I have made a potential energy scan however it is wobbly, and I plotted out the individual energies of E_vdwl, E_coul, E_long, E_tail and found that it was E_vdwl that contributed to the wobbliness of the graph and I am not sure how to go about this from here.
i have increased the cutoff value from 9.0 to 13.0 lj and coulombic parameters. I have changed the pairstyle: lj/cut/coul/long and lj/charmm/coul/long. But the PES graph is still wobbly, not a smooth line. I have also tried changing the kspace style: trying out pppm and ewald.
To the extent that anyone can hazard a guess, this looks like the tiniest manifestation of “egg-box error” in the long-range Coulombic solver.
Having said that, it’s still impossible for us to know what’s happening and why. You would have to tell us what you are modelling in the first place (let alone the script commands) or we won’t be able to tell you anything informative in return.
I cannot reproduce this, so it seems that your way of computing a PES scan may be flawed.
Neither of these modifications should have any impact on the potential energy from the LJ interactions (= E_vdwl), assuming that the x-axis in your graph represents the distance in angstrom and the y-axis E_vdwl in kcal/mol:
your displayed distance is between 2.2 and 3.9 and thus well within either of the cutoffs
the same is true to when the switching function of lj/charmm is applied instead of the straight cutoff of lj/cut
since you are looking at E_vdwl, the coulomb style and the resulting energy (E_coul+ E_long) are not included
