(I added a wrong title to my first e-mail, Gamma parameter in fix qeq/reax I apologise for this)
Dear lammps users and Ray Shan,
I am trying to use QEq to calculate charges of atoms. When the Taper upper radius is 10 A the charges are meaningless (I mean they are so high, +3, -5, etc). When I decrease its value the charges are better.
I also tried to reproduce the QEq charges of some molecules (H2O, CH4, HF) reported in the original paper of QEq (The Journal of Physical Chemistry 95 (1991) 3358-3363) and I got the similar results when the upper Taper radius is around 2 A.
I did all simulations with k-space switched off.
What should the upper Taper radius value be?
cheers.
baris
(I added a wrong title to my first e-mail, Gamma parameter in fix qeq/reax I apologise for this)
Dear lammps users and Ray Shan,
I am trying to use QEq to calculate charges of atoms. When the Taper upper radius is 10 A the charges are meaningless (I mean they are so high, +3, -5, etc). When I decrease its value the charges are better.
I also tried to reproduce the QEq charges of some molecules (H2O, CH4, HF) reported in the original paper of QEq (The Journal of Physical Chemistry 95 (1991) 3358-3363) and I got the similar results when the upper Taper radius is around 2 A.
That does not seem reasonable. Taper radius is essentially the cutoff and 2A is unreasonably short. Likely reasons for not getting the reported charges are 1) if the QEq parameters are inputted correctly, 2) structure, and 3) difference in pair_styles used. Mind you, with different pair_styles for short range interactions, there may be a difference in bond length and bond angles and these differences can result in different charges.
I did all simulations with k-space switched off.
But we don’t know the else.
What should the upper Taper radius value be?
It should be as long as the longest cutoff, which is 10A.
Ray