AI Approach Helps Identify Materials for More Reliable Wearable Electronics

Molecular single crystals are promising for such stretchable electronics, yet the microscale links between strain and charge transport remain poorly defined. This uncertainty limits the design of sensors that can endure daily motion without signal loss. Researchers have now developed an AI framework to map how mechanical deformation reshapes charge mobility in these crystals, guiding materials selection for reliable, skin-mounted systems.

The team from Tsinghua University (Beijing, China) and Nanyang Technological University (Singapore) reports an approach that integrates molecular dynamics (MD) with machine learning (ML). The method, published in Wearable Electronics, relates tensile strain to charge mobility in organic molecular single crystals relevant to flexible devices. The goal is to clarify microscopic mechanisms that determine electrical performance under deformation.

The researchers trained an ML model to predict electronic coupling, defined as the ease with which charge hops between adjacent molecules, from molecular arrangements generated by MD. They simulated stretching in pentacene, a common organic semiconductor, to capture how structural changes alter charge-transport pathways. This enabled a direct map from applied strain to predicted mobility.

The study found that a 5% tensile strain along the crystal a-axis increased charge mobility by 55% along the b-axis. In contrast, a 5% strain along the b-axis reduced mobility by 33%. Topological analysis indicated that a-axis strain draws molecules closer to create a more efficient transport network, whereas b-axis strain separates them and hinders charge flow.

These results clarify when simpler, static models fail by overlooking strain-induced molecular motion. The work supports rational design of robust organic crystals that preserve high electrical performance during stretching and bending typical of wearable use. The authors highlight applications in next-generation skin-mountable sensing and computing systems.

"Flexible electronic components are prone to mechanical stresses, such as stretching and bending, during operations. Ensuring their seamless functionality necessitates both excellent flexibility and reliable electrical performance," said Shuzhou Li, senior and corresponding author.

"The results indicated that strain along the b-axis significantly enhances molecular motion, broadening the distribution of electronic couplings and further limiting charge transport. This finding clarifies the boundaries of applicability for simpler models that neglect this dynamic effect, which can lead to inaccurate predictions," said Xi Chen, lead author.

Related Links
Tsinghua University
Nanyang Technological University