University of Southern California Dissertations and Theses

Chemisorption studies of hydrocarbons on clean and hydrogen precovered platinum(111) and tin/platinum(111) surface alloys

(USC Thesis Other)

Chemisorption studies of hydrocarbons on clean and hydrogen precovered platinum(111) and tin/platinum(111) surface alloys

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content NOTE TO USERS This reproduction is the best copy available. ® UMI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHEMISORPTION STUDIES OF HYDROCARBONS ON CLEAN AND HYDROGEN PRECOVERED Pt(l 11) & Sn/Pt(l 11) SURFACE ALLOYS by Haibo Zhao A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (CHEMISTRY) December 2004 Copyright 2004 Haibo Zhao Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3155501 INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 3155501 Copyright 2005 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements First of all, I would like to express my heartiest appreciation to my research advisor, Prof. Bruce E. Koel, for his support and guidance in the path of accomplishing my Ph.D. degree. Without him, it would not have been impossible to complete this thesis. His excellence, enthusiasm and encouragement will always be a source o f inspiration to me in both my professional and personal endeavors. I am specially thankful to Dr. Jooho Kim who initially trained me in surface science experiments and Dr. Michael Quilan who gave me a lot of advices on the vacuum technologies. Thanks also go to the helpful staff in the machine shop (Victor Jordan and Don Wiggins), the glassblowing shop (Jim Merritt), and the electronics shop (Frank Niertit and Ross Lewis) for their help in constructing and maintaining numerous devices. I would also like to express my appreciation to the former and current members o f the Koel group, including Dr. Shuchen Hsieh, Dr. Dmitri Jerdev, Dr. Chih-Sung Ho, Dr. David Beck, and Dr. Adrian Hightower. I would like to thank the scientists at PNNL, including Dr. Charles H. F. Peden, Dr. Russell G. Tonkyn, and Dr.Stephan E Barlow for their advice and discussion in the summer of 2003. I also want to thank committee members, Hanna Reisler and Edwards Goo who examined my thesis and provided thoughtful comments and directions. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Finally, I am very indebted to my loving wife Wendy and my loving daughter Karen. Wendy is always supportive and understanding through my life as a graduate student in USC. It is impossible for me to finish my Ph.D. degree without her support and encouragement. The funding by the Division of Chemical Sciences, Office of Basic Energy Sciences, US Department of Energy is gratefully acknowledged. m Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents Acknowledgements............................................................................................................ ii List of Figures....................................................................................................................viii List of Equations................................................................................................................ xii Abstract...............................................................................................................................xiii Chapter 1 Introduction....................................................................................................... 1 1.1 Adsorption and reaction of hydrocarbons on P t( lll) and Sn/Pt(l 11) alloys........................................................................................................... 1 1.2 Influence of preadsorbed H adatoms on adsorption and reaction of ethylene and butadiene on Pt(l 11) and Sn/Pt(l 11) alloys...................4 1.3 References....................................................................................................8 Chapter 2 Experimental Methods.....................................................................................11 2.1 Ultrahigh Vacuum Chamber and Sample Handling.............................. 11 2.2 Cleaning of Pt( 111) and preparation of Sn/Pt( 111) surface alloys.... 15 2.3 Hydrogen Atom Doser...............................................................................20 2.4 References.................................................................................................. 31 Chapter 3 Adsorption and reaction of bicyclic hydrocarbons at P t( lll) and S n /P t(lll) surface alloys: trans-decahydronaphthalene (CioHig) and bicyclohexane (C12H 22) .................................................................................. 32 3.1 Introduction................................................................................................ 34 3.2 Experimental M ethods..............................................................................37 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.3 R esults.........................................................................................................39 3.4 Discussion.................................................................................................. 58 3.5 Conclusions................................................................................................ 62 3.6 References...................................................................................................64 Chapter 4 Adsorption and reaction of 1,3-butadiene on P t(lll) and S n/P t(lll) surface alloys 4.1 Introduction................................................................................................ 68 4.2 Experimental M ethods..............................................................................70 4.3 R esults......................................................................................................... 72 4.4 Discussion.................................................................................................. 79 4.5 Conclusions.................................................................................................82 4.6 Appendix 1...................................................................................................83 4.7 References...................................................................................................85 Chapter 5 Reactivity of Chemisorbed Ethyl Groups on a (V3xV3)R30°-Sn/Pt(l 11) Surface Alloy....................................................................................................88 5.1 Introduction................................................................................................ 89 5.2 Experimental M ethods..............................................................................91 5.3 R esults......................................................................................................... 93 5.4 Discussion...................................................................................................99 5.5 Conclusions...............................................................................................103 5.6 References.................................................................................................105 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 Site-blocking effects of preadsorbed H on P t(lll) probed by 1,3-butadiene adsorption and reaction............................................................................... 109 6.1 Introduction.............................................................................................. 109 6.2 Experimental M ethods........................................................................... 113 6.3 R esults....................................................................................................... 115 6.4 D iscussion.................................................................................................121 6.5 Conclusions.............................................................................................. 128 6 .6 References.................................................................................................130 Chapter 7 Hydrogenation of 1,3-butadiene on two ordered Sn/Pt( 111) surface alloys................................................................................................................133 7.1 Introduction...............................................................................................135 7.2 Experimental M ethods............................................................................136 7.3 R esults....................................................................................................... 139 7.4 Discussion.................................................................................................155 7.5 Conclusions...............................................................................................161 7.6 References.................................................................................................164 Chapter 8 Influence of coadsorbed hydrogen on ethylene adsorption and reaction on a (V3xV3)R30°-Sn/Pt(lll) surface alloy...................................................166 8.1 Introduction...............................................................................................167 8.2 Experimental M ethods............................................................................169 8.3 R esults........................................................................................................171 v i Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8.4 Discussion................................................................................................ 180 8.5 Conclusions.............................................................................................. 183 8.6 References................................................................................................ 184 Chapter 9. Proposed future w o rk ....................................................................................187 9.1 Probing the C-H bond dissociation barrier........................................... 187 9.2 Synthesis and characterization of new surface intermediate species prepared by H-atom addition reactions.................................................188 9.3 Hydrogenation of C=C and C =0 double bonds under UHV conditions..................................................................................................190 9.4 Hydrogenation of butadiene over S n /P t(lll) alloys at high pressure.....................................................................................................191 9.5 References..................................................................................................192 Bibliography...................................................................................................................... 194 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Figure 2.1.1 Overall view from the front of the HREELS ultrahigh vacuum chamber................................................................................................................................. 12 Figure 2.1.2 Top views of the HREELS ultrahigh vacuum chamber showing cross- sections of the three main analysis levels.........................................................................13 Figure 2.1.3 Schematic drawing of the mounting of the Pt(l 11) crystal......................14 Figure 2.2.1 AES uptake plot for Sn deposited on Pt(l 11) at 100 K ........................... 16 Figure 2.2.2 AES spectra of clean P t( lll) and the (2x2)-Sn/Pt(l 11) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys...........................................................................17 Figure 2.2.3 LEED pictures taken at for an incident beam energy of 120 eV: (a) P t(lll); (b) (2 x2 )-S n /P t(lll) alloy; (c) (V3xV3)R30°-Sn/Pt(ll 1) alloy; (d) mixed alloy.......................................................................................................................................18 Figure 2.2.4 ^-butane adsorption and desorption provides a chemical method to characterize these surface alloys, n-butane TPD is shown after a monolayer saturation coverage of «-butane was adsorbed on the surface at 100 K .......................19 Figure 2.3.1 Schematic drawing of the platinum-tube hydrogen atom source........... 21 Figure 2.3.2a Side view of the Pt-tube H-atom doser....................................................22 Figure 2.3.2b Flange-view of the H-atom doser..............................................................22 Figure 2.3.2c Front view of the H-atom doser (Cu block only)................................... 23 Figure 2.3.2d Side view of the Cu block and Pt tube of the H-atom doser.................24 Figure 2.3.3 H coverages observed for a fixed exposure on the V 3 alloy as a function of tbe H-atom source temperature.................................................................................... 26 Figure 2.3.4 H2 TPD spectra after H atoms exposure on the (2x2) alloy at 100 K .. .28 Figure 2.3.5 H2 TPD spectra after H atoms exposure on the V3 alloy at 100 K .........29 viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.3.6 Water contamination resulting from several exposures using the H-atom doser. The bottom curves provides an H2O TPD spectrum after a H20 exposure that can be used to calibrate the amount of H20 codeposited during H-atom dosing 30 Figure 3.1 CioHig TPD spectra after trans-decahydronaphthalene (Ci0H |8 ) exposures on P t( lll) at 110 K ............................................................................................................. 41 Figure 3.2 H2 TPD spectra after trans-decahydronaphthalene (CioHig) exposures on P t( lll) at 110 K ...................................................................................................................42 Figure 3.3 AES spectra of clean P t( lll) (top), monolayer CioHig-covered P t( lll) (bottom), and the P t( lll) surface following a 0.6-L CioHig TPD experiment (middle).................................................................................................................................44 Figure 3.4 CioHig TPD spectra after trans-decahydronaphthalene (CioHig) exposures on a (2x2)Sn/Pt(l 11) surface alloy at 110 K ...................................................................45 Figure 3.5 CiqH| 8 TPD spectra after trans-decahydronaphthalene (CioHig) exposures on a (V3xV3)R30°Sn/Pt(l 11) surface alloy at 110 K ...................................................46 Figure 3.6 Ci2H22 TPD spectra after bicyclohexane (Ci2H2 2 ) exposures on Pt(l 11) at 110 K .................................................................................................................................... 48 Figure 3.7 H2 TPD spectra after bicyclohexane (Ci2H2 2 ) exposures on Pt(l 11) at 110 K ............................................................................................................................................50 Figure 3.8 AES spectra of clean P t( lll) (top), monolayer Ci2H22-covered P t( lll) (bottom), and the P t( lll) surface following a 1.2-L Ci2H22 TPD experiment (middle)................................................................................................................................. 51 Figure 3.9 Ci2H22 TPD spectra after bicyclohexane (Ci2H2 2 ) exposures on a (2x2)Sn/Pt(l 11) surface alloy at 110 K ............................................................................52 Figure 3.10 H2 TPD spectra after bicyclohexane (Ci2H2 2 ) exposures on a (2x2)Sn/Pt(l 11) surface alloy at 110 K ............................................................................54 Figure 3.11 Ci2H22 TPD spectra after bicyclohexane (Ci2H22) exposures on a (V3xV3)R30°Sn/Pt(l 11) surface alloy at 1 1 0 ...............................................................55 Figure 3.12 Influence of alloyed Sn on CioHig and Ci2H22 adsorption and reaction on Pt( 111) and two Sn/Pt(l 11) surface alloys..................................................................... 57 ix Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.1 Butadiene (C4H6) TPD spectra after 1,3-butadiene exposures on Pt(l 11) at 100 K ................................................................................................................................... 73 Figure 4.2 FF TPD spectra after 1,3-butadiene exposures on Pt(l 11) at 100 K 74 Figure 4.3 Butadiene (C4H6) TPD spectra after 1,3-butadiene exposures on the (2x2)- Sn/Pt(l 11) alloy at 100 K .................................................................................................. 76 Figure 4.4 Butadiene (C4FI6) TPD spectra after 1,3-butadiene exposures on the (V3xV3)R30°-Sn/Pt(l 11) alloy at 100 K ......................................................................... 77 Figure 4.5 Comparison of H2 desorption spectra resulting from 1,3-butadiene decomposition on P t( lll) and the (2x2)-Sn/Pt(lll) and (V3xV3)R30°-Sn/Pt(l 11) alloys..................................................................................................................................... 78 Figure 5.1 Ethylene (C2FI4) TPD spectra after increasing gas-phase H-atom exposures on a chemisorbed ethylene monolayer on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy......................................................................................................................... 95 Figure 5.2 Ethane (C2H6) TPD spectra after increasing gas-phase H-atom exposures on a chemisorbed ethylene monolayer on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy...................................................................................................................................... 96 Figure 5.3 H2 TPD spectra after increasing gas-phase H-atom exposures on a chemisorbed ethylene monolayer on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy 98 Figure 6.1 C4H6 TPD spectra after 0.24-L 1,3-butadiene exposures on H preadsorbed Pt(l 11) at 100 K .................................................................................................................117 Figure 6.2 H 2 TPD spectra after 0.24-L 1,3-butadiene exposures on H preadsorbed P t( lll) at 100 K. (dash curve is the H2 TPD spectra after 0.057-L 1,3-butadiene exposures on Pt(l 11) at 100 K ).......................................................................................118 Figure 6.3 AES spectra on P t( lll) following 0.24-L C4H6 TPD experiment on H preadsorbed Pt(l 11).......................................................................................................... 119 Figure 6.4 The influence of preadsorbed H to the C4H6 saturation coverage in the chemisorption layer.......................................................................................................... 121 Figure 6.5 Side view of 1,3-butadiene (a), ethylene (b) and benzene (c) adsorption on Pt( 111) and H preadsorbed Pt( 111)................................................................................ 126 x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.1 1,3-butadiene, C4H 6 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (2x2)Sn/Pt(lll) alloy at 100 K ....................................... 140 Figure 7.2 Butene, C4H8 TPD spectra after after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (2x2)Sn/Pt(lll) alloy at 100 K ....................................... 142 Figure 7.3 H2 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H- precovered (2x2)Sn/Pt(lll) alloy at 100 K ................................................................... 144 Figure 7.4 1,3-butadiene, C4H6 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (V3xV3)R30°Sn/Pt(lll) alloy at 100 K ........................,146 Figure 7.5 Butene, C4H8 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (V3xV3)R30°Sn/Pt(lll) alloy at 100 K ............................... 147 Figure 7.6 H2 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H- precovered (V3xV3)R30°Sn/Pt(lll) alloy at 100 K ..................................................... 148 Figure 7.7 Influence of H-adatom precoverage on the amount of 1,3-butadiene adsorption, desorption, and hydrogenation on the (2x2)Sn/Pt(lll) alloy................. 152 Figure 7.8 Influence of H-adatom precoverage on the amount of 1,3-butadiene adsorption, desorption, and hydrogenation on the (V3xV3)R30°Sn/Pt(l 11) alloy... 155 Figure 8.1 H2 (2 amu), ethylene (28 amu), and ethane (30 amu) TPD spectra following 0.24-L ethylene exposures on a (V3xV3)R30° S n /P t(lll).........................173 O Figure 8.2 H2 TPD spectra following 0.24-L ethylene exposures on a (V3xV3)R30 S n /P t(lll) surface alloy at 100 K that has been precovered with varying amounts of H adatoms...........................................................................................................................174 Figure 8.3 Ethylene TPD spectra following 0.24-L ethylene exposures on a (V3xV3)R30° S n/P t(lll) surface alloy at 100 K that has been precovered with varying amounts of H adatoms........................................................................................175 Figure 8.4 Influence of preadsorbed H adatoms on the monolayer (saturation) coverage of ethylene on a (V3xV3)R30° Sn/Pt(l 11) surface alloy at 100 K 177 Figure 8.5 AES spectra obtained following adsorption of a monolayer (saturation) coverage of ethylene on clean P t(lll) (bottom) and a 0.6-ML H precovered P t(lll) surface (top)....................................................................................................................... 179 xi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Equations 53912 — — ^ + 0.15 (4.1) ------------ ™ 2 + 0.15 0.24 0.18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract Platinum and bimetallic Pt-Sn catalysts are important in commercial hydrocarbon processing applications. These catalysts are complex materials systems, and a surface science approach to understanding aspects of this catalysis involves using well-defined model systems. Single-crystal P t( lll) and ordered S n /P t(lll) alloy substrates offer outstanding new opportunities for fundamental surface science and catalytic studies at the molecular level. The adsorption and reaction of a number of interesting hydrocarbon molecules was investigated on three substrates, i.e., P t( lll) and the (2 x2)-S n /P t(lll) and (V3xV3)R30°-Sn/Pt(lll) surface alloys, by using temperature programmed desorption (TPD) mass spectroscopy and Auger electron spectroscopy (AES). These surfaces were characterized by low energy electron diffraction (LEED), AES, and a chemical method involving w-butane TPD. In addition, an H-atom doser was constructed to cleanly produce gas-phase atomic hydrogen in order to probe new coadsorption reactions on these surfaces. Adsorption and desorption of trans-decahydronaphthalene (CioHig), bicyclohexane (C 12H22) and 1,3-butadiene (C4H6) were used to probe activation energies for C-H bond cleavage at Pt-Sn alloys. Improved estimates of the activation energy barriers to break aliphatic C-H bonds and vinylic C-H bonds on Sn/Pt alloys were obtained. Ethyl intermediates were synthesized by reaction of H atoms with adsorbed ethylene on a (V3xV3)R30°-Sn/Pt(l 11) alloy, and the activation energy xiii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. barrier for ethyl dehydrogenation was determined. Also, a lower limit was placed on the ethyl hydrogenation barrier on this alloy. One particular catalytic application of interest is to selectively remove butadiene in C4 alkene streams produced by steam cracking. Additional basic information on such chemistry is important for development of improved heterogeneous catalysts. An investigation was made of the influence of hydrogen coadsorption on 1,3-butadiene chemistry on P t( lll) and S n /P t(lll) alloys, and this is a key aspect to an improved understanding of catalytic selective hydrogenation on Pt and PtSn bimetallic catalysts. A strong site blocking effect of preadsorbed hydrogen adatoms was observed on P t( lll) and the two Sn/Pt(l 11) alloys. Importantly, less surface hydrogen was needed to fully block chemisorption of 1,3- butadiene on the alloys, and this decreased with increasing surface Sn concentration. Alloyed Sn opened a new hydrogenation reaction pathway, compared to P t(lll), such that hydrogenation of 1,3-butadiene was observed in TPD experiments. This occured with 100% selectivity to liberate butene (C4H8) in TPD and no deeper hydrogenation product (butane) was observed. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1. Introduction This dissertation covers two research topics: (1) Adsorption and reaction of two bicyclic hydrocarbons (trans-decahydronaphthalene (CioHig) and bicyclohexane (C12H22)), butadiene (C4H6), and surface-bound ethyl groups (C2H 5) on P t( lll) and two ordered, S n /P t(lll) surface alloys; and (2) The influence of preadsorbed H adatoms on adsorption and reaction of ethylene (C2H4) and butadiene on Pt(l 11) and two ordered, Sn/Pt(l 11) surface alloys. 1.1 Adsorption and reaction of hydrocarbons on P t(lll) and S n /P t(lll) alloys Platinum catalysts are very important in hydrocarbon conversion reactions involving making and breaking C-H and C-C bonds in hydrocarbon molecules and thus have been of great interest to scientists and engineers in academia and industry [1], Overshadowing supported Pt, bimetallic Pt catalysts, such as Pt-Sn, are also of great practical importance [2, 3]. Adding Sn to the catalyst usually increases both the selectivity of the catalyst and the resistance toward poisoning by coke (carbonaceous deposits). Catalytically inactive Sn has often been considered to dilute contiguous Pt surface sites to create smaller ensembles that are more favorable for nondegradative reactions and weaken the C-H and C-C bond breaking activity of Pt catalysts for reforming reactions. Sn modification of the catalytic properties of Pt has been investigated by many researchers [4-10]. We do not yet have a molecular level 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. understanding of the origin of the improved performance of these bimetallic supported catalysts. This is enormously difficult because of the great number of different phases and structures present on the catalyst, such as monometallic clusters, alloy clusters, metal particle-edge decoration, and variation of such features with dispersion, or particle size and shape. Even with a detailed knowledge of all of the reactive sites present in the catalyst, one needs to know how this is affected by the operating conditions of the reaction (pressure, temperature, oxidizing or reducing atmosphere) and to make detailed structure-reactivity correlations to identify the important active sites. It would seem that the only possible solution to this challenging problem is to take a surface science approach and work with well- defined model systems to establish a foundation for deeper understanding. Ordered Sn/Pt surface alloys, which were first reported by Paffett and Windham [11] after vapor deposition of Sn onto a P t( lll) single crystal, offer outstanding new opportunities for fundamental surface science and catalytic studies at a molecular level. Such substrates have been extensively used by Koel and his group to show that alloyed Sn strongly decreases the reactivity of many hydrocarbons and organic molecules on P t( lll) surfaces in surface science studies under UHV conditions [12- 21]. The values of activation energy barriers to C-H bond cleavage for alkanes, alkenes, and other hydrocarbons on Pt-Sn alloys remains a key fundamental question. This problem arises in part because alkanes (saturated hydrocarbons) 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interact weakly when molecularly adsorbed on low-Miller index Pt surfaces, and even more weakly on Pt-Sn alloys. Currently, the desorption activation energies (40- 58 kJ/mol) can be used to place a lower limit on the activation energy barrier E* to break primary C-H bonds in alkanes and cycloalkanes on Sn/Pt alloys. However, these numbers are likely to be much too small. In order to determine the value of E*, or at least place a more realistic bound on it, we need to use larger, more strongly adsorbed hydrocarbon molecules that will remain adsorbed at the alloy surface until temperatures are reached where decomposition occurs. Trans-decahydronaphthalene (CioHig) and bicyclohexane (C 12H22) gives us a better chance to observe C-H bond dissociation due to the increased desorption activation energy. CioHig and C 12H22 are of interest in their own right, because they are components of a series of diesel and jet fuel blends made from catalytic reforming processes [2 2 ], Experiments using 1,3-butadiene (C4H6) [23-25] on P t( lll) have also been carried out in part due to the industrial importance of partial hydrogenation of dienes and alkynes [26, 27]. Improved fundamental understanding of adsorption and reaction of 1,3-butadiene on well-defined Pt-Sn alloys would provide important information for understanding and improving Pt-Sn catalysts for use in diene hydrogenation. 1,3-butadiene contains two C=C double bonds, and thus is expected to have a much stronger interaction with these alloys. Stronger adsorption often leads to higher reactivity and so this is an interesting molecule to probe C-H bond cleavage barriers on these alloy surfaces. 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Simple alkyl species have been considered for a long time to be intermediates in the conversion of saturated hydrocarbons into more economically attractive products over transition metal catalysts [28, 29]. However, our knowledge of the chemistry of these species from surface science type experiments does not reflect their importance. The production of simple species such as ethyl (C2H5) from their fully saturated precursors is constrained heavily under UHV conditions by the significant energetic barriers to dissociative alkane adsorption. Perhaps the most versatile approach is one pioneered by Bent and coworkers [30-32] in which incident gas-phase H atoms undergo radical addition to the 7 r-bond in weakly adsorbed ethylene to form ethyl groups. Subsequent addition from gas-phase H atoms directly to the surface-bound alkyl is orders of magnitude slower (and any ethane formed would desorb immediately) and so the reaction is self-limiting and clean. Ethylene chemisorbs on most other metal surfaces retaining more double-bond character on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy and radical addition may occur with a useful, higher probability. By using this method, we studied the thermal chemistry of chemisorbed ethyl groups on this alloy surface without perturbation by other coadsorbed species. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.2 Influence of preadsorbed H adatoms on adsorption and reaction of ethylene and butadiene on P t (lll) and S n /P t(lll) alloys. Butadiene is an impurity in C4 alkene streams produced by steam cracking and it is desirable to remove this product from butadiene-butene mixtures. An ideal process would be to selectively convert butadiene to butene (C4H8) and avoid forming the completely hydrogenated product, i.e., butane (C4H 10). Hydrogenation of 1,3-butadiene over platinum catalysts produces a mixture of 7 7-butane, 1-butene and 2-butene [33-36]. Methods to study adsorbed layers and identify surface intermediates during hydrogenation at high pressures on metal single crystals are limited. One work-around has been to investigate coadsorption of hydrogen and unsaturated hydrocarbon molecules under UHV conditions in a surface science approach to study hydrogenation at the molecular level. The supported Pt-Sn catalyst system is very effective in selectively hydrogenating diolefin impurities [37]. Studies show that the presence of Sn in the catalysts prevents hydrogenation of alkenes [38-39], which is important in selectively removing diene impurities. Alloying Sn to Pt(111) reduced the interaction between 1,3-butadiene and Sn/Pt(l 11) surface alloys, and this may result in different hydrogenation reaction pathways in the presence of preadsorbed hydrogen. Lutterloh et al. [40], using a pyrolytic H atom source, recently found that chemisorption of benzene was completely blocked by high coverages of preadsorbed H adatoms on P t( lll) at 125 K. Because the absolute coverages in adsorbed 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hydrocarbon monolayers on P t( lll) can be quite low, 0h c = 0.25 - 0.1 ML, it may be important to study the influence of coadsorbed hydrogen at very high coverages otherwise islanding or ensemble effects can allow coadsorbed hydrocarbon molecules to simply avoid interacting directly with the coadsorbed H adatoms. In addition, the role of site-blocking effects by preadsorbed H should become more important at high values of Oh than for low ones. The hydrogenation of ethylene (C2H4) to ethane (C2H6) over Pt( 111) has been studied by a number of investigators as a model system for understanding the surface science of catalytic hydrogenation reactions [41-46], Somorjai and co-workers [47] reported that the relative hydrogenation rates of species adsorbed on P t( lll) at 295 K during ethylene hydrogenation at 1 atm are in the order of 7 1-bonded ethylene » di-a-bonded ethylene » ethylidyne. Alloying with Sn weakens the interaction between ethylene and Pt at the surface, and ethylene is less strongly chemisorbed and less rehybridized from the gas phase on both S n /P t(lll) surface alloys. One would thus predict that a higher relative hydrogenation rate may be obtained on Pt-Sn alloys due to a decreased sp3 rehybridization of chemisorbed ethylene and a weaker Pt-H bond on Pt-Sn alloys [48] which should create a more labile H adatom. In contrast to this prediction, the hydrogenation activity for ethylene is inhibited on Pt-Sn/AI2O3 catalysts [39]. In order to better understand this result, we have investigated the coadsorption of hydrogen and ethylene on the surface of a well-defined Sn-Pt alloy. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.3 References [1] J.H. Sinfelt, Catalysis: Science and Technology, J.R. Anderson and M. Boudart, Eds., (Springer-Verlag, Berlin, 1981), vol. 1, Ch. 5, p. 258. [2] J.M. Parera and N.S. Figoli, in: Catalytic Naphtha Reforming, eds.G.J. Antos, A.M. Aitani and J.M. Parera (Dekker, New York, 1995)p. 45. [3] V. Ponec and G.C. Bond, Catalysis by Metals and Alloys, Stud. Surf. Sci. Catal., Vol. 95 (Elsevier, Amsterdam, 1995). [4] J. VOlter, G. Lietz, M. Uhlemann and M. Hermann, J. Catal. 68 (1981) 42. [5] R. Burch and L. C. Garla. J. Catal. 71 (1981) 360. [6] K. Balakrishnan and J. Schwank. J. Catal. 127 (1991) 287. [7] H. Lieske and J. Volter. J. Catal. 90 (1984) 96. [8] Y. X. Li, J. M. Stencel and B. H. Davis. Appl. Catal. 64 (1990) 71. [9] F. M. Dautzenberg, J. N. Helle, P. Biloen and W. M. H. Sachtler. J. Catal. 63 (1980) 119. [10] T. Fujikawa, F. H. Ribeiro, and G. A. Somorjai, J. Catal. 178 (1998) 58. [11] M. T. Paffett and R. G. Windham, Surf. Sci. 208 (1989) 34. [12] C. Xu, B. E Koel, and M. T. Paffett, Langmuir 10 (1994) 166. [13] C. Xu, Y. Tsai, and B. E. Koel, J. Phy. Chem. 98 (1994) 585. [14] Y. Tsai and B. E Koel, Langmuir 14 (1998) 1290. [15] J.W. Peck, D.I. Mahon, and B.E. Koel, Surf. Sci. 410 (1998) 200. 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [16] J.W. Peck, D.I. Mahon, D.E. Beck and B.E. Koel, Surf. Sci. 410 (1998) 170. [17] C. Panja, N. Saliba, and B.E. Koel, Surf. Sci. 395 (1998) 248. [18] Y. Tsai, C. Xu, and B. E. Koel, Surf. Sci. 385 (1997) 37. [19] Y. Tsai and B. E. Koel, J. Phy. Chem. B 10 (1997) 2895. [20] J.W. Peck and B.E. Koel, J. Am. Chem. Soc. 99 (1995) 16670. [21] C. Xu and B.E. Koel, Surf Sci. 304 (1994) 249. [22] N. White PCTInt. appl. (1985) 56. [23] J.C. Bertolini, A. Cassuto, Y. Jugnet, J. Massardier, B. Tardy, and G. Tourillon, Surf. Sci. 349 (1996) 88. [24] G. Tourillon, A. Cassuto, Y. Jugnet, J. Massardier, J.C. Bertolini, J. Chem. Soc., Faraday Tran. 92 (1996) 92. [25] N.R. Avery, N. Sheppard, Proc. R. Soc. Lond. A 405 (1986) 1. [26] J.P. Boitiaux, P. Cosyns, M. Derrien and G. Leger. Hydrocarbon Proc. 64 (1985) 51. [27] M. Derrien. In: L. Cerny, Editor, Studies in Surface Science and Catalysis 27, Elsevier, Amsterdam (1986) p. 313. [28] G. C. Bond, Heterogeneous Catalysis, 2n d ed. (Clarendon, Oxford, 1987), chaps. 8 and 9. [29] G. A. Somorjai, Introduction to Surface Chemistry and Catalysis (Wiley, New York, 1994), Chap.6 and 7. [30] M. Xi and B.E. Bent, J. Vac. Sci. Technol. B 10 (1992) 2440. 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [31] C. J. Jenks, M. Xi, M.X. Yang, and B.E. Bent, J. Phys Chem. 98 (1994) 2152. [32] M.X. Yang and B.E. Bent, J. Phys Chem. 100 (1996) 822. [33] G.C. Bond, G. Webb, P.B. Wells and J.M. Winterbottom, J. Chem. Soc. A, (1965)3128. [34] P.B Wells and A.J. Bates, J. Chem. Soc. A, (1968) 3064. [35] C.M. Pradier, E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 43 (1988) 177. [36] C. Yoon, M.X. Yang, and G.A. Somorjai, Catal. Lett. 46 (1997) 37. [37] D.F. Rohr, D.M. Haskell, F.M. Brinkmeyer, Eur. Pat. Appl. EP 211381 A1 19870225 (1987). [38] M. Galvagno, P. Staiti, P. Antonucci, A. Donato, and R. Pietropaolo, J. Chem. Soc, Faraday Trans. 