Ti-Doped GaO_x Resistive Switching Memory with Self-Rectifying Behavior by Using NbO_x/Pt Bilayers

Area of Science:

• Materials Science
• Electrical Engineering
• Nanotechnology

Background:

• Resistive random-access memory (ReRAM) in crossbar arrays (CBAs) offers high-density storage but suffers from sneak paths, increasing power consumption and limiting scalability.
• Sneak paths arise from current leakage through unselected memory cells in the CBA architecture.
• Overcoming sneak paths is crucial for advancing high-density memory technologies.

Purpose of the Study:

• To propose and demonstrate a novel self-rectifying ReRAM structure to mitigate sneak path issues in CBAs.
• To enhance the performance characteristics of ReRAM devices, including operating voltage, speed, and reliability.
• To enable the development of larger and more power-efficient high-density memory arrays.

Main Methods:

• Fabrication of a Ti/GaOx/NbOx/Pt multilayer structure for ReRAM devices.
• Engineering interfacial reactions between Ti and GaOx layers to increase trap sites and reduce operating voltage.
• Incorporation of a NbOx/Pt stack to introduce self-rectifying behavior via a large Schottky barrier.
• Characterization of the device's resistive-switching properties, including operating voltage, on/off ratio, selectivity, speed, and stability.

Main Results:

• The proposed Ti/GaOx/NbOx/Pt structure exhibits self-rectifying resistive-switching behavior, effectively suppressing sneak paths.
• The device operates at a low voltage range (-2.8 to 2.5 V) with a very low forming voltage (∼3 V).
• Excellent performance metrics were achieved, including high on/off ratios (∼20), high selectivity (∼10^4), fast operating speeds (200-500 ns), and stable, uniform operation.

Conclusions:

• The developed self-rectifying ReRAM device successfully addresses the sneak path problem in CBAs.
• The unique material stack and engineered interfaces lead to significantly improved device performance and efficiency.
• This advancement holds promise for next-generation high-density memory applications.