3: To... , ”W. .1 .I . :3. . .suiyfiw... . . ‘ a! . fin v u “fill“ ‘ \ . 0 1:. .3...- _.1 1 s i... 1. V ,1 I is!) .3- up.“ JP; 3 can , ms mama I i ‘3 . 5': . fih’gfi" ”E i. .. an. . : Eb, :w\9wrr:aq‘ ! I m “whims...“ |<. .él .Pnuwkfiu . ‘ . ‘ . ,L. 3...:2 a. is > . x45)? Hi}.:iiihh... . . . ‘ we. ‘nfiv.d. 4.4V \ 14.9..“ fiSrafiFJvu-fi. v . . ; Lu“... .b...:m..m.. «tamiflu . . . . £1... . 19$» . . X! it... in .1... - “firetfl 2:333 23.5.: kl»; bu. : :1)?! I \vil xlv t; . . , x :1 . n t .it. ~ .I .21 .110 ”:10! iii i531“; I/iiiiiiflWii/‘ii‘l/iil This is to certify that the dissertation entitled THE INFLUENCE OF SURFACE STRUCTURE ON THE CATALYTIC ACTIVITY OF ALUMINA SUPPORTED TRANSITION METAL OXIDES: A STUDY OF CARBON MONOXIDE AND METHANE OXIDATION presented by Paul Worn Park has been accepted towards fulfillment of the requirements for PhD. degree in _£hemisin¥ v/mgfl]gww4 or professor Date WI?) M7! M5 U is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY Michigan State University PLACE IN RETURN BOX to roman this chockout from your record. To AVOID FINES return on or baton date duo. DATE DUE DATE DUE DATE DUE 7 i -_.____— THE INFLUENCE OF SURFACE STRUCTURE ON THE CATALYTIC ACTIVITY OF ALUMINA SUPPORTED TRANSITION METAL OXIDES: A STUDY OF CARBON MONOXIDE AND METHANE OXIDATION By PAUL WORN PARK A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1996 ABSTRACT THE INFLUENCE OF SURFACE STRUCTURE ON THE CATALYTIC ACTIVITY OF ALUMINA SUPPORTED TRANSITION METAL OXIDES: A STUDY OF CARBON MONOXIDE AND METHANE OXIDATION By PAUL WORN PARK The projects described in this thesis examine the effect of rare earth and transition metal oxide promoters on the structure and reactivity of base metal oxide/alumina catalysts for total oxidation reactions. The research involves the preparation and testing of alumina supported Single and mixed metal oxide catalysts designed for total oxidation reactions. The catalysts were prepared with various active phase loadings, promoter contents, and precursors to control the chemical state and structure of supported phases. The identification and quantification of various supported phases have been determined using surface (XPS), bulk (XRD, ESR, FTIR. TGA, BET), and microprobe (TEM) techniques. The information derived from these techniques was correlated with CO and CH4 oxidation activities to identify active sites and to examine the effect of promoter on the structure and reactivity of the catalysts. The effect of crystallinity on the photoreduction of Ce(IV) species during XPS analysis of CeOz samples and Ce/A1203 catalysts has been determined. The crystallinity of the samples was controlled by varying the calcination temperature or by using difierent Ce precursors (ammonium Ce(IV) nitrate and Ce(IV) methoxyethoxide). XPS, XRD, and TEM results indicated that the amorphous cerium oxide was reduced more extensively than the crystalline material during XPS analysis. XRD, XPS, and ESR results indicated that an isolated c0pper surface phase, an interacting copper surface phase, and large CuO crystallites are formed on Cu/A1203 catalysts, depending on Cu content. The crystalline CuO and isolated copper surface phase has been assigned to the active site for CO and CH4 oxidation, respectively. Cu/Ale3 catalyst prepared using Cu(II) ethoxide showed higher Cu dispersion, less crystalline CuO, and lower oxidation activity for CO and CH4 than the catalyst derived from Cu(II) nitrate. For Cu/Ce/A1203 catalysts, Ce had little effect on the dispersion and crystallinity of the copper species. However, Cu addition decreased the Ce dispersion and increased the amount of crystalline CeOz. Cerium addition dramatically increased the CO oxidation activity, but it had little effect on CH4 oxidation. This indicated that cerium strongly interacted with crystalline CuO, but not with copper surface phase. The characterization of Cr/A1203 catalysts indicated that Cr was present as a highly dispersed Cr6+ surface phase, Cr3+ clusters, and large Cr203 crystallites, depending on Cr content. The active phase for CH4 oxidation has been assigned to a Cr(III)-Cr(VI) interaction species. For the Cu/Cr/Ale3 catalysts, Cu addition decreased the dispersion of the chrOmium phase by reacting selectively with a dispersed Cr3+ species to form CuCr204. The Cu dispersion also decreased. with increasing Cr loading due to the formation of CuO and CuCrZOI. These phases contributed to the catalytic activity of CO oxidation with increasing Cr content up to Cr/Al = 0.054. For further Cr addition, the catalytic activity decreased due to the decrease in catalyst dispersion or to encapsulation of the active site by excess Cr species. ACKNOWLEDGMENTS I would like to thank Dr. Jeff S. Ledford, my research adviser, for lead me to the fields of surface science and catalysis which are my all time-favorite subjects. I greatly appreciate for his advice, correction, and encouragement whenever I needed. I also appreciate professor Thomas J. Pinnavaia, my official mentor, who understands and supports me constantly throughout my graduate school years at Michigan State University. There were many technical staffs in the department who deserve to get credits for my research. Dr. Atkinson: He have always been available when I needed to fix and to upgrade my computer which becomes the best PC in the chemistry building. The Glass Shop: Their glassblowing technique must be one of the best in the country. I cannot imagine without them how I could do my experiments with my easy-breaking hand. The Machine Shop: They designed lot of tools and saved me a lot of money. The Electronic Shop & The Electric Designer: Their magic fingers created and fixed my sophisticated instrument and saved me from the disaster. I have several friends whose friendship have certainly made my graduate school life enjoyable. I will remember the friends, Tom Curtis, Ed Thounsand, Jeff Rasimas, Mike Thelen, Mark Waner, Radu Craciun, and Dana Spence. Especially, I thank our final five orphans Kathy Severin, Greg Noonan, Lee Hoffman, Per Askeland for our discussion and encouragement each other. I wish you my best regards and to keep our friendship forever. This work could never be completed without the support of my Mothers and Fathers who are taking care of me and my family with endless and devoted love. Finally, this thesis will be dedicated to my lovable wife, Jongmi who sacrifices herself to support me and to take care of our kids. I will promise you to be a better husband and a father. TABLE OF CONTENTS List of Tables ..................................................................................... ix List of Figures .................................................................................... xi Chapter 1. Air Pollution and Pollution Control Catalysts 1.1. 1.2. 1.3. Air Pollution ....................................................................... 1 1.1.1. Introduction .............................................................. 1 1.1.2. Origin and Effect of Air pollutants ..................................... 2 1.1.3. Pollution Control. Techniques ........................................... 4 1.1.3.1. Recovery Techniques ......................................... 5 1.1.3.2. Combustion Techniques ...................................... 6 Pollution Control Catalysts ...................................................... 7 1.2.]. Introduction .............................................................. 7 1.2.1.1. Principle of Catalytic Incineration ........................... 7 1.2.1.2. Advantages of Catalytic Pollution Control ................. 8 1.2.1.3. Disadvantages of Catalytic Pollution Control .............. 8 1.2.2. Noble Metal Catalysts ................................................... 9 1.2.2.1. Advantages of Noble Metal Catalysts ....................... 9 1.2.2.2. Disadvantages of Noble Metal Catalysts ................... 10 1.2.3. Transition Metal Oxide Catalysts ..................................... 11 1.2.3.1. Advantages of Transition Metal Oxide Catalysts ......... 11 1.2.3.2. Principles of Transition Metal Oxide Catalysts ............ 12 1.2.4. Industrial Waste Treatment Catalysts ................................. 14 1.2.4.1. Volatile Organic Compounds (VOC) Emission ........... 14 1.2.4.2. NOx Emission ................................................. 16 1.2.5. Automobile Emission Control Catalysts .............................. 21 References ........................................................................ 28 Chapter 2. Catalysts Preparation and Characterization Techniques 2.1. Preparation of Catalysts ......................................................... 35 2.1.1. ‘Y-A1203 ................................................................... 36 2.1.1.1. Preparation of'y-A1203 ....................................... 36 vi 2.2. 2.3. 2.4. 2.5. Chapter 3. 3.1. 3.2. 3.3. 3.4. 3.5. Chapter 4. 4.1. 4.2. 4.3. 4.4. 4.5. 2.1.1.2. Structure ofy-A1203 .......................................... 36 2.1.1.3. Advantages of y-A1203 support ............................. 38 2.1.2. Catalysts Preparation Techniques ..................................... 40 2.1.2.1. Incipient wetness impregnation ............................. 40 2.1.2.2. Ion exchange .................................................. 41 2.1.2.3. Grafting ........................................................ 41 2.1.2.4. Precipitation ................................................... 42 Bulk Characterization ........................................................... 43 2.2.1. Surface Area Measurement (BET) .................................... 43 2.2.2. Electron Spin Resonance (ESR) ....................................... 48 2.2.3. X-ray Diffraction (XRD) ............................................... 51 2.2.3.1. Qualitative analysis of XRD ................................. 51 2.2.3.2. Quantitative analysis of XRD ............................... 54 2.2.3.3. Particle size determination ................................... 55 Surface Characterization ........................................................ 56 2.3.1. Ultrahigh Vacuum Chamber ........................................... 56 2.3.2. Transmission Electron Spectroscopy (TEM) ........................ 57 2.3.2.1. Instrumentation for TEM .................................... 60 2.3.2.2. Sample preparation for TEM ................................ 62 2.3.3. X-ray Photoelectron Spectroscopy (XPS) ........................... 63 2.3.3.1. Qualitative analysis of XPS .................................. 63 2.3.3.2. Quantitative analysis ofXPS ................................ 65 2.3.3.3. Particle size determination ................................... 66 2.3.3.4. XPS shake up satellite peaks ................................ 69 Catalytic Activity Measurement ................................................ 71 References ........................................................................ 74 The Effect of Crystallinity on the Photoreduction of Cerium Oxide: A Study of CeO; and CC/Ale3 Catalysts Abstract ........................................................................... 77 Introduction ...................................................................... 78 Experimental ..................................................................... 80 Results and Discussion .......................................................... 84 References ...................................................................... 100 The Influence of Surface Structure on the Catalytic Activity of Alumina Supported Copper Oxide Catalysts: Oxidation of Methane and Carbon Monoxide Abstract ......................................................................... 1 03 Introduction .................................................................... 104 Experimental ................................................................... 106 Results .......................................................................... l 1 1 Discussion ...................................................................... l 18 vii 4.6. 4.7. Conclusions ..................................................................... 122 References ...................................................................... 1 24 Chapter 5. Characterization and CH4 Oxidation Activity of Cr/Ale3 Catalysts 5.1. 5.2. 5.3. 5.4. 5.5. 5.6. Abstract ......................................................................... 1 27 Introduction .................................................................... 128 Experimental ................................................................... 1 3 0 Results and Discussion ........................................................ 134 Conclusions ..................................................................... 144 References ...................................................................... 145 Chapter 6. Characterization and CO Oxidation Activity of Cu/Cr/A1203 Catalysts 6.1. 6.2. 6.3. 6.4. 6.5. 6.6. Abstract ......................................................................... 149 Introduction .................................................................... 1 50 Experimental ................................................................... 1 52 Results and Discussion ........................................................ 157 Conclusions ..................................................................... 171 References ...................................................................... 1 72 Chapter 7. Influence of Cerium Promotion and Catalyst Precursors on the 7.1. 7.2. 7.3. 7.4. 7.5. 7.6. Structure and Total Oxidation Activity of Alumina Supported Copper Oxide Catalysts Abstract ......................................................................... 175 Introduction .................................................................... 176 Experimental ................................................................... 178 Results and Discussion ........................................................ 183 Conclusions ..................................................................... 194 References ...................................................................... 196 viii LIST OF TABLES Table ........................................................................................... Page Table 3.1. Particle Sizes of Cerium Oxide Species in Ce/A1203 Catalysts Determined from XRD Line Broadening and XPS Intensity Ratios Measurements ...................................................................... 97 Table 4.1. Concentration of Crystalline Phases in Cuy Catalysts Calculated from Quantitative XRD Data ......................................................... 113 Table 4.2. Particle Size of Copper Phases Determined from XPS Cu/Al Intensity Ratios and XRD Line Broadening Calculations .............................. 115 Table 4.3. Turn Over Numbers and Activation Energies of Cu/A1203 Catalysts for CO and CH4 Oxidations ..................................................... 117 Table 5.1. Concentration of Crystalline Phases in Cry Catalysts Calculated from Quantitative XRD Data ......................................................... 137 Table 5.2. XPS Cr 2p3a/A1 2p intensity ratios of Cr(VI) and Cr(III) Species in Cry Catalysts Determined Using Non-Linear Least-Squares Curve Fitting (NLLSCF) ................................................................ 139 Table 5.3. Particle Size and Dispersion of Chromium Phases Determined from XRD Line Broadening Calculations and XPS Cr/Al Intensity Ratios ...... 142 Table 5.4. Turn Over Numbers and Activation Energies of Cry Catalysts for CH4 Oxidation. ................................................................... 144 Table 6.1. XPS Cu 2pm Binding Energies and shakeup/main Peak Intensity Ratios Measured for CuCry Catalysts and Standard Compounds .......... 158 ix Table 6.2. Table 6.3. Table 6.4. Table 6.5. Table 6.6. Table 7.1. Table 7.2. Table 7.3. Distribution of Cr Oxidation States Determined from XPS Cr 2pm Spectra of CuCry Catalysts Measured by non-linear least-squares curve fitting (NLLSCF) ......................................................... 162 Concentration of Crystalline Phases in CuCry Catalysts Calculated from Quantitative XRD Data ................................................... 162 Particle Sizes of Chromium and Copper Species Determined from XPS Intensity Ratios ............................................................ 166 Particle Sizes of CuO, CT303, and CuCrZO. Phases Determined from XRD Line Broadening Calculations ........................................... 166 Turn Over Numbers and Activation Energies of CuCry Catalysts for CO Oxidation ..................................................................... 169 Concentration of Crystalline Phases in Cu/Ce/A1203 Catalysts Calculated from Quantitative XRD Data ...................................... 186 Particle Size of Cerium and Copper Phases Determined from XRD Line Broadening Calculations and XPS Intensity Ratios .................... 187 Turn Over Numbers and Activation Energies of Cu/Ce/Ale3 Catalysts for CO and CH4 Oxidation ........................................... 192 Figure ..... Figure 1.1. Figure 2.1. Figure 2.2. Figure 2.3. Figure 2.4. Figure 2.5. Figure 2.6. Figure 2.7. Figure 2.8. Figure 2.9. LIST OF FIGURES ..................................................................................... Page Simultaneous conversion of hydrocarbon (HC), CO, and NOx for TWC as a function of air/fuel ratio. (From Heck, R. M.; Farrauto, R. J. Catalytic Air Pollution Control: Commercial Technology; Van Nostrand Reinhold: New York, 1995) ........................................... 24 Schematic diagram of the formation of various alumina phases .............. 37 Lattice struture of spinel .......................................................... 39 The five types of adsorption isotherms (From Satterfield, C. N. Heterogeneous Catalysis in Practice; McGraw-Hill: New York, 1980) ................................................................................ 44 Energy level splitting of the electron produced by an applied magnetic field ....................................................................... 49 Constructive interference of scattered X-rays from crystalline Lattice ....... 53 Schematic drawings of XPS chamber ............................................ 58 Diagram of electron beam interaction with a solid sample .................... 59 TEM column (From Flegler, S. L.; Heckman, Jr. J. W.; Klomparens, K. L. Scanning and Transmission Electron Microscopy: An Introduction; W. H. Freeman and Company: New York, 1993 .............. 61 Diagram of XPS process .......................................................... 64 xi Figure 2.10. Figure 2.11. Figure 3.1. Figure 3.2. Figure 3.3. Figure 3.4. Figure 3.5. Figure 3.6. Figure 3.7. Figure 3.8. Figure 4.1. Figure 4.2. Figure 4.3. Figure 5.1. Kerkhof and Moulijn's model of the supported catalysts ...................... 67 Schematic of the catalytic reactor system ....................................... 72 Ce 3d XPS spectra measured for (a) CeOz and (b) Ce(III) acetylacetonate hydrate ........................................................... 85 TEM micrographs and SAED patterns obtained for (a) Ce200 and (b) Ce750 samples ..................................................................... 87 XRD patterns measured for (a) Ce200 and (b) Ce750 samples ............... 89 Ce 3d XPS spectra measured for the Ce200 sample after irradiation times of (a) 15 min., (b) 30 min, (c) 60 min., (d) 120 min, (e) 180 min., and (f) 300 min .............................................................. 91 Ce 3d XPS spectra measured for the Ce750 sample after irradiation times of (a) 15 min., (b) 30 min, (c) 60 min., (d) 120 min., (e) 180 min., and (f) 300 min .. ............................................................ 93 Fraction of CeOz reduced as a function of irradiation time for the Ce200 (O) and Ce750 (0) samples .............................................. 94 XRD patterns measured for (a) alumina support and (b) CeA and (c) CeN catalysts ....................................................................... 96 XPS Ce 3d spectra measured for (a) CeN and (b) CeA catalysts ............. 98 XRD patterns measured for (a) A1203, (b) Cu7.7, (c) Cu10, and (d) Cu13 ............................................................................... 112 Cu 2p33/A1 2p XPS intensity ratios measured for Cuy catalysts(.) plotted versus Cu/Al atomic ratio. Cu2p3/2/A12p intensity ratios calculated for monolayer dispersion (line) ..................................... 114 Normalized ESR intensity (1 mg CuO) of Cuy catalysts relative to the internal standard .................................................................. 1 l6 XRD patterns measured for (a) A1203, (b) Cr] 1, and (c) Crl3 ............. 136 xii Figure 5.2. Cr 2p XPS spectra obtained for (a) Cr1.3, (b) Cr2.7, (c) 054, (d) Cr8.0, (e) Cr] 1, and (f) Cr13 catalysts ........................................ 138 Figure 5.3. Cr 2p/Al 2p XPS intensity ratios measured for Cry catalysts (O) plotted versus Cr/Al atomic ratio. Cr 2p / A] 2p intensity ratios calculated for monolayer dispersion (line) ..................................... 141 Figure 6.1. XRD patterns measured for CuCry catalysts ( O alumina, I CuO, 0 cubic CuCrzO4, V tetragonal CuCrzO4, ‘il’ Cr203). (a) alumina (b) CuCrO, (c) CuCr1.3, (d) CuCr2.7, (e) CuCr5.4, (f) CuCr8.0, (g) CuCr10.7, and (h) CuCr13.4 ................................................... 159 Figure 6.2. Cr 2p3/2 XPS spectra measured for CuCry catalysts. (a) CuCr1.3, (b) CuCr2.7, (c) CuCr5.4, (d) CuCr8.0, (e) CuCr10.7, and (f) CuCr13.4 ..... 160 Figure 6.3. Cu 2p3,g/Al 2p XPS intensity ratios of CuCry catalysts plotted versus Cr/Al atomic ratio. Monolayer dispersion value = 2.28 ..................... 165 Figure 6.4. Cr 2p/Al 2p XPS intensity ratios of Cry (0) and CuCry(O) catalysts plotted versus Cr/Al atomic ratio. Cr 2p/Al 2p intensity ratios calculated for monolayer dispersion (line) ..................................... 167 Figure 7.1. XRD patterns measured for (a) alumina support, (b) CeA, (c) CeN, (d) CuA, and (e) C uN catalysts ................................................ 185 Figure 7.2. XRD patterns measured for (a) CuACeA (b) CuNCeA, (c) CuACeN, and (d) CuNCeN catalysts .......................................... 189 xiii Chapter 1 Air Pollution and Pollution Control Catalysts 1.1. Air Pollution 1.1.1. Introduction In recent years, there has been great concern about environmental problems which have resulted from the development of technologies. A tremendous amount of governmental and scientific effort has been exerted to improve the environment all over the world. While we still want to continue our economic growth which creates more products, more jobs, and a more comfortable and pleasurable lifestyle, at the same time, we also want to protect ourselves from an environmental crisis. One of the answers to this apparent paradox (economic growth and a clean environment) can be addressed by developments in catalytic science and technology [I]. The definition of air pollution was given by Seinfeld [2] as follows, ‘any atmospheric condition in which substances are present at concentrations above their normal ambient levels to produce a measurable effect (i.e., undesirable effect) on human beings, animals, plants, or materials’. By this definition, the substances can be any natural or artificial airborne chemicals which may exist in the atmosphere in gas, liquid, or solid 2 forms. Since it is impossible and unreasonable to eliminate all man-made emissions, it is more desirable to focus on the reduction of pollutants to the level of no harmful effects. The air pollution problems cannot be solved by the efforts of a few scientists. The study of emission sources, behavior of emissions in the atmosphere, and the effect of emissions on animals and plants requires a knowledge of the chemistry of combustion, engineering aspects of the equipment design, meteorology, mechanics, atmospheric chemistry, aerosol physics, physiology, medicine, plant pathology and so on. Since air pollution is also a global problem which requires international cooperation, not only science and technology fields but also social, economic, and political factors may play a major role in the abatement of air pollution problems. 1.1.2. Origin and Effect of Air Pollutants The origin of air pollution is an emission source. Major emission sources are transportation, electric power generation, refuse burning, industrial and domestic fuel burning, and industrial process. These emission sources produce air pollutants such as carbon monoxide, carbon dioxide, sulfur compounds, nitrogen compounds, alcohol, aldehyde, hydrocarbons, halogen compounds, particulate matter and radioactive compounds. These air pollutants cause many undesirable effects such as respiratory illness, visibility diminution, damage to animals and plants, and possibly catastrophic effects on the global scale [2-4]. 3 Carbon Monoxide and Dioxide(C0 and CO2): With the exception of C02, C0 is the most abundant air pollutant in the lower atmosphere especially in urban areas. The major source of C0 is the incomplete combustion of fossil fuels from automobile exhaust. Carbon monoxide is quite stable in the air and the rate of conversion to C02 is very low. Carbon monoxide is a poisonous inhalant because it has a strong affinity (250 times greater than oxygen) for hemoglobin and replaces oxygen. C02 is not normally regarded as an air pollutant according to the definition of air pollution and is an inevitable final product of the combustion process. However, although C02 effects may have been small to date, it may cause a significant effect in the fiIture when its concentration increases in the atmosphere. C02 absorbs infrared terrestrial radiation and can increase surface and atmospheric temperatures (green-house effect). Sulfur Oxides (SO/J5 The sources of atmospheric sulfur oxides (802, 803) are combustion of fossil fiJels (typically coal), decomposition and combustion of organic matter from electric power generation and industrial sources. SOx also appears in auto exhaust gases depending on the sulfur content of the fuel. S02 is highly soluble and absorbed in the upper respiratory system causing constriction of the airways. It also contributes to acid rain. . Nitrogen oxides (NOJ: The primary sources of nitrogen oxides (predominantly N0 and less than 5% N02) are high temperature combustion processes (particularly above 1500 °C) and the oxidation of fiJel containing nitrogen compounds. The combustion process fixes atmospheric nitrogen to produce first nitrogen monoxide (N0), which will be converted into nitrogen dioxide (N02). Under the action of solar radiation N02 4 dissociates into N0 and atomic oxygen which gives rise to the formation of ozone 03. The ozone produces a narrowing of the airways in the lung and accelerates an aging of lung tissue by oxidation. Organic molecules react with ozone to form free radicals which in turn act as a catalyst for the oxidation of N0 and hydrocarbons. Nitrogen oxides are known to cause bronchitis, pneumonia, susceptibility to viral infection and damage to the immune system. They also contributes to acid rain, urban smog, and ozone. Hydrocarbons: The combustion of gasoline is a major source of hydrocarbon emissions as a result of either simple evaporation before combustion or incomplete combustion. Benzene, its analogs, and aromatic polycyclic hydrocarbons are responsible for mutagenic or carcinogenic action. Hydrocarbons along with nitrogen oxides are a very important contributors to photochemical smog. Particles: Fine particles (mostly fly ash), which consist of inorganic and organic compounds of high molecular weight, are emitted from combustion systems (especially coal combustion). The particles, due to their small diameter, penetrate deep into the lungs and are retained for indefinite periods of time. Motor vehicles emit a number of metals such as Pb, Cr, Mn, Ba, V, Fe, A1, Cd, and Ni from engine exhaust, panel body alterations, tires, brakes systems, etc. Many of these metals are toxic, in particular Cd, Ni, Cr are carcinogenic, while Mn affects to the nervous system. 1.1.3. Pollution Control Techniques Control techniques for limiting pollutant formation and emission at the source can be classified into three groups: modification of the basic process, substitution of cleaner- burning fuels, and cleaning of the effluent gases before release to the atmosphere. The particular technique is selected depending on pollutants involved, the process responsible for pollutant formation, and the required degree of control [2]. Cleaning of effluent gases has received the most attention and is the most widely used among the three control techniques. The most frequently used technique to clean effluent gases from emission sources is the application of add-on devices. These devices are usually divided into two types: recovery and incineration devices. The suitable control device is selected on the basis of stream-specific characteristics and desired control efficiency. 1.1.3. 1. Recovery techniques One of the recovery techniques is the gas absorption method which uses a liquid solution capable of dissolving the pollutant gas molecules. Absorption is a diffusion process that involves the transfer of molecules from the gas state into the liquid state because of the contaminant concentration gradient between two phases. For best operation, the contaminants should be readily soluble in the liquid phase and undergo an irreversible reaction with a scrubbing reagent. An alternative recovery technique is to adsorb pollutants on a solid surface which can retain gas molecules by physisorption or chemisorption. Adsorption is a semi-continuous, diffusion-limited, surface area dependent process that involves percolation of a gas through a solid bed. Adsorption depends on selectivity and high capture efficiency which usually requires a significant mass-transfer gradient from the gas phase to the surface [2,5]. 1.1.3.2. Combustion Techniques The combustion method involves chemical alteration of the pollutant to non- pollutant molecules through thermal or catalytic incineration of the waste gases. Thermal incineration is currently the most widely used method to control emissions from industrial volatile organic compound (VOC) sources because it can provide high (> 99%) VOC destruction efficiencies when operated above their auto-ignition temperatures (usually 500 - 1000 °C) [6.7]. In addition, thermal incinerators are relatively simple and economical to install and operate. Thermal incinerators are applied principally when low concentrations of combustible materials are present in a waste stream or when a reliable supply of waste gas is available to use as a fuel for preheat burners. Thermal incinerators are generally used when organic aerosols, significant amounts of non-combustible, high boiling point compounds, inorganic materials, dusts, catalyst poisons or inhibitors are present in the waste gas stream. However, thermal incineration requires a high fuel cost since the waste gases need to be heated up to high operating temperatures for effective operation. The high operating temperature of thermal incinerators may produce undesirable secondary emissions such as N0x (via fixation of atmospheric nitrogen) and dioxins (via destruction of chlorinated hydrocarbons) and create difficulties in mechanical design [5-7]. 7 1.2. Pollution Control Catalysts 1.2.1. Introduction 1.2.1. 1. Principle of Catalytic Incineration The objective of introducing a catalyst in a combustor is to carry out heterogeneous oxidation on its surface. By choice of a suitable catalyst, one can ensure that an activation energy for a heterogeneous reaction is much lower than that for a purely homogeneous reaction. For instance, the activation energies for the homogeneous oxidation of propane are approximately 25-50 kcal/mol, while those for heterogeneous oxidation reactions are approximately 10—20 kcal/mol. Consequently, appreciable heterogeneous oxidation rates can be achieved for temperatures and fuel concentrations much lower than those required for the homogeneous reactions to proceed [8]. Therefore, catalytic oxidation is an efficient and effective method for burning low heating value gaseous fuels and for the destruction of trace organic contaminants in air [9]. To obtain optimal performance from a combustor, the catalyst system should possess several properties. First, the catalyst should be able to ignite a fuel/air mixture at the lowest possible temperature (low light off temperature). Second, the catalyst activity Should be sufficiently high to maintain complete combustion at the lowest temperature and the highest mass throughput. And finally, the support should have a large surface area, low pressure drop, and high thermal and shock resistance [8]. 1.2.1.2. Advantages of Catalytic Pollution Control Primary advantages of using catalysts for emission control are that pollutants can be completely oxidized at relatively low temperatures and consequently the formation of N0,, is greatly suppressed compared to thermal incineration [10]. For the thermal process, high temperature is normally maintained by burning extra find which represents up to 40 % of total Operating cost. Therefore, operating cost is lower than for thermal incineration. The decreased temperature also allows the use of corrosion resistance alloys for reactors rather than replaceable ceramic linings which corrode rapidly in the thermal incinerator [11]. The introduction of a heterogeneous catalyst also allows better control of oxidation over wider air/fuel ratios. If a fuel concentration is too low to maintain catalyst temperature, a catalyst bed can be heated effectively to the temperature at which oxidation is initiated. In addition, the use of a catalyst improves energy recovery by minimizing energy loss as visible light (i.e. non-flammable combustion) [12]. These advantages have led to widespread application of catalytic incineration for the control of waste gas emitted from industrial processes and automobile converters. 