1 MOLECULAR CHEMISORPTION OF O2 ON OXIDIZED Pd(111) By JOSE ANGEL HINOJOSA JR. A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007
2 2007 Jose Angel Hinojosa Jr.
3 To my wife Beverly, and my family.
4 ACKNOWLEDGMENTS I gratefully acknowledge Dr. He lena Hagelin-Weaver, Dr. Aravind Asthagiri, and Dr. Scott Perry for their guidance and interest in my edu cation. I thank my advisor, Dr. Jason Weaver for his patience, support, and words of encouragem ent. I also thank my research group members who have made my research a fulfilling experi ence. I also gratefully acknowledge financial support for this work provided by the Departme nt of Energy, Office of Basic Energy Science, Catalysis and Chemical Transformations Di vision through grant number DE-FG02-03ER15478 and the SEAGEP program.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF FIGURES................................................................................................................ .........6 ABSTRACT....................................................................................................................... ..............7 CHAPTER 1 INTRODUCTION................................................................................................................... .9 Literature Review.............................................................................................................. .......9 Experiment Equipment...........................................................................................................11 2 MOLECULAR CHEMISORPTION OF O2 ON OXIDIZED Pd(111)..................................14 Results and Discussion......................................................................................................... ..14 Oxygen Adsorption on 3 ML Oxide-covered Surface....................................................14 Mixed Isotope..................................................................................................................18 Oxygen Adsorption on Varied Oxide Coverages............................................................19 CO Oxidation...................................................................................................................22 Summary........................................................................................................................ .........26 LIST OF REFERENCES............................................................................................................. ..34 BIOGRAPHICAL SKETCH.........................................................................................................36
6 LIST OF FIGURES Figure page 2-1. O2 TPD spectra (heating rate = 1 K s-1) obtained for a clean Pd(111) surface and a 3 ML PdO surface. The initial PdO surface wa s prepared by an exposure to atomic oxygen at 500 K which was then exposed to a saturation coverage of molecular oxygen. The total desorption quantity from all channels below 400K for the clean and PdO surface are 0.42 ML and 0.26 ML respectively..................................................28 2-2. 18O2 TPD spectra (heating rate 1 K s-1) obtained by exposing Pd(111) to an atomic oxygen beam for 10 minutes at 500 K followed by various exposures of 18O2 at 85 K. The initial molecular oxygen adsorption va ried from 0.03 ML to 0.33 ML. The initial atomic oxygen coverage was cal culated to be 3.0 .20 ML............................................29 2-3. Mixed isotope O2 TPD spectra (heating rate 1 K s-1) obtained after exposing Pd(111) to an atomic oxygen beam at 500 K for 10 minutes. The sample was exposed to 0.015 L of 18O2 followed by a saturation of 16O2 at 85K. The calculated coverage for 16O2 and 18O2 below 500K was 0.14 ML and 0.06 respectively. A negligible amount of 16O18O was observed throughout the temperature range. Note the 16O2 spectrum has been scaled by a multiplication factor of 0.01 after 400 K..........................................30 2-4. Oxygen uptake curve obtained by plotting the molecular oxygen coverage as a function of the atomic oxygen coverage on Pd (111). Coverages were calculated from TPD spectra obtained by exposing Pd(111) to various amounts of atomic oxygen at 500 K followed by a saturation 16O2 exposure at 85 K......................................................31 2-5. 16O2 TPD spectra (heating rate 1 K s-1) obtained by exposing Pd(111) to various amounts of atomic oxygen at 500 K fo llowed by a saturation exposure of 16O2 at 85 K. Initial atomic oxygen coverages ra nge from 0.28 ML through 3.36 ML.....................32 2-6. TPRS traces (heating rate = 1 K s-1) obtained after initially exposing Pd(111) to an atomic oxygen beam for 10 minutes at 500 K. The oxidized surface was then exposed at 85 K to the following, A) a sa turation exposure of CO B) a 0.01 L of 18O2 followed by a saturation exposure of CO. The estimated initial atomic oxygen coverage was 3 ML for both TPRS traces.......................................................................33
7 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science MOLECULAR CHEMISORPTION OF O2 ON OXIDIZED Pd(111) By Jose Angel Hinojosa Jr. December 2007 Chair: Jason F. Weaver Major: Chemical Engineering In this work, we oxidized Pd(111) in ultr ahigh vacuum (UHV) using an O atom beam, and investigated the mol ecular chemisorption of O2 on the oxidized surface using temperature programmed desorption (TPD ). We find that the O2 molecules chemisorb readily on oxidized Pd(111) at 85 K, reaching a saturation covera ge of 0.33 ML on a surface covered with 3 ML (monolayer) of Pd oxide. The O2 TPD spectrum from the oxidized surface at O2 saturation exhibits four features centered at 118 K, 227 K, 275 K and 315 K associated with the desorption of chemisorbed O2 molecules. Comparison with O2 TPD from clean Pd(111) demonstrates that about half of the O2 molecules on oxidized Pd(111) are more strongly bound than on the metallic surface. Experiments with co-adsorbed 16O2 and 18O2 further reveal that O2 molecules dissociate negligibly on the oxidized surface. We also find that O2 molecules chemisorb only in small quantities (< 0.03 ML) on the p(2) and 2D oxide phases of atomic oxygen on Pd(111), indicating that these phases have much weaker binding affinities toward O2 than the 3D oxide (PdO) generated in our experiments. Finally, te mperature programmed reaction spectra of coadsorbed 18O2 and CO demonstrate that both PdO and molecularly adsorbed O2 are reactive in CO oxidation, with the molecular O2 exhibiting slightly higher re activity. The results of this study may have implications for understanding Pd oxidation catalysis at high pressures given
8 that we find relatively st rong binding states of O2 on oxidized Pd(111) and observe that these molecules are reactive toward CO.
