I am running calculations in vasp. I have checked the planewave cutoff convergence to 400 eV. However, I also use 550 eV to make sure that it is "safe". In geometry optimization, the one that has 550 eV has lattice parameter longer than experiment by 0.1 Angstrom and the one that has 440 eV has better agreement (0.005 A longer than experiment). How does planewave cutoff energy affect the geometry optimization?
To calibrate the cutoff value you have to perform a series of single-point calculations of identical model (geometry, simulation settings) and prepare a plot: energy vs cutoff. The most reasonable choice is in the area, where increasing the cutoff value does not induce much change in energy. One has to keep in mind, that an optimal value of cutoff has to be determined - too high cutoff will slow down the calculations rapidly, while too low values will make the results useless.
Good luck!
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics