Professor Christoph J. Brabec presented a thought-provoking talk titled “Material Discovery for Emerging-PV Technologies” at the SENECA Talks, a monthly seminar series on Solar Energy Conversion and Storage organized by EnergyVille/imec/UHasselt (imo-imomec). The talk focused on the challenges and advancements in material discovery for organic and perovskite photovoltaics. These materials hold immense potential for future solar cell technologies, but optimizing them requires navigating a complex, multi-dimensional parameter space. Dr. Brabec presented how machine learning techniques like Gaussian Process Regression (GPR) and Bayesian Optimization (BO) can efficiently predict new materials and optimize processing, significantly reducing experimental needs. For instance, in a four-dimensional space (solvent, donor-acceptor ratio, spin speed, concentration), just 30 samples were sufficient to find the optimum within a large parameter space explored using a 10% grid search. The talk extended the discussion beyond simple optimization, exploring the potential of BO for discovering entirely new molecules. Professor Brabec showcased the research campaign that successfully identified novel molecular semiconductors not previously reported in the scientific literature, outperforming the current state-of-the-art materials in terms of performance.
In a rather unique turn of events, the talk was interrupted by a safety alarm, leading to a building evacuation. Professor Brabec, ever the optimist, seized this opportunity to capture a memorable group photo with Dr. Sudhanshu Shukla, Dr. Yinghuan Kuang and Dr. Tom Aernouts from imec.
Professor Christoph J. Brabec presented a thought-provoking talk titled “Material Discovery for Emerging-PV Technologies” at the SENECA Talks, a monthly seminar series on Solar Energy Conversion and Storage organized by EnergyVille/imec/UHasselt (imo-imomec). The talk focused on the challenges and advancements in material discovery for organic and perovskite photovoltaics. These materials hold immense potential for future solar cell technologies, but optimizing them requires navigating a complex, multi-dimensional parameter space. Dr. Brabec presented how machine learning techniques like Gaussian Process Regression (GPR) and Bayesian Optimization (BO) can efficiently predict new materials and optimize processing, significantly reducing experimental needs. For instance, in a four-dimensional space (solvent, donor-acceptor ratio, spin speed, concentration), just 30 samples were sufficient to find the optimum within a large parameter space explored using a 10% grid search. The talk extended the discussion beyond simple optimization, exploring the potential of BO for discovering entirely new molecules. Professor Brabec showcased the research campaign that successfully identified novel molecular semiconductors not previously reported in the scientific literature, outperforming the current state-of-the-art materials in terms of performance.
In a rather unique turn of events, the talk was interrupted by a safety alarm, leading to a building evacuation. Professor Brabec, ever the optimist, seized this opportunity to capture a memorable group photo with Dr. Sudhanshu Shukla, Dr. Yinghuan Kuang and Dr. Tom Aernouts from imec.