Virtual Sensor for Chamber Cleaning Endpoint via Foreline Pressure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current endpoint detection methods for processing chamber cleaning in semiconductor manufacturing are inaccurate and costly, often due to sub-optimal metrology conditions, leading to inefficient cleaning times that can result in device defectivity or equipment degradation.
Innovation Solution
A method involving monitoring chamber foreline pressure at multiple time intervals during clean and unclean plasma cleaning processes to determine a clean endpoint by comparing traces, allowing for accurate detection of cleaning completion without external, expensive equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional endpoint detection methods (RGA, OES, NDIR) are used, then endpoint detection capability is provided, but measurement precision deteriorates due to sub-optimal metrology conditions and high cost
Solution Approach 1:
The system uses the chamber's own vacuum pump and existing pressure sensor to detect cleaning endpoints, eliminating the need for external expensive equipment. The vacuum pump serves dual purposes: maintaining vacuum and providing the detection mechanism through foreline pressure monitoring.
Solution Approach 2:
The foreline pressure acts as an intermediary parameter that indirectly indicates the cleaning state of the chamber. Instead of directly measuring cleaning endpoints with complex spectroscopy equipment, the system monitors foreline pressure changes caused by outgassing from residual materials during the cleaning process.
2Reliability
If clean time is extended, then cleaning thoroughness is improved, but productivity deteriorates due to reduced throughput
Solution Approach 1:
The system implements real-time feedback monitoring of foreline pressure during the cleaning process. By continuously tracking pressure changes and comparing them against reference traces, the system can automatically determine when the cleaning endpoint has been reached, preventing both insufficient and excessive cleaning.
Solution Approach 2:
The system performs preliminary characterization by establishing reference foreline pressure traces from clean chamber runs before production cleaning. These reference traces serve as benchmarks for determining when production cleaning has achieved the desired level of cleanliness, enabling accurate endpoint detection.
3Productivity
If clean time is reduced, then productivity is improved, but reliability deteriorates due to insufficient cleaning
Solution Approach 1:
Real-time feedback monitoring allows the system to precisely determine when cleaning objectives are met, preventing premature termination of cleaning while avoiding unnecessary extension. The feedback mechanism compares current foreline pressure traces against reference traces to objectively assess cleaning completion.
Solution Approach 2:
The system replaces subjective or indirect cleaning assessment methods with objective foreline pressure trace analysis. By substituting complex spectroscopy equipment with simple pressure monitoring and computational analysis, the system achieves more reliable and consistent endpoint detection.
4Reliability
If excessive clean time is used, then cleaning thoroughness is improved, but manufacturing precision deteriorates due to equipment degradation
Solution Approach 1:
Reference foreline pressure traces are established during preliminary clean chamber runs, capturing the characteristic pressure evolution during proper cleaning. These references serve as targets for production cleaning, ensuring that the same level of cleanliness is achieved without excessive cleaning time that could cause degradation.
Solution Approach 2:
Continuous feedback monitoring during production cleaning allows real-time comparison against reference traces, enabling precise termination of cleaning when the endpoint is reached. This prevents over-cleaning that could expose equipment to prolonged corrosive environments and cause degradation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides a more accurate and cost-effective method for determining cleaning endpoints, reducing chamber downtime and improving process reproducibility and film quality.
Implementation Method 1
performing a first plasma cleaning process in a clean chamber environment
Implementation Method 2
deposition processes... plasma-enhanced chemical vapor deposition (PECVD) processing chambers
Implementation Method 3
the processing chamber is coupled with a vacuum pump via a vacuum foreline
Data Source
AI summary
Implementations of the present disclosure generally relate to methods for cleaning processing chambers. More specifically, implementations described herein relate to methods for determining processing chamber cleaning endpoints. In some implementations, a “virtual sensor” for detecting a cleaning endpoint is provided. The “virtual sensor” is based on monitoring trends of chamber foreline pressure during cleaning of the chamber, which involves converting solid deposited films on the chamber parts into gaseous byproducts by reaction with etchants like fluorine plasma for example. Validity of the “virtual sensor” has been confirmed by comparing the “virtual sensor” response with infrared-based optical measurements. In another implementation, methods of accounting for foreline pressure differences due to facility design and foreline clogging over time.