Interface engineering in ZnO/CdO hybrid nanocomposites to enhanced resistive switching memory for neuromorphic...

Area of Science:

• Materials Science
• Nanotechnology
• Solid-State Electronics

Background:

• Resistive random-access memory (RRAMs) are crucial for embedded storage and neuromorphic computing.
• Metal chalcogenides offer tunable electronic states for advanced RRAM applications.

Purpose of the Study:

• To synthesize a ZnO-CdO hybrid nanocomposite for memristor applications.
• To evaluate the memristor's performance in emulating biological synaptic functions and its potential for artificial neural networks.

Main Methods:

• Hydrothermal synthesis of ZnO-CdO hybrid nanocomposite.
• Fabrication of Ag/ZnO-CdO/Pt memristor devices.
• Characterization of electrical properties, synaptic functions, and artificial neural network performance.

Main Results:

• The memristor demonstrated stable resistive switching with low SET/RESET voltages and a high R_ON/OFF ratio (~10⁵).
• Excellent retention stability, endurance (>10⁴ cycles), and multilevel storage were achieved.
• Successful emulation of synaptic functions (LTP, LTD, PPF) and 92.6% accuracy in handwritten digit recognition using an artificial neural network.

Conclusions:

• The ZnO-CdO hybrid nanocomposite memristors exhibit excellent non-volatile memory properties.
• These memristors effectively emulate biological synapses, paving the way for efficient neuromorphic computing architectures.