|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 DEVELOPMENT OF AN ULTRAHIGH VACUUM SYSTEM TO INVESTIGATE SURFACE CHEMICAL REACTIONS By SUHRITA MUKHERJEE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUI REMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Suhrita Mukherjee
3 To my family
4 ACKNOWLEDGMENTS I take this opportunity to thank my advisor Dr. Jason Weaver for guiding and supporting me throughout the course of this work. I would like to thank Feng Zhang for his patience and enthusiasm in answering all my questions no matter how silly they were. I friendship and support. I thank all my friends for making my graduate experience memorable. Last but not the least, I am grateful to my family for their continuous words of encoura gement and having faith in me at all times.
5 TABLE OF CONTENTS P age ACKNOWLEDGMENTS ................................ ................................ ............................ 4 LIST OF FIGURES ................................ ................................ ................................ .... 7 LIST OF ABBREVIATIONS ................................ ................................ ....................... 8 ABSTRACT ................................ ................................ ................................ ............... 9 CHAPTER 1 INTRODUCTION ................................ ................................ .............................. 10 Brief Background ................................ ................................ .............................. 10 Instrumentation and UHV Assembly ................................ ................................ 11 Techniques Used ................................ ................................ .............................. 14 Temperature Programmed Des orption and Reaction Spectroscopy .......... 14 Fourier Transform Infrared Spectroscopy ................................ ................... 16 2 TESTING THE NEW ULTRAHIGH VACUUM SYSTEM ................................ ... 21 Brief Background ................................ ................................ .............................. 21 Testing System for UHV Operating Conditions ................................ ................. 22 Testing Heating and Cooling System ................................ ................................ 24 Testing Plasma Source ................................ ................................ ..................... 26 Genera ting Oxygen Plasma ................................ ................................ ....... 27 Beam Composition of Plasma ................................ ................................ .... 31 3 TEMPERATURE PROGRAMMED DESORPTION (TPD) OF PdO(101) THIN FILMS ................................ ................................ ................................ ............... 34 Brief Background ................................ ................................ .............................. 34 Crystal Stru cture of Pd(111) and PdO(101) ................................ ...................... 35 Pd(111) ................................ ................................ ................................ ....... 35 PdO(101) ................................ ................................ ................................ .... 35 Mechanism of Oxidation of Pd(111) ................................ ................................ .. 36 Development of PdO(101) Thin Films on Pd(111) ................................ ...... 36 Adsorption of O 2 on PdO(101) ................................ ................................ .... 37 Experimental Details ................................ ................................ ......................... 38 Observations and Discussions ................................ ................................ .......... 40 Future Scope ................................ ................................ ................................ .... 42 Summary ................................ ................................ ................................ .......... 43 LIST OF REFERENCES ................................ ................................ ......................... 51
6 BIOGRAPHICAL SKETCH ................................ ................................ ...................... 53
7 LIST OF FIGURES Figure P age 1 1 UHV chamber assembly layout ................................ ................................ .... 18 1 2 UHV chamber at Larse n 125 used for all experiments ................................ 19 1 3 Example of O 2 TPD ................................ ................................ ...................... 20 2 1 Schematic diagram of the sample holder ................................ ..................... 33 2 2 Schema tic diagram of the setup used for generating oxygen plasma .......... 33 3 1 O 2 TPD spectra obtained from Pd(111) for higher exposure times of 40 minutes,30 min utes, 20 minutes and 10 minutes ................................ .......... 45 3 2 O 2 TPD spectra obtained from Pd(111) for lower exposure times of 5 minutes,3 minutes, 2 minutes, 1.5 minutes, 1minute and 30 seconds .......... 46 3 3 O 2 TPD spectra obtained from Pd(111) for all exposure times ..................... 47 3 4 Uptake plot obta ined for the while surface coverage of Pd(111) .................. 48 3 5 TPD spectrum at higher emission current about 3 minutes exposure to find 2D oxide coverage required for calibration ................................ ................... 49 3 6 IR spectral obtained from CO adsorption on PdO(101) at 100 K .................. 50
8 LIST OF ABBREVIATIONS DC Direct Current FCC Face Center ed Cubic FTIR Fourier Transform Infrared IR Infrared L Langmuir LEED Low Energy Electron Diffraction ML Monol ayer MS Mass Spectrometer PID Proportional Integral Derivative QMS Quadrupole Mass Spectrometer RAIRS Reflection Absorption Infrared Spectroscopy RF Radio Frequency STM Scanning Tunneling Microscope TPD Temperature Programmed Desorption TPRS Temperature Programmed Reaction Spectroscopy TSP Titanium Sublimation Pump UHV Ultrahigh Vacuum
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for th e Degree of Master of Science DEVE LOPMENT OF AN ULTRAHIGH VACUUM SYSTEM TO INVESTIGATE SURFACE CHEMICAL REACTIONS By Suhrita Mukherjee May 2013 Chair: Jason F. Weaver Major: Chemical Engineering In this work we tested the newly assembled ultrahigh vacuum (UHV) system along with the operating characteristics of the plasma source. These studies have been carried out by oxidizing a Pd(111) surface in UHV using atomic oxygen to generate a PdO(101) sur face. The PdO(101) surface has then been investigated to observe similar oxidation behavior as seen in previous studies by our group. Experiments have b een performed to calibrate the mass s pectrometer (MS) signal intensities to quantify the atomic oxygen c overage on the Pd(111) surface The results obtained from the temperature progra m med d esorption (TPD) spectroscopy of the P dO(101) thin film are compared to the ones obtained before to validate the plasma source
10 CHAPTER 1 INTRODUCTION Brief Background A catalyst is a substance which when added to a reaction, increases the rate of reaction or selectivity of the system by lowering the activation energy required for the reaction to occur. The characterization of catalysts are separated into two types namel y homogeneous and heterogeneous. A homogeneous catalytic reaction implies both the system and the catalyst are in the same phase while if the system and catalyst are in different phases the system is heterogeneous. The study of the chemical kinetics invol ving heterogeneous catalytic reactions is of great commercial importance especially those involving late transition metals like palladium, platinum, rhodium, iridium and ruthenium. These are primarily used for oxidation reactions such as combustion of natu ral gas, oxidation of organic compounds etc. They are expensive catalyst materials and so a deposition method is used to impregnate porous pellets with small amount of active material. Thus one gets a cata lyst with large surface area which reduce s the cost compared to having pure catalyst material. The catalysis starts with oxygen reacting with a metal to create a metal oxide layer which has very different chemical properties as compared to the metal, for example transition metal oxides used for me thane combustion. This results in significant effect in the performance of the catalyst. Hence, it is evident that the fundamental understanding of the development and properties of oxygen phases on Pd surfaces is very useful for predicting the catalytic b e havior of these materials [ 1 ; 2 ] Palladium (Pd) is used in pro cesses which require excellent oxidation properties like oxidation of carbon monoxide (CO) and methane that play vital role in lean gas