1.2.1.3. Disadvantages of Catalytic Pollution Control The primary disadvantage of using catalysts for emission control is that all catalysts are eventually deactivated physically and chemically. Physical deactivation results from mechanical attrition and sintering which lead to loss in active components and surface area, respectively. Chemical deactivation results from presence of impurities 9 which can lead to poisoning and accumulation of deposits which can lead to fouling. Thus, many catalysts require frequent and costly replacement of a catalyst bed and interruption of process operation [5,13]. Further disadvantages of catalytic combustors are that the volumetric heat release rates and mass transfer coefficients are very low compared to those of the thermal combuster [14]. In addition, destruction efficiency for a given compound may vary depending on the composition of a mixture in an emission stream [15]. 1.2.2. Noble Metal Catalysts 1.2.2. 1. Advantages of Noble Metal Catalysts Most commercial catalysts for total oxidation reactions utilize noble metals such as Pt and Pd. Although, in principle, any noble metal (such as Ag, Au, Ru, Os, Ir) may be used as a total oxidation catalyst, their application is limited due to sintering, volatility loss, and irreversible oxidation resulting from excess oxygen present at high temperature. Limited supply and high cost Of these metals also minimize their use [8]. Compared to base metal oxides, noble metals have many advantages such as high intrinsic activity for oxidation reactions, high thermal stability, and sulfur poisoning resistance [16]. In addition, Pt and Pd are readily prepared in a highly dispersed form on a number of support materials. For these reasons, only small amounts (0.1-0.5 wt.%) are typically necessary for the production of good combustion catalysts. The high intrinsic activity of these metals is related to their ability to activate H2, 02, C-H, and O-H bonds [8]. On noble metals, oxygen adsorption is relatively weak. This implies that the mechanism may 10 proceed through direct electron transfer between reactant and oxygen on metal surface. The general mechanism of oxidation over noble metals is suggested to involve the dissociative adsorption of oxygen [17]. 02 +1] -+ [02] —> ZIOI (H) where [] represents a surface active site. 1.2.2.2. Disadvantages of Noble Metal Catalysts While noble metal catalysts have many advantages, there are several problems in the operation of the catalysts. This is especially true for chlorinated hydrocarbon oxidation because such systems are susceptible to C12 and HCl poisoning, byproducts of the reaction [58,59]. Noble metals are deactivated by sintering at temperatures of between 500 ~ 900 °C [8]. Sintering reduces the active surface area and catalyst activity. It has been found that in oxidizing atmosphere, noble metals can disperse as oxides on A1203 surface at temperatures below the decomposition temperatures of the oxides (Pt02, 585 °C; PdO, 790 °C) [18]. Pt02 interacts weakly with y-AI2O3 surface so that the Pt dispersion decreases and large metallic crystallites form at temperatures greater than 600 °C under oxidizing conditions [16]. The prolonged exposure of Pd to oxygen has been observed to cause structural changes in Pd metal that result in loss of catalytic activity for methane oxidation at temperatures above 450 °C [19]. In addition, catalytic combustors using noble metal catalysts convert very high levels of fuel bound nitrogen to N0x (50~90 11 %) under lean condition [8]. The high price and limited supply are also major disadvantages for using noble metals. 1.2.3. Transition Metal Oxide Catalysts 1.2.3.1. Advantages of Transition Metal Oxide Catalysts Transition metal oxide catalysts have been considered as suitable substitutes for noble metal catalysts because of their natural abundance and low cost. Current issues are the development of contaminant resistant and thermally stable base metal oxide catalysts for automobile catalytic converters and industrial catalytic incinerators [20]. Transition metal oxide catalysts have greater resistance to certain poisons such as halogens, As, Pb, and P [21] and yield a lower level of fiIel bound nitrogen conversion to N0x (10~20 %), even in lean fuel Operation, than noble metal catalysts [8]. Thus, the use of oxide based catalysts such as Cr203-C03O4 [22] and NiO [23] appears to be very promising for N01, control from both thermal and fuel bound nitrogen. Metal oxide catalysts typically have lower catalytic activity and higher light-off temperatures than noble metal catalysts. However, for typical catalytic combustion, catalysts are operated under a diffusion control region where their activities are comparable to those of the noble metals and under such operation, the catalytic activity (kinetic control) plays a secondary role in determining the reaction rate [8]. 12 1.2. 3. 2. Principles of Transition Metal Oxide Catalysts Blazowski et al. [24] indicated that C0304, Cr203, CuO, NiO, Mn02 and V205 are potential catalysts for hydrocarbon oxidation. MnO2 and CuO are promising candidates for olefin oxidation [25,26] and C0304 possesses the highest activity of base metal oxides for paraffin or large molecular weight olefin oxidation reactions [27]. On the basis of the activity of catalyst per unit surface area, C0304 and CuO catalysts showed high CO oxidation activity [92]. The activity of the transition metal oxide catalysts for oxidation reactions may depend on redox behavior of metal ions [28], metal-oxygen bond strength [29], and types of oxygen involved in the reaction [30]. In addition, there should be an optimum level of interaction between active sites and reactants. If the interaction is too strong, the catalyst will be deactivated by irreversible chemisorption. If the interaction is too weak, only a small fraction of the surface is covered and the catalytic activity is very low [31]. Low temperature ESR studies have shown that the interaction of oxygen gas with an oxide surface proceeds the following steps [32]. 02 (gas) + e' —> 02' (adsorbed) (1.2) 02' + e' —> 20' (adsorbed) (1.3) 0' + e' —> 02' (lattice) (1.4) The mobility of oxygen species is also important for oxidation catalysts because highly mobile oxygen should result in a highly active and non-selective catalyst [9]. Both 13 adsorbed and lattice oxygen participate in a catalytic oxidation process. Haber [33] postulated that surface adsorbed oxygen may generally lead to complete oxidation, whereas lattice oxygen is used for partially oxidized products. This may be because surface adsorbed oxygen is more mobile than lattice oxygen. Though, in principle, either 02' or 0' may promote total oxidation, 0' may be more reactive than 02' due to its higher mobility. In addition, 0' is known to oxidize CO and H2 even at -196 °C [30]. Lattice oxygen becomes important when the oxidation processes are at high temperatures (> 500 °C). Mixtures of these oxides often exhibit greater stability and combustion activity than the single oxides [34-37]. Considerable improvement in thermal stability can be achieved by utilization of complex oxides such as rare earth perovskites [38] and mixed transition metal oxides [39]. Indeed, a binary transition metal oxide catalyst comprised of Cr203-CO304 supported on alumina has been successfully used for the catalytic combustion of lean propane-air mixtures [40]. Catalysts comprised of La203-Cr203 also appear to be very promising for oxidation at high temperature [41]. CuO-Mn02 (hopcalite) [42] and CuO-Cr203 [43] catalysts are well known as some of the most active oxidation catalysts among base metal oxide catalysts. These catalysts were also studied for the removal of air pollutants such as CO and N0 from exhaust gas [44,45]. The synergistic effect of mixed oxide catalysts compared to their individual oxide components may be due to the readily available multiple energy levels of the metals and their associated oxygen anions, which makes the organic reactant more accessible to the active oxygen 14 anions. This also may result in higher surface mobility of oxygen and/or the activated complex as well as electron transport through the lattice,[31]. 1.2.4. Industrial Waste Treatment Catalysts 1.2.4.1. Volatile Organic Compounds (VOC) Emission The control of organic emissions from industrial processes was first accomplished by catalytic incineration in the late 1940’s [46]. However, after thermal incineration was developed during 1960’s [13], the first catalytic system was installed for commercial use in 1975. Catalytic oxidation is widely used for trace contaminant removal in Europe due to the high cost of energy and relatively strict environmental regulations. The Clean Air Act 1990 in the US raised the number of toxic chemicals to be controlled by the year 2000 from 7 to 189 (70 % of these are VOCS) [47]. Both catalytic and thermal combustion are used for controlling VOCs. Some areas where catalysts are used beyond thermal combustion are in can, paper, coating, organic chemical and plywood manufacture, tire production, asphalt blowing, odor control, waste water plants, automotive exhaust, and to remove air contaminants in submarines [48]. For these processes, a highly active and non- selective catalyst is required. Pt shows high oxidation activity for saturated hydrocarbons and high molecular weight species [49]. Pd is preferred for CO, olefins, methane and low molecular weight olefin oxidation [50]. Base metal oxide catalysts are used less frequently for waste gas which is relatively free of contaminants or for fluidized bed processes in which the catalytic surface is refreshed by abrasion [51]. The preferred support for organic abatement catalysts is either ceramic (cordierite) or metallic (aluminum, stainless 15 steel) monolith or honeycomb. These support materials can minimize pressure drOp with parallel channels (200-400 channels per square inch). Most commonly 0.1-0.5 wt.% Pt on high surface area y-Al203 is used for a bead or washcoat on a monolith. In limited cases, a small amount of Pd or Rh is added to promote a particular reaction where the Pt may be deficient. If the feed contains a large amount of sulfur ( > 50 ppm S), less reactive carriers such as Ti02 or Ot-Al203 are used for Pt [10] to avoid formation of sulfates. An air stripper is used to remove VOCS from contaminated ground water. The emission from air strippers contains a wide range of compounds and concentrations of VOCS including chlorinated organics. Chlorinated organics cause byproduct formation in thermal oxidation systems. For instance, thermal destruction of polychlorinated biphenyls (PCBS) requires temperatures of at least 1200 °C and residence times of over 1 second. Standard municipal incinerators merely vaporize and release PCBs into the atmosphere [52] or produce even more toxic polychlorinated dibenzofiirans (PCDFs) and polychlorinated dibenzo-p-dioxins (PCDDS) [53-55]. Hence, there is a compelling need for development of economically viable, lower temperature processes capable of effectively treating incinerator flue gas or other effluents that may be contaminated with PCB vapors [56]. Catalytic combustion systems may be the answer for air stripper emission control due to its low temperature Operation, relatively humidity insensitive, and its compatibility with low gas concentrations [57]. Transition metal catalysts are suitable candidates for VOC control because noble metals (Pt, Pd) are susceptible to C12 and HCl poisoning from the destruction of chlorinated hydrocarbons [58-62]. Cr203/Al203 is considered to be a 16 suitable catalyst for chlorinated hydrocarbon oxidation because it produces low levels of C12 which is a more toxic compound than HCl. Weldon and Senkan [63] have determined the C12 selectivity for CH3CI oxidation to be in the 3-7 % range over Cr203/A1203, whereas a 19-42 % range of C12 selectivity was reported for copper based catalysts [64]. 1.2.4.2. N0, Emission N0,IL consists primarily of N0 and N02, which are produced in all combustion processes by the oxidation of N2 and fuel bound nitrogen. At temperatures higher than 1500 °C, the reaction proceeds at appreciable rates through the Zeldovich mechanism [65]. N2+O—>NO+N (1.5) N+02—)N0+0 (1.6) N+OH—>NO+H (1.7) The rate of the above reactions is controlled by reaction (1.5), which has a high activation energy and, as a result, thermal NO. formation is highly temperature dependent. If a compound in the combustion process has bound nitrogen (e.g., pyridine), NOx is readily formed at much lower temperature through an oxidation process [66]. Recently, the emissions of N20 from fiber production plants has received attention due to its potential to contribution to global warming (strong infrared absorptivity). 17 The major contribution to N0x generation is stationary sources. NOx emission from power plants and lean burn engines cannot be controlled by three-way catalysts due to the uncontrolled stoichiometric air/fuel ratio. The most effective NOx control strategies are catalytic decomposition of nitrogen oxides and selective catalytic reduction [67]. Catalytic NO, decomposition: The decomposition of N03‘ to nitrogen and oxygen is one of the most challenging approaches to remove N0x fiom exhaust gases. Thermodynamically, NO should be decomposed to its constituents elements at low temperature. N0 (g) —> 1/2N2 + 1/202 AG" = -207 kcal/mol (1.8) Accordingly, a catalytic N03 decomposition method could be the simplest and cheapest technique Since no added reductant is necessary. Silver promoted C0304 [68], perovskite type oxides [69,70], Pt/Al203 [71], Rh/Al203 [72], Cu-ZSM-S [73] have been reported to be active catalysts for N0,X decomposition. Cu-ZSM-S catalyst is considered to be the most active system [67]. However, at present these catalysts are not active enough under commercial conditions [74]. Furthermore, the Cu-ZSM-S catalyst is poisoned by 502 [75] and by oxygen in the feed gas or produced from N0 decomposition [76]. Catalytic reduction of N0,r by ammonia: Selective catalytic reduction (SCR) of NOx using NH3 was first discovered with Pt catalysts in excess amounts of oxygen [77]. The technology of selective catalytic reduction with ammonia was well developed in Japan and 18 Western EurOpe to control of NOx emissions from electricity generating power plants. The Clean Air Act 1990 will prompt widespread use of SCR technologies in USA. The desired reactions are: 4NH3 + 4N0 'l' 02 -> 4N2 + 61120 (1.9) 4N1‘13 ‘1' 2N02 '1‘ 02 —) 3N2 + 61120 (1.10) The undesired reactions are: 4N1'13 ‘1' 302 ---> 2N3 + 61120 (1.11) 4N1'13 + 502 —-)4NO + 61120 (1.12) Pt catalysts were not applicable because they have poor selectivity for N03‘ reduction at temperatures above 250 °C. Above 250 °C, a V205/Al203 catalyst was used first, but was restricted to sulfur free exhaust gases due to its high sensitivity to poisoning. 802 is converted to $03 which reacts to form sulfates (Al2(SO4)3) which plug catalyst pores and cause deactivation. Further, as the catalyst deteriorates, the activity at a fixed NH3/N03 ratio decreases and slippage of NH3 increases, with resultant discharge of NH3 into the atmosphere [67]. Therefore, an alternative method of NOx control has been developed using vanadia/titania catalysts. These catalysts facilitate reaction between NOx and NH3 or hydrocarbons in the presence of oxygen. Si02, W03, M03 can be used as promoters. A key to the success of the foi catalyst is that it is not poisoned by S02 19 because the T102 carrier is non-sulfating under operating conditions [78-80]. V205 based catalysts operate best in the temperature range between 260-450 °C. The maximum exposure temperature for this catalyst must not exceed 450 °C, because the active anatase phase of Ti02, with a surface area of 80-120 mZ/g, irreversibly converts to rutile, with a surface area of less than 10 mZ/g. Recently, Cu/Zeolite catalysts have been developed that function at higher temperatures over a broader range of temperatures than Pt and V205 based catalysts [81]. These catalysts have much lower tendency to oxidize ammonia to NOx and the maximum operating temperature is about 600 °C. Above 600 °C, the zeolite is deactivated by a de-alumination process whereby the Al ions migrate out of the Si02- A1203 structure [10]. The SCR technique using ammonia has received much attention because ammonia is a very selective reducing agent to NOx under oxidative conditions. The other reactants such as CO, H2, CH4 and hydrocarbons react readily with oxygen in the gas stream. There are several disadvantages of using ammonia as a reductant. NH3 is more expensive than hydrocarbons especially methane, it requires special handling and storage, and a sophisticated metering system is needed to avoid NH3 leakage [10]. Catalytic reduction of NO; by hydrocarbons: Reduction of NOx by natural gas or other hydrocarbon has recently attracted much attention as a possible new de-nitrification process for exhaust gases from power plants, diesel and lean-bum gasoline engines. The motivation to develop N0x reduction in lean environments using hydrocarbons is to improve fiIel economy. Operating an internal combustion engine under lean conditions 20 improves the combustion efficiency, power output, and decreases the emission of the greenhouse gas CO2. Additionally, in many new power plants, natural gas is commonly used as a fuel and is readily available [10,67,82]. CH4 + 2N0 + 02 —) N2 + 2H2O '1' C02 (113) Catalysts comprised of copper ion exchanged zeolite (Cu-ZSM-S) were the first discovered to be active and selective for N02, reduction with propane at 500 °C under oxidative conditions. Cu-ZSM-S catalyzed N02 reduction by C2H4, C3H6, C4113, C3Hg in the presence of 02 [83,84]. H-ZSM-S was a particularly good catalyst for the reduction of nitric oxide by C3113 in the presence of 02 [85]. Its activity is actually promoted by the presence of 02, but it is ineffective at lower temperatures and inhibited by H20. In contrast, the use of ethylene or propene reduces NOx at lower temperatures (160 -200 °C). An Fe ion exchanged mordenite catalyst was recently reported to be most active for N02 reduction by C2H4 in the presence of 02 at 200 °C [86]. Since CH4 is a major hydrocarbon emission from a gas-cogeneration system, such as a electric power plant, it would be advantageous to develop a catalyst which is active for reduction of NOx by CH4 in lean conditions [87,88]. Li and Armor [89] have found that a Co-ZSM-S catalyst showed potential for using natural gas as a reductant. However, the reaction rate was slow and was inhibited strongly by H2O. Noble metal catalysts are also regarded as promising. Alumina supported Pt catalysts show higher activity and durability than Cu-ZSM-S 21 catalysts under real diesel exhaust conditions [90]. However, at present, no commercial process has been developed for catalytic reduction ofNOx under oxidative conditions. 1.2.5. Automobile Emission Control Catalysts The development of catalytic reactors for cleaning exhaust gases of automobiles has been stimulated successfully by legislation motivated by environmental and political issues. In 1959 and 1960, laws on motor vehicle emission standards were first enacted and these standards could be met with the modification of engine parameters. The Clean Air Act of 1970 announced further tightening of the standards that went well beyond existing technology and catalytic research increased considerably. The catalytic converter has been considered as the only technology available for meeting the stringent automobile exhaust standards [91]. Automobile catalytic converters have been used in the US since 1974 (1975 model year vehicles) in order to meet emission standards. The initial objective was to reduce the emission of CO and hydrocarbons. The engine was operated in the fiiel rich side and the concentrations of C0 and hydrocarbons are relatively high, whereas the NO" concentration is low. The nitrogen oxide standard could be met by using exhaust gas recirculation which leads to the formation of lower levels of nitrogen oxides by reducing the combustion flame temperature in the engine [91]. The oxidation catalysts used until 1979 were a combination of Pt and Pd and operated in the temperature range of 250 - 600 °C with space velocities from 10,000 - 100,000 l/hr depending on the engine size and driving mode. Typical catalyst compositions were Pt and Pd in a 2.5:] or 5:1 ratio, 22 ranging from 1.6-3.1 g/car [10]. Many base metal oxide candidates were also investigated in this period, such as Cu, Cr, Ni, Mn [92,93]. Since 1979, the NOx standards tightened and exhaust gas recirculation alone was no longer sufficient to control nitrogen oxides emission. In the beginning, the dual bed system with separate reduction and oxidation catalysts was considered due to difficulties in controlling stoichiometric air/fiJel ratio. The engine was operated under fiIel rich conditions for the first bed so that a catalyst could reduce N03 with H2, C0 and hydrocarbons. Remaining exhaust gases would be oxidized in the second bed with air injection [94]. The activity of alumina supported noble metal catalysts for NO reduction by C0-H2 mixture was reported in the order Ru>Rh>Pd>Pt [95]. However, Ru, Pd, and Pt were not proper catalysts for N03, reduction beds. Ru was found to be volatile by forming Ru03 or Ru04 when the engine exhaust was oxidizing and the temperature exceeded about 700 °C [96]. In addition, Ru04 is a highly toxic compound. Pt and Pd catalysts are less useful as NO reduction catalysts due to large production of NH3 which would be reoxidized to N0, in second oxidation bed. Rh was suitable for the N01, reduction bed because it has a high activity for N0, reduction [95], is less inhibited by C0 and sulfur [97], and forms less NH; than Pt or Pd. It also has an extended N03 conversion window [98]. The next generation of catalysts named three-way catalyst (TWC) could catalyze all reactions simultaneously: hydrocarbons and carbon monoxide are oxidized to C02 and H20, while nitric oxides are reduced to N2. The noble metal Rh combined with Pt is able to do both sets of reactions if the engine exhaust is operated close to stoichiometric air/fuel ratio. A typical example of the influence of air/fuel ratio to the conversions is 23 given in Figure 1.1. On the rich side (airfuel), the CO and hydrocarbon conversions are high, but at the sacrifice of the NOx reduction. A special control system to maintain the stoichiometric air/fire] ratio at all times required an advance in this technology. This was made possible by the development Of the 02 sensor, which was positioned before the catalyst bed in the exhaust manifold [99]. The catalysts in use contain Pt, Pd, and Rh as major constituents supported on La203 (and/or BaO) stabilized (1-2 wt.%) alumina. The alumina is usually introduced as a washcoat (about 15 wt.% and 30-50 um thick) to a ceramic honeycomb monolith made of cordierite (2Mg0-2Al203'58i02) [100] containing 200-400 square channels per square inch. A converter typically contains 1-3 g Pt and 01-05 g Rh and 0-3 g Pd. Pt is an effective oxidation catalyst for CO and the complete oxidation of hydrocarbons. Pd is more active than Pt to promote the oxidation of CO and hydrocarbons but is more sensitive to poisoning and sintering than Pt in the exhaust environment. Both Pt and Pd promote the reduction of nitric oxide, but they are less effective than Rh. Rh also contributes to C0 oxidation, but is mostly required for N0, reduction. In addition to the noble metals, three way catalysts contain Ce02 and possibly other additives such as La, Ni, Fe. These base metal additives are believed to improve catalyst performance by extending conversion Figure 1.1. 24 Removal Efficiency (x) 100 40 Ctoionoinotric l Alr-Fuai Ratio I I I I I 14.3 14.4 14.5 14.8 14.7 14.8 14.9 Air-Fuel Ratio (wot/wot.) Simultaneous conversion of hydrocarbon (HC), CO, and NOx for TWC as a fiinction of air/fiael ratio. (From Heck, R. M.; Farrauto, R. J. Catalytic Air Pollution Control: Commercial Technology; Van Nostrand Reinhold: New York, 1995). 25 during the rapid air/fuel ratio perturbations and help to stabilize the alumina support against thermal degradation [91]. Ceria may be added at a loading of 10-30 wt.% to the washcoat of three-way catalysts to store oxygen, promote the water-gas shift reaction, stabilize the noble metals against thermal damage, and to alter CO oxidation kinetics [101,102]. Ce02 enhances NOx conversion to N2 by its ability of oxygen storage derived from N0x decomposition during lean condition. Stored oxygen is then available for reaction with C0 and hydrocarbons during subsequent rich condition [103]. This oxygen storage behavior broadens the operating air/fuel ratio window of the catalyst [104]. Ceria has been shown to enhance the decomposition of NO by extending the time before the noble metal catalyst is deactivated by the accumulation of surface oxygen derived from N0 decomposition. For instance, Rh/CeO2 is deactivated more slowly than is Rh/Al203 during N0 decomposition, probably due to oxygen spillover from the noble metals to the reduced ceria [105]. In addition, ceria addition to an alumina supported Rh catalysts was shown to enhance NO reduction activity at low temperatures by decreasing the apparent activation energy for the reaction of CO with N0 and by shifting to positive order the dependence of the rate on NO partial pressure [106]. The reactions are indicated below: Rich condition: 2Ce02 + CO -—> Ce203 + C02 (1.14) 2C602 + CxHy -) C8203 + mC02 + [11120 (1.15) 26 Lean condition: Ce203 + %02 (or NO)-—> 2Ce02 (1.16) Another benefit of Ce02 is that it is a good steam-reforming (or water gas Shift) catalyst and catalyzes the reaction of C0 and hydrocarbons with H2O to produce H2 in the rich condition. Then the H2 reduces a portion of the NO, to N2: C0 + H2O —) 1'12 '1' C02 (1.17) CxHy + Z1120 -) 11112 + mC02 (1.18) 2N0 + 2112 -) N2 + 21120 (1.19) Ce is the component mainly responsible for the release of H25 gas by reducing Ce(SOI)2 which is formed with 803 and C602 in the lean mode. Nickel has been found to be an effective scavenger of H25 [107]. Lanthanum is added to the alumina washcoat for thermal stabilization [108,109]. Recently, the replacement of Pt and/or Rh with Pd has been studied for TWC because Pd is relatively less expensive than Pt and Rh [1 10]. First, the Pd/Rh catalyst was developed for an automotive application [111,112] and then the Pt/Pd catalyst was developed to further reduce TWC cost [1 1 1]. Although three way catalysts have been developed which successfully control automobile emissions, they still need to be improved. A major challenge in this field is to develop a catalytic technology which removes N0,i under lean-bum conditions with the 27 objective of increasing fuel efficiency and environmental protection with a lower level of CO2 production [67]. 10. ll. 12. 13. 14. 15. 28 1.3. References Cusumano, J. A. In Science and Technology in Catalysis 1994 ; Izumi, Y., Arai, H.; Iwamoto, M., Ed.; Elsevier: Amsterdam, 1995; p 3. Seinfeld, J. H. Air Pollution Physical and Chemical Fundamentals ; McGraw-Hill: New York, 1975. Perkins, H. C. Air Pollution; McGraw-Hill: New York, 1974. Chiron, M. In Catalysis and Automotive Pollution Control; Crucq, A. Frennet, A. Ed.; Elsevier: Amsterdam, 1987; p1. Bethea, R. M. Air Pollution Control Technology: An Engineering Analysis point of View; Litton Educational Publishing: New York, 1978. “Control of Volatile Organic Emissions from Existing Stationary Sources, Volume 1: Control Methods for Surface Coating Operations”; (Prepared for US. Environmental Protection Agency.) Research Triangle Park, NC. EPA Publication No. 450/2-76-028. November 1976. Rolke, R.W. “Afterburner Systems Study” (Prepared for US. Environmental Protection Agency.) Research Triangle Park, NC. Publication No. EPAR272062. August 1972. Prasad, R.; Kennedy, L. A.; Ruckenstein, A. E. Catal. Rev. -Sci. Eng. 1984, 26, 1. Satterfield, C. N. Heterogeneous Catalysis in Practice ; McGraw-Hill: New York, 1980. Heck, R. M.; Farrauto, R. J. Catalytic Air Pollution Control: Commercial T echnology; Van Nostrand Reinhold: New York, 1995. Manning, M. P. Hazardous Waste 1984, l, 41. Trimm, D. L. Appl. Catal. 1983, 7, 249. Hardison, L. C .; Dowd, E. J. Chemical Engineering Progress 1977, 73, 31 Pfefferle, L. D.; Pfefferle, W. C. Catal. Rev. -Sci. Eng. 1987, 29, 219. US. EPA. "Parametric Evaluation of VOC/HAP Destruction Via Catalytic Incineration." EPA-600/2-85-041 (NTIS PBSS-l91187). April 1985. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25 26. 27. 28. 29. 30. 31. 32. 33. 29 Kummer, J. T. Prog. Energy Combust. Sci. 1979, 6, 177. Golodets, G. I. Heterogeneous Catalytic Reactions Involving Molecular Oxygen; Elsevier: Amsterdam, New York, 1983. Yao, H. C.; Sieg, M.; Plummer, H. K. Jr. J. Catal. 1979, 59, 365. Cullis, C. F.; Willatt, B. M. J. Catal. 1983, 83, 267. Yang, B. L., Chan, S. F., Chang, W. S., and Chen, Y. Z. J. Catal. 1991, 130, 52. Cimino, A.; De Angelis, B. A.; Luchetti, A.; Minelli, G. J. Catal. 1976, 45, 316. Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. Technol. 1981, 27, 45. B. A. Folsom et al., “Fuel nitrogen conversion-The impact of catalyst type”, Energy & Environmental Research Corporation, EPA-600/9-80-035 (1980). Blazowski, W. S. and Walsh, D. E. Combustion Science and Technology, 1975, 10, 233. Morovka, Y.; Ozaki, A. J. Catal. 1966, 5, 116. Morovka, Y.; Morikawa, Y.; Ozaki, A. J. Catal. 1967, 7, 23. Boreskov, G. K. Proc. 5th Int. Congress on Catalysis, Amsterdam, paper 71, 1972. ' Labinger, J. A; Ott, K. C. Catal. Lett. 1990, 4, 245. Bond, G. C. Heterogeneous Catalysis: Principles and Applications; Clarendon Press: Oxford, UK, 1974. Margolis, L. Y. Catal. Rev. 1973, 8, 241. Spivey, J. J. In Catalysis, vol 8 The Royal Society of Chemistry, Cambridge, 1989; p 157. Kon, M. Ya; Schvets, V, A,; Kazanskii, V. B. Kinet. Kat. 1972, 13, 635. Haber, J. Int]. Chem. Eng. 1975, 15, 21. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48 49. 50. 3O Dwyer, F. G., Catal. Rev. -Sci. Eng 1972, 6, 261. Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Catal. Rev-Sci. Eng. 1984, 26, 1. Weldon, J. and Senkan, S. M. Combust. Sci. and Tech. 1986, 4 7, 229. Dadyburjor, D. B., Jewur, S. S., and Ruckenstein, E. Catal. Rev.- Sci. Eng. 1979,19, 293. Coutures, J.-P.; et al. High Temp. Sci. 1980, 13, 331. Prasad, R.; et al. Comb. Sci. Techn. 1980, 26( 1&2), 51. Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. Technol. 1980, 22, 271. Tong, H.; Snow, G. O; Chu, E. K.; Chang, R. C. S.; Angwin, M. J.; Pessagno, S. L. NASA CR-l65396, Sep. 1981. Yoon, C.; Cocke, D. L. J. Catal. 1988, 113, 267. Severino, F.; Brito, J.; Carias, 0.; Laine, J. J. Catal. 1986, 102, 172. Dennott, J. M. C. in “Pollution Control Review”, vol 2, Noyes Data Corp., Park Ridge, IL, 1971. Stegenga, S.; van Soest, R.; Kapteijn, F.; Moulijn, J. A. Appl. Catal. B 1993, 2, 257. Hardison, L. C., “A Summary of the Use of Catalysts for Stationary Emissions Sources Control”, Fisrt National Symposium on the Use of Heterogeneous Catalysts for Control of Air Pollution, Philadelphia, Pa. Nov. 21-22, 1968. Chemical Engineering, 1990, 24. Armor, J. N. Appl. Catal. B, 1992, l, 221. Yao, Y.-F. l & EC Prod. Res. Dev. 1980, 19, 293. Stein, K.; Feenan, J .; Hofer, L.; Anderson, R. Bureau of Mins Bulletin, 1962, 608. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 31 Hardison, L.; Dowd, E. Chemical Engineering Progress, 1977, 75, 31. Farrell, J. B.; Salotto, B. V. Effect of Incineration on Metals, Pesticides and PCBs in Sewage Sludge; National Symposium on the Ultimate Disposal of Wastewaters, and Their Residuals; Raleigh: NC, 1973. Buser, H.R.; Bosshardt, H. P.; Rappe, C. Chemosphere 1978, 1, 109. Buser, H.R.; Bosshardt, H. P.; Rappe, C. Chemosphere 1978, 2, 165. Olie, K.; Verrnuelen, P. L.; Hutzinger, O. Chemosphere 1977, 8, 455. Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. Techno]. 1988, 22, 557. Agarwal, S. K.; Spivey, J. J. Environmental Progress 1993, 12, 182. Heyes, C. J.; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol. 1982, 32, 1034. Cullis, C. F.; Keene, D. E.; Trimm, D. L. J. Catal. 1970, 19, 378. Wang, Y.; Shaw, H.; Farrauto, R. ACS Symposium Series 45, 1992, 125. Yu, T.; Shaw, H.; Farrauto, R. ACS Symposium Series 45, 1992, 141. Simone, D.; Kennelly, T.; Brungard, N.; Farrauto, R. App]. Catal. 1991, 70, 87. Weldon, J.; Senkan, S. M. C ombust. Sci. and Tech. 1986, 47, 229. Laidig, G., Hoenicke, D.; Griesbaum, K. Erdol. Kohle. Erdgas Petrochem. 1981, 34, 329. Zeldovich, J. Acta Physiochim, 1946, 2], 577. Fenimore, C. Combustion and Flame 1972, 19, 289. Tamaru, K.; Mills, G. A. Catal. Today 1994, 22, 349. Hamada, H., Kuwahara, Y., Kindaichi, Y., Ito, T. Nat. Meeting Chem. Soc. Jpn., 1Va39, 1988. Shimada, H., Miyama, 8., Kuroda, H. Chem. Lett. 1988, 1797. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 32 Viswanathan, B. Catal. Rev. -Sci. Eng. 1992, 34, 337. Amimazmi, A.; Benson, J. E.; Boudart, M.; J. Catal. 1973, 30, 55. Hardee, J. R.; Hightower, J. W. J. Catal. 