Machine‐Learning‐Assisted Process Optimization for High‐Performance Organic Thermoelectrics

Jeehyun Jeong,Suyeon Park,Juhyung Park,Jeonghan Song,Jeonghun Kwak

DOI: https://doi.org/10.1002/aenm.202403431

IF: 27.8

2024-11-28

Advanced Energy Materials

Abstract:Machine learning‐assisted design of experiments is applied to optimize the processing conditions for enhancing the thermoelectric performance of a thiophene‐based semicrystalline polymer system. This method allows for efficient sampling and screening of various processing parameters across a large, multi‐dimensional parameter space, enabling the prediction of optimal conditions and maximizing the thermoelectric power factor. Optimizing the processing conditions of conjugated polymers is a key step in improving the performance of various organic electronic devices, including organic thermoelectric (TE) devices that are capable of harvesting energy from waste heat. However, the inherent structural and energetic disorders in these materials and their unpredictable property changes after processing complicate the optimization principles, requiring numerous trial‐and‐error experiments to maximize the TE performance. Here, a machine‐learning (ML)‐based design of experiments approach is introduced to address these challenges, and its effectiveness is demonstrated using a representative thiophene‐based doped polymer system for organic TE. This approach not only helps to quantitatively understand how each processing parameter affects the TE properties of the polymer but also facilitates the prediction of optimal processing conditions, leading to achieve maximum TE performance with minimal experiments within a large parameter space. Furthermore, the ML‐based approach is validated by revealing the origins of the power factor maximization at the ML‐predicted processing conditions through analysis of the morphology and electronic states of the doped polymer films. The proposed methodology is applicable to the majority of polymer systems for organic thermoelectric energy conversion and will provide guidelines for determining optimal process conditions and improving TE performance with minimal effort.

materials science, multidisciplinary,chemistry, physical,physics, applied, condensed matter,energy & fuels