Gas Delivery Timing Offsets for Precise Substrate Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing manufacturing systems face challenges in precisely controlling the timing of gas delivery and other processing actions, leading to delays and mismatches that affect substrate properties and processing efficiency.
Innovation Solution
A method that involves identifying a target substrate process operation start time and using a model to predict a preemptive time period for initiating gas delivery actions, thereby ensuring that process gases are delivered within a threshold time window of the start time. This method also includes training a machine learning model to generate time delay data based on gas transfer parameter data, allowing for adjustments to process recipes and equipment constants to optimize timing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If gas delivery actions are initiated at the scheduled substrate process operation start time, then the process recipe is simple and easy to follow, but the process gas arrives late causing delays and mismatches that affect substrate properties
Solution Approach 1:
The system performs preliminary action by calculating and applying a preemptive time period offset before the scheduled substrate process operation start time. This offset compensates for the delay between initiating gas delivery and the actual arrival of process gas at the substrate, ensuring that gas arrives precisely when needed without requiring complex real-time adjustments during the process.
Solution Approach 2:
The system dynamically adjusts the gas delivery timing by incorporating a calculated preemptive time period that varies based on specific process conditions. This dynamic adjustment allows the process recipe to adapt to different scenarios while maintaining precise timing control, balancing simplicity with accuracy.
2Productivity
If the system waits for substrate process operation start time to initiate gas delivery, then the control logic is simple, but dead time increases and processing efficiency decreases
Solution Approach 1:
The system performs preliminary calculation of the preemptive time period before gas delivery is initiated. This pre-calculated offset eliminates dead time by ensuring gas delivery is started at the precise moment needed, rather than waiting for the substrate process operation start time, thereby maximizing productivity without adding operational complexity.
Solution Approach 2:
The system skips the unnecessary waiting period by using the preemptive time period offset to initiate gas delivery earlier than the traditional approach. This rushes through the dead time period by calculating the exact moment gas should be delivered, eliminating delays and improving processing throughput.
3Reliability
If traditional timing control is used without predictive modeling, then the system is simpler to implement, but timing accuracy and reliability of substrate processing deteriorate
Solution Approach 1:
The system implements feedback by using a predictive model that continuously learns from historical data to refine the preemptive time period calculations. This feedback mechanism improves timing accuracy and reliability over time, as the model adapts to specific process conditions and variations, while the underlying system structure remains relatively simple.
Solution Approach 2:
The system performs preliminary predictive modeling to determine the optimal preemptive time period before actual gas delivery occurs. This pre-calculated timing information is then applied consistently, improving reliability without requiring complex real-time control systems during the actual processing operation.
Data Source
AI summary
A method includes identifying a target substrate process operation start time. The start time corresponds to a time of initiation of one or more substrate process actions. The method further includes providing to a model first one or more parameters of a gas transfer system associated with the substrate process operation. The method further includes obtaining first output from the model. The first output includes an indication of a first preemptive time period for initiation of first one or more gas delivery actions. The method further includes updating a process recipe. The process recipe is updated in accordance with the first preemptive time period. Updating the process recipe is to cause the first one or more gas delivery actions to deliver a first process gas to a process chamber within a threshold time window of the substrate process operation start time.