McSAS: software for the retrieval of model parameter distributions from scattering patterns

Area of Science:

• Materials Science
• Physical Chemistry
• Biophysics

Background:

• Small-angle scattering (SAS) is a powerful technique for characterizing nanoscale structures.
• Analyzing SAS data often involves complex, underdetermined problems requiring specialized software.
• Existing methods may impose limitations on the mathematical form of particle size distributions.

Purpose of the Study:

• To introduce McSAS, a user-friendly, open-source Monte Carlo regression package for SAS data analysis.
• To enable the analysis of scattering data using uncorrelated, shape-similar particles or contributions.
• To overcome limitations of traditional methods by allowing for complex, multimodal, or unusual parameter distributions.

Main Methods:

• The software employs a form-free Monte Carlo approach.
• Users select scatterer models from a comprehensive library and define parameter variation intervals.
• The package handles underdetermined problems when sufficient external information is provided.

Main Results:

• McSAS can extract complex parameter distributions without prior assumptions on their mathematical form.
• The software outputs model parameter distributions in absolute volume fraction with uncertainty estimates.
• It provides statistical measures of the distribution, such as mean and variance.

Conclusions:

• McSAS offers a flexible and powerful tool for evaluating SAS curves, accommodating diverse particle shapes and distributions.
• The software enhances the reliability of SAS analysis by providing uncertainty estimates.
• McSAS is suitable for integration into automated data reduction pipelines at laboratory instruments and synchrotrons.