Powder diffractometry and DANSE: powDANSE S.J.L. Billinge Dept. Physics and Astronomy and Center for Fundamental Materials Research Michigan State University.
Outline PowDANSE objectives Scientific frontiers Project plan: PowDANSE objectives revisited
PowDANSE objectives Baseline objectives (basic engineering) Advanced objectives (enabling new science)
PowDANSE objectives Baseline objectives (basic engineering) –On day 1 allow POWGEN3 to give real-time Rietveld’able data to users –On day 1 allow POWGEN3 to give real-time total-scattering S(Q) data (including an extensive overhaul of existing total- scattering/PDF data analysis codes) –On day 1 give the user an intuitive graphical data manipulation toolkit for basic real-time data interrogations –On day 1 give the instrument scientist a powerful suite of data interrogation tools for painless user support –Work towards day 1 support of future SNS powder diffractometers (disordered materials diffractometer etc.) –As a legacy, support all the world’s powder diffractometers with the same data analysis capabilities within the same software framework
PowDANSE objectives Advanced Objectives (enabling new science) –Prototype and test new codes that radically depart from existing data analysis methods –Optimize the use of precious neutron beam-time by simulating experiments before and during data-collection as well as after –Investigate ways in which access to TeraGrid type cyberinfrastructure will allow qualitatively new scientific problems to be addressed with powder diffraction –Extend scope of the codes to single-crystal diffraction and single-crystal total-scattering studies
Powder diffraction: brief tour Basic work-horse of materials science/ chemistry/ physics (from full-profile fitting methods) –Accurate atomic positions –Sample quality characterization (usually from in-house x-ray) –Sample phase analysis –Phase diagram determination –Higher level information: strain distributions, inhomogeneous strains, textures in polycrystals –Powder diffractometers are among the most heavily oversubscribed instruments at facilities despite high throughput –Large and diverse user base
Powder diffraction: brief tour (not strictly powders but…) Structural analysis of glasses and liquids –Progress being made on enormously complex but fundamental issues such as solvation structures and hydrophobic interactions –Higher level information now wanted from glasses such as strain distributions New horizons: structure solution from powder data New horizons: nanocrystallography –Solving structures from materials with nano-scale structural motifs Over the horizon: “Cooperative refinement” and “Total determination”
powDANSE main contributors Main contributors –me –Jason Hodges (powgen3) –Thomas Proffen (npdf) –Jim Richardson (gppd) –Chris Benmore (glad) –SNS disordered materials diffractometer IS Tightly coupled consultative roles (?) –Brian Toby (nist) –Angus Wilkinson (g tech) –Paolo Radaelli (isis) –Alan Soper/Spencer Howells/Robert McGreevy (isis) –… More loosely coupled consultative roles (?) –Takeshi Egami (jins) –Jim Jorgensen (anl) –Single crystal people: Si Moss/Lee Robertson/Ray Osborn/Richard Wellberry –…
Science Some Physics
Local vs. long-range structure: semiconductor alloy In 1-x Ga x As In 0.17 In 0.33 In 0.50 In 0.83 Average arsenic atomic probability distribution at different indium concentrations [Petkov et al., PRL. 83, 4089 (1999); Jeong et al., PRB (2001)] Behaves like: local structure average structure r = nm
What does the polaron look like? S.J.L. Billinge et al, Phys. Rev. Lett. 77, 715 (1996); S. J. L. Billinge, et al., Phys. Rev. B 62, 1203 (2000)] Mn 4+ Mn 3+ Mn 4+
Charge-stripes in correlated-electron oxides Long Cu-O bonds Short Cu-O bonds Strain will build up here Qualitatively we see that lattice strain will tend to break up the stripes into short segments!
