Query regarding using reduced LJ units for system with charges

LAMMPS LAMMPS Mailing List Mirror
1

I am simulating salt ions in implicit water using the restricted primitive model. The size of the ions is 4.25A. Along with electrostatics the ions also have an LJ potential truncated and shifted at the minima (1.12 sigma).

I am using reduced units in LAMMPS. What are the correct reduced parameters I should choose for this system for room temperature simulations (T=298K)? This is highly confusing as I am coming from an MC background where I could control the reference state used for non-dimensionalizing.

For example, I used to calculatE:

  1. T*=kT/epsilon where epsilon is the electrostatic energy at contact.
  2. epsilon = e^2/4 pi epsilon_o epsilon_r sigma
  3. Substituting the value of epsilon from 2 in 1 and by putting exact values ie
    e = 1.6 *10^(-19)
    sigma = 4.25 A
    epsilon_r = 78.5 ( for water)
    k = 1.3 10^(-23)
    T = 298 K.
    Using this I get T    =0.59

But LAMMPS tells T*=1.0 in reduced units for room temperature simulations as it uses epsilon_LJ to non-dimensionalize.

  1. The energy is reduced using epsilon; what is this epsilon in LAMMPS. Is it the epsilon_LJ?

Kindly help.

Thanks,
Mithun

2

why use reduced units for this and make your life needlessly complicated.
i would just use the real units settings and never look back.

axel

3

Dear Axel,
Thanks for your response. I agree with you, but could you still shed light on the query specifically

  1. What is the epsilon LAMMPS uses a) Is the epsilon_LJ or some arbitrary epsilon
  2. Does this epsilon change if I change epsilon_LJ?
  3. What is T* is should use for room temperature simulation in water?

We are using T*=0.59, epsilon_LJ =0.2, charge=±1, sigma=4.25A, and using reduced LJ in LAMMPS. with pppm.
Are we not modeling the system at room temperature?

4

Dear Axel,
Thanks for your response. I agree with you, but could you still shed light on the query specifically

  1. What is the epsilon LAMMPS uses a) Is the epsilon_LJ or some arbitrary epsilon

it is the epsilon that results in a value of 1.0 in the pair coeff command

  1. Does this epsilon change if I change epsilon_LJ?

no.

  1. What is T* is should use for room temperature simulation in water?

that is impossible to say, since it depends on what units of time, length and energy you would want to use to divide your parameters by in order to convert from absolute to reduced units.

We are using T*=0.59, epsilon_LJ =0.2, charge=±1, sigma=4.25A, and using reduced LJ in LAMMPS. with pppm.

sigma is not given in angstrom in reduced units but in multiples of your reference sigma

Are we not modeling the system at room temperature?

again, impossible to say since there is no fixed reference. you seem to be completely missing the point of reduced units. your are not simulating something at a fixed set parameters in parameter space but have releative parameters.

axel