Dear Xiaowang and Lammps user,
I have a question about the tersoff potential for GaN system:
In the GaN.tersoff parameter file the value of ‘beta’ is taken different from that reported in the paper of (J. Nord, K. Albe, P. Erhart and K. Nordlund, J. Phys.: Condens. Matter, 15, 5649(2003)). E.g., in the file beta is taken as 1.0 for some place with Ga-Ga interaction whereas in some other place it is 0. This is completely different the paper. Similar observation can be noted for either Ga-N and N-N interactions. This is confused.
Is there any other source from where the values are adopted.
Regards,
Rajib
I think Aidan has said this is explained in the Tersoff doc pages,
but maybe this is a new question ...
Steve
I suggest you spend some more time examining the paper and the LAMMPS
documentation. If you positively identify an error in the GaN file, let us
know. Mere confusion on your part does not constitute an error on ours.
Yeah. The beta defined in the paper is a different parameter from that in LAMMPS documentation. Xiaowang
Is there any specific arguments for different ‘beta’ values documented in LAMMPS, than the original value reported in the paper?
Rajib,
Suggest comparing potential formulae in the paper and the LAMMPS document. It is not the value, but rather, the definition of beta, different (different papers use different symbols). In fact, the beta used in Albe’s paper is an exponential coefficient, which is different from many other papers.
Xiaowang
Rajib,
Remember, this pair style is called TERSOFF, and is based on a particular paper by J. Tersoff, as is clearly stated in the documentation. There is no pair_style called ALBE. We have gone to considerable effort to carefully document the meaning of the Tersoff parameters. The file GaN.tersoff kindly provided by Dr. Zhou shows how one can specify the potentials published by Albe et al., using pair_style tersoff. Nowhere do we suggest that we have magically replaced Tersoff’s beta with Albe’s beta. If this is somehow implied, let us know, and we will modify the offending text. If you are bothered by the fact that the Tersoff beta is not the same as the Albe beta, I suggest you write a new pair_style albe. If it is properly validated and documented, we will be happy to distribute it with LAMMPS. In the mean time, please read the definitions for pair_style tersoff provided in the documentation.
Aidan