Photoelectron Spectroscopy (XPS, UPS) 광전자 분광법 Photoelectron Spectroscopy (XPS, UPS) 김 정 원 미래융합기술본부 소재게놈측정센터 (E-mail: jeongwonk@kriss.re.kr)
Introduction - 광전자분광법(PES)란 무엇인가 ? PES로 무엇을 할 수 있는가 ? PES의 간략한 역사 Principles of XPS Chemical shifts
Photoelectron Emission energy, polarization, angle Low energy electrons Vacuum UV UPS energy, angle, spin x-ray valence XPS core High energy electrons XPS: X-ray Photoelectron Spectroscopy (hv = 50~2000 eV) UPS: Ultraviolet Photoelectron Spectroscopy (hv = 6~80 eV)
Schematic Diagram of PES Process Electronic structure Solid UPS Atomic composition Chemical structure XPS Energy conservation EB: element & chemistry specific in case of core levels
Typical XPS spectra of Ag 3d5/2 Al Kα Mg Kα MNV MNV 150 3d3/2 kCounts [a.u.] 100 3p3/2 3p3/1 3s 50 4d (valence) 4s 4p 1200 1000 800 600 400 200 Binding Energy [eV] EB: constant irrespective of excitation source AES transition: varies with excitation energy (constant kinetic energy)
약어들 Abbreviations XPS: X-ray Photoelectron Spectroscopy ESCA(XPS): Electron Spectroscopy for Chemical Analysis UPS: Ultraviolet Photoelectron Spectroscopy PES: Photoelectron (Photoemission) Spectroscopy KE: Kinetic energy BE (EB): Binding energy ARXPS, UPS, PES: Angle-Resolved XPS, UPS, PES XPD: X-ray Photoelectron Diffraction PED: Photoelectron Diffraction IP(E)S: Inverse Photoemission Spectroscopy EELS: Electron Energy Loss Spectroscopy Electron Spectroscopy: AES, EELS, PES
PES로 무엇을 할 수 있는가 ? Non-destructive Elemental Identification except H and He (다른 방법? X-ray fluorescence, SIMS, AES) - Quantification (원소농도분석): ~ %정확도 - Chemical State Identification (e.g. Si, SiO2) - Surface/Adsorbate Structure - Electronic Structure valence band level positions band structure mapping work function many body effect, etc. - Microscopy with Chemical Sensitivity
Available? Where? XPS (ESCA): 서울? 대부분의 종합대 및 공대, 일부 실험실 (not open), KIST, 생기연 지방? 대부분의 국립 종합대, 일부 종합 사립대 기초과학지원연구원 (대전, 부산), 화학연 대부분의 대기업 분석실 (not open), 포항방사광 UPS (Ultraviolet Photoelectron Spectroscopy): 기초과학지원연구원 (대전, 부산), KAIST, KIST 많은 개별 실험실 (not open) vs. Evans Analytical Group (USA), Toray Research Center (Japan) Pros: cheap and relatively fast Cons: instrument-dependent, weak analysis
A Brief History of PES - 1887, H. Hertz: 광전효과(photoelectric effect) 발견 - 1899, J. J. Thompson: 전자 발견 - 1900, M. Planck: Quantum theory (quantized energy) - 1905, A. Einstein: Quantum theory로 광전효과 설명 - 1958, W. E. Spicer: UPS spectra와 DOS 관련주장 - 1967, K. Siegbahn: XPS (ESCA) 확립 - 1969, HP - Commercial XPS - 1970s: Synchrotron Radiation
History: Discovery of Electrons Photoelectric effect Metal plate in a vacuum, irradiated by ultraviolet light, emits charged particles (Hertz 1887), which were subsequently shown to be electrons by J.J. Thomson (1899). Classical expectations Light, frequency ν Vacuum chamber Electric field E of light exerts force F=-eE on electrons. As intensity of light increases, force increases, so KE of ejected electrons should increase. Collecting plate Metal plate Electrons should be emitted whatever the frequency ν of the light, so long as E is sufficiently large I For very low intensities, expect a time lag between light exposure and emission, while electrons absorb enough energy to escape from material Ammeter Potentiostat Lecture note: Fisher (Univ. College London )
History: Photons and Electrons Photoelectric effect Einstein Actual results: Einstein’s interpretation (1905): light is emitted and absorbed in packets (quanta) of energy Maximum KE of ejected electrons is independent of intensity, but dependent on ν For ν<ν0 (i.e. for frequencies below a cut-off frequency) no electrons are emitted Millikan An electron absorbs a single quantum in order to leave the material There is no time lag. However, rate of ejection of electrons depends on light intensity. The maximum KE of an emitted electron is then predicted to be: Verified in detail through subsequent experiments by Millikan Work function: minimum energy needed for electron to escape from metal (depends on material, but usually 2-5eV) Planck constant: universal constant of nature
광과 시료원자와의 상호작용 표면탈출: XPS 이차전자 증폭 발생: SEM 탄성 충돌: XRD, LEED, RHEED 비탄성 충돌: EELS X-선, UV (EDX/WDX) XES
