Lisa Hetzel
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Lisa Hetzel has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Christopher Johannes Stein
Department of Chemistry and Catalysis Research Center, TUM School of Natural Sciences Technische Universität München Garching Germany
4 shared publications - Anton Beiersdorfer
Department of Chemistry and Catalysis Research Center, TUM School of Natural Sciences, Technical University Munich, 85748 Garching, Germany
2 shared publications - Anton Abramov
University of Regensburg
1 shared publication - Carsten G. Staacke
Cellforce Group GmbH, 72138 Kirchentellinsfurt, Germany
1 shared publication - Carsten Staacke
Cellforce Group GmbH
1 shared publication - Christian Lucht
Leipzig University
1 shared publication
Latest Publications
- Open Access
- 21 January 2026
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 7 (1) , 015016
GAP vs. MACE: efficiency evaluation in a liquid electrolyte system
Machine learning interatomic potentials (MLIPs) have revolutionized molecular simulations, but as they evolve, so does the demand for advanced computing architectures, particularly graphics processing units (GPUs). However, the high cost of GPUs limits accessibility, making it crucial to compare GPU and central processing unit (CPU) based MLIPs under practical conditions. This study examines two popular MLIPs: the GPU-accelerated multi-atomic cluster expansion model and the CPU-based Gaussian approximation potentials model, applied to a battery electrolyte system known for its complex properties. By focusing on these models, our study evaluates differences in computational performance, resource efficiency, and accuracy in reproducing experimental properties. This rigorous benchmark provides insights into the trade-offs between GPU and CPU-based approaches in molecular simulations.
- 21 January 2026
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 7 (1) , 015016
Open AccessGAP vs. MACE: efficiency evaluation in a liquid electrolyte system
Machine learning interatomic potentials (MLIPs) have revolutionized molecular simulations, but as they evolve, so does the demand for advanced computing architectures, particularly graphics processing units (GPUs). However, the high cost of GPUs limits accessibility, making it crucial to compare GPU and central processing unit (CPU) based MLIPs under practical conditions. This study examines two popular MLIPs: the GPU-accelerated multi-atomic cluster expansion model and the CPU-based Gaussian approximation potentials model, applied to a battery electrolyte system known for its complex properties. By focusing on these models, our study evaluates differences in computational performance, resource efficiency, and accuracy in reproducing experimental properties. This rigorous benchmark provides insights into the trade-offs between GPU and CPU-based approaches in molecular simulations.
- Open Access
- 15 October 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Short-range machine-learning potentials for aqueous electrolyte solutions
Machine-learning potentials (MLPs) extend the time and length scales of atomistic simulations, enabling the study of complex systems such as electrolyte solutions. Yet most models face a trade-off between accuracy, computational cost, and the ability to capture long-range interactions. Large foundation models promise generality but often come with substantial overhead and energy demands. In contrast, compact, system-specific models may offer a more sustainable path for large-scale simulations. Here, we benchmark the MACE architecture on aqueous sodium chloride, systematically varying model size and the level of equivariance to assess their effect on accuracy, stability, and efficiency. We find that predictive accuracy of the investigated MLPs has little influence on physical observables but is crucial for stability, highlighting the potential of minimal, dedicated models for efficient simulations of electrolyte solutions.
- 15 October 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessShort-range machine-learning potentials for aqueous electrolyte solutions
Machine-learning potentials (MLPs) extend the time and length scales of atomistic simulations, enabling the study of complex systems such as electrolyte solutions. Yet most models face a trade-off between accuracy, computational cost, and the ability to capture long-range interactions. Large foundation models promise generality but often come with substantial overhead and energy demands. In contrast, compact, system-specific models may offer a more sustainable path for large-scale simulations. Here, we benchmark the MACE architecture on aqueous sodium chloride, systematically varying model size and the level of equivariance to assess their effect on accuracy, stability, and efficiency. We find that predictive accuracy of the investigated MLPs has little influence on physical observables but is crucial for stability, highlighting the potential of minimal, dedicated models for efficient simulations of electrolyte solutions.
