This page gathers links to miscellaneous norm-conserving pseudopotentials that can be used by ABINIT, but are not among the recommended ones. Some of these (e.g., HGH) have been tested, but many others not. Although not recommended, they might still be perfectly fine, particularly for selected usages not permitted by the recommended ones.
For some elements, a small core pseudopotential and a large core pseudopotential are available. These pseudopotentials do not include non-linear core corrections and can be used to perform mGGA calculations with the Tran-Blaha 09 XC functional, Phys. Rev. Lett. 102, 226401 (2009), also called modified Becke-Johnson. These pseudopotentials have been used in the publication by Lejaeghere et al, Science 351 (6280) (2016) and correspond to HGH/ABINIT. The pseudopotential for the Ta-Au series, Ru, Rh, Pd, Cr, and Mn should be avoided. Others are good or often very good.
These pseudopotentials have been used in the publication by Lejaeghere et al, Science 351 (6280) (2016) and correspond to FHI98pp/ABINIT. Some pseudopotentials are awful and completely screwed the result (W -especially-, Fe, Co, Ni, Ti, V, Tl, Ta, Re, Os, Ni).
These pseudopotentials do not include non-linear core corrections and can be used to perform mGGA calculations with the Tran-Blaha 09 XC functional, Phys. Rev. Lett. 102, 226401 (2009). They have been tested, but not as carefully as the ABINIT recommended pseudopotentials. Following the Delta project analysis for similar GGA pseudopotentials, the pseudopotential for the Ta-Au series, Ru, Rh, Pd, Cr, and Mn should be avoided. Others are good or often very good.
Pseudopotentials generated with FHI98PP have to be complemented by a header: How to modify pseudopotentials generated using the FHI (FHI98PP) code?
