Theoretical study of the Fluorine doped anatase surfaces Yanaris Ortega c , Oriol Lamiel-Garcia b , Daniel Fernandez Hevia c,d , Sergio Tosoni b,c , Jaime Oviedo a , Miguel Angel San-Miguel a , Francesc Illas b, ⁎ a Departamento de Química Física, Facultad de Química, Universidad de Sevilla, C/Profesor García González 1, E-41012 Sevilla, Spain b Departament de Química Física & Institut de Química Teòrica i Computacional (IQTCUB), Universitat de Barcelona, C/Martí i Franquès 1, E-08028 Barcelona, Spain c Departamento de Química, Universidad de Las Palmas de Gran Canaria, Campus Universitario de Tafira, 35017 Las Palmas de Gran Canaria, Spain d INAEL Electrical Systems S.A., C/Jarama 5, 45007 Toledo, Spain abstract article info Article history: Received 1 August 2013 Accepted 4 September 2013 Available online 17 September 2013 Keywords: Photocatalysis F doped Titania DFT DFT+U A systematic theoretical study based on periodic density functional theory (DFT) calculations using GGA and GGA+U approaches has been carried out to establish the thermodynamic stability of O by F substitution on the (001) and (101) surfaces of anatase. All calculations consistently predict that for both surfaces F implantation is more favorable at surface sites than at subsurface sites. However, the absolute value of the implantation energy has been found to largely depend on the density functional. This fact has strong implications in the study of doped oxides for those cases where accurate values of substitution energies are required. © 2013 Elsevier B.V. All rights reserved. 1. Introduction Titanium dioxide, generally referred to as titania, has surely become one of the most studied inorganic oxides in the last past years from theory to experiment to applications in several technologies. One of the main reasons behind this broad and ever growing interest lies in the very important properties that this material displays in the field of photocatalysis, a heterogeneously catalysed process where the catalyst is activated by incident light of a given frequency. In particular, the pos- sibility of taking advantage of sunlight to accelerate reactions like water splitting for hydrogen generation, oxidation of organic residues in drink- ing water or removal of pollutants and greenhouse gases in the atmo- sphere makes photocatalysis by titania and other semiconducting oxides such as ZrO 2 , ZnO or SrTiO 3 very attractive for future technological applications, as revealed in a series of recent reviews in the field [1–7]. The properties of TiO 2 under UV irradiation are known for many years and have extensively been reviewed [8–11], even though many details on the microscopic mechanisms concerning the excited states involved in the process are still a matter of debate [12]. Nevertheless, the basis of the photocatalytic process is rather simple and involves the generation of a hole in the valence band and of an electron in the conduction band. This hole–electron couple can act as a redox pair to- wards species adsorbed on the external surfaces of the material, provid- ed that their redox potentials are compatible with the energy levels of the photogenerated electron and hole [13]. Note, however, that dynamical aspects regarding the electron–hole recombination are also to be taken into account. Moreover, the electron–hole generation re- quires irradiation of light in a frequency range sufficient to overcome the band gap of the semiconducting oxide (or material) and this is one of the key aspects of the whole process. Precisely, at the state of the art, the main drawback of titania for photocatalytic applications is its rather large band gap (close to 3 eV). This value implies that, under sunlight, only 10% of the incoming photons are adsorbed and, hence, able to participate in the photocatalytic process, resulting in a rather low efficiency. Thus, most of the recent efforts are devoted to devise suitable strategies to reduce the band gap down to 1–1.5 eV which would allow improving the photocatalytic performance of titania under visible light irradiation [8,14]. Several approaches to shrink the band gap of titania have been proposed, spanning from the synthesis of ultrathin layers to the doping with different metals. However, a par- ticular promising approach seems to be the doping with non-metallic elements, like nitrogen [15] or fluorine [16–18]. The latter has additional advantages since, as elegantly shown by Yang et al. [19] treating TiO 2 anatase nanoparticles with F largely stabilizes the reactive (001) facets which should result in an additional enhanced chemical activity. Here, it is important to mention that the experiments on fluorine-doped tita- nia consist of impregnating TiO 2 nanoparticles with F-containing solu- tions [20,21]. This leads to rather complex structures, where domains of different titania polymorphs (anatase, brookite and rutile) coexist, including defects and doping at several possible sites. Therefore, some important questions concerning the physical nature of dopant-related electronic states, the stability of the doped structures and the distribu- tion of fluorine in the catalytic nanoparticles remain open. For instance, Surface Science 618 (2013) 154–158 ⁎ Corresponding author. Tel.: +34 944021229; fax: +34 934021231. E-mail address: francesc.illas@ub.edu (F. Illas). 0039-6028/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.susc.2013.09.010 Contents lists available at ScienceDirect Surface Science journal homepage: www.elsevier.com/locate/susc