[gpaw-users] water adsorption at various metals

[Jens Jørgen Mortensen]

On 12-12-2011 15:57, Vladislav Ivanistsev wrote:
> By trying to calculate a single water molecule adsorption at Pt slab, 
> I could not get a desirable value of approximately -0.3 eV, which 
> comes from works of Michaelides, Blanco et al, Meng etc.

Are they also using RPBE?  What orientation of the water molecule do 
they use?  Are your numbers converged with respect to number of k-points 
and number of layers?

Jens Jørgen

> For different set of parameters (k points, grid spacing, unit cell 
> size) I've got positive values of adsorption energy calculated in a 
> usual way per water molecule: Eads = Etot - Ewater - Eslab. Dipole 
> correction and application of vdW functional have not helped. In 
> attached Fig. 1 are shown results for calculations at slab. The best 
> result has been achieved for 3*3*1 k-points for 4*4*2 Pt slab. The 
> lowest energy is +0.1 eV. Reference energy for water molecule used 
> equals to -14,0 eV for default GPAW mode, and -13,0 for lcao mode with 
> dzp basis set.
> In Fig. 2 and Fig. 3 are shown results for calculations on slab and 
> cluster. This time potential energy curves has been calculated, i.e. 
> energy at distance of 1 nm is chooses as a reference. Firstly, very 
> small systems shows desirable result. However, a bit bigger systems 
> like Pt10 are totally disappointing. Mark, that in Fig. 3 the values 
> are close to zero. Also information on unit size, calculation mode and 
> cluster size is given in legend.
> Similar behavior is seen for Bi and Au surfaces, which were tested. 
> This points me to the fact, that I'm somewhere just wrong. 
> Unfortunately, I don't know where.
> Please correct me and point where I might be wrong.
>
> Gratefully,
> MSc Vladislav Ivanistsev
> University of Tartu, Estonia