First we worked on the preparation and on the study of models of nanostructures for a photosensitive semiconducting oxide: titanium dioxide (TiO2). We investigated nanoparticles (NPs) of increasing size (up to a diameter of 4.4 nm: about 4000 atoms, see Figure 2) and considered different shapes: decahedral (exposing facet and edges) and spherical shaped nanoparticles. This activity required a careful design for chemically stable stoichiometric systems and the recourse to some simulated thermal annealing through molecular dynamics runs, as a suitable approach to global optimization of such complex and large systems. The final optimized structures were fully relaxed at a quantum chemical level to obtain accurate electronic properties for comparison with bulk and with available experimental data.
Since nanoparticles typically experience water environments or biological aqueous media, as a next step we investigated the oxide/water interface, for both flat and curved TiO2 surfaces. This work was done in collaboration with the group of Prof. Y. Matsumoto in Kyoto (Japan), who can prepare nanoparticles of different shape and of small size (comparable to that we can simulate with first-principles calculations) and to gradually control the water partial pressure, going smoothly from ultra-high vacuum to high water partial pressure.
To make our studies more realistic, we aimed at increasing the nanoparticle size (in terms of number of atoms) and the number of water layers around the nanoparticles. However, calculations become quickly too costly. For this reason, we made an effort to provide an accurate benchmark of a computationally cheaper method (Density Functional Tight Binding, DFTB) with respect to reference Density Functional Theory (DFT) calculations for water multilayers on TiO2 slab models, as described by both static and dynamic simulations. The assessment of the validity of DFTB method for this type of investigations, derived from our study, in collaboration with Prof. G. Seifert from Dresden University, is an extremely important result for this project. With this powerful tool, we could address large nanoparticles in a multilayer of water. On top of that, we also developed a QM/MM interface (Figure 3) to include the presence of surrounding bulk water and we also compared results with the REAXFF approach.
As a further important step, we studied the photoirradiation process of TiO2 NPs by computing electronic transitions (excitation and emission) and spectroscopic properties, by defining the nature of the photoexcited excitons and by following the faith of the photogenerated electron/hole (Figure 4).
Surface coating with polyethyleneglycol (PEG) is a common strategy to improve stealth properties of NPs in vivo (Figure 5). We investigated how anchoring groups, chain length and environment affect the dynamical properties of the PEGylated TiO2 nanosystems, through a multiscale approach involving both QM and MM calculations.
Titania NPs are commonly functionalized to modify their absorption properties and to activate their tethering ability towards biologically active molecules (Figure 6). Based on QM calculations, we prepared reliable models of TiO2 NPs that are functionalized with bifunctional linker (TETT, DOPAC, dopamine, etc.) and performed a thorough characterization, based on DFT and TDDFT calculations to obtain not only structural and electronic properties but also simulated absorption and emission spectra.
We then used these QM models of functionalized NPs as a starting point for the investigation, by means of classical MD and docking techniques, of the biological functions of these nanohybrids in a physiological medium (Figure 7 and Figure 8): targeting abilities, drug transport, interaction with oligonucleotides or with proteins.
In parallel, we also similarly worked on a magnetic metal oxide of extremely high interest for nanomedical applications: magnetite (Fe3O4). We have first determined the proper methodology to describe the delicate electronic and magnetic properties of this complex system, then we model surface/water interface and differently shaped NPs of realistic size. Based on this work, we could develop a general formula to determine total magnetic moment of nanostructures to be used to determine the efficacy of the system in nanomedical applications. We also proposed an unconventional view on the underlying mechanism through which the coating with organic acids or other ligands affects the magnetic properties of the NPs.
