Strained Heterojunction Release via Sacrificial Etching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for forming high-quality strained heterojunction semiconductor structures face challenges in controlling strain and reducing dislocation defects in SiGe films, leading to poor carrier mobility and thermal conductivity, and are limited by the difficulty in achieving compliant substrates and maintaining strain states in adjacent devices.

Innovation Solution

A method involving the growth of multilayer films on a crystalline template layer supported by a sacrificial layer, allowing for controlled strain relaxation and release of the multilayer structure, which can be mounted on various substrates to form devices with improved carrier mobility and strain distribution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If thick graded SiGe films are grown to achieve strain relaxation, then strain control is improved, but dislocation density increases and thermal conductivity deteriorates

Engineering Contradiction:
Improvestrain controlVSAvoiddislocation density
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent segments the continuous graded SiGe buffer into multiple discrete SiGe layers with different Ge compositions (e.g., 0%, 5%, 10%, 15% Ge). Each layer is thin enough to remain strain-free while collectively providing the desired strain profile. This segmentation prevents dislocation formation that would occur in thick continuous graded layers, thus resolving the contradiction between strain control and dislocation density.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary strain engineering by growing thin SiGe layers with controlled Ge content before growing the final strained Si channel. The SiGe layers are prepared in advance with compositions that induce the desired tensile strain in the Si channel, eliminating the need for thick graded buffers and associated dislocations.

Inventive Principle:
Principle #10Preliminary action

2Manufacturing precision

If thick graded SiGe films are grown to achieve strain relaxation, then strain control is improved, but film thickness increases making adjacent devices difficult to form

Engineering Contradiction:
Improvestrain controlVSAvoidfilm thickness
Core Design Contradiction:
Manufacturing precisionVSLength of moving object

Solution Approach 1:

The patent replaces thick continuous graded buffers with multiple thin discrete SiGe layers. Each layer is only tens of nanometers thick, yet collectively they provide the same strain control function as a micrometer-thick graded buffer. This dramatically reduces the vertical space required, enabling adjacent devices to be formed in the same wafer.

Inventive Principle:
Principle #1Segmentation

3Manufacturing precision

If dislocation-driven relaxation is used, then strain relaxation is achieved, but threading dislocations extend through the strained-Si layer degrading carrier mobility

Engineering Contradiction:
Improvestrain relaxationVSAvoidcarrier mobility
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent performs preliminary strain engineering by growing thin SiGe layers with controlled Ge content before growing the final strained Si channel. The SiGe layers are prepared in advance with compositions that induce the desired tensile strain in the Si channel, eliminating the need for thick continuous graded buffers and associated dislocations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the composition parameter of SiGe layers from continuous grading to discrete steps (0%, 5%, 10%, 15% Ge). Each layer has a specific Ge content that provides controlled strain without exceeding the critical thickness for dislocation formation. This parameter change enables strain relaxation without threading dislocations penetrating the Si channel.

Inventive Principle:
Principle #35Parameter changes

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 approach enables precise control of strain in semiconductor layers, reducing dislocation defects and enhancing carrier mobility, facilitating the fabrication of high-performance devices such as modulation-doped field-effect transistors and imaging devices with increased speed.

Implementation Method 1

the final strain state is achieved by elastic strain sharing between the films

Methodology Applied
Scientific EffectElastic strain: Elasticity

Implementation Method 2

The sacrificial layer is then preferentially etched at the exposed areas thereof to release the multilayer structure

Methodology Applied
Scientific EffectSelective etching:

Data Source

PatentUS7973336B2Released freestanding strained heterojunction structures
Publication Date: 2011.07.05 WISCONSIN ALUMNI RES FOUND
  • US7973336B2 patent drawing
  • US7973336B2 patent drawing
  • US7973336B2 patent drawing

AI summary

Growth of multilayer films is carried out in a manner which allows close control of the strain in the grown layers and complete release of the grown films to allow mounting of the released multilayer structures on selected substrates. A layer of material, such as silicon-germanium, is grown onto a template layer, such as silicon, of a substrate having a sacrificial layer on which the template layer is formed. The grown layer has a lattice mismatch with the template layer so that it is strained as deposited. A top layer of crystalline material, such as silicon, is grown on the alloy layer to form a multilayer structure with the grown layer and the template layer. The sacrificial layer is preferentially etched away to release the multilayer structure from the sacrificial layer, relaxing the grown layer and straining the crystalline layers interfaced with it.