Active Learning Accelerates Design and Optimization of Hole-Transporting Materials for Organic Electronics

Area of Science:

• Materials Science
• Organic Electronics
• Computational Chemistry

Background:

• Data-driven methods, including machine learning (ML), are crucial for accelerating materials design in organic electronics.
• Developing accurate ML models requires extensive data management to cover the complexity of chemical spaces.
• Active learning (AL) offers an efficient strategy to navigate these complex search spaces by prioritizing data selection.

Purpose of the Study:

• To apply an active learning (AL) workflow for efficient identification of materials candidates for hole-transport layers (HTLs) in organic light-emitting diodes (OLEDs).
• To leverage AL to minimize the computational resources needed for exploring diverse materials libraries.

Main Methods:

• Implementation of an active learning (AL) workflow tailored for materials discovery.
• Integration of multiple optoelectronic parameters within the AL strategy.
• Systematic screening of a materials library to identify potential HTL candidates.

Main Results:

• Demonstrated the efficacy of the AL workflow in navigating complex materials' chemical space.
• Successfully identified promising materials candidates for HTLs in OLEDs.
• Reduced the number of required computations for materials exploration.

Conclusions:

• The developed AL workflow provides an efficient mechanism for screening materials for organic electronics.
• This approach enables the identification of high-performance materials prior to extensive experimental validation.
• Paves the way for accelerated discovery of superior materials for OLED applications.