9 CHAPTER 1 INTRODUCTION Literature Review Late transition metals are cr itical in industrial applica tions of oxidation catalysis. Palladium in particular is a highly effective catalyst for the oxidation of carbon monoxide [1-3] in automobile catalytic converters and methane ox idation [4-11] in lean gas turbines. There are two thermodynamically stable states for Pd depending on the environmental conditions. For instance, metallic Pd is predominate at high te mperature while the surface is oxidized at low temperature . Prior studies provide evidence that PdO is the more active state in the oxidation of methane [4-6,10]. For example, higher reaction rates are reported at the onset of the formation of PdO  as compared to the metal surface. Still other studies identify the metallic Pd surface as the active state [9,14,15]. These contradictory findings demonstrate the need for more detailed studies to a dvance the basic unders tanding of catalytic oxidation by Pd. The oxidation of late transition metal surfaces in ultrahigh vacuum (UHV) has proven to be challenging. Several approaches have been developed to produce high concentrations of atomic oxygen on transition metal surfaces in UHV. The traditional approach of using molecular oxygen provides a useful mechanism in studies that target the lo w coverage chemisorption range of oxygen on Pd. Coverages above this range could not be achieved on Pd by larger O2 doses . Stronger oxidatants, such as NO2 and O3, proved to be successful in producing coverages up to 2.4 ML on Pt  and 2.2 ML on Pd  Another approach, which utilizes a high pressure reaction cell, has the capability of pr oducing high oxygen covera ges in excess of 20 ML on a variety of Pd surfaces [ 17]. Atomic oxygen beams also produce high oxygen coverages on both Pd and Pt [18-21]. With these procedures further understanding of how oxidation occurs and effects on surface properties can be probed.
10 The oxidation of the Pd(111) surface has been studied in detail previously [12,13,17,2226]. Pd(111) oxidation first proceeds by chemisor ption of atomic oxygen to produce a p(2) structure at a surface coverage of 0.25 ML . At a coverage of 0.40 ML several stable states exist, of which one is the ( 67 67) R12.2 structure . A s light increase in the oxygen coverage causes a transformation into the two di mensional (2D) surface oxide, determined to be Pd5O4 . Transformation between 2D oxide (Pd5O4) and three dimensional (3D) oxide (PdO) is an active area of research. Current work in our laboratory has provided in sight on a previously unreported precursor state that leads to bulk oxidation. The precurs or state appear s to be oxygen chemisorbed on the 2D surface oxide, which is pr esent up to coverages as high as 2 ML [22,27]. Depopulation of the precursor state begins at 1.5 ML and leads to the onset of particles that resemble the 3D bulk oxide . Above this coverage range the surface is predominantly covered by 3D oxide . Establishing an oxidized Pd surface in UHV conditions provides the opportunity to further probe different characteristics and intera ctions of the surface. Fo r instance, real-world applications of Pd catalysis occur in envi ronments with higher pa rtial pressure of O2 than those accessible in UHV. At these conditions, O2 could adsorb molecularly on the oxidized surface and play an important role in subsequent surface ch emical reactions. Further characterization of the oxidized surface can provide additional insight into these effects. The experiments discussed in this work evaluated the chemisorption of mol ecular oxygen and the oxidized surface. Key results from this study are that O2 molecules bind more strongly on the oxide surface than on the metallic surface, that O2 dissociation is negligible on the oxidized surface, and that O2 chemisorbed on the oxide is reactive toward CO.
11 Experiment Equipment The following provides details of the equipmen t and procedures used in this study, with further information provided in previous st udies [18,21,28]. The UHV chamber is a three-level chamber that routinely reaches a base pressure below 2-10 Torr. Evacuation of the main chamber is provided by a combination of a 400 l/s ion pump, a 210 l/s turbo molecular pump, and a titanium sublimation pump contained wi thin a liquid-nitrogen -cooled cryoshroud. The chamber is equipped with instrumentation fo r X-ray photoelectron spectroscopy (XPS), Auger electron spectroscopy (AES), electron ener gy loss spectroscopy (EELS ), ion scattering spectroscopy (ISS), and low-ener gy electron diffraction (LEED). Th e chamber also contains a quadrupole mass spectrometer (QMS) used to m onitor the concentra tions of gases. The main chamber has an additional twostage differentially pumped chamber that contains an RF plasma source (Oxford Scientific Instruments). The plasma source is used to partially dissociate O2 (BOC gases 99.999%) into highly reactiv e O atoms used in the oxidation of Pd(111). The dissociation capability of th e plasma source has been estimated to be 20% . The first pumping stage cont ains oppositely-charged ion deflec tion plates and is evacuated with a 1200 l/s diffusion pump. A 3 mm conical skimmer connects the first and second pumping stages, directing the beam into a quartz tube 60 mm in length with an inner diameter of 6 mm. The second pumping stage is evacuated using a 70 l/s turbo molecular pump. A mechanical shutter attached at the end of the first pumping stage controls the exposur e of the surface to the atomic oxygen beam. The Pd (111) crystal used in this study is a circular disk 8 mm 1 mm purchased from Mateck GmbH. The crystal is polished on one si de and has a surface roughness of less than 3 m. The crystal is attached to a copper sample holder in thermal contact with a liquid nitrogen reservoir by spot-welding two tungsten wires to the backside of th e crystal. A K-type
