Output file

fatma

Hello!

I have done Pawley refinement using topass but i don't know how to infer information from output file. How much error is acceptable.
You should get a Rwp of ~ 14.6%. What does this mean.
Below is data i got from my output file. Is 4.345 is acceptable or not.
How to get to know accuracy of lattice parameters in my output file.
r_wp 4.3451529 r_exp 2.99643118 r_p 3.10328839 r_wp_dash 9.78581239 r_p_dash 9.641208 r_exp_dash 6.74832716
Please tell.

johnsoevans

There's no simple answer to this question as Rwp depends on many factors of your experiment such as data collection statistics and background levels. The issues are discussed in many of the texts on Rietveld refinement. You should also look at quantities such as "goodness of fit" and "chi²" which related your Rwp to the statistically expected value Rexp. They should approach 1 for a good model (chi=gof=Rwp/Rexp).
You could also try looking at the article by Brian Toby on agreement factors here.

fatma

Thank you for your reply.
I've seen the paper which stated that; 'when chi 2 is close to one , there is no guarantee that model is correct.'
Chi2 should never drop below one or equivalent. is it?
Can you please tell looking at my output file, is it acceptable or not, i took longer scan(not a quick scan) while collecting data.
r_wp 4.3451529 r_exp 2.99643118 r_p 3.10328839 r_wp_dash 9.78581239 r_p_dash 9.641208 r_exp_dash 6.74832716

Chi square in this case would be 1.45, which is not ideal but can be acceptable?
If you can please also explain Rwp of 14.6 percent mean what?
How to calculate its percentage
i'll be grateful

johnsoevans

The Rwp is defined in paragraph 4 of the Toby paper. It's also in the topas technical reference, or in the youtube Rietveld video here. You multiply by 100 to get the %.
In general, making sure the fit is visually correct is at least as important as looking at agreement factors. This is discussed around minute 17 of the video.

fatma

Hello
Thanks alot for detailed reply.
it helped me understand a lot.
i am also confused about :
1)For materials which is not present on ICSD project, where to take its alpha, beta, gamma, space group etc. so that it can be added on my input file of topass.
2) if i have fitted 2 phases of materials how to differentiate between them in origin too, so that it can be differentiated that which phase it is. And also please tell how to calculate ratio of two phases.

johnsoevans

If the structure isn't in the ICSD then you either have to start with a model from a compund you think is similar, or try and solve the structure of the material ab-initio.
If you type "weight_percent 0" in each str section of your INP file the "0" will be updated with the weight percentage at the end of refinement. In the gui you can see peaks from individual strs by hovering your mouse over the phase name in hte main plot window. "View/Phase Names" needs to be ticked for this to be possible. It's also helpful if you give each str a sensible name with "phase_name yournamechoice" in the str section of the INP file.

fatma

Thank you so much for replying,
In software it worked as I've given names in input file to differentiate between 2 files, but when i drew data in origin , i was unable to differentiate between 2 phases. If there is any relevant video or data which can elaborate further my questions ,would work best for me.
Please also comment on how to calculate ratio of 2 phases.

johnsoevans

Search the forum for eg “output individual phase”.

I dont understand your second question. My previous reply tells you how to work out weight percentages. Or you can compare scale factors.