Neural Network Photomask Generation for Accurate Pattern Transfer

Solution Overview

Problem

Existing photolithography techniques face inefficiencies and inaccuracies in compensating for optical effects during mask generation, leading to suboptimal pattern transfer on substrates, and current tools are time-consuming and resource-intensive.

Innovation Solution

Employing a neural network to process and generate lithographic masks by compressing and decompressing image representations formed of two-point elements, utilizing a trained neural network to accurately produce masks that account for optical distortions, thereby improving efficiency and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If traditional photomask generation tools are used to compensate for optical effects, then pattern transfer accuracy is improved, but processing time and resource consumption increase significantly

Engineering Contradiction:

Improvepattern transfer accuracyVSAvoidprocessing time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent replaces traditional mechanical/optical simulation-based photomask generation tools with a neural network-based system. The neural network is trained to predict optimal photomask patterns directly from circuit designs, substituting the computationally intensive optical simulation process with a faster machine learning inference process while maintaining pattern transfer accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The neural network is trained in advance on a large dataset of circuit patterns and their corresponding optimal photomask patterns. This preliminary training phase allows the network to learn the complex relationships between circuit designs and optical effects, enabling fast and accurate photomask generation during actual manufacturing without requiring time-consuming simulations at that stage.

Inventive Principle:

Principle #10Preliminary action

2Manufacturing precision

If traditional photomask generation tools are used to compensate for optical effects, then pattern transfer accuracy is improved, but processing resource consumption increases significantly

Engineering Contradiction:

Improvepattern transfer accuracyVSAvoidprocessing resource consumption

Core Design Contradiction:

Manufacturing precisionVSUse of energy by moving object

Solution Approach 1:

The patent replaces traditional mechanical/optical simulation-based photomask generation tools with a neural network-based system. The neural network is trained to predict optimal photomask patterns directly from circuit designs, substituting the computationally intensive optical simulation process with a faster machine learning inference process while maintaining pattern transfer accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Ease of manufacture

If existing lithographic mask design techniques are used, then pattern generation is achieved, but the process is cumbersome and expensive in terms of time and processing resources

Engineering Contradiction:

Improvemask generation processVSAvoidgeneration efficiency

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The patent replaces traditional mechanical/optical simulation-based photomask generation tools with a neural network-based system. The neural network is trained to predict optimal photomask patterns directly from circuit designs, substituting the computationally intensive optical simulation process with a faster machine learning inference process while maintaining pattern transfer accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The neural network system is designed to autonomously generate optimized photomask patterns from circuit designs without requiring manual intervention or complex iterative simulations. The system self-adjusts to account for optical effects based on the patterns it has learned during training, making the mask generation process more autonomous and efficient.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS12455502B2Device and method for generating photomasks

Publication Date: 2025.10.28 STMICROELECTRONICS (CROLLES 2) SAS

AI summary

The present description concerns a method that includes the compression, by a processor, of an image comprising first patterns by transforming the image into a first representation formed of two-point elements. The method also includes the execution, by a neural network, of an inference operation on the first representation to generate a second representation formed of two-point elements. The method further includes the generation of a lithographic mask based on the decompression of the second representation.