1984, 86, 137. Iwamoto, M. In Future Opportunities in Catalytic and Separation Technology; Misono, M.; Moro-oka, Y.; Kimura, 8., Ed; Elsevier: Amsterdam, 1990; p121. Catalytica, “Prospects for Catalytic Nitric Oxide Decomposition in Air Pollution Control: Multi-Client Study”, Catalytica, Mountain View, CA. 1993. Iwamoto, M.; Yahiro, H.; Shundo, S.; Yu-u, Y.; Mizuno, N. App]. C atal. 1991, 69, 5. Li, Y.; Hall, W. K. .1. Catal. 1991, 129, 202. Cohn, 6.; Steele, 0.; Andersen, H.; US. Patent 2,975,025 1961 Bosch, H.; Janssen, F. Catal. Today, 1988, 2, 369. Farrauto, R. J .; Heck, R. M.; Speronello, B. 1992, Chemical and Engineering News, 1992, 70, 34. Heck, R. M.; Chen, J. M.; Speronello, B. K.; Morris, L. in Environmeta] Catalysis; Armor, J. Ed.; ACS Symposium Series 552; American Chemical Society: Washington, DC, 1994; p 215. Heck, R. M.; Chen, J. M.; Speronello, B. K. Environmental Progress, 1994, 13, 221. Gopalakrishnam,R.; Stafford, P. R.; Davidson, J. E.; Hecker, W. C .; Bartholomew, C. H. Appl. Catal. B, 1993, 2, 165. Held, W.; Koenig, T. R.; Puppe, L. SAE Pasper 900496, 1990. Iwamoto, M.; Hamada, H. ('atal. Today, 1991, 10, 57. Hamada, H.; Kintaichi, Y.; Sasaki, M.; Ito, T.; Tabata, M. Appl. Catal. 1991, 64, L1. Sato, S.; Hirabayashi, H; Yahiro, H.; Mizuno, N.; Iwamoto, M. Catal. Lett. 1992,12, 193. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 33 Armor, J. N.; Li, Y. Prepr. Div. Pet. Chem. Am. Chem. Soc. 1994, 39, 104. Zhang, X.; Vannice, M. A.; Walters, A. B. Prepr. Div. Pet. Chem. Am. Chem. Soc. 1994, 39, 141. Li, Y.; Armor, J. Appl. Catal. B, 1992, 1, L31. Obuchi, A.; Ohi, A.; Nakamura, M.; Ogata, A.; Mizuno, K.; Ohuchi, H. Appl. Catal. B, 1993, 2, 71. Taylor, K. C. In Catalysis andAutomotive Pollution Control; Crucq, A.; Frennet, A. Ed.; Elsevier: Amsterdam, 1987; p 97. Yu-Yao, Y.-F. J. Catal. 1975, 39, 104. Kummer, J. In Catalyst for the Control of Automotive Pollutants; McEvoy, J. Ed.; ACS Series 143; American Chemical Society: Washington, DC, 1975; p 178. Truex, T. J.; Searles, R. A.; Sun, D. C. Platinum Metals Rev. 1992, 36, 2. Kobylinski, T.; Taylor, B. J. Catal. 1974, 33, 376. Shelef, M.; Gandhi, H. Platinum Met. Rev. 1974, 18, 2. Summers, J.; Baron, K. J. Catal. 1979, 57, 380. Taylor, K. C.; Schlatter, J. C. .l. Catal. 1980, 63, 53. Wang, T.; Soltis, R.; Logothetis, E.; Cook, J.; Hamburg, D. “Static Characteristics onrO2 Exhaust Gas Oxygen Sensors”, SAE 930352. Warrendale, Pa.: SAE International. 1993. Gulati, S. T. Society of Automotive Engineers Publ. No. 85013 and 900268. Taylor, K. C. CHEMTECH, 1990 Sept, 551. Nunan, J. G.; Robota, H. J.; Cohn, M. 1; Bradley, S. A. In Catalysis and Automotive Pollution Control I]; Crucq, A. Ed.; Elsevier: Amsterdam, 1991; p 221. Gandhi, H. S.; Piken, A. G.; Shelef, M.; Delosh, R. G. Society of Automotive Engineers Publ. No. 760201. 104. 105. 106. 107. 108. 109. 110. 111. 112. 34 Harrison, 8.; Cooper, B. J .; Wilkins, A. J. J. Plat. Metals Rev, 1981, 25, 14. Loof, P.; Kasemo, B.; Anderson, S.; Frestad, A. J. Catal. 1991, 130, 181. Oh, S. H. J. Catal. 1990, 124, 477. Golunski, S.; Roth, S. Catal. Today, 1991, 9, 105. Schaper, H., DoeSburg, E. B. M., and Van Reijen, L. L. Appl. Catal. 1983, 7, 211. Oudet, F., Courtine, P., and Vejux, A. J. Catal. 1988, 114, 112. Muraki, 111.; Yokota, K.; Fujitani, Y. App]. Catal. 1989, 48, 93. Lui, Y.; Dettling, J. “Evolution of a Pd/Rh TWC Catalyst Thechnology” SAE 930249, Warrendale, Pa: SAE International. 1993 Summers, J.; Williamson, W.; Henk, M. “Use of Palladium in Automotive Applications” SAE 920094, Warrendale, Pa: SAE International. 1988. Chapter 2 Catalysts Preparation and Characterization Techniques 2.1. Preparation of Catalysts Heterogeneous catalysts used in our work consist of two or three components: the active component, promoter, and support. The active component is responsible for a catalytic reaction. The promoter is catalytically inactive for a reaction, however, it enhances the catalytic performance through interaction with the active component. For Cu/Cr/Al203 and Cu/Ce/Al203 catalysts of interest in our study, Cu is the main active component for CO and CH4 oxidation and Cr and Ce are considered as promoters. The support plays a role of carrier for the active phase or promoter. The support provides a large surface area and disperses an active phase on it in order to maximize the surface area of the active phase. The y-alumina used in our study provides about 200 mZ/g surface area and 0.6 mL/g pore volume. Alumina is quite inactive for C0 and CH4 oxidation in our study. In other cases, alumina can be used as a catalyst by itself for alcohol dehydration, alkene isomerization, and deuterium exchange reaction [1]. 35 36 2.1.1. y-AI2O3 2. 1. 1. 1. Preparation of y-A I203 Aluminas exist as a various amorphous and micro-crystalline forms which can be distinguished by X-ray diffraction. y-AI2O3 is the most important catalyst support due to its thermally stability, controllable surface acidity, and high surface area [2]. Aluminum metal can be dissolved in both acidic and basic solvents because of its amphoteric character. Dissolved aluminum species will exist as solvated A131 ions in acidic solutions (< pH 2) and aluminate ions (A1045) in basic solutions (> pH 12). From using an aqueous acidic Al3+ solution (e.g. aluminum sulfate solution), aqueous basic A1045' solution (e.g. sodium aluminate solution), or mixture of the basic and the acidic solution, precipitation (aluminum hydroxide) can be obtained if the pH of the solution is properly controlled by addition of extra acid or base. The various structures of aluminas are generally prepared by subsequent dehydration (aging or heating) of the precipitated aluminum hydroxide. Since the structure transformation is determined by kinetic factors rather than thermodynamics, a mixture of different alumina phases are frequently obtained [3]. The transformation of aluminas observed upon aging and heating is Shown in Figure 2.1. 2.1.1.2. Structure of y-A I203 The oxygen arrangement in y-Al203 is cubic close packing, with occupation of the cations (Al3') located partially in tetrahedral and octahedral sites. This arrangement is similar to the spinel structure (e. g. MgAl2O4) which contains Mgzi ions in tetrahedral and Al3+ in octahedral interstices. The unit cell of a cubic close packed structure 7 3 83:; «£52m 25:9» Co SEES.“ 65 mo EEwEv usannom ._.~ 8sz cowacw MO? 339 38:3. .o._<-o Al 6.25 l 5.2-» l 2:538 05:5ch Al Esme... _< 0.8.x 98:58 9.22-me 953-com + wvzavo 228.38.“? :3 4 Al 6.2-x Alli 6.2-x All 0:236 9.3 meme O. 8 _ .A O. o: .58 O. 825m m m 15 O 2 In FREE mime :vmq w Invfl n V AON_<-o Al .O~_<-: Al 59? All .5328 ...O_< ’ A10 OmZA 8:995 oE—Eabu 2mm; mzoozc< . O. 8:52 O. 832 28 v.9. mama «vi. 5902 3:?“ 8:28 ...._< ”Os—e. mag—923‘ Al Em BEsoom o==_Smbooco_§ 41 228 302:5. Us Om _ _A Us coo omen 38 is shown in Figure 2.2. A face centered cube (fcc) contains four oxygen atoms, four octahedral and eight tetrahedral sites. The atomic ratio of metal to oxygen of a true spinel structure is 3:4, as in Mn304, Fe304 or C0304, while the atomic ratio of alumina is 2:3. Therefore, there are more cation vacant Sites in the A1203 structure (so called defect spinel) than in a true spinel [2]. The unit cell consists of 32 oxygen and 21-1/3 aluminum atoms, therefore 2-2/3 cation vacant sites per unit cell occur in the lattice [1]. These vacant sites sometimes cause formation of solid solutions with metal cations during catalyst preparation or aging at high temperature. The formula of the y-alumina is given formally as Ho,5Alo,5[Al2]O4 in which two out of four octahedral sites and one out of the eight tetrahedral sites are occupied by hydrogen and aluminum. However, hydrogen atoms are not located in tetrahedral or octahedral positions. Instead, they are at the surface as OH groups. One OH group is present for every two fcc units [3]. 2.1.1.3. Advantages of y-A ]_203 support Alumina support has been used the most widely in industry among variety of support materials due to its several advantages for practical use [2,4]. Alumina is cheap since its raw materials (Gibbsite or aluminum) are available in large amounts at low cost. Alumina is inert for many chemical reactions, has good mechanical strength (such as attrition resistance, hardness, compressive strength) and maintains high surface area at high temperature. Alumina is amphoteric so that its surface can be electrically charged either positively or negatively according to the pH of solution and can adsorb ions selectively. Active metal (Pt) is dispersed as small clusters on alumina due to the very low 39 . Octahedral site Oxygen . Tetrahedral site Figure 2.2. Lattice structure of spinel. 4O diffusion of the metal. In addition, alumina can be shaped with an accurate control of its porosity. 2.1.2. Catalyst Preparation Techniques 2.1.2.1. Incipient wetness impregnation Incipient wetness impregnation (or dry impregnation) is widely used in industry due to its simplicity. Accurate control of the amount of support and active ingredient can be achieved by this procedure. The support is contacted with a solution of appropriate active ingredient, corresponding in quantity to the total known pore volume of the support. Capillary force ensures intimate contact of the solution with the interior support surface (meso and micro-pores). The time required for incipient wetness impregnation can be calculated from the following equation, if the contact angle is assumed to be zero _ 411x yd (1) where r] is the liquid viscosity, x is the penetration distance into a capillary, y is the surface tension, and d is the pore diameter. A disadvantage of this method is the limited maximum loading in a single impregnation due to the solubility of the reagent [2]. In our work, all catalysts were prepared by incipient wetness technique with aqueous metal nitrate or alcoholic metal alkoxide solution. The binary component catalysts, such as CulCr/Al203 or CulCe/Al203, were prepared by step-wise incipient impregnation methods. The Cr or 41 Ce modified alumina catalysts were prepared first and then Cu was impregnated subsequently on the modified alumina. This Step-wise impregnation method allows us to study supported binary oxide catalysts in a simple, systematic way. 2. 1. 2. 2. Ion exchange Ion exchange is very suitable as a method for preparing atomically dispersed metal [5] or metal oxide [6] catalysts. The principle of this method is based on the fact that the support surface is polarized at the pH of the solution and counter ions are adsorbed by electrostatic attraction. Adsorption phenomena involving ion complexes are determined by the isoelectric point of the support, the pH of the aqueous solution, and the nature of the ion complex. The isoelectric point (or zero point of charge) of y-alumina is in the range of 7-9 (so called amphoteric point). Therefore its surface charge can be controlled by solution pH in order to adsorb cations or anions selectively. Several disadvantages of the ion exchange method must be considered, such as the adsorption capacity limit of the support, metal complex stability, and the solubility of the support material in acidic or basic solution [7]. 2. 1.2. 3. Grafting Grafting methods have been developed recently for preparing monolayer catalyst dispersions [8,9]. The method is based on condensation reaction between the metal alkoxide precursor and the hydroxyl group of the support. This chemical reaction yields the corresponding alcohol and a strong metal-support bond. Since alkoxide groups react 42 with water to be polymerized, nonaqueous solution such as alcohol derivative is commonly used. Several disadvantages of grafting methods are such as availability, solubility, moisture sensitivity of metal alkoxide, and reactivity with hydroxyl group of support. 2. 1. 2. 4. Precipitation Precipitation is one of the commonly used methods for making catalysts. The procedure involves precipitation from a mixture of more than two solutions or suspensions of material. The principle of precipitation is that the reaction between aqueous metal salt and basic solution (such as ammonium hydroxide, or ammonium carbonate) produces insoluble metal hydroxide or carbonate. Several advantages of precipitation method are: producing more homogeneous mixing of catalyst ingredients, homogeneous distribution of active phase throughout bulk and surface of catalyst, suitable for preparing high loading catalyst, and controllable pore size and distribution. However, there are disadvantages when two or more metal compounds are precipitated in different rate or sequence rather than simultaneously and some of the active phase will be unavailable to the surface due to being encapsulated by other material [2]. 43 2.2. Bulk Characterization 2.2.1. Surface Area Measurement (BET) The adsorption of a particular molecular species from a gas or liquid phase onto the surface of a solid is the principal method for measuring the total surface area of porous structures. If one can control the conditions of monolayer coverage and calculate the cross section of the molecule, then the surface area of a sample can be obtained from the volume of the adsorbed molecules. The molecules used for measuring surface area should be inert, small, spherical and easy to handle at the required temperature. Though not spherical, N2 is usually used since it is cheap and readily available in high purity. For measuring the surface area of solid materials, nonspecific physical adsorption is required. Physical adsorption isotherms have been classified into five groups by Brunauer, Deming, Deming, and Teller (called BDDT classification) [10] (Figure 2.3). Type I (Langmuir type) is commonly found for activated carbons, silica gels, and zeolites that contain micropores (< 10 A). Type II (sigmoid) is found for nonporous structures. Complete monolayer coverage is achieved at point B. Type III is found for systems having weak adsorption forces (e. g. water vapor on graphite). Type IV is a typically found for porous materials. The adsorption isotherm increases significantly at higher P/Po ratio due to pore condensation. Type V is similar to Type III with pore condensation at high P/Po ratio. Type III and V are relatively rare and difficult to use for surface area determination due to multilayer formation before completing monolayer coverage. 44 .83. 15> 262 “Eitsemvo—z 66.55% E hibesb Sentences»: .Z .0 63:03am EBB Eco—LEE .8398? go womb o>c 2:. omi .8388.“ 2523. o o. _ _ AGE .2335 9523. o. _ _ >— V m 0 n N c p m q m. MN Sawm— o o _ s — #— paqsospc tunomv 45 The most common method for measuring the surface areas of catalyst is that developed by Brunauer, Emmett, and Teller (called the BET method) [11]. They extended the Langmuir mechanism [12] to multimolecular layers. Langmuir assumed that an adsorption site on the solid surface accommodates one adsorbed molecule and the adsorption sites are uniform regardless of surface coverage, so that the adsorption probability (heat of adsorption) is the same at all sites. He also proposed that the rate of evaporation equals the rate of condensation for the first layer under equilibrium condition. For multimolecular layers, Brunauer, Emmett, and Teller assumed that the adsorption rate was proportional to the vacant sites of the lower layer and that the desorption rate was proportional to the adsorbed molecules in that layer. They also assumed that the heat of adsorption for all layers except the first one is equal to the heat of condensation of the molecule. The summation over an infinite number of adsorbed layers gives the following equation; P __ 1 + (C -1)P V(Po ‘- P) VmC VmCPo (2) where V is the volume of gas adsorbed at pressure P, V... is the volume of gas adsorbed in monolayer, P0 is the saturation pressure of adsorbate gas at the experimental temperature, C is a constant related exponentially to the heats of adsorption and condensation of the gas. The BET equation yields a straight line when P/V(P,,-P) is plotted versus P/Po. The BET plot is usually found to be linear in the range P/Po = 005-035, and this range is 46 usually used for surface area measurements. The deviation from linearity increases at higher P/P0 values (> 0.35) due to multilayer adsorption and/or pore condensation. The experimental error increases at lower P/P0 (< 0.05) due to the small amount of molecules adsorbed. The slope (S) is expressed as S = (C — 1) , WC and the intercept (I) is expressed as I = ] VmC Solving the equations (3) and (4) for Vm gives v...=—‘— S+I (3) (4) (5) The surface area of the catalyst can be calculated from Vm if the cross sectional area of an adsorbed molecule is known. Practically, the BET measurement can be performed with the value of the weight adsorbed on the sample (W & Wm) instead of volume (V & Va) [13]. From the PVT relationship the amount of adsorbate adsorbed on the sample can be calculated from calibrated integrator counts caused by thermal desorption of the adsorbate. 47 — _ Veal — (6) where W is the mass of adsorbate adsorbed on the sample, A is the sample integrator counts, A... is the calibration integrator counts, Veal is the calibration volume (cm3), P is the ambient pressure, M is the adsorbate molecular weight (g), R is the gas constant (82.1 mL atm deg'l mol’l). Therefore equation (2) and (5) can be rewritten as P = 1 + (C — 1)P (7) W(Po - P) WmC WmCPo W = .___l._ (3) m S + I The total surface area of the sample (8.) is determined from following equation; WmNAcs 51: M (9) where Wm is the weight of adsorbate adsorbed at a coverage of one monolayer, N is Avogadro’s number (602 x 1073), Acs is the cross sectional area of adsorbate molecule (N2 = 0.162 nmz), M is the adsorbate molecular weight (g). The specific surface area S is given by following equation lib-D pL , .27 48 s= . & no weight of sample (g) 2.2.2. Electron Spin Resonance (ESR) Electron spin resonance (electron paramagnetic resonance or electron magnetic resonance) can be applied only to paramagnetic species such as supported metal ions, surface defects, adsorbed molecules, and ions on heterogeneous catalysts. In ESR, different energy states arise from the interaction of the unpaired electron spin moment with the applied magnetic field (Zeeman effect). The Zeeman Hamiltonian (if) for the interaction of an electron with the magnetic field is given by following equation [14]; %=gBHsz on where g is the gyromagnetic value for a free electron (2.0023193), B is the electron Bohr magneton (eh/2mcc), SZ is the Spin operator, and If is the applied field Strength. In the absence of an applied magnetic field, the electrons are oriented at random. However, under applied magnetic field, the electron spin axes are oriented either in the same direction (+) or in the opposite direction (-) with respect to H. Therefore, there are two spin populations: electron spin of m, = -1/2 and m, = 1/2 has lower and higher energy state respectively (Figure 2.4). The transition energy difference is given as [14] 49 4+ E E = +1/2gl3H m8 = +1/2 / hv = gBH V rns = +1/2 0 m,=-1/2 \ m5 = -1/2 E = -1/2gBH Figure 2.4. Energy level splitting of the electron produced by an applied magnetic field. 50 AE=hv=gB% an The EPR experiment usually is performed by sweeping the magnetic field at fixed frequency. When the energy difference of two states (AE) matches with the hv, electrons spin flip from the antiparallel to parallel. This point is called the resonance condition at which energy absorption occurs. Since absorption curves are weak, EPR spectra are usually represented as a derivative curve. The microwave frequency region such as X- band (~ 9.5 GHz) and Q-band (~35 GHz) frequency ranges are commonly used in ESR. Water has to be avoid for ESR measurement due to Strong absorption of microwave. Quartz sample tube is preferred to Pyrex one because Pyrex absorb more microwave and also exhibit an IEPR signal. In this study, ESR has been used for semi-quantitative analysis for copper ions in Cu/Al203 catalysts. ESR selectively detects isolated Cu2+ in Cu/Al203 catalysts. Although clustered and crystalline CuO have paramagnetic Cuzi ions, their resonance absorption is too broad to be detected due to the interaction with neighbor copper ions. The number of spins can be obtained by the ratio of catalyst sample to the standard signal based on following equation [15]; = (Au/ASXSS/gu )2“s 55(55 +1) u PL Su(su +1) U” 51 where n“ and n5 are the number of spins per gram, A,' and A, are the areas of resonance absorption curves of sample and standard, gu and g. are the g values, Su and S. are total spins of sample and standard respectively, p is the mass per unit length of sample in the sample tube, and L is experimental factor. Therefore, the intensity ratio of sample vs. standard is proportional to the absolute number of spins per gram sample. Crystalline CuSO4.5H20 or phosphorous doped silicon embedded in polyethylene are usually used for intensity standards. In our work, a Ruby crystal standard was glued at the center of the inside cavity for improved accuracy, because each loading of the cavity will slightly alter the standing wave of the microwave field. However, the greatest source of error usually comes from double integration of the experimental derivative curve due to the difficulty in determinating the base line. Measuring peak to peak distance was adapted in this work rather than double integration since both methods showed very similar trends. 2.2.3. X-Ray Diffraction (XRD) 2. 2. 3. l . Qualitative analysis of X RD X-rays are sufficiently energetic to penetrate solid, and are well suited to probe their internal structure by diffraction. XRD is used to identify bulk crystalline phases, to quantify the amount of the crystalline phase, and to estimate particle sizes that are large enough to be detected. When an X-ray source target (e.g. Cu) is bombarded with high energy electron, the Cu Ku line (8.04 keV, 0.154 nm) is irradiated because a primary electron creates a hole in the K shell which is filled by an electron from the L shell with 52 emission of an X-ray photon. The KB line is also observed due to the electron transition from the M-shell to the K-shell. X-ray diffraction is the elastic scattering of X-ray photons by atoms in a periodic lattice. When X-rays are scattered by the ordered environment in a crystal, interference (both constructive and destructive) takes place among the scattered rays because the distances between the scattering centers are of the same order of magnitude as the wavelength of the radiation. Figure 2.5 shows how diffraction of X-rays by crystal planes allows one to derive lattice spacing by using the Bragg equation [16] nl=2dsin(O) (14) where n is the integer order of the reflection, A is the wavelength of the X-ray, d is the distance between two lattice planes, O is the angle between the incoming X-rays and the normal to the reflecting lattice plane. The XRD pattern is obtained with a stationary X-ray source and a movable detector. XRD peaks appear to be reflected from the crystal only if the angle of incidence for the X-ray satisfies the condition sinO = nMZd. At all other angles, destructive interference occurs. If one measures peak positions as the angle 28 (which is angle between the incoming and the diffracted beams) from a XRD pattern, the Bragg equation gives the corresponding lattice spacing, which are unique characteristic for a crystal compound. For powder sample such as catalyst sample, which randomly oriented crystalline particle, only a small portion of powder sample will be oriented by chance to provide a right angle with the incident beam for constructive interference. The limitation 53 in coming X-ray diffracted X-ray l e o d 9 pdsinO I Figure 2.5. Constructive interference of scattered X-rays from crystalline lattice. 54 of XRD is that diffraction peaks are only observed when the sample contains large enough crystalline particles. If the sample contains amorphous phase or very small crystalline particle size (less than 3 nm), no diffraction peaks will be observed due to destructive interference in scattering directions where the X-rays are out of phase [17]. 2. 2. 3. 2. Quantitative analysis of XRD In our work, quantitative X-ray diffraction data were obtained by comparing metal oxide (e.g. CuO, Cr203<104>, and CuCr204<311>) vs. support (Al2O3<400>) intensity ratios measured for catalyst samples with intensity ratios measured for the physical mixtures of pure CuO, Cr203, or CuCr2O4 and y-Al203. For the Ce/Al203 catalyst, the Ce02 <11 1> vs. AI2O3<440> intensity ratio was used because the A1203<440> peak is distorted severely by the Ce02 <220> peak at high loading. Pure CuO, Cr203, CeO2 or CuCr204 were finely grounded before measuring XRD intensities for preparation of calibration curves (intensity ratio vs. weight percent of standard compounds based on alumina). For maximum accuracy, the sample and the standard compound peak widths (full width at half maximum) should be comparable. This method assumes that metal oxides do not disrupt the y-Al203 spinel and do not affect the intensity of the A1203 XRD line. The error in this method is estimated to be i 20 %. Another quantitative XRD method compares the metal oxide (e.g. CuO) intensity relative to the metal oxide intensity for physical mixtures of pure CuO and y-Al203. This method does not use the A1203 peak as an internal standard. It assumes that the XRD intensity is proportional to the mass of CuO. This method may be more accurate when 55 supported metal oxide absorbs X-ray significantly, such as at high loading of transition metal or lanthanide element. 2. 2. 3. 3. Particle size determination The width of diffraction peaks carries information on the dimension of the reflecting planes. Perfect crystals produce very narrow diffraction lines. However, line broadening occurs for small size crystallites (< 100 nm) due to incomplete constructive or destructive interference. The mean crystallite Sizes (II) of the crystal particles can be detemtined from XRD line broadening measurements using the Scherrer equation [18]: d=KAchosO (15) where it is the X-ray wavelength (Cu K“ = 0.154 nm), K is the particle shape factor, taken as 0.9. The dimension d is defined as the effective thickness of the crystallite in a direction perpendicular to the reflecting planes. 0n the basis of this definition there is only a small dependence of K upon the crystallite shape. [3 is the full width at half maximum in radians. O is the angle between the beam and normal on the reflecting plane (half of 26). The detection limit of XRD technique depends on the X-ray source (target material). According to the Bragg (14) and Scherrer equations (15), if one use X-ray with smaller wavelength, the 3. and O terms become smaller. Therefore, smaller particles (d) can be detected. For example, XRD patterns from smaller particles can be obtained using a Ag (0.056 nm) or Mo (0.071 nm) target compared to Cu (0.154 am). Another way to 56 improve the XRD technique is to use synchrotron radiation source. The advantages are high intensity of radiation, better signal to noise, shorter collection time and varying wavelength of X-ray. The best resolved XRD patterns can be obtained by avoiding incident radiation energy close to absorption edge of element, because the scattering efficiency of an element decreases close to an absorption edge [17]. It must be noted that while X-ray line broadening measurements are useful as a relative measure of average crystallite size, the Scherrer formula should not be used for absolute determinations due to its failure to account for the effects of microstrain, the consequences of particle size distribution, and the uncertainty in correcting for instrumental broadening [19]. Hence, the values are meant to be used for relative comparisons only. 2.3. Surface Characterization 2.3.1. Ultrahigh Vacuum Chamber Electron spectrometers must operate under a vacuum of 10" Torr or lower. At pressures higher than 10'6 Torr, the electrons would be scattered in the path from the sample to the detector and the surface to be analyzed could be contaminated by gaseous impurities. A reasonable criterion for proper working pressure would be that no more than a few percent of an atomic layer of gaseous atoms should attach to the surface from the gas phase during measurement. 57 From the simple kinetics theory of gases, the rate of arrival can be expressed as follows [20], }/ 22 2 [ RT J =351x10 P (16) 211M JTM where r is particle flux (molecules cm'zs"), N is the number of gas molecules per cm3, R is the gas constant, P is a pressure (Torr), T is a temperature (K), and M is a molecular weight. If we assume that a monolayer consists of approximately 10ls atoms cm'2 on a solid surface and the gas molecule sticking coefficient is 1, at least 109 Torr is necessary to keep the surface clean for an hour. The ultrahigh vacuum (UHV) chamber used in our work (Figure 2.6) is pumped by a rough pump (Varian SD-90, 1.5 L/s pumping speed), a turbo molecular pump (Varian V-80, pumping speed 80 L/s), and an ion pump (Perkin-Elmer 207-0230, pumping speed 220 US) in combination with a titanium sublimation pump [20]. The typical base pressure of the chamber is normally approximately 1x103 Torr. 2.3.2. Transmission Electron Spectroscopy (TEM) The TEM instrument is very similar to an optical microscope, if one replaces the optical lenses by electromagnetic ones. TEM is able to observe the Size and shape of 58 sample manipulator top view X-ray source / \ t \x’; analyzer x p . , prep chamber V 1” \\ l; ‘1 side view pumping port————> e 6 ion gauge ——> o l t . 1.1 4 ion pump and tungsten sublimation pump Figure 2.6. Schematic drawings of XPS chamber. 59 Primary electron beam Secondary electrons Auger electrons X-rays Backscattered electrons Photons Diffracted electrons i Transmitted electrons Figure 2.7. Diagram of electron beam interaction with a solid sample. 60 particles directly. In TEM, a primary electron beam of high energy and intensity passes through condenser lenses to control the size of the beam that impinges on the sample, which is subsequently magnified by the Objective lenses to produce a image. The interaction of a high energy electron beam with a solid sample produces various electrons and X-rays (Figure 2.7). The intensity of transmitted electrons (unscattered electrons), which depends on the density and thickness of the sample, produces a two-dimensional image of the atomic mass. Diffracted electrons formed by the elastic scattering of the electron beam by the atoms in the specimen enable one to identify crystallographic phases of the sample as well as dark field images. If an area of a sample is crystalline, scattering occurs regularly, giving a pattern of sharp spots. Polycrystalline samples produce sharp lines consisting of spots. If an area of a sample is amorphous, the pattern will contain several diffuse rings. Backscattered electrons are formed from colliding electron beams with atoms in the sample and scattering the electrons backward. Secondary electrons are formed from several consecutive inelastic collisions. X-rays and Auger electrons are formed from the relaxation of core ionized atoms and reveal information on the sample composition. The emissions of photons ranging from UV to IR are mainly caused by recombination of electron-hole pairs in the sample. Typical operating conditions of a TEM instrument are 100-200 keV electrons, 106 Torr, 0.5 nm resolution and a magnification of 105-10" [17]. 2. 3. 2. 1. Instrumentationfm T EM Figure 2.8 shows the TEM column used in this study. The electron gun chamber commonly consists of a tungsten wire filament, shield, and anode. 61 \ High-voltage cable F Gun chamber Condenser-lens _ / aperture assembly Specimen / exchange port Condenser lens 1 Condenser lens 2 Cl 1 ) Ill Objective lens an] Objective lens aperture assembly DIfffaCth“ lens Diffraction-lens aperture assembly Intermediate lens , Specrmen Projector lens / traverse rods 0 O Specimen viewing f \ chamber ——i r—w {1‘ —-—Plate camera port Figure 2.8. TEM column (From Flegler, S. L.; Heckman, Jr. J. W.; Klomparens, K. L. Scanning and Transmission Electron Microscopy: An Introduction; W. H. Freeman and Company: New York, 1993. 62 The electrons produced by the heated tungsten filament are accelerated by the potential difference between the filament and anode, which is normally 20 - 100 kV. In this study, TEM images and selected area diffraction patterns were obtained with a JEOL 100CX2 using 120 kV primary voltage. The condenser-lens system is used to control electron illumination on the sample and viewing screen. The first condenser lens can condense the 50 um electron beam, which is the diameter of electron beam at crossover in electron gun chamber, to 1 um. The second condenser lens is used to control the beam brightness for image viewing and photography. The objective lens, the diffraction lens, the intermediate lens, and the projector lens all contribute to image magnification. The objective lens is the first magnifying lens and the sample is inserted into it. The diffraction lens is used for imaging diffraction patterns. The intermediate lens actually magnifies image by changing intermediate lens current. The projector lens projects the final magnified the image onto the viewing screen. Apertures are a piece of metal with a small hole, typically 30 - 1000 um, depending on their function. The various apertures are used to control the electron beam diameter or to remove stray or widely scattered electrons [21]. 2. 3. 2. 