Discovering hidden relations in organic semiconductor composites – a concept to accelerate the development of organic photovoltaics
Brabec C (2021)
Publication Type: Conference contribution, Conference Contribution
Publication year: 2021
Original Authors: Christoph Brabec
Event location: Online, Spanien
DOI: 10.29363/nanoge.hopv.2021.087
Abstract
While c-Si appears untouchable as leading PV main stream technology for a longer time, many of the new applications which rely on flexibility, transparency, colour management, integrability or simply elegant appearance require novel photovoltaic materials and technologies.
Organic photovoltaics (OPV), like other emerging thin film PV technologies, are not yet part of this global TW scenario. The first printed PV products were launched in 2008/2009 for portable chargers at an efficiency of about 2 %. Despite the rather low performance at that time, these first products already showed the characteristic “OPV features” like flexibility, transparency and colour variability. Since then, the printed PV community has concentrated on developing novel material systems for higher efficiency – a development which was outstandingly successful. In the last 10 years. Organic solar modules with close to 13 % efficiency were certified in 2020, while single junction cells already reach 18 % power conversion efficiency. Despite the great process, OPV is still facing multiple challenges, from lifetime to cost and to environmental aspects. The most urgent question therefore is – how can we accelerate the development of organic solar cells towards a market ready product.
In this talk I will discuss concepts and strategies to speed-up the development of OPV technologies towards an earlier deployment on the market. Automated, robot-based research lines with shared interfaces to multi-objective machine learning based optimization routines are introduced as a powerful concept to accelerate the development of new materials towards markets. Despite the commonly accepted understanding that machine learning algorithms can accelerate optimization, we demonstrate that physical model-based AI is superior in discovering hidden relations. We will demonstrate that in specific cases it is becoming possible to predict efficiency or even stability of organic solar cells based on a simple absorption measurement.
Authors with CRIS profile
How to cite
APA:
Brabec, C. (2021). Discovering hidden relations in organic semiconductor composites – a concept to accelerate the development of organic photovoltaics. In Proceedings of the 13th Conference on Hybrid and Organic Photovoltaics (HOPV21). Online, Spanien, ES.
MLA:
Brabec, Christoph. "Discovering hidden relations in organic semiconductor composites – a concept to accelerate the development of organic photovoltaics." Proceedings of the 13th Conference on Hybrid and Organic Photovoltaics (HOPV21), Online, Spanien 2021.
BibTeX: Download