Insights into synaptic functionality and resistive switching in lead iodide flexible memristor devices

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Traditional von Neumann architecture faces limitations in efficiency and power consumption.
• Neuromorphic computing offers a promising alternative with parallel processing and low power needs.
• Memristor devices are key components for emulating synaptic functions in neuromorphic systems.

Purpose of the Study:

• Investigate resistive switching mechanisms in lead iodide (PbI2) memristors.
• Correlate device performance with morphological features for flexible neuromorphic applications.
• Elucidate the factors influencing the stability and synaptic behavior of PbI2 memristors.

Main Methods:

• Fabrication and characterization of PbI2-based memristor devices.
• Analysis of resistive switching mechanisms, including charge transport and ion migration.
• Evaluation of device stability under mechanical strain and high humidity.
• Testing of synaptic plasticity (potentiation/depression) and STDP.
• Performance assessment using MNIST dataset for pattern recognition.

Main Results:

• Identified a highly reliable unipolar switching mechanism in PbI2 memristors.
• Demonstrated stability under mechanical strain (4 mm bending radius) and 75% relative humidity without encapsulation.
• Observed synaptic behaviors with potentiation/depression over 2x10^4 cycles, indicative of STDP.
• Achieved 95.06% accuracy on MNIST pattern recognition with 30 training epochs.
• Established correlations between device performance and morphological features.

Conclusions:

• PbI2-based flexible memristors show significant potential for neuromorphic computing applications.
• The devices exhibit robust and reliable resistive switching, mimicking synaptic functions.
• Mechanical and electrical resilience make PbI2 memristors suitable for advanced information processing.