[lammps-users] How to add a numeric term to Morse potential function in LAMMPS

[mamiri]

Hello all,
I need to use a potential developed in the form below, which basically consists of two terms; a BKS (Coulombic and Buckingham) and a Morse term.

I am going to use this potential for simulation of Ta2O5 oxide. I am using hybrid/overlay pair_style to add the BKS and Morse terms for Ta-O interactions. The problem I have is that as we can see the Morse term, D_ij(1-exp(-a_ij(r_ij-r_e))² = D_ij + D_ij [exp(-2a_ij(r_ij-r_e)) –2exp(-a_ij(r_ij-r_e**))], has an extra Dij term as compared to what is mentioned in LAMMPS documentation for Morse potential. As such, when I give the required constants (Dij, aij and re) for the morse term it does not include the extra Dij term in my simulation.

I am using the following command lines to use this potential in my simulation. However, this is obviously not including the Dij. Is there a way to incorporate that term in the simulation?

pair_style hybrid/overlay morse 6.8 buck/coul/cut 6.8 10
pair_coeff 1 1 buck/coul/cut 0 1 0 6.8 10 #Ta-Ta
pair_coeff 1 2 morse 0.3789 1.6254 2.5445 6.8 #Ta-O
pair_coeff 1 2 buck/coul/cut 100067.01 0.1319 6.05 6.8 10 #Ta-O
pair_coeff 2 2 buck/coul/cut 1388.77 0.3623 175.00 6.8 10 #O-O

FYI, the table below provides the constants for the Ta-O pair for both the BKS and Morse terms.

Thank you,
Best regards,
Mahsa Amiri

[akohlmey]

There is no need to add this constant term to the morse potential since a) the absolute energy in classical potentials is arbitrary, only energy differences for different geometries matter and b) it will vanish when taking the derivative and thus has no impact on the forces or stress.

Axel.