Recently, I was confused about primitive cell and unit cell which were used in maps and mcsqs. I used primitive cell for FCC structure in maps previously. However, a professor who has a lot of experience in DFT told me what I did was wrong, because the alloy of supercell expanded from FCC primitive cell broke the FCC symmetry. Actually, I didn’t agree his opinion but I cannot persuade him. In my opinion, starting from primitive can include all structures. Am I right? Is there any literates that I can refer to insist my opinion?
On the other hand, when I generate a SQS structure of wurtizite, if I used primitive cell in mcsqs, it usually generate a non wurtzite structure. In this case, it’s hard to compare the band structure between the parent materials and alloys. To compare the band structure, I use mcsqs with a wurtzite supercell as start and keep the atoms number fixed. For example, for wurtzite MgxZn1-xO alloy, I generate a 221 supercell from primitive cell with 16 atoms, and then mcsqs -n=16 to generate SQS. Is it correct? Or I should try other better method?
It’s difficult to see who right because I don’t know which property you are trying to model.
The SQS formalism works for scalar function of configuration (e.g. energy or even electronic density of state). For those properties, the lower symmetry of the SQS is irrelevant.
If you want elastic constants or band structures (which depend on direction), then you need to symmetrize the SQS.
In the case of elastic constant, see https://dx.doi.org/10.1103/PhysRevB.72.144109 for how it’s done.
There is also a cluster expansion formalism generalized to account for orientation-dependence, e.g. https://dx.doi.org/10.1038/nmat2200.
If you want to plot band structures (e.g. in the fcc Brillouin zone) using an SQS, then I don’t have an easy answer.
I use maps to to find intermediate ground states from FCC A to FCC AB3. I want to find out B sites in A lattice with various B composition. Since A has space group 225 and occupies 4a sites, I put B on 4b and 8c sites and make B substituted by Vac. What we argue is that I insist the intermediate ground states, eg. AB, with the lowest formation energy is correct. But the professor said, since I input FCC AB3, then no matter of Vac substitution, the material will be AB3 instead of AB. On other word, he means AB can not be obtained from AB3 through replacing B by Vac. Sorry for the confusion, because I confuse too much now.
I don’t have enough information to answer.
I would need the actual structure (in ATAT format):
lat.in and the */str.out files.
Otherwise, I don’t know if one reports the correct symmetry and wyckoff positions.