June 10, 2003

Frontiers in Powder Diffraction

The theme of the workshop, “Frontiers in Powder Diffraction,” held on May 19th at the 2003 NSLS Users’ Meeting, was the growing practice and utility of powder diffraction and related techniques in a variety of contexts. Speakers covered work done with x-rays and neutrons, performed at the National Synchrotron Light Source, Advanced Photon Source, European Synchrotron Radiation Facility, ISIS (spallation neutron source at Rutherford Appleton Lab, UK), the Intense Pulsed Neutron Source (Argonne National Lab), and the Institut Laue-Langevan, as well as laboratory x-ray instruments.

The first speaker was Cam Hubbard of Oak Ridge National Lab, who spoke on in situ powder diffraction measurements at high temperatures. He discussed a variety of experimental systems, studied both in the High Temperature Materials Lab at ORNL and at the NSLS in which the ability to follow phase transformations at high temperatures, under synthetic conditions, was key to solving practical problems in ceramics and other high performance materials. Richard Harlow (of Harlow, Inc.) followed with discussions of work performed at the APS on Fe metal catalysts that are used by du Pont in commercial scale manufacturing. These catalysts are activated at high temperature and pressure at the beginning of the process batch, and there was inadequate understanding of the chemical basis for the observed lot-to-lot variation of their performance. High energy x-rays were necessary to penetrate the stainless steel tube used to house the catalyst under process conditions, and high angular resolution was required to distinguish the processes of interest in the catalyst. Insights gained from the study of the state of the activated catalyst have given information useful to optimize the process in the chemical plant.

The next two talks addressed an extension of the domain of powder diffraction that is becoming increasingly important, pair distribution analysis. Briefly, this technique transforms the entire diffraction pattern into a radial distribution function. Instead of analyzing only the Bragg peaks to learn the periodically repeating component of the crystal structure, pair distribution analysis reveals the distribution of local environments throughout the sample. Accordingly, it is particularly valuable in materials that are only partially crystalline, such as nanoscale phases. Valeri Petkov of Central Michigan University provided an introduction to the technique, and discussed recent results from studies of nanophase LiMoS₂, Ag_0.4MoS₂, (NH₄)_0.5V₂O₅, magnetic GdAl₂, and Cs intercalated into zeolite. Jonathan Hanson (Brookhaven National Laboratory, Chemistry Department) continued with the theme of radial distribution structural refinements, combined with “conventional” Bragg peak analysis of diffraction patterns in studies of the reduction of (nominal) CuO and CeO₂, with measurements performed in situ at high temperature. This work shows the complementary information available from the two techniques, and the importance of both in unraveling complicated behavior in mixed phase materials with partial occupancy of several crystallographic sites.

After lunch, Bill David (Rutherford Appleton Laboratory, UK) woke the audience up with some startling new comments on a concept taken for granted by most practitioners: least squares analysis. While that would be the correct approach if the data errors obeyed a normal probability distribution governed by counting statistics and the hypothesized model was a correct description of the sample diffraction properties, these conditions are often not met. Starting with a formal description of least-squares analysis, David reviewed principles of experimental design to meet those criteria. He then presented some new results on techniques to deal with problems frequently observed: unknown impurities in a powder diffraction pattern handled with a new minimization criterion, and a maximum likelihood approach to analyze incomplete structures in which some atoms have not been located.

The two following talks covered various perovskite-related materials in which the interplay of structural distortions, charge ordering, and magnetism require complementary application of neutron and x-ray powder diffraction. El’ad Caspi of Argonne National Laboratory discussed the phase diagram of the colossal magnetoresistance system (Ca²⁺_1-xCe⁴⁺_x)MnO₃. This is a two-electron doped system (in contrast to more familiar one-electron doped systems such as (Ca²⁺,Bi³⁺)MnO₃), and the faster change of electronic charge with ion substitution leads to a much more complicated interplay among charge ordering, orbital ordering, and spin ordering, which in turn causes phase separation over a much larger range than 1e systems. Patrick Woodward of Ohio State University opened his talk with several demonstrations that neutrons are often superior to x-rays in ab initio structure solutions of oxides and fluorides, even though the latter are much more widely used. He then discussed several neutron and x-ray experiments: Fe charge disproportionation in CaFeO₃, Mn orbital ordering in NdSrMn₂O₆, and the Verway transition on oxygen deficient double perovskites, RBaFe₂O_5+w (R = Rb, Y, Ho, and Nd).

In the last session, Tom Vogt (Brookhaven National Laboratory, Physics Department) discussed work on the high pressure chemistry of zeolites. The theme of his talk was the surprising discovery of materials that expand under pressure, due to increased incorporation of water into the zeolite cavities. Sodium aluminosilicate natrolite undergoes a reversibly pressure-induced lattice expansion, whereas a synthetic analog, potassium gallosilicate natrolite expands irreversibly, retaining the expanded high pressure phase upon return to ambient pressure. Vogt presented structure determinations showing the role of non-framework metal ions in distinguishing the two cases. Finally, Peter Stephens presented a talk largely prepared by Robert Von Dreele (Los Alamos and Argonne National Labs) on their work applying high resolution x-ray powder diffraction to proteins.

In all, the broad range of powder diffraction, pair distribution function, and single crystal analysis, and the large and growing user community at synchrotron and neutron facilities points towards increasing growth at the frontiers of powder diffraction. This in turn indicates continuing demand for improved instruments as well as improved access to the current generation of operating instruments.

FOR MORE INFORMATION
Peter Stephens
Affiliations: Department of Physics and Astronomy, Stony Brook University
National Synchrotron Light Source, BNL
Department of Physics and Astronomy
SUNY at Stony Brook
Stony Brook, NY 11794-3800
Tel: (631) 632-8156
Fax: (631) 632-8774
Email: pstephens@notes.cc.sunysb.edu