79 (1983) 2605. [39] A. Palazov, Ch. Bonev, D. Shopov, G. Lietz, A. Sarkany and J. Yolter, J. Catal. 103 (1987) 249. [40] C. Lutterloh, J. Biener, K. Pohlmann, A. Schenk, and J. Kiippers, Surf. Sci. 352(1996) 133. [41] A. I. Boronin, V. I. Bukhtiyarov, R. Kvon, Y. V. Chesnokov, R. A.Buyanov, Surf. Sci. 258 (1991)289. [42] A. Casuto, M. Mane, J. Jupille, Surf. Sci. 249 (1990) 8. [43] D. Godbey, F. Zaera, R. Yeates, G. A. Somorjai, Surf. Sci. 167 (1986) 150. 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [44] T. S. Marinova, D. V. Chakarov, Surf. Sci. 192 (1987) 275. [45] G. A. Somorjai, Z. Phys. Chem. (Munich) 197 (1996) 1. [46] F. Zaera, T. V. W. Janssens, H. Ofner, Surf. Sci. 368 (1996) 371. [47] P. S. Cremer, X. Su, Y. R. Shen, G. A. Somorjai, J. Am. Chem. Soc. 118 (1996) 2942. [48] M. R. Voss, H. Busse, B. E. Koel, Surf. Sci. 414 (1998) 330. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2 Experimental Methods 2.1 Ultrahigh Vacuum Chamber and Sample Handling All experiments were conducted within a custom-designed, ultrahigh vacuum (UHV) chamber made of ANSI Type 316 stainless steel (fabricated by MDC). This chamber is shown schematically in Figure 2.1.1 The chamber was pumped by a 400 1/s ion pump (Perkin-Elmer, Model 2070420) along with a 240 1/s turbomolecular pump (Pfeiffer-Balzers, Model TPU 240) backed with a dual stage, oil-driven, rotary vane mechanical pump (Varian IPO 301). A Ti sublimation pump with cryopanel (Perkin-Elmer, Model 2140411) was also used occasionally to improve the vacuum. The chamber and associated electronics racks were connected at a single point to earth ground. The chamber has three analysis levels, as shown in Figure 2.1.2. The first level was equipped with a double-pass cylindrical mirror analyzer (CMA) (Perkin- Elmer) that was used for AES, X-ray and UV photoelectron spectroscopy (XPS and UPS). The ion gun was also mounted on this level for cleaning the sample by Ar+ - ion sputtering. The second level was equipped with optics for LEED and a quadrupole mass spectrometer (QMS) that was used for TPD. The tin doser was mounted on a flange located between the first and second levels. The third level contained a LK2000 spectrometer (LK Technologies) for high-resolution electron energy loss spectroscopy (HREELS) studies. 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CMA QMS [ HREELS Ion Pump Turbo Pump Figure 2.1.1 Overall view from the front of the HREELS ultrahigh vacuum chamber 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CMA Level 1 LEED QMS HREELS Level 2 ] Level 3 Figure 2.1.2 Top views of the HREELS ultrahigh vacuum chamber showing cross- sections of the three main analysis levels. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A dual-stage sample manipulator (Vacuum Generators, U.K.) was installed for crystal mounting and manipulation, and provided sample translation in three dimensions along with sample rotation. The sample holder was connected to a long stainless steel tube, which could be filled with liquid nitrogen (LN2) for effective cooling of the sample. A Pt(l 11) crystal (Atomergic; 10-mm dia., 1.5-mm thick) was used as the sample and mounted on tantalum rods (0.126-in) that were imbedded in LN2-cooled Cu blocks at the bottom of the manipulator, as shown in figure 2.1.3. Connected to manipulater Liquid N2 Reservoir Cu Blocks " A 7 therm ocouple Wires Sapphire Cr-AI 0.015" Ta Wire Pt(l 11) / 0.126" Ta Rod Figure 2.1.3 Schematic drawing of the mounting of the Pt(l 11) crystal. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tantalum wire (0.015-in) was spot-welded to the top and bottom edges of the crystal and spot-welded to Ta rods of the sample holder in order to provide for resistive heating of the sample by passing a DC current. A chromel-alumel (Cr-Al, type K) thermocouple (0.005-in) was spot-welded on the side of the crystal. The sample could be quickly cooled to 90 K by LN2 or resistively heated to 1200 K. 2.2 Cleaning of P t (lll) and preparation of S n /P t(lll) surface alloys The P t( lll) crystal was cleaned by Ar+ -ion sputtering (1-keV beam voltage and 30-mA emission current gave a sample current of 10 pA at PA r = 5x1 O ' 5 torr) for 5 minutes at 500 K and then annealing at 1200 K for 60 sec in UHV. In order to remove residual surface carbon, the crystal was exposed to 5 xl0'7 -torr O2 at 900 K for 2 minutes followed by flashing the sample to 1200 K for a few seconds in UHV. This procedure gave a clean spectrum using AES and a sharp (lx l) pattern in LEED. The Sn doser was made by using an enclosed boat constructed from 0.13-mm Ta foil containing a small amount of Sn foil (6N purity). Prior to deposition, the doser was outgassed by passing a current of 13.5 A through the boat. This was continued until the pressure read on the ion gauge returned into the 10"'°-torr range. AES was used to determine the Sn coverage in the first several Sn deposition experiments. All of the Sn deposition experiments were done with the temperature of P t( lll) crystal held at 100 K. At a constant heating current of 13.5 A, the Pt(237) 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and Sn(430) peak-to-peak heights in AES varied with the Sn deposition time as shown in Figure 2.2.1. A break in the slopes of these curves occurs after a 120-sec exposure and this was taken to signal the completion of a Sn monolayer. At that point, the Sn(430)/Pt(237) AES ratio was equal to 3.4. 1.0 0.8 c o c CD c o 0.6 co L LI < T3 C D n . N 0.4 z 0.2 0.0 Figure 2.2.1 AES uptake plot for Sn deposited on Pt(l 11) at 100 K. The (2 x2 )-Sn/Pt(l 11) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys were prepared by evaporating one monolayer of Sn (13.5 A for 120 sec) onto the P t( lll) crystal 16 1 M L Sn Sn(430) Pt(237) Sn(430)/Pt(237) = 3.36 100 150 200 250 300 50 0 Sn exposure time (sec) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. surface and subsequently annealing the sample for 20 s to 1000 and 830 K, respectively. AES and LEED are used to characterize the two S n /P t(lll) surface alloys after annealing. AES spectra from P t( lll) and the (2x2)-Sn/Pt(l 11) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys are shown in Figure 2 .2 .2 . AES Sn(430)/Pt(237) Pt(111) (2x2) alloy 1.53 LD ~ 0 a/3 alloy 2.05 Sn 150 200 250 300 350 400 450 500 550 Kinetic Energy (eV) Figure 2.2.2 AES spectra of clean Pt(l 11) and the (2x2)-Sn/Pt(l 11) and (V3xA/3)R30°-Sn/Pt(l 11) surface alloys. The corresponding LEED patterns are given in Figures 2.2.3a-c, respectively. The Sn(430)/Pt(237) ratio in AES was 1.5-1.6 for the (2x2)-Sn/Pt(l 11) surface alloy and 2.05-2.15 for the (A/3xV3)R30°-Sn/Pt(l 11) surface alloy. In addition, a LEED pattern for a mixed-phase S n /P t(lll) surface alloy with Sn(430)/Pt(237) = 1.75, 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. which was occasionally made in our experiments, is shown in Figure 2.2.3d. In addition to LEED and AES, butane adsorption and desorption was used to characterize the Sn/Pt(l 11) surface alloys as shown in Fig 2.2.4. We found that the S n /P t(lll) surface composition was characterized effectively by the desorption peak temperature of adsorbed butane. This chemical method was more sensitive than either AES or LEED to the presence of defective alloy regions at the surface. (a) P t( lll) (b) (2x2) alloy (c) a /3 alloy (d) mixed alloy Figure 2.2.3 LEED pictures taken at for an incident beam energy of 120 eV: (a) Pt(l 11); (b) (2x2)-Sn/Pt(l 11) alloy; (c) (A/3xA/3)R30°-Sn/Pt(l 11) alloy; (d) mixed alloy. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TPD 147 0.2 L n-Butane on different surfaces (V3xV3)R30°Sn-Pt(111) oo LO M 156 Mixed surface alloy (2x2)Sn-Pt(111) 179 150 200 250 300 100 Temperrature (K) Figure 2.2.4 ^-butane adsorption and desorption provides a chemical method to characterize these surface alloys, n-butane TPD is shown after a monolayer saturation coverage of n-butane was adsorbed on the surface at 100 K. Prior work has already established that for samples prepared in this manner, Sn is incorporated substitutionally into primarily only the surface layer to form an ordered alloy or intermetallic compound with 0sn =O.25, with a composition corresponding to the (111) face of a bulk PfjSn crystal, and 0sn =0.33, with a composition corresponding to a Pt2Sn surface. These surface alloys are relatively “flat” but Sn atoms protrude 0.02 nm above the surface-Pt plane [1,2]. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For brevity, we will refer to the (2 x2 )-Sn/Pt(l 11) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys as the (2x2) and a/3 alloys, respectively, in the rest of this thesis. 2.3 Hydrogen Atom Doser A Pt-tube doser was constructed, based on the design of Engel and Rieder [3] as a pyrolytic source of gas-phase hydrogen atoms. This source is drawn schematically in Figure 2.3.1. Two OFHC Cu blocks were silver soldered onto a 3- tube feedthrough (Insulator Seal, cat. #9462011) through some custom-designed adaptors and Cu-tube extensions. The two Cu blocks were electrically isolated by a sapphire plate and fastened together with two screws that were electrically isolated by ceramic spacers. The principal component of the doser was a bent Pt tube (Goodfellow; 99.95%, 1.0 mm O.D., 0.8 mm I.D, 10 cm length) into which a hole of 0.1 mm diameter was mechanically drilled (The bit used was purchased from www.store.yahoo.com/drillcity, Catalog #05M010). This Pt tube was silver soldered to both Cu blocks in the last step of construction. Figures 2.3.2a-d illustrate the structures of this source in detail. The tube can be resistively heated up to 1275°C while the Cu blocks are kept cold via recirculating water from an ice bath using a peristaltic pump. The doser was pretreated daily with O2 at a chamber background pressure of 5x1 O ' 8 torr with the doser at 800 °C for 5 minutes and then it was flashed to 1000 °C 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. before use. The temperature of the Pt tube was directly measured by an optical pyrometer (Omega) that was calibrated by the temperature of the P t( lll) crystal sample, as measured by a Cr/Al thermocouple. The estimated relative accuracy of the pyrometer reading was ±5°C. H atom doser H2 0 out’ H2 0 in\ Screw with ceramic spacer ^/S apphire spacer Cu Platinum tube Figure 2.3.1 Schematic drawing of the platinum-tube hydrogen atom source. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Od 6.45 mm Cu tube Od J .0 mm, id O.Jj mm tube with a hole in the center ( D ^ 0.1mm) | Od 6.45 m m Cu tube Cu block Tube adaptor Cu tube feedthrough 164 mm 2.75” conflate flange 228.6 mm Cham ber center line Figure 2.3.2a Side view of the Pt-tube H-atom doser. Figure 2.3.2b Flange-view of the Fl-atom doser. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T hrough to gas mm Hole diameter = 1. 1mm 6.0 m m Sapphire spacer ] nun ram .I.A -I.ffl!5 . 1 . Center point B lind 0.2 m m deep hole 5.9 mm 6.8 mm I + ■ 25.4 mm Figure 2.3.2c Front view of the Fl-atom doser (Cu block only). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Zoom in of front side 0.1 mm diam eter hole at the top For illustration 9.9 mm Curvature = D/2 =2.4 mm C enter line Front view Figure 2.3.2d Side view of the Cu block and Pt tube of the H-atom doser. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Because H2 molecules can not dissociate under UHV conditions to produce adsorbed H adatoms on the (V3xV3)R30°-Sn/Pt(l 11) alloy, we used this alloy to optimize the operation temperature of the H-atom doser. At a chamber background pressure of P = 5x1 O ' 8 torr with the sample temperature Tsu rf = 100 K, H atoms were exposed to the V3 surface alloy for 100 sec under different operation temperatures of the doser. The H2 desorption peak area in the following TPD experiments was used to calculate the H coverage on the V3 surface alloy. The amount of H2 was determined by comparison of this H2 TPD peak area to a reference H2 TPD spectrum obtained after ethylene (C2H4) exposures on P t( lll) at 300 K to produce 0.25-ML ethylidyne (CCH3) [4], in which complete decomposition produces 0.375-ML H2, i.e. 6^ii=0.75 ML. This value was also checked by comparison to a reference H2 TPD spectrum obtained after a 300-L H2 exposure on Pt(l 11) at 110 K, in which 0h was reported to be 0.8 ML [5], The amount of H2 obtained by using the second method was 10% higher than that obtained by using the first method. The influence of doser temperature on 0h after fixed H exposures on the V 3 alloy at 100 K is shown in Figure 2.3.3. 0h increased with increasing doser temperature until -1073 K (800 °C) and then decreased slightly at higher doser temperatures. A temperature of 800 °C was chosen as the operation temperature in the rest of our experiments. The sharpness of the LEED pattern and the 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sn(430)/Pt(237) AES ratio of the V 3 alloy prior to H exposure were not noticeably affected by H-atom dosing under these conditions. 0.5 H on (V3xV3)R30°-Sn/Pt(111) P = 5.0x1 O '8 torr Exposure time = 100 sec surf 0.4- ^ 0.3- 0) D ) 2 g 0.2- o o X 0.0 200 400 600 800 1000 1200 0 Source Temperature (K) Figure 2.3.3 H coverages observed for a fixed exposure on the V 3 alloy as a function of tbe H-atom source temperature. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In an experiment designed to calibrate the flux of H atoms coming from the source, H atoms were exposed for 20 sec to P t( lll) at 100 K with the Pt tube at o 800°C and a background pressure of 5x10' Torr in the UHV chamber. Assuming that the initial sticking coefficient of H atoms on P t( lll) at 100 K was unity, and using the H2 desorption peak area in the subsequent TPD experiment to determine the value of 9h, the flux of H atoms under these conditions was estimated to be 3 x l0 1 3 atoms cm'2-s'!. TPD spectra of H2 from the (2 x 2 ) and V3 alloys after exposure to H atoms are shown in figure 2.3.4 and 2.3.5 , respectively. H2 (Matheson; 99.99%) was introduced via a variable leak valve (Granville- Phillips) into the Pt-tube doser after passing through a liquid-nitrogen cooled, U-tube trap. Operation of the doser with the sample at 100 K places severe constraints on the amount of water codeposited on the sample during dosing. The water condensation from this source was monitored in several experiments that produced different values for Or. Figure 2.3.4 shows water TPD spectra on Pt(l 11) after preadsorbing different amounts of H adatoms on the surface. H2O TPD following H2O adsorption on Pt(l 11) was used to calibrate the amount of H2O desorption in Figure 2.3.6. These experiments established that the amount of water coadsorption in the subsequent preadsorbed-H experiments was small, ranging from 0.03-0.08 ML for all but the largest H atom precoverage of 6h=0.9 ML where 9 h20 = 0.13 ML. 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (2x2) alloy 347 245 0 H preads < M L) zs E C D C M 0.476 (/) c 0 c 0.324 0.159 200 300 400 500 600 700 800 100 Temperature (K) Figure 2.3.4 H2 TPD spectra after H atoms exposure on the (2x2) alloy at 100 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 alloy 254 Ho T P D 267 213 preads E CD CM >> '(/> c 0 -i— > 0.31 c: 0.26 0.23 0.13 200 300 100 400 500 600 Temperature (K) Figure 2.3.5 H2 TPD spectra after H atoms exposure on the V3 alloy at 100 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. P t(1 11) preads (M l-) water 0.13 0 .9 0 0.08 0.75 0.07 0.59 0.03 = 5 E to GO background 0.01 co £Z C D a 0.14 L H20/Pt(111) 1.34 120 140 160 180 200 T e m p e ra tu re (K) Figure 2.3.6 Water contamination resulting from several exposures using the H-atom doser. The bottom curves provides an H2O TPD spectrum after a H2O exposure that can be used to calibrate the amount of H2O codeposited during H-atom dosing 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.4 References [1] S. H. Overbury, D. R. Mullins, M. T. Paffett, B. E. Koel, Surf. Sci. 254 (1991)45 [2] A. Atrei, U. Bardi, G. Rovida, M. Torrini, E. Zanazzi, P. N. Ross, Phys. Rev. B 46(1992) 1649 [3] T. Engel, K.H. Rieder, in: G. Hohler (Ed.), Structural Studies o f Surfaces, Springer, Berlin, 1982, Vol. 91, p. 55. [4] R.G. Windham, M.E. Bartram, B.E. Koel, J. Phys. Chem. 92 (1988) 2862 [5] K. Christmann, G. Ertl, T. Pignet, Surf. Sci. 54 (1976) 365 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 3 Adsorption and reaction of bicyclic hydrocarbons at P t(lll) and S n /P t(lll) surface alloys: trans-decahydronaphthalene (CioH]8 ) and bicyclohexane (C 12H22) Abstract Adsorption and desorption of trans-decahydronaphthalene (CioHig) and bicyclohexane (C12H22) can be used to probe important aspects of non-specific dehydrogenation leading to surface carbon accumulation and establish better estimates of activation energies for C-H bond cleavage at Pt-Sn alloys. This chemistry was studied on P t( lll) and the (2x2)-Sn/Pt(lll) and (V3xV3)R30°- S n /P t(lll) surface alloys by using temperature programmed desorption (TPD) mass spectroscopy and Auger electron spectroscopy (AES). These hydrocarbons are reactive on Pt(l 11) surfaces and fully dehydrogenate at low coverages to produce H2 and surface carbon during TPD. At monolayer coverage, 87% of adsorbed CioHig and 75% C 12H22 on P t( lll) desorb with activation energies of 70 and 75 kJ/mol, respectively. Decomposition of CioHig is totally inhibited during TPD on these Sn/Pt(l 11) alloys and decomposition of C 12H22 is reduced to 10 % of the monolayer coverage on the (2x2)-Sn/Pt(lll) alloy and totally inhibited on the (V3xV3)R30°- S n /P t(lll) alloy. CioHig and C12H22 are more weakly chemsorbed on these two alloys compared to P t( lll) and these molecules desorb in narrow peaks characteristic of each surface with activation energies of 65 and 73 kJ/mol on the (2x2) alloy and 60 and 70 kJ/mol on the (V3xV3)R30°-Sn/Pt(l 11) alloy, respectively. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Alloyed Sn has little influence on the monolayer saturation coverage of these two molecules, and this is decreased only slightly on these two Sn/Pt(l 11) alloys. The use of these two probe molecules enables an improved estimate of the activation energy barriers E* to break aliphatic C-H bonds in alkanes on Sn/Pt alloys; E* = 65 - 73 kJ/mol on the (2x2)-Sn/Pt(lll) alloy and E* > 70 kJ/mol on the (V3xV3)R30°- Sn/Pt(l 11) alloy. 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.1 Introduction Bimetallic Pt-Sn catalysts are commercially important for suppressing the hydrogenolysis that can occur on pure-Pt catalysts [1,2]. Practical economic issues and superior performance at low pressures makes alumina-supported Pt-Sn catalysts an important commercial system in naphtha reforming [3], In addition, Pt-Sn catalysts are promising systems for efficient reforming that is needed for hydrocarbon fuel cells and are known to catalyze a variety of selective dehydrogenation and hydrogenation reactions. For example, Cortright and Dumesic have reported a highly active and selective Pt/Sn/K-L zeolite catalyst for isobutene dehydrogenation in which Pt-Sn alloy particles are formed [4-6], Recent developments by Schmidt and coworkers [7-9] have shown that thick metal films used in millisecond-partial oxidation reactions may be useful in natural gas oxidation to produce CO and H2 and direct oxidation of ethane to ethylene. For example, thick Pt-Sn alloy films on alumina monoliths have shown high selectivities (85%) and activities (70% conversion) for direct oxidation of ethane to ethylene. The surface remains almost clean after these reactions. Addition of tin in reforming catalysts plays an important role in selectivity and catalyst lifetimes [10]. Sn is considered to reduce C-H and C-C bond breaking activity of Pt catalysts in hydrocarbon reforming reactions, and this influence presumably extends to these newer applications as well. Specifically, Pt-Sn alloy phases are relevant to the action of many of these catalysts and alloyed Sn has been 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. shown to strongly decrease the reactivity of many hydrocarbons and organic molecules on Pt(l 11) surfaces in surface science studies under UHV conditions [11- 20], A quantitative elucidation of hydrocarbon thermochemistry on these Pt-Sn alloys remains an elusive goal to better understand catalytic reaction mechanisms on these catalysts. Such diagrams exist for Pt(111) [21,22], However, on Pt-Sn alloys, the value for the activation energy barrier to C-H bond cleavage in alkanes, alkenes, and other hydrocarbons remains a key fundamental question. This problem arises in part because alkanes (saturated hydrocarbons) interact weakly when molecularly adsorbed on low-Miller index Pt surfaces, and even more weakly on Pt-Sn alloys. Thus, small saturated hydrocarbons have low desorption activation energies and desorb at low temperatures which leads to reversible adsorption of these molecules under UHV conditions, i.e., no decomposition occurs. For example, butane and isobutane adsorb reversibly on two S n /P t(lll) alloy surfaces with desorption energies below 40 kJ/mol [11]. Cs-Cg cycloalkanes are all reversibly adsorbed on two Sn/Pt(l 11) alloy surfaces (a trace amount of dehydrogenation was observed on the (2x2)Sn/Pt(lll) surface, possibly due to defects) [12, 13]. Currently, these desorption activation energies (40- 58 kJ/mol) can be used to place a lower limit on the activation energy barrier E* to break primary C-H bonds in alkanes and cycloalkanes on Sn/Pt alloys. However, these numbers are likely to be much too small. In order to determine the value of E*, or at least place a more realistic bound 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. on it, we need to use larger, more strongly adsorbed hydrocarbon molecules that will remain adsorbed at the alloy surface until temperatures are reached where decomposition occurs. Trans-decahydronaphthalene (CioHix) and bicyclohexane (C12H22), as shown in Scheme I. Both molecules are comprised of two rings and should interact significantly stronger with the surface than those molecules mentioned above. This gives us a better chance to observe C-H bond dissociation due to the increased desorption activation energy. C10H 18 and C 12H22 are of interest in their own right, because they are components of a series of diesel and jet fuel blends made from catalytic reforming processes [23]. However, there have been no previous surface science studies reported for CioHig and C 12H22 adsorption on metal surfaces. Surface science studies of adsorption and desorption of these molecules on Pt and Pt-Sn alloys provides basic information that is helpful in understanding catalytic reforming processes. trans-decahydronaphthalene (C1 0 H 1 8 ) bicyclohexane (C1 2 H22) Scheme I 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In the present study, we investigated the interaction of Ci0 H i8 and C 12H22 with P t( lll) and two Sn/Pt(111) alloys under UHV conditions by temperature- programmed desorption (TPD) and Auger electron spectroscopy (AES). We used two, well-defined, ordered, Pt-Sn surface alloys denoted as the (2x2)Sn/Pt(l 11) and (V3xV3)R30°Sn/Pt(ll 1) alloys prepared by evaporating Sn onto a P t( lll) surface and annealing the sample to 1000 K, as first reported by Paffett and Windham [24], Subsequently, these surfaces have been extensively characterized by LEISS [25], LEED I-V [26], XRD [27], and STM [28, 29], These Sn/Pt(l 11) surface alloys offer outstanding new opportunities for fundamental surface science and catalytic study at the molecular level. 3.2 Experimental methods Experiments were performed in a three-level UHV chamber as described earlier [30], The P t( lll) crystal (Atomergic; 10-mm dia., 1.5-mm thick) was prepared by j H using 1-keV Ar -ion sputtering and oxygen exposures (5x10' -torr O2, at 900 K for 2 min) to give a clean spectrum using Auger electron spectroscopy (AES) and a sharp (lx l) pattern in low energy electron diffraction (LEED). The (2x2)-Sn/Pt(l 11) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys were prepared by evaporating two monolayers of Sn onto the P t( lll) crystal surface and subsequently annealing the sample for 20 s to 1000 and 830 K, respectively. Sn is incorporated substitutionally into primarily only the surface layer to form an ordered 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. alloy or intermetallic compound with A,n =0.25, with a composition corresponding to the (111) face of a bulk Pt3Sn crystal, and 6(n =0.33, with a composition corresponding to a PtaSn surface. These surface alloys are relatively “flat”, but Sn atoms protrude 0.02 nm above the surface-Pt plane at both surfaces [25], In the (2x2) structure, pure-Pt three-fold reactive sites are present, but no adjacent pure-Pt three fold sites exist. All pure-Pt three-fold sites are eliminated in the (V3xV3)R30° structure, and only two-fold pure-Pt sites are present. For brevity throughout this paper, we will refer to the p(2x2)-Sn/Pt(lll) and (V3xV3)R30°-Sn/Pt(lll) surface alloys as the (2x2) and V 3 alloys, respectively. Trans-decahydronaphthalene (CioHig, Aldrich Chem. Co., 99%) and bicyclohexane (C12H22, Aldrich Chem. Co., 99%) were placed in a glass reservoir attached to a stainless-steel dosing line and used as supplied after degassing by multiple freeze-pump-thaw cycles. The gases were exposed on the Pt crystal by a microcapillary array doser connected to the gas line through a variable leak valve. All of the exposures reported here are given simply in terms of the background pressure in the UHV chamber as measured by an ion gauge. No attempt was made to correct for the flux enhancement of the doser or ion gage sensitivity. The mass spectrometer in the chamber was used to check the purity of the gases during dosing. For all TPD experiments, the heating rate was 3.6 K/s and all exposures were given with the surface temperature at 110 K or below. AES measurements were made with a double-pass cylindrical mirror analyzer (CMA) using a modulation 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. voltage of 4 eV. The electron gun was operated at 3-keV beam energy and 1.5-p.A beam current. 3.3 Results Trans-decahydronaphthalene (CioHig). Possible decomposition products produced during TPD following adsorption of trans-decahydronaphthalene (CioHig) were examined by monitoring 24 signals, including those at 138, 136, 134, 132, 130, 128, 78, 80, 82, 67, and 2 amu. Based on analysis of the shape and peak temperature of all of these signals during TPD, only the desorption of H2 and the parent molecule, C 10H 18, were detected from all three surfaces studied. A series of TPD spectra for C 10H 18 desorption after CioHig adsorption on Pt(l 11) at 110 K is shown in Figure 3.1. The signal at 67 amu is due to the primary cracking peak from CioHig and was used to monitor CioHig desorption. A clear separation occurs between a high-temperature peak at 275 K due to a chemisorbed state and a low-temperature peak at 199 K arising from desorption from a physisorbed layer. Increasing the coverage further causes a new peak at 189 K to appear. The peak at 199 K eventually saturates in intensity with increasing exposure, but the peak at 189 K does not. We assign the peaks at 199 and 189 K as due to desorption of second-layer and multilayer species, respectively. H2 evolution, as shown in Figure 3.2, monitors complete CioHig decomposition on P t(lll). The amount of H2 desorption saturates after 0.3-L CioHig exposure, prior to 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reaching monolayer coverage. Five H2 desorption peaks can be identified at 288, 318, 466, 516, and 602 K, which indicates that several elementary steps are needed to complete the decomposition process. The total amount of H2 evolved in TPD after large exposures of CioFfis is 0.20 ML FI2, correspond to the complete decomposition of 0.0224-ML CioHig. The amount of H2 was determined by comparison of this H2 TPD peak area to a reference FL TPD spectrum obtained after ethylene (C2H4) exposures on Pt(l 11) at 300 K to produce 0.25-ML ethylidyne (CCH3) [31], in which complete decomposition produces 0.375-ML FI2, i.e. 6h=0.75 ML. This value was also checked by comparison to a reference H2 TPD spectrum obtained after a 300-L FI2 exposure on P t( lll) at 110 K, in which was reported to be 0.8 ML [32], The amount of FL obtained by using the second method was 10% higher than that obtained by using the first method. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pt(111) 199 275 189 E 03 F- C D t/> C B c 0.9 L 0.6 L 0.45 L 0.3 L 100 200 300 400 500 600 700 800 T em p eratu re (K) Figure 3.1 CioHig TPD spectra after trans-decahydronaphthalene (CioHig) exposures on P t( lll) at 110 K. 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pt(111) 288 318 466 516 ZD £ (D C M 602 co c < 1 ) 0.9 L c 0.6 L 0.45 L 0.3 L 100 200 300 400 500 600 700 800 Temperature (K) Figure 3.2 H2 TPD spectra after trans-decahydronaphthalene (CioHis) exposures on Pt(l 11) at 110 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The C (272 eV)/Pt (237 eV) peak-to-peak ratio in AES was used to determine the monolayer coverage of CioHig on Pt(l 11) and the amount of residual carbon left on the surface following heating to 800 K in TPD, as shown in Figure 3.3. The AES spectrum for the CioHig monolayer was taken after dosing 0.45-L CioHig on Pt(l 11) at 110 K and annealing at 210 K for 5 sec. The amount of carbon present on the surface decreased after TPD due to desorption of reversibly adsorbed CioHis. We can estimate the amount of surface carbon from an AES calibration using 0.25-ML ethylidyne (CCH3) on Pt(l 11) at 300 K [31], which corresponds to 0c = 0.5 ML and gives C(272)/Pt(237) = 0.35. We find that 6c = 1.71 ML and #C i()//|g = 0.17 ML in the CioHig monolayer and 6c = 0.29 ML was left on the surface following TPD. This corresponds to the complete decomposition of 0C m H n = 0-029 ML or 16.8% of the amount of CioFIjg monolayer. This is consistent with the extent of decomposition calculated independently by using the H2 TPD area, which gives 6 * ^ = 0.0224 ML corresponding to 13.1% decomposition A similar series of TPD spectra for CioFLg desorption from the (2x2) and V 3 alloys are shown in Figures 3.4 and 3.5, respectively. In each case, a high- temperature desorption peak was assigned to desorption from a chemisorbed state and low-temperature desorption peaks was assigned to desorption from physisorbed states. No distinct peak from desorption of second-layer molecules was observed 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between the monolayer and multilayer features. Physisorbed molecules in the condensed phase, multilayer desorb in a peak at 186 K on both of these two alloys Pt(111) A E S Clean Pt(111) C/Pt = 0 After TPD L U T3 C/Pt = 0.20 Monolayer C/Pt = 1.20 200 240 160 280 320 K in etic E n e rg y (eV ) Figure 3.3 AES spectra of clean Pt(l 11) (top), monolayer CioHig-covered Pt(l 11) (bottom), and the Pt(l 11) surface following a 0.6-L CioHig TPD experiment (middle). 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 186 (2x2) alloy 256 E C O P - C O (/) C £ c 0.9 L 0.45 L 0.3 L 0.15 L 100 200 300 400 500 600 700 800 Temperature (K) Figure 3.4 CioHig TPD spectra after trans-decahydronaphthalene (CioHis) exposures on a (2x2)Sn/Pt(l 11) surface alloy at 110 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 alloy 186 242 1.2 L -s-5 E C O h - co C O a < D 175 0.51 L 0.3 L 200 300 400 500 600 700 800 Tmeperature (K) Figure 3.5 CioHigTPD spectra after trans-decahydronaphthalene (CiqHix) exposures on a (V3xV3)R30°Sn/Pt(l 11) surface alloy at 110 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. after large exposures. Desorption from the monolayer appeared at 256 K on the (2x2) alloy and 242 K on the V3 alloy. The peak shapes and absence of a shift with increasing CioHig coverage on both of these surface alloys and on Pt(l 11) indicate first-order desorption kinetics. Assuming first-order desorption kinetics with a preexponential factor of 101 3 s'1 , the Redhead method [33] gives an estimate of the desorption activation energy E & of 70, 65, 60 kJ/mol for CioHig adsorbed in the monolayer on Pt(l 11) and the (2x2) and a/3 alloys, respectively. No appreciable amount of H2 was detected during TPD from the two Pt-Sn alloys following CioHig exposures and no carbon was detected by AES following TPD. Thus, CioHig is completely reversibly adsorbed on these two alloys under these conditions and no decomposition occurs during TPD. Bicyclohexane (C 12H 22). C 12H 12 and H2 were the only two desorption products detected in TPD following C 12H 12 exposures on P t( lll) at 110 K. The signal at 82 amu is the major cracking peak of C 12H 12 and was used to monitor the desorption of C12H 12. TPD spectra for C 12H22 desorption from P t( lll) are shown in Figure 3.6. After exposures less than 0.4 L, no C 12H 12 desorption was detected during heating in TPD. When the coverage was increased after larger exposures, a single peak at 291 K was observed initially due to desorption from a chemisorbed state and then a low-temperature peak at 210 K arose from desorption from a 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. condensed, physisorbed layer. No clearly resolved C 12H 12 desorption peak from second-layer physisorbed species was observed. Pt(111) 1210 1 5 E C O C sJ 00 291 >> +j to c (U 2.4 L a 0.4 L 100 200 300 400 500 600 700 800 Temperature (K) Figure 3.6 C 12H22 TPD spectra after bicyclohexane (C 12H22) exposures on Pt(l 11) at 110 K. 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H2 evolution from C 12H22 decomposition on Pt(l 11) during TPD is shown in Figure 3.7. For increasing C 12FI22 exposures, the amount of H2 desorption increased until after a 1.2-L C 12H22 exposure. Four desorption features can be identified at 305, 334, 417, and 486 K corresponding to sequential decomposition steps. The maximum yield in the Ft2 TPD curves from C 12H22 exposures in Figure 3.7 corresponds to 6a = 0.89 ML. Using an ethylidyne reference [31], this indicates that 0.040-ML C 12H22 decomposed during heating in TPD. Figure 3.8 shows AES spectra from P t( lll) after C 12H22 experiments. A C 12H22 monolayer was obtained by dosing 1.2-L C 12H22 on Pt(l 11) at 110 K and then annealing to 220 K for 5 sec. By using the C(272)/Pt(237) AES peak ratio from ethylidyne [31], we determined that 6c = 1.92 ML, or 6CnHi= 0.16 ML in the monolayer, and 0c = 0.60 ML left on the surface following TPD. This corresponds to 0 ‘ ''2 ht i =0.050 ML, or 31 % of the C 12H22 monolayer. This is fairly consistent with the value of 25 % of the C12FI22 monolayer that decomposed deduced from the H2 TPD measurements in Figure 7. On the (2x2) surface alloy, some irreversible C 12H 12 adsorption occurred during TPD experiments, but no desorption products other than C 12H 12 and H2 were detected. Figure 3.9 shows TPD spectra for C 12H 12 desorption following C 12H 12 dosing on the (2x2) surface alloy at 110 K. The peak maximum for desorption from the chemisorbed monolayer occurs at 286 K. This value varies only slightly with 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. changes in the coverage. A desorption peak at 206 K, the same temperature as that observed on P t(lll), forms after large exposures due to desorption of physisorbed species. P t(1 1 1 ) 3 05 334 4 86 417 2 .4 L 3 E 05 C M >> to c < D -I— » c 0 .4 L 100 2 00 300 4 0 0 500 6 00 700 800 T em p erature (K) Figure 3.7 H2 TPD spectra after bicyclohexane (C12H22) exposures on Pt(l 11) at 110 K. 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pt(111) AES Clean Pt(111) C/Pt = 0 After TPD C/Pt = 0.42 Monolayer C/Pt = 1.35 200 240 280 320 160 Kinetic Energy (eV) Figure 3.8 AES spectra of clean P t( lll) (top), monolayer CnHn-covered P t( lll) (bottom), and the P t( lll) surface following a 1.2-L C 12H22 TPD experiment (middle). 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (2 x 2 ) alloy 206 3 E oj C M oo 286 (/) c 0 1.71 L £Z 1.35 L 0.6 L 0.3 L 100 200 300 400 500 600 700 800 Temperature (K) Figure 3.9 C 12H22 TPD spectra after bicyclohexane (C12H22) exposures on (2x2)Sn/Pt(l 11) surface alloy at 110 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FL TPD spectra shown in Figure 3.10 were used to monitor C 12H22 decomposition on the (2x2) alloy. The H2 desorption yield saturated after a 1.35-L C 12H22 exposure. H2 desorption initially had a feature near 291 K and a broad peak evolved near 370 K with increasing exposures. The maximum amount of H2 evolved in Figure 3.10 corresponds to 6ft = 0.35 ML, using the same ethylidyne reference as before, which indicates that 6 l !c c ,, =0.016 ML. The amount of C 12H22 that was 5 M 2 " 22 desorbed in TPD was 0=0.145 ML, as determined by comparison to the C 12H22 TPD area and amount of C 12H22 desorbed from Pt(l 11), and so the monolayer coverage of 0.161 ML could be obtained by adding the amount of C 12H22 desorbed and decomposed. This amount is almost same as that on P t(lll). About 10 % of the C 12H22 monolayer on the (2x2) alloy decomposed during heating in TPD. C 12H22 reversibly adsorbs and desorbs molecularly from the V3 alloy without any decomposition. This was deduced by the absence of any detectable signal at 2 amu in TPD due to H2 desorption and by the lack of any carbon signal in AES following TPD. A series of TPD spectra obtained at 82 amu for C 12H22 desorption from the V3 alloy are shown in Figure 3.11. Desorption of C 12H 12 from the monolayer on this alloy occurs in a peak at 275 K that does not shift significantly with increasing coverage. Desorption from physisorbed species produced after large exposures occurs in a peak at 206 K, which is same as that on the (2x2) alloy and close to that on Pt(l 11). 