I am not sure, but using the MECP geometry between the two spin states may be a meaningful choice, at least it is not biased to any state.

I would like to estimate the magnetic exchange coupling parameter "J", and I was confused about which geometry should be used for the calculations. Should I be using the structure that corresponds to the high-spin state or the low-spin state, or can I use whichever structure was the lowerst in energy of the two? I am using ORCA for the calculations.

Many thanks,

Jano.