Pseudo potential question

[rishikanta_m]

Hi,

Would it be feasible to utilize the Pb POTCAR in lieu of the Pb_d POTCAR, given that it aligns well with the experimental lattice parameters and band gap values?

[marie-therese.huebsch]

Hi,

it depends on the properties that you want to compute. The Pb POTCAR is probably a good choice for computing the Young's modulus or running an ab-initio molecular-dynamics simulation. It is computationally cheaper than the Pb_d POTCAR as it has fewer valence electrons and requires a lower cutoff energy (ENCUT).

However, if you are interested in electronic properties that may emerge due to the d electrons, you need the Pb_d POTCAR as it treats the semi-core d electrons as valence states.

Generally, VASP uses a frozen-core approximation: We expect that many chemical and physical properties are not influenced by electrons that are very tightly bound to one nucleus in the material. Therefore, we call them core electrons and treat them as part of the external potential. The remaining electrons are called valence electrons, although they might actually be somewhat far below the Fermi energy. The valence electrons are treated within the electronic minimization as Kohn-Sham orbitals. There is no exact rule on which electrons should be considered as valence and which as core states. It depends on the chemical and physical property of interest whether the treatment of these states affect the result. The electronic states that are somewhat tightly bound but do contribute to some chemical or physical properties are referred to as semi-core states.

Does this answer your question?
Best regards,
Marie-Therese

[rishikanta_m]

Thanks a lot, Marie for the insights!!!