Kawamura Three-body potential in MD simulation

[Jin You Lu]

Hello~

I am trying to implement a three-body potential, as shown the equation below or in the attached file.
V_ijk3(r_ji,r_jk,θ)=-f{cos(2(θ_ijk-θ_o))-1}(k₁k₂)^1/2
k1=1/(exp[g_r(r_ij-r_m)]+1)
k2=1/(exp[g_r(r_jk-r_m)]+1)

https://www.researchgate.net/publication/29773387_Physical_Properties_of_Clay_Minerals_and_Water_By_means_Molecular_Dynamics_Simulations?enrichId=rgreq-dfafdda917aaf583a0706dddf4c0f460-XXX&enrichSource=Y292ZXJQYWdlOzI5NzczMzg3O0FTOjI2MDg1Nzk2NzkzNTQ4OEAxNDM5MjA0Nzk0NDM5&el=1_x_3

Is there internal potential which can be used to describe the equation?
Thanks

[Julian Schneider]

Unfortunately, such a potential term is currently not implemented in ATK-Classical.
One could add it as a python hook function, though.
From the paper I see that this term is only used to model H-O-H interactions. Is that true for your system, as well?

[Jin You Lu]

Dear Julian Schneider
For my current system H2O on calcite (CaCO3),
I need three body potential to describe, like H-O-H and C-O-C, and the parameters i can get from the supporting Information published in J. Phys. Chem. C, 2014, 118 (6), pp 3078–3087
http://pubs.acs.org/doi/suppl/10.1021/jp411151u/suppl_file/jp411151u_si_001.pdf

For other two-body potentials, i would try to use potential terms implemented in ATK-classical^^.
Thanks

[Julian Schneider]

Ok, I'll see if I can prepare a script that includes these interactions for your system. Might take a day or two, though.

[Jin You Lu]

Dear Julian Schneider

Do you have any update for that?
Thank you very much.

[Julian Schneider]

It's unfortunately a bit more complicated than expected.
I will need about a week more to implement it.