Model Building Engine for Combined X-Ray and Optical Metrology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current metrology systems face challenges in accurately measuring complex three-dimensional semiconductor structures and parameters, particularly those smaller than 10 nm and those employing opaque materials, due to limitations in optical radiation penetration and increased parameter correlation.
Innovation Solution
A combined analysis system that integrates optical and x-ray metrology techniques, using a model building and analysis engine to identify shared geometric parameters and reduce correlations by performing sequential, parallel, or combined fitting of optical and x-ray measurement data, with the inclusion of field enhancement elements to enhance measurement sensitivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If optical metrology techniques are used to measure semiconductor structures, then throughput is maintained high, but measurement precision deteriorates for structures smaller than 10 nm and those with opaque materials due to limited radiation penetration
Solution Approach 1:
The patent combines optical metrology and x-ray metrology into a unified measurement system. Optical metrology maintains high throughput while x-ray metrology provides enhanced penetration for opaque materials and better precision for sub-10nm structures. The two techniques are merged at the data processing level through a common model fitting engine that integrates measurements from both sources to achieve both high throughput and high precision simultaneously.
2Adaptability or versatility
If complex three-dimensional structures with diverse materials are measured, then structural complexity increases, but parameter correlation increases making it difficult to decouple parameters with available measurements
Solution Approach 1:
The patent introduces a unified physical model as an intermediary that connects optical and x-ray measurements to structural parameters. This model serves as a mediator that translates measurements from both techniques into decoupled structural parameters. By using a common model framework that incorporates the physics of both optical and x-ray interactions with matter, the system can distinguish between parameters that affect optical responses versus those that affect x-ray responses, thereby reducing parameter correlation.
3Measurement precision
If multiple measurement technologies are combined to characterize complex structures, then measurement accuracy improves, but analysis complexity increases
Solution Approach 1:
The patent creates a universal analysis engine that handles both optical and x-ray measurements through a single unified framework. This multi-functional engine can process measurements from either technique or both simultaneously using the same model fitting algorithms. By designing the analysis system to be universal rather than requiring separate analysis pipelines for each technique, the patent reduces analysis complexity while maintaining improved measurement accuracy from combined measurements.
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 improves measurement sensitivity, throughput, and accuracy by leveraging the complementary nature of x-ray and optical techniques, allowing for precise determination of structural and material characteristics in semiconductor fabrication processes.
Implementation Method 1
x-ray radiation can penetrate layers constructed of these materials
Implementation Method 2
measuring multiple Mueller matrix elements in addition to the more conventional reflectivity or ellipsometric signals
Data Source
AI summary
Structural parameters of a specimen are determined by fitting models of the response of the specimen to measurements collected by different measurement techniques in a combined analysis. Models of the response of the specimen to at least two different measurement technologies share at least one common geometric parameter. In some embodiments, a model building and analysis engine performs x-ray and optical analyses wherein at least one common parameter is coupled during the analysis. The fitting of the response models to measured data can be done sequentially, in parallel, or by a combination of sequential and parallel analyses. In a further aspect, the structure of the response models is altered based on the quality of the fit between the models and the corresponding measurement data. For example, a geometric model of the specimen is restructured based on the fit between the response models and corresponding measurement data.