DOC PREVIEW
UW-Madison G 777 - Electron Probe Microanalysis

This preview shows page 1-2-3-24-25-26 out of 26 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 26 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Electron probe microanalysis EPMAWhat’s the point?Bulk vs particleThe Good, the Bad, the UglySize: “Mass Effect”Size: Absorption EffectSize - Detector-sample GeometryEDS vs WDSHow Do People Try to Do EPMA of Particles?Two approachesArmstrong-Buseck correctionOther approach: Peak/BackgroundOther approaches: Peak/BackgroundOther approaches: Monte Carlo SimulationsSlide 15DTSA-IIPENEPMAOther Approaches to Particle AnalysisPowerPoint PresentationSlide 20Other concerns regarding particlesDispersion Technique-1Dispersion Technique-2Dispersion Technique-3Inclusions = particlesConclusionVersion: November 16, 2009EPMA of ParticlesElectron probe microanalysisEPMAWhat’s the point?EPMA is traditionally done for bulk material. What are the issues for particles? How precise/accurate are such analyses?Bulk vs particleNormal EPMA assumes that the electron beam is exciting a homogeneous volume, i.e. there is no difference either laterally or vertically. Thus, the matrix correction is being applied in a uniform matter, and there is one applicable (z) profile for each element.However, particles createmany difficulties, including:• Their size (small)• Their venue (location, surroundings)• Their surface (not flat nor smooth)The Good, the Bad, the UglyMost operators of electron microprobes, using WDS, understand the difficulties of trying to do EPMA on non-flat surfaces (particles).However, the real problem is that there are probably 100 EDS systems on SEMs for every 1 WDS on an electron microprobe. And with SEMs and EDS systems so easy to operate, a large number of users make major errors without knowing it, assuming whatever the software spits out as a composition is really what it is.After these slides, you should know the difference between the GOOD, the BAD, and the UGLY… particle analyses.Size: “Mass Effect”Goldstein et al, 1992, p. 479, 481Consider a 2 um diameter sphere (say of NIST glass K412). Mass effect/error: electrons escape from sides of small particles if the E0 is great enough so that electrons scatter out of the body before using up all their energy.Size: Absorption Effect Absorption effect of non-flat upper surface: there is a different path length from the normal flat geometry. In this case, the emergent x-rays will travel ~50% shorter distance in the material, and thus have ~50% of the absorption, and all else being equal, will yield a K-ratio twice as large as it would be if it were larger and had a flat polished surface.Goldstein et al, 1992, p. 479, 481Size - Detector-sample GeometryEDS (shown to right): Variable effect of geometry of trajectory between beam impact area on non-uniform surface and the location of the detector. K-ratios from point 1 could be significantly different than those from point 3, despite the sphere being of uniform composition.WDS: Could be as bad, with up to 5 spectrometers positioned at 5 different orientationsGoldstein et al, 1992, p. 479, 481EDS vs WDSFor all the problems cited, EDS has an advantage over WDS for “particle analysis” in that the spectrometer is in a constant position relative to the sample and beam “landing spot”, whereas WDS spectrometers would be at different positions, creating various different absorption path lengths to each spectromter.WDS 1WDS 2How Do People Try to Do EPMA of Particles?1. Take whatever values come off the EDS with no correction other than built in ‘normalization to 100 wt%’;2. Overscan the sample;3. Put the beam on a top “flat” surface and hope for the best;4. Do #3 but apply Armstrong and Buseck’s geometric correction factors;5. Peak-to-background method;6. Do Monte Carlo simulations with DTSA-2 or PENEPMA and compare with actual K-ratios (time consuming);7. If large enough, mount, polish flat and do traditional WDSTwo approaches• Traditional approach: normalize numbers – but this is not very good (above left table)• Armstrong and Buseck (1975) developed a procedure based upon a regular geometric shape factor, where the different path length and other effects could be used. Method is based on bracketing particle and beam overscanning during collection of spectrum by EDS and modeling of electron path and x-ray propagation out through several shapes – sphere, hemisphere, squared pyramid, and rectangular, tetragonal, cylindrical and right triangular prisms. Correction factors are given in terms of predicted k-factors for pairs of particular elements, vs particle thickness along e beam. This is not easy, takes much trial and error, but apparently can yield fairly good results (see table above).Goldstein et al, 1992, p. 488, 489Armstrong-Buseck correctionGoldstein et al, 1992, p. 490Here is another useful way to view a lot of data: error histograms. The top one shows the “conventional” particle analysis procedure to be very inaccurate, whereas using their geometric-correction approach, the errors shrink greatly. Though there still are problems, which the analyst must recognize and report with any writeup.Other approach: Peak/BackgroundIt is known that whereas geometric effects can create large differences in measured x-ray intensities, the ratio of a characteristic x-ray to the nearby background (=continuum of same energy range) intensity is less sensitive to geometric effects. Thus, an ‘automatic internal normalization factor’ can be created for both a standard and an unknown. This has been shown to work well. The issue then is to find software that has this implemented.This approach is discussed in Goldstein et al, 2nd edition, pp. 490-493.(It is also used in biological EPMA, where the beam can destroy the cellular material, so mass “disappears”.)Other approaches: Peak/BackgroundResults from using P/B approach for 2 minerals. Excellent results for the pyrite, better but not within the 1% error we would like to see.Goldstein et al, 1992, p. 491Other approaches: Monte Carlo Simulations Another approach is to create a similar geometric shape, give it a composition which is what you believe the unknown to be (based upon initial EPMA results). In the model, you place the beam on a particular spot corresponding to where the beam hits it, and then “run the simulation” to create x-ray counts (and similarly for a standard), and then compare the resulting simulated K-ratio with the experimental K-ratio. If the K-ratios match, then the composition is correct; if not, modify the composition and run another simulation. Iterate until the


View Full Document

UW-Madison G 777 - Electron Probe Microanalysis

Documents in this Course
Load more
Download Electron Probe Microanalysis
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Electron Probe Microanalysis and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Electron Probe Microanalysis 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?