Faculty seminar "From Atomistic Interfaces to Functional Materials: Multiscale Modeling and Machine Learning for Energy Applications"
The Faculty of Chemistry at Gdańsk University of Technology cordially invites you to a seminar that will take place on April 23, 2026, at the Faculty of Chemistry, Gdańsk University of Technology, at 12:15 in classroom 013, Chemistry C Building.
Title: From Atomistic Interfaces to Functional Materials: Multiscale Modeling and Machine Learning for Energy Applications
Speaker: David Dell’Angelo, Unité de Catalyse et de Chimie du Solide (UCCS), University of Lille, University of Artois
Abstract: Understanding and controlling matter at the atomic scale is essential for the design of functional materials for energy applications. In this seminar, I will present a multiscale modeling framework that bridges atomistic insights at interfaces with the predictive design of complex materials, combining first-principles simulations, advanced electronic structure methods, and machine learning techniques.
I will first discuss the electronic and transport properties of nanostructured materials, including low-dimensional systems and π-stacked frameworks, where quantum confinement, defects, and structural anisotropy play a central role in tuning band structure and charge dynamics. These effects are particularly relevant for applications in photocatalysis and energy conversion. I will then briefly show how atomistic simulations can capture key interactions at solid–liquid interfaces, providing insights into structure, energetics, and selectivity, and enabling more rational approaches to molecular design.
Finally, I will discuss emerging data-driven strategies for materials discovery, including machine learning-assisted screening, active learning approaches, and the development of machine-learned interatomic potentials. Combined with automated workflows and conceptual density functional theory, these methods enable the efficient exploration of complex chemical spaces. Applications to hydrogen production, dynamic photocatalysis exploiting ferroelectric effects, and plasmonic nanocomposites will illustrate how the integration of physics-based modeling and machine learning opens new avenues for the design of next-generation energy materials.