Unify how we pass env variables to LAMMPS

[WardLT]

It's via the Runner for UFF and via Parsl for MACE. 🤦🏼

[WardLT]

Let's go with passing env variables via the Runner because that gives us the most flexibility with running the workflow

• Add a lammps_env attribute to MACERunner and use it when calling LAMMPS like we do in LAMMPSRunner (which, confusingly, only runs UFF): https://github.com/globus-labs/mof-generation-at-scale/blob/main/mofa/simulation/lammps.py#L162 https://github.com/globus-labs/mof-generation-at-scale/blob/main/mofa/simulation/mace.py#L276
• Pass the lammps_env argument from the HPC config when instantiating the MACE runner on start: https://github.com/globus-labs/mof-generation-at-scale/blob/main/run_parallel_workflow.py#L180