The University of Oregon
Micro Analytical Facility

Cameca SX50 Operations Page

Table of Contents

What do you want to do?

If the configuration of the microprobe is the result of a well thought out analytical procedure it is much less likely you'll waste analytical time (... and dollars ...), and much less likely you'll have to return for analyzing something else . . . or analyzing it over again.
Thinking about what you want to analyze goes beyond the research question you want to answer. For example, your research may simply want an evaluation of AB:AN:OR endmembers or perhaps thin film composition which requires more thought and planning.  
However, which elements will the probe need measure? . . . Do you know if you're analyzing alkali-feldspar or plagioclase? . . . What other elements are important to feldspar chemistry? . . . How accurate do your feldspar analyses need to be? . . . Given the list of elements, what is the most efficient spectrometer configuration?

The analysis of a specific phase

Specific phases (...e.g., mineral or glass ...) demand specific attention.   For example, not enough is known about modelling the electron interaction and the subsequent x-ray generation to simply use metal standards for silicate minerals.   It is for this reason we always try to use specific minerals as the most appropriate standards.
Still . . . we can not provide as many mineral standards as we'd like . . . you'll be asked to compromise with dis-similar standards for elements heavier than sulfur, . . . this can be justified with the sophistication of today's computer modelling.   For lighter elements there exists likely problems associated with differences in peak locations and peak shapes . . . for these elements a similar standard is likely to exist.   Feel free to consult with your EPMA facility manager for appropriately choosing your standards.

Choosing the elements

Choosing the elements you need to analyze with this technique can be as easy a looking at someone else's analysis (... for example, the elements for pyroxene as listed in Deer, Howie and Zussman ...).   The list of elements might also include special elements associated with your research question (... e.g., Mg in lunar feldspar, or Al and Fe in detrital quartz ...).   Keep in mind however EPMA is not the best trace element technique ... electron induced continuum background, reasonable count times and possible beam damage limit elemental sensitivities to greater than 100ppm. Still, if in situ measurement of trace elements is required, EPMA may be the only technique.
The list of elements may also need include elements not actually measured.   Oxygen is an example of an element which is present and must be included because it is a major constituant and therefore must be included in the matrix model for proper correction of the other elemental x-ray intensities (... e.g., the corrections for average atomic number and x-ray absorption ...).   Assumptions are therefore made regarding the amount of oxygen present.   It is our practice to assume stoichiometric oxygen (... e.g., 2 oxygens for every silicon atom measured ...), however oxygen associated with ferric iron, after an initial assumption of ferrous (... which is the accepted practice ...), may need to be recalculated based on mineral stiochiometry.   Also, be aware oxygen by difference is also an option and may be sometimes more appropriate.
However, your list of elements shouldn't include elements you really don't need.   Just because someone else analyzed for Ca in olivine at a sensitivity of 100ppm doesn't mean you need to.   Consider your financial budget and the fact Ca at 0.03 weight percent will add 3 minutes to every analysis and possibly $50 to $100 to your invoice.   (On the other hand, you don't want, at a later time, to be adding calcium to your olivine definition and re-analyzing.)


copyrights 1995-2006,
john donovan: March, 2006
Table of goto home page Contents