I’m curious about how AMSET calculates phonon frequencies using DFPT. Does it necessitate a supercell to obtain these frequencies? Or does it compute them based on the unit cell, presuming that phonon frequencies at the Gamma point remain consistent regardless of supercell size? What’s your take on this?

Hi,
The DFPT calculation was done with a unit cell.
Here’s how I understand it:
AMSET reads the output to get the gamma center modes and calculate an averaged polar optical phonon frequency by weighting each mode with Born effective charge.
The code then use this frequency to calculate the Frohlich term.
They also assume that the frequency changes slowly near gamma point, so it can be treated as a constant (independent with q vector).