Hi,
I am using LAMMPS for modelling crosslinked DGEBA(Epoxy) and DETA(Resin). 80 DGEBA and 32 DETA molecules were packed in a 46A box using packmol. The CHARMM force field was used.The topology was updated using VMD. The formal charges were assigned manually.
The simulation box was then minimized and then equillibrated using 50ps NVT and then 400ps NPT at 600K.
Then,The fix bond/create command was used to crosslink the molecules.
From the literature it can be known that the charges of the molecules change after bonding.
How should the charges be updated after crosslinking or bonding has taken place?
Is there any software that can calculate and assign charges to molecules by reading and writing out a lammps data file?
Or should the lammps data file be changed to any other format?
Hi,
I am using LAMMPS for modelling crosslinked DGEBA(Epoxy) and DETA(Resin). 80
DGEBA and 32 DETA molecules were packed in a 46A box using packmol. The
CHARMM force field was used.The topology was updated using VMD. The formal
charges were assigned manually.
The simulation box was then minimized and then equillibrated using 50ps NVT
and then 400ps NPT at 600K.
Then,The fix bond/create command was used to crosslink the molecules.
From the literature it can be known that the charges of the molecules change
after bonding.
How should the charges be updated after crosslinking or bonding has taken
place?
this is a question for the charmm forums. the rules for how to do this
would be determined by the charmm forcefield conventions. LAMMPS
doesn't care, it uses what you feed it.
Is there any software that can calculate and assign charges to molecules by
reading and writing out a lammps data file?
(partial) charges and LJ parameters belong together, so wherever the
parameters for your molecules have been derived and documented, you
have to follow those rules and make sure they remain consistent. in
the most extreme case, you may need to reparameterize some of them
according to the same process the original set was derived. that is
standard procedure for all (first gen classical) force fields.
Or should the lammps data file be changed to any other format?
to what purpose?
axel.
Hi Axel,
Now, let us assume that the charges after and before the the reaction are known to me; the problem is that not only the charges of the atoms taking part in the crosslinking(whose charges can be easily manipulated since fix bond/create changes them to another atom type) changes but that of the nearby atoms(bonded to the atoms taking part in crosslinking) also changes.
The question now is that is there a tool that can work with lammps data file and assign known charges by may be searching for the atom with the new type; find its bonded neighbours; assign them charges?
Hi Axel,
Now, let us assume that the charges after and before the the reaction are
known to me; the problem is that not only the charges of the atoms taking
part in the crosslinking(whose charges can be easily manipulated since fix
bond/create changes them to another atom type) changes but that of the
nearby atoms(bonded to the atoms taking part in crosslinking) also changes.
The question now is that is there a tool that can work with lammps data file
and assign known charges by may be searching for the atom with the new type;
find its bonded neighbours; assign them charges?
none that i know of and there is nothing i know that would work automatically.
basically, you need to retype atoms which would also require to retype
bonds, angles, dihedrals and impropers, if present.
since you have somewhat simple polymers and thus little variation in
atoms types (compared to, say, proteins), one could consider writing a
script that does this by looping over all atoms and all bonds attached
to that atom and so on. since i am quite fluent in Tcl scripting and
using VMD, that is what i would use for that. There are some simple
examples for that using OPLS-AA on my homepage. But since this is very
much dependent on the force field and specific situation, there is
little hope to find/write a generic script. OPLS-AA is particularly
suitable for such scripting, since it has a simple increment system
for determining partial charges in alkanes. Commercial simulation
tools, OTOH, often have support for such automated atom typing, but
they often also require custom force fields that are more suitable for
that.
axel.
Hi,
Thank you for your help…
It was really helpful… I had used autohotkey scripting language… but later when I used kspace pppm, it gave a warning that system charge not neutral… may be there was some problem with the script…
anyways thanks again…
Dhritiman