Selective UV Sensing for Energy-Efficient UV-A Artificial Synapses Using a ZnO/ZnGa2O4 Heterojunction Diode

Abstract

As neuromorphic computing systems, which allow for parallel data storage and processing with high area and energy efficiency, show great potential for future storage and in-memory computing technologies. In this article, a high-performance UV detector for artificial optical synapse applications is demonstrated that can selectively detect UV-A and UV-C, with a responsivity of 407 A W−1. The pyrophototronic effect increases photocurrent dramatically under UV-A irradiation due to heat accumulation in the ZnO layer and ZnGa2O4’s low thermal conductivity. In context of synaptic device, it's shown that a ZnO/ZnGa2O4 heterostructure can be used as a light-tunable charge trapping medium to create an electro-photoactive synapse. The photogating effect enables via pyrophototronic, which traps photogenerated electrons within the ZnO/ZnGa2O4 interface, and drives synaptic activity, as proven by electrical techniques based on UV-A stimuli. This phenomenon results in a selective detection capability for UV-A over UV-C. Thermally produced pyrophototronic effect synaptic plasticity, simulating biological synapse activity. Persistent photoconductivity under 380 (UV-A) nm UV light mimics synaptic processes, with low thermal conductivity enhancing synaptic weight updates during learning and forgetting. These findings show the possibility of using ZnO/ZnGa2O4 heterostructures into artificial optoelectronic synapse systems controlled by thermal dynamics.