analysis of photoelectron spectra - fhi · analysis of photoelectron spectra ... *„handbook of...
TRANSCRIPT
Analysis of photoelectron spectra
Modern Methods in Heterogeneous Catalysis
Research
Axel Knop
OutlinePhotoelectron Spectroscopy: General Principle
Surface sensitivity
Instrumentation
Background subtraction
PES peaks, loss features
Binding energy calibration
Chemical state
Peak fitting
Quantitative analysis
High pressure XPS
examples
General Principle
- Detection of photoelectrons fromthe valence band region and corelevels
- Detection of Auger electrons and X-Ray Fluorescence
Where do the electrons come from?
Distance electron can travelin solids depends on:
• Material• Electron kinetic energy
Measure attenuation of electrons by covering surfacewith known thickness of element
Sampling Depth
Sampling depth:
For normal takeoff angle, cosΘ=1:
• When d=λ: -ln(I/I0)=0.367, i.e. 63.3% from within λ• When d=2λ: -ln(I/I0)=0.136, i.e. 86.4% from within 2λ• When d=3λ: -ln(I/I0)=0.050, i.e. 95.0% from within 3λ
Disregarding elastic scattering:
Mean Free Path of Electrons in Solids
• IMFP is average distance between inelastic collisions• Minimum λ of 5-10Å for KE ~ 50-100 eV
Maximum surface sensitivity
Background Correction
Background substraction
Choose suitable energy range for substractionBaseline on high KE side
Background Correction
Several ways to substract background:
- Stepwise- Linear- Method of Tougaard
Most common:
Method of Shirley et al.:
Always use same backgroundsubstraction method for all peaks!
D.A. Shirley, Phys. Rev. B, 5, S. 4709, 1972.
worst
best
Spectral features: PE Peaks
Orbital peaks used to identifydifferent elements in sample
Observation: • s orbitals are not spin-orbit splitted singlet in XPS• p, d, f.. Orbitals are spin-orbit splitted doublets in XPS
Spectral features: Auger Peaks
Auger peaks:
Result from excess energy of atom during relaxation (after corehole creation)
• always accompany XPS• broader and more complex structure than PES peaks
KE energy independent of incident hν
Spectral features: Loss Peaks
After Emission of PE the remaining electronsrearrange (Secondary peaks):
• Relaxation by excitation of valence electron to higher state (shake up)
loss of kinetic energy of PE leads to new peaksshifted in BE with respect to main peak
• Relaxation by ionisation of valence electron(shake off)
broad shoulder to main peak
• Plasmon losses
Spectral features: Loss Peaks
Loss peaks can be usedto identify chemicalcomposition
Problem:References needed(theoretical orexperimental)
Relationship between the degree of core levelasymmetry and the density of states at the Fermi
level (BE=0)
Identification of FE:
Easy for high density of states near EF, butmay be ambigeous for samples with lowDOS near EF, band gap or charged samples
Analyzing the data: Calibration of Binding Energies
Line profile modification by charging
Mo oxide on silica
real catalyst is powdersample after impregnationand calcination.
Analyzing the data: Calibration of Binding Energies
Identification of FE:
In case of low DOS near EF, band gap or charged samples (spectrum will move in this case!)
Look for reference peaks with known BE (be sure about that!!!)Works for any PE or Auger peak, e.g. peaks originating from substrate
Chemical Shift
BE of core electrons depends on the electron density of theatom (screening of the core electrons), affected by theelectronegativity of neighboring atoms
For unknown BE:
• Calculation of Pauling´s Charge (easy, but not always true)• More elaborate calculations (theorists, also not always true)
References:
• Use of data bases (NIST, Handbook of PhotoelectronSpectroscopy etc.*)
• Clean materials (pure metals, well definedsubstrates, „untreated“samples etc.)
