Noise Map Segmentation for Wafer Inspection Sensitivity

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current semiconductor inspection methods are inadequate in sensitivity due to subjective noise area selection and static noise statistics, leading to inefficient defect detection and process control, especially as design rules shrink, causing increased detection of both yield-relevant and nuisance defects.

Innovation Solution

A system and method that generate noise maps to segment regions based on intensity statistics, using these maps as segmentation masks to suppress noise during wafer inspection, correlating them with design information and care areas to improve sensitivity and automate recipe setup.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If operator experience or design information is used to select inspection areas, then noise avoidance is attempted, but inspection sensitivity is reduced due to subjectivity and inaccuracy

Engineering Contradiction:

Improveinspection sensitivityVSAvoidsubjectivity of noise area selection

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system performs self-service by automatically generating noise maps and identifying noise regions without operator intervention. The processor analyzes intensity statistics from multiple measurements to create dynamic noise maps that automatically segment the wafer into noise and non-noise regions, eliminating subjective operator selection while maintaining high inspection sensitivity.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system changes parameters by using dynamic intensity statistics from multiple measurements to create adaptive noise maps. Instead of static care areas, the system continuously updates noise region identification based on measured intensity variations, allowing optimal detection thresholds to be applied dynamically across different wafer regions.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If static noise statistics are used in inspection, then process is simplified, but inspection sensitivity is reduced due to inability to adapt to varying noise levels

Engineering Contradiction:

Improveinspection sensitivityVSAvoidcomplexity of noise statistics processing

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary action by generating noise maps before the actual inspection process. Multiple intensity measurements are taken and analyzed in advance to create the noise map, which then guides the inspection process. This preliminary noise characterization enables optimized detection thresholds to be established before defect detection begins.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system applies segmentation by dividing the wafer into distinct noise and non-noise regions based on the generated noise map. This spatial segmentation allows different detection parameters and thresholds to be applied to different regions, improving sensitivity in low-noise areas while maintaining robustness in high-noise areas.

Inventive Principle:

Principle #1Segmentation

3Productivity

If design rules are shrunk to increase device density, then productivity is improved, but defect detection becomes more difficult due to smaller defect sizes

Engineering Contradiction:

Improvedevice densityVSAvoiddefect detection difficulty

Core Design Contradiction:

ProductivityVSDifficulty of detecting and measuring

Solution Approach 1:

The system applies local quality by using spatially-varying detection parameters based on the noise map. Different regions of the wafer receive customized inspection sensitivity settings according to their local noise characteristics. This allows maximum detection sensitivity to be achieved in each local region, improving the ability to detect small defects in high-density designs.

Inventive Principle:

Principle #3Local quality

4Reliability

If all detected defects are corrected, then yield is improved, but manufacturing cost increases due to expensive process corrections for nuisance defects

Engineering Contradiction:

ImproveyieldVSAvoidmanufacturing cost

Core Design Contradiction:

ReliabilityVSEase of manufacture

Solution Approach 1:

The system extracts and separates nuisance defects from yield-relevant defects by comparing detected defects against the generated noise map. Defects that coincide with high-noise regions are identified as nuisance defects and excluded from correction actions. This extraction process reduces unnecessary process corrections while maintaining yield improvement from genuine defects.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS11295432B2Broad band plasma inspection based on a nuisance map

Publication Date: 2022.04.05 KLA CORP

AI summary

A noise map is used for defect detection. One or more measurements of intensities at one or more pixels are received and an intensity statistic is determined for each measurement. The intensity statistics are grouped into at least one region and stored with at least one alignment target. A wafer can be inspected with a wafer inspection tool using the noise map. The noise map can be used as a segmentation mask to suppress noise.