Dr. Valdes-Morales, Hector

Anatase nanoparticles containing surface oxygen vacancies (VO) and Ti3+ are of great importance for applications in photocatalysis, batteries, catalysis, sensors among other uses. The properties of VO and their dependence on the size of nanoparticles are of great research interest and could allow obtaining advanced functional materials. In this work, a complete set of oxygen vacancies in an anatase nanoparticle of size 1.1 nm was investigated and compared to those of a twice larger nanoparticle, having the same shape and surface hydroxylation pattern. It turned out that the decrease in the size of the anatase nanoparticle strongly facilitated creation of surface oxygen vacancies and Ti3+. After their creation, oxygen vacancies undergo three transformation paths — (1) small repulsion of surrounding Ti cations with retention of the vacancy, (2) transfer of oxygen anion, leading to the movement of oxygen vacancy to a more stable position, and (3) collapse of oxygen vacancy accompanied by structure deformation towards Magneli-like phase.

Background: Understanding the structures and properties of photocatalysts requires the developing of computational quantum models. The present study is devoted to calculating structures, bands positions and location of the HOMO/LUMO orbitals in anatase titanium dioxide (TiO2) nanoparticles comprising of exposed (001) and (100) surfaces of various sizes having different extent of hydroxylation at their edges. The (001) surface was left intact or was reconstructed by introducing an additional row of atoms every four unit cells. Two computational approaches were compared: self‐consistent charge density‐functional tight‐binding (SCC‐DFTB) and PM6. Results: The SCC‐DFTB method was found to be, for most cases, superior to the PM6 method in terms of structure, band positions and electronic orbitals. Based on the SCC‐DFTB approach it was concluded that the presence of the (1 × 4) reconstruction was essential for keeping the (001) surface flat. Otherwise the surface undergoes anisotropic shrinking and bending, which contradicts experimental data. The band gap of the de‐hydroxylated or partially hydroxylated nanoparticles was always smaller than that of nanoparticles with hydroxylated edges. No clear quantum size effects were found for these nanoparticles. Photogenerated non‐thermalized holes were found to localize around (001) facets and at their edges, while electrons tended to concentrate over the central parts of the (100) facets. Conclusion: Efficient separation of charge carriers is predicted for anatase nanoparticles having (001) and (100) external surfaces. This conclusion, and moreover, the approach of using SCC‐DFTB calculations to study faceting effects, is likely to be relevant to the developing of new, highly active, photocatalysts as well as for fundamental studies of adsorption.

Anatase (001)nanosheets have recently attracted great attention as very active catalysts and photocatalysts. These graphene analogs have very high surface area and unique surface properties. In the present paper, very thin two-layer anatase nanosheets are investigated computationally in the form of quantum dots of various size. Quantum size effect (QSE)was clearly observed for nanosheets with fully hydroxylated edges and size up to 14 nm and the ultimate band gap is around 3.4 eV. Dehydroxylation of nanosheets obscured QSE, decreased band gap and induced visible light absorption. Therefore, contradictory trends reported in experimental studies for anatase QSE can be ascribed to different degree of hydroxylation of the TiO 2 samples surface. All anatase nanosheet quantum dots retained their flat graphene-like shape. These findings demonstrate that dehydroxylated anatase nanosheet quantum dots are prospective visible-light active photocatalysts even if their inherent band gap is considerably larger than for bulk anatase.

Cationic doping of TiO2 anatase with sulphur represents a facile method to improve catalytic and photocatalytic activity for hydrogen production and extend the action spectrum of TiO2 into the visible light region. However, there is a lot of misunderstanding when trying to explain the experimental findings and suggest theoretical models. In the present computational research work, novel theoretical models are put forward representing fully hydroxylated small anatase nanoparticles with S(IV) and S(VI) doping in various surface positions and in the bulk. It was found that sulfur in the doped anatase nanoparticles preserves its typical coordination geometries of trigonal pyramid for S(IV) and tetrahedron for S(VI). Doping in the anatase surface is much more energetically favorable compared to doping in the bulk. Doping with S(IV) causes decrease of the band gap from 3.22 to 2.65 eV while S(VI) doping could decrease Eg only to 2.96 eV. Location of photogenerated electrons and holes depends strongly on the position of dopant atoms and their valent state. Contrary to some experimental works, no strong and extended visible light absorption bands could be found with cationic doped hydroxylated anatase nanoparticles. However, improved charges separation is observed indeed and causes improved photocatalytic hydrogen production.