I am using two software like MAUD and Full Prof suite if I will take simple XRD data not like neutron or synchrotron XRD data for TiO2 (White) and Reduced TiO2-x whether it is reliable to do refinement two compare density value or occupancy factor because lighter elements like oxygen cannot be appropriate for proper reflection in XRD two give occupancy in a unit cell? so please somebody suggest me is it right ..?
How to fit peak if shift is there in peak when to use B anisotropy and when B isotropy data
even I fit the above data but in density value for Anatase it is showing 7.66cm3/g which is wrong although chi2 is coming near around 1.13
Please find the attachment for that also see thedensity difference for two files? in Fullprof
I think, the major problem is that you are increasing the number of interfering parameters. Rietveld refinement has been developed in order to refine structure data. Already this is a very critical procedure since you have to make a lot of assumptions. Only because a software enables to refine everything it is really not recommended to do this. Commonly refine everything step by step. Start with data you understand and keep everything else fixed. Of course, if you play with B (anyway which kind) your specific measurement (noise,misalignment etc.) will change B in order to get a lower mathematical difference between experiment and simulation. However, it does not mean that the result is physically more meaningful. The more parameter you start to refine the worse is the physical meaning but mathematically it will improve the fit. It becomes even worse if you now even work with two or more phases since the number of parameters is immediately much bigger. Why programs offer this option? Since for specific mixtures it works (if you don't touch the structure parameter!). This was the idea of full pattern refinement based on the approach of H. Rietveld. The mathematics is in fact the same, why not use the same program? But you should do both at the same time. There are many people fail in structure refinement, why do you believe you can do structure refinement on a mixture of phases?
I can only recommend to refine only parameters you are familiar with. It makes no sense to refine them with the aim to get the lowest R values, or which kind of reliability values you prefer. I worse R-value can match the truth better than a o lower one. And don't use anisotropic B values for powder patterns. You cannot see them as in a scan of a single crystal?
I think, the major problem is that you are increasing the number of interfering parameters. Rietveld refinement has been developed in order to refine structure data. Already this is a very critical procedure since you have to make a lot of assumptions. Only because a software enables to refine everything it is really not recommended to do this. Commonly refine everything step by step. Start with data you understand and keep everything else fixed. Of course, if you play with B (anyway which kind) your specific measurement (noise,misalignment etc.) will change B in order to get a lower mathematical difference between experiment and simulation. However, it does not mean that the result is physically more meaningful. The more parameter you start to refine the worse is the physical meaning but mathematically it will improve the fit. It becomes even worse if you now even work with two or more phases since the number of parameters is immediately much bigger. Why programs offer this option? Since for specific mixtures it works (if you don't touch the structure parameter!). This was the idea of full pattern refinement based on the approach of H. Rietveld. The mathematics is in fact the same, why not use the same program? But you should do both at the same time. There are many people fail in structure refinement, why do you believe you can do structure refinement on a mixture of phases?
I can only recommend to refine only parameters you are familiar with. It makes no sense to refine them with the aim to get the lowest R values, or which kind of reliability values you prefer. I worse R-value can match the truth better than a o lower one. And don't use anisotropic B values for powder patterns. You cannot see them as in a scan of a single crystal?
In general it is better to not refine the anisotropic thermal factors from powder data, particularly if the quality of the pattern is low as in your case.
TNT-BK, the Anatase occ of Ti/O = 0.43886/0.95865, that gave a density of 4.029.
TNT-WT, the Anatase occ of Ti/O = 0.31701/0.61709, that gave a density of 7.663.
Once change to Ti/O = 0.4/0.8 than the density 3.885.
There is also mix Biso and Baniso in same phase, also negative. Try to refine thermal parameter at very last.
The comparisson of both diffractograms shows a very different quality.
what will be the intial value of occupancy that has been used during Reitveld refinement of TIO2-rutile structure?
Yes, you can play with all the parameter if the process is step wise and accurately. I also did the same and easily find the result. Try to refine B and occupancy at last. Initially at occupancy position put 1 for Ti and 2 for Oxygen and then refine it which easily take accurate value as in your sample it should be.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Nanomaterials