Theory of Photoemission (binding energy) frozen orbital approximation (Koopman’s theorem) + e (KE) initial state A(N) final state A+(N-1) Considering energy conservation relativistic energy kth orbital energy by Hatree-Fock calculation first approximation correlation energy Relaxation energy
Element-specific Core Levels
All core levels undergo same chemical shift Chemical Shifts (Initial state effect?) valence shell (charge qi) All core levels undergo same chemical shift (approximation) Surroundings (charge qj) core shell electrons rij similar to Madelung potential Charge potential model in ionic bonds EB ↑ as q ↑ (note sign) (Fig. 3.3) for neutral atom Coulomb interaction between core electron and nucleus screened by valence electron charge Dependent on chemical environment such as oxidation state electronegativity of neighboring atoms number of surrounding atoms J.C. Vickerman and I.S. Gilmore, Surface Analysis, The principal Techniques, 2nd ed. (Wiley, 2008)
Chemical Shifts of S 1s and S 2p Formal oxidation state is good indication of EB It is only valid in ionic bonds If there is covalent/ionic bond character mixing, Charge density on an atom is the best criterion
A Example of Chemical Shifts Not Always Possible Tabulated Values Vary Widely
Final state consideration hot electron electron ejection e- Ei Δt hole hole Ef valence shell final state Instant excitation Relaxation (dynamic screening, solvation…) initial state Ei(N) + = Ef(N-1) + KE (e-) BE = - KE (e-) = Ef(N-1) – Ei(N) ≈ orbital energy But, final-state relaxation is much dependent on its atomic environment Relaxation energy –> binding energy (time-dependent) Relaxation time –> peak width
Spin-orbit splitting Unpaired electron after ionization J=L+S 3d5/2 Ag 3d Unpaired electron after ionization J=L-S 3d3/2 Ag 3d: (3d)10 + → (3d)9 + e- L=2, S=1/2, J=L+S,,, L−S = 5/2, 3/2 2D5/2, gJ = 2×(5/2)+1 = 6 2D3/2, gJ = 2×(3/2)+1 = 4 Branching ratio = 6:4 = 3:2 p, d, f core levels split into two doublet peaks BE (J=L−S) > BE (J=L+S) considering final state energy Splitting ↑ as Z ↑ What about p and f levels?
Final State Effects Initial state effect: Koopman’s theorem Final state effect: : 1~10 eV -the created core hole after photoionization affects the energy distribution of the emitted electrons in different ways. Relaxation effect Multiplet splitting Multielectron excitations -shake up and shake off satellites -electron-hole excitation (continuous satellite): asymmetric line shape Plasmon loss peaks Vibrational effects Ref. Electron spectroscopy, theory, techniques, and applications Vol.2 (Academic Press. 1978) Chap. 1, C.S. Fadley, Basic concepts of x-ray photoelectron spectroscopy
Relaxation effect Atomic relaxation (gas phase) by Franck-Condon principle Extra-atomic relaxation by surroundings (sold phase) As Ef ↓ , BE ↓ Instantaneous electronic transition (A) but relaxation afterward (F)
Typical shake-up in C 1s
Instrumentation - Vacuum System - Photon Source Electron Energy Analyzer Resolution Binding Energy Referencing Charging compenstation
Universal Curve Inelastic electron mean free path (IMFP) ~ Intensity attenuation length IMFP
Other consideration (Escape depth) Comparison of AES and EDX analysis volume Hard X-ray PES (HAXPES) HAXPES SX-PES Conventional VUV Probing depth (3λ) up to > 10 nm Accessing bulk or buried interface Depth profile of thin film Lecture note: R. Claessen (Univ. Wϋrzburg)
Photon Source Intensity, Focus, Monochromatic, energy selection dual anode, monochromatic, discharge lamp, SR
Characteristic X-ray Lines Width (eV) 2.6 3.8 3.0 0.77 0.47 0.85 0.7 Line Cu Kα Cu Lα Cr Lα Zr Mζ Y Mζ Al Kα Mg Kα Energy (eV) 8048.0 929.7 572.8 151.4 132.3 1486.6 1253.6 1.0 Si Kα 1739.5 Satellites of Mg Kα Relative Intensity Line Separation (eV) 100 α12 0.0 8.0 α3 8.4 0.55 α5 17.5 4.1 α4 10.2 0.45 α6 20 0.5 β 48.5
Al K Monochromatic X-ray Source의 구조 PHI Quantera
C 1s: Effect of monochromatization 10 ~ 40 times intensity reduction Focusing at the sample (up to 10 ㎛ recently) - Intense beam - Imaging
Resonance Lines of Rare Gas Discharge UV sources : each monochromator available Intensity (%) 15 2 100 Resonance Line Kr I Ne I He II He I Ar I Energy (eV) 10.0321 16.6704 40.8136 23.0865 21.2175 11.6233 16.8474 50 11.8278 9.5695 10.6434 Xe I 8.4363
Synchrotron Radiation - High Intensity and resolution, energy tunability, focused Beam, polarization, pulsed beam - Big Facility means Big Money ! – Shared facility Pohang Acceleration Lab.
