[Pw_forum] frequency of NO molecule

Nicola Marzari marzari at MIT.EDU
Thu Sep 7 18:58:09 CEST 2006

We have done this in the past for CO on Pt (since that
Pt is very heavy, only 3 force constants are neeeded
(Pt-C, C-O, Pt-O)), and you diagonalize a 2x2 matrix
(instead of 6x6, for a molecule standing upright, with only
z as a degree of freedom). Compared to a full phonon calculation,
the discrepancies on the stretching mode of CO were of 50-70 cm-1,
i.e. 5%.

Of course, it depends on how heavy the metal is, and how
things are bonded, as Paolo points out.


Paolo Giannozzi wrote:
> On Thursday 07 September 2006 16:01, Axel Kohlmeyer wrote:
>> how about including several atoms in the neighborhood of the 
>> adsorbed molecule? this is how it works in CPMD: you pick an
>> atom (or pick a reference coordinate) and set a radius and then 
>> do the vibrational analysis for those atoms and keep the rest fixed. 
>> this way you include at least the contributions from the near 
>> neighbors, have the rest as 'static background', and still save
>> a lot of (cpu) time. 
> the problem I see in this approach is that the vibrations of 
> surface atoms close to the adsorbed molecule may be 
> strongly correlated to vibrations of other atoms in the 
> surface; i.e., taking only a few surface atoms close to 
> the molecule will produce rather bogus displacement
> patterns for those atoms. 
> Paolo

Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu

More information about the users mailing list