[Octopus-users] Periodic Boundary Conditions.

[Jose R. Valverde]

On Wed, 1 Oct 2008 09:54:46 +0200 (CEST)
alberto.castro at tddft.org (Alberto Castro) wrote:
> octopus, contrary to vasp, you are using a grid, not a plane wave
> basis set, so there are not periodic boundary conditions in any
> case. You do not need a supercell (the cell has to be large enough to
> completely hold the system, though).

I confess my ignorance, but wouldn't it still be possible/sensible
to add PBC to Octopus?

My problem is of a different nature: I am trying to analyze a highly
exotermic process, e.g.

> ********************* Time-Dependent Simulation **********************
>   Iter           Time        Energy     Elapsed Time
> **********************************************************************
> 
>       1       0.005000   -837.507274        23.940
>       2       0.010000 527159.908675       116.150
>       3       0.015000**************        14.748
>       4       0.020000           NaN        18.270

This results in a large release of energy which thermally fires the 
molecules very far away apart. In a "normal" situation with many 
molecules this need not be a problem: either their crashing into 
each other keeps them close or brownian movement brings others close.

Without PBC the initial energy release produces escape velocities from
intermolecular interactions precluding further steps to be followed.
And at escape velocities there is no way to avoid molecules leaving
the simulation box:

> ** Warning:
> **   Atom     1 is outside the box.
> 
> 
> ** Warning:
> **   Atom     2 is outside the box.
etc...

I've seen similar situations with other codes, which I could correct
either with PBC or forcing an NPT (or NET) ensemble with quick P, T
quenching. Quenching allows the whole process to be followed, but
the energy dissipation distorts calculations...

So, would it be sensible to add PBC to Octopus? Or barring that,
may I choose the kind of MD ensemble to be used so as to absorb
the released energy before it turns into excessive kinetic energy?

I would also note in passing that tuning the MD ensemble is a just
hack as it will remove energy (e.g. by coupling to a thermal bath)
that might hypothetically be used in later steps of the process.

				j
-- 
			EMBnet/CNB
		Scientific Computing Service
	Solving all your computer needs for Scientific
			Research.

		http://bioportal.cnb.csic.es
		  http://www.es.embnet.org