Process-Function Data Mining for the Discovery of Solid-State Iron-Oxide PV

Area of Science:

• Materials Science
• Data Science
• Renewable Energy

Background:

• Data mining tools are valuable for analyzing large material datasets from high-throughput methods.
• Traditional structural descriptors for analysis are often difficult to obtain and optimize.
• Developing efficient photovoltaic materials requires advanced analytical techniques.

Purpose of the Study:

• To investigate the use of deposition process parameters as descriptors for analyzing photovoltaic data.
• To demonstrate that deposition parameters can effectively model and predict photovoltaic performance.
• To enable the discovery and fabrication of high-performance photovoltaic materials using data mining.

Main Methods:

• Fabrication of iron oxide solar cell libraries using varied deposition parameters.
• Measurement of photovoltaic performance for each fabricated cell.
• Development of predictive models using genetic programming and stepwise regression with deposition parameters.

Main Results:

• Models successfully identified critical deposition parameters for enhancing photovoltaic performance.
• Fabrication of an iron oxide library based on model predictions yielded superior cell performance.
• Demonstrated the efficacy of deposition parameters in predicting and optimizing solar cell efficiency.

Conclusions:

• Deposition process parameters are effective descriptors for data mining in photovoltaics.
• This method facilitates the discovery and fabrication of high-performance solar cells.
• The approach offers a promising pathway for accelerating materials discovery in renewable energy.