Nano-cone resistive memory for ultralow power operation

Area of Science:

• Materials Science
• Electrical Engineering
• Nanotechnology

Background:

• Resistive random-access memory (RRAM) is a promising next-generation non-volatile memory technology.
• Silicon nitride (SiNₓ) is a key material for RRAM fabrication due to its CMOS compatibility.
• Improving the performance of SiNₓ-based RRAM is crucial for advanced computing applications.

Purpose of the Study:

• To fabricate and characterize SiNₓ-based nano-structure resistive memory using a novel nano-cone silicon bottom electrode (BE).
• To investigate the impact of the nano-cone BE structure on resistive switching characteristics.
• To explore the potential of this device for ultra-high-density crossbar memory arrays.

Main Methods:

• Fabrication of SiNₓ-based RRAM devices using a silicon CMOS-compatible process.
• Integration of a specifically designed anisotropic etching process for creating a nano-cone silicon BE.
• Characterization of bipolar resistive switching behavior in both nano-cone and flat BE structures.
• Systematic device simulations to analyze the electric field distribution and performance enhancement.

Main Results:

• The nano-cone BE structure significantly reduced switching current and voltage compared to a flat BE.
• Device simulations confirmed that electric field concentration at the cone tip is the primary reason for performance improvement.
• The nano-cone resistive memory cell exhibited greatly improved nonlinearity.
• The fabricated devices demonstrated bipolar resistive switching characteristics.

Conclusions:

• The nano-cone SiNₓ-based resistive memory cell offers superior performance metrics, including lower switching current and voltage.
• The enhanced electric field concentration in the nano-cone BE is key to the improved device characteristics.
• This technology holds significant promise for the development of ultra-high-density crossbar memory arrays.