Dual-Electrolyte Neuromorphic Transistor for Risk Detection and Image Processing

The human brain is a highly efficient structure that can easily perform various complex tasks, such as shape recognition, presentation, and classification, while consuming minimal energy and occupying only a small volume. This study introduces a bio-inspired electrolyte-gated neuromorphic transistor that mimics the functionality of the human brain. A dual-electrolyte structure combining lithium phosphorus oxynitride and lithium silicate achieves the best performance, with a mobility of 3.1 cm² V⁻¹ s⁻¹, a paired-pulse facilitation index of 162.6%, and nonlinearity coefficients of 0.02 and 0.03 (for potentiation and depression, respectively). Further, risk pre-detection and image recognition are successfully demonstrated using the developed dual-electrolyte synaptic transistors. A test conducted on the Modified National Institute of Standards and Technology database indicates an accuracy of 91.0%. Thus, the device has the potential to advance artificial vision systems.