Integrated Floating-Gate Organic Electrochemical Transistor for In-Memory Sensing

Traditional sensing technologies rely on data transfer between sensor, memory and processing units. Traditional architectures encounter the problem of the memory wall, which restricts the processing speed to the rate at which data can be accessed or stored in memory. As data-intensive applications continue to grow, bio-inspired sensing systems have recently been developed to tackle these limitations. Biological neurons can indeed integrate sensing, memory and processing in a single element, making them extremely efficient in terms of energy, time and area. In-memory sensing technologies emulate this architecture, moving towards local and autonomous operation of point-of-care sensing devices. Furthermore, this all-in-one architecture prevents data from being uploaded to external servers, ensuring safer storage of the sensitive clinical data acquired by the sensor.<br/>This work presents a floating-gate organic electrochemical transistor (FG-OECT) able to work both as a sensor and as a memory.<br/>Organic mixed ionic-electronic conductors (OMIECs) are optimal candidates for in-memory sensing, as they have already been employed for both biosensing and neuromorphic computing. Organic electronic devices are notably characterized by low voltage of operation, high conductivity and excellent tunability, making them extremely versatile devices relevant to different fields.<br/>Thanks to the floating-gate architecture, the memory can be decoupled from the sensing region, preventing contamination of the channel material with the analyte and allowing optimal materials to be employed for the two purposes. In particular, the memory gate is made of poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS), granting state retention to the system thanks to the electrochemical doping/dedoping process of both the gate and the channel. Differently, the sensing gate is made of gold, providing an easier functionalization of the electrode with artificial bioreceptors for capacitive sensing. Figures of merit for both biosensing and memory properties have been addressed, including transconductance, state retention and number of states.