ExtSym (Extinction Symbol) program

Home

Summary: This page is the main access point to the computer program: ExtSym.

This web page is structured as follows:

• Purpose of program
• WARNING
• Required ExtSym input
• Copyright
• Downloads
• Changed to the code
• Errata to Ref. [1]
• History of this document
• References

Purpose of program

Its purpose it to help determine the extinction symbol of a crystalline material from powder diffraction data; ExtSym does not distinguish between space groups which have the same extinction symbol.

It returns a ranked list of the probabilities of the extinction symbols belonging to a crystal system (which may e.g. be the orthorhombic or cubic crystal system). Of the 55 dataset tested in Ref [4] ExtSym ranks the published extinction symbol as the most probable symbol in 52 cases. Hence, the top ranked symbols from a ExtSym output table is very likely to include the correct extinction symbol which your crystalline powder belongs to.

Notice, this program does not index the unit cell. Therefore, before using ExtSym it is assumed the pattern has been successfully indexed by a program such as Dicvol, Ito or Treor.

To browse a collection of ExtSym test results click here.

WARNING

ExtSym was specifically made for the purpose of testing the algorithm presented in Ref. [1] and not much effort was put in to make it user friendly. Therefore, if you adventure into trying it out, be prepared to be patient and perhaps write a bit of code yourself!

Required ExtSym input

Extsym requires the input data to be in the form of Lorentz-polarisation corrected, extracted intensities (for more information see here). Which means that before you use ExtSym you must perform a Pawley or le Bail refinement in a space group that has no systematic absences (for instance, P222 for an orthorhombic indexed crystal, p2 for a monoclinic indexed crystal and so on).

There exist no common format for the outputs returned by Pawley and le Bail programs, and these tend to be different from the input format required by ExtSym (described here). Hence, scripts have been written for making it easier to use ExtSym with various programs. See the Download section below.

Also, the ExtSym program is shipped with the commercial software: CCDC DASH software and HighScore from PANalytical (in the Plus option). These include graphical user interfaces from which ExtSym can be straightforwardly executed. Before using the ExtSym software through another software (or as standalone) it is recommended to read this page.

Finally, another strategy is not to use the ExtSym program as described on these pages and write your own version of the space-group algorithm in Ref. [1]. The company Accelrys (now BIOVIA) has done this and their implementation is part of the Reflex software.

Copyright

Copies of any binary version of ExtSym can be used for any purpose. The code makes heavy use of the Numerical Recipies library [2] (and also makes a couple of calls to functions from [3]). The source code minus the Numerical recipies libraries is now downloadable from here. The main references for ExtSym are Ref. [1] and [4].

Downloads

ExtSym can be downloaded in 'raw format' or bundled together with various support scripts that make it easier to use ExtSym with a number of existing powder decomposition programs; note these scripts come with zero warranty. The source code for these are provided and these may be used as a starting point for writing improved scripts.

• ExtSym bundled together with script for use with PRODD.
• ExtSym bundled together with script for use with Topas.
• ExtSym bundled together with script for use with GSAS.
• ExtSym bundled together with script for use with FullProf.
• To download ExtSym in its raw form, without any helper scripts.

If your favorite powder decomposition program is not listed above, instructions on how to write a support script from scratch can be found here.

Changes to the code

Since the submission of the paper [1] June 2000 the following changes have been made:

• 2/4-2001: Added code to allow ExtSym to accept .hkl data files containing any multiplet.
• 14/2-2002: Added code to allow for .hkl filenames that contain spaces.
• 15/10-2006: Added code that allow an additional powder data file format, that is the .hkl_weight format that is described here. Updated the text printed to the screen to say "ExtSym version 2006".
• 30/7-2007: Simplified the table.asc output to just containing the ranked table.
• 29/7-2008: Added option to allow combined search for trigonal and hexagonal extinction symbol (as noted would be useful by referee to paper [4]).
• 7/8-2008: Replaced 'Laue class' with 'crystal system' in parameter input file.
• 9/8-2008: Spotted and corrected for subtle error in way the systematic absences were handled for some of the hexagonal/trigonal extinction symbols.
• 8/4-2015: Removed repeated orthorhombic extinction symbol label 'P c c n' in table output (one of which should have read 'P c n n') as reported by Thomas Dortmann.

For the most recent version of the source code now see here

Errata to Ref. [1]

Errata to Acta Cryst. A57, 47 (2001):

• In the author list: J. C. Johnson should have been J. C. Johnston (many apologizes for this mistake).
• There is a misprint in Eq. (7). This equation should read Cij / (Cii*Cjj)1/2 not Cij / (Cii*Cjj)-1/2.
• Replace all occurancies of 'Laue class' with 'crystal system'.

See also corrigendum: Markvardsen et al, Acta Cryst. A68, 780 (2012), http://dx.doi.org/10.1107/S0108767312038305.

Updates to this document includes

• The first version of this document 13th Nov. 2006.
• Added ref to follow up ExtSym paper. 24th November 2008.
• Added link to PANalytical Highscore and link to source code. 11th March 2016.

References

1. A. J. Markvardsen, W. I. F. David, J. Johnston and K. Shankland, Acta Cryst. A57, 47 (2001), https://doi.org/10.1107/S0108767300012174; Further a small extension to this paper is presented in the 2004 conference paper https://aip.scitation.org/doi/abs/10.1063/1.1835216: extending the prior probability expression to the case where three or more reflections completely overlap, thereby allowing information from such data points to also be including in the analysis. This is prior probability expression used by ExtSym.
2. Numerical Recipies in C by Press et al., 2nd edition, 1992.
3. Atlas for computing mathematical functions by W. J. Thompson (John Wiley & Sons, Inc. 1997).
4. A. J. Markvardsen, K. Shankland, W. I. F. David, J. Johnston, R. M. Ibberson, M. Tucker, H. Nowell and T. Griffin, J. Appl. Cryst. 41, 1177 (2008), https://doi.org/10.1107/S0021889808031087.