Artificial synapse based on Co3O4 nanoflowers with picojoule energy consumption, microsecond time constant, and visual learning capability

Abstract Metal oxide nanomaterials are promising candidates for neuromorphic systems due to their remarkable physical properties, including atomic-scale thickness and high ionic activity. In this work, Co 3 O 4 nanoflowers were incorporated into artificial synaptic devices, exhibiting both short-term plasticity (e.g. paired-pulse facilitation) and long-term plasticity (e.g. spike-timing-dependent plasticity) with picojoule energy consumption and microsecond time constants. Utilizing the long-term potentiation and depression characteristics of the Co 3 O 4 nanoflower-based artificial synapses, the visual learning capability was demonstrated. X-ray photoelectron spectroscopy depth profiling analysis revealed that the synaptic behavior is primarily governed by the migration of oxygen vacancies. These findings underscore the potential of metal oxide nanoflower-based artificial synapses for use in energy-efficient neuromorphic circuits and complex cognitive systems.