Perovskite solar cells (PSCs) have competitive efficiency but suffer from stability issues. The energy loss processes involve pinholes, defects, or poor device quality, which may both lower efficiency and make the device more prone to instability. A recent study published in Nature Communications,(1) analyzes a large dataset of Maximum Power Point Tracking (MPPT) operational ageing…

Written by

×

Stability Issues in Perovskite Solar Cells solved by Machine Learning

Perovskite solar cells (PSCs) have competitive efficiency but suffer from stability issues. The energy loss processes involve pinholes, defects, or poor device quality, which may both lower efficiency and make the device more prone to instability.

A recent study published in Nature Communications,(1) analyzes a large dataset of Maximum Power Point Tracking (MPPT) operational ageing data to understand stability. There is a correlation between the maximum power conversion efficiency (PCE) and the PCE loss after 150 hours of ageing. It revealed a simple conclusion that more efficient cells tend to have higher stability.

MPPT is a technique used in solar power systems to optimize the efficiency of converting sunlight into electrical power. MPPT is achieved by continuously adjusting the electrical load to ensure that the solar panel operates at its maximum power point, where the maximum power output is obtained. This is important because the optimal operating point for a solar panel can vary depending on factors such as the intensity of sunlight, temperature, and electrical load.

A machine learning method called self-organising map (SOM), which is a clustering method to classify different degradation curves, is used. The analysis identified four distinct degradation curve shapes: initial gain, slow exponential decay, medium exponential decay, and fast-exponential decay. There is a correlation between the occurrence of specific shape clusters and the maximum reached PCE. Clustering is a type of machine learning method, where no training data is used, and all the data is used for extracting the information.

A dataset of 2,245 MPPT curves of various device architectures, layers, and perovskite composition materials is analyzed. The ageing tests were performed under controlled conditions in a custom-built High-throughput Ageing System. The dataset includes devices with different electron transport materials, hole transport materials, and top electrodes.

The study analyzed a dataset of MPPT-ageing data to investigate the relationship between the maximum power conversion efficiency (PCE) reached during the first 150 hours of operational MPPT ageing and the relative loss after 150 hours (ΔPCE) as a metric for perovskite solar cells (PSCs) stability. The findings suggested that higher maximum PCE values in the first 150 hours of testing were associated with lower mean ΔPCE, indicating that efforts to improve PSC efficiency could enhance stability.

As solar cells with higher maximum efficiency also offer a lower loss in efficiency in the first 150 hours, indicating higher stability. The study clustered the MPPT curves based on their shape and found that the degradation curve shape was related to both the cells’ maximum PCE group and stability. Cells with an initial gain degradation curve shape showed lower ΔPCE and were more frequently observed in cells with higher maximum PCE, suggesting that this shape type could serve as an early indicator for stability.

(1) Hartono, N.T.P., Köbler, H., Graniero, P. et al. Stability follows efficiency based on the analysis of a large perovskite solar cells ageing dataset. Nat Commun 14, 4869 (2023). https://doi.org/10.1038/s41467-023-40585-3

Leave a comment