GSAS-II provides two forms of CIF files, which could be characterized as “quick-and-dirty” and “comprehensive.” The former type of CIFs are intended for communication of information between software programs and are used as would be any other exporter. One selects the exporter, selects a phase or histogram and the file is written. The process of creating a comprehensive CIF report from a GSAS-II project can be a much more complex task. The reason for this is that there is considerable information that the IUCr recommends for inclusion in a deposited CIF that you would need to supply. The CIF exporter will help you provide this information. Importantly, this information is placed into template files so that should the refinement be updated and the CIF then needs to be regenerated, the templates are reused and no additional work is needed. Since some of the information that is placed in the template files is likely to be useful for other projects, these files can be reused.
When a Rietveld refinement contains a single histogram and a single phase, the entire CIF to document this can be a single CIF block. However, when more than one histogram or phase is used, then the CIF must have multiple blocks. When this is the case, GSAS-II will write a CIF block for each phase and a block for each histogram and finally a summary block that provides links between all the CIF blocks. Each CIF block is identified with a block tag. This block tag is intended to be completely unique and contains a timestamp, the CIF author name and the instrument name as well as the name of the project. If you include the institution name with the instrument, as well as use your name, it becomes quite unlikely that anyone else will duplicate the block tags.
The process of writing a comprehensive CIF is outlined here. When the menu command for a project-based CIF export is first entered, you will first be asked to supply a name for the file to be written.
Every powder histogram must have an instrument name supplied, since this is used to make the block tags unique. (This name can be edited on the histogram’s “Sample Parameters” data tree entry.) If any are blank, the window shown in Fig. 22.1 is seen.
You will then be asked to supply atomic radii for each phase, as seen in Fig. 22.2. These are used to search for bond distances and angles that will be reported; the defaults for these values will probably be fine, but should you find that important distances are not being computed, it will be simple enough to include change these values later so that more (or fewer) values are generated. These radii values are saved in the project, so you will not be presented with these window(s) again. Once this has been completed, the main window for CIF generation is shown as in as seen in Fig. 22.3.
Setting the CIF author name is strongly recommended, since this is also used to make the block tags unique. For a multi-histogram export, the “Edit Instrument Name(s)” button will open the window seen in Fig. 22.4. Note that the “down arrow” button allows a value to be copied to all the entries below, so the instrument name only needs to be entered once. In a sequential refinement, the instrument names must be the same, and only the first value is shown.
In the main CIF creation window (as seen in Fig. 22.3), there are buttons in each section that reference a CIF template. There are three types of templates, for the overall project, for phases and for histograms, where the fields in the template are taken from IUCr guidelines for reporting powder diffraction results (https://journals.iucr.org/services/cif/powder.html). An example of the window that appears when a template is edited is shown in Fig. 22.5. As will be clear from looking at these items, they describe the instrument used for data collection. This is likely to be the same for every CIF generated with this instrument, so considerable time will be saved by reusing this content once the template is prepared. Save the template file before pressing the “Use” button and the next time that a CIF is prepared, you can read the saved template file in place of the default. Note two features of this template window. There is a “help button” (labeled with a question mark; on Windows this appears in yellow) for every CIF data name that has a definition in the CIF dictionaries that are provided as part of the GSAS-II installation. Pressing the help button causes the definition to be shown, as seen in Fig. 22.6. Also, the data entry boxes in this window can be resized by dragging the right-angle symbol in the lower right of the data entry box. Increasing this size is helpful when lengthy values will be entered. This is also shown in Fig. 22.6.
Associated with each phase is a table of bond distances and angles; each value in that table has a “publish” flag. IUCr Journals will generate tables from this information, but will only include the values that have been flagged with the “publish” flag set as true (which is the default). The “Set distance/angle publication flags” window, where an example is shown as 22.7. This window serves two purposes. The publication flags can be set here, though I think this is commonly not done anymore. More significantly, you can see if the radii are set in the right range to pick up the desired distances and angles, but not non-bonding contacts, unless they are of interest. In the example shown here, the desired bond distances were found, but the angle radii needed to be lengthened so that the bond angles were included.
When the CIF has been customized by entering information into the templates, atomic radii and distance/angle flags, press the “Create CIF” button and the CIF will be created. Note that all this customization information is saved in the GSAS-II project (.gpx) file. This means that if the project export CIF menu command is used again, the revised CIF will retain all of the information you entered, but the cell dimensions, coordinates, distances, etc. will all be updated. Thus, all the work invested in creating a well-documented CIF will not be wasted should the refinement be updated.
