Semiconductor Feature Measurement Using Prior Data Dimensionality Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Scatterometry-based inspection and metrology for semiconductor wafers face challenges in speed, precision, and accuracy due to increasing complexity of features, which results in prohibitively large computational processing times and parameter correlation issues, limiting the ability to accurately determine feature dimensions and electromagnetic scattering properties.
Innovation Solution
The technique reduces the dimensionality of feature models by reusing data from prior measurements, using a database to store and retrieve relevant data for feed-forward processing, thereby reducing computational complexity and parameter correlation, and incorporating this data into the measurement algorithm to enhance measurement precision and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of parameters in the model increases to account for feature complexity, then measurement precision is improved, but computational processing time becomes prohibitively large
Solution Approach 1:
The patent applies preliminary action by performing measurements and computations at test locations before the actual measurement at the device location. The scatterometry measurement is first performed at a test location to obtain diffraction data, and then this data is used to determine parameters for the device location, avoiding the need to perform full complex computations at every measurement point.
Solution Approach 2:
The patent segments the measurement process into distinct test locations and device locations. Test locations are used for initial measurements and parameter determination, while device locations utilize these pre-determined parameters for faster measurement. This segmentation allows the system to handle complexity by dividing the overall measurement task into manageable parts.
2Measurement precision
If the number of parameters in the model increases to account for multiple layers and shapes, then measurement precision is improved, but the computational complexity increases nearly exponentially
Solution Approach 1:
The patent performs the computationally intensive preliminary measurements and model fitting at test locations before measuring at device locations. By pre-determining parameters such as layer thicknesses and material properties at the test location, the system avoids repeating these complex computations at each device location, thereby reducing overall computational complexity.
Solution Approach 2:
The patent uses test locations as copies or representations of the device locations. The test location features are designed to be similar to device location features, allowing the measurements and parameter determinations at the test location to be replicated or applied to the device locations, thus avoiding redundant complex computations.
3Productivity
If parameter correlation is present between height and width, then measurement speed is improved by reducing computational complexity, but measurement precision deteriorates due to similar diffraction spectra
Solution Approach 1:
The patent applies local quality by performing specialized measurements at test locations that are designed to be sensitive to specific parameters. The test location features are configured to provide diffraction data that is particularly informative for determining parameters like layer thickness and material properties, allowing for accurate parameter determination even when correlations exist between geometric parameters.
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 significantly increases the speed of model-based measurements, reduces parameter correlation, and enhances precision and accuracy by leveraging prior data to simplify the computational complexity of scatterometry measurements for complex features.
Implementation Method 1
Scatterometry refers to an optical technique that analyzes diffraction to deduce structural details of a diffracting sample
Implementation Method 2
data that describe the electromagnetic scattering function of an underlying layer
Data Source
AI summary
The present application discloses a new technique which reduces the dimensionality of a feature model by re-use of data that has been obtained by a prior measurement. The data re-used from the prior measurement may range from parameters, such as geometrical dimensions, to more complex data that describe the electromagnetic scattering function of an underlying layer (for example, a local solution of the electric field properties).