The Role of Surface and Subsurface Point Defects for Chemical Model Studies on TiO2: A First-Principles Theoretical Study of Formaldehyde Bonding on Rutile TiO2(110)

We report a systematic investigation of the effects of different surface and subsurface point defects on the adsorption of formaldehyde on rutile TiO2(110) surfaces using density functional theory (DFT). All point defects investigated-including surface bridging oxygen vacancies, titanium interstitials, and subsurface oxygen vacancies-stabilize the adsorption significantly by up to 56 kJ mol(-1) at a coverage of 0.1 monolayer (ML). The stabilization is due to a decrease of the coordination (covalent saturation) of the surface Ti adsorption sites adjacent to the defects, which leads to a stronger molecule-surface interaction. This change in the Ti is caused by the removal of a neighboring atom (oxygen vacancies) or substantial lattice relaxations induced by the subsurface defects. On the stoichiometric reference surface, the most stable adsorption geometry of formaldehyde is a tilted eta(2)-dioxymethylene (with an adsorption energy E-ads=-125 kJ mol(-1)), in which a bond forms to a nearby bridging O atom and the carbonyl-O atom in the formaldehyde binds to a Ti atom in the adjacent fivefold coordinated lattice site. The eta(1)-top configuration on five-coordinate Ti4+ is much less favorable (E-ads=-69 kJ mol(-1)). The largest stabilization is exerted by subsurface Ti interstitials between the first and second layers. These defects stabilize the eta(2)-dioxymethylene structure by nearly 40 kJ mol(-1) to an adsorption energy of -164 kJ mol(-1). Contrary to popular belief, adsorption in a bridging oxygen vacancy (E-ads=-86 kJ mol(-1)) is much less favorable for formaldehyde compared to the eta(2)-dioxymethylene structures. From these results we conclude that formaldehyde will bind in the eta(2)-dioxymethylene structure on the stoichiometric surface as well as in the presence of Ti interstitials and bridging oxygen vacancies. In the light of these substantial effects, we conclude that it is essential to include all the types of point defects present in typical, reduced rutile samples used for model studies, at realistic concentrations to obtain correct adsorption sites, structures, energetic, and chemi-physical properties.