Grouping sites

Alloy Theoretic Automated Toolkit (ATAT)
1

Hi,

Is it possible to group sites when we generate structures? For example, if we have 4 sites where there are Li or Vac, and in a structure file generated by running mmaps (str.out), I want only one of the first two sites to be occupied (inside a unit cell and not the supercell), and one of the last two sites to be occupied. This type of calculations seem to be necessary when ions (say Li) form complexes with certain atoms in the sub-lattice (say O), and while there are more than one ‘site’ around ‘O’ where ‘Li’ can adsorb, there is only one Li per O atom.

If there is only one ‘O’ in the lat.in file, I think I can trivially achieve this by setting inequalities in the crange.in file such as 1.0O - 1.0Li >= -0.1 and 1.0O -1.0Li <=0.1, so that they remain equal, but if there are more than one type of ‘O’ positions, this problem appears to be non trivial.

Any suggestions?

Thanks,
Ganesh (forwarded by “avgjoe”)

2

It’s tricky but one way to do it is to have your job script not run any structure with two ions or more in the same group . Maps won’t be able to fit ecis connecting sites in same group and set them to zero. Once the fit is done, you can manually set these Eci to large value to penalize multiple occupations in the group.