<%BANNER%>

Nanoparticle Oxides as Catalyst Supports for the Oxidative Coupling of 4-Methylpyridine over Palladium

Permanent Link: http://ufdc.ufl.edu/UFE0022849/00001

Material Information

Title: Nanoparticle Oxides as Catalyst Supports for the Oxidative Coupling of 4-Methylpyridine over Palladium
Physical Description: 1 online resource (140 p.)
Language: english
Creator: Neal, Luke
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bipyridine, catalyst, methylpyridine, nanoparticle, oxidative, palladium, picoline
Chemical Engineering -- Dissertations, Academic -- UF
Genre: Chemical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The use of nanoparticles as supports for palladium catalysts was studied for application to the oxidative coupling of 4-methylpyridine via C-H activation and C-C coupling. The product of this oxidative coupling is 4,4?-dimethyl-2,2?-bipyridine which is a useful but expensive chelating agent. Although palladium on alumina are traditionally poor catalysts in this reaction, an excellent palladium on nanoparticle alumina catalyst was developed which exceeds the activities of the traditional palladium on carbon catalyst. This nanoparticle catalyst only gives high activities when prepared by precipitation. Other nanoparticle oxide supports were found that give highly active palladium catalyst for the oxidative coupling reaction, including nanoparticle magnesia and zirconia. Additionally, some traditional porous supports were found to result in active catalysts, including gamma-alumina, titania, and silica, but only the porous titania had activities comparable to the best performing nanoparticle-supported catalysts. It was also determined that zirconia and ceria are effective additives in several of the palladium catalysts. In general, oxides with strong palladium-support interactions, and/or very high surface areas were the most active. This is attributed to electrophilic palladium oxide particles that can form on supports with strong metal-support interactions or to the numerous low-coordination edge or corner sites of the high surface area supports. The electrophilicity of these particles was determined by x-ray photoelectron spectroscopy (XPS). It was also shown that basic sites on the support are not necessary for forming an active catalyst, but that a high concentration of acidic support sites tends to give more active catalysts. Optimization studies of the catalysts were also performed. It was determined that 2.5, 5, and 10% loadings of palladium on nanoparticle alumina give the same product yield per unit weight of catalyst, which indicates a structure sensitive reaction. Lowering the Pd loading to 1% decreases the yield palladium loadings of ~ 2.5% were, thus, determined to be optimal on the nanoparticle alumina support. Additionally it was shown that titration to a pH of 11 gives more consistent results compared with titration to a simple stoichiometric excess (based on the amount of palladium(II) nitrate used).
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Luke Neal.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Hagelin-Weaver, Helena E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022849:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022849/00001

Material Information

Title: Nanoparticle Oxides as Catalyst Supports for the Oxidative Coupling of 4-Methylpyridine over Palladium
Physical Description: 1 online resource (140 p.)
Language: english
Creator: Neal, Luke
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bipyridine, catalyst, methylpyridine, nanoparticle, oxidative, palladium, picoline
Chemical Engineering -- Dissertations, Academic -- UF
Genre: Chemical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The use of nanoparticles as supports for palladium catalysts was studied for application to the oxidative coupling of 4-methylpyridine via C-H activation and C-C coupling. The product of this oxidative coupling is 4,4?-dimethyl-2,2?-bipyridine which is a useful but expensive chelating agent. Although palladium on alumina are traditionally poor catalysts in this reaction, an excellent palladium on nanoparticle alumina catalyst was developed which exceeds the activities of the traditional palladium on carbon catalyst. This nanoparticle catalyst only gives high activities when prepared by precipitation. Other nanoparticle oxide supports were found that give highly active palladium catalyst for the oxidative coupling reaction, including nanoparticle magnesia and zirconia. Additionally, some traditional porous supports were found to result in active catalysts, including gamma-alumina, titania, and silica, but only the porous titania had activities comparable to the best performing nanoparticle-supported catalysts. It was also determined that zirconia and ceria are effective additives in several of the palladium catalysts. In general, oxides with strong palladium-support interactions, and/or very high surface areas were the most active. This is attributed to electrophilic palladium oxide particles that can form on supports with strong metal-support interactions or to the numerous low-coordination edge or corner sites of the high surface area supports. The electrophilicity of these particles was determined by x-ray photoelectron spectroscopy (XPS). It was also shown that basic sites on the support are not necessary for forming an active catalyst, but that a high concentration of acidic support sites tends to give more active catalysts. Optimization studies of the catalysts were also performed. It was determined that 2.5, 5, and 10% loadings of palladium on nanoparticle alumina give the same product yield per unit weight of catalyst, which indicates a structure sensitive reaction. Lowering the Pd loading to 1% decreases the yield palladium loadings of ~ 2.5% were, thus, determined to be optimal on the nanoparticle alumina support. Additionally it was shown that titration to a pH of 11 gives more consistent results compared with titration to a simple stoichiometric excess (based on the amount of palladium(II) nitrate used).
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Luke Neal.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Hagelin-Weaver, Helena E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022849:00001


This item has the following downloads:


Full Text

PAGE 1

NANOPARTICLE OXIDES AS CA TALYST SUPPORTS FOR THE OXIDATIVE COUPLING OF 4-METHYL PYRIDINE OVER PALLADIUM By LUKE MICHAEL NEAL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 Luke M Neal 2

PAGE 3

To my Mom and Dad 3

PAGE 4

ACKNOWLEDGMENTS Many people have helped me arrive at this point in my research. First, I thank my advisor, Dr. Helena Hagelin-Weaver, for her technical guidance and patient support of my research. My committee Dr. Gar Hoflund, Dr. Da vid Hahn, and Dr. Kirk Ziegler also provided important guidance in my research and writi ng. Group members Sam Jones, Mike Everett, Justin Dodson and Daniel Hernandez provided valu able assistance to me in the lab over the years. The American Chemical Society Petrol eum Research Fund supported portions of this research. I am grateful to Prof. Angela Lindner (Environmental Engineering Sciences, EES) for her fruitful collaboration and fo r providing equipment used to initiate experiments during this research. Dr. David Mazyck (EES ) provided an activated carbon sa mple used during the research project. Melani Withana and Charlotta Brodin pe rformed the initial work on this research project. The Particle Engineering Research Ce nter and the Major Analytical Instrumentation Center at the University of Florida and their staff, particularly Gill Brubaker and Dr. Kerry Siebein provided valuable assistance. I also th ank my former colleagues and supervisors at the USDA National Center for Agriculture Utilizat ion Research (NCAUR) in Peoria, Illinois for their encouragement in my pursuit of additional studies and resear ch in the field of chemical engineering. Finally, I thank my parents for their unwavering support throughout my years of study. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................10 1 INTRODUCTION................................................................................................................. .12 1.1 Overview...........................................................................................................................12 1.2 Background.......................................................................................................................12 1.2.1 Nanoparticle Oxide Supports.................................................................................12 1.2.2 Palladium Catalysts................................................................................................13 1.2.3 Bipyridines: Applications and Synthesis................................................................14 1.2.4 Catalyst Characte rization Techniques....................................................................15 1.2.4.1 Brunauer-Emmett-Teller (BET) surface area measurements.......................15 1.2.4.2 Chemisorption measurements......................................................................16 1.2.4.3 X-ray photoelectron spectroscopy (XPS).....................................................16 1.2.4.4 X-ray diffraction (XRD)...............................................................................17 1.3 Objectives.........................................................................................................................18 2 C-H ACTIVATION AND C-C COUPLI NG OF 4-METHYLPYRIDINE USING PALLADIUM SUPPORTED ON NANOPARTICLE ALUMINA.......................................25 2.1 Introduction............................................................................................................... ........25 2.2 Experimental............................................................................................................... ......28 2.2.1 Catalysts Preparation..............................................................................................28 2.2.2 Reaction Conditions...............................................................................................29 2.2.3 Catalyst Characterization........................................................................................30 2.3 Results and Discussion..................................................................................................... 31 2.3.1 Effects of Catalyst Preparation Method.................................................................31 2.3.2 Catalyst Support Effects.........................................................................................32 2.3.3 Palladium Surface Areas........................................................................................34 2.3.4 Impact of Reactant Quality.....................................................................................36 2.3.5 Homogeneous Versus Heterogeneous Catalysis....................................................38 2.3.6 Precipitation Base...................................................................................................40 2.4 Conclusions.......................................................................................................................40 5

PAGE 6

3 NANOPARTICLEAND POROUS-META L-OXIDE-SUPPORTED PALLADIUM CATALYST FOR THE OXIDATIVE COUPLING OF 4-METHYLPYRIDINE................48 3.1 Introduction............................................................................................................... ........48 3.2 Experimental............................................................................................................... ......51 3.2.1 Catalyst Preparation................................................................................................51 3.2.2 Reaction Conditions and Product Recovery...........................................................52 3.2.3 Catalyst Characterization........................................................................................53 3.3 Results and Discussion..................................................................................................... 55 3.3.1 Catalytic Activity....................................................................................................55 3.3.2 Catalyst and Metal Surface Areas..........................................................................56 3.3.3 Acidic and Basic Sites............................................................................................59 3.3.4 Estimated Cost of Selected Catalyst.......................................................................60 3.4 Conclusions.......................................................................................................................61 4 USE OF ZIRCONIA, CERIA, AND ZINC OXIDE AS ADDITIVES IN NANOPARTICLE-OXIDE-SUPPORTED PALLADIUM CATALYST FOR THE OXIDATIVE COUPLING OF 4-METHYLPYRIDINE.......................................................72 4.1 Introduction............................................................................................................... ........72 4.2 Experimental............................................................................................................... ......75 4.2.1 Catalyst Preparation................................................................................................75 4.2.2 Reaction Conditions...............................................................................................76 4.2.3 Carbon Monoxide Chemisorption..........................................................................76 4.3 Results and Discussion..................................................................................................... 77 4.3.1 Catalytic Activity....................................................................................................77 4.3.2 Additives.................................................................................................................77 4.3.3 Catalyst Preparation Method..................................................................................79 4.3.4 Catalyst and Metal Surface Area............................................................................80 4.4 Conclusions.......................................................................................................................80 5 CHARACTERIZATION OF PALLADIUM CATALYSTS SUPPORTED ON NANOPARTICLE METAL OXIDES FOR THE OXIDATIVE COUPLING OF 4METHYPYRIDINE...............................................................................................................84 5.1 Introduction............................................................................................................... ........84 5.2 Experimental............................................................................................................... ......87 5.2.1 Catalyst Preparation and Reaction..........................................................................87 5.2.2 Chemisorption........................................................................................................87 5.2.3 X-ray Photoelectron Spectroscopy.........................................................................88 5.2.4 X-ray Diffraction....................................................................................................88 5.2.5 Transmission Electron Microscopy........................................................................88 6

PAGE 7

5.3 Results and Discussions.................................................................................................... 89 5.3.1 Proposed Reaction Mechanism..............................................................................89 5.3.2 Transmission Electron Microscopy........................................................................94 5.3.3 X-ray Diffraction....................................................................................................95 5.4 Conclusions.......................................................................................................................97 OPTIMAZATION OF METAL-OXIDE-SUPP ORTED PALLADIUM CATALYST FOR THE OXIDATIVE COUPLING OF 4-METHYLPYRIDINE.............................................110 6.1 Introduction............................................................................................................... ......110 6.2 Experimental............................................................................................................... ....113 6.2.1 Catalysts preparation............................................................................................113 6.2.2 Reaction Conditions.............................................................................................114 6.2.3 Palladium Dispersion............................................................................................114 6.3 Results and Discussion...................................................................................................11 5 6.3.1 Palladium Loading................................................................................................115 6.3.3 Reactant-to-Catalyst Ratios..................................................................................116 6.3.4 Titration pH..........................................................................................................117 6.3.5 Pre-reaction Treatment.........................................................................................118 6.4 Conclusions.....................................................................................................................119 CONCLUSIONS.................................................................................................................... ......124 7.1 Summary of Results........................................................................................................124 7.2 Nanoparticles Oxides as Palladium Support...................................................................125 7.2.1 Viability................................................................................................................125 7.2.2 Traditional Porous Supports.................................................................................126 7.2.3 Nanoparticle vs. Porous Supports.........................................................................126 7.3 Factors in Catalyst Activity and Deactivation................................................................127 7.3.1 Palladium Dispersion and Surface Area...............................................................127 7.3.2 Structure Sensitivity.............................................................................................128 7.3.3 Support Properties................................................................................................129 7.3.4 Deactivation..........................................................................................................130 7.4 Final Remarks.................................................................................................................131 7.4.1 Future Work..........................................................................................................131 7.4.2 Broader Impact.....................................................................................................132 7.4.3 Best Catalyst for Oxidative Coupling of 4-Methylpyridine.................................133 LIST OF REFERENCES.............................................................................................................134 BIOGRAPHICAL SKETCH.......................................................................................................140 7

PAGE 8

LIST OF TABLES Table page 2-1 Reported literature yields for 4,4-dim ethyl-2,2-bipyridine forming reactions.....................43 2-2 Product yields and catalyst surface areas of various catalysts prepared and tested in the coupling reaction of 4-met hylpyridine. Reaction condi tions are given in the Experimental Section.........................................................................................................44 2-3 BET surface areas of the supports us ed in the catalyst preparations.......................................45 2-4 Results from CO chemisorption measurements of selected catalysts prepared and tested in the coupling reaction of 4-methylpyridine.....................................................................46 2-5 Results from experiments at different reaction times to probe homogeneous versus heterogeneous catalysis......................................................................................................47 3-1 Suppliers, properties and price of the supports used in the study............................................64 3-2 Catalyst pretreatment temp eratures and catalytic activities for all prepared catalysts............65 3-3 Catalyst surface areas a nd other support properties................................................................66 3-4 Adsorption of NH3 and CO2 on the different catalyst supports...............................................67 3-5 Catalyst cost.............................................................................................................. ...............68 4-1 Catalyst activity.......................................................................................................... .............82 4-2 Promoted catalyst dispersion............................................................................................... ....83 6-1 Effects of palladium loading on catalys t properties and catalytic activities..........................120 6-2 The effects of catalyst reactant ratio on the catalytic activities of Pd/ n-Al2O3 and Pd/pTiO2..................................................................................................................................121 6-3 Effects of titration pH on the act ivities of selected catalysts.................................................122 6-4 Effects of catalyst pretreatments on the ac tivities and dispersion of 5% Pd precipitated on nano alumina (+) or porous titania..............................................................................123 8

PAGE 9

LIST OF FIGURES Figure page 1-1 Oxidative coupling reaction scheme......................................................................................2 0 1-2 Brunauer-Emmett-Teller (BET) measurement and BET isotherm........................................21 1-3 ChemBET chemisorption measurement................................................................................22 1-4 X-ray photoelect ron spectroscopy (XPS)..............................................................................23 1-5 X-ray Diffraction (XRD).................................................................................................. .....24 2-1 Oxidative coupling of 4-methylpyridine using a palladium catalyst.....................................42 3-1 Support surface area effects,..................................................................................................69 3-2 Palladium disper sion vs. product yield..................................................................................70 3-3 Support pH.............................................................................................................................71 5-1 Proposed oxidative c oupling reaction mechanism..............................................................100 5-2 Palladium 3d spectra of palladium/nano-alumina catalysts................................................101 5-3 Oxygen 1s spectra of pa lladium/nano-alumina catalysts....................................................102 5-4 Palladium 3d spectra of palladium/titani a catalysts............................................................103 5-5 Palladium 3d spectra of palladium on nanozirconia and ceria catalysts...........................104 5-6 Oxygen 1s spectra of palladium on nanozirconia and ceria catalysts...............................105 5-7 TEM of select catalysts........................................................................................................106 5-8 Powder XRD of n-Al2O3(+) catalyst...................................................................................107 5-9 Powder XRD of palladium on ceria catalyst.......................................................................108 5-10 Powder XRD of spent and fr esh palladium on porous titania.............................................109 9

PAGE 10

Abstract of Dissertation Presented to the Graduate School of the Un iversity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy NANOPARTICLE OXIDES AS CA TALYST SUPPRTS FOR THE OXIDATIVE COUPLING OF 4-METHYL PYRIDINE OVER PALLADIUM By Luke Michael Neal December 2008 Chair: Helena E. Hagelin-Weaver Major: Chemical Engineering The use of nanoparticles as supports for palladium catalysts was studied for application to the oxidative coupling of 4-methylpyridine via C-H activation and C-C coupling. The product of this oxidative coupling is 4,4-dimethyl-2,2-bipyridine which is a useful but expensive chelating agent. Although palladium on alumina are traditionally poor catalysts in this reaction, an excellent palladium on nanoparticle alumin a catalyst was developed which exceeds the activities of the traditional palladium on carbon cata lyst. This nanoparticle catalyst only gives high activities when prepared by precipitation. Other nanoparticle ox ide supports were found that give highly active palladium catalyst for the oxidative coupling reaction, including nanoparticle magnesia and zircon ia. Additionally, some trad itional porous supports were found to result in active catalysts, including gamma-a lumina, titania, and silica, but only the porous titania had activities comparable to the best performing nanopart icle-supported catalysts. It was also determined that zirconia and ceria are effective additive s in several of the palladium catalysts. In general, oxides with strong palladium-s upport interactions, and/ or very high surface areas were the most active. This is attributed to electrophilic palladium oxide particles that can form on supports with strong me tal-support interactions or to the numerous low-coordination 10

PAGE 11

edge or corner sites of the high surface area supports. The el ectrophilicity of these particles was determined by x-ray photoelectron spectroscopy (XPS). It was also shown that basic sites on the support are not necessary for forming an activ e catalyst, but that a high concentration of acidic support sites tends to give more active catalysts. Optimization studies of the catalysts were also performed. It was determined that 2.5, 5, and 10% loadings of palladium on nanoparticle alumina give the same product yield per unit weight of catalyst, which indicates a structure sensitive reaction. Lowering the Pd loading to 1% decreases the yield palladium loadings of ~ 2.5% were, thus, determined to be optimal on the nanoparticle alumina support. Additionally it was shown that titration to a pH of 11 gives more consistent results compared with titration to a simple stoichio metric excess (based on the amount of palladium(II) nitrate used). 11

PAGE 12

CHAPTER 1 INTRODUCTION 1.1 Overview The oxidative coupling of 4-methylpyridine over palladium via C-H activation and C-C coupling (Figure 1-1) is a useful system for the study of nanopart icle oxide catalyst supports. Early research has shown that palladium on carbon (Pd/C) is a reasonably good catalyst for this reaction [1-4], and that palladium on alumina is a poor catalyst [1]. The reason for the large differences in activities between the carbon and al umina supported catalyst is not fully explained in literature, as the carbon supports advantage is mainly high surface area with concurrent propensity to give high metal disp ersion. It has been suggested th at palladium oxide is the active phase [5] and commercial alumina catalysts are ofte n prereduced. As such, it was predicted that palladium supported on nano-alumina would be activ e due to the oxidized palladium of prepared catalysts, and the unique propertie s of nanoparticles, including re latively high surface areas. In this research, the use of nanoparticle alumina and other nano-oxides, as well as porous oxides, as supports for palladium catalysts was studied in the oxidative c oupling reaction. Additionally the properties of these catalysts were probed to determine what contributes to the activity of a palladium catalyst for this reaction system. 1.2 Background 1.2.1 Nanoparticle Oxide Supports Nanoparticles are particles with diameters of less than 100 nm particularly those with diameters in the size range of 0.5-20 nm. Due to their small size, they can have very high surface areas per unit mass. Add itionally, they have a high density of edge and corner sites, which have low coordination [6]. With these lo w coordination sites, na noparticle oxide supports are expected to interact more st rongly with the active catalyst phase (s) than the traditionally used 12

PAGE 13

porous oxides. These interactions may result in high dispersions and/or significantly alter the electronic environment of the palladi um/palladium oxide particles. 1.2.2 Palladium Catalysts In heterogeneous catalysis palladium (Pd) is typically dispersed onto a high surface area support. This increases the metal surface ar ea available to catalyze the reaction and helps prevent sintering of the small palladium particle s. This is particularly important due to palladiums high cost. However, in addition to maintaining a high palladium surface area, the support can also interact strongl y with palladium. Therefore, different supports can give palladium catalysts with somewhat different properties. Consequently, palladium catalysts can be tailored to a wide array of reaction system s. Palladium on carbon (Pd/C), one of the most widely used Pd catalysts, is efficient in hydrogenation reactions. There are, however, numerous palladium catalysts studied for oxidation reactions in particular low temperature and selective oxidations, that may be better suited for oxidativ e coupling. Palladium on alumina is a typical oxidation catalyst [7], and is of particular inte rest for oxidation of methane [8] as the mechanism involves the activation of methanes very strong C-H bonds [9-11]. Pd/tin oxide is also an effective low-temperature methane oxidation cata lyst [12,13]. Pd/titania is used in selective hydrogenation such as partial hydroge nation of alkadienes to alkene s [14]. Titania is also of interest as a support as it has b een reported to have strong metal support interactions in palladium catalysts which change the CO and H2 uptake of the palladium phase, with the strength of the interaction depending upon the titani a structure (anatase or rutile) and the heat treatment of the catalyst [14]. Reactions activated by acidic sites such as oxidati on of ethane [15] proceed well on palladium supported on silica and modified sili ca. Pd/ceria and Pd/ceria-zirconia have interesting redox properties that make them good catalysts for CO to CO2 oxidation [16-18]. Pd/zirconia and Pd/magnesia are also of interest in a coupling reaction as they catalyze the 13

PAGE 14

aromatization of hexane to aromatic compounds [ 19]. If the high dispersion properties of Pd/C can be obtained on a nanoparticle oxide, there w ould be great potential to find a more suitable support for the oxidative coupling reaction. 1.2.3 Bipyridines: Applications and Synthesis Bipyridines are excellent chelating agents th at form useful transition metal complexes [20]. The interesting properties of these comple xes make them suitable for a range of catalytic and photochemical systems. Ruthenium/bipyridin e complexes adsorb visible light and exhibit electron transfer and chemiluminescence propert ies [21,22,], making them promising for use in artificial photosynthesi s [23] and chemiluminescence detecti on systems [24]. Copper/bipyridine catalysts are widely used in atom transfer ra dical polymerization (ATRP) reactions [25]. Iron and cobalt/bipyridine complexes can catalyze the reduction of CO2 [26]. Palladium/bipyridine complexes are applied in numerous organic reac tions including oxidative carbonylations [27], the Kumada-Corriu reaction [28], and the Su zuki cross-coupling reaction [29]. Of the bipyridines, 4,4-dimethyl-2,2-bipyrid ine is particularly useful as it can be derivatized in the 4 positions [30]. However, it is also prohibitively e xpensive for industrial scale use. The methods of synthesizing this bipyridine are expensive and time consuming, which results in very high prices (more than $5,200/kg, [31]). Some bipyrid ine synthesis methods include building a second pyridine ring onto a substituted pyridine precursor, and coupling of halogenated pyridines precursors have also been used [32-34]. Unfortunately, these processes are complex, use expensive and environmental unfriendly starting materials, and require involved subsequent purification steps. Of the synthesis pathways, the coupling of pyrid ine is the simplest. This coupling can be accomplished with over Raney nickel or palladium [1-4]. Although Raney nickel (or skeletal nickel) is a common and relativel y inexpensive catalyst, it requi res complicated pretreatments 14

PAGE 15

and special handling due to its pyrophoric nature, making it undesira ble as catalyst. Oxidative coupling of 4-methylpyridine to 4,4-dimethyl-2 ,2-bipyridine over palladium, on the other hand is a simple one-step reaction (F igure 1-1), in which the only produc ts are water, the bipyridine, and a small amount of terpyridine byproduct. The Pd catalyst is re latively safe and requires no special handling. However, this reaction is slow, and the cataly st loses activity during reaction [5]. Finding an improved palla dium catalyst for this reaction would make use of 4,4-dimethyl2,2-bipyridine more viable economically. 1.2.4 Catalyst Characterization Techniques In the research presented here numerous techniqu es were used to characterize the supports. Detailed descriptions of the technique can be found in several textbooks and hand books such as Masel [35], Cullity [36] and the Handbook of X-ra y Photoelection Spectros copy [37]. However, a brief overview is in order. 1.2.4.1 Brunauer-Emmett-Teller (BET) surface area measurements The surface areas of the catalysts and catal yst supports were determined using BET adsorption isotherms, which are described in detail in Masel [35]. Briefly, in the instrument used in these studies (Quantachrome Nova 1200) nitrogen is adsorbed onto the catalyst or support surface by pulsing nitrogen gas into an evacuated sa mple cell that contains the catalyst or support and is immersed in liquid nitrogen (Figure 1-2). An isotherm (constant temperature ~123 K) plot that relates the cell pressure (P) to the volume of gas pulsed into the cell (V) is created (Figure 13). By finding the slope of the isotherm plot a nd its zero pressure intercept, the volume of nitrogen necessary to adsorb one-monolayer ( VM) of nitrogen molecules on the surface can be calculated: Intercept Slope VM 1 [36] 15

PAGE 16

Because the cross-sectio nal area of nitrogen can be determined (16.2 2) [35] the amount of area one monolayer of nitrogen covers can be calculated. This area is approximately equal to the surface area of the adsorben t (catalyst/support) surface. 1.2.4.2 Chemisorption measurements Masel [35] also describes the chemisorption phenomena in detail. Chemisorption measurements are powerful techniques for probi ng the sample surface. Carbon monoxide (CO) chemisorption is commonly used to determine the surface area of palladium metal on a catalyst support. CO adsorbs to metal atoms on surfaces a nd if the stoichiometry of CO to Pd surface atoms is known then the number of Pd atoms on the surface can be calculated. Additionally with estimates of the shapes of Pd particles on the support surface, the surface area of palladium and palladium crystallite sizes can be calculated. These calculations are desc ribed in a detailed in chapter 3. By using an acidic or basic gas, such as CO2 or NH3, the number of basic and acidic sites on a support can also be measured. In the studies presented here chemisorpti on measurements were performed on a ChemBET 3000. This instrument works by titrating a known amount of adsorbate gas (such as CO, CO2 or NH3) gas into a stream of inert makeup gas (such as helium) (Figure 1-3). The titrated stream then flows over the sample surface, where the ac tive gas is adsorbed onto the surface until the surface is saturated. The amount of unabsorbed titrate gas in the makeup stream is then measured with a thermal conductivity detector (TCD). 1.2.4.3 X-ray photoelectron spectroscopy (XPS) An excellent overview of XPS can be f ound in The Handbook of X-pay Photoelectron Spectroscopy [37]. In this technique a solid surf ace is exited with an X-ray source (Figure 1-4). Electrons from the surface and near surface region of a solid are ejected into a vacuum by this excitation. If these ejected photoe lectrons escape without other effects, they have an energy 16