- Open Access
- 27 August 2025
- Published by American Chemical Society (ACS) in ChemRxiv
GAP vs. MACE: Efficiency evaluation in a liquid electrolyte system
Machine learning interatomic potentials (MLIPs) have revolutionized molecular simulations, but as they evolve, so does the demand for advanced computing architectures, particularly graphics processing units (GPUs). However, the high cost of GPUs limits accessibility, making it crucial to compare GPU and central processing unit (CPU) based MLIPs under practical conditions. This study examines two popular MLIPs: the GPU-accelerated Multi-Atomic Cluster Expansion (MACE) model and the CPU-based Gaussian Approximation Potentials (GAP) model, applied to a battery electrolyte system known for its complex properties. By focusing on these models, our study evaluates differences in computational performance, resource efficiency, and accuracy in reproducing experimental properties. This rigorous benchmark provides insights into the trade-offs between GPU and CPU-based approaches in molecular simulations.
- 27 August 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessGAP vs. MACE: Efficiency evaluation in a liquid electrolyte system
Machine learning interatomic potentials (MLIPs) have revolutionized molecular simulations, but as they evolve, so does the demand for advanced computing architectures, particularly graphics processing units (GPUs). However, the high cost of GPUs limits accessibility, making it crucial to compare GPU and central processing unit (CPU) based MLIPs under practical conditions. This study examines two popular MLIPs: the GPU-accelerated Multi-Atomic Cluster Expansion (MACE) model and the CPU-based Gaussian Approximation Potentials (GAP) model, applied to a battery electrolyte system known for its complex properties. By focusing on these models, our study evaluates differences in computational performance, resource efficiency, and accuracy in reproducing experimental properties. This rigorous benchmark provides insights into the trade-offs between GPU and CPU-based approaches in molecular simulations.
- Open Access
- 24 June 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Photocatalytic Eosin Y Disproportionation as Driving Force for Terpene Cyclizations
A mechanistic paradigm for terpene cyclization driven by photocatalytic disproportionation of eosin Y in fluorinated alcohol media was discovered. By employing a 1:1 mixture of HFIP and PFTB, the neutral form of eosin Y (EH₂) undergoes efficient intersystem crossing to its triplet state, initiating hydrogen atom transfer (HAT) with ground-state EH₂. This generates two radicals, EH• and EH₃•, with EH• identified as the key species triggering polyene cyclizations. Spectroscopic and computational analyses revealed that hydrogen bonding with fluorinated alcohols stabilizes the otherwise unfavored open carboxylic acid form of eosin Y, enabling visible-light absorption and reactivity. The study clarifies the long-debated mechanism of eosin Y-mediated cyclizations, introducing a dual-radical generation strategy from a single photocatalyst molecule. These findings introduce a new paradigm in photoredox catalysis, offering a mild, selective, and sustainable approach to complex terpene synthesis and paving the way for broader applications in synthetic organic chemistry.
- 24 June 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessPhotocatalytic Eosin Y Disproportionation as Driving Force for Terpene Cyclizations
A mechanistic paradigm for terpene cyclization driven by photocatalytic disproportionation of eosin Y in fluorinated alcohol media was discovered. By employing a 1:1 mixture of HFIP and PFTB, the neutral form of eosin Y (EH₂) undergoes efficient intersystem crossing to its triplet state, initiating hydrogen atom transfer (HAT) with ground-state EH₂. This generates two radicals, EH• and EH₃•, with EH• identified as the key species triggering polyene cyclizations. Spectroscopic and computational analyses revealed that hydrogen bonding with fluorinated alcohols stabilizes the otherwise unfavored open carboxylic acid form of eosin Y, enabling visible-light absorption and reactivity. The study clarifies the long-debated mechanism of eosin Y-mediated cyclizations, introducing a dual-radical generation strategy from a single photocatalyst molecule. These findings introduce a new paradigm in photoredox catalysis, offering a mild, selective, and sustainable approach to complex terpene synthesis and paving the way for broader applications in synthetic organic chemistry.