Aerial Image Registration Error Determination via Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current position measuring devices in lithography are unable to achieve the required accuracy for registration error determination, particularly with metrological requirements exceeding 1 nm, due to limitations in existing methods for subpixel-accurate position determination and the influence of distortions in aerial image generation.
Innovation Solution
The method involves capturing a first aerial image and simulating a second aerial image based on pattern specifications of the mask, accounting for distortion effects such as proximity effects, lens distortions, and mask topography, to accurately determine the registration error by comparing the positions of features in both images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional subpixel-accurate position determination methods are used, then measurement capability is provided, but measurement precision is insufficient for requirements below 1 nm
Solution Approach 1:
The patent creates a simulated aerial image that copies the expected pattern appearance based on mask design data, then compares this simulated copy with the actual captured aerial image to determine registration errors. This copying approach allows distortion compensation by referencing the known design geometry.
Solution Approach 2:
The patent introduces a simulation model as an intermediary between the mask design and the actual measurement. This simulation intermediary accounts for optical distortions and proximity effects, serving as a mediator that translates design intent into expected measurement appearances, enabling accurate registration determination despite physical distortions.
2Measurement precision
If distortion effects are not accounted for in aerial image analysis, then measurement process is simpler, but measurement precision deteriorates due to distortion artifacts
Solution Approach 1:
The patent performs preliminary simulation of the aerial image based on mask design data before actual measurement. This preliminary action pre-calculates the expected pattern appearance including all distortion effects, so that when the actual measurement is taken, the comparison can directly yield accurate registration errors without complex real-time distortion correction.
Solution Approach 2:
The patent replaces complex mechanical or optical distortion correction mechanisms with a computational simulation approach. Instead of physically compensating for distortions through hardware adjustments, the system uses software-based aerial image simulation to model and account for all distortion effects mathematically.
3Measurement precision
If high-resolution pattern specifications are used for simulation, then measurement precision improves, but computational requirements increase
Solution Approach 1:
The patent applies simulation only to the specific regions and features relevant to registration measurement, rather than simulating the entire mask at maximum resolution. This partial action approach focuses computational resources on the critical measurement areas, achieving sufficient precision without the full computational burden of complete high-resolution simulation.
Data Source
AI summary
A method for determining a registration error of a feature on a mask, including providing a first aerial image that was captured by means of a position measuring device and includes at least the feature, simulating, from pattern specifications of the mask, a second aerial image that includes at least the feature, taking into account at least one effect that causes distortion of the first aerial image, and determining the registration error of the feature as the distance of the position of the feature in the first aerial image from the position of the feature in the second aerial image. Also provided is a method for simulating an aerial image from pattern specifications of a mask and a position measuring device for carrying out the method.