GAA Metal Gate Stack Tuning for Threshold Voltage Scaling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The challenge of tuning threshold voltage in field effect transistors (FETs) is exacerbated by scaling constraints, particularly in gate structures of nanostructured FETs, where increasing the thickness of work function layers like TiAl or TiAlC is limited by shrinking spacing between channel regions, leading to poor gap-fill and resistance variations.

Innovation Solution

The implementation of multi-layered gate structures with adjustable work function layers, including titanium nitride and tantalum nitride, allows for independent tuning of threshold voltages in n-type and p-type FETs on the same substrate, using aluminum doping to achieve different work function values and threshold voltages, even in constrained geometries.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the thickness of work function layers (TiAl or TiAlC) is increased to tune threshold voltage, then threshold voltage tuning capability is improved, but manufacturing reliability deteriorates due to poor gap-fill and resistance variations in scaled devices

Engineering Contradiction:
Improvethreshold voltage tuning capabilityVSAvoidgap-fill quality and resistance uniformity
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The work function layer is segmented into multiple thinner layers (e.g., TiN and TaN alternating layers) instead of using a single thick layer. This segmentation allows better control of gap-fill quality and reduces resistance variations while maintaining the ability to tune threshold voltage through the combined work function of multiple layers.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent uses composite work function layers combining different materials (TiN and TaN) with different work function values. By adjusting the thickness ratio of different materials in the composite structure, threshold voltage can be tuned while achieving better gap-fill and reduced resistance variations compared to single-material thick layers.

Inventive Principle:
Principle #40Composite materials

2Productivity

If device dimensions are scaled down to manufacture smaller devices, then device density and performance are improved, but threshold voltage tuning capability deteriorates due to limited adjustment range in work function layer thickness

Engineering Contradiction:
Improvedevice density and performanceVSAvoidthreshold voltage tuning range
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

Instead of relying solely on varying the thickness of a single work function layer (one-dimensional adjustment), the patent introduces material composition as an additional dimension for tuning. By combining different materials with different work function values in alternating layers, the threshold voltage can be adjusted over a wider range without increasing the overall layer thickness, thus accommodating scaled device dimensions.

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

Solution Approach 2:

The composite structure of TiN and TaN layers provides an extended tuning range for threshold voltage. The different work function values of TiN and TaN allow fine-adjustment of the overall work function by controlling the thickness ratio of the two materials, enabling threshold voltage tuning in scaled devices where single-material layers would be too limited.

Inventive Principle:
Principle #40Composite materials

3Adaptability or versatility

If multi-layered gate structures with different work function materials are implemented, then threshold voltage tuning flexibility is improved, but device complexity increases

Engineering Contradiction:
Improvethreshold voltage tuning flexibilityVSAvoidgate structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The gate structure is segmented into multiple thin alternating layers of TiN and TaN rather than using a single thick layer of one material. This segmentation achieves better gap-fill and reduced resistance while maintaining manufacturing feasibility. The repeated pattern of thin layers simplifies the fabrication process compared to attempting to deposit and pattern single thick layers.

Inventive Principle:
Principle #1Segmentation

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 reliable and cost-effective manufacturing of FETs with varied threshold voltages, ensuring functional integrity and compliance with Moore's law by overcoming scaling limitations in nanostructured channel regions.

Implementation Method 1

using aluminum doping to achieve different work function values and threshold voltages

Methodology Applied
Scientific EffectDoping: Dopants

Data Source

PatentUS20250366015A1Metal gate structures of semiconductor devices
Publication Date: 2025.11.27 TAIWAN SEMICONDUCTOR MANUFACTURING CO LTD
  • US20250366015A1 patent drawing
  • US20250366015A1 patent drawing
  • US20250366015A1 patent drawing

AI summary

A method of fabricating a semiconductor device includes forming first and second nanostructured layers arranged in an alternating configuration on a substrate, forming first and second nanostructured channel regions in the first nanostructured layers, forming first and second gate-all-around structures wrapped around each of the first and second nanostructured channel regions. The forming the GAA structures includes depositing first and second gate barrier layers having similar material compositions and work function values on the first and second gate dielectric layers, forming first and second diffusion barrier layers on the first and second gate barrier layers, and doping the first and second gate barrier layers from a dopant source layer through the first and second diffusion barrier layers. The first diffusion barrier layer is thicker than the second diffusion barrier layer and the doped first and second gate barrier layers have work function values and doping concentrations different from each other.