Multi-Beam Writer Blur Correction for CD Uniformity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Charged-particle multi-beam lithography systems face challenges in achieving uniform critical dimension (CD) and global CD uniformity (GCDU) due to variations in blur and target elevation, leading to unwanted local CD signatures and deviations from desired pattern edges.
Innovation Solution
A method to compute an exposure pattern that accounts for target elevation and blur variations by determining local blur values and applying a convolution kernel to correct for these effects, ensuring uniform blur across the exposure area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the target is exposed using a multi-beam writer with nominal beam direction, then the exposure pattern is written efficiently, but blur variation occurs due to target elevation changes affecting critical dimension uniformity
Solution Approach 1:
The patent applies preliminary action by computing and applying a corrected exposure pattern that anticipates and compensates for elevation-induced blur variations before the actual exposure process. The correction pattern is calculated based on predicted blur variations across the exposure area, and this pre-computed correction is integrated into the exposure data before writing, thereby preventing CD uniformity issues rather than correcting them after exposure.
Solution Approach 2:
The patent implements parameter changes by modifying the exposure pattern parameters (dose distribution, beam positioning) based on calculated blur variations. The system adjusts exposure parameters dynamically across different regions of the target to compensate for elevation changes, transforming a uniform exposure approach into a spatially varying exposure strategy that maintains CD uniformity despite topography variations.
2Area of stationary object
If the target surface has elevation variations, then the exposure area covers complex topography, but blur varies across the exposure area degrading pattern fidelity
Solution Approach 1:
The patent applies local quality by computing spatially varying correction patterns tailored to specific regions of the exposure area. Instead of using a uniform correction approach, the system calculates local blur variations for different regions based on their respective elevations and applies region-specific correction parameters. This allows each local area to be exposed with optimized parameters that compensate for its particular topography, maintaining pattern fidelity across complex surfaces.
3Device complexity
If conventional exposure pattern computation is used, then the writing process is simple, but elevation-dependent blur causes non-uniform critical dimensions
Solution Approach 1:
The patent implements preliminary action by performing blur variation calculations and correction pattern generation before the actual exposure writing process. The system pre-computes the elevation-dependent blur map and derives correction patterns in advance, integrating these corrections into the exposure data structure before writing begins. This approach adds computational steps but maintains a streamlined writing process by preparing all corrections beforehand.
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
Enhances CD uniformity and GCDU by compensating for elevation-induced blur variations, improving pattern fidelity and resolution in multi-beam writing processes.
Implementation Method 1
a particle beam is generated in an illumination system, is directed to and illuminates a pattern definition device comprising an aperture array composed of a plurality of blanking apertures through which the particle beam penetrates and is then imaged by a projection optics system onto a target
Implementation Method 2
calculating a convolution kernel using the local blur value, which kernel describes a mapping from an image element of the graphical representation to a group of pixels... by convolution of the graphical representation with the convolution kernel
Data Source
Figure 1
Figure 2
Figure 3A~3D
AI summary
In order to compensate for undesired effects of varying elevation (h) of a target (141) with respect to a nominal target plane (150), during writing a desired pattern on the target in a charged-particle beam apparatus, the pattern is re-calculated in each of a number of segments of the target plane by: determining an elevation (h) of the target in the segment from the nominal target plane (150); determining a local blur value which represents the actual value of blur corresponding to the elevation (h), with regard to a dependence of the blur upon the elevation of the target; calculating a convolution kernel which represents a point spreading function realizing a local blur value; and re-calculating a nominal exposure pattern by applying the kernel to the pattern. The convolution kernel corresponds to introducing an additional blur into the pattern in the segment, increasing the blur to a given target blur value which is uniform to all segments.