PAGE 17

equal to the exciting x-ray s ource less the energy required to remove the electron (binding energy) from the solid and the work function of the spectrometer. This binding energy is partially dependent upon the energy of the near surface atoms before (initial state) and after (final state) the electron is rem oved. The binding energies of elect rons from various orbitals are different not only for different elements, but for di fferent oxidation states of the same element, as well as the electronic environment of the atom In XPS the kinetic energy of the ejected photoelectron is then measured. Since the en ergy of the exciting x-ray is known (typically magnesium or aluminum K 1,2) and the work function of the sp ectrometer is corrected for by use of a standard, the binding energy of the electron can be calculated. This allows for the elemental composition of the surface as well as the oxidation state of the surface atoms to be determined. Since this is a surface technique it is a very powerful tool for probing catalysts, where their behavior is determined by surface atoms. However, this is an ultra high vacuum (UHV) (pressures less than 10-9 Torr) technique, requiring special equipment. 1.2.4.4 X-ray diffraction (XRD) X-ray diffraction (XRD) is covered in detail in numerous textbooks such as Cullity [36]. In XRD an x-ray beam is directed at a solid sample. If the sample is crystalline than a diffraction pattern will emerge due to the planes of the cr ystal. The angel of diffraction is dependent upon the wavelength ( ) of the incident light and the di stance between the parallel planes ( dhkl) of diffraction (Figure 1-5) as de scribed by the Bragg equation: )sin(2 hkldn (Where n must be an integer fo r a reflection to be observed) If the composition of the sa mple is known then the crystal phases present can be determined by comparing the diffraction pattern to known standards. Additionally in powder 17

PAGE 18

XRD where the sample is polycrystalline the average diameter of the crystals ( dav) can be determined from the peak widths (full width at half maximum or FWHM ) and the angle of the reflection ( ) diffraction patterns by th e Scherrer equation: ))cos( (/ FWHM Kdav ( K is a prefactor determined by the instrument but is typically assigned a value of unity) However, care must be taken when using XRD to characterize catalysts. The attenuation depth of diffracted x-rays in the samples can be several micrometers, which is several orders of magnitude greater than the thickness of surface atom layers. Powder XRD is thus a bulk technique in terms of catalysts Additionally, the ac tive surface phase can be amorphous and would thus not be detected by XRD. Conseque ntly it may be impossible to probe the active surface phase in many catalysts with XRD. A nd the XRD patterns obtained may be for inactive phases far removed from the cata lyst surfaces. Nonetheless, know ing these limitations and when used in conjunction with surface techniques, powder XRD can give important information. 1.3 Objectives In this work palladium on nanoparticle oxide supports are studied for the use in the synthesis of 4,4-dimethyl-2,2-bipyrid ine. Prior to this research the highest yields reported for a catalyst (5% Pd/C) is ~2 g product per gram of ca talyst, i.e. 40 g/g Pd [38], with lower yields being typical [4,39,40]. In the research presented here a palladium on nanoparticle alumina catalyst was developed and studied for this reaction as a proof of c oncept. Additional nanoparticle and porous oxide supports were then studied to find the most active and economical support materials. Once prospective supports were develope d, the use of promoters was studied to increase yields further. Additionally an extensive charac terization of these catalysts was undertaken to determine the 18

PAGE 19

important properties responsible for catalyst ac tivity. Finally, optimization was undertaken to develop the best possible catalys t by the simplest means. The objectives of this resear ch were: 1) showing that nanoparticle oxide s are viable supports for palladium catalysts in the production of 4,4-dimetyl-2,2-bipyridine via oxidative coupling of 4-methylpyridine, 2) determining what support/catalyst properti es are responsible for activity, and 3) developing an ec onomically viable catalyst. 19

PAGE 20

N N N C H 3 2 CH3 CH3 + H 2 O O2 Palladium Figure 1-1 Palladium-catalyzed oxidativ e coupling of 4-me thylpyridine. 20

PAGE 21

)1(b bXV X bX Saturation bP P X Figure 1-2 Brunauer-Emmett-Teller (BET) measurement and BET isotherm 21

PAGE 22

TCD He Makeup Gas To Vent Cell Catalyst Adsorbate Gas Titration Valve Signal to Computer Loop Figure 1-3 ChemBET chemisorption measurement 22

PAGE 23

+ + Mg k 1,2 X-ray 1 25 6 3 eV P h o t o e l e c t r o n K E = 1 2 5 6 3 e V B E To Detector Figure 1-4 X-ray photoelect ron spectroscopy (XPS) 23

PAGE 24

Figure 1-5 X-ray Diffraction (XRD) 24

PAGE 25

CHAPTER 2 C-H ACTIVATION AND C-C COUPLING OF 4-METHYLPYRIDINE USING PALLADIUM SUPPORTED ON NANOPARTICLE ALUMINA 1 2.1 Introduction Bipyridines are receiving increasing attention in the literature due to their ability to coordinate to transition metal cations and form complexes with interestin g properties [20]. The importance of metal complexes containing bipyri dine ligands has been revealed in several reviews that have been published in the past fifteen years on the synthesis of bipyridines, as well as on properties of metal-bipyridine complexe s and their applications [21,22,41-43]. Of the transition metals, bipyridine complexes of ru thenium are by far the most commonly studied systems [21,43]. This is due to their unique phot oand electrochemical properties. Rutheniumbipyridine complexes can absorb photons in the visible light regi on and have unique redox properties that can lead to el ectron transfer and chemilumines cence, which make them suitable for application in solar energy conversion (e.g. in solar cells and artifici al photosynthesis systems [43,44]), in organic light-emitting diodes [23] and in chemilumines cence detection systems [24]. Bipyridines are also commonly used as ligands to metals in various catalyst systems. For example, many copper-based atom transfer radi cal polymerization (ATR P) catalysts contain bipyridine units [25]. Iron and cobalt complexes of bior ter-pyridines have been shown to catalyze the reduction of CO2 and O2 [26]. Palladium bipyridine complexes have been used as catalyst in several reactions, su ch as oxidative carbonylation [ 27], the Kumada-Corriu reaction [30], and the Suzuki cross-coupling reaction [29]. Of the bipyridines, 4,4-dimethyl-2,2bipyridine is of particular interest, since this compound can easily be modified by reactions with 1 Reprinted from Journal of Molecular Cataly sis A: Chemical Vol. 284 pp. 141, Luke M. Neal, Helena E. Hagelin-Weaver (2008) with permission from Elsevier 25

PAGE 26

the methyl groups in the 4-positions. However, due to the poor yields in the coupling reaction of 4-methylpyridine, the production of 4,4 dimethyl -2,2-bipyridine is prohibitively expensive for large scale processes. A kilogram quantity sells for more than $4,500, [45] while smaller quantities sell for significantly hi gher prices per unit weight. Bipyridines can be formed vi a a number of pathways. So me include building the second pyridine ring from a substituted pyridine, while other methods rely on the coupling of halogenated pyridines us ing transition metal catalysts [32-34] The disadvantages with these processes are the low yields of multi-step processes, the cost of the halogenated precursors and the environmental impact of the waste streams of halide salts and other byproducts that result from these reactions. Furthermore, these pr ocesses require a high level of subsequent purification for applications that are sensitive to halogens, such as catalysis. Consequently, a one-step process in which the bipyridine is fo rmed directly from the pyridine reactant is desirable. The oxidative coupling of 4-met hylpyridine to 4,4-dimethyl -2,2-bipyridine using palladium on carbon as the catalyst meets this crite rion (Figure 2-1). This reaction requires only a catalyst plus the reactant and the only by-produc ts of the reaction are water and the terpyridine (Figure 2-1). Furthermore, compared to its halogenated derivatives (the most active bromoderivative is available through reaction with commercially available 2-amino-4-methylpyridine at ~$200/kg [34, 46]), 4-methylpyridine is relativ ely inexpensive (less than $40/kg [47]) with lower environmental impact. The disadvantages of this reaction are the slow reaction rate and the deactivation of the catalyst [5]. Early research has shown that 2,2-bipyridines can be formed via coupling of pyridine derivatives over catalysts such as Raney nick el and palladium on carbon (Pd/C) [1-4]. The results from these early experiments reveal that while palladium on carbon is a reasonable 26

PAGE 27

catalyst in the coupling reaction of pyridine de rivatives, palladium on alumina exhibits poor activity [1]. No explanation as to why alumin a is an inferior support has been given in the literature. The early research also indicated that low-valent palladium is the active form of the Pd/C catalysts. However, more recent results reveal ed that the active catalyst actually contains a Pd(II) species [5] and that the variation in catalyt ic activity between batches of Pd/C catalysts could be reduced by simply oxidizing the catalyst before reaction. However, despite previous improvements of the catalyst, the reaction is slow and suffers from catalyst deactivation. One of the major limitations of the reaction appears to be reoxidation of Pd in these solution experiments, since the catalyst af ter exposure to the reaction conditions is in a reduced form [5]. The maximum isolated yield reported for a 5% Pd/C is ~2 g product per gram of catalyst, i.e. 40 g/g Pd [38]. However, yields of 1.5-2 g/g fo r a 10% Pd/C catalysts are more common [5, 39, 40]. Comparing these product yields to reactio ns using a homogeneous catalyst complex or halogenated precursors in solution is difficult. Th e yields reported as converted reactant tend to be higher for reactions in solution with halogenated precursors [33, 34, 48], while the yield per gram of palladium generally is higher in reactions where 4-methyl pyridine is the reactant and no solvent is used (See summary of literature yields in Table 2-1). Furthermore, the higher conversion of reactant for the reactions usi ng solvent and halogenated precursors can be outweighed by the lower yields per gram of palla dium, the formed byproducts (halide salts) and the use of solvent. Furthermore, the unreacted 4-methylpyridine can easily be recovered in the reactions with no solvent and a heterogeneous catalyst and there is a potential for recycling and regeneration of the heterogeneous catalyst. Th erefore, there are sign ificant advantages of synthesizing 4,4-dimethyl-2,2-bipyridine usi ng a heterogeneous catalyst and no solvent, particularly if the yields can be increased. Naturally, the conver ted reactant yield can also be 27

PAGE 28

increased in reactions with no solvent by simply increasing the catalyst-to-reactant ratio, although more terpyridine byproduct will be formed in these cases. Our hypothesis is that palladi um supported on nanoparticle oxide supports has potential to be a very efficient catalyst. This hypothesis is based on the fact that nanoparticles have a large surface area compared to their bulk analogues. Fu rthermore, nanoparticles have a high degree of low coordination sites, such as corners and edge s [6]. These low coordination sites may cause a stronger interaction between the support oxide and the ac tive metal deposited onto the nanoparticles compared with more conventional supports. In addition to potentially higher dispersions of the active metal, these interactions may result in unique cataly tic properties. It is possible that this is the reason for the high activity observed at low temperatures in the palladium-catalyzed oxidation of methane when na noparticle oxides are used as supports [49]. Consequently, not only the large surface area, but also the intrinsic properties of nanoparticle oxides can result in unique catalytic activit ies of nanoparticle-supported catalysts. The main objective in this work is to de termine if palladium supported on nanoparticle alumina can be an efficient catalyst in the c oupling of pyridines desp ite previous research showing that commercial palladium on alumina is a poor catalyst in this reaction. Part of the objective is also to determine the effects of catalyst preparation on th e catalytic activity and if the active species is likely to be a dissolved homogeneous complex instead of heterogeneous palladium surface species. 2.2 Experimental 2.2.1 Catalysts Preparation The catalysts were prepared using commercia lly available alumina nanoparticles [50]. Two catalyst preparation methods were used; we t impregnation and prec ipitation. In the wet impregnation method the support po wders were dispersed in an aqueous solution of palladium 28

PAGE 29

nitrate (Fluka or Alfa Aesar). The water wa s then boiled off until a paste consistency was achieved. This paste was dried in a muffle furnace at 105 C overnight. The dried samples were ground and calcinated at 450 C for 3 hours to decompose the palladium nitrate and form palladium oxide on the support. In the precipitation method, the supp ort was dispersed into a soluti on of palladium nitrate. The mixture was then titrated with a NaOH solution, which formed Pd(OH)2 on the support [51]. The amount of NaOH used in these experiments corresponds to 50% stoichiometric excess. The resulting mixture was aged overnight at room temp erature before it was filtered. The recovered catalyst was rinsed by stirring in water overnig ht, followed by another filtration. As for the catalysts prepared via the wet impregnation me thod, the precipitated sa mples were dried over night at 105 C and calcinated at 450 C for 3 hours. Several commercial catalyst supports and catalysts were also used for catalyst preparation, or used as recei ved, and tested for activity in the coupling reaction of 4methylpyridine. These include 5% Pd/C (Alfa Aesar, surf ace area [SA]: 695 m2/g), 5% Pd/Al2O3 (Alfa Aesar Pd/ -alumina, SA: 155 m2/g), -Al2O3 (Alfa Aesar high surface area bimodal, SA: 260 m2/g) and activated carbon (Calgon F 400, SA: 765 m2/g). 2.2.2 Reaction Conditions The 4-methylpyridine (Aldrich or Across) was distilled over KOH or NaOH prior to use. In a typical reaction run 1 g of catalyst was pla ced in a round bottom flask along with 10 g of the distilled 4-methylpyridine. The reaction mixture was evacuated and an oxygen atmosphere introduced before it was heated to the boiling point (145C). The reaction proceeded under reflux for 72 hours. After a complete reaction th e flask contents were filtered using a glass 29

PAGE 30

micro-fiber filter and washed with chloroform to dissolve the product. The chloroform, water and unreacted 4-methylpyridine were rem oved using a rotary evaporator. Selected samples were purified via sublimation. These experi ments indicate that at least 75% of the raw yield is the desired product. This, however, is a low number since only about 85% of the original sample mass is recovered in the sublimate and residue due to the difficulty in removing all of the product from the sublimatio n apparatus. Conseque ntly, the yields are reported as raw yields in the paper. The sublimate, residue and raw products were characterized using NMR. The only significant product found was the bipyridine. The residue did not redissolve well into the chloroform. The soluble fraction of the sublimati on residue showed little evidence of organic compounds ot her than the product. Based upon NMR it is likely that the major portion of the non-sublimated impurities is inorganic residues from the catalysts. 2.2.3 Catalyst Characterization Brunauer-Emmett-Teller (BET) surface area measurements were performed on a Quantachrome NOVA 1200 instrument. Fresh catalysts were outgassed under vacuum for 3 hours at room temperature before the measurements Catalysts that had been stored for longer times were outgassed at 105C for at least an hour before the BET analysis. The N2 adsorption was performed over five isotherms, which gave roughly linear fits. To determine the dispersion of Pd on the catalysts carbon monoxide chemisorption experiments were performed. The fresh catalysts were reduced for 2 hours at 170 C using a 5% hydrogen in nitrogen gas mixture. The samples we re then out-gassed for 1 hour in helium before the pulse titration experiments with CO at 25 C and 1 atm. The low reduction temperature was used to avoid excessive sintering or spalling of the Pd partic les during the reductive treatment and was chosen to be close to reaction temperature. 30

PAGE 31

2.3 Results and Discussion 2.3.1 Effects of Catalyst Preparation Method In the first set of experiments 10% (by we ight) of palladium was deposited onto the NanoActive Aluminum Oxide Plus support [Pd/nano-Al2O3(+)] using the wet impregnation method. This catalyst was tested for activity in the coupling reaction of 4-methylpyridine. As expected for a palladium on alumina catalyst, this nano-particle supported catalyst did not exhibit any significant activity in the reaction (Table 2-2, Entry 1). When th e precipitation method was used, however, a 10% Pd/nano-Al2O3( +) catalyst yielded a signif icant amount of the 4,4dimethyl-2.2-bipyridine (Table 2-2, Entries 2 and 3). In fact, the 20-25 g product per g of palladium corresponds to 2-2.5 g product per gram of catalyst, which is equal to or higher than the maximum yield reported from a palladium on car bon catalyst [5]. It was also shown that the Pd content can be reduced to 5% Pd on alumin a without substantially reducing the product yield (Table 2-2, Entry 4). In other words, the yi eld per gram of palladium can be doubled by going from a 10% to a 5% Pd/nano-Al2O3(+). Consequently, a yield of 50 g raw product per gram palladium can be obtained with th is catalyst. The 5% Pd/nano-Al2O3( +) catalyst is more active than the commonly used Pd/C catalyst despit e the inactivity reported for commercial Pd/Al2O3 catalysts. It is also interesting to note that th e catalytic activity is very dependent on the catalyst preparation method. While the precipitation method yiel ds a catalyst with the highest observed catalytic activity to date, the impregnation techni que results in a very low activity or an inactive catalyst. Both the impregnated and the precipitat ed catalysts have specific surface areas in the range of 150-200 m2/g (Table 2-2, Entries 1-4). Theref ore, the support surface areas cannot explain the differences in activities between thes e catalysts. As can be seen in Table 2, the results are reproducible (Entries 4, 11 and 26). In fact, the yields obtained from different preparations of precipitated Pd/nano-Al2O3(+) catalysts are more repr oducible than the yields 31

PAGE 32

obtained from different runs on the same batch of commercial Pd/C (Table 2-2, Entries 5 and 25). This is likely due to variations in the Pd O concentration on the as received Pd/C catalysts [5]. Since our catalysts are ca lcined in air and not reduced before reaction, the amount of PdO on the Pd/nano-Al2O3( +) catalysts is almost certainly more constant compared to the commercial Pd(0)/C. To test whether or not the differences in th e catalytic activity between catalysts prepared by the impregnation and precipitation methods onl y applies to alumina supports, impregnated and precipitated palladium on activated carbon catalysts wher e prepared. The precipitated catalyst gave a modest yield (15.1 g/g Pd, Entry 22, Table 2-2), which is comparable to the commercial Pd/C catalyst, while the impregnated Pd/C catalyst gave a low yield (6.6 g/g Pd, Entry 23, Table 2-2). This indicates that the pr ecipitation method is superior to the impregnation method regardless of the support used, although the difference is more drastic for the nanoparticle alumina support. 2.3.2 Catalyst Support Effects As expected, the commercial 5% Pd/Al2O3 catalyst exhibited poor ac tivity and resulted in little, if any, product under the reac tion conditions of the experiments. This is most likely due to the fact that the commercial 5% Pd/Al2O3 catalyst consists mainly of Pd(0), in contrast to commercial Pd/C catalysts which appear to have a relatively high, albeit varying, surface Pd(II) content. The yield of the commercial 5% Pd/C catalyst agrees with previous results [5]. It is lower than the highest yield observe d using this catalyst, due to the fact that the catalyst was used as received without oxidation treatment. To de termine if it is the natu re of the nanoparticle alumina support or the preparation method that is responsible for the high cat alytic activity, a 5% Pd/Al2O3 catalyst was prepared with a commerci al alumina support using the precipitation method. The yield for this cataly st varied widely (Table 2-2, Entries 15, 21 and 24). On average 32

PAGE 33

the catalyst does exhibit a fair activity, but it is not as active as th e nanoparticle-supported Pd/Al2O3 catalyst. Another catalyst was prepared using the precipitation method and a second nanoparticle alumina sample [NanoActive Aluminum Oxide, nano-Al2O3(-)] as support. The surface area of this alumina sample (275 m2/g) is lower than the NanoActive Aluminum Oxide Plus sample (695 m2/g), but on the same order as the commercial -alumina support (260 m2/g) [See Table 3]. The catalytic activ ity of this catalyst is significantly lower than that obtained on the commercial Pd/C catalyst and the palladium on the NanoActive Aluminum Oxide Plus or commercial -alumina supports. Consequently, there is a significant difference in the catalytic activities of the catalysts prep ared with the two NanoActive Al uminum Oxide supports. It is interesting to note that the catalyst supported on the nano-Al2O3(-) gives a lower yield than the catalyst prepared using the commercial -alumina despite the fact that the surface areas of these supports are similar. The catalysts prepared via precipitation onto thes e three alumina supports all have surface areas between 150 and 205 m2/g. These results indicate that neither the surface area of the bare support nor the fina l surface area of the prepared catalysts is the sole determining factor of the catalytic activity. A further indication that the initial support surfa ce area is not the main factor in determining the catalytic activ ity can be seen when comparing alumina and carbon supported catalysts. While a catalyst prepared via precip itation of palladium ont o NanoActive Aluminum Oxide Plus (695 m2/g, Table 2-3) gives a produc t yield of ~50 g/g Pd, a catalyst prepared using the same method but with an activat ed carbon support (surface area: 765 m2/g, Table 2-3) only gives a yield of 15 g/g Pd. Thus, despite the slig htly higher surface area of the activated carbon it does not result in a catalyst as active as the one supported on nano-Al2O3(+). 33

PAGE 34

Aside from giving the best activity of any of the s upports tested, the nano-Al2O3(+) catalysts exhibited better reproducibility than an y of the catalysts including the commonly used commercial palladium on carbon. The only major ac tivity variations seen for the catalysts supported on nano-Al2O3(+) can be attributed to reactant quality. 2.3.3 Palladium Surface Areas The palladium surface areas were determined after reduction of the PdO on the catalyst surfaces using CO adsorption and the volume of ad sorbed CO is given in Table 4 for selected catalysts. Performing CO titration measurements on reduced catalysts do not necessarily result in a good measure of the catalytically active surf ace area since the active pha se on these catalysts is PdO and not Pd metal. This is particularly the case when the catalyst before reduction consists of Pd metal or a mixture of PdO and Pd, as in the case of the commercial Pd/C and Pd/Al2O3 catalysts. However, in cases where the original catalysts consist solely of PdO (as is the case for the nanoparticle-supported catalysts) and if the same mild reduction conditions are used for all catalysts, it should be possible to observe trends and obtain qualitative results from the Pd surface area measurements. CO chemisorption measurements on supported Pd catalysts are further complicated by the dependence of the Pd :CO stoichiometry on the dispersion since CO can adsorb in linear, bridge and hollow bindi ng modes on the surface [52]. Literature data indicate that a stoichiometry of two Pd surface atoms per CO mol ecule adsorbed is appropriate for high dispersions and a surface atom density (Cm) of 1.42 x 1015 /cm2 based on the cubooctahedral geometry is reasonable for small Pd crystallites [53]. While the exact Pd:CO stoichiometry may deviate slightly from the 2: 1 ratio used here, the assumptions made in calculating the Pd surface area should give suffi cient accuracy for comparisons between the different catalysts in the study, pa rticularly considering that the active phase is PdO rather than Pd on these catalysts. 34

PAGE 35

The 5% precipitated nano-Al2O3(+) has a very high Pd surface area compared to the other catalysts in this study (Table 2-4). In fact, the Pd surface area of this cata lyst is more than two times that of the commercial 5% Pd on activ ated carbon catalyst, which is the commercial catalyst with the highest Pd surf ace area. The commercial 5% Pd/Al2O3 catalyst also has a high dispersion but, as mentioned, this catalyst exhibits a low activity due to a low PdO concentration rather than a lack of Pd surface area. The 5% impregnated nano-Al2O3(+) has a surprisingly high Pd dispersion considering its low activity. The high yields per Pd surface area for the impregnated palladium on carbon and the palladiu m precipitated on nano-alumina(-) catalysts is likely an artifact of impurities in the product and the low dispersi ons of the catalysts. For low yields (~0.1 g for 0.7 g catalyst or 2.9 g/g Pd) tr ace impurities from the reactant and catalyst can be a significant portion of the raw product mass and due to the nanoparticle supports used, it is difficult to separate these impurities from the produc ts. Even though the same is partly true for the precipitated palladium on car bon catalyst, i.e. the reported yield per surface area is unrealistically high, it appears that this cataly st does give a decent yield despite the low dispersion. The Pd surface areas of th e precipitated Pd/C and Pd/nano-Al2O3(-) are on the same order, but the yield is higher for the Pd/C catalyst. In fact, the data suggest that the palladium on the surface of the precipitated Pd/C catalyst is more active than the palladium on the surface of the commercial Pd/C. Comparing the data in Tabl es 2 and 4 it is evident that the modest yields obtained from the precipitated Pd/nano-Al2O3(-) and the precipitated and impregnated Pd/C catalysts can be explained in part by low dispersions. Another su rprising result is the relatively high Pd surface area of the 10% precipitated nano-Al2O3(+). It has roughly twice the metal surface area of the 5% catalyst but exhibits no increase in yield. The most striking result, however, is that the Pd surface area of the 5% Pd precipitated on commercial -Al2O3 is very low 35

PAGE 36

despite the relatively good yield. The low disper sion of this catalyst is the main reason for the very high yield reported per palladium surface area. These results strongly indicate a structure sensitive reaction, i.e. only a portion of the surface Pd atoms are active. Consequently, the catalytic activity does not correlate with the measur ed Pd surface areas. Th is can be a result of the fact that the surface areas are determined on reduced catalysts, while the active catalyst is the unreduced form. While the support is important to give catalysts with high Pd surface areas, it is evident that this is not a suffici ent criterion for an active catalyst. The Pd, or rather PdO, on the surface must also have the correct structure. 2.3.4 Impact of Reactant Quality As mentioned, the catalysts pr epared using the nanoparticle alumina(+) support for the most part exhibited good reproducibility. There we re slight variations in the specific surface areas and yields for repeated samples as seen in Table 2-2 (Entries 4, 11, 13 and 26). However, the results are very sensitive to the quality of the 4-methylpyridine used. The 4-methylpyridine must be distilled over KOH (Distillates 1 and 3-5 in Table 2-2) to give g ood results, preferably doubly distilled. The product yields obtained with a reactant that had been distilled over a small amount of NaOH (Distillate 2) were markedly lowe r than the yields for reactants distilled over KOH (compare Entries 8 and 11 in Table 2-2). It was noticed that Distillate 2 discolored more quickly over time compared with th e other distillates. In addition, Distillate 3, which was treated with KOH over night and then distilled over the same KOH, turned cloudy or turbid over time. This may explain the lower yields obtained for Entries 15 and 16 in Table 2-2, since those were the last two runs using Distillate 3. The best a nd most reproducible resu lts are obtained if the 4methylpyridine is treated with KOH over night then decanted and distilled over fresh KOH. This treatment results in a clear liquid, which will not discolor or turn opaque over a reasonable time (on the order of months). These results tend to indicate that an impurity is present in the lower 36

