Optical Metrology Grating Anomaly Reduction via Parameter Subsets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical metrology systems face challenges in accurately measuring semiconductor structures due to increased complexity and sensitivity to grating anomalies, which lead to computational inefficiencies and systematic errors, especially when dealing with small feature sizes and multi-parameter correlations.
Innovation Solution
The system reduces sensitivity to grating anomalies by selecting a subset of available system parameter values for measurement analysis, using a measurement model that optimizes precision, sensitivity, accuracy, and computational effort, and incorporates optical system models to correct errors at arbitrary azimuth angles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spectroscopic scatterometry measurements are performed at multiple angles of incidence and multiple azimuth angles to increase measurement sensitivity and diversity, then measurement precision and accuracy are improved, but the risk of triggering significant grating anomalies increases
Solution Approach 1:
The system performs preliminary calculations of Rayleigh wavelengths and identification of grating anomaly locations before executing the actual measurement. This allows the measurement wavelengths to be selected in advance to avoid anomalous regions, thus preventing grating anomalies from affecting measurement precision while still enabling multi-angle measurements for enhanced sensitivity
Solution Approach 2:
The system dynamically adjusts measurement parameters (wavelengths, angles of incidence, azimuth angles) based on calculated grating anomaly locations. By changing these parameters to avoid anomalous regions while maintaining measurement diversity, the system preserves measurement precision without triggering significant grating anomalies
2Productivity
If simplified models of grating anomalies (such as reflection Rayleigh manifold) are used, then computational effort is reduced, but the models fail to sufficiently capture observed grating anomalies in spectroscopic measurements
Solution Approach 1:
The system segments the spectral range into different regions based on calculated Rayleigh wavelengths and grating anomaly locations. By dividing the spectrum into safe regions (free from anomalies) and anomalous regions, the system can selectively use data from safe regions for regression analysis, achieving both computational efficiency and accurate anomaly capture
Solution Approach 2:
The system introduces an intermediary calculation layer that computes Rayleigh wavelengths and identifies grating anomaly locations before the main measurement regression. This intermediary step provides accurate anomaly information without requiring complex full-wave simulations during the actual measurement, thus maintaining computational efficiency while improving anomaly capture accuracy
3Reliability
If the spectral range where grating anomalies arise is excluded from measurement, then regression results suffer from systematic errors is avoided, but measurement diversity and sensitivity are reduced
Solution Approach 1:
The system applies local quality by treating different spectral regions differently based on their anomaly characteristics. Instead of uniformly excluding entire spectral ranges, the system identifies specific wavelength regions free from grating anomalies and selectively uses only those regions for regression analysis. This localized approach preserves measurement diversity within safe regions while ensuring regression accuracy by avoiding anomalous regions
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 and efficient optical critical dimension measurements with reduced computational burden, improving measurement accuracy and throughput by minimizing the impact of grating anomalies.
Implementation Method 1
spectroscopic scatterometry measurements performed at multiple angles of incidence (AOI) and multiple azimuth angles
Implementation Method 2
spectroscopic scatterometry measurements performed at multiple angles of incidence (AOI) and multiple azimuth angles
Implementation Method 3
incorporates optical system models to correct errors at arbitrary azimuth angles
Data Source
AI summary
Methods and systems for performing broadband spectroscopic metrology with reduced sensitivity to grating anomalies are presented herein. A reduction in sensitivity to grating anomalies is achieved by selecting a subset of available system parameter values for measurement analysis. The reduction in sensitivity to grating anomalies enables an optimization of any combination of precision, sensitivity, accuracy, system matching, and computational effort. These benefits are particularly evident in optical metrology systems having large ranges of available azimuth angle, angle of incidence, illumination wavelength, and illumination polarization. Predictions of grating anomalies are determined based on a measurement model that accurately represents the interaction between the measurement system and the periodic metrology target under measurement. A subset of available system parameter values is selected to reduce the impact of grating anomalies on measurement results. The selected subset of available system parameters is implemented on a configurable spectroscopic metrology system performing measurements.