Scatterometry Weighting Function for Undesired Diffraction Noise

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Scatterometry metrology systems face challenges in accurately measuring sample attributes due to noise introduced by undesired diffraction orders, which can contaminate data and lead to inaccuracies in feature dimension measurements, especially when using large numerical apertures that capture unwanted diffraction orders.

Innovation Solution

A metrology system and method that generate a weighting function to de-emphasize spectroscopic scatterometry data associated with undesired diffraction orders, allowing for accurate attribute measurement by balancing contamination impact across a wide spectral range, using techniques like ray-tracing and rigorous coupled-wave analysis to estimate contamination intensity and generate weights.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a large numerical aperture is used to collect light for metrology measurements, then the measurement precision is improved, but noise from undesired diffraction orders increases

Engineering Contradiction:
Improvemeasurement precisionVSAvoidnoise from undesired diffraction orders
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent segments the spectral data by applying wavelength-dependent weighting functions that separate desired diffraction orders from undesired ones. The weighting function divides the spectral range into regions where different diffraction orders dominate, allowing selective emphasis on clean spectral regions while suppressing contaminated regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter of spectral weighting across different wavelengths. By varying the weight assigned to each wavelength based on the predicted contamination level, the system optimizes the contribution of each spectral region to the regression analysis, thereby improving measurement precision while accounting for diffraction order contamination.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If spectroscopic scatterometry data across a wide spectral range is used, then the measurement accuracy is improved, but the impact of contamination from undesired diffraction orders increases

Engineering Contradiction:
Improvemeasurement accuracyVSAvoiddata integrity
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent implements dynamic weighting of spectral data points based on their susceptibility to contamination. The weighting function is calculated to reflect the varying reliability of different wavelength regions, allowing the regression analysis to dynamically adjust the influence of each spectral point based on predicted contamination levels from undesired diffraction orders.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The weighting function serves as an intermediary between the raw spectroscopic scatterometry data and the regression analysis. It mediates the contribution of each wavelength by applying appropriate weights that account for contamination, thereby protecting the measurement accuracy from the harmful effects of undesired diffraction orders while still utilizing the full spectral range.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If all captured light is used for regression analysis, then the signal-to-noise ratio is improved, but noise from undesired diffraction orders contaminates the data fitting

Engineering Contradiction:
Improvesignal-to-noise ratioVSAvoidfeature dimension measurement accuracy
Core Design Contradiction:
Measurement precisionVSManufacturing precision

Solution Approach 1:

The patent applies partial action by selectively weighting only the portions of the spectral data that are least contaminated by undesired diffraction orders. Rather than using all captured light equally, the method applies differential weighting to optimize the signal-to-noise ratio while preventing contamination from degrading the feature dimension measurement accuracy.

Inventive Principle:
Principle #16Partial or excessive action

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 solution enables highly accurate measurements by effectively reducing the impact of noise from undesired diffraction orders, maintaining data integrity across a wide spectral range, and improving the fit of scatterometry data to the model, thus enhancing the precision of critical dimension and other feature measurements.

Implementation Method 1

Light incident on a sample is reflected, scattered, and/or diffracted at various angles based on the structure and composition of features on a sample

Methodology Applied
Scientific EffectReflection: Reflection

Implementation Method 2

Light incident on a sample is reflected, scattered, and/or diffracted at various angles based on the structure and composition of features on a sample

Methodology Applied
Scientific EffectScattering: Scattering

Implementation Method 3

Light incident on a sample is reflected, scattered, and/or diffracted at various angles based on the structure and composition of features on a sample

Methodology Applied
Scientific EffectDiffraction: Diffraction

Implementation Method 4

generates a weighting function for the model to de-emphasize portions of the spectroscopic scatterometry data associated with one or more wavelengths within the selected range of wavelengths at which light captured by the scatterometry tool when measuring the target is predicted to include one or more undesired diffraction orders

Methodology Applied
Scientific EffectDiffraction: Diffraction

Data Source

PatentEP3891489B1Scatterometry modeling in the presence of undesired diffraction orders
Publication Date: 2024.03.06 KLA CORP
  • EP3891489B1 patent drawingFigure 1A
  • EP3891489B1 patent drawingFigure 1B
  • EP3891489B1 patent drawingFigure 1C

AI summary

A metrology system may receive a model for measuring one or more selected attributes of a target including features distributed in a selected pattern based on regression of spectroscopic scatterometry data from a scatterometry tool for a range of wavelengths. The metrology system may further generate a weighting function for the model to de-emphasize portions of the spectroscopic scatterometry data associated with wavelengths at which light captured by the scatterometry tool when measuring the target is predicted to include undesired diffraction orders. The metrology system may further direct the spectroscopic scatterometry tool to generate scatterometry data of one or more measurement targets including fabricated features distributed in the selected pattern. The metrology system may further measure the selected attributes for the one or more measurement targets based on regression of the scatterometry data of the one or more measurement targets to the model weighted by the weighting function.