Advancements in artificial synapses: The role of fluorite–structured ferroelectrics

Abstract

In today's world, the rise of big data demands a new computing paradigm beyond the von Neumann architecture to handle massive datasets effectively. Neuromorphic computing, inspired by the synaptic plasticity of biological synapses, has emerged as a solution. Artificial synapses (ASs) in neuromorphic systems replicate synaptic functions like potentiation/depression, short-/long-term plasticity, and spike-time-dependent plasticity. Initial research on fluorite-structured ferroelectrics focused on understanding ferroelectricity and improving device performance. Since the discovery of ferroelectricity in hafnium-zirconium oxide in 2011, these materials have gained attention for their scalability and compatibility with CMOS technologies. This review explores advances in fluorite-structured ferroelectric ASs, including two-terminal switchable diodes, ferroelectric-tunnel junctions, three-terminal field-effect transistors, and thin-film transistors. Additionally, we examine future challenges and prospects for developing ferroelectric-based ASs for neuromorphic computing.