QMMM in GROMACS using wB97x/6-31G

[Lukas Müllender]

Hi all!

I'm currently trying to perform a QMMM simulation of a simple system of phenol in water, just to get some reference values to compare with results obtained with neural network potentials. These potentials were trained on data obtained using the wB97-X functional and 6-31G basis sets.

I'm using the CP2K interface in GROMACS to perform these simulations, which by default supports PBE and BLYP functionals. Now, I'm very new to CP2K and QMMM calculations in general, so I struggled quite a bit to adapt the CP2K input file to use a different functional and basis set. After some failed attempts, where I got running simulations but wrong results, and consulting other posts on this forum, I think I arrived at something close to correct. However, before starting another time- and resource-consuming calculation, I could use another pair of (more experienced) eyes to tell me if I'm using sensible parameter values. In particular, I'm very unsure about the XC and HF sections, where even after reading the [original paper](https://pubs.aip.org/aip/jcp/article/128/8/084106/71104/Systematic-optimization-of-long-range-corrected) I couldn't figure out if I'm using the correct parameters or missing something. 

I would appreciate any comments or advice!

Cheers,

Lukas