The Department of Materials Science and Engineering welcomed I. Cevdet Noyan from Columbia University as part of the MSE Seminar Series. Noyan presented a seminar titled “Diffraction Analysis of Nanocrystalline Powder Samples.” Below is the abstract:
Most formalisms utilized for diffraction analysis of the structure of nanocrystalline powder samples are based on basic concepts developed over the past century. Commercial or open-source programs based on these formalisms are included with most laboratory-based diffraction systems. In most of these codes, the algorithms used for the propagation of errors associated with the computed structural parameters are not explicitly described but, in general, these are based on photon counting statistics and peak-fits. This approach can cause problems in determining the degree of uncertainty of published results, since for non-ideal samples both photon statistics and Bragg sampling influence the measured Bragg peak intensities.
In the past decade, we have used rigorous forward modeling of diffraction from powder samples to investigate the diffraction process. Our simulations showed that the sampling rules and statistics for powder diffraction, and concepts like reflection multiplicity, the “Lorentz factor”, sampling statistics, etc. are different when the diffraction signal originates from nanocrystalline powders. In this lecture, we will discuss these results and show that quantifying uncertainty budgets for powder diffraction experiments through rigorous forward modeling is essential for understanding the data from x-ray diffraction experiments.