New Mexico Mineral Symposium — Abstracts

Electron microprobe analysis is one of the most important techniques, if not the major one, used in mineral research today. It is the only technique that combines the desirable qualities of being quantitative, being micro (having high spatial resolution), and being nondestructive to the material analyzed. Like any technique, it does have some distinct limitations. In this talk I want to discuss the kinds of information (signals) we can obtain from a microprobe and how they are produced. These include secondary electron, back-scattered electron, and x-ray
information. I will also discuss many of the shortcomings, disadvantages, and limitations of microprobe analysis.

Secondary electrons (SE) are produced when primary electrons, those that have been accelerated down the column of the instrument toward the sample, "collide" with electrons in the shells forming the outer portions of atoms and knock them out of orbit. Prodigious numbers of these low-energy, secondary electrons are produced, and they are attracted to the secondary-electron detector by a charged grid. Because the energy of these electrons is very weak (a few 10's of volts at best), the efficiency with which they travel is highly affected by the topography of the sample. Contrast between high- and low-efficiency areas gives rise to the secondary electron image, which is a faithful rendering of the sample topography.

Back-scattered electrons (BSE) are produced when primary electrons "bounce" or "reflect" off the surface of the sample. In fact, the primary electrons occasionally try to collide with nuclei of the atoms in the sample, but tremendous nuclear forces repel the electrons and they are scattered back elastically. This is much like the case of a billiard ball--the electrons have almost the same energy or speed as before but they are going in a different direction. The back-scatter coefficient of a material is related mainly to its average atomic number. Minerals that contain even modest amounts of lead or rare-earth elements would be very bright on a BSE photograph, as compared to normal silicates, while graphite or diamond would be very dark. With modern detectors it is easy to detect the compositional differences between such similar materials as olivine, orthopyroxene, and clinopyroxene or even zoning within a crystal.

X-rays are produced on those rare occasions when primary electrons knock inner-shell electrons out of their orbits. Electron vacancies in the inner shells are very unstable, and the atoms will respond immediately to remedy the situation. This is accomplished by the transfer of an electron from the next higher shell of the atom. Because the electron has to have more energy to be in a higher shell, it must lose energy to go to the lower shell. It can do this by producing a photon (a particle of electromagnetic energy) that is, in most cases, part of the x-ray spectrum. Because the energy differences between electron shells are constant and well known for all elements, it is possible to determine the composition of a sample by examining all of the x-rays (the x-ray spectrum) coming from the material. Quantitation can be accomplished by counting under constant conditions, comparing with suitable standards, and mathematically correcting for the effects that the elements in the mineral have on the x-rays produced.

Microprobe analysis is an important and powerful technique, but like all analytical techniques it has limitations. Electron microprobes cannot detect hydrogen, helium, or lithium and have great problems with most elements lighter than sodium. Probe analysis is an elemental technique; it doesn't tell how the elements are combined, their valences, or the structure of the mineral. Pyrite, marcasite, pyrrhotite, troilite, melanterite, etc., would all show lots of iron and sulfur. It requires careful technique to distinguish pyrite/marcasite from the rest, but there is no probe technique available that will distinguish pyrite from marcasite. Heat is generated in the mineral by the electron beam. Some delicate minerals can be damaged or destroyed by this heat, and even when the effect is slight, the probe analysis may be perturbed. Further damage to mineral samples occurs during preparation. Quantitative analyses require a flat and polished surface. In most cases that means the mineral is embedded in epoxy, and a flat surface is ground and polished through the sample (good-bye world-class specimen!). Probe analyses use electrons, which are electrically charged particles. The excess electrons have to be drained from the surface being analyzed or the area will charge up and repulse the electron beam. Which all means that most minerals being electrical insulators, have to be coated with carbon and that pretty much destroys the sample's beautiful colors and appearance. Some might consider cost a limitation (a modern, fully automated microprobe with all the whistles and bells costs a tad more than a Hastings triplet). And there are other problems. But, all things considered, probe analysis is still one of the most powerful analytical techniques for looking at minerals.

* This work was performed at Sandia National Laboratories and was supported by the U.S. Department of Energy under Contract no. DEAC04-76DP00789.