Dual-phase virtual metrology for semiconductor process control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the semiconductor industry, as device dimensions shrink, traditional run-to-run advanced process control methods become inadequate for ensuring workpiece-to-workpiece control due to increased metrology delays and the need for tighter process control, leading to complications in advanced process capability and performance.
Innovation Solution
A dual-phase virtual metrology method is developed, utilizing historical process data and measurement values to establish conjecture, reference, and statistical distance models, which provide prompt and accurate virtual metrology values through phase-one and phase-two calculations, along with reliance and similarity indexes, to enable workpiece-to-workpiece control without actual measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If real measurements are performed on each workpiece to achieve workpiece-to-workpiece control, then measurement accuracy is improved, but production cycle time increases significantly
Solution Approach 1:
The patent applies preliminary action by performing virtual measurements on all workpieces in advance using process data and statistical models before actual measurements are taken. This allows the system to predict workpiece characteristics and prepare control decisions beforehand, eliminating the need to wait for sequential physical measurements of each workpiece, thereby reducing production cycle time while maintaining measurement accuracy through model-based predictions.
Solution Approach 2:
The patent uses copying by creating virtual copies of measurement data through statistical models and process data analysis. Instead of physically measuring each workpiece, the system generates virtual measurement values that replicate the information obtained from actual measurements. This virtual measurement approach provides sufficient accuracy for workpiece-to-workpiece control without the time penalty of physical measurement, effectively copying the essential measurement information in advance.
2Loss of time
If virtual metrology is used to provide measurement values without actual measurements, then production cycle time is reduced, but measurement accuracy deteriorates
Solution Approach 1:
The patent implements feedback by continuously comparing actual measurement values with virtual measurement values and using the differences to refine and update the statistical models. The system uses process data and actual measurement results to recalibrate the virtual measurement models, ensuring that virtual measurements remain accurate over time. This feedback mechanism maintains measurement precision while preserving the time efficiency of virtual metrology by avoiding repeated physical measurements.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting model parameters and statistical distributions based on process data and actual measurement results. The system modifies the virtual measurement models to reflect changing process conditions, material variations, and equipment drift. This adaptive parameter adjustment ensures that virtual measurements maintain high accuracy despite process variations, resolving the trade-off between speed and precision.
3Device complexity
If traditional run-to-run control is used, then device complexity is reduced, but process control capability deteriorates for workpiece-to-workpiece precision
Solution Approach 1:
The patent applies preliminary action by pre-calculating virtual measurement values and control parameters for all workpieces before production begins. The system uses process data and statistical models to predict workpiece characteristics and determine optimal control settings in advance. This preliminary preparation enables workpiece-to-workpiece precision control without requiring complex real-time measurement and adjustment systems, maintaining relatively simple device complexity while achieving high manufacturing precision.
Data Source
AI summary
A dual-phase virtual metrology method is disclosed for considering both promptness and accuracy by generating dual-phase virtual metrology (VM) values, wherein a Phase-I conjecture step emphasizes promptness by immediately calculating the Phase-I virtual metrology value (VMI) of a workpiece once the entire process data of the workpiece are completely collected; and a Phase-II conjecture step intensifies accuracy, which does not re-calculate the Phase-II virtual metrology values (VMII) of all the workpieces in the cassette until an actual metrology value (required for tuning or re-training purposes) of a selected workpiece in the same cassette is collected. Besides, the accompanying reliance index (RI) and global similarity index (GSI) of each VMI and VMII are also generated.