SLM Pixel Rasterization via Global Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional maskless lithography techniques face challenges in achieving accurate rasterization with minimal light loss, particularly for spatial light modulator (SLM) pixels using various modulation principles, as they often result in scaled or dimmed images due to limitations in approximating diffraction fields and constraints imposed by SLM designs.
Innovation Solution
The global optimization rasterization technique determines the states of SLM pixels to match the diffraction orders of an ideal mask, accounting for constraints such as modulation capabilities and pixel design, allowing for precise control of light distribution in the pupil field to reproduce the desired pattern with minimal light loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If conventional rasterization techniques are used to configure SLM pixels, then the desired pattern can be reproduced, but significant light loss occurs due to approximations of diffraction fields and SLM design constraints
Solution Approach 1:
The patent transforms the rasterization problem from a local pixel-state determination into a global optimization problem by changing the parameters being optimized. Instead of determining pixel states directly, the invention optimizes a cost function that measures the difference between the actual and desired aerial images, subject to SLM constraints. This parameter transformation enables finding pixel configurations that minimize light loss while maintaining pattern accuracy.
Solution Approach 2:
The invention implements a feedback mechanism by using aerial image simulation and comparison. The actual aerial image produced by proposed pixel states is simulated and compared against the desired aerial image, with the difference feeding back into the optimization process. This feedback loop allows iterative refinement of pixel states to achieve both accurate pattern reproduction and minimal light loss.
2Manufacturing precision
If image optimization is performed to achieve precise control of desired pattern properties, then manufacturing precision is improved, but computational cost increases significantly due to repeated aerial image evaluations
Solution Approach 1:
The patent applies preliminary action by pre-computing the diffraction field of the ideal mask and storing it as a reference. This pre-computed reference is then used during the optimization process to avoid repeated full aerial image simulations. By preparing the desired pattern characteristics in advance, the invention reduces the computational burden of iterative comparisons while maintaining precision in controlling pattern properties.
3Productivity
If off-grid filter methods are used for fast rasterization, then productivity is improved through real-time execution, but light loss increases due to local approximations and filter constraints
Solution Approach 1:
The invention introduces dynamics by making the rasterization process adaptive rather than static. The global optimization framework allows the pixel states to be dynamically adjusted based on the specific pattern being rasterized and the SLM constraints, rather than applying a fixed filter. This dynamic approach enables the system to find optimal solutions for different patterns, minimizing light loss while maintaining computational efficiency through the structured optimization methodology.
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
This approach significantly reduces light loss and can achieve light gain in certain patterns, enabling the reproduction of ideal or near-ideal patterns that were previously prohibitively expensive or impossible with conventional mask-based lithography, and allows for new applications in maskless lithography.
Implementation Method 1
The SLMs consist of many pixels, with each pixel having an ability to vary its optical properties in a controllable manner. Typically, each pixel is sub-resolution in size and can assume one of many possible states. Some SLM designs utilize the physical principles of light modulation that are not necessarily equivalent to the ones utilized in a design of traditional masks.
Implementation Method 2
determining diffraction orders associated with an ideal mask of a pattern to be printed by the lithography system, and then configuring the states of the SLM pixels to match all the diffraction orders that are relevant in the image formation
Data Source
AI summary
Provided are a method and system for determining states of spatial light modulator (SLM) pixels in a lithography system configured to print a desired pattern. The method includes determining diffraction orders associated with an ideal mask of a pattern to be printed by the lithography system, and then configuring the states of the SLM pixels to match all the diffraction orders that are relevant in the image formation.