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 370 (2x2)aHOy H0 TPD 291 =3 £ O J C M >> 1.35 L c 0.6 L 0.3 L 100 200 300 400 500 600 700 800 Temperature (K) Figure 3.10 H2 TPD spectra after bicyclohexane (C 12H22) exposures on (2x2)Sn/Pt(l 11) surface alloy at 110 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 alloy 209 Zi E 03 CM 00 275 3.6 L C /5 C CD c 1.71 L 0.9 L 0.6 L 600 700 800 200 300 400 500 100 Tmeperature (K) Figure 3.11 C 12H22TPD spectra after bicyclohexane (C 12H22) exposures on a (V3xV3)R30°Sn/Pt(l 11) surface alloy at 110 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Redhead analysis using the C12H22 desorption peak maxima of TPD peaks at moderate coverages in the chemisorbed monolayer on the three surfaces provides desorption activation energies of 75, 73, and 70 kJ/mol on Pt(l 11) and the (2x2) and V3 alloys, respectively. Figure 3.12 summarizes the influence of alloyed Sn on the monolayer coverage, desorption activation energy, and amount of decomposition of these two closely related molecules. The coverage in the chemisorbed monolayer and the amount of decomposition are given on the left-hand scale of Figure 3.12. The saturation coverage of chemisorbed CioHig and C 12H22 are nearly the same on each surface, and the coverage of each molecule slightly decreases only slightly with increasing Sn concentration in the alloys. Flowever, the reactivity of the alloy surfaces, as probed by the decomposition of both molecules, decreases noticeably with increasing Sn concentration. CuTIis decomposition (13% on P t( lll)) is completely inhibited on both surface alloys while C 12H22 decomposition (25% on P t(lll)) decreases on the (2x2) alloy to 10% and is completely inhibited on the V3 alloy. The adsorption energy, which is equal to the desorption activation energy (Ed) in the case of non-activated adsorption (as is the case here), decreases for both molecules only slightly with increasing Sn concentration in the S n /P t(lll) alloy surface. However, this change is non-linear, which illustrates the importance of 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. “reactive sites” and specific surface geometries rather than a generalized scaling or homogeneous influence of 0.20 - 90 sat 0 .1 0 -Q . < 1 > O ) C t 3 a) > o O - 70 - 6 0 0.02 - 50 0.4 0.3 0.2 Figure 3.12 Influence of alloyed Sn on C]0Hi8 and C 12H22 adsorption and reaction on Pt( 111) and two Sn/Pt(l 11) surface alloys 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.4 Discussion Adsorption and desorption of alkanes [11, 34] and cycloalkanes [12, 13] on P t( lll) and two Sn/Pt(111) alloys under UHV conditions have been studied previously, and many aspects of the influence of alloyed Sn on this chemistry can be described. Linear alkanes are reversibly adsorbed on Pt(l 11), up to at least n-hexane (C6H 14), and do not decompose during heating in TPD. However, the cycloalkanes are more reactive. While n-hexane [34] has the same desorption activation energy, Ed = 58 kJ/mol, as cyclohexane (c-CeH^) on P t( lll) [12], dehydrogenation of cyclohexane also occurs with desorption during TPD on P t( lll) [12, 35], This dehydrogenation takes place in a step-wise fashion, first to form adsorbed cyclohexyl groups with a barrier of E*= 42 kJ/mol, eventually forming benzene [22], and then adsorbed benzene dehydrogenates further to release H2 at higher temperature in TPD. Methylcyclohexane (C-C7H 14), even though it has the same adsorption energy, Ed = 56-61 kJ/mol [36, 37], as cyclohexane, contains tertiary (3°) C-H bonds that are weaker and more reactive [38]. Consistently, 45% of the adsorbed monolayer decomposes on P t( lll) during TPD [36] via a barrier that is predicted to be lower than that for cyclohexane dehydrogenation. As expected, CioHig (trans-decahydronaphthalene) and C 12H22 (bicyclohexane) both decomposed on P t( lll) during TPD. Presumably CioHig dehydrogenates step-wise by losing 10 H atoms to form adsorbed naphthalene as one 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. eventual intermediate. Adsorbed naphthalene continues to dehydrogenate and liberate additional H2 in TPD. This view is supported by the fact that the H2 TPD area below 400 K and that above 400 K in Figure 3.2 has a ratio close to 5:4. Similarly, adsorbed biphenyl may act as the intermediate during the dehydrogenation of C 12H22 since the H2 TPD area below 400 K and that above 400 K in Figure 3.2 has a ratio close to 6:5. Although both C 12H22 and CioHjg have bicyclic ring structures, different carbon numbers and structures lead to a small difference in the adsorption energies, i.e., 75 and 70 kJ/mol, respectively. The larger adsorption energy of C 12FI22 leads to a higher reactivity and the percentage of C 12II22 that decomposes (25%) in the monolayer on P t( lll) during TPD is almost twice that of CuTTs (13%). We expect that the C-FI bond dissociation barrier on Pt(l 11) for these cycloalkanes, which have secondary (2°) and tertiary (3°) C-H bonds, is the same or lower than that for cyclohexane, which has only secondary (2°) C-H bonds, and similar to that of methylcyclohexane. Thus, no new information is gained about the C-H bond dissociation barrier E* on P t(lll). No information about the barrier for C-C bond scission was obtained either, because no carbon-containing decomposition products were detected in TPD (other than the parent molecules) and no spectroscopic evidence was obtained for the surface species formed during TPD. The replacement of 25% or more of the surface Pt atoms by Sn, 9sn >0.25, in the surface alloys suppressed CioHig dehydrogenation completely under these 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conditions. This was observed previously for cyclohexane [12]. C 12H22 is more strongly adsorbed and forming the V3 surface alloy with = 0.33 was required to totally stop dehydrogenation during TPD. On the (2x2) alloy, the desorption energy of C 12H22 (73 kJ/mol) is 8 kJ/mol higher than that of CioHig (65 kJ/mol) and this allows dehydrogenation to compete with desorption at low coverages. Thus, we can estimate that the 3°-C-H bond dissociation barrier £*=65-73 kJ/mol on the (2x2) alloy, and £*>70 kJ/mol on the V 3 alloy. This can be compared to that on P t( lll) where £*<42 kJ/mol. The barrier to breaking the first C-H bond in adsorbed cycloalkanes on the V 3 alloy is much higher than that to break the next C-H bond in a subsequent reaction. This assertion comes from results using surface-bound cyclohexyl species that were created on all three of these surfaces to explore dehydrogenation reactions [12]. On the V3 alloy, cyclohexene that was produced from cyclohexyl dehydrogenation desorbed at 208 K (and even this was thought to be desorption-rate-limited). Thus, C-H bond scission in cyclohexyl dehydrogenation occurs at much lower temperatures than 275 K where C12H 22 desorbed molecularly without any decomposition. An understanding of the origin of the increase in the C-H bond-breaking barrier upon alloying Pt with Sn is still forthcoming. The electronic structure of the two S n /P t(lll) surface alloys studied here was characterized previously by using UPS [39] and XPS [40]. XPS showed no significant chemical shifts of the Pt core 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. level peaks, and only small changes occurred in the UPS spectra upon alloying. However, UPS spectra of the alloys were not a simple addition of Pt and Sn spectra, and some new features were attributed to either a strongly modified band of a single component or some kind of common band of Sn and Pt. Electronic structure calculations of bulk PtaSn alloys by Pick [41] showed that bonding interactions between Pt and Sn in the alloy changed the local electronic structure at Pt and Sn sites. Hybridization between Pt-d and Sn-/? electrons leads to a lowering of the LDOS at the Fermi level and a downward shift of the Pt local d-band. Depopulation of the Pt 5d band, and the loss of adjacent pure-Pt 3-fold hollow sites, in the (2x2) alloy evidently leads to the formation of a significant activation energy barrier for dissociation reactions of adsorbed molecules. Further electronic changes such as lowering of the LDOS at the Fermi level, as observed directly with STM [28], and/or structural changes involving the loss of all pure-Pt 3-fold hollow sites, upon forming the a/3 alloy decrease the reactivity of this alloy even further. Ci0 H 1 8 and C 12H22 desorb in a single, well-defined, characteristic peak for each of the three surfaces, like that observed previously for cyclohexane and other alkane molecules. Thus, Ci0 H |8 and C12H22 desorb with little sensitivity to the presence of individual Pt and Sn atoms in the surface layer of these Pt-Sn alloys. C 10H 18 and C 12H22 adsorption energies on P t( lll) are decreased, but only a small amount, upon alloying with Sn. For example, these energies are lowered by 14 and 6 %, respectively, when comparing Pt(l 11) to the a/3 alloy with 0sn = 0.33. Changes in 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. electronic structure have weak effects on saturated hydrocarbons because they interact with metal surfaces primarily via van der Waals or polarization interactions. In contrast, the ethylene adsorption energy is decreased by 40 % comparing P t( lll) to the V3 alloy due to the strong interaction between fully rehybridized, di-a-bonded ethylene molecules and the Pt(l 11) surface [42]. We note that the adsorption energy of C|oH18 on Pt(l 11) decreased upon alloying more than that of C12H22. This may be caused by the difference in molecular geometry. The a bond between the two rings in C 12H22 allows this molecule slightly greater flexibility in adjusting the adsorption geometry on the alloy surfaces. Finally, we point out that there is only a small decrease in the CioFlig and C 12H22 monolayer saturation coverage in comparing Pt(l 11) with the two Sn/Pt(l 11) alloys, and this fact should be considered in discussing and modeling catalytic reactions on these alloy surfaces: up to one-third of a monolayer of alloyed Sn does not effectively block sites for adsorption of such molecules. 3.5 Conclusions CioHig (trans-decahydronaphthalene) and C12H22 (bicyclohexane) are partially reversibly adsorbed on Pt(l 11) at 110 K, with some decomposition occuring during TPD. The only gas-phase decomposition product detected was H2. CioHig desorbs from Pt(l 11) at 275 K with a desorption activation energy of E & = 70 kJ/mol, and 13 % of the adsorbed monolayer fully dehydrogenates to produce surface carbon 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. during TPD. C 12H22 desorbs from Pt(l 11) at 291 K with E & = 15 kJ/mol, and 2 5 % of the adsorbed monolayer fully dehydrogenates to produce surface carbon during TPD. The presence of alloyed Sn in the surface layer of the two surface alloys, the (2x2)- Sn/Pt(l 11) with 9 s n = 0 . 2 5 and (V3xV3)R30°-Sn/Pt(lll) with 6^n =0.33, caused only a small decrease in the monolayer coverages and a relatively small decrease in the adsorption energies of these molecules. C10H 18 and C 12H22 desorb at lower temperatures from these S n /P t(lll) surface alloys in single, narrow characteristic peaks. CioHig is completely reversibly adsorbed on both of the two Sn/Pt(l 11) alloys and no decomposition occurs upon heating in TPD. The decomposition of C 12H22 during TPD is reduced to 10 % of the monolayer on the (2x2) surface alloy and eliminated on the V3 surface alloy. In addition to providing additional insight into the chemistry of bicyclic hydrocarbons on Pt and Pt-Sn alloy surfaces, these measurements enable us to improve our estimates of the activation energy barriers for breaking aliphatic C-H bonds in hydrocarbons adsorbed on Pt-Sn alloy surfaces: E*=65-73 kJ/mol on the (2x2)-Sn/Pt(l 11) surface alloy and E* > 70 kJ/mol on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.6 References [1] Z. Karpinski and J. K. Clarke, J. Chem. Soc. Faraday Trans. /, 71 (1975) 893. [2] B. H. Davis, J. Catal. 46 (1977) 378. [3] H. Verbeek and W. M. H. Sachtler, J. Catal. 42 (1976) 257. [4] R.D. Cortright and J. A. Dumestic, Appl. Catal. A 129 (1995) 101. [5] R.D. Cortright, P.E. Levin, and J. A. Dumestic Ind. Eng. Chem. Res. 37 (1998) 1717. [6 ] J.M. Hill, R.D. Cortright, and J. A. Dumestic, Appl. Catal. A-Gen. 168 (1998) 9. [7] D. A. Hickman and L. D. Schmidt, Science 259 (1993) 343. [8] A. S. Bodke, D.A. Olschki, and L.D Schmidt, Science 285 (1999) 712. [9] L.S. Liebmann and L.D Schmidt Appl. Catal. A-Gen. 179 (1999) 93. [10] V. Ponec and G. C. Bond, Stud. Surf. Sci. Catal. 95 (1995) 477. [11] C. Xu, B. E Koel, and M. T. Paffett, Langmuir 10 (1994) 166. [12] C. Xu, Y. Tsai, and B. E. Koel, J. Phy. Chem. 98 (1994) 585. [13] Y. Tsai and B. E Koel, Langmuir 14 (1998) 1290. [14] J.W. Peck, D.I. Mahon, and B.E. Koel, Surf. Sci. 410 (1998) 200. [15] J.W. Peck, D.I. Mahon, D.E. Beck and B.E. Koel, Surf. Sci. 410 (1998) 170. [16] C. Panja, N. Saliba, and B.E. Koel, Surf. Sci. 395 (1998) 248. [17] Y. Tsai, C. Xu, and B. E. Koel, Surf. Sci. 385 (1997) 37. [18] Y. Tsai and B. E. Koel, J. Phy. Chem. B 10 (1997) 2895. 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [19] J.W. Peck and B.E. Koel, J. Am. Chem. Soc. 99 (1995) 16670. [20] C. Xu and B.E. Koel, Surf. Sci. 304 (1994) 249. [21] E. A. Carter and B. E. Koel, Surf. Sci. 226 (1990) 339. [22] B.E. Koel, D.A. Blank, and E.A Carter, J. Mol. Catal. A: Chemical 131 (1998)39. [23] N. White PCTInt. appl. (1985) 56. [24] M. T. Paffett and R. G. Windham, Surf Sci. 208 (1989) 34. [25] S.H. Overbury, D.R. Mullins, M.T. Paffett, and B.E. Koel, Surf. Sci. 254 (1991)45. [26] A. Atrei, U. Bardi, G. Rovida, M. Torrini, E. Zanazzi, and P.N. Ross, Phy. Rev. B 46 (1992) 1649. [27] M. Galeotti, A. Atrei, U. Bardi, G. Rovida, and M. Torrini, Surf. Sci. 313 (1994)349. [28] M. Batzill, D. E. Beck, and B. E. Koel, Surf. Sci. 466 (2000) L821. [29] J. Kuntze, S. Speller, W. Heiland, A. Atrei, I. Spolveri, and U. Bardi, Phy. Rev. B 58 (1998) R16005. [30] H. Zhao, J. Kim, and B. E. Koel, Surf. Sci, 538 (2003) 147. [31] R.G. Wndham, M.E. Bartram, and B.E. Koel, J. Phy. Chem. 92 (1988) 2862. [32] K. Christmann, G. Ertl, and T. Pignet, Surf. Sci. 54 (1976) 365. [33] P. A. Redhead, Vacuum 12 (1962) 203. [34] H. Zhao and B.E. Koel, to be published. 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [35] J.A. Rodriguez and C.T. Campbell, J. Phy. Chem. 93 (1989) 826. [36] C. Xu, B.E. Koel, M.A. Newton, N.A. Frei, and C.T. Campbell, J. Phy. Chem. 99 (1995) 16670. [37] L.Q. Jiang, A. Avoyan, B.E. Koel, and J. L. Falconer, J. Am. Chem. Soc. 115 (1993) 12106. [38] R.T. Morrison and R.N. Boyd, Organic Chemistry, (Allyn and Bacon, Boston, 1973). [39] M. T. Paffett, S.C Gebhard, R. G. Windham, and B.E. Koel, J. Phy. Chem. 94 (1990) 6831. [40] Y. Li, M. R. Voss, N. Swami, Y. Tsai, and B. E. Koel, Phy. Rev. B 56 (1997) 15982. [41] S. Pick, Surf. Sci. 436 (1999) 220. [42] M. T. Paffett, S.C Gebhard, R. G. Windham, and B.E. Koel, Surf. Sci. 223 (1989) 449. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 4 Adsorption and reaction of 1,3-butadiene on P t(lll) and S n/P t(lll) surface alloys A bstract Temperature-programmed desorption (TPD), Auger electron spectroscopy (AES), and low-energy electron diffraction (LEED) were used to study the chemistry of 1,3-butadiene (H2C=CHCH=CH2, C4H6 ) on P t( lll) and p(2x2)-Sn/Pt(lll) and (V3xV3)R30°-Sn/Pt(lll) surface alloys. All chemisorbed 1,3-butadiene completely dehydrogenated to H2 and surface carbon on P t(lll). Alloying Sn on P t( lll) can completely inhibit this decomposition and 1,3-butadiene reversibly adsorbs and desorbs from the two S n /P t(lll) alloys under UHV conditions. The desorption activation energy of 1,3-butadiene on the (2 x2 ) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys is 88 and 75 kJ/mol, respectively. These values are good estimates of the adsorption energies, and also place lower limits on the activation energy barrier for dissociating vinylic C-H bonds on the (2x2) and V3 surface alloys. Even though 1,3- butadiene is much more strongly chemisorbed than 1-butene (H2C=CHCII2CH3, C4H 8) on the (2x2)-Sn/Pt(lll) alloy, 1,3-butadiene is less reactive than 1-butene because there are no allylic (3-CH bonds in 1,3-butadiene as there are in 1-butene. 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.1 Introduction Chemisorption of alkenes on Pt(l 11) has been investigated rather extensively because of the fundamental interest in the chemistry of these molecules as reactants, products and intermediates in hydrocarbon conversion over Pt-based catalysts [1-5]. Surface science studies of the chemistry of acetylene (HC=CH, C2H2) [6-13] and 1,3-butadiene (H2C=CHCH=CH2, C4H6 ) [4, 14, 15] on P t( lll) have also been carried out in part due to the industrial importance of partial hydrogenation of dienes and alkynes [16-17]. The activity and selectivity for such reactions strongly depend on the metal used as a catalyst. Two metallic systems, Pd and Pt, have been the focus of most fundamental studies of the hydrogenation of hydrocarbons [18]. For example, butadiene is preferentially hydrogenated to form butane (H3CCH2CH2CH3, C4H 10) over Pt [19]. Generally, the addition of Sn to supported Pt catalysts for hydrocarbon conversion decreases the catalytic activity, but increases the selectivity for unsaturated hydrocarbon products and reduces coking, a poisoning of the catalyst that decreases its useful lifetime due to carbon accumulation from non-specific dehydrogenation. Bimetallic Pt-Sn catalysts are particularly promising for a variety of selective dehydrogenation and hydrogenation reactions. For example, Cortright and Dumesic [20-22] have reported a highly active and selective Pt/Sn/K-L zeolite catalyst for isobutene dehydrogenation in which Pt-Sn alloy particles are clearly formed. The reduced reactivity of Pt-Sn alloy surfaces provides a simple explanation 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for why bimetallic Pt-Sn catalysts may have an increased selectivity for producing alkenes from the hydrogenation of dienes over that of Pt catalysts. Fundamental understanding of adsorption and reaction of 1,3-butadiene on well-defined Pt-Sn alloys would provide important information for understanding and improving Pt-Sn catalysts for use in diene hydrogenation. Adsorption of 1,3-butadiene on P t( lll) has been studied experimentally by temperature programmed desorption (TPD) [4], high-resolution electron energy loss spectroscopy (HREELS) [4, 14,15] , and near-edge X-ray absorption fine structure (NEXAFS) [14, 15]. Avery et al. [4] concluded that chemisorption occurred via di-a- bonding of both C=C bonds in the molecule in a tetra-a-bonding configuration without any overall hydrogen loss at 300 K. This picture is supported by recent DFT calculations[23, 24]. However, Bertolini et al. [14,15] argued that only the carbon atoms at the ends of the molecule form c t bonds to the Pt( 111) surface, in a 1,4-di-cr- bonding configuration, and chemisorption is accompanied by the formation of a new C=C bond at the center of the molecule. This is consistent with an older calculation by Baetzold [25], Alloying Sn in a Pt(l 11) surface greatly changes the reactivity of the surface and weakens the interaction with adsorbed molecules [26-28]. Under UHV conditions, decomposition of alkenes is totally inhibited on the (V3xV3)R30°- S n /P t(lll) alloy and greatly decreased on the (2x2)-Sn/Pt(l 11) alloy [26-28]. 1,3- butadiene contains two C=C double bonds, and thus is expected to have a much 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. stronger interaction with these alloys (barring severe steric constraints due to Sn). Stronger adsorption often leads to higher reactivity and so this is a interesting molecule to probe C-H bond cleavage barriers on these alloy surfaces. In this paper, we investigated adsorption, desorption, and dehydrogenation of 1,3-butadiene on P t(lll), the (2x2)-Sn/Pt(lll) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys. We examined the influence of alloy structure on the chemistry of 1,3- butadiene and compared the reactivity of this molecule, as a prototype diene, with that of the alkenes, ethylene (H2C=CH2, C2H4) and 1-butene (H2C=CHCH2CH3, C4H8 ). 4.2 Experimental methods Experiments were performed in a three-level UHV chamber as described earlier [29]. The P t( lll) crystal (Atomergic; 10-mm dia., 1.5-mm thick) was prepared by using 1-keV Ar+-ion sputtering and oxygen exposures (5xl0'7 -torr O2, at 900 K for 2 min) to give a clean spectrum in Auger electron spectroscopy (AES) and a sharp (lx l) pattern in low energy electron diffraction (LEED). The (2 x2 )S n /P t(lll) and (V3xV3)R30°Sn/Pt(l 11) surface alloys were prepared by evaporating one monolayer of Sn onto the P t( lll) crystal surface and subsequently annealing the sample for 20 s to 1000 K and 830 K, respectively. Sn is incorporated substitutionally into primarily only the surface layer to form an ordered alloy or intermetallic compound with 9sn =0.25, with a composition corresponding to 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the (111) face of a bulk PtsSn crystal, and 0sn =O.33, with a composition corresponding to a Pt2Sn surface. These surface alloys are relatively “flat”, but Sn atoms protrude 0.02 nm above the surface-Pt plane at both surfaces [30], In the (2x2) structure, pure-Pt three-fold reactive sites are present, but no adjacent pure-Pt three fold sites exist. All pure-Pt three-fold sites are eliminated in the (a/3xa/3)R30° structure, and only two-fold pure Pt sites are present. For brevity throughout this paper, we will refer to the p(2 x2 )-S n/P t(lll) and (V3xV3)R30°-Sn/Pt(lll) surface alloys as the (2x2) and a/3 alloys, respectively. 1,3-butadiene (C4H6, Matheson, 99.5%) was used without additional purification. The gas was exposed on the Pt crystal by a microcapillary array doser connected to a gas line through a variable leak valve. All of the exposures reported here are given simply in terms of the background pressure in the UHV chamber as measured by an ion gauge. No attempt was made to correct for the flux enhancement of the doser or ion gage sensitivity. The mass spectrometer in the chamber was used to check the purity of the gases during dosing. For all TPD experiments, the heating rate was 3.6 K/s and all exposures were given with the surface temperature at 100 K. AES measurements were made with a double-pass cylindrical mirror analyzer (CMA) using a modulation voltage of 4 eV. The electron gun was operated at 3-keV beam energy and 1.5-pA beam current. Coverages 0, reported in this paper are referenced to the surface atom density of Pt(l 11) such that 0pt =1.0 ML is defined as 1.505 x l0 15cm'2. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.3 Results Figure 4.1 shows butadiene, C4H6, TPD spectra after different 1,3-butadiene exposures on P t( lll) at 100 K. At low exposures, no molecular C4H6 desorbs upon heating. Increasing butadiene exposures leads to desorption from the physisorbed layer at 130 K. At any coverage, the chemisorbed layer of butadiene decomposes and only H2 is detected in the TPD spectra. The dehydrogenation of butadiene leads to the build up of surface carbon on Pt(l 11), which is detected by AES following TPD experiments. Our results agree with the TPD results of Avery et al. [4], H2 evolution, as shown in Figure 4.2, monitors complete butadiene decomposition on P t(lll). The amount of H2 desorption saturates at 0.18-L exposure, where the butadiene monolayer saturation coverage was reached as determined by monitoring C4H6 desorption in Figure 4.1. Three H2 desorption peaks can be identified at 370, 398 and 585 K, which indicates that several steps are needed to complete the decomposition process. At low coverage, the first H2 desorption peak appears at 339 K and no peak at 398 K is detected. This shift indicates that co adsorbed hydrocarbon fragments produced from dehydrogenation increases the C-H bond cleavage barrier. The total amount of H2 evolved in TPD after large exposures of butadiene is 0.44 ML H2, corresponding to the complete decomposition of 0.15- ML C4H6. This value was determined by comparison of this H2 TPD peak area to a reference H2 TPD spectrum obtained after ethylene (C2H4) exposures on P t( lll) at 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 300 K to produce 0.25-ML ethylidyne (CCH3)[31], in which complete decomposition produces 0.375-ML H2, i.e. 0.75 ML H. Since complete dehydrogenation is the only reaction pathway for chemisorbed C4H6, the saturation coverage of C4H6 on Pt(l 11) is 0.15 ML. £ C O 10 < 7 > £Z 0> "c 100 200 300 400 500 600 700 800 T e m p e r a t u r e ( K ) Figure 4.1 Butadiene (C4FL) TPD spectra after 1,3-butadiene exposures on Pt(l 11) at 100 K. 130 P t ( 1 1 1 ) 1,3-butadiene 0.24 L 0.18 L 0.12 L 0.057 L 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H0 TPD 398 370 Pt(1 11) 339 1,3-butadiene 585 0.24 L = 3 E co CM 0.18 L 0.12 L 0.057 L background 100 200 300 400 500 600 700 800 Tem perature (K) Figure 4.2 H2 TPD spectra after 1,3-butadiene exposures on Pt(l 11) at 100 K. 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In Figure 4.3, a series TPD spectra are shown for C4H6 following 1,3- butadiene dosing on the (2x2) surface alloy at 100 K. Butadiene from the chemisorbed layer desorbs at 341 K at low coverage and this shifts to 330 K at monolayer coverage. Desorption from the physisorbed layer is observed at 135 K at high coverage. Assuming first-order desorption kinetics with a preexponential factor of 101 3 s"1 , the Redhead method [32] was used to estimate a desorption activation energy E & of 88 kJ/mol for C4H6 adsorbed on the (2x2) alloy at low coverage. In Figure 4.4, a series of C4H6 TPD spectra are shown following 1,3- butadiene dosing on the V3 alloy at 100 K. Butadiene from the chemisorbed layer desorbed at 292 K at low coverage and this shifted to 275 K at monolayer coverage. Desorption from a physisorbed layer was observed at 121 and 145 K at high coverages. A value for E < \ of 75 kJ/mol on the V3 alloy was obtained by Redhead analysis using the peak at low coverage. We note that the desorption peak width of chemisorbed butadiene on the V3 alloy is appreciably narrower than that on the (2 x2 ) alloy. This might mean that butadiene chemisorbs on the (2x2) alloy in more than one bonding configuration. H2 evolution shown in Figure 4.5 monitors butadiene decomposition on P t( lll) and the (2x2) and V3 alloys at monolayer coverage. In contrast to the complete dehydrogenation of butadiene on Pt(l 11), no significant H2 desorption was observed (or at any butadiene coverage used in these experiments) on the two S n /P t(lll) surface alloys. The small H2 desorption features are mostly due to 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cracking fractions from butadiene desorption and the residual yield is almost at the same level as that in “background” TPD experiments where no exposure to butadiene was given. This is consistent with AES which detected no carbon on the two alloys after any TPD experiment. Therefore, butadiene thermal decomposition was completely suppressed under these conditions upon forming the two alloys. 1,3- Butadiene reversibly adsorbs and desorbs on these two Sn/Pt(l 11) surface alloys. 135 (2 x2 ) alloy Z3 E T O ■ 't f - LO 330 341 1,3-butadiene 0.24 L c 0.18 L 0.12 L 0.076 L 0.057 L 500 700 800 100 200 300 400 600 T e m p e ra tu re (K) Figure 4.3 Butadiene (C4H6) TPD spectra after 1,3-butadiene exposures on the (2x2)- Sn/Pt(l 11) alloy at 100 K. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121 a/3 alloy 275 3 £ CO N T LO 292 145 ,3-butadiene >» ■+— < co c C D 0.36 L c 0.24 L 0.18 L 0.114 L 0.057 L 800 400 500 600 700 200 300 100 Tem perature (K) Figure 4.4 Butadiene (C4H6) TPD spectra after 1,3-butadiene exposures on the (V3xV3)R30°-Sn/Pt(l 11) alloy at 100 K. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 .1 8 L 1 ,3 - b u ta d ie n e 13 E CD CM P t(1 1 1 ) CO a ( 2 x 2 ) a llo y V3 a llo y 100 200 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 Temperature (K) Figure 4.5 Comparison of H2 desorption spectra resulting from 1,3-butadiene decomposition on Pt(l 11) and the (2x2)-Sn/Pt(l 11) and (V3xV3)R30°-Sn/Pt(111) alloys. Using the relationship between C4H6 TPD peak area and butadiene coverage deduced from the data in Figures 1 and 2, as described in Appendix 1, we can calculate the unknown butadiene monolayer coverage on each alloy by using the C4H6 TPD peak area in spectra from the chemisorbed monolayer on the two alloys. This approach gives values of 0.15 and 0.17 ML on the (2x2) and a/3 alloy, respectively. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.4 Discussion At 300 K, 1,3-butadiene is believed to chemisorb on P t( lll) in a tetra-a- bonding configuration [4, 23, 24], without any dehydrogenation of the molecule [4]. Figure 4.2 shows that no H2 evolution occurs before 300 K, which is an experimental result consistent with that reported by Avery et al. [4], An older theoretical study by Baetzold [25] predicted that 1,3-butadiene would decompose via loss of two terminal hydrogens to form a (CH)4 metallacycle, but the data in Figure 2 shows that this decomposition mechanism is not correct. At low exposure (0.057-ML C4H6), the H2 desorption peak area ratio at 339 K to 585 K is larger than 4:2, rather than the 2:4 ratio predicted by the theory. The presence of Sn in the surface layer of P t( lll) greatly weakened the interaction between 1,3-butadiene and the alloy surface. Increasing the Sn concentration from 25% to 33% leads to a decrease in the desorption activation energy Ed from 88 to 75 kJ/mol. This is also a good estimate of the adsorption energy in this case where there is no appreciable activation energy barrier to adsorption. This effect of Sn on the adsorption energy has been observed for alkenes [26-28] and many other molecules as well. The adsorption energy of 1,3-butadiene on the (2x2) and V3 alloys is roughly 1.5 times that of ethylene [26] and 1-butene [27] on the same alloy. Ethylene chemisorbs on P t( lll) and the two S n /P t(lll) alloys via di-a- bonding, but the carbon atoms in chemisorbed ethylene are less rehybridized toward sp3 with increasing Sn in the surface layer [26]. A comparison of adsorption energies 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between 1,3-butadiene and 1-butene or ethylene suggests that all four carbon atoms in 1,3-butadiene should have strong interactions with the alloy surfaces. We propose that 1,3-butadiene chemisorbs on the two S n /P t(lll) alloys in a tetra-a-bonding configuration, as on P t(lll), but with the carbon atoms less sp3 -rehybridized than that on P t(lll). The presence of Sn in the surface layer of P t( lll) has no effect, however, on the chemisorbed monolayer coverage of 1,3-butadiene. This coverage is identical on Pt(l 11) and the (2x2) alloy, and even slightly higher on the a/3 alloy. Such behavior has also been observed in studies of other alkene and alkane molecules [27, 28, 33]. This is probably due to a small relaxation in the bonding geometry/site requirements that result from slightly weaker bonding interactions between the chemisorbed molecule and the a/3 alloy surface. All of the butene isomers, along with propene, undergo an appreciable amount of decomposition (7-10%) on the (2x2) alloy. Even though the adsorption energy of 1,3-butadiene is 1.5 times that of 1-butene or 2-butene [27], the decomposition of 1,3-butadiene is completely inhibited on the (2x2) alloy, like that observed for ethylene [26]. This result which is suprising at first glance can be explained simply by the presence of relatively weak allylic |3-CH bonds in propene and butene. Only vinylic CH bonds exist in ethylene and 1,3-butadiene, and these bonds have a higher bond dissociation energy than allylic p-CH bonds [34], So, these vinylic CH bonds are appreciably less reactive on P t( lll) and these alloys. The 1,4- di-a-bonded configuration of 1,3-butadiene on P t( lll) proposed by Bertolini et al. 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [14,15] may not be appropriate, particularly on the (2x2) alloy, because this configuration would create allylic P-CH bonds on the end carbon atoms and this would pretty clearly lead to decomposition on the surface. Like all of the small alkene molecules, 1,3-butadiene reversibly adsorbs and desorbs on the V3 alloy. The origin of this decreased reactivity compared to that on Pt(l 11) was explained originally for ethylene as an increase in the activation energy for C-H bond scission, E*, on the alloys. However, this was never quite clear because the ethylene adsorption energy also decreased on these two S n /P t(lll) alloys, and this could also lead to increased desorption and lack of reaction under UHV conditions. Now, because of the much larger desorption activation energy for 1,3-butadiene compared to ethylene on these two S n /P t(lll) alloys, we can exclude this latter rationale which uses the decrease in desorption activation energy by alloying Sn to P t(lll). We now show that ethylene decomposition is inhibited because of the increase in E*, the C-H bond breaking barrier, on the alloys. Also, now we can place $ a much better lower limit for E of vinylic CH bonds of 88 and 75 kJ/mol on the (2x2) and V3 alloys, respectively. The increase in this barrier may be caused by site- blocking or by electronic modifications of the Pt( 111) surface by alloying with Sn. In the (2x2) structure, pure-Pt three-fold reactive sites are present, but no adjacent pure- Pt three-fold sites exist. All pure-Pt three-fold sites are eliminated in the (V3xV3)R30° structure, and only two-fold pure Pt sites are present. Alloying also changes the local electronic structure at Pt sites. Calculations by Pick [35] and 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Delbecq and Sautet [36] consider that a shift of the Pt d-band center resulting from charge transfer from Sn to Pt decreases the bond strength between adsorbed molecules and Sn-Pt alloys. Such changes could also leads to the formation of a significant activation energy barriers for bond dissociation reactions of adsorbed molecules. 4.5 Conclusions On P t( lll), chemisorbed 1,3-butadiene completely decomposed to H2 and surface carbon upon heating. This decomposition pathway is completely inhibited on two S n /P t(lll) surface alloys. Alloyed Sn also reduces the chemisorption bond strength of 1,3-butadiene compared to that on P t(lll). The desorption activation energy Ed for 1,3-butadiene on the (2 x2 ) and (V3xV3)R30°-Sn/Pt(l 11) alloys is 88 and 75 kJ/mol, respectively. However, alloyed Sn has no effect on the coverage of 1,3-butadiene in the chemisorbed monolayer, which is 0.15, 0.15, and 0.17 ML on P t( lll) and the (2 x2) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys, respectively. We propose that the lack of reactivity of ethylene and 1,3-butadiene on these alloys is due to lack of allylic P-CH bonds. Because Ed for 1,3-butadiene on these alloys is 1.5 times that of ethylene or 1-butene on the same surface, we can now set a more useful lower limit for the vinylic C-H bond breaking barrier E* on these alloys of 88 kJ/mol. