Multimodal Neuromorphic Sensory Device for Intelligent Perception

Cross-modal integration in biological systems arises from physically coupled neural processes rather than from parallel signal fusion. Emulating such associative behavior in solid-state hardware remains challenging, particularly for sensory modalities governed by fundamentally different physical domains. Here, we present a visual-gustatory neuromorphic device based on an asymmetric CuO-Cu₂O nanowire film that intrinsically couples solid-state optoelectronics with liquid-phase ionics. The heterostructured nanowire network exhibits a pronounced surface photovoltaic effect and spontaneous charge redistribution at the solid-liquid interface, enabling bias-free operation in aqueous environments. Ionic adsorption associated with the chemical response forms an electric double layer that dynamically modulates surface band bending and the majority carrier density. This ionic modulation is intrinsically coupled to the photoresponse, resulting in synergistic amplification of the electrical output. The coupled physical mechanism gives rise to characteristic time-dependent optoelectronic responses that provide a functional basis for the multimodal integration. Importantly, pairing optical and chemical stimuli establishes a persistent cross-modal association, allowing purely visual inputs to evoke gustatory-specific responses after conditioning. This work demonstrates device-level cross-modal associative learning emerging directly from a multimodal sensory device.