Computer modelling of metal oxide semiconductors for photo-catalysis

由于矿物燃料的枯竭和环境污染的加剧，寻找替代和可持续能源的需要正在增加。由于金属氧化物半导体如NaNbO3、LaMnO3和WO3具有光催化将水分解为氢分子和氧分子的能力，它们已成为当前研究的热点。然而，光催化剂表面的氧化是一个不希望的副作用。虽然许多反应机制尚未完全了解，但精心设计的含有掺杂剂的材料很可能更稳定，并保持其催化活性。在这个计算项目中，要研究的基本步骤包括：1。研究已知成分的金属氧化物半导体之间形成的固态溶液的混合热力学。以最稳定的组成和原子分布将额外的掺杂剂、缺陷和手性配体掺入到固体溶液的体和界面中。评估掺杂剂的潜在偏析，例如向体或界面的迁移。在与环境分子相互作用时产生表面缺陷，如空位、台阶和（水合）氧。通过调整掺杂剂、手性配体、空位以及表面缺陷和异质界面的类型和浓度来控制可见光光催化活性。光催化水分解和二氧化碳转化为有机小分子的机理途径的模拟。开发一种机器学习算法，以预测在可见光存在下具有更大活性的新型金属氧化物半导体。

Due to the depletion of fossil fuels and the intensification of environmental pollution, the need to find alternative and sustainable energy sources is increasing. Since metal oxide semiconductors such as NaNbO3, LaMnO3 and WO3, have the ability to split water photo-catalytically into molecular hydrogen and oxygen, they have become a topic of intense current research. However, the oxidation of the surfaces of the photo-catalysts is an undesired side-effect. Although many reaction mechanisms are not yet fully understood, carefully designed materials containing dopants may well be more stable and retain their catalytic activity.In this computational project, the fundamental steps to be investigated include:1. Study of the mixing thermodynamics of the solid-state solution formed between metal oxide semiconductors of known compositions.2. Incorporation of additional dopants, defects and chiral ligands into the bulk and at the interfaces of the solid-state solutions with the most stable composition and atomic distribution.3. Evaluation of potential segregation of the dopants, e.g. migration towards the bulk or interfaces.4. Creation of surface defects, such as vacancies, steps and (hydr-)oxo species upon interaction with environmental molecules.5. Control of the visible-light photocatalytic activity via tuning of the type and concentration of dopants, chiral ligands, vacancies as well as surface defects and hetero-interfaces.6. Simulation of the mechanistic pathways for the photo-catalytic water splitting and carbon dioxide conversion into small organic molecules.7. Development of a machine-learning algorithm to predict new metal oxide semiconductors with larger activities in the presence of visible light.