3D DRAM Structure Using HAR Holes and Selective SiGe Etching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Two-dimensional memory structures are reaching theoretical limits in density, and existing three-dimensional memory devices face challenges in scalability and processing, making it difficult to achieve memory densities comparable to two-dimensional DRAM without significant cost increases.

Innovation Solution

The use of alternating heteroepitaxy layers of crystalline silicon and silicon germanium, combined with selective etching processes, allows for the formation of three-dimensional dynamic random-access memory (3D DRAM) structures with scalable dimensions, featuring gate all-around transistors and precise control over feature sizes, enabling efficient construction of memory arrays with high density.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If two-dimensional memory structures are used, then manufacturing process is simple, but memory density reaches theoretical limit

Engineering Contradiction:
Improvemanufacturing process simplicityVSAvoidmemory density
Core Design Contradiction:
Ease of manufactureVSQuantity of substance

Solution Approach 1:

The patent transitions from two-dimensional planar memory structures to three-dimensional vertically stacked memory structures. This dimensional change allows multiple memory layers to be stacked above a single substrate area, dramatically increasing memory density without requiring proportional increases in substrate area, thereby resolving the contradiction between manufacturing simplicity and memory density.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Quantity of substance

If three-dimensional memory structures are implemented, then memory density increases, but processing complexity and cost increase

Engineering Contradiction:
Improvememory densityVSAvoidprocessing complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent employs preliminary patterning actions where through-holes are formed first, followed by sequential deposition of alternating c-Si and c-SiGe layers. The selective etching of c-SiGe sacrificial layers is performed in advance to create voids that will later receive electrode materials. This preliminary structuring simplifies subsequent processing steps and reduces overall manufacturing complexity despite the three-dimensional architecture.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The memory structure is segmented into distinct functional layers: alternating c-Si active layers and c-SiGe sacrificial layers, with through-holes segmented into different regions for different purposes (capacitor formation, word line formation, bit line formation). This segmentation allows independent optimization and processing of each layer type, reducing overall processing complexity.

Inventive Principle:
Principle #1Segmentation

3Manufacturing precision

If alternating c-Si and c-SiGe layers are deposited, then selective etching capability is achieved, but manufacturing process complexity increases

Engineering Contradiction:
Improveselective etching capabilityVSAvoidmanufacturing process complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent introduces local quality differentiation by using c-SiGe layers specifically as sacrificial materials in certain regions while maintaining c-Si layers as active memory elements in other regions. This local differentiation enables selective etching processes that remove only c-SiGe, creating precise three-dimensional structures with high manufacturing precision. The localized function assignment justifies the increased process complexity.

Inventive Principle:
Principle #3Local quality

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 method enables the production of 3D DRAM structures with densities beyond current capabilities, achieving approximately 1300 um² per megabit, reducing manufacturing costs and time while maintaining structural integrity and performance.

Implementation Method 1

depositing alternating layers of crystalline silicon (c-Si) material and crystalline silicon germanium (c-SiGe) material using an heteroepitaxy process onto a substrate

Methodology Applied
Scientific EffectHeteroepitaxy: Epitaxy

Implementation Method 2

etching of a pattern of holes with at least one high aspect ratio (HAR) hole into the substrate

Methodology Applied
Scientific EffectChemical etching:

Implementation Method 3

a liner of a first dielectric material which is removable by exposure to hydrogen fluoride (HF)

Methodology Applied
Scientific EffectChemical etching by HF:

Data Source

PatentUS12581637B2Methods and structures for three-dimensional dynamic random-access memory
Publication Date: 2026.03.17 APPLIED MATERIALS INC
  • US12581637B2 patent drawing
  • US12581637B2 patent drawing
  • US12581637B2 patent drawing

AI summary

Methods for forming three-dimensional dynamic random-access memory (3D DRAM) structures that leverage a grid pattern of high aspect ratio holes to form subsequent features of the 3D DRAM. The method may include depositing alternating layers of crystalline silicon (c-Si) and crystalline silicon germanium (c-SiGe) using an heteroepitaxy process onto a substrate and HAR etching of a pattern of holes into the substrate. The holes configured to provide chemistry access to laterally etch or deposit materials to form 3D DRAM features without requiring subsequent HAR etching of holes to form the 3D DRAM features.