12 thermocouple spot-welded to the backside of the cr ystal is used to measure sample temperature. Resistive heating, controlled by a PID controller connected to a DC power supply, provides a linear temperature ramp from 85 K to 925 K, utilized in TPD. The sample holder is attached to a XYZ manipulator that allows for accurate positioning of the sample. Initial sample cleaning entails sputtering with Ar+ ions at 600 eV and a sample temp erature of 900 K and then annealing to 1100 K. Routine cleaning consisted of exposing the sample at 856 K to the atomic oxygen beam for 30 minutes, followed by annealing to 92 5 K. The sample is positioned approximately 50 mm from the end of the quart z tube within the UHV chamber at a 45 rotation to provide uniform exposure across the crystal surface. Th is position is monitored frequently to avoid nonuniform oxidation of the surface. The sample was considered clean when CO was no longer detected during an annealing cycle. The focus of this study was molecu lar oxygen adsorption on an oxidized Pd(111) surface. The experiments conducted were oxygen ad sorption on a 3 ML oxide-covered surface, oxygen adsorption on a variety of atomic oxygen coverages, mixed isotope adsorption, and oxygen co-adsorption with CO. These experi ments required molecular oxygen and CO exposures as well as exposure to atomic oxyge n. Atomic oxygen exposures were conducted at 500 K for a specified duration to produce a desi red coverage, for example a 10 minute exposure routinely produced a 3 ML coverage. Molecu lar gases, exposed at 85 K had a typical 18O2 exposure and 16O2 exposure of less than 0.10 L and 61 ML respectively. TPD and TPRS measurements were obtained by placing the sa mple within 10 mm of the QMS ionizer and increasing the sample temperatur e at a linear rate of 1 K s-1. Data processing for TPD experiments entailed a straight line base li ne subtraction and 5 point adjacent averaging smoothing. Coverage ca libration spectra were obtained by saturating
13 the clean surface at room temperature. The first calibration spectrum invol ved a triple cycle of saturation exposures using both 16O2 and 18O2 gas. The areas under the curve of these spectra were equated to a 0.25 ML coverage , from which area-coverage factors could be extracted for 16O2, 16O18O and 18O2. The other calibration cove rage involved only using 16O2 gas to saturate the surface at room temperature and this calibra tion factor was used in experiment that only implemented 16O2.
14 CHAPTER 2 MOLECULAR CHEMISORPTION OF O2 ON OXIDIZED PD(111) Results and Discussion Oxygen Adsorption on 3 ML Oxide-covered Surface Shown in Figure 2-1 are O2 TPD spectra obtained from clean and oxide-covered Pd(111) after exposing these surfaces to saturation doses of O2 at 85 K. The oxide-covered surface had an initial atomic oxygen coverage of 3 ML. Considering differences in sample heating rates, our TPD spectrum from the O2-exposed metal surface agrees well w ith those reported in prior studies [30,31]. For the metal surface, we observe tw o sharp desorption features, labeled as 2 and 1, centered at 120 K and 179 K, respectively, as well as a third broader feature (not shown) at about 750 K. A previous study establis hed that the low temperature 1 and 2 features originate from molecularly chemisorbed O2 on Pd(111), while the desorption feature near 750 K results from the recombinative desorption of chemisorbed oxygen atoms . At 85 K, only molecularly chemisorbed O2 exists on Pd(111), but a fraction of th ese species dissociate as the sample is heated during the TPD experiment, giving rise to the recombinative desorption feature near 750 K. From our data, we estimate a total oxygen coverage of 0.66 ML (O/Pd) at O2 saturation of the metal surface at 85 K, with 0.42 ML evolving belo w 400 K. These values agree well with prior work by Guo et al . Notice also that we obtain a coverage of 0.24 ML of oxygen atoms produced by O2 dissociation, in excellent agreement with the known O-atom saturation coverage of 0.25 ML on Pd(111) obtained with O2 as the oxidant . After the O2 exposure at 85 K, the O2 TPD spectrum obtained from the oxide-covered surface exhibits two prominent features ( 2 and 1) with maxima at 118 K and 227 K, respectively, and shoulders, labeled as 1 and 2, centered at about 275K and 315 K, respectively. As discussed in more detail below, experiments using 18O2 show that these features originate
15 from molecularly chemisorbed O2, and that a negligible amount of the adsorbed O2 molecules dissociates on the Pd oxide during heating. The O2 TPD spectrum below 400 K is qualitatively similar to that obtained from the metal surface in that two main features are observed in each case. The 2 feature is less intense than the analogous 2 feature obtained from the metal surface, but both features appear at about 120 K in the TP D spectra and each is rath er narrow. In contrast, the 1 feature is much broader than the 1 peak, and reaches a maximum at a temperature nearly 50 K higher than the 1 state. This latter observation is intriguing since it implies that O2 molecules chemisorbed in the 1 state on PdO are more strongly bound than O2 chemisorbed on metallic Pd(111). Finally, the data shown in Fi gure 2-1 yields an oxygen coverage of 0.26 ML (O/Pd) for O2 that desorbs below 400 K from the oxide-cov ered surface. This va lue is lower than that for the metal surface, but it is a relatively la rge coverage in an absolute sense. As such, the strong O2 binding in the 1 state is difficult to rationalize in terms of O2 adsorption on oxide defect sites such as oxygen vacancies. Prior work using high resolution electr on energy loss spectroscopy (HREELS) reveal distinct differences in the O2 binding states on Pd(111) . Specifically, O2 chemisorbed in the 1 and 2 states exhibit O-O stretching frequencies of 650 cm-1 and 800 cm-1, respectively . The lower O-O stretching frequency of O2 in the 1 state suggests that the O-O bond is weaker for the O2 molecules that are more strongly bound to the surface. Imbihl et al.  proposed that binding in the 1 and 2 states involves each O-atom of an O2 molecule bonding to a different Pd atom versus the same Pd atom, respectively. Sinc e the O-O stretching frequencies fall within the range expected for peroxo linkages, these author s referred to the binding configurations in the 1 and 2 states as peroxo-II and peroxo-I, respectivel y. A superoxo state of O2 on Pd(111) has also been identified us ing HREELS  and is found to desorb at temperatures below the 2
16 peak maximum . However, we do not observe ev idence for this state in TPD, which suggests possibly that O2 molecules do not populate the superoxo state under the conditions we examined or that such species convert to the more strongly bound states before they can desorb during TPD. The data obtained in the present study does not provide information for determining the O2 binding configurations on the oxidized Pd su rface. However, considering the qualitative similarities between the O2 TPD spectra obtained from the metal and oxidized surfaces, it is conceivable that the 1 and 2 states are analogous to the pe roxo-II and peroxo-I configurations identified for O2 on Pd(111). If this is co rrect, the TPD data suggests that the peroxo-I species has a similar binding strength on the metal and oxi de surfaces, while the peroxo-II species is more strongly bound on the oxide. This would be an interesting difference between O2 chemisorbed on the metal versus oxide surfaces if O2 does actually bind on the oxide in the peroxo configurations. Spectroscopic work is cl early needed to further characterize the bonding of O2 on the oxidized surface. Interestingly, O2 chemisorbed on the oxide does not measurably dissociate during heating ev en though the strong O2-surface bond suggests a weakened O-O bond in the 1 state. Most likely, O2 dissociation is highly activat ed on the oxide. Finally, the small features ( 1 and 2) in the O2 TPD spectra may originate from a small concentration of O2 molecules that are strongly bound to defects on the oxide surface. Figure 2-2 shows 18O2 TPD spectra obtained after exposi ng oxidized Pd(111) held at 85 K to various doses of 18O2. The resulting 18O2 coverages are stated in the figure. Prior to the 18O2 exposures, we generated 16O coverages of 3 0.2 ML by exposing the Pd(111) sample to an 16O atom beam with the surface held at 500 K. As seen in Figure 2-2, 18O2 initially deso rbs in a broad feature centered at 275 K. This desorption featur e appears to consist of about three components
