Many thanks, Roberto Köferstein! 

I am wondering how accurately would be the calculation using these software. if so, is any of these software has advantage than the other?

I am asking this because the last option, that you have provided, would not be feasible in my method (unless I buy nanoparticles of known phases). 

Thank you. 

Sincerely, 

Pradip

It seems like the default language of the software is not English (and I did not find an option to change it when run the software). So, can you please, send me the link (if you can find and different than the one you provided before), i.e. compatible with windows 7, 64-bit)?

Anyways, I am thinking to test experimental results for Degussa P25 (of known anatase to rutile ration) by comparing the estimated ratios of two different phases. 

Thank you. 

Sincerely, 

Pradip

Thanks a lot, Dr. Köferstein!

Unfortunately, the software is not compatible with my computer (and got the same error message before when I was trying to use from the link you sent me). I have attached the screen shot; do not worry, I will try it in another PC.

And regarding the procedure with DegussaP25: I was trying to isolate both the peaks correspond to respectively Anatase and Rutile (if the peak width is the same meaning the crystallite sizes comparable, in the mixture of known ratio!) and compare the results to estimate their ratio (with company provided value(s)).

Please, let me know if this is something wrong in my assumption.

Thank you.

P.S. I have not yet read the paper that you sent; and you can ignore this note if answer/explanation is there.

Try to run it in a compatibility mode of Windows 7 or 8 - this usually works well.

Dear Dr. Köferstein, 

I am afraid to inform you that I could not install and run the powdercell yet; and hence I have not yet run the scan for Degussa P25 (i.e. to compare the results).

However, I am trying to use the results from texture measurements (to avoid the issue with structure factors (as you mentioned that it could be a problem). However, I am not 100 % sure whether we can compare the peak intensities for pure Cu (cubic, with 0 intensities for Cu2O and CuO), pure Cu2O (cubic) and also same way for pure CuO (monoclinic; and without other impurities. If we get the expected results (from the intensity values, i.e. obtained for different phi, and Psi), then I will test the CuxO samples. 

Please, let me know if you have any comments/suggestions. 

By the way, I will definitely let you know if I solve this issue.

Thank you.

Sincerely,

Pradip  

download GSAS, it is free and works on all platforms, and there are tutorials available with sample data to refine phase fractions.

There are other open soruce programs which may be helpful to you,  like FULLPROF ( with Manual and examples) and MAUD program. for whole profile fitting and quantitative phase analysis. 

I agree with Prof Roberto's suggestion, but, in your case, you used three JCPDS files to identify the peaks.Further, at   200 deg C, peaks show some "chemical inhomogenity" (see "Bridge" type of profile between peaks at  2theta range of 35 deg to 40 deg). Therefore, whole profile fitting based on Rietveld refinement is the best option according to my opinon as there are free and open source codes are available for whole profile fitting.  

Many thanks to all of you!

Based on your suggestions and a short review of literature, I used Match software (which can be downloaded, here: http://www.crystalimpact.com/match/download.htm#download_fp).

And regarding the estimation and accuracy, what I found/figured out are the following:

1) the estimated quantity(%), of each phases in the mixture, is "SEMI-QUANTITATIVE".

2) accuracy depends on the chosen parameters while using the software, i.e. for the "CANDIDATE PHASE" and the associated parameters, viz: I/I_c, I/I_0, scale factor etc. (and the input variable is only the wavelength of X-ray used to get the data). Then, running Rietveld refinement (FullProf) would be an easy task: nothing to worry about, just run it to get the results for Quantity% (of matched phases).

3) the output, Quantity%, is nothing to do with the values of Phi and Psi when record the XRD data (I was assuming these values will come into account as we are accounting the I/I_c (intensity ratios in the recorded XRD data; and which, I think, should depends on those Phi and Psi values chosen?)

Anyways, I have tested the method for Degussa P25 (which is supposed to be a mixture of ~80% Anatase and ~20% Rutile, company prediction!). The XRD data were recorded at Phi, Psi = 0, 0 and omega = 0.5, using Cu Ka1 of wavelength = 1.5405980 A. And I found Anatase to be 85.2 % and Rutile to be 14.8 % (two screen shots are attached!). I have attached a .dat file if you are interested  (your feedback would be much appreciated!). 

Now on, I will use the same procedure to analyze the data for my CuxO samples. 

Thank you. 

Regards, 

Pradip

HI Amaresh, 

I did not get your point, in the line 3: can you explain about "chemical inhomogenity"?

Actually, that was one of the mixed phases (obtained at T = 200 C) that I wanted to estimate/separate Cu2O and CuO (their amounts!); and we can clearly see an extra peak that is emerging in between 35 and 40 degrees (2 theta), which I believe some part of Cu2O is oxidized into CuO after annealing at T = 170 C.

Thank you.

Regards, 

Pradip

I used the word "Chemical in-homogenity" to describe the mixed phases co-existing  (i used "bridge" to describe the small peak in between two large peaks); such profiles are difficult and often gives erratic results when quantitative analysis is done by reference intensity ratio method (i.e. comparing the mixture of Cu2O and CuO in this present case - as suggested by Prof Roberto). Therefore, in my opinion  whole profile fitting method is best option.

Sorry , if i have confused you (My english language may not be that good) 

Thanks for the explanation, Amaresh! 

I have just started using the software; and I was afraid of using erroneous results, obtained by using the false treatment(s) (that's why I was trying to understand your comments if there was something I needed to consider while doing the estimation or even beyond).

By the way, I am trying to improve the results by setting the parameters manually (i.e. using advanced and expert skill level for profile parameters, crystal structures and strain/stress/preferred orientation etc.). I am not sure whether I can get the results improved.

I would be happy if I get consistent results with error less than 5% (as Dr. Köferstein mentioned) for the P25 that I have shown the calculation for. You can try using the data (if you are interested!).  

Thank you. 

I can try in my spare time and i normally use MAUD and FULLPROF for analysis, i will let you know results of analysis (after my attempt for analysis).

thank you,