Transparent Photo-Driven Neurotransmitter Using 2D SnS Photodetectors and Artificial Synapses

Neuromorphic computing, an emerging field driven by artificial intelligence (AI), has the potential to revolutionize electronics and transform everyday life in unprecedented ways. Central to this field is the development of artificial synapses capable of detecting, storing, and processing information while simultaneously exhibiting both electronic and photonic synaptic behaviors. However, traditional photodetectors, which convert light stimuli into electrical signals, lack the memory functions of the human visual cortex. To address this limitation, researchers have attempted to integrate nonvolatile memory devices into photodetectors using the von Neumann architecture. In this architecture, data shuttles back and forth between memory storage and the processing unit, which significantly increases energy consumption, processing time, and device temperature. Consequently, fabricating a single device that can simultaneously gather and process information for neuromorphic electronics remains a significant challenge.
To overcome this, we propose an artificial synapse with neurotransmitter-like functionality based on a Schottky junction model, fabricated using a vapor transfer method for the uniform and large-area growth of 2D tin sulfide (2D-SnS). This synapse is composed of a van der Waals (vdW) semiconductor (PVT-SnS) and a transparent conducting oxide platform (ITO). The unique atomic structures and tunable optoelectronic properties of 2D materials have garnered growing interest as promising candidates for large-area fabrication of artificial synapses in neuromorphic computing.
The PVT-SnS-based synaptic device demonstrates its ability to perform elementary logic operations, including Boolean logic gates such as OR, NOT, and AND. Unlike conventional logic gates, which rely on a combination of transistors and resistors, the PVT-SnS-based device eliminates the need for these components. Instead, photonic stimuli serve as the input parameters, and the outputs are generated based on specific conditions. The device’s neurotransmitter-like modulatory capability is particularly significant, as it mimics the role of neurotransmitters in biological neurons, which adjust how cells communicate at synapses and influence the behavior of other chemical messengers. This functionality allows for precise control over logic gate operations.
Artificial synapses with neurotransmitter-like functionality could drive advances in neuromorphic computing, unlocking potential applications in AI-related fields such as humanoids. These devices could enhance the efficiency and intelligence of humanoid robots by endowing them with neurotransmitter-like capabilities, paving the way for next-generation AI systems with improved adaptability and functionality.