17 ( 1, 1, 2) centered at 250, 275 and 325 K, respectiv ely, in the TPD spectrum obtained from 0.03 ML 18O2. The 1 feature, initially centered at 250 K, inte nsifies considerably and shifts toward lower temperature as the 18O2 coverage increases, its maximum appearing at 233 K at a coverage of 0.16 ML. The 1 and 2 desorption features also intensify with coverage, but to a lesser extent than the 1 feature. In the spectrum obtained from 0.16 ML 18O2, the 2 peak is evident as a distinct shoulder at 325 K wh ile superposition between the 1 and 1 features appears to produce a single broad feature that is sk ewed toward high temperatures. Interestingly, the shift of the 1 maximum from 250 to 233 K suggests that O2 binding to the surface weakens with O2 coverage. For first order desorption, the TPD peak temperat ure is independent of the coverage if the activation energy is constant. As the coverage increases beyond 0.16 ML, a new desorption peak, labeled as 2, appears at 120 K. Unlike the 1 feature, the 2 peak temperature does not shift appreciably with increasing coverage, suggesting that interactions among chemisorbed O2 molecules have a negligible influence on desorption of the 2 state. Both the 1 and 2 peaks intensify concurrently with increasing coverage until the 18O2 layer saturates at 0.33 ML. No tice that this coverage is higher than that shown in Figur e 2-1, indicating that the oxidiz ed surface was not saturated in those experiments. The development of the hi gh and low temperature desorption features is similar to that observed previously on cl ean Pd(111) [30,31]. On the metal surface, the 1 state initially populates with incr easing coverage, and then the 1 and 2 states populate simultaneously beyond a critical coverage. While this may suggest similarities in the O2 binding states on the metal and oxidized Pd surfaces, th e high temperature TPD feature obtained from the oxidized surface is clearly compos ed of three, possibly more, distinct features. This suggests
18 greater variability in the bonding envi ronments or configurations for O2 on the oxidized surface compared with the metal. Mixed Isotope To determine whether the low temperature O2 desorption features arise from molecularly or atomically adsorbed species, we c onducted TPD experiments using co-adsorbed 16O2 and 18O2. In these experiments, we first dosed 18O2 onto the oxidized surface at 85 K to generate an 18O2 coverage slightly below 0.16 ML. We then exposed the surface to a relatively large 16O2 dose and performed TPD while monitoring masses 32, 34, and 36 amu. Representative TPD spectra are shown in Figure 2-3. The evolution of 16O18O is immeasurable below 400 K, confirming that the low temperature desorption feat ures indeed arise from molecularly adsorbed O2. The figure also shows the desorption signals above 500 K. Decomposition of the oxide produces the sharp 16O2 feature centered at 760 K as detailed in other work from our group . The small amount of 16O18O desorbing as the 3 ML oxide decomposes corresponds to approximately 0.005 ML of 18O. This is consistent w ith the natura l abundance of 18O2, and hence indicates that O2 molecules chemisorbed on oxidized Pd(111) dissociate to a ne gligible extent during heating. This is not surprising consid ering that Pd(111) cannot be oxidized with O2 at partial pressures typical of dosing in UHV. The TPD data also reveals that O2 molecules interchange among the various adsorbed states at temperatures as lo w as 120 K. Specifically, from integration of the TPD data, we estimate that the initi al exposures produced 18O2 and 16O2 coverages of 0.06 ML and 0.14 ML, respectively. Based on the data shown in Figure 2-2, the initial 0.06 ML of 18O2 molecules populate only the 1, 1 and 2 states, yet a fraction of the 18O2 from the mixed layer desorbs in the 2 peak (Figure 2-3). Thus, the addition of 16O2 to the initial 18O2 layer causes 18O2 molecules
19 to populate the 2 state. Notice also that the intensity of the 1 feature relative to the features is higher in the 18O2 spectrum of Figure 2-3 than that s een in the TPD spectrum obtained from a layer containing only 0.06 ML of O2 (Figure 2-2). This sugges ts that a fraction of the 16O2 molecules also displace 18O2 molecules from the states. The interchange of O2 among adsorbed states appears to be relatively facile si nce some interchange must occur below the 2 desorption temperature of 120 K. However, differences in the 1 to 2 peak ratios in the 16O2 and 18O2 TPD spectra demonstrate that the in terchange is not rapid enough to cause complete isotopic mixing. Interestingly, similar isotopic mixing behavior has been reported for O2 on clean Pd(111) [30,31]. Oxygen Adsorption on Varied Oxide Coverages To examine how the initial atomic oxygen phases on Pd(111) influence molecular O2 chemisorption, we conducted a series of O2 TPD experiments on surf aces with varying amounts of atomic oxygen. For these experiments, we first generated an atomic oxygen coverage on Pd(111) at 500 K using an 16O atom beam, and then exposed the 16O-covered surface held at 85 K to an 16O2 exposure of 61 ML. We found that a 61 ML exposure was more than enough to saturate the 3 ML oxide on Pd(111) with chemisorbed O2. The use of 16O2, rather than 18O2, for populating the molecular chemisorbed states was motivated largely by experimental convenience. In the experiments with 18O2, it was necessary to stop the 16O2 flow to the plasma source to avoid uptake of 16O2 prior to the 18O2 dose. By using 16O2 as the molecular adsorbent, the residual 16O2 uptake prior to the 16O2 dose does not affect the TPD measurements and enabled us to stabilize a plasma at the start of the day, and ma intain it for the duration of the experiments.