11 turbines and autom obile catalytic converters [ 3 ] Like the other transition metals, studies show that the process begins with the formati on of the palladium oxide (PdO) layer and the high activity of palladium oxide (PdO) inst igates a detailed investigation of the chemistry of PdO surfaces. However, the study of Pd oxidation has been limited by the technical difficulties in carrying out the process in ultrahigh vacuum (UHV).The UHV ensures atomically clean surfaces during the experiment and helps us observe the development of PdO(101) thin film grown on a Pd(111) surface. Despite this limitation, recent studies has shown significant insights for understanding the growth and properties of PdO and other high concentration oxygen phases on single crystal Pd surfa ces [ 4 6 ] .In this thesis work I will focus on developing and testing a UHV system that allows fundamental studies of the growth and surface che mistry of a PdO(101) thin film and other oxide films Instrumentation and UHV Assembly The UHV chamber we are working with h as base pressures i n the order of 10 10 Torr. This is achieved by a series of pumps used continuously. The main chamber we have is evacuated by a combination of roughing pumps, turbo pumps, ion pumps and titanium sublim ation pump (TSP) The ion pump and TSP are two types of entrainment pumps which utilizes a mechanism that captures gas molecules and binds them to the s urface of the pump or chamber [ 7 ] The ion pump generates an electric field to accelerate particles into the wall of the pump. On the other hand, the TSP functions by generating a thin coating of titanium inbuilt to the walls of the UHV chamber which creates a temporary active surface site for the gas molecules to attach. The chamber is also equipped with instruments like quadrupole mass s pectrometer (QMS) to monitor the concentrations of gases and Fourier transformation
12 infrared s pectrometer (FTIR) for observing chemical structure and adsorbate b onding. T here is a provision to install scanning tunneling m icroscope (STM) in future. The main chamber has a radio frequency ( RF ) plasma source. The plasma source is used to dissociate ultra high pure oxygen (O 2 ) into highly reactive O atoms which are use d in the oxidation of Pd(111) as well as for plasma cleaning of the Pd(111) surface. Figure 1 1 shows a brief layout of the setup used f or performing experimental study The Pd (111) crystal used throughout this study is a circular disk 10 mm 1 mm purcha sed from Mateck GmbH. The crystal is polished on one side and the surface thermal contact with a liquid nitrogen reservoir. The setup is also equipped with a K type thermocouple which is spot welded to the backside of the crystal and measures the sample temperature. There is also a proportional integral derivative ( PID ) controller connected to a direct current ( DC ) power supply which is used to provide a linear temper ature rise from 100 K to 930 K while performing experiment s like temperature programmed desorption (TPD) and temperature programmed reaction s pectroscopy (TPRS). In order to be able to move the sample to different positions for dosing and spectroscopy the sample holder is attached to a XYZ manipulator. In order to ensure the UHV has a controlled environment, every day we do a background scan with MS and observe the presence of gases with in the chamber. In a real UHV system there is always a small amount of background gases that are difficult to eliminate. To begin experiments, the first step is sample cleaning by exposing it to atomic oxygen from the plasma source at 850 K for 40 minutes and then flashing the
13 sample temperature to 930 K for annealing. The d istance between the sample and the plasma 0 mm and the sample is positioned at a 30 angle towards the tube to provide uniform exposure across the sample surface. The position is changed occasionally in case we observe non un iform oxidation of surface. Figure 1 2 is shows the assembly of the UHV chamber used for all experiments in this work. The main purpose of using a UHV system can be understood by calculating the number of molecular collisions that occur on a specific surfa ce over a period of time. From the kinetic theory of gas, the number of surface collisions can be correlate to the below relationships [ 7 ] : (1 1) (1 2) (1 3) W here, F = number of particles striking the surface; N = density of gas = average velocity of particles; K = Boltzmann constant T = temperature; m = molecular weight; P = pr essure On substituting the value of average velocity and gas density into the particle flux equation we can simplify to get: (1 4) Equation 1 4 is known as the Hertz Knudsen equation. It correlates the particle surface collision to the mass of the particle as well as the pressure and temperature of the environment. In order to get accurate detection in surface science techniques, one
14 requires the mean free path to be very large which definitely requires low environmental pressures. In order to analyze and compare the results one uses the unit of Langmuir (L) and monolayer (ML) which are equivalent to 10 6 Torr s and 1.5310 15 particles /cm 2 respective ly for a Pd(111) surface [ 8 ; 9 ] .The mean free path of a particle can be calculated using the following equation: (1 5) W here ean free path; d = molecular diameter From literature one can observe that the mean free path is sufficiently large at lower pressures which enables us to maintain a contamination free surface. Also the low gas flux at the sample helps us to keep it clean Thus, one uses UHV environment to maintain a clean surface throughout the experimental procedures. Techniques Used Temperature Programmed Desorption and Reaction Spectroscopy The technique that is used extensively in our UHV chamber is temperature programmed desorption (TPD) or temperature programmed reaction spectrometry (TPRS). If the resultant spectrum is from the species which are products of a surface reaction then it is called TPRS while if they are a result of just desorption without reaction it is called TPD. At first the sample surface is prepared by dosing O atoms through the plasma source. Then the temperature of the sample is increased at a rate of 1K/second in every experiment and the species desorbing from the sample surface is monitored using a mass spectrometer ( MS) The TPD spectrum thus obtained shows the rate of desorption with change in temperature. The shape of the peak gives us
15 insight on the binding energy of the desorbed species on the surfac e. Also kinetic parameters, such as activation e nergies, can be deduced from desorption spectra since the peak temperature shows the amount of energy required to break the surface bond. The method that is used for the analysis is called Redhead hich states that desorption kinetics follows the power law given by: (1 6) W here, r d = rate of desorption; N s = concentration of surface sites = species coverage; k 0 = pre exponential of the rate constant t = time; n = desorption order k = Boltzmann constant; E a = activation energy T = temperature The maximum desorption ra te can be determined and related to the activation energy by differentiating Equation 1 6 [ 9 ] (1 7) W here, ption rate; p = species coverage at the maximum peak temperat ure T p = maximum peak temperature We observe from equation 1 7 that we can learn details about the activation energy and surface coverage from the desorption spectra. For a first order desorp tion rate with respect to the concentration, the peak temperature T p would be independent of the surface concentration and would be expected to remain constant wit h increasing surface coverage [ 10 ] .For a second order desorption rate, the process will be dependent on the initial coverage thereby causing a shift in the peak desorption temperature.
