We are excited to welcome Prof. Jonas Elm as our next speaker in the Atomistic Modeling Seminar!

Prof. Jonas Elm is an Associate Professor at the Department of Chemistry at Aarhus University. His research focuses on computational chemistry, atmospheric chemistry, molecular clusters, and new particle formation. He has received several competitive research grants, including an ERC Starting Grant (2021), and currently serves as President of the European Aerosol Assembly (EAA) and the Nordic Society for Aerosol Research (NOSA).

In his talk “Machine Learning–Enhanced Quantum Chemical Modeling of Atmospheric Molecular Cluster Formation”, Prof. Elm will address the formation of molecular clusters as a critical step in atmospheric new particle formation and discuss the challenges associated with identifying the contributions of different vapors. He will explain how quantum chemical calculations combined with kinetics modeling can be used to analyze cluster formation pathways, while also pointing to the computational demands of these approaches and the size gap between modeled clusters and experimentally observed particles.

Prof. Elm will further present their recent work on the screening of atmospherically relevant cluster systems and show how machine learning methods can accelerate the exploration of configurational space. In addition, the talk will present ongoing work on the composition of freshly nucleated particles and discuss how these efforts help bridge the gap between theoretical and experimental studies in atmospheric aerosol modeling.

Date: Tuesday, May 5, 2026, 10:30 am

Location: MIBE Lecture Hall

Abstract:

The formation of molecular clusters is a critical step in atmospheric new particle formation, yet the contributions of different vapors remain highly uncertain. Quantum chemical (QC) calculations combined with kinetics modeling enable detailed mapping of cluster formation pathways, but these approaches are computationally demanding. Moreover, a significant size gap exists between clusters that can be accurately modeled (~1 nm) and particles routinely measured in field and laboratory studies (~2 nm). As a result, the transition from small clusters to freshly nucleated particles (FNPs) remains largely unexplored.

In this presentation, recent work on the comprehensive screening of atmospherically relevant cluster systems will be described, providing direct insight into their formation potential and ability to grow to larger sizes. It will also be demonstrated how machine learning models can accelerate the exploration of configurational space in cluster formation studies. Finally, ongoing work on the composition of FNPs will be presented. Together, these efforts help bridge the gap between theoretical and experimental studies, enabling molecular-level cluster formation processes to be incorporated into atmospheric aerosol models.