Indexing

Files needed: none (or results from peak fitting)

Learning outcomes: How to index a powder pattern in topas. See Alan's j. appl. cryst. paper for details of the algorithms used.

Topas advantages: relatively robust to impurity peaks and (in my experience) has indexed everything that other routines have been able to index

1. Under the "Indexing" menu click on "Create Indexing File".

2. Paste 2-theta and peak intensity values from the TiO2 peak fitting tutorial into the section of the file labelled (i.e. below where it says "paste here"!). If you haven't done the peak fitting tutorial then either use the values below or read the data file into topas and use the "View/Search Peaks" facility.

27.4933274 1483.03892
36.1417160 600.374011
39.2491072 87.1466545
41.2995694 293.361291
44.0974357 111.974602
54.3775231 852.552752
56.6779687 263.740365
62.8179591 114.115219
64.0981291 125.848734
65.6797591 8.31017226
69.0535316 311.102223
69.8553749 142.043952
72.4689461 16.9511527
76.5935592 31.9086674
79.8803068 17.1994137
82.3774808 68.5855316
84.2921991 48.4586528
87.5124851 17.9694065
89.5947678 115.718833

3. Save the file as e.g. d5_05005_index_01.inp. Click the "Save/send to topas" icon. Launch topas and run the file.

4. Topas will then try and index in the crystal systems specified in the input file. It will produce a graph showing the best solutions. The whole run will take ~ 1 minute.

5. Openthe file d5_05005_index_01.ndx which contains the results in jedit. Topas will have found a huge number of possible cells with very high figures of merit (>100). It's possible to view these various results graphically in topas. See the tutorial linked here on PbSO4. For this example it's clear that best results are pseudo-tetragonal. Try running the input file again with all but cubic and tetragonal sytems excluded. i.e. the top of the input file should look like:

Bravais_Cubic_sgs
'Bravais_Trigonal_Hexagonal_sgs
Bravais_Tetragonal_sgs
'Bravais_Orthorhombic_sgs
'Bravais_Monoclinic_sgs
'Bravais_Triclinic_sgs

6. The best cell should now be:

P42nm a=4.5946 c=2.9587

7. Extra work: Try a Pawley fit to the data using this cell and space group (instructions here); try indexing any of the other data files from the tutorial sets.

8. Beware. Topas gives a huge number of suggestions. Many of them are equivalent cells. You could use a package such as delred to investigate this. For complex systems lots of cells may "almost" fit the data. Only one of them would allow you to solve/refine the structure correctly. Indexing is one of the hardest stages of structure solution.

9. Extra work: if you're interested in indexing then try the same sort of protocol on any of the other data sets in the school.

 

[Modified 01-Oct-2017 by John S.O. Evans. Pages checked for Google Chrome.]