For a beginner of the vasp package user, this question may arise. Which pseudopotential should I chose?
Here is the VASP official site page describing this qustion.
Here is the VASP official site page describing this qustion.
Which one should I use?
Point #1:
The short answer is, it is based on your calculation. Yes. What are you calculating first of all?
For example, in conductivity studies, importance should be given to the band curvatures. Therefore, always check a known property of the system before doing the actual calculations.
Point #2
You need to decide which core electrons should be treated with pseudo-potentials. The less and less number of core-electrons you replace with pseudo-potentials, more accurate (generally) your results. However, it is at the expense of computational cost. You should have an intuition on which pseudo-potential should be used (where the accuracy and the cost are balanced within your requirements)
I am striving to make this site as a useful site for the beginners in computational researchers in physics, chemistry and materials science. Please leave comments to improve or more points to add here. Thanks for your support.
The short answer is, it is based on your calculation. Yes. What are you calculating first of all?
For example, in conductivity studies, importance should be given to the band curvatures. Therefore, always check a known property of the system before doing the actual calculations.
Point #2
You need to decide which core electrons should be treated with pseudo-potentials. The less and less number of core-electrons you replace with pseudo-potentials, more accurate (generally) your results. However, it is at the expense of computational cost. You should have an intuition on which pseudo-potential should be used (where the accuracy and the cost are balanced within your requirements)
I am striving to make this site as a useful site for the beginners in computational researchers in physics, chemistry and materials science. Please leave comments to improve or more points to add here. Thanks for your support.
No comments:
Post a Comment