Dear all. I’m uesing lammps to simulate SiO2/Ag NPs，now I have encountered the problem of potential setting. I want to use EAM in Ag, Tersoff potential function in SiO2 and LJ potential function between this two. I don’t know whether this code is correct, like:
pair_style hybrid/overlay tersoff eam lj/cut 8
pair_coeff * * tersoff SiO.tersoff Si O NULL #Si-O
pair_coeff 1 3 lj/cut 0.21413 3.39 #Si-Ag lj
pair_coeff 2 3 lj/cut 0.04755 2.85 #Ag-O lj
pair_coeff 3 3 eam Ag_u3.eam #Ag-Ag
( atom type 1-Si,2-O,3-Ag)
Thanks!
What do your mean by “correct”?
It is easy to test if the syntax is correct: just run it and see if LAMMPS complains.
Conceptually, there is the issue that you really want to use pair style hybrid and not hybrid/overlay (overlay is for adding multiple potentials for the same atoms) for your kind of system, however, your input uses settings where both options would give the same result. So this is only a formality at the moment.
However, if you want to know whether this is a meaningful model, with adequate parameters and thus suitable for the simulations you want to perform, then there is no alternative to making test simulations and comparing the results to reference data, e.g. from other simulations/publications or experimental properties.
Another step to do is to check is whether the parameters that you have chosen are suitable for the material, the kind of geometry (some potentials are only suitable for bulk systems, for example, and not for nanoparticles) and the conditions. That you would do by looking up the publications describing the potential parameters and how they were devised and then looking at the descriptions and possible applications that were discussed by the authors of those publications.
Sorry for my improper use of words. Actually I want to ask if the syntax is used correctly.
As you said, should hybrid/overlay be changed to hybrid? I also wonder if the syntax of EAM is used correctly?
You can change it, but it won’t make a difference given the pair_coeff settings you have.
Like I said, just run the input and see whether LAMMPS complains. That would be much faster than writing and reading all those forum messages.