16 One can also use TPD to calculate unknown surface coverage by correlating it to known coverage of a spectrum, since desorption rate is proportional to the initial coverage. This is done by integrating the area under the whole desorption rate curve for an initial surface coverage and using that to find the surface coverage at different exposures. This process is called the surface coverage calibration. Howev er, one needs to know the initial surface coverage to begin with, which is usually taken to be the saturation coverage. There are a number of ways in which a saturation coverage is measured and then used to relate to the other spectra. For example, a p(22 ) adsorbate structure of palladium oxide was observed at room temperature by LEED and STM thus conclusive of the fa ct that the 0.25 monolayer (ML) is the saturation coverage of O atoms when using O 2 oxidant on Pd (111) under UHV conditions [ 11 ; 12 ] Using this data once can directly correlate the other surface coverage to the known saturation coverage of 0.25ML and compare the data to see how much more adsorption occurred between different exposures. Figure 1 3 shows a typical example of a TPD spectrum. Fourier Transform Infrared Spectroscopy F ourier transform infrared spectros copy (F TIR ) is the preferred method of infrared spectroscopy. The term FTIR originates from th e mathematical steps used i.e. Fourier transform which is used to convert the raw data into the spectrum analyzed. At first the infrared (IR) radiation is passed through a sample of interest upon which some of the IR radiation is absorbed by the sample and some of it is transmitted and sensed by a detector. This results in a spectrum representing the molecular absorption and transmission and since no two unique molecular structures produce the same infrared spectrum IR spectroscopy is very useful for several types of analysis like identi fication of unknown materials and gases absorbed on a surface etc. In modern software
17 algorithms, IR is an excellent tool for quantitative analysis. In addition, the size of the peaks in the spectrum is a direct indication of the amount of material prese nt IR is emitted from a polychromatic infrared source which is a black body radiator. This collimated beam then passes through an aperture which controls the amount o f energy focused on the sample. Then the beam enters the Michelson interferometer. Through a motorized mirror the beam is periodically blocked and transmitted. Thus different wavelengths are modulated at different rates, which results in a beam with different spectrum. After exiting the interferometer the beam then focused into the sample from w here it is transmi tted through or reflected off the surface of the sample. This results in absorption of the specific frequencies of energy which are uniquely characteristic of the sample. Finally the resultant beam passes to the detector which mea sures th e interferogram signal. The measured signal is converted and sent to the computer where the Fourier transformation process takes place thereby presenting to the user the spectrum for interpretation and any further analysis.
18 Figure 1 1 UHV chamber a sse mbly l ayout
19 Figure 1 2 UHV chamber at Larsen 125 used for all experiments
20 Figure 1 3 Example of O 2 TPD
21 CHAPTER 2 TESTING THE NEW ULTRAHIGH VACUUM SYSTEM Brief Background In order to study the surface chemical reactions in the chamber, one has to achieve u lt ra clean and ultra high vacuum environment with a wide variety of com patible deposition sources and associated control modules In this section I will describe the significant challenges we f aced during the development of ultrahigh v acuum (UHV) system and attempts made to resolve them in order to prepare our UHV system for the future study of surface analysis. Our system for the current experiments has been primarily designed to carry out catalysis reactions of a model palladium crystal in UHV conditions. It is constructed of stainless steel and is equipped with surface analyzing tools. In order to study the reactions one needs to generate a clean crysta l surface followed by deposition of thin films if necessary and thereby study the ongoing surface chemical reactions. However a step by step testing in required to satisfy the operating conditions required for each procedure. To be able to start our experi ments we had to prepare our setup by rigorous testing. The important testing stages are as follows: Testing system for UHV operating conditions Testing heating and cooling system Testing plasma source Each of these testing phases has been described in th e subsequent sections with the ir importance in experiments, details of testing proc edures and measures taken to acquire the optimum conditions.
22 Testing System for UHV Operating C onditions The ultrahigh vacuum (UHV) chamber can be divided into two sections namely the main chamber and the beam chamber. The main chamber and the beam chamber are both pumped continuously by vacuum pumps. The main chamber is equi pped with the mass spectrometer ( MS) and potassium bromate windows for performin g spectroscopy and als o a XYZ manipulator which holds the model crystal of the desired sample in a three wedge sample holder. This manipulator can move the sample in all the three directions as well as rotate it 360. The beam chamber consists of the plasma gun and is different ially pumped by a turbo pump which is backed by a roughing pump. The pressures in both the chambers are monitored by hot filament ion gauges. The main chamber is also provided with ion pump and titanium sublimation pump ( TSP) to further reduce pressure. A small turbo is also set up to pump down the gas manifolds as and when necessary to recharge the lines with same or different gases. Since our system is meant to operate in ultra clean environment, maintaining ultra high vacuum is essential as it not only e liminates the possibility of contaminants, but also controls the flux of atoms when dozed into the chamber thereby ensuring uniform thin film deposition. Whenever a new vacuum system is assembled, it is critical to follow steps to reach the vacuum pressure s. Once the system is completely sealed, the roughing pumps are turned on first to bring the pressure from atmosphere to 10 3 Torr and once that pressure is attained we can turn on the turbo pumps backed by the roughing pumps to reduce the pressure to abou t 10 9 Torr. However, pumping alone is not sufficient to bring down the pressures in UHV range. We had to perform several bake out s of the chamber and turbo pumps through its heating jacket. Initially the system is baked using heating straps around the whole
23 exterior stainless steel body of the chamber at about 15 0C for a day or two to remove water and traces of gases that stick to t he surface of the chamber. The turbo pump heating jackets are turned on about 6 hours prior to the termination of bake outs. Outgassing from the surface of a chamber is a common problem in UHV systems. Outgassing from chamber wall is minimized by careful selection of construction material which is usually one with low vapor pressures. For laboratory high vacuum systems stainless steel is the preferred material. It has high yield strength and is easy to fabricate. The 300 series stainless steel is also know n to be corrosion resistant and non magnetic. These steels contain some percentage of chromium which forms chromium oxide layer on its surface which lowers outgassing rate and also increases oxidation resistance. There are also mathematical models availabl e to predict the outgass ing from stainless steel [ 7 ] However a t extremely low pressures, more gas molecules are adsorbed on the walls tha n those which are floating in the chamber. Thus, the total surface area inside a chamber is more important than its volume for reaching UHV. Water is a significant source of outgassing because a thin layer o f water vapor rapidly adsorbs on the chamber surf ace whenever the chamber is opened to atmospheric conditions Water evaporates from chamber surface too slowly to be fully remove d at room temperature, but fast enough to present a continuous level of background contamination while performing experiments Hence r emoval of water and similar gases which includes hydrocarbons generally require baki ng the UHV system above 150 C while vacuum pumps are running t o pump out the vapor from the system Another alternative could be to cool down the ch amber walls by liquid nitrogen. In our chamber we have use d bake outs to remove water vapor and trace gases
24 Even in well baked UHV systems, hydrogen and carbon monoxide ( CO) are the most common background gases observed while running back ground scan using mass spectrometer. Both h ydrogen and CO diffuse out from the grain boundaries in stainless steel. Helium could diffuse through the steel and glass from the outside air, but since He is usually negligible in th e atmosphere it is seldom observed in background scans Hydrogen and water vapor peaks are monitored every day in our chamber first thing before starting any experiment. If the peaks are way beyond the permissible limits, we sought to bake outs to reduce t he pressure and reduce the presence of these gases inside the chamber. Testing Heating and Cooling S ystem Once the system has attained the high vacuum conditions next we need to test the heating and cooling system of the sample. As mentioned before the man ipulator has the sample held on it and is equipped wi th features allowing sam ple heating and cooling Sample heating can be accomplished by electron bombardment or thermal radiation. In thermal radiation, the filament in the 3 wedge s ample holder is heated to high temperature s as a result of which the sample heats up through radiation For electron bombardment, the filament is used to emit electrons when biased at a high negative potential provided by an external source The sample is heated up due to the impact of the electrons bombarding the sample at high energy. Figure 2 1 shows a schematic drawing of the 3 wedge sample holder used in our UHV system to mount the palladium crystal. The sample holder is particularly designed to perform in situ STM analysi s in future. A sample holder is of key importance in our system since it provides the path way for supplying the high and low temperatures necessary during experiments without degrading itself. Our s ample holder is made of
25 oxygen free copper and gold plate d to resist oxidation. It holds single crystal of palladium currently. It is equipped with 6 electrical contacts 2 for temperature, 2 for heating, 2 for auxiliary use such as for high voltage supply to aid e beam heating. For our sample holder, the tungsten filament provides the radiative and e beam heating while the sample temperature is measured by an integral thermocouple which is in direct contact with the sample. Sample is mounted between sapphire wash ers, ensuring min imal thermal transfer to the sample holder during heating but maximizing thermal transfer duri ng cooling. The sample is also electrically isolated for e beam heating required for sample preparation and other testing experiments The cooling is regulated by copper braids which are in contact with liquid nitrogen flowing continuously with in the manipulator. Once the desired vacuum conditions are obtained, we need to eliminate the contaminants from the sample surface which might have been adsorbed on the surf ace while being exposed to the atmosphere prior to sealing the chamber. These contaminates could be water vapor hydrogen, carbon monoxide or other gases present in trace amount in the atmosphere. This is done using highly ionized oxygen plasma at around 8 50 K since at this temperature, oxygen will knock of the contaminants from the sample surface without any oxide formation as oxygen is observed to desorb from the palladium surface till 750 K from studies conducted by our research group [ 13 ] The plasma is generated from the plasma gun using grade four oxygen, details of which are describe d in the next section. Hence on e needs to heat the sample surface to 850 K and then reduce it back to desired temperature for depositing the thin surface oxide films on the sample. Thus second stage o f testing the UHV system is to test the
26 tungsten filament in order to ensure the desired temperat ures required for plasma cleaning and plasma oxide deposition are att ainable. This was done by varying the shape and size of filament constructed by tungsten w ire s of different diameter followed by testing them to see their behavior over high temperatur es as well as low temperatures. Apart from the filament breakage one other challenge faced during the testing was the deposition of tungsten on the sapphire surfa ce, generating a surface which could conduct heat. As a result there were h eat losses in the sample holder preventing it from reaching the desired temperatures as well as causing grounding issues due to expansion of metal surface in the sample holder. The filament constructed out o f 6 mil diameter tungsten wire with 4 coils was finally concluded to work best in our operating conditions. Optimizing the liquid nitrogen flow rate by pressurizing the liquid nitrogen container with gaseous nitrogen ensured cool ing rapidly to de sired temperature was possible. Also tuning the PID controller helped in controlling the linear ramp ups and downs of temperature. The sample can be heated linearly till about 750 K with the low voltage power supply and higher temperatures are obtained using high voltage supply which provides potential required for e beam heating. Currently our sample is capable to operating between temperatures of 92 K to 1200 K. Testing Plasma S ource The plasma source in our assembly is used for two purp oses firstly for plasma cleaning and secondly for generating the palladium oxide surface. In order to begin the study of surface chemical reactions, it is essential to obtain an ultra clean surface free from contaminants. As mentioned before the contamin ants could be water vapor, carbon monoxide, carbon dioxide etc. which might be present in minute amounts even
27 in a well pumped UHV system. These molecules tend to adsorb to the crystal surface resulting in blocking of the active catalysis sites as well as creating an uneven surface for any further studies. Hence to start our experiments, we needed to generate a clean crystal surface. This is done using the plasma source designed by Dr. Jason Weaver which uses molecular oxygen and radio frequency ( RF ) power source to generate a beam of atomic oxygen which is used for cleaning at a sample temperature of 850 K and deposition at temperatures below 700 K [ 13 ] If the gas used is oxygen, the plasma is an effective, economical, environm entally safe method for ultra high vacuum cleaning. Once a clean surface is generated, the sample is ready for further deposition experiments. Our chamber is built to study variety of surface reactions especially those involving alkanes. Since PdO(101) has been found to be one of the active phases of cataly tic combustion on palladium surface, our research als o requires the formation of PdO(101) films on Pd surface, upon which various CO and alkane reactions can be studied by dosing them in various amounts on active PdO(101) surface [ 13 ] This is achieved by using the same plasma source g enerating atomic oxygen plasma and particular sample tempe rature. However, to form the oxide surfaces, an elaborate testing of the plasma source for generating the desired plasma and beam composition of the generated plasma is required In the following s ections details of testing the plasma source has been described. Generating Oxygen P lasma A plasma is a highly ionized gas consisting of atoms, molecules, ions, electrons, free radicals, metastable and photons in the short wave ultraviolet range. As the a toms and molecules relax to their normal, lower energy states they rele ase a photon of light
28 which resu lts in the characteristic glow of a plasma of a particular gas The oxygen plasma generate d in our chamber has a light purple glow when fully developed. The plasma is generated in the plasma gun present in the beam chamber which is differentially pumped by a turbo pump Differential pumping allows to maintain the beam chamber pressure three order of magnitude higher in the ionization region of the gun tha n main chamber. For example in our system the beam chamber pressure during plasma cleaning is around 10 6 Torr while main chamber pressure is around 10 9 Torr when the plasma is operational. The plasma gun is equipped with RF power source, an auto tune controller, an ion trap controller and a matching network. Due to the evanescent wave coupling there are no electrodes present in the plasma The plasma i s entirely surrounded by a metallic shie ld. Hence the source is also suitable for use with reactive gases such as oxygen and hydroge n All joints are welded. The power source can be easily removed for bake ou t s at temperatures in excess of 1 00C The plasma is generated by inductively coupled pl asma source in which the energy i s supplied by the RF source which are produced by time varying magnetic fields also called ele ctromagnetic induction. The induction coil consists of several spiral turns, depending on t he RF power source characteristics. Co il parameters which are importan t for proper impedance include the coil diameter, number of coil turns and radius of each turn. In our plasma gun the coils are wrapped in a confinement tube with in which the induction plasma is generated along its cylindri cal axis. The coil is also supplied by cold water from a chiller running through it for cooling to mitigate high operating temperatures of the coils that
29 result from the high electrical currents required during operation of the plasma while plasma cleaning and deposition During our plasma operation, the power generator supplies an alternating current of radio frequency to the copper coil which induces an alternating magnetic field inside the coil according to Ampe (2 1) W here, B is the flux of magnetic field 0 7 Wb/A.m) I c is the coil current n is the number of coil turns per unit length r 0 is the mean radius of the coil turns. variation in magnetic f ield flux will induce a voltage or electromagnetic force: (2 2) Where, N is the number of coil turns, and the item in parenthesis is the rate at which the flux is changing. The plasma is extremely conductive by nature The electro magnetic force E will thus drive a current of density j in closed loops while the energy transferred to the plasma is dissipated via Joule heating, j 2 R according to e R is the resistance of plasma. While testing the pl asma source, the biggest challenge encountered was to obtain the plasma in the system by proper impedance match ing using the matching network. Several factors were required to be tested to finally obtain an operational
30 plasma for our experiments. A proper synchronization of the impedance matching along with power and oxygen pressure adjustments were done to finally obtain the plasma. Our RF generated oxygen plasma operates in two modes namely E and H modes [ 14 ; 15 ] The E mode is observed at low input po wer of about 180 W, where plasma ignites and is sustai ned by a capacitiv e coupling, while the H mode is observed at higher i nput powers of about 230 W, where the inductive coupling prevails. T he E mode at low input power is characterized by a relati vely low electron density and also low light emission, while the H mode operates at higher input power and is cha racterized by about two order higher e lectron density and mu ch brighter emission The transition fr om E to H mode in our system depends on the incoming oxygen pressure. For the E mode of oxygen plasma we observe dim white light while the H mode generates a bright purple glow. By increasing the input p ower, a transition from th e E to the H mode is seen In our system the E mode operates at 180 W and about 15 mTorr oxygen pressure while H mode operates at 23 0 W power and 7 mTorr oxygen pressure. All the cleaning and deposition experiments have been performed in the H mode of the oxygen plasma. Plasma is created by an RF generator coupled to t he coil through a matching network. In order to adjust the impedance of th e plasma gun to the output impedance of the power generator a matching network is used. Impedance matching is required mainly to provide maximum power transfer between the RF source and its load The second reason to use impedance matching is to ensure device protection since an imperfect match gives rise to reflected power. This reflect ed power builds standing waves on the transmission line between the source and load.
