Mechanism discrimination in multilevel RRAM using state-resolved impedance evolution

Area of Science:

• Materials Science
• Electrical Engineering
• Semiconductor Device Physics

Background:

• Multilevel operation in Resistive Random Access Memory (RRAM) necessitates reliable discrimination between distinct switching mechanisms.
• Different switching pathways in RRAM lead to varied state evolution and frequency responses, complicating device analysis.

Purpose of the Study:

• To employ state-resolved impedance evolution for analyzing and distinguishing filamentary-type and interfacial-type RRAM devices.
• To systematically examine the evolution of resistive and capacitive contributions across multiple programmed resistance states.
• To link impedance spectra changes to underlying physical switching mechanisms.

Main Methods:

• Tracking impedance spectra at each programmed resistance state in RRAM devices.
• Utilizing Nyquist and Bode analyses to characterize state-dependent impedance responses.
• Constructing and implementing equivalent circuit models in SPICE to simulate and validate experimental findings.

Main Results:

• State-resolved impedance evolution clearly differentiates between filamentary and interfacial switching mechanisms.
• Resistive and capacitive contributions show distinct evolutionary patterns linked to the switching type.
• SPICE simulations successfully reproduced measured impedance trends and validated physical mechanism interpretations.

Conclusions:

• State-resolved impedance evolution provides a practical framework for mechanism analysis in multilevel RRAM.
• This approach enables reliable discrimination between filamentary and interfacial switching behaviors.
• The combined experimental and modeling strategy offers robust insights into RRAM device physics.