PAGE 37

quality reactants (Distillate 2 and 3) and that this impurity inhib its the reaction. As the primary purpose of adding KOH to the sample during dist illation is for drying, it seems possible that water is the impurity. To test this hypothesis, parallel reactions were run with the same catalyst preparation [5% Pd precipitated onto nano-Al2O3(+)] and the same reactant distillate. To one of these reactions 0.3 g of water pe r g of catalyst was added to the reactant. The water-free specimen gave a typical yield of 52 g/g Pd (Ent ry 26, Table 2-2). The sample with the added water yielded 31 g/g catalyst (Entry 27, Table 2-2). While this re veals that water exhibits an inhibiting effect on the reaction, it does not explai n the considerably lower yields (18.2 g/g Pd) obtained from Distillate 2 (cf Entries 8 and 27 in Ta ble 2-2). Furthermore, at a product yield of 50 g/g Pd the amount of water formed is slight ly lower than the adde d 0.3 g/g catalyst. Consequently, the presence of wate r cannot be solely responsible for the inhibiting effects seen in the lower quality distillates. The palladium nitrate source also had an e ffect on the catalytic activity. Catalysts prepared using the palladium nitrate from Fluka resulted in higher ca talytic activities than catalysts prepared using the palladium nitrate from Alfa Aesar. During the catalyst preparation it was observed that the palladium nitrate from Alfa Aesar did not dissolve as well in the deionized water as the Fluka Pd(NO3)2. In fact, the residual solids from the Alfa Aesar palladium nitrate did not dissolve even af ter addition of acid (HNO3 to a pH of 1.0). XRD analysis of both materials indicated the presence of a palladium oxide or related phase in the palladium nitrate from Alfa Aesar. The Fluka Pd(NO3)2 did not contain these phas es. Consequently, it is important to check the catalyst pr ecursor quality before preparing supported palladium catalysts. 37

PAGE 38

2.3.5 Homogeneous Versus He terogeneous Catalysis With palladium-pyridine systems there is a natural question of whether supported catalysts are truly heterogeneous, or if the actual activ e species is dissolved into the solution. Palladium(II) ions coordinate easily to the nitr ogen of 4-methylpyridine and perhaps even more strongly to the bipyridine product formed in the reaction. Conse quently, it is possible that the surface palladium is simply a precursor, or a source, to active palladium ions in solution. Other palladium-catalyzed coupling reactions, such as the Heck reaction, do exhibit palladium leaching when heterogeneous catalysts are used [54,55], a nd there is still a debate as to whether these heterogeneous palladium catalysts are active catalysts or simply precursors to a dissolved active species [56]. Therefore, a set of experiments was designed to probe if a dissolved palladium species is active in this reacti on. If palladium is dissolved in to the reaction mixture and this palladium is catalytically active, it would be expected that the reaction proceeds after the heterogeneous catalyst (support) has been removed. One reaction wa s taken out of the oil bath after 24 hours and the reaction mixture was filtered hot to remove the solid catalyst. The filtrate was then returned to reflux under oxygen for an additional 72 hours. The product yield recovered from a 24-hour experiment is lower th an the product yield afte r a 72-hour reaction (17 g/g Pd versus 50 g/g Pd, Entry 1, Column 3 in Tabl e 2-5). This reveals a slow reaction rate and the need for a 72-hour reaction time. Furthermor e, the product recovered after 24 hours with solid catalyst plus 72 hours without solid catalyst (i.e. after returning th e filtrate to the reaction conditions) is very close to the yi eld of a 24-hour reaction (22 g/g Pd). This is evidence that the reaction is heterogeneous, or at least that the presence of a heterogeneous catalyst is necessary for the reaction to proceed. A set of similar experiments were run by reac ting a catalyst-reactant mixture for 24 hours and then use the recovered catalyst in a subse quent reaction. The recovered catalyst from a 2438

PAGE 39

hour reaction was washed with chloroform, dried a nd then reloaded with fresh 4-methylpyridine. After 72 hours at the reaction conditions, 24 g of pr oduct was recovered per gram of the reloaded catalyst [Pd/nano-Al2O3(+), Entry 1, Column 4 in Table 2-5] This is significantly less product compared to a fresh catalyst, however, the combined product yield from the 24-hour and the 72hour reactions is consistent with an uninterrupted 72-hour experime nt (there is very little product formed after the initial 72 hours at the reaction co nditions). While the resu lts clearly show that a recovered catalyst is still active, it is evident that the recovered catalyst has a lower activity compared to a fresh catalyst. One of the main reasons for the lower activity of a catalyst after exposure to the reaction conditi ons for 24 hours or more, is pr obably reduction of the active Pd(II) species. However, it is also possible th at some leaching does occur, even though any palladium in solution does not appear to be an active catalyst. From the experiments on commercial Pd/C it is evident th at the 24-hour product yield is hi gher (~30 g/g Pd) than for the Pd/nano-Al2O3(+) catalyst (~ 20 g/g Pd), even though the 72-hour yield is considerably higher for the nano-Al2O3(+)-supported catalyst. However, the r ecovered catalyst is much less active (Table 2-5). This indicates that the reaction is faster on Pd/C, but the catalyst also deactivates faster than the Pd/nano-Al2O3(+) catalyst. Even though the above results imply that no catalytically active species is present in solution, there are indications of palladium leach ing from the support. Previous results have indicated that the palladium surface concentration is lower after reaction [5]. In these reaction runs a palladium mirror could be observed in some experiments. However, generally the active Pd/nano-Al2O3(+) catalyst prepared via precipitation displayed little or no Pd mirror on the reaction flask after a completed r eaction. In contrast, significant Pd mirrors were seen on several of the poorly performing catalysts, such as the Pd/nano-Al2O3(+) catalyst prepared via 39

PAGE 40

impregnation. Some mirroring was observed also in the commercial Pd/C catalyst. If the catalytic action of these catalyst s was reliant solely upon a dissolved palladium species, it might be expected that palladium mirro ring (which is evidence of palladium leaching from the support) would accompany a significant catalyst activity. In contrast, palladium dissolution is observed mainly for catalysts with poor activity. Thus, this can be taken as another indication that the active species is on the surface of the heterogeneous catalyst and not dissolved into the solution. 2.3.6 Precipitation Base NaOH or KOH were used as the base to fo rce palladium precipitation onto the support. Both these bases resulted in hi ghly active catalysts w ith similar yields (Entries 11 and 13, Table 2-2). To determine if the presence of trace amounts of base was responsible for the activity of the precipitated samples, KOH was added to a re action with an impregnated catalyst. This catalyst system showed no significant increase in ac tivity. If the base is cr ucial for the activity it is possible the Na or K is required in the deposition stage, i.e. a closer interaction between the Pd and the Na or K is necessary. 2.4 Conclusions While the palladium impregnated on nano-Al2O3(+) exhibits little or no activity in the coupling reaction of 4methylpyridine, nano-Al2O3(+)-supported catalysts prepared via the precipitation method give signifi cant yields. Despite the f act that commercial Pd/Al2O3 is not an active catalyst in this reaction, the nano-Al2O3(+) was shown to be a viable support for the coupling reaction. The yield obtained from the palladium precipitated onto nano-Al2O3(+) is significantly higher, the highest reported for this reaction system, and more reproducible compared to the yields obtained from commerci al Pd/C catalysts, whic h are the commonly used catalysts in this reaction. Additionally, it was found that traditional al umina-supported catalysts can be active for this reaction if prepared by precipitation. Ho wever, the yields are not as 40

PAGE 41

reproducible or as high as those obtained for the nano-Al2O3(+) support. The sensitivity to catalyst preparation indicates that the activity of this ca talyst is not based solely upon the surface area of the support. The lack of correlation between the measured Pd surface and the catalyst yields strongly suggests a structure sensitive reac tion. Furthermore, experiments reveal that the differences in activities between catalysts prepar ed via precipitation and impregnation is not due to the presence of bases (NaOH or KOH) in the r eaction mixture. To ensure high product yields the 4-methylpyridine must be distilled over KOH. Purification by sublimation followed by NMR measurements verifies that the product is 4,4 dimethyl 2,2 bipyridine a nd that the raw yields are at least 75% or more of the desired product. Experiments also reveal that a heterogeneous phase is necessary for the reaction to proceed at a significant rate. 41

PAGE 42

N N N N N N CH3 2 CH3 CH3 CH3 CH3 CH3 + H2O+ O2Pd/nano-Al2O3(+) Figure 2-1 Oxidative coupling of 4-methyl pyridine using a palladium catalyst. 42

PAGE 43

Table 2-1 Reported literature yields for 4,4-dimethyl-2,2-b ipyridine forming reactions Starting material Amount Reactant [mol] Catalyst Amount Catalyst [g] Yield [%] Yield [g/g Pd] Reference 2-Br-4methylpyridine 2-3 5% Pd/C, Zn 0.3 19 2.3 [34] 2-Br-4methylpyridine 2-3 5% Pd/C, Zn 0.11 19 6.6 [34] 2-Br-4methylpyridine + 23-3 Pd(PPh3)4 0.95 67 32.5 [35] 2-SnBu3-4methylpyridine 27-3 [35] 2-Br-4methylpyridine 5-3 NiBr2 0.07 47 10.5 [Ni] [50] 2-Br-4methylpyridine 5-3 NiBr2 0.3 93 4.9 [Ni] [50] 4-methylpyridine 1.03 10% Pd/C 4 7 17.3 [40] 4-methylpyridine 2.06 5% Pd/C 8.93 9 40.0 [39] 4-methylpyridine 7.19 10% Pd/C 28 6 14.3 [41] 4-methylpyridine 75-3 5% Pd/C 0.7 26 52 This work [Ni] Nickel Catalyst Used 43

PAGE 44

Table 2-2 Product yields and catalys t surface areas of vari ous catalysts prepared and tested in the coupling reaction of 4-met hylpyridine. Reaction condi tions are given in the Experimental Section. Entry Catalyst Description a 4-MethylPyridine b Raw Product [g/g Pd] c Specific Surface Area [m2/g] 1 10% Pd Impregnated on nano-Al2O3(+) 1 <1 d 155 2 10% Pd Precipitated on nano-Al2O3(+) 1 25.4 205 3 10% Pd Precipitated on nano-Al2O3(+) 1 20.0 165 4 5% Pd Precipitated on nano-Al2O3(+) 1 48.6 180 5 5% Pd/C Commercial 1 16.4 695 6 5% Pd/Al2O3 Commercial 1 2.6 d 155 7 5% Pd Impregnated on nano-Al2O3(+) 2 1.4 d 170 8 5% Pd Precipitated on nano-Al2O3(+) 2 18.2 170 9 5% Pd/C Commercial 2 7.6 695 10 5% Pd/Al2O3 Commercial 2 2.4 d 155 11 5% Pd Precipitated on nano-Al2O3(+) 3 52.6 165 12 5% Pd/C Commercial 3 14.4 695 13 5% Pd Precipitated (KOH) on nano-Al2O3(+)e 3 55.2 165 14 5% Pd Precipitated on nano-Al2O3(-) 3 8.2 155 15 5% Pd Precipitated on Commercial Al2O3 f 3 4.2 200 16 5% Pd Precipitated on activated carbon 3 2.4 17 5% Pd Impregnated on nano-Al2O3(+) +KOHg 3 2.6 18 5% Pd Precipitated on nano-Al2O3(+) (Alfa)h 4 14.0 19 5% Pd Precipitated on nano-Al2O3(+) 4 44.0 20 5% Pd Precipitated on nano-Al2O3(-) 4 5.8 21 5% Pd Precipitated on Commercial -Al2O3 f 4 24.0 22 5% Pd Precipitated on Activated Carbon 4 15.1 23 5% Pd Impregnated on Activated Carbon 4 6.6 24 5% Pd Precipitated on Commercial -Al2O3 f 4 34.0 25 5% Pd/C Commercial 5 36.0 26 5% Pd Precipitated on nano-Al2O3(+) 5 52.0 27 5% Pd Precipitated on nano-Al2O3(+) + H2Oi 5 31.0 a nano-Al2O3(+): nanoparticle alumina (NanoActive Aluminum Oxide Plus), nano-Al2O3(-): nanoparticle alumina (NanoActive Aluminum Oxide). Unless otherwise stat ed the palladium(II) nitrate source is Fluka. b 4-Methylpyridine distillates: 1: 4-methylpyridine distilled over KOH, 2: 4-methylpyridi ne distilled over NaOH, 3: 4-methylpyridine treated with KOH over night before di stilling over the same KOH, 4 and 5: 4-methylpyridine treated with KOH over night then decanted and distilled over fresh KOH. c Product: 4,4-dimethyl-2,2-bipyridine. Raw yield contains a small amount of the terpyridine byproduct: 4,4,4trimethyl-2,2,6-terpyridine as the only byproduct. d There is very little 4,4-dimet hyl-2,2-bipyridine product in these runs. The solids recovered appear to be mostly catalyst and organic byproducts. e Catalyst prepared via precipitation using KOH instead of NaOH. f Alfa Aesar high surface area bimodal -alumina. g 0.09 g of KOH was added to the reaction mixture. h Palladium(II) nitrate source was Alfa Aesar not Fluka. i 0.3 g of H2O was added to the reaction mixture. 44

PAGE 45

Table 2-3 BET surface areas of the supports used in the catalyst preparations. Catalyst Support Surface Area [m2/g] Nano Active Alumina Plus 695 Nano Active Alumina 275 Commercial -Alumina bimodal (Alfa Aesar) 260 Activated Carbon (Calgon F 400) 765 45

PAGE 46

Table 2-4 Results from CO chemisorption meas urements of selected catalysts prepared and tested in the coupling reacti on of 4-methylpyridine. Catalyst CO Adsorbed ( l/g cat) Pd Surface Areaa (m2/g catalyst) Product Yield (g/m2 Pd) Nano-Al2O3(+) 0 0.0 N/A 10% Pd Precipitated on nano-Al2O3(+) 8480 29.4 0.08 5% Pd Precipitated on nano-Al2O3(+) 4550 15.8 0.17 5% Pd Impregnated on nano-Al2O3(+) 1530 5.3 0.06 5% Pd Precipitated on nano-Al2O3(-) 280 1.0 0.29 5% Pd/Al2O3 Commercial 1880 6.5 0.02 5% Pd/C Commercial 2530 8.8 0.20 5% Pd Precipitated on Activated Carbon 310 1.1 0.68 5% Pd Impregnated on Activated Carbon 240 0.8 0.41 5% Pd Precipitated on Commercial -Al2O3 f 140 0.5 3.4 a The Pd surface area has been calculated assumi ng a Pd:CO stoichiometry of 2:1 and a surface atom density of 1.42 x 1015 atoms/cm2. 46

PAGE 47

Table 2-5 Results from experiments at different reaction times to probe homogeneous versus heterogeneous catalysis. Entry Catalyst Description a Raw Yield [g/g Pd] b 1st Day Raw Yield [g/g Pd]b Recovered Catalyst c 1 5% Precipitated Pd/nano-Al2O3(+) (1st day and 4th day) 17 24 2 5% Pd/C Commercial (1st day and 4th day) 32 5 Recovered Reactant Mixture d 3 5% Precipitated Pd/nano-Al2O3(+) (1 day with and 3 da ys without catalyst) N/A 22 4 5% Pd/C commercial (1 day with and 3 days without catalyst) N/A 30 a nano-Al2O3(+): nanoparticle alumina (NanoActive Aluminum Oxide Plus). b Raw yield of 4,4-dimethyl-2,2-bipyridine, which contains a small amount of the terpyridine byproduct: 4,4,4trimethyl-2,2,6-terpyridine as the only byproduct. c. The 1-day yield is the product recovered during filtration after the first 24 hours. The 4-day yield is the product obtained from a recovered 1-day catalys t after reaction with fresh 4-methylpy ridine for three additional days. d Reaction was run for 24 hours with catalyst. The reac tion mixture was filtered and the catalyst removed. The filtrate was then returned to the reaction conditions for an additional 72 hours. 47

PAGE 48

CHAPTER 3 NANOPARTICLEAND POROUS-META L-OXIDE-SUPPORTED PALLADIUM CATALYST FOR THE OXIDATIVE COUPLING OF 4-METHYLPYRIDINE 2 3.1 Introduction Palladium is a transition metal that has the ability to induce C-H activations in hydrocarbons and aromatic systems [57]. The effi cacy with which palladium inserts into the CH bond of methane, which is the strongest of C-H bonds, is evident in the large number of publications involving palladium-catalyzed me thane oxidation [9,11,12]. While partial and complete oxidation reactions of hydrocarbons are ve ry important, reactions that can lead to C-C coupling after a C-H activation step are of particular interest [ 58]. Naturally, C-H activation and C-C coupling of CH4 would be a highly desirable method to produce chemicals directly from methane. However, C-C coupling of aromatic compounds after a direct C-H activation step is also of significance in the synthesis of fine chemicals and pharmaceuticals. An important example is the oxidative coupling of 4-methylpyrid ine to 4,4-dimethyl-2,2-b ipyridine. This is a simple one-step process in which the bipyridine is formed directly from the pyridine reactant and the only by-products are water a nd the terpyridine. Consequentl y, this is an environmentally friendly reaction since no solven t or halogenated compounds are n eeded. In addition to reducing halogenated byproduct salts or compounds, halogenate d derivatives are usually more expensive, with 2-bromo-4-methylpyridine available through reaction with commercially available 2-amino4-methylpyridine at ~$200/kg [46], comp ared to the non-halogenated analogue, 4methylpyridine, which is relativel y inexpensive at less than $40/kg [47]. Th e disadvantages of 2 Luke M Neal, Daniel Hernandez, Helena Weaver, Manuscript in preparation to be submitted in the near future, Unpublished copyright (2008). 48

PAGE 49

this reaction are the slow reaction rate and the deac tivation of the catalyst [5]. The low yields of bipyridine product over the palladium (or Raney ni ckel) catalysts [1-4] or the high prices of halogenated reactants results in prices of 4,4 dimethyl-2,2-bi pyridine in excess of $5,300 [31]. This prohibits large scale use of the 4,4 dimethyl-2,2-bipyrid ine, and its transition metal complexes. This is unfortunate since transiti on metal complexes with bipyridine ligands have many interesting photoand electrochemical [ 20,21,23,24,43,45] or catalytic [25-29] properties. The oxidative coupling of 4-methylpyridine is th us a useful probe reaction on which to test new catalyst formulations to find more efficient catalysts. In heterogeneous catalysts th e active metals are normally supported on some type of high surface area support. This is done to increase the surface area of the ac tive metal and to reduce sintering of the active metal. It has also been shown that the support can interact with the active metal and alter the catalytic activity of the result ing catalyst. Consequent ly, a large number of different supports have been used for palladium catalyst systems and so me important palladiumsupport interactions have been identified. For example, carbonsupported palladium catalysts are commonly used for hydrogenation reactions, whil e alumina-supported catalysts are normally used for complete oxidation reactions [7], incl uding methane oxidation [8]. Silica and modified silica supports are utilized for reactions dependent upon acidic sites such as ethane oxidation to acetic acid [15] and NO reduction [60]. Pd/ZrO2 and Pd/MgO have been used for aromatization of hexane [19]. Palladium on TiO2 has been studied for selective oxidation of acetylene to ethylene [61] and the selective hydrogenation of al kadienes [14]. The type of phase of titania (anatase or rutile) has been re ported to affect the CO and H2 uptake of palladium catalyst with similar Pd sizes [14]. Pd/ZnO has been studied fo r use in steam reforming applications [62], and has been shown to have longer ca talyst life than Cu catalysts in methanol reformation [63]. 49

PAGE 50

Pd-Sn catalysts are used for denitration of contaminated water [64] and Pd/SnO2 have been studied in the low-temperature oxi dation of methan e [12,13]. Pd/CeO2 are efficient lowtemperature CO oxidation catalysts [16,17]. This is likely due to strong Pd-CeO2 interactions, which can supply oxygen to Pd at Pd-CeO2 interfaces and affect the reduction and oxidation properties of Pd by lowering the temperature of Pd oxidation. As a result oxidation reactions are facilitated on Pd/CeO2 catalysts compared to palladium supported on non-reducible oxides [65]. The use of ZrO2 and ZrO2/CeO2 in Pd catalysts has demonstr ated similar properties [11,18,19]. The oxygen mobility of reducible oxides may be very beneficial in the oxidative coupling of 4methylpyridine, since the catalysts are known to re duce during the reaction [5]. Considering the number of supports that have b een used effectively in palladi um-catalyzed reactions, we are interested in how the properties of the catalysts can be altered by using different catalyst supports. In particular, we are interested in nanopart icle oxides supports and if these supports lead to catalytic properties that are different from traditional high surface area, porous supports. While early research indicated that palladium on alumina exhibits poor activity [1], more recent results have demonstrated that palladi um precipitated onto alumina nanoparticles [nAl2O3(+)] is not only active, but one of the best catalysts found for this reaction [66]. The maximum isolated yield reported for a commercial 5% Pd/C is ~2 g product per g of catalyst [38], with yields of 1.5-2 g/g for 10% Pd/C catalysts being mo re common [5,39,40]. In contrast, the 5% Pd/n-Al2O3(+) catalyst gave yields in excess of 2.5 g/g of catalyst [5]. In addition to the excellent n-Al2O3(+)-supported catalyst, previous work re vealed that 5% Pd precipitated onto a traditional porous alumina support is an active cata lyst but with only half the yield compared to the n-Al2O3(+) catalyst. Our hypothesis is that the high activity of the 5%Pd/n-Al2O3 catalyst is due to the high number of low coordination site s, such as corners and edges [6] that many 50

PAGE 51

nanoparticles and some traditional high surf ace area, porous supports have. These low coordination sites may cause a stronger interact ion between the support oxide and the deposited active metal than with most conventional supports, which in turn can alte r the catalytic activity of the resulting catalyst. In the aluminasupported catalysts, no favor able strong palladiumsupport interactions are e xpected. We therefore decided to prepare palladium catalysts supported on nanoparticle ZrO2, CeO2, ZnO, CuO, SnO2, since these oxides have revealed favorable interactions with palladium (see preceding literature review). Traditional supports (porous SiO2 and TiO2) were also included as well as their nanoparticle analogue s and MgO plus CaO nanoparticle oxides, to probe different acidic and basic support prope rties as well as the differences between traditional and nanoparticle oxide supports. Considering that the nanoparticle alumina is a relatively expensiv e support at a price of nearly $700/kg on a kg purchase basis (NanoScale NanoActive Alumina Plus), it is also important to search for a less expensive support that can give comparable yields to find an economically viable catalyst. The main objectives in this work are to 1) determine the effect s of the support on the catalytic activity of palladium supported on porou s and nanoparticle oxides, 2) determine if the acidic or basic properties of the supports are important for the preparation of an active catalyst, and 3) identify any catalyst s upports competitive with the n-Al2O3(+) support in terms of cost and/or activity. 3.2 Experimental 3.2.1 Catalyst Preparation The catalysts were prepared using commercia lly available nanoparticle oxides supplied by NanoScale Materials Inc. [67] and Nanostruc tured & Amorphous Materials, Inc (NanoAmor) [68] as well as some commercially available trad itional supports [69]. The properties of these support oxides are given in Table 3-1. 51

PAGE 52

The catalysts where prepared by precipitati on. In this method, the support was dispersed into a solution of palladium nitr ate. Porous oxides pellets we re ground before dispersion. The mixture was then titrated with a NaOH solution, which formed Pd(OH)2 on the support [51]. The amount of NaOH used in these experiments corresponds to 50% stoichiometric excess based on the amount of palladium nitrate used. The resulting mixture was aged overnight at room temperature before it was filtered. The recove red catalyst was rinsed by stirring in water overnight, followed by another filtration. The cata lyst was then dried overnight at 105C and calcined at 350C or 450C for 3 hours. 3.2.2 Reaction Conditions and Product Recovery The 4-methylpyridine (Aldrich or Across) was doubly distilled over KOH prior to use. In a typical reaction run 0.7 g of catalyst was placed in a round bottom flask along with 7 g of the distilled 4-methylpyridine. The reaction mixture was evacuated and an oxygen atmosphere introduced before it was heated to the boiling point (145C). The reaction proceeded under reflux for 72 hours. After a complete reaction th e flask contents were filtered using a glass micro-fiber filter and washed with chloroform to dissolve the product. The chloroform, water and unreacted 4-methylpyridine were rem oved using a rotary evaporator. The standard deviation was estimated from 5% Pd precipitated onto n-Al2O3(+) (6 samples) and p-TiO2 (5 samples). These were found to have average yields of 2.5 0.15 g / g catalyst for n-Al2O3(+) and 2.6 0.25 g / g catalyst for p-TiO2 (from this and previous work (See chapter 2). This amounts to a % standard de viation of ~6.5% and ~%10 respectively. These numbers are typical for the catalys ts used in this study, except for the commercial Pd/C catalyst. In the case of the Pd/C catalyst, the variation was closer to 20%, which in this case is due to the fact that the PdO:Pd ratio varies on this catalyst Since the 6.5% and 10% are typical values of the standard deviation the more conservative 10% value was used in the research. The variation 52

PAGE 53

in yield is believed to be caused by variations in palladium precursor and reactant distillate qualities, as well as (to a lesser extent) variations in support and extraneous variables, such as varying chlorine contents. It was found in previous work (See chapter 2) that the palladiumnitrate precursor and the quality of the 4-methlypyridine distil late had large effects upon the yields, and are, perhaps, the most important parameters to assure reproducible results. 3.2.3 Catalyst Characterization The surface areas of supports as received and the prepared catalysts were characterized by multipoint BET isotherms using a Quantachro me Nova 1200 instrument. Chemisorption measurements to characterize act ive surface area of the catalysts and the number of acidic and basic sites of the supports were performed on a ChemBET 3000 from Quantachrome Instruments. Reduction in hydrogen followed by CO titration was used to characterize the active palladium surface area. Since PdO is believed to be the active phase, or a necessary precursor for this reaction [5], measuring the Pd surface area after reduction of the PdO phase is not necessarily a good measure of the active surface area. This is pa rticularly the case when the surface consists of a mixture of Pd and PdO phases. However, in our prepared catalysts the palladium phase after calcination is PdO in all cases and no partially reduced PdO has been observed. Furthermore, reductions are performed using mild conditions (170 C in a stream of 5% hydrogen) to minimize sintering of the formed Pd phase. A temperature of 170 C was chosen to be close to the temperatures e xperienced by the catalyst during reaction. The Pd particle size was calculated using: [53] ma mm avg avN CkV SV1 Sav Average stoichiometry; CO/Pd = 1. This valu e was selected to be conservative and report a lower Pd surface area (or dispersion) rather than one that may be higher than the true value. 53

PAGE 54

Some research has indicated that a 1:2 CO:Pd stoichiometry may be more accurate for small particles [53], and this would give higher surface areas than t hose reported here. k Shape factor. A value of 5 was used in the current study. It corresponds to a cube with one side attached to the support. This is re ported to be a reasonable approximation for Pd crystals in this size range [53]. Vm Molar volume Na Avogadros number m Metal density Vg Volume of gas adsorbed av Average particle size Cm Surface density of metal atoms. Values between 1.27 1015 and 1.44 1015 atoms/cm2 have been reported in the lit erature [53]. The 1.44 1015 atoms/cm2 assumes roughly spherical particles and was thus used in the current study. The formula can be rearranged to give metal surface area: amm avg mNCV SV A The basic sites of the supports were probed by titration with CO2, and acidic sites were probed by titration with anhydrous ammonia in a str eam of helium. Prior to titration, the support samples were outgassed in a flow of nitrogen at 105 C for an hour. Although some of the more basic supports may have adsorbed atmospheric CO2 that does not desorb at 105 C, the possibility of sintering precludes th e use of higher temperatures during out-gassing. 54

PAGE 55

3.3 Results and Discussion 3.3.1 Catalytic Activity All prepared catalysts were subjected to cat alytic activity measurements. The results from these experiments are summarized in Table 3-2. Entries 1 and 2 have been added for comparison. Entry 1 reports a typical yield for th e commercial Pd/C catalyst and Entry 2 is the result from a nanoparticle alumina catalyst, and is the best yield reported for this reaction to date [66]. As can be seen in Table 3-1, the surface ar eas of the various supports included in this study covers a wide range from ~700 down to 35 m2/g. If the only function of the support is to provide the surface area onto which the palladium is disperse d, the yields would be expected to decrease with the surface area of the support. This is indeed observed for some of the supports used in this study (see Figure 3-1A). In particular, catalysts supported on n-Al2O3(+), n-MgO, n-SiO2, Al2O3 and n-SnO2 give yields that appear to be a linear function of the support surface area. Fitting this data to a strai ght line gives a regression coe fficient of 0.99 (Line equation: Yx = 0.058 0.004 + 7.9 2.2SAx, where Yx is the yield obtained over support x and SAx is the surface area of that support). The n-TiO2 support also is also close to this line and including the yields obtained from the Pd/n-TiO2 catalyst gives a linear regression coefficient of 0.98. Evidently, there are also a number of catalysts that result in higher yields than expected from the support surface areas. Th is may be expected in cases where there are strong palladiumsupport interactions, since favorab le interactions can induce, for example, higher Pd surface areas, or a more catalytically active Pd speci es, compared to noninteracting supports. Consequently, strong palladium-support interacti ons are most likely the reason why the yields obtained from the catalysts supported on n-ZrO2+CeO2, n-ZrO2, n-ZnO and n-CeO2 are high, or reasonably high, despite their lo w support surfac e areas (70 m2/g and below). These oxides have been shown to exhibit strong interactions with palladium in other catalysts [11,19,65]. In the 55

PAGE 56

light of this, it is unexpe cted that the CuO and SnO2 do not result in catalysts with higher activities, since Cu(II) and Sn(II) are often added as co-catalysts to palladium in homogeneous reaction systems [70-73]. In f act, the Sn(II)/Pd(II) pair has been shown to be effective in homogeneous C-H activation and C-C coupling reactions [74]. The yields obtained from the catalysts supported on n-CaO and the n-Al2O3(-) are also unexpectedly low, which appears to indicate some unfavorable interactions between the palladium and these supports. In contrast, the yields obtained from the catalysts supported on nAl(OH)3 and the traditio nal supports p-SiO2 and p-TiO2 are higher than predicted from their corresponding surface areas according to the equation derived above. In this case, it appears that there again are some favorable interactions betw een the palladium and the support. However, in the case of SiO2 and Al(OH)3 no specific palladium-support inte ractions have to our knowledge been reported. 3.3.2 Catalyst and Metal Surface Areas Since the support surface area does not n ecessarily correlate with the amount of palladium available at the surface, it is important to determine the palladi um surface areas of the catalysts. For reactions that are not structure sensitive it is expected that there is a linear correlation between the catalytic activity and the palladium surface area. Carbon monoxide adsorption measurements were therefore perf ormed on reduced catalysts to determine the palladium surface areas of all catalysts. The re sults are presented in Table 3-3. From these measurements the dispersions of the catalysts, i.e. the fraction of Pd atoms at the surface (compared to all palladium atoms on the catalyst), were calculated. The palladium dispersions obtained on several of the prepared catalysts are high, ranging from 20 to over 40% (Table 3-3). By comparison, the dispersion on th e commercial 5% Pd/C is 23%. It is noteworthy that such high dispersions (relative to the commercial Pd /C catalyst) can be obtained from a simple 56

PAGE 57

precipitation method. This indicates strong promise for use of nanoparticles as catalyst supports also in other reaction systems. When analyzing the data it is important to k eep in mind that PdO and not Pd metal is the active site on these catalysts. This is not n ecessarily detrimental since the temperatures were kept low during the reductive pretreatment to avoi d palladium particle growth. However, it is possible that different oxide s upports could exhibit different amounts of palladium sintering during the reduction process. This together with structure sensitivity can lead to deviations from a linear correlation between the activ ity and the Pd surface area. When plotting the Pd dispersion versus th e support surface areas (Figure 3-1B) and comparing this to the plot of yield versus suppo rt surface areas (Figure 3-1A), some of the high and low catalyst yields can be explained. For ex ample, the low yields obtained from catalysts supported on CuO, CaO and n-Al2O3(-) are evidently due to very low palladium surface areas on these catalysts. The palladium surface area on the n-SnO2 support is also lowe r than expected if this support indeed would induce metal-support interactions. As expected, the n-ZrO2+CeO2, nZrO2, n-ZnO and n-CeO2 supports do result in catalysts w ith high palladium surface areas, compared to the respective support surface area. The palladium surface area as a fraction of the total surface area ranges from 10-19% on these catalysts. By comparison the palladium surface areas of the n-Al2O3(+)-supported catalyst and the commerci al palladium on carbon catalyst are 10% and 1% of the total surface areas, respectively. While the reported dispersion on the Pd/nCeO2 appears abnormally high, this is likely the result of strong metalsupport interactions. Since the CeO2 is a reducible oxide, it is possible that some of the CeO2 is reduced in the reduction process due to the Pd-CeO2 interactions. Evidence of oxygen mobility on CeO2 and transfer of oxygen from CeO2 to palladium has been observed in, for example methane oxidation 57

PAGE 58

over Pd/CeO2 [18]. The Pd/CeO2 system could also be similar to CuO/CeO2 catalysts where it has been reported that surface CeO2 can be reduced at low temperat ures (less than 200C) [75]. It was verified that no CO is adsorbed on a n-CeO2 support after reductive treatment if palladium is not present on the surface. If the Pd induces low-temperature CeO2 reduction, it is likely that CO adsorbs on these reduced CeOx sites, in addition to Pd sites during the CO titration process. This would result in very high CO adsorption values, and consequently abnormally high Pd surface areas (or dispersions). A high Pd dispersion on the porous TiO2 support can also explain the high yield obtained from this catalyst. Whilst the high dispersion on this support is likely due to Pd-TiO2 interactions, it is evident that th ese interactions differ between p-TiO2 and n-TiO2 supports. The Pd dispersions are close to th e same on these two supports, while the support surface areas are considerably different, 505 (n-TiO2) versus 120 (p-TiO2) m2/g. Some other unexpected results include the low Pd surface areas on the p-SiO2 n-SiO2, MgO supports. Considering the reasonably high yields obtained from palladium on these supports, it would be expected that the palladium surface areas are higher than the measured values. Thus, either the palladium sinters more on these supports compared to the other supports, or a greater fraction of the palladium on the surface of these catalysts is active compared to the palladium on the surface of for example the n-Al2O3 (+) support (see Figure 3-2 and Ta ble 3-3). From Figure 3-2 it is evident that there is a correlati on between the catalytic activity (product yield) and the palladium surface area (dispersion). In gene ral, a higher dispersion results in a higher yield. However, there are some deviations from this relationshi p and the correlation is definitely not linear (Figure 3-2). It appears that se veral supports inte ract with the palladium and alter the catalytic properties of these catalysts. 58

PAGE 59

3.3.3 Acidic and Basic Sites In order to further probe the how the prope rties of the supports affect the palladium dispersion and the catalytic activity of the catalysts, the amounts of acidic and basic sites of the supports were investigated. This wa s done by measuring the adsorption of NH3 and CO2 on the different catalysts (see Table 3-4). As is evident in Figures 3A and 3B there is no simple correlation between the catalyst dispersion and the acidic or ba sic sites on the supports. As expected the n-CaO had a large uptake of CO2 (due to facile CaCO3 formation). The amount of CO2 adsorbed by the calcium (4.4 ml/g) is more consistent with surface adsorption rather th an bulk carbonate formation (~400 ml/g). While moderately basic supports (such as -Al2O3 and MgO) evidently can give reasonable product yields (Figure 3-1 A), it appear s that the Pd dispersions are lower on these compared to less basic or neutral supports (Figur e 3-1 B). Considering that the Pd dispersion on the CaO support is very low, avoiding moderately to highly basic supports when preparing welldispersed Pd catalysts using the ba se precipitation method may be advi sable. It also appears that avoiding highly acidic supports is recommended when preparing well-dispersed palladium catalysts. The support with the highest ammonia adsorption (n-SiO2) resulted in a catalyst with a palladium surface area below that expected for it s support surface area (Figure 3-1 B). However, despite a lower palladium surface area on the n-SiO2 support, the yield obtained from this catalyst is in the range e xpected for this catalyst considering its support surface area (Figure 3-1 A). Consequently, the average activity of the pa lladium on the surface of th is catalyst is higher than the activity of for example the palladium on the n-Al2O3(+) support (see the turnover numbers in Table 3-3). The sec ond most acidic support, the n-Al(OH)3, also results in a catalyst with higher turnover numb er than the Pd/n-Al2O3(+) catalyst, and the same is true for the most 59

PAGE 60

basic supports (Table 3-3). Cons equently, highly acidic or basi c supports appear to result in lower Pd dispersions, but the palladium that is on the surface is more active than the palladium on our reference Pd/n-Al2O3(+) catalyst. It is easier to see how an acidic support (electron acceptor) can lead to more electrophilic, and thus more reactive, palladium, than how the basic support leads to more active Pd species. For ex ample, a density functional calculation study of Pd atoms on -Al2O3 [110] found that single Pd atoms intera ct strongly with tetragonal Lewis acid sites [76]. Evidently, there are some co mplex interactions between palladium and the supports on the catalysts included in this inve stigation. Consideri ng the turnover numbers reported in Table 3-3, more active catalysts should result if the dispersion on the -Al2O3 support can be increased. 3.3.4 Estimated Cost of Selected Catalyst In the search for a more effective catalyst the economics must also be considered, i.e. cost of the catalyst compared to the yield obtained. For example, this reaction does proceed over Raney nickel, which is a considerably less expe nsive catalyst. However, the yields over the Raney nickel catalyst are also low per gram of catalyst [1-4] and the reaction is more involved since the catalyst must be act ivated and the resulting catalys t after reaction is pyrophoric. Consequently, palladium would be the active meta l of choice if higher yields and economically viable catalysts can be develope d. Although the main cost of the catalysts is due to palladium, the cost of nanoparticle oxide supports can be signi ficant, as can be seen in Table 1. Since the nAl2O3(+) support is the most expensiv e of the oxide supports used in this study, it is possible that other supports can be competitive with the Pd/n-Al2O3(+) catalyst if the cost of the catalyst per g of product is considered. To allo w for variations in the price of palladium, as well as the cost of preparing palladium nitrate, a more expensive palladium cost basis ($450/oz) was used for calculating the cost of catalyst per gram of product (Table 4-5), even though the price of 60

PAGE 61

palladium has recently been as lower than $250/oz [77]. At these prices the Pd/p-TiO2, Pd/nZrO2, and Pd/n-ZnO are catalysts competitive with the Pd/n-Al2O3(+) catalyst. At a commercial scale the amount of catalyst used and the conversi on of starting material also comes into the economic analysis equation, since a larger amount of catalyst and a lower conversion would result in larger reactor sizes (higher capital costs) for the same production rate. This naturally favors the most active catalyst by minimizing reac tor size in addition to palladium usage. Consequently, the lower yields of the Pd/ZrO2 and Pd/n-ZnO catalysts likely offset any cost savings from the support compared to the Pd/n-Al2O3(+) catalyst. Conse quently, the Pd/p-TiO2 is likely the only support that has potential to exhibit better economics than the Pd/n-Al2O3(+) catalyst. This said, some of the other support oxides with moderate price and relatively high activities, such as the n-MgO, n-TiO2, n-CeO2, n-ZrO2 and n-ZrO2+CeO2 may make their corresponding catalysts competitive with the n-Al2O3(+) support through optimization and, perhaps, promoter use. 3.4 Conclusions Palladium supported on n-ZrO2, n-ZrO2(+10% CeO2), n-CeO2, n-ZnO, n-MgO, n-TiO2 and p-TiO2 supported catalysts all dem onstrated high catalytic activ ities. While none of the catalysts outperformed the activity of the precipitated 5% Pd/n-Al2O3(+) catalyst developed in previous research, several of th e catalysts outperformed the hi ghest 2g/g yields reported for activated carbon catalysts (Table 3-2). In general there appeared to be two groups of supports gi ving highly active catalysts; 1) supports with high surface areas and 2) suppor t oxides known to resu lt in strong Pd-support interactions. For example, supports with ve ry high initial surface ar eas (more than 450 m2/g), such as n-TiO2, n-SiO2, n-MgO and n-Al2O3(+), were generally quite active (>35 g product/g Pd) and high surface area supports (450 > SA > 200 m2/g), such as porous SiO2 and Al2O3 gave 61

PAGE 62

moderate yields (32 and 24 g/g Pd respectively). The other group of active catalysts includes, for example, palladium supported on n-ZrO2, n-ZrO2+n-CeO2, n-CeO2, and n-ZnO, which all have relatively low surface areas. As a result of stro ng metal-support interactions these supports resulted in relatively high Pd surface areas co mpared to the surface area of the support (10-19% of the total SA), which in turn gave higher yields than expected from the support surface areas (Figure 3-1A). It is also po ssible that these supports can result in electronic Pd-support interactions that positively affect the catalytic activity. This w ould explain why the activities for some of the catalysts are higher than what would be expected fr om the Pd dispersion on these catalysts. No simple correlation between the catalytic activity and the support surface area or the palladium dispersion could be iden tified. It is also evident that there is no correlation between the acid or basic sites of the supports and the palladium disper sion or the catalytic activity. Despite the lack of correlation between activity and aci dity or basicity, th e CaO support with a high CO2 uptake (basic support) resu lted in a catalyst with ve ry low dispersion. Using a precipitation procedure where a base is added to palladium nitrat e solution it may therefore be advantageous to avoid highly basic supports. Al so, using highly acidic support, such as n-SiO2, did not result in poor catalysts. While the n-SiO2 support resulted in a palladium dispersion lower than what would be expected from its surface area, this catalyst had a high turnover frequency. The lack of linear correla tion between the catalytic activity and the palladium surface area indicates that the reaction is structure sensitive, i.e. some palladium species are more active than others. This is par ticularly evident in the n-Al2O3(+)-supported catalysts, where there was no significant difference in produc t yield between the 5% and the 10% Pd loadings [66]. Our 62

PAGE 63

hypothesis is that the most active palladium species are very small Pd/PdO particles, which may form at corner and edge sites on the supports due to Pd-support interactions. Nanoparticle oxide supports, and to some extent more traditional high surface area supports (such as bimodal Al2O3, p-SiO2 and p-TiO2), would have a large number of th ese corner and edge sites. Once these support sites are saturate d with palladium, any additional palladium would be deposited on non-active support sites, or on top of the active palladium. Th e number of corner and edge sites on the support may be more important on non-interacting supports. On, for example, reducible oxides there ar e already active support sites, wh ich can result in high Pd dispersions and highly active Pd species. In summary, the search for more active and economically competitive catalysts, identified a few promising supports. In particular, the n-ZnO and p-TiO2 result in catalysts that are significantly less expensive but with comparable activity to Pd/n-Al2O3(+) catalysts. The high surface area n-MgO support also result in a catalyst with reasonabl y high activity. This support may, thus, prove useful in future catalyst development where the effects of additives are investigated. This is particul arly important in cases where the promoters are known to form inactive aluminates with alumina supports. 63

PAGE 64

Table 3-1 Suppliers, properties and pri ce of the supports us ed in the study Oxide Supplier a SA m2/g b Crystallite Diameter c [nm] Cost d $/kg n-Al2O3(+) NanoScale 695 NA f 694.10 n-MgO NanoScale 685 4 424.60 n-TiO2 NanoScale 505 NA f 277.45 n-SiO2 NanoAmor 490 15 180.00 n-Al2O3(-) NanoScale 275 NA f 68.45 -Al2O3 Alfa Aesar 260 NA 115.00 p-SiO2 Alfa Aesar 240 NA f 116.00 p-TiO2 Alfa Aesar 120 NA 112.00 n-CaO NanoScale 100 20 65.80 n-ZnO NanoScale 70 10 87.70 n-CeO2 NanoScale 60 7 197.30 n-ZrO2+10%CeO2 NanoAmor 45 20-30 450.00 n-Al(OH)3 NanoAmor 40 15 320.00 n-ZrO2 NanoAmor 35 29-68 395.00 n-SnO2 NanoAmor 35 55 270.00 n-CuO NanoScale 35 8 Unavailable e a Suppliers; NanoScale, NanoActive compounds [ 66]; NanoAmor: [68]:3.Alfa Aersar: [69] b Determined by BET (Nova 1200) c As specified by supplier (determined from XRD) NA = not available/amorphous substance. d Price FOB for 1 kg quantities 3/13/08 e Discontinued Product f Amorphous 64

PAGE 65

Table 3-2 Catalyst pretreatment temperatures and cat alytic activities for all prepared catalysts Entry Catalyst: 5% Pd on Calc. T [C] Yield [g/g cat] % Yield [g/g Pd] 1 Activated C 1,2 NA 1.74 36 2 n-Al2O3(+) 2 350 2.5 52 3 n-MgO 350 2.3 46 4 n-TiO2 350 1.6 31 5 n-SiO2 350 1.6 35 6 n-Al2O3(-) 2 350 0.3 5.8 7 -Al2O3 2 350 1.5 24 8 p-SiO2 350 1.6 32 9 p-TiO2 350 2.6 52 10 n-CaO 350 0.15 2.8 11 n-ZnO 350 2.1 44 12 n-CeO2 350 1.5 29 13 n-ZrO2+10%CeO2 450 2.7 58 14 n-ZrO2+10%CeO2 350 2.2 44 15 Al(OH)3 350 0.8 17 16 n-ZrO2 350 2.3 47 17 n-SnO2 350 0.5 10 18 n-CuO 350 Fail 3 NA 1 Commercial catalyst. 2 Results from previous work [66] 3 No measurable amount of product could be recovered 4 Yield % 65

PAGE 66

Table 3-3 Catalyst surface areas and other support properties Catalyst 5% Pd Catalyst SA [m2] CO Adsorbed [ mol/g cat] % Dispersion1 Pd SA1 [m2/g] Pd % SA Pd Diameter [nm] Product/Pd SA [mg/m2] Turnover Number [mol prod. / mol surf. Pd] C (commercial)1 695 110 23.4 5.2 0.7 4 345 90 n-Al2O3(+)2 180 205 44 9.7 5 2.0 165 70 n-MgO 85 105 22 5.0 6 4.0 45 120 n-TiO2 210 185 39 8.8 4 2.5 180 50 n-SiO2 120 65 14 3.1 3 7 585 140 n-Al2O3(-)2 155 15 3 0.7 0.5 30 420 110 g-Al2O3 2 200 6.5 1 0.3 0.2 68 3870 1000 p-SiO2 230 20 4 0.9 0.4 22 1685 430 p-TiO2 115 195 41 9.2 8.0 2.0 280 70 n-CaO 30 4.5 1 0.2 0.7 100 715 180 n-ZnO 35 75 16 3.6 10 6 620 150 n-CeO2 60 245 52 12 19 1.5 145 30 n-ZrO2+CeO2 40 145 31 6.9 17 3.0 420 110 Al(OH)3 55 25 5 1.2 2 18 1090 170 n-ZrO2 35 95 20 4.5 13 4.5 355 130 SnO2 20 15 3 0.7 4 30 705 180 CuO 30 1 0.2 0.05 0.2 440 0 0 1 The Pd dispersion, surface area and average partic le diameter and SA have been calculated using a conservative value of 1:1 as the CO:Pd stoichiometry and a surf ace atom density of 1.27 1015 atoms/cm2. Some research has indicated that 2:1 stoichiometry with surface atom density of 1.44 1015 and may be more accurate for small particles [53] 2 Results from previous work [66] 66

PAGE 67

Table 3-4 Adsorption of NH3 and CO2 on the different catalyst supports Support NH3 [sccm1/g support] CO2 [sccm1/g support] n-Al2O3(+) 9.0 0.75 n-MgO 3.4 2.0 n-TiO2 4.0 0.6 n-SiO2 53 0.0 n-Al2O3(-) 8.3 0.6 -Al2O3 1.6 1.75 p-SiO2 5.7 0.0 p-TiO2 9.9 1.0 n-CaO 2.8 4.4 n-ZnO 14 0.2 n-CeO2 7.3 0.8 n-ZrO2+10%CeO2 2.6 1.1 Al(OH)3 25 0.1 n-ZrO2 4.7 0.3 n-SnO2 2.3 0.02 n-CuO 3.3 0.4 1Standard cubic centimeter: 1 ml at standard temperature and pressure 67

PAGE 68

Table 3-5 Catalyst cost Support Support Cost [$/ g Product] 1 Cost of product [$ cat./g product] 2 n-Al2O3(+) 0.29 0.57 n-MgO(+) 0.18 0.49 n-TiO2 0.16 0.62 -Al2O3 0.09 0.69 p-TiO2 0.04 0.32 n-ZnO 0.04 0.37 n-CeO2 0.11 0.54 n-ZrO2+10% CeO2 0.20 0.53 n-ZrO2 0.16 0.48 1 Price FOB for 1 kg quantities 3/13/08 2 At a price of palladium at $450/oz. Calculation excludes cost of de rivation to a nitrate precursor 68

PAGE 69

0100200300400500600700800900 0 10 20 30 40 50 60 An-CuO p-SiO2n-AlOH3 n-CeO2 n-ZnO n-ZrO2n-Al2O3(-) p-TiO2 -Al2O3n-CaO n-SiO2n-ZrO2+CeO2n-MgO Product Yield [g/g Pd]Support Surface Area [m2/g] n-Al2O3(+) n-TiO2 n-SnO2 0100200300400500600700800900 0 10 20 30 40 50 60 B n-CuO p-SiO2n-AlOH3 n-CeO2 n-ZnO n-ZrO2n-Al2O3(-) p-TiO2 -Al2O3n-CaO n-SiO2n-ZrO2+CeO2n-MgO Pd Dispersion [%]Support Surface Area [m2/g] n-Al2O3(+) n-TiO2 n-SnO2 Figure 3-1 Support surface area effects, A) Product yields, B) Pd dispersions 69

PAGE 70

0102030405060 0 10 20 30 40 50 60 n-CuO p-SiO2n-TiO2 n-AlOH3n-ZnO n-ZrO2n-Al2O3(-) p-TiO2-Al2O3n-CaO n-SiO2n-ZrO2+CeO2n-MgO Product Yield [g/g Pd]Dispersion [%] n-Al2O3(+) AC n-CeO2 n-SnO2 Figure 3-2 Palladium dispersion vs. product yield 70

PAGE 71

012345 0 10 20 30 40 50 60 An-CuO p-SiO2n-AlOH3 n-CeO2 n-ZnO n-ZrO2n-Al2O3(-) p-TiO2-Al2O3n-CaO n-SiO2n-ZrO2+CeO2n-MgO Dispersion [%]CO2 Adsorption n-Al2O3(+) n-TiO2 n-SnO2 0102030405060 0 10 20 30 40 50 60 Bn-CuO p-SiO2n-AlOH3 n-CeO2 n-ZnO n-ZrO2n-Al2O3(-) p-TiO2-Al2O3n-CaO n-SiO2n-ZrO2+CeO2n-MgO Dispersion [%]Ammonia Adsorption n-Al2O3(+) n-TiO2 n-SnO2 Figure 3-3 Support pH vs. pa lladium dispersion A) CO2 adsorption B) Ammonia adsorption 71

