Lecture Notes on X-Ray Photoelectron Spectroscopy | PHYS 3340, Study notes of Physics

Download Lecture Notes on X-Ray Photoelectron Spectroscopy | PHYS 3340 and more Study notes Physics in PDF only on Docsity! Advanced Lab Spring 2007 X-ray Photoelectron Spectroscopy (XPS) Purpose: In this lab you will study the photoelectron spectra of silicon and silicon dioxide using a commercial XPS instrument running in ultra-high vacuum (UHV) conditions. You will make use of the surface sensitivity as well as the chemical selectivity of XPS to estimate the thickness of a very thin SiO2 layer on top of a Silicon wafer. Introduction: The process of photoelectron spectroscopy was first explained by Einstein and garnered him the Nobel prize in physics according to his explanation that a photon transfers all of its energy (hv) upon being absorbed. Photoemission also was the subject of the physics Nobel given to the Swede Kai Siegbahn for his development of the technique of ESCA (Electron Spectroscopy for Chemical Analysis) which we usually now just call XPS. Siegbahn showed how different valence configurations of an atom give rise to slightly different energy levels (chemical shifts) observed in XPS. Very briefly, a monochromatic beam of photons (X-rays) impinges on the sample to be studied, exciting an electron in the sample to a higher energy state. If the excited energy state is higher than the “vacuum level” of the sample (the Fermi energy plus the work function), the electron may be ejected into the vacuum and collected and energy analyzed by the electron spectrometer. With knowledge of the final state (collected) electron energy and the energy of the x-ray, energy conservation then allows us to precisely determine the initial state energy of the electron before it was ejected from the sample, using the equation Ekin=hv-BE-φ where Ekin is the kinetic energy of the ejected electron, hv is the photon energy, BE is the binding energy of the electron in the solid, and φ is the work function (typically a few eV). The peaks as a function of energy thus tell us the chemical elements in the sample that the electrons were ejected from. The strength or intensity of the peaks are proportional to the number of atoms of any kind from which the electrons are ejected, so that the stoichiometry of a “mystery” sample can be identified, as well as the valence of chemical state of the internal elements. Finally, the surface sensitivity (on the order of 5-50 Angstroms) allows a unique view of the chemistry and physics of a materials surface. ESCA or XPS has grown into a very widely used and indispensable spectroscopy that is utilized in most major companies, universities, and national labs, and also from special analytical services companies such as Evans Analytical Group (http://www.eaglabs.com/en-US/services/esca.html) or RBD Enterprises (http://www.surfaceanalysislab.com/). These companies typically charge upwards of $250/hour to study a sample, and the instruments themselves have replacement value over $500,000. So have fun, but also be careful with our instrument! Jason Underwood, who has been performing his Ph.D. thesis using the XPS instrument, will be available to help you with this experiment (he is being partially supported to do this). He can be reached at underwjm@gmail.com or on his cell phone at 720-470-0564. 1 Outline of the experiment: Begin by reading the attached pages from the “Handbook of XPS” written by the instrument manufacturer Physical Electronics (abbreviated Phi) and answer the pre-lab questions. You will then clean your Si and SiO2 samples for UHV conditions and mount each sample to a transferable sample puck. You will load a puck into the fast-entry vacuum load-lock, which you will pump out to below 10-6 torr using a turbo pump. You will then transfer the puck (and sample) into the UHV (10-9 torr) ion- pumped analysis chamber and (carefully) turn on all electronics. You will take both a broad-energy survey scan of each sample and a higher resolution higher statistics “multiplexed” scan of a few individual regions. You will perform some data analysis routines on your spectra, including a removal of “satellite” peaks due to additional weak x-ray lines, a removal of the background of scattered electrons, and finally a calculation of the area under the individual peaks so that you may determine the stoichiometry of each sample. For your Silicon sample you should see a Si 2p peak due both to pure silicon and due to silicon which has been bonded to oxygen (in SiO2). You will use the included fitting routine to separate out the contribution from each of these peaks and determine the relative strength of Si and SiO2 in your spectra. From this and a knowledge of the photoelectron mean free path in SiO2 you will make an estimate of the thickness of the SiO2 layer on top of your Si sample. Your lab report will include the answers to the included prelab questions as well as plots of your raw data, reduced data, and your analysis and results. The instrument: The two key pieces of the