Concentric Hemispherical Analyzer : CHA Hemispherical Energy Analyzer Concentric Hemispherical Analyzer : CHA
Microchannel Plate (MCP) Electron Detection Channeltron (CEM) Microchannel Plate (MCP)
Position-sensitive Energy Analyzer Channeltron array MCP + CCD (or DLD)
Energy Resolution of Analyzer Small Epass Large Eretard Large Epass Small Eretard High resolution Low throughput (low count rates) Low resolution High throughput (high count rates)
Total Energy Resolution ΔETotal = √(ΔEanal.2 + ΔEphoton2 + ΔEthermal broad2 + ΔEinhomogen.2 + ..) ΔEanal.= Epass(d/2Ro + α2/4) for HEA type d= slit width Ro=mean radius α=a half acceptance angle ΔEthermal broad ~ 3/2kT <Practical measurement of total resolution>
Differential surface charging Peak Broadening or shift
Charging Compensation
Charging Compensation With a Magnetic lens system (Cratos) Nearly complete charging compensation !
Modern ESCA Features 가격 $1M? Monochromatic x-ray (1486 eV) Angle-resolved analyzer UV discharge lamp (optional) Automatic sample transfer Charge compensation 가격 $1M?
Data Processing - Smoothing - Background and X-ray Satellite 제거 - Peak Fitting and Deconvolution Practical Programs
Smoothening - NOT recommended, but - 너무 오래 걸리거나, 변하는 시료, 또는 미분하기 전에 - Savitsky-Golay method - Adjacent Averaging
Background Removal Background Removal Methods Inelastic Energy Loss Mechanism Complicated Process Sample Dependent Geometry Dependent Instrument Dependent Must be Removed for Quantification Removal Methods Linear, Shirley, Tougaard
Linear Background Removal
Shirley Method integrated background Key point: 어떤 점 x에서 BG 는 x 보다 높은 K. E.를 가진 전자들의 total intensity에 비례한다
Shirley BG Removal in Action
Peak Fitting and Deconvolution Lorentzian Gaussian Voigt Lorentzian width: core hole life time: ΔE·τp≥ ħ Gaussian width: instrumental resolution, system inhomogeneity
Least Square Peak Fitting
Peak Fitting in action C 1s urea formaldehyde/epoxy coating J.F. Watts and J. Wolstenholme, An Introduction to Surface Analysis by XPS and AES (Wiley, 2003) C 1s urea formaldehyde/epoxy coating Any principles? Proper background subtraction should be made. Initial guess is important. Any peak should have physical meaning. Fitting is just fitting (no result from nothing) Peak splitting within an system resolution
Fitting Programs Non-commercial programs CasaXPS http://www.casaxps.com/ Unifit http://www.uni-leipzig.de/~unifit/ NIST Database for the Simulation of Electron Spectra for Surface Analysis (SESSA) http://www.nist.gov/srd/nist100.htm Non-commercial programs Fitt-win http://escalab.snu.ac.kr/ XPSPEAK 4.1 http://www.uksaf.org/software.html FitXPS Compro http://www.sasj.jp/COMPRO/index.html
Applications - Elemental Identification Quantitative Analysis Chemical Shifts Work function Measurement Angle-resolved Techniques Depth profile Microscopy
Elemental Identification
정량분석: Basic Concept 시료 중 원소 A의 농도 시료에서 측정된 원소 A의 XPS 피크 세기 원소 A의 XPS 피크 세기 (Atom Sensitivity Factor) 값은 알기 어려우나, 는 비교적 알기 쉬움
정량분석: RASF 이용 다른 조건들이 동일할 때 원소들 피크의 상대적 세기 Analyzer, Source, Geometry dependent parameter Measured Intensity Element RASF 6443.2 Sr 3d 1.843 6009.1 Ti 2p 2.001 1080.6 C 1s 0.296 5339.7 O 1s 0.771 Concentration 20.4 17.5 21.3 40.8
RASF Relative elemental sensitivity 3d 4f 2p 4d 1s
정량분석: complicated problem http://www.nist.gov/srd/surface.cfm Material-dependent inelastic mean free path (IMFP) Probing different volumes in mixed (A and B) materials Correction must be made !!