PAGE 72

CHAPTER 4 USE OF ZIRCONIA, CERIA, AND ZINC OXIDE AS ADDITIVE S IN NANOPARTICLEOXIDE-SUPPORTED PALLADIUM CATALYST FOR THE OXIDATIVE COUPLING OF 4METHYLPYRIDINE 1 4.1 Introduction Bipyridines possess the ability to coordina te to transition metal cations and form complexes with distinct photochemical a nd catalytic properties [20,23,42,43]. Bipyridine complexes of ruthenium are especially interesting for applications in organic light-emitting diodes [23] and chemiluminescence detection systems [21,23,24]. Various ca talyst systems also use transition metal complexe s with bipyridine ligands [25,26], including oxidative carbonylation [27], the Kumada-Co rriu reaction [28], and the Su zuki cross-coupling reaction [29]. Their widespread, large-scale use, howev er, is likely limited by th e cost of bipyridine compounds. For instance, 4,4-dim ethyl-2,2-bipryidine costs in excess of $5,200 per kilogram [31]. Consequently, finding more economical synthesis pathways to bipyridines is desirable. When developing new processes it is also important to consider environmentally friendly reactions. Reactions using no solvents with no byproduct formation, such as halide salts, are preferred. The oxidative coupling of 4-methylpyridine to 4,4-dimethyl-2,2-bipyridine over a palladium is a simple one-step process that uses no solvents nor halogenated precursors (Figure 1-1) with only water and terpyrid ine as by-products. This reacti on thus meets the criteria of an environmentally friendly process. However, the reaction rate is slow, and the catalyst undergoes deactivation during reaction, which limit the produc t yields [5]. Early research focused on 1 Luke M Neal, Justin Dodson, Helena Weaver. Manuscript in preparation, used by permission, unpublished copyright (2008) 72

PAGE 73

palladium on carbon (Pd/C) catalysts with yiel ds for 5 wt% and 10 wt% Pd loadings varying between 1.5 to 2.0 g/g catalyst [ 5, 38-40]. In our previous resear ch it was shown that a 5 wt% palladium precipitated onto nanoparticle alumina [n-Al2O3(+)] gives yields in excess of 2.5 g/g catalyst, which is the best performing palladium catalyst reported to da te [66]. Several other oxide supports, including nanoparticle magnesia (n-M gO) and nanoparticle TiO2, give catalysts with reasonable yields after 5 wt% palladium depo sition, i.e. yields simila r to or higher than the yields obtained from Pd/C catalysts. In addi tion, catalysts supported on nanoparticle ceria (nCeO2), nanoparticle zinc oxide (n-ZnO ), and nanoparticle zirconia (n-ZrO2) were found to produce moderate yields (> 1.5 g/g catalyst) de spite the relatively lo w surface areas of the supports (< 70 m2/g). The high dispersions on the latter nanoparticle oxides indicate favorable metal-support interactions. In this paper, the effects of additives on na noparticle catalysts are investigated. A few selected nanoparticle oxides and promoter oxides have been chosen for the study. The nAl2O3(+), n-TiO2 and n-MgO supports were selected due to their very high surface areas (> 500 m2/g), and, consequently, strong potential for producing catalysts with high palladium dispersions. The addition of promoter oxide s to palladium supported on these nanoparticle oxides could potentially improve the product yields. The strong metal-support interactions of CeO2, ZnO, and ZrO2 indicated by previous studies [see Chap ter 3] suggest that these materials have potential as efficient additives for pallad ium catalysts. These materials are commonly used as catalyst promoters or catalyst supports in a number of systems. For example, Pd/CeO2 catalysts have been studied in CO oxidation reac tions [22,42] and in the methane reforming with CO2 [78]. It has been suggested that CeO2 interacts strongly in Pd cat alysts due to formation of palladium hydroxide at basic site s on the ceria surface [79]. CeO2 also has oxygen storage 73

PAGE 74

abilities and can serve as an oxygen source to palladium [31]. The formation of Pd-O-Ce bonds promotes high metal dispersion and favors activit y at interfacial metalsupport sites on Pd-CeO2 catalysts [80-82]. ZnO promoted Pd catalysts have been used in the water-gas shift reaction with a palladium catalyst [83] and Pd/Z nO are excellent catalyst for th e steam reforming of methanol [21,43,84]. In these catalysts, a Pd-Zn alloy ap pears to form, leadi ng to good metal-support interaction [83] and to improved thermal stabil ity [84]. Pd/ZnO-based methanol reforming catalysts has been shown to have a longer catalyst life than Cu catalysts [4 1]. Using zirconia as a promoter improves the activity and thermal stabi lity in methane combustion [18], and three-way catalysts (TWC) utilizes both ZrO2 and CeO2 to increase the oxygen storage capacity of palladium catalysts [80]. A nother application of Pd/ZrO2 catalysts includes the aromatization of hexane [20]. ZrO2-supported palladium catalys ts have also been applied to dichloromethane oxidation because of their thermal stability and th eir unique surfaces with acidic, basic, reducing, and oxidizing characteristics [85]. Pd/CeO2 and Pd/ZrO2 catalysts have similar oxidizing surface properties [23] and the oxygen storage and mobi lity, as well as the redox effects make these catalysts particularly interesting for oxidative co upling as the catalytic activity of this reaction appears to be limited by the reox idation of palladium. Evidence suggests the PdO is the active phase whilst Pd0 is inactive [5]. Since the catalysts after reaction are r ecovered with surface palladium in a reduced state and very little, if any, ac tivity, assistance in o xygen transfer between the support and the palladium ha s potential to result in highly effective catalysts. When preparing promoted supported catalysts, the additive can be deposited on the support with the active metal, or before the active metal (two step deposition). Therefore, this study also focuses on whether coor sequent ial precipitation of additive a nd palladium on the nanoparticle oxides is the preferred catalyst preparation technique for this reaction. Co-precipitation forms a 74

PAGE 75

uniform distribution of active sp ecies [86]. However, the added oxide may cover the active metal and thus hinder the reaction. Se quential precipitation ensures the active metal gets deposited on the externals surface of the additive/support. The disadvantage of sequential precipitation results mainly from the reduction in the surface area of the promoted support versus the base support. Dependent on the additive/suppor t identity this can lead to inhomogeneous active metal dispersion [86]. Furthermore, two calcination treatments are likely to cause damage to the support via sintering The main objectives in this work were 1) to determine if CeO2, ZnO, and ZrO2 are effective additives for n-Al2O3(+)-, n-MgO-, and n-TiO2-supported palladium catalysts and if the promoting effects are dependent on the nanoparticle oxide used, 2) determine which is the more effective preparation method, coor sequential pr ecipitation of the additive and palladium, and 3) determine if a catalyst with an activity greater than the best catalyst to date, i.e. the 5wt% Pd precipitated onto n-Al2O3(+), can be prepared via promoter addition. 4.2 Experimental 4.2.1 Catalyst Preparation The catalysts were prepared using comme rcially available na noparticles supplied by NanoScale Materials Inc 2 [66]. The precipitatio n method was used to deposit the palladium onto the supports. Using this method, the support was di spersed into an aqueous solution of metal nitrate(s). The mixture was titrated with NaOH to form metal hydroxide( s) on the support [51]. The amount of NaOH corresponded to a 50% stoi chiometric excess based on the amount of metal nitrate(s) used. The resulting mixture was ag ed overnight at room te mperature before being 2 Surface areas: n-TiO2 = 505 m2/g, n-Mg O = 685m2/g, and n-Al2O3(+) = 695m2/g, 75

PAGE 76

filtered. The filtered material was rinsed in deionized water overnight and filtered again. Then the material was then dried at 105C ove rnight and calcined at 350C for 3 hours. All the catalysts had loadings of 5% palladium and 5% additive by weight on metal basis, unless otherwise noted. Two different precipitation methods were used to determine the effect of the additives, co-precipitation a nd sequential precipitation. In the co-precipitation method, the cerium nitrate, zirconium nitrate, or zinc nitrate was dissolved together with the palladium nitrate in deionized water before the support was added. The metals were then precipitated together onto the support using NaOH. In the sequential precipitation method, the metal oxide additive was deposited first by precipitati on onto the support, aged, rinsed and calcined, and then the process was repeated to deposit the palladium. In the sequential precipitation, the amount of palladium is based upon the mass recovere d additive/support after calcination. 4.2.2 Reaction Conditions The reactant, 4-methylpyridine (Acros) was doubly distilled over KOH prior to use. The reaction was run in a 1:10 ratio of catalyst to reactant by weight in a r ound bottom flask. For all reactions 0.7 g of catalyst was used. The reaction mixture was evacuated followed by introduction of an oxygen atmosphere. The reacti on was heated to the boiling point (145C) under continuous agitation. After refluxing for 72 hour s, the flask contents were filtered using a glass micro-fiber filter and washed with chloroform to dissolve the product. The product was obtained by removing the chloroform, water, an d unreacted 4-methylpyridine with a rotary evaporator. 4.2.3 Carbon Monoxide Chemisorption The support surface areas, as received, and the prepared catalyst were performed by multipoint Brunauer-Emmett-Teller (BET) isothe rms on a Quantachrome Nova 1200 instrument as described in previous work [66]. Chem isorption measurements were performed in a 76

PAGE 77

Quantachrome ChemBET 3000 instrument and used to characterize active catalyst surface area. The catalysts were reduced with hydrogen at 170 C for 1 hour, and then outgassed in nitrogen at 170 C for another hour, followed by titration with CO to characterize the palladium dispersions of the catalysts. A detailed description of the procedure and calculations is giving in previous work [See chapter 3]. A very mild reduction was us ed to limit sintering of the Pd particles on the surface. This is very important since PdO is believed to be the active phase or a necessary precursor for this reaction. 4.3 Results and Discussion 4.3.1 Catalytic Activity All prepared catalysts were subjected to activ ity measurements in th e coupling reaction of 4methylpyridine. The results are summarized in Table 4-1. The Pd/n-Al2O3(+) catalyst from previous research has been added for comparison. Included is also a catalyst in which the nanoalumina support was calcined before deposition of palladium to determine the effects of the first calcination treatment in the case of the sequentia l precipitation method. As is evident from Table 4-1, calcining the n-Al2O3(+) support before deposition of pallad ium resulted in a catalyst with significantly reduced product yield. 4.3.2 Additives Adding ZnO to the supported palladium catalys ts under investigation did not increase the product yields by either deposition me thod. In the case of the Pd/n-Al2O3(+) addition of ZnO had little effect, only a slight, insignificant d ecrease was observed using the co-precipitation technique. In contrast ZnO appeared to reduce the catalytic activities of the Pd/n-TiO2 and Pd/nMgO catalysts. It has been reported that zinc readily forms an aluminate (ZnAl2O4) with the nAl2O3(+) support under conditions similar to the one s used in the current study [87]. Mixed metal oxides with ZnO may also be expected for the n-TiO2 and n-MgO, since ZnTiO4 [88] and 77

PAGE 78

MgxZn1-xO [89] have been observed. Naturally, compound formation between the support and the additive would reduce the favorab le effects of the added metal oxi de. In fact, if the additive only interacts with the support a nd there are no favorable interactions between the added metal oxide and the active metal (palladium), then it ma y be expected that the additive results in inferior activities. In this case, the additive can block some of the palladium on the surface and it will reduce any favorable palladium-support interac tions. Furthermore, it has been observed that ZnO combined with oxide supports can induce reduction of palladium [36]. These phenomena would explain the low activities of the ZnO-promoted catalysts. Addition of CeO2 and ZrO2 to the supported palladium cat alysts could in most cases induce a promoting effect. However, the effect of the additive is evidently very sensitive to the preparation method, i.e. whether th e Ceor Zr-precursor is added before or at the same time as the active metal. The only exception to the promoting effects of the additives was the CeO2 and n-Al2O3(+) combination for which a decreased ca talytic activity was observed with CeO2 addition irrespective of preparation method. There is no evident trend in the effects on the catalytic activity of CeO2 and ZrO2 addition as well as the pr eparation method between the supports. Sequentially precipitated CeO2 appeared to promote the n-MgOand n-TiO2-supported catalysts with yield increases of approximately 15% and 35 % respectively. However, the coprecipitation method yielded Pd/CeO2/n-MgO and Pd/CeO2/n-TiO2 catalysts that had lower activities compared to the cases without additives (see Table 4-1). The lack of promoting effect of CeO2 on the alumina catalysts may be due to CeO2-Al2O3 interactions as in the case of ZnO and ZnAl2O4 formation. This is supp orted by the fact that CeO2 has been reported to induce damage to some supports, particularly alumina [31]. 78

PAGE 79

While sequential addition of ZrO2 resulted in catalysts with lower activities than the catalysts without zirconia for the n-Al2O3(+) and n-MgO supports, this pr eparation method resulted in a more active Pd/n-TiO2 catalyst. Evidently, co-precipitation of ZrO2-and PdO-precursors resulted in less active Pd/nTiO2 and Pd/MgO catalysts. In contrast, ZrO2 proved to be an excellent promoting additive when co-precipitated with PdO on the n-Al2O3(+). The co-precipitated Pd/nZrO2 catalyst gave yields in excess of 3.3 g/g catalyst. This is ~30% more than the highest yield reported to date for this reaction sy stem. The promoting effects of ZrO2 and CeO2 strongly indicate that decreasing the extent of PdO re duction, and/or enhancing the supply of oxygen to the reaction surface increase(s) the catalytic activity. 4.3.3 Catalyst Preparation Method The results in Table 4-1 indicate that the catalytic activities of these catalysts are very dependent on the preparation method (sequentialversus co-precipitation). It appears that, in general, the co-precipitation method re sults in more active catalysts on n-Al2O3(+) supports, compared to the sequential precipitation method. In contrast, it appears that the sequential precipitation method results in mo re active catalysts on the n-TiO2 and n-MgO supports. As pointed out in Section 3.1, calcination of the n-Al2O3(+) support before palladium deposition results in a significantly less active catalyst comp ared to the catalyst prepared using the as received n-Al2O3(+). This would explain the lower activities of the sequentially precipitated nAl2O3(+)-supported catalysts. In fact, compari ng the yield of the sequentially precipitated Pd/ZrO2/n-Al2O3(+) catalyst, it is evident that the ZrO2 indeed has a promoting effect compared to the catalyst prepared using the precalcined n-Al2O3(+). For the n-TiO2 and n-MgO supports, the first calcination step does not appear to have the same effect as on the n-Al2O3(+). The lower activities of the co-p recipitated catalysts, compared to the sequentially precipitated catalysts, may in these cases be due to partial coverage of the surface 79

PAGE 80

palladium by the additives. An unusual eff ect was noted for the co-precipitated Pd/CeO2/n-TiO2 catalyst; significant palladium mirrors were observed on the reaction flask after a completed reaction. This is evidence of palladium being dissolved into the reaction solution from the support and indicates week pallad ium-support interactions. This is usually only observed in cases where the catalyst has very low activity. Ceria may interact with Pd ions strongly enough to undermine support interactions du ring co-precipitation, whereas CeO2 may provide good surface sites for Pd ions once it has been coated on to the surface. In general it appears that the sequential precipitation must be avoided on supports that ar e significantly damaged during calcination and this may be the main reas on for using the co-precipitation method. 4.3.4 Catalyst and Metal Surface Area Consistent with previous work, it is observe d that, in general high dispersions correspond to high activities but that the co rrelation is not always direct, and does not hold across different catalyst types (Table 4-2). Generally the more active catalyst had higher dispersion than the catalyst with lower activity for a given support but among the higher activit y catalyst there is no direct correlation of Pd surface area (SA) or crystallite size and activity. We have hypothesized that while the active Pd/PdO crysta llites are quite small, the active Pd species in the catalyst are only a fraction of the total surface Pd. The sequentially prepared n-Al2O3(+)-supported palladium catalysts consistently had lower Pd surface areas (%50-70) than the corresponding coprecipitated catalysts. This s upports the hypothesis that the su pport is damaged in the first calcination. 4.4 Conclusions Several imported observations can be made about the use of ZnO, ZrO2, and CeO2 as additives in oxide supported palla dium catalyst. When the support is more thermally stable (nTiO2, n-MgO) sequential deposition a ppears to work better. This may result from the additive 80

PAGE 81

partially covering active Pd/PdO phase during co-deposition. In the case of the n-Al2O3(+) coprecipitation worked better than the sequential me thod. This is attributed to damage to the support during calcinations, as a 5% Pd/n-Al2O3 demonstrates similar decreases in activity if the support is pre-calcined before palladium deposition. As an additive, ZnO generally decreased the activity of the catalysts. This is attributed to strong Zn-Pd interactions that either undermine Pd/PdO-support interactions, or disrupt the Pd/PdO phase due to alloying. CeO2 appeared to promote the reaction in some cases, but did not produce any catalyst significantly superior to the 2.5.16 g/g catalyst. The ZrO2 also promoted some of the catalysts (coprecipitated Pd/n-TiO2 and sequentially pr ecipitated Pd/n-Al2O3(+)). In the case of sequentially precipitated Pd/ZrO2/n-Al2O3(+), the highest observ ed activity per gram catalyst was observed (3.3.3 g/g catalyst). 81

PAGE 82

Table 4-1 Catalyst activity Entry Method Support Additive Yield [g/g catalyst] Yield [g/g Pd] 1 1 Precipitation n-Al2O3(+) 2.5 50 2 Precipitation (precalcined) n-Al2O3(+) 1.5 30 3 Co-precipitation n-Al2O3(+) CeO2 2.1 42 4 Sequential precipitation n-Al2O3(+) CeO2 1.6 30 5 Co-precipitation n-Al2O3(+) ZnO 2.4 47 6 Sequential precipitation n-Al2O3(+) ZnO 1.6 30 7 Co-precipitation n-Al2O3(+) ZrO2 3.5 69 8 Sequential precipitation n-Al2O3(+) ZrO2 2.3 45 9 1 Precipitation n-TiO2 1.6 31 10 Co-precipitation n-TiO2 CeO2 1.2 25 11 Sequential precipitation n-TiO2 CeO2 2.4 48 12 Co-precipitation n-TiO2 ZnO 0.9 17 13 Sequential precipitation n-TiO2 ZnO 1.2 24 14 Co-precipitation n-TiO2 ZrO2 0.8 16 15 Sequential precipitation n-TiO2 ZrO2 2.0 40 16 1 Precipitation n-MgO 2.3 46 17 Co-precipitation n-MgO CeO2 0.9 17 18 Sequential precipitation n-MgO CeO2 2.9 54 19 Co-precipitation n-MgO ZnO 0.6 12 20 Sequential precipitation n-MgO ZnO 1.0 20 21 Co-precipitation n-MgO ZrO2 1.9 38 22 Sequential precipitation n-MgO ZrO2 1.8 36 1 Results From previous work 82

PAGE 83

Table 4-2 Promoted catalyst dispersion Catalyst [5/5/90%] Deposition Metod L CO/g Pd Ce/Pd/n-Al2O3(+) sequential precipitate 3690 Ce/Pd/n-Al2O3(+) co-precipitate 5060 Zn/Pd/n-Al2O3(+) sequential precipitate 2800 Zn/Pd/n-Al2O3(+) co-precipitate 4490 Zr/Pd/n-Al2O3(+) sequential precipitate 2230 Zr/Pd/n-Al2O3(+) co-precipitate 4490 Ce/Pd/n-TiO2 sequential precipitate 5430 Zn/Pd/n-TiO2 sequential precipitate 680 Zn/Pd/n-TiO2 co-precipitate 2220 83

PAGE 84

CHAPTER 5 CHARACTERIZATION OF PALLADIUM CATALYSTS SUPPORTED ON NANOPARTICLE METAL OXIDES FOR THE OXIDATIVE COUPLING OF 4METHYPYRIDINE 1 5.1 Introduction Palladium is one of the more versatile active metals since it can serve as a catalyst in a large number of reactions [9,90-92]. It is a ve ry efficient hydrogenation catalyst that is commonly used in the reduction of many organic species. Palladium is also an excellent oxidation catalyst that has been used in the complete oxidatio n of CO [92,93], CH4 [9-11] and volatile organic compounds (VOC)[95-97]. Sel ective oxidations can also be performed, e.g. conversion of hexane to benzen e [18], as well as oxidation of alcohols to aldehydes or ketones [98]. Other applications include steam reforming of methanol and ethanol [62] and conversion of acetylene to ethylene [61]. One important reaction that has received relatively limited attention is the C-H activation and C-C coupling of aromatic syst ems over palladium catalysts. Of these, the palladium-catalyzed oxidative coupl ing of 4-methyl pyridin e to 4,4-dimethyl-2,2bipyridine is of particular interest. The bipyr idine product has the abili ty to coordinate to transition metal cations and form complexes with interesting properties [2]. For example, some bipyridine complexes have unique photoand electrochemical properties [20,21,23,24,43,44], while other organometallic bipyridine complexes are very efficient catalyst systems [25-29]. The 4,4-dimethyl-2,2-bipyridine is a particularly important bipyridine, since it can be easily modified in the 4 position [23,30]. For example, the oxidation of the methyl groups is facile resulting in 2,2-bipyridine-4,4-d icarboxylic acid [30], which can then be reacted further and used to tether the bipyridine, and then later a coordinated transition me tal, to a support [99]. 1 Luke M Neal, Samuel D. Jones, Michael Ev erett, Gar B. Hoflund, Helena Hagelin-Weaver, Manuscript in preparation, unpublished copyright (2008). 84

PAGE 85

Moreover, the oxidative coupli ng of 4-methylpyridine to 4,4 -dimethyl-2,2-bipyridine over palladium catalysts is a simple, environmenta lly friendly, one-step process (Scheme 1), where water and the terpyridine are the only byproducts. This is thus an important alternative to routes utilizing expensive halogenated precursors to form the bipyridine. However, the coupling reaction is slow with relatively low yields [5] so significant improvements would be desirable. According to the literature, the most commonl y used palladium catalyst for the coupling of 4-methylpyridine is palladium on carbon (Pd/ C), while commercial palladium on alumina (Pd/Al2O3) catalysts had been shown to be practically inactive [1-4]. In contrast, recent research studies have demonstrated that 5% palladium precipitated onto alumina nano particles (NanoScale NanoActive alumina plus) or porous titania (Alfa Aesar TiO2 catalyst support) results in the highest yields reported to date for this palladium-cat alyzed reaction, 2.5 g of product / g of palladium or 50 g / g Pd [66, see Chapter 2]. Previous studies have also shown that catalysts prepared via the wet impregnation method, or via precipita tion of palladium onto a nanoparticle alumina support with larger particle sizes, were not act ive. In contrast, palladium precipitated onto a traditional -alumina support was shown to exhibit some, albeit limited, activity [66, see Chapter 2]. Other supports, such as n-MgO, n-ZrO2, n-ZrO2 doped with 10% CeO2, and n-ZnO, can also be used to prepare activ e catalysts [see Chapter 3]. This represents a wide range of supports with vary ing properties. To this point, no systematic correlation between Pd dispersion and catalytic activity has been found. Low dispersions most often explain a low yield, but a high dispersion is not necessarily a requisite for a high activity. The most active catalysts were prepared using high surface area supports (>400 m2/g) or were prepared using oxides which have been shown to exhibit strong metal-support interactions with palladium. 85

PAGE 86

Additionally most of the active cat alysts had a high number of aci dic sites, but not necessarily any strong basic sites, i.e. little or no CO2 uptake. Since no distinct correlation between catalyst properties and catalytic activity has been found, a more thorough catalyst charact erization study is therefore need ed to determine so called structure-activity relationships, wh ich in turn is necessary to be able to further improve these catalysts. Our hypothesis is that palladium supported on a nanopartic le oxide support has potential to be a very efficient catalyst fo r the reaction due to the high number of low coordination sites, such as corners and edges of this support [6]. These low coordination sites could result in strong metal suppo rt interactions and result in high dispersions and unique catalytic properties. We have found that high surface area supports, both nanoparticle oxides or traditional porous supports, can indeed result in high dispersions. Additionally, supports with strong palladium-support interactions can also give high dispersions and high catalytic activities. To further probe these catalysts they have been subjected to the following catalyst characterization techniques: I) X-ray photoele ctron spectroscopy (XPS) to probe the surface composition and chemical states of the Pd phase in the fresh and spent catalyst, II) transmission electron microscopy (TEM) with energy dispersi ve spectroscopy (EDS) to image the structure and dispersion of prepared catalysts, as well as III) X-ray diffraction (X RD) to determine the particle sizes of the catalyst supports (and to determine if the palladium on the surface can be detected with XRD). The objectives of this work were to character ize the Pd/PdO crystallites in fresh, reduced, and spent catalys t for 1) crystallite size/dis persion, 2) oxidation state and electronic binding energy, and 3) compare binding energies of fresh catalysts with varying activities. 86

PAGE 87

5.2 Experimental 5.2.1 Catalyst Preparation and Reaction The commercial catalysts, 5% Pd/C (dry, reduced) and 5% Pd/Al2O3 were obtained from Alfa Aesar and used (as received or dried at 105C for 1hr) in the reactions and catalyst characterizations. A number of 5% Pd catalysts precipitated onto nanopart icle oxide supports or porous supports, as described in previous work [66], were also carefully characterized. The reactions were carried out by placing fr esh calcined catalyst and doubly distilled 4methylpyridine under reflux for 72 hours, as describe d in previous work [66]. After a complete reaction the catalyst was recovered using a glass micro-fiber filter and washed with chloroform to dissolve the product. The recovered spent cata lyst was stirred in additional chloroform and filtered, followed by brief drying at room temp erature before it was put in the XPS system. Reduced catalysts were prepared by reducti on in 5% hydrogen in nitrogen for 1 hour at 170C and, outgassed in nitrogen for at 170 C an additional hour and then cooled to room temperature with continued nitrogen flow. The sample was kept sealed in nitrogen until immediately before XPS sample preparation 5.2.2 Chemisorption Fresh catalysts were reduced in 5% H2 in nitrogen at 170C for one hour and then outgassed in nitrogen at 170C. Spent catalysts were outgassed in nitrogen at 170C for one hour. Selected spent catalysts were also subj ected to the same reduction applied to fresh catalysts. These catalysts were then subjected to CO adsorption measurements to determine the Pd surface area. Estimates of the Pd particle sizes were made from these CO adsorption measurements. The details of the calculations for Pd dispersion, surface area and crystallite size are given in previous work [see Chapter 3]. 87

