UT theoretical chemistry code forum

Hi Everyone,
I am simulating oxygen interstitial diffusion in fluorite structures. The dominant mechanism is the interstitialcy mechanism but I am having trouble figuring out the saddle point configuration. As I understand it, the configuration should be 2 interstitials with a vacancy in the middle. But I do not know the final position to do a NEB calculation. Can somebody help me figure this out ? Any help would be appreciated.
Thank You.

Besides trying to determine a final state from your intuition about the reaction mechanism, you could either run dynamics at a high temperature to find nearby minima, or run min-mode following saddle point searches from your initial state.

Thank You for the suggestion. I tried optimizing the structure with one interstitial in the vacant octahedral position at a higher temperature using GULP. But the interstitial goes back to the vacant site. I will now try the dimer method to get the diffusion path.