Large-Spot Spectral Sensing for Wafer Uniformity Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced etch and deposition processes in semiconductor manufacturing result in non-uniform features across wafers due to variations in critical dimensions and etch depths, requiring time-consuming metrology and process adjustments to compensate for these variations.
Innovation Solution
A large beam spot metrology tool integrated with a process chamber or wafer handling tool for rapid sampling of wafer surfaces, combined with a control system for scanning and multiplexed capture of metrology samples, and a computational system for denoising optical metrology data using neural networks to reduce noise and predict process settings for uniformity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional metrology tools are used to identify non-uniformities on wafer surfaces, then measurement precision is improved, but productivity deteriorates due to time-consuming sequential sampling
Solution Approach 1:
The wafer surface is divided into multiple discrete locations that are sampled independently. The metrology tool captures data from numerous locations across the wafer surface, allowing comprehensive uniformity assessment while maintaining rapid acquisition at each point, thus resolving the contradiction between thorough measurement and processing speed.
Solution Approach 2:
The metrology tool employs periodic pulsing of the light beam at high rates (30-300 flashes/second) to rapidly acquire spectral data from multiple wafer locations. This periodic action enables fast sequential sampling across the wafer surface, improving productivity while maintaining measurement precision through repeated rapid measurements.
2Measurement precision
If comprehensive metrology sampling is performed over large wafer areas, then measurement precision is improved, but loss of time increases due to extensive scanning requirements
Solution Approach 1:
The system performs preliminary rapid scanning to identify regions of interest or non-uniformities before conducting more detailed measurements. This preliminary action allows the system to focus subsequent measurement efforts on critical areas, reducing total measurement time while maintaining comprehensive spatial characterization precision.
Solution Approach 2:
The system captures a large number of metrology samples (300-100,000) across the wafer surface, using excessive sampling density in certain regions to ensure comprehensive coverage. This partial or excessive action guarantees precise spatial distribution characterization while the rapid acquisition rate minimizes the time penalty.
3Productivity
If rapid beam spot scanning is used to increase sampling rate, then productivity is improved, but measurement precision may deteriorate due to reduced integration time
Solution Approach 1:
The light beam is pulsed periodically at high rates (30-300 flashes/second) with each pulse providing sufficient integration time for precise spectral measurement. This periodic action maintains measurement precision by ensuring adequate signal collection during each pulse while achieving high overall sampling rates through rapid repetition.
Solution Approach 2:
The metrology system maintains continuous scanning and measurement action across the wafer surface without interruption. The beam spot continuously moves from location to location, capturing spectral data in an unbroken sequence, which maintains productivity while the continuous nature of the action ensures consistent signal quality through uninterrupted integration at each point.
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
Enables rapid and accurate assessment of wafer uniformity, reducing the time and resources needed for process adjustments by providing real-time feedback for achieving target spatial distributions of wafer structure parameters.
Implementation Method 1
a large beam spot metrology tool configured to (i) direct an incident light beam onto a wafer with a beam spot of at least about 1 mm on the wafer's surface, and (ii) detect a metrology signal from the wafer in response to illumination with the incident light beam; The metrology samples may comprise reflectance spectra
Data Source
AI summary
A large beam spot spectral reflectometer system for measuring a substrate is provided. Hardware components for collecting in situ large beam spot optical signals is disclosed. Machine learning models for denoising large beam spot optical signals are disclosed. Machine learning models for interpreting in situ optical data and facilitating process control are also disclosed.