SDPD Internet Course

Week 4

Advanced methods for extracting structure factors, part 1. 


Read deeply chapter 3.2, including 3.2.1 and 3.2.2 of the SDPD tutorial (CD) or access to the Kunming Workshop (CD) conference text (beginning of part 3, same chapters). Read the conference "The Practice of "|Fobs|" Extraction from Powder Diffraction Data" (CD). Try to reproduce the structure factor extractions which are inside the SDPD tutorial (CD: Na2C2O4, [Pd(NH3)4]Cr2O7, t-AlF3, beta-BaAlF5 and cimetidine C10H16N6S) before to undertake the exercice given at the bottom of this page.

You may also consider to look at the book chapters : 

B3 : Chapter 14 and 15, pages 254-275 and 276-292. Chapter 14 is devoted to powder pattern decomposition methods, either by individual profile fitting or by the Pawley method. Chapter 15 is devoted to ab initio structure solution and cites some works where the Le Bail method is used but the chapter is not clear on that point. It should be realized that the book was published in 1993 but was mainly the result of a 1989 workshop. 

Much more is inside :

B7- Structure Determination from Powder Diffraction Data 
W.I.F. David, K. Shankland, L.B. McCusker, and Ch Baerlocher, 
IUCr Monographs on Crystallography, Vol 13, 
Oxford Science Publications, 2002. 

In chapter 8 (pages 136-161), David and Sivia compare the Pawley and Le Bail methods. The conclusion is : do not use a negative background with the Le Bail method ;-). Those methods are cited and commented in several other chapters of the book.

If you want even to learn more about the Le Bail method, more is available (CD).

Software to download

In a sense, you already extracted inaccurate intensities in precedent weeks, because intensities are a by-product of peak position estimation software. But extraction of accurate intensities is the subject of this session, using specific computer programs. Transforming those intensities in structure factors needs the cell knowledge and the attribution of hkl Miller indices to each extracted amplitude, and thus, cell-constraint is applied. The list of softwares used in powder structure determination for this purpose is quite long, but all programs are not available.

Le Bail method (LB) is more accessible than Pawley method (P), embedded in commercial programs (DASH for instance), however the Cambridge Fortran ressources (CCSL) may provide the Pawley program Fortran source. Many programs below are Rietveld programs including the Le Bail method algorithm which consists in iterating the Rietveld decomposition formula, starting from a set of arbitrarily equal "|F|" instead of "|Fcalc|" obtained from the starting structure model. The full list of programs used for this purpose of pattern decomposition and which were used in SDPD, is online (CD) in the SDPD-Database. Unfortunately, few of them are available directly on Internet. If you should download only one to three of the programs below, they could be those listed first, they will certainly do the job :

Le Bail method

Fullprof by Juan Rodriguez Carvajal (CD : inside of already installed) (Le Bail method, renamed "pattern matching" or "profile matching", in fact, you will not find any reference to the algorithm origin in the manual), Rietveld program built on a DBW basis (archetype of modern Rietveld method software). (CD: see tutorials).

Gsas by R. von Dreele (CD: gsaskit.exe) (Rietveld program including the Le Bail method). See a tutorial (CD) about structure factors extraction by using Gsas at the CCP14 Web site (CD).

WinMprof by Alain Jouanneaux (CD: instalWMP.exe) (Rietveld program including the Le Bail method) is a Windows and updated version of MPROF (CD).

Expo that you should already have downloaded last week. Expo is the last avatar of a package including Extra and Sirpow (CD).

Extrac by Ch. Baerlocher, stand alone program using the Le Bail algorithm, included in the XRS-82 suite.

LHPM/Rietica another DBW avatar including the Le Bail method (CD).

XND by J. F. Berar (no reference too in that program about the Le Bail algorithm in the manual, though the program uses it) (CD).

Arit by Armel Le Bail (CD) (open Fortran source, but quite obsolete, not recommended, excepted if you want to look at the code - Rietveld program and first code including the Le Bail algorithm).

You may also have a look at the CCP14 page on the Le Bail method subject. (CD)

Pawley method

The ISIS Pawley program versions (CAILS, ...(CD)) is included into the CCSL Cambridge Crystallography Subroutine Library (very interesting compendium of legacy Fortran codes). We have also an old Fortran source code of the Pawley program (CD) that compiled on VAX computers and of which you may be interested in, but no executable code for PC (ask me for the manual, existing only in paper form). The ALLHKL code is also available at the CCP14 Web site (CD). In this family, you will also find :

Simpro by Harald Ritter

WPPF by Hideo Toraya (CD).

Other useful software

Other programs which may be useful here are those which can display a powder pattern and which will allow to estimate interactively a background. This is in fact generally the best way to proceed first, since trying to refine the background too soon may lead to instabilities when using those pattern decomposition methods. One program which may allow you to do that is :

Dmplot is a shareware by H. Marciniak (CD: (can also display results from Rietveld refinements, and so can display the results of Le Bail method structure factors extraction when dealing with Fullprof or other software). Care that you may need the special version extended to more points than 9600 ( on the DMPLOT FTP site).

You may also try to estimate a space group automatically, but you will have to verify manually the result :

Absen by P. McArdle (part of Oscail).

Equiv by M. Prencipe.


Verify the indexing proposed two weeks ago for the Na5Cr3F14 compound (probable formula) by performing a Le Bail or Pawley method (like you want or like you can) full profile decomposition with cell constraint. You may estimate the background by using WinPLOTR or DMPLOT (for instance) directly on the na5.rit datafile (14001 points). Preliminary patterns have shown that there was no reflection before 10°(2-theta). Use a space group without any extinction at the beginning. Then look accurately at the possible extinctions and try to conclude to a space group for this compound. Finally, perform a Pawley or Le Bail fit with this selected space group (or tell if several space groups are possible). As results, send by email the list of the first 50 extracted "|Fobs|", together with the summary of the fit (the .sum file if you used Fullprof, and the beginning of the .fou file).

Software selected for the correction

WinPlotr or DMPLOT for background estimation; Fullprof for extracting "|Fobs|".

The next week (or when you will be ready) will be reserved for
Advanced methods for extracting structure factors, part 2.
Good luck !