31 There are three basic matching networks use d in RF designs and T configuration. Each of them have their advantages and disadvantages. Most commonly used matching network in plas ma processing is the L circuit and the matching network in our chamber is also L circuit. It is mostly used because of its simplicity to build auto matching networks. It has only two components to be controlled for adjusting the real and imaginary part of the i mpedance. It can also be used for wide tuning ranges. By using capacitors and inductances we can achieve impedance matching with min imal po wer loss after several trials The most critical component in the matching network is the inductor. We had to modify the inductor coil by changing the coil turn number and the coil diameter with in the matching network in order to minimize power los ses and have good impedance. The RF power dissipated on the coil is much lower than the nominal power of the RF generator. There is our RF generator indicating the effectiveness of the power usage and it should be as low as possib le and usually close to zero. The RF generator operates at the standard frequ ency of 13.56 MHz and the impedance ratio is 50 Ohms. Figure 2 2 shows the schematic interiors of our plasma set up. Once the plasma is g enerated, it is focused through a quartz tube into the sample surface. Also the sample is placed at a certain angle in order to ensure homogeneity of deposition as well as cleaning. Beam Composition of P lasma After the plasma was fully functional, we tested the beam composition of the generated plasma. Since the plasma generated contain s both ions and neutral components, it is critical to see the compositio n of the beam to ensure it is uniform over time with respect to the neutral components which include atomic oxygen and molecular
32 oxygen. We al so wanted to check the molecular to atomic oxygen ratio at all times in order to check the efficiency of plasma formation in the confined area over time. Since we eventually wanted to form PdO(101) films, it was critical to get a uniform atomic oxygen comp osition in the incoming beam Initially, when we started the atomic oxygen content was less in the plasma beam, as a result the atomic flux was very less to give rise to a uniform thin oxide film. So we eventually added an ion trap to remove the residual ion content from the be am. Also we tested several sample positions in order to ensure that the incoming beam flux is hitting the sample and the deposition and cleaning is uniform. Sample is mounted about 20 mm away from the end of the quartz tube aperture and is said to be downstream of the plasma source and away from the most energetic species in order to minimize surface damage Ion energies are defined by the intrinsic pl asma potential and are around 13.62 eV for oxygen plasma
33 Figure 2 1. Schema tic diagram of the sample holder Figure 2 2. Schema tic diagram of the setup used for generating oxygen plasma
34 C HAPTER 3 TEMPERATURE PROGRAMMED DESORPTION (TPD) OF PdO(101) THIN FILMS Brief Background There are various therm odynamically stable states for p alladium depending on the oxygen environmental and temperature conditions used to perform the oxidation. It has been observed that at low temperature the surface undergoes oxidation while the metallic Pd dom inated at higher temperature [ 16 ] .There are evidences statin g that oxidized palladium, PdO, shows higher reaction rates as compared to metal surface and is the more active state in the oxi dation of methane [ 17 21 ] More recent in situ c haracterization experiments have also shown evidence of PdO as the activ e phase on Pd surface during oxidation of methane in the millibar pressure range [ 22 ; 23 ] .Hence a fundamental knowledge of the phase evolution of oxygen and the enhanced surface activity of PdO is needed to understand the catalytic oxidation by Pd. As said before, experimental challenges in oxidizing late transition metals like Pd in ultrahigh vacuum ( UHV) have limited investigations of the surface properties of PdO and other oxides. There are many methods to produce high concentrations of atomic oxygen on transition metal surfaces however the most common method invo lves the use of molecular oxygen. But this limits the study to chemisorbed range of oxygen on Pd and cannot be extended to other coverage ranges by just incr easing the exposure duration [ 16 ] Anot her method involves the use of high pressure reaction cell which produces oxygen coverages in ex cess of 20 ML on Pd surfaces [ 24 ] One can use atomic oxygen beams to produce high oxyge n coverages on Pd [ 24 ; 25 ] Also the use of stronger oxidants like NO 2 and O 3 has shown coverages up to 2.2ML on Pd
35 The mechanism of oxidation of Pd(111) has the following steps. At first the chemisorbed atomic oxygen pro duces a p(22) structure with a surface coverage of 0.25 ML [ 16 ] At 0.40 ML there are several stable states, one of w R12.2 structure. Overshooting slighting the oxygen coverage one observes the formation of the two dimensional (2D) oxide phase determined to be Pd 5 O 4 [ 26 ] and later the three d i mensional (3D) oxide (PdO). There is extensive research being done to explain the t ransformation of 2D oxide to 3D oxide. Our research group has provided insight on the precursor state which leads to bulk oxidation. The presence of precursor state is obser ved till surface coverage as high as 2 ML [ 27 ; 28 ] Above this range the surface is predomina ntly covered by 3D oxide [ 29 ] Crystal Structure of Pd(111) and PdO(101) Pd(111) All experiments in the present study were performed on a palladium single crystal purchased from MaTeck GmbH. The single crystal was cut with an accuracy of 0.1 of the (111) plane and then polished t o obtain a s urface roughness of less than reported that Pd forms a single crystal in a face center ed cubic (FCC) structure with the length of elementary cube within 3.86 and 3.95 [ 30 ] The expected step height or planar distance of a Pd(111) surface based on the elementary cube length is between 2.23 and 2.28 and the expected nearest neighbor distance is between 2.73 and 2.79 PdO(101) From earlier studies it has been found out that oxidation of Pd(111) under UHV condition with an atomic oxygen beam forms a ordered PdO(101) thin film The PdO crystal can be described as a tetragonal cooperate structure in space group 4 9 where
