I have a simple case containing water molecules and Zn2+ and SO42- ions. The system was equilibriated, and now I’m using NVT for data collection and statistical analysis (File attached).
I do not see movement for the ions in the dump file. While, the water molecules undergo movement. I use the command: “compute msd”. At first I used “com yes”, then I removed it and recalculated the MSD with the rerun command. The MSD of all species looks normal, but the MSD of Zn2+ looks very strange. I searched in MATSCI with ion movement as a keyword, but could not find a possible cause. Can anyone tell me what details I’m missing or what keywords I can change to search?
Dr. Kohlmeyer previously pointed out that my LAMMPS version is too old (LAMMPS (31 Aug 2021)). I’m sorry that I can’t upgrade it at the moment due to some limitations.
Any comments will be appreciated.
I have attached the documents that may be required.
These two statements are contradicting each other:
It seems that your ions are moving.
What I would do is compare diffusion coefficients from the MSDs for all species with some expected values from the literature. Note that there are best-practice that must be followed when measuring MSD, such as minimum system size, proper thermalization/equilibration at a given pressure/volume, ect., and I have no guaranties here that you followed them here…
You are using a TIP3P water potential, presumably rigid? But you have not rigidified your water molecules in any way.
More importantly you are running charge-based simulations without long-ranged electrostatics, which is highly unlikely to give you physical results.
Have you attempted to visualise your trajectory? The Zn MSD is reminiscent of step motion but at 300K the TIP3P force field should give you liquid water. (Then again any pathological trajectory behaviour would be attributable to the issues above.)
