Hi Dr Ian,

Well Topas dont have these facility yet. I spoted the issue when i saw all Np is 8, at Origin1 all coordinate in x,y,z (pic 1).

"I notice that is different to the Giuseppetti et al. files I was trying - OR have you edited it??" I just edited x and y. O ya I did for Thernadite as well.

The Beq values are different and have greater precision. What is the source for that cif? It's from from ICSD (pic2). I attached the original file saved from HS+.

"Should this occur again I need to know how to spot the problem;"  The easiest way to see the Np, is it the same as the .CIF.

"why is a shift of precisely -1/4, -1/4, 0 selected? Is this done manually" It's from the origin choice (pic3).

O so sorry I refined by Topas5. Here I attached for version 4.2. The quantification is not the same as v5, because i refined all microstructure, Beq, Occ and PO.

Unfortunatelly I lackof ur instrument configuration about the OD or PSD and most of the slits so I simply use the common configuration.

I hope it help.

When using cif files on TOPAS you must substitute some atoms coordinates by a function, otherwise it may cause imprecise results. As an example "0.666" by "=2/3". Should it be the problem?

Hello William,

yes I did all that. There are no fractions of 1/3, 2/3, there are some at 1/4 and some 3/4

As in several other space groups, there are two origin settings for Pmnm. I do not have TOPAS now, so I cannot check which one it uses. GSAS offers to add a shift vector (-0.25; 0; 0.25) to all coordinates if the intensities are wrong. The attached plot 1 is before shift and 2 is after shift.

It appears that the two y values of O3 and O5 have been manipulated, as commented in the supplied link.

I did a quick comparison with the structures in the ICSD. (sorry I cannot post the data here)

It turned out that a minus was wrongly added to the y values of O3 and O5.

If you delete the minus the structure looks sound.

Regards

Dominique

Alexander, I cannot find any origin shifting feature so for in TOPAS. However it is not only the intensities but also the position that are wrong. Please see examples below.

Dominique, thank you for your suggestions. I tried removing the -ve values but no difference. The cifs you posted are the same as I am using.

I have posted a series of images here to show you my problem.

1 - a qualitative scan in EVA to show the presence of burkeite in the assemblage matching PDF-2 024-1134. Note positions of the peaks and the relative intensities largely match the experimental data.

2 - the cell parameters as loaded from the cif. I have put in some limiting values.

3 - the sites from the cif with added occupancies. With the -ve y values for O3 and O5 changed to +ve

4 - the resulting structure as seen in the TOPAS structure viewer. Note some of the atoms appear to be on the same site.

5 - the structure from the cif in JSmol on AMCSD - This is what I was expecting.

6 - a close up of the problem in the TOPAS viewer. Two Na atoms on the same site and ST, Na and O nearly on the same site.

7 - an image after refinement using this structure. Note the peaks are now in different positions, and the intensity match is poor.

I should also say that I spotted the two atoms at the same site CT1 and ST1 and tried this (image), but it does not solve the issue.

Dear Dr Ian, would you mind to attach ur files ( .par and raw data)

Hello Maykel,

here is the .raw file. I don't have a .par file for this. I have not used BGMN, the issue is in TOPAS.

Hi Dr Ian,

Please find the attachment for TOPAS .pro file and Burkiete.CIF. Modification in coordinat is by shift to -1/4, -1/4, 0 to get the correct Np.

Hello Maykel,

That is perfect! Thank you very much. This is the solution that Alexander proposed, but I still don't know how to implement it in TOPAS. The structure now looks good, the Np values are now correct and the pattern refines correctly.  I was unable to load your .pro file (see error below) so I don't really know how it is done. I used your .cif file, and I notice that is different to the Giuseppetti et al. files I was trying - OR have you edited it?? The SG is now pmmn as opposed to pmnm, the cell parameters for abc are transposed. The Beq values are different and have greater precision. What is the source for that cif?

Should this occur again I need to know how to spot the problem; why is a shift of precisely -1/4, -1/4, 0 selected? Is this done manually? 

So many questions, but I am indebted to you for your help.

Ian

*** Error loading sstring_in at

{ is_amor }

unknown or misplaced keyword.

at LINE 3640

See log file C:\Topas4-2\topas.log

Hi Dr Ian,

Well Topas dont have these facility yet. I spoted the issue when i saw all Np is 8, at Origin1 all coordinate in x,y,z (pic 1).

"I notice that is different to the Giuseppetti et al. files I was trying - OR have you edited it??" I just edited x and y. O ya I did for Thernadite as well.

The Beq values are different and have greater precision. What is the source for that cif? It's from from ICSD (pic2). I attached the original file saved from HS+.

"Should this occur again I need to know how to spot the problem;"  The easiest way to see the Np, is it the same as the .CIF.

"why is a shift of precisely -1/4, -1/4, 0 selected? Is this done manually" It's from the origin choice (pic3).

O so sorry I refined by Topas5. Here I attached for version 4.2. The quantification is not the same as v5, because i refined all microstructure, Beq, Occ and PO.

Unfortunatelly I lackof ur instrument configuration about the OD or PSD and most of the slits so I simply use the common configuration.

I hope it help.