Setting up the Rietica-LHPM input file

Select under "Phases" that you wish to do an extraction and unless you already have the synchrotron diffractometer defined as an "instrument", stick to the defaults and we will change these in the input file later.

Select Model, General and setup the general information as shown in the following screen dump. Number of cycles, automatic Marquadt damping, data file format, we wish to obtain a EXTRA/Sirpow97 HKL file, and that we are using Synchrotron data.

This would be a good time to save and name the file. Select File, Save and browse to the directory that holds the Powder Diffraction data for cimetidine that you are using.

Select Model, Phases and input the basic information. That we want to do a Le Bail Extraction, Spacegroup and cell constants. the cell constants and two-theta offset should be from your unit-cell refinement so the Rietveld program should be right on the spot for starting the extraction. Le Bail fitting with an inaccurate cell can put you in a world of pain as the Le Bail fitting via the Rietveld software goes beserk.

Select Model, Histograms and input the basic information. Setting the start, step and stop to zero tells LHPM to use the full pattern for the extraction. A Chebyshev function should be OK for fitting the background. Set the background function to refine.

Select Model, Sample and input the sample information. FJC Pseudo-Voight can be a flexible function to use that should work. As this is synchrotron data, set a thin width. Normally it would be good to fit a standard such as NIST/NBS Silicon or NIST/NBS Cerium Oxide to get good starting width and shape values. But if we don't have this information, starting off with a constant width should be OK.

Now start the refinement by selecting Start then Step. Don't be intimidated if the fit is very bad to start with. This is due to the initial eqipartition of starting intensities. Bad initial fits can also be due to problems width and shape. This will improve as we refine these options.

(be careful to watch for weirdess in the refinement. You will notice that in the following Rietveld plot, not all the HKLs are matching up with observered lines. This is because an error was made in the selection of the spacegroup (P 21/N instead of P 21/A). These errors are easy to make so always keep an eye out for errors of this sort. The screen dumps after this were made with the correct spacegroup.

Following is the fit after unitcell, 2-theta offset, V, W and Gamma 1 shape have been released.

Completed fit, all width and shape parameters, and one assymetry parameter.

Completed Le Bail fit

Using File, Export, you can create a starting Sirpow/EXPO file ready for structure solution. To use the following EXPO file created by LHPM-Rietica, all you have to do is name the FOU file, and put in the cell contents.

%window
%struct rietica
%init
%job Cimetidine                                                                  
%data
wavelength  1.52904
cell   10.3950   18.8207    6.8254   90.0000  106.4380   90.0000
space P 21/A              
ref2 xxxxxxx.fou - enter LeBail filename here
content  XX XX XX XX - enter expected cell contents here.
%normal
%continue