guidelines on implementing special kind of collective variable

Peter pmam... at
Wed Feb 9 16:29:09 UTC 2011

Thank you for reply, Teo.

Looks like it is much more complicated task than I hoped it would be.
Probably I'll search for some more technically feasible approaches to
my problem.


On Feb 9, 4:35 pm, Teodoro Laino <teodor... at> wrote:
> Dear Peter,
> it is by far not an immediate task: you need to run 2 SCF calculations for a same geometry (which is not available at the moment).
> In principle, with some modification you may use a fake mixed_env (2 calculations, using only 1 (the neutral I guess) for propagation).
> But you still need to implement the possibility of storing/retrieving the energy/forces of the two systems independently.
> Feasible but not immediate: You need to be extremely well acquainted with the code to implement something like that.
> Or alternatively one could implement the possibility to run an additional SCF (without messing up the extrapolation guess) during an MD: in your case you would propagate the neutral system and compute for each geometry an energy/forces for the negative one.
> Best,
> Teo
> On Feb 9, 2011, at 2:20 PM, Peter wrote:
> > Dear CP2K developers!
> > I would like to try to implement a special type of collective
> > variable: it is a difference between potential energy of a system and
> > potential energy of the system in the same geometry with an extra
> > electron. In other words, I want to use vertical ionization energy as
> > a collective variable for e.g. metadynamics.
> > If there are any caveats regarding this task? E.g. with potential
> > derivatives calculation.
> > Could you please give some guidelines on implementing this in cp2k?
> > Thanks,
> > Peter
> > --
> > You received this message because you are subscribed to the Google Groups "cp2k" group.
> > To post to this group, send email to cp... at
> > To unsubscribe from this group, send email to cp2k+uns... at
> > For more options, visit this group at

More information about the CP2K-user mailing list