Cations pairing during Molecular Dynamics

[Stagias]

Hello,

I am studying aqueous solvated ions and I’ve seen strange behavior with xTB-gfn2 (v6.4.1 on a 20 core Ubuntu PC). When I put two (or more) identical cations in a water droplet for a molecular dynamics simulation, they move together and stick to each other (notably K+), which is surprising because they have the same charge… I have also done the test at 600K in gas phase (with fewer water molecules) and the K+ ions also stick together. I have tried to switch back to gfn1 but it does not converge. I have also tried to use the alpb command to test if the water solvent model might solve this but it didn't.

I am therefore wondering if this is a problem with xTB or on my side, and if anybody has any suggested solutions or workarounds (I have seen in the documentation that it’s possible to keep the distance between certain atoms constant but this seems a bit tricky and wouldn’t really help if I want to add more ions I think…).

Attached input and output files for the gas phase calculation (liquid droplet available on request, I haven’t attached because it is much larger).

Any help or suggestions would be greatly appreciated,

Lucas
Masters Student, Nantes Ecole Centrale

Calculation with the gas phase :
-command line : xtb gas.xyz --charge +2 -- input md.inp --omd > outputmd.txt
-input :
gas.xyz.txt
md.inp.txt
-output :
outputmd.txt
trajectories.txt

Calculation with the liquid phase (unfinished):
-command line : xtb drop.gfn2.xyz --charge +2 -- input md.inp --omd > output.txt
-input :
drop.gfn2.xyz.txt
md.inp.txt
-output :
output.txt
trajectories.txt