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.6 Appendix 1. Herein we derive the relationship between butadiene TPD peak area and butadiene coverage by using data from Figure 1 in order to calculate the monolayer coverage of butadiene on the alloys from the areas of the TPD peaks obtained from the two alloy surfaces. Surface coverage 9 is related to gas exposure s by 9=Ss, where S is the sticking coefficient. Assuming that the butadiene sticking coefficient is a constant value over the monolayer and multilayer regime on P t( lll) at 100 K, as has been shown for similar molecules many times before, then there is a simple relationship 02/0 i=£2/£i between two different coverages produced by two different exposures. TPD peak areas are also related to surface coverages by a factor A that is unknown but is a constant and has units of M L'1 . This factor for butadiene can be determined from the butadiene TPD curves from P t( lll) shown in Figure 1, along with information that the monolayer coverage of butadiene 0m o n o is 0.15 ML, as derived from the H2 TPD data of Figure 2. Butadiene exposures of 0.18 and 0.24 L produce a butadiene coverage on the surface that exceeds one monolayer, and the total amount is given by 0m u iti + © m o n o - Butadiene desorbs from the two multilayers produced by these two exposures to yield desorption peak areas of 7842 and 53912, respectively. The value of 0m u iti is given by 7842/A and 53912/A for exposures of 0.18 and 0.24 L, respectively. Substitution into the simple relationship between coverage and exposure given above yields the following equation 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53912 n u + 0.i5 0 2 4 A ----------------= v ^ t ™ “ + 0.!5 °’18 ‘ X that can now be solved for X A value o fX = 869120 ML' 1 is obtained. All of the butadiene in the monolayer desorbed molecularly on the (2x2) and V3 alloy to give butadiene desorption peak areas of 129931 and 150978, respectively. These peak areas are converted to monolayer coverages by dividing by X to give 0.15 and 0.17 ML on the (2x2) and V3 alloy, respectively. 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.7 References 1. M. Salmeron and G. A Somorjai, J. Phys. Chem. 8 6 (1982) 341. 2. R.J. Koestner, J.C. Frost, P.C. Stair, M A. Van Hove, and G.A Somorjai, Surf. Sci. 116(1982) 85. 3. R.J. Koestner, M.A. Van Hove, and G.A Somorjai, J. Phy. Chem. 87 (1983) 203. 4. N.R. Avery, N. Sheppard, Proc. R. Soc. Lond. A 405 (1986) 1. 5. K.M. Ogle, J.R. Creighton, S. Akhter, J.M. White, Surf. Sci. 169 (1986) 246. 6 . N.R. Avery Langmuir 4 (1988) 445. 7. W.H. Weinberg, H.A. Deans and R.P. Merill Surf Sci. 41 (1974) 312. 8 . L.L. Kesmodel, P.C. Stair, R.C. Baetzold and G.A. Somorjai Phys. Rev. Lett. 36 (1976) 1316. 9. L.L. Kesmodel, L.H. Dubois and G.A. Somorjai J. Chem. Phys. 70 (1979) 2180. 10. H. Ibach and S. Lehwald J. Vac. Sci. Technol. 15 (1978) 407. 11. P.S. Cremmer, X. Su, Y.R. Chen and G.A. Somorjai J. Phys. Chem. B 101 (1997) 6474. 12. P. Skinner, M. Howard, I. Oxton, S. Kettle, D. Powell and N. Sheppard J. Chem. Soc. Faraday Trans. 15 (1978) 407. 13. O. Nakagoe, N. Takagi, and Y. Matsumoto Surf. Sci. 514 (2002) 414. 14. J.C. Bertolini, A. Cassuto, Y. Jugnet, J. Massardier, B. Tardy, and G. 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tourillon, Surf Sci. 349 (1996) 8 8 . 15. G. Tourillon, A. Cassuto, Y. Jugnet, J. Massardier, J.C. Bertolini, J. Chem. Soc., Faraday Tran. 92 (1996) 92. 16. J.P. Boitiaux, P. Cosyns, M. Derrien and G. Leger. Hydrocarbon Proc. 64 (1985) 51. 17. M. Derrien. In: L. Cerny, Editor, Studies in Surface Science and Catalysis 27, Elsevier, Amsterdam (1986), p. 313. 18. H. Arnold, F. Dobert and J. Gaube, In Handbook o f Heterogeneous Catalysis G. Ertl, H. Knozinger and J. Weitkamp, Eds, Wiley-YCH: Weiheim, Germany, 1997. 19. A.J. Bates, Z.K. Leszcynski, J.J. Phillipson, P.B. Wells and G.R. Wilson, J. Chem. Soc. A (London) (1970) 2435. 20. R.D. Cortright and J. A. Dumesic, Appl. Catal. A 129 (1995) 101. 21. R.D. Cortright, P.E. Levin, J. A. Dumesic Ind. Eng. Chem. Res. 37 (1998) 1717. 22. J.M. Hill, R.D. Cortright, J. A. Dumesic, Appl. Catal. A-Gen. 168 (1998) 9. 23. F. Mittendorfer, C. Thomazeau, P. Raybaud, and H. Toulhoat, J. Phy. Chem. B. 107(2003)12287. 24. A. Valcarcel, A. Clotet, J.M. Ricart, F. Delbecq, P. Sautet, Surf. Sci. 549 (2004)121. 25. R.C. Baetzold, Langmuir 3 (1987) 189. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26. M.T. Paffett, S.C. Gebhard, R.G. Windham and B.E. Koel, Surf. Sci. 223 (1989) 449. 27. Y. Tsai and B.E. Koel, J. Phy. Chem. B 101 (1997) 1895. 28. Y. Tsai, C. Xu and B.E. Koel, Surf. Sci. 385 (1997) 37. 29. H. Zhao, J. Kim, B. E. Koel, Surf. Sci., 538 (2003) 147. 30. S.H. Overbury, D.R. Mullins, M.T. Paffett, B.E. Koel, Surf. Sci. 254 (1991) 45. 31. R.G. Windham, M.E. Bartram, B.E. Koel, J. Phy. Chem. 92 (1988) 2862. 32. P. A. Redhead, Vacuum 12 (1962) 203. 33. C. Xu, B.E. Koel, and M.T. Paffett, Langmuir 10 (1994) 166. 34. J. Berkowitz, G. Ellison, and D. Gutman, J. Phy. Chem. 98 (1994) 2774. 35. S. Pick, Surf. Sci. 436 (1999) 220. 36. F. Delbecq and P. Sautet, J. Catal. 220 (2003) 115. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 5 Reactivity of Chemisorbed Ethyl Groups on a (V3xV3)R30°- Sn/Pt(l 11) Surface Alloy A bstract Surface-bound alkyl groups have often been implicated as important intermediates in hydrocarbon conversion reactions over transition metal catalysts. In particular, chemisorbed ethyl (CH3CH2) groups are important intermediates in catalytic ethylene hydrogenation. In order to probe the thermal stability and reactivity of such species over Pt-Sn alloys, which model aspects of commercial bimetallic Pt-Sn catalysts for catalytic hydrogenation, we synthesized chemisorbed ethyl groups by the addition reaction of incident gas-phase H atoms with adsorbed ethylene (C2H4) on a well-defined (V3xV3)R30°-Sn/Pt(lll) surface alloy. Ethyl groups are stable on this alloy until 376 K where they react to evolve ethane (C2H6), ethylene, and H2 in temperature programmed desorption (TPD). The activation energy for ethyl dehydrogenation or C-H bond-cleavage in this species on the (W3xV3)R30°-Sn/Pt(l 11) surface alloy is estimated to be T ^d eh y d * 97 kJ/mol, which is twice that reported on P t(lll). In addition, a lower limit of /%dr* > 70 kJ/mol is placed on the activation energy for ethyl hydrogenation on this alloy surface. The thermal stability and reactivity of ethyl groups that we have measured on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy help to explain reactions of alkyl intermediates in alkane dehydrogenation on supported Pt-Sn catalysts. 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.1 Introduction Simple alkyl species have been considered for a long time to be intermediates in the conversion of saturated hydrocarbons into more economically attractive products over transition metal catalysts [1,2]. However, our knowledge of the chemistry of these species from surface science type experiments does not reflect their importance. The production of simple species such as ethyl (C2H 5) from their fully saturated precursors is constrained heavily under UHY conditions by the significant energetic barriers to dissociative alkane adsorption. Most commonly, this problem has been surmounted by using thermal, photochemical or electron-induced dissociation (EID) of ethyl halides (C2H5X, X=C1, Br, I) adsorbed at low temperatures on metal single crystals [3,4]. In this manner, studies of ethyl chemistry have been reported on copper [5-9], silver [10, 11], nickel [12, 13], platinum [14-17], and rhodium [18], A disadvantage of this approach is the introduction of coadsorbed halogen adatoms on the surface. Metal-ethyl compounds, like triethylbismuth [19], can also be used as thermal or photochemical precursors, but the problem of co adsorbed metal adatoms remains. Several other approaches are available that eliminate any doubts about the perturbing effects caused by co-adsorbed adatoms. EID of physisorbed hydrocarbons is useful for synthesis of symmetrical alkyl and cycloalkyl species [20-26], but condensation of ethane requires a LHe-cooled probe. In heroic efforts, supersonic molecular beams [27] and hyperthermal collisions [28] have been used to cleanly create adsorbed ethyl species on P t(lll). Perhaps the 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. most versatile approach is one pioneered by Bent and coworkers [29-31] in which incident gas-phase H atoms undergo radical addition to the 7 1-bond in weakly adsorbed ethylene to form ethyl groups. Subsequent addition from gas-phase H atoms directly to the surface-bound alkyl is orders of magnitude slower (and any ethane formed would desorb immediately) and so the reaction is self-limiting and clean. P t( lll) is the worst substrate for using this approach, because it nearly completely rehybridizes ethylene to form a di-a-bonded species containing essentially a C-C single bond. However, ethylene chemisorbs on most other metal surfaces retaining more double-bond character and radical addition may occur with a useful, higher probability. We decided to try this approach using Pt-Sn alloy substrates. Bimetallic Sn/Pt supported catalysts are commercially important for suppressing the hydrogenolysis that can occur on pure-Pt catalysts [32, 33]. In surface science studies, we have used well-characterized Sn/Pt(l 11) surface alloys as models for aspects of this catalysis involving Pt-Sn alloy phases. While we have some information on the reactivity of cycloalkyl species on these alloys [23], no information is available for ethyl groups. In related studies, we were able to use CH3I as a precursor and thermal decomposition to form methyl (CH3) groups on a (2x2)-Sn/Pt(lll) alloy, but CH3I was reversibly adsorbed on the (V3xV3)R30°- S n /P t(lll) alloy with no decomposition during heating under UHV conditions [34], This latter observation and the relatively high temperature for activation of the C-I 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bond on the (2x2)-Sn/Pt(l 11) alloy prompted us to turn to the use of methyl radical beams, from a pyrolytic source using azomethane, to prepare methyl groups on these surfaces. However, no such mechanism can be used to make clean beams of ethyl radicals. Herein, we report on the synthesis of chemisorbed ethyl groups on the (V3x a/3)R30° Sn/Pt(l 11) alloy by utilizing the radical addition reaction between gas- phase H atoms and adsorbed ethylene. We investigated the thermal chemistry of chemisorbed ethyl groups on this Pt-Sn alloy surface without perturbation of other coadsorbed species. 5.2 Experimental methods Experiments were performed in a three-level UHV chamber as described earlier [35]. The Pt(lll) crystal (Atomergic; 10 mm dia., 1.5 mm thick) was prepared by 1-keV Ar+ -ion sputtering and oxygen treatments (5x1 O ' 7 torr O2, 900 K, 2 min) to give a clean spectrum using Auger electron spectroscopy (AES) and a sharp (lxl) pattern in low energy electron diffraction (LEED). The (V3xV3)R30°-Sn/Pt(l 11) surface alloy was prepared by evaporating one monolayer of Sn onto the Pt(lll) crystal surface and subsequently annealing the sample to 830 K for 20 s. Sn on this surface, prepared as above, is substitutionally incorporated into this surface layer at Pt-atom positions to form an alloy with 9 s n = 0 .3 3 , corresponding to a Pt2Sn surface. Because of the larger size of Sn than Pt, 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and the importance of relieving stress in the surface layer, Sn atoms protrude 0.02 nm above the surface-Pt plane for this surface alloy [36, 37]. All pure-Pt three-fold sites are eliminated in the (V3xV3)R30° structure. For brevity throughout this paper, we will refer to the (V3xV3)R30°-Sn/Pt(l 11) surface alloy as the V3 alloy. A Pt-tube doser was constructed, based on the design of Engel and Rieder [38] as a pyrolytic source of hydrogen atoms. The principal component is a bent Pt tube (1mm O.D., 0.8 mm I.D.) into which a hole of 0.1mm diameter was mechanically drilled. The tube was resistively heated to 1275°C, and water-cooling kept the adjacent Cu block cold. The temperature of the Pt tube was directly measured by an optical pyrometer that was calibrated by the temperature of the Pt(111) crystal sample, as measured by a Cr/Al thermocouple. The estimated relative accuracy of the pyrometer reading was ±5°C. The flux of H atoms obtained from this source with the Pt tube at 800°C and the UHV chamber pressure of 5x1 O '8 Torr was 3xl0 1 3 atoms cm'2-s'1 . This value was obtained by assuming that the initial sticking coefficient of H atoms on Pt(lll) at 100 K was unity and using the hydrogen coverage produced from the well-known decomposition of ethylene on Pt(lll) to give an absolute calibration for the H2 yield in TPD [39]. The normal operating temperature of the Pt-tube doser during these experiments was 800°C. This corresponds to an energy (kT) of -0.09 eV for the incident H atoms, although it is known that similarly designed thermal atom sources do not necessarily produce atoms in a Boltzmann energy distribution [40]. 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H2 (Matheson; 99.99%) was introduced via a variable leak valve (Granville- Phillips) into the Pt-tube doser after passing through a liquid-nitrogen cooled, U-tube trap. C2H4 (Matheson; 99.99%) was introduced via a microcapillary array doser connected to the gas line through a leak valve. All of the exposures reported here are given simply in terms of the pressure measured by the ion gauge in the UHV chamber. No attempt was made to correct for the flux enhancement of the doser or ion gauge sensitivity. Mass spectrometry performed in the UHV chamber showed no appreciable concentration of impurities in the source gases. For all TPD experiments, the heating rate was 3.6 K/s and all exposures were given with the surface temperature at 100 K. AES measurements were made with a double-pass cylindrical mirror analyzer (CMA). The electron gun was operated at 3- keV beam energy and 1.5-pA beam current. Coverages 0j reported in this paper are referenced to the surface atom density of Pt(l 11) such that 0pt =1.0 ML is defined as 1.505 x l0 15cnf2. 5.3 Results Ethylene chemisorbs on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy as a di-a- bonded species with a coverage of 0.27 ML in the monolayer and desorbs at 183 K in TPD [41, 42], No decomposition to produce H2 nor hydrogenation to produce ethane occurs in TPD. In our experiments reported herein, an ethylene exposure was given to form a monolayer (saturation) coverage on the V 3 alloy at 100 K, and this 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. monolayer was then exposed to incident gas-phase H atoms for increasing amounts of time. TPD experiments were then carried out. The mass spectrometer was multiplexed to monitor many different signals simultaneously, including those at 58, 56, 54, 30, 28, 26 and 2 amu. Only ethane (30 amu), ethylene (28 amu) and H2 (2 amu) were identified as desorption products. In particular, no C4 products that might arise from C-C coupling reactions were detected. Ethylene desorption spectra are shown in Figure 5.1. The bottom curve reproduces previous reports [41, 42] for ethylene adsorption and desorption in the absence of H-atom reactions. TPD spectra show that with increasing amounts of subsequent H atom exposure, the ethylene desorption peak below 200 K, corresponding to chemisorbed ethylene, decreases in intensity while a new peak at 380 K appears and grows in size. By monitoring all of the appropriate mass signals, we conclude that this high-temperature peak is due to the combination of ethylene desorption from ethyl decomposition and a 28-amu cracking fraction arising from ethane desorption. Figure 5.2 shows ethane desorption spectra in these experiments. In the absence of H-atom exposure, the bottom curve shows that no ethane desorption occurs after ethylene adsorption on the V3 alloy. Ethane desorption in a peak at 376 K appears and then increases in size with increasing H-atom exposures. This indicates that the amount of adsorbed ethylene that is converted to ethane increases with increasing exposures to gas-phase H atoms. This is also supported by the 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ethylene TPD spectra in Figure 5.1, where the ethylene desorption peak below 200 K corresponding to chemisorbed ethylene decreased in intensity with increased H-atom exposure time. 3 7 6 TPD H a t o m e x p o s u r e t i m e Z3 E ro oo cxi 3 0 m i n t o 4 0 0 s c r a > 1 0 0 s 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 5 0 0 T em perature (K) Figure 5.1 Ethylene (C2H4) TPD spectra after increasing gas-phase H-atom exposures on a chemisorbed ethylene monolayer on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy. 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 376 Z3 E 05 O CO H atom exposure time to a CD -f— » c 30 min 400 s 100 s 100 150 200 250 300 350 400 450 500 Temperature (K) Figure 5.2 Ethane (C2H6) TPD spectra after increasing gas-phase H-atom exposures on a chemisorbed ethylene monolayer on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy. 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although there is no information on the desorption temperature of ethane on the a/3 alloy, it must be quite low, below 1 0 0 K, and ethane adsorption must be very weak. Adsorbed ^-butane desorbs at 1 5 3 K on the V 3 alloy [ 4 3 ] , and butane should interact about twice as strongly as ethane with the V3 alloy. This means that ethane desorption at 3 7 6 K in Figure 5 . 2 must be reaction-rate limited, i.e., ethane desorbs promptly as soon as it is formed at the surface and the ethane desorption peak reveals information on the kinetics of ethane production at the surface. The only possible precursor to produce ethane at 3 7 6 K in TPD is adsorbed ethyl groups. These species are generated by the reaction between adsorbed ethylene and gas- phase H atoms. Figure 5.3 shows H2 desorption spectra in these experiments. In the absence of H- atom exposure, the bottom curve shows that little H2 desorption occurs after ethylene adsorption on the a/3 alloy. Ethylene is nearly reversibly adsorbed on this alloy. The peak below 200 K arises primarily from a cracking fraction of desorbed ethylene. With increasing H-atom exposures, an H2 desorption peak appears at 275K and a peak at 380 K appears and then increases in size. The H2 desorption peak between 250 and 300 K in Figure 5.3 is from recombination of H adatoms, and this shows that incident H atoms can still chemisorb on the a/3 alloy, even though it is covered with coadsorbed ethylene and ethyl species. There are two possible origins of the H2 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. peak at 380 K in addition to the cracking fraction of desorbed hydrocarbons (ethane and ethylene) that are products from ethyl dehydrogenation. h U T P D 376 H atom exposure tim e 30 min Z 3 E a s C M 4 0 0 s c z a> cz 100 s 2 00 2 5 0 3 00 3 50 4 0 0 4 5 0 500 100 150 T e m p e r a t u r e ( K ) Figure 5.3 H2 TPD spectra after increasing gas-phase H-atom exposures on a chemisorbed ethylene monolayer on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy. Adsorbed hydrogen adatoms recombine and desorb as H2 below 300 K on the V 3 alloy [44] and so this excludes a pathway for coadsorbed H adatoms and ethyl to combine to produce ethane at 376 K. Adsorbed ethyl groups must obtain another 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. additional H atom from other ethyl groups to form ethane. Adsorbed ethyl species can lose one H through (3-H elimination to produce ethylene and hydrogen adatoms, and this is a common pathway on most transition metal surfaces [4], This is the most likely mechanism for the origin of the H2 peak at 380 K. The other mechanism is for a strict disproportionation of two adjacent ethyl groups to directly yield ethane and ethylene. Ethylene does not further decompose on the V 3 alloy. The peak at 380 K in Figure 5.3 has contributions from both ethane and ethylene produced by ethyl decomposition. No significant H2 desorption at temperatures higher than 380 K was detected, which indicates that full decomposition of ethyl does not occur to any appreciable extent on the V3 alloy. Consistent with these H2 TPD results, no carbon was detected by AES following TPD. Of note, no ethyl-coupling product was observed in TPD experiments by monitoring the signal at 58 amu (C4FI10). 5.4 Discussion There is no reaction observed between coadsorbed H adatoms and ethylene molecules on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy [45], The nature of gas- phase atomic hydrogen (a radical) increased the reactivity compared to surface- bound atomic hydrogen for reaction with adsorbed ethylene on the V3 alloy. Ethane formation shown in Figure 5.2 indicates that adsorbed ethyl groups form on the V3 alloy at 100 K by radical addition of FI atoms and adsorbed ethylene. Adsorbed ethyl 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. species were also created in this fashion on Cu(lll) [29, 30] and Cu(100) [31]. The estimated cross section for hydrogen addition to ethylene on Cu(l 11) was 18 A 2 [29]. Gas-phase H atoms have also been shown to react with di-a-bonded cyclohexene to produce adsorbed cyclohexyl species on Ni(100) [46], even though 7 i-bonded alkenes should have larger cross sections for radical addition reactions than di-a-bonded alkenes. These H-radical addition reactions occur through an Eley-Rideal mechanism [4] utilizing the 218 kJ/mol of potential energy per mole of H atoms to induce reactions. Our results extend the illustrations of the utility of this process to cleanly create adsorbed alkyl species at low temperatures and thus enable studies of their subsequent chemistry. The thermal stability and chemistry of adsorbed ethyl species on this Pt-Sn alloy is dramatically different from that on Pt(lll). Adsorbed ethyl species created by dissociation of ethyl iodide (CH3CH2I) on Pt(lll) immediately undergo a dehydrogenation step at 200 K to form chemisorbed ethylene (C2H4) and hydrogen adatoms via fi-H elimination [15]. Although alkyl reaction to form alkenes and alkanes has been reported on Cu [6], Pt [1, 3, 14], and A1 surfaces [47], these reactions are not commonly thought of as strict disproportionation reactions because the alkene is the dominant product. Coupling and disproportionation products were detected below 200 K upon heating adsorbed alkyl iodides on Cu(l 11) [7], but alkyl radicals and not surface-bound alkyl groups were proven to be the precursors to these 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reactions. Surface-bound alkyl groups only produce alkenes and hydrogen on Cu(l 11) as a result of P-H elimination. Alloying Sn into Pt(lll) greatly stabilizes adsorbed ethyl groups, and these can exist on the V3 alloy until 376 K. As discussed above, strict disproportionation reactions of adsorbed ethyl groups to produce ethane and ethylene is not observed on most metal surfaces. Disproportionation usually happens in the gas phase and in solution and this is accompanied by coupling reactions [7]. Since no coupling reactions were observed on the V3 alloy, the disproportionation reaction may not be a major reaction pathway to produce ethane and ethylene. A reaction mechanism for dehydrogenation of adsorbed ethyl groups on the V3 alloy can be proposed as follows. As the first step, surface-bound ethyl dehydrogenates at 376 K to produce ethylene and H adatoms through P-H elimination. In the second step, some of the H adatoms generated in the first step immediately hydrogenate adsorbed ethyl groups to release ethane into the gas phase while the rest of the H adatoms recombine to liberate. H2 from the surface as a product. The first step of P-H elimination is the rate-determining step and desorption of all products are reaction-rate limited. Ethyl groups are more stable on the V 3 alloy than any other hydrocarbon molecule reported to date, but have a stability similar to that of cyclohexyl groups on this surface [23]. The P-H elimination reaction of adsorbed cyclohexyl to produce coadsorbed cyclohexene and H adatoms was also observed on the V 3 alloy at 345 K 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [23], but no cyclohexane was produced by hydrogenation of cyclohexyl. The reason for this apparent difference in hydrogenation reactivity between ethyl and cyclohexyl groups on the V3 alloy is not clear, but it may be due to the low cyclohexyl coverage in those experiment or the richer chemistry of cyclohexyl ligands, e.g. isomerization and/or H-shifts. The TPD spectra in Figure 5.2 provide a good estimation for the C- H bond breaking barrier jE dehyd* in the dehydrogenation of surface ethyl groups. We calculate that .E dehyd* - 97 kJ/mol by assuming first-order kinetics and a 1-3 t preexponential factor of 10 s' . This barrier is almost twice that on Pt(lll) [15]. This huge increase in the C-H bond breaking barrier is due either to the general decrease in the LDOS at the Fermi energy at Pt atoms caused by alloying Sn to Pt( 111) [48] or perhaps to a loss of the specific surface electronic structure at pure-Pt 3-fold sites due to the loss of these sites on this alloy surface. These results have important implications to understanding the surface science of catalysis. SFG studies of ethylene and propene hydrogenation on Pt(l 11) suggest that ethyl and propyl groups are reaction intermediates [49, 50]. Because supported Pt-Sn catalysts are not active as catalysts in alkene hydrogenation [51, 52], the poor activity of supported Pt-Sn catalysts may result from an increased barrier to form alkyl groups and/or an increased barrier to hydrogenation of these alkyl intermediates. Our studies have shown that coadsorbed H does not react with surface ethyl groups and FI simply recombines and desorbs as H2 from the V3 alloy at 275 K with an activation energy E & of 70 kJ/mol. This places a lower limit on the 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hydrogenation activation energy barrier £ilydr* = 70 kJ/mol for surface surface-bound ethyl groups on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy. In contrast, supported Pt-Sn catalysts are widely used in selective dehydrogenation of alkanes to alkenes [53-56], Cortright et al. [57] have proposed that dehydrogenation of isobutane over Pt/Sn/SiC> 2 catalysts is through a Horiuti- Polanyi mechanism and the first step to form adsorbed isobutyl groups is the slowest and rate-determining step. They measured a barrier of £hy dr* = 85 kJ/mol to hydrogenate isobutyl with adsorbed H, and there must be a barrier of at least -E d eh y d * = 118 kJ/mol to dehydrogenate isobutyl to form isobutene gas. Their studies also indicate that alkyl is a stable intermediate on these supported Pt-Sn catalysts. Our studies of the thermal stability of ethyl groups and the reactivity of these species with and without coadsorbed H on the V3 Sn-Pt alloy support so far the reaction mechanism and activation energies deduced for elementary steps in the reaction of alkyl species as intermediates in alkane dehydrogenation on supported Pt-Sn catalysts. 5.5 Conclusions Reaction between gas-phase H atoms and adsorbed ethylene produced surface-bound ethyl groups on a (V3xV3)R30°-Sn/Pt(l 11) surface alloy at 100 K. These ethyl groups do not dehydrogenate and are stable on the surface much above room temperature, until 376 K. C-H bond dissociation then produces ethylene and 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hydrogen, and this hydrogen immediately reacts to desorb as H2 and hydrogenate coadsorbed ethyl groups to release ethane into the gas phase. The reaction-rate limited desorption of ethane can be used to provide a good estimate of the activation energy E* for breaking C-H bonds in alkyl species bound on the a/3 alloy. This value of -E dehyd* = 97 kJ/mol is almost twice that on Pt(l 11). Ethyl groups on the V 3 alloy do not react with coadsorbed H and recombination of H adatoms leads to H2 desorption at 275 K. This observation enables us to place a lower limit on the barrier for ethyl hydrogenation on the a /3 alloy of £hydr* > 70 kJ/mol. Our studies of the thermal stability of ethyl groups and the reactivity of these species with and without coadsorbed H on the a/3 Sn-Pt alloy support so far the reaction mechanism and activation energies deduced for elementary steps in the reaction of alkyl species as intermediates in alkane dehydrogenation on supported Pt-Sn catalysts. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.6 References [1] G. C. Bond, Heterogeneous Catalysis, 2n d ed. (Clarendon, Oxford, 1987), Ch. 8,9. [2] G. A. Somorjai, Introduction to Surface Chemistry and Catalysis (Wiley, New York, 1994), Ch.6 , 7. [3] F. Zaera, Acc. Chem. Res. 25 (1992) 260. [4] B. E. Bent, Chem. Rev. 96 (1996) 1361. [5] J.-L. Lin and B. E. Bent, Chem. Phys. Lett. 194 (1992) 208. [6 ] J.-L. Lin and B. E. Bent, J. Phys. Chem. 96 (1992) 8529. [7] J.-L. Lin and B. E. Bent, J. Am. Chem. Soc. 115 (1993) 6943. [8] J.-L. Lin, C.-M. Chiang, C. J. Jenks, M. X. Yang, T. H. Wentzlaff and B. E. Bent, J. Catal. 147 (1994) 250. [9] J.-L. Lin, A. Y. Teplyakov and B. E. Bent, J. Phys. Chem. 100 (1996) 10721. [10] Z.-M. Liu, X.-L. Zhou and J. M. White, Chem. Phys. Lett., 615 (1992) 615. [11] X.-L. Zhou, P. M. Blass, B. E. Koel and J. M. White, Surf. Sci. 271 (1992) 453. [12] S. Tjandra and F. Zaera, Surf. Sci. 289 (1993) 255. [13] S. Tjandra and F. Zaera, Surf. Sci. 140 (1995) 140. [14] K. G. Lloyd, B. Roop, A. Campion and J. M. White, Surf. Sci. 214 (1989) 227. [15] F. Zaera, Surf. Sci. 219 (1989) 453. [16] F. Zaera, J. Phys. Chem. 94 (1990) 8350. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [17] H. Hoffmann, P. R. Griffiths and F. Zaera, Surf. Sci. 262 (1992) 141. [18] F. Solymosi, L. Bugyi and A. Oszko, Langmuir 12 (1996) 4145. [19] M. E. Pansoy-Hjelvik, R. Xu, Q. Gao, K. Weller, F. Feher, and J. C. Hemminger, Surf. Sci. 312 (1994) 97. [20] X. Zhou, P.M. Blass, B.E. Koel, and J.M. White, Surf. Sci. 271 (1992) 427. [21] X. Zhou, P.M. Blass, B.E. Koel, and J.M. White, Surf. Sci. Ill (1992) 452. [22] X. Chen and B.E. Koel, Surf. Sci. 292 (1993) L803. [23] C. Xu, Y. Tsai, and B.E. Koel, J. Phy. Chem. 98 (1994) 585. [24] D. Syomin and B.E. Koel, Surf. Sci. 492 (2001) L693. [25] Y. Tsai and B.E. Koel, J. Phy. Chem. B 101 (1997) 4781. [26] Y. Tsai and B.E. Koel, Langmuir 14 (1998) 1290. [27] H. E. Newell, M. R. S. McCoustra, M. A. Chesters and C. D. L. Cruz, J. Chem. Soc., Faraday Trans., 94 (1998) 3695. [28] D.J. Oakes, H. E. Newell, F.J.M. Rutten, M. R. S. McCoustra, and M. A. Chesters, J. Vac. Sci. Technol. A 14 (1996) 1439. [29] M. Xi and B.E. Bent, J. Vac. Sci. Technol. B 10 (1992) 2440. [30] C. J. Jenks, M. Xi, M.X. Yang, and B.E. Bent, J. Phys Chem. 98 (1994) 2152. [31] M.X. Yang and B.E. Bent, J. Phys Chem. 100 (1996) 822. [32] Z. Karpinski and J. K. Clarke, J. Chem. Soc. Faraday Trans. I, 71 (1975) 893. [33] B. H. Davis, J. Catal. 46 (1977) 378. [34] C. Panja, E.C. Samano, N. A. Saliba, and B.E. Koel, Surf. Sci. 553 (2004) 39. 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [35] H. Zhao, J. Kim, B. E. Koel, Surf. Sci. 538 (2003) 147. [36] S.H. Overbury, D.R. Mullins, M.T. Paffett, B.E. Koel, Surf. Sci. 254 (1991) 45. [37] W.C.A.N. Ceelen, A.W. Denie van der Gon, M.A. Reijme, H.H. Brongersma, I. Spolveri, A. Atrei, and U. Bardi, Surf. Sci. 406 (1998) 264. [38] T. Engel, K.H. Rieder, in: G. Hohler (Ed.), Structural Studies o f Surfaces (Springer, Berlin, 1982) Vol. 91, p. 55. [39] R.G. Wndham, M.E. Bartram, B.E. Koel, J. Phy. Chem. 92 (1988) 2862. [40] C.T. Rettner. J. Chem. Phys. 101 (1994) 1529. [41] M. T. Paffett, S.C Gebhard, R. G. Windham, and B.E. Koel, Surf. Sci. 223 (1989) 449. [42] T. Tsai, C. Xu, and B.E. Koel, Surf Sci. 385 (1997) 37. [43] C. Xu, B.E. Koel, and M.T. Paffett, Langmuir 10 (1994) 166. [44] M.R. Voss, H. Busse, and B. E. Koel, Surf. Sci. 414 (1998) 330. [45] E l. Zhao and B. E. Koel, to be published. [46] K.A. Son and J. L. Gland, J. Phys. Chem. B 101 (1997) 3540. [47] B. E. Bent, R.G. Nuzzo, B.R. Zegarski, L.H. Dubois, J. Am. Chem. Soc. 113 (1991) 1137. [48] S. Pick, Surf. Sci. 436 (1999) 220. [49] P.S. Cremer, B.J. Mcintyre, M. Salmeron, Y.R. Shen, and G.A. Somojai, Catal. Lett. 34(1995) 11. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [50] K.R. McCrea and G.A. Somorjai, J. Mol. Catal. A: Chemical 163 (2000) 43. [51] M. Galvagno, P. Staiti, P. Antonucci, A. Donato, and R. Pietropaolo, J. Chem. Soc, Faraday Trans. 79 (1983) 2605. [52] A. Palazov, Ch. Bonev, D. Shopov, G. Lietz, A. Sarkany and J. Volter, J. Catal. 103 (1987) 249. [53] O.A. Barias, A. Holmen, and E.A. Blekkan Catal. Today 24 (1995) 361. [54] R.D. Cortright, J.M. Hill, and J.A. Dumesic Catal. Today 55 (2000) 213. [55] O.A. Barias, A. Holmen, and E.A. Blekkan J. Catal 158 (1996) 1. [56] J. M. Hill, R.D. Cortright, J.A. Dumesic, Appl. Catal. A 168 (1998) 9. [57] R.D. Cortright, E. Bergene, P. Levin, M. Natal-Santiago, J.A. Dumesic, Stud. Surf. Sci. Catal. 101 (1996) 1185. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 Site-blocking effects of preadsorbed H on Pt(l 11) probed by 1,3-butadiene adsorption and reaction A bstract The influence of hydrogen coadsorption on hydrocarbon chemistry on transition metal surfaces is a key aspect to an improved understanding of catalytic selective hydrogenation. We have investigated the effects of H preadsorption on adsorption and reaction of 1,3-butadiene (H2C=CHCH=CH2, C4H6) on Pt(lll) surfaces by using temperature-programmed desorption (TPD) and Auger electron spectroscopy (AES). Preadsorbed hydrogen adatoms decrease the amount of 1,3- butadiene chemisorbed on the surface and chemisorption is completely blocked by the hydrogen monolayer (saturation) coverage (0h = 0.92 ML). No hydrogenation products of reactions between coadsorbed H adatoms and 1,3-butadiene were observed to desorb in TPD experiments. This is in strong contrast to the copious evolution of ethane (CH3CH3, C2H6) from coadsorbed hydrogen and ethylene (C H 2= C H 2, C2H4) on Pt(lll). Hydrogen adatoms effectively (in a 1:1 stoichiometry) remove sites from interaction with chemisorbed 1,3-butadiene, but do not affect adjacent sites. The adsorption energy of coadsorbed 1,3-butadiene, is not affected by the presence of hydrogen on Pt(lll). The chemisorbed 1,3-butadiene on hydrogen preadsorbed Pt(lll) completely dehydrogenates to H2 and surface carbon upon heating without any molecular desorption detected, which is identical to that 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. observed on clean Pt(lll). In addition to revealing aspects of site-blocking that should have broad implications for hydrogen coadsorption with hydrocarbon molecules on transition metal surfaces in general, these results also provide additional basic information on the surface science of selective catalytic hydrogenation of butadiene in butadiene-butene mixtures. 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.1 Introduction Butadiene (H2C=CHCH=CH2, C4H6) is an impurity in C4 alkene streams produced by steam cracking and it is desirable to remove this product from butadiene-butene mixtures. An ideal process would be to selectively convert butadiene to butene (C4H8) and avoid forming the completely hydrogenated product, i.e., butane (C4H 10). The hydrogenation of butadiene has been widely studied over supported Pt catalysts [1-6] and Pt single crystals [4, 7-12], Over supported Pt catalysts, hydrogenation of 1,3-butadiene yields n-butane (H3CCH2CH2CH3, C4H 10), 1-butene (H2C=CHCH2CH3, C4H8 ) and 2-butene(H3CCH=CHCH3, C4H8 ) simultaneously [2-4], The selectivity of forming butene over butane is greatly affected by the support material [6], Hydrogenation of 1,3-butadiene over a Pt(l 11) single crystal also produces a mixture of n-butane, 1-butene and 2-butene. Two issues of interest in these catalytic studies are the selectivity toward butene formation and product distribution between butene isomers. Methods to study adsorbed layers and identify surface intermediates during hydrogenation at high pressures on metal single crystals are limited. One work-around has been to investigate coadsorption of hydrogen and unsaturated hydrocarbon molecules under UHV conditions in a surface science approach to study hydrogenation at the molecular level. Coadsorption of hydrogen and ethylene on Pt(l 11) has been studied nicely by Zaera and coworkers [13-16], In those investigations, however, the amount of hydrogen present on the surface, 6h, was always far away from monolayer 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (saturation) coverages and no comments were made on any influence of H adatoms blocking sites at the surface. Hemminger and coworkers [17,18] studied the dehydrogenation of cyclohexane on Pt(l 11) that had been precovered by a saturation coverage of adsorbed hydrogen produced by large H2 exposures. They observed that preadsorbed H weakened the interaction between Pt(lll) and cyclohexane and the initial dehydrogenation activation energy was lowered by 20%. They reported that H adatoms had no effect on the sticking coefficient of cyclohexane on Pt(l 11) at 135 K and did not block cyclohexane adsorption. However, Lutterloh et al. [19] using a pyrolytic H atom source, recently found that chemisorption of benzene was completely blocked at high coverages of preadsorbed H adatoms on Pt(lll) at 125 K. It is interesting to note that the maximum value for O r on Pt(lll) obtained by exposing H2 has been reported to be 0.8 ML [20], and this coverage is difficult to obtain practically (requiring exposures of several hundreds of Langmuirs) because of the low dissociative sticking probability of H2 at large values of 0\\. This is probably why the amount of preadsorbed H used in most previous studies did not include a saturation (monolayer) coverage. The saturation coverage of H adatoms on Pt(lll) can be increased to 6h=0.95 ML by using a pyrolytic H atom source [21]. Because the absolute coverages in adsorbed hydrocarbon monolayers on Pt(l 11) can be quite low, 6hc = 0.25 - 0.1 ML, it may be important to study the influence of coadsorbed hydrogen at very high coverages otherwise islanding or ensemble effects can allow coadsorbed hydrocarbon molecules to simply avoid interacting directly with the 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. coadsorbed H adatoms. In addition, the role of site blocking effects of preadsorbed H should become more important at high values of 0\\than for low ones. Herein, we describe the results of our investigations of adsorption and reaction of 1,3-butadiene on Pt(lll) surfaces that contain varying amounts of preadsorbed H adatoms up to 9^=0.9 ML by using a pyrolytic H atom source. 