Computational Models based on First-Principles Calculations for In-Silico Evaluation of Nano-compounds Toxicity

Identificador:  TDX:4521

Handle: https://hdl.handle.net/20.500.11797/TDX4521

Autores:  Çetin, Yarkin Aybars

Resumen:
This thesis explores metal oxide (MO) toxicity-related surface reactivity through studies that integrate theoretical and computational methodologies. The compendium is structured around four research manuscripts. Each chapter focuses on distinct aspects of MO surface reactivity and its implications across various applications. The first chapter explores the reactivity of titanium dioxide (TiO2), a widely used mate- rial. Using Density Functional Theory (DFT) and Density Functional Tight Binding (DFTB) methods, the chapter analyses oxygen vacancies' influence on the TiO2 surface and compares these methods. By computing chemical descriptors for anatase and rutile TiO2 surfaces, the study provides valuable insights into reducibility properties for understanding toxicological mechanisms. The second chapter revisited water adsorption on TiO2 and zinc oxide (ZnO) surfaces through Self-Consistent-Charge Density Functional Tight Binding (SCCDFTB) molecu- lar dynamics simulations. By examining geometric and electronic structure dynamics in an MO-water interphase, the research elucidates surface interactions and their implications for potential toxicity. It also explores crucial molecular-level insights into dissolution processes. The third chapter investigates the genotoxicity of TiO2 surfaces, focusing on an adsorp- tion model of guanine on oxygen-deficient TiO2 surfaces. Through a combined SCC-DFTB Molecular Dynamics (SCC-DFTB/MD) and DFT strategy, novel adsorption modes are characterized, highlighting the significance of oxygen-deficient surfaces in genotoxic effects. The fourth chapter introduces a machine-learning approach to predict the electronic density of states in MO-guanine adsorption models. By utilizing neural networks, the study proposes a computationally efficient alternative to traditional DFTB calculations.