Dear all,
we are planning to calculate the optical properties (the dielectric function) for a supercell of 100-200 atoms of a semiconducting material. We would appreciate some comments on
- the efficiency
- and accuracy
or any other aspect that comes to mind comparing VASP and Wien2k codes in that matter.
plz explain how can we get the dielectric functions by using WIEN2k and what is the procedure for obtaining the graphical representation.
Are you aware that a standard DFT calculation does neither give you the correct band structure nor does the linear response calculation give you the true excitation spectrum? For quantitative comparison with experiment, you need a GW calculation of the electronic structure and a Bethe-Salpeter calculation of the dielectric function. This is computationally super-expensive for large unit cells. You might want to have a look at https://berkeleygw.org/, which I haven't used, but it seems capable of doing this. Before thinking about accuracy differences between different state-of-the-art codes, I would rather consider what level of physical approximation you're willing to accept. If you're happy with standard DFT+linear response dielectric function, go for VASP, abinit, GPAW.... Personally, I prefer elk (elk.sourceforge.net), but a 200-atom cell is going to be cumbersome.
Hello Magdalena Laurien
Could you please tell me that what computer do you have or how many cores or processors does your computer has to calculate the optical properties (the dielectric function) for a supercell of 100-200 atoms of a semiconducting material.
- Badges
- Science topic
- Similar topics
- Filing