• Literature about similartopics
Analyzing the data: Chemical States
*„Handbook of Photoelectron Spectroscopy“, Perkin-Elmer, 1992G. Ertl, J. Küppers;´´Low Energy Electrons and Surface Chemistry´´; VCH Verlag, 1985S. Hüfner; ´´Photoelectronspectroscopy´´; Springer, 1995.D. Briggs, M.P. Seah;´´Practical Surface Analysis´´, Vol. 1; Wiley + Sons, 1990.
Analyzing the data: Peak Fitting
Several influences on peakshape:
• Broadening mainly due to excitating light (natural), structural and thermal effects
• Asymmetry due to final stateeffects
• Species with similar BE
These effects have to be considered when fitting by reasonablechemical/physical model of the sample!!!
Analyzing the data: Peak Fitting
Asymmetric Peak shapes modeled by Doniach-Sunjicfunctions (convoluted with Gaussian profiles)
Most fitting programs provide useful tools:
Levenberg-Marquardt algorithm to minimize the χ2
M: measured spectrumN: energy valuesS: synthesized spectrumP: parameter values
β: peak parameterM: mixing ratioα: asymmetry parameterh: peak heightα 0: Lorentzian profile
Convolution of Lorentz (or D.S.) Gaussian profiles suited best!!
Voigt function
Strategy:
• Use of references forasymmetry and broadening
• Reasoning of mostlikely chemicalcomposition
Analyzing the data: Peak Fitting
Peak broadening and asymmetry should bethe same for samecomponent in different spectra
pure Au
graphitic
CNT
disordered
But: both may differ for different componentsdiscrepancy for different peaks should always be based on physical reasons
Quantitative Analysis: First Steps
In general:
IX = B × σ × λtot × T × nX
B: all instrumental contributionsσ: ionisation cross section for given photon energyλtot: total escape depthT: transmission through surfacenX: atomic density of analyzed species in sample
σ= σtot × f(X,α)
with f(X,α) = 1+(β(X)/4) (1-3cos2α)*σtot: total ionisation cross sectionf: form function accounting for asymmetry of peakβ: asymmetry parameterα: angle between photon beam and emitted electron(different for standard x-ray source and synchrotron)
*Yeh and Lindau, Atomic data nucl. data tables32(1985)1
λtot = E(X) × 1/a[ln(E(X)-b)*
E: KE of electrona, b: parameters dependent on dielectric function and concentration of host
*Penn, J.of Electr.Spectr. And Rel. Phen. 9(1976)29
Cross section changes with energyof incident beam!!!
Quantitative Analysis: Cross Section
Calculated changes of Cross section beam energy(Yeh, Lindau; Atomic data and nuclear data tables 32 (1985) 1)
Quantitative Analysis
In most cases the sample surface is quite heterogeneous
Need of models to extract an accurate approximation of composition out of spectral intensities
RuO2
RuxOy
Ru 3d5/2-map of the chemical states on Ru(0001) after pre-treatment with Operformed with Scanning Photoelectron Microscopy (ELETTRA):
Quantitative Analysis: Useful Examples
with the formulasabove follows
andaA: „radius of A“
Mol fraction
A. Heterogeneous mixture (e.g. alloy):
IA,B0: reference of
clean materialfollows
Quantitative Analysis: Useful Examples
B. Partial Coverage (e.g. adsorbat):
Contribution of B:(attenuated by A) (direct)
For signals of A and B follows
ΘA: coverageof A on B
Θ: angle betweensurface normal and emitted electron
Which givesIA,B
0 : referenceof clean material
Quantitative Analysis: Useful Examples
C. Thin layer of A on B (e.g. oxide):
Contribution of B:
Contribution of A:
Which gives
Special case:
i.e.
