SOT MRAM Beta Tungsten Composite Layer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Scaling magnetoresistive random access memory (MRAM) devices to high densities is hindered by large currents required for switching and disturbances to neighboring cells during writing, leading to erroneous data storage in spin transfer torque (STT) and spin orbit torque (SOT) MRAM devices.
Innovation Solution
The SOT MRAM cell incorporates a magnetic tunnel junction with a free layer, a reference layer, a tunnel barrier layer, and a nonmagnetic spin Hall effect layer comprising an alternating stack of beta phase tungsten layers and noble metal dusting layers, along with a hafnium layer, to reduce the critical switching current and stabilize the beta phase tungsten material, ensuring efficient and accurate data storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional STT or SOT MRAM devices are used, then data storage capability is achieved, but large currents are required for switching and disturbances occur to neighboring cells
Solution Approach 1:
The patent employs a composite spin Hall effect layer comprising multiple materials (beta-phase tungsten, platinum, cobalt-iron-boron) arranged in a layered structure. This composite configuration optimizes the spin Hall angle and reduces the critical switching current density while maintaining stable magnetization, thereby resolving the contradiction between achieving reliable data storage and reducing switching current requirements
Solution Approach 2:
The patent modifies key material parameters including the thickness of each layer (e.g., beta-phase tungsten layer thickness of 1-3 nm, platinum layer thickness of 0.3-1 nm), the composition ratios, and the magnetic anisotropy energy. These parameter optimizations enable lower switching currents while preventing thermal instability and cross-cell disturbances, thus improving reliability without excessive energy consumption
2Productivity
If device density is increased, then storage capacity improves, but cross-cell disturbances increase causing erroneous writing
Solution Approach 1:
The patent implements local quality optimization by configuring the spin Hall effect layer with specific material compositions and thicknesses in different regions of the magnetic tunnel junction. The beta-phase tungsten/platinum/Cobalt-iron-boron structure provides localized spin polarization that confines the switching effect to the target cell, reducing cross-cell disturbances while enabling high-density storage configurations
Solution Approach 2:
The spin Hall effect layer acts as an intermediary that converts charge current to spin current, which then exerts torque on the magnetization of the free layer. This intermediary mechanism allows for more precise control of the switching process, reducing spurious effects on neighboring cells while maintaining high storage density capability
3Use of energy by moving object
If beta phase tungsten is used in the spin Hall effect layer, then switching current is reduced, but thermal instability occurs at elevated temperatures
Solution Approach 1:
The patent creates a composite structure where beta-phase tungsten is combined with platinum and cobalt-iron-boron layers. The platinum layer serves as a stabilizing matrix that maintains the beta phase of tungsten at elevated temperatures, while the cobalt-iron-boron layer provides the necessary spin polarization. This composite approach preserves the low switching current benefit of beta-phase tungsten while eliminating its thermal instability
Solution Approach 2:
The patent employs thin layers (beta-phase tungsten layer thickness of 1-3 nm, platinum layer thickness of 0.3-1 nm) that are sufficient to provide the desired spin Hall effect and thermal stability. These thin films maintain the beta phase structure through the deposition and annealing processes while minimizing the total material volume that could undergo phase transformation, effectively using minimal stable structures to overcome the inherent instability of thicker beta-phase tungsten
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
The solution effectively lowers the switching current and enhances data storage accuracy by maintaining the beta phase of tungsten, reducing thermal instability, and minimizing disturbances between cells, thereby improving the density and reliability of MRAM devices.
Implementation Method 1
a nonmagnetic spin Hall effect layer. The spin Hall effect layer includes an alternating stack of beta phase tungsten layers and noble metal nonmagnetic dusting layers
Implementation Method 2
Spin transfer torque (STT) MRAM devices are similar to conventional MRAM devices except that the write current paths pass through the magnetic layers of each memory element. The free layer is set via the spin transfer torque from the spin polarized current passing through the reference magnetic layer
Implementation Method 3
The spin Hall effect layer includes an alternating stack of beta phase tungsten layers and noble metal nonmagnetic dusting layers. Each of the beta phase tungsten layers has a thickness in a range from 0.2 nm to 1.2 nm
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A spin orbit torque magnetoresistive random access memory (SOT MRAM) cell includes a magnetic tunnel junction that contains a free layer having two bi-stable magnetization directions, a reference magnetic layer having a fixed magnetization direction, and a tunnel barrier layer located between the free layer and the reference layer, and a nonmagnetic spin Hall effect layer. The spin Hall effect layer may include an alternating stack of beta phase tungsten layers and noble metal nonmagnetic dusting layers. Alternatively or in addition, a hafnium layer may be located between the nonmagnetic spin Hall effect layer and the free layer.