) before and (b) after the catalytic test (500× magnification)], of PdCMOS3 after the catalytic test [(c) at 500× magnification and (d) at 3000× magnification; the asterisk shows the ceramic monolith and the circle shows a catalytic particle aggregate], and of PdCMDC3 after the catalytic test [(e) at 500× magnification and (f) at 5000× magnification; the asterisk shows the ceramic monolith and the circle shows a catalytic particle aggregate].

ceramic support (marked with an asterisk (*)) and the catalyst (circles) (Figure 6d,f) are observed. Figure 7a of the fresh metallic monolith shows a plain and quite smooth surface. Analysis carried out on the sample indicates no surface rugosity compared to the surface of the ceramic monoliths (Figure 6). However, the morphology of the surface of the metallic monoliths changes drastically after the catalytic test as shown by Figure 7b. The micrograph of the tested monolith shows the presence of scratches and small spherical particles. SEM micrographs of Pd/Fe-Al oxide structured on metallic monolith by orbital stirring (PdMMOS3) and by dip coating (PdMMDC3) are presented in Figure 7c,d. There are no significant morphological changes of the surface of PdMMOS3 and PdMMDC3 compared to the micrograph of the tested monolith without catalyst. Amount of Deposited Catalyst. If the amount of deposited catalyst on metallic and ceramic monoliths, using the same technique (OS or DC) and the same number of cycles (three cycles), is taken into account [PdCMOS3 (0.0538 g) and PdMMOS3 (0.0701 g); with PdCMDC3 (0.1217 g) and PdMMDC3 (0.1325 g)], it can be noticed that the catalyst mass deposited on the ceramic monoliths is lower than that deposited on the metallic ones (Table 4). If the amount of catalyst per unit of surface of monolith was taken into account, the same trend was observed. Catalytic ActiWity. The evolutions of the catalytic activities of the ceramic and metallic monoliths impregnated by orbital

stirring or dip coating are presented in Figure 8. For the ceramic monoliths (PdCMOS3, PdCMDC3) there are no significant differences between the catalytic performances of the two structured catalysts. However, beyond 813 K, the best catalytic performances are obtained by the monolith with the lowest catalyst amount, namely PdCMOS3 (0.0538 g). In the case of the metallic monoliths, the catalytic performances seem to be the same for both structured catalysts. When both ceramic and metallic monoliths are compared, metallic monoliths display better catalytic performances at low temperatures and the ceramic ones are more active at higher temperature. This behavior can be explained by the difference in the support’s thermal conductivity. The conductivity of the metallic monoliths is higher than that of the ceramic monoliths and would allow development of a homogeneous temperature profile through the monolith to decrease the ignition temperature. It is interesting to compare the catalytic results obtained for powder and monolithic catalysts (Figure 8). At low temperature, structured Pd(2 wt %)/Fe-Al oxide coated on ceramic monolith (PdCMOS3) has a lower catalytic activity compared to the corresponding powder catalyst. At higher temperature, PdMCOS3 has the highest activity. If the metallic monoliths are considered, at low temperature, the catalytic performances of both structured metallic monoliths are higher than that of the powder catalyst. At high temperature, catalytic performances of metallic monoliths are lower than that observed for Pd(2 wt %)/Fe-Al oxide.

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts

J. Phys. Chem. C, Vol. 113, No. 37, 2009 16511

Figure 7. SEM micrographs of a metallic monolith without the catalyst [(a) before and (b) after the catalytic test (500× magnification)] and of Pd(2 wt %)/Fe-Al oxide metallic monoliths [(c) PdMMOS3 (500× magnification) and (d) PdMMDC3 (2000× magnification)].

TABLE 4: Amount of Deposited Catalyst (g) and Amount of Deposited Catalyst per Unit of Surface (g m-2) deposited catalyst LaMnO3

monoliths

notation

amt of catal (g)

amt of catal per unit of surface of monolith (g m-2)

ceramic

LaCMOS3 LaCMDC3 LaCMDC1 LaMMOS11 LaMMOS3 LaMMDC9 LaMMDC3 PdCMOS3 PdCMDC3 PdMMOS3 PdMMDC3

0.2280 0.4415 0.2822 0.2724 0.0881 0.2698 0.0881 0.0538 0.1217 0.0701 0.1325

1.9487 × 10-3 3.8060 × 10-3 2.3714 × 10-3 7.7829 × 10-2 3.5240 × 10-2 2.6980 × 10-2 8.8100 × 10-2 4.6101 10-4 1.0176 × 10-3 3.8944 × 10-2 6.0227 × 10-2