2. Sample Preparationfin' TIE/14 Generally, a thin copper grid (3 mm diameter, 50 pm thick) is used as a standard sample support disk for TEM measurements. The powder form of a catalyst sample is dispersed in water or organic solvent by an ultrasonic vibrator. In this work, methanol was used as a liquid media to prepare samples. A drop of the suspension was deposited on a holey carbon film coated copper grid and allowed to adsorb for about a minute. The 63 bulk of the solution was blotted off the edge of the drop with filter paper. The sample was ready for viewing after drying excess solution in the air. 2.3.3. X-Ray Photoelectron Spectroscopy (XPS) X-ray Photoelectron Spectroscopy (XPS or ESCA) is a powerfirl surface sensitive techniques which provides chemical or physical information about surface atoms within about 50-100 A thickness. From XPS measurements of heterogeneous catalysts, we can determine the oxidation state of surface atoms, dispersion of catalyst on the support, and quantitative analysis of all elements except H and He [22]. 2. 3. 3. 1. Qualitative Analysis by XPS The primary process of XPS is an ionization phenomenon brought about by photons, the so-called photoelectric effect (Figure 2.9). A photon with energy (hv) penetrates the surface and is absorbed by atoms. An electron with binding energy (Eb) below the Fermi level escapes from the solid surface with a discrete kinetic energy (Ek). Ek=hv-Eb-¢ (17) where (b is the work fiinction of the spectrometer. From the binding energy, which is characteristic for each element, the elements in a sample can be detected by XPS [23]. 64 2|: _._O_Q_O_ 23 L! // electrons hv hole 18 LOW A+hv——a A++e' Figure 2.9. Diagram of XPS Process. 65 2. 3.3.2. Quantitative analysis of XPS XPS can be used as a quantitative tool for the determination of the chemical composition of the surface region of a solid. We can determine the concentration of impurities in a host material (e.g. supported catalyst) of known chemical composition. The signal intensity (1,) observed in XPS can be expressed as follows [24], 11 =10111017~T(€I)D(81) (18) where I0 is the X-ray flux, r], is the density of atoms of i, O, is the photo excitation probability (photo ionization cross section), lfiei) is the mean free path (escape depth) of an excited electron with kinetic energy (Si) in the host, and D(ei) is the fraction of electrons detected by the analyzer (analyzer-detection efficiency). From the above equation, we can find that the relative impurity vs. host material density as follows [25], £111.): lieutenants.» 19 T1h I11917~T(81)D(€1) ( ) where n, and D11 are the density of impurity and host atoms. The relative energy dependence of D(8h)/D(8i) can be determined (usually D(e) is proportional to 1/8) and photoionization cross section (0’) have been reported by Scofield [26]. A method to estimate escape depths in elemental solids and compounds has been reported by Penn [25]. 66 A calculation of M48) for free-electron like materials (also valid for transition, noble, and rare earth metals) is as follows, Me) =[ 8 (20) aane+bfl where a and b depend on the electron concentration of the host material. However, in order to obtain III/71h for eqn.(18), we need only the relative energy dependence map/meg. Penn [25] has suggested a simple equation which is independent of a and b. Met): 810“: '23) (21) A1611) 82(ln81- 2'3) This eqn.assumes (with i 5 "/0 error) that the mean free path ratio shows little or no dependence on the density parameters a and b, but is dependent on eh and Si (for eh , a, 2 200 eV). Therefore, we can determine the relative impurity concentration. 2.3.3.3. Particle size determination It has been shown by Defosse et al. [27] that one may calculate the theoretical intensity ratio I°p/I°, expected for a supported phase (p) atomically dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and Moulijn [28] has been used in the present investigation. They proposed that catalysts consisted of sheets of .1 I I e //////////////////////Z//// 2 WW I I I5 I %////////////////////////////A 1+1 .. Figure 2.10. E. # of layers It. \\ Kerkhof and Moulijn's model of the supported catalysts. 68 support (dimension t) with cubic crystallites (dimension c) in between the layer structure (Figure 2.10). The photoelectron cross sections and the mean escape depths of the photoelectrons used in these calculations were taken from Scofield [26] and Penn [25] respectively. In the case of a crystalline sample, the intensity ratio can be expressed as i— s D(e, )o,2a(1 — e'B’ ) [1,] =(B-j ”(e,).,s,a_.-«)a..-B=) <22) 5 cryst b where II, is the intensity of the electrons from element p in the catalyst, Is is the intensity of the electrons from element 5 in the support, 0,, and OS are photoelectron cross sections, (p/s)b is the bulk atomic ratio of the promoter and support, [31 and B2 are the dimensionless support thickness, and D are the detector efficiencies which may be a function of the kinetic energy (e) of the electrons. on is dimensionless crystallite size parameter which can be calculated from following equation, 01 = c / A”, (23) where A", is escape depth of electron from the element p in the catalyst traveling through the element p. In the case of a monolayer catalyst, or is small. Therefore the intensity ratio can be expressed as follows, 69 I_, _(£) D(e,)o,(3,(1+e’9=) (24) I mono- S 1’ D(£3)O,2(I'CTBZ) B1 and [32 can be calculated from the following equations, Bl=t/1,,, [La/1,, (25) where A“ is the escape depth of electron from support traveling through the support. A, is the escape depth of electron from the promoter traveling through support. We can calculate the crystalline size of supported materials with the following eqn. and eqn. (24), if it", is known. The following eqn. is obtained from eqn. (22) and eqn. (23). U 1‘ I 5 CM, (1— e‘“) = (26) s mono where (Ip/I‘)m Is a experimental data, (IP/Is)mono is a theoretical monolayer Intensrty ratio F—II ~ '3 —1 calculated from eqn. (24). 70 2.3.3.4. HS shake up satellite peaks Shake-up processes arise when the photoelectron excites an outer electron to a bound excited state in the atom or molecule. The photoelectron loses an amount of kinetic energy corresponding to the excitation and that appears at a higher binding energy in the spectrum. The origin of the satellite peaks is considered to be the promotion of 3d electrons to 45 and/or 4p levels [29-32] or alternatively to be a charge transfer of ligand electrons to unfilled 3d orbitals (e.g. 0 2p —> Cu 3d in case of CuO) [33-38]. Such transitions are not seen for Cull or Cu0 compounds. It is well known [39] that transition metal ions with unfilled 3d orbitals and completely empty 3d orbitals show satellite peaks due to shake-up, that the satellite peaks are not observed when the 3d orbitals are completely filled (Zn2+ and CuT), and that the transition is a monopole excitation, conserving molecular symmetry and parity. Taking into consideration the above facts, the explanation based on the ligand charge transfer seems more reasonable than that based on the 3d —> 4s, 4p transitions. In this study, we discussed the shake up satellites associated with the Cu(2p3/2) level on the basis of the monopole charge transfer mechanism. Thus, the satellite Structure provides information about the bonding nature of Cuzi ion in the catalyst. In order to estimate the nature of the distorted copper oxide structure, a correlation between the satellite intensity of the Cu(2p3/2) level and the symmetry of the Cu2+ ion is required. The detailed structure of the satellite peaks depended on the catalyst composition. Frost et al. [29] reported that Cu2+ ion in a distorted octahedral symmetry Cu(acac.)H2O, Cu(NO3)23H20, Cu(OAc)2, CuC03 and Cu(OH)2 show generally stronger 71 and well resolved satellite peaks compared with those in a distorted square planar symmetry (CuO, CuCl2, CuBr2). Strohmeier et al. [40] reported that the Cu 2p3/2 satellite structure of bulk CuAl204 (Cu ions in 60 % T4 and 40 % 02, sites) was more intense than that of CuO. The surface Cu2+ ions in CulAl203 catalysts at low loading (distributed 90 % in a tetragonally distorted octahedral environment and 10 % in tetrahedral environment) exhibit satellite peaks similar to CuO rather than CuAl204 [40,41,42]. It has been suggested [40] that this did not necessarily mean that well dispersed or amorphous CuO was present on the catalysts, but simply that the copper oxide surface spinel species is chemically different from CuO and CuAl204. 2.4. Catalytic Activity Measurement The catalytic reactor system built for this study is shown in Figure 2.11. Measurement of C0 and CH4 oxidation activity was performed in a flow microreactor. Approximately 30 and 100 mg of catalysts was supported on a glass frit (70 - 100 pm) for C0 and CH4 oxidation, respectively. The temperature was measured with a K-type thermocouple located just above the catalyst bed. The reactor was heated by a tube fumace (Lindberg) with temperature being controlled within 1 °C by an Omega CN 1200 temperature controller. Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers. Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD and interfaced to a Hewlett-Packard 3394A integrator. Samples (0.5 and 5 cc) were collected by Valco sample loops and electric actuator for C0 and CH4 72 Zeolite Mass Flow Controllers He 02 C0/02/He CHJOJHe \> Bellows Valves Controller Temperature 1 Union Tee Furnace : Thermocouple Glass Fritz Vent 3-way Valve Sample Loop Water Trap Inte grater Figure 2.11. Schematic of the catalytic reactor system. 73 oxidation, respectively. Reaction products were separated on a 6 ft 60/80 mesh Carbosieve S-II column. Reaction temperature control, sample injection into GC, integration of GC response were controlled automatically by a controller constructed in house. Prior to the first activity measurement, the catalyst was pretreated with a mixture of 5% O2/He (99.5 % purity for 02, 99.995% purity for He, AGA Gas Co.) stream (143 cc/min) at 350°C for 1 hr to remove any impurities adsorbed on the surface during catalyst preparation and storage. CO oxidation reactions were performed with a 80 cm3/min flow of 4.8% CO/9.8% 02/85.4% He gas mixture (AGA, purity > 99.99%) at 100-310°C. Methane oxidation reactions were performed with a 15 cm3/min flow of 0.98% CHI/5.25% 02/93.77% He gas mixture (AGA, purity > 99.99%) at 310-450 °C. All activity measurements were obtained under steady-state conditions at conversions less than 15%. Water produced during methane oxidation was frozen downstream from the reactor in a trap maintained below -40 °C with a mixture of ethanol and dry ice. The TCD response factor for each molecule was obtained using a standard gas mixture (Scotty II Analyzed Gases 1% CH4, C0, C02, H2, 02 in N2, Supelco). The relative thermal response factors for C0 and CH4 based on CO2 are 1.14 and 1.33 respectively, which agree well with the reported values [43]. IO. 11. 12. l3. 14. 15. 74 2.5. References Knozinger, H.; Ratnasamy, P. Catal. Rev. Sci. Eng. 1978, 17, 31. Satterfield, C. N. Heterogeneous Catalysis in Practice; McGraw-Hill: New York, 1980. Doesburg, E. B. M. and van Hooff, J. H. C. In Catalysis: An Integrated Approach to Homogeneous, Heterogeneous and Industrial Catalysis; Mouli j n, J. A.; van Leeuwen, P.W.N.M.; van Santen, RA. Ed.; Elsevier: Amsterdam, 1993; pp 309. Nortier, P.; Soustelle, M. in “Catalysis and Automotive Pollution Control”, Crucq, A; Frennet, A. Eds. Elsevier, Amsterdam 1987 pp275. Benesi, H. A.; Curtis, R. M.; Studer, H. P. J. Catal. 1968, 10, 328. Scierka, S. J .; Houalla, M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1995, 99, 1537. Brunelle, J. P. in “Preparation of Catalysts II” Delmon, B.; Grange, P.; Jacobs, P., Poncelet, G., Ed. Elsevier, Amsterdam pp 21 1, 1979. Kijenski, J.; Baiker, A.; Glinski, M.; Dollenmeier, P.; Wokaun, A. J. Catal. 1986, 101, 1. Okabe, K.; Sayama, K.; Matsubayashi, N.; Shimomura, K; Arakawa, H. Bull. Chem. Soc. Jpn. 1992, 65, 2520. Brunauer, S.; Deming, L. S.; Deming, W. S.; Teller,E. J. Am. Chem. Soc, 1940, 62, 1723. Brunauer, S.; Emmett, P.H.; Teller,E. J. Am. Chem. Soc, 1938, 60, 309. Langmuir, I. J. Amer. Chem. Soc. 1916, 38, 2221. Manual onuantasorb Jr. Quantachrome 1985. Drago, R. S. “Physical methods for Chemists” 2nd edition, Saunders College Publishing, New York 1992. Delgass, W. N.; Haller, G. L.; Kellerman, R.; Lunsford, J. H. “Spectroscopy in Heterogeneous Catalysis” Academic Press, New York 1979. 16. 1’7. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 2.8- 29_ 30- 31. 32. 33. 75 Bragg, W. L. Proc. Camb. Phil. Soc. 1912, 17, 43. Niemantsverdriet, J. W. “Spectroscopy in Catalysis” VCH, Weinheim, 1993. Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline and Amorphous Materials, lst.Ed. Wiley, New York, 1954. Cohen, J. B. Ultramicroscopy 1990, 34, 41. O’Hanlon, J. F. “A User’s Guide to Vacuum Technology” 2nd. edition, Wiley, New York, 1989. Flegler, S. L.; Heckman, Jr. J. W.; Klomparens, K. L. Scanning and Transmission Electron Microscopy: An Introduction; W. H. Freeman and Company: New York, 1993. Hercules, D. M.; Hercules, S. H. J. Chem. Educ. 1984, 61, 483. Briggs, D. and Seah, M. P. Practical Surface Analysis by Auger and X-Ray Photoelectron Spectroscopy, 2nd ed.; Wiley: New York, 1990. Fadley, C. S., Baird, R. J., Siekhaus, W., Novakov, T., and Bergstrom, S. A. L. J. Electron. Spectrosc, 1974, 4, 93. Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29. Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129. C. Deffosse, D. Canesson, P. G. Rouxhet and B. Delmon, J. Catal. 51(1978)269. Kerkhof, F. P. J. M. and Moulijin, J. A. J. Phys. Chem. 1979, 83, 1612. Frost, D. C.; Ishitani, A.; McDowell, C. A. Mo]. Phys. 1972, 24,861. Rosencwaig, A. J. Phys. Rev.Lett. 1971, 27, 479. Yin, L. Chem. Phys. Lett. 1974, 24, 81. Fiermans,L. Surf Sci. 1975,47, l. Hufner, S. Phys. Rev.B. 1973, 7, 5086. 34. 35 36. 3'7. 38. 39. 40. 41, 42. 76 Kim, K. S. J. Electron Specsc. Relat. Phenom. 1974, 3, 217. Kim, K. S. J. Electron Specsc. Relat. Phenom. 1972 1, 253. Wallbank, B. J. Electron Specsc. Relat. Phenom. 1974, 5, 259. Wallbank, B. J. Phys. C. 1973, 6, 493. Carlson, T.A. J. Electron Specsc. Relat. Phenom. 1974, 5,247. Carlson “Photoelectron and Auger Spectroscopy” 1975. Strohmeier, B. R.; Leyden, D. E.; Field, R. S.; Hercules, D. M.; Petrakis, L. J. Catal. 1985, 94, 514. Friedman, R. M.; Freeman, J. J.; Lytle, F. W. J. Catal. 1978, 55, 10. Tikhov, S. F.; Sadykov, V. A.; Kryukova, G. N.; Paukshtis, E. A.; Popovskii, V. V.; Starostina, T. G.; Kharlamov, G. V.; Anufrienko, V. F.; Poluboyarov, V. F.; Razdobarov, V. A.; Bulgakov, N. N; Kalinkin, A. V. J. Catal. I992, 134, 506. McNair, H. M.; Bonelli, E. J. “Basic Gas Chromatography” Varian Aerograph, 1969. Chapter 3 The Effect of Crystallinity on the Photoreduction of Cerium Oxide: A Study of CeOz and Ce/AIZO3 Catalysts 3.1. Abstract The effect of crystallinity on the photoreduction of cerium oxide (Ce4+ to Cesi) d uring X-ray photoelectron spectroscopy (XPS) analysis of CeOz and Ce/A1203 catalysts has been determined. Cerium oxide samples were prepared by calcining cerium (IV) methoxyethoxide in air at different temperatures. The structure of the resulting metal Oxides was determined using X-ray diffraction (XRD) and transmission electron microscopy (TEM). The oxide obtained from calcination at 200 °C consisted primarily of amorphous cerium oxide, while the material produced at 750 °C contained large CeOz cI3/stallites. XPS Ce 3d spectra obtained for the samples showed that the amorphous cerium oxide sample was reduced more extensively than the crystalline material during XPS analysis. Ce/A1203 catalysts (5.1 wt.% CeOz) were prepared by incipient wetness impregnation using ammonium Ce(IV) nitrate and Ce(IV) methoxyethoxide precursors. ){RD and XPS analyses of the Ce/A1203 catalyst derived from the nitrate precursor (designated “CeN”) indicated that most of the cerium was present as poorly dispersed CeO; crystallites. The catalyst prepared using the alkoxide precursor (designated “CeA”) 77 78 also contained poorly dispersed cerium oxide; however, XRD data suggested that very little of the cerium oxide phase was crystalline. The XPS Ce 3d spectrum obtained for the CeN catalyst was typical of Ce(IV) oxide while the spectrum collected for the CeA catalyst was similar to a Ce(III) compound. These results have been attributed to enhanced photoreduction of the amorphous cerium oxide present in the CeA catalyst. 3.2. Introduction Cerium oxide is an important component of superconductors [1,2], thin film 0 ptical devices [3,4], ceramics [5], gas sensors [6], and heterogeneous catalysts [7-13]. In In any applications, knowledge of the cerium oxidation state is necessary for understanding ‘2 he properties of the material. For example, Tranquada et al. [1] have reported that the d etermination of cerium valence in cerium doped copper oxide superconductors is i m portant for understanding the details of electron-pairing in these systems. The oxygen Storage capabilities of CeOz/A1203 catalysts are commonly attributed to reversible changes i r) the cerium oxidation state [7-9, 12]. X-ray photoelectron spectroscopy (XPS) is frequently used to determine the Oxidation state of surface species. However, it is well known that many types of Compounds are not stable under the conditions used for XPS measurements [14,15]. Several factors such as X-ray flux [16], integrated X-ray dose [17], secondary electrons from the X-ray source [18], sample charging [19], temperature [20], and high vacuum [21] are known to be responsible for sample damage. Cerium oxide photoreduction - 79 during XPS measurements has been reported by several researchers [19,22-25]. Paparazzo et al. [24,25] attributed cerium oxide photoreduction primarily to intense heating of the sample surface. In contrast, Laachir et al. [26] found that CeOz samples vvith both low and high surface area (5 mZ/g and 115 m2/g, respectively) were stable for up to four hours during XPS analysis. Dausher et al. [22] reported that a Ti02-Ce02 mixed oxide prepared using the sol-gel method was more reduced during XPS measurement than t he physical mixture of the two oxides. These seemingly inconsistent results may be due t O a difference in the structures of cerium oxides analyzed. We are interested in the surface chemistry of cerium oxide materials that can be u sed as chemical sensors [27] and heterogeneous catalysts [28]. XPS is widely utilized to d etermine the effect of preparation method and sample pretreatment on the distribution of c erium oxidation states in such materials [10,13,29,30]. 1E] Fallah et a1. [19] demonstrated 1: hat ceria photoreduction could be limited by controlling X-ray beam power, irradiation t i me, and sample charging. However, extended analysis times are sometimes necessary to obtain reasonable quality spectra for materials with low cerium contents. This may cause a Change in cerium oxidation state and invalidate the results. Therefore, it is important to kn ow which types of materials are susceptible to photoreduction during XPS analysis. In this study, we have examined the effect of crystallinity on the photoreduction of cerium Oxide. Amorphous and crystalline CeOz materials were prepared by calcining cerium(IV) methoxyethoxide in air at 200 °C. and 750 °C, respectively. X-ray diffraction (XRD) and transmission electron microscopy (TEM) have been used to determine the structure of these oxide materials. In addition, Ce/A1203 catalysts were prepared using ammonium 8O Ce(IV) nitrate and Ce(IV) methoxyethoxide precursors. XRD and XPS have been used to determine the crystallinity, dispersion, and oxidation state of the alumina supported cerium oxides. The structural information obtained for CeOz samples and Ce/A1203 catalysts has been correlated with the extent of cerium photoreduction observed during XPS measurements. 3.3. Experimental Standard Materials. Cerium(III) acetylacetonate hydrate was obtained from Aldrich Chemical Company, Inc. CeOz was prepared by calcining ammonium cerium(IV) nitrate (Analytical Reagent, Mallianrodt, Inc.) in air at 800 °C for 16 h. The XRD pattern of the CeOz standard compound matched the appropriate Powder Diffraction File E: 3 l]. Cerium Oxide Sample Preparation. Cerium oxide samples were prepared by allowing cerium(IV) methoxyethoxide solution (18-20 % cerium alkoxide in methoxyethanol; Gelest, Inc.) to dry on quartz slide substrates at room temperature for 5 days in a nitrogen purged glove bag. These samples were subsequently calcined in air at 200 °C and 750 °C for 36 h and 16 h, respectively. TGA and FTIR analyses indicated that Cerium methoxyethoxide decomposition was complete afier calcination steps. Samples prepared at 200 0C and 750 °C are designated as Ce200 and Ce750, respectively. Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y- alumina (Cyanamid, y-A1203, surface area = 203 mz/g, pore volume = 0.6 mL/g). The 81 alumina support was finely ground (< 230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation. Catalysts were prepared using a deionized water solution of ammonium cerium(IV) nitrate (Mallinckrodt Inc., Analytical Reagent) or an ethanol solution of cerium(IV) methoxyethoxide (Gelest Inc., 18-20 "/0 cerium alkoxide in methoxyethanol). The catalyst derived from the cerium nitrate precursor (designated “(IeN”) was dried in air at 120 °C for 24 h and calcined in air at 500 °C for 16 h. The catalyst prepared using the cerium alkoxide precursor (designated “CeA”) was impregnated and subsequently dried at room temperature for 48 h in a N2 purged glove 1) ag prior to firrther drying in air at 120 °C for 24 h and calcination in air at 500 °C for 16 h - The Ce loading was held constant at 5.1 wt.% CeOz. BET Surface Area. Catalyst surface areas were determined using a QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 g of a catalyst sample was outgassed i n a Ny/He mixture (5% N2) at 350 °C for 1 h prior to adsorption measurements. The measurements were made using relative pressures of N; to He of 0.05, 0.08, and 0.15 (N2 811 rface area = 0.162 nmz) at 77 K. Electron Microscopy. Transmission electron microscopy (TEM) images and Selected area electron diffraction (SAED) patterns were obtained with a JEOL 100CX2 using 120 kV primary voltage. Cerium oxide samples were ultrasonically dispersed in methanol and deposited on a holey carbon film supported on a conventional copper grid (3 00 mesh, 3 mm). X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a Rigaku XRD diffractometer employing Cu Kat radiation (2» = 1.5418 A). The X-ray was 82 operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scan rate of 0.25 deg/min. with divergence slit and scatter slit widths of 1°. Cerium oxide samples were run as powders mounted in the cavity of a circular silicon sample holder (C-12, Dow Corning). Catalysts were run as powders packed into a glass sample holder having a 20 x 1 6 x 0.5-mm cavity. The mean crystalline sizes ((3 ) of the CeOz particles were determined from XRD line broadening measurements using the Scherrer equation [32]: d=IOt/Bc059 (1) Where is. is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the fill width at half maximum (fwhm), in radians, of CeOz <1 11> and <220> lines measured for Ce/A1203 and CeOz samples, respectively. Quantitative X-ray diffraction data for Ce/A1203 catalysts were obtained by comparing Ce02 <1 1 1>/A1203<440> intensity ratios measured for catalyst samples with peak ratios measured for physical mixtures of CeOz standard and y-A1203 support. This method assumes that cerium addition does not disrupt the y-A1203 spinel and subsequently afiect the intensity of the A1203 peak. The error in this method was estimated to be i 20 0/0 . XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science instrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20 mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The 3:? V] (In I“) 83 instrument typically operated at pressures near 1 x 10'8 Torr in the analysis chamber. The take-off angle between the sample and analyzer was approximately 45° for all measurements. The distance between the X—ray source and sample was maintained at approximately 3/4 inch for all analyses. Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer. CeOz samples were analyzed as prepared while catalyst samples were analyzed as powders dusted onto double-sided sticky tape or spray coated onto a quartz slide using a methanol suspension of the catalyst. XPS binding energies were measured with a precision of :t 0.2 eV, or better. Binding energies for the st andard samples were referenced to the C ls line (284.6 eV) of the carbon overlayer. S i mce the charging observed for the Ce200 and Ce750 samples changed during our experiments, we were concerned that the C Is charge reference may be less accurate than a Ce 3d derived charge correction. Thus, we have charge corrected the Ce 3d spectra Obtained for the cerium oxide samples using the Ce 3d u'" peak (916.7 eV). Using the Ce 3 d u'" reference energy, the binding energy of the C 15 peak measured for the CeOz Standard was 284.6 eV. The Ce 3d u'" peak has been commonly used to compensate for Sample charging in the ceria photoreduction studies [19,24,25]. Binding energies for the catalyst samples were referenced to the Al 2p peak (74.5 eV) of the alumina support. The Ce 3d spectra of Ce200 and Ce750 samples were obtained at 15 min. intervals up to 300 min. The Ce 3d region of CeN and CeA catalysts were measured for 600 min. to obtain reasonable quality spectra. The extent of cerium reduction was determined using composite spectra [33] derived from the Ce 3d spectra measured for CeOz and Ce(III) 84 acetylacetonate hydrate standard compounds. We estimate the error in this method is less than i 5 %. Quantitative XPS Analysis. It has been shown by Defosse et al. [34] that one may calculate the theoretical intensity ratio I°p/1°. expected for a supported phase (p) atomically dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and Moulijn [35] has been used in the present investigation. The photoelectron cross sections and the mean escape depths of the photoelectrons used in these calculations were taken from Scofield [36] and Penn [37], respectively. For a phase (p) present as discrete particles, the experimental intensity ratio (Ip/Is) is given by the following expression: Ip/r, = 1°,,/1°.[1-exp(-d/x,)]/ cm.,, (2) where (I°,/I°,) is the theoretical monolayer intensity ratio, d is the length of the edge of the cubic crystallites of the deposited phase and 3.? is the mean escape depth of the photoelectrons in the deposited phase. 3.4. Results and Discussion XPS Analysis of Standard ( 'omponnds. The XPS Ce 3d spectra of cerium compounds are known to be complicated due to hybridization of the Ce 4f with ligand orbitals and fractional occupancy of the valence 4f orbitals [38-44]. Figure 3.1 shows the 85 “ w " \ (b) l l l 920 900 880 Binding Energy (eV) Figure 3.1. Ce 3d XPS spectra measured for (a) CeO2 and (b) Ce(III) acetylacetonate hydrate. 86 Ce 3d spectra measured for CeOz and Ce(III) acetylacetonate hydrate standard compounds. The peak shapes are similar to those reported for Ce(IV) and Ce(III) compounds [10,29,45,46]. The Ce 3d spectrum measured for CeOz contains three main 3d5/2 features at 882.6 eV (v), 889.3 eV (v"), and 898.6 eV (v"') and three main 3d3,2 features at 900.8 eV (u), 907.8 eV (u"), and 916.7 eV (u'"). The high binding energy doublet v'"(u"') has been assigned to a final state with primarily 4f“ configuration. The v" and v (u" and u) doublets are generally assigned to final states with strong mixing of 40 and 4f2 configurations. These states arise from the core hole potential in the final state and 4f hybridization in the initial state [38-43]. The Ce 3d spectrum measured for Ce(III) acetylacetonate hydrate contains two main 3d5,2 peaks at 881.4 eV (v) and 885.2 eV (v') and two main 3d3a features at 899.5 eV (u) and 903.6 eV (u'). These doublets have been assigned to final states where 4fl and 4f2 configurations are strongly mixed [44]. Structure of Ce200 and Ce750. TEM images and SAED patterns of Ce200 and Ce750 samples are shown in Figure 3.2. The micrograph obtained for the Ce200 sample consists of thick agglomerates with no preferred shape or orientation; however, small grains (~ 3 nm diameter) are observed at the edges of agglomerates. The SAED pattern of the Ce200 sample shows seVeral diffuse rings indicative of an amorphous phase. The TEM image obtained for the Ce750 sample shows faceted particles. The SAED pattern shows several sets of sharp spots characteristic of a crystalline phase. Figure 3.3 shows the XRD patterns obtained for Ce200 and Ce750 samples. The most intense CeOz peak (CeOz <1 1 1>) was excluded from this figure due to overlap with a broad silica peak from the sample substrate. The XRD pattern obtained for the Ce200 87 (a) 88 (b) Figure 3.2. TEM micrographs and SAED patterns obtained for (a) Ce200 and (b) Ce750 samples. 89 <220> <311> 1000 <222> JL J (b) 100 l l l 4O 50 60 Degrees (20) Figure 3.3. XRD patterns measured for (a) Ce200 and (b) Ce750 samples. ~ gmdes thank pads l C6505 REE m Oi Im 90 sample shows broad, weak peaks characteristic of poorly crystallized CeOz. The diffraction pattern measured for the Ce750 sample shows more intense, sharper CeOz peaks. In addition, the CeOz <222> peak can be observed in the diffraction pattern of the Ce750 sample. CeOz particle sizes determined using XRD line broadening calculations are significantly larger for the Ce750 sample (34.8 nm) than for the Ce200 sample (2.2 nm). For the Ce200 sample, the XRD determined CeOz particle size is similar to the grain size observed in TEM. In summary, TEM and XRD results indicate that calcination of cerium alkoxide leads to the formation of cerium oxide. The extent of crystallinity is affected by the calcination temperature. The Ce200 sample consists primarily of large, amorphous cerium oxide particles that contain small CeOz crystallites. The Ce750 sample consists primarily of large CeOz crystallites. No evidence for the presence of an amorphous cerium oxide phase has been observed for the Ce750 sample. Photoreduction of Ce200 and C e 750. Figure 3.4 shows the Ce 3d spectra measured for the Ce200 sample as a function of irradiation time. The spectrum obtained in the first 15 min. of analysis has features similar to those observed for the CeOz standard; the Ce 3d5/2 binding energy (882.7 eV) is identical, within experimental error, to the value measured for Ce02 (882.6 eV). As the analysis time increases, new features characteristic of Ce3‘ appear in the Ce 3d spectra. For intermediate irradiation times (30 and 60 min), this is most evident as a loss of resolution between the v and v" peaks of Ce4+ due to the growth of the v' peak characteristic of C e3 ’. We believe that the new features appearing in the Ce 3d spectrum are due to the photoreduction of Ce4+ to Ce‘ii during the XPS analysis. This is consistent with numerous reports in the literature [19,22-25]. We also note that 91 (f) (d) (C) (b) (a) l l l 920 900 880 Binding Energy (eV) Figure 3.4. Ce 3d XPS spectra measured for the Ce200 sample after irradiation times of (a) 15 min, (b) 30 min, (c) 60 min, (d) 120 min., (e) 180 min, and (f) 300 min. 92 increasing analysis time broadens the Ce 3d5/2 region on the low binding energy side. Paparazzo [24] has reported that such broadening can be attributed to the superposition of “Ce203” components and CeOz features. The spectrum measured afier 300 min. clearly shows features characteristic of Ce3+ and Ce“. The Ce 3d5,r2 binding measured for the Ce3+ v' component (885.4 eV) in the irradiated sample is close to values measured for Ce(III) acetylacetonate hydrate (885.2 eV) and reported for Ce203 (885.3 eV [25] and 885.8 eV [47]) compound. In addition, Le Normand et al. [48] reported that a v' peak (885.5 eV) appeared in the XPS Ce 3d spectra measured for C602 reduced in H2 at 900 °C. The XPS Ce 3d spectra measured for the Ce750 sample are shown as a fimction of irradiation time in Figure 3.5. One can clearly see that these spectra also change with time; however, the growth of Ce?" features is less dramatic than those observed for Ce200 sample. The XPS Ce 3d binding energies measured for the Ce750 sample are identical, within experimental error, to the values obtained for the Ce200 sample. For the Ce750 sample, well resolved v and v" peaks are observed for the duration of the experiment. Figure 3.6 shows the extent of cerium reduction with respect to irradiation time for Ce200 and Ce750 samples. As anticipated from Figures 3.4 and 3.5, the extent of reduction increases rapidly with time up to approximately 180 min. For longer analysis times, the change in fraction reduced is less pronounced. This indicates that the rate of reduction decreases with irradiation time. Similar results have been reported by Wallbank et al. [49] in the case of Cqu photoreduction (Cu2+ to Cu"). Figure 3.6 also shows that the cerium oxide calcination temperature affects the extent of reduction. - As noted above, 93 (f) (6) (d) (C) (b) (a) l l l 920 900 880 Binding Energy (eV) Figure 3.5. Ce 3d XPS spectra measured for the Ce750 sample after irradiation times of (a) 15 min, (b) 30 min., (c) 60 min., (d) 120 min, (e) 180 min, and (f) 300 min. 94 35 30- O O 25- O 20- O 15- 0 Fraction Reduced (%) O O O 10- 0 54a O-‘I I I 1 1 l I I I r I I I I I I I O 50 100 150 200 250 300 350 400 450 Irradiation Time (min) Figure 3.6. Fraction of CeOz reduced as a function of irradiation time for the Ce200 (O) and Ce750 (0) samples. 95 TEM and XRD analyses have shown that the Ce200 sample consists of small CeOz crystallites imbedded in large amorphous cerium oxide particles, while the Ce750 sample contains only large CeOz crystallites. Thus, we believe the photoreduction results indicate that amorphous cerium oxide is more readily reduced than crystalline CeOz during the XPS measurement. Structure of Cerium/Alumina Catalysts. The surface areas of CeN and CeA catalysts (195 mZ/g) are identical and slightly lower than the value obtained for the A1203 support (203 mz/g). Figure 3.7 shows the )GKD patterns obtained for CeN and CeA catalysts. The XRD pattern of the CeN catalyst shows intense peaks characteristic of CeOz; however, only weak CeOz lines are observed in the diffraction pattern of the CeA catalyst. Quantitative XRD measurements indicate that the CeN catalyst contains 4.6 wt.