set dielectric constant with the bjerrum length for coarse grain model

Hi dear users,
I want to simulate a large ion (using coarse grain model) between other negative and positive ions in a watery space. So as i know I must use coulombic potential in this form for my large ion:

Vcoulomb® =KbTℓb*qiqj/r^2

So I must define dielectric constant for coulombic interaction of large ion and other particles as C/(KbTℓb)

But there is two questions,

1- Can I define different dielectric constants in lammps? (One for interaction between large ion and other particles, and the other for the interaction between other ions)

2- I use real units, so I have my parameters as below:

C=about 336

Kb*T=0.596 Kcal/mol for T=300K

ℓb=about 7.1A for watery space

Therefore I determine dielectric constant equal to 79.4 which decreases the coulombic force between the large ion and other particles very much! So is it true? If no, what is wrong in my computations?

Thanks alot for your advices

Ly Xiang

Hi dear users,
I want to simulate a large ion (using coarse grain model) between other
negative and positive ions in a watery space. So as i know I must use
coulombic potential in this form for my large ion:

Vcoulomb(r) =Kb*T*ℓb*qiqj/r^2

So I must define dielectric constant for coulombic interaction of large ion
and other particles as C/(Kb*T*ℓb)

But there is two questions,

1- Can I define different dielectric constants in lammps? (One for
interaction between large ion and other particles, and the other for the
interaction between other ions)

no. you would need explicit potentials (eg. from tabulation) for that.
the dielectric constant is for the implicit solvent and thus global.

2- I use real units, so I have my parameters as below:

C=about 336

Kb*T=0.596 Kcal/mol for T=300K

ℓb=about 7.1A for watery space

Therefore I determine dielectric constant equal to 79.4 which decreases the
coulombic force between the large ion and other particles very much! So is
it true? If no, what is wrong in my computations?

have a look at:
http://lammps.sandia.gov/threads/msg02972.html
which gives a details description.

axel.

Dear Axel,
Thanks alot for your answer,
And what about when I simulate water explicitly in my model (for example spc moldel)?
In this case what is the value of dielectric constant? Is it 80 which is for watery space or not I must consider the coulombic interaction as in the vacuum space?
thanks alot

Dear Axel,
Thanks alot for your answer,
And what about when I simulate water explicitly in my model (for example spc
moldel)?
In this case what is the value of dielectric constant? Is it 80 which is for

1

watery space or not I must consider the coulombic interaction as in the
vacuum space?

the latter, since you have explicit solvent.

or to put it crude and simple: setting the
(relative) dielectric constant in practice means
to downscale all coulomb interactions with it.

axel.