Hello everyone,
I am trying to calculate the bond energy in respect to bond distance between two carbon atoms in a simple hydrocarbon molecule.
I am familiar with the fact that reaxff does not explicitly defines bonds so you cant use compute bond/local and that the compute [name] all pair reax/c computes the energy for the group “all” so it does no difference if i define the two atoms in a group and then use compute pair reax/c. I have also read the previous threads of the documentation and I have not found a proper anwer to my problem but the discussion is pretty old.
I was wandering if there is something new on the topic. I would like to extract a curve of bond energy vs distance for the two atoms in a molecule.
Thank you all in advance!
This kind of projection would be something you would have to program yourself into the source code.
But I think what you are trying to do is based on a flawed concept. Unlike with classical force fields where the different contributions can be separated and added, this is not the case with ReaxFF. There the potential function has a bond term (so you could actually compute that based simply of the functional form and the parameters from the parameter file; no need to use LAMMPS), but the bond topology itself is based on the bond order assignment (which changes with changing positions) and the interactions also contain Coulomb interactions that are subject to charge equilibration (which will also change with changing positions), so there are more contributions that change when you move two atoms in a bond apart. In a way this is not very different from doing a quantum mechanical calculation, where you also always compute the interaction of all electrons with all electrons and do not have a specific decomposition into individual components and making projections is difficult, because the change in potential energy between two atoms crucially also depends on the (relative) positions of all the other atoms.
ok ok i get what you say. Thank you Axel!