20 Figure 2-4 shows the saturation O2 coverage obtained at 85 K as a function of the initial O atom coverage. Briefly, the evolution of atom ic oxygen states on Pd(111) is known to initially involve O atoms arranging into a p(2) struct ure . Increasing the O atom coverage beyond 0.25 ML then causes a so-called two-dimensional (2 D) surface oxide to develop, which saturates at approximately 0.7 ML[ 23,25]. Prior studies show that the 2D oxide consists of a single layer of Pd and O atoms arranged on top of the Pd(111) surface [23,25]. The 2D oxide has a Pd5O4 stoichiometry, is incommensurate with the underlyi ng (111) lattice, and its structure does not match any lattice plane of crystalline PdO  Recently, our group has found evidence that O atoms adsorb on top of the 2D oxide as the oxygen coverage increases beyond 0.7 ML, and that these adsorbed O atoms react with the 2D oxi de to produce three-dimensional (3D) oxide particles that resemble bulk PdO . The 3D oxide particles grow on Pd (111) and coexist with O-atom covered 2D oxide domains as the co verage increases above 0.7 ML. Once the oxygen coverage reaches about 3 ML, the surface is predominantly covered by the 3D oxide, with 2D oxide covering, at most, a very small area of th e surface . Oxidation at 500 K effectively ceases at 3.4 ML for the O atom beam flux (~0.02 ML/s) employed in our experiments. However, we are able to generate higher oxygen atom coverages by oxidizing at surface temperatures above 500 K. Figure 2-4 shows that the O2 coverage obtained at 85 K depe nds sensitively on the initial atomic oxygen coverage. Initially increasing the coverage of chemisorbed O atoms to about 0.25 ML causes the O2 coverage to drop precipitously from 0.66 ML to about 0.03 ML. The O2 coverage remains low as the in itial O atom coverage increase s to about 1 ML, corresponding to growth of the 2D oxide a nd initial adsorption of O at oms on the 2D oxide. The O2 coverage increases nearly linearly with in itial O-atom coverage as 3D oxi de particles grow on the surface
21 from 1 to 2.7 ML. The O2 coverage then appears to rise more slowly in the O-atom coverage range from 2.7 to 3.4 ML, reaching a maximum of about 0.29 ML on the 3.4 ML oxide. Further increasing the atomic oxygen coverage apparently causes the saturation O2 coverage to decrease again. Specifically, we find that the saturation O2 coverage on surfaces with 8 and 19 ML of atomic oxygen is 0.20 ML, nearly 35% less than that obtained on the 3 ML oxide. We note, however, that the 8 and 19 ML oxi des were prepared at surface temperatures of 650 and 700 K, respectively, so it is unclear if the decrease in O2 uptake is associated with the preparation temperature, the initial O-atom coverage or both. For example, oxides prepared at higher surface temperature may be smoother and th ereby provide less surface area for O2 uptake. Focusing on the data obtained below 3.5 ML, it is clear that the 2D and 3D oxides have distinct affinities toward binding O2 molecules. This result indicat es that the 2D and 3D oxides offer chemically distinct binding si tes, and suggests that these oxide s are likely to exhibit distinct catalytic properties as well. In terestingly, our group has recently characterized the 2D and 3D oxides using ion scattering spectro scopy (ISS), and find that the ra tio of O/Pd ISS signals is about twice as high for the 3D oxide (at 3 ML) th an the 2D oxide . Assuming similarities in ion scattering cross sections a nd neutralization probabilities on the 2D and 3D oxides, this finding indicates that the concentr ation of exposed Pd atoms on the 3D oxide is roughly half that on the 2D oxide. Since the Pd concentration of the 2D oxide is 0.67 ML according to several prior studies [23,25], our ISS results predict a concentrati on of exposed Pd atoms of 0.33 ML for the 3D oxide at 3 ML total oxygen coverage. This value is remarkably cl ose to the saturation O2 coverage of ~0.3 ML (O/Pd) obtained on the 3 ML oxide, and hence implies that one surface Pd atom is available for each O-atom of the molecularly chemisorbed O2 at saturation.