36 Z = 2 [ 30 ] The calculated l attice dimensions of the bulk terminated PdO(101) surface are 3.043 and 6.143 in the  and [ 101] directions [ 31 ] The expected step height or planar distance of a PdO(101) surface based on the lattice is found to b e 2.68 T he structure of the PdO(101) surface is composed of fourfold and threefold coordinated palladium and oxygen atoms where the palladium and oxygen surface atom density is calculated to be 0.70 ML and 0.35 ML corresponds to the concentration of uns aturated (cus) atoms. Recent studies indicate that this unsaturated atoms are expec ted to be more active sites Mechanism of Oxidation of Pd(111) Development of PdO(101) Thin F ilms on Pd(111) The bulk crystalline PdO has a tetragonal unit cell and consists of square planar units of Pd atoms coordinated with oxygen atoms. As mentioned before, t he stoichiometric PdO(101) surface consists of alternating rows of threefold or fourfold coordinated Pd or O atoms that run parallel. For UHV studies the O 2 dissociation on Pd(111) results in the formation of 2D (single ML) oxide which is also called the subsurface oxygen without the oxide layer progressing t o a multilayer PdO structure [ 32 ] The bulk PdO required O 2 pressures about 10 3 Torr for which one requires a high pressure cell attached to the UHV chamber however conta mination could be a problem Our research gro up found that PdO(101) thin film can be generated on Pd(111) in UHV by oxi dizing Pd at moderate temperature using ato mic beam from a plasma source Using a surface temperature of 500 K, it has been observed that the uptake of atomic oxygen on Pd(111) stops at coverage between ~3 and 4 ML and the PdO(111) thin film generated has a total thickness between ~11 and 15 This is the most common film used for experimental
37 study in our group however one can also generate thicker PdO(101) films by oxidizing Pd(111) at temperatures above 650 K using atomic oxygen beams. Adsorption of O 2 on PdO(101) Experimental studies have shown that oxygen chemisorbs on Pd(111) relatively strongly in molecular form but dissociates until O atom coverage is 0.25 ML durin g heating in UHV Oxygen atoms chemisorb on Pd(111) up to 0.25 ML where 1 ML is def ined as surface atom density of Pd(111) of 1.5 10 15 cm 2 [ 27 ; 33 ] Above 0.25 ML, 2D oxide develops on Pd(111) and saturates the surface at an oxygen atom density of about 0.70 ML. After the surface oxide phase saturates, further increasing the oxygen coverage results in the formation of the PdO precursor phase that transforms into a bulk PdO(1 01) thin film From temperature programmed d esorption (T PD ) measurements in Figure 3 2 one can observe that the precursor phase decomposes to give an oxygen desorption feature at 600 K but the 2D oxide phase on Pd(111) produces a sharp O 2 peak at ~750 K. D ecomposition of the PdO film initiates at about 630 K and the desorption rate reaches a maximum at 765 K for a heating rate of 1 K/s. It has been observed through experimental studies that PdO seeds influence the kinetics of oxide growth on Pd(111). Our re search group has identified two kinetic regimes for the oxidation of Pd(111). The main difference between the two regimes is the formation and growth of the PdO phase which could occur below or above the surface temperature at wh ich the PdO seeds are stabl e [ 31 ] It has been observed before that the oxidation at a sample temperature of 500 K produces stable PdO seeds which reach a maximum concentration at ~1.3 ML, and coexist with bulk PdO(101)
38 phase up to an oxygen coverage of ~2 ML. The PdO(101) film on Pd(111) reaches an effective saturation coverage of ~3 to 4 ML during oxidation at 500 K, and is characterized by large, flat crystalline domains. Compared to the oxidation at other surface temperature it has been observed t hat oxidation at 500 K produces a stable regime which leads to a more uniform PdO(101) thin film. It also allows the formation of surfaces containing stable PdO seeds in coexistence with either 2 D oxide or 3D PdO(101) oxide In this thesis, I will be focus ing on testing newly developed UHV chamber to perform Pd(111) oxida tion at varying oxygen coverage and at a surface temperature of 500 K and thereby conduct investigations of oxygen coverage on palladium crystal I will also show results to calibrate our m ass spectrometer ( MS ) signal intensities to quantify atomic oxygen coverages which will be required for future Fourier transform infrared spectroscopy ( FTIR ) experiments on PdO(101). Experimental Details As mentioned in the introduction, the UHV chamber used for all experiments in a horizontal chamber and the sample used is Pd(111) crystal which is a circular disk (10 mm ~1 mm) which is inserted into a three wedge sample holder. The sample was heated thr ough radiation and electron bombardment using a coiled tungsten filament and cooled through a thin copper braid attached to a liquid nitrogen cooled reservoir. A K type thermocouple is spot welded on the back side of the crystal which allows us to measure the sample temperature. A PID temperature controller attached to a DC power supply was used to maintain and ramp the temperature from 92K to 1200 K which is essential during the TPD measurements.
39 The sample is initially cleaned by atomic oxygen beam gener ated by the RF plasma so urce at sample temperature of 85 0 K for 40 minutes followed by flashing the sample to 930 K to desorb oxygen, carbon dioxide as well as other contaminants that may be present on the surface. We restrict the sample temperature to 930 K primarily to maintain oxygen saturation in the subsurface reservoir and also to ensure reproducibility in preparing the PdO(101) thin films us ed throughout this study After the cleaning procedure, the sample was exposed to oxygen at different exposures The initial starting point was to form the bulk like oxide by exposing the surface to an atomic oxygen beam at a sample temperature of 500 K, which was generated by dissociating research grade O 2 (Praxair 99.999%) The coverage experiments were carried o ut in the UHV chamber described in the introduction and the other equipment which is used in every experiment is the quadrupole mass spectrometer (QMS) to collect the desorption spectra. The oxygen atom beams were generated using a RF plasma source. The pl asma source was placed in single stage differentially pumped chamber such that the main chamber pressure remained three orders of magnitude lower than that in the beam chamber while the plasma source was operated. The plasma source has an ion trap which is basically two oppositely charged ( 10 kV) parallel ion plates fabri cated at the outlet of a quartz tube which removes ions and electrons from the beam. A mechanical shutter is attached at the end of the plasma source in order to control the beam exposure time during each experiment. Exposure s were carried out at different times between 30 seconds to 40 minutes as a result of which an oxide thickness ranging from ~0.1 ML to ~4 ML was obtained.
