Dear researchers:
I think the pi-pi interaction may be understood as an interaction caused by electron-exchange and electrostatic effects (even so on).
But how can MD simulate this interaction without calculating wavefunctions when the electron-exchange really matters? Is it just by adding special potential functions?
Answers will be appreciated. :)
Yike Huang
An example of a system with pi-pi interactions would be graphene. The Tersoff-Brenner potential is traditionally used for carbon-carbon interactions in this system. You can look at this paper regarding parameters of this potential used to simulate phonon dispersion in CNTs and graphene:
https://journals.aps.org/prb/pdf/10.1103/PhysRevB.81.205441
If you look at the functional form of Tersoff potential, there is no explicit notion of hybridization. The only different between the sp3 carbons without the pi-pi interaction and the sp2 with pi-pi interaction is through the b_ij angle-dependent term. So, essentially, your guess is right. You would manipulate the potential function to distinguish between the sp2 and sp3 carbons.
This would only take care of the difference in stiffness of such bonds. If you are interested in electronic effects arising from the pi orbitals, or you want to be accurate about varying bond-order for example, you need to use ab initio MD which includes a representation of the electronic wave functions, and doesn't use empirical potentials such as Tersoff-Brenner.
- Badges
- Science topic