Vortrag im Rahmen der Chemisch Physikalischen Gesellschaft
Understanding and predicting organic reactivity across molecular scales remains a central challenge in chemistry. We demonstrate how a synergistic interplay between experiment and theory enables mechanistic insight and rational reaction design.
Using representative case studies, including novel sulfonium rearrangements, unconventional Diels–Alder or photoredox-catalyzed C–H monoalkylation reactions, we combine density functional theory (DFT) with wave function-based methods to elucidate reaction mechanisms, rationalize experimental observations, and identify more efficient reaction pathways.
To address the growing complexity of modern catalytic systems, we further introduce a computational strategy for large molecular systems (~250 atoms), where conventional approaches become impractical. By integrating GPU-accelerated energy and gradient evaluations with robust geometry optimization algorithms, we achieve accurate and efficient exploration of realistic reaction models without resorting to oversimplified representations.
Beyond quantum chemical modeling, we present a data-driven framework for interpreting reaction outcomes using the Buchwald–Hartwig amination as a case study. By employing statistically grounded models tailored to the structure of high-throughput experimental data, we extract interpretable relationships between reaction components and yields, providing mechanistically meaningful insight and improving predictive capability.
Together, these approaches define an iterative workflow in which computation and experiment inform and refine one another. This closed-loop strategy not only explains existing observations but also guides the discovery and optimization of new organic reactions, advancing a predictive and mechanistically grounded paradigm in chemistry.
