Elucidating the role of carrier transport layers on perovskite photovoltaics' stability through automated experimentation and a machine-learning-assisted analysis

Despite recent clean-energy initiatives, the US currently ranks second in global yearly total CO2 emission, totaling 5.01 billion metric tons in 2022. As such, developing stable and low-cost solar cells is critical. An emerging class of material, named halide perovskites, have the potential to deliver high-performing and low-cost solar cells or be combined with silicon, which dominates ~90% of the market, to boost performance.

However, device degradation under environmental stressors (e.g., humidity, oxygen, and temperature) still precludes their commercialization. Concomitantly, the implementation of low-cost polymer materials to help collecting the current produced by the devices while securing long lifetime is urgently needed. Therefore, a primary research goal of this project is to advance the state-of-knowledge of halide perovskite solar cells by developing high-performance devices with enhanced stability upon exposure to environmental stressors.

The methodology to be implemented combines automated experiments with a machine learning driven analysis that will help identifying polymers that can enhance device stability and performance. Broadly, the research component of this project will impact the development of future solar cells by resolving the ideal combination of stressors. The outreach impacts will help female students securing leading positions in STEM by provide mentoring and training, including research experiences to graduate and undergraduate students.

The Materials Science and Engineering at UC Davis curriculum will be enhanced by adding new lectures into a core course for undergraduates. All ML codes will be made available in GitHub and the scientific findings will be disseminated through peer-reviewed publications.

This research aims to investigate three challenges related to HP solar cells: (1) determine the combined effects of environmental stressors on perovskites’ stability; (2) quantify hole transport material/halide perovskite (HTM/HP) interface stability using novel, state-of-the-art conducting polymers that enable matching the energetics of band alignment required for high-performing photovoltaics; and (3) demonstrate solar cells with >95% performance retention for >1,000 h. Because controlled stability tests conventionally require extremely time-consuming tasks, automated and high-throughput experiments will be used.

The information acquired will, in turn, inform machine learning algorithms to forecast stable HTM/HP interfaces. First, in situ optical measurements will be performed under distinct environmental conditions to elucidate the individual and combined effects of humidity, oxygen, and temperature on HP degradation. Second, the HPs will be interrogated via in situ X-ray diffraction to resolve how structural changes correlate with and affect device performance.

Third, the effects of the abovementioned stressors on the HTM/HP interfaces using six cutting-edge, novel HTMs will be quantified while gathering valuable data for training machine learning algorithms. Fourth, this information will identify the most stable HTM/HP interfaces through predictive, physics-informed models. The results generated will guide the future development of application-dependent encapsulation strategies for HP solar cells with prolonged lifetime.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.