metallic

Pd(2 wt %)/Fe-Al oxide

ceramic metallic

Discussion Slurry coating is probably the most widely used method to incorporate a catalytic active phase with known structure and properties into a monolith. Monoliths used in the present work are hydrophobic, making it difficult to coat the catalysts. To be suitable, the slurry must be stable enough, its viscosity must be in the right range, and the amount of catalysts to be structured should be sufficient to allow a high catalytic activity. The concomitant measurable properties such as viscosity, heat uptake, crystalline structure, composition, and size of particles are the experimental parameters which have to be controlled rigorously, in order to understand the mechanism occurring during the structuration of the slurry. On one hand, the effect depends on the treatment (thermal), rheological properties (viscosity, capillary flow), and physical properties (temperature, time, pressure), and on physicochemical (concentration, pH, ionic strength), electrochemical (zero point charge), and acid-base properties. On the other hand, the geometries of the supports, which in the present case are different, could play also a role in the amount of structured catalysts. The geometry of a support can modify the hydrodynamics of the flow into the cells and the mass transport inside them. Coating techniques depend also on the air blowing of excess materials which have been deposed from the slurry, a step which had a high importance in the final structured catalysts. The literature is scarce concerning the physicochemical parameters which are involved in the deposition

of catalytic phases into monoliths. No systematic studies have been realized to clarify the role of these parameters. Research works give principally the protocol of the deposition of the active phases into monoliths and the application of such structured catalysts in well-defined reactions. Discussions concern principally the catalytic activity of monoliths under different operational conditions and the modeling of the mass transfer of reactants and the diffusional limitations inside the channels of the monoliths. The present results are a contribution which, we expect, aims to underline particularly some physicochemical aspects of the deposition process of an active phase in a monolith. Obviously it is beyond our scope to study all the parameters underlined above simultaneously. In the present work, our objective is to study principally the influence of the nature of the slurry and some of the final physicochemical and catalytic properties of the final structured catalysts. These parameters will be discussed at the light of the experimental information obtained in this work. In our opinion, no similar works have been presented previously in the literature. Amount of Deposited Catalyst in the Structured Catalyst and Influence of the Nature of Structured Catalysts. In our previous work,47 we underlined two parameters which influence the amount of deposited catalyst on the monolith: (i) the nature of the monolith and (ii) the coating technique. Present results underline the fact that the nature of the slurry used to impregnate the monolith influences the amount of deposited catalyst.

16512

J. Phys. Chem. C, Vol. 113, No. 37, 2009

Arendt et al.

Figure 8. Specific activity as a function of reaction temperature for the catalysts coated on ceramic monoliths and on metallic monoliths by orbital stirring (PdCMOS3 and PdMMOS3) and dip coating (PdCMDC3 and PdMMDC3). The results for the powder catalyst (Pd(2 wt %)/Fe-Al oxide) are also shown.

Whatever the coating technique (OS or DC), the amount of Pd(2 wt %)/Fe-Al oxide catalysts deposited on the ceramic support is lower than that deposited on the metallic ones. This is an opposite behavior, compared to that observed in our previous work,47 which deals with the structuration of LaMnO3 perovskite catalysts on the same ceramic and metallic monoliths. Taking into account the results obtained with Pd(2 wt %)/Fe-Al oxide and LaMnO3 slurry, it can be noticed that the nature of monolith (ceramic or metallic) definitely has an effect on the amount of deposited catalyst but the amount deposited depends on the nature of the slurry used to structure the catalysts. If the ceramic monolith is considered, the amount of deposited catalyst, using the Pd(2 wt %)/Fe-Al oxide slurry, is always lower than that deposited using the slurry of LaMnO3 (Table 4).47 This is independent of the method used to structure the catalysts (dip coating or orbital stirring). In the case of the metallic monoliths, we do not observe the same tendency. The catalyst mass deposited on PdMMOS3 (0.0701 g) is slightly lower than that deposited on LaMMOS3 (0.0881 g). However, the amount of catalyst deposited on the metallic monolith with the dip coating method and using the Pd(2 wt %)/Fe-Al oxide slurry [PdMMDC3 (0.1217 g)] is higher than that deposited on the same monolith using LaMnO3 slurry [LaMMDC3 (0.0881 g)].47 These results show that the nature of slurry has an important effect on the amount of deposited catalyst. According to the monolith nature, this influence is more or less significant. In the case of the ceramic monoliths, the catalyst mass deposited strongly differs according to the slurry nature: the amount of deposited catalyst, using the Pd(2 wt %)/Fe-Al oxide slurry, is always lower than that deposited using the slurry of LaMnO3. In the case of the metallic monoliths, the difference is less outstanding. In a study dedicated to the deposition of γ-Al2O3 layers on ceramic and metallic supports for the preparation of structured catalysts, Valentini et al.52 present data concerning the effects of the major preparation variables (HNO3 and H2O concentrations in the dispersion, withdrawal velocity, drying temperature, number of dipping cycles, calcination temperature) on the deposited coating load of the wash coat. They reported that the most critical variable turns out to be the HNO3/Al2O3 ratio in the starting aqueous dispersion as the apparent viscosity of the γ-Al2O3 dispersions is strongly influenced by the HNO3 concentration. In fact, they reported that the same trend is obtained both by plotting coating load or viscosity values versus HNO3/Al2O3 ratios. This correlation points out the key role of

