Salt aggregration during simulation

GROMACS version: Lates
GROMACS modification: No

Hi I am simulating a bi layer structure of decane oil-water with AOS surfactant presence at the interface. The middle water layer has a NaCl concentration of approximate 12.11 wt%.
Force fields used is gromos54a7 as many research have cited this to be a good fit for my application.
After the 30 ns production run I am observing that Na and Cl atoms has formed form aggregates near the interface.

( green and blue dots in the middle are Cl and Na)

I am not sure why this is happening? Have I exceeded the maximum salt concentration due to which simulation has failed?
I am attaching my mdp file
pdr.mdp (2.4 KB)

Sodiums coordinating with the AOS at the interface looks reasonable. As for the NA / CL aggregation, appears that the interactions at this concentration is such that it is a more stable state for it to alternate Na - Cl - Na and not be hydrated.

1 Like

In a polymer-bring system, we observed a similar aggregation of Na molecules using the OPLSAA force field too. At higher salinity (salt concentration), aggregates of salts tends to form

If the force field has not been validated against scattering or osmotic pressure data at higher concentration, the parameters are likely not suitable for such simulations. Most biomolecular force fields parametrize ions based on hydration free energy and coordination number/structure and stop there since most biomolecules are not solvated in very high concentrations of salt so aggregation is not usually a concern. This has obviously not been a successful strategy in many force fields without substantial refinement. I do not know of any study that has looked at GROMOS parameters in the context of osmotic pressure.

1 Like

Thank you for reply.
I had ready your comments somewhere about the same and have seen in few papers too. Most of all Gromacs gives us a warning for this force field during mdrun too.
But I had ready many papers related to application to the petroleum industry where, people had used it at high concentration and have published results. I just don’t know how it worked for them.

A lot of people use force fields without really checking for crazy behavior. Beware, just because it’s published, doesn’t mean it’s right…

1 Like

Dear Sir
Can I try this approach mentioned by you in this thread using a different force field?
After reading the paper you mentioned and the post, It seems the modification is already included in the paper and it should work for high salt concentration for the charm force field, provided I can calibrate my system with this force field.

Assigning pair-specific nonbonded parameters is an approach that can be used with any force field. Simulations with high NaCl or KCl concentrations already work “out of the box” with CHARMM but not necessarily other force fields, for which parameters would need to be developed.

1 Like