Single Crystal Refinement

Files needed: glycine_cr.cif; glycine.hkl

Learning outcomes: How to do a simple single crystal refinement in topas starting from a shelx hkl file and a .cif starting model. How topas displays single crystal data as a "pseudo powder pattern".

Warning: when trying to find/refine H atom position it's best to use the same scattering factors used by e.g. shelx/crystals. In your topas directory click on "use_9f0.bat". To switch back to the default set click on "use_11f0.bat".

1. Save the files listed above in your directory.

2. Work through the menus under "Single Crystal Refinement" to set up the input file.

3. Click on "Select hkl Data File" and navigate to find glycine.hkl.

4. Click on "Instrument/Wavelength" and select "Molybdenum Radiation".

5. Click on "Structure - cif" and navigate to find glycine_cr.cif.

6. Click on "Save/send to topas" icon. Launch topas if not already running. Run in topas.

7. You should get Rwp of around 32%.

8. Examine the quality of the fit in topas. You should see that low angle reflections are very poor. You might find the topas representation easier to visualise if you click on the filename (02srv022merged.xy) in the topas grid window then click on the "display" tab. Change the data point size to 5. Each point represents a single hkl intensity.

9. To improve the fit try refining temperature factors on each atomic site. At the end of each non H atom type the word "adps". For hydrogens try typing "beq @ 1" to refine isotropically. The atoms lines should now look like:

site O1 x 0.65421(17) y 0.35764(7) z 0.39353(16) occ O 1.0000 adps
site O2 x 0.19789(18) y 0.40630(7) z 0.26383(16) occ O 1.0000 adps
site N3 x 0.2040(2) y 0.41119(9) z 0.75949(19) occ N 1.0000 adps
site C4 x 0.4299(2) y 0.37485(9) z 0.4341(2) occ C 1.0000 adps
site C5 x 0.4416(2) y 0.3543(1) z 0.7140(2) occ C 1.0000 adps
site H31 x 0.208(3) y 0.4862(15) z 0.730(3) occ H 1.0000 beq bvalue 1
site H32 x 0.030(3) y 0.3864(13) z 0.647(3) occ H 1.0000 beq bvalue 1
site H33 x 0.208(4) y 0.4041(14) z 0.927(3) occ H 1.0000 beq bvalue 1
site H51 x 0.620(3) y 0.3807(12) z 0.844(3) occ H 1.0000 beq bvalue 1
site H52 x 0.429(3) y 0.2732(13) z 0.739(3) occ H 1.0000 beq bvalue 1


10. Run this in topas. You should get an Rwp of ~ 6.25%. In the input file the "adps" words will have been replaced by uij's according to the site symmetry of the atom.

11. You might want to look at some of the items in the "Single crystal specific" jedit menu.

12. (Advanced) You might want to compare these refinements directly with e.g. crystals/shelx. To do this you should tell topas to use the same scattering factors as crystals/shelx. In the main topas folder click on the command file "use_9f0.bat". These scattering factors will be used next time you launch topas. Remember to switch back to the default with "use_11f0.bat" afterwards. You should find that with these scattering factors all H's can be refined isotropically independently.

13. Try refining coordinates by putting "@" symbols in front of each coordinate. The Rwp should come down to ~6.16%; coordinates shouldn't shift much (the .cif file comes from an equivalent crystals refinement).

14. In most cases H atoms won't have been found in the structure solution process you used (e.g. charge flipping or direct methods). Try deleting the H atoms from this refinement (e.g. put a ' at the start of the site line) and finding them again by a difference Fourier map. You'll need the lines below in your inp file:

fourier_map 1
fourier_map_formula = Fobs - Fcalc;
load f_atom_type f_atom_quantity
H = 20;


at the end of the refinement a structure viewing window should appear showing the difference map and any H atoms found. There should be an NH3 and a CH2 group present.

[Modified 31-Jul-2017 by John S.O. Evans. Pages checked for Google Chrome.]