the slurry flow behavior in determining the thickness of the coating layer (see below). Thus, it is reliable to suggest that the amount of deposited catalysts could be explained by taking into account the pH at which the impregnation has been performed. The key role of the pH is also pointed out by Perez-Cadenas et al.,53 who investigated the preparation of R-Al2O3 coated monoliths. They reported that the amount of deposited R-Al2O3 on cordierite monoliths increased with increasing pH and increasing alumina concentration in the suspension. Gonzalez-Velasco et al.54 have studied the influence of crushing and acid addition in the deposition of a catalyst on a cordierite monolith. It was found that a good wash coating of these materials is favored by particle size distributions preferably below 10 µm. Nitric acid at pH 5 was preferred among different acids and resulted in a uniform wash coat. Small particles are also advantageously used for nonporous substrates.55 In our results, the Pd(2 wt %)/Fe-Al oxide slurry is more acidic (pH 3.5) than the LaMnO3 slurry (pH 5.6). The acidity of the slurry could influence the amount of deposited catalyst because the impregnation directly depends on the electrical charge of the support (zero isoelectric point) and the electrical charge of the powder catalyst particles. Nevertheless, the pH of the medium seems to have less influence on the metallic monoliths in which the low adherence between the metal matrix and the catalytic slurry is also extremely important. It cannot be excluded that, in the case of ceramic monolith, the acidity of the medium could also modify the nature of the surface of the monolith (a dissolution of the portion of the surface of the monolith, for example) and could result in a modification of the electrical properties of the monolith leading to a smaller amount of catalyst deposited than that deposited on metallic monolith. These results lead to the conclusion that the amount of powder catalyst deposited on the monolith support depends on the nature of the slurry solution, the charge of the support, and the charge of the particles in the slurry. These observation are in agreement with the results obtained in refs 52 and 53. The fine control of these parameters can allow modulating the amount of the powder catalyst deposited on the support. Structured Catalyst. The powder catalyst used in this work is palladium supported on a mixed Fe-Al oxide support. This catalyst has the same composition as an FeAl12O19 hexaaluminate but has not been calcined at a temperature high enough to produce the hexaaluminate phase transformation. γ-Al2O3 phase was found by both XRD and XPS analysis. This phase was

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts

J. Phys. Chem. C, Vol. 113, No. 37, 2009 16513

TABLE 5: Methane Conversion as a Function of Temperature from LaMnO3, Fe-Al Oxide, and Pd(2 wt %)/Fe-Al Oxide Powders: Results Expressed in Terms of T10, T30, T50, T70, and T90 Values Which Are the Temperatures To Be Reached To Obtain 10, 30, 50, 70, and 90% Methane Conversion, Respectively LaMnO3 Fe-Al oxide Pd(2 wt %)/Fe-Al oxide

T10 (K)

T30 (K)

T50 (K)

T70 (K)

T90 (K)

593 798 623

691 853 683

727 893 728

751 928 758

785 958 798

also reported by Lietti et al., who investigated the phase composition of the various hexaaluminates by XRD at different temperatures.56 No Fe oxide phase was observed, either with XRD or with a sensitive surface technique as XPS. Moreover, in diffractogram patterns (Figure 2) there is a peak about 2θ equal 57°. This peak is probably due to the presence of some Fe (or/and Pd) compounds. These compounds could be at the origin of a specific layered structure which could be attributed to a germination of a hexaaluminate phase. The catalyst (Pd(2 wt %)/Fe-Al oxide) used for the monolith coating does not undergo any textural, structural, or crystallinity modification during the deposition procedure. The same PdO and γ-Al2O3 phases are observed both in the powder and in structured catalysts (Figures 4 and 5). Moreover, there are no significant differences between the diffractograms of the fresh catalyst and tested structured monoliths. The values of the atomic Pd/(Fe + Al) ratio for the coated Pd(2 wt %)/Fe-Al oxide catalysts on all monoliths are lower than those of the corresponding tested powder catalyst. However, since XPS peaks for aluminum and iron are present in monoliths, it is normal to find values for Pd/(Fe + Al) ratios which are smaller than the value obtained for the powder catalyst. Concerning the Fe/Al atomic ratio, the values for structured metallic monoliths are similar to the value observed for the tested powder. In the case of ceramic monoliths, an increase of the ratio was noticed. This can be explained by the presence of iron in the monolith used to structurate the catalyst. Nevertheless, it seems that the surface of the structured powder catalyst maintains the same composition during the coating procedure. The powder catalysts have not been modified during the deposition process. Coating a catalyst on ceramic monoliths led to a specific surface area (SBET) decrease when compared to the specific surface of the ceramic monolith itself. This could be explained because the structured catalyst covers part of the surface of the pores of the monolith. The fact that the coatings blocks the porosity of the monolith is also reported by Perez-Cadenas.53 On the contrary, coating a catalyst on metallic monoliths leads to a specific surface area increase when compared to the pure monolith. This could be explained by a more significant contribution of the powder catalyst to the overall specific surface area of the monolith, in agreement with previous data.47 Catalytic Performances. If the T10, T30, T50, T70, and T90 values are considered (Table 5), it is observed that LaMnO3 displays higher catalytic activities. LaMnO3 catalysts displayed 10% methane conversion at 593 K. These temperatures are higher than those reported by Cimino et al.,57 whose catalyst displayed a 10% methane conversion at 472 K and reached 90% at 646 K. Temperatures observed with LaMnO3 are nevertheless lower than those found for Fe-Al oxide and Pd(2 wt %)/Fe-Al oxide powder catalysts. However, the conversion of the LaMnO3 catalyst reaches 50% at 727 K, a temperature similar to the one

