Researchers at Xi’an Jiaotong University, the University of Hong Kong and Xi’an University of Science and Technology have developed a dual-function organic transistor, both a sensor and processor.
Electronics engineers strive to create efficient brain-inspired hardware for artificial intelligence (AI) models. Current hardware focuses on sensing, processing, or storing data, but some teams aim to integrate all three functions into one device.
Researchers at Xi’an Jiaotong University, the University of Hong Kong and Xi’an University of Science and Technology have introduced an innovative organic transistor that serves as both a sensor and processor. Traditional AI hardware employs distinct data sensing, processing, and memory storage systems. The separation causes high energy consumption and delays as data is converted between hardware components and analog signals. Pioneering studies emphasize organic electrochemical transistors’ impressive sensing and analog memory abilities (OECTs).
The researchers aimed to create a dual-function OECT as a sensor and processor to enhance AI hardware efficiency. OECTs are thin film devices that function as transistors, holding promise for smart bioelectronics and neuromorphic hardware. The OECT features distinct sensing and processing modes enabled by selective ion doping of the crystalline-amorphous channel within the device. In the sensing mode, migrating ions driven by a physiological signal enter the crystal structure but can diffuse back, keeping low conductance. In the processing mode, these ions are ‘trapped’ by the crystal structure, maintaining high conductance. This dual functionality enhances the uniqueness and efficiency of our OECT device.
The researchers employed cost-effective techniques like thermal evaporation, solution blade coating, thermal annealing, and reactive ion etching to produce their OECT array, enabling large-scale fabrication. The device is a sensor for diverse signals like electrophysiology, chemicals, light, and temperature. Additionally, as a memory unit, it stores 10-bit analog states, exhibits low switching randomness, and retains states for over 10,000 seconds. Our OECT device is truly versatile in the field of AI. The team conducted experiments to evaluate their device’s ability to switch operating modes. They discovered effective modulation, enabling it to function as a sensor and processor. As a sensor, it detects various stimuli, including ions and light. As a processor, it handles 10-bit analog states while retaining them well.
In the future, this transistor could advance neuromorphic devices for data collection and processing. The researchers demonstrated its real-time cardiac disease diagnosis capability and planned to explore more applications.
Reference: Shijie Wang et al, An organic electrochemical transistor for multi-modal sensing, memory and processing, Nature Electronics (2023). DOI: 10.1038/s41928-023-00950-y