22 Desorption of O2 below 400 K is shown in Figure 2-5 as a function of the initial O-atom coverage. For O-atom coverages between 0.25 and about 1 ML, only a small feature at about 120 K is evident in the TPD spectra. This feature is broader than the 2 peak that populates at high coverage on the 3D oxide, though they appear at a similar temperat ure. Considering that such a small amount of O2 desorbs in this regime, this O2 may originate from defects in the chemisorbed O-atom layer or the 2D oxi de. It is also possible that this O2 evolves from the edges or backside of the crystal. As the amount of 3D oxide increases with increasing O-atom coverage from 1 to 3.4 ML, O2 TPD features centered at 120 and 230 K simultaneously intensify. In fact, above an atomic oxygen coverage of 1.8 ML, the 1 to 2 intensity ratio remains approximately constant. Since 3D oxide particles grow in th e presence of 2D oxide domains, the available surface area of the 3D oxide incr eases with total oxygen coverage. Thus, the TPD data suggests that O2 molecules populate the same bi nding states at saturation of 3D oxide particles of varying size, and that the total O2 coverage scales with the surface area of the 3D oxide particles present at a given O-atom coverage. The dependence of the O2 saturation coverage on the initial O-atom coverage may provide insights for characte rizing the surface morphological changes during growth of the 3D oxide. CO Oxidation The present study was motivated largely by the idea that molecularly chemisorbed O2 could be a reactive species in cat alytic processes that occur on oxidized Pd at commercially relevant pressures, including th e catalytic oxidation of CO and CH4. Our finding that O2 molecules, in certain states, bind more strongly on PdO than Pd(111) supports the idea that molecularly chemisorbed O2 can exist in appreciable concen trations on oxidized Pd at high pressure and temperature, and is therefore a viable reactant. To initially explore the reactivity of
23 chemisorbed O2 on PdO, we examined the oxidation of CO on the oxidized Pd(111) surface using TPRS. For this study, we first exam ined the reaction of only CO with the 16O oxidized surface, and then studied the re action of CO co-adsorbed with 18O2 on the 16O oxidized surface. For each experiment, we first oxidized the Pd(111) sample at 500 K with an 16O beam to reach an 16O coverage of ~3 ML. We then exposed the oxidized surface held at 85 K to 0.01 L of 18O2, followed by a 26 ML dose of C16O. The same CO exposure was used in the experiments with and without co-adsorbed 18O2, but resulted in different CO coverages as discussed below. Finally, after preparing the adsorb ed layers, we heated the sample at a linear rate of 1 K s-1 while monitoring the partial pressures of C16O, 16O2, 18O2, C16O2, C18O16O and C18O2. Figure 2-6 shows representative TPRS traces ob tained in these experiments. The top panel shows the C16O2 desorption trace during TPRS with only adsorbed C16O, while the bottom panel shows the TPRS traces of C16O2, C18O16O and 18O2 obtained during experiments with coadsorbed C16O and 18O2. Before discussing the TPRS data it is important to clarify our quantitative analysis of this da ta and the rate scale used in Figure 2-6. As discussed above, we express the measured O2 desorption signals in un its of ML of O-atoms using the integrated signal obtained from Pd(111) with an assumed covera ge of 0.25 ML of O-atoms, which we prepare using a well established proce dure. To estimate the CO and CO2 yields in ML of these species, we rescaled calibration factors determined in a recent study of CO oxidation on Pt(100) conducted in our laboratory . Specifically, we assumed that the relationships among the O2, CO and CO2 scaling factors are the same in the pres ent study as in our wo rk with Pt(100), and then rescaled the CO and CO2 factors using the value that we more recently measured for O2. This procedure may introduce error in absolute CO and CO2 yields, but such error should largely cancel when considering relative values. Using thes e scaling factors, we estimate that the 26 ML
24 CO exposures produce initial CO coverages of 0.24 ML and 0.44 ML, respectively, on the oxidized surface with and without co-adsorbed 18O2. Thus, the adsorbed 18O2 appears to suppress CO adsorption on the oxidized surface, possibl y by blocking surface site s. Finally, to better compare the relative reactivity of th e oxidized surface with and without 18O2, we scaled the TPRS curves shown in Figure 2-6 by the initial CO coverages stated above. Figure 2-6 shows that both C16O2 and C18O16O evolve from the 18O2 covered surface, indicating that both PdO and chemisorbed 18O2 react with CO duri ng TPRS. Recall that chemisorbed O2 dissociates to a negligible extent on the oxidized surface (Figure 2-3) so the presence of co-adsorbed CO must facilitate cl eavage of the O-O bond, resulting in the addition of 18O to adsorbed C16O. Importantly, we find th at the evolution of C18O2 is negligible in these experiments, which implies that CO oxidati on on oxidized Pd(111) does not involve C-O bond cleavage. The C16O2 desorption trace obtaine d without co-adsorbed 18O2 exhibits three main features centered at about 117, 355, and 520 K (Fi gure 2-6A). In prior studies, we have found that desorption from the sample mounting wires as well as the initially high, transient heating rate are largely responsible fo r the sharp peak at 117 K [18,33] Nevertheless, reaction between CO and the oxide likely contributes to the trailing edge of this si gnal. The feature at 355 K most likely originates from reactions between CO and O atoms of PdO, while the feature at 520 K is characteristic of reaction between CO and O at oms adsorbed on metallic sites produced as the oxide is reduced. In addition to the initial sharp peak, the C16O2 spectrum obtained from the surface with adsorbed 18O2 exhibits a broad feature centered at 320 K, and only low intensity above 500 K (Figure 2-6B). Since the initial CO coverage was lower in this experiment, the majority of the CO molecules likely reacted below 500 K ther eby resulting in the diminution of the CO2 signal
