Filamentary Resistive Memory Device with Silver Filament Formation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current non-volatile memory devices face challenges such as dielectric breakdown, material incompatibility with CMOS manufacturing, and poor scalability and reliability as device sizes shrink, particularly with Flash memories and other types of non-volatile RAM devices.

Innovation Solution

The development of non-volatile resistive switch memory devices with a highly doped intermediate region between an undoped amorphous silicon layer and a doped polysilicon or silicon germanium layer, where a silver filament forms upon voltage application, enhancing data retention through a negative charge attraction mechanism.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If device sizes are reduced to scale memory devices, then memory density increases, but dielectric breakdown and reliability problems occur

Engineering Contradiction:
Improvememory densityVSAvoiddevice reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent changes the switching mechanism from charge-based (voltage threshold) to filament-based resistance change. By forming silver filaments through electrochemical migration at lower voltages, the device achieves reliable switching at scaled dimensions without dielectric breakdown, as the filament formation occurs through controlled ion migration rather than high-field stress.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical/electrical field-based switching mechanism of traditional FETs with an electrochemical filament formation mechanism. Silver ions migrate and form conductive filaments through the solid electrolyte, creating a resistance-based switching system that is more suitable for scaled dimensions and eliminates dielectric breakdown issues.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Speed

If high voltage is applied for programming Flash memories, then programming speed improves, but dielectric breakdown occurs

Engineering Contradiction:
Improveprogramming speedVSAvoiddielectric breakdown
Core Design Contradiction:
SpeedVSObject-affected harmful factors

Solution Approach 1:

The patent replaces high-voltage charge injection with low-voltage electrochemical filament formation. Silver ions migrate through the solid electrolyte under low voltage to form conductive filaments, achieving fast switching without the dielectric breakdown associated with high-voltage programming in Flash memories.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the programming mechanism from high-voltage electrical field injection to low-voltage electrochemical ion migration. The switching action is achieved through controlled silver ion movement and filament formation, enabling fast programming at reduced voltages that eliminate dielectric stress.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If new materials are used for next generation memory devices, then operational characteristics improve, but compatibility with CMOS manufacturing decreases

Engineering Contradiction:
Improveoperational characteristicsVSAvoidCMOS compatibility
Core Design Contradiction:
ReliabilityVSEase of manufacture

Solution Approach 1:

The patent uses a multi-layer structure where each layer serves multiple functions: the solid electrolyte provides both the ion transport medium and the switching medium, the metal layers serve as both electrodes and silver ion sources, and the doped semiconductor layers provide both structural support and electrical contacts. This universal approach enables CMOS compatibility while achieving superior operational characteristics.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent employs a composite structure combining solid electrolyte, metal layers, and doped semiconductor layers. This composite material approach integrates the benefits of each material type: ion transport from the solid electrolyte, conductive pathways from metal layers, and structural/electrical properties from doped semiconductors, achieving both improved operation and CMOS compatibility.

Inventive Principle:
Principle #40Composite materials

4Productivity

If device dimensions are reduced for scaling, then memory density increases, but manufacturing precision requirements increase

Engineering Contradiction:
Improvememory densityVSAvoidfabrication precision
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The patent segments the memory cell into distinct functional layers: solid electrolyte layer, metal layers, and doped semiconductor layers. This segmentation allows each layer to be optimized and fabricated independently using standard CMOS processes, reducing the overall manufacturing precision requirements while enabling high-density scaling.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent designs layers that serve multiple functions to reduce the number of fabrication steps and precision requirements. For example, metal layers serve as both electrodes and silver ion reservoirs, and doped semiconductor layers provide both structural support and electrical contacts, simplifying the manufacturing process for scaled devices.

Inventive Principle:
Principle #6Universality (Multi-functionality)

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

This configuration improves data retention and scalability of memory devices by forming a stable silver filament, reducing power consumption and maintaining the filament's longevity, thus addressing the limitations of existing memory technologies.

Implementation Method 1

an interface region between the non-metallic conductive layer and the switching media having a negative ionic charge... positive metal (e.g. silver) ions from the silver metal migrate into the amorphous silicon

Methodology Applied
Scientific EffectIon attraction: Ion Repulsion/Attraction

Implementation Method 2

conductive filaments comprising positive metallic ions... a silver 'filament' is formed, similar to a short-circuit through the a-Si

Methodology Applied
Scientific EffectElectrical conduction: Conduction (electrical)

Implementation Method 3

dopant impurities from the highly doped region typically migrate to the lower doped region. The migration of holes from the intermediate region, in various embodiments, result in the intermediate region having a negative charge

Methodology Applied
Scientific EffectDopant migration: Diffusion

Data Source

PatentUS8796658B1Filamentary based non-volatile resistive memory device and method
Publication Date: 2014.08.05 CROSSBAR INC
  • US8796658B1 patent drawing
  • US8796658B1 patent drawing
  • US8796658B1 patent drawing

AI summary

A resistive memory device includes a first metallic layer comprising a source of positive metallic ions, a switching media having an upper surface and a lower surface, wherein the upper surface is adjacent to the first metallic layer, wherein the switching media comprises conductive filaments comprising positive metallic ions from the source of positive metallic ions formed from the upper surface towards the lower surface, a semiconductor substrate, a second metallic layer disposed above the semiconductor substrate, a non-metallic conductive layer disposed above the second metallic layer, and an interface region between the non-metallic conductive layer and the switching media having a negative ionic charge.