PAGE 88

5.2.3 X-ray Photoelectron Spectroscopy The fresh, spent and reduced catalyst powders were pressed into aluminum cups prior to insertion into the ultra-high vacuum (UHV) chamber (base pressure 1-10 Torr). XPS was performed using a double pass cylindrical mirror analyzer (PHI model 25-270 AR). Spectra were taking in retarding mode with a pass energy of 50 eV for surveys and 25 eV for high resolution spectra using a Mg K X-ray source (PHI 04-151). Data was collected and then digitally smoothed using a computer interface. A value for the C1s binding energy of 284.6 eV was assigned to correct for static charging [100]. 5.2.4 X-ray Diffraction The XRD data was gathered on a Philips powder X-ray diffractometer using BraggBrentano geometry with Cu-K radiation ( = 1.54 ). Diffraction pa tterns were obtained for selected calcined, reduced and sp ent catalysts. The catalyst powders were secured onto a glass side with double-sided sticky tape. Average pa rticle sizes were calculated from the linebroadening of the XRD peaks us ing the Scherrer equation ))cos( ( FWHM K d In this equation K is a constant generally taken as unity, is the wavelength of the incident radiation, FWHM is the full width at half max and is the peak position. 5.2.5 Transmission Electron Microscopy TEM grids we prepared by dispersing the cata lyst into water by ultasonication and then placing a drop of the dispersion on to lacy carbon grids. Micr ographs and EDS spectra were taking on a JEOL TEM 2010F, w ith a 200 kV electron source. 88

PAGE 89

5.3 Results and Discussions In the discussion of the characterization data obt ained it is important to have an idea of a likely reaction mechanism. Therefore, a suggest ed reaction mechanism is presented here. 5.3.1 Proposed Reaction Mechanism A reasonable description of the catalytic cy cle for the palladiumcatalyzed coupling of pyridine is presented in Figure (5-1 ). The first step is most li kely coordination of the pyridine nitrogen to a palladium atom on the surface. Cons idering that there is a significant difference in yield between pyridine and 4-methyl pyridine in these reactions [5], this coordination appears to be very important for the overall reaction. Th e C-H activation is probably facilitated by an oxygen atom in close proximity to the palladium on the surface. This step is likely the ratedetermining step in these reactions. The resu lt of the C-H activation is a hydroxyl group and a pyridine bonded to palladium via a carbon rather than the nitrogen. After two consecutive C-H activations and insertions, a reductive eliminatio n gives a coordinated bi pyridine [not shown in Figure (5-1)] which later desorbs to give the produc t. Also, water is formed from the two surface hydroxyl groups and desorbs leaving an oxygen vacan cy on the surface. To close the catalytic cycle the surface oxygen must be regenerated. This step is likely not trivial in the reaction system, since the oxygen is in gas phase above th e dispersion of catalyst and liquid reactant. 5.3.2 X-ray Photoelectron Spectroscopy Even though XPS is considered a surface techni que, it must be interpreted carefully for catalysts. This is due to the f act that the photoelectr on attenuation length in a solid is ~ 20-40 [101], dependent on the kinetic energy of the el ectron (and thus also its binding energy). Consequently, the XPS probing depth is often greate r than 10 atomic layers down into the solid. As catalysis takes place only on the top layer of a catalyst, most of the XPS signal originates from non-active near surface species. Furthermore, the near surface region of a catalyst is very 89

PAGE 90

inhomogeneous both laterally and vertically. Conse quently, it is challengi ng to obtain qualitative data of the surface composition of a catalyst. Ho wever, knowing the limitations of the technique very useful information can be obtained from X PS data collected from different catalysts. The Pd 3d XPS spectra obtained from the palladium catalysts supported on the various alumina supports are presented in Figure (5-2). The commercial Pd/C and Pd/Al2O3 catalysts have been included for comparison. The major sp ecies on most catalysts, including Pd/C, is located at a binding energy of 336.1 eV, which is reasonably consiste nt with reference values for PdO (336.3 eV [36]). In fact, values between 335.6 and 337.1 eV have been reported for the Pd 3d5/2 peak of PdO [102]. In contra st, the major species on the Pd/Al2O3 catalyst is located at 334.5 eV. This binding energy is lower than that reported for bulk Pd metal (334.9 eV [100 ] ), but is in agreement with values observed for surface Pd0 typically seen in high-dispersion catalysts [78]. This observation supports the notion that PdO is the active phase on these catalysts. It also explains why the commercial Pd/Al2O3 catalyst exhibits little if any activity in these reactions, while the Pd/C is an active catalyst. As the commercial Pd/Al2O3 catalyst, the Pd 3d5/2 peak obtained from the spent Pd/n-Al2O3(+) catalyst is located at a low binding energy (334.9 eV), which is indicative of Pd metal as has been observed previously for spent catalysts [5]. As pointed out previously, this is likely the main reason for catalyst deactivation in these systems. While the major species is PdO on the prepared fresh catalysts, the Pd 3d peaks obtained from the catalysts supported on the nanoparticle Al2O3(+) are considerably broader than those obtained from the n-Al2O3(-)and -Al2O3-supported catalysts. The Pd/n-Al2O3(+) catalyst prepared via the precipitation method not only has a distinct shoulder at 337.8 eV, it also exhibits a small shoulder at 334.5 eV. This suggests that PdO2 and Pd0, as well as PdO are present on the surface of this catalyst. While the Pd 3d peaks obtained from the Pd/n-Al2O3(+) catalyst 90

PAGE 91

prepared via impregnation also are broad co mpared to those obtained from the Pd/Al2O3(-) and Pd/ -Al2O3 catalysts, the contri butions from the PdO2 and Pd0 states are considerably smaller on this catalyst compared to the precipitated one. This may explain the observed differences in activity between the impregnated and the precipitated Pd/n-Al2O3(+) catalysts (A summary of the product yields obtained from the catalysts are presented in Chap ter 3 Table 3-2). However, it does not easily explain the differences in ca talytic activity between the impregnated Pd/nAl2O3(+) and the precipitated Pd/ -Al2O3 catalysts. Evidently, these systems are very complex and it may be that only certain palladium speci es, perhaps palladium particles below a certain size are active in these catalysts. The main difference between the Pd 3d peaks obtained from the Pd/Al2O3(-) and Pd/ -Al2O3 catalysts appears to be a smal l shoulder at low binding energy on the Pd/ -Al2O3 catalyst. There may also be an indication of a PdO2 state on the Pd/ -Al2O3 catalyst, but this species is clos e to the noise level. Since it is known that the dispersions are very low on the Pd/Al2O3(-) and Pd/ -Al2O3 catalysts, it may be di fficult to discern the active species on these catalysts. This may be pa rticularly true in the case of the Pd/Al2O3(-), which apparently has a very low numb er of active sites [66]. The O 1s binding energy regi on is dominated by the Al2O3 oxygen at 531.0 eV on these catalysts (Figure 5-3). The O 1s peak obtained from PdO and the Pd 3p3/2 peaks cannot be detected on the alumina-containing catalysts, which means that no differences in the O 1s peaks between the catalysts can be observed. Howeve r, even on the Pd/C cat alyst, where the support has a limited number of oxygens, this region is co mplex. This is due to the close proximity of the Pd 3p3/2 and the O 1s peaks. For example, the Pd 3p3/2 peak of PdO is located at 533.8 eV, while that of Pd metal is locat ed at 532.3 eV and the O 1s peak of PdO is at 530.0 eV [103]. Therefore, on the Pd/C catalyst, which consist of a mixture of Pd metal and PdO, the O 1s from 91

PAGE 92

PdO and the Pd 3p3/2 peaks from PdO and Pd result in a broa d peak that cannot be resolved into the different constituents. As for the alumina-supported catalysts, the titania-supported catalyst with the broader Pd 3d peaks is the most active (Figure 5-4). In this case it is the catalyst supported on porous TiO2 rather than the Pd/n-TiO2, which is the more active catalyst. It is a little surprising that the main difference between the Pd 3d peaks on the Pd/p-TiO2 and Pd/n-TiO2 catalysts is a more pronounced shoulder at lower binding energies for the Pd/p-TiO2 catalyst. The spent Pd/n-TiO2 catalyst has a very low binding energy, belo w the 334.5 eV reported for dispersed surface Pd0. One catalyst was therefore subjected to a reduc tive treatment, the same reduction treatment as before the CO titration experiments. The Pd 3d5/2 binding energy of the reduced catalyst, 335.1 eV, is in agreement with the values reported for Pd metal. Consequently, the low binding energy of the spent Pd/p-TiO2 may be attributed to differential charging caused by the presence of a species covering the Pd surface. This is supported by the fact that sputtering of the spent catalysts, i.e. removal of this surface species, results in a Pd 3d5/2 binding energy of 334.9 eV, i.e. very close to the binding energy reported for the reduced catalyst. These results further indicate the surface of the spent catalysts do contai n decomposed products (Pd fouling) or reactants/products that strongly in teract with the palladium (Pd poi soning), which could lead to catalyst deactivation as has been suggested previously [5]. The Pd 3d and O 1s peaks obtained from the fresh catalysts supported on n-ZrO2, nZrO2+10%CeO2, and n-CeO2 are presented in Figures 5-5 and 5-6. The binding energies of the Pd 3d5/2 peaks are located at 336.6 eV for the Pd/n-ZrO2 and Pd/n-ZrO2+10%CeO2, and at 337.4 eV for the Pd/n-CeO2 catalyst. These Pd 3d5/2 binding energies are higher than those observed on the Pd/Al2O3 and Pd/TiO2 catalysts (336.1-336.3 eV, Table 51). The high binding energies 92

PAGE 93

indicate the presence of electr on deficient Pd(II) species on the su rface, which is likely due to strong palladium-support interactions. As pointed out in Section 5.3.1, the coordination strength of 4-methylpyridine to a surface palladium atom likely has a significant effect on the reaction. A more electrophilic palladium, i.e. electron defici ent, will have a higher af finity to the electron lone pair of the pyridine nitrogen which leads to a stronger coordination. After coordination, the higher charge on the pyridine nitrogen could facil itate abstraction of the proton to form the Cbonded pyridine on the surface. If this is the ca se, then increasing the electrophilicity of the palladium(II) on the surface would be expected to increase activity. Although the Pd/CeO2 catalyst with a Pd 3d5/2 binding energy of 337.4 eV, is not more active than the Pd/n-ZrO2 and Pd/n-ZrO2+10%CeO2 catalysts with binding energies of 336. 6 eV, it appears that catalysts with high Pd 3d binding energies and broad peaks in ge neral are more active than catalysts containing less electrophilic palladium(II) and narrower Pd 3d peaks. In the case of the reduced Pd/CeO2 catalyst, the Pd 3d5/2 binding energy is shifted to lower values (Figure 5-5). However, the binding ener gy is slightly higher (335.3 eV) than that observed for the reduced Pd/TiO2 catalyst (335.1 eV). In addition, there is a larg e shoulder at 336.7 eV, indicating the presence of PdO on this catal yst. While this may in part be due to the brief air exposure after the reduction, it does indicate that either the PdO is difficult to completely reduce, since CeO2 can assist with oxygen, or any Pd metal on the surface is easily re-oxidized to PdO. Either one of these would be very beneficial to the reaction system under investigation. The same ma y be true for the Pd/n-ZrO2 and Pd/n-ZrO2+10%CeO2 catalysts, but in this case the Zr 3p peaks overlap with the Pd 3d peaks and obstruct the interpretation of the spectra. 93

PAGE 94

The O 1s binding energy region is natura lly dominated by the oxygen atoms from the support (Figure 5-6). In our case th e O 1s peaks obtained from the CeO2 and ZrO2 nanoparticles are located at 529.1 and 529.7 eV. The value for CeO2 is in agreement with the literature, while the binding energy for the ZrO2 is lower than the commonly reported 530.5 eV [100]. However, a value of 529.9 eV has been reported for ZrO2 [102], which means that using the C 1s peak as the charge reference is reasonable. Furthermore, considering that the O 1s peaks line up for the ZrO2 nanoparticle, Pd/ZrO2 and the reduced Pd/ZrO2 it does not appear likely that the shift is due to differential charging. In general the O 1s peaks obtained from the catalysts are slightly broader than those obtained from the pure s upports, due to the presence of the Pd 3p3/2 peaks (and perhaps the O 1s peak from PdO). It is interesting to note the O 1s peak obtained from the Pd/CeO2 catalyst is significantly broader than that obtained from the CeO2 support. It appears that there is a substantial cont ribution from oxygen-containing sp ecies with binding energies of 534 eV and higher on this catalyst. This is proba bly from the electron deficient PdO species and supports the high binding en ergy observed for the Pd 3d5/2 peak on this catalyst. 5.3.2 Transmission Electron Microscopy The micrographs of the alumina catalysts s how a variety of support structures (Figure 57). In the n-Al2O3(+) catalyst the support a ppears to be composed of agglomerated flakes (Figure 5-7 a). The bimodal -alumina ( -Al2O3) contains fine rod/needle structures (Figure 5-7 B). The n-Al2O3(-) has larger and more round struct ure compared to the flakier n-Al2O3(+). Also, the palladium particles appear much larger on the Pd/n-Al2O3(-) compared to the Pd/ nAl2O3(+). The presence of a fine st ructure in the moderately active -Al2O3 would be consistent with the presence of some low-coordination ed ge sites, whilst highl y active, flaky nano-Al2O3(+) would be expected to have a highe r concentration of corner and edge sites. It is also expected 94

PAGE 95

that the largely inactive nano-Al2O3(-) with more spherical particles would have very few lowcoordination sites. These results are consiste nt with our hypothesis that the number of low coordination sites on a support correlates with the catalytic activity of the corresponding catalyst. Comparing the precipitated Pd/n-Al2O3(+) and a commercial Pd/C reference, small particles (>2nm) of PdO (identified by EDS) can be observed. It appears that the PdO particles are smaller on the Pd/n-Al2O3(+), which is consistent with the CO adsorption data. It is difficult to find Pd on the -Al2O3 catalyst and in cases where they are visible they appear larger than on the Pd/n-Al2O3(+) and Pd/C catalysts. This correlates well with the observed low dispersion on the Pd/ -Al2O3. 5.3.3 X-ray Diffraction Caution is necessary in interpreting XRD data obtained from catalysts. First of all, it is considered a bulk technique and ha s attenuation depths on the orde r of micrometers. XRD thus probes several orders of magnitude deeper than th e thickness of the catalyst surface layers. This is a serious limitation, since th e reaction over a heterogeneous catalyst occurs at the surface. Furthermore, there must be long-range order in a crystal for a phase to be detected with XRD. XRD is, consequently, insensitiv e to amorphous phases and the dete ction limit for small particles is around one nm. For catalysts where a few percen t of an active phase is present at the surface of a support, the active phase can be, and often is, invisible with XRD. Despite these limitations, useful information can be obtained from XRD. Th is is particularly the case when the catalytic activity depends on the structure of the support. Even though the surface structure is likely different from the bulk, the bulk structure can infl uence surface structure and thus lead catalytic 95

PAGE 96

activities that are depend ent on the support structure 2 XRD can also be used to probe trends between catalysts and can provide useful inform ation when the active surface particles can be resolved. As the palladium oxide is visible on some of the catalyst suppo rts used in this study, XRD was performed on a few selected catalysts. Despite the high dispersions and the low palladium loadings, the presence of PdO in the fresh 5% Pd/n-Al2O3(+) is evident in the XRD spectra obtai ned from this catalyst (Figure 5-8). The XRD spectrum obtained from the n-Al2O3(+) support is include d in the figure for comparison. It was known from the supplier that the n-Al2O3(+) is a poorly crys talline (actually stated to be amorphous) [67]. However, there is evidently some structure present in the nAl2O3(+) sample, and it is not -Al2O3 3 which is often the phase pr esent in high surface area alumina supports. The phase present on the n-Al2O3(+) support is iden tified as bohemite (aluminum oxide hydroxide) 4 This phase is observed in cat alysts calcined under 350C whilst only -Al2O3 is observed in samples calcined at 450C. The low palladium oxide signal and overlapping features with -Al2O3 or bohemite makes it difficult to determine the peak widths of the palladium oxide. Consequently, calculating the particle sizes using th e Scherrer equation is challenging at best for these catalysts. Howeve r, on the precipitated sample calcined at 450C a PdO particle size of ~4.0 nm was calculated using the PdO 5 (122) peak. The low signal to noise of this peak limits the accuracy of the peak fitting, but consider ing the XRD limits in catalysis this is at least a number that can be us ed for comparison with other catalysts. 2 A classic example is the influe nce of monoclinic versus tetragona l ZrO2 in the methanol steam reforming or methanol synthesis over Cu-based catalysts [104-105]. 3 JCPDS # 10-0425 4 JCPDS # 21-1307 5 JCPDS # 43-1024 96

PAGE 97

The same difficulties in determining PdO partic le sizes using the Scherrer equation arise for the fresh Pd/n-CeO2 catalyst on which the main PdO peak s (101) and (002) is obscured by the (200) peak of the CeO2 6 support (Figure 5-10) and no other p eak due to PdO can be resolved. Of the catalysts analyzed using XRD, the p-TiO2-supported catalysts were the only ones where determining the Pd and PdO particle sizes using the Scherrer equation was trivial (Figure 5-11). On fresh and spent Pd/p-TiO2 catalysts, the estimated PdO and Pd particle sizes were ~4.0 and 4.6 nm respectively. As expected, th is is a little higher than the particle sizes calculated based on the measured Pd surface areas, i.e. 2.4 nm for reduced Pd/p-TiO2 catalysts. This is expected since XRD measurements normally give larger partic le sizes compared to ot her techniques. It is due to the fact that amorphous phases cannot be de tected with XRD and that the detection limit of XRD is particle sizes on the or der of one nm. Of particular in terest is that no PdO is observed in the spent catalyst. While the presence of amorphous PdO cannot be excluded, this indicates that the reaction over thes e catalysts reduces not only the surface PdO, but also the bulk PdO. This is a very important observation since complete reduction of PdO to Pd metal on the catalyst surface likely results in a species that is much more difficult to reoxidize than a PdO particle with a Pd surface layer. Again, this explains why supports with mobile oxygens, such as CeO2 and ZrO2, can result in very active catalysts. 5.4 Conclusions The correlation between the fine structures ob served in TEM and the catalytic activity is consistent with the hypothesis that low coordination sites on the s upport results in high catalytic activities. The particle sizes observed in TEM/E DS correlate well with the measured dispersions of the catalyst. XRD gives slightly larger particle sizes but is still reasonably consistent with the 6 JCPDS# 43-1002 97

PAGE 98

chemisorption measurements. This indicates th at the CO chemisorption data gives reasonable estimates of dispersion and particle size. XPS and XRD suggest that the PdO on the fres h catalyst is completely (both surface and the crystalline bulk) reduced to Pd metal at the end of the reaction. This is consistent with deactivation of the catalyst by complete reduc tion of the active PdO phase. Additionally, differential charging in the XPS spect ra obtained from the spent p-TiO2 catalyst indicates the presence of a contaminant/surface species on the palladium surface. The differential charging is diminished by sputtering, which indicates removal of the surface species. This in turn is consistent with deactivation by Pd/PdO su rface fouling or poisoning by an unidentified byproduct, the desired product itself, or an unknown contaminant. In XPS the most active catalysts exhibited broadening of the Pd 3d peaks (Pd/n-Al2O3(+) and Pd/p-TiO2) and/or binding energies higher than expected for PdO (Pd 3d3/2 greater than 336.3 eV) (Pd/n-ZrO2 and Pd/n-CeO2). This is consistent with more electrophilic Pd(II) being responsible for activity. 98

PAGE 99

Table 5-1 Palladium 3d binding energy fo r fresh, reduced and spent catalyst Catalyst Pd 3d5/2 O 1s Fresh Palladium 5% n-Al2O3(+) 450C 336.3 531.0 n-Al2O3(+) 450C (impregnated) 336.3 531.0 n-TiO2 336.4 529.9 n-MgO 336.2 531.0 n-Al2O3(-) 336.1 531.0 -Al2O3 336.2 531.0 p-SiO2 336.1 532.1 p-TiO2 336.3 529.8 n-ZnO 336.1 529.8 n-CeO2 337.4 529.1 n-ZrO2 336.6 529.7 n-ZrO2+10%CeO2 336.6 529.4 Reduced Palladium p-TiO2 335.1 529.3 n-ZnO 334.5 529.9 n-ZrO2 334.6 529.7 Alfa Aesar 334.5 531.0 Spent Palladium n-Al2O3(+) 350C 334.7 531.1 p-TiO2 334.1 529.2 p-TiO2 (sputtered) 334.9 529.4 n-ZrO2+10%CeO2 334.6 529.3 99

PAGE 100

N N H Pd O Pd O N N Pd OH Pd OH N N Pd O Pd OH2 Figure 5-1 Proposed oxidative coupling reaction mechanism 100

PAGE 101

350345340335330325 3d3/23d5/2Pr -Al2O3Pr n-Al2O3(-) Pd/Al2O3 Com. Pr n-Al2O3(+) Sp. Pd/C Pr n-Al2O3(+) Im n-Al2O3(+)Pd 3dN(E) (arbitrary units)Binding Energy (eV) Figure 5-2 Palladium 3d spectra of palladium/nano-alumina catalysts 101

PAGE 102

545540535530525520 Pr. -Al2O3Pr n-Al2O3(-) Pd/C Al2O3Im. n-Al2O3(+) Pr n-Al2O3(+) Pr. n-Al2O3(+) Sp. Pd/Al2O3 Com. O 1sN(E) (arbitrary units)Binding Energy (eV) Figure 5-3 Oxygen 1s spectra of pa lladium/nano-alumina catalysts 102

PAGE 103

350345340335330 3d5/23d3/2Pure PdO Spent, Sputtered Reduced Nano Fresh Fresh Spent Pd 3dN(E) (arbitrary units)Binding Energy (eV) Figure 5-4 Palladium 3d spectra of palladium/titania catalysts 103

PAGE 104

350345340335330 Zr 3p1/2Zr 3p3/2Pd 3d3/2Red. CeO2CeO2Red. ZrO2Spent ZrO2 (CeO2) Fresh ZrO2Fresh ZrO2 (CeO2)Pd 3d5/2N(E) (arbitrary units)Binding Energy (eV) Figure 5-5 Palladium 3d spectra of palladi um on nanozirconia and ceria catalysts 104

PAGE 105

540535530525520 As Received As Received n-CeO2n-ZrO2Fresh Fresh Fresh Reduced Spent Pd/n-ZrO2 (CeO2) Pd/n-ZrO2Pd/n-ZrO2 (CeO2) Pd/n-ZrO2Pd/n-CeO2 O 1sN(E) (arbitrary units)Binding Energy (eV) Figure 5-6 Oxygen 1s spectra of palladiu m on nanozirconia and ceria catalysts 105

PAGE 106

A B C D Figure 5-7 TEM of select catalysts A) Pd/n-Al2O3(+), B) Pd/ -Al2O3, C) Pd/n-Al2O3(-) and D) Pd/C. 106

PAGE 107

30 40 50 60 70 80 (112) (110) PdO (002),(101) -Al2O3 BohemitePrec. 450 Imp. 350 Prec. 350 2Intensity [arbitrary units] Figure 5-8 Powder XRD of n-Al2O3(+) catalysts 7 7 PdO(JCPDS # 43-1024), bohemite (JCPDS # 21-1307), -Al2O3 (JCPDS # 10-0425) 107

PAGE 108

3032343638404244 CeO2 (200) PdO (002),(100) PdO (200)Intensity [arbitrary units]2 Figure 5-9 Powder XRD of palladium on ceria catalyst 108

PAGE 109

3032343638404244 TiO2Pd Metal PdO Fresh p-TiO2Spent p-TiO2Intensity [arbitrary units]2 Figure 5-10 Powder XRD of spent a nd fresh palladium on porous titania 109

PAGE 110

CHAPTER 6 OPTIMAZATION OF METAL-OXIDE-SUPPOR TED PALLADIUM CATALYST FOR THE OXIDATIVE COUPLING OF 4-METHYLPYRIDINE 1 6.1 Introduction Bipyridines are important compounds since they can coordinate to transition metal ions and form complexes with interesting properties [20-22,41,42]. Some of these complexes have been studied for their photochemical propertie s, such as Ru-based bipyridine complexes [20,23,42-44], while other transition metal complexe s have been used in various catalyst systems, e.g. Pd, Rh and Mo based systems [2538,99]. However, bipyridines, such as the commonly used 4,4 dimethyl-2,2-bipyridine, are prohibitively expensive for large-scale commercial use. They cost in excess of $5,200/ kg [31], with smaller quantities selling for significantly higher prices per uni t weight ($13-20/g depending on quantity [5-50 g batches] and purity [31,47,106]. Consequently, more efficien t routes to bipyridi nes using relatively inexpensive starting materials are desired. The simplest pathway to 4,4-dimethyl-2,2-b ipyridine is the oxida tive coupling of 4methylpyridine using palladium catalysts [5] or Raney nickel [1-4]. Over both catalysts this is a one-step process in which the bipyridine is formed directly from the pyridin e reactant without the use of solvents or costly halogenated precursor s (Scheme 6-1). However, the reactions are slow and in both cases the catalyst deactivates, yiel ding only a few grams of product per gram of catalyst at best [1-5]. While Ra ney nickel is less expensive compar ed to palladium catalysts, it is more hazardous to handle as it is pyrophoric when dry and releases toxic vapors when ignited [107]. Additionally the potential for recycling after reaction is low for the Raney nickel. 1 Reproduced from Manuscript in preparation Luke M Neal, Helena Weaver, Unpublished copyright (2008) 110