25 observed at 520 K in the expe riment without co-adsorbed 18O2. It is interesting that the intermediate C16O2 feature appears at 320 K in the presence of 18O2, but is centered at 355 K and is more asymmetric when CO is adsorbed alon e on the oxide. Neglecting differences in the CO coverages, this may indicate that the presence of chemisorbed 18O2 influences the reaction of CO with the oxide. The C18O16O spectrum also exhibits a broad f eature at about 320 K as well as a less intense feature below 200 K. In this case, the intensity below 200 K, while small, is broad and overlaps the range of temperature over which chemisorbed 18O2 desorbs, suggesting that this CO2 originates mainly from reaction on the samp le surface rather than the mounting wires. Finally, we note that the highest C18O16O desorption rate occurs at 320 K, and mainly overlaps the trailing edge of the 18O2 desorption feature. Estimates of the product yields show that the oxidation of CO is relatively efficient on both surfaces studied. For example, on the surface without co-adsorbed O2, we estimate that 57% of the adsorbed CO molecules react with the oxide to produce CO2. In the presence of adsorbed 18O2, approximately 40% of the CO molecules react with the oxide, while 15% react with 18O to produce C18O16O. Interestingly, the total fr action of CO molecules that react is nearly equal on both surfaces for the conditions examined. The yiel ds show that fewer CO molecules react with 18O than 16O of the oxide. However, to put this in pe rspective, it is necessary to estimate the amount of 16O and 18O atoms that are present at the surf ace. From the desorption yields of C18O16O and 18O2, we estimate that the initial 18O coverage from 18O2 molecules was 0.18 ML, whereas the reaction occurred on an oxide with 3 ML of 16O atoms. However, only a fraction of the oxygen atoms are present at the surface of the oxide particles. Our re cent ISS results show that the 3 ML oxide has about twice the surface concen tration of atomic oxygen than the 2D oxide for which the O coverage is 0.58 ML acco rding to recent studies [23,25]. Thus, based on
26 the ISS results, we estimate that the 3 ML oxi de has an affective surface oxygen concentration near 1.2 ML. Most likely, this concentration exceeds 1 ML becau se the oxidized surface is rougher than the original Pd(111) surface. These estimates suggest that the surface concentration of 16O atoms of the oxide was more than si x times greater than the coverage of 18O in the molecular oxygen chemisorbed on the oxide duri ng the TPRS experiments. For comparison, the yield of C16O2 was approximately 2.7 times greater than the C18O16O yield, indicating that the probability for CO to react with 18O2 is higher than the CO reac tion probability with oxygen atoms of PdO for the conditions studied. These initial experiments cl early demonstrate that O2 molecules chemisorbed on oxidized Pd(111) are reactive toward CO. However, more thorough studies are needed to elicit the detailed aspects of this reactive interaction. For example, it is conceivable that co-adsorbed CO and O2 react directly on the oxide surface, possibly via a CO3 intermediate, to produce CO2 and an O atom. Another possibility, however, is that the reaction of CO w ith the oxide creates oxygen vacancies that are eff ective in activating the O2 bond, even though O2 dissociation on the oxide is otherwise negligible (v ide supra). Overall, the findings of these preliminary experiments provide substantial motivation for conducting fu rther investigations of the reactivity of O2 molecules on oxidized transition metal surfaces, particularly consideri ng the possibility that chemisorbed O2 is an active catalytic species in commercial applications. Summary In this work, we utilized TPD and TPRS to investigate the chemisorption and reactivity of O2 on oxidized Pd(111). The TPD data shows that O2 binds more strongly in certain states to the oxidized surface than to the clean Pd surface, reaching a saturation co verage of 0.33 ML on Pd(111) covered with 3 ML of oxi de. Experiments with co-adsorbed 16O2 and 18O2 reveal that O2 molecules dissociate negligibly on the oxidized surface. The desorption of molecularly
27 chemisorbed O2 gives rise to two main features cente red at 118 K and 227 K as well as smaller features at 275 and 315 K. Interestingly, we find that O2 molecules chemisorb only in small quantities on the 2D oxide on Pd(111), demonstra ting that the 2D and 3D oxides are chemically distinct at least with respect to binding O2. Finally, TPRS spectra pr ovide clear evidence that both PdO and molecularly chemisorbed O2 on the oxide are reactive toward CO, with the molecular species being slightly more reactive. This final, previously unreported observation warrants further investigation as it suggests that chemisorbed O2 molecules could play a role in catalytic oxidation processes occurring at high pressure.
28 Figure 2-1. O2 TPD spectra (heating rate = 1 K s-1) obtained for a clean Pd(111) surface and a 3 ML PdO surface. The initial PdO surface wa s prepared by an exposure to atomic oxygen at 500 K which was then exposed to a saturation coverage of molecular oxygen. The total desorption quantity from a ll channels below 400K for the clean and PdO surface are 0.42 ML and 0.26 ML respectively.
29 Figure 2-2. 18O2 TPD spectra (heating rate 1 K s-1) obtained by exposing Pd(111) to an atomic oxygen beam for 10 minutes at 500 K followed by various exposures of 18O2 at 85 K. The initial molecular oxygen adsorption va ried from 0.03 ML to 0.33 ML. The initial atomic oxygen coverage was cal culated to be 3.0 .20 ML.
30 Figure 2-3. Mixed isotope O2 TPD spectra (heating rate 1 K s-1) obtained after exposing Pd(111) to an atomic oxygen beam at 500 K for 10 minutes. The sample was exposed to 0.015 L of 18O2 followed by a saturation of 16O2 at 85K. The calculated coverage for 16O2 and 18O2 below 500K was 0.14 ML and 0.06 respec tively. A negligible amount of 16O18O was observed throughout the te mperature range. Note the 16O2 spectrum has been scaled by a multiplication factor of 0.01 after 400 K.
