Perovskite on silicon (Si) tandem solar cells are predicted by the ITRPV to be the first Si-based tandem solar cells to enter the market. Their superior power conversion efficiency promises a step change in module efficiency that is not possible with silicon only technologies, as shown by Oxford PV’s recent certified world record of 28.6% in a full area cell. With its integrated pilot line at Brandenburg, Germany, Oxford PV is currently ramping up to high-volume production of commercial M6 tandem wafer product.

Perovskite tandem commercialisation has taken place in a very short period of time compared to other photovoltaic technologies. This rapid progress has been possible due to the development of in-house protocols and the statistical analysis of the large datasets generated. Some of the advanced characterisation at Oxford PV and how they inform decision making on process and technology designs in view of cell efficiency and longevity targets will be presented. The measurements can be categorised according to their speed, the complexity of data analysis and the value of the measured parameters – increasing measurement speed and analysis allows for more insights on more samples more frequently, and so a key part of industrial solar R&D involves innovations on measurement protocols. In this talk, there will be examples of the advanced characterisation suite that have developed at Oxford PV to maximise the learning for as many samples as possible whilst focusing on the most important parameters, and eliminating redundancies. Upon generation of high quality datasets, machine learning tools are used to improve our statistical understanding of trends we see with development of the Oxford PV tandem cell technology. As well as cell efficiency enhancements, these complimentary research methods are just as suited for improving cell reliability, for which the perovskite field has made good steps in the last 12 months in aligning on predictive accelerated stressing protocols. The big data approach is also suitable for identifying process sensitivities in view of designing scalable processes, a critical part of R&D in an industrial setting.

Finally, sustainability, both in financial and ecological terms, is a crucial aspect of manufacturing at large scale. Leveraging advanced characterisation and machine learning can expedite from small scale proof of concepts towards achieving comparable KPIs with high-TRL production ready processes. Oxford PV is committed to bring to the market the most sustainable technology in the most sustainable way; examples of the initiatives Oxford PV has implemented to make that happen will be presented.