Re: simulating graphene sheets (with bonds across boundaries) with water

From: upayan baul (
Date: Tue Sep 15 2015 - 06:57:20 CDT

Dear Abhishek,

Thank you for your response.

Using larger PBC did not cure the issue. Increasing PBC by 2A along X and Y
directions resulted in
patch grid error (Periodic cell has become too small for original patch
grid). The PME grid sizes were
also set up in accordance with system dimensions (60 along all axes with
system dimensions < 55 along
each of the axes).

I had tried holding the graphene atoms fixed but that had resulted in the
following errors:
Fixed atoms without constant area : simulations crashed (owing to fixed
atoms at boundaries?)
Fixed atoms with constant area and flexible cell : system expands unusually
along Z.

I tried restraining the graphene atoms (along Z) as suggested by you using

constraints on
consexp 2
consref fixed_GRA_beta_50.pdb
conskfile fixed_GRA_beta_50.pdb
conskcol B
selectConstraints on
selectConstrZ on

and this seems to be working. Further, the sheets are uniformly flat
without impropers, but slightly rough
with the same. So I shall be simulating without improper interactions.


On Tue, Sep 15, 2015 at 5:58 AM, Abhishek TYAGI <>

> Dear Upayan,
> In your work you may try fixing the graphene sheet i hope you know how to
> do it or you may restrain them, and for the error, it seems like that your
> box is small, try to use bigger pbc than your original dimensions, have a
> look on PME too.
> Thanks
> Abhi
> Abhishek Tyagi
> PhD Student
> Chemical and Biomolecular Engineering
> Hong Kong University of Science and Technology
> Clear Water Bay, Hong Kong
> ------------------------------
> *From:* <> on behalf of
> upayan baul <>
> *Sent:* Monday, September 14, 2015 7:59 PM
> *To:* Namd Mailing List
> *Subject:* namd-l: simulating graphene sheets (with bonds across
> boundaries) with water
> Dear NAMD users,
> Sorry for the long mail. I wished to let you know the things I have
> already tried.
> I wish to simulate a system with stacked graphene sheets (with bonds
> across
> boundaries in XY) in water in an NPT ensemble. The initial configuration
> of the
> system is shown in.
> As an ideal test case, I wish to retain the planarity of the graphene
> sheets to as great a
> degree as possible.
> I am using atom type CA and corresponding parameters for bond, angle,
> dihedral
> and nonbonded interactions from CHARMM protein parameters
> (par_all22_prot). I
> have tried small simulations
> without impropers : (based on Kenno's suggestion that impropers only
> over-rigidify 6
> membered ring structures
> )
> which resulted in a configuration as shown (periodic) in:
> with impropers: (based on Axel's suggestion (though this might be for the
> finite-sheet case,
> I might have misunderstood) to include impropers for planarity
> )
> which resulted in a configuration as shown in:
> in which I have used an additional improper interaction of
> !atom types Kpsi psi0
> CA CA CA CA 10.0 0 0.00
> Neither of the two cases are resulting in highly planar structures.
> Increasing Kpsi is resulting
> in more puckered structures (for Kpsi > 20, the simulations are crashing
> with statements like
> ERROR: Atoms moving too fast; simulation has become unstable (3 atoms on
> patch 1 pe 0).
> How can I obtain a more planar configuration for graphene? I have tried
> fixed atom simulations.
> In NPT, without constant area such simulations crash owing to fixed atom
> at boundaries. With
> constant area, the system expands in the Z- direction, causing the water
> to move away from the
> surface.
> Thanking you,
> Upayan
> S.R.F., Physics
> IMSc

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:22:03 CST