PyFitIt

Decipher your X-ray absorption spectra standing on the shoulders of machine learning methods

Whenever you measure a single spectrum of unknown compound or series of spectra in operando regime the question appears about the reliable structural model behind. Machine learning methods open a new perspective to quantify the charge state, local symmetry and structural parameters along with their uncertainties. The PyFitIt project combines best existing approaches to the quantitative analysis of your data in a user-friendly way.

Single spectrum: compare with the references

Upload experimental spectrum and compare it to the finely selected references of bulk and molecular compounds. The classical linear combination fit, or more sophisticated machine learning analysis will provide information about parameters of the local atomic and electronic structure around absorbing atom

Series of spectra: PCA analysis and MCR

The concentrations of individual phases vary upon synthesis or chemical reaction. X-ray absorption spectra, acquired in situ, thus consist of linear combinations of the pure phases. Principal component analysis (PCA) is applied to determine the number of mathematically independent components. Further target transformation or multicurve resolution (MCR) reconstructs meaningful spectra of each phase.

Explore the database

We have collected spectra of well-defined single component species, including bulk and molecular compounds. The structure of each compound is verified by X-ray diffraction, the energy is calibrated using metal foil and all details of measurement are documented. Visit the page with all entries and explore existing databases relevant for your research

Tools Available