TPP-2M equation for IMFP By Tanimura, Powell, Penn Surf. Interface Anal. 21, 165 (1994) (Å) density (g/cm3) Band gap (eV) atomic or molecular weight free electron plasmon energy (eV) number of valence electrons
Valence band spectra final state XPS UPS EF initial state core levels Zn 3d XPS UPS EF initial state core levels
Photoionization intensity (cross section) http://ulisse.elettra.trieste.it/services/elements/WebElements.html Lecture note: S. Kim (KAIST)
Atomic Cross-section UPS XPS
UPS의 상대적인 특징 1. Peak is broad Why? Due to electronic band structure or orbital hybridization 2. Feels more surface-sensitive than with XPS Why? Low electron attenuation length & high O atomic photoioinization cross-section at low photon energy 3. Very difficult to quantify Why? This is not atomic character but bonding character. It means that you need more clear and physical idea of your specimen in simplified way 4. Resolution is better Why? Gas discharge source provides a sharp fluorescent light
Example of XPS analysis-1 Pigment from Mummy Artwork
Example of XPS analysis-2 Valence electron state of Pt in PtPc 4+ 2+
Metal-semiconductor Contact Energy Level Diagram
Organic-Metal contact in OLED Device performance depends on the balanced carrier injection and transport To maximize device efficiency : Enhance charge injection
Energy levels and UPS spectra interface dipole UV (X-ray) ionization potential work function HOMO level S. Braun et al., Adv. Mat.(2009)
Angle-Resolved PES Better Surface Sensitivity SiO2 SiO2 Si Si
XPS에 의한 SiO2 박막 두께 측정 L = Attenuation length of Si 2p q cos L e Sample Analyzer x-ray (hν) Si L = Attenuation length of Si 2p correct measurement of SiO2 thickness from R0 and L
XPS에 의한 SiO2 박막 두께 측정 결과 SiO2 K. J. Kim, Thin Solid Films 500, 356 (2006) K. J. Kim, Surf. Interface Anal. 39, 512 (2007) a-Si 0 offset and the same linearity in the sub-nm region
Solid in periodic potential
Angle-Resolved UPS Band Mapping using Valence Spectra
Band Structure of Graphite
ESCA sputter depth profile 단점: ① sample damage ② very slow
Sputtering conditions
Photoemission Electron Microscopy SPEM PEEM
Photoelectron Spectromicroscopy Contaminated Coronary Stent Secondary Electron Image XPS Spectrum of Contaminated Area 100 µm F F O O C Cl Si Si F 1000 800 600 400 200 Binding Energy (eV)
Contaminated Coronary Stent Secondary Electron Image Carbon Spectrum of Contaminated Area 100 µm CF3 CF2O C-C CFO O=C-O C-O CF 294 290 286 282 Binding Energy (eV)
Contaminated Coronary Stent Fluorocarbon Map Hydrocarbon Map 100 µm 100 µm 500 x 500um
Summary XPS Characteristics UPS Characteristics Surface sensitive Accurate quantification Detection of all elements above He Provides chemical state information Can be applied to both inorganic and organic materials Can be applied to both conductors and non-conductors UPS Characteristics More surface sensitive No quantification Information on chemical bonds and electronic band structure
References 1. Electron spectroscopy, theory, techniques, and applications I-IV, edited by C.R. Brundle (Academic, NY, 1977) 2. Photoemission in Solids I&II, edited by M. Cardona and L. Ley (Springer-Verlag 1978) 3. G. Ertl and J. Kuppers, Low energy electrons and surface chemistry (VCH, 1985) 4. D.P. Woodruff and T.A. Delchar, Modern techniques of surface science (Cambridge, 1986) 5. Practical surface analysis, Vol.I, edited by D. Briggs and M.P.Seah (Salle & Sauerlander) 6. S. Hufner, Photoelectron Spectroscopy (Springer 1996) 7. J.F. Watts and J. Wolstenholme, An Introduction to Surface Analysis by XPS and AES (Wiley 2003) 8. J.C. Vickerman and I.S. Gilmore, Surface Analysis, The principal Techniques, 2nd ed. (Wiley, 2008) 9. Website http://www.xpsdata.com/