I am interested in calculating some redox potentials of organic compounds (H,C,N,O,F,Cl,Br) in organic solvents such as MeCN. I do not need absolute agreement with experiments since this is just a first pass to identify some potential candidates for experiments, so I am looking for the best simple approach to this problem. An accuracy of +/- 0.5V would be sufficient for example, if this is doable.
There are several protocols. Since you mention that you're mostly looking for trends, I'll give the faster method, e.g. by Jason Gilmore:
"Expanding and Testing a Computational Method for Predicting the Ground State Reduction Potentials of Organic Molecules on the Basis of Empirical Correlation to Experiment" J. Org. Chem. 2012, 77, 15, 6423–6430
- Optimize the geometry of the neutral and charged species in gas phase (e.g., a faster optimization)
- Optimize the geometry of the neutral and charged species in solution (e.g., using a polarizable continuum model for your solvent) with a good density functional and basis set.
The 2012 paper from Gilmore mentions optimizing in the gas phase and taking single point energies in solution.
I'd recommend at least:
- Using a "better" basis set (e.g, from the def2 or pc- series basis sets)
- Using dispersion correction (this likely won't make much difference but will give you better energies)
- Using, e.g. $\omega$B97X-D3 or some newer functional (although B3LYP/6-31G* still works remarkably well)
Anyway, the main point is this: Take the difference in total energy between the optimized neutral molecule ($S_0$) and the radical cation or radical anion ($D_0$) in the solution calculation.
This is the ionization potential for that species in solution.
It's not a full redox potential - for that you'll want a full thermodynamic cycle but it yields great trends that correlate very well for many organic redox potentials in acetonitrile.