Re: Paper on Genetic Algorithms for Powder Diffraction

Armel Le Bail ( armel@fluo.univ-lemans.fr )
Tue, 27 Jan 1998 22:56:46 +0100

Lachlan wrote :

>Though I would also be
>interested in references related to genetic
>programming and genetic algorithms to
>Powder Diffraction and Crystallography.

Here is one :- Crystal structure determination from powder
diffraction data by the application of a genetic algorithm
Shankland K. David WIF. Csoka T.
Zeitschrift fur Kristallographie. 212 (1997) 550-552.

I have tested yesterday once more the "problematic step of
extracting the intensities of individual reflections from
the powder data" : no problem ;-). The structure was then
revealed by SHELXS86 or 97 with all the 12 P + O atoms in
general position. Fourier difference map allowed to locate
Li atoms. Nothing exotic here. Monte Carlo attempts by
FULLPROF failed (maybe by lack of expertize)...

Are genetic algorithms frequently used to solve small molecule
structures from single crystal data ? If not, I am afraid that
the same will occur when using powder data. I don't think that
practices in structure determination from powder or single
crystal data will ever completely diverge (very special cases
excluded).

Armel