Title: Unlocking Extra Functionalities: Exploring Energy Storage Materials for Iontronic Applications with Li-Based Materials

Deep learning has demonstrated remarkable success in various applications, but the energy consumption of conventional CMOS circuitry remains a challenge, limiting their implementation to large data centers rather than end-user devices. In response, neuromorphic computing has emerged as a promising solution for highly efficient computation, drawing inspiration from the human brain.

At the core of neuromorphic computing architectures are artificial synapses that mimic the behavior of biological synapses. These synapses must modulate their resistance in an analog manner with minimal energy consumption to adjust synaptic weights. To achieve this, two- and three-terminal devices that utilize ions instead of electrons have been proposed, aiming to reduce energy consumption while improving linearity and symmetry of conductance changes.

Significant progress has been made at the materials level, exploring various ions and material systems, including lithium (Li⁺), oxygen (O^2-), and hydrogen (H⁺). Li-based materials offer several advantages, such as typically higher diffusivities at room temperature, low-energy intercalation potentials, and a rich library of electrochemically diverse materials already developed for Li-ion batteries. Additionally, the growing industry involving the thin-film processing of Li-based materials for microbattery applications demonstrates the industrial viability of fabricating these materials.

This presentation will provide a historical perspective on the development of Li-based iontronics and highlight recent advancements in two-terminal -memristive or resistive switching devices- and three-terminal -ionic transistors- approaches for electrochemical synapses in neuromorphic computing. This presentation will address as well the challenges faced in this field and discuss future research directions aimed at improving device performance and reliability. The integration of energy storage materials not only enhances the energy efficiency of these architectures but also opens new avenues for their applications in diverse fields, ranging from artificial intelligence to robotics and edge computing in which both their main energy source and the active neuromorphic elements are made with the material class.