Indexing (ReX.Cell)

This tutorial provides a step-by-step guide on how to perform powder diffraction indexing with ReX.Cell. The example is pretty simple, but extension to more complicated cases should be straightforward. To begin, start the ReX.Cell program by double clicking on the program icon (see legacy installation instructions for troubleshooting). The main application window will show several “views”: on the top left, the “Dataset” view reports the list of the loaded diffraction patterns; below, the “Peaks List” view displays the diffraction peaks found in the active diffraction pattern; on the right, the currently selected pattern is displayed in a xy plot; finally, the “Solutions” view will contain the list of the solutions found after the indexing run.



Now, download the ReX examples archive and unzip it in a folder of your choice. Click on the “Load data” button  located on top of the Dataset view, then select the 88.dat file located in the “indexing” subfolder (the data corresponds to Armel le Bail’s Powbase repository entry number 88); after a confirmation dialog, the diffraction pattern should appear in the Plot view:



Before performing the peak search, we are going to restrict the data interval to a smaller region. To do this, select the diffraction pattern (88.dat) in the Dataset view and double click on it (or alternatively click on the “Edit…” button). In the “Edit pattern” dialog, set 10° and 40° for the x min and x max used values, respectively; after closing the window, the pattern plot is updated according to the new interval.



Now let’s start the peak search wizard. Make sure the diffraction pattern is still selected in the Dataset view and click on the “Peak search…” button , which starts the wizard.



The first page of the wizard simply asks which diffraction pattern to work on; just click on the “Next” button, which brings us to the Data processing page. Here, we have the chance to apply basic pre-processing operation to the diffraction data, namely background subtraction and radiation stripping; in this case, only background subtraction is needed, which is automatically applied when checking the “Subtract background from pattern” checkbox. The background subtracted pattern is displayed in blue in the plot window; changing the type and parameters of background model (combobox and “Edit…” button nearby) changes the pattern accordingly.



After subtracting the background, click the “Next” button to proceed to the peak search page. The peak finding is based on a two-step procedure: first the pattern is processed through a Savitzky-Golay method, then peak maxima are found using a second derivative search. Both steps can be configured in the graphical user interface, in the “Data smoothing” and the “Peak detection” sub-panels, respectively.



Initially, the pattern plot shows the peak positions automatically found with the standard option values applied:



In our case, only the “Average FWHM” value needs to be slightly increased to obtain a satisfying results (e.g. 0.07 or 0.08). You may want to experiment a little bit with the option values and see in real time how this affect the peaks list; notice that you can zoom and pan the plot at any time to observe a particular region of the pattern. After you are reasonably satisfied with the results, click “Finish” to end the wizard. The Peaks List view shows now the found diffraction peaks along with their relative intensities and positions (in both 2-theta and d-spacing).

At this point, we can fine-tune the peaks list by deleting wrongly positioned peaks or adding new ones. To delete an unwanted peak, simply select it in the peaks list view (it will turn blue in the plot window) then click on the “Delete”  button. To add a new peak, position the cursor in the plot window at the wanted position and double click; a new peak will appear in both the plot view and the peaks list. Alternatively, you can click on the “Add new peak” button  and manually enter the peak data in the dialog that shows up.



At this point we can proceed with the indexing. In the Peaks List view, locate the “Run indexing” button  and press it; a new window will appear, showing the indexing options. In the combo box, make sure “Dicvol 06” is selected; the press the “Options…” button. In the Options panel, which opens add the “Monoclinic” symmetry by selecting the corresponding checkbox:



Close the options panel and click on the Run “button”; this starts the Dicvol indexing program, which completes pretty quickly (unless the peak selection was incorrectly carried out, and one or more “impurity” peaks are present). If the indexing is successful, one or more entries should appear in the “Solutions” view (below the Plot view); in our case, the first (and probably only) solution is the correct one.
It is possible to compare the indexed solutions with the original peaks list simply by clicking on the solution entries: the corresponding calculated peaks list appear in blue below the experimental peaks list, displayed in green. It is also possible to directly examine the program output by clicking on the “Output” button.



At this point, you may choose to perform the indexing again with different options; manual edit of the input file is also supported. Please notice however, that any change in the graphical option panel after the manual edit override any previous manual input.
You can finally export the selected solution by clicking the “Export solution…” button  ; support for different structural format is provided, among which the *.cif format is probably the most convenient.

This concludes the indexing tutorial; you may want to experiment again with this example by choosing different options in the peak search wizard and others indexing programs, and see how these choices affects the indexing results.