Introduction:

This study explores the potential of Metal-Insulator-Semiconductor (MIS) type Resistive Random Access Memory (RRAM) as an electronic synapse. The research focuses on a fabricated RRAM stack with the structure n++Si/HfO2​/Ti/Al, specifically investigating its performance and synaptic behavior in the inversion regime. 

Objectives

Examine Switching Behavior: Investigate the switching characteristics of the MIS RRAM stack, particularly when operating in the inversion regime.

Analyze Interface Effects: Understand the role of the oxide interfacial layer (IL) at the HfO2​/Si interface and the oxygen scavenging layer of Ti metal over the HfO2 layer in filament formation and rupture mechanisms.

Emulate Synaptic Behavior: Demonstrate the ability of the RRAM to emulate synaptic behavior by achieving multiple resistive states through controlled sweeping.

Validate Learning Mechanism: Employ a physical model to explain the switching of the MIS RRAM and validate synaptic learning using Spike-Timing-Dependent Plasticity (STDP), showing changes in device conductance as a function of spiking order.

Assess Feasibility for Neuromorphic Applications: Corroborate that the MIS type RRAM exhibits bio-synaptic behavior similar to MIM stacks and evaluate its potential for neuromorphic applications.

Results

Inversion Regime Switching: The RRAM stack demonstrated relevant switching properties exclusively in the inversion regime, attributed to the interfacial layer and oxygen scavenging effects.

Interface and Vacancy Impact: The role of the oxide interfacial layer and variation in oxygen vacancy density were critical in filament formation and rupture mechanisms.

Synaptic Emulation: Multiple resistive states were achieved, emulating synaptic behavior effectively through controlled sweeping processes.

STDP Learning Validation: The study confirmed that the device conductance varied as a temporal function of spiking order, demonstrating successful emulation of synaptic learning via STDP.

Bio-Synaptic Behavior: Experimental results confirmed that the MIS-type RRAM shows bio-synaptic behavior comparable to MIM stacks and is feasible for deployment in neuromorphic systems.

Valleytronics, leveraging two-dimensional (2D) materials, offers novel opportunities for information processing, with the valley polarizer emerging as a fundamental building block. Existing methods for creating valley polarization—such as strain engineering, line defects, and static magnetic fields—often face challenges like low transmission efficiency and limited polarization directionality. 

Objectives

1. Design an All-Electrical Valley Polarizer: Develop a valley polarizer using zigzag edge graphene nanoribbons in a multiterminal setup.

2. Enable Dual Operational Modes: Implement gate-tuned functionality for both terminal-specific and parity-specific valley filtering.

3. Explore Multimode Physics: Investigate how multimode operation affects valley polarization and analyze factors like angle-selective transmission and edge state localization.

4. Ensure Disorder Resilience: Assess and confirm the device’s robustness against Anderson short-range disorder.

5. Optimize for Maximum Polarization: Fine-tune the device geometry to achieve optimal valley polarization.

Results

Innovative Design Achieved: Successfully developed a multiterminal valley polarizer using zigzag edge graphene nanoribbons.

Dual Mode Functionality: Demonstrated the device's ability to operate in both terminal-specific and parity-specific valley filtering modes.

Significant Multimode Insights: Revealed how multimode operation influences valley polarization, with key findings on angle-selective transmission and edge state behavior.

Robust Against Disorder: Confirmed that the device performs well under Anderson short-range disorder, showcasing its durability.

Optimized Polarization: Achieved maximum valley polarization through refined device geometry.

This study addresses the challenge of generating valley contrast in graphene by introducing a tilted PN junction. This approach creates anisotropic chiral transport, effectively inducing valley splitting and maintaining high fermion mobility. Our research also explores optimal parameters and demonstrates resilience to disorder, paving the way for similar techniques in other isotropic Dirac systems. 

Objectives

Introduce Valley Contrast: Develop a method to generate discernible valley contrasts in graphene by leveraging anisotropic chiral transport through a tilted PN junction.

Optimize Transport Properties: Explore the effects of the tilt angle and transition width on valley-resolved chiral transport and pseudospin-conserved modes.

Assess Experimental Setup: Optimize the experimental setup, including doping sequences and junction parameters, to achieve effective valley splitting.

Evaluate Disorder Resilience: Test the system’s resilience to Anderson short-range edge disorder.

Explore Broader Implications: Investigate the potential for inducing anisotropic chiral transport in isotropic Dirac systems, similar to characteristics found in tilted Dirac-Weyl semimetals.

Results

Effective Valley Contrast: Successfully introduced valley contrast in graphene through anisotropic chiral transport enabled by a tilted PN junction.

Enhanced Transport Properties: Observed valley-resolved chiral transport with significant improvements in mobility and transmission, attributed to specular edge scattering.

Optimized Experimental Conditions: Identified optimal conditions for tilt angle and transition width, confirming enhanced transmission with increased transition width.

Resilience to Disorder: Demonstrated that the system remains effective and resilient against Anderson short-range edge disorder.

Transformative Potential: Showed that analogous anisotropic chiral transport can be achieved in isotropic Dirac systems, paving the way for similar behaviors in Dirac-Weyl semimetals.