40 The bulk oxide PdO(101) was observed to grow from 2.7 ML to 4 M L which is comparabl e to the data obtained before [ 13 ] Another fact taken into consideration is that after the PdO(101) coverage was reached, further oxygen exposures only increased the oxide thickness. After the surface was prepared, TPD experiments were conducted by positioning the s ample such that it faces the QMS and temperature of the sample is ramped at a rate of 1K/second from 500 K to 930 K to observe the desorption spectra. The distance between the sample and the QMS was 7 mm approximately. Another aim of the study is to develo p the scale of calibration for experiments which will be carried out in future. I n order to fix the scale, it is required to find the 2D saturation surface coverage which wa s observed at an exposure of about 3 minutes. Note that the 2D oxide saturates at 0 .7 ML and the precursor peak then starts to develop. To verify the results a number of low exposures were done around 3 minutes and QMS was done at higher emission current to improve the signal to noise ratio as seen in Figure 3 5. To confirm the 2D oxid e saturation a surface that slightly overshoots 2D oxide saturation was prepared at different exposures between 2.5 minutes and 3 minutes. Then the surface was annealed to 625 K and dropped to 500 K fo llowed by a normal TPD. Figure 3 5 shows the evolution of the precursor peak at high emission current and low exposures which confirmed that the 2D oxide saturates as 2.75 minutes. This exposure has been used to set the coverage scale throughout the experiment. Observations and Discussions Figure 3 3 shows the O 2 spectra obtained from oxide covered Pd(111) after exposing the surface to atomic oxygen for different exposure s at 500 K. The main characteristics of the TPD spectra p resented in Figure 3 3 agree well with those
41 reported in prior studies of Pd(11 1) oxidation [ 13 ; 27 ; 33 ] The data is comparable to the previous studies and results obtained by Heywood et al. in our group which shows the presence of the precursor features in the TPD spectrum This precursor state corresponds to small oxide particles, re ferred to as PdO seeds, which mediate the formation of bulk PdO on the Pd(111) surface. We can observe that the precursor state shows a TPD peak at 600 K which rise s maximum to a tota l coverage of 1.2 ML in Figure 3 2 Earlier studies have reported the max imum intensity at total oxygen coverage of 1.3 ML. The evolution of atomic oxygen states on Pd(111) is known to initially involve O atoms arran ging into a p(22) structure [ 34 ] This can be observed in the TPD feature between 675 and 800 K. Further increasing the atomic oxy gen coverage above 0.25 ML (saturation of p(22) structure) then causes a so calle d 2D surface oxide to develop, which satura tes at approximately 0.7 ML [ 35 ] This c an be observed from the Figure 3 2 right before the formation of the precursor peak. Surface oxide decomposition is responsible for the sharp O 2 TPD peak seen in Figure 3 1. Prior studies have shown that the 2D oxide consists of a single layer of Pd and O atoms arranged on t op of the Pd(111) surf ace [ 35 ] The 2D oxide has a Pd 5 O 4 stoichiometry, and its structure does not match any latt ice plane of crystalline PdO Our group has also found out before that after the 2D saturation the 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 oxid e to produce 3D oxide do mains that resemble bulk PdO The 3D oxide domains grow on Pd(111) and coexist with oxygen atom covered 2D oxide domains as the coverage increases above 0.7 ML. Bulk like
42 PdO begins to form on Pd(111) above 0.70 ML, causing the explosive desorption feature to intensify and shif t to higher temperature [ 13 ; 3 6 ] .Once the oxygen coverage reaches about 5+ ML the surface is predominantly c overed by 3D oxide. These results are comparable to the data obtained before where the 3D oxide dominated when the coverage reached about 3 ML with 2D oxide covering ver y small area of the surface Oxidation at 500 K results in the effective growth till 4 ML for the O atom beam flux (~0.0 056 ML/s) employed in our experiments. Thus the results from TPD spectra indicates that the 2D and 3D oxides offer chemically distinct binding sites, and suggests that these oxides are likely to exhibit distinct catalytic p roperties as well. Overall, the TPD spectra and the uptake kinetics observed in the present study compare well with that reported previously for Pd(111) oxidation at 500 K. The uptake plot for the prese nt study can be seen in Figure 3 4. The implication is that the oxygen phase evolution obtained in the present st udy is similar to that obtained previously by our group Future Scope The present study was done to test our newly developed UHV chamber and operating characteristics of the plasma source to observ e similar Pd(111) oxidation behavior as seen previously. However future scope of this work lies in the catalytic oxidation of CO on the PdO(101) surface. Our research group found previously that O 2 molecules, in certain states, bind more strongly on PdO th an Pd(111) supports the idea that molecularly chemisorbed oxygen can exist in appreciable concentrations on oxidized Pd at high pressure and temperature, and is therefore a viable reactant. From prior studies conducted by the group, it can be concluded tha t the activity of CO cannot be characterized solely based on the spectra. Hence our future studies are going to be
43 done using Fourier transform infrared s pectroscopy (FTIR). This technique allows us to perform in situ reflection absorption infrared spectro scopy (RAIRS) on CO molecule adsorbed on palladiu m as well as PdO(101) surface. The absorbance spectra will provide the bond vibrational information of the adsorbed CO molecules which can be used to find out the binding configurations and energies of the C O molecules on the palladium and its oxide surface. Figure 3 6 shows a typical infrared ( IR ) spectrum obtained from our IR setup using the same UHV chamber described in the introduction Further goal is to install a scanning tunneling m icroscope (STM) and use it to characterize the surface of p alladium along with oxides and adsorbed species. The previously unreported detailed behavior of CO adsorption on PdO(101) surface motivates further investigation in this area. Summary We have inves tigated the evolution of the oxygen phases on Pd(111) surface to test our system which shows comparable results from previous studies by our group. Our results revealed the rate of oxygen desorption from the PdO(101) thin film undergoes a variety of change s indicating multiple mechanisms for decomposition. We have focused our interests in the reactivity of the PdO(101) surface primarily because bulk termination of the fully oxidized Pd(111) surface is PdO(101) and also as dis cussed in the present study the higher coverages of the oxidized Pd(100) is also PdO(101) Furthermore, it has been found that the surface oxide on Pd(100) does not have the reactive properties of the bulk PdO(101) although it possesses the same (101) structure [ 36 ] PdO(101) surface has been found to possess t he PdO seeds and shows kinks while the Pd(111) surface is flat. Hence PdO surface offers more sites promoting reactivity and thereby favoring catalytic reactions. From prior studies done by our group,
44 it has been observed that PdO(101) is the main site pro moting activity of methanol, propane, hydrogen, carbon monoxide etc. They are conclusive of the fact that the local geometric and electronic configurations of the adsorption sites are one of the most important features of the catalyst surface for heterogen eous catalysis. One can easily point out based on observations that the vari ous phase evolution of PdO(101) proves that the surface promotes oxidation of molecules. One can use the UHV experimental studies to predict the kinetics of the catalysis phenomena under oxygen rich environment for units operating under atmospheric conditions. Another particular aspect to be noted is the ability to regenerate the PdO(101) domains which enables the autocatalytic process to evolve smoothly as the total amount of desor bed oxygen increases beyond the concentration of oxygen atoms in a single PdO(101) layer.
