Biomimetic Designs of Organic Electrochemical Transistors for Human-Interfaced Biosensing and Neuromorphic Computing

The use of electronic devices for acquiring biological information and delivering therapeutic interventions relies on two aspects: the effective acquisition of physiological information through direct contact with soft bio-tissues, and the high-throughput processing of the acquired data using machine learning. To ensure high-quality signal transductions, the interfaces between bioelectronic devices and bio-tissues must combine signal amplification with stable and conformable contact. Organic electrochemical transistors (OECTs) have been developed as one of the most advanced technologies for high-performance bio-sensing and neuromorphic computing. Moreover, the fastest and most energy-efficient data processing through machine learning is directly on the human body. However, the rigid mechanical properties and the lack of tissue/skin adhesion from transistors largely prevent the formation of such intimate and long-term stable bio-interfaces. In the first part of my talk, I will introduce our material and device designs for introducing three highly important biomimetic properties onto OECT-based biosensors—stretchability, tissue-like softness, and bioadhesive properties. Our rationale designs from the material to the device level allow the realization of these properties with state-of-the-art electrical performance. I will also introduce the strategies and advantages of using these new biomimetic properties in bioelectrical and biochemical sensing. In the second part of my talk, I will introduce our research in imparting intrinsic stretchability onto neuromorphic devices that can provide state-of-the-art computing performance. I will also show the practical applicability of this device for implementing machine-learning computing and algorithms for health data analysis, when the computing hardware is under human-body-induced deformation.