A new spin on pair distribution function analysis : Introducing magnetic PDF

Orateur : Benjamin A. FRANDSEN (Department of Physics, Columbia University) & Simon J. L. BILLINGE (Department of Applied Physics and Applied Mathematics, Columbia University - Condensed Matter Physics and Materials Science Department, Brookhaven National Laboratory)

Significant efforts have been made in recent years to experimentally investigate short-range magnetic correlations through the analysis of magnetic diffuse neutron scattering. The development of an experimental methodology to accurately measure and model such diffuse scattering is of great importance for condensed matter physics and materials science research topics that involve local magnetic structure, including geometrically frustrated spin systems, unconventional superconductors, multiferroics, diluted magnetic semiconductors, and others. Reverse Monte Carlo techniques that iteratively fit randomly generated structural models in momentum space have been used successfully, demonstrating that diffuse magnetic scattering can be rich in information. [1-2] Recently [3-4], we developed an alternative real-space approach to investigating local magnetic correlations, which we call magnetic pair distribution function (mPDF) analysis in analogy to the more familiar atomic pair distribution function. This experimentally accessible and intuitive quantity reveals magnetic correlations directly in real space and places diffuse and Bragg scattering on an equal footing, thereby gaining sensitivity to both short- and long-range magnetic order. In this talk, I will present the basic theory behind mPDF analysis, provide simulated and experimental examples of the mPDF from representative magnetic structures, and give a demonstration of the newly available “mPDFfit” software package for calculating and refining magnetic PDFs. The mPDFfit software is an open-source package based on the DiffPy library of python packages created to facilitate the analysis of diffraction data. mPDFfit will soon be freely available from the DiffPy website at http://www.diffpy.org/index.html.

Télécharger dans mon agenda