Hi all
I have a question regarding the cut off used for long range electrostatics.
System Description:
I am trying to study the interaction of an electrolyte with a negatively charged interface.
Two charged interfaces parallel to the x-y plane are separated by a distance Z. The space between the interfaces is filled with an electrolyte consisting of positively and negatively charged ions.
The system is charge neutral and periodic in the x,y direction.
There is a reflecting wall on either end of the box in the Z direction.
System Dimensions :
83Å in the X and Y direction and 300Å in the Z direction.
Potential :
All particles interact using the following force field parameters.
pair_style lj/cut/coul/long 3.0 10.0 # 40.0 or 80.0
dielectric 78.0
kspace_style ewald 1.0e-6
kspace_modify slab 3.0
QUESTION :
For the same initial configuration of the system when I change the ewald cut off from 10 to 80, the ion distribution along z direction is remarkably changed.
I am confused as to why this happens? Why should the cut off affect the ion distribution at all? Which one do I regard as the correct one?
Please find a plot attached which shows the results of the positive ion distribution (top) and the negative ion distribution (bottom) along the z axis. Results are for the same system with different cutoffs.
(I can provide with the input scripts if required)
You should know that the precision you are setting (1e-6) is valid given certain assumptions and even though LAMMPS tries to choose good Ewald parameters, it might also fail to do so. If those assumptions break down you might have to set the ewald sum parameters yourself (via kspace_modify)and make sure that the ewald sums converge.
Hi all
I have a question regarding the cut off used for long range electrostatics.
System Description:
I am trying to study the interaction of an electrolyte with a negatively
charged interface.
Two charged interfaces parallel to the x-y plane are separated by a distance
Z. The space between the interfaces is filled with an electrolyte consisting
of positively and negatively charged ions.
The system is charge neutral and periodic in the x,y direction.
There is a reflecting wall on either end of the box in the Z direction.
System Dimensions :
83Å in the X and Y direction and 300Å in the Z direction.
Potential :
All particles interact using the following force field parameters.
pair_style lj/cut/coul/long 3.0 10.0 # 40.0 or 80.0
dielectric 78.0
why do you set dielectric to 78?
i suspect that ewald doesn't support this properly currently.
Hi all
I have a question regarding the cut off used for long range electrostatics.
System Description:
I am trying to study the interaction of an electrolyte with a negatively
charged interface.
Two charged interfaces parallel to the x-y plane are separated by a distance
Z. The space between the interfaces is filled with an electrolyte consisting
of positively and negatively charged ions.
The system is charge neutral and periodic in the x,y direction.
There is a reflecting wall on either end of the box in the Z direction.
System Dimensions :
83Å in the X and Y direction and 300Å in the Z direction.
Potential :
All particles interact using the following force field parameters.
pair_style lj/cut/coul/long 3.0 10.0 # 40.0 or 80.0
dielectric 78.0
why do you set dielectric to 78?
i suspect that ewald doesn't support this properly currently.
strike that. i misread the code.
sorry for the confusion.
Ewald and PPPM should give the same answer,
always. And they both use the dielectric constant
(as a simple scale factor). Whether your script
should set it to 78 is a different question.