Exploring the Use of Crystal-Chemistry Based Manipulations of Digital X-Ray Maps as a Method of Discriminating among Soil Samples

Determining whether one soil sample is significantly different from another can be an analytical challenge depending on the resources available to the investigator. In some cases bulk chemical assays are sufficient, while in cases involving the same soil type more specific descriminators are required. Examination of polished rock thin sections, with optical microscopy and with digital X-ray maps collected using wavelength (WDS) or energy dispersive (EDS) spectrometers attached to an electron microprobe or a scanning electron microscope is easily accomplished. This presentation will describe the efforts required to use crystal-chemical based computational manipulations of X-ray maps, on a spatially registered x by y, pixel by pixel basis, to distinguish different soils.

A small selection of soil samples was prepared for characterization using our group's ASPEX instrument. As a comparison method, I explored the use of X-ray maps acquired with more common electron beam instruments. Digital X-ray maps can be manipulated using software tools like IDL to evaluate the chemistry of individual grains. Crystal chemistry based criteria are used to sort the grains by mineral type, accomplishing a modal determination based on tens or hundreds of thousands of data points. If required, the modal data can be used to calculate a bulk composition. Importantly though, this method allows for the characterization of unique or minor chemical features, of even individual soil grains, that cannot be observed with bulk chemical assays.

While this method works well with sections of rock (closed or known system) where the mineralogy is determined prior to X-ray mapping, it requires extra effort to use with soil (open or unknown system) samples.