Visible Light-Stimulated Artificial Synapse Based on an Organic Field-Effect Transistor for Imitating Human Emotions and Mood Swings

Abstract

The development of a synaptic transistor stimulated by visible light has drawn significant attention in recent years due to its high operating speed, low energy consumption, wide bandwidth, high transmission rate, and relatively low noise. These properties make them potential devices for understanding the behavior of the human brain and artificial intelligence-related applications. In this work, we report a low-voltage operating photosynaptic transistor based on an organic semiconductor poly(2,5bis(3-alkylthiophen-2-yl) thieno[3,2-b] thiophene) [PBTTT-C14]. Basic neurobiological synapse-like behaviors such as excitatory post-synaptic current (EPSC), pair pulse facilitation (PPF), short-term plasticity (STP), long-term plasticity (LTP), and learningforgetting-memorizing nature like the human brain are successfully explained. Photoresponse parameters like photoresponsivity (R), detectivity (D*), and photo and dark current ratio are estimated and interpreted. These devices successfully demonstrated the conversion from STP to LTP with the variation of initial photoexcitations. We attribute the human brain-like behavior of these photosynaptic devices to the interfacial charge-trapping effect at the interface of the semiconducting channel and dielectric layer. Further, we investigated the impact of interfacial traps on the photosynaptic behavior using photocurrent measurements at different light intensities. Moreover, these devices also show a human emotion-tunable and mood swing-dependent learning and memory behavior. Additionally, we have successfully demonstrated the utilization of these devices for "OR" logic gate operation. This study shows the photosynaptic behavior using the most simple OFET architecture, which can be tuned by improvising trap density at semiconductor/dielectric interfaces using a self-assembled monolayer. The results presented in this paper support the potential of such devices as neural units for constructing low-power-operating artificial neural networks.