found for Pd(2 wt %)/Fe-Al oxide but still lower than the one found for Fe-Al oxide. When we compared both Fe-Al oxide base catalysts, it could be observed that Pd(2 wt %)/Fe-Al oxide displays quite a higher activity. This clearly showed the beneficial effect of the support for the stabilization of the PdOx active phase, in agreement with data reported by Baylet et al.38 The authors reported that the pure oxide samples (i.e., without palladium) present low catalytic activity for methane total oxidation compared to a reference Pd/Al2O3 catalyst. In our previous work, we showed that catalytic activity in methane combustion is higher for those monoliths containing a lower amount of catalyst.47 The results obtained with Pd(2 wt %)/Fe-Al oxide structured monoliths are different: there are no significant differences (Figure 8). For Pd(2 wt %)/Fe-Al oxide ceramic monoliths, there is no difference below 813 K. Above, the best catalytic performances are obtained for the one containing a lower amount of catalyst (PdCMOS3, 0.0538 g). However, for the Pd(2 wt %)/Fe-Al oxide metallic monoliths, better catalytic performances seem to be obtained by the one with the highest catalyst mass (PdMMDC3, 0.1325 g). Furthermore, it could be noticed that the amount of deposited catalyst on a monolith, using the Pd(2 wt %)/Fe-Al oxide slurry, is lower than that deposited using LaMnO3 slurry (Table 4). These results suggest that the catalytic performance increases when the amount of deposited catalyst increases. However, if the amount of deposited catalyst is too high, the activity decreases. Hence, there is a critical amount of catalyst that should be taken into account in order to optimize subsequently the catalytic performances. It is safe to assume that the initial amount of deposited catalyst has an influence on the catalytic activity. The high activity stand for PdCMOS3 could be related to the dispersion of the catalysts on the surface of the monolith. It is important to underline that the high Fe/Al ratio is explained by the content of Fe in the ceramic monolith. In fact, XPS results show an Fe/Al ratio of 0.086 for PdCMOS3 compared to 0.075 for PdCMDC3. This reveals that the surface of PdCMOS3 is rich in Fe. In the case of the metallic structured catalyst, the dispersion of Fe seems to be not too different (XPS atomic intensity ratio Fe/Al is 0.42 and 0.40 for PdMMOS3 and PdMMDC3, respectively). In this case, the difference could be attributed to a high oxidation state of Pd. Binding energy values could indicate that during the reaction the palladium on PdMMDC3 is a little more oxidized (337.7 compared to 337.3). The binding energy after the test for the powder catalyst is only 337.1 eV. This increase in the oxidation state of palladium could be related to a support effect. The surface of the metallic monolith promotes this oxidation. When the catalytic results obtained for powder and monolithic catalysts are compared, it could be observed that, at low temperature, the catalytic performances of Pd(2 wt %)/Fe-Al oxide metallic monoliths are higher than those of the powder catalysts. Nevertheless, at low temperature Pd(2 wt %)/Fe-Al oxide ceramic monoliths have a lower catalytic activity compared to the powder catalysts. At high temperature, PdCMOS3 has the highest activity. These results confirm the fact that the surface of the support plays an important role in the catalytic performance of structured catalysts, as concluded with LaMnO3 perovskite catalysts on ceramic and metallic monoliths.47 Coating Pd(2 wt %)/Fe-Al oxide on the ceramic monoliths leads to an increasing activity at higher temperature compared to the corresponding powder catalyst. On the contrary, coating Pd(2 wt %)/Fe-Al oxide on the metallic monoliths leads to a better activity at low temperature.

