Optimizing Thermoelectric Performance of Graphene Antidot Lattices Via Quantum Transport and Machine-learning Molecular Dynamics Simulations

Strategic Insights -

Researchers from Bohai University have made a groundbreaking discovery in the field of machine learning, specifically in the optimization of thermoelectric performance of graphene antidot lattices (GALs). According to the study, the introduction of antidots in GALs effectively decouples lattice and electronic transport, leading to a significant violation of the Wiedemann-Franz law. The research team used a combination of machine-learning molecular dynamics simulations and quantum transport calculations to optimize the performance of GALs, with the goal of achieving high-performance thermoelectric energy conversion.

Key Takeaways:

  • The researchers found that the introduction of antidots in GALs decouples lattice and electronic transport, leading to a significant violation of the Wiedemann-Franz law.
  • The optimal thermoelectric performance of GALs occurs at intermediate lattice side length (L) and antidot radius (R) values, closely correlated with peak power factor values.
  • The research team used a combination of machine-learning molecular dynamics simulations and quantum transport calculations to optimize the performance of GALs.
  • The study showed that the maximal thermoelectric figure of merit (ZT) values approach 2 at room temperature, highlighting GALs as promising candidates for high-performance thermoelectric energy conversion.
  • The research has been peer-reviewed and published in Physical Review Materials.
  • The study's lead author is Zheyong Fan from Bohai University, with additional authors including Yang Xiao, Yuqi Liu, Zihan Tan, Bohan Zhang, Ke Xu, Haikuan Dong, Shunda Chen, and Shiyun Xiong.

Statistics:

  • The optimal thermoelectric performance of GALs occurs at intermediate lattice side length (L) and antidot radius (R) values, with L=20-30 nm and R=7-10 nm.
  • The maximal thermoelectric figure of merit (ZT) values approach 2 at room temperature.
  • The study used a combination of machine-learning molecular dynamics simulations and quantum transport calculations to optimize the performance of GALs.
  • The research team simulated 100 GALs with different L and R values, with the goal of identifying the optimal configuration.

Sources:

  • NewsRx, "Findings in Machine Learning Reported from Bohai University," Journal of Engineering, October 20, 2025, p 915.
  • Research paper: "Optimizing Thermoelectric Performance of Graphene Antidot Lattices Via Quantum Transport and Machine-learning Molecular Dynamics Simulations," Physical Review Materials, 2025;9(8).
  • Authors: Zheyong Fan, Yang Xiao, Yuqi Liu, Zihan Tan, Bohan Zhang, Ke Xu, Haikuan Dong, Shunda Chen, and Shiyun Xiong.