% crystalline CeO; phase, while the CeA catalyst contains only 0.3 wt.% crystalline C602. Variation in the particle size of the cerium oxides determined from XRD line broadening calculations and XPS Ce/Al intensity ratios are shown in Table 3.1 for both CeN and CeA catalysts. The CeOz particle size determined for the CeN catalyst using XRD is slightly larger than the value obtained for the CeA catalyst. However, the C602 particle size calculated from XRD analysis of the CeA catalyst must be considered as approximate due to the low intensity of the CeOz <1 1 1> peak. The cerium oxide particle sizes determined for CeN and CeA catalysts using XPS intensity ratios are identical and smaller than the values obtained from XRD. 96 f r r F ' CeO2 C602 <200> (c) w <111> (b) A1203 <311> A1203 (a) <220> A1203 <222> r r r I 25 30 35 40 Degrees (20) Figure 3.7. XRD patterns measured for (a) alumina support and (b) CeA and (c) CeN catalysts. 97 In summary, XRD and XPS analyses indicate that the cerium oxide dispersion is similar for catalysts derived from nitrate (CeN) and alkoxide (CeA) cerium precursors. However, the CeN catalyst consists primarily of CeOz crystallites while most of the cerium oxide present in the CeA catalyst is amorphous [28]. Table 3.1. Particle Sizes of Cerium Oxide Species in Ce/A1203 Catalysts Determined from m Line Broadening and XPS Intensity Ratios Measurements. mm Samples l XRD XPS CeN 5.1 1.6 CeA 4.5 1.6 Photoreduction of Cerium/A lununa Catalysts. Figure 3.8 shows the Ce 3d spectra obtained for CeN and CeA catalysts. The XPS Ce 3d spectrum measured for the CeN catalyst shows Ce“ and Ce‘i’ features similar to those observed for the photoreduced Ce750 sample. For the CeA catalyst, the Ce 3d spectrum has v' and u' lines similar to those observed for the Ce(IIl) acetylacetonate hydrate standard compound as well as the u'" peak associated with Ce" species. The presence of Ce3‘ in calcined CC/Al203 catalysts is ofien attributed to the formation of bulk CeAlO; or a Ce-A1203 surface phase. We do not believe that the CeN and CeA catalysts examined in this study contain either of these species since they were prepared using Ce“ precursors and were calcined at a relatively low temperature. Shyu et 98 (b) r 1 1 920 900 880 Binding Energy (eV) Figure 3.8. Ce 3d XPS spectra measured for (a) CeN and (b) CeA catalysts. 99 al. [29] have reported that high temperature (1000 °C), long reaction times (80 h), and a reducing atmosphere were required to form bulk CeAlO3 in Ce/A1203 catalysts. Formation of a Ce3i—A1203 surface phase would seem to require a Ce“ precursor. Further, the relatively poor Ce dispersion noted for CeN and CeA catalysts is not consistent with the formation of a Ce-A1203 surface phase since surface phases are normally considered to be highly dispersed (particle size < 1 nm). Thus we believe that the Ce3+ species observed by XPS arise from cerium oxide photoreduction. The greater extent of photoreduction shown by the CeA catalyst can be attributed to the predominance of amorphous cerium oxide in the catalyst. 10. ll. 12. 13. I4. 15. 16. 17. 100 3.5. References Tranquada, J. M.; Heald, S. M.; Moodenbaugh, A. R.; Liang, G.; Croft, M. Nature 1989, 337, 720. Merchant, P.; Jacowitz, R. D.; Tibbs, K; Taber, R. C.; Landerrnan, S. S. Appl. Phys. Lett. 1992, 60, 763. Al-Robaee, M. S.; Rao, K. N.; Mohan, S. J. Appl. Phys. 1992, 7!, 2380. Al-than, Z. T.; Hogarth, C. A.; Riddleston, N. Phys. Stat. Sol. (b) 1988, 145, 145. Kleebe, H.-J.; Cinibulk, M. K. J. Mater. Sci. Lett. 1993, 12, 70. Butta, N.; Cinquegrani, L.; Mugno, E.; Tagliente, A.; Pizzini, S. Sensors and Actuators B, 1992, 6, 253. Su, E. C.; Montreuil, C. N.; Rothschild, W. G. Appl. Catal. 1985, 17, 75. Herz, R. K. Ind. Eng. Chem. Prod. Res. Dev. 1981, 20, 451. Yao, H. C.; Yu-Yao, Y. F. J. Catal. 1984, 86, 254. Le Normand, F.; Hilaire, L.; Kili, K.; Krill, G.; Maire, G. J. Phys. Chem. 1988, 92, 2561. Jin, T.; Okuhara, T.; Mains, G. J.; White, J. M. J. Phys. Chem. 1987, 91, 3310. Cho, B. K.; Shanks, B. H.; Bailey, J. E. J. Catal. 1989, 115, 486. Bak, K.; Hilaire, L. Appl. Surf Sci. 1993, 70/71, 191. Copperthwaite, R. G. Surf. Interface Anal. 1980, 2, l7. Storp, S. Spectrochim. Acta, B, 1985, 40, 745. Copperthwaite, R. G.; Lloyd, J. .1. Electron Spectrosc. Relat. Phenom. 1978, 14, 159. Batista-Leal, M.; Lester, J. E.; Lucchesi, C. A. J. Electron Spectrosc. Relat. Phenom. 1977, 11, 333. 18 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 101 Johnson, C. E. In Electronic States of Inorganic Compounds: New Experimental Techniques, Nato Advanced Study Institutes, University of Oxford, Sept. 1974; Day, P., Ed.; D. Reidel Publishing Company, Dordrecht-Holland, 1975; p 409- 424. E1 Fallah, J.; Hilaire, L.; Romeo, M.; Le Normand, F. J. Electron Spec. Rel. Phenom. 1995, 73, 89. Hirokawa, K.; Honda, F .; Oku, M. J. Electron Spectrosc. Relat. Phenom. 1975, 6, 333. Bumess, J. H.; Dillard, J. G.; Taylor, L. T. J. Amer. Chem. Soc. 1975, 97, 6080. Dauscher, A.; Hilaire, L.; Le Normand, F.; Muller, W.; Maire, G.; Vasquez, A. Surf Interface Anal. 1990, I6, 341. Al-than, Z. T.; Hashemi, T.; Hogarth, C. A. Spectrochimica Acta. 1989, 441), 205. Paparazzo, E. Surf Sci. 1990, 234, L253. Paparazzo, E.; Ingo. G. M.; Zacchetti, N. J. Vac. Sci. Techrrol. A. 1991, 9, 1416. Laachir, A.; Perrichon, V.; Badri, A.; Lamotte, J.; Catherine, E.; Lavalley, C.; El Fallah, J.; Hilaire, L.; Le Normand, F .:, Quéméré, E.; Sauvion, G. N.; Touret, O. J. Chem. Soc. Faraday Trans. 1991, 87, 1601. Askeland, P. A.; Ledford, J. S. to be published. Park, P. W.; Ledford, J. S. to be published. Shyu, J. 2.; Weber, W. 1-1.; Gandhi, H. S. J. Phys. Chem. 1988, 92, 4964. Graham, G. W.; Schmitz, P. J.; Usmen, R. K.; McCabe, R. W. Catal. Lett. 1993, 17,175. Powder Diffraction File, Inorganic Phases, JCPDS, 1983. Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline and Amorphous Materials, lst.ed.; Wiley: New York, 1954. XPS data manipulation performed using sofiware provided by Dr. Andrew Proctor, University of Pittsburgh, PA. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 102 Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269. Kerkhof, F. P. I. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612. Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129. Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29. Fujimori, A. Phys. Rev. B 1983, 28, 2282. Fujimori, A. Phys. Rev. B 1984, 53, 2518. Kotani, A.; Mizuta, H.; Jo, T.; Parlebas, J . C. Solid State Commun. 1985, 53, 805. Wuilloud, E.;Delley, B.; Schneider, W.-D.; Baer, Y. Phys. Rev. Lett. 1984, 53, 2519. Wuilloud, E; Delley, B.; Schneider, W.-D.; Baer, Y. Phys. Rev. Lett. 1984, 53, 202. JO, 1.; Kotani, A. Phys. Scrp. 1987, 35, 570. Nakano, T.; Kotani, A.; Parlebas, J. C. J. Phys. Soc. Jpn. 1987, 2201. Creaser, D. A.; Harrison, P. G.; Morris, M. A.; Wolfindale, B. A. Catal. Lett. 1994, 23, 13. Burroughs, P.; Hamnett, A.; Orchard, A. F; Thornton, G. J. Chem. Soc., Dalton Trans. 1976, 1686. Praline, G.; Koel, B. E.; Hance, R. L.; Lee, H.-I.; White, J. M J. Electron. Spectrosc. Relat. Phenom. 1980, 21, 17. Le Normand, F.; El Fallah, J.; Hilaire, L.; Legare, P.; Kotani, A.; Parlebas, J. C. Solide State Communications 1989, 7], 885. Wallbank, B.; Johnson, C. E.; Main, 1. G. J. Electron Spectrosc. Relat. Phenom. 1974, 4, 263. Chapter 4 The Influence of Surface Structure on the Catalytic Activity of Alumina Supported Copper Oxide Catalysts: Oxidation of Carbon Monoxide and Methane 4.1. Abstract X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD) and electron spin resonance (ESR) have been used to characterize a series of CU/A1203 catalysts. The information obtained from surface and bulk characterization has been correlated with CO and CH4 oxidation activity of the catalysts. For catalysts with Cu/Al atomic ratios 3 0.051, XPS data indicated that most of the Cu was present as a dispersed surface phase. ESR results showed that the ratio of isolated/interacting copper surface phase decreased with increasing Cu content. For catalysts with C u/Al atomic ratio 2 0.077, large CuO crystallites were detected by XRD. The turn over number for CO oxidation increased with increasing Cu content. This has been attributed to an increase in the amount of crystalline CuO present in the catalysts. The specific activity for CH4 oxidation decreased with increasing Cu content up to Cu/Al = 0.051. XPS results indicate that Cu dispersion decreased with increasing Cu content. Since the turn over number has been normalized by Cu dispersion obtained from XPS measurement, the decrease in CH4 103 Lfiiéial f0! 1h: 104 oxidation activity could not be attributed to a decrease in Cu dispersion. Instead, we propose that the isolated Cu surface phase is more active for CH4 oxidation than the interacting copper surface phase or crystalline CuO. CH4 oxidation activities were similar for the catalysts with Cu/Al atomic ratio 2 0.077. 4.2. Introduction Copper oxide is one of the most active transition metal oxide catalysts for emission control reactions [1-9]. Catalysts based on copper oxide are useful for the total oxidation of CO [5,6,10-12], hydrocarbons [2,5,6,13], chlorinated hydrocarbons [l4], and alcohols [15,16]. NO,i [7,17-19] and SO; reduction [19] reactions are also catalyzed by copper oxide. Kummer [2] has reported that copper oxide catalyst exhibits a CO oxidation activity per unit surface area similar to that of noble metal catalysts. Thus, supported copper oxide catalysts have been considered as suitable substitutes for noble metal-based emission control catalysts. The structure of Cu/A1303 catalysts depends on both the metal loading and calcination temperature [20-32]. For low copper loadings (< 8~10 wt.% Cu on 200m2/g y-Ale3 support), copper is present as a well-dispersed surface phase. Magnetic susceptibility and ESR measurements have been used to distinguish between two species in the copper surface phase: isolated and interacting copper species [20-25]. The isolated CUPric ions contribute to both magnetic susceptibility and resonance absorption, whereas the interacting copper species (or ‘clustered copper ions’) do not give a detectable ESR 105 signal. The c0pper surface spinel is believed to result either from the diffusion of copper ions into an alumina lattice [26] or from the interaction of a Cu2+ ion complex with alumina hydroxyl groups [28]. Strohmeier et al. [26] have reported that the chemical state of the spinel is different from both bulk CuO and CuA1204. Indeed, diffuse reflectance spectroscopy results have shown that cupric ions in the surface spinel predominantly occupy tetragonally distorted octahedral sites (> 90%) with only a small fraction located in tetrahedral sites [21,29,32]. *Bulk CuAle4 (60 % tetrahedral Cuzi and 40 % octahedral Cu”) can be formed only afier high temperature calcination (> 700 °C) [20,21,26]. For high copper loadings (> 10 wt.% Cu), crystalline CuO is formed on the alumina support [21,26]. The CO and CH4 oxidation activities of CU/N203 catalysts also depend on copper content. Mooi and Selwood [33] have reported that the CO oxidation activity is the highest for a 7 wt.% CuO/AIZO3 catalyst (supported on 200 mz/g y-AIZO3). Severino et al. [10] reported that a 5 wt.% CuO/A1303 catalyst (supported on 100 mZ/g y-Ale3) shows ,. - the maximum CO oxidation activity. The authors attributed the activity of the low loading catalysts to copper ions dispersed on the alumina surface. However, Pierron et al. [34] have reported that CO oxidation occurs on crystalline copper oxide phases. Tsikoza et al. [35] and Marion et al. [363 7] reported that the CH4 oxidation activity per unit mole or gram of copper decreases with increasing copper loading. Tsikoza et al. [35] suggested that CH4 oxidation proceeds primarily on the well-dispersed copper surface phase. Marion et al. [36,37] also suggested that the catalytic activity for methane oxidation depends on the amount of well-dispersed copper ions. In addition, these authors a? tutu 106 attributed the decrease in CH4 oxidation activity with increasing Cu loading primarily to a decrease in copper oxide dispersion. Several previous studies of Cu/A1203 catalysts have focused on the structure or activity of the catalyst. Little effort has been devoted to investigating systematically the relationship between the surface structure and CO and CH4 oxidation activity of Cu/Ale3 catalysts. The present work is part of a broad study to investigate structure-reactivity correlations for transition metal oxide based emission control catalysts. In this paper, X- ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), and electron spin resonance (ESR) have been used to determine the effect of Cu loading on the chemical state and dispersion of copper species supported on y-alumina. The information derived from these techniques is correlated with the CO and CH4 oxidation activity in order to develop a more complete understanding of Cu/Al203 catalysts. 4.3. Experimental Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y- alumina (Cyanamid, surface area = 203 mz/g, pore volume = 0.6 mL/g) using solutions of copper(II) nitrate (Columbus Chemical Industries Inc, ACS Grade). The alumina was finely ground (<230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation. The impregnated samples were dried in air at 120 °C for 24 h and calcined in air at 500 °C for 16 h. The copper content of the CU/Aleg, series was varied from a Cu/Al atomic 1'3 ti o of 0 to 0.128 (0 to 20 wt.% CuO). Cu/A1203 catalysts will be designated by “Cuy”, respec match Quam outga meaSL 05, COME Opera 0ft»: 35p0 107 where y is the Cu/Al atomic ratio (x102) of the catalysts. The catalyst colors change with increasing copper content from pale blue (Cu0.3 and Cu0.6), to pale green (Cu1.3 and Cu2.6), to green (Cu5.1), to dark green (Cu7.7), to gray (Cu10), to dark gray (Cul3). Standard Materials. CuO was prepared by calcining copper(II) nitrate in air at 500 °C for 16 h. CuA1204 was prepared by calcining stoichiometric amounts of the respective nitrates in air at 1000 °C for 24 h. XRD patterns of the standard compounds matched the appropriate Powder Diffraction File [3 8]. BET Surface Area. Surface area measurements were performed using a QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 g of catalyst were outgassed in a Nz/He mixture (5% N2) at 350 °C for 1 h prior to adsorption measurements. The measurements were made using relative pressures of N; to He of 0.05, 0.08, and 0.15 (N; surface area = 0.162 nmz) at 77 K. Increasing the copper content of the catalysts decreases the BET surface area by a maximum of 20 %. X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a Rigaku XRD diffractometer employing Cu K, radiation (7x = 1.5418 A). The X-ray was operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scanning rate of 0.5 deg/min (in 29) with divergence slit and scatter slit widths of 1°. Samples were run as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity. The mean crystallite sizes (3) of the CuO particles were determined from XRD line broadening measurements using the Scherrer equation [39]: Z=Kmpcose (1) there iii wi Cu0< measr app: the A 1mm [FLA ] instn Spec 108 where it is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the full width at half maximum (fwhm), in radians, ofthe Cuo line. Quantitative X-ray diffraction data were obtained by comparing CuO/A1203<400> intensity ratios measured for catalyst samples with intensity ratios measured for physical mixtures of pure CuO and y—A1203. This method assumed that copper addition did not disrupt the y-A1203 spinel and subsequently affect the intensity of the A1203<400> line. The error in this method was estimated to be i 20 %. XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science instrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20 mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The instrument typically operates at pressures near 1 x 10’8 Torr in the analysis chamber. Spectra were collected using a PC 137 board interfaced to a Zeos 386SX computer. XPS spectra were also obtained using a VG Microtech spectrometer to confirm Cu 2pm binding energies. The VG spectrometer is equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20 mA) and a Clam2 hemispherical analyzer operated with a pass energy of 50 eV. The instrument typically operates at pressures below 1 x 10'8 Torr in the analysis chamber. Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the catalyst samples and standard compounds which contained Al Were referenced to the Al 2p peak (74.5 eV). The binding energy for CuO was referenced to the C Is line (284.6 eV) of the carbon overlayer. Detector efficiency was corrected 109 using Cu 2p and Cu 3p peak intensities measured for copper metal foil which was cleaned with Ar ion bombardment. Since the Cu 3p line overlaps the Al 2p line for the catalyst samples, estimated Cu 3p intensities (from the ratio of Cu 3p to Cu 2p) were subtracted from Al 2p intensities prior to calculating the Cu/Al intensity ratio. XPS binding energies were measured with a precision of i 0.2 eV, or better. Reduction of copper species during XPS experiments has been reported [26,40,41] and attributed to several factors such as X-ray flux, X-ray dose, temperature, and pressure. In order to minimize the effect of photoreduction on the results, all samples were analyzed using the same distance between the X-ray source and sample (~ 2 cm) and minimum data acquisition time (5 min). No significant photoreduction was observed using these experimental conditions. Quantitative XPS Analysis. It has been shown by Defosse et al. [42] that one may calculate the theoretical intensity ratio I°,,/I°s expected for a supported phase (p) atomically dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and Moulijn [43] has been used in the present investigation. The photoelectron cross sections and the mean escape depths of the photoelectrons used in these calculations were taken from Scofteld [44] and Penn [45], respectively. For a phase (p) present as discrete particles, the experimental intensity ratio l,,/1s is given by the following expression: U], = 1°p/I°,[l-exp(-d/?tp)]/ cm, (2) 110 where I°P/I°. is the theoretical monolayer intensity ratio, d is the length of the edge of the cubic crystallites of the deposited phase and A? is the mean escape depth of the photoelectrons in the deposited phase. The Cu species particle sizes were determined by solving the Eq.(2) for d. The relative Cu dispersion to the monolayer was estimated from the ratio of VI. to 1°./1°, ESR Study. ESR spectra were obtained using a Varian E-4 spectrometer with a rectangular TE 10; cavity operated at a microwave frequency of 9.06 GHz and a power of 20 mW. Ruby crystal mounted inside the cavity at fixed orientation was used as an intensity standard for determination of relative ESR intensities. The relative intensities per unit mg CuO were obtained from the ratio of the peak intensity of catalyst sample to the value of Ruby crystal. Powder samples were dried at 120 °C for 24 hr prior and run at 98 K in a quartz ESR tube. C 0 Oxidation Activity. Measurement of CO oxidation activity was performed in a flow microreactor. Approximately 0.03 g of catalyst were supported on a glass frit (70 - 100 um) and the temperature was measured with a K-type thermocouple located just above the catalyst bed. The reactor was heated by a tube furnace (Lindberg) with temperature being controlled within 1 °C by an Omega CN 1200 temperature controller. Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers. Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated on a 6 fi 60/80 mesh Carbosieve S-Il column. Prior to the first activity measurement, catalysts were pretreated with a mixture of 5% Oz/He (99.5 % purity for 02, 99.995 % pumy impuri reactic coast: punt} were late trap r (C0 than mole XPS 111m loadi 111 purity for He, AGA Gas Co.) stream (143 cc/min) at 350 °C for 1 hr to remove any impurities adsorbed on the surface during catalyst preparation and storage. CO oxidation reactions were performed with a constant flow rate (80 cm3/min) of 4.8% CO/9.8°/o 07/85.4% He gas mixture (AGA, purity > 99.99 %) in the temperature range of 150-400 oC_ CH4 Oxidation Activity. Methane oxidation reactions were performed with a constant flow (15 cm3/min) of 0.98% CH4/5.25% O;/93.77% He gas mixture (AGA, purity > 99.99 %) in the temperature range of 360-430 °C. Approximately 0.1 g catalyst were charged into the same type of microreactor used for CO oxidation measurements. Water produced during methane oxidation was frozen downstream from the reactor in a trap maintained < -40 °C with a mixture of ethanol and dry ice. All activity measurements (CO and CH4 oxidation) were obtained under steady-state conditions at conversions less than 15 %. The turn over number (TON) was expressed as mole CO or CH; oxidized per mole of surface-bound c0pper. The surface bound copper content was determined by XPS from the relative intensity ratios of Cu 2pm and Al 2p peaks. 4.4. Results State and Dispersion of Copper. XRD patterns obtained for Cuy catalysts with Clat/Al atomic ratios 3 0.051 (s 8 wt.% CuO) showed only lines characteristic of the alumina carrier (Figure 4.1a). XRD patterns obtained for catalysts with higher Cu 10adings (Cu/Al atomic ratios 2 0.077) contain peaks characteristic of CuO (Figure 4.1b- 112 CuO Cuo <111> CuO C110 <110> <262> (d) w (b) AIZO, < > 3 l 1 A120: A120, . ‘ <222> <220> (a) A120, <400> L l l l l l 25 30 35 40 45 50 Degrees (20) l:igure 4.1. XRD patterns measured for (a) A1203, (b) Cu7.7, (c) CulO, and (d) Cu13. 113 Table 4.1. Concentration of Crystalline‘ Phases in Cuy Catalysts Calculated from Quantitative XRD Data. Catalysts Crystalline Phase (wt.%) As prepared Measured Cu 0.3 0.5 0 Cu 0.6 1 0 Cu 1.3 2 0 Cu 2.6 4 0 Cu 5.1 8 0 Cu 7.7 12 5.7 Cu 10 16 10.8 Cu 13 20 14.6 a- Valid for crystalline phases with particle sizes > 3.0 nm. d)- The intensities of the CuO XRD peaks increase with increasing Cu content. Quantitative XRD measurements (Table 4.1) show that the amount of crystalline CuO increases from 5.7 to 14.6 wt.% as the Cu/Al atomic ratio increases from 0.077 to 0.13. The XPS Cu 2133/2 binding energies measured for the Guy catalysts (935.1 i 0.2 eV) were independent of Cu content and similar to the Cu 2pm binding energy measured for CuAhOa (935.0 eV) and higher than the value measured for CuO (933.9 eV). The SEltellite/main peak ratio measured for the Cuy catalysts (0.47 i 0.1) is similar to that measured for CuO (0.45) and lower than the value for CuAlea (0.71). Figure 4.2 shows the variation of the Cu 2p3;2/Al 2p intensity ratio measured for the Cuy catalysts as a fiJnction of Cu/Al atomic ratio. The theoretical line calculated for monolayer dispersion [43] is shown for comparison. The Cu/Al intensity ratio increases with increasing Cu/Al atomic ratio up to 0.051. For higher Cu loadings, the Cu/Al Intensity ratio measured for the catalysts levels off. Variation in the particle size of the XPS Intensity Ratio (Cu2p3/2/A12p) I“igure 4.2. 0 114 monolayer line 0 catalysts .. O r 1 J W I I l U I I I I l U r I 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Atomic Ratio (Cu/A1) Cu 2P3/2/Al 2p XPS intensity ratios measured for Cuy catalysts(0) plotted versus Cu/Al atomic ratio. Cu 2p3n/Al 2p intensity ratios calculated for monolayer dispersion (line). 115 Table 4.2. Particle Size of Copper Phases Determined from Cu/Al XPS Intensity Ratios and XRD Line Broadening Calculations. Catalysts Particle size (nm) XPS XRD Cu 0.3 0.4 -‘ Cu 0.6 0.6 -' Cu 1.3 0.8 -‘ Cu 2.6 0.8 -' Cu 5.] 0.9 -" Cu 7.7 1.7 30 Cu 10 2.3 30 Cu 13 3.0 29 a- No CuO XRD peaks detected for these catalysts copper species determined from Cu/Al XPS intensity ratios using Eq. (2) is given as a function of the Cu loadings in Table 4.2. The mean particle size of the copper species Calculated from XPS intensity ratios increases from 0.4 to 3.0 nm with increasing Cu/Al atomic ratio. Particle sizes determined from XRD line broadening calculations using Eq. ( 1 ) are also shown in Table 4.2. For catalysts with Cu/Al atomic ratios 2 0.077, XRD line broadening calculations show that the CuO crystallite sizes are independent of Cu loading and significantly larger than the values determined using XPS. Figure 4.3 shows the variation of the relative ESR signal intensity measured for the Guy catalysts as a function of Cu/Al atomic ratio. For catalysts with Cu/Al atomic ratios 3 0- 051, relative ESR signal intensity decreases significantly with increasing Cu content. I“Iowever, for catalysts with higher Cu/Al atomic ratios, the relative ESR signal intensities are independent of Cu content. 116 18 16 _. A 'e u- _fi/ 14 / .~ L m c: A 3 12 .5 G4 m "__i LU 10 lllllll B m 16% 230a“? 4000 48m -521 v-g 8 0 Z 6- 4- 2- U U 0.1.1.1.111oi. 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Atomic Ratio (Cu/A1) I:igure 4.3. Normalized ESR intensity (1 mg CuO) of Cuy catalysts relative to the internal standard. 117 Table 4.3. Turn Over Numbers‘ and Activation Energies” of Cu/A1203 Catalysts for CO and CHa Oxidations. CO Oxidation CH4 Oxidation Catalysts Activation Energy TON (x104) Activation Energy TON (x104) Cu 0.3 25.8 3.5 21.8 11.8 Cu 0.6 25.7 6.8 22.1 8.3 Cu 1.3 19.2 33.6 22.6 6.5 Cu 2.6 19.4 76.0 23.4 4.7 Cu 5.1 19.2 242 21.9 2.8 Cu 7.7 13.1 390 22.3 3.2 Cu 10 13.7 694 23.6 2.9 Cu 13 13.2 1010 21.4 2.9 a- Estimated from Arrhenius plots (kcal/mol) within 15 "/0 conversion. b- Calculated at 260 and 390 0C for CO and CH4 oxidation respectively (5") using exposed Cu species determined from XPS. CO Oxidation Activity. Table 4.3 shows the specific turn over number (TON) for CO oxidation over Cuy catalysts calculated at 260 °C using XPS estimates of the Cu dispersion. The TON increases by nearly a factor of 300 as the Cu/Al atomic ratio increases from 0.003 to 0.13. The activation energies calculated for CO oxidation are also SFlown as a function of Cu content in Table 4.3. The activation energy decreases from 25-8 to 13.1 kcal/mol as the Cu/Al atomic ratio increases from 0.003 to 0.077. For cEltrdlysts with Cu/Al atomic ratios 2 0.077, the activation energies are identical within experimental error (13.3 i 0.3 kcal/mol). CH4 Oxidation Activity. Table 4.3 shows the TON for CH4 oxidation over Cuy canalysts calculated at 390 °C using XPS estimates of the Cu dispersion. The TON decreases by nearly a factor of 4 as the Cu/Al atomic ratio increases from 0.003 to 0.051. I:llt‘ther addition of Cu has little effect on the TON for CH4 oxidation. Table 4.3 also 118 shows that the activation energies calculated for CH4 oxidation are similar within experimental error (22.4 :t 1.0 kcal/mol). For catalysts with Cu/Al atomic ratio 3 0.006, the CO; selectivity measured at 390 °C increased from 63 % to 80 % with increasing copper content. No partial oxidation product was observed for catalysts with Cu/Al atomic ratio 2 0.013. 4.5. Discussion Structure of Cooper/Alumina Catalysts. For catalysts with Cu/Al atomic ratio 3 O - 051, the absence of XRD peaks characteristic of discrete CuO and the small particle size calculated from Cu/Al XPS intensity ratios suggest that copper exists as a well-dispersed COpper surface phase. These results are consistent with previous work by Friedman et a1. [2 l ], which showed that the support capacity of several y-aluminas for copper surface phase is approximately 4 wt.%Cu/lOOmZ/g support. Strohmeier et al. [26] also reported tl'lat 9-10 wt.% copper (for 195 mZ/g y-alumina) is the maximum copper loading that can be accommodated as a surface phase. The Cu 2pm XPS spectra measured for the Cuy catalysts show Cu 2pm binding erlergies characteristic of CuAlea and satellite/main peak intensity ratios close to the Value obtained for CuO. It is well known [21,26] that Cu/A1203 catalysts contain Cu2+ SL1 rface phase and discrete CuO particles. In principle, photoelectrons from both discrete CUO particles and Cuzi surface phase will contribute to the XPS signal. However, due to 119 the high dispersion of the surface phase relative to the CuO crystallites, the qualitative features of the Cu 2pm XPS spectra are expected to be more representative of the surface phase. Consequently, the observation of satellite/main peak intensity ratios typical of CuO is surprising. However, this apparent discrepancy has been reported by previous researchers [26,30]. Strohmeier et al. [26] suggested that this did not necessarily mean that well-dispersed or amorphous CuO was present on the catalysts, but simply that the copper surface species is chemically different from bulk CuO and CuA1204. The significant decrease in the relative ESR intensity as a function of Cu content is consistent with the result from previous work [21,22]. The ESR result indicates that isolated/interacting copper surface phase ratio decreases as Cu loading is increased in the range of 0 < Cu/Al 3 0.051. The former contributes the resonance absorption, which can be assigned as tetragonally distorted, octahedrally-coordinated cupric ions [23]. The latter contributes less ESR signal due to interaction with neighboring cupric ions [22]. The decrease in Cu dispersion for catalysts with Cu/Al g 0.051 can be attributed to the formation of interacting copper surface phase or copper oxide clusters. For Cuy catalysts with Cu/Al atomic ratio 2 0.077, the increase in XPS determined COpper oxide particle size with increasing Cu loading is consistent with the appearance of X~ray diffraction lines characteristic of CuO. These results are similar to previous work [2 1,26] which indicated that at a critical metal loading, the support becomes saturated with copper surface phase and further addition of the metal results in formation of CuO crystallites. The difference in CuO particle sizes determined using XPS and XRD may be attributed to the limitations of X-ray diffraction. It is well known that XRD determined 120 particle sizes are skewed to large values since XRD does not detect highly dispersed species. For the Cu rich catalysts, a significant fraction of the copper is present as a surface phase or small CuO particles (d < 3.0 nm) that are detected by XPS, but not XRD. Thus, smaller average particle sizes are expected from XPS calculations. Effect of Catalyst Structure on C0 Oxidation Activity. The significant increase in CO oxidation TON as a function of Cu content observed for the Cuy catalysts suggests that CuO is the active phase for CO oxidation. In addition, when one considers that the crystalline CuO is poorly dispersed compared to the copper surface phase, it is clear that CuO is significantly more reactive than the copper surface phase. The increase in CO oxidation activity with increasing Cu loading observed in this study is not consistent with previous studies of CO oxidation over alumina supported copper oxide catalysts. The optimum copper contents have been reported as approximately 5 wt.% Cu/95 m2/g y- A1203 [10] and 7 wt.% Cu on 200 mZ/g y-Ale3 [33]. For the CO oxidation over alumina supported copper oxide catalysts, it is generally considered that CO oxidation involves redox phenomena on the surface of copper oxide. Jernigan and Somorjai [46] have recently proposed that CO oxidation proceeds by a redox mechanism involving CuO and CuzO and that the reduction of CuO by CO is the rate determining step. Since the oxidation reaction may require adsorption of CO at oxygen vacancies, the availability of oxygen vacancies is important for copper oxide based CO oxidation catalysts [11]. On the basis of CO TPR results, Severino et a1. [10] have reported that the CuO crystalline phase is reduced at a lower temperature than the copper surface phase. Therefore, the high CO oxidation activity of the CuO phase can be 121 attributed to more facile generation of oxygen vacancies during the CO oxidation reaction compared to the copper surface phase. Effect of Catalyst Structure on CH4 Oxidation Activity. The decrease in CH4 oxidation activity with increasing Cu loading has been reported in previous studies [35,36]. Tsikoza et al. [35] suggested that CH4 oxidation occurred primarily on the copper surface phase. In addition, Garbowski and Primet [28] have proposed that octahedrally coordinated surface copper ions are accessible to reactants involved in catalytic combustion reactions. Marion et al. [36] have reported that CH4 oxidation activity for CuO/Ale3 catalysts is proportional to the number of surface copper species and that the number of active centers decreased with increasing copper loading due to the decrease in Cu dispersion. In this study, the decrease in CH4 oxidation activity for Cuy catalysts cannot be attributed to a change in Cu dispersion, since the turn over number was obtained by normalization with Cu dispersion. Instead, we propose that the isolated Cu2+ species are more reactive for CH4 oxidation than the interacting copper surface species. For the high loading catalysts (Cu/Al 2 0.077), the CH4 oxidation activity is independent of Cu content, which can be ascribed to the formation of poorly dispersed CuO crystallites. The mechanism of CH4 oxidation differs in some respects from C0 oxidation. First, the rate determining step of CH4 oxidation involves C-H bond breakage [47]. Second, oxidation of CH4 produces water which may poison catalysts [11]. Therefore CH4 oxidation requires a higher reaction temperature, and thus the influence of oxygen vacancies will be less important due to high mobility of lattice oxygen [13]. Liu and 122 Flytzani-Stephanopoulos [48] have proposed that acidic and basic sites on the catalyst surface play an important role for C-H bond breakage and formation of intermediate species. Isolated copper species may provide acidic/basic sites with more suitable strength and geometry for CH; oxidation than interacting copper species or crystalline CuO. CH4 TPR studies have been performed with CuO/A1203 catalysts. Kartheuser et al. [49] reported a broad, featureless peak in the range of 500-700 °C for a 5 wt.