Reduced charge (LJ) and FENE bonds in NPT ensemble

Hi,

I’m trying to simulate a polyelectrolyte polymer gel which I previously crosslinked using fix bond/create. I have added extra atoms (counterions) and have successfully read in all Atoms and Bonds. I have also put one charge per atom on the polymer backbone and one unit of opposite charge on the counterions.

Just to double check, +1e of charge in reduced units is equivalent to 1.602e-19/((43.148.85e-12)^0.5) LJ units?

I now want to equilibrate the gel in a water bath so I use a fix

fix 1 all npt temp 300.0 300.0 100.0 iso 0.0 0.0 1000.0
I start the simulation without charges just to equilibrate since the extra atoms were added at random and there may be an overlap, but when I turn on the chcharges and let the simluation run, I get first a FENE warning saying that the bonds are too long, then a FENE error saying that it's lost bonds (the system has only swollen 1% of it's original size at this point.)
Any ideas?

Sam

Hi,

I'm trying to simulate a polyelectrolyte polymer gel which I previously
crosslinked using fix bond/create. I have added extra atoms (counterions)
and have successfully read in all Atoms and Bonds. I have also put one
charge per atom on the polymer backbone and one unit of opposite charge on
the counterions.

Just to double check, +1e of charge in reduced units is equivalent to
1.602e-19/((4*3.14*8.85e-12)^0.5) LJ units?

I now want to equilibrate the gel in a water bath so I use a fix

fix 1 all npt temp 300.0 300.0 100.0 iso 0.0 0.0 1000.0

I start the simulation without charges just to equilibrate since the extra
atoms were added at random and there may be an overlap, but when I turn on
the chcharges and let the simluation run, I get first a FENE warning saying
that the bonds are too long, then a FENE error saying that it's lost bonds
(the system has only swollen 1% of it's original size at this point.)

Any ideas?

if you suddenly add the charges, you add a lot of potential energy
and move the system far away from equilibrium. remember that
coulomb is a rather strong and long-ranging interaction.

you, you probably need run a minimization to quench the system
down, do some MD with a reduced time step and perhaps a few
more quenches until everything is wiggled into place. if you still
run into problems, you may need to check, whether you have the
right balance between coulomb and lennard-jones.

cheers,
    axel.

The FENE error means a bond stretched
close to the hard FENE limit on bond length
(the formula blows up), which is typically due
to bad dynamics.

Steve