I am trying to do a potential energy scan of a BTA (N,N',N''-trismethyl-benzene-1,3,5-tricarboxamide) dimer. I defined the coordinates of the COM (I choose 6 carbons of benzene ring for this) using GIC (generalized internal coordinates) and then doing a scan. For first few steps, when the COM-COM distance is less than 5 Angstrom it works fine, but as soon as distance increases the dimer adopts a random configuration (which I think it does to make sure to have stabilization via H-bond, not sure). I am not able to find solution to avoid this. Can anyone please help me?