% CuO/325m2/g A1203 catalyst. They assigned this feature to the reduction of CuO or Cu- A1203. H2 TPR results for Cu/A1203 catalysts showed that the reduction of copper surface phase occurs at a lower temperature than for CuO [37,50]. These H2 TPR results are usefiil for understanding the high CH4 oxidation activity of isolated copper species since both H2 and CH4 utilize hydrogen to reduce copper species and produce water. 4.6. Conclusions The combined use of several techniques to investigate the effect of Cu loading on the structure and catalytic activity of Cu/A1203 catalysts for CO and CH4 oxidation leads to the following conclusions. 1. For low Cu loadings (Cu/Al 5 0.051), copper exists as a dispersed copper surface phase. ESR results indicate that the fraction of interacting Cu phase increases with increasing Cu content. For Cu rich catalysts (Cu/A1 2 0.077), large CuO crystallites are formed. 123 2. The increase in CO oxidation activity observed with increasing Cu content can be attributed to an increase in concentration of crystalline CuO. The copper surface phase shows very low CO oxidation activity. This has been explained with the ease of CuO redox behavior compared to copper surface phase during the CO oxidation reaction. 3. The CH4 oxidation activity decreases with increasing Cu loading. We propose that the isolated copper surface phase is more active for CH; oxidation than the interacting copper surface phase or crystalline CuO. For higher Cu loadings, the CH4 oxidation activity is independent of Cu loading, which can be attributed to the formation of poorly dispersed CuO crystallites. 10. 11. 12. 13. 14. 15. 16. 17. 18. 124 4.7. References Prasad, R., Kennedy, L. A. and Ruckenstein, E. Catal. Rev. -Sci. Eng. 1984, 26, l. Kummer J. T., Prog. Energy Combust. Sci. 1980, 6, 177. Dwyer, F. G. Catal. Rev-Sci. Eng. 1972, 6, 261. Zwinkels, M. F. M.; Jarés, S. G.; Menon, P. G. Catal. Rev. -Sci. Eng. 1993, 35 319. Yao, Y. F. Yu; Kummer, J. T. J. Catal. 1977, 46, 388. Yao, Y. F. Yu J. Catal. 1975, 39, 104. Kapteijn, F.; Stegenga, S.; Dekker, N. J. J.; Bijsterbosch, J. W.; Moulijn, J. A. Catal. Today 1993, I6, 273. Klimisch, R. L. General Motors Res. Publ. GMR-842 (1968). Heyes, C. J .; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol. 1982, 32, 1025. Severino, F.; Brito, J.; Cari’as, O. Laine, J. J. Catal. 1986, 102, 172. Boon, A. O. M.; van Looij, Fr, Geus, J. W. J. Mol. Catal. 1992, 75, 277. Lopez Agudo, A.; Palacios, J. M.; Fierro, J.L.G.; Laine, J.; Severino, F. Appl. Catal. 1992, 91,43 . Boon, A. O. M.; Huisman, H. M.; Geus, J. W. J. Mol. Catal. 1992, 75, 293. Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. Technol. 1988, 22, 557. Ozkan, U. S.; Kueller, R. F.; Moctezuma, E. Ind. Eng. Chem. Res. 1990, 29, 1136. Rajesh, H.; Ozkan, U. S. Ind. Eng. Chem. Res. 1993, 32, 1622. London, J. W.; Bell, A. T. J Catal. 1973, 31, 96. Huang, T.-J.; Yu, T.-C. Appl. Catal. 1991, 7], 275. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 125 Goetz, V. N.; Sood, A.; Kittrell, J. R. Ind. Eng. Chem. Prod. Res. Develop. 1974, I3, 110. Wolberg, A.; Roth, J. F. J. Catal. 1969, 15, 250. Friedman, R. M.; Freeman, J. J.; Lytle, F. W. J. Catal. 1978, 55, 10. Matsunaga, Y. Bull. Chem. Soc. Jpn. 1961, 34, 1291. Berger, P. A.; Roth, J. F. J. Phys. Chem. 1967, 71, 4307. Lumbeck, H.; Voitlander, J. J. Catal. 1969, 13, 117. Centi, G.; Perathoner, S.; Biglino, D.; and Giamello, E. J. Catal. 1995, 151, 75. Strohmeier, B. R.; Leyden, D. E; Field, R. S.; Hercules, D. M.; Petrakis, L. J. Catal. 1985, 94, 514. Ertl, G.; Hierl, R.; Knozinger, H.; Thiele, N.; Urbach, H.-P., Appl. Surf Sci. 1980, 5, 49. Garbowski, E.; Primet, M. J. Chem. Soc. Chem. Commun. 1991, 11. Lo Jacono, M.; Cimino, A.; Inversi, M. J. Catal. 1982, 76, 320. Tikhov, S. F.‘, Sadykov, V. A.; Kryukova, G. N.; Paukshtis, E. A.; Popovskii, V. V.; Starostina, T. G.; Kharlamov, G. V.; Anufrienko, V. F.; Poluboyarov, V. F.; Razdobarov, V. A.; Bulgakov, N. N.; Kalinkin, A. V. J. Catal. 1992, I34, 506. Summers, J. C. and Klimisch, R. L. Proc. Int. Cong): Catal. 51h 1972 vol 2, 293. Freeman, J. J .; Friedman, R. M. J. Chem. Soc. Faraday Trans. 1. 1978, 758. Mooi, J.; Selwood, P. W. J. Am. Chem. Soc. 1952, 74, 2461. Pierron, E. D.; Rashkin, J. A.; Roth, J. F. J. Catal. 1967, 9 38. Tsikoza, L. T.; Torasova, D. V.; Ketchik, S. V.; Maksimov, N. G.; Popovskii, V. V. Kinet. Katal, 1981, 22, 1022. Marion, M. C; Garbowski, E.; Primet, M. J. Chem. Soc, Faraday Trans. 1990, 86, 3027. 37. 38. 39. 40. 41. 42. 43. 44. 45.‘ 46. 47 48. 49. 50. 126 Marion, M. C.; Garbowski, E.; Primet, M. J. Chem. Soc, Faraday Trans. 1991, 87, 1795. Powder Diffraction File, Inorganic Phases, JCPDS, 1983. Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline and Amorphous Materials, lst.Ed. Wiley, New York, 1954. Rosencwaig, A. and Wertheim, G. K. J. Electron. Spectrosc. Relat. Phenom. 1972, l, 493. Wallbank, B., Johnson, C. E., and Main, 1. G. J. Electron. Spectrosc. Relat. Phenom. 1974, 4, 263. Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269. Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612. Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129. Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29. Jernigan, G. G.; Somorjai, G. A. J. Catal. 1994, 147, 567. Otto, K. Langmuir 1989, 5, 1364. Liu W. and Flytzani-Stephanopoulos, M. J. Catal. 1995, 153, 304. Kartheuser, B.; Hodnett, B. K.; Riva, A.; Centi, G.; Matralis, H.; Ruwet, M.; Grange, P.; Passarini, N. IndEng. Chem. Res. 1991, 30, 2105. Dumas, J. M.; Geron, C.; Kribii, A.; Barbier, J. Appl. Catal. 1989, 47, L9. Chapter 5 Characterization and CH4 Oxidation Activity of Cr/A1203 Catalysts 5.1. Abstract X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been used to characterize a series of Cr/Ale3 catalysts (designated “Cry”, y = Cr/AJ atomic ratio). The information obtained from surface and bulk characterization has been correlated with CH4 oxidation activity of the Cry catalysts. For catalysts with a Cr/Al atomic ratio y = 0.013, XPS results indicated that most of the Cr was present as a highly dispersed Cr6+ surface phase. Catalysts with intermediate Cr loadings (y = 0.027 ~ 0.080) showed no XRD peak characteristic of chromium oxide. However, XPS data indicated that the Cr dispersion decreased and the concentration of Cr3+ species increased with increasingCr content. For catalysts with high Cr loadings (Cr/Al atomic ratio 2 0.107), large Cr203 crystallites were detected by XRD. The specific activity for CH4 oxidation increased with increasing Cr content up to Cr/Al = 0.107. This has been attributed to an increase in the amount of Cr(IIl)-Cr(VI) cluster present in the catalysts. A decrease in CH.. oxidation activity observed for a Cr rich catalyst (Cr/Al = 0.13) has been ascribed to the formation of poorly dispersed Cr203. 127 128 5.2. Introduction Alumina supported chromium oxides have been studied extensively due to their potential application in emission control catalysis [1-3]. Desirable chemical and physical properties, such as low temperature efficiency for hydrocarbon [4-6] and chlorinated hydrocarbon oxidation [7-10], low selectivity for C12 formation [8,11], high resistance to HCl poisoning [12] during chlorinated hydrocarbon oxidation, NO reduction with CO [13] and NH3 [14], high thermal stability [6,15] and mechanical durability, [12] make Cr/A1203 catalysts suitable candidates for automotive emission control and industrial waste treatment. Chromium oxide is also an effective promoter for copper oxide catalysts used for CO and hydrocarbon oxidation and NO reduction [13,16,17]. The surface structure of alumina supported chromium oxide catalyst depends on the metal loading, calcination temperature and hydration/dehydration conditions [18-26]. For low chromium loadings (s 10 wt.% Cr203/Al203 200 mZ/g), most of the chromium is present as a highly dispersed Cr(Vl) surface species. XPS studies indicate that the proportion of Cr(VI) on the alumina support decreases with increasing Cr content [18,19,21] or with increasing calcination temperature [21,27]. This has been attributed to the formation of Cr(IIl) species on the support surface [19,21,27]. Ellison and coworkers [28-32] studied Cr/Al2O3 catalysts using magnetic susceptibility, DRS, SIMS and ESR techniques. They proposed that a mixed oxidation state cluster (-Cr6i-02'-Cr3+-Oz'-Cr6+-; y-phase) forms on the alumina support even at 1 wt.% Cr on 157 mz/g A1203. Wachs and 129 coworkers [22,33] reported the monolayer coverage of chromium oxide (12 wt.% CrO3/Al2O3 180 mZ/g) based on Raman, XRD, and FTIR data. This is in good agreement with the theoretical monolayer coverage calculated from the cross section of K2Cr2O7 [34] and ISS results [35]. For high chromium loadings (> 10 wt.% Cr203/Al2O3 200 mZ/g), crystalline 020:. is observed using Raman spectroscopy and X-ray diffraction. Bulk Cr(III)-Al2O3 forms after high temperature calcination (2 800 °C) of the Cr/Al2O3 containing crystalline Cr2O3 [22]. Despite the importance of Cr/Al203. catalysts in emission control applications, the study of CH; oxidation over Cr/Al2O3 catalysts has been relatively limited [36,37]. Little effort has been devoted to investigating systematically the relationship between the surface structure and the CH4 oxidation activity of Cr/AJ2O3 catalysts. Anderson et a1. [36] have studied CH4 oxidation over various alumina supported transition metal oxide catalysts and reported that chromium oxide shows the highest methane oxidation activity. Kuznetsova et al. [37] have reported that CH4 oxidation activity increases and levels off with increasing Cr content. They proposed that Cr(VI) is the main active component for the CH4 oxidation reaction. The present work is part of a broad study to investigate structure-reactivity correlation for transition metal oxide based emission control catalysts. In this paper, X- ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been used to determine the effect of Cr loading on the chemical state and dispersion of chromium oxide Phases supported on y-alumina. The information derived from these techniques is 130 correlated with CH. oxidation activity to develop a more complete understanding of Cr/Al2O3 catalysts. 5.3. Experimental Catalyst Preparation. The Cr-modified alumina carriers were prepared by pore volume impregnation of y-alumina (Cyanamid, surface area = 203 mZ/g, pore volume = 0.6 mL/g) using solutions of chromium (III) nitrate (Mallinckrodt, Analytical Reagent). The alumina was finely ground (<230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation. The impregnated samples were dried at 120 °C for 24 h and calcined at 500 °C for 16 h. The chromium content of the Cr/Al2O3 series was varied from a Cr/Al atomic ratio of O to 0.134 (0 to 20 wt.% Cr2O3). Cr/A1203 catalysts will be designated by "Cry", where y is the Cr/Al atomic ratio (x 102) of the catalysts. The catalyst colors change with increasing chromium content from pale yellow (Cr1.3), to yellow (Cr2.7), to dark yellow (Cr5.4), to brown (Cr8.0), to dark brown (Crl 1 and Crl3). Standard Materials. Cr203 was prepared by calcining chromium (III) nitrate at 500 °C for 16 h in air. CrO_~. (99.9 %) was obtained from Aldrich Inc. XRD patterns of the standard compounds matched the appropriate Powder Diffraction File [3 8]. BET Surface Area. Surface area measurements were performed using a QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 grams of catalyst Were outgassed in a N2/He mixture (5% N2) at 350 °C for l h prior to adsorption measurements. The measurements were made using relative pressures of N2 to He of 131 0.05, 0.08, and 0.15 (N2 surface area = 0.162 nmz) at 77 K. Increasing the chromium content of the catalysts decreases the BET surface area by a maximum of 10%. X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a Rigaku XRD diffractometer employing Cu K.ll radiation (it = 1.5418 A). The X-ray was operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scanning rate of 0.5 deg/min (in 20) with divergence slit and scatter slit widths of 1°. Samples were run as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity. The mean crystallite sizes (El) of the Cr2O3 particles were determined from XRD line broadening measurements using the Scherrer equation [39]: d = K)» / [3 cos 9 (l) where A is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and [3 is the full width at half maximum (fwhm), in radians, of the Cr2O3<104> line. Quantitative X-ray diffraction data were obtained by comparing Cr2O3<104>/AJ2O3<400> intensity ratios measured for catalyst samples with intensity ratios measured for physical mixtures of pure Cr2O3, and y-Al2O3. This method assumed that chromium addition did not disrupt the y-Al2O3 spinel and subsequently affect the intensity of the A1203<400> line. The error in this method was estimated to be i 20%. XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science il’lStrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20 mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The Insttument typically operates at pressures near 1 x 10'8 torr in the analysis chamber. rep pre WEI min usir Sign; p051 obta La) 9) lists film 53131} 132 Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer. Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the catalyst samples were referenced to the A1 2p peak (74.5 eV). The binding energies for standard compounds that did not contain Al were referenced to the C ls line (284.6 eV) of the carbon overlayer. XPS binding energies were measured with a precision of i 0.2 eV, or better. Reduction of chromium species [18,27,40] during XPS experiments has been reported and attributed to several factors such as X-ray flux, X-ray dose, temperature, and pressure. In order to minimize the effect of photoreduction on the results, all samples were analyzed using the same distance between X-ray source and sample (~3/4 inch) and minimum data acquisition time (< 20 min.) No significant photoreduction was observed using these experimental conditions. The oxidation state distribution of chromium was determined by non-linear least- squares curve fitting (NLLSCF) using the Cr2p3/2 envelope [41]. Chromium peak positions were allowed to float slightly (i 0.1 eV) in order to fit the Cr2p3/2 envelope obtained for each sample. The full width at half maximum (fwhm) of Cr(III) was fixed at 3 -3 eV (determined using Cr203 as a reference compound). Since the fwhm of Cr(V1) measured for Cry catalysts did not agree with the value obtained for the 003 standard Compound (2.] eV) and increased with decreasing Cr content, the fwhm of Cr(VI) was allowed to float near 3.0 eV (i 0.3 eV). The fwhm of standard compounds and Cry catalysts agree well with the values reported [42]. Scierka et al. [42] have examined the ta. dis an; fret par ilou 100 ll abO't'e 133 Cr2p envelope of Cr/Al2O3 catalysts with principal component analysis and target testing. The authors concluded that only Cr61 and Cr3+ components are observed on the Cr/Al203 catalyst. Quantitative )G’S Analysis. It has been shown by Defosse et al. [43] that one may calculate the theoretical intensity ratio I°,/I°, expected for a supported phase (p) atomically dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and Moulijn [44] has been used in the present investigation. The photoelectron cross sections and the mean escape depths of the photoelectrons used in these calculations were taken from Scofield [45] and Penn [46], respectively. For a phase (p) present as discrete particles, the experimental intensity ratio Ip/I. is given by the following expression: m. = 1°./1°.r1-expt-d/xm/ cm». (2) where I°,,/1°s is the theoretical monolayer intensity ratio, d is the length of the edge of the cubic crystallites of the deposited phase and XP is the mean escape depth of the photoelectrons in the deposited phase. The Cr species particle sizes were determined by solving the equation for d. The relative Cr dispersion to the monolayer was estimated from the ratio of Ip/Is to 1°P/I°,. CH4 Oxidation Activity. Measurement of CH; oxidation activity was performed in a flow microreactor. Approximately 0.1 g of catalyst were supported on a glass frit (70 - 1 00 pm) and the temperature was measured with a K-type thermocouple located just above the catalyst bed. The reactor was heated by a tube fumace (Lindberg) with 134 temperature being controlled within 1°C by an Omega CN 1200 temperature controller. Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers. Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated on a 6 ft 60/80 mesh Carbosieve S-II column. Prior to the first activity measurement, the catalyst was pretreated with a mixture of 5% 02/He (99.5 % purity for 02, 99.995% purity for He, AGA Gas Co.) stream (143 cc/min) at 350 °C for 1 hr to remove any impurities adsorbed on the surface during catalyst preparation and storage. CH4 oxidation reactions were performed with a constant flow rate (15 cm3/min) of 0.98 % CHa/5.25 % O2/93.77 % He gas mixture (AGA, purity > 99.99%) in the temperature range of 3 10-450 °C. Water produced during methane oxidation was frozen downstream from the reactor in a trap maintained < -40 °C with a mixture of ethanol and dry ice. All activity measurements were obtained under steady-state conditions at conversions less than 15 %. CO was observed as a partial oxidation product. The selectivity to CO2 was in the range of 73-83 % over Cry catalysts at 10 % conversion. Turn over number (TON) was calculated with CH; oxidation rate at 390 °C normalized by surface chromium atoms determined from Cr content and dispersion. 5.4. Results and Discussion Chemical State of Cr/A lumina Catalysts. XRD patterns obtained for Cry catalysts with Cr/Al atomic ratios 3 0.080 showed only lines characteristic of the alumina carrier 135 (Figure 5.1). For higher Cr loadings (Cr/Al atomic ratios 2 0.107), peaks characteristic of Cr2O3 were observed. The intensities of the 020; XRD peaks increase with increasing Cr content. Quantitative XRD measurements (Table 5.1) show that the amount of crystalline Cr2O3 increases from 0.9 to 5.5 wt.% as the Cr/Al atomic ratio increases from 0.107 to 0.134 (16 to 20 wt.% Cr2O3). Previous researchers have suggested that the support becomes saturated with surface chromium oxide at a critical metal loading and further Cr addition leads to the formation of crystalline Cr2O3 [22,47]. XRD results for the Cry catalysts indicate that the critical chromium loading is approximately 15 wt.% Cr2O3. Zaki et a1. [47] have reported that crystalline Cr2O3 forms at 10.2 wt.% Cr2O3 on 135 m2/g A1203 support. This corresponds to 15.3 wt.% Cr2O3 after normalization for the surface area of alumina used in our study. Figure 5.2 shows the Cr 2p XPS spectra obtained for the Cry catalysts. The average Cr 2p3/2 binding energies measured for the catalysts (579.7 3: 0.1 eV) are close to the value determined for 00;. (579.4 eV) and higher than the values obtained for Cr2O3 (576.4 eV). The Cr 2p3,2 spectra also contain a shoulder at lower binding energy (577.0 1‘ 0.2 eV) which is close to the value measured for Cr2O3. The intensity of the shoulder increases with increasing Cr loading. The binding energies measured for the standard compounds and catalysts agree well with values reported previously [l8,21,27,35,48]. Comparison of binding energies measured for Cry catalysts and standard compounds indicates that the major peak and shoulder can be assigned to Cr(Vl) and Cr(III), reSpectively. This is consistent with previous XPS studies of Cr/Al2O3 catalysts [18-21] which showed that Cr(VI) species was present predominantly at low Cr concentration (< 5 wt - °/o Cr2O3 on 200 mZ/g A1203) and the concentration of Cr(IIl) species increased with 136 C1303 CrZO, <110> <104> Cr,03 <012> C1303 _ <113> ‘ (C) , " A120, (b) <3 l> ' . . A120, A120, . 2) <220> r A120, (a) <400> l l l l L l Figure 5.1. 25 30 35 4o 45 50 Degrees (20) XRD patterns measured for (a) A1203, (b) Cr] 1, and (c) Cr13. 137 Table 5.]. Concentration of Crystalline' Phases in Cry Catalysts Calculated from Quantitative XRD Data. Crystalline Phase (wt%) Catalysts As prepared Measured Cr 1.3 2 3’ Cr 2.7 4 -b Cr 5.4 8 3’ Cr 8.0 12 3’ Cr 11 16 0.9 Cr 13 20 5.5 a - Valid for crystalline phases with particle sizes > 3.0 nm. b - No 0203 peaks detected forthese catalysts. increasing Cr content. The variation of the Cr 2p3/2/Al 2p XPS intensity ratio of Cr(VI) and Cr(III) species for the Cry catalysts determined using non-linear least-squares curve fitting are shown in Table 5.2. The Cr(V1)/Al intensity ratio increases with increasing Cr/Al atomic ratio up to 0.054. For higher Cr loadings, the Cr(VI)/Al intensity ratios measured for the catalysts are independent of Cr content. This indicates that the alumina support becomes saturated with Cr(VI) species at the catalyst with Cr/Al atomic ratio = 0.054 (8 wt.% Cr203) and further addition of chromium only leads to the formation of Cr(III) species. Wachs and coworkers [22,33] have used the absence of Raman and XRD peaks of 0203, the suppression of FTIR peaks of 0H, and C02 chemisorption to show that monolayer coverage was reached at 12 wt.% Cr03 (for 180 mz/g A1203). This corresponds to 10 wt.% Cr203 after normalization for chromium oxide molecular formula and surface area of alumina used in our study. The discrepancy between the XRD and 3338 data for saturation point (15 wt.% and 8 wt.% Cr203/Al203, respectively) is due to the different detection limits (selectivity) of the techniques. The saturation point 138 (f) (d) (C) (b) (a) 595 590 585 580 575 Binding Energy (eV) Figure 5.2. Cr 2p XPS spectra obtained for (a) Cr1.3, (b) Cr2.7, (c) Cr5.4, (d) Cr8.0, (e) Cr] 1, and (f) Crl3 catalysts. 139 Table 5.2. Cr 2p3/2/Al 2p XPS Intensity Ratios of Cr(VI) and Cr(III) Species in Cry Catalysts Determined Using Non-Linear Least-Squares Curve Fitting (NLLSCF). Intensity Ratio Catalysts Cr(VI)/Al Cr(III)/Al Cr 1.3 0.21 0.05 Cr 2.7 0.31 0.13 Cr 5.4 0.57 0.25 Cr 8.0 0.54 0.45 Cr 11 0.52 0.64 Cr 13 0.51 0.75 d at ermined by XRD is based on large Cr203 crystallites (d > 3.0 nm), while well dispersed Q r(VI) species are used for the determination derived from XPS. The Cr(III) species are Ob served from the Cr1.3 catalyst and the Cr(III)/Al )G’S intensity ratio increases with i h creasing Cr content (Table 5.2). The Cr(III) species on the low Cr content catalysts (below monolayer coverage) has been detected by ESR [28-32] and XPS [18-20,42] t echniques. The presence of Cr(IIl) species has been attributed to cluster formation [30] 61‘ calcination induced reduction [21]. The relative distributions of Cr(VI) and Cr(III) Sli>ecies for the low Cr loading catalysts (e.g. 19 % Cr(III) for Crl.3) are similar to the Values reported previously [19,42]. Scierka et al. [42] have found 17 % Cr(III) species for the 2.5 wt.% Cr/Al203 (170 mz/g) catalyst which corresponds to 4.4 wt.% Cr203 for alumina support used in this study. The firrther increase in Cr(III) content for Cr rich e21talysts(Cr11 and Crl3) can be attributed to the formation of crystalline Cr203. 140 Dispersion of Cr/Alumina Catalysts. Figure 5.3 shows the variation of the Cr 2 p / A1 2p intensity ratio measured for the Cry catalysts as a function of the Cr/Al atomic rat i o. The theoretical line calculated for monolayer dispersion [44] is shown for co m parison. The Cr/Al intensity ratio measured for the Crl.3 catalyst is identical, within exp erimental error, to the value predicted for monolayer dispersion. For higher Cr 1<>a.c1ings, the Cr/Al intensity ratios measured for the catalysts are lower than the monolayer values. The increasing deviation of the Cr/Al intensity ratio from the monolayer line with increasing Cr loading indicates that the Cr dispersion decreases e0tttinuously with increasing Cr content [44]. We ascribed the decrease in Cr dispersion t9 the formation of chromium oxide cluster or crystalline on the alumina support as a fixt‘rction of Cr content. Variation in the particle size of the chromium phase determined from Cr/Al XPS i “tensity ratios using Eq. (2) is given as a function of the Cr loading in Table 5.3. The mean particle size of the chromium phase calculated from XPS intensity ratios increases fl‘om 0.5 to 1.9 nm with increasing Cr content. Particle sizes determined from XRD line broadening calculations using Eq. (1) are also shown in Table 5.3. For catalysts with Cr/Al atomic ratios 2 0.107, XRD line broadening calculations show that the 0203 particle sizes are significantly larger than the values determined using XPS. The difference i h Cr203 particle sizes determined using XPS and XRD may be attributed to the limitations of X-ray diffraction. It is well known that XRD determined particle sizes are skewed to l arge values since XRD does not detect highly dispersed species. For the Cr rich catalysts, a significant fraction of the chromium is present as a surface phase or small Cr203 particles 141 A l monolayer —’ l u.) l N j I O O C XPS Intensity Ratio (Cr2p/A12p) u—d l O 1 1 U l r I I l l j I j 1 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Cr/Al Atomic Ratio I:igure 5.3. Cr 2p/Al 2p XPS intensity ratios measured for Cry catalysts (0) plotted versus Cr/Al atomic ratio. Cr 2p/Al 2p intensity ratios calculated for monolayer dispersion (line). 0nd load mos Cr(l inCTl 142 Table 5.3. Particle Size and DiSpersion of Chromium Phases Determined from XRD Line Broadening Calculations and Cr/Al XPS Intensity Ratios. Particle size (nm) Catalysts XRD XPS Dispersionc (%) Cr 1.3 J 3’ 100 Cr 2.7 -' 0.5 75.8 Cr 5.4 -' 0.8 67.6 Cr 8.0 -‘ 1.2 55.7 Cr 11 46 1.5 50.8 Cr 13 29 1.9 42.4 a - No Cr203 peaks detected for these catalysts. b - Value corresponds to monolayer dispersion. c - Calculated from XPS data. ((1 < 3.0 nm) that are detected by XPS but not XRD. Thus, smaller average chromium QDrizide particle sizes are expected from XPS calculations. Based on the XRD and XPS results presented above, the effect of chromium 1‘ Qading on Cry catalyst structure can be summarized as follows: (1) for low Cr loadings, l"lfiost of the Cr is present as a highly dispersed Cr(VI) surface phase, (2) the fraction of Cr(III) species present in the catalysts increases and the Cr dispersion decreases with i l’Icreasing Cr content, and (3) Cr rich catalysts contain large Cr203 particles. Effect of Catalyst Structure on CH4 Oxidation Activity. Table 5.4 shows the SIZJeCific turnover number (TON) for CH4 oxidation over Cry catalysts calculated at 390 °C using XPS estimates of the Cr dispersion. The TON increases approximately by a factor of 3 as the Cr/Al atomic ratio increases from 0.013 to 0.11. Further addition of Cr decreases the TON for CH4 oxidation. The activation energies for CH4 oxidation (Table 54) cal kcaltnc upon moteasi Colll) loading Ctlll) enhanc: several PlOPOSI ttachcn additio: for Ht decreas ElIItOUn irate f Nllanr X0 rec the (hp: 143 5 4 ) calculated for the Cry catalysts are similar within experimental error (20.2 i 0.9 1.: c alimol). The variation in TON for CH4 oxidation as a function of Cr loading can be ex p l ained in terms of the surface structure of the Cry catalysts. The increase in TON with increasing Cr/Al ratio over the range 0.013 to 0.107 parallels a significant increase in the Cr< III) content of the catalysts (Table 5.2). The inorease in TON as a function of Cr 10 ading up to Crll catalyst can be attributed to an increase in the formation of Cr(III)- C—‘—'t‘(\/I) clusters that may be more active for CH4 oxidation than Cr(VI) species. The e‘1'111anced activity of an alumina supported Cr(III)-Cr(VI) species has been reported for S"F—E‘weral different reactions [28-32,47-52]. Ellison and coworkers [28,31,32] have I) roposed that the mixed oxidation state cluster (y-phase) has high catalytic activity for t‘e-Ezlctions that require both oxidation and reduction process (Zener mechanism). In zi-Cldition, Zaki and coworkers [47,49,50] have reported that the y-phase was the active site er H202 and propanol decomposition over chromium catalysts. For the Crl3 catalyst, the Cl'Ecrease in TON relative to the Crll catalyst may be correlated to an increase in the a~1'Tlount of large Cr203 particles present in the Cr rich catalyst. Large 020:, particles will l”)ave fewer Cr(III)-Cr(VI) interfacial interactions compared to small Cr203 particles. 1\Tiiyama et al. [51,52] have reported that the y-phase was more active than bulk Cr203 for NO reduction with NH3. They proposed that the metal - surface oxygen bond strength of t he y-phase was weaker than that of bulk Cr203. The c firUCtl Surfac decree mstal Cilmm Conn Poorly 144 Table 5.4. Turn Over Numbers1 and Activation Energiesb of Cry Catalysts for C114 Oxidation. Catalysts TON (s’lxlOJ') Activation Energy Cr 1.3 2.8 21.7 Cr 2.7 3.4 20.4 Cr 5.4 4.5 20.3 Cr 8.0 5.7 19.8 Cr 11 7.3 20.2 Cr 13 5.6 18.8 a - Calculated at 390 °C using exposed Cr species determined from XPS. b - Estimated from Arrhenius plots. 5.5. Conclusions The combined use of several techniques to investigate the effect of Cr addition on the S‘tnicture and CH4 oxidation activity of Cry catalysts leads to the following conclusions. 1. XPS data indicate that most of the Cr is present as a highly dispersed Cr(VI) Starface phase in Crl.3 catalyst. The fraction of Cr(III) increases and the Cr dispersion decreases with increasing Cr loading. Cr-rich catalysts (Crll and Crl3) contain large Qtystallites of Cr203. 2. The TON for CH4 oxidation increases by a factor of 3 with increasing c}u'omium loading up to the Crl 1 catalyst. This has been attributed to the formation of a CI‘(III)-Cr(VI) cluster. Further addition of Cr decreases the TON due to the formation of poorly dispersed Cr203 crystallites. 11. 12- 13- 14- 15- 145 5.6. References Kummer, J. T. Prog. Energy Combust. Sci. 1980, 6, 177. Armor, J. N. Appl. Catal. B 1992, 1, 221. Zwinkels, M. F. M.; Jaras, S. G.; Menon, P. G., Griffin, T. A. Catal. Rev. -Sci. Eng. 1993, 35, 319. Dmuchovsky, B.; Freerks, M. C.; Zienty, F. B. J. Catal. 1965, 4, 577. Kang, Y.-M.; Wan, B.-Z. Appl. Catal. A 1994, 114, 35. Prasad, R.; Kennedy, L. A.; Ruckenstein, E. Combust. Sci. and Tech. 1980, 22, 271. Manning, M. P. Hazardous Waste 1984, l, 41. Weldon, J .; Senkan, S. M. Combust. Sci. and Tech. 1986, 47, 229. Kolaczkowski, S. T.; Beltran, F.; Crittenden, B. D.; Jefferies, T. M. Process Safety and Environmental Protection 1990, 68, 49. Drago, R. S.; Petrosius, S. C.; Grunewald, G. C., and Brendley, Jr., W. H. In Environmental Catalysis; Armor, J. N. Ed.; ACS Symposium Series 552; American Chemical Society: Washington, DC, 1994; Chapter 28. Agarwal, S. K; Spivey, J. 1.; Butt. J. B. Appl. Catal. 1992, 82, 259. Johnson, A. J.; Meyer, F. G.; Hunter, D. I., Lombardi, E. F. Incineration of polychlorinated biphenyls using a fluidised bed incinerator, RFP-3271, (Rockwell Int.) 1981. Shelef, M., Otto, K., Gandhi, H. J. Catal. 1968, 12, 361. Wong, W. C.; Nobe, K. Ind. Eng. Chem. Prod. Res. Dev. 1986, 25, 179. Simon, S.; van der Pol, A.; Reijerse, E. J.; Kentgens, A. P. M.; van Moorsel, G.-J. M. P.; de Boer, E. J. Chem. Soc, Faraday Trans. 1995, 91, 1519. Dwyer, F. G. Catal. Rev. 1972, 6, 261. 17- ll 1E3: .. TIL SEQ! 2C) 146 Kapteijn, F., Stegenga, S., Dekker, N. J. J., Bijsterbosch, J. W., Moulijn, J. A. Catal. Today 1993, 16, 273. Okamoto, Y.; Fujii, M.; Imanaka, T.; Teranishi, S. Bull. Chem. Soc. Jpn. 1976, 49, 859. Jagannathan, K.; Srinivasan, A.; Rao, C. N. R. J. Catal. 1981, 69, 418. Gorriz, O. F.; Corberén, V. C.; Fierro, J. L. G. Ind. Eng. Chem. Res. 1992, 3], 2670. Rahman, A.; Mohamed, M. H.; Ahmed, M.; Aitani, A. M. Appl. Catal. 1995, 121, 203. Vuurman, M. A.; Hardcastle F. D.; Wachs, I. E. J. Mol. Catal. 1993, 84, 193. Hardcastle, F. D.; Wachs, I. E. J. Mol. Catal. 1988, 46, 173. Vuurman, M. A.; Wachs, 1. E; Stufltens, D. J .; Oskam, A. J. Mol. Catal. 1993, 80, 209. Vuurman, M. A.; Wachs, I. E. J. Phys. Chem. 1992, 96, 5008. Weckhuysen, B. M.; Schoonheydt, R. A.; Jehng, J.-M.; Wachs, I. E.; Cho, S. J.; Ryoo, R.; Kilistra, S.; Poels, E. J. Chem. Soc. Faraday Trans. 1995, 91, 3245. Cimino, A.; De Angelis, B. A.; Luchetti, A.; Minelli, G. J. Catal. 1976, 45, 316. Ellison, A.; Oubridge, J. O. V.; Sing, K. S. W. Trans. Faraday Soc. 1970, 66, 1004. Ellison, A.; Sing, K. S. W. J. Chem. Soc., Faraday Trans. 1978, 74, 2017. Ellison, A. J. Chem. Soc, Faraday Trans. 1, 1984, 80, 2567. Ellison, A.; Sing, K. S. W. J. Chem. Soc. Faraday Trans. 1978, 74, 2807. Ellison, A J. Chem. Soc, Faraday Trans. 1, 1984, 80, 2581. Turek, A. M.; Wachs, I. E.; DeCanio, E. J. Phys. Chem. 1992, 96, 5000. Vuurman, M. A.; Stuflcens, D. J.; Oskam, A.; Moulijn, J. A.; Kapteijn, F. J. Mol. Catal. 1990, 60, 83. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 147 Scierka, S. J .; Houalla, M.; Proctor, A.; Hercules, D. M. J. Phys. Chem. 1995, 99, 1537. Anderson, R. B.; Stein, K. C.; Feenan, J. J .; Hofer, L. J. E. Ind. Eng. Chem. 1961, 53, 809. Kuznetsova, L. L.; Paukshtis, E. A.; Shkurina, G. P.; Shkrabina, R. A.; Koryabkina, N. A.; Arendarskii, D. A.; Barannik, G. B.; Ismagilov, Z. R. Catal. Today 1993, 17, 209. Powder Diffraction File: Inorganic Phases, Joint Committee on Powder Diffraction Standards, PA, 1983. Klug, H. P.; Alexander, L. E., X-Ray Diffraction Procedures for Polycrystalline and Amorphous Materials. lst.Ed. Wiley, New York, 1954. De Angelis, B. A. J. Electron Spectrosc. Relat. Phenom. 