What are the differences between Tersoff, AIREBO, SW (Stillinger Weber), and EDIP interatomic potentials? If a MD simulation provides reasonable results with all of these interatomic potential, which one would be the best to work with? Are there any inherent advantages of these interatomic potentials over others?
Hi Tousif,
Mohamed has summarized the essence of the logic in identifying the correct interatomic potential. However, another criteria that needs to be examined is looking at the performance of these interatomic potentials for the basic material properties (thermal, mechanical and predicting correct phase) that are of interest to your problem.
Hope this helps you.
Ilaksh
Yes, there are many differences not only in accuracy, but also in the physics or the chemistry behind the problem needed to be solved. for example if you would simulate chemical reaction at the surface of a given material you need to use a reactive potential. if your material is covalent it appears that Tersoff . BOP or SW can describe its properties in a good manner ( vibrational, thermal, and structural prperties).
so, this depends on what you need to simulate.
Thanks so much for the answer. Any good reference to have a deeper knowledge on these?
Hi Tousif,
Mohamed has summarized the essence of the logic in identifying the correct interatomic potential. However, another criteria that needs to be examined is looking at the performance of these interatomic potentials for the basic material properties (thermal, mechanical and predicting correct phase) that are of interest to your problem.
Hope this helps you.
Ilaksh
Thanks. Actually I'm getting into MD and atomistic studies very recently. I appreciate all helps. :)
I would agree with both of the previous answers, as it really depends on your application and material. My laboratory, for example, is currently comparing the energies of gases scattering from silicon at a wide range of temperatures using different empirical potentials, which has certain implications on the potentials we can use. Given that you are mostly considering bond-order (or similar) potentials, you are probably considering either silicon or carbon. For silicon, the following papers (and the references therein) have some discussion of the differences between the potentials which you may find useful:
Article Comparative Study of Silicon Empirical Potentials
Article Comparison of silicon potentials for cluster bombardment simulations
Article Two-phase simulation of the crystalline silicon melting line...
For metals, you should also probably consider the embedded-atom model (EAM)-like class of potentials. But, again, it really depends on the material and application.
You are welcome! I know it was a bit late, but I hope it was still helpful! Let me know if you have further questions. We are currently working on comparing certain aspects of these potentials as well.
- Badges
- Science topic