Substitute Sample for Semiconductor Process Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for optimizing semiconductor processing conditions require a large amount of expensive actual product samples and time-consuming tests, leading to high costs and inefficiencies due to the shortage of skilled engineers and the expense of using actual product samples for process development.
Innovation Solution
A substitute sample system is introduced, featuring a first surface and a second surface with inflow parts allowing particles to flow between them, simulating the processing state of semiconductor samples, which can be measured by a measurement device to optimize processing conditions at reduced cost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large amount of learning data is obtained by performing a large number of processing tests using actual product samples, then the accuracy of optimizing processing conditions improves, but the cost for process development increases and the time required increases
Solution Approach 1:
The patent creates a substitute sample that copies the essential processing characteristics of the actual semiconductor product sample. The substitute sample includes a first surface, a second surface, and a plurality of inflow parts that allow particles to flow between surfaces, simulating the processing state. By using this copy instead of the original expensive actual product samples, the system can perform numerous processing tests without consuming valuable production samples, thereby obtaining large amounts of learning data efficiently.
Solution Approach 2:
The substitute sample serves as a disposable or reusable inexpensive object that can be used for process development testing without the cost constraints of actual product samples. The measurement system measures the substitute sample to obtain processing test results, which can then be used for machine learning optimization. This approach allows extensive testing with cheap substitutes rather than expensive actual products.
2Measurement precision
If a large amount of learning data is obtained by performing a large number of processing tests using actual product samples, then the accuracy of optimizing processing conditions improves, but the cost for process development increases
Solution Approach 1:
The patent creates a substitute sample that copies the essential processing characteristics of the actual semiconductor product sample. The substitute sample includes a first surface, a second surface, and a plurality of inflow parts that allow particles to flow between surfaces, simulating the processing state. By using this copy instead of the original expensive actual product samples, the system can perform numerous processing tests without consuming valuable production samples, thereby obtaining large amounts of learning data efficiently.
Solution Approach 2:
The substitute sample serves as a disposable or reusable inexpensive object that can be used for process development testing without the cost constraints of actual product samples. The measurement system measures the substitute sample to obtain processing test results, which can then be used for machine learning optimization. This approach allows extensive testing with cheap substitutes rather than expensive actual products.
3Productivity
If processing conditions are optimized using machine learning techniques, then the productivity of process development improves, but the complexity of the measurement system increases
Solution Approach 1:
The measurement system acts as an intermediary between the substitute sample and the machine learning algorithm. It measures specific parameters of the substitute sample after particle processing and provides this data to the machine learning system for optimization. This intermediary measurement approach enables automated data collection for machine learning without requiring overly complex measurement equipment, maintaining a balance between productivity improvement and system complexity.
Data Source
AI summary
A sample simulates a processing state of a semiconductor sample and is measured by a measurement device. The sample includes: a first surface formed at a first height when viewed from a sample surface; a second surface formed at a second height higher than the first height; and a plurality of inflow parts which allow a particle for performing processing on the first surface to flow between the first surface and the second surface. The processing by the particle flowing from the inflow parts is superimposed in at least a part of a region to be processed on the first surface, and the region where the processing is superimposed on the first surface is measured by the measurement device.