Machine-Learning-assisted design of materials for energy

by Prof. Silvana Botti (RU Bochum)

Thursday 7 Nov 2024, 16:15 → 17:30 Europe/Berlin

D 114 (ENC)

In this presentation, I explore the use of materials databases to design new functional materials, particularly for energy applications. Leveraging high-throughput calculations with density functional theory (DFT) and the capabilities of advanced supercomputers, we can automate the screening of numerous hypothetical materials to address current technological challenges. Additionally, integrating machine learning methods with DFT significantly accelerates the materials discovery process. We summarize our recent efforts in discovering, characterizing, and understanding inorganic compounds, with a specific focus on energy materials.

While characterizing the electronic properties of bulk crystalline materials is essential, it is often insufficient for the development of electronic devices. Quantum processes at interfaces play a crucial role in, e.g., transistors, light-emitting diodes, and solar cells. Shaping the band diagrams at these interfaces presents opportunities for electron manipulation and the creation of new functionalities. However, designing and comprehensively understanding electronic structures at interfaces poses challenges that exceed current methodologies. We discuss recent advances aimed at overcoming these challenges.

Einladender: Mario Agio