Short Bio

Johannes Schörghuber is a PhD student and FWF-funded project assistant in the Theoretical Materials Chemistry research group led by Prof. Georg K. H. Madsen at TU Wien. He is part of both the FWF Cluster of Excellence MECS and the iCAIML doctoral college. His research is concerned with the electrochemical interfaces, integrating machine-learned interatomic potentials in order to achieve simulations at scale. Johannes earned a BSc in Technical Chemistry in 2020 and a Dipl.-Ing. degree in Computational Science and Engineering in 2023, both completed at TU Wien. He already started his work in the field of computational electrochemistry with his master’s thesis titled “Electrostatic interactions in neural network force fields”.

PhD Project - Atomistic Simulations of Electrochemical Interfaces

Supervised by Georg K. H. Madsen

In his research, Johannes simulates electrochemical interfaces at the atomic scale, utilizing machine-learning interatomic potentials (MLIPs) to scale to the required system sizes and simulation length scales while maintaining the accuracy of the underlying ab-initio calculations. Since generating training data is computationally intensive, efficient sampling is important. For this reason he initially focused on uncertainty quantification for MLIPs and the development of active learning procedures. Using these methods the project now focuses on the modelling of interfaces under applied electric fields and using MLIPs to simulate interfaces under such conditions.

Publications and Conferences

Journal Papers

• Heid E., Schörghuber J., Wanzenböck R. and Madsen G. K. H. “Spatially Resolved Uncertainties in Machine Learning Potentials“ J. Chem. Inf. Model., 2024, 64, 6377-6387.
• Mai X., Wanz Z., Pan L., Schörghuber J., Kovács P., Carrete J., Madsen G. K. H. “Computing Anharmonic Infrared Spectra of Polycyclic Aromatic Hydrocarbons Using Machine-Learning Molecular Dynamics“ arXiv preprint arXiv:2503.05120.
• Schörghuber J., Bučková N., Heid E., Madsen G. K. H. “From flat to stepped: active learning frameworks for investigating local structure at copper-water interfaces“ Phys. Chem. Chem. Phys., 2025, 27, 9169-9177.
• Bučková N., Unglert N., Schörghuber J., Heid E., Berland K. Madsen G. K. H. “The density isobar of water: A comparative study of vdW-DF-cx and RPBE-D3“ ChemRxiv preprint 2025-23fzc.

Posters (without proceedings)

• Schörghuber J., Carrete J., Madsen G. K. H. “Electron-Passing Neural Networks”, presented at Joint TACO-NanoCat Conference 2023, Vienna, Austria.
• Schörghuber J., Bučková N., Heid E., Madsen G. K. H. “Active learning for modelling the Cu-H2O interface”, presented at CoE MECS Kickoff event 2024, Vienna, Austria.

Presentations

• “Active learning for modelling the Cu-H2O interface”. CoE MECS Retreat. Contributed Talk. Waidhofen a.d. Ybbs, 2024.