Thermochemistry and solvation-energy parsing gaps (Gaussian, NWChem)

Part of the battery-research feature set. A prerequisite for any ΔG-based property.

What exists

ORCA: EZEROPT, ETHERM, ENTHALPY, ENTROPY and EGIBBS (orca.desc:282-320, orca.thermo).
MOPAC: H, S, Cp and ZPE, but not G.
NWChem: only ZPE (nwchem.desc:277).
Gaussian-16: none.
Solvation free energy is parsed for no code.

Gaussian-16 and -09 thermochemistry ("Sum of electronic and thermal Free Energies", the ZPE/H/G lines, T and P) into the existing keys.
NWChem H, S and G from its thermal analysis.
A new property for the solvation energy (ΔG_solv, or the gas/solvated split, where the code prints it): Gaussian SMD/PCM, ORCA CPCM/SMD, NWChem COSMO.
tests/parsers cases from real runs of each code, with values checked against the code's own summary lines.

2–3 days, including runs on niobium (g16, ORCA and NWChem are installed).

Overlay test: count the shaded blue the viewer draws for Mulliken cha…

Gaussian 16/09: ZPE, thermal energy, H, T·S, G and the temperature from a frequency run, in the keys ORCA uses. Checked against real g16 water outputs (HF/STO-3G, B3LYP/6-31G(d), SMD water). The G09 scripts are copies of the G16 ones, not run on a G09 output.
NWChem: ZPE, thermal energy, H, T·S, G read from the output file (the ecce_print trace has none of them); checked on a real scf freq run. NWChem prints corrections only; they are added to the energy printed after the Hessian.
MOPAC prints no Gibbs energy (checked with PM7 FORCE THERMO); unchanged.
Solvation: g16 and ORCA print no gas-phase total in an SMD/PCM run, so one job cannot give ΔG_solv. Two properties: ESOLV (E_sol − E_gas) where both are printed, so far NWChem COSMO; ESOLVCDS, the SMD cavity-dispersion term, for g16 and ORCA. ΔG_solv for g16/ORCA needs a gas-phase run beside the solvated one (a two-calculation result, see Multi-calculation results: an energy table and reaction/voltage profile across a project #175).
Not done: ORCA's CPCM dielectric term.