Spin-Orbit Torque Memory With Composite Free Layer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spin-transfer torque RAM (STTRAM) faces challenges such as rapid tunnel barrier degradation and inefficient electron spin use due to its current-perpendicular-to-plane geometry, leading to high write energy and limited switching speed, while existing memory technologies like DRAM suffer from high power consumption and refresh energy requirements.
Innovation Solution
The development of spin-orbit torque (SOT) devices with a composite free layer comprising a high-anisotropy ferromagnetic layer, a non-magnetic transition metal layer, and an ultra-low damping magnetic insulator, where spin-orbit torque is induced by passing a current through a heavy metal region, allowing for efficient switching of magnetization and reducing write energy through exchange coupling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If current-perpendicular-to-plane geometry is used in STTRAM, then switching speed is improved, but write energy increases and tunnel barrier degrades rapidly
Solution Approach 1:
The patent introduces a heavy metal layer as an intermediary between the current path and the ferromagnetic layer. The current flows through the heavy metal layer which generates spin-orbit torque to switch the magnetization of the ferromagnetic layer, rather than passing current directly through the tunnel barrier as in STTRAM. This intermediary approach enables efficient magnetization switching with lower energy consumption while maintaining fast switching speed.
Solution Approach 2:
The patent replaces the direct spin-transfer torque mechanism (which requires high current through the tunnel barrier) with a spin-orbit torque mechanism utilizing the spin Hall effect or Rashba effect in heavy metal layers. This substitution changes the physical mechanism from direct momentum transfer through the barrier to orbital-to-spin angular momentum conversion at the heavy metal/ferromagnetic interface, achieving lower write energy with comparable or improved switching speed.
2Speed
If current-perpendicular-to-plane geometry is used in STTRAM, then switching speed is improved, but tunnel barrier degrades rapidly
Solution Approach 1:
The heavy metal layer serves as a mediator that decouples the current path from the tunnel barrier. Current flows laterally through the heavy metal layer adjacent to the tunnel barrier, generating spin-orbit torque that acts on the ferromagnetic layer without requiring high current density through the barrier itself. This protects the tunnel barrier from degradation while maintaining fast switching performance.
Solution Approach 2:
The patent segments the current path from the magnetization switching path. The current flows through the heavy metal layer (conductive channel) while the spin-orbit torque is transferred to the ferromagnetic layer through their interface. This spatial segmentation allows the current to bypass the tunnel barrier entirely, preventing barrier degradation while achieving the desired switching speed through the spin-orbit interaction at the heavy metal/ferromagnetic interface.
3Quantity of substance
If DRAM is used for high density memory, then capacity is improved, but power consumption and refresh energy increase
Solution Approach 1:
The patent utilizes phase transitions in the sense of magnetic state transitions (parallel vs. anti-parallel magnetization alignment) to store binary information. The SOT-MRAM device switches between stable magnetic states using spin-orbit torque, enabling non-volatile storage without requiring continuous refresh operations. This allows high-density memory implementation with dramatically reduced power consumption compared to volatile DRAM that requires periodic refreshing.
Solution Approach 2:
The SOT-MRAM structure achieves non-volatile storage where the magnetic state is maintained without external energy input. The heavy metal layer and ferromagnetic layer structure enables the device to maintain its state indefinitely without refresh operations, making the memory self-sufficient in maintaining data. This eliminates the refresh energy requirement that plagues DRAM systems, achieving both high density and low power consumption.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
SOT devices achieve a write energy of 10 aJ/bit for a 10 nm cell, significantly reducing energy consumption by a factor of 103 compared to DRAM and 105 compared to DRAM refresh energies, while enabling ultra-high-density memory with low power consumption and improved thermal stability.
Implementation Method 1
inducing spin orbit torque by passing a current through a heavy metal region of a conductive channel adjacent a composite free layer
Implementation Method 2
The example technique includes inducing spin orbit torque by passing a current through a heavy metal region
Implementation Method 3
The high-anisotropy ferromagnetic layer is exchange-coupled to the ultra-low damping magnetic insulator
Data Source
AI summary
An example article includes a composite free layer and a conductive channel. The composite free layer includes a high-anisotropy ferromagnetic layer, a non-magnetic transition metal layer adjacent to the high anisotropy ferromagnetic layer, and an ultra-low damping magnetic insulator. The non-magnetic transition metal layer is between the ultra-low damping magnetic insulator and the high-anisotropy ferromagnetic layer. An example spin-orbit torque (SOT) stack may include the example article. Techniques for forming and switching example articles and SOT stacks are described.