Dear Zajdel, I am literally touched, thanks for your kind reply!
I sent my question to the fullprof support team as you told me. I will share the reason and solution as soon as I receive. Thanks for your expertise again. Best regards, Do-Yeon, Kim ============================================ 김도연 (Kim Do Yeon) Pohang Accelerator Laboratory (PAL) 80 Jigokro-127-beongil, Nam-gu, Pohang, Gyeongbuk, 37673 Republic of Korea E-mail: kdo01...@postech.ac.kr, kdo01...@gmail.com Tel: +82-054-279-1613, Mobile : +82-010-7350-2858 ________________________________ 보낸 사람: Paweł Zajdel <pawel.zaj...@us.edu.pl> 보낸 날짜: 2022년 8월 11일 목요일 오전 4:39 받는 사람: 김도연(에너지환경연구팀) <kdo01...@postech.ac.kr> 제목: Re: R 3 m:R in fullprof Dear Kim, Some good news and some bad news 1. Please check that all of your phases have correct crystallographic SOFs. Phases 2 and 3 are off by a factor of 8. This is the bad news. The good news is that FP seems corrects this in the analysis but please check in future. This is clear when you see: "The given occupation factors have been obtained mutiplying m/M by 8.0000" The relative occupations within each phase are OK. 2. You should use similar isotropic B for the same phase. This does not change a lot in the final result for strong phases. Changing Biso can correlate with the scale factor for weak phases (and this is your Phase 2) and can change results for them a lot. 3. There is a free origin along the z-axis for both cases of Phase 1. You have fixed it for x=y=z Ba at 0 for the rhombohedral case so you can also fix z_Ba = 0 in the hex axes setting. 4. From the first peaks of Phase 2 (100) and (110) it looks like Phase 2 has wrong stoichiometry. But it will not change your overall discrepancy between rhombo and hex axes. 5. Do not refine at the same time individual Bs and the overall B like in the phase 1 in the hex setting. Pick up either individual or overall. 6. Now for the bad news. Even taking all of the above into the account your original problem remains. Sorry that I did not help much in this. Keep asking the maintainers of Fullprof. It does not look 30% of the cubic Phase 2. 5% or less is more likely result. Stay with the result in the trigonal group. Best Pawel On 8/10/2022 7:00 PM, 김도연(에너지환경연구팀) wrote: Dear Zajdel, Hello, thanks for your skillful reply!! I must keep in mind with your comments, but I think it can not solve my problem. I attach my data and pcr files as I posted on the choice of hexagonal and rhombohedral system. Thus, if you feel free and have spare time, I hope you can check my results, why this weight fraction difference happens. I really appericate to be interested in my question ! Best regards, Do-Yeon, Kim ============================================ 김도연 (Kim Do Yeon) Pohang Accelerator Laboratory (PAL) 80 Jigokro-127-beongil, Nam-gu, Pohang, Gyeongbuk, 37673 Republic of Korea E-mail: kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>, kdo01...@gmail.com<mailto:kdo01...@gmail.com> Tel: +82-054-279-1613, Mobile : +82-010-7350-2858 ________________________________ 보낸 사람: Paweł Zajdel <pawel.zaj...@us.edu.pl><mailto:pawel.zaj...@us.edu.pl> 보낸 날짜: 2022년 8월 11일 목요일 오전 1:50 받는 사람: 김도연(에너지환경연구팀) <kdo01...@postech.ac.kr><mailto:kdo01...@postech.ac.kr> 제목: Re: R 3 m:R in fullprof Dear Kim, Mass % fraction in Fullprof is calculated on the basis of several assumption. 1. First of all, please check you *.out file and find a line that says: " => IT IS ASSUMED THAT THE FIRST GIVEN SITE IS FULLY OCCUPIED " Under this line there will be several lines that may lead to different mass % depending on the way you have introduced the new structure into your PCR file. If you have solid solution - make sure the first atom is always the fully occupied one. 2. Make sure all your SOFs relative to the first atom are correct. There is a line that says: "=> The normalized site occupation numbers in % are:" If correct - the numbers should reflect the chemical occupancy of each site in the cell. The calculated crystallographic density also should be close to real one. This means that the first atom in the atom list should say 100% and the remaining sites should be set relative to this one. In the ideal case - 100% for all. 3. Then check if they have correct absolute crystallographic occupancies. There is a line that says: "The given occupation factors have been obtained mutiplying m/M by 1.0000" If the number at the end is not 1.000 - there is a multiplicative factor missing from occupancies given in the pcr file. This is irrelevant in a single phase case but crucial in multiphase systems and quantitative results. 4. If you change structure make sure to recalculate your ATZ. There is a line: "The given value of ATZ is XXXX the program has calculated: XXXX" It should list the same number. The more technical detail are given in Fernado's reply. Hope this helps Pawel On 8/10/2022 6:26 PM, 김도연(에너지환경연구팀) wrote: Dear Igoa, I am touched with your kind reply!! I agree with your opinions as you told me perfectly, the Z difference bewtween hexa and rhombo system, and hexagonal system is more reliable results. That is why I was really confused to refine, and I thought I had to appy the hexagonal system to refine regardless of rhombohderal phase. Is it possible to earn different results depending on the choice of axe, even though it is same structure..?! Anyways, I am going to send our discussion to the fullprof support team on the fullprof homepage. Also, if you feel free and have spare time, you can ask this issue to anyone. I will share any opinion from the support team as soon as I receive. Thanks for your expertise again. Best regards, Do-Yeon, Kim ============================================ 김도연 (Kim Do Yeon) Pohang Accelerator Laboratory (PAL) 80 Jigokro-127-beongil, Nam-gu, Pohang, Gyeongbuk, 37673 Republic of Korea E-mail: kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>, kdo01...@gmail.com<mailto:kdo01...@gmail.com> Tel: +82-054-279-1613, Mobile : +82-010-7350-2858 ________________________________ 보낸 사람: Fernando Igoa <fer.igoa.1...@gmail.com><mailto:fer.igoa.1...@gmail.com> 보낸 날짜: 2022년 8월 10일 수요일 오후 7:21 받는 사람: 김도연(에너지환경연구팀) <kdo01...@postech.ac.kr><mailto:kdo01...@postech.ac.kr> 제목: Re: R 3 m:R in fullprof Hello Kim, Thank you very much for sharing your files. I took a look at your pcr, I don't see anything wrong in your inputs, so I tried to break down the problem: See, the phase fraction (Wj) is calculated as: Wj = Sj.Vj.ATZj/Sum(i)[Si.Vi.ATZi] Where Sj is the scale factor, Vj the cell volume and ATZ = Z.Mw.f^2 / t Where Z is the number of formula units per cell, Mw the molecular weight, f are the transformed occupancies I talked you about before (f = occ*M/m with occ = chemical occupancy, m = site multiplicity and M = general multiplicity). Your ATZ are coherent, for the Hexagonal system you have 738.7585 and for the rhombohedral you have 246.2509, that is exactly 3 times smaller, as it should be from the difference in Z (Z = 1 for R, Z = 3 for H). Now this factor of 3 is never compensated back as far as I can tell. According to the equation of Wj, the only way to compensate it, is with the scale factor Sj, but you have the same scale factor for both phases. I took the liberty to calculate what would you obtain with a scale factor 3 times larger for BaHfTiO3 and 3 times smaller for the other 2 phases, and indeed the results become much more similar than to the hexagonal system, still not the same because the scale factor does not necessarily has to be 3 times larger to accurately compensate because you have the other 2 scale factor of the other two phases). Anyway, I think is not you is the program, and I think that this is a question to take to Rodriguez-Carvajal, if you are interested you can write him yourself or if you allow me I can write on your behalf as this interests me as well. In any case, if you are in a hurry, trust the Hexagonal system result, your pattern is clearly 96 % pure and not 73 %. Apologies I could not be of further help, if you figure something out also let me know please Best, Fernando Igoa PhD student Laboratoire Chimie de la Matière Condensée de Paris Sorbonne Université 4 place Jussieu 75005 PARIS ORCID: orcid.org/0000-0002-3960-4720<https://orcid.org/0000-0002-3960-4720> ResearchGate: researchgate.net/profile/Fernando_Igoa<https://www.researchgate.net/profile/Fernando_Igoa> [https://68ef2f69c7787d4078ac-7864ae55ba174c40683f10ab811d9167.ssl.cf1.rackcdn.com/twitter-icon_24x24.png]: twitter.com/figoa96<https://twitter.com/figoa96> El mié, 10 ago 2022 a la(s) 08:14, 김도연(에너지환경연구팀) (kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>) escribió: Dear Igoa, I really appreciate your contact and reply !! I attached my pcr files, and you can check my results as I posted. In details, I am researching the perovskite structure on BaTiO3, and I doped Hf into Ti site. While doing refinement, I found that the weight fraction was severe different depending on the choice of axes. Just I want to know and check that we can earn different refinement data depending on the choice of axes. In addition, if I refine hexagonal system, I have to use the crystal structure information from standardized crystal structure data in ICSD? (I attached BaTiO3 rhombohedral phase, ICSD information) or just something wrong in fullprof program? Thus, also if you deal with some materials which are the hexagonal structure, could you try them like me? Thanks for your kind and expertise, Best regards, Do-Yeon, Kim ============================================ 김도연 (Kim Do Yeon) Pohang Accelerator Laboratory (PAL) 80 Jigokro-127-beongil, Nam-gu, Pohang, Gyeongbuk, 37673 Republic of Korea E-mail: kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>, kdo01...