16514

J. Phys. Chem. C, Vol. 113, No. 37, 2009

TABLE 6: Apparent Activation Energies Ea (kJ mol-1) PdCMOS3 PdCMDC3 PdMMOS3 PdMMDC3 Pd(2wt %)/Fe-Al oxide

107.3 83.8 12.2 14.6 42

Arendt et al.

S(monolith) + S(catalyst) - S(covered) ) S(structured catalyst) (2) fraction of coverage )

T(coating) ) To go a step further, the results of activity versus temperature are reported with the Arrhenius plot form (ln(activity) vs 1000/T (kelvin)). In Figure 9, Arrhenius plots of the catalysts are shown, and apparent activation energies calculated from these plots are given in Table 6. It can be observed that the Ea’s of the metallic monoliths are lower than both the Ea of the powder catalyst and the Ea of the ceramic monoliths. The difference in the nature of the support could explain that the Ea’s of the metallic monoliths are lower than the Ea of the ceramic monoliths. This fact confirms previous work,47 namely that the support plays a key role in the catalytic combustion. In addition, metallic monoliths could allow development of a homogeneous temperature profile through the monolith which could decrease the ignition temperature. This can explain that the Ea is lower than for the powder. Concerning the Ea of the ceramic monoliths, we can observe that the Ea is higher than that of the powder. The ceramic support is a porous support (SBET ) 140 m2/g), so it can be suggested that the mass transfer or pore limitation could be more important. If the reaction is limited by diffusion with the ceramic monoliths, the apparent activation energies would be different compared to powder. As a consequence, the difference of the Ea’s could be explained by the presence of the support. Indeed, the support could play a key role for the stabilization of the PdOx active phase.58 In other words, the increase in the oxidation state of palladium could be related to a support effect. The support could facilitate the adsorption of reactants or the mobility of chemical process species (i.e., oxygen spillover) which could be important in the reactivity or in the control of the more adequate oxidation state of catalytic site. Fraction Coverage and Thickness of the Coating. Since the catalyst loading differs in monolithic catalysts, the surface of the monolith covered by the powder catalyst [S(covered)], the fraction of coverage of the surface of the monolith, and the thickness of the coating [T(coating)] could be expressed according to the following equations:

S(covered) S(monolith)

V(powder) S(covered)

(3)

(4)

where S(monolith) (m2) represents the BET specific surface area of the monolith tested, without catalyst (m2/g) (Table 2), times the weight of the monolith in 1 g of structured catalyst (g); S(catalyst) (m2) is the BET specific surface area of Pd(2 wt %)/ Fe-Al oxide tested (m2/g) (Table 2) times the loading of catalyst in 1 g of structured catalyst (g); S(covered) (m2) is the surface of the monolith covered by the catalyst (m2); S(structured catalyst) (m2) is the BET specific surface area of structured catalyst (Table 2) for 1 g of structured catalyst; V(powder) is the volume occupied by the powder Pd(2 wt %)/Fe-Al oxide catalyst (cm3) (for Pd(2 wt %)/Fe-Al oxide powder, 0.9560 g occupied 1 cm3); and T(coating) represents the thickness of the coating (µm). Table 7 displays the results for other monoliths. The fraction of coverage and the thickness of coating on metallic monolith were not evaluated because the BET specific surface areas are significantly smaller than those of the ceramic ones (1.5 compared to 140.0 m2 g-1 in Table 2) and the variations of the specific surface areas due to coverage of the monolith by the powder catalysts could lead to a misinterpretation. It can be observed that the orbital technique led to a smaller thickness: for the same number of cycles (three), the thickness of the coating for PdCMOS3 (0.0006 µm) is smaller than the thickness for PdCMDC3 (0.0016 µm). This is in agreement with our previous work, but it seems that the thicknesses obtained with LaMnO3 slurry are higher than those obtained with Pd(2 wt %)/Fe-Al oxide slurry. This shows that the nature of the slurry also influences the thickness of the coating. The higher the amount of catalyst, the higher the thickness of the coating (i.e., the thickness of the coating for PdCMDC3 (0.1217 g) is 0.0016 µm, this value is higher than that observed for PdCMOS3 (0.0538 g), 0.0006 µm (Table 7)); this is in agreement with results reported in ref 52. On the other hand, the fraction of coverage of the surface of ceramic monoliths with Pd(2 wt %)/Fe-Al oxide seems to be

Figure 9. Arrhenius plots for the catalyst coated on ceramic monoliths and on metallic monoliths by orbital stirring (PdCMOS3 and PdMMOS3) and dip coating (PdCMDC3 and PdMMDC3). The result for the powder catalyst (Pd(2 wt %)/Fe-Al oxide) is also shown.

