flo fsch... at
Mon Apr 7 09:13:12 UTC 2008

Hi Ali,

it's a bit of a misunderstanding.

At first what you will calculate using this keyword is the Hab
parameter in Marcus theory. Hab is of course related to a special
electron transfer event (an electron "localized" at either the one or
the other side of your system). In the approach we use for
calculating, we try to find the two diabatic states related to this ET
event. As an simple example imagine a system containing two Zn (total
charge +1) atoms far apart. DFT will return fractional charge on both
Zn, which is clearly not a diabatic state.
For this system the diabatic states are either the one or the other Zn
with a charge of +1. For to describe these states you can apply a
constraint on the charge, electron density,....
With the wavefunctions for both states you can calculate Hab (for
details see Wu, Van Voorhis, JCP, 125, 164105 (2006)).

With this section you can't force any event. This section is more or
less a Runtype on it's own (only to be used with RUN_TYPE ENERGY). But
what will happen is, that two Energy calculations will be performed to
optimize the value for your applied constraints. And out of the
results of this two optimizations you calculate Hab.

That's how it should work... But it's still only an experimental
feature which works pretty good in system with well defined diabatic
states (such as described above), but in more complicated system like
ET from one molecular site to another it fails since it's rather
sensitive to the definition of the involved diabatic states and in
such a system it's hard to estimate what it should look like.

I don't know what your system looks like, but if you want to have a
try using this method jsut contact me, and I will explain you a bit
more in detail how to use this keyword.


More information about the CP2K-user mailing list