FinFET T-Shaped Channel Etching for Higher Mobility Density

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The semiconductor industry faces challenges in increasing integration density and mobility of electronic components due to limitations in feature size reduction and channel width in traditional transistor designs.

Innovation Solution

The formation of vertically stacked, self-aligned 'T' shaped channel regions through a self-limited etching process, which increases effective channel width and mobility by utilizing crystal directions, and is compatible with CMOS fabrication processes at a relatively low cost.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional transistor designs are used, then manufacturing simplicity is maintained, but integration density and mobility are limited

Engineering Contradiction:
Improveintegration densityVSAvoidtransistor structure complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent transitions from planar 2D channel structures to three-dimensional T-shaped and vertically stacked channel regions. This dimensional change increases the effective channel width and surface area without proportionally increasing the footprint area, thereby improving integration density while managing structural complexity through self-aligned fabrication processes

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

Solution Approach 2:

The channel region is segmented into multiple T-shaped or vertically stacked segments rather than a single continuous region. This segmentation increases the total effective channel width and allows for better control of carrier mobility by utilizing different crystal orientations in different segments, while the self-aligned process keeps manufacturing complexity manageable

Inventive Principle:
Principle #1Segmentation

2Productivity

If feature size is reduced to increase integration density, then more components fit in given area, but mobility and performance deteriorate

Engineering Contradiction:
Improveintegration densityVSAvoidcarrier mobility
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

Instead of further reducing the lateral footprint dimensions, the patent extends the channel into the third dimension with T-shaped and vertically stacked structures. This maintains or reduces the footprint area for higher integration density while providing longer effective channel paths and multiple crystal orientations that enhance carrier mobility and device performance

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

Solution Approach 2:

Different portions of the channel region are engineered with different crystal orientations and geometries to optimize local carrier transport properties. The T-shaped and stacked structures create regions with favorable crystallographic orientations for high mobility, while maintaining overall compact dimensions for high integration density

Inventive Principle:
Principle #3Local quality

3Reliability

If channel width is increased to improve mobility, then device area increases, but integration density decreases

Engineering Contradiction:
Improvecarrier mobilityVSAvoiddevice footprint area
Core Design Contradiction:
ReliabilityVSArea of stationary object

Solution Approach 1:

The patent achieves increased effective channel width by extending the channel into the vertical dimension with stacked structures and T-shaped cross-sections. This provides larger effective channel area for improved carrier mobility while keeping the lateral footprint dimensions compact, thereby maintaining high integration density

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

Solution Approach 2:

Multiple channel regions are nested vertically in a stacked configuration, with each stack containing multiple T-shaped channel regions at different heights. This nesting approach consolidates multiple high-mobility channels into a compact vertical structure, improving mobility without increasing the lateral device footprint area

Inventive Principle:
Principle #7Nested doll (Nesting)

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 enhances mobility and integration density by optimizing channel dimensions and manufacturing compatibility, while maintaining cost-effectiveness.

Implementation Method 1

performing a self-limited anisotropic etching process to form T-shaped channel regions

Methodology Applied
Scientific EffectAnisotropic etching:

Implementation Method 2

performing a self-limited solution based etching process to form T-shaped channel regions... performing a second etching process to form a second portion of the first recess... performing a third etching process to remove the first dielectric layer

Methodology Applied
Scientific EffectCrystal direction-dependent etching:

Data Source

PatentUS11887985B2Semiconductor device and method
Publication Date: 2024.01.30 TAIWAN SEMICONDUCTOR MANUFACTURING CO LTD
  • US11887985B2 patent drawing
  • US11887985B2 patent drawing
  • US11887985B2 patent drawing

AI summary

A method includes etching a substrate to form a semiconductor fin, forming a gate stack on a top surface and sidewalls of the semiconductor fin, and forming a first recess in the semiconductor fin on a side of the gate stack, wherein forming the first recess comprises, performing a first etching process to form a first portion of the first recess, depositing a first dielectric layer on sidewalls of the gate stack and the first portion of the first recess, performing a second etching process to form a second portion of the first recess using the first dielectric layer as a mask, wherein the second portion of the first recess extends under the gate stack, and performing a third etching process to remove the first dielectric layer.