Plasma Reactor ML Control for Wafer Process Drift Compensation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Plasma reactors in semiconductor processing face challenges in maintaining consistent processing states due to physical changes and process drift, leading to inefficiencies and increased costs from frequent wet clean operations and lengthy seasoning processes.

Innovation Solution

Implementing a method that uses data streams from sensors and machine learning algorithms to continuously monitor and adjust the tuning knobs of plasma reactors, enabling real-time compensation for deviations from desired processing states and reducing the need for prolonged seasoning operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If wet clean operations are performed frequently to maintain processing consistency, then processing reliability is improved, but productivity deteriorates due to frequent downtime

Engineering Contradiction:

Improveprocessing consistencyVSAvoidwafer processing throughput

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system continuously monitors processing parameters and chamber conditions, using feedback loops to detect drift and automatically adjust process parameters or trigger maintenance operations only when necessary, rather than following a fixed schedule

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system dynamically adjusts process parameters based on real-time chamber conditions and processing state, allowing the reactor to maintain optimal performance by changing parameters adaptively rather than requiring frequent physical cleaning

Inventive Principle:

Principle #35Parameter changes

2Reliability

If lengthy seasoning operations are performed after wet clean to restore chamber conditions, then processing reliability is improved, but loss of time increases significantly

Engineering Contradiction:

Improvechamber condition stabilityVSAvoidseasoning time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary adjustments and conditioning actions during or immediately after wet clean operations, preparing the chamber in advance to reduce or eliminate the need for extended seasoning periods

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces the traditional mechanical/time-based seasoning process with an intelligent control system that uses sensors, data analysis, and automated parameter adjustment to rapidly restore chamber conditions without requiring prolonged idle time

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If traditional seasoning processes are used to establish steady state conditions, then processing reliability is improved, but use of energy increases due to prolonged operation

Engineering Contradiction:

Improvesteady state stabilityVSAvoidenergy consumption during seasoning

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system uses real-time monitoring and feedback control to detect when steady state conditions are achieved, allowing the chamber to exit seasoning mode as soon as target conditions are met, rather than following a predetermined time schedule

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system transitions from a static, time-based seasoning process to a dynamic, condition-based process that continuously adapts chamber parameters and monitors conditions, enabling rapid achievement of steady state when conditions permit

Inventive Principle:

Principle #15Dynamics

4Manufacturing precision

If complex multivariate processing is implemented with machine learning, then manufacturing precision is improved, but device complexity increases

Engineering Contradiction:

Improveprocessing state control accuracyVSAvoidsystem architecture complexity

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The system introduces an intermediary intelligent control layer that sits between the physical chamber and the control system, using sensors, data processing algorithms, and machine learning models to translate complex multivariate data into actionable control decisions, simplifying the overall system architecture while enabling advanced processing

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentEP3512977B1Method and process of implementing machine learning in complex multivariate wafer processing equipment

Publication Date: 2023.11.08 LAM RES CORP

Figure 1
Figure 2
Figure 3

AI summary

Methods and systems for controlling processing state of a plasma reactor to initiate processing of production substrates. The method initiate processing of a substrate in the plasma reactor using settings for tuning knobs of the plasma reactor that are approximated to achieve desired processing state values. A plurality of data streams are received and are used to identify current processing state values. The method includes generating a compensation vector that identifies differences between the current processing state values and the desired processing state values. The generation of the compensation vector uses machine learning to improve and refile the identification and amount of compensation needed, as identified in the compensation vector. The method further includes transforming the compensation vector into adjustments to the settings for the tuning knobs and then applying the adjustment to the tuning knobs of the plasma reactor.