Acoustic Wafer Defect Detection via Resonance Frequency Shifts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional wafer defect inspection systems are inadequate in detecting systematic defects and are prone to missing hairline fractures and internal defects, particularly due to their sensitivity to surface texture and lighting variations, and require lengthy measurement times, while existing ultrasound techniques are sensitive to transducer placement and material uniformity.
Innovation Solution
The use of acoustic signals to induce vibrations in wafers, measuring linear and nonlinear frequency response metrics to identify defects, employing bismuth telluride (Bi2Te3) wafers and a piezoelectric transducer with point contact excitation, allowing for simultaneous crack detection throughout the wafer without requiring uniform dimensions or material properties, and integrating multiple damage metrics for rapid identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If optical microscopy is used for wafer inspection, then surface defects can be detected, but hairline fractures and internal defects are missed and a significant number of flawed wafers pass through
Solution Approach 1:
The patent replaces optical inspection methods with acoustic resonance techniques. A transducer generates acoustic waves that excite the wafer, and a sensor detects changes in resonance frequency caused by defects. This mechanical/acoustic substitution enables detection of internal defects, hairline fractures, and delaminations that are invisible to optical methods, while maintaining detection of surface defects.
2Measurement precision
If conventional ultrasound burst techniques are used, then cracks can be detected, but the method requires precise transducer placement and uniform wafer geometry
Solution Approach 1:
The patent employs mechanical vibration at resonance frequencies to detect cracks. By exciting the wafer at its natural resonant frequencies and measuring frequency shifts, the system can detect cracks without requiring precise transducer placement. The resonance-based approach inherently accounts for variations in wafer geometry and material properties, eliminating the need for uniform wafer dimensions.
3Measurement precision
If resonance ultrasound techniques are used, then standing waves can be established for crack detection, but measurement times become prohibitively long
Solution Approach 1:
The patent accelerates the inspection process by measuring only the fundamental resonance frequency and a limited number of higher modes, rather than performing exhaustive frequency sweeps. The system rapidly excites the wafer and quickly captures the resonance response, reducing measurement time from minutes to seconds while maintaining high detection sensitivity for cracks and defects.
4Measurement precision
If current ultrasound techniques are used, then cracks can be interrogated, but the methods mistake nonlinearities within materials for cracks and require uniform material properties
Solution Approach 1:
The patent monitors changes in resonance frequency as the primary indicator of cracks, rather than relying on amplitude or time-of-flight measurements that are sensitive to material nonlinearities. By focusing on frequency shifts caused by crack-induced stiffness reductions, the system can distinguish actual cracks from material heterogeneities, allowing inspection of wafers with non-uniform material properties.
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 rapid, noninvasive, and efficient detection of wafer defects, including those that would fracture due to processing-induced stresses, with high accuracy comparable to optical microscopy, preventing flawed wafers from progressing downstream and reducing inspection time to less than 3 minutes.
Implementation Method 1
employing bismuth telluride (Bi2Te3) wafers and a piezoelectric transducer with point contact excitation
Implementation Method 2
exciting a wafer using an acoustic signal to cause the wafer to exhibit vibrations, measuring one or more of linear frequency response metrics or nonlinear frequency responses metrics associated with the vibrations
Data Source
AI summary
Techniques are provided for detecting wafer defects. Example techniques include exciting a wafer using an acoustic signal to cause the wafer to exhibit vibrations, measuring one or more of linear frequency response metrics or nonlinear frequency responses metrics associated with the vibrations, and identifying any defects in the wafer based at least in part on one or more of the linear frequency response metrics or nonlinear frequency responses metrics. In embodiments, the wafer includes bismuth telluride (Bi2Te3).