The AMC Seminar is back! After the summer break, our Winter Term Seminar Series on Atomistic Modeling-related topics will resume, taking place every Tuesday.
This term’s second speaker, Oliver Hofmann, is Associated Professor at the TU Graz. His research group focuses on the computational discovery of novel materials and material combinations, with a particular emphasis on organic thin films and inorganic/organic interfaces using machine learning techniques.
Polymorphism critically influences charge-carrier mobility in organic semiconductors, yet high-performance crystal structures are often metastable and experimentally elusive. Hofmanns talk on “Computationally Predicting Targeted Growth of High-Mobility Metastable Phases” introduces a machine-learning-accelerated structure search algorithm (SAMPLE) that maps polymorph landscapes and uses Monte Carlo growth simulations with control theory to guide processing conditions for selectively stabilizing high-mobility phases.
Date: Tuesday, October 21, 2025, 3:30 pm
Location: MIBE Lecture Hall
Abstract:
Polymorphism plays a decisive role in defining the charge-carrier mobility and overall performance of crystalline organic semiconductors. However, the crystal structures that exhibit the most desirable charge-transport properties are rarely the thermodynamically most stable ones, making them notoriously difficult to access experimentally. This challenge limits both fundamental studies and the rational design of next-generation organic electronic devices.
In this contribution, we present a computational and theoretical framework to systematically identify and stabilize high-mobility polymorphs. Our machine-learning-accelerated structure search algorithm (SAMPLE) delivers a quasi-exhaustive map of the potential polymorph space for molecular layers on surfaces, including metastable structures. To bridge computation and experiment, we combine kinetic Monte Carlo growth simulations with optimal-control theory to derive practical processing guidelines. These include recommendations for substrate choice, temperature ramps, and deposition conditions that maximize the likelihood of obtaining a desired metastable phase. We illustrate our approach with case studies where the predicted high-mobility phase can be selectively stabilized, providing a proof-of-concept for rationally steering crystal growth.