Electron probe microanalysis EPMAWhat’s the point?TopicsWhat makes it difficult?PowerPoint PresentationSlide 6Slide 7Slide 8Slide 9Slide 10Slide 11Slide 12Slide 13Slide 14Slide 15Slide 16Slide 17Slide 18Slide 19Slide 20Slide 21Slide 22Slide 23Slide 24Slide 25Slide 26Slide 27Slide 28Electron probe microanalysisEPMA Difficult/Tricky Materials for EPMAWhat’s the point?Some materials are somewhat to extremely difficult to do EPMA on. Guidance and passed on experience is very helpfulAnd: How do you know something is tricky?Topics• Thin films and particles (discussed separately)• Alkali-rich glasses• Hydrous alkali-rich glasses• Carbonates• Hydrous phases (clays, micas, hydroxides, zeolites…)• Apatite and similar non-isometric materials• Porous materials• Magnetic materials• “Hot” (radioactive) materials• “Hard to handle” (matrix correction) element pairsWhat makes it difficult?• Counts change over time– Due to heating– Due to charging– Due to oxidation• Counts change depending on crystallographic orientation • Difficult to determine the element abundance• Impossible to detect the elementDrop in Na Ka with counting time: weathered Ethiopian volcanic glassThe prior slide shows an example of “mobility” of some elements in insulating samples, in response to high energy electron bombardment in the electron probe. Alkali elements in glasses are particularly mobile. Jack Lineweaver at Corning in 1963, in studying the degradation on the inside surfaces of TV tubes, collected oxygen released due to the electron bombardment, seeing that it changed with time.J. Lineweaver (1963) Oxygen outgassing caused by electorn bombardment of glass, J. Applied Physics, 34, 1786-1791.Lineweaver hypothesized that the 10 to 28 keV electrons travel thru the Al coating on the glass surface and are implanted into the glass at some depth, which then creates an electrical field which the Na atoms migrate toward. J. Lineweaver (1963) Oxygen outgassing caused by electorn bombardment of glass, J. Applied Physics, 34, 1786-1791.Non-bridging oxygen atoms then are freed to migrate to vacated positions nearer the surface and/or lose their electrons to the Al “electrode” and be released into the vacuumG. Morgan and D. London, 1996, Optimizing the electron microprobe analysis of hydrous alkali aluminosilicate glasses, American Mineralogist, 81, 1176-1185. Morgan and London examined hydrous aluminosilicate glasses, which many volcanologists study, as “melt inclusions” trapped in minerals. Here, the dramatic drop in Na Ka counts versus time is shown. It is worst at 20 nA, and less at 2 nA (though still creates errors).Potassium also drops, though not as dramatically.NaHydrous alkali glasses: worst case scenarioThe related phenomenon of “grow in” is shown to the right, with Al Ka showing the increase in counts with time more than Si Ka.Presumably, as the alkali atoms migrate deeper into the material toward implanted electrons, Si and Al atoms either migrate upward -- or maybe there just are just less “other” atoms for the beam to interact with.SiAlG. Morgan and D. London, 1996, Optimizing the electron microprobe analysis of hydrous alkali aluminosilicate glasses, American Mineralogist, 81, 1176-1185.There are several approaches:• Defocus the beam (reduce the impact) and do no correction (but you can’t do this with small melt inclusions)• Use a low current (reduce the impact) and do no correction (statistics will be poor)• Use a focused beam and high current and do a time dependent correction (but you have to have the software).NaSo how do you deal with this???• In a later paper*, Morgan and London further consider this; one conclusion being that 20 kV is better than 15 kV.* G. Morgan and D. London, 2005, Effect of current density on the electron microprobe analysis of alkali aluminosilicate glasses, American Mineralogist, 90, 1131-1138.OK, but how to quantify the Water (in the Glass)?* B. Nash, 1992, Analysis of oxygen with the electron microprobe: applications to hydrated glass and minerals, American Mineralogist, 77, 453-457. There are several ways to go: don’t use EPMA, use FTIR or SIMS. But if you want EPMA, there are two options:• “Water by difference” where you assume the difference from 100 wt% is water + other volatiles (CO2, F, Cl and any other unanalyzed elements). This is not very precise and embodies all errors.• Measure oxygen, as suggested by Nash. However, this requires careful attention to carbon coating (must have same thickness as standards) plus problems with 2nd order Na Ka overlap on the O Ka peak.Stormer et al. also showed that Ca Ka intensities in apatite also vary with time if the electron beam was perpendicular to the c axis, though in a somewhat more complicated way.J. Stormer, M. Pierson and R. Tacker, 1993, Variation of F and Cl X-ray intensity due to anisotropic diffusion in apatite during electron microprobe analysis; American Mineralogist, 78, 641-648.Sample Orientation Intensity Variations: F in ApatiteCarbonates - DifficultiesEPMA of carbonates have always been difficult: • since C is not measured (nor O), there always is a very low total and so any errors are not immediately obvious. However, Probe for EPMA software does away with that issue.• beam currents of 10-20 nA will immediately cause apparent changes to the surface of calcite (in particular), e.g. black irregular area; whether this is a hole or build up of contamination, is not clear to me (I suspect the former)• I have always “gotten by” with using very low beam currents (1 nA) and a defocussed (10-20 um) beam which seemed to give satisfactory results for calcite• However, I have always wondered “why”…Carbonates - DifficultiesBut with the TDI ability with Probe for EPMA, I have started to investigate this -- here 15 kV with 20 nA…hmmm….not at all what I expected, which would have been dropping Ca…but it is increasing…and the O has a periodic behavior, dropping over the first 20 seconds then oscillating a bit…Carbonates - DifficultiesDolomite seems somewhat similar, though less extreme at first glance, both cations Ca and Mg increase (though some of the up and down behavior), and O drops as before then rises … so similar in general.Obviously only by using a very similar standard to the unknown, would you have any chance of getting a decent EPMA result.• Beam sensitive samples: require care, such as lower
View Full Document
Unlocking...