1976, 9, 81. Software provided by Dr. Andrew Proctor, University of Pittsburgh, Pittsburgh, PA. Scierka, S. J.; Proctor, A.; Houalla, M.; Fiedor, J. N.; Hercules, D. M. Surf Interface Anal. 1993, 20, 901. Deffosse, C .; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269. Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612. Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129. Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29. Zaki, M. 11.; Fouad, N. E.; Leyrer, J.; Knézinger, H. Appl.Catal. 1986, 21, 359. Griinert, W.; Shpiro, E. S.; Feldhaus, R.; Anders, K. Antoshin, G. V.; Minachev, KH. M. J. Catal. I986, 100, 138. Fahim, R. B.; Zaki, M. I.; Gabr, R. M. Surf Technol. 1981, 12, 317. Fahim, R. B.; Zaki, M. 1.; Gabr, R. M. Appl. Catal. 1982, 4, 189. 148 51. Niiyama, H.; Murata, K.; Ebitani, A.; Echigoya, E. J. Catal. 1977, 48, 194. 52. Niiyama, H.; Murata, K.; Echigoya, E. J. Catal. I977, 48, 201. Chapter 6 Characterization and CO Oxidation Activity of Cu/Cr/Al2O3 Catalysts 6.1. Abstract X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been used to characterize a series of Cu/Cr/Al203 catalysts. The catalysts were prepared by stepwise incipient wetness impregnation of chromium first, followed by copper. The copper loading was held constant at 8 wt.% CuO and chromium loadings were varied from Cr/Al atomic ratios from 0 to 0.0134 (0 to 20 wt.% Cr203). The information obtained from surface and bulk characterization has been correlated with CO oxidation activity of the Cu/Cr/Al203 catalysts. The copper addition decreased the dispersion of chromium phase by reacting selectively with a dispersed Cr3i species to form large crystalline CuCr2Oa. The Cu dispersion also decreased with increasing Cr loading. For low Cr loading catalysts (Cr/Al g 0.027), CO oxidation activity increased with the addition of Cr due to the formation of crystalline CuO on the chromium modified alumina. This has been attributed to the prevention of Cu ion diffusion to alumina lattice vacancies by highly dispersed chromium species. The Cu/Cr/Al203 catalyst of Cr/Al = 0.054 showed the highest C0 oxidation activity due to the formation of an active CuCr204 phase. For further addition of Cr (Cr/Al 2 0.080), the catalytic activity for C0 oxidation decreased. 149 150 This might be due to the decrease in active phase dispersion and encapsulation of the active site with excess Cr species. 6.2. Introduction The combination of copper and chromium oxide was recognized as a promising active component for emission control reactions in the early years of pollution control catalysis research [1-5]. The Cu-Cr catalyst system has been studied extensively for various reactions, including CO oxidation [6-21], hydrocarbon oxidation [6-10,22-23], N0 reduction [19-20,24], alcohol and aldehyde oxidation [25-26], and sulfumated [27] and chlorinated [28] hydrocarbon oxidation. Shelef et al. [4] and Kapteijn et al. [20] have performed a screening study of transition metal oxide catalysts for C0 oxidation and N0 reduction. They have reported that Cu-Cr catalysts showed consistently higher catalytic activity than single oxide catalysts based on Cu, Ni, Co, Fe, Mn, Cr, or V. Barnes [6] studied automotive pollution control catalysts for exhaust treatment such as CO and hydrocarbon oxidations. He reported that the alumina supported 8 wt.% Cu-7 wt.% Cr catalyst showed similar activity to 0.3 wt.% Pt catalyst. In addition, Stegenga et al. [29] have found that monolith supported 10 wt.% Cu-Cr/Al203 catalyst showed a three-way catalytic behavior (CO and hydrocarbon oxidation, and N0 reduction, simultaneously) comparable to noble metal catalysts Operated under the same conditions. In the copper chromite catalyst system, copper oxide has been proposed as an active component for C0 oxidation reaction, while chromium oxide has been considered 151 as a promoter [13, 19,23,30]. The roles of Cr promoter to the Cu species are believed to limit catalyst reduction [12-14,24,31], prevent catalyst poisoning [10,12,15], inhibit bulk copper aluminate formation [30], improve thermal stability [10], and increase catalyst dispersion [17,30]. Furthermore the mixed oxide catalyst shows greater activity and stability than the single oxide catalysts [32,33]. The synergistic effect of Cu-Cr catalysts has been attributed to the electronic interaction between copper and chromium species [13,17,23]. Despite a considerable amount of effort devoted to the study of Cu-Cr catalyst, the understanding of Cu-Cr catalyst system is still not clear. Stegenga et al. [29] reported that the optimum metal composition of Cu-Cr/Al203 catalyst was Cu/Cr = 2 for CO oxidation. However, Severino et al. [12] found that Cu/Cr = 1 was a optimal metal composition of alumina supported catalyst for C0 oxidation. Also, uncertainty has arisen with regard to the pretreatment for the most efficient catalyst. Severino et a1. [12] reported that catalyst prereduction enhanced catalytic activity efficiently for CO oxidation. However, Chien et al. [23] showed that oxidative pretreatment of Cu-Cr catalyst was more beneficial than reductive pretreatment for CO oxidation. Much of the previous research on the Cu-Cr catalysts have focused on the understanding of catalytic properties such as activity or stability. Little effort has been devoted to the relationship between surface structure of Cu/Cr/Al203 and its catalytic activity. The present work is a part of a broad study to investigate the effect of base metal oxide promoters on the structure and reactivity of copper oxide based emission control catalysts. In this paper, X-ray photoelectron spectroscopy (XPS) and X-ray diffraction 152 (XRD) were used to determine the effect of Cr loading on the chemical state and dispersion of chromium and c0pper oxide phases supported on y-alumina. The information derived from these techniques has been correlated with C0 oxidation activity to develop a more complete understanding of the Cu/Cr/Al203 catalyst. 6.3. Experimental Standard Materials. Cr03 (99.9 %) was obtained from Aldrich Chemical Company, Inc. CuO and Cr203 were prepared by calcining metal nitrates in air at 500 °C for 16 h. CuCr2Oa and CuAl204 were prepared by calcining stoichiometric mixtures of the respective nitrates in air at 1000 °C. for 24 h. XRD patterns of the standard compounds matched with the appropriate Powder Diffraction File [34]. Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y- alumina (Cyanamid, y-A1203, surface area = 203 mZ/g, pore volume = 0.6 mL/g). The alumina support was finely ground (< 230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation. Catalysts were prepared using deionized water solutions of chromium(III) nitrate (Mallinckrodt, Analytical Reagent) and copper(II) nitrate (Columbus Chemical Industries Inc, ACS Grade). The impregnated catalysts were dried in air at 120 °C for 24 h and subsequently calcined in air at 500 °C for 16 h. Mixed metal Oxide catalysts (Cu/Cr/Al203) were prepared by the step wise impregnation of chromium first, followed by Cu. Chromium impregnated alumina (Cr/A1203) was dried and calcined 153 prior to the introduction of copper. Catalysts with added copper were dried and calcined again under the same conditions. The copper content was held constant at 8 wt.% (as CuO) of the alumina support. The chromium content was varied from an Cr/Al atomic ratio of 0 to 0.134 (0 to 20 wt.% Cr203). Catalyst samples will be designated by "Cry" and "CuCry", where y is the Cr/Al atomic ratio (x 102). BET Surface Area. Catalysts surface area were determined using a QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 g of catalyst was outgassed in a N2/He mixture (5% N2) at 350 0C for 1 h prior to adsorption measurement. The measurement was made using relative pressures of N2 to He of 0.05, 0.08, and 0.15 (N2 surface area = 0.162 nmz) at 77 K. The BET surface areas of catalyst samples decreased with increasing catalyst loading by a maximum of 20 %. X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a Rigaku XRD diffractometer employing Cu Ka radiation (A = 1.5418 A). The diffractometor was operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scan rate of 0.5 deg/min. with divergence slit and scatter slit widths of 1°. Samples were run as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity. The mean crystalline sizes (d) of the CuO, Cr203. and CuCr204 particles were determined from XRD line broadening measurement using the Scherrer equation [3 5]: d=Kiu/Bc059 (1) 154 where A is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the fill] width at half maximum (fwhm), in radians, of the CuO, Cr203<104>, or CuCr204<31l> lines. Quantitative X-ray diffraction data were obtained by comparing CuO, Cr203<104>, or CuCr2Oa<3 1 1> vs A1203<400> intensity ratios measured for catalyst samples with intensity ratios measured for the physical mixtures of pure CuO, Cr203, or CuCr204 and y-Al203. This method assumes that copper and chromium addition does not disrupt the y-Al203 spinel and subsequently affect the intensity of the AJ2O3<400> line. The error in this method was estimated to be i 20 %. XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science instrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20 mA) and a 10-360 hemispherical analyzer operated with a pass energy of 50 eV. The instrument typically was operated at pressures near 1 x 10'8 torr in the analysis chamber. Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer. Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the catalyst samples and CuAl2Oa were referenced to the Al 2p peak (74.5 eV). The binding energies for standard compounds with no A1 were referenced to the C Is line (284.6 eV) of the carbon overlayer. XPS binding energies were measured with a precision of i 0.2 eV, or better. Reduction of copper [36-38] and chromium [39-41] species during the XPS experiment has been reported and attributed to several factors such as X-ray flux, X-ray dose, temperature, and pressure. In order to minimize the effect of photoreduction on the 155 results, all samples were analyzed using the same X-ray flux, a constant distance between the X-ray source and the sample, and a minimum data acquisition time. Quantitative XPS Analysis. It has been shown by Defosse et al. [42] that one may calculate the theoretical intensity ratio I°,/I°. expected for a supported phase (p) atomically dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and Moulijn [43] has been used in the present investigation. The photoelectron cross sections and the mean escape depths of the photoelectrons used in these calculations were taken from Scofield [44] and Penn [45], respectively. For a phase (p) present as discrete particles, the experimental intensity ratio IP/I, is given by the following expression: 1px], = I°,,/1°,[1-exp(-d/r.,)]/ do.p (2) where I°,,/I°, is the theoretical monolayer intensity ratio, d is the length of the edge of the cubic crystallites of the deposited phase and 2,, is the mean escape depth of the photoelectrons in the deposited phase. Curve Fitting. XPS data manipulation performed using ‘GOOGLY’ PC software [46]. The Cr 2p3/2 range was used to measure Cr binding energies and a relative oxidation state distribution by non-linear least-squares curve fitting (NLLSCF). The Cr peak positions were allowed to float slightly (i 0.1eV) in order to fit into the Cr 2p3/2 envelope of each sample. The full width at half maximum (FWHM) of Cr(IIl) was fixed as 3.3 eV, a value that was determined from the line width for Cr203. The Cr(VI) FWHM of Cr/Al203 catalysts did not agree with the value for 003 (2.1 eV). Also, the width 156 increased with decreasing Cr content. The FWHM of Cr(VI) was allowed to float from 3.0 :l: 0.3 eV. Those FWHM of standard compounds and Cr/Al203 catalysts are well agreed with the values reported by Scierka et a1 [41]. Catalytic Activity Measurement The measurement of CO oxidation activity was performed in a flow microreactor (1/2 inch glass tube). Approximately 0.03 g of catalyst was supported on a glass frit (70 - 100 um) and the temperature was measured with a K- type thermocouple located just above the catalyst bed. The reactor was heated by a tube furnace (Lindberg) with temperature being controlled within 1 °C by an Omega CN 1200 temperature controller. Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers. Product gases were analyzed with a Varian 920 gas chromatography equipped with a TCD and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated on a 6 ft 60/80 mesh Carbosieve S—II column. Prior to the first activity measurement, the catalysts were pretreated with a mixture of 5% 02/He (99.5 % purity for 02, 99.995% purity for He, AGA Gas Co.) stream (143 cc/min) at 350 °C for 1 hr to remove any impurities adsorbed on the surface during catalyst preparation and storage. CO oxidation reactions were performed with a 80 cm3/min flow of 4.8% C0/9.8% 02/85.4% He gas mixture (AGA, purity > 99.99%). The reaction rate was calculated from the TCD sensitivity corrected CO and CO2 chromatographic response of the reactor output. All activity measurements were obtained under steady-state conditions at conversions less than 15 %. The specific turn over number (TON) was calculated from the C0 oxidation rate at 240 °C which was 157 normalized with respect to Cu dispersion as estimated by XPS data. The activation energy was determined from an Arrhenius plot for CO oxidation. 6.4. Results and Discussion Chemical Composition of Cu/Cr/AI2O3 Catalysts Standard Compounds. The XPS Cu 2p3/2 binding energies and satellite/main peak intensity ratios for standard compounds are shown in Table 6.1. CuA1204 and CuCr204 have higher binding energies and satellite/main peak intensity ratios than the values obtained for CuO. Those binding energies and satellite/main peak intensity ratios of the standard compounds are consistent with previously reported values [3 8,47,48]. Ctr/A1203 Catalyst. The XPS Cu 2pm binding energy measured for the CuCrO catalyst is higher than the value obtained for CuAlgOa and CuO, but closer to the value of CuAl204 (Table 6.1). Figure 6.1 shows the XRD patterns obtained for alumina support and CuCry catalysts. The XRD pattern of the CuCrO catalyst shows only lines characteristic of alumina. The XPS and XRD results indicate the predominant presence of dispersed copper ‘surface spinel’ phase on the alumina support [38,48]. However, the satellite/main peak ratio measured for the CuCrO catalyst is closer to the value of CuO than to that of CuAl2Oa. This result conflicts with XRD data which indicates the absence of crystalline CuO for the CuCrO catalyst. Strohmeier et al. [38] have explained this discrepancy that well dispersed or amorphous CuO was not necessarily present on the 158 Table 6.1. XPS Cu 2p3/2 Binding Energies and Shakeup/Main Peak Intensity Ratios Measured for CuCry Catalysts and Standard Compounds. Cr/Al Atomic Ratio (x103) BE of Cu 2pm shakeup/main peak ratio 0.0 935.4 0.42 1.3 935.3 0.48 2.7 935.5 0.56 5.4 ‘ 935.5 0.62 8.0 ' 935.5 0.63 10.7 935.5 0.69 13.4 935.5 0.72 CuO 933.9 0.45 CuAlgOa 935.0 0.71 CuCrgOa 934.5 0.70 catalyst, but the surface spinel copper species might be chemically different from both bulk CuO and CuAl2Oa. Cu/Cr/AlgO3 Catalysts- The average XPS Cu 2p3,2 binding energy measured for CuCry catalysts (935.5 : 0.1. eV) is identical, within experimental error, to the value obtained for CuCrO catalyst (Table 6.1). This indicates that the addition of chromium had little effect on the XPS Cu 2p32 binding energies measured for the catalysts. Figure 6.2 shows the Cr 2p XPS spectra of CuCry catalysts. The Cr 2p3/2 binding energies of a main peak (580.0 : 0.1 eV) and a shoulder (577.4 i 0.1 eV) measured for the catalysts has been assigned to the Cr(VI) and Cr(III) species, respectively [49]. The relative distribution of Cr(VI) and Cr(III) oxidation states in CuCry catalysts indicates that Cr(III) species increase with increasing Cr content (Table 6.2). The XPS binding energies 159 + + + + (h) (g) (f) (e) ' . (d) Figure 6.1. ' 25 30 35 40 45 50 Degrees (20) XRD patterns measured for CuCry catalysts ( O alumina, l CuO, 0 cubic CuCr204, V tetragonal CuCr2Oa, ‘f‘r’ Cr203). (a) alumina (b) CuCrO, (c) CuCr1.3, (d) CuCr2.7, (e) CuCr5.4, (f) CuCr8.0, (g) CuCr10.7, and (h) CuCr13.4. 160 (f) (6) (d) (0) (b) 595 590 585 580 575 Binding Energy (eV) Figure 6.2. Cr 2pm XPS spectra measured for CuCry catalysts. (a) CuCr1.3, (b) CuCr2.7, (c) CuCr5.4, (d) CuCr8.0, (e) CuCr10.7, and (f) CuCr13.4. 161 and relative oxidation state distribution of Cr species for the CuCry catalysts are consistent with the results obtained for Cry catalysts within experimental error [49]. This indicates that the post-addition of Cu does not affect Cr oxidation states in CuCry catalysts. Our previous study [49] indicated that the chromium oxide structure of Cry catalysts depended on the Cr content. Three levels of Cr loading will be discussed the structure of CuCry catalysts: low Cr loading (0.013 S Cr/Al atomic ratios 3 0.027), intermediate Cr loading (0.054 s Cr/Al g 0.080) and high Cr loading (Cr/Al 2 0.107). For CuCry catalysts with low Cr loading, the XRD patterns show peaks characteristic of CuO (Figure 6.1). Table 6.3 shows the amount of crystalline CuO, tetragonal CuCr2Oa, and Cr203 present in the CuCry catalysts as a function of Cr/Al atomic ratio. The amount of crystalline CuO increases from 0.4 to 1.9 wt.% as the Cr/Al atomic ratio increases from 0.013 to 0.027. The formation of CuO phase can be explained in terms of a decrease in interaction between copper species and alumina support [50]. A significant number of the lattice vacant sites normally accessible to the Cu2+ ions during catalyst preparation steps are blocked by pre-loaded chromium oxide species (well dispersed Cr6’ species). Thus, Cui‘ ions are not able to enter the alumina lattice to form the copper surface spinel phase. Upon calcination, these Cu} ions agglomerate to form large particles of CuO. The formation. of a CuO phase on the Cr-modified alumina indicates that the copper species does not readily interact with the Cr°+ species to form CuCr204 phase. This is consistent with the result of the relative distribution of Cr(VI) and Cr(III) oxidation states in CuCry catalysts which were similar to the values of Cry catalysts [49]. For CuCry catalysts with intermediate Cr loading, the XRD patterns show 162 Table 6.2. Distribution of Cr Oxidation States Determined from Cr 2p322 XPS Spectra of CuCry Catalysts Measured by Non-Linear Least-Squares Curve Fitting (NLLSCF). Cr/Al Atomic Ratio Relative concentration (%) (x103) Cr(VI) Cr(III) 1.3 79 21 2.7 77 23 5.4 73 27 8.0 57 43 10.7 54 46 13.4 47 53 Table 6.3. Concentration of Crystalline“ Phases in CuCry Catalysts Calculated from Quantitative XRD Data. Cr/Al Atomic Weight Percent of Crystalline Phase Ratio (x103) Cr203 CuO CUCT304 0.0 b 0.0 b 1.3 0.0 0.4 0.0 2.7 0.0 1.9 0.0 5.4 0.0 0.0 0.3 8.0 0.0 0.0 3.4 10.7 1.3 0.0 3.4 13.4 8.1 0.0 3.1 a - Valid for crystalline phases with particle sizes > 3.0 nm. b - Catalyst does not contain chromium. 163 peaks characteristic of tetragonal CuCr204 (Figure 6.1). In addition, a peak observed at 359° in the diffraction pattern of the CuCr5.4 catalyst can be assigned to a cubic CuCr204 <311>. The crystalline tetragonal CuCr204 content of the catalysts increases from 0.3 to 3.4 wt.% as the Cr/Al atomic ratio increases from 0.054 to 0.080 (Table 6.3). For CuCr5.4 and CuCr8.0 catalysts, the XRD patterns indicate that CuCr204 is the major crystalline phase. This is consistent with the increase of the Cu 2p3/2 XPS satellite/main peak ratio up to 0.63 (Table 6.1). For CuCry catalysts with high Cr loading, the peaks characteristic of crystalline Cr203. are observed in the XRD patterns (Figure 6.1). The crystalline Cr203 content of the catalysts increases from 1.3 to 8.1 wt.%, however, the crystalline CuCr204 content is independent of the Cr addition as the Cr/Al atomic ratio increases from 0.107 to 0.134 (Table 6.3). The Cu 2p3,2 XPS satellite/main peak intensity ratios for the high Cr loading catalysts (Table 6.1) are close to the value obtained for CuCr204. This result indicates that most of the Cu reacts with Cr species to form CuCr204 instead of interacting with the alumina support. The comparable Cr203 content obtained for the CuCry and the analogous Cry catalysts [49] suggests that little CUCT304 phase is formed by the interaction of Cu with large Cr203 crystallites. This is readily understood when one considers that large Cr2O3 crystallites have relatively less surface area available to react with Cu species than small Cr203 crystallites. 164 Dispersion of Cu/Cr/Al203 Catalysts. Copper species. Figure 6.3 shows the variation of the Cu ZpyMAl 2p XPS intensity ratios measured for the CuCry catalysts as a function of Cr/Al atomic ratio. The decrease in the XPS intensity ratios with increasing Cr content indicates a decrease of Cu species dispersion in CuCry catalysts as a function of Cr loading [43]. However, it should be noted relatively high Cu 2p322/Al 2p XPS intensity ratio observed for CuCr5.4 catalysts. This is probably due to the formation of CuCr204 phase from well dispersed Cr oxide species on the alumina support. The variation in copper particle size as a function of Cr content determined from Cu/Al XPS intensity ratios and XRD line broadening is shown in Table 6.4 and 6.5, respectively. The particle size of Cu species in CuCr0 catalyst (1.0 nm) calculated fi'om XPS data indicates that Cu is well dispersed over alumina carrier. This is consistent with XRD result which does not show any crystalline phase. The Cu phase particle sizes calculated from XPS data increase from 1.0 to 2.1 nm with increasing Cr loadings. This result agrees with XRD data that the formation of large CuO and CuCr204 crystallites is observed in the CuCry catalysts with increasing Cr content. CuO and CuCr204 particle sizes determined using XRD line broadening calculations are significantly larger than the particles sizes determined from XPS data. The difference in particle size determined from XPS and XRD results may be understood in terms of the detection limit of XRD. Since XRD detects large particles (> 3 nm) exclusively, the particle sizes calculated from XRD line broadening result have tendency to be large values. Chromium species. Figure 6.4 shows the variation of the Cr 2p/Al 2p XPS intensity ratios measured for CuCry catalysts (open circles) as a firnction of Cr/Al atomic ratio. Cr 2p/Al 2p XPS intensity ratios obtained for the analogous Cry catalysts (solid 165 1.1 1.0 C 2.5- 0.9 - 3 Q 0 a: It '2 g 0 8 - . .9. 0 35? 0.7 - U) fi 8 E. ‘ 0 0.6 - ' o 0.5 U T U l I I I l I ' U I I 0.00 0.02 0. 04 0.06 0.08 0. 10 0.12 0.14 Cr/Al Atomic Ratio Figure 6.3. Cu 2p3,2/Al 2p XPS intensity ratios of CuCry catalysts plotted versus Cr/Al atomic ratio. Monolayer dispersion value = 2.28. 166 Table 6.4. Particle Sizes of Chromium and Copper Species Determined from XPS Intensity Ratios. Cr/Al Atomic Particle Size (nm) Ratio (x103) Chromium Copper 0.0 a 1.0 1.3 b 1.3 2.7 0.7 1.5 5.4 1.2 1.4 8.0 2.0 1.8 10.7 2.0 2.1 13.4 2.5 2.1 a - Catalyst does not contain chromium. b - Catalyst dispersion is close to monolayer. Table 6.5. Particle Sizes of CuO, Cr203. and CuCrgO. Phases Determined from XRD Line Broadening Calculations. Cr/Al Atomic Particle Size(nm) Ratio (x103) CuO Cr203 CUCl'gos 0.0 a b b 1.3 17 a a 2.7 16 a a 5.4 a a 31 8.0 a a 29 10.7 a 44 30 13.4 a 33 28 a - No diffraction lines detected. b - Catalyst does not contain chromium. 167 5 4- a . monolayer —’ N 3‘ o. o 3- C.) v .2 3 Ei Dd €2- o r: C 2 E. . O O 0 g o 1- 0 O 0 0 I I I l I l I l I I I I fl 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Cr/Al Atomic Ratio Figure 6.4. Cr 2p/Al 2p XPS intensity ratios of Cry (0) and CuCry(O) catalysts plotted versus Cr/Al atomic ratio. Cr 2p/Al 2p intensity ratios calculated for monolayer dispersion (line). 168 circles) [49] and the theoretical line calculated from the Kerkhof and Moulijn model [43] for monolayer dispersion are also shown for comparison. The Cr/Al intensity ratios measured for the CuCry catalysts with Cr/Al atomic ratio 3 0.027 are similar to the values measured for Cry catalysts. The absence of XRD peaks characteristic of discrete Cr phases and the similarity of the Cr/Al intensity ratios measured for the catalysts with the values calculated for monolayer dispersion indicate that Cr is highly dispersed over the alumina carrier. For the CuCry catalysts with Cr/Al atomic ratio 2 0.054, the Cr/Al intensity ratios are lower than the values obtained for the analogous Cry catalysts. This indicates that Cu addition decreases the dispersion of the Cr species. This is consistent with XRD results which showed large crystalline CUCT304 and G203 phases for the catalysts with Cr/Al atomic ratio 2 0.054. The variation in Cr phase particle size as a function of Cr loading determined from Cr/Al XPS intensity ratios and XRD line broadening is shown in Tables 6.4 and 6.5, respectively. Cr particle sizes calculated from XPS data generally increase with increasing Cr loading. As for the copper species, the Cr2O3 particle sizes determined for CuCry catalysts using XRD line broadening calculations are significantly larger than the values determined using XPS. Effect of Catalyst Structure on C0 Oxidation Activity. Table 6.6 shows the specific turn over numbers (TON) and activation energies of CuCry catalysts for C0 oxidation as a function of Cr/Al atomic ratio. The specific TON increases by a factor of 8 with increasing the Cr/Al atomic ratio from 0.0 to 0.054. However, further addition of Cr decreases the specific TON for CO oxidation. The 169 Table 6.6. Turn Over Numbers' and Activation Energies” of CuCry Catalysts for CO Oxidation. Cr/Al Atomic TON Activation Energy Ratio (x102) (s'lx102) (kcal/mol) 0.0 4 17.9 1.3 15.3 16.5 2.7 24.3 14.4 5.4 31.5 15.8 8.0 16.2 16.2 10.7 19.7 17.5 13.4 15.2 16.3 a— Determined from reaction rate at 240 °C normalized by Cu dispersion. b- Estimated from Arrhenius plots within 15 % conversion. activation energies of CuCry catalysts are independent of Cr content (16.4 i 1.1 kcal/mol). The C0 oxidation activity can be attributed predominantly to the Cu species in CuCry catalysts, since Cr species show a little CO oxidation activity compared to Cu catalysts [6,13,19,23]. In this study, Cry catalysts showed negligible CO oxidation activity compared to CuCr0 catalyst at the same reaction condition. Addition of Cr up to a Cr/Al atomic ratio of 0.027 increases the TON by a factor of 6 compared to the CuCrO catalyst. We attributed this to the formation of an active CuO phase which was observed in XRD patterns for CuCrl.3 and CuCr2.7 catalysts. In our previous work [51], we concluded that crystalline CuO was more active for C0 oxidation reaction than Cu surface phase. This has been attributed to that redox ability of CuO/Al2O3 was more favorable than dispersed Cu surface phase during CO oxidation reaction [12]. 170 It should be noted that CuCr5.4 catalyst shows the highest specific TON among the CuCry catalysts, although Cu and Cr dispersion decreases continuously as a function of Cr content. This has been correlated with the formation of CuCr204 phase on the CuCr5.4 catalyst which was detected by XRD. This is consistent with that satellite/main peak ratio of CuCr5.4 catalyst increases to 0.62 which is close to the value of pure CuCr204 (0.70) (Table 1). Therefore we attributed the activity improvement of CuCr5.4 catalyst to the formation of active phase (CuCr204). CuCr204 has been known more active catalyst than CuO for CO oxidation [4,13,19]. Severino and Laine [13] have studied alumina supported and unsupported Cu-Cr catalysts for CO oxidation. They suggested that chromium might enhance the catalytic activity by providing structure oxygen (02') to copper species and electron transfer from copper to chromium. Kapteijn et a1. [19] have also proposed recently that the Cu-Cr oxide catalyst was more active than a single oxide for CO oxidation due to more available of surface oxygen on the mixed oxide catalyst. The further addition of Cr (Cr/Al atomic ratio 2 0.080) decreases C0 oxidation activity for CuCry catalysts. XPS and XRD data indicate that the dispersions of both Cu and Cr are decreased significantly and a large CuCr204 crystalline phase is formed on the catalyst. In addition, excess Cr203 phase, which is not active for CO oxidation reaction, is present on the catalyst surface. Thus, low dispersion and covering active sites by excess Cr phase may decrease catalytic activity for high Cr content CuCry catalysts. 171 6.5. Conclusions The combined use of several techniques to investigate the effect of Cr addition on the structure and C0 oxidation activity of Cu/Cr/Al203 catalysts leads to the following conclusions. 1. Cu addition decreased Cr dispersion in CuCry catalysts (Cr/Al 2 0.054) compared to the Cr dispersion in analogue Cry catalysts. The dispersion of Cu species in CuCry catalysts also decreased with increasing Cr content. The decrease in dispersion of both Cu and Cr species has been attributed to the formation of large crystalline CuO, CuCr2Oa, and Cr203 on the alumina support. 2. For CuCry catalysts with low Cr contents (Cr/Al g 0.027), the presence of highly dispersed Cr6+ phase inhibited the formation of copper surface phase and enhanced the formation of crystalline CuO over the alumina support. The CO oxidation activity increased in these catalysts (CuCrl.3 and CuCr2.7) has been attributed to an increase in the amount of crystalline CuO. 3. The CuCr5.4 catalyst showed the highest CO oxidation activity in this study. This is due to the formation of active CuCr204 phase. Post loaded copper species interacted with well dispersed Cr?" species to form CuCr204 rather than with highly dispersed Créi species or large Cr203 crystallites. 4. The CuCry catalysts with high Cr contents (Cr/Al 2 0.08) showed the decreased CO oxidation activity. This has been attributed to a decrease in the dispersion of Cu and Cr species and encapsulation of the active site with excess Cr203. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 172 6.6. References Frazer, J. C. W. US. Pat. 2031475, 1936. Roth, J. F.; Doerr, R. C. Ind. Eng. Chem. 1961, 53, 293. Hofer, L. J. E; Gussey, R; Anderson, R. B. J. Catal. 1964, 3, 451. Shelef, M.; Otto, K.; Gandhi, H. J. Catal. 1968, 12, 361. Dwyer, F. G. Catal. Rev.- Sci. Eng. 1972, 6, 261. Hertl, W.; Farrauto, R. J. J. Catal. 1973, 29, 352. Yu Yao, Y. F.; Kummer, J. T. J. Catal. 1977, 46, 388. Barnes, G. J. Adv. Chem. Series 1975, 143, 72. Farrauto, R. J.; Hoekstra, K. E; Shoup, R. D. U. S. Patent 3,870,658; 1975. Yu Yao, Y. F. J. Catal. 1975, 39, 104. Kummer, J. T. Adv. Chem. Series 1975, 143, 178. Severino, F.; Brito, J .; Carias 0.; Laine,J. J. C atal. 1986, 102, 172. Severino, F .; Laine, J. Ind. Eng. Chem. Prod. Res. Dev. 1983, 22, 396. Laine, J .; Albomoz, A.; Brito, J .; Carias, 0.; Castro, G.; Severino, F.; Valera, D. In Catalysis and Automotive Pollution Control; Crucq, A. and Frennet, A. Ed.; Elsevier, Amsterdam, 1987; pp 387-393. Laine, J.; Severino, F. Appl. Catal. 1990, 65, 253. L6pez Agudo,A., Palacios, J. M., Fierro, J .L.G., Laine, J ., and Severino, F. Appl. Catal. 1992, 91, 43. Bijsterbosch, J. W.; Kapteijn, F .; Moulijn, J. A. J. Mol. Catal. 1992, 74, 193. Dekker, F. H. M.; Dekker, M. C .; Bliek, A.; Kapteijn, F.; Moulijn, J. A. Catal. Today 1994, 20, 409. Stegenga, S.; van Soest, R.; Kapteijn, F.; Moulijn, J. A. Appl. Catal. B 1993, 2, 257. 20. 21. 22. 23. 24. 25. 26. 27 28. 29. 30. 31. 32. 33. 34. 35. 36. 173 Dekker, N. J. J.; Hoom, J. A. A.; Stegenga, S.; Kapteijn, F.; Moulijn, J. A. AIChE J. 1992, 38, 385. Kapteijn, F.; Stegenga, S.; Dekker, N. J. J .; Bijsterbosch, J. W.; Moulijn, J. A. Catalysis Today 1993, 16, 273. Rastogi, R. P.; Singh, G.; Dubey, B. L.; Shukla, C. S. J. Catal. 1980, 65, 25. Chien, C.-C.; Chuang, W.—P.; Huang, T.-J. Appl. Catal. A. 1995, I31, 73. Tarasov, A. L.; Osmanov, M. 0.; Shvets, V. A.; Kazanskii, V. B. Kinet. Catal. 1990, 31, 565. McCabe, R. W.; Mitchell, J. Ind. Eng. Chem. Prod. Res. Dev. 1983, 22, 212. Rajesh, H.; Ozkan, U. S. Ind. Eng. Chem. Res. 1993, 32, 1622. Heyes, C. J .; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol. 1982, 32, 1034. Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. T echnol. 1988, 22, 557. Stegenga, S.; Dekker, N.; Bijsterbosch, J .; Kapteijn, F.; Moulijn, J .; Belot, G.;Roche, R. In Catalysis and Automotive pollution Control I]; Crucq, A. Ed.; Elsevier, Amsterdam, 1991; pp 353-369. Stegenga, S. Ph. D. Thesis, University of Amsterdam, The Netherlands, 1991. Fattakhova, Z. T.; Ukharskii, A. A.; Shiryaev, P. A.; Berrnan, A. D. Kinet. Katal. 1986, 27, 884. Prasad, R.; Kennedy, L. A.; Ruckenstein, A. E. Combust. Sci. Technol. 1980, 22, 271. Dadyburjor, D. B; Jewur, S. S.; Ruckenstein, A. E. Catal. Rev.- Sci. Eng. 1979, 19, 293. Powder Diffraction File, Inorganic Phases, JCPDS, 1983. Klug, H. P.; Alexander, L. E., X -Ray Diffraction Procedures for Polycrystalline and Amorphous Materials, lst.Ed. Wiley, New York, 1954. Rosencwaig, A. and Wertheim, G. K. J. Electron. Spectrosc. Relat. Phenom. 1972, 1, 493. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 174 Wallbank, B., Johnson, C. E, and Main, 1. G. J. Electron. Spectrosc. Relat. Phenom. 1974, 4, 263. Strohmeier, B. R.; Leyden, D. E; Field, R. S.; Hercules, D. M. J. Catal. 1985, 94, 514. Okamoto, Y.; Fujii, M.; Imanaka, T.; Teranishi, S. Bull. Chem. Soc. Jpn. 1976, 49, 859. ' De Angelis, B. A. J. Electron Spectrosc. Relat. Phenom. 1976, 9, 81. Scierka, S. J.; Proctor, A.; Houalla, M.; Fiedor, J. N.; Hercules, D. M. Surf Interface Anal. 1993, 20, 901. ’ Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269. Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612. Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129. Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29. ‘GOOGLY’ written by Dr. Andrew Proctor, University of Pittsburgh, PA. Capece, F. M.; Di Castro, V.; Furlani, C.; Mattogno, G.; Fragale, C.; Gargano, M.; Rossi, M. J. Electron Spec. Rel. Phenom. 1982, 27, 1 l9. Friedman, R. M.; Freeman, J. 1.; Lytle, F. W. J. Catal. 1978, 55, 10. Park, P. W.; Ledford, J. S. submitted to J. Phys. Chem. Lo Jacono, M.; Schiavello, M. in “Preparation of Catalysis” Delmon, E; Jacobs, P. A.; Poncelet, G. Eds. pp 474-487. Elsevier, Amsterdam, 1976. Park, P. W.; Ledford, J. S. to be published. Chapter 7 The Influence of Surface Structure on the Catalytic Activity of Cerium Promoted Copper Oxide Catalysts on Alumina: Oxidation of Carbon Monoxide and Methane 7.1. Abstract X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) have been used to characterize a series of Cu/Ce/Al203 catalysts. Catalysts were prepared by incipient wetness impregnation using metal nitrates and alkoxides precursors. Catalyst loadings were held constant at 12 wt.% CuO and 5.1 wt.% CeO2. Mixed oxide catalysts were prepared by impregnation of cerium first, followed by copper. The information obtained from surface and bulk characterization has been correlated with CO and CH4 oxidation activity of the catalysts. Cu/Al203 catalysts prepared using Cu(II) nitrate (CuN) and Cu(II) ethoxide (CuA) precursors consist of a mixture of copper surface phase and crystalline CuO. The CuA catalyst shows higher dispersion, less crystalline CuO phase, and lower oxidation activity for CO and CH4 than the CuN catalyst. For Cu/Ce/Al203 catalysts, Ce has little effect on the dispersion and crystallinity of the copper species. However, Cu impregnation decreases the Ce dispersion and increases the amount of crystalline CeO2 present in the catalysts, particularly in catalysts prepared using the CeA 175 176 support. Cerium addition dramatically increases the C0 oxidation activity, however, it has little effect on C114 oxidation. 7.2. Introduction Copper oxide is a well known component of catalysts for CO [1-6], hydrocarbon [1-3,7], chlorinated hydrocarbon [8] and alcohol oxidation [9,10] as well as for NO" [11- 13] and $02 reduction [13]. These air purification reactions using heterogeneous catalysts have been important topics due to their application to automobile exhaust converters and industrial waste incinerators. Copper oxide based catalysts have been considered as suitable substitutes for noble metal catalysts in emission control applications due to their high catalytic activity for the purification reactions, tolerant to 802 [3] and refractory to high temperature [14]. Several researchers reported that copper oxide based catalysts show similar activity to noble metal catalysts for CO oxidation [2], butanal and mercaptan oxidation [15], and reduction ofNO by CO [13]. Rare earth oxide additives have been widely applied as textural and structural promoters for supported noble metal catalysts. Cerium addition prevents y-alumina from sintering at high temperature and improves the dispersion and thermal stability of noble metals [16-20]. In addition, cerium oxide has been added to noble metal catalysts for treatment of automobile exhaust to promote the water-gas shift reaction [21,22], to suppress the CO inhibition effect [23-25], and to remove C0, hydrocarbons, and N0x simultaneously using its oxygen storage ability [17,18,26-29]. 177 Previous studies of rare earth promoters have focused primarily on their effects on the structure and activity of noble metal-based catalysts. The study of the promoters effect on supported transition metal oxide catalysts has been relatively limited. Recently cerium oxide has been studied both as a promoter [30] and as a support [31] for oxidation reactions over copper oxide based catalysts. Agarwal et a1. [30] found that a ceria promoted hopcalite catalyst has a higher initial activity for hydrocarbon oxidation than a Pt catalyst. Bedford and LaBarge [3 1] showed that copper-chromium oxide impregnated on high surface area ceria has three-way catalyst behavior and retains catalyst activity after aging at high temperature. Flytzani-Stephanopoulos and coworkers [32-34] reported recently that a Cu-Ce-O composite catalyst shows high activity for CO and CI-Ia oxidation and high resistance to water poisoning. They attributed the enhanced catalytic activity and stability to the strong interaction of copper and cerium oxides. The effect of a CeO2 promoter on alumina supported copper oxide catalysts has been reported similar to the effect it has on noble metal catalysts [35,36]. The authors reported that the addition of Ce02 enhances redox behavior of copper ion, increases dispersion of copper oxides, provides surface oxygen, and improves oxygen storage capacity of catalyst. While, cerium oxide promoted copper oxide based catalysts have recently drawn an attention of many researchers, the effect of cerium oxide on the surface structure and reactivity of copper catalysts has not been studied extensively. The present work is part of a broad study to investigate structure-reactivity correlation for rare earth oxide promoted transition metal oxide based emission control catalysts. In this paper, X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) have been used to determine the effect 178 of catalyst precursors and cerium promotion on the chemical state and dispersion of copper oxide phases in a series of Cu/Ce/Al203 catalysts. The information derived from these techniques is correlated with CO and CH4 oxidation activity to develop a more complete understanding of Cu/Ce/A12O3 catalysts. 7.3. Experimental Catalyst Preparation. Catalysts were prepared by pore volume impregnation of y- alumina (Cyanamid, surface area = 203 mz/g, pore volume = 0.6 mL/g). The alumina was finely ground (< 230 mesh) and calcined in air at 500 °C for 24 h prior to impregnation. The Ce/Al2O3 catalysts were prepared using a deionized water solution of ammonium cerium(IV) nitrate (Mallinckrodt Inc, Analytical Reagent) or an ethanol solution of cerium(IV) methoxyethoxide (Gelest Inc., l8-20% cerium methoxyethoxide in methoxyethanol). The impregnated sample derived from the aqueous cerium nitrate solution (designated “CeN”) was dried in air at 120 °C for 24 h and calcined in air at 500 °C for 16 h. The sample prepared using the ethanol cerium methoxyethoxide solution (designated “CeA”) was impregnated and subsequently dried at room temperature for 48 h in a N2 purged glove bag prior to further drying in air at 120 °C for 24 h and calcination in air at 500 °C for 16 h. The Ce loading was held constant at 5.1 wt.% CeO2. The Cu/Al203 catalysts were prepared with a deionized water solution of copper(II) nitrate (Columbus Chemical Industries Inc, ACS Grade) or an aqueous diethylenetriamine solution (DETA/water ratio = 1/3 by volume) of copper(II) ethoxide. DETA was 179 necessary due to the limited solubility of c0pper(II) ethoxide in water and common organic solvents. Since an aqueous DETA solution was used to dissolve copper(II) ethoxide, a Cu-DETA complex should be formed in the impregnating solution [37]. The impregnated samples derived from the aqueous copper nitrate solution (designated “CuN”) and catalysts prepared with the aqueous DETA copper ethoxide solution (designated “CuA”) were dried and calcined under the same conditions as used for CeN and CeA samples respectively. The Cu loading was held constant at 12 wt.% CuO. Mixed metal oxide catalysts were prepared by step-wise impregnation of Ce first, followed by Cu. The cerium modified alumina carrier was dried and calcined prior to introduction of copper. After the addition of copper, the catalysts were dried and calcined again at the same conditions described above depending on the choice of precursors. Catalysts prepared using nitrate and alkoxide precursors will be symbolized using "N" and "A", respectively. Standard Materials. CuO and CeO2 were prepared by calcining copper(II) nitrate and ammonium cerium(IV) nitrate in air at 500 °C for 16 h. CuAl2Oa was prepared by calcining stoichiometric amounts of the respective nitrates in air at 1000 °C for 24 h. CeAlO3 was prepared by reduction of a dried CeA sample at 800 °C in H2 (AGA, 99.99%) for 8 h. XRD patterns of the standard compounds matched the appropriate Powder Diffraction File [3 8]. BET Surface Area. Surface area measurements were performed using a QuantaChrome Quantasorb Jr. Sorption System. Approximately 0.1 g of catalyst were outgassed in a N2/He mixture (5% N2) at 350 0C for l h prior to adsorption 180 measurements. The measurements were made using relative pressures of N2 to He of 0.05, 0.08, and 0.15 (N2 surface area = 0.162 nmz) at 77 K. The addition of copper and/or cerium to the alumina carrier decreased the BET surface area by a maximum of 20 %. X-Ray Diffraction. X-ray powder diffraction patterns were obtained with a Rigaku XRD diffractometer employing Cu Ka radiation (2» = 1.5418 A). The X-ray was operated at 45 kV and 100 mA. Diffraction patterns were obtained using a scanning rate of 0.5 deg/min (in 20) with divergence slit and scatter slit widths of 1°. Samples were run as powders packed into a glass sample holder having a 20 x 16 x 0.5-mm cavity. The mean crystallite sizes (d) of the CuO and CeO2 particles were determined from XRD line broadening measurements using the Scherrer equation [39]: d: KK/BcosG (1) where it is the X-ray wavelength, K is the particle shape factor, taken as 0.9, and B is the fill] width at half maximum (fwhm), in radians, ofthe CuO <111> or CeO2 line. Quantitative X-ray diffraction data were obtained by comparing Cu0<111>/Al203<400> or Ce02 <1 1 l>/A1203<440> intensity ratios measured for catalyst samples with intensity ratios measured for physical mixtures of pure CuO or CeO2 and y-Al203. This method assumed that copper or cerium addition did not disrupt the 7- 181 A1203 spinel and subsequently affect the intensity of the A1203 lines. The error in this method was estimated to be i 20 %. XPS Analysis. XPS data were obtained using a Perkin-Elmer Surface Science instrument equipped with a magnesium anode (1253.6 eV) operated at 300 W (15 kV, 20 mA) and a 10-360 hemispherical analyzer Operated with a pass energy of 50 eV. The instrument typically operates at pressures near 1 x 10‘8 torr in the analysis chamber. Spectra were collected using a PC137 board interfaced to a Zeos 386SX computer. Samples were analyzed as powders dusted onto double-sided sticky tape or spray coated onto a quartz slide using a methanol suspension of the catalyst. Binding energies for the catalyst samples and standard compounds which contained Al were referenced to the Al 2p peak (74.5 eV). The binding energies for standard compounds that did not contain Al were referenced to the C ls line (284.6 eV) of the carbon overlayer. XPS binding energies were measured with a precision of i 0.2 eV, or better. Reduction of copper [40-42] and cerium [43-46] species during XPS experiments has been reported and attributed to several factors such as X-ray flux, X-ray dose, temperature, and pressure. In order to minimize the effect of photoreduction on the results, all‘samples were analyzed using the same distance between X-ray source and sample (~3/4 inch) and minimum data acquisition time (5 min. scan for Cu 2p3r2 spectra). No significant photoreduction of copper species was observed using these experimental conditions. However, Ce 3d region had to be scanned for extended time (~ 600 min.) in order to obtain reasonable quality spectra. Therefore Ce XPS data were used only for quantitative analysis to measure Ce dispersion on alumina. 182 Quantitative XPS Analysis. It has been shown by Defosse et al. [47] that one may calculate the theoretical intensity ratio 1°./1°, expected for a supported phase (p) atomically dispersed on a carrier (5). An extension of the Defosse model proposed by Kerkhof and Moulijn [48] has been used in the present investigation. The photoelectron cross sections and the mean escape depths of the photoelectrons used in these calculation were taken from Scofield [49] and Penn [50], respectively. For a phase (p) present as discrete particles, the experimental intensity ratio I,/I. is given by the following expression: 141. = 1°./1°. rI-exp (-d/7~p)] / do. (2) where 1°P/I°, is the theoretical monolayer intensity ratio, d is the length of the edge of the cubic crystallites of the deposited phase and it? is the mean escape depth of the photoelectrons in the deposited phase. CO Oxidation Activity. Measurement of CO oxidation activity was performed in a flow microreactor. Approximately 0.03 g of catalyst were supported on a glass hit (70 - 100 um) and the temperature was measured with a K-type thermocouple located just above the catalyst bed. The reactor was heated by a tube fumace (Lindberg) with temperature being controlled within 1°C by an Omega CN 1200 temperature controller. Reactant gas flow rates were held constant with Brooks 5850 mass flow controllers. Product gases were analyzed with a Varian 920 gas chromatograph equipped with a TCD and interfaced to a Hewlett-Packard 3394A integrator. Reaction products were separated on a 6 ft 60/80 mesh Carbosieve S-II column. Prior to the first activity measurement, the 183 catalyst was pretreated with a mixture of 5% O2/He (99.5 % purity for 02, 99.995% purity for He, AGA Gas Co.) stream (143 cc/min.) at 350 °C for 1 hr to remove any impurities absorbed on the surface during catalyst preparation and storage. CO oxidation reactions were performed with a constant flow rate (80 cm3/min) of 4.8% CO/9.8% 02/85.4% He gas mixture (AGA, purity > 99.99%) in the temperature range of 100-220 °C. A11 activity measurements were obtained under steady-state conditions at conversions less than 15%. CH; Oxidation Activity. Methane oxidation reactions were performed with a constant flow (15 cm3/min) of 0.98% CHa/5.25% 02/93.77% He gas mixture (AGA, purity > 99.99%) in the temperature range of 360-420 °C. Approximately 0.1 g catalyst were charged into the same type of microreactor used for C0 oxidation measurements. Water produced during methane oxidation was frozen downstream from the reactor in a trap maintained < -40 °C with a mixture of ethanol and dry ice. No partial oxidation product was observed. All activity measurements were obtained under steady-state conditions at conversions less than 15%. 7.4. Results and Discussions Chemical State and Dispersion of Cerium/A lumina Catalysts. The cerium oxide chemical state and dispersion of CeN and CeA catalysts have been reported previously [46]. In summary, the Ce dispersion for the CeN and CeA was found to be poor and Similar each other (Table 7.2). For the CeA, only 0.3 wt.% cerium oxide was detectable 184 by XRD (Figure 7.1b), indicating that most of the cerium oxide is present as an amorphous phase. A large amount of crystalline Ce02 was observed for CeN (Figure 7.1c), although a low loading ceria and high surface area alumina were used in this study. This result does not agree with the results of other researchers [20,51]. For cerium catalysts with < 2.6 umol ceria/m2 alumina, which is corresponding to < 9 wt.% Ce0j200 m2 A1203, Ce species were present as CeA103 dispersed surface phase and/or as small CeO2 crystallites which were not detectable by XRD. This discrepancy may be readily understood when one considers that the cerium precursor used in this study (Ce‘f) was different from the one used in other studies (Ce3i). The limited interaction between the Ce4+ precursor and alumina support may produce less Ce3+ ion occupancy in cation vacancies in alumina lattice due to charge conflict. This leads to the formation of CeO2 during sample preparation. Chemical State of Copper/Alumina Catalysts. The XRD patterns obtained for CuA and CuN catalysts are shown in Figure 7.1d and 7.1e. The XRD pattern of the CuN catalyst shows intense peaks characteristic of CuO, however, weak CuO lines are observed in the diffraction pattern of the CuA catalyst. Quantitative XRD measurements indicate that the CuN catalyst contains 5.5 wt.% crystalline CuO phase, while the CuA catalyst contains only 0.8 wt.% crystalline CuO (Table 7.1). For alumina supports with surface area comparable to that used in this study, previous results [40,52] suggest that up to 8 wt.% Cu can be incorporated into the alumina lattice as a Cu-Al203 surface phase when copper nitrate precursor is used to prepare the catalyst. Thus, our finding that the 185 CuO <1 1 1) CuO CuO hv*hd (d) (C) #22. :fi‘: ‘2‘ ; W) Figure 7.1. A120, <311> 15.1203 <220> <222> l l l l 25 30 35 40 Degrees (20) CuA, and (e) CuN catalysts. XRD patterns measured for (a) alumina support, (b) CeA, (c) CeN, (d) 186 Table 7.1. Concentration of Crystalline‘ Phases in Cu/Ce/A12O3 Catalysts Calculated from Quantitative XRD Data. Crystalline Phase (wt.%) Catalysts Ce02 CuO CeN 4.6 b CeA 0.3 b CuN b 5.5 CuA b 0.8 CuNCeN 4.7 6.0 CuACeN 4.1 0.6 CuNCeA 3 .4 5.5 CuACeA 1.5 0.8 a - Valid for crystalline phases with particle sizes > 3.0 nm. b - Catalysts do not contain the component. presence of 6.5 wt.% of CuO as a dispersed phase for CuN catalyst is in good agreement with previous work. In addition, the XPS Cu 2p3/2 binding energies measured for CuN and CuA catalysts (935.1 eV) are consistent with the value measured for CuAl204 (935.0 eV) and higher than the value measured for CuO (933.9 eV). Thus, the XRD and XPS results indicate that copper is present predominantly as a dispersed copper surface phase on the alumina support. Dispersion of Copper/Alumina Catalysts. Particle sizes determined from XRD line broadening calculations and from Cu/Al XPS intensity ratios for CuN and CuA catalysts are shown in Table 7.2. CuO particle sizes determined using XRD line broadening calculations are identical for CuN and CuA catalysts. While, the copper species particle size determined for the CuA catalyst using XPS is smaller than the value 187 Table 7.2. Particle Size of Cerium and Copper Phases Determined from XRD Line Broadening Calculations and XPS Intensity Ratios. Particle Size (nm) Catalysts Cerium Copper XRD XPS XRD XPS CeN 5.1 1.6 -' -‘ CeA 4.5 1.6 -' -‘ CuN -“ -‘ 30 1.9 CuA -“ -" 30 1 . 1 CuNCeN 5.1 3.0 32 2.1 CuACeN 5.4 3.6 32 1.1 CuNCeA 4.1 4.1 32 2.2 CuACeA 4.6 4.7 30 1.1 a - Catalysts do not contain the component. obtained for the CuN catalyst. XPS and XRD analyses of the CuA catalyst indicate the formation of less crystalline and more dispersed copper species on the alumina support compared to CuN catalyst. From quantitative XRD data, approximately 11.2 wt.% CuO in CuA which is not detected by XRD can be assigned as a dispersed copper species on the alumina surface. However, the relatively poor Cu dispersion compared to monolayer noted for CuA catalyst indicates that it is not entirely due to a Cu-Al203 surface phase, since copper surface phase is normally considered to be highly dispersed. It can be assumed that a poorly dispersed amorphous copper oxide phase is partially formed on CuA catalyst, which is not detectable by XRD. Uchikawa and Mackenzie [37] have proposed that a polymer structure of gel containing Cu-DETA unit was obtained afier a 188 solution of copper ethoxide-aqueous DETA was heated at 90 °C. Therefore, the polymer structure is possibly obtained during the drying of aqueous copper ethoxide-DETA solution and then an amorphous CuO phase may be formed subsequently in calcination step. It is also possible that Cu-DETA complexes aggregate less during catalyst preparation than the copper nitrate precursor. The CuO particle sizes of all copper catalysts determined with XRD are significantly larger than the values obtained from XPS (Table 7.2). This difference may be attributed to the limitations of X-ray diffraction. It is well known that XRD determined particle sizes are skewed to large values since XRD does not detect highly dispersed species. For the copper catalysts, a significant fraction of the Cu is present as a copper surface phase or small CuO particles (d < 3.0 nm) that are detected by XPS but not by XRD. Thus, smaller average particle sizes are expected from XPS calculations. Chemical State of Copper/Cerium’Alumina Catalysts. Figure 7.2 shows the XRD patterns obtained for Cu/Ce/Al203 catalysts. The XRD patterns show only peaks characteristic of CuO and CeO2. This is in agreement with studies have found that copper oxide is immiscible with cerium oxide [33]. For the Cu/Ce/Al203 catalysts prepared using the CeN (i.e. CuNCeN and CuACeN), Cu addition has little effect on the intensity of Ce02 XRD peaks. This can be attributed to that a large amount of poorly dispersed crystalline CeO2 (4.1 - 4.7 wt.%) is already present in CeN. For the Cu/Ce/A12O3 catalysts derived from the CeA (i.e. CuNCeA or CuACeA), Cu addition increases the intensity of CeO2 XRD peaks. The Cu addition effect on the Ce02 peak intensity is observed more Figure 189 CuO <111> CuO <11 l> CeO2 <111> O <110> ‘ (d) CeO2 Q00> (C) l (b) (a) 1 1 r 1 25 30 3 5 40 Degrees (20) Figure 7.2. XRD patterns measured for (a) CuACeA, (b) CuNCeA, (c) CuACeN, and (d) CuNCeN catalysts. 190 strongly in the diffraction pattern of CuNCeA catalyst than that of CuACeA catalyst. Quantitative XRD measurements (Table 7.1) show that Cu addition increases the amount of crystalline Ce02 from 0.3 wt.% (CeA) to 1.5 wt.% (CuACeA) and 3.4 wt.% (CuNCeA). Since cerium oxide in CeA initially was a amorphous phase [46], post loaded copper species may act as a seed to form a crystalline Ce02 during sample preparation. We attribute the more pronounced increase in crystalline Ce02 content observed for the CuNCeA catalyst to the dissolution of cerium oxide in an acidic copper nitrate solution (pH 2.2) followed by agglomeration during subsequent drying and calcination steps of the catalyst preparation. Cerium oxide is insoluble in basic impregnating solution (pH 12.4) used for CuA derived catalysts. For the Cu/Ce/Al203 catalysts, quantitative XRD measurements show that the CuO crystallinity is identical within experimental error to the value measured for analogous CuN and CuA catalysts. In addition, Ce has little effect on the XPS Cu 2p3/2 binding energies (935.0 i 0.1 eV) which were identical to the value measured for CuAl2Oa and higher than the value of CuO. This indicates that chemical state of copper species is independent of Ce promoter in this study. Dispersion of Copper/Cerium/A lumina Catalysts. The particle sizes of Cu and Ce species calculated using XRD line broadening and XPS intensity ratios are shown in Table 7.2. XRD line broadening calculations show that CeO2 particle sizes of Cu/Ce/A12O3 catalysts are not changed by Cu addition and are similar to those of analogous CeN and CeA supports. However, the cerium particle size calculated from Ce/Al XPS intensity ratios increases significantly after Cu addition. This indicates that the addition of copper 191 Species dramatically decreases the dispersion of cerium species. Since the addition of Cu increases the crystallinity of cerium oxide, we can attribute the decrease of dispersion to the agglomeration of cerium species during catalyst preparation step. In addition, we can not exclude an encapsulation phenomena involving the deposition of copper species on the surface of the cerium oxide phase, which would lead to a decrease in the Ce/Al XPS intensity ratio and dispersion subsequently. Such encapsulation effect would be more prominent for CuA due to its higher dispersion than CuN. The particle sizes of copper oxide calculated from XRD line broadening calculations are identical regardless of cerium promotion and copper precursors. The particle sizes of copper species calculated from Cu/Al XPS intensity ratio depend only on the precursors used for catalyst preparation. The absence of a significant effect of Ce addition on the CuO crystallinity and dispersion can be understood in terms of the Ce/A12O3 surface structure. Since the CeN and CeA contain a low concentration of poorly dispersed cerium oxide, most of the y-Al203 surface is exposed to the Cu introduction. This exposed alumina dominates the alteration of the copper structure in the catalysts. Effect of Catalyst Structure on CO Oxidation Activity. Table 7.3 shows the specific turnover numbers (TON) and activation energies for CO oxidation over Cu/Al203 and Cu/Ce/Al203 catalysts calculated at 160 °C using XPS estimates of the Cu dispersion. The results of CeN and CeA catalysts were not shown here due to their negligible CO oxidation activity compared to the copper catalysts. The CuN catalyst shows higher TON and lower activation energy than the values of the CuA catalyst. Comparison of the CO 192 Table 7.3. Turn Over Numbers‘ and Activation Energies” of Cu/Ce/A1203 Catalysts for CO and CH4 Oxidation. CO Oxidation CHa Oxidation Catalysts TON Activation Energy TON Activation EnergL CuN 1.1 12.0 3.7 22.7 CuA 0.3 15.4 1.2 21.6 CuNCeN 8.7 10.9 3.7 22.7 CuACeN 3.0 12.1 1.3 23.8 CuNCeA 13.7 12.6 4.5 21.9 CuACeA 1.5 12.8 1.4 22.6 a - Calculated at 160 °C and 390 °C for CO and CH4 oxidation respectively using exposed Cu species determined from XPS (s'1 x 102). b - Estmiated from Arrhenius Plots (kcal/mol) within 15 % conversion. oxidation activities measured for the CuN and CuA catalysts indicates that the catalyst containing more crystalline CuO (i.e. CuN) is more active than catalysts containing less CuO phase (CuA). This is consistent with our previous study [53], which showed that crystalline CuO was more active for C0 oxidation than a dispersed copper surface phase. The Ce promoted catalysts show increased TONs by a factor of 5 to 12 compared to unpromoted copper catalysts. The activation energies for CO oxidation calculated for cerium promoted catalysts are similar within a experimental error (12.1 i 0.9 kcanol). The higher CO oxidation TON for cerium promoted copper catalysts indicates that cerium oxide is very effective promoter when one considers that only a small amount of cerium oxide (5.1 wt.%) was used in this study. Changing crystallinity and dispersion of cerium oxide species after Cu addition indicates that copper oxide interacts strongly with cerium oxide during catalyst preparation. Therefore, cerium species included in the CuO particles 193 may lead to increased CO oxidation activity. Liu and Flytzani-Stephanopoulos [33,34] reported that Cu-Ce-O composite catalyst shows high CO oxidation activity, comparable to that of the Pt/alumina catalyst. They attributed the enhanced catalytic activity to the strong interaction of copper and cerium oxides. It also has been reported recently [3 5,36] that ceria promoter enhances the redox behavior of copper ions and produces surface oxygen ions over CuO/CeO2/y-alumina catalysts. The Cu/Ce/Al203 catalysts derived from CuN show higher CO oxidation activity than the catalyst prepared with CuA. This is consistent with the result of CuN and CuA catalysts, which shows higher CO oxidation activity for the catalyst containing more crystalline CuO. Effect of Catalyst Structure on CH4 Oxidation Activity. Table 7.3 shows the specific turnover numbers (TON) and activation energies for CH4 oxidation calculated at 390 °C. The results of CeN and CeA catalysts were not shown here due to their negligible CH4 oxidation activity compared to the copper catalysts. The TONs of the catalysts derived from CuN is about 3 times higher than the catalysts prepared with CuA. The Ce addition has little effect on the CH4 oxidation TON for copper catalysts. The activation energies are identical within experimental error (22.6 i- 0.8 kcal/mol). Since the active site of copper oxide catalyst for CH4 oxidation reaction has been assigned to the isolated copper surface phase [53], the catalyst containing more isolated copper surface phase should be more effective. In our study, CuA catalyst may contain less isolated copper surface phase than CuN catalyst, although Cu dispersion of CuA is higher than that of CuN. The formation of less isolated copper surface phase can be attributed to the formation of Cu-DETA polymer structure during catalyst preparation 194 step [37]. We propose that this polymer type Cu-DETA complex leads to the formation of an interacting copper surface phase (i.e. clustered c0pper species) and/or amorphous copper oxide which may be a less active phase than isolated c0pper species for CH4 oxidation. The no cerium promoter effect on CH4 oxidation reaction for Cu/Ce/Al203 catalysts can be explained by little interaction between poorly dispersed cerium oxide and isolated copper surface phase. 7.5. Conclusions The combined use of several techniques to investigate the effect of cerium promoter and metal precursors on the structure of Cu/Ce/A12O3 catalysts leads to the following conclusions. 1. The CuA catalyst shows enhanced Cu dispersion compared to CuN catalyst. However, CuA catalyst is less active for CO and CH4 oxidation than CuN catalyst. This has been attributed to that CuA catalyst contain less active phases than the CuN catalyst for CO and CH4 oxidation which are crystalline CuO and isolated copper surface species, respectively. 2. Ce has little effect on the dispersion and crystallinity of copper species. Whereas Cu addition decreases the Ce dispersion and increases the amount of crystalline Ce02 present in the Cu/Ce/Al203 catalysts particularly in catalysts prepared using the CeA. 195 3. Ce addition increases the CO oxidation activity of the copper catalysts. This has been attributed to that cerium oxide interacts strongly with CuO crystalline which is active phase for CO oxidation. While, the Ce may not promote the isolated copper surface species which is active site for CH.; oxidation. 10. 11. 12. 13. 14. 15. 16. 17. 18. 196 7.6. References Yao, Y. F. Yu; Kummer, J. T. J. Catal. 1977, 46, 388. Kummer J. T., Prog. Energy Combust. Sci. 1980, 6, 177. Yao, Y. F. YuJ. Catal. 1975, 39, 104. Severino, F.; Brito, J.; Cari’as, O. Laine, J. J. Catal. 1986, [02, 172. Boon, A. Q. M.; van Looij, F.; Geus, J. W. J. Mol. Catal. 1992, 75, 277. Lopez Agudo, A.; Palacios, J. M.; Fierro, J .L.G.; Laine, J.; Severino, F. Appl. Catal. 1992, 91, 43. Boon, A. Q. M.; Huisman, H. M.; Geus, J. W. J. Mol. Catal. 1992, 75, 293. Subbanna, P.; Greene, H.; Desal, F. Environ. Sci. T echnol. 1988, 22, 557. Ozkan, U. S.; Kueller, R. F.; Moctezuma, E. Ind. Eng. Chem. Res. 1990, 29, 1136. Rajesh, H.; Ozkan, U. S. Ind. Eng. Chem. Res. 1993, 32, 1622. Kapteijn, F.; Stegenga, S.; Dekker, N. J. J.; Bijsterbosch, J. W.; Moulijn, J. A. Catal. Today 1993, 16, 273. Huang, T.-J.; Yu, T.-C. Appl. Catal. 1991, 71, 275. Goetz, V. N.; Sood, A.; Kittrell, J. R. Ind. Eng. Chem. Prod. Res. Develop. 1974, I3, 110. R. Prasad, L. A. Kennedy and E. Ruckenstein, Catal. Rev-Sci. Eng, 1984, 26, l. Heyes, C. J.; Irwin, J. G.; Johnson, H. A.; Moss, R. L. J. Chem. Tech. Biotechnol. 1982, 32,1025. Yao, Y. F. Yu; Kummer, J. T. J. Catal. 1987, 106, 307. Su, E. C.; Montreuil, C. N.; Rothschild, W. G. Appl. Catal. 1985, 17, 75. Gandhi, H. S.; Piken, A. G.; Shelef, M.; Delosh, R. G. SAE paper No. 760201, 1976. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 197 Dictor, R.; Roberts, S. J. Phys. Chem. 1989, 93, 5846. Le Normand, F .; Hilaire, L.; Kili, K.; Krill, G.; Maire, G. J. Phys. Chem. 1988, 92, 2561. Kim, G. Ind. Eng. Chem. Prod. Res. Dev. 1982, 21, 267. Herz, R. K.; Sell, J. A. J. Catal. 1985, 94, 166. Oh, S. H. J. Catal. 1990, 124, 477. Yao, Y. F. YuJ. Catal. 1984, 87, 152. Oh, S. H.; Eickel, C. C. J. Catal. 1988, 112, 543. Yao, H. C.; Yu-Yao, Y. F. J. Catal. 1984, 86, 254. Herz, R. K. Am. Chem. Soc. Symp. Ser. 1982, 178, 59. Jin, T.; Okuhara, T.; Mains, G. J.; White, J. M. J. Phys. Chem. 1987, 91, 3310. Cho, B. K.; Shanks, B. H.; Bailey, J. E. .I. Catal. I989, 115, 486. Agarwal, S. K.; Spivey, J. J.; Butt, J. B. Appl. Catal. 1992, 81, 239. Bedford, R. E; LaBarge, W. J. US. Patent 5,063,193 Nov. 5, 1991. Liu, W.; Sarofim, A. F; FIytzani-Stephanopoulos Appl. Catal. B. 1994, 4, 167. Liu, W.; Flytzani-Stephanopoulos, M. J. Catal. 1995, 153, 304. Liu, W.; Flytzani-Stephanopoulos, M. J. C atal. 1995, 153, 317. Lu, G.; Wang, R. Cuihua Xuebao 1991, I2, 314. Tian, Y.; Fu, Y.; Lin, P. Cuihua Xuebao 1994, I5, 189. Uchikawa, F .; Mackenzie, J. D. J. Mater. Res. 1989, 4, 787. Powder Diffraction File, Inorganic Phases, JCPDS, 1983. Klug, H. P.; Alexander, L. E, X-Ray Diffraction Procedures for Polycrystalline and Amorphous Materials, lst.Ed. Wiley, New York, 1954. 40. .41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 198 Strohmeier, B. R.; Leyden, D. E; Field, R. S.; Hercules, D. M. J. Catal. 1985, 94, 514. Wallbank, B., Johnson, C. E, and Main, 1. G. J. Electron. Spectrosc. Relat. Phenom. 1974, 4, 263. Rosencr-s' g, A. and Wertheim, G. K. J. Electron. Spectrosc. Relat. Phenom. 1972, 1. -, . 3. Al-than, Z. T.; Hashemi, T.; Hogarth, C. A. Spectrochimica Acta. 1989, 44b, 205. Paparazzo, E. Surf Sci. 1990, 234, L253. Paparazzo, E; Ingo, G. M.; Zacchetti, N. J. Vac. Sci. T echrrol. A. 1991, 9, 1416. Park P.W.; Ledford J. S. accepted to be published in Langmuir Deffosse, C.; Canesson, D.; Rouxhet, P. G.; Delmon, B. J. Catal. 1978, 51, 269. Kerkhof, F. P. J. M.; Moulijn, J. A. J. Phys. Chem. 1979, 83, 1612. Scofield, J. H. J. Electron. Spectrosc. Relat. Phenom. 1979, 8, 129. Penn, D. R. J. Electron. Spectrosc. Relat. Phenom. 1976, 9, 29. Shyu, J. 2.; Weber, W. H.; Gandhi, H. S. J. Phys. Chem. 1988, 92, 4964. Friedman, R. M.; Freeman, J. J.; Lytle, F. W. J. Catal. 1978, 55, 10. Park, P.W.; Ledford. J. S. to be published HIGH ‘1 c N STATE UNIV. LIBRARIES llllllllllllllllllllllIlllllllllllllllllllllllllllllllll 1293014217313 I 3