Lock-in Averaging for Semiconductor Defect Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor diagnostics methods, such as lock-in thermography, face challenges in enhancing the contrast of images and accurately localizing defects within semiconductor devices, often resulting in suppressed finer details and reduced thermal enhancement of hotspot signals.
Innovation Solution
A method and system that utilize lock-in averaging by switching the semiconductor device and an irradiating device on and off at different frequencies, while acquiring images with a camera, to obtain images with improved topographical contrast and amplitude signals indicative of defect locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If histogram modification methods are used to enhance thermogram appearance, then image contrast is improved, but finer details of the image are suppressed
Solution Approach 1:
The patent segments the image processing into multiple frequency components through Fourier transformation. By separating the thermogram into different spatial frequencies, the method can selectively enhance contrast in specific frequency ranges without suppressing finer details that exist in other frequency components. This resolves the contradiction by allowing contrast enhancement while preserving information across the full frequency spectrum.
Solution Approach 2:
The patent applies parameter changes by modifying the frequency domain representation of the thermogram. Through selective filtering and phase manipulation in the frequency domain, the method enhances contrast while maintaining the integrity of fine details. This approach transforms the problem from direct spatial domain enhancement to frequency domain parameter adjustment, avoiding information loss.
2Measurement precision
If high-performance infrared cameras are used to improve LIT measurement sensitivity, then defect detection quality is improved, but device cost increases
Solution Approach 1:
The patent creates a frequency domain copy of the spatial thermogram through Fourier transformation. By analyzing and processing this frequency domain representation, the method extracts defect information without requiring high-performance hardware. The frequency domain copy contains all necessary information for defect detection, allowing standard cameras to achieve enhanced measurement precision through computational processing rather than hardware upgrades.
Solution Approach 2:
The patent replaces the mechanical/hardware approach of using high-performance cameras with a computational/software-based frequency domain processing approach. By substituting hardware complexity with algorithmic processing, the method achieves improved defect detection quality while using standard, lower-cost infrared cameras. The frequency domain analysis acts as a computational substitute for hardware performance enhancement.
3Ease of operation
If conventional thermography is used for defect detection, then the process is simple, but thermal enhancement of hotspot signals is reduced
Solution Approach 1:
The patent applies periodic action by modulating the excitation signal at a specific frequency and using synchronous detection in the frequency domain. This periodic modulation enhances the hotspot signals by concentrating thermal energy at the modulation frequency, making defect detection more sensitive. The method maintains operational simplicity by using standard lock-in thermography procedures combined with frequency domain processing, achieving thermal enhancement without complicating the basic操作流程.
Solution Approach 2:
The patent introduces asymmetry in the frequency domain processing by applying selective filtering and phase manipulation that asymmetrically enhances certain frequency components associated with defect hotspots. This asymmetric processing amplifies the thermal signals from defects while suppressing background thermal noise, achieving improved hotspot enhancement while maintaining the simplicity of the overall measurement process.
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 proposed method achieves improved contrast in topographical thermograms and reduces noise sensitivity, allowing for more accurate localization of defects and reduced thermal budget in failure analysis processes.
Implementation Method 1
When a defect is present in a semiconductor device and the semiconductor device is operational (i.e. it is turned on), this will typically lead to an increased temperature around the defect in comparison with a device which does not have this defect. This is illustrated in FIG. 1 which shows the defect at the center of the semiconductor device and the thermal radiation resulting from this defect.
Implementation Method 2
switching an irradiating device, for irradiating the semiconductor device, on and off at a second frequency
Data Source
AI summary
Examples include a method for localizing one or more defects in a semiconductor device. The method includes switching the semiconductor device on and off at a first frequency and switching an irradiating device on and off at a second frequency. The method also includes acquiring images of the semiconductor device and lock-in averaging the images with a first reference signal having the first frequency and a first phase, to obtain amplitudes indicative of temperatures at a surface of the semiconductor device and/or phase signals indicative of a depth location of the one or more defects in the semiconductor device. The method also includes lock-in averaging the images with a second reference signal having the second frequency and a second phase to obtain a topography of the surface of the semiconductor device. The first frequency is different from the second frequency and/or the first phase is different from the second phase.