Improving multifunctional monolithic nano-device by single (Al,Ga)N nanowire/graphene van der Waals heterostructure for neuromorphic computing, photodetection and imaging

Integrating neuromorphic computing, photodetection and imaging in single devices remains challenging due to the inherent trade-off between the transient photoresponse speeds of artificial synapses and photodetectors. This study develops a dual-mode monolithic device using a (Al,Ga)N nanowire/graphene heterojunction, operating as a photodetector under negative bias and a neuromorphic sensor under positive bias. Graphene strengthens the built-in electric field, enhancing carrier separation and photocurrent for both functions. The device consumes ultralow energy (3.19 × 10−11 J) with demonstrated synaptic plasticity features like spike-dependent learning and accelerated memory reinforcement. Leveraging this synaptic plasticity, the device achieves over 90% accuracy in image processing tasks. This work introduces a multifunctional integration strategy that advances neuromorphic computing efficiency and optoelectronic device design, demonstrating the feasibility of simultaneous imaging and brain-inspired computation in compact systems. The integration of multiple functionalities, such as neuromorphic computing, photodetection and imaging, within a single optoelectronic device is challenging due to the trade-off between the transient photoresponse speeds of artificial synapses and photodetectors. Here, a multifunctional monolithic nano-device based on the heterojunction between a single (Al,Ga)N nanowire and graphene is proposed, providing a dual-mode operation that integrates photodetection and neuromorphic visual sensing.