Convergence of energies

Materials Project
#1

Hi, energies of materials on MP are typically reported to the fourth decimal place. For example NaCl: -3.3881 eV. (found here: https://materialsproject.org/materials/mp-22862/#calculation-summary)

Does this mean a numerical precision of -3.3881 ± 0.0001 eV/atom ?

My doubts come from the Materials Project Documentation (https://docs.materialsproject.org/methodology/total-energies/):

In the “Total energy convergence” chapter it says that “96% of compounds tested were converged to within 15 meV / atom.”

Does this mean that I have to assume for every energy reported on MP a worst-case numerical precision of 15 meV?

In the case of NaCl this would give -3.38 ± 0.015 eV/atom ?

#2

Hello Victor,

Well, the numerical precision might be up to 4 digits (i.e., the precision you get if you reran the exact same calculation twice), however, what matters is the convergence accuracy of DFT (i.e., to what extent is the energy value converged w.r.t. the plane wave energy cutoff, number of k-points, etc.). And as you already mentioned, this is explained in the link you shared:

Using a 500 pra k-point mesh, the numerical convergence for most compounds tested was within 5 meV/atom, and 96% of compounds tested were converged to within 15 meV/atom. Results for the new parameter set will be better due to the denser k-point mesh employed. Convergence will depend on chemical system; for example, oxides were generally converged to less than 1 meV/atom.

-Peter

#3

Alright, thank you!