| Soft x-ray spectromicroscopy provides spectral data on the chemical speciation of light elements at sub-100 nanometer spatial resolution. The high resolution imaging places a strong demand on the microscope stability and on the reproducibility of the scanned image field, and the volume of data necessitates the need for improved data analysis methods. This dissertation concerns two developments in extending the capability of soft x-ray transmission microscopes to carry out studies of chemical speciation at high spatial resolution. One development involves an improvement in x-ray microscope instrumentation: a new Stony Brook scanning transmission x-ray microscope which incorporates laser interferometer feedback in scanning stage positions. The interferometer is used to control the position between the sample and focusing optics, and thus improve the stability of the system. A second development concerns new analysis methods for the study of chemical speciation of complex specimens, such as those in biological and environmental science studies. When all chemical species in a specimen are known and separately characterized, existing approaches can be used to measure the concentration of each component at each pixel. In other cases (such as often occur in biology or environmental science), where the specimen may be too complicated or provide at least some unknown spectral signatures, other approaches must be used. We describe here an approach that uses principal component analysis (similar to factor analysis) to orthogonalize and noise-filter spectromicroscopy data. We then use cluster analysis (a form of unsupervised pattern matching) to classify pixels according to spectral similarity, to extract representative, cluster-averaged spectra with good signal-to-noise ratio, and to obtain gradations of concentration of these representative spectra at each pixel. The method is illustrated with a simulated data set of organic compounds, and a mixture of lutetium in hematite used to understand colloidal transport properties of radionuclides. Also, we describe here an extension of that work employing an angle distance measure; this measure provides better classification based on spectral signatures alone in specimens with significant thickness variations. The method is illustrated using simulated data, and also to examine sporulation in the bacterium Clostridium sp. |