6.2 Experim ental methods Experiments were performed in a three-level UHV chamber as described earlier [22]. The Pt(lll) crystal (Atomergic; 10 mm dia., 1.5 mm thick) was prepared by 1-keV Ar+ -ion sputtering and oxygen treatments (5x1 O ' 7 torr O2, 900 K, 2 min) to give a clean spectrum using Auger electron spectroscopy (AES) and a sharp (lxl) pattern in low energy electron diffraction (LEED). A Pt-tube doser was constructed, based on the design of Engel and Rieder [23], and used as a pyrolytic source of gas-phase H atoms. The principal component is a bent Pt-tube (1-mm O.D., 0.8-mm I.D.) into which a hole of 0.1-mm diameter was mechanically drilled. The Pt tube was resistively heated to 1275 °C, and water- cooling kept the adjacent Cu block cold. The temperature of the Pt tube was directly measured by an optical pyrometer that was calibrated by the temperature of the Pt(lll) crystal sample, as measured by a Cr/Al thermocouple. The relative accuracy of the pyrometer reading was estimated to be ±5 °C. The flux of H atoms obtained from this source at 800 °C and a pressure of 5x1 O ' 8 Torr in the UHV chamber was 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3xl0 1 3 atoms e rn e s'1 . This value was obtained by assuming that the initial sticking coefficient of incident H atoms on Pt(lll) at 100 K was unity and using the H coverage produced from the well-known decomposition of ethylene on Pt(lll) to give an absolute calibration for the H2 yield in subsequent TPD measurements [24]. H2 (Matheson; 99.99%) was introduced via a variable leak valve (Granville- Phillips) into the Pt-tube doser after passing through a liquid-nitrogen cooled, U-tube trap. 1,3-butadiene (C4H6, Matheson; 99.5%) was introduced via a microcapillary array doser connected to the gas line through a leak valve. All of the exposures reported herein are given simply in terms of the pressure measured in the UHV chamber by the ion gauge. No attempt was made to correct for the flux enhancement of the gas doser or ion gage sensitivity. Mass spectrometry performed in the UHV chamber showed no appreciable concentration of impurities in the source gases. The heating rate was 3.6 K/s in all TPD experiments, and all exposures were given at a sample temperature of 100 K. AES measurements were made with a double-pass cylindrical mirror analyzer (CMA) using a modulation voltage of 4 eV. The electron gun was operated at 3-keV beam energy and 1.5-pA beam current. Coverages 0, reported in this paper are referenced to the surface atom density of Pt(l 11) such that O pt =1.0 ML is defined as 1.505xl01 5 cm'2. 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.3 Results TPD was used to probe adsorption and reaction of 1,3-butadiene on clean Pt(lll) at 100 K and surfaces that were precovered with H adatoms following exposures from the pyrolytic H-atom doser. The exposure of 0.24-L 1,3-butadiene was determined in other experiments (not shown) to produce about 1.3 monolayers of 1,3-butadiene on the Pt(lll) surface at 100 K [25]. Signals at 2, 54, 56, 58, 18, 108, 110 amu were recorded to monitor possible reaction products in TPD, but only H2 (2 amu) and C4H6 (54 amu) were identified as desorbed products. No butene or butane desorption due to hydrogenation of 1,3-butadiene by preadsorbed H on Pt(l 11) was observed. Desorption spectra for C4H6 and H2 are shown in Figures 6.1 and 6.2, respectively. We also note here that we monitored water coadsorption in all of these experiments because of the low temperature adsorption requirements. TPD measurements (not shown) that directly compared H2O TPD following H2O adsorption on Pt(lll) were used to establish that the amount of water coadsorption in these preadsorbed-H experiments was small, ranging from 0.03-0.08 ML for all but the largest H atom precoverage of 0 h = 0 . 9 ML where 0 h2o = 0.13 ML. In the absence of preadsorbed H, as shown in the bottom curve of Figure 6.1, the 1,3-butadiene monolayer is completely irreversibly adsorbed and no molecular desorption of chemisorbed species occurs. Physisorbed molecules present in the second layer desorb in a low-temperature peak at 130 K. On H-precovered Pt(lll) surfaces, no chemisorbed species desorb and only desorption of physisorbed C4FL 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. species in a peak at 147 - 139 K was observed. Butadiene molecules that cannot be accommodated in the monolayer form condensed multilayer clusters and desorption from this phase occurs at low coverages in a peak at 107 K. As shown in the bottom curve of Figure 6.2, with no preadsorbed H on the Pt(l 11) surface, evolution of H2 from decomposition of the 1,3-butadiene monolayer occurs in a large peak at 339 K and a smaller feature at 610 K [25], Preadsorbing H on the Pt( 111) surface induces a new broad H2 desorption peak at 278 K that arises from desorption rate-limited recombination of H adatoms to desorb as H2, but does not alter the higher temperature hydrocarbon decomposition processes that liberate additional H2. The peak intensity at 339 K in Figure 6.2 decreased with an increase in preadsorbed H and this indicates that the amount of hydrocarbon decomposition decreased. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pt(111) 139 After 0.24-L CM 107 147 preads (ML) 0.90 M L Z3 E C O LO C O a C D 0.75 M L 4-> C 130 0.59 M L 0 M L 120 160 200 240 280 320 360 400 Temperature (K) Figure 6.1 C4H6 TPD spectra after 0.24-L 1,3-butadiene exposures on H preadsorbed Pt(lll) at 100 K. 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pt(111) 278 After 0.24-L CM 339 H preads (^L) 13 £ co C M 0.90 M L • <— > '( / ) c 0 -I— < 0.75 M L 0.59 M L c 0 M L 100 200 300 400 500 600 700 800 Temperature (K) Figure 6.2 H2 TPD spectra after 0.24-L 1,3-butadiene exposures on H preadsorbed Pt(lll) at 100 K. (dash curve is the FL TPD spectra after 0.057-L 1,3-butadiene exposures on Pt(l 11) at 100 K). 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Consistent with these results, the concentration of surface carbon remaining after a TPD experiment, detected by AES, also decreased with an increase in amount of preadsorbed hydrogen. This is shown in Figure 6.3, and the carbon AES peak after TPD is nearly eliminated after TPD following 0.24-L 1,3-butadiene on Pt(l 11) with 0.90-ML H preadsorbed. Considering the results of Figure 6.1 and that no other hydrocarbons desorbed, we explain these observations by a decrease in amount of chemisorbed 1,3-butadiene, rather than a decrease in the fraction of chemisorbed 1,3- butadiene that dehydrogenates during TPD, with increasing amounts of preadsorbed hydrogen. 150 200 250 300 350 400 450 500 550 Kinetic Energy (eV) Figure 6.3 AES spectra on Pt(lll) following 0.24-L C4FE TPD experiment on H preadsorbed Pt(l 11). Pt(111) After 0.24-L C4H6 TPD nrearis (ML) AES 0.90 M L t v\A'Vvv\y'w Y )7 0.75 M L i9 , 0.59 M L 0 M L 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Because all chemisorbed 1,3-butadiene molecules decompose to liberate H2 and form surface carbon, the amount of H2 generated from decomposition of 1,3- butadiene in TPD can be used to determine the coverage of chemisorbed 1,3- butadiene. This requires a comparison of the measured H2 TPD peak area to that of a reference H2 TPD spectrum that corresponds to a known value of 0H . We utilized the H2 TPD spectrum after ethylene (C2H4) exposures on Pt(lll) at 300 K to produce 0.25-ML ethylidyne (CCH3) [24], in which complete decomposition produces 0.375- ML H2, i.e., Oh = 0.75 ML. The H2 TPD area resulting from the decomposition of chemisorbed 1,3-butadiene was obtained by subtracting the H2 TPD area recorded without adding coadsorbed 1,3-butadiene from the corresponding H2 TPD area in Figure 6.2. These results are shown in Figure 6.4, which includes the 1,3-butadiene monolayer (saturation) coverage Bc^] ( = 0.15 ML on clean Pt(lll) [25]. Figure 6.4 also gives the result of using the C(272 eV)/Pt(237 eV) peak-to-peak ratio in AES after TPD (Figure 3) for determining the C4H6 coverage that decomposes during TPD. A C(272 eV)/Pt (237 eV) ratio of 0.89 after TPD of 0.24-L 1,3-butadiene on clean Pt(lll) corresponds to B c - 0.60 ML. Thus, with consistent results from H2 TPD and AES, the coverage of 1,3-butadiene 0C H decreased almost linearly with an increase in precoverage of hydrogen 0H . Chemisorption of 1,3-butadiene is totally blocked on Pt(l 11) at 100 K by preadsorption of 0h = 0.92 ML. 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.15 Pt(111) 0 .1 2 - ■ From H 2 TPD o From AES data _i 1 0.0 9 - 0 D) (0 u . 0 o 0.06 - o to X " 3 - o 0.0 3 - 0.00 0.2 0.4 0.6 0.8 1.0 0.0 H preadsorption ( ML) Figure 6.4 The influence of preadsorbed H to the C4H6 saturation coverage in the chemisorption layer 6.4 Discussion The two most important findings in this study were that no butene nor butane desorption due to hydrogenation of 1,3-butadiene by preadsorbed H on Pt(lll) during TPD was observed, and that chemisorption of 1,3-butadiene was totally 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. blocked on Pt(lll) at 100 K by preadsorption of 0H = 0.92 ML. These two observations will be discussed in turn. In order to understand the different hydrogenation activity for alkenes and dienes, it is instructive to compare in detail the chemisorption of ethylene, 1 -butene, and 1,3-butadiene on Pt(lll). Chemisorbed ethylene is di-c-bonded with an adsorption energy of 17 kcal/mol [26], During heating in TPD, 62% of the ethylene monolayer desorbed, 36% decomposed, and 2% formed ethane [24]. Under UHV conditions, chemisorbed ethylene on Pt(lll) is hydrogenated to ethane at 302 K, corresponding to an activation energy of 18 kcal/mol, in the presence of low hydrogen coverages [27]. Self-hydrogenation [27] can occur in the absence of any specific attempt to cause H coadsorption, and the ethane yield is about 2% of a monolayer. However, this is likely due to small amounts of hydrogen formed from the early stages of ethylene decomposition, low temperature decomposition at steps, or small amounts of coadsorption from background H2 adsorption. Addition of hydrogen to form a coadsorbed layer with ethylene on Pt(lll) greatly increased the formation of ethane (up to 50 % of the coadsorbed ethylene can be hydrogenated) and decreased the temperature at which ethane desorption occurred in TPD to 252 K [27]. This corresponds to an activation energy for ethylene hydrogenation or ethane formation of 15 kcal/mol (The authors in ref. [27] report this value to be 6 kcal/mol). Chemisorbed 1-butene is also di-c-bonded with an adsorption energy of 17 kcal/mol, 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and has chemistry similar to that of ethylene. During heating in TPD, 53% of the 1- butene monolayer desorbed, 47% decomposed, but no butane was formed [28]. These results for alkenes are in contrast to the behavior that we observed for chemisorbed 1,3-butadiene on Pt(lll) during TPD studies under UHV conditions. 100% of the 1,3-butadiene monolayer dehydrogenated on the Pt(l 11) surface during TPD and no molecular butadiene desorption or self-hydrogenation reaction was observed [25]. The specific addition of coadsorbed hydrogen on Pt(lll) still does not induce any hydrogenation reactions that lead to butene or butane desorption. This qualitatively different behavior from the alkenes can be explained by considering the difference in bonding at the surface. The adsorption configuration of chemisorbed 1,3-butadiene on Pt(l 11) at 300 K is still unclear, and proposals have been made for tetra-a-bonding [25, 29, 30] and a 1,4-di-a-bonding configuration in which a new C=C bond is formed in the center of the molecule [312-33]. Because no butadiene desorption is observed in TPD, there are no experimental measurements of the desorption activation energy or the adsorption energy. One can easily estimate that tetra-a-bonding would lead to an adsorption energy twice that of ethylene or 36 kcal/mol. This strong bonding to the surface would lead to irreversible chemisorption because of the relatively low energetic barrier to dehydrogenation, which should be about the same as that for ethylene or 17 kcal/mol. It is simple to estimate that the adsorption energy for tetra-a-bonding is much larger than that for the 1,4-di-a-bonding configuration, which would be the same as that for di-a-bonded 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ethylene plus an additional 7 1-bonding interaction of about 10 kcal/mol for an adsorption energy of 28 kcal/mol. Even this additional stability over that of ethylene could still explain irreversible adsorption. Hydrogenation of ethylene to form ethane presumably would occur via an ethyl intermediate [15]. The barrier for hydrogen addition to chemisorbed butadiene should be about the same as that for ethylene, but the similar barrier sizes for hydrogenation and dehydrogenation makes it much more difficult for the intermediates from butadiene hydrogenation to further hydrogenate to form butene and butane without undergoing dehydrogenation reactions that lead back to butadiene and irreversible H2 desorption from H adatom recombination. 1-Butene dehydrogenates prior to 345 K and 2-butene dehydrogenates prior to 352 K on Pt( 111) at small coverages [29]. It is obvious that if the barriers to hydrogenation and dehydrogenation of the intermediates involved are similar, but the adsorption energy of the reactant molecule is larger, then thermal reactions will disfavor molecular desorption. Forming a complete H-adatom monolayer on P t( lll) completely blocks subsequent chemisorption of benzene [19]. Consistently, we also recently observed that adsorption of ethylene was greatly decreased by the presence of 0.6-ML H on P t( lll) [34], 1,3-butadiene has a much stronger interaction with P t( lll) than ethylene (18 kcal/mol) and comparable or stronger interaction than benzene (28 kcal/mol) and its chemisorption is still blocked by preadsorbed H. These results 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. predict that chemisorption of most hydrocarbon molecules will be blocked by high coverages of preadsorbed H. Coadsorbed H adatoms may block chemisorption of 1,3-butadiene, and other hydrocarbon molecules, by simple repulsive interactions. Figure 6.5 shows a schematic drawing showing side views of 1,3-butadiene, ethylene, and benzene adsorption on an H-precovered P t( lll) surface and clean P t(lll). In constructing this drawing, the van der Waals radii for H adatoms and the hydrocarbon molecules were used. The vertical positions of the adsorbed species were determined by using either experimentally measured values or theoretical calculations of these numbers. Low-energy recoil scattering [35] indicates that H adatoms occupy fee 3-fold hollow sites and are located 1.2 A above the first layer of Pt atoms. The C-Pt distances in ethylene [36] and benzene [37] were obtained from recent DFT calculations. The value for 1,3-butadiene was assumed to be the same as that of ethylene for Figure 6.5. This schematic drawing, despite its limitations, clearly shows that non-bonding, repulsive, van der Waals interactions between coadsorbed H adatoms and any of these unsaturated hydrocarbon molecules will not allow for a molecule-Pt distance that is required for covalent overlap and strong chemisorption bonding with Pt atoms at the surface. This contact leads to an increase in the adsorbate-surface distance on H-preadsorbed P t( lll) by 1.3-1.4 A in comparison with that on clean P t(lll). This increased distance makes it impossible for the hydrocarbon adsorbates to form chemical bonds with Pt atoms at the surface. However, there still remains an 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. appreciable polarization interaction that stabilizes these hydrocarbons in the adsorbed layer far over that involved in the hydrocarbon condensed phase, and this leads to desorption of 1,3-butadiene at 140-150 K in Figure 6.1. Distance: A _________ . >•' 3.45 2 .05. (A) 1,3-butadiene/Pt(111) 3.34 2.05 (B) Ethylene/Pt(111) — H 3 ,54, (C) Benzene/Pt(111) Figure 6.5 Side view of 1,3-butadiene (a), ethylene (b) and benzene (c) adsorption on Pt(111) and H preadsorbed Pt(l 11). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. At low coverages of coadsorbed H, islanding of the hydrogen or hydrocarbons into small domains will allow these coadsorbates to simply avoid interacting directly with each other. Displacement of the coadsorbed H adatoms to adjacent adsorption sites will also enable these coadsorbates to avoid large repulsive interactions. Both of these phenomena should be dependant on adsorbate coverages and the temperature. Thus, strong site-blocking effects by H adatoms are expected to be less important at low coverages and high temperatures. A change in the electronic structure of P t( lll) may also play a role in the elimination of chemisorption of 1,3-butadiene and other hydrocarbons by coadsorbed H. The Pt-H bond dissociation energy D(Pt-H) for chemisorbed H adatoms on P t( lll) is 61.5 kcal/mol [15]. This is a strong polar-covalent bond and this suggests that preadsorbed H could have a significant effect on the electronic structure of Pt(l 11). Lutterloh et al. [19] argued that preadsorbed H eliminates the empty metal d states on P t( lll), which are critical for a- and Tc-donation, and that this explained why preadsorbed H blocked the chemisorption of benzene. While this undoubtedly play a role, Figure 6.5 illustrates that a simple site-blocking effect of preadsorbed H on P t( lll) is all that is needed to explain these results. This should be a general effect with broad applicability to the chemisorption of other organic molecules on other transition metal surfaces. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.5 Conclusions Coadsorbed H adatoms have a strong influence on the coverage of chemisorbed 1,3-butadiene dC iH t.on P t( lll) at 100 K. 0C i„( i decreases linearly with increasing coverages of preadsorbed H on P t( lll) and chemisorption of 1,3- butadiene is completely blocked at O n = 0.92 ML. This site-blocking effect of preadsorbed H can be explained by a simple argument involving repulsive interactions between coadsorbed H and the hydrocarbon coadsorbate, and thus this observation should be a general one applicable to other organic molecules on other transition metal surfaces. However, other than reducing the amount of chemisorbed species, the presence of coadsorbed H adatoms did not alter the chemistry of chemisorbed 1,3- butadiene on P t(lll). 1,3-butadiene completely dehydrogenates on P t( lll) to liberate only H2 and form surface carbon upon heating in TPD. Significantly, no hydrogenation reaction was observed between coadsorbed H adatoms and 1,3- butadiene, even though extensive hydrogenation of coadsorbed layers of H and ethylene on Pt(l 11) has been reported. This can be explained by a stronger bonding interaction between 1,3-butadiene and P t( lll) that makes it much more difficult for the intermediates from butadiene hydrogenation to further hydrogenate to form butene and butane without undergoing dehydrogenation reactions that lead back to butadiene and irreversible H2 desorption from H adatom recombination. 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. These results also provide additional basic information on the surface science of selective catalytic hydrogenation of butadiene in butadiene-butene mixtures. For example, the Pt surface bonds dienes too strongly to be a good diene hydrogenation catalyst, and so Pt should be modified to reduce this interaction. This commonly occurs by alloying Pt with a second, less-reactive metal, or surface carbonaceous deposits formed under actual catalytic conditions may also play this role. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.6 References [1] G.C. Bond, G. Webb, P.B. Wells and J.M. Winterbottom, J. Catal. 1 (1962) 4. [2] G.C. Bond, G. Webb, P.B. Wells and J.M. Winterbottom, J. Chem. Soc. A, (1965)3128. [3] P.B Wells and A.J. Bates, J. Chem. Soc. A, (1968) 3064. [4] C.M. Pradier, E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 43 (1988) 177. [5] M. Primet, M. El Azhar and M. Guenin, Appl. Catal., 58 (1990) 241. [6] A. Sarkany, G. Stefler and J.W. Hightower, Appl. Catal. A: Gen., 127 (1995) 77. [7] J. Massardier, J.C. Bertolini, P. Ruiz, P. Delichere, J .Catal. 112 (1988) 21. [8] J. Qudar, S. Pinol, and Y. Berthier, J . Catal. 107 (1987) 434. [9] C.M. Pradier, E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 31 (1987) 243. [10] T. Ouchaib, J. Massardier, and A. Renouprez, J .Catal. 119 (1989) 517. [11] C.M. Pradier and Y. Berthier, J .Catal. 129 (1991) 356. [12] C. Yoon, M.X. Yang, and G.A. Somorjai, Catal. Lett. 46 (1997) 37. [13] F. Zaera and G.A. Somorjai J. Am. Chem. Soc. 106 (1984) 2288. [14] F. Zaera, J. Am. Chem. Soc.I l l (1989) 4240. [15] F. Zaera, J. Phys. Chem. 94 (1990) 5090. [16] T.V.W. Janssens, D. Stone, J.C. Hemminger, F. Zaera, J. Catal. 177 (1997) 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 284. [17] D.A. Perry and J.C. Hemminger, J. Am. Chem. Soc. 122 (2000) 8079. [18] M.E. Pansoy-Hjelvik, P. Schnabel, and J.C. Hemminger, J. Phys. Chem. B. 104 (2000) 6554. [19] C. Lutterloh, J. Biener, K. Pohlmann, A. Schenk, and J. Kuppers, Surf. Sci. 352 (1996) 133. [20] K. Christmann, G. Ertl, and T. Pignet, Surf. Sci. 54 (1976) 365. [21] M.R. Voss, H. Busse, and B.E. Koel, Surf. Sci. 414 (1998) 330. [22] H. Zhao, J. Kim, B. E. Koel, Surf. Sci. 538 (2003) 147. [23] T. Engel, K.H. Rieder, in: G. Hohler (Ed.), Structural Studies o f Surfaces, Springer, Berlin, 1982, Vol. 91, p. 55. [24] R.G. Windham, M.E. Bartram, B.E. Koel, J. Phys. Chem. 92 (1988) 2862. [25] H. Zhao and B.E. Koel, Surf. Sci. in press [26] M. T. Paffett, S. C. Gebhard, R.G. Windham, and B.E. Koel, Surf. Sci. 223 (1989) 449. [27] D. Godbey, F. Zaera, R. Yeates, and G.A. Somorjai, Surf. Sci. 167 (1986) 150. [28] Y. Tsai and B.E. Koel, J. Phy. Chem. B 101 (1997) 2895. [29] N.R. Avery, N. Sheppard, Proc. R. Soc. Lond. A 405 (1986) 1. [30] F. Mittendorfer, C. Thomazeau, P. Raybaud, and H. Toulhoat, J. Phys. Chem. B. 107 (2003) 12287. [31] J.C. Bertolini, A. Cassuto, Y. Jugnet, J. Massardier, B. Tardy, and G. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tourillon, Surf. Sci. 349 (1996) 8 8 . [32] G. Tourillon, A. Cassuto, Y. Jugnet, J. Massardier, J.C. Bertolini, J. Chem. Soc., Faraday Tran. 92 (1996) 92. [33] R.C. Baetzold, Langmuir 3 (1987) 189. [34] H. Zhao and B. E. Koel, to be published. [35] K. Umezawa, T. Ito, M. Asada, S. Nakanishi, P. Ding, W.A. Lanford, and B. Hjorvarsson, Surf. Sci. 387 (1997) 320. [36] G.W. Watson, R.P.K. Wells, D. J. Willock, and G.J. Hutchings, J. Phys. Chem. B 104 (2000) 6439. [37] M. Saeys, M.F. Reyniers, and G.B. Marin, and M. Neurock, J. Phys. Chem. B 106 (2002) 7489. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 7 Hydrogenation of 1,3-butadiene on two ordered S n /P t(lll) surface alloys A bstract Adsorption and reaction of 1,3-butadiene (C4H6) on two ordered Pt-Sn surface alloys precovered with hydrogen adatoms was studied by using temperature programmed desorption (TPD) mass spectroscopy and Auger electron spectroscopy (AES). The two alloys investigated were the (2x2)Sn/Pt(lll) and (V3xV3)R30°Sn/Pt(lll) surface alloys, with 25% and 33% Sn alloyed in the surface layer, respectively, formed by vapor deposition of Sn onto a P t( lll) single crystal. Alloyed Sn opens a new hydrogenation reaction pathway, compared to P t(lll). Butadiene hydrogenation by coadsorbed hydrogen occurs with 100% selectivity to liberate butene (C4H8) in reaction rate-limited peaks in TPD and no deeper hydrogenation product (butane) was observed. The activation energy barrier for hydrogenation of strongly bound 1,3-butadiene is estimated to be 91 and 72 kJ/mol on the (2 x2 ) and (V3xV3)R30° alloys, respectively. Butadiene conversion was highest on the (2x2) alloy, reaching 100 % at high hydrogen precoverages. Strong site-blocking effects of preadsorbed H adatoms were observed for 1,3-butadiene chemisorption on both alloys under these conditions; butadiene chemisorption was eliminated by 0// = 0.49 ML on the (2x2) alloy and 0h = 0.34 ML on the V3 alloy. These studies addressing the influence of alloyed Sn on the reaction barrier to 1,3- 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. butadiene hydrogenation and the effect of surface Sn concentration on hydrogenation activity provide observations of several novel phenomena and may aid in the development of heterogeneous catalysts to selectively remove dienes in alkene streams. 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7.1 Introduction Development of heterogeneous catalysts to selectively remove butadiene in C4 alkene streams produced by steam cracking is of considerable interest. The ideal catalyst and process would convert butadiene selectively to butene but not lead to further hydrogenation of butene to butane. Various platinum surfaces, including supported catalysts [1-6] and single crystals [4, 7-12], have been investigated regarding the hydrogenation of butadiene. Unfortunately, mixtures of butene isomers and butane are apparently usually produced over platinum catalysts. Supported Pt-Sn bimetallic catalysts have been reported to be effective in selectively hydrogenating diolefin impurities [13]. Generally, adding Sn to Pt catalysts used for hydrocarbon conversion reactions results in decreased catalytic activity, increased selectivity for unsaturated hydrocarbon products, and reduced coking, which prolongs the lifetime of the catalyst. Specifically, several reports show that the presence of Sn in bimetallic Pt catalysts prevents alkene hydrogenation, which is important to selectively removing diene impurities. Adding Sn to a supported Pt catalyst caused a dramatic decrease in catalytic activity for propene hydrogenation [14]. In addition, hydrogenation activity for ethylene and 1- hexene was inhibited on Pt-Sn/AI2O3 catalysts [15]. Although the reactivity of a Pt(111) single-crystal increased slightly for ethylene hydrogenation when 0.1-ML Sn was added, it decreases quickly with further increase in the Sn coverage [16]. 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experiments investigating coadsorption of hydrogen and ethylene on the (V3xV3)R30°-Sn/Pt(l 11) surface alloy under UHV conditions show that hydrogenation is completely inhibited by alloying Sn into P t( lll) to form this alloy [17] These results confirm that alloyed Sn acts as an inhibitor to alkene hydrogenation. Other recent experiments have shown that chemisorbed 1,3- butadiene completely decomposes on both clean and H-precovered P t( lll) surfaces during TPD, with no hydrogenation reaction observed [18]. This presumably is due to the very strong bonding interactions between 1,3-butadiene and P t(lll). Previously, we have shown in 1,3-butadiene chemisorption studies that alloying Sn to P t( lll) decreases this interaction and the decomposition pathway is totally blocked on the (2 x2) and (V3xV3)R30° S n /P t(lll) surface alloys [19]. It is natural to inquire whether this reduced interaction between 1,3-butadiene and the Sn/Pt(l 11) surface alloys may lead to hydrogenation reactions in the presence of coadsorbed hydrogen. In this paper, we report on studies using temperature programmed desorption (TPD) of adsorption and reaction of 1,3-butadiene on hydrogen-precovered (2x2)- S n /P t(lll) and (V3xV3)R30°-Sn/Pt(l 11) surface alloys. In particular, we address the influence of alloyed Sn on the reaction barrier to 1,3-butadiene hydrogenation and the effect of surface Sn concentration on hydrogenation activity. 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7.2 Experimental methods Experiments were performed in a three-level UHV chamber as described earlier [20], The P t( lll) crystal (Atomergic; 10-mm dia., 1.5-mm thick) was prepared by using 1-keV Ar+-ion sputtering and oxygen exposures (5xl0’7 -torr O2, at 900 K for 2 min) to give a clean spectrum using Auger electron spectroscopy (AES) and a sharp (lx l) pattern in low energy electron diffraction (LEED). The (2 x2 )S n/P t(lll) and (V3xV3)R30°Sn/Pt(l 11) surface alloys were prepared by evaporating one monolayer of Sn onto the P t( lll) crystal surface and subsequently annealing the sample for 20 s to 1000 K and 830 K, respectively. Sn is incorporated substitutionally into primarily the surface layer to form an ordered alloy or intermetallic compound with 0sn =O.25, with a composition and structure corresponding to the ( 1 11) face of a bulk PtsSn crystal, and for the latter situation, 0 S n = O .3 3 , with a composition corresponding to a Pt2Sn surface. These surface alloys are relatively “flat”, but Sn atoms protrude 0.02 nm above the surface-Pt plane at both surfaces [21]. In the (2x2) alloy, pure-Pt three-fold sites are present, but no adjacent pure-Pt three-fold sites exist. All pure-Pt three-fold sites are eliminated on the (V3xV3)R30° alloy, and only two-fold pure-Pt sites are present. For brevity throughout this paper, we will refer to the (2 x2 )-S n/P t(lll) and (V3xV3)R30°- Sn/Pt(l 11) surface alloys as the (2x2) and V3 alloys, respectively. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A Pt-tube doser was constructed, based on the design of Engel and Rieder [22] as a pyrolytic source of hydrogen atoms. The principal component is a bent Pt tube (1-mm O.D., 0.8-mm I.D.) into which a 0.1-mm diameter hole was mechanically drilled. The tube was resistively heated up to 1275°C, and water- cooling kept the adjacent Cu block cold. The temperature of the Pt tube was directly measured by an optical pyrometer that was calibrated by the temperature of the Pt(l 11) crystal sample, as measured by a Cr/Al thermocouple. The estimated relative accuracy of the pyrometer reading was ±5°C. The flux of H atoms obtained from this source operating at 800°C with a background pressure rise in the chamber of 5x10' 8 Torr was 3 x l0 1 3 atoms-cm"2-s’1 . This value was obtained by assuming that the initial sticking coefficient of H atoms on P t( lll) at 100 K was unity and using a calibration for the H coverage, 0h, that was given by the hydrogen yield in TPD from the well-known decomposition of ethylene on Pt(l 11) [23], H2 (Matheson; 99.99%) was introduced via a variable leak valve (Granville- Phillips) into the Pt-tube doser. 1,3-butadiene, C4H6, (Matheson, 99.5%) was used without additional purification. 1,3-butadiene was exposed on the alloy surface by a microcapillary array doser connected to the gas line through a variable leak valve. All of the exposures reported here are given simply in terms of the background pressure in the UHV chamber as measured by an ion gauge. No attempt was made to correct for the flux enhancement of the doser or ion gauge sensitivity. The mass spectrometer in the chamber was used to check the purity of the gases during dosing. 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For all TPD experiments, the heating rate was 3.6 K/s and all exposures were given with the surface temperature at 100 K. AES measurements were made with a double-pass cylindrical mirror analyzer (CMA) using a modulation voltage of 4 eV. The electron gun was operated at 3-keV beam energy and 1.5-p.A beam current. Coverages 0t reported in this paper are referenced to the surface atom density of Pt(l 11) such that 6pt =1.0 ML is defined as 1.505 x l0 15cm'2. 7.3 Results Hydrogen adatoms were preadsorbed on the (2x2) alloy and then adsorption and reaction of a monolayer of 1,3-butadiene (C4H6) on this surface at 100 K was investigated by TPD. An exposure of 0.24-L 1,3-butadiene was used, which produces a coverage of about two monolayers on the two clean alloy surfaces. Desorption spectra for 1,3-butadiene, C4H6 (54 amu), butene, C4H8 (56 amu), and H2 (2 amu) obtained in these experiments are shown in Figures 7.1, 7.2, and 7.3 respectively. Several other masses, including butane, C4H10 (58 amu), were also monitored during heating in TPD, but no significant signals were detected in the TPD spectra. Figure 7.1 shows the thermal desorption of molecular 1,3-butadiene from the clean (bottom curve) and H-precovered (2x2) alloy surface. Desorption from the chemisorbed monolayer on the clean (2 x2 ) alloy produces a wide peak centered at 334 K, which is consistent with our previous study [19]. There, an estimate was 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. made of the desorption activation energy Ed of 88 kJ/mol by using the Redhead method [24] and assuming first-order desorption kinetics with a preexponential factor of 1 0 1 3 s’1 . (2 x 2 ) alloy After 0.24-L C.H =3 E C D LO preads 100 150 334 200 > 4-» 283 0.48 < /> c C D c 0.32 0.16 100 200 300 400 500 600 700 800 Temperature (K) Figure 7.1 1,3-butadiene, C4H6 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (2x2)Sn/Pt(lll) alloy at 100 K. 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Preadsorbed H adatoms reduce the amount of 1,3-butadiene desorption in this peak, and this peak disappears following preadsorption of 0.476-ML H. Desorption of the most strongly chemisorbed 1,3-butadiene is affected first by preadsorbed H. This is shown by the shift of the 1,3-butadiene desorption peak to lower temperature and the appearance of a peak at 283 K (Ed = 73 kJ/mol) in the TPD spectra taken for 0.159-ML H. The inset shows the low temperature region of the desorption traces in Figure 7.1 on an expanded scale. As shown in the inset, the desorption peak of physisorbed (second-layer) 1,3-butadiene shifts from 130 K on the clean (2x2) alloy to 140 K on the 0.476-ML FI surface. Some clustering occurs apparently on the H-precovered surfaces, and low-temperature desorption onset characteristic of the multilayer appears near 105 K. One can calculate values of Ed of 33 and 35 kJ/mol for 1,3- butadiene adsorbed in the second layer on the clean and 0.476-ML H precovered (2 x2 ) alloy, respectively. Desorption of butene, C4H8, is shown in Figure 7.2. This is a hydrogenation product produced by surface reactions. 1,3-butadiene adsorption on the clean (2x2) alloy leads to a small amount of desorption (bottom curve) in a peak at 376 K. Because there is no coadsorption possible from H2 in the background gas, i.e., H2 does not dissociatively chemisorb on the (2x2) alloy under UHV conditions [25], this must originate from hydrogenation reactions utilizing hydrogen liberated by 1,3- butadiene dehydrogenation. 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (2x2) alloy After 0.24-L C,H 354 preads (^L) 0.48 1 3 E C D CD L O (/) c £ c 0.32 0.16 376 300 400 500 600 700 800 100 200 Temperature (K) Figure 7.2 Butene, C4H8 TPD spectra after after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (2x2)Sn/Pt(lll) alloy at 100 K. 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The amount of butene desorption is initially increased by H preadsorption and then decreased by the largest H precoverages. The decreased hydrogenation yield at high values of 6h is at least partially due to the low coverage of 1,3-butadiene that result from the site-blocking effects of preadsorbed H. The butene desorption peak shifts down to 354 K (Ed = 91 kJ/mol) for 0.16-ML H and then to 345 K (Ed = 89 kJ/mol) for 6h = 0.32 and 0.48 ML. Because butene desorption is reaction-rate limited, as discussed below, this shift arises from faster reaction kinetics at higher Oh- Figure 7.2 shows that some butene is produced below 200 K. The butene TPD peaks near 130 and 145 K have shapes similar to the low temperature 1,3- butadiene desorption traces shown in the inset to Fig. 7.1, and thus these are TPD artifacts in the butene spectra. Flowever, for O h = 0.32 and 0.48 ML, the butene peak area at 145 K (Ed = 36 kJ/mol) increases significantly (and even a new peak arises at 170 K) while those for 1,3-butadiene do not, and so butene is indeed produced under these conditions. At these hydrogen precoverages, desorption of chemisorbed 1,3- butadiene near 300 K was nearly eliminated. This indicates that the low-temperature butene yield was due to facile hydrogenation of weakly adsorbed, 7t-bonded species. Desorption of H2 associated with these experiments is shown in Figure 7.3. The bottom curve shows the FL yield from 1,3-butadiene dehydrogenation and decomposition on the clean (2x2) alloy. This is consistent with previous studies [19] that showed only a small amount of dehydrogenation occurred. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (2x2) alloy u J p Q 347 After 0.24-L C ^H 425 3 E (0 C N I preads 0.48 co c < D H— ' C 0.32 0.16 600 700 800 200 300 400 500 100 Temperature (K) Figure 7.3 H2 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H- precovered (2x2)Sn/Pt(lll) alloy at 100 K. 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In such cases, it is difficult to distinguish between a small reactivity for the alloy and a small number of reactive defect sites. H-preadsorption results in a new or larger H2 desorption peak at 347 K (Ed = 90 kJ/mol) and a sharpening of the peak at 425 K (Ed = 111 kJ/mol). Increasing preadsorbed O h to 0.48 ML shifts the primary peak to a slightly lower temperature of 338 K and results in two new desorption peaks at 214 and 254 K. A comparison of these H2 TPD spectra in Figure 7.3 to those obtained after the same hydrogen exposure on the (2x2) alloy without 1,3-butadiene post adsorption shown in Figure 2.3.4 establishes that the peaks at 214, 254, and 338-347 K in Figure 7.3 are from desorption of preadsorbed H, and furthermore, no shifts of these peaks were observed due to coadsorbed 1,3-butadiene. The peak at 425 K is similar to that observed in TPD for the decomposition of 1-butene and 2-butene at low coverages, and 11 K higher than that from the full monolayers, on the (2x2) alloy [26], Consistent with this H2 desorption yield at 425 K, surface carbon was detected by AES after TPD experiments on the H precovered (2x2) alloy. TPD experiments were also carried out after 0.24-L exposures of 1,3- butadiene on the clean and H-precovered a/3 alloy at 100 K. Only 1,3-butadiene, C4H6 (54 amu), butene, C4H8 (56 amu), and H2 (2 amu) were detected in TPD and these curves are shown in Figures 7.4, 7.5, and 7.6, respectively. 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 alloy After 0.24-L C.H u E 03 UO 100 150 200 preads ^ L ) 307 285 361 177 0.31 </) c < D 0.26 c 0.23 0.13 10 0 200 300 400 500 600 Temperature (K) Figure 7.4 1,3-butadiene, C4H6 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (V3xV3)R30°Sn/Pt(l 11) alloy at 100 K. 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 alloy 348 After 0.24-L CM 295 183 H preads 0.31 3 E C D CD L O 0.26 282 to c 0 c 0.23 0.13 300 400 500 600 100 200 Temperature (K) Figure 7.5 Butene, C4H8 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H-precovered (V3xV3)R30°Sn/Pt(lll) alloy at 100 K. 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 alloy After 0.24-L C^H 13 £ (U CNI preads (^L) 0.31 c n c 0 C = 0.26 0.23 0.13 200 100 300 400 500 600 Temperature (K) Figure 7.6 H2 TPD spectra after 0.24-L 1,3-butadiene exposures on the clean and H- precovered (V3xV3)R30°Sn/Pt( 111) alloy at 100 K. 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.4 shows the influence of preadsorbed H on the desorption of molecular 1,3-butadiene from the a/3 alloy. Chemisorbed 1,3-butadiene desorbs in relatively narrow peak at 285 K (bottom curve) from the clean a/3 alloy. This peak decreases in intensity with increasing H preadsorption and shifts up slightly to 3 0 7 K (Ed = 79 kJ/mol) on the a/3 alloy with f% r= 0 .3 1 ML. In addition, two new, weak desorption features arise at 1 7 7 and 361 K on the H precovered surfaces. While the peak at 177 K (Ed = 45 kJ/mol) is probably from desorption of a weakly bound state of 1,3-butadiene in the monolayer, the peak at 361 K (Ed = 93 kJ/mol) seems to be necessarily from the dehydrogenation of some surface intermediate formed by hydrogenation reactions at lower temperatures. The intensity decrease with increasing 6h occurs primarily from site-blocking effects of preadsorbed hydrogen on 1,3-butadiene chemisorption. The inset to Figure 7.4 highlights the H-induced changes in desorption of physisorbed 1,3-butadiene. Desorption from the physisorbed, second layer shifts from 121 K on the clean a/3 alloy to 134 K on the surface with #^=0.31 ML. This corresponds to a small increase in the desorption activation energy from 30 to 34 kJ/mol. Figure 7.5 shows butene TPD spectra that result from 1,3-butadiene hydrogenation reactions. On the clean a/3 alloy, no butene desorption occurred. No coadsorbed hydrogen is possible from FL adsorption from the background [25] and 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. no decomposition or self-hydrogenation occurs [19]. The low-temperature butene TPD peaks from the clean V 3 alloy are thought to be TPD artifacts in the butene spectra, as discussed above. Preadsorbed H immediately leads to butene production and desorption, and two broad butene TPD peaks were observed at 282 and 348 K from the surface with B h= 0.13 ML. The onset for butene desorption appears to be as low as 170 K (Ea = 43 kJ/mol). The peak at 282 K (Ed = 72 kJ/mol) shifts higher in temperature to 295 K (Ed = 76 kJ/mol) with increasing 0H , but no large change occurs in the peak at 348 K (Ed = 90 kJ/mol). As occurred on the the (2x2) alloy, for O h > 0.23, butene is produced and desorbs near 134 K (Ed = 34 kJ/mol) due to facile hydrogenation of weakly adsorbed, Tt-bonded species. Increasing the H precoverage to 0.31 ML results a new butene desorption peak at 183 K (Ed = 46 kJ/mol). LL TPD spectra generated simultaneously in these experiments are shown in Figure 7.6. The bottom trace of Figure 7.6 was obtained following 1,3-butadiene adsorption on the clean V3 alloy. The result that there was no significant FL desorption is consistent with our previous study showing that 1,3-butadiene did not decompose on the V 3 alloy during TPD [19]. There is a peak at 272 K (Ed = 70 kJ/mol) at 9h =0.13, which shifts to 260 K (Ed = 67 kJ/mol) at 0h =0.31 ML, and a peak at 217 K (Ed = 55 kJ/mol) for 0h = 0.31 ML. Signals below 200 K are derived from butadiene desorption and do not reflect Ffe desorption. The observation that no new H2 TPD peaks arose and the close resemblance of the H2 TPD peaks in Figure 7.6 with those peaks obtained after the same H-atom exposures without any 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. subsequent 1,3-butadiene adsorption indicate that no 1,3-butadiene dehydrogenation or decomposition occurs under these conditions. AES measurements after each TPD experiment detect no surface carbon and thus give a consistent picture. Furthermore, coadsorbed 1,3-butadiene had no significant effect on the H2 TPD spectra from H- atom exposure,. In these experiments, we were unable to identify the nature of the butene molecules, i.e., as 1-butene, cis- or trarz.s-2 -butene, or isobutene, that were desorbed as shown in Figures 7.2 and 7.5. Previously, we studied the adsorption of 1-butene, cis-2-butene and isobutene on P t( lll) and these two S n /P t(lll) surface alloys [26], We note that the conversion of the C4H8 TPD peak areas in Figures 7.2 and 7.5 to the amount or coverage of desorbed butene is complicated by this fact. In general, this is a simple conversion determined by comparing the unknown TPD peak area to that from a calibrated yield of one of the butenes from P t(ll 1). However, if we do the conversion assuming that 1-butene is the desorbed product, and thus comparing directly with the yield of chemisorbed 1-butene from the 1-butene monolayer on Pt(l 11), we obtain about twice the yield that we would calculate by using 2-butene as the calibration. The butene yields reported below in Figures 7.7 and 7.8 were obtained by using 1-butene as the calibration. Figure 7.7 summarizes and quantifies the influence of preadsorbed H adatoms on the yields of products measured in the TPD data from the (2x2) alloy. We plot in Figure 7.7, as a function of the H precoverage 9h on the (2x2) alloy, the 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. amount of reversibly chemisorbed 1,3-butadiene (0CH), amount of butene desorbed after formation by hydrogenation (0CH ), and total amount of chemisorbed 1,3- butadiene (0 C iH < .+ c H li )• The conversion (0 C i,,x / ^c4 //6 +c4 //8) 's also plotted on the right- hand axis. 0.18 2 x 2 alloy 0.15- - 0.8 0.12 - C K + C„H, _i 2 - 0.6 C,H, C D 0.09 - O) C O © > o 0.06 - -0 .4 o - 0.2 0.03- C H 0.00 0.0 0.0 0.1 0.2 0.3 0.4 0.5 preads Figure 7.7 Influence of H-adatom precoverage on the amount of 1,3-butadiene adsorption, desorption, and hydrogenation on the (2x2)Sn/Pt(lll) alloy. 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The total coverage of chemisorbed 1,3-butadiene (0C H c H ) is decreased, roughly linearly, by preadsorbed H and we find that 1,3-butadiene chemisorption is eliminated when O h reaches 0.49 ML. This is a significant site-blocking influence exerted by H adatoms because 1,3-butadiene is potentially a strongly chemisorbed species. The other main effect that is observed is that the yield of butene from hydrogenation reactions increases strongly. The 0C H curve increases from near zero on the clean alloy to a broad maximum near O h =0.25 ML. The decrease observed at larger O h values could easily be due simply to the decrease in the initial amount of 1.3-butadiene available for reaction. Of course, the amount of reversibly adsorbed 1.3-butadiene 0C H i decreases quickly with increasing O h because of the combined effects of increased site-blocking and the propensity for hydrogenation reactions. The conversion of 1,3-butadiene to butene increases almost linearly with increasing O h and reaches a value of 1.0 at O h = 0.47 ML. But, this hydrogenation is in competition with other processes on the surface during heating in TPD. Two H atoms are needed to hydrogenate one 1,3-butadiene molecule to produce one butene molecule. Although 1,3-butadiene is in surplus on the (2x2) alloy at small O h, some H atom recombination occurs to desorb H2, and this competes with hydrogenation. On the (2x2) alloy with large O h, some 1,3-butadiene desorbs without being hydrogenated even though H adatoms are oversupplied, and this competes with hydrogenation. 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The influence of preadsorbed H on adsorption and hydrogenation of 1,3- butadiene on the V3 alloy is shown in Figure 7.8. The site-blocking effects of preadsorbed H initially decrease the total amount of chemisorbed 1,3-butadiene #c # +c // in a very similar fashion to that on the (2x2) alloy. For example, the total coverage is reduced to 0.085 ML on both alloys for 0H = 0.23 ML. However, for O h > 0.25 ML, site blocking is much more effective on this alloy than on the (2x2) alloy and 1,3-butadiene chemisorption is completely blocked at O h - 0.34 ML. The yield of desorbed butene 0C increases from zero on the clean alloy to a broad maximum with increasing H precoverage until, at O h =0.26 ML, it starts to decrease with additional 0H because of the sharp decline in the amount of chemisorbed 1,3- butadiene available. Increasing 0h also decreases the amount of desorbed 1,3- butadiene 0C H but in this case the curve much more closely tracks that of the decrease in the total amount of chemisorbed 1,3-butadiene. Preadsorbed H adatoms certainly increase the conversion of 1,3-butadiene to butene on the V3 alloy, almost linearly with increasing O h from a value of zero to 0.3 when O h = 0.34 ML. However, this can be compared to a conversion of 65% on the (2x2) alloy at this value of 0H - Conversion never climbs higher than 0.3 on the V3 alloy. As on the (2x2) alloy, hydrogenation on the V3 alloy is in competition with other processes on the surface during heating in TPD. H2 desorption when O h < 2x 0C H indicates that hydrogenation competes with H recombination. 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hydrogenation also competes with 1,3-butadiene desorption, which occurs even in cases of O h > 2x 0C H . 0.18 V3 alloy 0.15- - 0.8 C K + C H, 0 .1 2 - - 0.6 CH, (D 0.09 - CD CD a 5 > 0.06- O -0 .4 o - 0.2 0.03- CH, 0.0 0.00 0.3 0.5 0.0 0.1 0.2 0.4 . (M L) H preads v 7 Figure 7.8 Influence of H-adatom precoverage on the amount of 1,3-butadiene adsorption, desorption, and hydrogenation on the (V3xV3)R30°Sn/Pt(l 11) alloy. 7.4 Discussion The addition of alloyed Sn in Pt surfaces weakens the bonding of hydrocarbons, such as alkenes and dienes, and this has been known for some time. 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The hydrogen coadsorption experiments reported herein greatly extend our understanding of the influence of alloyed Sn on the chemistry and catalysis that occurs on Pt surfaces by revealing how this weakened bonding alters the influence of coadsorbed hydrogen on adsorption, desorption, and hydrogenation rates. Blocking of strong adsorption sites by preadsorbed H adatoms on Pt(l 11) has been reported previously [27]. We also observed this H-adatom site-blocking effect for 1,3-butadiene adsorption on H-precovered P t( lll) and have proposed why this could be a general phenomena for other hydrocarbons [18]. Such a site-blocking effect of preadsorbed H was also observed in this work for 1,3-butadiene chemisorption on both S n /P t(lll) surface alloys. What is interesting is to realize how alloyed Sn increases the importance of this effect, i.e., how much smaller amounts of hydrogen are needed to block chemisorption on the alloy surfaces. Even though the monolayer saturation coverage of chemisorbed 1,3-butadiene (0.15 ML) is nearly the same on Pt(l 11) and the (2x2) and a/3 alloys, the amount of preadsorbed hydrogen needed to complete block 1,3-butadiene chemisorption on P t( lll) is 0ff= 0.91 ML, but only <9^=0.49 and 0.34 ML on the (2x2) and V3 alloy, respectively. The implications of this behavior are clear. One H adatom per unit cell on Pt(l 11), adsorbed in a 3-fold hollow site, produces a saturation coverage of 1 ML. Each Pt at the surface has a nearest-neighbor H adatom, and this evidently passivates the surface against additional H atom uptake and strong 1,3-butadiene chemisorption. On the (2x2) alloy, two H adatoms per unit cell, adsorbed in pure-Pt, 3-fold hollow 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sites, produces a saturation coverage of 0.5 ML. The two H adatoms occupy one fee and one hep site, with each Pt at the surface having a nearest-neighbor H adatom, and one Pt per unit cell having two nearest-neighbor H adatoms. This evidently passivates the surface against additional H atom uptake and strong 1,3-butadiene chemisorption. Only one H adatom per unit cell on the V3 alloy is required for this passivation. Each H adatom is presumed to be adsorbed in a pure-Pt, 2-fold bridge site at the center of each unit cell to produce a saturation coverage of 0.33 ML. Each Pt at the surface has a nearest-neighbor H adatom. Obviously, electronic structure calculations are needed to validate and establish such proposals, but we hope these arguments stimulate such work. Another significant difference in the chemistry of coadsorbed hydrogen on Pt(l 11) compared with that on the two Sn/Pt(l 11) alloy surfaces is the reactivity for hydrogenating 1,3-butadiene. On P t(lll), chemisorbed 1,3-butadiene is completely irreversibly adsorbed and completely decomposes to liberate LL and form surface carbon [19]. This chemistry is unaffected by coadsorbed hydrogen, and specifically no butene desorption occurs [18]. This may indicate that dehydrogenation occurs more easily on P t( lll) than hydrogenation, or possibly that the hydrogenation barrier of strongly chemisorbed 1,3-butadiene is larger than the LL desorption barrier for coadsorbed H adatoms. The adsorption energy of 1,3-butadiene is decreased on the two S n /P t(lll) alloy surfaces, and alloying has an even stronger affect on 1,3- butadiene decomposition, i.e., decomposition is nearly eliminated on the (2x2) alloy 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and completely eliminated on the V3 alloy [19]. This is due to an increase in the C-H bond breaking barrier. On both alloys containing preadsorbed hydrogen, significant amounts of butene desorption from the hydrogenation of 1,3-butadiene occurs. This activity is much larger on the (2x2) alloy, and 100% conversion of 1,3-butadiene to butene is achieved at high values of 6h (near 0.5 ML). We did not determine which butene isomer (1-butene, cis- or /ram'-2-butene, or isobutene) was produced in our TPD experiments. Both 1-butene and cz'.s-2-butene desorb at low coverage on the (2x2) alloy at 270 K and on the V3 alloy below 220 K [26], Thus, butene desorption at temperatures higher than these in Figures 7.2 and 7.5 is rate-limited by hydrogenation reactions. Butene produced at 170-183 K by hydrogenation of weakly adsorbed, 7t-bonded species appears to be reaction rate- limited, but butene evolution at 140-145 K occurs at about the same temperature as butene desorbs from physisorbed layers and so we do not distinguish the limiting kinetics in this case. Reaction rate-limited butene desorption peaks in TPD can be used to estimate the hydrogenation reaction barrier Ea for strongly chemisorbed 1,3-butadiene on the two surface alloys, i.e., Ecf=Ea. The single butene desorption peak at 354 K on the (2x2) alloy with 6h =0.16-ML H corresponds to a hydrogenation reaction barrier Ea=91 kJ/mol. On the V3 alloy with 0H =0.13-ML H, the two butene desorption peaks at 282 and 348 K correspond to Ea =12 and 90 kJ/mol, respectively. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Surface-bound butenyl groups, either 1-butenyl (-CH2CH2CHNCH2) or 2 - butenyl (-CH(CH3)CH=CH2), are considered to be reaction intermediates in 1,3- butadiene hydrogenation on Pt surfaces at high pressure [12]. Although no spectroscopic evidence exists at this time for butenyl groups on the H-precovered Sn/Pt(l 11) alloys, it is reasonable to assume that hydrogenation occurs on these alloys through such intermediates. Our TPD data provide some clues that an additional path exists on the the V3 alloy. We observed two, reaction-rate limited butene desorption peaks and a high-temperature 1,3-butadiene desorption peak. In previous studies, surface-bound ethyl groups (CH3CH2-) were observed to produce ethylene and ethane on the a/3 alloy at 376 K through self-hydrogenation [28]. Surface-bound butenyl groups would be expected to react via [3-H elimination to produce simultaneously butadiene and butene. On the H-precovered V3 alloy, all hydrogen desorbs before 330 K and is thus not available for hydrogenation reactions at higher temperatures. This strongly suggests that the butene desorption peak at 348 K is from self-hydrogenation of surface-bound butenyl groups. This conclusion is also supported by observation of 1,3-butadiene desorption at 361 K. Apparently, high coverages of H adatoms are able to hydrogenate weakly bound, 7 t-bonded 1,3-butadiene species on both alloys. On the a/3 alloy at 9h=0.31 ML, a particularly weakly bound state of H adatoms exists on the surface, and these species recombine and desorb as H2 at 217 K, as presented in Figure 7.6. These species cause a new, weakly bound, 7t-bonded 1,3-butadiene species to form that 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. desorbs at 177 K, as shown in Figure 7.4, and are also implicated in the hydrogenation of these 1,3-butadiene species to form butene which desorbs in a new peak at 183 K. Both Sn/Pt(l 11) surface alloys have 100% selectivity in conversion of 1,3- butadiene to butene versus butane; butane was never detected as a desorbed product in these experiments. This observation is promising for applications of Sn-Pt bimetallic catalysts to selectively remove dienes from alkene feeds. This selectivity is caused in part by the decrease in the butene adsorption energy when Sn is alloyed to P t(lll). Butene desorbs before any hydrogenation reaction occurs because the butene desorption activation energy barrier is lower than that of the hydrogenation reaction barrier, based on our previous studies of ethylene hydrogenation [17]. Reactions on the two Sn/Pt(l 11) surface alloys at low O h can be summarized by the following scheme: (2x2) alloy C4H6 (ad) ------------ ► C4H6 (g ) 2H (ad) ► H 2 (g) H (ad) + C4H6 (ad) ^ C4H7 (ad) H (ad) + C4H7 (ad) C4H8 (g) V3 alloy 2H (ad) ► H 2 (g ) H (ad) + C4H6 (ad) --------------- ► C4H7 (ad) H (ad) " t" C4H7 (ad) ^ C4H8 (g) 160 283 K (1) 347 K (2) <354 K (3) 354 K (4) 272 K (5) <282 K (6) 282 K (7) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C 4H 6(ad )--------------- ► C 4H 6 (g) 285 K (8) 2C 4H 7(ad) --------------- ► C 4H6 (g) + C 4Hg (g) 348-361 K (9) Although 9C is larger on the V3 alloy than that on the (2x2) alloy at the same value of O h, the 1,3-butadiene conversion to butene is smaller on the V3 alloy because the H2 desorption activation energy on the V3 alloy is a little smaller than that for 1,3-butadiene. This causes more of the coadsorbed H to recombine and desorb as H2. On the (2x2) alloy, recombinative H2 and 1,3-butadiene desorption have similar activation energies, and this results in more extensive hydrogenation reactions. These arguments lead to the conclusion that hydrogenation reactions are maximized when 1,3-butadiene and recombinative H2 desorption have the same activation energies. In practical catalysts, this synergy might be exploited by controlling the Sn concentration in the supported Sn-Pt catalysts. 7.5 Conclusions Alloying Sn to Pt( 111) opens a new pathway for hydrogenation, compared to chemistry on P t(lll), to selectively produce butene during heating of coadsorbed layers of H and 1,3-butadiene in TPD. This hydrogenation reaction was observed on both of the two ordered, S n /P t(lll) surface alloys investigated. No further hydrogenation reactions to produce butane were observed. The activation energy barriers to hydrogenation of strongly chemisorbed 1,3-butadiene were estimated to 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. be 91 and 72 kJ/mol on the (2x2) alloy and V3 alloys, respectively, by considering the butene desorption that occurs at 354 and 282 K. 1,3-butadiene hydrogenation occurs in competition with 1,3-butadiene desorption and hydrogen adatom recombination to desorb H2, and the relative rates vary with O h and depends on the alloy surface. The highest conversion from 1,3-butadiene to butene was observed on the (2x2) alloy, and this is likely due to the nearly matching desorption activation energies of H2 and 1,3-butadiene on the (2x2) alloy, compared to that on the V 3 alloy. This leads to high coverages of both reactants at the highest temperatures, and this synergy might be exploited in practical catalyst by controlling the Sn concentration. Hydrogenation of 1,3-butadiene presumably proceeds through surface-bound butenyl groups as intermediates on both alloys to produce reaction rate-limited butene desorption at 282-354 K on both alloy surfaces. In addition, on the V3 alloy, self hydrogenation of surface butenyl groups also occurs during heating in TPD and produces butene desorption at 348 K. We also observed a small amount of facile, low-temperature hydrogenation of weakly bound, 7 1-bonded 1,3-butadiene to butene, which desorbed in reaction rate-limited peaks below 200 K corresponding to a hydrogenation activation barrier of 46 kJ/mol. Site-blocking effects by H adatoms on 1,3-butadiene adsorption were observed on both of the two S n /P t(lll) alloys, as we saw previously on P t(lll). However, the amount of preadsorbed hydrogen needed to complete block 1,3- 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. butadiene chemisorption on the two alloys was much less than that needed on Pt(l 11), i.e., 6*//=0.49 and 0.34 ML on the (2x2) and V 3 alloy, respectively, compared to 6ff= 0.91 on P t(lll). We propose a simple model for hydrogen adsorption to explain these changes. 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7.6 References [1] G.C. Bond, G. Webb, P.B. Wells and J.M. Winterbottom, J. Catal. 1 (1962) 4. [2] G.C. Bond, G. Webb, P.B. Wells and J.M. Winterbottom, J. Chem. Soc. A, (1965) 3128. [3] P.B Wells and A.J. Bates, J. Chem. Soc. A, (1968) 3064. [4] C.M. Pradier, E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 43 (1988) 177. [5] M. Primet, M. El Azhar and M. Guenin, Appl. Catal., 58 (1990) 241. [6] A. Sarkany, G. Stefler and J.W. Hightower, Appl. Catal. A: Gen., 127 (1995) 77. [7] J. Massardier, J.C. Bertolini, P. Ruiz, P. Delichere, J .Catal. 112 (1988) 21. [8] J. Qudar, S. Pinol, and Y. Berthier, J .Catal. 107 (1987) 434. [9] C.M. Pradier, E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 31 (1987) 243. [10] T. Ouchaib, J. Massardier, and A. Renouprez, J .Catal. 119 (1989) 517. [11] C.M. Pradier and Y. Berthier, J .Catal. 129 (1991) 356. [12] C. Yoon, M.X. Yang, and G.A. Somorjai, Catal. Lett. 46 (1997) 37. [13] D.F. Rohr, D.M. Haskell, F.M. Brinkmeyer, Eur. Pat. Appl. EP 211381 A1 19870225 (1987). [14] M. Galvagno, P. Staiti, P. Antonucci, A. Donato, and R. Pietropaolo, J. Chem. 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Soc, Faraday Trans. 19 (1983) 2605. [15] A. Palazov, Ch. Bonev, D. Shopov, G. Lietz, A. Sarkany and J. Volter, J. Catal. 103 (1987) 249. [16] Y. Park, F.H. Ribeiro, and G.A. Somorjai, J. Catal 178 (1998) 66. [17] H. Zhao and B.E. Koel, accept to Langmuir [18] H. Zhao and B.E. Koel, submitted to Surf. Sci. [19] H. Zhao and B.E. Koel, Surf. Sci. in press [20] H. Zhao, J. Kim, B. E. Koel, Surf. Sci., 538 (2003) 147. [21] S.H. Overbury, D.R. Mullins, M.T. Paffett, B.E. Koel, Surf. Sci. 254 (1991) 45. [22] T. Engel, K.H. Rieder, in: G. Hohler (Ed.), Structural Studies o f Surfaces, Springer, Berlin, 1982, Vol. 91, p. 55. [23] R.G. Wndham, M.E. Bartram, B.E. Koel, J. Phy. Chem. 92 (1988) 2862. [24] P. A. Redhead, Vacuum 12 (1962) 203. [25] M. T. Paffett, S.C Gebhard, R. G. Windham, and B.E. Koel, J. Phy. Chem. 94 (1990) 6831. [26] Y. Tsai and B.E. Koel, J. Phy. Chem. B 101 (1997) 2895. [27] C. Lutterloh, J. Biener, K. Pohlmann, A. Schenk, and J. Kiippers, S u rf Sci. 352 (1996) 133. [28] H. Zhao and B.E. Koel, Catal. Lett., accepted. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 8 Influence o f coadsorbed hydrogen on ethylene adsorption and reaction on a (V3xV3)R30°-Sn/Pt(l 11) surface alloy A bstract The effect of surface-bound hydrogen adatoms on adsorption, desorption, and reaction of ethylene (CI l2=CH2) on a (V3xV3)R30°-Sn/Pt(l 11) surface alloy with 0sn = 0.33 was investigated by using temperature-programmed desorption (TPD) and Auger electron spectroscopy (AES). Preadsorbed H decreased the saturation coverage of chemisorbed ethylene and less H was required to completely block ethylene chemisorption on this alloy than that on Pt(l 11). This is also the first report of extensive H site-blocking of ethylene chemisorption on P t(lll). Preadsorbed H also decreased the desorption activation energy of ethylene on the alloy surface. The reaction chemistry of ethylene on this Sn/Pt(l 11) alloy is dramatically different than on the Pt(l 11) surface: the H-addition reaction channel taking ethylene to ethane on Pt(l 11) is totally inhibited on the alloy. This is important information for advancing understanding of the surface chemistry involved in hydrogenation and dehydrogenation catalysis. 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8.1 Introduction Hydrogenation of ethylene (C2H4) to ethane (C2H6) over P t( lll) has been studied by a number of investigators as a model system for understanding the surface science of catalytic hydrogenation reactions [1-6]. On P t( lll) at 52 K under UHV conditions, ethylene chemisorbs molecularly via nearly complete rehybridization from extensive backbonding to eliminate the C-C 7 i-bond and form two Pt-C a- - 3 bonds [2]. This di-a-bonded ethylene species is sp -hybridized [7] and resides in a fee 3-fold hollow site on the P t( lll) surface, as determined by LEED [8]. Heating the surface causes di-a-bonded ethylene to begin to dehydrogenate to form surface- bound ethylidyne (CCH3) species at 240 K [9] via a pathway that includes isomerization to ethylidene (CHCH3) as first proposed by Koel and coworkers [10, 11] and later supported by others [9, 12]. Coadsorption of hydrogen adatoms and C2H4 on P t( lll) weakens the chemisorption bonding of ethylene to the surface and greatly increases the production of ethane in TPD experiments [3,13]. Somorjai and coworkers [14] have studied ethylene hydrogenation at 300-370 K with the total pressure of 110 torr and proposed that the reaction mechanism of ethylene hydrogenation over Pt( 111) does not involve ethylidyne, which is a spectator species or is involved in a decomposition pathway. They also reported that relative hydrogenation rates of species adsorbed on P t( lll) at 295 K during ethylene hydrogenation at 1 atm are in the order of 7r-bonded ethylene » di-a-bonded ethylene » ethylidyne [8]. 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Commercially, other metal or metals are added as promoters to supported Pt catalysts. Pt-Sn bimetallic catalysts using a neutral support have been reported to exhibit high selectivity and catalyst stability for the dehydrogenation of light paraffins at elevated temperatures [15-17], Model studies of ethylene hydrogenation on unsupported Pt-Sn bimetallic surfaces reported that the reaction rate, i.e., ethylene turnover frequency (TOF), at 15 torr and 300 K reached a maximum at 0sn = 0.1. This corresponded to an activity enhancement of 75 % over that of clean P t( lll) [18]. Verbeek and Sachtler [19] reported years ago that ethylene chemisorption was weakened on polycrystalline Pt-Sn alloys compared to that on Pt. Later, surface science studies probed ethylene adsorption on well-defined, ordered Pt-Sn alloys [20, 21]. Alloying Sn and Pt weakens the interaction between ethylene and Pt at the surface, and ethylene is less strongly chemisorbed and less rehybridized from the gas phase on both a (2x2)Sn/Pt(lll) surface alloy with 9sn =0.25, with a composition corresponding to the (111) face of a bulk Pt3Sn crystal, and a (V3xV3)R30°Sn/Pt(lll) surface alloy with O sn 0.33 and a composition corresponding to a Pt2Sn surface. While the nature of the chemisorbed species can still be regarded as di-a-bonded ethylene, no decomposition of ethylene occurs on either of these two alloys during heating in TPD [20]. There are no reports of ethylene and hydrogen coadsorption on Pt-Sn alloys, but one might expect copious production of ethane on the two S n /P t(lll) surface alloys in coadsorption experiments. This could arise from a confluence of the large 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. increase in ethane production observed in TPD after H + C2H4 coadsorption on P t(lll), the higher relative hydrogenation rate reported for rc-bonded ethylene over that of di-c-bonded ethylene on Pt(l 11) along with the decrease in rehybridization of chemisorbed ethylene on Pt-Sn alloys back toward that in the gas phase, and the slightly weaker Pt-H bond on Pt-Sn alloys [22] which should create a more labile H adatom. In order to probe experimentally the influence of chemisorbed hydrogen on ethylene chemistry on Pt-Sn alloys, we have investigated the effect of hydrogen adatoms on adsorption, desorption, and reaction of ethylene on a (V3xV3)R30°- S n /P t(lll) surface alloy by using temperature-programmed desorption (TPD) and Auger electron spectroscopy (AES). We report these results herein. Because H2 does not adsorb on this alloy under UHV conditions [22], an H-atom source was used to generate the hydrogen-precoverages investigated. 8.2 Experim ental methods Experiments were performed in a three-level UHV chamber as described earlier [23]. The P t( lll) crystal (Atomergic; 10 mm dia., 1.5 mm thick) was prepared by using 1-keV Ar+ -ion sputtering and oxygen treatments (5xl0"7 torr O2, 900 K, 2 min) to give a clean spectrum using AES and a sharp (lxl) pattern in low energy electron diffraction (LEED). 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The (V3xV3)R30°Sn/Pt(lll) surface alloy was prepared by evaporating two monolayers of Sn onto the Pt(lll) crystal surface and subsequently annealing the sample to 830 K for 20 s. On this surface prepared as above, Sn is incorporated substitutionally into the surface layer to form an ordered alloy with 9sn =0.33, corresponding to a Pt2Sn surface. This surface is relatively “flat”, but Sn atoms protrude 0.02 nm above the surface-Pt plane [24, 25], Importantly, forming this alloy structure eliminates all pure-Pt, three-fold sites. For brevity herein, we will refer to the (V3xV3)R30°-Sn/Pt(l 11) surface alloy as the V3 alloy. A Pt-tube doser was constructed, based on the design of Engel and Rieder [26] as a pyrolytic source of gas-phase hydrogen atoms. The principal component is a bent Pt tube (1 mm O.D., 0.8 mm I.D.) into which a hole of 0.1 mm diameter was mechanically drilled. The tube was resistively heated up to 1275°C while water- cooling kept an adjacent Cu block cold. The temperature of the Pt tube was directly measured by an optical pyrometer that was calibrated by the temperature of the Pt(111) crystal sample, as measured by a Cr/Al thermocouple. The estimated relative accuracy of the pyrometer reading was ±5°C. The flux of H atoms obtained from this source with the Pt tube at 800°C and a background pressure of 5x10'8 Torr in the IT 9 1 UHV chamber was estimated to be 3x10 atoms cm' -s' . This value was obtained by assuming that the initial sticking coefficient of H atoms on Pt(l 11) at 100 K was unity and using the H adatom coverage, and subsequent H2 yield in TPD, produced from the well-known decomposition of ethylene on Pt( 111) [10]. 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H2 (Matheson; 99.99%) was introduced via a variable leak valve (Granville- Phillips) into the Pt-tube doser after passing through a liquid-nitrogen cooled, U-tube trap. C2H4 (Matheson; 99.99%) was introduced via a microcapillary array doser connected to the gas line through a leak valve. All of the exposures reported here are given simply in terms of the pressure measured by the ion gauge in the UHV chamber. No attempt was made to correct for the flux enhancement of the doser or ion gage sensitivity. Mass spectrometry performed in the UHV chamber showed no appreciable concentration of impurities in the source gases. For all TPD experiments, the heating rate was 3.6 K/s and all exposures were given with the surface temperature at 100 K. AES measurements were made with a double-pass cylindrical mirror analyzer (CMA). The electron gun was operated at 3- keV beam energy and 1.5-pA beam current. Coverages 0i reported herein are referenced to the surface atom density of Pt(l 11) such that 0pt =1.0 ML is defined as 1.505 xlO’W 2. 8.3 Results TPD was used to probe monolayer (saturation) coverages of ethylene on a clean and H-precovered a/3 alloy at 1 0 0 K. Possible gas-phase reaction products formed during TPD were monitored at 2, 26, 28, 3 0 amu, but only the desorption of H2 (2 amu) and ethylene (28 amu) were identified, as shown in Figure 8.1. The most 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significant finding of this investigation is that the reaction pathway that is so prominent on Pt(lll) [3,13] between coadsorbed H adatoms and ethylene resulting in the evolution of gas-phase ethane upon heating is totally blocked on the V3 alloy. Desorption spectra for the H2 and C2H4 products are shown in Figure 8.2 and 8.3. The desorption spectra of H2 was not perturbed to any significant extent by the subsequent coadsorption of 0.24-L ethylene, as shown in Figure 8.2. This ethylene exposure exceeds by at least 30% the amount required to saturate a “clean” V 3 alloy surface, i. e., with Or = 0, at 100 K. H2 desorption at low 0h occurs at 265 K in a peak that shifts to 253 K with increasing 0h- Upon saturation of this peak, a second peak at 215 K evolves at higher 0H . The features below 200 K are in part from the cracking of ethylene in the mass spectrometer. The H2 desorption peaks are about 15 K lower than D2 desorption peaks reported previously on the V3 alloy [22], which is attributed to kinetic isotope effects (in addition to different heating rates). 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 Sn/Pt(111) alloy TPD after 0.24-L C9H -i— » to c 0 * • — » c = 0.14 ML H preads 28 amu 2 amu 30 amu 400 500 600 300 100 200 Temperature (K) Figure 8.1 FL (2 amu), ethylene (28 amu), and ethane (30 amu) TPD spectra following 0.24-L ethylene exposures on a (V3xV3)R30° Sn/Pt(l 11). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V3 Sn/P t(111) alloy |-| T P D 253 after 0.24-L CQ H 265 215 Z D £ 03 CM preads (ML-) 0.30 if) d (D C 0.27 0.19 0.14 0.09 100 200 300 400 500 Tem perature (K) Figure 8.2 H2 TPD spectra following 0.24-L ethylene exposures on a ( V 3 x a/ 3 ) R 3 0 ° Sn/Pt(l 11) surface alloy at 100 K that has been precovered with varying amounts of H adatoms 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V 3 S n/P t(111) alloy C0hL TPD 162 after 0.24-L C0H 180 preads 0.30 Z 5 E CD CO C N 0.27 0.19 > > c/5 a CD 0.14 £ Z 0.09 100 200 300 400 500 Tem perature (K) Figure 8.3 Ethylene TPD spectra following 0.24-L ethylene exposures on a (V3xV3)R30° Sn/Pt(l 11) surface alloy at 100 K that has been precovered with varying amounts of H adatoms Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As shown in Figure 8.3 for small precoverages of H adatoms, subsequent ethylene exposures produced ethylene desorption in TPD in a peak at 170 K, which is 10 K lower than that from a clean V 3 alloy. With increasing 9h, ethylene desorption shifted to lower temperatures and eventually ethylene desorbed in a peak at 162 K for 0h> 0.19 ML. An important observation shown in Figure 8.3 is that the intensity of the ethylene desorption peak decreases with increasing 0 H , and we will expand on this observation below. The small desorption feature observed between 350 and 400 K in Figure 8.3 is from CO desorption from background CO adsorption. Assuming 13 1 first-order kinetics with a preexponential factor of 10 s' , the Redhead method [27] can be used to estimate a desorption activation energy Ed for ethylene. A value of Ed = 45.5 kJ/mol was found for C2H4 adsorbed on the clean V3 alloy, and this was decreased to 40.8 kJ/mol on the larger H precoverages on the V3 alloy. In cases such as this where no appreciable barrier to adsorption is expected to exist, the desorption activation energy is equal to the adsorption energy and one sees that hydrogen weakens slightly ( - 10%) the bonding interaction of coadsorbed ethylene to the alloy surface. 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -0.25 0.8 - 0.20 1 0.6 -0.15 g 0.4 - 0.10 CD 0.2 -0.05 0.0 0.00 0.1 0.0 0.2 0.3 0.4 0M (ML) H preads ' ' Figure 8.4 Influence of preadsorbed H adatoms on the monolayer (saturation) coverage of ethylene on a (V3xV3)R30° Sn/Pt(l 11) surface alloy at 100 K. Figure 8.4 summarizes the influence of preadsorbed H adatoms on the ethylene monolayer (saturation) coverage on the V3 alloy, as obtained by integration of the ethylene TPD peak areas. The saturation coverage of ethylene on the clean V3 alloy was set to 1.0 on the left axis, and this corresponds to an actual coverage of 0.27 ML [21]. The actual ethylene coverages are provided on the right-hand axis. 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ethylene coverage was decreased linearly with increasing amounts of preadsorbed H adatoms. For 0h = 0.30 ML, the ethylene saturation coverage is reduced to only 18% of that on the clean a/3 alloy. Extrapolation of the line given in Figure 8.4 indicates that the chemisorption of di-a-bonded ethylene will be completely blocked when 0.35 ML of H adatoms is preadsorbed on the a/3 alloy surface. In order to understand the quantitative aspects of the decrease in the saturation coverage of ethylene with increasing coverages of preadsorbed H on the a/3 alloy, we also carried out one coadsorption experiment on Pt(l 1 1 ) . AES was used to measure the amount of adsorbed ethylene on the surface with and without preadsorbed hydrogen. Figure 8.5 shows two AES spectra, one obtained after a saturation exposure of ethylene on clean Pt(lll) (bottom) and the other after the same ethylene exposure on a Pt(l 11) surface that had been precovered by 0.6-ML H (top). The C(272):Pt(237) ratio shown for the bottom spectrum corresponds to 0c = 0.50 ML (0 = 0.25 ML [10]) on clean Pt(l 11). This ratio decreases dramatically C2 H 4 from that on Pt(lll) to that of C(272):Pt(237) = 0.17 for ethylene adsorbed on 0.6- ML H precovered Pt(lll). This ratio corresponds to 0c = 0.15 ML (6 = 0.07 ('2 4 ML). This data suggest that H adatoms effectively block ethylene access to Pt sites on both Pt(lll) and the a/3 alloy. Such site blocking effects of preadsorbed H have also been observed for benzene [28] and 1,3-butadiene [29] chemisorption on Pt(lll). Preadsorbed hydrogen on Pt(lll) also blocks ethylene chemisorption sites, 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. but more hydrogen than on the alloy is required to eliminate chemisorbed ethylene on Pt(lll) at 100 K. On the V 3 alloy, preadsorbed 0h = 0.30 ML blocked 82% of ethylene adsorption sites while twice as much hydrogen, 0H = 0.60 ML, only blocked 70% of the ethylene adsorption sites on Pt(lll). This means that ethylene cannot chemisorb on Sn sites or on many of the remaining Pt sites on the V3 alloy. On the V 3 alloy, Sn only occupies 33% of the surface area and 67% is still available as Pt sites. Our results show that only 0.35 ML of preadsorbed H, which occupies one-half the available Pt sites, will totally block the adsorption of ethylene. This indicates that two Pt atoms are required for ethylene adsorption. AES after 0.24-L C,bL Pt(111) = 0.6 ML H preads C/Pt = 0.17 Clean C/Pt = 0.57 300 150 200 250 350 400 450 500 Kinetic Energy (eV) Figure 8.5 AES spectra obtained following adsorption of a monolayer (saturation) coverage of ethylene on clean Pt(l 11) (bottom) and a 0.6-ML H precovered Pt(l 11) surface (top). 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8.4 Discussion Hydrogen coadsorption weakens the interaction between ethylene and the (V3xV3)R30°-Sn/Pt(l 11) alloy surface. One effect is to cause a decrease in the adsorption energy of coadsorbed di-a-bonded ethylene, but this perturbation is rather small. The most significant, and unexpected, effect of hydrogen coadsorption is to cause a nearly 1:1 “site-blocking” of ethylene chemisorption. Preadsorption of 0.3- ML H is sufficient to eliminate population of di-a-bonded ethylene on the V3 alloy surface. There is a direct, large repulsive interaction between coadsorbed H adatoms and ethylene molecules on the V3 alloy [29]. Such a geometric or site-blocking effect has often been invoked to explain the action of a second metal atom, such as Sn or Bi, or other adatoms such as S, O, or C, on transition metal chemistry, but such an effect is rarely used in explaining the action of the relatively “small” H adatoms. The site-blocking behavior of H on the V3 alloy and Pt(l 11) is similar, although one might expect small differences to arise from the weaker, and thus longer, Pt-H bond on the alloy which allows the H adatom to sit higher at the surface or perhaps have a little additional negative charge, and thus larger diameter, interfering with the close approach of ethylene to Pt atoms at the surface needed to facilitate strong bonding interactions. The large difference in the amount of H needed on the V3 alloy and P t( lll) to completely eliminate sites for ethylene to chemisorb arises from the 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. influence of alloyed Sn. These surface Sn atoms directly eliminate Pt sites by their physical presence and also affect the electronic structure of neighboring Pt sites [30, 31]. Another possible influence is that preadsorbed H exerts an electronic or ligand effect in which H adatoms change the local density of states at Pt sites on the V3 alloy, and this inhibits backbonding so severely that di-a bonded ethylene cannot be formed. There is support for this notion. RAIRS has been used to show that di-cr- bonded ethylene on P d (lll) is converted to 7 t-bonded ethylene when the surface is presaturated by H adatoms [32], First-principles calculation of ethylene and H coadsorption on Pd(l 10) [33] shows that the presence of coadsorbed H increases the 4-electron repulsion between the ethylene tc orbital and the 2 band of the metal and decreases the back-donation interaction between the metal d 2 orbital and the ethylene tC orbital. Because ethylene is already less rehybridized on the V3 alloy than on Pt(l 11), the addition of coadsorbed hydrogen could reduce the backbonding just enough to convert di-a-bonded ethylene to 7t-bonded ethylene, which is only physisorbed on the V3 alloy and therefore not adsorbed at 100 K (desorbing at 85 K on K/Pt(l 11) [34]). The other major finding of this work is the large difference in hydrogenation activity in these experiments between the P t( lll) surface and the V3 alloy. Coadsorption of H adatoms and ethylene on P t( lll) forms mostly ethane upon heating [3,13], This reaction pathway is totally blocked on the V 3 alloy. This is 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unexpected. Alloyed Sn in Pt(l 11) shifts the sp -hybridization of carbon atoms in di- a-bonded ethylene in the direction of sp2■ -hybridized carbon in 7 r-bonded ethylene [20] and it has been reported that 7 r-bonded ethylene hydrogenates to ethane much faster than di-a-bonded ethylene under catalytic conditions on Pt(l 11) [8], However, Domen’s recent IRAS findings [35] excluded the existence of 7 t-bonded ethylene under similar catalytic conditions and they concluded that the rate-determining step of hydrogenation does not involve forming either di-a-bonded or 7i-bonded ethylene (They proposed that activation of hydrogen plays an important role in ethylene hydrogenation under catalytic conditions). Our results have important implications for modeling and understanding catalytic hydrogenation over bimetallic Pt-Sn catalysts. Adding Sn to supported Pt catalysts causes a dramatic decrease in catalytic activity for the propene (C3H6) hydrogenation reaction [36]. The hydrogenation activity for ethylene and 1-hexene (CeHn) is also inhibited on Pt-Sn/AI2O3 catalysts [37]. Our surface science results are consistent with these observations, even though Pt-Sn alloy phases have not been identified as catalytically active in these catalysts. There has been one report in a model study that the addition of 0.1-ML Sn increased the reactivity of P t( lll) for ethylene hydrogenation [18]. However, these authors themselves pointed out the difficulty of distinguishing a direct effect from changes that occurred in the amount of carbon deposition, and reaction rate decreased quickly with further increases of Sn. We conclude that adding Sn to platinum catalysts decreases the alkene 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hydrogenation activity by decreasing the interaction between alkenes and Pt, but also more importantly by decreasing the reactivity of adsorbed alkenes at the catalyst surface. 8.5 Conclusions The hydrogenation reaction pathway so prevalent on P t( lll) between ethylene and coadsorbed hydrogen is completely inhibited on a (V3xV3)R30°- S n /P t(lll) surface alloy with 0sn = 0.33. Coadsorbed H on the alloy slightly weakens (-10%) the adsorption energy for ethylene on the this alloy. However, the chemisorbed monolayer saturation coverage of ethylene decreased linearly with increasing Oh on the alloy due to substantial site-blocking effects, which totally eliminate ethylene chemisorption at Oh = 0.35 ML. This is much less hydrogen than is needed to completely block the adsorption of ethylene on P t(lll). These observations represent important information for advancing the surface science of hydrogenation and dehydrogenation catalysis, and aid in explaining the decreased activity for alkene hydrogenation observed in supported bimetallic Pt-Sn catalysts. 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8.6 References [1] A. I. Boronin, V. I. Bukhtiyarov, R. Kvon, V. V. Chesnokov, and R. A. Buyanov, Surf. Sci. 258 (1991) 289. [2] A. Casuto, M. Mane, and J. Jupille, Surf. Sci. 249 (1990) 8 . [3] D. Godbey, F. Zaera, R. Yeates, and G. A. Somorjai, Surf. Sci. 167 (1986) 150. [4] T. S. Marinova and D. V. Chakarov, Surf. Sci. 192 (1987) 275. [5] G. A. Somorjai, Z. Phys. Chem. (Munich) 197 (1996) 1. [6] F. Zaera, T. V. W. Janssens, and H. Ofner, Surf. Sci. 368 (1996) 371. [7] H. Ibach and S. Lehwald, Surf. Sci. 117 (1982) 685. [8] P. S. Cremer, X. Su, Y. R. Shen, and G. A. Somorjai, J. Am. Chem. Soc. 118 (1996) 2942. [9] P. S. Cremer, C. Stanners, J. Niemantsverdriet, Y. Shen, and G. A. Somorjai, Surf. Sci. 328(1995) 111. [10] R.G. Windham, M.E. Bartram, and B.E. Koel, J. Phy. Chem. 92 (1988) 2862. [11] E.A. Carter and B.E. Koel, Surf. Sci. 226 (1990) 339. [12] R. Deng, E. Herceg, and M. Trenary, Surf. Sci. 560 (2004) L195. [13] P. Berlowitz, C. Megiris, J. B. Butt, and H. K. Kung, Langmuir 1 (1986) 206. [14] F. Zaera and G. A. Somorjai, J. Am. Chem. Soc. 106 (1984) 2288. [15] S. J. Miller, U.S. Patent 4,727,216 (1986). 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [16] F.M. Brinkmeyer and D.F. Rohr, U.S. Patent 4,866,211 (1987). [17] T. Imai and C. W. Hung, U.S. Patent 4,430,517 (1983). [18] Y. K. Park, F. H. Ribeiro, and G. A. Somorjai, J. Catal. 178 (1998) 6 6 . [19] H. Verbeek and W.M.H. Sachtler J. Catal. 42 (1976) 257. [20] M. T. Paffett, S. C. Gebhard, R.G. Windham, and B.E. Koel, Surf. Sci. 223 (1989) 449. [21] T. Tsai, C. Xu, and B.E. Koel, Surf. Sci. 385 (1997) 37. [22] M.R. Voss, H. Busse, and B. E. Koel, Surf. Sci. 414 (1998) 330. [23] H. Zhao, J. Kim, B. E. Koel, Surf. Sci. 538 (2003) 147. [24] S.H. Overbury, D.R. Mullins, M.T. Paffett, B.E. Koel, Surf. Sci. 254 (1991) 45. [25] A. Atrei, U. Bardi, G. Rovida, M. Torrini, E. Zanazzi, and P.N. Ross, Phy. Rev. B 46(1992) 1649. [26] T. Engel, K.H. Rieder, in: G. Hohler (Ed.), Structural Studies o f Surfaces (Springer, Berlin, 1982) Vol. 91, p. 55. [27] P. A. Redhead, Vacuum 12 (1962) 203. [28] C. Lutterloh, J.Biener, K. Pohlmann, A. Schenk, and J. Kuppers, Surf. Sci. 352 (1996)133. [29] H. Zhao and B.E. Koel, Surf. Sci. in press [30] S. Pick, Surf. Sci. 436 (1999) 220. [31] M. Batzill, D.E. Beck, and B.E. Koel, Surf. Sci. 466 (2000) L821. 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [32] D. Stacchiola, S. Azad, L. Burkholder, and W. T. Tysoe, J. Phys, Chem. B, 105 (2001) 11233. [33] J.S. Fihol, D. Simon, and P. Sautet, J. Phys. Chem. 107 (2003) 1604. [34] A. Cassuto, M. Touffaire, M. Hugenschmidt, P. Dolle, and J. Jupille, Vacuum 41 (1990) 161. [35] T. Ohtani, J. Kubota, J.N. Kondo, C. Hirose, and K. Domen, J. Phys. Chem. 103 (1999) 4562. [36] M. Galvagno, P. Staiti, P. Antonucci, A. Donato, and R. Pietropaolo, J. Chem. Soc, Faraday Trans. 79 (1983) 2605. [37] A. Palazov, Ch. Bonev, D. Shopov, G. Lietz, A. Sarkany, and J. Volter, J. Catal. 103 (1987) 249. 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 9. Proposed future work 9.1 Probing the C-H bond dissociation barrier. Determination of activation energies for breaking C-H bonds on Pt-Sn alloys is a necessary part of understanding the thermochemistry of hydrocarbon conversion on these surfaces. Because no alkane or cycloalkane molecules investigated so far were found to decompose on the (V3xV3)R30°-Sn/Pt(l 11) alloy, these values are not available yet and it is important to probe C-H bonding breaking barriers on this alloy surface with much heavier molecules than bicyclohexane (C 12H22). Larger alkane molecules have low vapor pressure, and a special sample transfer rod has to be constructed to deliver these molecules in UHV to the alloy. This is feasible and could be accomplished fairly easily. Our studies showed no reactivity of ethylene and 1,3-butadiene on the two S n /P t(lll) alloy surfaces. We explained that this is due to a lack of allylic (3-CH bonds. Only vinylic CH bonds exist in ethylene and 1,3-butadiene, and these bonds in gas-phase molecules have higher bond dissociation energies than allylic (3-CH bonds. One approach that could probe the C-H bond dissociation barrier for vinylic CH bonds would be to use some linear conjugated molecule containing only vinylic CH bonds, e.g., CftPL (CH2=CH-CH=CH-CH=CH2), in future experiments. Such molecules would be more strongly chemisorbed on the surface and may be stable to a sufficiently high temperature where thermal activation of C-H bond breaking could occur. Such experiments would let us measure for the first time the C-H bond 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dissociation barrier in vinylic and allylic CH bonds on these two S n /P t(lll) alloy surfaces and determine if there are any substantial differences due to the changing alloy structure. 9.2 Synthesis and characterization of new surface intermediate species prepared by H-atom addition reactions. Surface-bound vinyl (-CH=CH2) species have been proposed as important intermediates in ethylene decomposition [1-7] and acetylene hydrogenation [8 , 9] on transition metals. Yet, we currently have little or no knowledge about the structure and stability of such species. Addition reactions of gaseous H-atoms with adsorbed acetylene on P t( lll) and two S n /P t(lll) alloys provide an opportunity to cleanly synthesize and study surface vinyl groups by TPD and vibrational spectroscopy using either HREELS or IRAS. This information would greatly advance the understanding of the mechanism of ethylene decomposition and acetylene hydrogenation on these surfaces. The electrooxidation of methanol (CH3OH) over Pt or Pt-alloy anodes is at the heart of the operation of direct methanol fuel cells. Fundamental studies of methanol decomposition on well-defined Pt single-crystal surfaces in UHV, which serve as model anodes, have been performed to elucidate possible reaction pathways. Early studies reporting methanol dissocation on P t( lll) [10] were affected by defects and more recent work on better crystals show completely reversible methanol 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. adsorption on P t( lll) [11, 12]. Surface-bound methoxy (OCH3) species have been identified as intermediates and often considered to be the products of the first step in methanol decomposition on Pt(llO) [13], Pt(110)-(lx2)[14, 15], and oxygen- precovered P t( lll) and Pt(100) [16] under UHV conditions. However, recent DFT calculations indicate that C-H activation should be significantly easier than O-H activation and surface-bound hydroxymethyl (CH2OH) was predicted to be the first dehydrogenation intermediate on P t( lll) [17-18], This hydroxymethyl intermediate is more often observed in electrochemical environments than methoxy during methanol dehydrogenation [10], but it has never been observed or investigated on Pt under UHV conditions. Utilizing incident gaseous H atoms, we should be able to carry out H addition reactions with the double bond in adsorbed formaldehyde (CH2=0 ). These reactions will form surface hydroxymethyl species or surface methoxy species, depending on whether the incident H atom bonds to the O or C atom in the formaldehyde molecule. Due to the strong O-H bond, we believe that we have a good chance to synthesize and study the chemistry of surface hydroxymethyl species, which has never been done before. We plan to use IRAS in this work to be able to use the high resolution inherent in this technique to probe the mixture of species that may result on the surface and to unambiguously identify hydroxymethyl species. Once identified, the thermal stability of this species can be followed by IRAS and TPD. In 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. addition, surface hydroxymethyl species and methoxy species are two possible intermediates in the hydrogenation of formaldehyde [19]. Thus, synthesis and characterization of surface hydroxymethyl species, and additional studies of methoxy species, are important for advancing our understanding of the mechanisms of catalytic hydrogenation reactions. 9.3 Hydrogenation of C=C and C = 0 double bonds under UHV conditions. Our experiments have shown that alloying Sn into the P t( lll) surface eliminates the hydrogenation reaction between ethylene and coadsorbed H adatoms, which occurs copiously on P t(lll). Similar studies that could be carried out for the coadsorption of formaldehyde and H adatoms on P t( lll) and the two S n /P t(lll) alloys would be very helpful for understanding the mechanism of C=0 double-bond hydrogenation on these surfaces. The selective hydrogenation of a , p-unsaturated aldehydes to the corresponding unsaturated alcohols is an important industrial reaction, but little is known about the fundamental aspects of this chemistry on well- defined surfaces. Some results have emerged recently, but more studies are needed. The activity of two Sn/Pt(l 11) alloys was found to be two times higher than that of P t( lll) in the hydrogenation of crotonaldehyde (CH3CH=CHCHO) at 5 torr, but little difference in the selectivity was observed [20]. Coadsorption studies of crotonaldehyde or 2-propenal (CH2=CHCHO) and H adatoms on Pt(l 11) and the two 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sn/Pt(l 11) alloys would be useful for understanding these catalytic measurements and possibly predicting how to improve catalytic performance. 9.4 Hydrogenation of butadiene over S n /P t(lll) alloys at high pressure. Our surface science studies show that there is 100 % selectivity in forming gaseous butene, i.e., no butane is evolved, in the hydrogenation of butadiene on two Sn/Pt(111) alloys under UHV conditions. These observations suggest that the application of Sn-Pt alloy catalysts is promising for selectively removing dienes from olefin feedstreams, i.e., diene hydrogenation to form olefins without the detrimental loss of olefins due to olefin hydrogenation to form alkanes. Advancing these ideas and testing such a proposal requires new experiments on the hydrogenation of butadiene over the two S n /P t(lll) alloys at higher pressures. Such studies are readily possible using the UHV system that is interfaced directly to a high-pressure (up to 1 atm.) reaction chamber in our laboratory. In addition, use of the GC on the high-pressure reactor would allow for the complete separation and quantification of the butene isomer products. These were not resolved in our UHV studies. 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9.5 References [1] F. Zaera and R.B. Hall, Surf. Sci. 180 (1987) 1. [2] F. Zaera and R.B. Hall, J. Phys. Chem. 91 (1987) 4318. [3] X.Y. Zhu, M.E. Castro, S. Akhter, J.M. White, J.E. Houston, Surf. Sci. 207 (1988) 1. [4] F. Zaera, D.A. Fischer, R.G. Carr and J.L. Gland, J. Chem. Phys. 89 (1988) 5335. [5] T.E. Fischer and S.R. Kelemen, Surf. Sci. 69 (1977) 485. [6] T.A. Clarke, I.D. Gay, B. Law, and R. Mason, Faraday Disc. Chem. Soc. 60 (1975) 119. [7] E.M. Stuve, R.J. Madix., and C.R. Brundle, Surf. Sci. 152/153 (1985) 532. [8] S. Lehwald and H. Ibach, Surf. Sci. 89 (1979) 425. [9] J.E. Parmeter, M.M. Hills, and W.H. Weinberg, J. Am. Chem. Soc. 109 (1987) 72. [10] B. A. Sexton, K.D. Rendulic, and A. E. Hughes, Surf. Sci. 121 (1982) 181 [11] K. D. Gibson, and L. H. Dubois, Surf. Sci. 223 (1990) 59 [12] C. Panja, N. Saliba, and B. E. Koel, Surf. Sci. 395 (1998) 248 [13] K. Franaszczuk, E. Herrero, P. Zelenay, A. Wieckowski, J. Wang, and R.I. Masel, J. Phys. Chem. 96 (1992) 8509. [14] J. Wang and R.I. Masel, Surf Sci. 235 (1991) 199. 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. [15] J. Wang and R.I. Masel, J. Vac. Sci. Technol. A 9 (1991) 1879. [16] N. Kizhakevariam and E.M. Stuve, Surf. Sci. 286 (1993) 246. [17] S.K. Desai, M. Neurock, and K. Kourtakis, J. Phys. Chem. B 106 (2002) 2559. [18] J. Greeley, and M. Mavrikakis, J. Am. Chem. Soc. 126 (2004) 3910 [19] R. Hirschl, A. Eichler, and J. Hafner, J. Catal. 226 (2004) 273. [20] D.I. Jerdev, A. Olivas, and B.E. Koel, J. Catal. 205 (2002) 278. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bibliography Arnold, H.; F. Dobert and J. Gaube, In Handbook o f Heterogeneous Catalysis G. Ertl, H. Knozinger and J. Weitkamp, Eds, Wiley-VCH: Weiheim, Germany, 1997. Atrei, A.; U. Bardi, G. Rovida, M. Torrini, E. Zanazzi, P. N. Ross, Phys. Rev. B 46 (1992)1649 Avery. N. R. Langmuir 4 (1988) 445. Avery. N. R.; N. Sheppard, Proc. R. Soc. Lond. A 405 (1986) 1. Baetzold, R.C.; Langmuir 3 (1987) 189. Balakrishnan, K.; J. Schwank. J. Catal. 127 (1991) 287. Barias, O. A.; A. Holmen, and E.A. Blekkan Catal. Today 24 (1995) 361. Barias, O. A.; A. Holmen, and E.A. Blekkan J. Catal 158 (1996) 1. Bates, A. J.; Z.K. Leszcynski, J.J. Phillipson, P.B. Wells and G.R. Wilson, J. Chem. Soc. A (London) (1970) 2435. Batzill, M.; D. E. Beck, and B. E. Koel, Surf. Sci. 466 (2000) L821. Bent, B. E.; Chem. Rev. 96 (1996) 1361. Bent, B. E.; R.G. Nuzzo, B.R. Zegarski, L.H. Dubois, J. Am. Chem. Soc. 113 (1991) 1137. Berkowitz, J.; G. Ellison, and D. Gutman, J. Phy. Chem. 98 (1994) 2774. Berlowitz, P.; C. Megiris, J. B. Butt, and H. K. Kung, Langmuir 1 (1986) 206. Bertolini, J. C.; A. Cassuto, Y. Jugnet, J. Massardier, B. Tardy, and G. Tourillon, Surf. Sci. 349 (1996) 8 8 . Bodke, A. S.; D.A. Olschki, and L.D Schmidt, Science 285 (1999) 712. Boitiaux, J. P.; P. Cosyns, M. Derrien and G. Leger. Hydrocarbon Proc. 64 (1985) 51. 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bond, G. C. Heterogeneous Catalysis, 2n d ed. (Clarendon, Oxford, 1987), chaps. 8 and 9. Bond, G. C.; G. Webb, P.B. Wells and J.M. Winterbottom, J. Catal. 1 (1962) 4. Bond, G. C.; G. Webb, P.B. Wells and J.M. Winterbottom, J. Chem. Soc. A, (1965) 3128. Boronin, A. I.: V. I. Bukhtiyarov, R. Kvon, V. V. Chesnokov, R. A.Buyanov, Surf Sci. 258(1991)289. Brinkmeyer, F. M.; D.F. Rohr, U.S. Patent 4,866,211 (1987). Burch, R.; L. C. Garla. J. Catal. 71 (1981) 360. Carter, E. A.; B. E. Koel, Surf. Sci. 226 (1990) 339. Cassuto, A.; M. Mane, J. Jupille, Surf. Sci. 249 (1990) 8 . Cassuto, A.; M. Touffaire, M. Hugenschmidt, P. Dolle, and J. Jupille, Vacuum 41 (1990) 161. Ceelen, W.C.A.N.; A.W. Denie van der Gon, M.A. Reijme, H.H. Brongersma, I. Spolveri, A. Atrei, and U. Bardi, Surf. Sci. 406 (1998) 264. Chen, X.; B.E. Koel, Surf. Sci. 292 (1993) L803. Christmann, K.; G. Ertl, T. Pignet, Surf. Sci. 54 (1976) 365 Clarke, T. A.; I. D. Gay, B. Law, and R. Mason, Faraday Disc. Chem. Soc. 60 (1975) 119. Cortright, R. D.; E. Bergene, P. Levin, M. Natal-Santiago, J.A. Dumesic, Stud. Surf. Sci. Catal. 101 (1996) 1185. Cortright, R. D.; J. A. Dumestic, Appl. Catal. A 129 (1995) 101. Cortright, R. D.; J.M. Hill, and J.A. Dumesic Catal. Today 55 (2000) 213. Cortright, R. D.; P.E. Levin, and J. A. Dumestic Ind. Eng. Chem. Res. 37 (1998) 1717. 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cremer, P. S.; B.J. Mcintyre, M. Salmeron, Y.R. Shen, and G.A. Somojai, Catal. Lett. 34(1995) 11. Cremer, P. S.; C. Stanners, J. Niemantsverdriet, Y. Shen, and G. A. Somorjai, Surf. Sci. 328(1995) 111. Cremer, P. S.; X. Su, Y.R. Chen and G.A. Somorjai J. Phys. Chem. B 101 (1997) 6474. Cremer, P. S.; X. Su, Y. R. Shen, G. A. Somorjai, J. Am. Chem. Soc. 118 (1996) 2942. Dautzenberg, F. M.; J. N. Helle, P. Biloen and W. M. H. Sachtler. J. Catal. 63 (1980) 119. Davis, B. H. J. Catal. 46 (1977) 378. Delbecq, F.; P. Sautet, J. Catal. 220 (2003) 115. Deng, R.; E. Herceg, and M. Trenary, Surf Sci. 560 (2004) L I95. Derrien. M. In: L. Cerny, Editor, Studies in Surface Science and Catalysis 27, Elsevier, Amsterdam (1986) p. 313. Desai, S. K.; M. Neurock, and K. Kourtakis, J. Phys. Chem. B 106 (2002) 2559. Engel, T.; K.H. Rieder, in: G. Hohler (Ed.), Structural Studies o f Surfaces, Springer, Berlin, 1982, Vol. 91, p. 55. Fihol, J. S.; D. Simon, and P. Sautet, J. Phys. Chem. 107 (2003) 1604. Fischer, T. E.; S.R. Kelemen, Surf. Sci. 69 (1977) 485. Franaszczuk, K.; E. Herrero, P. Zelenay, A. Wieckowski, J. Wang, and R.I. Masel, J. Phys. Chem. 96 (1992) 8509. Fujikawa, T.; F. H. Ribeiro, and G. A. Somorjai, J. Catal. 178 (1998) 58. Galeotti, M.; A. Atrei, U. Bardi, G. Rovida, and M. Torrini, Surf. Sci. 313 (1994) 349. 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Galvagno, M.; P. Staiti, P. Antonucci, A. Donato, and R. Pietropaolo, J. Chem. Soc, Faraday Trans. 79 (1983) 2605. Gibson, K. D.; L. H. Dubois, Surf. Sci. 223 (1990) 59 Godbey, D.; F. Zaera, R. Yeates, G. A. Somorjai, Surf. Sci. 167 (1986) 150. Greeley, J.; M. Mavrikakis, J. Am. Chem. Soc. 126 (2004) 3910 Hickman, D. A.; L. D. Schmidt, Science 259 (1993) 343. Hill, J. M.; R.D. Cortright, and J. A. Dumestic, Appl. Catal. A-Gen. 168 (1998) 9. Hirschl, R.; A. Eichler, and J. Hafner, J. Catal. 226 (2004) 273. Hoffmann, H.; P. R. Griffiths and F. Zaera, Surf. Sci. 262 (1992) 141. Ibach, H.; S. Lehwald J. Vac. Sci. Technol. 15 (1978) 407. Ibach, H.; S. Lehwald, Surf. Sci. 117 (1982) 685. Imai, T.; C. W. Hung, U.S. Patent 4,430,517 (1983). Janssens, T.V.W.; D. Stone, J.C. Hemminger, F. Zaera, J. Catal. 177 (1997) 284. Jenks, C. J.; M. Xi, M.X. Yang, and B.E. Bent, J. Phys Chem. 98 (1994) 2152. Jerdev, D. I.; A. Olivas, and B.E. Koel, J. Catal. 205 (2002) 278. Jiang, L. Q.; A. Avoyan, B.E. Koel, and J. L. Falconer, J. Am. Chem. Soc. 115 (1993) 12106. Karpinski, Z.; J. K. Clarke, J. Chem. Soc. Faraday Trans. /, 71 (1975) 893. Kesmodel, L. L.; L.H. Dubois and G.A. Somorjai J. Chem. Phys. 70 (1979) 2180. Kesmodel, L. LP.C. Stair, R.C. Baetzold and G.A. Somorjai Phys. Rev. Lett. 36 (1976)1316. Kizhakevariam, N.; E.M. Stuve, Surf. Sci. 286 (1993) 246. Koel, B. E.; D.A. Blank, and E.A Carter, J. Mol. Catal. A: Chemical 131 (1998) 39. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Koestner, R, J.; J.C. Frost, P.C. Stair, M A. Van Hove, and G.A Somorjai, Surf. Sci. 116(1982) 85. Koestner, R. J.; M.A. Van Hove, and G.A Somorjai, J. Phy. Chem. 87 (1983) 203. Kuntze, J.; S. Speller, W. Heiland, A. Atrei, I. Spolveri, and U. Bardi, Phy. Rev. B 58 (1998) R16005. Lehwald, S.; H. Ibach, Surf. Sci. 89 (1979) 425. Li, Y. X.; J. M. Stencel and B. H. Davis. Appl. Catal. 64 (1990) 71. Li, Y.; M. R. Voss, N. Swami, Y. Tsai, and B. E. Koel, Phy. Rev. B 56 (1997) 15982. Liebmann, L. S.; L.D Schmidt Appl. Catal. A-Gen. 179 (1999) 93. Lieske, H.; J. Volter. J. Catal. 90 (1984) 96. Lin, J.-L; B. E. Bent, Chem. Phys. Lett. 194 (1992) 208. Lin, J.-L; B. E. Bent, J. Am. Chem. Soc. 115 (1993) 6943. Lin, J.-L; B. E. Bent, J. Phys. Chem. 96 (1992) 8529. Lin, J.-L; C.-M. Chiang, C. J. Jenks, M. X. Yang, T. H. Wentzlaff and B. E. Bent, J. Catal. 147 (1994) 250. Lin, J.-L; A. V. Teplyakov and B. E. Bent, J. Phys. Chem. 100 (1996) 10721. Liu, Z.-M.; X.-L. Zhou and J. M. White, Chem. Phys. Lett., 615 (1992) 615. Lloyd, K. G.; B. Roop, A. Campion and J. M. White, Surf. Sci. 214 (1989) 227. Lutterloh, C.; J. Biener, K. Pohlmann, A. Schenk, and J. Kiippers, Surf Sci. 352 (1996) 133. Marinova, T. S.; D. V. Chakarov, Surf. Sci. 192 (1987) 275. Massardier, J.; J.C. Bertolini, P. Ruiz, P. Delichere, J .Catal. 112 (1988) 21. 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. McCrea, K. R.; G.A. Somorjai,,/. Mol. Catal. A: Chemical 163 (2000) 43. Miller, S. J. U.S. Patent 4,727,216 (1986). Mittendorfer, F.; C. Thomazeau, P. Raybaud, and H. Toulhoat, J. Phy. Chem. B. 107 (2003) 12287. Morrison, R. T.; R.N. Boyd, Organic Chemistry, (Allyn and Bacon, Boston, 1973). Nakagoe, O.; N. Takagi, and Y. Matsumoto Surf. Sci. 514 (2002) 414. Newell, H. E.; M. R. S. McCoustra, M. A. Chesters and C. D. L. Cruz, J. Chem. Soc., Faraday Trans., 94 (1998) 3695. Oakes, D. J.; H. E. Newell, F.J.M. Rutten, M. R. S. McCoustra, and M. A. Chesters, J. Vac. Sci. Technol. A 14 (1996) 1439. Ogle, K. M.; J.R. Creighton, S. Akhter, J.M. White, Surf. Sci. 169 (1986) 246. Ohtani, T.; J. Kubota, J.N. Kondo, C. Hirose, and K. Domen, J. Phys. Chem. 103 (1999) 4562. Ouchaib, T.; J. Massardier, and A. Renouprez, J .Catal. 119 (1989) 517. Overbury, S. H.; D. R. Mullins, M. T. Paffett, B. E. Koel, Surf. Sci. 254 (1991) 45 Paffett, M. T.; S.C Gebhard, R. G. Windham, and B.E. Koel, J. Phy. Chem. 94 (1990) 6831. Paffett, M. T.; S.C Gebhard, R. G. Windham, and B.E. Koel, Surf. Sci. 223 (1989) 449. Paffett, M. T.; R. G. Windham, Surf Sci. 208 (1989) 34. Palazov, A.; Ch. Bonev, D. Shopov, G. Lietz, A. Sarkany and J. Volter, J. Catal. 103 (1987) 249. Panja, C.; N. Saliba, and B.E. Koel, Surf. Sci. 395 (1998) 248. Panja, C.; E.C. Samano, N. A. Saliba, and B.E. Koel, Surf. Sci. 553 (2004) 39. Pansoy-Hjelvik, M. E.; P. Schnabel, and J.C. Hemminger, J. Phys. Chem. B. 104 (2000) 6554. 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pansoy-Hielvik, M. E.; R. Xu, Q. Gao, K. Weller, F. Feher, and J. C. Hemminger, Surf. Sci. 312 (1994) 97. Parera, J. M.; N.S. Flgoli, in: Catalytic Naphtha Reforming, eds.G.J. Antos, A.M. Aitani and J.M. Parera (Dekker, New York, 1995)p. 45. Park, Y. K.; F. H. Ribeiro, and G. A. Somorjai, J. Catal. 178 (1998) 6 6 . Parmeter, J. E.; M.M. Hills, and W.H. Weinberg, J. Am. Chem. Soc. 109 (1987) 72. Peck, J.W.; B.E. Koel, J. Am. Chem. Soc. 99 (1995) 16670. Peck, J.W.; D.I. Mahon, D.E. Beck and B.E. Koel, Surf Sci. 410 (1998) 170. Peck, J.W.; D.I. Mahon, and B.E. Koel, Surf. Sci. 410 (1998) 200. Perry, D. A.; J.C. Hemminger, J. Am. Chem. Soc. 122 (2000) 8079. Pick, S. Surf. Sci. 436 (1999) 220. Ponec, V.; G.C. Bond, Catalysis by Metals and Alloys, Stud. Surf. Sci. Catal., Vol. 95 (Elsevier, Amsterdam, 1995). Ponec, V.; G. C. Bond, Stud. Surf. Sci. Catal. 95 (1995) 477. Pradier, C. M.; Y. Berthier, J .Catal. 129 (1991) 356. Pradier, C. M.; E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 31 (1987) 243. Pradier, C. M.; E. Margot, Y. Berthier and J. Qudar, Appl. Catal., 43 (1988) 177. Primet, M.; M. El Azhar and M. Guenin, Appl. Catal., 58 (1990) 241. Qudar, J.; S. Pinol, and Y. Berthier, J .Catal. 107 (1987) 434. Redhead, P. A. Vacuum 12 (1962) 203. Rettner, C. T. J. Chem. Phys. 101 (1994) 1529. Rodriguez, J. A.; C.T. Campbell, J. Phy. Chem. 93 (1989) 826. 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Rohr, D. F.; D.M. Haskell, F.M. Brinkmeyer, Eur. Pat. Appl. EP 211381 A1 19870225 (1987). Saeys, M.; M.F. Reyniers, and G.B. Marin, and M. Neurock, J. Phys. Chem. B 106 (2002) 7489. Salmeron, M.; G. A Somorjai, J. Phys. Chem. 8 6 (1982) 341. Sarkany, A.; G. Stefler and J.W. Hightower, Appl. Catal. A: Gen., 127 (1995) 77. Sexton, B. A.; K.D. Rendulic, and A. E. Hughes, Surf. Sci. 121 (1982) 181 Sinfelt, J. H.; Catalysis: Science and Technology, J.R. Anderson and M. Boudart, Eds., (Springer-Verlag, Berlin, 1981), vol. 1, Ch. 5, p. 258. Skinner,P; M. Howard, I. Oxton, S. Kettle, D. Powell andN. Sheppard J. Chem. Soc. Faraday Trans. 15 (1978) 407. Solymosi, F.; L. Bugyi and A. Oszko, Langmuir 12 (1996) 4145. Somorjai, G. A.; Introduction to Surface Chemistry and Catalysis (Wiley, New York, 1994), Chap.6 and 7. Somorjai, G. A.; Z. Phys. Chem. (Munich) 197 (1996) 1. Son, K. A.; J. L. Gland, J. Phys. Chem. B 101 (1997) 3540. Stacchiola, D.; S. Azad, L. Burkholder, and W. T. Tysoe, J. Phys. Chem. B, 105 (2001) 11233. Stuve, E.M; R.J. Madix., and C.R. Brundle, Surf. Sci. 152/153 (1985) 532. Syomin, D.; B.E. Koel, Surf. Sci. 492 (2001) L693. Tjandra, S.; F. Zaera, Surf. Sci. 140 (1995) 140. Tjandra, S.; F. Zaera, Surf. Sci. 289 (1993) 255. Tourillon, G.; A. Cassuto, Y. Jugnet, J. Massardier, J.C. Bertolini, J. Chem. Soc., Faraday Tran. 92 (1996) 92. Tsai, Y.; B. E Koel, Langmuir 14 (1998) 1290. 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tsai, Y.; B. E. Koel, J. Phy. Chem. B 10 (1997) 2895. Tsai, Y.; B.E. Koel, J. Phy Chem. B 101 (1997) 1895. Tsai, Y.; B.E. Koel, J. Phy Chem. B 101 (1997) 4781. Tsai, Y.; C. Xu, and B. E. Koel, Surf. Sci. 385 (1997) 37. Umezawa, K.; T. Ito, M. Asada, S. Nakanishi, P. Ding, W.A. Lanford, and B. Hjorvarsson, Surf. Sci. 387 (1997) 320. Valcarcel, A.; A. Clotet, J.M. Ricart, F. Delbecq, P. Sautet, Surf Sci. 549 (2004) 121. Verbeek, H.; W. M. H. Sachtler, J. Catal. 42 (1976) 257. VOlter, J.; G. Lietz, M. Uhlemann and M. Hermann, J. Catal. 68 (1981) 42. Voss, M. R.; H. Busse, B. E. Koel, Surf. Sci. 414 (1998) 330. Wang, J.; R.I. Masel, J. Vac. Sci. Technol. A 9 (1991) 1879. Wang, J.; R.I. Masel, Surf. Sci. 235 (1991) 199. Watson, G. W.; R.P.K. Wells, D. J. Willock, and G.J. Hutchings, J. Phys. Chem. B 104 (2000) 6439. Weinberg, W. H.; H.A. Deans and R.P. Merill Surf. Sci. 41 (1974) 312. Wells, P. B.; A.J. Bates, J. Chem. Soc. A, (1968) 3064. White, N.; PCTInt. appl. (1985) 56. Windham, R. G; M.E. Bartram, B.E. Koel, J. Phys. Chem. 92 (1988) 2862 Xi, M.; B.E. Bent, J. Vac. Sci. Technol. B 10 (1992) 2440. Xu, C.; B.E. Koel, Surf. Sci. 304 (1994) 249. Xu, C.; B.E. Koel, M.A. Newton, N.A. Frei, and C.T. Campbell, J. Phy. Chem. 99 (1995) 16670. 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Xu, C.; B. E Koel, and M. T. Paffett, Langmuir 10 (1994) 166. Xu, C.; Y. Tsai, and B. E. Koel, J. Phy. Chem. 98 (1994) 585. Yang, M.X.; B.E. Bent, J. Phys Chem. 100 (1996) 822. Yoon, C.; M.X. Yang, and G.A. Somorjai, Catal. Lett. 46 (1997) 37. Zaera, F. Acc. Chem. Res. 25 (1992) 260. Zaera, F. J. Am. Chem. Soc. I l l (1989) 4240. Zaera, F.; D.A. Fischer, R.G. Carr and J.L. Gland, J. Chem. Phys. 89 (1988) 5335 Zaera, F.; R.B. Hall, Surf. Sci. 180 (1987) 1. Zaera, F.; R.B. Hall, J. Phys. Chem. 91 (1987) 4318. Zaera, F. J. Phys. Chem. 94 (1990) 5090. Zaera, F. J. Phys. Chem. 94 (1990) 8350. Zaera, F. Surf. Sci. 219 (1989) 453. Zaera, F.; T. V. W. Janssens, H. Ofner, Surf. Sci. 368 (1996) 371. Zaera, F.; G.A. Somorjai J. Am. Chem. Soc. 106 (1984) 2288. Zhao, H.; J. Kim, and B. E. Koel, Surf. Sci, 538 (2003) 147. Zhao, H.; B.E. Koel, Catal. Lett., accepted. Zhao, H.; B.E. Koel, Langmuir, accepted. Zhao, H.; B.E. Koel, submitted to Surf. Sci. Zhao, H.; B.E. Koel, Surf. Sci. in press Zhao, H.; B.E. Koel, to be published. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Zhou, X.; P.M. Blass, B.E. Koel, and J.M. White, Surf. Sci. 271 (1992) 427. Zhou, X.; P.M. Blass, B.E. Koel, and J.M. White, Surf. Sci. I l l (1992) 452. Zhou, X.-L.; P. M. Blass, B. E. Koel and J. M. White, Surf. Sci. I l l (1992) 453. Zhu, X.Y.; M.E. Castro, S. Akhter, J.M. White, J.E. Houston, Surf. Sci. 207 (1988) 1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Chemisorption studies of oxygen-containing molecules on Au(211), Pt(111) and Sn-modified Pt(111) surfaces

PDF

Chemisorption on the (111) and (100) faces of platinum-tin bimetallic surfaces

PDF

Characterization of ordered tin-platinum surface alloys and tin-oxide overlayers on platinum(111) using ultra-high vacuum-scanning tunneling microscopy

PDF

Crotonaldehyde hydrogenation on platinum(111)-based model catalysts

PDF

Ion imaging studies of the spectroscopy and photodissociation dynamics of chloromethyl radical and nitric oxide dimer

PDF

Developments of chromium-free conversion coatings on galvanized steels

PDF

Deformation of bulk nanocrystalline aluminum alloys

PDF

Corrosion protection of high-copper aluminum alloys and stainless steels by surface modification

PDF

Electrochemical test methods for advanced battery and semiconductor technology

PDF

Evaluation of the superplastic potential in commercial aluminum alloys through equal -channel angular pressing

PDF

Development of steel foam processing methods and characterization of metal foam

PDF

Creep deformation in superplastic yttria -stabilized zirconia ceramics and yttria-stabilized zirconia /alumina composites

PDF

Application of electrochemical methods in corrosion and battery research

PDF

EFAB: A novel, high-aspect-ratio true three-dimensional microfabrication process for rapid, low-cost desktop micromachining of MEMS

PDF

A measurement of the temperature dependent work functions of alkali metal clusters and implications of the scaling law

PDF

Corrosion protection of anodized aluminum alloys and aluminum metal matrix composites and stainless steels

PDF

Dispersive properties in photonic crystals applications to beam-steering and wavelength demultiplexing

PDF

Adsorption, orientation and reaction intermediates for nitrogen dioxide-containing molecules on gold(111) surfaces

PDF

I. Layered nano fabrication. II. Adhesion layers for hippocampal neurons

PDF

Dispersive and nonlinear effects in high-speed reconfigurable WDM optical fiber communication systems

Asset Metadata

Creator Zhao, Haibo (author)

Core Title Chemisorption studies of hydrocarbons on clean and hydrogen precovered platinum(111) and tin/platinum(111) surface alloys

School Graduate School

Degree Doctor of Philosophy

Degree Program Chemistry

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry (physical chemistry),chemistry, analytical,engineering, materials science,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Koel, Bruce (committee chair), Goo, Edward (committee member), Reisler, Hanna (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-330593

Unique identifier UC11335729

Identifier 3155501.pdf (filename),usctheses-c16-330593 (legacy record id)

Legacy Identifier 3155501.pdf

Dmrecord 330593

Document Type Dissertation

Rights Zhao, Haibo

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA