I am relaxing the plastic crystal Li2SO4 (space group Fm-3m) using the Projector Augmented Wave (PAW) method and a revised PBE for solids (PBEsol). The INCAR I used can be found below. When relaxing the structure, the volume of Li2SO4 increases about ~30% compared to experimentally found values. I have also tried using ISIF = 2 and ISIF = 7 before using ISIF = 3, but this did not make a difference. I also relaxed the structure by performing an NpT simulation and found the volume to be much closer to experimental values.
Another plastic crystal (gamma-Na3PS4. Space groud Fddd) that I relaxed does not have this issue using the same INCAR, however it does have a much less densely packed unit cell. What could be causing this volume expansion?
INCAR:
GGA = PS
ENCUT = 600
ISTART = 0
ICHARG = 2
PREC = MEDIUM
NSW = 1000
ISIF = 3
LREAL = Auto
IBRION = 2
ISMEAR = -5
NPAR = 4
LWAVE = .FALSE.
LCHARG = .FALSE.
Is it possible that the experimental value is giving in Angstrom but the software requires input in atomic units (Bohr radii)? I mean, the deviation by that swap would be more than 30%, but it would at least match the order of magnitude.
The initial volume of my relaxation matches the cube of my lattice parameter in Angstrom so I am confident that it is not a unit issue.
- Badges
- Science topic
- Similar topics
- Correction