Dark-Field Inspection Apparatus Haze Signal Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Dark-field inspection apparatuses face challenges in accurately detecting defects on semiconductor wafers with high haze levels due to superimposed shot noise, leading to incorrect defect sizing and sensitivity issues, which are operator-dependent and time-consuming to resolve.
Innovation Solution
An inspection apparatus and method that utilize the correlation between surface roughness and scattered light to automatically determine optimal sensitivity settings, separating haze and defect signals, and adjusting sensitivity to prevent saturation, thereby improving detection precision and reducing operator influence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dark-field inspection is performed on wafers with high haze levels, then defect detection capability is maintained, but shot noise from haze superimposes on defect signals causing incorrect defect sizing and measurement errors
Solution Approach 1:
The inspection signal is segmented into two distinct components: a low-frequency component representing haze scattered light and a high-frequency component representing defect signals. By applying frequency separation techniques, the system isolates the defect detection capability from the haze interference, allowing accurate defect sizing even on high-haze wafers.
Solution Approach 2:
The system intentionally allows the sensor to operate in a partially saturated state where the DC component (haze signal) approaches but does not exceed the sensor's dynamic range limit. By setting the DC offset removal level appropriately, the system maximizes the utilization of the amplifier's dynamic range for AC defect signals while preventing complete saturation, thereby maintaining measurement precision without sacrificing defect detection capability.
2Reliability
If inspection sensitivity is increased to detect small defects, then detection capability improves, but sensor output signal saturates on wafers with haze
Solution Approach 1:
Before performing defect inspection, the system preliminarily measures the haze level of the wafer surface and calculates the appropriate DC offset removal level based on this measurement. This preliminary action sets the inspection sensitivity and amplifier gain to optimal values that prevent signal saturation while maintaining high detection capability for small defects on high-haze wafers.
Solution Approach 2:
The system dynamically changes the inspection parameters (amplifier gain, DC offset removal level, sensitivity threshold) based on the measured haze level of each wafer. By adjusting these parameters in response to varying haze conditions, the system maintains optimal detection capability without causing signal saturation across different wafer types.
3Measurement precision
If inspection conditions are manually optimized for each new process step, then detection accuracy can be adapted, but the process requires significant time and operator expertise
Solution Approach 1:
The inspection system performs self-optimization by automatically measuring the haze level of each wafer, calculating the optimal inspection conditions (sensitivity, gain, threshold), and configuring these parameters without operator intervention. This self-service capability eliminates the time-consuming manual optimization process while maintaining high detection accuracy adapted to each specific wafer's surface conditions.
Solution Approach 2:
The system implements a feedback mechanism where the measured haze level from preliminary inspection feeds into the automatic determination of optimal inspection conditions. This closed-loop feedback process enables the system to adaptively optimize detection accuracy for each new process step or wafer type without requiring manual operator input or extensive setup time.
4Measurement precision
If DC offset removal is applied to expand dynamic range, then small defect detection improves, but incorrect DC offset removal level causes loss of signal information
Solution Approach 1:
The system performs preliminary haze measurement and uses this information to calculate the correct DC offset removal level before defect detection. By determining the appropriate offset removal amount in advance based on actual wafer conditions, the system expands the dynamic range for small defect detection while preserving the integrity of the defect signal information.
Solution Approach 2:
The system dynamically adjusts the DC offset removal level parameter based on the measured haze characteristics of each wafer. By changing this parameter to match the specific wafer conditions rather than using a fixed value, the system maximizes small defect detection capability while preventing information loss that would result from incorrect offset removal.
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 solution enables precise detection of defects on wafers with high haze levels by optimizing sensitivity based on haze data, ensuring accurate defect sizing and efficient inspection conditions without relying on operator expertise or lengthy trial-and-error processes.
Implementation Method 1
the illumination with laser light causes light scattered by the haze
Implementation Method 2
a dark-field optical system which detects scattered light generated by the illumination light with which the specimen is illuminated
Implementation Method 3
a photoelectric converter which converts the scattered light detected by the dark-field optical system, into an electric signal
Data Source
AI summary
This application relates to an inspection apparatus including: a stage which holds a specimen; an illumination optical system which illuminates a surface of the specimen held on the stage, with illumination light; a dark-field optical system which detects scattered light generated by the illumination light with which the specimen is illuminated; a photoelectric converter which converts the scattered light detected by the dark-field optical system, into an electric signal; an A/D converter which converts the electric signal obtained by conversion by the photoelectric converter, into a digital signal; a judgement unit which determines the dimension of a foreign substance on the surface of the specimen on the basis of a magnitude of the scattered light from the foreign substance; and a signal processor which determines an inspection condition by use of information on the scattered light from the specimen surface.