Structuration of Pd(2 wt %)/Fe-Al Oxide Catalysts

J. Phys. Chem. C, Vol. 113, No. 37, 2009 16515

TABLE 7: Fraction of Coverage of the Surface of Monolith and Thickness of the Coatinga

PdCMOS3 PdCMDC3 PdMMOS3 PdMMDC3 LaCMOS3 LaCMDC3 LaCMDC1 LaMMOS11 LaMMOS3 LaMMDC9 LaMMDC3

S(monolith) (m2

S(catalyst) (m2)

S(structured catalyst) (m2)

138.6 136.5 1.5 1.5 133.1 127.2 131.6 1.4 1.5 1.4 1.5

0.34 0.75 0.41 0.76 0.9 1.7 1.1 1.0 0.3 1.0 0.3

116.7 119.6 1.8 2.2 117.0 116.0 119.0 3.5 2.5 1.0 1.0

S(covered) (cm-3) 22.24 17.65 0.11 0.06 17.0 12.8 13.7

-

fraction of coverage 0.16 0.13 0.13 0.10 0.10

-

-

V(powder) (cm-3) 0.01 0.03 0.02 0.03 0.7 1.2 0.8 0.7 0.3 0.7 0.25

T(coating) (µm) 0.0006 0.0016 0.04 0.10 0.06

-

-

a S(monolith) (m2) represents the BET specific surface area of the monolith tested, without catalyst (m2/g) (Table 2) times the weight of the monolith in 1 g of structured catalyst (g). S(catalyst) (m2) is the BET specific surface area of Pd(2 wt %)/Fe-Al oxide tested (m2/g) (Table 2) times the loading of catalyst in 1 g of structured catalyst (g). S(covered) (m2) is the surface of the monolith covered by the catalyst (m2). S(structured catalyst) (m2) is the BET specific surface area of structured catalyst (Table 2) for 1 g of structured catalyst. V(powder) is the volume occupied by the powder Pd(2 wt %)/Fe-Al oxide catalyst (cm3) (for Pd(2 wt %)/Fe-Al oxide powder, 0.9560 g occupied 1 cm3). T(coating) represents the thickness of the coating (µm). The specific surface areas of ceramic and metallic monoliths are 140 and 1.5 m2 g-1, respectively. Results obtained with LaMnO3 structured catalysts are also displayed.

higher than the fraction of coverage with LaMnO3 slurry. For monoliths prepared by orbital stirring, the fraction of coverage is 0.16 for PdCMOS compared to 0.13 for LaCMOS3, and for monoliths prepared by dip coating, the fraction of coverage is 0.13 for PdCMDC3 compared to 0.10 for LaCMDC3. To point out the diffusion limitation, the thickness of the coating and the specific catalytic activity of the ceramic monoliths could be taken into account. On rapid kinetics, as in the case of catalytic combustion of methane, the reaction could be limited by the diffusion of the reactants inside the pores. The diffusion process will depend on the thickness of the coating. In the present case, for ceramic monoliths, the thickness obtained with Pd/Fe-Al oxide slurry is lower than that observed with LaMnO3 slurry. This indicates that the diffusion limitation process will play a less significant role when Pd/Fe-Al oxide is coated. The fraction coverage measurements confirm this observation. Obviously, these are only indications. A definitive conclusion concerning the influence of diffusion limitation must be performed during complete mass transfer and limitation studies of the monoliths. These studies are outside the scope of the present research. Conclusion The results obtained in this study confirm that the amount of deposited catalyst depends on the nature of the monolith, on the coating technique used, and on the nature of the slurry. Besides, the acidity of the slurry could modify the surface electric charge of the monolith and, as a consequence, could influence the amount of deposited catalyst. These indicate that the pH of the slurry is an important parameter to be controlled in order to optimize the amount of the coating. Comparisons between structured Pd(2 wt %)/Fe-Al oxide and structured LaMnO3 on ceramic and metallic monoliths show that the amount of deposited catalysts on ceramic monoliths using Fe-Al oxide is lower than that deposited using LaMnO3 (previous work). In the case of metallic monoliths, the difference is less significant. The Fe-Al oxide thickness of the coated catalysts on the monoliths with the Fe-Al oxide seems to be less important than with LaMnO3. This indicates that the diffusion limitation in the Fe-Al oxide could be less important. Moreover, the nature of the monolith and the nature of the slurry strongly affect the catalytic activity of the monolithic catalysts. On one hand, higher catalytic activities were observed for Pd(2

