Alloy electrode engineering in memristors for emulating the biological synapse

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Conductive bridging random access memory (CBRAM) is crucial for artificial synaptic devices in neuromorphic computing.
• Simple electrodes in CBRAM often lead to unstable conductive filaments and poor device performance.

Purpose of the Study:

• To investigate the impact of different alloy electrode ratios on HfO2-based CBRAM device performance.
• To enhance the functionality of artificial synaptic devices through electrode composition engineering.

Main Methods:

• Systematic investigation of HfO2 devices with varying silver-copper (Ag-Cu) electrode ratios.
• Characterization of device performance metrics including formation voltage, set voltage, uniformity, response speed, and power consumption.
• Evaluation of biosimulation capabilities, including paired-pulse facilitation (PPF), post-tetanic potentiation (PTP), spike-rate-dependent plasticity (SRDP), and spike-timing-dependent plasticity (STDP).

Main Results:

• Devices with a 63:37 Ag-Cu ratio demonstrated lower formation and set voltages, improved set voltage uniformity, faster response speeds, and reduced power consumption.
• The optimized CBRAM devices successfully emulated key biosynaptic functions (PPF, PTP, SRDP, STDP).
• Demonstrated associative learning (Pavlov's dog) and aversion therapy without complex external circuitry.

Conclusions:

• Electrode composition engineering in CBRAM devices significantly boosts memristor properties.
• Optimized Ag-Cu electrodes in HfO2 CBRAM offer a promising pathway for advanced neuromorphic computing systems.
• This research lays the groundwork for the development of more efficient and capable biomimetic neural computing systems.