Title: Organic and Perovskite Solar Cells: Upscaling Challenges and the Research Revolution via AI-Driven Automatic Experiments
Time：August 30th 10:00am－11:00am
Reporter: Prof. Doojin Vak
Host: Prof. Luo Qun
Location: Conference Room A718
Abstract：
Organic photovoltaics (OPV) and organic-inorganic hybrid perovskite photovoltaics (PePV) are promising PV technologies that can be manufactured using industrial roll-to-roll (R2R) printing which is a widely used mass-production technique for low-cost products. These emerging PV technologies have been making exciting progress toward commercial applications and the efficiencies of the laboratory cells (19.2 % for OPV and 26.1 % for PePV) are already high enough to enter the PV market. However, the efficiency of R2R-produced PV still lags behind those achieved for champion laboratory cells. This is attributed to the materials, processes and device configurations developed for research purposes not being readily translatable to R2R printing, with significant material and process optimisation required to achieve compatibility with scalable R2R processes. The time-consuming optimisation process has delayed the translation of these technologies to the marketplace, necessitating a new revolutionary research method.
Therefore, an automated R2R research platform has been developed to accelerate the progress of R2R-fabricated solar cell technologies. A bespoke R2R coater was developed to optimise formulations and fabrication parameters including deposition conditions, coating speed, and annealing temperature. An in-situ formulation technique was introduced to fabricate over 10,000 unique cells a day via unmanned operation, and an automated R2R PV measurement unit has also been developed to test this number of cells in a single day. This revolutionary approach has enabled the rapid progress of R2R-fabricated solar cells, resulting in vacuum-free R2R-fabricated PePV and OPV devices achieving PCEs of 15% and 11%, respectively, both of which are record PCEs in their class. All processing and testing parameters are digitalised to be used as training data for machine learning (ML). The recent progress and the potential of ML-assisted research will be discussed in this talk.
Doojin Vak is a principal research scientist who works on printed photovoltaics (PV) at the Commonwealth Scientific and Industrial Research Organization (CSIRO) in Australia. He received his PhD in Materials Science and Engineering from the Gwangju Institute of Science and Technology (GIST), where he focused on conjugated polymers for optoelectronic applications. He began his research career on printing-based PV fabrication at the University of Melbourne as a Korea Research Foundation (KRF) postdoctoral fellow. Since joining CSIRO in 2010, he has been exploring various industrial upscaling techniques, especially with the roll-to-roll process, to fabricate organic and perovskite PV with the goal of commercialization. He has also been exploring machine learning (ML)-assisted research to accelerate the development of printed PV technologies
downloadFile