Dear LAMMPS Users,
For TIP4P water model, we always use lj/cut/tip4p/long, which can describe water dynamics very well. However, when I add ions, something strange happens. The ions aggregate together, instead of homogeneous solution.
are you certain, that you have assigned the correct charges and potential parameters for your ions?
After checking, I think the problem is the implement of interactions between the massless site and charged ions.
what did you check, and why do you come to your conclusion?
The following is my code, and I tried two types, all are completely wrong.
please be careful with such bold statements. you are talking about a part of LAMMPS that has been used a LOT by quite a few people. thus the chances that this pair style would be incorrect is extremely low. people would have noticed. it is by far more likely, that you are overlooking something.
Code 1real units
pair_style lj/cut/tip4p/long 4 1 1 1 0.1546 12.0 10.0 #otype htype btype atype qdist cutoff
kspace_style pppm/tip4p 1.0e-7
kspace_modify order 7
providing only the LAMMPS input script is insufficient to document a problem. there could be mistakes in the data file as well.
[…]
Is there anyone who has experience about tip4p water solutions? From my understanding, the implement of pair_style for M-Ions is important, which will affect the dynamics of solution. How should we realize this in LAMMPS? It seems that there is no inf. about this point in MANUAL.
the info in the manual is sufficient for using LAMMPS for the systems you have in mind, if you want to know how exactly the handling of the M site is realized, you must read the source code.
BTW: you can avoid the tip4p pair styles completely, by building an explicit 4-site water model, give the M site a miniscule mass (1.0e-10) and then use fix rigid/small for time integration.
axel.