PAGE 111

Therefore, Pd/C has been the catalyst of choi ce in the literature. Compared to the commonly used Pd/C our previous research has produced catalysts with higher and more consistent activities [66, see chapter 2]. Th e best catalysts to date for unpromoted catalysts are a 5% palladium precipitated onto alumina nanoparticles [n-Al2O3(+)] and a 5% palladium precipitated onto porous titania [p-TiO2], which both gave yields in ex cess of 2.5 g of product per g of catalyst or 50 g/g Pd [66, see chapters 2 and 3], with yields in excess of 3.3 g/g catalyst (66 g/g Pd) being observed when 5% ZrO2 (metal basis) is added to 5% Pd/n-alumina by co-precipitation [see Chapter 4]. This is to be compared with the maximum isolated yiel d of ~2 g/g of catalyst (40 g/g pd) reported for a 5% Pd/C, [37]. Wh ile the prepared catalys ts supported on p-TiO2 and n-Al2O3 are superior to the traditional Pd/C catalyst, it is our belief that further optimization is possible. Since the major cost of palladium catalysts is the high price of the palladium, the most cost-effective catalyst is likely the one giving the highest yield per gram of palladium. The price of the support can also influence the cost-effectiven ess of the palladium catal ysts, particularly in the case of nanoparticle oxide ca talysts in which the support can cost up to $700 per kg [68]. Consequently, to optimize the cost-efficiency of these catalysts it is important to optimize the yield per unit weight of palladium and to investigate supports other than the n-Al2O3(+), which has proven to be the best support to date. Fo r this reason supports, such as n-MgO, n-TiO2, and p-TiO2, were included in the optimization study. In addition to the selection of active metal a nd support, there are a number of factors that can influence the catalytic activities of heteroge neous catalysts. These include, for example, 1) catalyst composition, i.e. loading of active metal on the support or addition of promoters, 2) catalyst preparation method, such as technique and order of depositi ng the active metal and 111

PAGE 112

additive on the support, as well as calcination temperatures or other pretreatments, and 3) reaction conditions, which includes reactant to catalyst ratios and reaction time. Previous results have demonstrated that 10% and 5% loadings on n-Al2O3(+) give approximately the same yield per gram of catalys t despite similar dispersions and significantly higher Pd surface areas per gram of catalyst [66]. In fact, it has been shown that the Pd surface areas of the various catalysts tested are not strongly correlated to catalytic activit y. Thus, it is expected that lower loadings of Pd have potenti al to produce similar yield per g catalyst, which corresponds to significantly higher yields per g Pd and a more cost-effective catalyst. Previous research has revealed that the precipitation method is superior to the impregnation method. Of these the precipitati on method involves more steps to be optimized. For example, in the early work a simple 50% stoichiometric exce ss of NaOH base (based on the amount of metal nitrate in solution) was added to precipi tate Pd(II) out of solution as Pd(OH)2 onto the support. It is possible that the resulting pH of the mixture after ti tration can affect the catalytic properties. This is particularly true for hi ghly acidic or basic supports. Th e base used for precipitation may also influence the catalytic activity. In addition to the titration, the current procedure for catalyst precipitation requires multiple days to age and rinse the catalyst. The goal of the current study is to determine the effects of final precipitation pH as well as the agi ng and rinsing time on the catalytic properties. In the earlier studies a reactan t catalyst ratio of 10:1 (by weight) was used [66]. The bipyridine product is predicted to poison the catalyst, and, consequently, it is expected that a higher ratio of reactant to catalyst would increas e the reaction yield per gram of palladium by diluting the concentration of inhi biting products. The disadvantage with this is that the actual conversion of reactant to product is likely to be lower at the higher reacta nt-to-catalyst ratio. 112

PAGE 113

This means that in a commercial process the r eactor cost would prohibit very high reactant to catalyst ratios due to the increase in reactor size. However, since it is pos sible that a higher ratio could significantly increase the yield, it is importa nt to optimize this ratio Another variable to optimize is the reaction time. The current pro cedure calls for a 72-hr reaction. With a more active catalyst, it is possible that such a long re action time is not necessary to reach a reasonable yield. The main objectives in this wo rk is to 1) optimize the Pd lo ading for selected supports, 2) optimize precipitation methodology and 3) optimize reaction conditi ons, i.e. reactant-to-catalyst ratios and reaction time. 6.2 Experimental 6.2.1 Catalysts preparation The catalysts were prepared using the prec ipitation method and commercially available nanoparticles supplied by NanoScale Materials Inc [67] or traditional porous supports from Alfa Aesar [69]. These supports were characterized in previous work [Chapter 3, Table 3-1 and Table 3-4]. In the precipitation method, the s upport was dispersed in to a solution of palladium nitrate. The porous oxide pellets were gr ound before dispersion. The mixt ure was then titrated with a NaOH solution, which formed Pd(OH)2 on the support [51]. The amount of NaOH used in for initial optimization corresponded to 50% stoichiometric excess based on the amount of Pd nitrate. For the optimization of the final precipitation pH, the dispersions were titrated with NaOH until a pH of 9 or 11 was obtained. The pH was measured with an Orion pH meter equipped with a ROSS Sure-Flow probe. For most catalysts the resulting mixture was aged overnight at room temperature before it was filtered. The recovered ca talysts were rinsed by stirring in water overnight, followed by another fi ltration. For some catalysts the aging and/or 113

PAGE 114

rinsing step was omitted. The catalysts were then dried overnight at 105C and calcined at 350C for 3 hours. Additionally, the calcination step was omitted for select catalysts to determine if it is necessary to heat the catalyst at higher temperatures before reaction to obtain reproducible results. Other catalysts were calcined at temperatures higher than 350C. 6.2.2 Reaction Conditions The 4-methylpyridine (Acros) was doubly distil led over KOH prior to use. In a typical reaction run 0.7 g of catalyst was pl aced in a round bottom flask al ong with 7 g of the distilled 4methylpyridine for a reactant-to-catalyst ratio of 10:1. The reaction mixture was evacuated and an oxygen atmosphere introduced before it was he ated to the boiling point (145C). The reaction proceeded under reflux for 72 hours. For some catalysts 1 g of catalyst and 5 g of 4methylpyridine or 0.5 g catalyst and 10 g of 4-met hylpyridine were used to obtain reactant-tocatalyst ratios of 5:1 or 20:1 (by weight) resp ectively. For reaction time optimization the reaction mixture was removed from reflux after 24 or 48 hours. Af ter reaction the flask contents were filtered using a glass microfiber filter and washed with chlo roform to dissolve the product. The chloroform, water and unreacte d 4-methylpyridine were removed using a rotary evaporator. 6.2.3 Palladium Dispersion Chemisorption measurements to characterize active surface area of the catalysts and the number of acidic and basic sites were pe rformed on a ChemBET 3000 instrument from Quantachrome Inc. The palladium surface area was measured by CO titration after reduction in hydrogen. The reductions were perf ormed using mild conditions (170 C in a stream of 5% hydrogen) to minimize sintering of the formed Pd phase and to be close to the temperatures experienced by the catalysts duri ng reaction. An explanation of the calculations used for Pd surface area and crystallite size are provided in previous work [see Chapter 3]. The surface areas 114

PAGE 115

of supports, as received, were characterized in previous work [ 66] using a Quantachrome Nova 1200 instrument. 6.3 Results and Discussion 6.3.1 Palladium Loading It was previously shown that reducing the palladium loading from 10% to 5% has little effect on the yield per g catalyst for the n-Al2O3(+) [66]. As is evident from Table 6-1 this is also true when going from palladium loadings of 5% to 2.5%. As a resu lt, the product yield per gram of palladium is doubled every time the load ing is reduced by half. Consequently, a 2.5% Pd loading gives four times as much product per gram of palladium as a 10% Pd loading. As is evident from Table 6-1 the palladium surface area decreases with the palladium loading. This means that the catalyst with the highest palla dium surface area also gives the lowest product yield per gram of palladium, and consequently the lowest turnover fre quency, i.e. the lowest amount of product formed per surface palladium for this set of catalysts. The turnover frequencies increase with decr easing palladium loading on thes e catalysts down to ~2.5% Pd loading (Table 6-1). The tur nover frequency of the 2.5% Pd/n-Al2O3(+) catalyst (130) is the highest observed on these catalysts and is consider ably higher than those of the 10%, 5% and 1% Pd/n-Al2O3(+) catalysts (35, 70, and 55 resp ectively). While the yield per gram of catalyst is reasonably constant between palladium loadings of 2.5-10%, a 1% palladium loading results in a dramatic decrease in the yield per gram of catalyst. In this case, the turnover frequency is also considerably reduced from the highest values observed for the 2.5% Pd loading (Table 6-1). Evidently, at the low 1% loading the amount of active palladium available at the surface is decreased compared to the available surface pallad ium on the 2.5% and 5% catalysts. Again it is evident that there is no direct correlation betw een the measured Pd surf ace area and the catalytic activity. Even though PdO is the active phase, it is not expected that the mild reduction used 115

PAGE 116

before the CO titration measurements to de termine the Pd surface area would lead to discrepancies of this size. Also, considering th e high dispersions on all th ese catalysts, it is not likely that there are large differences betw een bridged and linearly bound CO on the surface, which would lead to difficulties in accurately determining the Pd su rface areas using our instrument. The observed results may be due to a particle size dependence of the catalytic activity on these catalysts. If only palladium particles below a certain particle size are active, then as larger particles are formed the catalytic activity per Pd surface area would decrease with higher palladium loadings. Another explana tion would be a limited number of activating support sites, i.e. palladium interacting with these specific support site s would result in much more active palladium species than palladium intera cting with other sites on the support. If this is the case, then the 1% Pd loading would not be sufficient to fill up (saturate) all these active support sites on the n-Al2O3(+). It appears that Pd lo adings of ~2.5% for the n-Al2O3(+) support is optimal. This catalyst produces the highest yields of desired product per g of Pd. In contrast to the Pd/n-Al2O3 catalyst, the p-TiO2-supported palladium catalyst does have some sensitivity to loading. This indicates that some supports either do not exhibit a clear saturation or that the saturati on Pd loading is higher on this s upport. The turnover number still significantly decreases as loading increases. Ho wever, it appears that optimal loadings, may vary from supports. 6.3.3 Reactant-to-Catalyst Ratios As expected, the low 5:1 reactant-to-catalyst ratios resulted in decreased yields per gram of catalyst (palladium), while the 20:1 ratio had lim ited improvement in some cases (Table 6-2). However, since less reactant is used per gram of catalyst for the 5:1 ratio compared to the 20:1 ratio, the conversions were 2 to 3 times higher for the 5:1 compared to the 20:1 ratio. If the sole or major source of catalyst deactivation was poisoning of active sites by the product or by116

PAGE 117

product (the terpyridine) it would be expected that the reactions were more sensitive to the reactant-to-catalyst ratios. While some effect is indeed observed, since the yield of product per gram of palladium is different for the various reac tant-to-catalyst ratios, this is evidently not the major source of deactivation in these reactions. The results from the reactant-to-catalyst ra tios indicate that a ra tio of ~10:1 appears reasonable unless unreacted started ma terial is recycled at the end of the reaction, in which case a ratio of 20:1 or higher can be used to increase the yields. Howeve r, if the goal is to produce more product per gram of catalyst, it is likely better to adjust another parameter that has a larger influence on the product yield. 6.3.4 Titration pH When preparing a 5% Pd catalyst on 1.9 g of n-Al2O3(+) or p-TiO2 a typical pH of the resulting dispersion after titration with 50% stoichiometric excess base (based on the palladium(II) nitrate) was ~10.5-11 (Table 6-3, Entrie s 1-2). In the case of nano-silica supports (n-SiO2) the resulting dispersion had a pH of ~9.0-9.5 after titrati on (Table 6-3, Entry 3). The pH of the n-Al2O3(+) dispersion increases relatively slowly near the end of its titration. In contrast, the pH of the p-TiO2 and n-SiO2 dispersions increase rapidl y near the end of their titrations, if the amount of base corresponds to a 50% excess of the palladium(II) nitrate in the solution. This indicates that th e acidity of the suppor t significantly influe nces the pH of the solution for these supports. Consequently, titration based upon a 50% excess NaOH based on the stoichiometric amount needed to precipitate Pd(II) as Pd(OH)2 may not be adequate for reproducibility on p-TiO2 and n-SiO2 supports. This is particularly true when using hygroscopic materials, such as sodium hydroxide and pallad ium nitrate dihydrate, for which the compositions of commercial preparations va ry. Consequently, weighing of these materials is inherently imprecise, which is problematic since small changes in the concentrations of these materials can 117

PAGE 118

lead to large variations in the final pH for dispersions in which there is a steep slope in the pH near the end of the titration. For low Pd load ings on acidic supports m onitoring the pH during precipitation is even more important, since the final pH may be significantly lower than the desired pH to complete the Pd(OH)2 precipitation. This may explain the lower yields obtained from a 2.5% loading on p-TiO2 when the pH is not monitored. Titration to a pH of 11 duri ng the preparation of the n-SiO2 catalyst did increase the yield modestly and improved reproducibility (Table 63). Additionally, no discernable brown color was observed in the filtrates during the recovery and rinsing processes. The pale brown tint due to palladium compounds was often seen in the fi rst and/or second filtrate when the amount of NaOH was based on a 50% stoichiometric excess re quired for the Pd(II) concentration used in the preparation of the Pd/n-SiO2 catalysts. Monitoring the pH is also important at higher pH values. At a pH of 12 a large portion of the p-titania catalyst cannot be recovered from the mixture supports by filtration or centrifugation. It is possible that at very high base concentrations the sup port is etched, i.e. it begins to dissolve into the basic solution. Anothe r possibility is that the high pH is far from the isoelectric point, with the resu lt that aggregates are broken up and the suspension produced contains particles too small to be filtered. These phenomena make pr ocessing difficult at this pH. 6.3.5 Pre-reaction Treatment It was found that unaged catalysts gave incons istent results with large differences in activity. Aging and rinsing the ca talysts did increase both Pd disp ersion and yield (Table 6-4). The aging process may be important in allowing Pd(OH)2 to be transported on the surface of the support from the solution. For the Pd/n-Al2O3(+) catalyst calcination of the sample did improve activity moderately with temperatures up to 350 C, compared to the yields obtained from the catalyst dried at 105C. Calcination of up to 550C induced only minor deactivation of the 118

PAGE 119

catalyst (Table 6-4). The small increase in yiel d with calcination temperat ure (between drying at 105C and calcining at 350C) ma y due to the removal of further water from the catalyst from either a Pd(OH)2 or a PdOH2O phase. It could also be removal of water from the alumina support, since powder XRD measurements have shown that the n-Al2O3(+) contains a bohemite (AlOOH) phase before calcinations (See chapter 5). For the Pd/p-TiO2 dried at 105 C, significant mirroring of the react ant flask and a dark brown pr oduct color is observed, which indicates leaching of the palladium from the s upport. Calcination at 450C gives a moderate decrease in activity while calcina tions at 550C gives a greatly reduced activity of less than one half the 2.6 g/g seen for 5% Pd/p-TiO2 calcined at 350C. 6.4 Conclusions In this study it was determined that calci nations up to 350C increases activity and reproducibility of the catalysts over uncalcined samples. In the case of Pd/p-TiO2 calcination at 550C gave approximately half the yield of a catalyst calcined at 350C, while higher calcination temperatures did not have a si gnificant effect on the n-Al2O3(+) catalyst. Additionally it was found that higher reactantto-catalyst ratios incr eased activity based on product yield per unit weight of catalyst, but above 10:1 by mass the incr ease is not significant. It was also found that the yield per gram of catalyst obtained from the Pd/n-Al2O3(+) is insensitive to loading between 2.5% and 10% on a metal basis, while Pd/p-TiO2 does exhibit an increase in yield per gram of catalyst with increased Pd loading. While furthe r work is likely to produce improved catalysts, the 2.5% Pd precipitation onto n-Al2O3(+) and the 5% Pd on p-TiO2 are the best catalysts to date with activities of 2.7.27 (105% g/g Pd) and 2.6.26 g product /g catalyst respectively. 119

PAGE 120

Table 6-1 Effects of palladium loading on catal yst properties and catalytic activities. Pd/n-A l2O3( +) Loading g product/g catalyst g Product /g Pd % Dispersion Pd SA [m2/g] Turnover Frequency2 n-Al2O3(+) 1.0% 0.66 66 72 2.8 55 n-Al2O3(+) 2.5% 2.7 108 48 4.7 130 n-Al2O3(+)1 5.0% 2.5 49 43 8.4 70 n-Al2O3(+) 10.0% 2.4 24 38 15 35 p-TiO2 1.0% 0.9 90 56 2.2 95 p-TiO2 2.5% 1.6 64 44 4.3 85 p-TiO2 1 5.0% 2.6 52 44 8.6 70 p-TiO2 1 10.0% 3.2 32 35 16 50 1Result from previous work [See chapter 3] 2Defined at mol product/mol surface Pd 120

PAGE 121

Table 6-2 The effects of catalyst reactant ratio on the catalytic activities of Pd/ n-Al2O3 and Pd/pTiO2. Catalyst Reactant : Catalyst ratio g product/g palladium % conversion 5% pd on n-Al2O3 5:1 32 23 10:1 51 25 20:1 63 16 5% Pd on pTiO2 5:1 34 34 10:1 46 23 20:1 49 13 121

PAGE 122

Table 6-3 Effects of titration pH on the activities of selected catalysts. Entry Catalyst Titration pH Product yield [g/g Pd] 1 5% Pd/n-Al2O3 1 10.4 52 2 5% Pd/p-TiO2 1 11.1 52 3 5% Pd/n-SiO2 1 9.5 32 4 5% Pd/n-Al2O3 11.0 43 5 5% Pd/p-TiO2 11.0 46 6 5% Pd/n-SiO2 11.0 40 7 5% Pd/n-SiO2 9.0 27 1Results from previous work [66] 122

PAGE 123

Table 6-4 Effects of catalyst pretreatments on the activities and dispersion of 5% Pd precipitated on nano alumina (+) or porous titania. Catalyst g product/g pa lladium % Dispersion nano-Al2O3(+) Pretreatment: Unaged 51 Unaged (repeat) 27 20 Unrinsed 25 Dried at 105C 45 43 Calcined 350C 53 35 Calcined 450C 49 40 Calcined 550C 47 p-TiO2 Dried at 105C 54* Calcined 350C 52 44 Calcined 450C 48 41 Calcined 550C 25 12 *Evidence of Pd leaching and/or catalyst contamination in product 123

PAGE 124

CHAPTER 7 CONCLUSIONS The research presented here was performed to study an important type of catalyst (supported palladium) and its use for producing a usef ul chemical (4,4-dimet yl-2,2-bipyridine). Additionally, the use of nanopart icles for enhancing catalytic activities was studied. The objectives of this work, included: 1) showing that nanoparticle oxides are viable supports for palladium catalysts in the produc tion of 4,4-dimetyl2,2-bipyridine via oxid ative coupling of 4methylpyridine, 2) determining what support/catal yst properties are responsible for activity, 3) developing an economically viable catalyst. 7.1 Summary of Results In the initial study, the use of two nanoparticle aluminas (n-Al2O3) and a traditional porous alumina as supports were studied. The we t impregnation method of deposition on the n-Al2O3(+) gave poor activities. The yield obtained from a 5% loading of palladium precipitated onto n-Al2O3(+) is significantly higher than the sa me loading on carbon (49 4 vs. 33 6 g/g Pd). In addition to nanoparticle alumina, severa l other nanoparticle catalys ts were determined to be active for the oxidative coupling react ions including nanoparticle ceria (n-CeO2), titania, silica, magnesia, zirconia (n-ZrO2), and ceria-doped zirconia, with activities ranging from (30-54 g/g Pd), while nanoparticle calcium oxide, coppe r oxide, aluminum hydroxide and tin oxide had little or no activity (less than 17 g/g Pd). Additionally porous silica, titania (p-TiO2) and alumina ( -Al2O3), were also active. The porous titania catalyst was of particular interest as it gives activities similar to the n-Al2O3(+) support (49 4 g/g Pd). This indicates that a porous oxide with strong Pd/support interactions and/or hi gh surface area (and, perhaps, a relatively high density of corner and edge sites) may be able to match nanoparticle oxid e supports in activity for the oxidative coupling reaction. 124

PAGE 125

The two main factors in catalyst optimization are cost and activity. The n-Al2O3(+) as well as n-ZrO2 and p-TiO2 catalyst were shown to have excellent activities compared to the standard palladium on carbon catalyst. However the n-ZrO2 and p-TiO2 are significantly less expensive than the n-Al2O3(+). It was determined that low loadings of palladium (2.5%) can be as active as higher loadings (5% and 10%) on the n-Al2O3(+) support. As the cost of palladium is the major cost in catalyst preparation, a lower loading with the same yield as a higher loading, represents a significantly cheaper catalyst. Also, it was determined that the use of ZrO2 and CeO2 as additives could boost the activity of a catalyst by up to 40% (up to yields of 3.1 0.3 g/g catalyst for 5/5/90% coprecipitated Pd/Zr/Al2O3). 7.2 Nanoparticles Oxides as Palladium Support 7.2.1 Viability At the outset of this research it was hypothe sized that nanoparticle alumina oxides may be more active supports for palladium compared to traditional porous supports when used in the oxidative coupling of 4-methylpyridine. Palladium dispersed onto a high quality nanoparticle alumina [n-Al2O3(+)] via precipitation indeed gives exce llent yields; 2.5 0.15 g/g catalyst at a 5% Pd loading with a maximum yield of 105 g/g Pd at a loading of 2.5% Pd (metal basis) on the [n-Al2O3(+)]. This exceeds the highest reported activitie s for this catalyst in literature of (40 g/g Pd on Carbon [38] see Table 2-1). Additionall y, palladium on other nanoparticle supports including nanomagnesia, zirconia, zirconia (w ith 10% ceria doping), and zinc oxide were found to have activities that exceeded reported values for Pd/C. Thus, in term s of activity per mass of catalyst and mass of palladium na noparticle oxides were found to be viable catalyst supports. 125

PAGE 126

7.2.2 Traditional Porous Supports In addition to nanoparticle oxides, it was found that palladium dispersed on traditional porous supports were also active, with 5% Pd/p-Al2O3, Pd/p-SiO2, and Pd/p-TiO2 prepared by precipitation giving yields of 1.2,1.6, and 2.6 % g/g catalyst respectively. (It is particularly noteworthy that Pd/alumina catalysts had been reported as inactive in literature [1]. However, this may be attributed to the us e of pre-reduced catalysts). While only the activity of the p-TiO2 was significant (in that it greatly exceeds th e activity of commercial Pd/C; 2.6 0.25 vs. 1.6 0.3 g/g Pd), further study of porous catalyst supports is needed. 7.2.3 Nanoparticle vs. Porous Supports The high activity of at least one porous oxide support catalyst ( 2.6.25 g/g catalyst for Pd/p-TiO2) combined with the low activity of several nanoparticle supports raises the question of whether nanoparticles have a significant advantage over porous oxides. At current market prices porous oxides can be significantly cheap er than nanoparticles ($112.0/kg for p-TiO2 [69] vs. $694.10/kg for n-Al2O3(+)[66]). Thus, at 5% loadings the p-TiO2 is more cost effective than nAl2O3(+), since the yields are almost the sa me: 2.6 0.15 g/g and 2.5 .25 g/g catalyst, respectively). At a 2.5% loading of Pd the Pd/n-Al2O3(+) catalyst is more active per unit weight of catalyst compared with the 2.5% Pd/p-TiO2. For Pd market prices of ~$15/g, the Pd content alone would cost ~ $0.75/ gram cat alyst at 5% loading and $0.38/g catalyst at 2.5% which is less than the difference in support cost. However, th e palladium nitrate precu rsors (for a 50g or less purchases) sell for large premiums over the mark et value of the actual Pd metal content ($60/g palladium 1 ). Additionally, it seems likely that the cost of producing nanoparticle oxides will decrease as the technology becomes more deve loped and the scale of production increases. 1 Based upon 50 g of palladium nitrate dihydrat e at $1,170 (39% Pd) from Alfa Aesar #11035 [69] 126

PAGE 127

Consequently, while porous oxides supports certainly deserve further st udy, nanoparticles oxides have greater promise for making the most cost effective catalyst. 7.3 Factors in Catalyst Activity and Deactivation 7.3.1 Palladium Dispersion and Surface Area Perhaps the most striking resu lt in this work is the lack of strong correlation between measured Pd dispersion and/or surface area. This seems to hold across multiple catalyst supports and within different loadings on the same support. If these results are reasonable measurements of the PdO surface area, then this has some impor tant implications in that they indicate a structure sensitivity (which will be discussed sh ortly). However, the limitations of the CO chemisorption measurement must, certainly, be taken into account. It is po ssible that the CO to Pd ratio is not constant in the range of Pd particle sizes measured. If, for example the Pd:CO stoichiometry changes from near 2:1 to near 1:1 for palladium particles when going from loadings of 2.5% to 10% on the n-Al2O3(+) support, then the disper sions would be 96% vs. 38% Pd rather than the 48 8 % vs. 38 6% reported. Furthermore, it is possible that the palladium dispersion is altered significan tly during reduction due to sinter ing, and that the degree of sintering could be much higher on some supports than others. Mo re detailed studies, possibly incorporating a combination of XRD and/ or TEM and possibly ISS (Ion Scattering Spectroscopy) would be useful in future studies to address thes e uncertainties. However, it seems unlikely that sintering and stoichiometry can fully account for the wi de variances seen in the dispersion of active catalyst. For example, catalysts with measured dispersions of less than 5% (such as 5% Pd/p-SiO2 and -Al2O3) were shown to give yields th at are as high as 1/3 to 1/2 of the yields obtained from catalysts with dispersions and Pd surface area s that are more than eight times greater. 127

