Calculations in the ground electronic state are probably the simplest cases where defaults in terms of charge (typically zero) and spin (singlet) are chosen by the calculation software. However, when dealing with reaction intermediates we encounter charged species with differing spin multiplicity. This radical may even correspond close to the first or second electronic state of the parent molecule, or a fragment and may require specified entry of the electronic state (as in CASSCF, RASSCF, etc).
What would be some good (general) resources to learn about modelling reaction intermediates (such as radicals). What are some of the caveats in such calculations. I am interested in methods such as DFT, MCSCF, MPn etc. for learning purposes.