instrument are the hemispherical electron energy analyzer, which is a fully electrostatic instrument (no magnetic fields) and the dual-anode (Mg and Al) x-ray source, as well as the associated control/detection electronics with each unit. The following website has some information about the design of a typical XPS instrument, including some key concepts such as electron pass energy, etc. http://www.casaxps.com/help_manual/XPSInformation/XPSInstr.htm The vacuum system. It is important that you understand the details of the vacuum system before you begin to use the chamber. The main vacuum chamber is pumped by a large 110 liter/second ion pump, which is hanging below the chamber. When in operation the ion pump supply (Varian VacIon pump control) should be almost 6 KV, with the ion pump current proportional to the gas pressure in the chamber. For our chamber at ~ 3x10-9 torr the ion pump current will be around 250 µA. Please note this value in your log book. You also should note the pressure in the main vacuum chamber, which is read off of the Granville Phillips ion gauge controller in the top position of the right rack. You should always keep an eye on the ion gauge controller, with the vacuum staying below a few x 10-7 torr at all times. The ion gauge itself is a nude gauge with a center pin (collector) and 5 pins around the outer circle. The load lock is pumped by a turbo pump which is backed by an oil-lubed rough pump. The load lock is separated from the turbo by a pneumatic angle valve and separated from 2 λλ λ // 0 11)0( dx d edxedP −− −==< ∫ where the proportionality factor 1/λ comes from requiring that P(x) be normalized – integrating from 0 to infinity should yield unity, as the electron must have come from somewhere. Setting d=λ we see that 63% of the detected electrons have come from within one attenuation length λ of the surface and 95% come from within 3λ. Prelab questions. 1) Why do we need to clean our parts well before entering them into the UHV chamber? 2) What do the labels 3/2, ½ in the Si 2p core level stand for? What is the ideal peak area ratio of the two peaks within a Si 2p core level? How about inside a 4f level? 3) What is a “chemical shift” in XPS? What is a typical order-of-magnitude of this size of effect? 4) Why is there an increasing background signal as you go to increasing binding energy? 5) What is the Auger process and how could one operationally distinguish an XPS peak from an Auger peak? 6) Briefly describe how the electron energy analyzer works. Briefly describe how the x- ray source works. 7) What does it mean when we discuss the analyzer pass energy? What is the effect of varying this on the spectrum that you take? 8) In a uniform material what percentage of electrons that you detect at normal emission are emitted from within the first two attenuation lengths? 9) You wish to study a silver core level, but it is covered by a 10 Angstrom thick layer of gold at the surface. Assuming a photoemission escape depth of 20 Angstroms, how much weaker will the silver core level be than if the gold layer was absent? What if we are at glancing emission of 75 degrees? 10) What happens if your sample charges slightly? Detailed Procedure and explanations. a) Clean your samples for UHV. There should be acetone (first) and isopropyl or ethyl alcohol (to finish) for this purpose, as well as an ultrasonic cleaner. You should use rubber gloves to keep all fingerprints off of your sample and the sample holder. Note that latex gloves will not hold up to acetone and so you should use nitryl gloves when using acetone. 5 b) Mount your sample (it is recommended that you start with the thermally-grown SiO2) to one of the transferable sample holders (the round stainless steel discs with two grooves along the sides). Ideally you should just be able to gently wedge the sample underneath one of the beryllium-copper springs, which will make physical as well as electrical contact to your sample (electrical contact is necessary so that the ejected electrons will be replenished. If they are not your sample will charge and the kinetic energies you measure will be incorrect). c) At this point you may wish to begin warming up the voltage supplies which set the voltages for the electron detector. This will give them time to stabilize while you are getting things ready. Do this only for the top 5 pieces of Leybold brand electronics. There is a round push button on the left of each of these units. BEFORE doing this you should check that the electron multiplier high voltage dial is off (fully CCW) so that there will be no high voltage applied. You do not want this HV on, in case there are any vacuum problems in the next steps (this could cause arcing and serious damage to the in- vacuum detector assembly). d) Following the included flow chart (and initially reviewing with Jason, Professor Dessau, or one of your instructors), load the sample onto the xyz manipulator in the vacuum chamber. Briefly, you will begin by first loading it into the transfer arm, grabbing the sample puck on the lower groove with the sample fork (the xyz manipulator should grab the puck at the top groove). You should pump on the transfer arm with the turbo for about 20-30 minutes, bringing the transfer arm vacuum to about 1x10-6 torr or better. Following that you will shut the angle valve to isolate the transfer arm from the UHV chamber, open the gate valve, checking that the main chamber vacuum does not go past the low 7 scale vacuum. Transfer the sample puck to the xyz manipulator, grabbing it on the top groove, and then remove the arm and shut the gate valve. You should then reopen the angle valve to keep the transfer arm pumping. As a starting point, you should expect the transfer position to be at about (x,y,z) = (12.5,13,10) and the analysis position to be (12.5,8.5,10), where you are reading the black scale, and the units are mm. z=50 is having the manipulator fully retracted from the chamber center, which is the safest location when moving the transfer arm in. e) You have a sample loaded into the UHV chamber. Congratulations! You now want to slowly turn on the x-ray source, after confirming that the vacuum in the UHV chamber is still good (8 scale or better). To turn on the x-rays, you should first turn on the cooling water using the white push-button switch on the x-ray controller (it is helpful to hold this in for a few seconds). You should wait a few seconds to let all air bubbles in the water lines disappear. After the water is on, check that the large HV (high voltage) knob is at the minimum (fully CCW) and the “emission regulation x-ray source” is in “standby” mode. Then push the HV push button on. You should now see about 0.16 KV high voltage, 0 mA of emission (both on the digital meters) and about 3 amps of filament current (on the analog meter). 6 f) You now want to SLOWLY begin turning up the x-ray source high voltage using the large knob on the panel. You should shoot for about 10 KV, taking about 1 minute to get there. You will begin to see first one and then 2 mA of current on the emission current digital display. This is not actually emission current, but is a slight bit of leakage current through the cooling water lines (the water is deionized but is still not infinitely resistive). When you have reached 10 KV you can start to get a real electron emission current. These emitted electrons will strike the Al or Mg target, creating the x-rays you need. You should check that the knob on the left is set to an emission current setting of only 20 mA. Then you should turn the “emission regulation” knob from “standby” to “operate” and you should see that you have 20 mA of emission current on the digital display. You now have x-rays! Typically we run the system with the Mg anode. If you wish to change from the Mg to the Al anode it is imperative that you first turn the x-ray source to “standby” (no emission current) and then turn the HV knob slowly down to below 5KV before switching over. Afterwards, bring HV back to 10 KV and go from “standby” to “operate” to get emission current. g) You now need to turn up the voltage on the electron multiplier, which is an avalanche- style device inside the vacuum (at the exit of the hemisphere) which amplifies a single electron which makes it around the hemisphere into a pulse of about 1 million electrons. This pulse of electrons is then further amplified with a preamp and a post-amp (both outside the vacuum) and then shaped into a TTL pulse which is counted by the computer. We are doing pulse-counting here, as opposed to measuring an electric current. Pulse counting is preferable for low signal levels, while analog current measurements are preferable at high signal levels (as the counting electronics may not be able to keep up). Check that the detection mode is “XPS” and then slowly turn the electron multiplier supply to 3.8 KV using the potentiometer. h) You are ready to take data! Launch “AugerScan 2.4” which is available by double- clicking on the icon on the computer desktop on the small black computer desk. There is a data acquisition card in this computer with a DAC (digital to analog converter) which puts out a single 0-10V analog signal which goes to the analyzer control to set the hemisphere voltages, determining the electron kinetic energy being studied. NO other voltages can be set from the computer – neither the analyzer pass energy, multiplier voltage, etc. The computer also has a counter/timer facility which counts the number of TTL pulses coming from the ratemeter (photoelectron counts). AugerScan will repeatedly ramp the kinetic energy over a predetermined range for you, counting the photoelectron signal and averaging the new sweep with the previous scans. It is recommended that you start with a broad energy (survey) scan of your sample, which we usually do using a pass energy of 150 eV (this is called “transmission energy” on the Leybold electron energy analyzer power supply). Under the “file” menu select “new” and then select “survey”. Then, under the “acquisition” menu select “settings” which you 7