EPMA SEMEDS Warren Straszheim PhD EPMA Ames Lab 227 Wilhelm SEMEDS MARL 23 Town Engineering wesaiaiastateedu 5152948187 With acknowledgements to John Donovan of the University of Oregon ID: 1030798
Download Presentation The PPT/PDF document "Electron Beam Analysis (" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. Electron Beam Analysis (EPMA, SEM-EDS)Warren Straszheim, PhDEPMA, Ames Lab, 227 WilhelmSEM-EDS, MARL, 23 Town Engineeringwesaia@iastate.edu 515-294-8187With acknowledgements to John Donovan of the University of Oregon
2. Instrumental TechniquesExcitemeasure characteristic responsequantify by comparison to standards
3. Bulk or microanalysisCan excitation be focused?Can detector be focused?
4. Electron beam microanalysisExcitation: focused electron beamSample interactionssecondary electronsbackscattered electronsauger electronscathodoluminescenceabsorbed currentX-rays
5. Precise x-ray intensitiesHigh spectral resolutionSub-micron spatial resolutionMatrix/standard independentAccurate quantitative chemistryElectron-Sample Interactions
6. characteristic emissionsBe and heavier elements background (bremsstrahlung)X-rays
7. X-ray Lines - K, L, MKa X-ray is produced due to removal of K shell electron, with L shell electron taking its place. Kb occurs in the case where K shell electron is replaced by electron from the M shell.La X-ray is produced due to removal of L shell electron, replaced by M shell electron.Ma X-ray is produced due to removal of M shell electron, replaced by N shell electron.
8. Ranges and interaction volumesIt is useful to have an understanding of the distance traveled by the beam electrons, or the depth of X-ray generation, i.e. specific ranges. For example: If you had a 1 um thick layer of compound AB atop substrate BC, is EPMA of AB possible?
9. Differences between SEM and EPMAMany shared componentsResulting from intent - imaging vs. analysisStability (higher for EPMA)Current capability (higher for EPMA)Spatial resolution (higher for SEM) via smaller spot and limited aberration correctionattached analyzer (EDS vs. WDS)
10. EDS vs. WDStechnology – solid state crystal vs. wavelength spectrometerResolution~126 eV vs 20eVP/B ratio Detection limitcount rate limitations 500 kcps in total vs. 70 kcps/elementparallel vs. serial operation
11. Spectral ResolutionWDS provides roughly an order of magnitude higher spectral resolution (sharper peaks) compared with EDS. Plotted here are resolutions of the 3 commonly used crystals, with the x-axis being the characteristic energy of detectable elements.Note that for elements that are detectable by two spectrometers (e.g., Y La by TAP and PET, V Ka by PET and LIF), one of the two crystals will have superior resolution (but lower count rate). Reed, 1995, Fig 13.11, in Williams, Goldstein and Newbury (Fiori volume)
12. Spectrometer EfficiencyThe intensity of a WDS spectrometer is a function of the solid angle subtended by the crystal, reflection efficiency, and detector efficiency. Reed (right) compared empirically the efficiency of various crystals vs EDS. However, the curves represent generation efficiency (recall overvoltage) and detection efficiency. Reed, 1996, Fig 4.19, p. 63Reed suggests that the WDS spectrometer has ~10% the collection efficiency relative to the EDS detector.How to explain the curvature of each crystal’s intensity function? At high Z, the overvoltage is presumably minimized (assuming Reed is using 15 or 20 keV). Low Z equates larger wavelength, and thus higher sinq, and thus the crystal is further away from the sample, with a smaller solid angle.
13. Effect of voltage Excitation volume goes as V1.7Available X-ray lines25kV 5um15kV 2.5um10kV 1.3um5kV 0.4um
14. Typical steel spectrum, 15 kV
15. Lines available at low kVNote overlap of V, Cr, Mn, and Fe. Also, O has its line at 0.53 keV.
16. Effect of currentspatial resolution reduced with high currentsgreater sensitivity with high currentsdetectabilityprecision/repeatability
17. Overlap considerationsSmaller issue for WDS – effects background choicesDeconvolution option for EDS if statistics permitStatistics become problematic if trace element on major element background
18. EDS Overlap: S, Mo, HgHgS stdLine TypeWt%Wt% SigmaAtomic %SK series13.380.1449.15HgM series86.620.1450.85Total100.00100.00Stoichiometry is on-the-mark - in this case.
19. WDS “overlap”: HgS, PbS, MoNote that signal drops to background in between most peaks. Mo tail interferes with S.
20. Rare earths by EDS and WDSPr peak fits between Ce La and Lb peaks. ErDyTb
21.