@gmail.com<mailto:kdo01...@gmail.com> Tel: +82-054-279-1613, Mobile : +82-010-7350-2858 ________________________________ 보낸 사람: Fernando Igoa <fer.igoa.1...@gmail.com<mailto:fer.igoa.1...@gmail.com>> 보낸 날짜: 2022년 8월 10일 수요일 오후 2:48 받는 사람: 김도연(에너지환경연구팀) <kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>> 참조: rietveld_l@ill.fr<mailto:rietveld_l@ill.fr> <rietveld_l@ill.fr<mailto:rietveld_l@ill.fr>> 제목: Re: R 3 m:R in fullprof Dear Kim, Are you sure you are deleting/adding the atoms generated by the change of system? Otherwise did you check if the occupations remain the same for the atoms after refining in each system? Like the true chemical occupations, not the occupation number refined in fullprof. Because in fullprof, the refined occupation number is calculated as the chemical occupancy x site multiplicity. Now this may or may not be normalised to the multiplicity of the general position of the group (literally in the manual it says "can be normalised to the multiplicity of the general position of the group"). So it may be that if this is not normalized, and you keep the same occupation numbers, this will change your chemical occupation, hence your weight fraction. Other than that I'm out of clue, if you could share the pcr file it would be of use. Hope it helps, let me know. Best, Fernando Igoa PhD student Laboratoire Chimie de la Matière Condensée de Paris Sorbonne Université 4 place Jussieu 75005 PARIS ORCID: orcid.org/0000-0002-3960-4720<https://orcid.org/0000-0002-3960-4720> ResearchGate: researchgate.net/profile/Fernando_Igoa<https://www.researchgate.net/profile/Fernando_Igoa> [https://68ef2f69c7787d4078ac-7864ae55ba174c40683f10ab811d9167.ssl.cf1.rackcdn.com/twitter-icon_24x24.png]: twitter.com/figoa96<https://twitter.com/figoa96> El mié, 10 ago 2022 a la(s) 02:29, 김도연(에너지환경연구팀) (kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>) escribió: Hello, all experts I figured out the excited update in fullprof, which was we can do refine the primitive cell of rhombohedral group immediately. (https://www.ill.eu/sites/fullprof/php/FullProf_News_2022.htm#2007) Fullprof News - 2022<https://www.ill.eu/sites/fullprof/php/FullProf_News_2022.htm#2007> Full Year 2022----- >> 2 August 2022 : Automatic renumbering of codes used in linear restraints ----- - If modifications of VARY/FIX commands to refine or fix some parameters were used, the renumbering of explicit codes used in linear restraints instructions was not applied. www.ill.eu<http://www.ill.eu> My material is R 3 m (#160), and I am trying to refine it using R 3 m:R with the rhombohedral structure information. Let me suppose my material is A, the second impurity is B. (my data have only two phases) When I refine using R 3 m with the trigonal structure, I earn the weight fraction A:B=80%:20% But, when I refine using R 3 m:R with the rhombohedral structure, I earn the weight fraction A:B=50%:50%. When I tried both cases, I put all paramteters with same values in the second impurity like same scale factor, same profile values.. That is why I am getting in the problem on the weight fraction, and I am so confused.. Please share your expertise... Thanks, Do-Yeon, Kim ============================================ 김도연 (Kim Do Yeon) Pohang Accelerator Laboratory (PAL) 80 Jigokro-127-beongil, Nam-gu, Pohang, Gyeongbuk, 37673 Republic of Korea E-mail: kdo01...@postech.ac.kr<mailto:kdo01...@postech.ac.kr>, kdo01...@gmail.com<mailto:kdo01...@gmail.com> Tel: +82-054-279-1613, Mobile : +82-010-7350-2858 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Please do NOT attach files to the whole list <alan.he...@neutronoptics.com><mailto:alan.he...@neutronoptics.com> Send commands to <lists...@ill.fr<mailto:lists...@ill.fr>> eg: HELP as the subject with no body text The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Please do NOT attach files to the whole list <alan.he...@neutronoptics.com><mailto:alan.he...@neutronoptics.com> Send commands to <lists...@ill.fr><mailto:lists...@ill.fr> eg: HELP as the subject with no body text The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Please do NOT attach files to the whole list <alan.he...@neutronoptics.com> Send commands to <lists...@ill.fr> eg: HELP as the subject with no body text The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++