Multifunctional organic synaptic transistors for tissue-equivalent dosimetry

Pervading the modern electronics landscape, organic synaptic transistors (OSTs) have emerged as a forefront device candidate for advancement in artificial intelligence (AI) systems. This paper reports OSTs as highly sensitive dosimeters which also emulate human tissue properties and cognitive functions. Fabricated OSTs exhibit excellent p-channel transistor characteristics with field-effect mobility of 0.21 (±0.03) cm²V⁻¹s⁻¹ and on-off current ratio on the order of 10³ for −10 V operation. The OSTs demonstrate short-term plasticity (STP) through behaviors such as pulse paired facilitation (PPF) and spike number dependent plasticity (SNDP). Notably, the relaxation time constants derived from PPF behavior, combined with an energy consumption per stimuli of ∼10 pJ, closely mimic those observed in human synapses. Moreover, by integrating the synaptic weights derived from the fabricated devices, artificial neural network (ANN) achieves a handwritten digit recognition accuracy exceeding 99.4 %. These OSTs exhibit negligible changes in on current and mobility after being irradiated, but a linearly varying shift in threshold voltage, suitable for detecting γ-radiation exposure. An impressive sensitivity (∼5 mV/rad) to γ-radiation exposure under radiation amount similar to the values used for treatment of tumors. Our results indicate that these flexible OSTs have the potential to be utilized as smart and intelligent radiation sensors for applications such as medical imaging, radiation therapy, and portable dosimeters for emergency responders and armed forces.