Evidence for Charge inhomogeneities: La 2-x A x CuO 4 (A=Sr,Ba) In-plane Cu-O PDF peak width broadens with doping (then sharpens) Bozin et al. Cond-mat/ UnderdopedOverdoped Fermi-liquid like Inhomogeneous Charges
Affect of misfit strain on stripe microstructure Collaboration with Phil Duxbury at MSU Model: lattice gas with strain Results in stripes at various doping Strain terms
Science Some Chemistry
Diffuse scattering: Underneath the Bragg- peaks
Nanocrystallography: Beyond Crystallography Crystallography fails in nanocrystalline materials: Nanocrystalline V 2 O 5.nH 2 O xerogel Crystalline V 2 O 5
Structure of xerogel Xerogel has bilayers of edge-shared VO 6 octahedra separated by water molecules Notice loss in peak amplitude above 11.5 Å => turbostratic disorder
Crystals and nanocrystals In crystals, the oscillation amplitude in G(r) is independent of r In Nanocrystals, the amplitude falls off with increasing r Thanks to Valentin Levashov and MFT for the plot.
“Nanostructure” in the xerogel Turbostratic disorder seen in the PDF consistent with bent and tangle fibres V. Petkov, et. al., J. Am. Chem. Soc. 121, (2002).
Atomic order in disordered carbon Pyrolize Polyfurfuryl alcohol at high temperature in an inert atmosphere The resulting carbon is nanoporous and disordered. PDF reveals atomic order evolving with process T High temperature processing Low temperature processing V. Petkov et al., Philos. Mag. B 79, 1519 (1999).
Total scattering then and now
Alumino-silicates (Si,Al)O 4 tetrahedral networks Important catalysts: zeolites, microporous materials Cannot study AlO 4 and SiO 4 separately (Si and Al have similar x-ray and neutron scattering lengths) R Si = 1.61A, R Al = 1.75A, R = 0.14A x-ray data from Advanced Photon Source V. Petkov et al., Phys. Rev. Lett., 85, 3436 (2000).
Chemical specificity Using anomalous scattering we can get a chemical specific PDF - in this case it is the In-DDF High resolution total-PDF is compared with (low resolution) chemically resolved differential Both data-sets can be co-refined using PDFFIT This is the PDF equivalent of XAFS but higher neighbor information is present Petkov et al. J. Appl. Phys. 88, 665 (2000). Correlations from: Indium - all all-all
Structure of intercalants: inorganic electride Zeolite ITQ-4 has 1D channels of ~7Å diameter Cs is intercalated X-ray data from NSLS- X7A Cs forms Cs + in zig-zag pattern Electrons are counter-ions
Experimental data: APS 1-ID, 80 keV Model data for a rigid molecule C C-C Fe-C ring-ring Ferrocene Fe(C 5 H 5 ) 2 Fe
Rapid Acquisition PDFs Fast neutron PDFs (powGEN3) Fast x-ray PDFs –Four orders of magnitude decrease in data collection time! –Nickel data, 1s collection time, Q max 28 Å -1
RAPDF Low-Z materials possible: AlF 3 Good reproducibility BiVO
Summary Frantic overview of current scientific questions in the Billinge-group as an unrepresentative taste of what can be done Strawman proposal for powDANSE objectives
Acknowledgements Valeri Petkov (former post-doc, now at CMU) Xiangyun Qiu (MSU student) Thomas Proffen (former post-doc, now staff at LANL0 Il-Kyoung Jeong (former MSU student now postdoc at LANL) Emil Bozin (post-doc and former student) Group of Mercouri Kanatzidis Group of Jim Dye Pete Chupas and group of Clare Grey Other Billinge group members involved –Matthias Gutmann –Pete Peterson Facilities –IPNS, MLNSC, ISIS and people therein Funding: NSF-DMR , CHE , DOE-DE- FG02-97ER45651
Obtaining the PDF Raw data Structure function PDF
Observing Domains in the PDF r 1 << r 2 ~ r1r1 r2r2 Intra-domain structureInter-domain structure
What is the PDF? Sit on an atom and look at your neighborhood G(r) gives the probability of finding a neighbor at a distance r PDF is experimentally accessible PDF gives instantaneous structure.
Crystallographic bond- lengths: Caveat Emptor Silica quartz at the to transition at 846K Crystallographic bond lengths shorten Real bonds (obtained directly from PDF) lengthen modestly The explanation Work by Dave Keen and Martin Dove