UT theoretical chemistry code forum

Dear all,

I'm doing CI-NEB calculations on a CO2 molecules diffusing LINEARLY along an axis, ie O2-C-O1 at the starting point becomes O2-C-O1 at the ending point and would like to do the same calculation but with the CO2 molecules rotating between the starting and ending points, ie O2-C-O1 becomes O1-C-O2.
I have thus inversed O2 and O1 in the ending point and ran the nebmake.pl script, nonetheless when looking at the resulting movie (from the POSCARs with nebmovie.pl 0), the CO2 molecule is not rotating, the oxygens remain on the axis; for the first half of the trajectory, the CO2 molecule remain O2-C-O1 and the distance between atoms decreases then in the middle, O2, O1 and C are overlapping then in the second half, the molecule becames inverse and then distances increase to give O1-C-O2 in the ending point. I've attached a small figure to explain the following.
I've also used the nebavoid.pl but it is correcting the distances but not implementing a rotation.

is there a way around in order to create images so the CO2 molecule is rotating?

best regards
Hub

A simple and crude solution is you make your full path via two half paths through the intermediate in which the molecule is vertical. You can then stitch them together, manually, to make your full path.

A much cleaner solution is to work in internal coordinates. The group of Hannes Jonsson has recently done this and has provided code within ASE. It should have the name NEB-TR.

thank you for you help.
Regards