follows
General problem here: proper background substraction
Fundamental limit:elastic and inelasticscattering of electronsin the gas phase
Technical issues: - Differential pumping to keep analyzer in high vacuum- Sample preparation and control in a flow reactor
In situ XPS: obstaclesIn situ XPS: obstacles
30x10-21
25
20
15
10
5
0
Ioni
zatio
n cr
oss
sect
ion
(m2 )
101
102
103
104
105
106
Electron kinetic energy (eV)
in situSEM in situ
TEM
in situEXAFSin situ
XPS
O2 exp. data from Schram et al. (1965) extrapolation of Schram's data
• Photons enterthrough a window
• Electrons and a gasjet escape throughan aperture tovacuum
In situ XPS: basic conceptIn situ XPS: basic concept
H. Siegbahn et al., J. Electron Spectrosc. Relat. Phenom. , 319 (1973)2
• H. Siegbahn et al. (1973- )• M.W. Roberts et al. (1979)• M. Faubel et al. (1987) • M. Grunze et al. (1988)• P. Oelhafen (1995)
H. Siegbahn et al., J. Electron Spectrosc. Relat. Phenom. , 205 (1981)24
In situ XPS instruments: previous designsIn situ XPS instruments: previous designs
In situ XPS using differentially pumped electrostatic lensesIn situ XPS using differentially pumped electrostatic lenses
D.F. Ogletree, H. Bluhm, G. Lebedev, C.S. Fadley, Z. Hussain, M. Salmeron, Rev. Sci. Instrum. 73 (2002) 3872.
to pump to pump
X-rays from synchrotron
Gas phase composition canbe measured by XPS.gas phase signal: 1 torr·mm ~ a few monolayers
CloseClose--up of sampleup of sample--first aperture regionfirst aperture region
gas
z
d
hν
e-
p0
1.0
0.5
0.0-2 -1 0 1 2
z [d]
p/p 0
z=2 mm
d=1 mm
In situ XPS systemIn situ XPS system
Experimental cellsupplied by gas lines (p0)
X-rays enter the cell at 55° incidence through an SiNx window(thickness ~ 1000 Å)
Analyzerinput lens
Focal point of analyzerinput lens
First differentialpumping stage (10-4p0)
Second differentialpumping stage (10-6p0)
Third differentialpumping stage (10-8p0)
Hemisphercalelectronanalyzer(10-9p0)
mass spectrometerand additional pumping
Partial oxidation: CH3OH + ½O2 CH2O + H2O
Total oxidation: CH3OH + O2 CO2 + 2H2O
Cu
Cu32
What is the state of the surface under reaction conditions?
Application of in situ XPS to catalysis: methanol oxidation on CApplication of in situ XPS to catalysis: methanol oxidation on Cuu
Reaction profile isidentical at 1 atm.
Onset of catalyticactivity at T>200 °C
-0,1
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
0 50 100 150 200 250 300 350 400 450
Y CH2O
Y CO2
Temperature / °C
Yie
ld
CO +H O2 2
CH O+H2 2O
O+ O
CH OH3
CH OH3
CH OH3
CO+H2
CH O+H2 2
O2
Cu0
Oxsurf
Oxbulk
Subox
Ovol
A. Knop-Gericke et al., Topics Catal. 15, 27 (2001).
In situ NEXAFS
I.E. Wachs & R.J. Madix, Surf. Sci. 76, 531 (1978); A. F. Carley et al., Catal. Lett. 37, 79 (1996).UHV XPS
O-(a) + 2CH3OH(g) 2CH3O-(a) + H2O(g)
CH3O-(a) CH2O(g) + H+(a)
Questions for in situ XPS: - Quantitative analysis of surface species- Carbon species on the surface- Depth-dependent analysis
CH3OH + O2 ~ 0.5 mbar
suboxide phase: - only present in situ
Partial oxidation of methanolPartial oxidation of methanol
Experimental conditionsExperimental conditions
sample: polycrystalline Cu foil
Variations of mixing ratios: CH3OH : O2 = 1:2, 3:1, 6:1; T = 400 °C; p = 0.6 mbarTemperature series: gas mixture at room temperature: CH3OH : O2 = 3:1;
p = 0.6 mbar; temperature: 25 °C → 450 °Cflow rates: 10 ... 20 sccm
XPS measurements Beam line U49/2-PGM1 at BessyEnergy range 100...1500 eVtotal spectral resolution 0.1 eV @ 500 eV
O 1s, C 1s, Cu 3p, Cu 2p: KE ~ 180 eVValence Band: KE ~ 260 eV
Depth profiling with KEs 180 eV, 350 eV,500 eV, 750 eV
8 6 4 2 0 -2Binding energy (eV)
VB
Δ ~ 2 eV
160
140
120
100
80
60
40
20
0
Nor
mal
ized
cou
nt ra
te (c
ps/m
A)
542 540 538 536 534 532 530 528 526
Binding energy (eV)
CH
3OH
(g)
CH
2O (g
)H
2O (g
)
CO
2(g
)
O2 (g) Cu2O
?
