Special Physical Chemistry Seminar : Interplay of Digitization and Simulation in the Quest for New Functional Materials

Abstract:

Functional materials design is increasingly limited not by ideas, but by the accessibility, provenance,
and reusability of the data we already generate. I will present an ontology-driven digital
infrastructure that converts heterogeneous experimental and simulation outputs into a coherent,
machine-actionable knowledge space. The GlasDigital ontology captures the full composition–
processing–structure–properties chain and enables interoperable data integration across synthesis
and characterization workflows.1
On this digital backbone, I will highlight a closed-loop workflow for AI-driven glass discovery
coupling semantic data spaces with high-throughput modelling and targeted synthesis. Highthroughput
molecular dynamics maps composition–property landscapes, while hybrid and iterative Δ-
learning reduces simulation bias using focused experimental measurements.2 Candidate compositions
are validated via an in-house robotic melting workflow, tightening the feedback loop from data to
models to experimentally realizable materials.1,2
Finally, I will discuss advances in Density Functional Embedding Theory (DFET) for molecular
and periodic systems, implemented efficiently with Gaussian basis sets and compatible with correlated
wave-function methods for accurate local descriptions at tractable cost.3,4 Extending embedding into
the time domain, real-time TDDFT embedding (RT-TDDFET) simulates linear and nonlinear optical
responses by time-propagating the embedded electron density under an explicit time-dependent
external field, enabling absorption spectra and strong-field phenomena at reduced computational
expense.3,4
References
[1] Chen, Y.-F.; Arendt, F.; Bornhöft, H.; de Camargo, A. S. S.; Deubener, J.; Diegeler, A.;
Gogula, S.; Contreras Jaimes, A. T.; Kempf, S.; Kilo, M.; Limbach, R.; Müller, R.; Niebergall,
R.; Pan, Z.; Puppe, F.; Reinsch, S.; Schottner, G.; Stier, S.; Waurischk, T.; Wondraczek, L.;
Sierka, M. Ontology-Based Digital Infrastructure for Data-Driven Glass Development. Adv.
Eng. Mater. 2025, 27, 2401560. https://doi.org/10.1002/adem.202401560.
[2] Arendt, F.; Waurischk, T.; Reinsch, S.; Sierka, M. AI-Driven Glass Discovery: From Semantic
Data to Autonomous Synthesis. Adv. Eng. Mater. 2026, submitted.
[3] Sharma, M.; Sierka, M. Efficient Implementation of Density Functional Theory-Based Embedding
for Molecular and Periodic Systems Using Gaussian-Type Orbitals. J. Chem. Theory
Comput. 2022, 18, 6892–6904. https://doi.org/10.1021/acs.jctc.2c00380.
[4] Sharma, M.; Franzke, Y. J.; Holzer, C.; Pauly, F.; Sierka, M. Density Functional Theory for
Molecular and Periodic Systems in TURBOMOLE: Theory, Implementation, and Applications.
J. Phys. Chem. A 2025, 129, 9062–9083. https://doi.org/10.1021/acs.jpca.5c02937.
1

Orgenizer: Prof. Barak Hirshberg