High-Reliability and Self-Rectifying Alkali Ion Memristor through Bottom Electrode Design and Dopant Incorporation

Abstract

Ionic memristor devices are crucial for efficient artificial neural network computations in neuromorphic hardware. They excel in multi-bit implementation but face challenges like device reliability and sneak currents in crossbar array architecture (CAA). Interface-type ionic memristors offer low variation, self-rectification, and no forming process, making them suitable for CAA. However, they suffer from slow weight updates and poor retention and endurance. To address these issues, the study demonstrated an alkali ion self-rectifying memristor with an alkali metal reservoir formed by a bottom electrode design. By adopting Li metal as the adhesion layer of the bottom electrode, an alkali ion reservoir was formed at the bottom of the memristor layer by diffusion occurring during the atomic layer deposition process for the Na:TiO 2 memristor layer. In addition, Al dopant was used to improve the retention characteristics by suppressing the diffusion of alkali cations. In the memristor device with optimized Al doping, retention characteristics of more than 20 h at 125 °C, endurance characteristics of more than 5.5 × 10 5 , and high linearity/symmetry of weight update characteristics were achieved. In reliability tests on 100 randomly selected devices from a 32 × 32 CAA device, device-to-device and cycle-to-cycle variations showed low variation values within 81% and 8%, respectively.

Interface-Type Ionic Memristor for Energy-Efficient Neuromorphic Hardware

“…Devices exhibiting resistance changes by migrating oxygen anions (or vacancies) or metal cations within transition metal oxide (TMO) thin films are generally classified as ionic memristors. Although various mechanisms have been reported, these can primarily be classified into filament-type and interface-type, as shown in Figure . ,, Interface-type RS behavior is commonly observed in TMO films that easily form multivalence states, but both filament-type and interface-type can occur in the same material. This is typically determined by the initial resistance state of the TMO films; a high initial resistance leads to filament-type behavior after a soft breakdown, whereas a lower resistance film (not too low) exhibits interface-type characteristics.…”

“…(d) Typical resistance change in interface-type memristor devices observed during DC sweep measurements: RS behavior observed at a symmetric interface (left), self-rectifying RS behavior caused by an asymmetric interface (right). Reproduced from ref . Copyright 2024, American Chemical Society.…”

“…Recently, efforts have been made to address the issues of conventional oxygen ion-based memristors by utilizing alkali ions with high mobility within solids as the primary agents for RS. ,,− At the early stage, reported research was limited due to difficulties in controlling lithium composition, contamination issues in the deposition process, and poor retention characteristics expected by the high mobility of Li ions in solids. The main question was whether Li-ion-based memristors could achieve good retention characteristics.…”

“…(h) Cumulative probability plot extracted from endurance data for C2C variation. Reproduced from ref . Copyright 2024, American Chemical Society.…”

“…This device type inherently follows an interface-limited conduction regime, making it challenging to use analog inputs for neuromorphic hardware due to its high nonlinear characteristics in DC conditions. However, interface-type devices, which do not require a forming process and exhibit lower D2D and C2C variations, are advantageous for multibit implementation . Typical DC RS curves of interface-type, as shown in Figure d, generally show gradual switching behavior.…”