PAGE 128

7.3.2 Structure Sensitivity The low apparent correlati on between the measured pa lladium surface area and the catalytic activity very strongly suggest that the reaction is structure sensitive 2 as a structure insensitive reaction would exhibit a direct corr elation between the numbe r of surface palladium atoms and the catalytic activity. In this work, it was not possible to conc lusively identify what structures are responsible for ac tivity. At first glance, the lo w correlation between activity and dispersion or palladium surface area suggests that the structure sensitivity is not related to particle size (which varies with dispersion), but this may be misleading. The turnover number varies significantly between catal ysts, and in some low dispersion catalysts is very high (1000 vs. 70 for 5% Pd/ -Al2O3 vs Pd/n-Al2O3(+). It is likely that only a small fraction of the surface Pd/PdO is active in any given catalyst. Because the measured dispersion can only give an average particle size, the presence of a minor frac tion (in terms of SA) of smaller particles in the low dispersion catalyst can not be discounted. Th at the activity does seem to increase with dispersion, in the higher dispersion ra nge (>10%), suggests that if there is a size dependence, it is the smaller Pd/PdO particles that are responsible for activity. Structure sensitivity could also be related to the support (as discussed in the next section). The proposed mechanism (see Chapter 5) does s uggest several factors th at could contribute to structure sensitivity. The cat alyst would need to hold oxygen a nd Pd atoms in a configuration that allows for formation of a Pd-N bond / CH-O bond transition state. The structure should also be stable enough to allow oxygen to be removed and added. Additionally the active structure must be large enough to accommodate two adsorbed pyridines Using FTIR with 2 A detailed discussion of structure sensitivity a nd its various modes can be found in Masel [36] 128

PAGE 129

pyridine 3 as an adsorbate, could prove useful in id entifying active catalyst sites and their related structures. Use of nanoparticle-s upported catalysts, were the active Pd/PdO phase is expected to be on the surface (rather than in pores) would be particularly suited for such a study. 7.3.3 Support Properties The study of the wide range of active and inac tive catalyst supports can give some insight into the properties that lead to catalytic activity. Firstly, ve ry high surface ar ea supports are likely to give high catalytic activities. The hi gher surface areas may be i ndirect indications of higher concentrations of edge and surface sites, which could be more active than oxide surfaces with lower surface areas. The interactions with low coordination sites could either influence the Pd/PdO structure or result in st rong coordination of Pd ions to these sites and consequently forming more electrophilic palladium species. If these low coordinati on sites are interacting strongly with the Pd/PdO in this catalyst, these support sites would be exp ected to adsorb probe molecules such as CO2 and pyridine. Consequently, the use of FTIR/TPD would likely give valuable insights into the concentration and nature of such sites. Such a study would also be useful in more fully probing acid and basic sites on the bare supports. The concentrations of these acid and base sites alone, did not give much insight into activity or dispersion. However, the effectiv eness of precipitation vs. impregnation strongly implies that acid/base propert ies are important (in that Pd+2/Pd(OH)2 is attracted to the support at some pH). The lack of any measured strong basic sites on the porous and nano silica together with their relatively high ac tivities (1.6 0.15 and 1.7 0.2 g product/ g catalyst at a 5% Pd loading, which is comparable to Pd/C), would su ggest that any interact ing sites are acidic in nature. In addition to FTIR studies, laboratory preparation of nanoparticles, in which the 3 Pyridine is a common adsorbate molecule in FTIR used for probing acid sites. Pyridine would seem likely to give useful information a bout PdO adsorption sight s in this reaction. 129

PAGE 130

particles properties could be adjusted slightl y, would be useful in pr obing the interactions between palladium and low-coordination and/or acid sites. Supports such as ZnO, TiO2, ZrO2, and CeO2 reveal strong interactions with Pd/PdO. These interacting oxides may be active due to an ability to supply oxygen to the reaction and/or they may also cause electronic in teractions with Pd/PdO at the surface. These interactions may be similar to the edge and corner sites-Pd/PdO interactions on the high surface area supports. The XPS indicates relatively high binding energy Pd2+ states on some high surface areas oxides, as well as on oxide with strong metal support interactions (SMSI) (In the case of Pd/n-Al2O3(+) a Pd4+ like binding energy is observed). This may indicate similar behaviors of these SMSI and high surface area support surfaces. However, the effects of ZrO2 and CeO2 on metal redox and oxygen storage properties reported in literature [23,80,85], along with their abilities to enhance the activity of the Pd catalyst when used as addi tives, would suggest that for these two oxides, the ability to supply oxygen is related to their activity promotion. 7.3.4 Deactivation Several possibilities for deactivation were suggest ed in this work. It was determined that adding water to the reaction mixture decreased activ ity. This suggests that water can inhibit the reaction, possibly by competing with pyridine for catalyst sites. However, the diminished activity was not sufficient to attr ibute deactivation solely to the presence of water. Observation of differential charging in XPS of the spent p-TiO2 catalyst that is diminished by sputtering indicates the presence of a contaminant on the palladium surface. This points to fouling or poisoning by either an unidentified byproduct, or possibly the desired pr oduct itself. XPS and XRD also suggest that the PdO on the fresh catalyst is completely reduced to Pd metal at the end of the reaction. Since pre-reduced catalysts have been shown to be ineffective in this reaction [5], it seems likely that a completely redu ced Pd phase would not be active. Water 130

PAGE 131

contamination, carbon fouling or product poisoni ng, and Pd reduction could be addressed by regeneration of the catalyst in oxidative condi tions. For such a treatment, Pd/metal oxide supported catalyst, and part icularly those with ZrO2 and CeO2 as supports or additives in this reaction are very promising, although the thermal stability of some of the nanoparticle supports would be a concern. 7.4 Final Remarks In general, it has been shown that nanoparticle oxides can be used to make highly active catalysts, although some traditional oxide supports may be competitive. Additionally, important insight into the activity of nanopa rticle catalysts has been obtaine d. Either a high concentration of corner and edge sites or strong oxide-palladium interactions give active catalysts. This activity may result from more electrophilic pall adium/palladium oxide pa rticles as a result of strong metal-support interactions on these types of catalysts. More broadl y, excellent alternative catalysts to palladium on carbon have been developed for the oxidative coupling of 4methylpyridine. This includes both a nanoparticle catalyst (2.5% Pd/n-Al2O3+) and a traditional porous support catalyst (5% Pd/p-TiO2) (2.7 0.27 and 2.6 0.26 g/g catalyst). 7.4.1 Future Work The work presented here has highlighted se veral avenues for future study. Certainly, further optimization of catalyst formulation and production is necessary to fully exploit the economic possibilities of this system. For example, further study of the CeO2 and ZrO2 as supports and additives would be in order. Given their effectiveness as additives in 5% Pd/nAl2O3 and 5% Pd/n-TiO2, their use in the hi ghly active 2.5% Pd/alu mina and 5% Pd/p-TiO2 is very promising. Furthermore, now that several promising catalysts have been developed, a study focusing on the reactivation of catalyst can be performed. This would include for example 131

PAGE 132

heating in an oxygen-rich environment, steam ox idation, or the use of an oxygen plasma to burn off the carbon on the surface and reoxidize the Pd metal to PdO. Additionally, several, more expansive, studies into fundamentals of this system seem inorder. Production of nanoparticles in which sizes could be controlled, would allow more controlled studies of the importa nce of support structure. Use of a combination of extensive XRD, ISS, and/or TEM would be useful in obtaining better estimates of Pd dispersion, particularly in the unreduced PdO state. Perhap s the most interesting prospect, though, is the use of FTIR. The use of CO2 and pyridine as probe molecules in FTIR/TPD could give valuable insight into the nature of support sites, and help determine what is responsible for the activities of palladium deposited on supports with these sites. Furthermore, pyridine may be used to directly probe the adsorption sites on Pd catalyst. Nanoparticle-supported Pd catalysts may be particularly well suited for this, as it seems likely that a relatively high fraction of Pd/PdO particles would be on the surface relative to a porous catalyst. 7.4.2 Broader Impact As discussed in numerous parts of this work, the oxidative coupling of 4-metlypyridine is important, as bipyridines have a wide variety of applications [20-29]. Re ducing their cost would have a great impact on the economic viability of many catalyst systems. Additionally, a catalyst active for oxidative coupling 4-methylpyridine could be appl ied to other systems such as the coupling of unsubstituted and substituted pyridin es, benzenes, as well as other heterocyclic systems, which are all important for the synthesis of fine chemicals. Because of the simple nature of the reaction and the omission of solven ts and halogens, these catalysts could have a great impact on environment as well as the cost of some fine chemicals. The impact, however, is not limited to oxidative coupling reactions. Pd catalyst are used in a wide number of systems [7-19], and their use in other oxidation reactions (particularly the low132

PAGE 133

temperature oxidation of CO and CH4 [12,13,16-18]), as well as methanol reforming and water gas shift reactions [21,43,62,63,84] are particularly important. The catalysts developed in this work increase our knowledge of these important Pd/P dO systems, but this work may have a more practical application. Many of th e catalysts were found to have excellent dispersions, despite relatively simple methods of preparations, and th eir high activities in this reaction may, possibly, correlate to higher activities in other systems, particularly oxidation reac tions. More broadly, the use of precipitation on nanoparticles may be applicab le to a wide range of active metals and their corresponding reaction systems. 7.4.3 Best Catalyst for Oxidativ e Coupling of 4-Methylpyridine The practical goal of this work was to devel op a more economically vi able catalyst for the oxidative coupling of 4-methylpyridine. While further work is likely needed to produce improved catalysts, the 2.5% Pd precipitation onto n-Al2O3(+) and the 5% Pd on p-TiO2 are the best catalysts to date with activities of 2.7.27 g product /g ca talyst (105% g/g Pd) and 2.6.26 g product /g catalyst(52g/g Pd) respectively. While the viability of one vs. the other is related to the cost of the Pd precursor, in small batch quantit ies, of palladium nitrate, the 2.5% Pd/n-Al2O3(+) catalyst is currently the most cost effective. 133

PAGE 134

LIST OF REFERENCES 1. G.D.F. Jackson, W.H.F. Sasse, C.P. Whittle, Aust. J. Chem. 16 (1963) 1126 2. P.E. Rosevear, W.H.F. Sasse, J. Heterocycl. Chem. 8 (1971) 483. 3. P.E. Rosevear, W.H. Sasse, Patent App. No. 40930/72, 1973, p. 1. 4. S. Munavalli, M. Gratzel, Chem. Ind. (1987) 722 5. H. Hagelin, B. Hedman, I. Orabona, T. k ermark, B. kermark, C.A. Klug, J. Mol. Catal. A 164 (2000) 137. 6. L. Guczi, A. Beck, A. Horvth and D. Horvth, Top. Catal. 19 (2002) 157 7. L. S. Escandn, S. Ordez, A. Vega, F. V. Dez, J. Haz. Mat. 153(2008) 742 8. M. Schmal, M. Souza, V. Alegre, M. da S ilva, D. Cesar,C. Perez, Catal. Today 118 (2006) 392. 9. P. Glin and M. Primet, Appl. Catal. B 39 (2002) 1-37, and references therein. 10. T.V. Choudhary, S. Banerjee and V.R. Choudhary, Appl. Catal. A 234 (2002) 1-23, and references therein. 11. D. Ciuparu, M.R. Lyubovsky, E. Altman, L.D. Pf efferle, A. Datye, Catal. Rev. Sci. Eng. 44 (2002) 593-649. 12. K. Sekizawa, H. Widjaja, S. Maeda, Y. Ozawa, K. Eguchi, Catal. Today 59 (2000) 69. 13. K. Sekizawa, H. Widjaja, S. Maeda, Y. Ozawa, K. Eguchi, Appl. Catal. A 200 (2000) 211. 14. Y. Li, Y. Fan, H. Yang, B. Xu, L. Feng, M. Yang, Y. Chen, Chem. Phys. Lett. 372 (2003) 160. 15. T. Kawakami, Y. Ooka, H. Hattori, W. C hu, Y. Kamiya,T. Okuhara Appl. Catal. A (2007) (In press) 16. O. Pozdnyakova, D. Teschner, A. Wootscha, J. Krhnert, B. Steinhaue r, H. Sauer, L. Tothc, F.C. Jentoft, A. Knop-Gericke, Z. Pal, R. Schlg, J. Catal. 237 (2006) 17. 17. H. Zhu, Z. Qin, W. Shan, W. Shen, J. Wang, J. Catal. 233 (2005) 41. 18. R Zhou, B Zhao, B. Yue, Appl. Surf. Sci. 254 (2008) 4701. 19. D. Ciuparu, A Ensuque, F. Bozon-Ve rduraz, Appl. Cata l. A 326 (2007) 130. 20. C. Kaes, A. Katz and M.W. Hosseini, Chem. Rev. 100 (2000) 3553. 134

PAGE 135

21. V. Balzani, G. Bergamini, F. Marchioni and P. Ceroni, Coord. Chem. Rev. 250 (2006) 1254-1266. 22. C.D. Clark, M.Z. Hoffman, Coord. Chem. Rev. 159 (1997) 359-373. 23. R.C. Evans, P. Douglas and C.J. Wins com, Coord. Chem. Rev. 250 (2006) 2093-2126. 24. L. Gmiz-Gracia, A.M. Garc a-Campaa, J.J. Soto-Chinchilla, J.F. Huertas-Prez and A. Gonzlez-Casado, Trends Anal Chem. 24 (2005) 927-942. 25. N.M.L. Hansen, K. Jankova and S. Hvilsted, Eur. Polym. J. 43 (2007) 255-293. 26. D.L. Feldheim, C.J. Baldy, P. Sebring, S.M. Hendrickson, C.M. Elliott, J. Electrochem. Soc. 142 (1995) 3366. 27. H.Ishii, M Goyal, M Ueda, K. Takeuchi and M. Asai, Applied Catalysis A 201(2000) 101. 28. F. Tsai, B. Lin, M. Chen, C. Mou and S. Liu, Tetrahedron 63 (2007) 4304. 29. W. Wu, S. Chen and F. Tsai, Tetrahedron Lett. 47(2006) 9267. 30. A. Launkonis, D. Lay, A. Mau, A. Saragesor and W. Sasse, Aust. J. Chem. 39 (1986), 1053-1062. 31. 4,4-dimethyl 2,2-bipyridine, GFS Chemicals: http://gfschemicals.com/statics/product details/cHJvZHVjdGRldGFpbHMxMjc=.html accessed 10/2/2008 32. L.A. Summers, Adv. Hetero cycl. Chem. 35 (1984) 281. 33. G. Cravotto, M. Beggiato, A. Penoni, G. Palmis ana, S. Tollari, J. Lvque, W. Bontrath, Tetrahedron Lett. 46 (2005) 2267. 34. M. Heller, U Schubert, J. Org. Chem. 67 (2002) 8269. 35. R. I. Masel, Principles of Adsorption and Reaction on Surfaces. Wiley: New York (1996). 36. B. D. Cullity, S. R. Stock, Elements of X -ray Diffraction, 3rd Ed. Prentice-Hall: New York (2001). 37. J.F. Moulder, W.F. Stickel, P.E. Sobol, K.D Bomben, Handbook of X-ray Photoelectron Spectroscopy, Physical-Electronics Inc.: Eden Prairie, MN (1995). 38. P.K. Ghosh and T.G. Spiro, J. Am. Chem. Soc. 102 (1980) 5543. 39. P.A. Adcock, F.R. Keene, R.S. Smythe, and M.R. Snow, Inorg. Chem. 23 (1984) 2336. 135

PAGE 136

40. G. Sprintschnik, H.W. Sprint schnik, P.P. Kirsch, and D.G. Whitten, J. Am. Chem. Soc. 99 (1977) 4947. 41. G.R. Newkome, A.K. Patri, E. Holder, U. S. Schubert, Eur. J. Org. Chem. (2004) 235. 42. U.S. Schubert and C. Eschbaumer, Angew. Chem. Int. Ed. 41 (2002) 2892-2926. 43. V. Balzani, A. Juris, M. Venturi, S. Campagna, S. Serroni, Chem. Rev. 96 (1996) 759. 44. K. Kalyanasundaram, Coord. Chem. Rev. 1982, 46, 159-244. 45. 4,4-dimethyl 2,2-bipyridine, GFS Chemicals: http://gfschemicals.com/chemicals/gfschem-127.asp accessed 07/13/07. 46. 2-amino-4-methylpyridine, Fl uka >98%, Sigma-Aldrich: https://www.sigmaaldrich.com/catal og/search/ProductDetail/FLUKA/08630 accessed 08/30/07. 47. 4-Methylpyridine 98% Fluka, Sigma-Aldrich: http://www.sigmaaldrich.com/catal og/search/ProductDetail/FLUKA/80240 accessed 08/30/07. 48. K.W.R. de Frana, M. Navarro, Lonel, M. Durandetti, and J.-Y. Ndlec, J. Org. Chem. 67 (2002) 1838. 49. G.B. Hoflund, Z. Li, W.S. Epling, T. Gbel, P. Schneider and H. Hahn, React. Kinet. Catal. Lett. 70 (2000) 97. 50. NanoActive Aluminum Oxide Plus and Na noActive Aluminum Oxide, NanoScale: http://nanoactive.interka n.net/content/nanoac tive_materials/nanoactive_home.asp accessed on 03/07/07. 51. B. Didillon, E. Merlen, T. Pags, D. Uzio, Stud. Surf. Sci. Catal. 118 (1998) 41-54. 52. T. Lear, R. Marshall, J.A. Lopez-Sanchez, S.D. Jackson, T.M. Klaptke, M. Bumer, G. Rupprechter, H.-J. Freund and D. Lennon, J. Chem. Phys. 133 (2005) 174706. 53. P. Canton, G. Fagherazzi, M. Battagliarin, F. Menegazzo, F. Pinna, and N. Pernicone, Langmuir 18 (2002) 6530. 54. R.G. Heidenreich, J.G.E. Krauter, J. Piet sch and K. Khler, J. Mol. Catal. A 182-183 (2002) 499. 55. F. Zhao, K. Murakami, M. Shirai an d M. Arai, J. Catal. 194 (2000) 479. 56. A. Biffis, M. Zecca, M. Basato, Eur. J. Inorg. Chem. (2001) 1131. 57. G. Dyker, Angew. Chem. Int. Ed. 38 (1999) 1698-1712. 136

PAGE 137

58. D. Schrder and H. Schwarz, Angew. Chem. Int. Ed. 34 (1995) 1973-1995. 59. G.M. Badger and W.H.F. Sasse, Adv. Heterocycl. Chem. 2 (1963) 179. 60. J Consul, I. Costilla, C. Gigola, I. Baibich, Appl. Catal. A 339 (2008) 151. 61. J. Hong, W. Chu, M. Chen, X. Wang, T. Zhang, Catal. Commun. 8 (2007) 593. 62. A. Casanovas, J. Llorca, N.S. Homs, J. Fierr o, P. de la Piscina, J. Mol. Catal. A 250 (2006) 44. 63. S. Liu, K. Takahashi, K. Fuchigami, K. Uematsu, Appl. Catal. A 299 (2006) 58. 64. A. Garron a, K Lazar F. Ep ron/App Cat B: Env. 59 (2005) 57-69. 65. M. Garcia,1 A. Martinez-Arias, L Salamanca, J. Coronado, J. Anderson, J. Conesa, J. Soria, J. Catal. 187 (1999) 474. 66. L.M. Neal, H.E. Hageline-Weaver, J. Mol. Ca tal. A, 284 (2008) 141. (included, in part, in chapter 2) 67. NanoScale Materials Inc., http://nanoactive.interka n.net/content/nanoac tive_materials/nanoactive_home.asp accessed on 10/01/08. 68. Nanostructured & Amorphous Materials Inc., http://www.nanoamor.com/__nanoparticles accessed on 10/01/08. 69. Alfa Aesar, http://www.alfa.com/alf/index.htm accessed on 10/01/08. 70. E.D. Park, S.H. Choi, J.S. Lee, J. Phys. Chem. B 104 (2000) 5586, 71. A.D. Silva, M.L. Patitucci, H.R. Bizzo, E. D'Elia and O.A.C. Antunes, Catal. Commun. 3 (2002) 435. 72. D.R. Bryant, J.E. McKeon, B.C. Ream, J. Org. Chem. 33 (1968) 4123. 73. T. Ohishi, J. Yamada, Y. Inui, T. Saka guchi, M. Yamashita, J. Org. Chem. 59 (1994) 7521. 74. H. Hagelin, J.D. Oslob, B. kermark, Chem. Eur. J. 5 (1999) 2413. 75. T. Caputo, L. Lisi, R. Pirone, G. Russo, Appl. Catal. A 348 (2008) 42. 76. A. Marquez, J. Sanz, Appl. Surf. Sci. 238 (2004) 82. 77. 5y Palladium Price, Kitco, http://www.kitco.com/charts/popup/pd1825nyb_.html accessed 10/01/08 137

PAGE 138

78. C.E. Gigola, M.S. Moreno, I. Costilla, M.D. Sanchez, Applied Surface Science 254 (2007) 325. 79. W. Shen, Y. Matsumura, J. Mol. Catal. A. 153 (2000) 165. 80. J. Kaspar, P. Fornasiero, M. Grazi ani, Catalysis Today 50 (1999) 285. 81. J.M. Padilla, G. Del Angel, J. Na varrete, Catalysis Today 133-135 (2008) 541. 82. A. Trovarelli, C. de Leitenburg, M. Boar o, G. Dolcetti, Catal. Today 50 (1999) 353. 83. L. Bollmann, J. L. Ratts, A. M. Joshi, J. of Catal. 257 (2008) 43. 84. S. Liu, K. Takahashi, H. Eguchi, K. Uematsu, Catal. Today 129 (2007) 287. 85. L.M. Martinez T, C. Montes de Correa, J.A. Odriozola, M.A. Centeno, Catal. Today 107108 (2005) 800. 86. F. Pinna, Catal. Today 41 (1998) 129. 87. S. Jones, L.M. Neal, H. E. Hagelin-Weaver, App. Catal. B 84 (2008) 631. 88. C.T. Wang J.C. Lin, App. Surf. Sci. 254 (2008) 4500. 89. R.N. Gayen, S.N. Das, S. Dalui, R. Bhar A.K. Pal J. Crystal Growth 310 (2008) 4073 90. P.M. Maitlis, The Organic Chemistry of Palladium, Academic Press, New York, 1971. 91. J.Tsuji, Palladium Reagents and Catalysis: Innovation in Organic Synthesis, Wiley, Chichester, UK (1995). 92. J.-L. Malleron, J.-C. Fiaud and J.-Y. Legros, Handbook of Palladium-Catalyzed Organic Reactions, Academic Press Limited, London, UK, 1997. 93. A.K. Santra and D.W. Goodman, Electrochim. Acta 47 (2002) 3595-3609. 94. S.-H. Oh and G.B. Hoflund, J. Catal. 245 (2007) 35-44. 95. M. Hosseini, S. Siffert, H.L. Tidahy, R. Cousin, J.-F. Lamonier, A. Aboukais, A. Vantomme, M. Roussel, B.-L. S u, Catal. Today 122 (2007) 391-396. 96. P. Papaefthimiou, T. Ioannides, and X.E. Ve rykios, Appl. Catal. B 13 (1997) 175-184. 97. E.M. Cordi and J.L. Falconer, J. Catal. 162 (1996) 104-117. 98. M. Besson, P. Gallezot, Ca tal. Today 57 (2000) 127-141. 99. C. A. Pez, N. J. Castellanos, F. Martnez O ., F. Ziarelli, G. Agrifoglio, E A. Pez-Mozo, and H. Arzoumanian, Catal. Today 620 133 (2008) 619. 138

PAGE 139

100. C.D. Wagner, W.M., Riggs, L.E Davis, J. F. Moulder and G.E. Muilenberg, Eds., Handbook of X-ray Photoelectron Spectroscopy, Pe rkin-Elmer Corp.: Eden Prairie, MN, (1979). 101. J.C. Rivire, S. Myhra, Eds., Handbook of Surface and Interface Analysis, Marcel Dekker: New York, (1998), 57-158. 102. National Institute of Standards and T echnology, NIST on-Line XPS Database http://srdata.nist.gov/xps/intro.aspx [Accessed 10/28/08] 103. M. Brun, A. Berthet, J.C. Bertolini, J. Electron Spectrosc. Relat. Phenom. 104 (1999) 55 104. K. T. Jung, A. T. Bell, Catal. Lett. 80(2002)63-68 105. M.D. Rhodes, A.T. Bell, J. of Catal. 233 (2005) 198. 106. 4,4-dimethyl 2,2 bipyridine http://www.sigmaaldrich.com/catalog/s earch/ProductDetail/ ALDRICH/245739 accessed 09/19/08 107. Raney nickel, Fluka # 8344 bipyridine http://www.sigmaaldrich.com/MSDS accessed 09/19/08 139

PAGE 140

BIOGRAPHICAL SKETCH Luke Michael Neal was born in Peoria, Illinoi s, to George and Sue Neal in 1981. He was raised in the town of Morton, Illinois where he attended Morton Community High School. After graduating high school in 1999 he attended the University of Illinois at Urbana-Champaign, where he pursued a bachelors degree in ch emical engineering. During his undergraduate education, he worked summers at the USDA National Center for Agriculture Utilization Research (NCAUR) in Peoria, Illinoi s. While there, he assisted in research into the purification of soy hull proteins and the use of the corn-eth anol byproduct, zien as a bioplastic. After the graduating from the University of Illinois in 200 3, he continued to work as a science aide for NCAUR while advancing his education with coursework at Bradley University in Peoria. In fall 2004, he came to Gainesville, Florida, to pursue a doctorate of philosophy in chemical engineering at the University of Florid a. His research focused on use of nanoparticle oxides in catalyst. Much of this work was devoted to nanopa rticle supported palladium oxide for use as an oxidative coupling cataly st. Additionally he contributed to research into nanoparticle nickel catalystss for biomass to hydrogen conversion and the use of nanoparticle iron and cobalt oxides for the catalytic decomposition of carbon monoxide into carbon nanotubes. 140