Now I am learning GSAS (GENERAL STRUCTURE ANALYSIS SYSTEM).
But I feel confusing about the setup of space group R3m. To my limited knowledge, R3m can be written in GSAS as "R 3 m R" or "R 3 m". The latter one is actually Hexoganal and is preferred in GSAS.
However, I found that the calculated intensity was very different from the observed one. And I checked my space group and atom sites and compared them with the cif files that have similar structures. Unfortunately, I still could not find what was wrong with my refinement.
The structure of my sample is pervoskite (ABO3), similar to BaTiO3, PbTiO3 etc. Pb and Bi atoms take the A site with the occupancy of 0.5:0.5. Ti and Sc atoms take the B site with the occupancy of 0.5:0.5.
The files attached here is the unit cell and atom setting. But from the liveplot, the calculated relative intensities of peaks is very different from the observed pattern.
Can someone tell me what is wrong with my refinement?
Thanks!
At first sight your coordinates are wrong for your lattice dimensions. If you input your coordinates in any structure plot software (e.g. DrawXTL as suggested in EXPGUI) you'll find your octahedra are very distorted and disconnected from each-other and the relative position of Ti and Pb are clearly wrong. With this cell (where c is exactly the body diagonal of the cubic perovskite cell) your Pb/Bi atoms should be half-way the c axis.
If you try
Pb/Bi 0, 0, ~0.5
Ti/Sc 0, 0, ~0 (that one is correct)
O ~0.34 ~0.17 ~0.16
youll get the correct intensity relations and the connectivity of a perovskite. This model however shows significant disorder and the true occupancies of all sites should be half of what you give in order to have reasonable metal-metal distances and density.
Maybe an ordered model with c~14Å is a better choice, I'm not expert in these kind of ferroelectric materials.
Check COD ID: 1501514 for an example in your same space group and very similar system.
Good luck!
are you sure your specimen is really the R3M phase? this is just one of the polytypes of the materials you mention!
At first sight your coordinates are wrong for your lattice dimensions. If you input your coordinates in any structure plot software (e.g. DrawXTL as suggested in EXPGUI) you'll find your octahedra are very distorted and disconnected from each-other and the relative position of Ti and Pb are clearly wrong. With this cell (where c is exactly the body diagonal of the cubic perovskite cell) your Pb/Bi atoms should be half-way the c axis.
If you try
Pb/Bi 0, 0, ~0.5
Ti/Sc 0, 0, ~0 (that one is correct)
O ~0.34 ~0.17 ~0.16
youll get the correct intensity relations and the connectivity of a perovskite. This model however shows significant disorder and the true occupancies of all sites should be half of what you give in order to have reasonable metal-metal distances and density.
Maybe an ordered model with c~14Å is a better choice, I'm not expert in these kind of ferroelectric materials.
Check COD ID: 1501514 for an example in your same space group and very similar system.
Good luck!
Can u paste ur data and or ur GSAS refinement here or to box? The reflection list is in the right position, maybe the real occupancy is not like what u expected, there is preferred orientation for bulk sample or u enter the phase using CIF database (cause some had 2 origin).
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction
- XRD Analysis