Parameterization for SPC/Fw water model

User discussions
1

Dear all,
I’d want to use the SPC/Fw water model to model an air-water interaction. To begin, I must construct a well-equilibrated cubic box containing some water molecules. However, when I try to minimise the box’s energy , I am unable to do so.
I am using the following parameters to do:
define = -DFLEXIBLE
integrator = steep
nsteps = 100000
nstenergy = 500
nstlog = 500
nstxout-compressed = 1000
constraint-algorithm = lincs
constraints = h-bonds
cutoff-scheme = Verlet
coulombtype = PME
rcoulomb = 0.7
vdwtype = Cut-off
rvdw = 0.7
DispCorr = EnerPres

Can someone please guide me about the parameters for this step step.

2

What do you mean with “unable to do so”?

Minimizing the energy of any liquid is impossible in practice. The real minimum would be a solid.

3

Hi @hess
I meant that the potential energy of my system doesn’t converge on using these parameters.
Initially, I tried energy minimization using the mentioned parameters. The potential energy plot obtained from this run is shown below.

energy1
energy1845×790 7.16 KB

In the next step, I tried to equilibrate by using an NVT ensemble, but this step is not producing any data. (nvt.tpr file doesn’t contain any info about energy). The parameters for the NVT equilibration run are given below.
integrator = md
dt = 0.002 ; 2 fs
nsteps = 50000 ; 100 ps
nstenergy = 200
nstlog = 2000
nstxout-compressed = 10000
gen-vel = yes
gen-temp = 298.15
cutoff-scheme = Verlet
coulombtype = PME
rcoulomb = 0.7
vdwtype = Cut-off
rvdw = 0.7
DispCorr = EnerPres
tcoupl = Nose-Hoover
tc-grps = System
tau-t = 2.0
ref-t = 298.15
nhchainlength = 1

4

As I said, energy minimization of a liquid will never converge in practice.

And you can not integrate a model with non-constrained bonds involving hydrogens with a 2 fs time step. You will likely need a 0.5 fs time step.

5

Thanks hess