?
Methanol oxidation on Cu: O1s spectraMethanol oxidation on Cu: O1s spectra
400 °C, 0.5 torr
CH3OH:O2
1:2
3:1
6:1
O 1s
O1s depth profilingO1s depth profiling
1.3
1.2
1.1
1.0
0.9
0.8
0.7529.
9 eV
/ 53
1.5
eV p
eak
area
ratio
14121086Inelastic mean free path (Å)
I529.9/I531.5=n529.9/n531.5·exp[-(z531.5-z529.9)/λ]
Δz = 3 Ǻ, n529.9/n531.5 = 1.6
14
12
10
8
6
Inel
astic
mea
n fre
e pa
th in
Cu
(Å)
800600400200Kinetic energy (eV)
from: Tanuma at al., SIA 17, 911 (1991).
CH3OH : O2 = 3:1
Nor
mal
ized
cou
nt ra
te (a
.u.)
534 532 530 528 526Binding energy (eV)
sub-surfaceoxygen
surfaceoxygen
A
180 eV
350 eV
500 eV
750 eV
Methanol oxidation on Cu: C1s spectraMethanol oxidation on Cu: C1s spectra
292 288 284 280Binding energy (eV)
CO
2(g
)
CH
3OH
(g)
CH
2O (g
)
Camorph
C 1s160
140
120
100
80
60
40
20
0
Nor
mal
ized
cou
nt ra
te (c
ps/m
A)
542 540 538 536 534 532 530 528 526
Binding energy (eV)
CH
3OH
(g)
CH
2O (g
)H
2O (g
)
CO
2(g
)
O2 (g) Cu2Osub-surface
oxygensurfaceoxygen
400 °C
CH3OH:O2
1:2
3:1
6:1
O 1s
MetastabilityMetastability of the Subof the Sub--Surface OxygenSurface Oxygen
536 534 532 530 528
Binding energy (eV)
CH
3OH
(g) sub-surface
oxygensurfaceoxygen
H2O
(g)
CH
2O (g
)
CH3OH:O2=6:1
CH3OH only
difference
536 534 532 530 528
Binding energy (eV)
CH
3OH
(g) sub-surface
oxygensurfaceoxygen
H2O
(g)
CH
2O (g
)
CH3OH:O2=6:1
CH3OH only
difference
Correlation of catalytic activity and surface speciesCorrelation of catalytic activity and surface species
CH2O yield vs sub-surface oxygen peak area
0.20
0.15
0.10
0.05
0.00
543210
normalized sub-surface oxygen peak area
1:2
3:1
6:1
CH
2O p
artia
l pre
ssur
e(m
bar) mixing ratio series
(T = 400 °C)
250 °C
300 °C
400 °C
450 °C
400 °C
150 °C
200 °C
temperature series(CH3OH:O2=3:1)
Open questions: What is the nature of thesub-surface oxygenspecies?What is its role in thecatalytic reaction?
