From Simulation to Autonomous Laboratory Preparation: Machine Learning-Driven Discovery of Porous Materials and Their Composites
Find out more about subscribing to add all events.
From Simulation to Autonomous Laboratory Preparation: Machine Learning-Driven Discovery of Porous Materials and Their Composites
Tue, 10/06/2025 - 13:00 to 14:00
Location:
Speaker:
Dr Maciej Haranczyk
Affiliation:
IMDEA Materials Institute (Spain)
Synopsis:
Nanoporous materials—especially metal–organic frameworks (MOFs) and natural clays—are emerging as key platforms for chemical separations and catalysis. MOFs provide virtually unlimited, designable architectures through modular coordination chemistry, while clay performance can be fine-tuned post-synthetically by thermal or chemical treatments. The central challenge is selecting, or inventing, the right material for a given application.
My research tackles this problem with machine-learning strategies that accelerate discovery across both data-rich and data-poor regimes. In the data-rich case, millions of hypothetical MOF structures require rapid screening. By blending computational geometry with topological data analysis, we create multiscale descriptors of porosity, then train machine-learning models on simulated adsorption data to isolate the most promising candidates in seconds.
When data are scarce, new experiments become essential. To address this, we have built a Materials Acceleration Platform that closes the loop between synthesis and prediction. Focused on polymer–clay composites, the platform integrates compounding, injection moulding, automated mechanical testing, and machine-learning-guided experiment selection, allowing us to converge on optimal formulations in only a handful of iterations. Together, these approaches compress the timeline from concept to deployment and transform a vast search space into a guided pathway toward high-performance porous materials.
Biography:
Dr. Maciej Haranczyk received a PhD degree in Chemistry from University of Gdansk in Poland in 2008. During his graduate studies, he received numerous research fellowships to collaborate with researchers at Pacific Northwest National Laboratory (M. Gutowski), University of Southern California (A. Warshel), University of Sheffield (P. Willett, J. Holliday), and others. After his graduate school, he moved to Lawrence Berkeley National Laboratory, which had offered him a Glenn T. Seaborg Postdoctoral Fellowship. He was then hired into a tenured Research Scientist position (2010) and subsequently promoted to a Staff Scientist position (2014). He joined IMDEA Materials Institute as a Senior Researcher in late 2015 supported by Ramon y Cajal award. He currently leads the Accelerated Materials Discovery Group, managing a robot-equipped materials research laboratory. His research activities are focused on data-driven materials discovery. His work effectively combines novel materials informatics and machine learning approaches with traditional computational material science techniques such molecular simulations as well as experimentation. These techniques are applied to discover materials for a wide spectrum of applications from energy, through separations in petrochemical and animal feed industry to sustainable and safe plastics.