31 Figure 2-4. Oxygen uptake curve obtained by pl otting the molecular oxygen coverage as a function of the atomic oxygen coverage on Pd (111). Coverages were calculated from TPD spectra obtained by exposing Pd(111) to various amounts of atomic oxygen at 500 K followed by a saturation 16O2 exposure at 85 K.
32 Figure 2-5. 16O2 TPD spectra (heating rate 1 K s-1) obtained by exposing Pd(111) to various amounts of atomic oxygen at 500 K fo llowed by a saturation exposure of 16O2 at 85 K. Initial atomic oxygen coverages range from 0.28 ML through 3.36 ML.
33 Figure 2-6. TPRS traces (heating rate = 1 K s-1) obtained after initially exposing Pd(111) to an atomic oxygen beam for 10 minutes at 500 K. The oxidized surface was then exposed at 85 K to the following, A) a satura tion exposure of CO B) a 0.01 L of 18O2 followed by a saturation exposure of CO. The estim ated initial atomic oxygen coverage was 3 ML for both TPRS traces.
34 LIST OF REFERENCES  H. Gabasch, A. Knop-Gericke, R. Schlogl, M. Borasio, C. Weilach, G. Rupprechter, S. Penner, B. Jenewein, K. Hayek, B. Klot zer, Phys. Chem. Chem. Phys. 9 (2007) 533.  K.W. Kolasinski, F. Cemic, A. Demeije re, E. Hasselbrink, Surf. Sci. 334 (1995) 19.  I. Nakai, H. Kondoh, T. Shimada, A. Resta, J.N. Andersen, T. Ohta, J. Chem. Phys. 124 (2006) 224712.  J.N. Carstens, S.C. Su, A.T. Bell, J. Catal. 176 (1998) 136.  A.K. Datye, J. Bravo, T.R. Nelson, P. Atan asova, M. Lyubovsky, L. Pfefferle, Appl. Catal. A. 198 (2000) 179.  R.J. Farrauto, J.K. Lampert, M.C. Hobson, E.M. Waterman, Appl. Catal. B. 6 (1995) 263.  H. Gabasch, et al., J. Phys. Chem. C. 111 (2007) 7957.  G.B. Hoflund, H.A.E. Hagelin, J.F. Weaver, G.N. Salaita, Appl. Surf. Sci. 205 (2003) 102.  M. Lyubovsky, L. Pfefferle, Catal. Today 47 (1999) 29.  J.G. McCarty, Catal. Today 26 (1995) 283.  F.H. Ribeiro, M. Chow, R.A. Dallabetta, J. Catal. 146 (1994) 537.  G. Zheng, E.I. Altman, Surf. Sci. 462 (2000) 151.  H. Gabasch, et al., Surf. Sci. 600 (2006) 2980.  R.F. Hicks, H.H. Qi, M.L. Y oung, R.G. Lee, J. Catal. 122 (1990) 280.  S.H. Oh, P.J. Mitchell, R.M. Siewert, J. Catal. 132 (1991) 287.  N. Saliba, Y.L. Tsai, C. Panja, B.E. Koel, Surf. Sci. 419 (1999) 79.  J.Y. Han, D.Y. Zemlyanov, F.H. Ribeiro, Surf. Sci. 600 (2006) 2752.  A.L. Gerrard, J.F. Weaver, J. Chem. Phys. 123 (2005) 224703.  R.B. Shumbera, H.H. Kan, J. F. Weaver, Surf. Sci. 601 (2007) 235.  J. Wang, Y. Yun, E.I. A ltman, Surf. Sci. 601 (2007) 3497.  J.F. Weaver, J.J. Chen, A.L. Gerrard, Surf. Sci. 592 (2005) 83.  H.H. Kan, R.B. Shumbera, J.F. Weaver, in preparation.
35  H. Gabasch, W. Unterberger, K. Hayek, B. Klotzer, G. Kresse, C. Klein, M. Schmid, P. Varga, Surf. Sci. 600 (2006) 205.  J.Y. Han, D.Y. Zemlyanov, F.H. Ribeiro, Surf. Sci. 600 (2006) 2730.  E. Lundgren, G. Kresse, C. Klein, M. Borg, J.N. Andersen, M. De Santis, Y. Gauthier, C. Konvicka, M. Schmid, P. Varga, Phys. Rev. Lett. 88 (2002) 246103.  D. Zemlyanov, et al., Surf. Sci. 600 (2006) 983.  R.B. Shumbera, H.H. Kan, J. F. Weaver, Surf. Sci. 601 (2007) 4809.  A.L. Gerrard, J.J. Chen, J.F. Weaver, J. Phys. Chem. B. 109 (2005) 8017.  H.H. Kan, R.B. Shumbera, J.F. Weaver, J. Chem. Phys. 126 (2007) 134704.  X.C. Guo, A. Hoffman, J.T. Yates, J. Chem. Phys. 90 (1989) 5787.  T. Matsushima, Surf. Sci. 157 (1985) 297.  R. Imbihl, J.E. Demuth, Surf. Sci. 173 (1986) 395.  R.B. Shumbera, H.H. Kan, J.F. Weaver, J. Phys. Chem. C. submitted.
36 BIOGRAPHICAL SKETCH Jose A. Hinojosa Jr. was born on July 22, 1982, in Houston, Texas, to Jose A. Hinojosa and Adelaida Hinojosa. He graduated from South Houston High School in 2000 and continued his education at the University of Houston. In May 2005 he received his Bachelors of Science in chemical engineering. Jose and Beverly Brooks were married in July 2005 before continuing his education at the University of Florida. In December 2005 he began working for Dr. Jason F. Weaver, performing surface science research.