wt %)/Fe-Al oxide on the ceramic monoliths at higher temperatures. On the other hand, coating Pd(2 wt %)/Fe-Al oxide on the metallic monoliths leads to a better activity at low temperature compared to the corresponding powder catalyst. These differences could be explained either by a better dispersion of Pd(2 wt %)/Fe-Al oxide or by a higher oxidation state of Pd. Monolithic catalysts are promising systems to replace catalysts in conventional multiphase reactors. Pilot-scale experiments deserve further investigation. Acknowledgment. E.A. thanks the Universite´ catholique de Louvain and the teaching assistantsPh.D. student position. The authors wish to thank the “Direction Ge´ne´rale des Technologies, de la Recherche et de l’Energie of the Re´gion Wallonne” (Belgium) and the “Fonds National de la Recherche Scientifique” (Belgium) for the acquisition of the XPS equipment and the “Interuniversity attraction pole (IAP), phase VI, Belgian Science Policy ” “advanced complex inorganic materials by a novel bottom-up nanochemistry approach: processing and shaping”, INANOMAT, for financial support. The Unite´ de catalyse et chimie des mate´riaux divise´s is also involved in the “EMMI” European Multifunctionnal Material Institute built from the previous “FAME” Network of Excellence of the EU 6th FP, and in the Cost Action D41 sustained by the European Science Foundation. References and Notes (1) Hicks, R. F.; Qi, H.; Young, M. L.; Lee, R. G. J. Catal. 1990, 122, 280–294. (2) Pfefferle, L. D.; Pferfferle, W. C. Catal. ReV.sSci. Eng. 1987, 29, 219–267. (3) Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Catal. ReV.sSci. Eng. 1984, 26, 1–58. (4) Trimm, D. L. Appl. Catal. 1983, 7, 249–282. (5) Choudhary, T. V.; Banerjee, S.; Choudhary, V. R. Appl. Catal. A: Gen. 2002, 234, 1–23. (6) Arai, H.; Fukuzawa, H. Catal. Today 1995, 26, 217–221. (7) Zwinkels, M. F. M.; Jaras, S. G.; Menon, P. G. Catal. ReV.sSci. Eng. 1993, 35, 319–358. (8) Forzatti, P.; Groppi, G. Catal. Today 1999, 54, 165–180. (9) Arai, H.; Machida, M. Catal. Today 1991, 10, 81–94. (10) Ge´lin, P.; Primet, M. Appl. Catal. B: EnViron. 2002, 39, 1–37. (11) Saint-Just, J.; der Kinderen, J. Catal. Today 1996, 29, 387–395. (12) Burch, R.; Urbano, F. J. Appl. Catal. A: Gen. 1995, 124, 121–138. (13) Lyubovsky, M.; Pfefferle, L. Catal. Today 1999, 47, 29–44. (14) Arai, H.; Machida, M. Appl. Catal. A: Gen. 1996, 138, 161–176.

16516

J. Phys. Chem. C, Vol. 113, No. 37, 2009

(15) Church, J. S.; Cant, N. W.; Trimm, D. L. Appl. Catal. A: Gen. 1993, 101, 105–116. (16) Ciuparu, D.; Lyubovsky, M. R.; Altman, E.; Pfefferle, L. D.; Datye, A. Catal. ReV. 2002, 44, 593–649. (17) Yashnik, S. A.; Ismagilov, Z. R.; Kuznetsov, V. V.; Ushakov, V. V.; Rogov, V. A.; Ovsyannikova, I. A. Catal. Today 2006, 117, 525–535. (18) Pena, M. A.; Fierro, J. L. G. Chem. ReV. 2001, 101, 1981–2017. (19) Arai, H.; Yamada, T.; Eguchi, K.; Seiyama, T. Appl. Catal. 1986, 26, 265–276. (20) Isupova, L. A.; Alikina, G. M.; Snegurenko, O. I.; Sadykov, V. A.; Tsybulya, S. V. Appl. Catal. B: EnViron. 1999, 21, 171–181. (21) Tejuca, L. G.; Fierro, J. L. G.; Tascon, J. M. D. AdV. Catal. 1989, 36, 237–328. (22) Baran, E. J. Catal. Today 1990, 8, 133–151. (23) Yamazoe, N.; Teraoka, Y. Catal. Today 1990, 8, 175–199. (24) Song, K.-S.; Cui, H. X.; Kim, S. D.; Kang, S.-K. Catal. Today 1999, 47, 155–160. (25) Ciambelli, P.; Cimino, S.; DeRossi, S.; Lisi, L.; Minelli, G.; Porta, P.; Russo, G. Appl. Catal. B: EnViron. 2001, 29, 239–250. (26) Oliva, C.; Forni, L.; Pasqualin, P.; D’Ambrosio, A.; Vishniakov, A. Phys. Chem. Chem. Phys. 1999, 1, 355–360. (27) Marti, P. E.; Baiker, A. Catal. Lett. 1994, 26, 71–84. (28) Marchetti, L.; Forni, L. Appl. Catal. B: EnViron. 1998, 15, 179– 187. (29) Civera, A.; Negro, G.; Specchia, S.; Saracco, G.; Specchia, V. Catal. Today 2005, 100, 275–281. (30) Ciambelli, P.; Cimino, S.; DeRossi, S.; Faticanti, M.; Lisi, L.; Minelli, G.; Porta, P.; Russo, G.; Turco, M. Appl. Catal. B: EnViron. 2000, 24, 243–253. (31) Seiyama, T. Catal. ReV.sSci. Eng. 1992, 34, 281–300. (32) Machida, M.; Eguchi, K.; Arai, H. J. Catal. 1989, 120, 277–286. (33) Machida, M.; Eguchi, K.; Arai, H. J. Catal. 1987, 103, 385–393. (34) Chu, W.; Yang, W.; Lin, L. Catal. Lett. 2001, 74, 139–144. (35) Eguchi, K.; Arai, H. Catal. Today 1996, 29, 379–386. (36) Jang, B. W. L.; Nelson, R.; Spivey, J. J.; Ocal, M.; Oukaci, R.; Marcelin, G. Catal. Today 1999, 47, 103–113. (37) Sekizawa, K.; Widjaja, H.; Maeda, S.; Ozawa, Y.; Eguchi, K. Catal. Today 2000, 59, 69–74. (38) Baylet, A.; Royer, S.; Mare´cot, P.; Tatiboue¨t, J. M.; Duprez, D. Appl. Catal. B: EnViron. 2008, 77, 237–246.

