2D Materials for Selectorless Resistive Switching Memory

Invention Description
Selectorless resistive random-access memory is crucial for achieving high-density crossbar scaling, however, suppressing sneak path currents (SPCs) without external selector components is still a major challenge. Conventional von Neumann architectures, which transfer data between the process unit towards memory, struggle to keep pace with AI demands. Traditional approaches that attempt to solve this bottle neck improve latency and bandwidth but still require data movement between on-chip memory.
 
Researchers at Arizona State University have developed novel Cu0.5Ag0.5InP2S6 (CAIPS) van der Waals materials for non-volatile memory arrays that intrinsically enable nonlinear switching and multilevel resistive states. These materials suppress sneak path currents without requiring external selector components. By combining the characteristics of CuInP2S6 and AgInP2S6, these materials exhibit asymmetric out-of-plane diffusion barriers enhancing self-rectifying transport and improving device performance. The resulting devices demonstrate stable bipolar resistive switching, high intrinsic nonlinearity, and low switching voltages suitable for scalable and energy-efficient memory and computing applications.
 
This technology introduces novel materials exhibiting intrinsic selectorless behavior which supports energy-efficient operation and offer new opportunities for compact memory and neuromorphic architectures in edge computing applications.
 
Potential Applications
  • High-density non-volatile memory arrays for data storage
  • Neuromorphic computing and artificial intelligence hardware
  • Edge computing devices requiring low power and high efficiency
  • Energy-efficient memory modules for wearable and IoT applications
  • Advanced computing architectures integrating scalable memory solutions
Benefits and Advantages
  • Eliminates need for external selector components, simplifying memory array design
  • High intrinsic nonlinearity and large memory window for reliable data storage
  • Stable bipolar resistive switching with low variability and robust endurance
  • Low operational switching voltages enable energy-efficient operation
  • Material properties confirmed via first-principles DFT, ensuring performance consistency
  • RRAM devices based on CAIPS achieve a high nonlinearity factor (>10), large memory window (>9×), and low resistance variability (CV% as low as 5.1%)
For more information about this opportunity, please see
Patent Information: