[lammps-users] pure ethanol

system · November 10, 2008, 5:17pm

Dear lammps-users,

To match the result of a Materials Studio simulation of pure ethanol
using PCFF with the results of a LAMMPS simulation, I translated the MS
files into a LAMMPS data file using msi2lmp. With this data file, I
first ran a minimize step to find its lowest energy, and then did an NPT
run for 100 ps at 298 K and 1.0 atm. The final box size of the
simulation was 22.12 Angstroms, which gives a density of 0.902 g/cm3 for
the solution. At this temperature and pressure, the density of ethanol
should be closer to 0.789 g/cm3.

Does anyone know if PCFF is supposed to accurately simulate the
properties of ethanol at this temperature and pressure. If not, what
corrections may be made to the force field to improve the simulations?

Thank you,

Paul Harten

_Vikas_Varshney2 · November 10, 2008, 6:25pm

Dear Paul,
Does material studio gives you the correct density?
Just on the side note which may help you. For PCFF force field, you need to incorporate 1-4 interactions as non-bonded in lammps. That is what I found out when I tried to match energies in MS and LAMMPS.

Regards,
Vikas

system · November 10, 2008, 8:58pm

Dear Vikas,

A colleague of mine who did a Materials Studio MS NPT run on pure
ethanol got a density close to 0.78 g/cm3 at 298 K and 0.0 atm.

In your side note, you suggest that for the PCFF force field I need to
incorporate 1-4 interaction as non-bonded in LAMMPS. What are the 1-4
interactions you are referring to?

Thanks,

Paul

             "Vikas Varshney"
             <[email protected]...
             > To
                                      Paul Harten/CI/USEPA/[email protected]...
             11/10/2008 01:25 cc
             PM [email protected]...
                                      t
                                                                Subject
                                      Re: [lammps-users] pure ethanol

Dear Paul,
Does material studio gives you the correct density?
Just on the side note which may help you. For PCFF force field, you need
to incorporate 1-4 interactions as non-bonded in lammps. That is what I
found out when I tried to match energies in MS and LAMMPS.

Regards,
Vikas

_Vikas_Varshney2 · November 11, 2008, 3:34am

Dear Paul,
By 1-4 interactions, I mean dihedral interactions, (between 1st and 4th atom along the chain)… they are evaluated as non-bonded interactions (columbic and van der waals, both) on the top on bonded interactions.

Regards,
Vikas

system · November 13, 2008, 2:59pm

Dear Vikas,

For a particular atom, aren't all atoms involved in bonded interactions
excluded from non-bonded interactions? How were you able to add those
atoms involved in dihedral interactions to the pairwise neighbor list?

Thank you for your suggestions and your explanations,

Paul

             "Vikas Varshney"
             <[email protected]...
             > To
                                      Paul Harten/CI/USEPA/[email protected]...
             11/10/2008 10:34 cc
             PM [email protected]...
                                      t
                                                                Subject
                                      Re: [lammps-users] pure ethanol

Dear Paul,
By 1-4 interactions, I mean dihedral interactions, (between 1st and 4th
atom along the chain)... they are evaluated as non-bonded interactions
(columbic and van der waals, both) on the top on bonded interactions.?

Regards,
Vikas

Dear Vikas,

A colleague of mine who did a Materials Studio MS NPT run on pure
ethanol got a density close to 0.78 g/cm3 at 298 K and 0.0 atm.

  In your side note, you suggest that for the PCFF force field I need to
  incorporate 1-4 interaction as non-bonded in LAMMPS. What are the 1-4
  interactions you are referring to?

Thanks,

Paul

              "Vikas Varshney"
              <[email protected]...
              > To
                                       Paul Harten/CI/USEPA/[email protected]...
              11/10/2008 01:25 cc
              PM [email protected]...
                                       t
                                                                 Subject
                                       Re: [lammps-users] pure ethanol

  Dear Paul,
  Does material studio gives you the correct density?
  Just on the side note which may help you. For PCFF force field, you
  need
  to incorporate 1-4 interactions as non-bonded in lammps. That is what
  I
  found out when I tried to match energies in MS and LAMMPS.

Regards,
Vikas

_Budzien_Joanne_Loui · November 13, 2008, 3:50pm

Dear Paul,

Use the special bonds command to add bonded atoms to the pairwise neighbor list.

Joanne Budzien

system · November 13, 2008, 6:44pm

Dear Joanne and Vikas,

From Sun's paper about COMPASS referenced in the LAMMPS User Manual,

PCFF and COMPASS force fields were constructed allowing atoms separated
by two or more atoms to have pairwise interactions. So, using
"special_bonds 0 0 1" to allow the end atoms of (1-4) dihedral
interactions to contribute to pairwise interactions sounds most
appropriate!

Thank you both for your assistance,

Paul

             "Budzien, Joanne
             Louise"
             <[email protected]... To
             > Paul Harten/CI/USEPA/[email protected]...,
                                      "Vikas Varshney"
             11/13/2008 10:50 <[email protected]...>
             AM cc
                                      "[email protected]...
                                      et"
                                      <[email protected]...
                                      >
                                                                Subject
                                      Re: [lammps-users] pure ethanol

Dear Paul,

Use the special bonds command to add bonded atoms to the pairwise
neighbor list.

Joanne Budzien

Dear Vikas,

      For a particular atom, aren't all atoms involved in bonded
      interactions
      excluded from non-bonded interactions? How were you able to add
      those
      atoms involved in dihedral interactions to the pairwise neighbor
      list?

Thank you for your suggestions and your explanations,

Paul

                   "Vikas Varshney"
                   <[email protected]...
                   >
      To
                                            Paul Harten/CI/USEPA/[email protected]...
                   11/10/2008 10:34
      cc
                   PM
      [email protected]...
                                            t

      Subject
                                            Re: [lammps-users] pure
      ethanol

      Dear Paul,
      By 1-4 interactions, I mean dihedral interactions, (between 1st
      and 4th
      atom along the chain)... they are evaluated as non-bonded
      interactions
      (columbic and van der waals, both) on the top on bonded
      interactions.?

Regards,
Vikas

Dear Vikas,

        A colleague of mine who did a Materials Studio MS NPT run on
      pure
        ethanol got a density close to 0.78 g/cm3 at 298 K and 0.0 atm.

        In your side note, you suggest that for the PCFF force field I
      need to
        incorporate 1-4 interaction as non-bonded in LAMMPS. What are
      the 1-4
        interactions you are referring to?

Thanks,

Paul

                    "Vikas Varshney"
                    <[email protected]...
                    >
      To
                                             Paul Harten/CI/USEPA/[email protected]...
                    11/10/2008 01:25
      cc
                    PM
      [email protected]...
                                             t

      Subject
                                             Re: [lammps-users] pure
      ethanol

        Dear Paul,
        Does material studio gives you the correct density?
        Just on the side note which may help you. For PCFF force field,
      you
        need
        to incorporate 1-4 interactions as non-bonded in lammps. That is
      what
        I
        found out when I tried to match energies in MS and LAMMPS.

Regards,
Vikas