(ECNS) -- Researchers at the University of Science and Technology of China (USTC), working with colleagues from Southern University of Science and Technology and City University of Hong Kong, have developed an artificial intelligence (AI)-guided framework to identify promising interface materials for perovskite solar cells, USTC said Tuesday.
The framework integrates AI, high-throughput density functional theory (DFT) screening and chemical expertise to identify self-assembled monolayer (SAM) materials that are both practical to synthesize and stable at the interface. The findings were recently published in Nature Synthesis.
Designing molecules for solar cell interfaces requires balancing several factors, including energy-level alignment, interfacial interactions, synthetic accessibility and compatibility with solution processing. A molecule that scores well on one computational measure may not be suitable for practical devices. Conventional trial-and-error approaches are time-consuming and often struggle to optimize several competing properties.
To address this challenge, the researchers incorporated synthetic accessibility into the screening process, linking computational design with experimental requirements from the outset.
Starting with carbazole derivatives listed in the PubChem database, the team extracted 308 functional fragments, which could be combined into more than 90,000 potential molecular structures. They selected 24 fragments and used modular assembly to construct 300 candidate molecules.
An initial screening based on energy levels narrowed the pool to 124 candidates. The researchers then ranked them with an integrated scoring system based on eight molecular descriptors, including synthetic accessibility, molecular energy levels, hole reorganization energy, aromaticity and heteroatom content, to decide which to test experimentally.
Using this framework, the team selected and synthesized three highly ranked molecules. Their ranking by performance in solar cell devices closely matched the scoring system's predictions, which the researchers said validated the approach.
The highest-ranked molecule enabled a perovskite solar cell to achieve a power conversion efficiency of 27.2%. The device kept about 95% of its initial efficiency after 2,000 hours of operation at 65 degrees Celsius (149 degrees Fahrenheit) under continuous illumination and maximum power point tracking (MPPT) conditions.
The researchers said the modular framework could be extended to other interface material systems, offering a new approach to the design and discovery of efficient and stable optoelectronic materials.
(By Evelyn)
京公网安备 11010202009201号