Crystal
structures are solved by EXPO2004 with default settings. In case of failure
the user can modify the default pathway in order to improve the success rate.
The number of options (some of them supported by graphic interface) is quite
large. The optional strategies concern the improvement of both the decomposition
process (for carrying out more reliable reflection intensities and Direct Methods
performances) and the structure model. We give some of them.
About the first target:
The use of the 'RANDOM' directive in the EXTRACTION module activates the procedure for overcoming the Le Bail method tendency to equipartition the intensity of a group of strongly overlapped reflections. Based on a Monte Carlo approach, it may provide a set of more statistically meaningful intensities.
If the default run has revealed pseudosymmetry effects and, in particular, if the percentage of electron density affected by pseudotranslation is large, it may be very useful to exploit that information in the extraction process.
In case of heavy atom structure, the use of prior information obtainable from the Patterson map can improve the quality of extraction (directive 'INVERSION' in the PATTERSON module).
In some cases, the use of prior information from the structure invariant estimates can optimize the extracted intensities (directive 'ESTIMATE' in the INVARIANT module).
If a partial structure fragment has been located by the default run, it may be recycled in the extraction process (a graphic option must be selected).
About the optimisation of Direct Methods procedure:
The CFOM figure of merit is in practice an estimate of the probability that a trial solution is correct. If the highest CFOM does not correspond to the correct solution, subsequent CFOM's may be successful. This is particularly true when several nearly equivalent CFOM's are available for different trials. The Fourier map corresponding to sets different from the first one must be explored (directive 'SET' followed by the set number in the FOURIER module).
Magic integers procedure
is an effective and simple way for starting a multisolution process.
A random approach may be alternatively chosen in the phasing process (directive
'RANDOM' in the PHASE module).
Second representation formula for triplet invariants proved much more powerful than first representation formula therefore it is used as a default for estimating triplets. In case of failure, the user can choose P3 formula (directive 'COCHRAN' in the INVARIANT module) and/or conveniently increase the number of symbols in the phasing step (directive 'SYMBOLS' followed by the number of symbols to be used in the 'PHASE' module).
A check for possible pseudotranslational symmetry is always made by EXPO: if PSEUDOTRANSLATION keyword is not activated the information is given but not used in the normalization routine and in the phasing process. The user can decide to use it (directive 'PSEUDOTRANSLATION' in the NORMAL module).
High (or low) resolution reflections may occasionally play a too important role in the first steps of convergence and divergence process. Fixing a thermal factor lower (or larger) than that automatically provided by the normalization routine may successfully change convergence and divergence process (directive 'BFACTOR' followed by the isotropic thermal factor in the NORMAL module).
An alternative space
group should be carefully considered (directive 'SPACEGROUP' in the DATA
module).
About the optimization
of the structure model provided by Direct Methods:
Sometimes the structure model is partial. Successive combined Fourier map calculations (Fo-Fc, 2Fo-Fc or Fo) and Rietveld refinements may be conveniently selected by graphic options.
The peak labelling of the map may be not correct. In case of heavy atom structures, if the heavy atom coordination is known, the new labelling procedure is advisable (command %CHANGELABELLING (see Example section) or by graphically selecting DIRECT --> Peak labelling).
Especially for the heavy
atom structures, the location of the light atoms is not straightforward.
The POLPO procedure (command %POLYHEDRA (see Example section) or by graphically
selecting DIRECT --> POLPO) is efficient for adding the missing atoms. In
particular, it may be combined with the peak labelling procedure. POLPO1
if all the heavy atoms have been located and POLPO2 if only some heavy atoms
have been recognized.
The EDM procedure (directive 'EDMO' in FOURIER module) can be an effective alternative to the Fourier recycling approach when low value of the ratio 'number of independent observations/number of parameters to refine' occurs.