H2 chemisorption on W(100) and W(110) surfaces

Chemisorption of H and H-2 on clean W(100) and W(110) surfaces is investigated from extensive density functional theory (DFT) calculations within the generalized gradient approximation (GGA). We obtain properties of the clean surfaces [e.g., the well-known reconstructed structure of W(100) below 200 K] as well as adsorption energies and geometries of H atoms chemisorbed on both faces of W in very good agreement with available experimental data. From our DFT-GGA results, we build accurate six-dimensional potential energy surfaces (PESs) which are used to compute dissociative sticking probabilities of H-2 on both faces of W through classical trajectory calculations. For both systems, our theoretical results agree with the available experimental data at low impact energies and are larger than experiments at higher energies. The main signatures of the sticking probabilities found in molecular beam experiments are reproduced, at least qualitatively, by our calculations: for instance, (i) the larger values for W(100) than for W(110), (ii) the nonmonotonic and monotonic energy variation for W(100) and W(110), respectively, and (iii) the dependence on incidence angle as a function of W face and impact energy. Our calculations show that dissociative adsorption of H-2 on both W(100) and W(110) is a nonactivated process. In contrast with the widely accepted assumption that H2 chemisorption on W(110) is a purely direct process, we obtain that, at very low energies, adsorption takes place through an indirect (dynamic trapping) and a direct mechanism on both surfaces. Thus, the qualitatively different behavior of the sticking probability at low energies arises from a smaller contribution of dynamic trapping in the case of H-2/W(110), due to slight differences between the corresponding PESs far from the surface in the entrance channel.