Using ISIF=2 for Surface Calculations

[abhardwaj73]

Hello everyone,

I'm doing surface slab calculations for Calcite, Dolomite, and other related materials. I found out that various input set generators for these calculations like MVLSlabSet
uses ISIF=2 for relaxation. Is this the correct choice? I'm not sure since the stresses are not relaxed using ISIF=2.

One reason I could think of is the underlying bulk material will not allow the surface to relax fully (or at least the stresses).

Please let me know if using ISIF=2 for surface slab calculations is the correct choice.

[DanielYang59]

uses ISIF=2 for relaxation. Is this the correct choice? I'm not sure since the stresses are not relaxed using ISIF=2.

I'm afraid you might have misread the VASP manual because "ISIF=2" does calculate stress tensors.

For ISIF ≥ 2, the stress tensor is calculated.

I'm by no means an expert, but what I usually do is using "ISIF=3" to relax the cell first (for slab with vacuum level), and then using "ISIF=2" to relax ionic positions. For a slab though, I usually fix the cell along the vacuum axis during cell relaxation. Alternatively you might try the Birch–Murnaghan equation of state.