Lattice parameter wrong by 8% with FR PBEsol for KTaO3

[bretzel15]

Dear pseudo-dojo people,

thank you very much for your work here! I just found one problem with one of the pps from your project:

The PBE pseudopotentials for K, Ta and O seem to be fine for lattice parameter calculations, but trying the PBEsol versions, I found out that the FR pps minimize the energy at a lattice parameter of around 4.3 A, which is almost 8% off the experimental one of 3.9885 A (see attached figure showing the total energy difference versus different lattice parameters), calculated with QE 6.3.

Calculations with the SR PBEsol pps end up at 3.977 A, while both PBE pps end up around 4.02 A (see attached images, which is in reasonable agreement with the experimental one.

Do you maybe know why the FR PBEsol are off that far? Is this a known problem?

Best,
Tobias

---

PBEsol FR with problems

---

PBEsol SR

---

PBE FR