Discrepancy between TOPAS and FullProf peak intensities

SWright1

Hello,

I am having a strange problem when trying to run a powder neutron diffraction refinement in TOPAS.

In TOPAS, I cannot fit the intensity of the Cu sample holder peaks well, without using preferred orientation. The pyrochlore phase fits OK, but requires further work. In Full Prof, the Cu peak intensities fit quite well. However, the pyrochlore peak intensities are fit poorly. I tried changing the Fd-3m space group settings in both TOPAS and Full Prof, but that didn't help.

Also, which peak shape functions in TOPAS correspond to the S_L and D_L asymmetry parameters in Full Prof?

I suspect that that those parameters may have something to do with the issue, since those are the only peak shape parameters which I wasn't sure about how to transfer over.

This experiment was performed on HB2A: a constant wavelength neutron diffractometer.

Here is a Google Drive link to the data and input files: https://drive.google.com/drive/folders/1RY3vLMOSAl5ZABoqarhAdzmBlXbpL3Xp?usp=share_link.

Thanks a lot for the help,
Sebastian

rowlesmr

I can't see any technical reasons why the Cu peaks aren't being fit.

Checking out the Full prof manual for S_L and D_L:

Asymmetry parameters corresponding to the L. Finger formulation of the axial divergenc

and

S_L is source width/detector distance

so, I'm guessing that

Finger_et_al(4.3, 16) ' S_L = 4.3, D_L = 16

What could help is looking at the structure factors and stuff between FP and TOPAS.

I smooshed this macro from things available in topas.inc. Maybe it could help see what is being calculated differently?

{
               out file
      Out_String("/")
      Out_String("*")
      Out_String("H --K --L --M -F_Real +ve -F_Real -ve -F_Imag +ve -F_Imag -ve -----F2 +ve -----F2 -ve --F2_Merged -------scale ---all_scale -I_no_scale I_after_scale")
      Out_String(" *")
      Out_String("/\n")
   phase_out file append load out_record out_fmt out_eqn
   {
       "%3.0f" = H;
      " %3.0f" = K;
      " %3.0f" = L;
      " %3.0f" = M;
      " %11.5f" = F_Real_positive;
      " %11.5f" = F_Real_negative;
      " %11.5f" = F_Imaginary_positive;
      " %11.5f" = F_Imaginary_negative;
      " %11.5f" = F2_positive M .5;
      " %11.5f" = F2_negative M .5;
      " %11.5f" = F2_Merged M;
      " %11.5e" = Get(scale);
      " %11.5e" = Get(all_scale_pks);
		" %11.5f" = I_no_scale_pks;
		" %13.5f\n" = I_after_scale_pks;
   }
}

You'd bung it in the str as Out_Details("Cu_out.txt"), or the like.

johnsoevans

I couldn't see anything either. Just as a sanity check I simulated the neutron pattern of Cu in other software. Your INP file would fit the simulated data and the preferred orientation refined to 0.9995. With only two Cu peaks it might be that another 2-theta dependant correction is affecting things?

SWright1

Hello again,

Looks like the correct expression for S_L = 0.043, D_L = 0.043 in Full Prof is

Finger_et_al(4.3, 4.3)

in TOPAS.

I have attached the Cu_out.txt file. I did find a mistake in my Full Prof input file, the occupancy for Cu should be occ = special pos multiplicity/general pos multiplicity. This correction only changed the intensity scale though.

Attached is Cu_out.txt file generated by TOPAS.

cu-out.txt

459B

Cheers

SWright1

Upon further thought, I think that I need to add a transmission geometry correction. I am going to try the macros found here: https://topas.awh.durham.ac.uk/doku.php?id=flat-plate_transmission.