Arendt et al. (39) McCarty, J. G.; Gusman, M.; Lowe, M.; Hildenbrand, D. L.; Lau, K. N. Catal. Today 1999, 47, 5–17. (40) Pocoroba, E.; Johansson, E. M.; Ja¨ras, S. G. Catal. Today 2000, 59, 179–189. (41) Sohn, J. M.; Kang, S. K.; Woo, S. I. J. Mol. Catal. A: Chem. 2002, 182, 135–144. (42) Burgos, N.; Paulis, M.; Montes, M. J. Mater. Chem. 2003, 13, 1458– 1467. (43) Cybulski, A.; Moulijn, J. Catal. ReV.sSci. Eng. 1994, 36, 179– 270. (44) Avilla, P.; Montes, M.; Miro, E. E. Chem. Eng. J. 2005, 109, 11–26. (45) Irandoust, S.; Andersson, B. Catal. ReV.sSci. Eng. 1988, 30, 341– 392. (46) Geus, J. W.; VanGiezen, J. C. Catal. Today 1999, 47, 169–180. (47) Arendt, E.; Maione, A.; Klisinska, A.; Sanz, O.; Montes, M.; Suarez, S.; Blanco, J.; Ruiz, P. Appl. Catal. A: Gen. 2008, 339, 1–14. (48) Blanco, J.; Avila, P.; Suarez, S.; Yates, M.; Martin, J. A.; Marzo, L.; Knapp, C. Chem. Eng. J. 2004, 97, 1. (49) Moulder, J. F.; Stickel, W. F.; Sobel, P. E.; Bomben, K. D. Handbook of X-ray Photoelectron Spectroscopy; Chastain, J., Ed.; PerkinElmer Corporation: Eden Prairie, MN, 1992. (50) Schmitz, P. J.; Otto, K.; deVries, J. E. Appl. Catal. A: Gen. 1992, 92, 59–72. (51) Otto, K.; Haack, L. P.; deVries, J. E. Appl. Catal. B: EnViron. 1992, 1, 1–12. (52) Valentini, M.; Groppi, G.; Cristiani, C.; Levi, M.; Tronconi, E.; Forzatti, P. Catal. Today 2001, 69, 307–314. (53) Perez-Cadenas, A. F.; Kapteijn, F.; Moulijn, J. A. Appl. Catal. A: Gen. 2007, 319, 267–271. (54) Gonzalez-Velasco, J. R.; Gutierrez-Ortiz, M. A.; Marc, J. L.; Botas, J. A.; Gonzalez-Marcos, M. P.; Blanchard, G. Ind. Eng. Chem. Res. 2003, 42, 311–317. (55) Meille, V. Appl. Catal. A: Gen. 2006, 315, 1–17. (56) Lietti, L.; Cristiani, C.; Groppi, P.; Forzatti, P. Catal. Today 2000, 59, 191–204. (57) Cimino, S.; Lisi, L.; Pirone, R.; Russo, G.; Turco, M. Catal. Today 2000, 59, 19–31. (58) Demoulin, O.; Navez, M.; Gaigneaux, E. M.; Ruiz, P.; Mamede, A.; Granger, P.; Payen, E. Phys. Chem. Chem. Phys. 2003, 5, 4394–4401.

JP901056Z