Towards a robust ad-hoc data correction approach that yields reliable atomic pair distribution functions from powder diffraction data

TL;DR

This paper proposes a largely automated, variational correction method for powder diffraction data that reliably produces atomic pair distribution functions, comparable to existing methods, enabling faster and more automated data analysis.

Contribution

It introduces a variational approach for data correction that automates the process of obtaining reliable PDFs from raw diffraction data, reducing user input and improving efficiency.

Findings

The method yields PDFs comparable to existing programs.

It enables rapid, automated processing of raw diffraction data.

The approach is applicable to x-ray, neutron, and electron data.

Abstract

We examine the equations to obtain atomic pair distribution functions (PDFs) from x-ray, neutron and electron powder diffraction data with a view to obtaining reliable and accurate PDFs from the raw data using a largely \emph{ad hoc} correction process. We find that this should be possible under certain circumstances that hold, to a reasonably good approximation, in many modern experiments. We describe a variational approach that could be applied to find data correction parameters that is highly automatable and should require little in the way of user inputs yet results in quantitatively reliable PDFs, modulo unknown scale factors that are often not of scientific interest when profile fitting models are applied to the data with scale-factor as a parameter. We have worked on a particular implementation of these ideas and demonstrate that it yields PDFs that are of comparable quality to those obtained with existing x-ray data reduction program PDFgetX2. This opens the door to rapid and highly automated processing of raw data to obtain PDFs.