Water piercing tube as temperature increases

User discussions
#1

GROMACS version:
GROMACS modification: No
Here post your question

Hello all

I have a tube of iron neutral atoms and a block of water inside. At 100K simulations the water stays inside the tube, but at 300K the water will break through the tube and escape. I thought to change something about the parameters itp file but am not sure where to start. What needs to be changed so the atom wall becomes a barrier to the water?

Thanks

#2

What parameters are you using for iron? Can you provide your complete .mdp file? You’re not excluding any interactions, are you?

#3

; ----
; Built itp for single.mol2
; by user vzoete Tue Jun 23 16:11:59 CEST 2020
; ----
;

[ atomtypes ]
; name at.num mass charge ptype sigma epsilon
FEP3 26 55.8470 0.0 A 0.115817 0.000000

[ pairtypes ]
; i j func sigma1-4 epsilon1-4 ; THESE ARE 1-4 INTERACTIONS

[ moleculetype ]
; Name nrexcl
single 3

[ atoms ]
; nr type resnr resid atom cgnr charge mass
1 FEP3 1 LIG FE 1 0.0000 55.8470

[ bonds ]
; ai aj fu b0 kb, b0 kb

[ pairs ]
; ai aj fu

[ angles ]
; ai aj ak fu th0 kth ub0 kub th0 kth ub0 kub

[ dihedrals ]
; ai aj ak al fu phi0 kphi mult phi0 kphi mult

[ dihedrals ]
; ai aj ak al fu xi0 kxi xi0 kxi

#ifdef POSRES_LIGAND
[ position_restraints ]
; atom type fx fy fz
1 1 100000 100000 100000
#endif

This is the itp being referenced by the topology file

#4

mdp here:

define = -DPOSRES_LIGAND
constraints = all-bonds
integrator = md
dt = 0.002 ; 2fs
nsteps = 50000
nstxout-compressed = 1000
nstlist = 50
ns_type = simple
cutoff-scheme = verlet
comm_mode = linear
coulombtype = cut-off
vdwtype = cut-off
rlist = 2
rcoulomb = 2
rvdw = 2
pbc = xyz

;Generate velocites
gen_vel = yes
gen_temp = 10

Tcoupl = nose-hoover
ref_t = 100
tau_t = 0.04 ; changed to match pull rate
tc-grps = system
nsttcouple = 1

;Pressure coupling
Pcoupl = NO

#5

Using ε = 0 means there is no LJ potential between the iron and anything else. And since you’re modeling them as atoms, not ions, they aren’t charged, so the water molecules never feel any influence of the iron. At low temperature, the kinetics are simply slower but given enough time, will behave exactly as the higher temperature system.

Looks like you’re co-opting the CHARMM heme iron parameters for this. That’s not appropriate. The heme iron atom type is explicitly for heme, in which the iron is charged and the electrostatic interactions dominate all interactions (and are generally designed to be covalently bonded to six ligands).

#6

Thank you I will work with that!