Deposition Model for Metrology Target Design in Semiconductor Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor manufacturing processes face challenges in accurately predicting and managing deposition profiles, which affects the precision of metrology target designs and subsequent photolithography processes.
Innovation Solution
A method is introduced that utilizes a deposition model to predict deposition profiles, enhance metrology target designs, and determine adjustments for photolithography apparatuses, thereby improving overlay and alignment precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional geometric models are used to predict deposition profiles, then the manufacturing process is simple, but the prediction accuracy is insufficient
Solution Approach 1:
The patent transforms the deposition prediction problem from a geometric model to a parameter-based machine learning model. The system learns optimal parameters (weights and biases) that map process inputs to deposition profiles, achieving higher prediction accuracy without increasing operational complexity. This resolves the contradiction by changing the fundamental approach from geometric reasoning to parameter optimization.
Solution Approach 2:
The patent creates a virtual copy of the deposition process through a machine learning model that replicates the behavior of the physical deposition system. This digital twin allows accurate prediction of deposition profiles without requiring complex physical measurements or interventions, resolving the accuracy-complexity trade-off by using a simplified computational replica.
2Manufacturing precision
If deposition profiles are not accurately predicted, then the manufacturing process is simple, but the metrology target design precision deteriorates
Solution Approach 1:
The patent applies deposition prediction and asymmetry compensation in advance during the metrology target design phase. By predicting the deposition profile before manufacturing and pre-compensating for asymmetries, the system ensures high manufacturing precision without adding complexity to the actual deposition process. The compensation parameters are calculated beforehand and integrated into the target design.
Solution Approach 2:
The patent introduces a machine learning-based deposition model as an intermediary between the manufacturing process and metrology target design. This intermediary layer provides accurate deposition predictions that inform target design decisions, improving manufacturing precision while keeping the physical manufacturing process itself simple and unchanged.
3Measurement precision
If asymmetry in deposition profiles is not compensated, then the process remains simple, but the overlay and alignment precision deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the machine learning model continuously learns from actual deposition data to improve asymmetry prediction and compensation. The system uses measured deposition profiles to update model parameters, creating a closed-loop system that automatically improves overlay and alignment precision without requiring manual intervention or complex adjustment procedures.
Solution Approach 2:
The patent compensates for deposition asymmetries by dynamically adjusting process parameters based on machine learning predictions. The system modifies deposition parameters in real-time to counteract predicted asymmetries, achieving high overlay precision through parameter optimization rather than through complex mechanical adjustments or manual compensation procedures.
Data Source
AI summary
A method for applying a deposition model in a semiconductor manufacturing process. The method includes predicting a deposition profile of a substrate using the deposition model; and using the predicted deposition profile to enhance a metrology target design. The deposition model can be calibrated using experimental cross-section profile information from a layer of a physical substrate. In some embodiments, the deposition model is a machine-learning model, and calibrating the deposition model includes training the machine-learning model. The metrology target design may include an alignment metrology target design or an overlay metrology target design, for example.