Mask Blank Substrate Surface Morphology Control for Defect Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mask blank substrates used in lithography face challenges in detecting critical defects due to the high detection sensitivity of defect inspection apparatuses, which leads to an excessive number of pseudo defects, concealing critical defects and resulting in defective semiconductor device production.
Innovation Solution
A mask blank substrate with specific surface morphology specifications, including power spectral density and root mean square roughness, is developed to reduce the number of detected defects, particularly by managing the spatial frequency of roughness components to inhibit pseudo defect detection and enhance critical defect visibility during highly sensitive defect inspections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If highly sensitive defect inspection apparatuses are used to detect critical defects on mask blank substrates, then detection sensitivity is improved, but the number of pseudo defects increases excessively, concealing critical defects
Solution Approach 1:
The patent applies parameter changes by precisely controlling surface morphology parameters including power spectral density (PSD) at spatial frequency 1×10^-2 μm^-1 to 1 μm^-1 not more than 4×10^6 nm^4, and PSD at spatial frequency not less than 1 μm^-1 not more than 10 nm^4. This quantitative control of surface roughness parameters reduces pseudo defect generation while preserving critical defect detectability
Solution Approach 2:
The patent implements preliminary action by performing surface processing on the mask blank substrate before defect inspection to control surface roughness characteristics. By pre-managing the spatial frequency distribution of surface irregularities, the substrate is prepared in advance to minimize pseudo defect interference during subsequent high-sensitivity inspections
2Manufacturing precision
If surface roughness is reduced to improve optical properties and defect quality, then manufacturing precision is improved, but detection of critical defects becomes more difficult due to increased sensitivity of inspection apparatuses
Solution Approach 1:
The patent resolves this contradiction by changing the parameters of surface roughness characterization from simple RMS values to power spectral density distributions across different spatial frequencies. By controlling PSD at specific spatial frequency ranges, the patent achieves surface smoothness for optical quality while maintaining detectability of critical defects through appropriate spatial frequency management
Solution Approach 2:
The patent transitions from single-parameter surface roughness measurement to multi-dimensional spectral analysis. By evaluating surface morphology in the spatial frequency domain rather than just amplitude domain, the patent can differentiate between benign surface variations and critical defects, resolving the detection difficulty
Data Source
AI summary
An object of the present invention is to provide a mask blank substrate and the like that enables critical defects to be reliably detected as a result of reducing the number of detected defects, including pseudo defects, even when using highly sensitive defect inspection apparatuses that use light of various wavelengths. The present invention relates to a mask blank substrate that is used in lithography, wherein the power spectral density at a spatial frequency of 1×10−2 μm−1 to 1 μm−1, obtained by measuring a 0.14 mm×0.1 mm region on a main surface of the mask blank substrate on the side of which a transfer pattern is formed at 640×480 pixels with a white-light interferometer, is not more than 4×106 nm4, and the power spectral density at a spatial frequency of not less than 1 μm−1, obtained by measuring a 1 μm×1 μm region on the main surface with an atomic force microscope, is not more than 10 nm4.