Wednesday, July 15, 2009

Some differences between analyzing minerals in natural rocks and minerals grown in experiments

For my PhD research I worked with Tasmanian metamorphic rocks. The schists I studied had major minerals in the 2 mm to 2 cm size range (though, of course, most of them were closer to 2 mm than 2 cm in size). As a result of these large sizes the approach to microprobe time for mineral analysis was to find sets of minerals near one another (in textural equilibrium) and select points for analysis. I’d take photographs of the selected minerals, make marks upon the printouts showing the locations of the selected points, and then leave the microprobe to do all of the analyses after I’d gone home for the day. Then I’d use ArcMap to align the photographs with the x-y coordinates from the microprobe so that I’d have a good record of where each analysis point was located.

Now that I’ve started my first post-doctoral position things will be somewhat different; I’ve signed on for an experimental position. This means that instead of analyzing mineral assemblages in natural rocks and doing calculations to try to determine at what temperature and pressure they must have formed, I shall be growing minerals at specific temperatures and pressures and analyzing them to see what their compositions are at those temperatures. (The results of these experiments will, in theory, be used by others who wish to determine the temperatures and pressures at which their minerals in natural rocks formed.)

I have not yet gotten to the point of being able to do my own experiments, but today I have spent some time in the microprobe lab with one of my colleagues as she analyzed the minerals which she grew in one of her experiments. The capsule in which she grew the minerals is only a few millimeters long, and the minerals present are all very small. The microprobe here is set up to use a 1 micron beam size, so the minerals need to be at least 1 mm in diameter in order to obtain analyses which are from a single mineral (the one I used in Australia was set up to analyze minerals that are at least 10 microns in diameter—remember that there are 1000 microns in every millimeter!).

Because the minerals here are so small they don’t tend to use the microprobe in automatic mode—instead one stays with the machine while it does the analyses. The procedure is to find a mineral you wish to analyze, analyze it, look at the results, and if it is “good” (has an appropriate proportions of elements for that mineral) move on to a new location to begin the next analysis. If it is “bad” (contains other elements than should be in that mineral and/or contains the correct elements in an unexpected proportion) it is mostly likely due to contamination from a neighboring mineral—sometimes the grain isn’t large enough and the electron beam also analyses the grain next to it, sometimes the grain isn’t thick enough, and the analysis includes the underlying grain. In this case you also move on to the next analysis, but you keep track of the number of “good” vs “bad” data, so that you can obtain sufficient “good” results to be able to state with confidence the composition of each of the minerals in the sample.

The approach is very different to what I have been used to, and requires one to become accustomed to looking at the raw microprobe data (hitherto I simply took the report generated by my microprobe operator and ran the results through the program Ax to convert it to number of cations per oxygen, and then dealt with the numbers only in that form. I am very much looking forward to starting my own experiments—I think that the procedures used here will help me to better understand the microprobe and how it operates, and it all sounds like fun. But first, there is much background reading to do…

No comments: