john201004
Hi
During multiphase Rietveld analysis after introduce "do_errors" I got extremely big errors for some parameters, for example:
weight_percent alpha_weight_percent 11.613`_14459073.775
cell_mass alpha_cell_mass 303.210`_205189276.638
scale alpha 3.31945616`_4.58e+006
Phase_Density_g_on_cm3( 4.59785`_3111466.76337)
all other parameter's error looks quite reasonable. I don't know this means my fitting is very wrong or this is how Topas reports error for these four parameters listed above?
Sincerely
mfisch
Wt.-% is correlated with cell mass. Cell mass also has a huge error in your case. Do you refine any atomic positions and if yes, what are their values and esd's? Try using a known structure (from databases) for your phase "alpha" and see what happens.
john201004
Hi, thank you for your reply!
I did not refined any atom positions and I also did not refined any peak shape parameters. However I refined unit cell parameters, scale, zerro_error and bkg.
Sincerely
mfisch
You have to find out why cell_mass is not constant but refined (I suppose it is as there is an esd on the value). I am pretty sure that this is the key. Normally, cell_mass is M*Z, where M is the sum of wt. of the atomic content as listed in the desrription of the structure.
Try setting cell_mass alpha_cell_mass 303.210`_205189276.638 to cell_mass 0 and run again, or delete cell_mass and weight_percent and replace with the MVW(0, 0, 0) macro