Modeling at finite temperatures

Alloy Theoretic Automated Toolkit (ATAT)
1

Hi Axel,

I wondered what difference it makes if one includes the temperature effects via phonon calculations on the DFT structures or if one just used the 0 K CE to proceed with the MC code to extrapolate to higher temperatures. Essentially, as I understand it, his would be 2 different approaches, isn’t it?

I guess it is much more computationally feasible to use 0 K CE with MC instead of constructing a CE with phonon contribution.

If I completely misunderstood these 2 concepts, please enlighten me.

2

Monte Carlo on a lattice model alone only gives you configurational contributions to the free energy.
Running actual phonon calculations (e.g. via fitfc or fitsvsl) is the only way to make the CE (and therefore the lattice model MC (emc2)) aware of vibrational contributions to the free energy.