Hi everyone,
I am calculating the out-of-plane polarization of trilayer WSe₂ using QuantumATK. I have generated several structures with different relative sliding configurations between the layers, and I would like to determine an appropriate vacuum size for the slab calculations.
To test this, I repeated selected structures using different vacuum thicknesses. However, I found that the calculated out-of-plane polarization is sensitive to the unit-cell height. My understanding is that QuantumATK reports the polarization as a 3D Berry-phase polarization normalized by the full periodic cell volume, including the vacuum region. Therefore, increasing the vacuum changes the reported \(P_z\).
I would like to ask: what is the recommended way to perform a vacuum convergence test for out-of-plane polarization in a 2D/slab system in QuantumATK?
I also checked the relaxed geometry as a function of vacuum size by monitoring the W–W interlayer distances. These distances are mostly stable, suggesting that the geometry is close to vacuum converged, but the polarization-derived quantities still show noticeable vacuum dependence.
Any advice on the correct QuantumATK workflow for vacuum convergence of slab polarization would be greatly appreciated.
Thank you.
