Multiple spiking functionalities in annealing-optimized Ag/Hf0.5Zr0.5O2-based memristive neurons

Rapid progress of artificial neural network applications in recent years has led to the issue of unprecedented energy consumption. It can be solved by the implementation of energy-efficient hardware based on non-von Neumann architectures, which requires the development of electronic components emulating the behavior of synapses and neurons. While research of synaptic elements is vast, the technology for the fabrication of scalable and highly reproducible neuronal elements is far less developed. In this work, we demonstrate an artificial neuron with multiple functionalities based on filamentary switching Ag/Hf0.5Zr0.5O2 (HZO) memristors. To improve the parameters of the memristors, we propose a two-step annealing method, which allows for better control of the crystallization of the functional dielectric layer (HZO) as well as of the diffusion of the active electrode (Ag) atoms. Furthermore, we demonstrate leaky integrate-and-fire neuronal behavior in multiple spiking modes: time-to-first-spike, number of spikes, and firing rate coding. Moreover, the neuron operation does not require the additional electronic overhead and is supported solely by an Ag/HZO memristor with a current limiting resistor connected in series. The presented results pave the way for the development of next-generation energy-efficient neuromorphic hardware operating on the principles of spiking neural networks.