[CPMD-list] Tool for CPMD-input creation

[Jan-Willem Handgraaf]

On Thu, 2004-04-29 at 09:57, Holger Hesske wrote:
> Dear all,
> 
> I investigate in proton-transfer reactions with CPMD. So far so well. But if
> I restart a MD calculation the EKINC is always set to zero (or to the
> TEMPCONTROL value given in the Input) and there is no smooth trajectory.
> Especially the first derivation of the proton coordinates shows discontinous
> behaviour.
> I try RESTART with allmost all options (GEOFILE / VELOCITIES / WAVEFUNCTION
> / ACCUMULATORS / ALL) but still the same result. Am`I missing something? 
> Any help or suggestions are welcome and would be gratefully accepted.
> 
> Thanks in advance 
> Holger

Dear Holger Hesske,

This discontinuity of the derivative sounds very strange to me. I have
performed proton transfer reactions, but never found any discontinuity.

Have you considered the following:

If you use some reaction constraint, you should probably not take too
large steps between consecutive constraint values because when you 
change for example the position of the proton manually it will disturb
your system too much. Also afterwards you have of course to bring you
system back to the ground state. Better is to change the velocities of
the ions manually in the right direction, and after a short MD run, fix
the reaction coordinate to the new value.

Hope this helps,

Jan-Willem

P.S. I think that it should not be a problem when the fictitious kinetic
energy is set to zero if you restart your MD run, since it will adopt
very quickly to the relatively slow motion of the ions.


--
Jan-Willem Handgraaf
Dipartimento di Chimica "G. Ciamician"
Università di Bologna
Via F. Selmi 2
40126 Bologna, Italy

Tel: +39-051-2099596
Email: jwh at ciam.unibo.it
Web: http://www.ciam.unibo.it/sitcon/