The goal of this project is to utilize simple model systems to elucidate possible differences between Abinit and VASP computations. The first use of VASP will be on the VO3 chains synthesized by the Norquist lab from Spring 2010-present. As such, the first model system is KVO3, which consists of potassium cations between VO3 chains which are infinite polymers of vanadium tetrahedra with a bridging oxygen between tetrahedra.
KVO3 submitted to the queue using Abinit as the computational package with the following parameters:
4 nodes, 8 ppn
10 10 10 Monkhorst-Pack grid
ecut 35 Hartrees
ngfft 54 120 60
Total energy converged after 14 SCF cycles
etot -2.1870788255E+02 Hartrees
total time :59:07:10 (6652 seconds per processor)
Hirshfeld-I calculations using the Abinit data are forthcoming.
Began setting up the VASP version of the KVO3 computation. I still need to talk to Josh about how VASP handles symmetry operations and how to convert CIF a,b,c,alpha,beta,gamma values into lattice vectors and a scaling constant. However, I believe that the INCAR, POSCAR, and POTCAR files are ready. The important parameters are listed below.
2 2 2 Monkhorst-Pack grid
Initial charge density from overlapping atoms
25 SCF cycles
Planewave cutoff 952 eV (35 Hartrees)
EDIFF 10-8 eV
Pseudopotentials used: K_pv, V_pv, and O.