22. EDS Atomic fractionCompoundFeYCePrNdGdTbDyHoErLuD5 Y2Fe1788.4911.51B4 Ce2Fe1789.0510.95B5 Pr2Fe1788.3911.61C1 Nd2Fe1788.8111.19C2 Gd2Fe1790.129.88C5 Tb2Fe1786.2113.79D1 Dy2Fe1788.4311.57D2 Ho2Fe1787.5912.41D3 Er2Fe1784.6915.31E2 Lu2Fe1789.7710.232/19 = 10.53%
23. Suitable samplessolid/rigidstable under beamconductive (while under beam)nonconductive samples can be coated with C or metal (e.g., Au, Pt, Ir)(coating obscures features and elements but only a little)
24. Samples includeMetalsGeologic samplesCeramicsPolymersExperimental materials
25. Quantitative ConsiderationsHomogeneous (within excitation volume)Thick (enclosing interaction volume);therefore, problems with layered samplesKnown geometry (preferably “flat” compared to excitation volume; thus, polished); therefore problems with rough samplesBe smart with construction (e.g., glass vs. Si)Standards collected each time vs.Standardless and normalization
26. Matrix effectsZ-A-F or Phi-Rho-Z corrections accounting for penetration depth, absorption, secondary fluorescenceAccuracy depends on well known curvature. Alternatively, need standard in region for better results.
27. Range of Quantitation100% down to 0.05% (500 ppm) EDS, 0.001% (10s of ppm) WDS Limited by statistics, differentiation from backgroundMore counts help!
28. Mapping and Line-scansPoint analysis are most sensitive to concentration differences (30s/point)Line scans are next (500 ms/pixel)Mapping is least sensitive (12 ms/pixel)Graphics convey much information quickly(i.e., a picture is worth a thousand words)
29. Digital image showing regions of analysis and line-scan
30. Mg portion of overlapped peak
31. Ge portion of overlapped peak
32. Line-scan using typical windowsGe-Mg overlap causes problems
33. Line-scan using deconvolutionGe contribution is stripped from Mg profile
34. Mapping using deconvolution
35.
36. EDS-WDS comparison
37. “Harper’s Index” of EPMA1 nA of beam electrons = 10-9 coulomb/sec1 electron’s charge = 1.6x 10-19 coulombergo, 1 nA = 1010 electrons/secProbability that an electron will cause an ionization: 1 in 1000 to 1 in 10,000 ergo, 1 nA of electrons in one second will yield 106 ionizations/secProbability that ionization will yield characteristic X-ray (not Auger electron): 1 in 10 to 4 in 10.ergo, our 1 nA of electrons in 1 second will yield 105 x-rays.Probability of detection: for EDS, solid angle < 0.03 (1 in 30). WDS, <0.001ergo 3000 x-rays/sec detected by EDS, and 100 by WDS. These are for pure elements. For EDS, 10 wt% = 300 X-rays; 1 wt% = 30 x-rays; 0.1 wt % = 3 x-ray/sec.ergo, counting statistics are very important, and we need to get as high count rates as possible within good operating practices.From Lehigh Microscopy Summer School
38. Raw data needs correctionThis plot of Fe Ka X-ray intensity data demonstrates why we must correct for matrix effects. Here 3 Fe alloys show distinct variations. Consider the 3 alloys at 40% Fe. X-ray intensity of the Fe-Ni alloy is ~5% higher than for the Fe-Mn, and the Fe-Cr is ~5% lower than the Fe-Mn. Thus, we cannot use the raw X-ray intensity to determine the compositions of the Fe-Ni and Fe-Cr alloys. (Note the hyperbolic functionality of the upper and lower curves)
39.
40. n = principal quantum number and indicates the electron shell or orbit (n=1=K, n=2=L, n=3=M, n=4=N) of the Bohr model. Number of electrons per shell = 2n2l = orbital quantum number of each shell, or orbital angular momentum, values from 0 to n –1Electrons have spin denoted by the letter s, angular momentum axis spin, restricted to +/- ½ due to magnetic coupling between spin and orbital angular momentum, the total angular momentum is described by j = l + sIn a magnetic field the angular momentum takes on specific directions denoted by the quantum number m <= ABS(j) or m = -l… -2, -1, 0, 1, 2 … +lRules for Allowable Combinations of Quantum Numbers:The three quantum numbers (n, l, and m) that describe an orbital must be integers. "n" cannot be zero. "n" = 1, 2, 3, 4... "l" can be any integer between zero and (n-1), e.g. If n = 4, l can be 0, 1, 2, or 3. "m" can be any integer between -l and +l. e.g. If l = 2, m can be -2, -1, 0, 1, or 2. "s" is arbitrarily assigned as +1/2 or –1/2, but for any one subshell (n, l, m combination), there can only be one of each. (1 photon = 1 unit of angular momentum and must be conserved, that is no ½ units, hence “forbidden transitions)No two electrons in an atom can have the same exact set of quantum numbers and therefore the same energy. (Of course if they did, we couldn’t observably differentiate them but that’s how the model works.)One slide Schrödinger Model of the Atom
41. Origin of X-ray Lines for K and L Transitions