Hi!
I’m trying to create a LAMMPS datafile using the OPLS-AA force field for an Ionic Liquid by using the parameters from this paper.
The torsional part of this ff is given as:
The improper coefficients in the SI of the paper are given in this form:
How do I interpret the X in these “X-C/N-X-X” improper? and what improper style should I use for these kinds of improper?
X in this context usually means any atom type. The non-X atom type would be the central atom
You could either define them as dihedrals using dihedral style opls or use improper style cvff with d=-1 and n=2 and K=\frac{1}{2}V_2 since the other fourier coefficients are zero.
While writing the impropers in the datafile as “n improper_type I J K L”; would the atom I be CW/CR/NA?
Also, I there are 3 atoms (J,K,L) bonded to I, does the order of writing them?
Please carefully study the documentation for the improper style in question.
Hi Alex, could I double check that is the 1/2 term for avoiding double counting, e.g. atom pairs 1234 and atom pair 4321? So in such that the special_bonds lj/coul for OPLS scaling is not applied to improper terms? Many thanks, Catherine
Pairs???
Where does it say that improper terms are subject to special bonds scaling factors?
4 atoms considered in dihedral group I meant. No just thinking about where the 1/2 term comes from
What 1/2 term?
You didn’t answer my other question.
No, just thinking about where the 1/2 term comes from in the (K=1/2 V2). While now got it, it’s from the torsion energy expression and the cvff style doesn’t not include 1/2.
As is discussed in the documentation of most such styles it is entirely a matter of convention whether the this 1/2 factor is explicitly included in the functional form of the potential or implicitly as part of the force constant. Both styles include this factor, just handle it differently, which then means that the force constant must be scaled accordingly if there is a difference between explicit or implicit use of the factor.
Yes it’s clear to me now. Thanks
