Update GTH pseudopotential database

[mkrack]

• Script efficiency (speed) improved
• Full sync with current GTH-PP database performed

[mfherbst]

Thanks @mkrack !

[mfherbst]

By the way we are not hung up on the naming convention. In particular long-term the plan is to move to a central pseudopotential registry in the Julia ecosystem (https://github.com/JuliaMolSim/PseudoLibrary). So if there are more/other GTH-type potentials worth having available, over there would be a good place to put them.

[mkrack]

@mfherbst Thanks for merging.

The GTH pseudopotentials in https://github.com/cp2k/cp2k-data/tree/master/potentials/Goedecker/cp2k have been generated in continuation of the work of C. Hartwigsen et al. on the LDA parameter sets based on the same recipes and codes for other exchange-correlation functionals, especially GGA (see Krack 2005). The corresponding GTH parameter files including the spin-orbit coupling (SOC) parameters have been release only recently, but, if I am not mistaken, SOC is not yet a feature supported by DFTK, but might be of relevance in the future.

Alternative versions of GTH PP parameter sets are indeed available which have been optimized using the atomic code ATOM integrated as a module in CP2K.

It makes sense to keep the parameter files in a separate registry. I suggest just to mirror the relevant xcf sub-folders of the cp2k-data repository in the case of the GTH (HGH) potentials.