Lecture
Van Marum Colloquium: New frontiers in modeling nanoporous materials and their applications at the crossroads of quantum mechanics, statistical physics and machine learning
- Date
- Thursday 4 September 2025
- Time
- Location
-
Gorlaeus Building
Einsteinweg 55
2333 CC Leiden - Room
- CM.3.23
Biography
Veronique Van Speybroeck is full professor at the Ghent University and head of the Center for Molecular Modeling, a multidisciplinary research center composed of about 45 researchers. She was trained as an engineer in Physics and obtained her PhD in 2001 from the Ghent University. She made significant contributions to the field of modeling nanoporous materials for catalysis, adsorption, separations; all applications are inspired and performed in close synergy with experimental groups. The research is driven by the ambition to model as close as possible realistic materials/processes. She played a pioneering role in development of molecular dynamics methods to simulate catalytic reactions at operating conditions. Currently, she is extending the horizon to integrate machine learning methods within molecular modeling of industrial processes to resolve complex catalytic cycles bridging length and time scales. She received three ERC grants, including a recently granted ERC AdG grant to explore the time dimension as a powerful design parameter for next-generation nanoporous materials important for catalysis, separation and sensing. The research was recognized by numerous recognitions and prizes, such as the Dr. Karl Wamsler innovation award in 2023 and the Francqui prize in exact sciences in 2024. She is also an elected member of the Royal (Flemish) Academy for Science and the Arts of Belgium (KVAB).
Abstract
Nanoporous materials are omnipresent in the fields of catalysis, sorption, sensing and crucial for future technologies. Modelling is pivotal to understand the function of nanoporous materials. Ideally it would be possible to design the right material with atomic scale precision for the desired macroscopic function. Modeling realistic functional nanomaterials poses significant challenges. Firstly, nanostructured materials used in applications are far from perfect, they possess a broad range of heterogeneities in space and time extending over several orders of magnitude. Spatial heterogeneities from the subnanometer to the micrometer scale in crystal particles with a finite size and specific morphology, impact the material’s dynamics. Secondly, the material’s functional behaviour is largely determined by the operating conditions. Currently, there exists a huge length-time scale gap between attainable theoretical length-time scales and experimentally relevant scales.1-2
A first important ingredient for modelling nanostructured materials is an accurate representation of the interatomic interactions. Ideally very accurate quantum mechanical methods are used for this purpose. Despite the availability of powerful high-performance computers and the development of advanced methods and algorithms, solving the quantum mechanical many-body problem directly for system sizes comparable to experimental systems remains infeasible.
A second crucial aspect of the modeling exercise is the sampling problem of the multidimensional potential energy surface. Identifying interesting regions in phase space is challenging due to the many degrees of freedom involved. To address this, advanced enhanced sampling methods have been developed, however often these rely on chemical intuition.
Recently new avenues have emerged in modelling nanostructured materials thanks to methodologies integrating concepts from machine learning, quantum mechanics and statistical physics. One notable development is the ability to determine energies and forces using numerical Machine Learning Potentials (MLPs) derived from underlying quantum mechanical data.
To fully integrate MLPs into the catalysis or materials design workflow, it is essential to efficiently generate training data that accurately represent the highly dimensional Free Energy Surface while maintaining chemical accuracy. Addressing this challenge requires methodological advancements that enable the coupling of MLPs with kinetic and sampling models to describe complex dynamical phenomena across a broad range of length and time scales. Within this talk, I will show some of our recent endeavours in this direction.3-6 As will become clear realistic nanomaterials requires a multidisciplinary vision between physics, chemistry, material science, engineering and machine learning. The methods will be illustrated on timely applications in the fields of heterogeneous catalysis, adsorption and sensing using nanoporous materials.7
References
- Van Speybroeck, V., Challenges in modelling dynamic processes in realistic nanostructured materials at operating conditions. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 2023, 381 (2250), 20220239.
- Van Speybroeck, V.; Bocus, M.; Cnudde, P.; Vanduyfhuys, L., Operando Modeling of Zeolite-Catalyzed Reactions Using First-Principles Molecular Dynamics Simulations. ACS Catal 2023, 13 (17), 11455-11493.
- Vandenhaute, S.; Cools-Ceuppens, M.; DeKeyser, S.; Verstraelen, T.; Van Speybroeck, V., Machine learning potentials for metal-organic frameworks using an incremental learning approach. npj Comput. Mater. 2023, 9 (1), 19.
- Bocus, M.; Vandenhaute, S.; Van Speybroeck, V., The Operando Nature of Isobutene Adsorbed in Zeolite H-SSZ-13 Unraveled by Machine Learning Potentials Beyond DFT Accuracy. Angew Chem Int Ed Engl 2025, 64 (1), e202413637.
- Bocus, M.; Goeminne, R.; Lamaire, A.; Cools-Ceuppens, M.; Verstraelen, T.; Van Speybroeck, V., Nuclear quantum effects on zeolite proton hopping kinetics explored with machine learning potentials and path integral molecular dynamics. Nat. Commun. 2023, 14 (1), 1008.
- Dobbelaere, P.; Vandenhaute, S.; Van Speybroeck, V., Cluster-Based Machine Learning Potentials to Describe Disordered Metal–Organic Frameworks up to the Mesoscale. Chemistry of Materials 2025, 37 (15), 5696-5709.
- Lamaire, A.; Wieme, J.; Vandenhaute, S.; Goeminne, R.; Rogge, S. M. J.; Van Speybroeck, V., Water motifs in zirconium metal-organic frameworks induced by nanoconfinement and hydrophilic adsorption sites. Nat Commun 2024, 15 (1), 9997.