Substrate Profile Mapping for Process Chamber Wear Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing technologies for detecting conditions of chamber components in processing chambers are inadequate, leading to degraded product quality and scrapped substrates due to wear and damage of chamber components over time.
Innovation Solution
A method and system that utilize a substrate measurement system to generate profile maps of substrates processed in a process chamber, and then process these maps using a trained machine learning model to estimate the conditions of chamber components such as the mesa condition, lift pin location, seal band condition, and process kit ring condition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If chamber components are used over time for substrate processing, then productivity is improved through continuous operation, but the chamber components wear out and become damaged leading to degraded product quality and scrapped substrates
Solution Approach 1:
The system performs preliminary detection of chamber component conditions by analyzing substrate profile maps before critical failures occur. The machine learning model estimates component wear and damage early in the usage cycle, enabling proactive maintenance scheduling that prevents both productivity loss from unplanned downtime and quality degradation from failed components.
Solution Approach 2:
The system establishes a feedback loop where substrate measurement data is continuously fed into the machine learning model, which then provides real-time estimates of chamber component conditions. This feedback enables dynamic adjustment of maintenance schedules and process parameters, resolving the contradiction by allowing continuous operation while maintaining component reliability through informed monitoring and timely intervention.
2Measurement precision
If traditional detection methods are used for chamber component conditions, then device complexity is minimized, but measurement precision of component conditions is insufficient leading to undetected wear and damage
Solution Approach 1:
The system introduces an intermediary approach by using substrate profile maps as indirect indicators of chamber component conditions. Rather than directly measuring component wear with complex sensors, the machine learning model infers component status from substrate measurement data, achieving high measurement precision while avoiding the complexity of direct component monitoring systems.
Solution Approach 2:
The invention replaces potential mechanical or direct physical measurement systems with a computational approach. The machine learning model substitutes complex hardware-based detection with software-based analysis of substrate profiles, achieving precise component condition assessment without increasing physical device complexity.
3Loss of time
If chamber components are not monitored, then device complexity is reduced, but loss of time occurs due to unplanned downtime and scrapped substrates
Solution Approach 1:
The system implements self-service monitoring where the substrate measurement system automatically generates profile maps that are processed by the machine learning model to assess chamber component conditions. This automated self-diagnosis capability reduces time loss from unplanned downtime and scrapping by enabling early detection of component issues without adding complex external monitoring infrastructure.
Data Source
AI summary
A method includes processing a substrate in a process chamber according to a recipe, wherein the substrate comprises at least one of a film or a feature after the processing. The method further includes generating a profile map of the first substrate. The method further includes processing data from the profile map using a first model, wherein the first model outputs at least one of an estimated mesa condition of a substrate support for the process chamber, an estimated lift pin location condition of the substrate support an estimated seal band condition of the substrate support, or an estimated process kit ring condition for a process kit ring for the process chamber. The method further includes outputting a notice as a result of the processing.