My Community

hi,

I have been trying to obtain the initial and final positions for a NEB calculation. However, I am encountering some problems.

The final state of my computations should result in the separation of the adsorbed molecule. However, I have observed that in some metallic surfaces, after molecular separation the adsorbed atoms don't remain on top of surface, the atoms move into the slab. Why this occurs?

Also, if the adsorbed atoms are very close to the surface, the metallic atoms of the slab move very drastically after a simple ionic relaxation. So my questions are:

1) In adsorption-energy computations, why do the adsorbed atoms move into the slab?

2) Why the metallic atoms drastically change their positions if the adsorbed atoms are in close proximity to the surface atoms?

thanks

This is the matter of reactivity of both atoms and the surface. In the first step I would try the dissociationof the molecule on a less reactive surface.

Dear Admin,

Thanks for the response. I agree with your view.

However, I would like to know whether or not there are other practical approaches (i.e., specif keywords in the INCAR file) to avoid: a) the fact that the adsorbed atoms move into the slab; and b) the drastic change of atomic postions of the slab atoms upon simple ionic relaxations.

For example, I have been able to converge NEB computations on various metallic surfaces; and for those metallic substrates, for which, I have either/or problems a) and b) (see above), I have obtained initial and final configurations by: 1) freezing all the atoms of the slab or 2) by placing the adsorbed molecule at a different adsorption site (e.g., fcc site insted of the on-top site).

What other approaches can one try to avoid problems like the migration of the adsorbed atoms into the metallic slab?

thanks

Dear sigi14,

it would improve the quality of our hints if you could post more details, please. What kind of metal, surface and atoms are you studying?

Cheers,

alex