Depth profilingDepth profiling
0.6
0.5
0.4
0.3
0.2
0.1
0.0sub-
surfa
ce o
xyge
n : C
u pe
ak ra
tio
14121086Inelastic mean free path (Å)
CH3OH:O2 = 3:1CH3OH:O2 = 6:1
(calculated from Cu 3p and sub-surface O 1s)
Reducing conditions
Open questions: What is the nature of the sub-surface oxygen species?What is its role in the catalytic reaction?
52
IntroductionIntroduction
Literature carbon laydown selective hydrogenation“similar” catalysts different activity & selectivity
(structure sensitivity?)
Selectivity issue: what defines selectivity?
[ ]+ H2
InIn--situ XPS: situ XPS: Pd 3d depth profilingPd 3d depth profiling
Not only adsorbate-induced surface core level
shift !
But on-top location!
InIn--situ XPS: situ XPS: C1s (Switching off experiments)C1s (Switching off experiments)
1: reaction mix.
2: C5 off
3: H2 off
4: C5 gas-phase
InIn--situ XPS: situ XPS: Pd 3d (Switching off experiments)Pd 3d (Switching off experiments)
1: reaction mix.
2: C5 off
3: H2 off
InIn--situ XPS: situ XPS: Pd vs. C depth profilingPd vs. C depth profiling
100 200 300 400 500 600 700 80020
25
30
35
40
45
50
55
60
65
70A
Palladium Inelastic Mean Free Path [Å]
Car
bon
Con
tent
[%]
Electron Kinetic Energy [eV]
Experiment: 358 K Model: 358 K Model: 3.5 ML Experiment: C5 off Model: C5 off
5.25 7.15 9.05 15.15
20
25
30
35
40
45
50
55
60
65
70
75100 200 300 400 500 600 700 800
22.3513.2510.25
Electron Kinetic Energy [eV]
Car
bon
cont
ent [
%]
Carbon Inelastic Mean Free Path [Å]
Epxeriment: 523K Model: 523K Model: 4.2 ML
7.25
B
0.3 nm
1.4 nm
14 nm
core statesatom specificquantitativecomplex final state effectschemical shift concepttheoretically difficult accessiblecan be applied in the mbar rangesurface sensitivedepth profiling
Summary
1. W. Göpel, Chr. Ziegler: Struktur der Materie: Grundlagen, Mikroskopie undSpektroskopie, Teubner Verlagsgesellschaft, Stuttgart-Leipzig, 19912. M. Henzler, W. Göpel: Oberflächenphysik des Festkörpers, TeubnerVerlagsgesellschaft, Stuttgart-Leipzig, 19913. W. Göpel, Chr. Ziegler : Einführung in die Materialwissenschaften, TeubnerVerlagsgesellschaft, Stuttgart-Leipzig, 19964. D. Briggs, M. P. Seah: Practical Surface Analysis, Volume 1: Auger and X-RayPhotoelectron Spectroscopy, 2. Auflage, John Wiley & Sons, Chichester, 19905. C. D. Wagner, W. M. Riggs, L. E. Davis, J. F. Moulder, G. E. Muilenberg: Handbookof X-Ray Photoelectron Spectroscopy, Physical Electronics Division, Perkin-ElmerCorporation, Eden Prairie, Minnesota, 19796. H. Lüth: Surfaces and Interfaces of Solid Materials, 3. Auflage, Springer Verlag,Berlin, 19957. G. Ertl, J. Küppers: Low Energy Electrons and Surface Chemistry, VCHVerlagsgesellschaft, Weinheim, 19858. K. Siegbahn, C. Nording et al.: ESCA Applied to Free Molecules, North-Holland,Amsterdam, 19719. M. Cardona, L. Ley: Photoemission of Solids, Topics in Applied Physics, Band 26und 27, Springer Berlin, 197810. M. Grasserbauer, H. J. Dudek, M. F. Ebel: Angewandte Oberflächenanalyse mitSIMS, AES und XPS, Springer Berlin, 197911. D. Briggs and J. T. Grant: Surface Analysis by Auger and PhotoelectronSpectroscopy, Surface Spectra and IM Publications 2003
Literature