45 Figure 3 1. O 2 TPD spectra obtained from Pd(111) for higher exposu re times of 40 minutes,30 minutes, 20 minutes and 10 minutes. Each coverage was generated by exposing Pd(111) to an O atom beam with the surface held at 500 K. A constant heating rate of 1 K/s was used to obtain each TPD spectrum.
46 Figure 3 2 O 2 TPD spectra obtained from Pd( 111) for lower exp osure times of 5 minutes,3 minutes, 2 minutes, 1.5 minutes, 1minute and 30 seconds. Each coverage was generated by exposing Pd(111) to an O atom beam with the surface held at 500 K. A constant heating rate of 1 K/s was used to obtain each TPD spectrum.
47 Figure 3 3 O 2 TPD spectra obtained from Pd(111) for all exposure times. Each coverage was generated by exposing Pd(111) to an O atom beam with the surface held at 500 K. A constant heating rate of 1 K/s was used to obtain each TPD spe ctrum.
48 Figure 3 4 Uptake plot obtained for the while surface coverage of Pd(111). Each coverage was generated by exposing Pd(111) to an O atom beam with the surface held at 500 K. A constant heating rate of 1 K/s was used to obtain each TPD spectrum
49 Figure 3 5. TPD spectrum at higher emission current about 3 minutes exposure to find the 2D oxide coverage required for calibration. Each coverage was generated by exposing Pd(111) to an O atom beam with the surface held at 500 K. A constant heating rate of 1 K/s was used to obtain each TPD spectrum.
50 Figure 3 6 IR spectral obtained from CO adsorption on PdO( 101) at 100 K Exposure time was 7 minutes. The different spectra are collected with and without annealing the sample after dosing CO. PdO(101) has been prepared by exposing the sample at for 40 minutes.
51 LIST OF REFERENCES  M.S. Peters, K.D. Timmerh aus, R.E. West, Plant design and economics for chemical engineers, McGraw Hill, New York, 2003.  I. Nakai, H. Kondoh, T. Shimada, A. Resta, J.N. Andersen, T. Ohta, Journal of Chemical Physics 124 (2006).  M. Lyubovsky, L. Pfefferle, Catalysis Today 4 7 (1999) 29 44.  J. Wang, Y. Yun, E.I. Altman, Surface Science 601 (2007) 3497 3505.  J.Y. Han, D.Y. Zemlyanov, F.H. Ribeiro, Surface Science 600 (2006) 2730 2744.  J.Y. Han, D.Y. Zemlyanov, F.H. Ribeiro, Surface Science 600 (2006) 2752 2761.  J.F. O'Hanlon, A user's guide to vacuum technology, Wiley, New York, 1989.  J.A. Hinojosa, H.H. Kan, J.F. Weaver, Journal of Physical Chemistry C 112 (2008) 8324 8331.  R.I. Masel, Principles of adsorption and reaction on solid surfaces, Wiley, New Y ork, 1996.  E.M. McCash, Surface chemistry, Oxford University Press, Oxford ; New York, 2001.  T. Dellwig, J. Hartmann, J. Libuda, I. Meusel, G. Rupprechter, H. Unterhalt, H.J. Freund, Journal of Molecular Catalysis a Chemical 162 (2000) 51 66. [12 ] M. Baumer, H.J. Freund, Progress in Surface Science 61 (1999) 127 198.  H.H. Kan, J.F. Weaver, Surface Science 602 (2008) L53 L57.  R. Zaplotnik, A. Vesel, M. Mozetic, Epl 95 (2011).  I.M. El Fayoumi, I.R. Jones, M.M. Turner, Journal of Physi cs D Applied Physics 31 (1998) 3082 3094.  G. Zheng, E.I. Altman, Surface Science 462 (2000) 151 168.  H. Gabasch, et al., Surface Science 600 (2006) 2980 2989.  J.N. Carstens, S.C. Su, A.T. Bell, Journal of Catalysis 176 (1998) 136 142.  A.K. Datye, J. Bravo, T.R. Nelson, P. Atanasova, M. Lyubovsky, L. Pfefferle, Applied Catalysis a General 198 (2000) 179 196.
52  R.J. Farrauto, J.K. Lampert, M.C. Hobson, E.M. Waterman, Applied Catalysis B Environmental 6 (1995) 263 270.  J.G. McCarty Catalysis Today 26 (1995) 283 293.  H. Gabasch, et al., Journal of Physical Chemistry C 111 (2007) 7957 7962.  A. Hellman, et al., Journal of Physical Chemistry Letters 3 (2012) 678 682.  A.L. Gerrard, J.F. Weaver, Journal of Chemical Physics 123 (2005).  J.F. Weaver, J.J. Chen, A.L. Gerrard, Surface Science 592 (2005) 83 103.  E. Lundgren, G. Kresse, C. Klein, M. Borg, J.N. Andersen, M. De Santis, Y. Gauthier, C. Konvicka, M. Schmid, P. Varga, Physical Review Letters 88 (2002).  H .H. Kan, R.B. Shumbera, J.F. Weaver, Surface Science 602 (2008) 1337 1346.  R.B. Shumbera, H.H. Kan, J.F. Weaver, Surface Science 601 (2007) 4809 4816.  C.J. Zhang, P. Hu, Journal of the American Chemical Society 123 (2001) 1166 1172.  K.C. Has s, A.E. Carlsson, Physical Review B 46 (1992) 4246 4249.  H.H. Kan, J.F. Weaver, Surface Science 603 (2009) 2671 2682.  E. Lundgren, A. Mikkelsen, J.N. Andersen, G. Kresse, M. Schmid, P. Varga, Journal of Physics Condensed Matter 18 (2006) R481 R49 9.  H. Gabasch, W. Unterberger, K. Hayek, B. Klotzer, G. Kresse, C. Klein, M. Schmid, P. Varga, Surface Science 600 (2006) 205 218.  V.A. Bondzie, P.H. Kleban, D.J. Dwyer, Surface Science 465 (2000) 266 276.  G. Ketteler, D.F. Ogletree, H. Bluh m, H.J. Liu, E.L.D. Hebenstreit, M. Salmeron, Journal of the American Chemical Society 127 (2005) 18269 18273.  A. Hellman, R.G.S. Pala, Journal of Physical Chemistry C 115 (2011) 12901 12907.
53 BIOGRAPHICAL SKETCH Su hrita Muk herjee was born in Calcutta, India to Somen Mukherjee and Mallika at Maharaja Sayajirao University of Baroda, India (MSU) and earned Bachelor of Engineering degree in chemical engineering in 2009. Later the same year she started working as Assistant Systems Engineer at Tata Consultancy Services (TCS), Ahmedabad, India and Resource Management Lead for TCS Hyderabad, India for 18 months. In August 2011 she enr olled at University of Florida Gainesville and received her Master of Science in chemical engineering in 2013.