Interfacial\nElectronic\nCharge Trapping and Photonic\nCarrier Excitation Coupling in Solution-Processed Zinc–Tin\nOxide Thin-Film Transistors Applied for Logic Gate Design and Quantized\nNeural Network

Components needed\nin Artificial Intelligence with a higher\ninformation\ncapacity are critically needed and have garnered significant attention\nat the forefront of information technology. This study utilizes solution-processed\nzinc–tin oxide (ZTO) thin-film phototransistors and modulates\nthe values of <i>V</i><sub>G</sub>, which allows for the\nregulation of electron trapping/detrapping at the ZTO/SiO<sub>2</sub> interface. By coupling the excited photonic carrier and electronic\ntrapping, logic gates such as “AND,” “OR,”\n“NAND,” and “NOR” can be achieved. With\nthe exponential growth in data generation, efficient processing and\nstorage solutions are imperative. However, extensive data transfer\nbetween computing units and storage limits the level of artificial\nneural networks (ANNs). Consequently, quantized neural networks (QNNs)\nhave gained interest for their reduced computational resource requirements\nand lower consumption. In this context, we introduce an optimized\nternary logic circuit based on ZTO devices. By utilizing optical modulation\nto adjust the turn-on voltage of the single device, we demonstrate\nthe achievement of ternary current states, thereby providing three\ndistinct discrete states. This configuration can be extended to QNN\ncomputing, demonstrating multilevel quantized current values for in-memory\ncomputation. We achieved a handwriting digit recognition rate of 91.6%,\nthereby demonstrating reliable QNN hardware performance. This robust\nQNN performance indicates that the metal oxide phototransistor shows\nsignificant potential for future ternary computing systems.