If you know the initial and final states then you can use the NEB go calculate the pathway. Just make sure that the gas phase molecules are separated in the initial and/or final states. Also, since the pathway may be long, in terms of the total atomic motion, make sure to use a sufficient number of images to resolve the reactive part of the path. Finally, use a conservative optimizer such as IBRION / IOPT = 3 at the start of the calculation to relax the path until the forces are below 0.5 eV/Ang.
You can use the dimer method when you have qualitatively identified the saddle. Don't use it away from the saddle because the gas phase system will have zero modes and the dimer needs a negative mode to find the saddle.
Using NEB method with gas-phase species
Moderator: moderators