Virtual Metrology Feedback for Plasma Recipe Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current plasma process recipes in semiconductor fabrication are often developed by trial and error, lacking precise control and requiring numerous iterations, as they do not effectively account for chamber conditions and their impact on plasma parameters and surface properties.
Innovation Solution
A method involving the construction of virtual metrology models that predict wafer characteristics based on plasma parameters, combined with control models that relate these parameters to recipe parameters, allowing for iterative optimization until the desired wafer characteristics are achieved, incorporating in-situ sensor data to account for chamber conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If plasma process recipes are developed by trial and error, then the process can be implemented without complex modeling, but the number of iterations required increases significantly and precision is reduced
Solution Approach 1:
The patent applies preliminary action by building virtual metrology models and control models before actual plasma processing. These models predict wafer characteristics and optimize recipe parameters in advance, allowing the system to start with pre-optimized parameters rather than relying on trial-and-error iterations. The VM model predicts outcomes based on plasma parameters, and the control model translates these predictions into optimized recipe settings before processing begins.
Solution Approach 2:
The patent implements feedback by measuring actual wafer characteristics after plasma processing and using these measurements to calibrate and update both the virtual metrology model and control model. This closed-loop feedback system continuously improves prediction accuracy and recipe optimization, enabling the system to adapt to chamber condition variations and maintain high precision without increasing operational complexity.
2Device complexity
If traditional trial and error methods are used, then the system complexity remains low, but the time required for recipe optimization increases
Solution Approach 1:
The patent reduces optimization time by performing preliminary modeling and prediction work before actual processing. The virtual metrology model is built once to establish relationships between plasma parameters and wafer characteristics, and the control model is pre-configured to translate predictions into recipe parameters. This preliminary setup enables rapid iteration and optimization without requiring complex real-time adjustments during processing.
Solution Approach 2:
The patent replaces the mechanical trial-and-error process with computational modeling and prediction systems. Instead of physically iterating through multiple recipe versions and measuring results manually, the system uses virtual metrology models to predict outcomes and control models to optimize parameters computationally, significantly reducing the time required for recipe development while maintaining manageable system complexity.
3Ease of operation
If chamber condition variations are not accounted for, then the process is simpler to control, but the reliability of wafer characteristics decreases
Solution Approach 1:
The patent addresses chamber condition variations through feedback by measuring actual wafer characteristics and using these measurements to recalibrate the virtual metrology and control models. This feedback mechanism allows the system to detect and compensate for drift in chamber conditions, maintaining reliable and consistent wafer characteristics without requiring complex real-time control adjustments during processing.
Solution Approach 2:
The patent handles chamber condition variations by dynamically adjusting model parameters based on measured wafer characteristics. When chamber conditions drift, the system updates the VM and control model parameters to reflect current conditions, enabling the models to maintain accuracy despite environmental changes. This parameter adaptation allows the system to maintain reliability while keeping the control approach relatively simple.
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 reduces the number of trial-and-error iterations, improves precision in plasma process control, and maintains robustness against chamber condition variations, enabling efficient optimization of plasma recipes for semiconductor fabrication processes.
Implementation Method 1
a remote plasma source generates an ionized gas which contains ions, electrons, radicals and neutral particles
Data Source
AI summary
A method of optimizing a recipe for a plasma process includes (a) building a virtual metrology (VM) model that predicts a wafer characteristic resulting from the plasma process based on a plasma parameter and (b) building a control model that describes a relationship between the plasma parameter and a recipe parameter. (c) The wafer characteristic is measured after performing the plasma process according to the recipe. (d) Whether the wafer characteristic is within a predetermined range is determined. (e) The VM model and the control model are calibrated based on the wafer characteristic. (f) The recipe is optimized by updating the plasma parameter based on the wafer characteristic using the VM model and updating the recipe parameter based on the plasma parameter using the control model. (c), (d), (e) and (f) are repeated until the wafer characteristic is within the predetermined range.