Semiconductor Defect Rate Estimation Using Virtual Distributions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for forming recombination centers in semiconductor devices using charged particles to adjust carrier lifetime face challenges in accurately estimating defect rates due to variations in carbon concentration and irradiation amounts, making it difficult to determine non-defective products.
Innovation Solution
An analysis method that involves acquiring measurement values of specific characteristics, generating measurement distributions, and simulating virtual distributions to calculate defect rates accurately by considering covariance and variance, allowing for precise estimation of semiconductor device quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If charged particle irradiation is used to form recombination centers and adjust carrier lifetime, then semiconductor device characteristics can be controlled, but defect rates increase due to variations in carbon concentration and irradiation amounts
Solution Approach 1:
The patent implements a feedback mechanism by acquiring measurement values of semiconductor device characteristics, comparing them against target values, and adjusting irradiation conditions based on the deviations. This closed-loop control system continuously refines the irradiation process to compensate for variations in carbon concentration and other parameters, thereby maintaining manufacturing precision while reducing defect rates.
Solution Approach 2:
The patent systematically varies irradiation parameters (irradiation amount, energy levels) and analyzes their effects on device characteristics. By establishing relationships between irradiation parameters and outcome characteristics through experimental data, the system optimizes parameter combinations to achieve desired carrier lifetime control while minimizing defect formation associated with carbon concentration variations.
2Productivity
If measurement data from limited samples is used to estimate defect rates, then analysis can be performed quickly, but estimation accuracy decreases
Solution Approach 1:
The patent creates virtual copies of measurement data through simulation techniques. By generating synthetic datasets that replicate the statistical properties and correlations of actual measurements, the system expands limited sample data into larger virtual datasets. This allows accurate defect rate estimation without requiring extensive physical measurements, thus maintaining both productivity and measurement precision.
Solution Approach 2:
The patent performs preliminary analysis to identify and extract key correlation patterns from available measurement data before conducting full defect rate estimation. By pre-processing data to establish relationships between variables (such as carbon concentration and device characteristics), the system creates a framework that enables accurate rapid estimation without needing large sample sizes, balancing speed and accuracy.
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
The method enables accurate estimation of defect rates in semiconductor devices by accounting for variations in carbon concentration and irradiation amounts, improving the determination of non-defective products and enhancing manufacturing efficiency.
Implementation Method 1
a technique of forming a recombination center of a carrier by irradiating a semiconductor substrate having formed a semiconductor device such as an Insulated Gate Bipolar Transistor (IGBT) with a charged particle such as a helium ion
Data Source
AI summary
Provided is an analysis method including acquiring measurement values of a characteristic of a plurality of semiconductor devices of a measurement group in which a concentration of a first impurity and an irradiation amount of a charged particle beam are included in a set range, generating a measurement distribution showing a distribution, in the measurement group, of the measurement values of the characteristic, generating a virtual distribution in which samples of the characteristic are distributed in a range that is wider than the measurement distribution by simulating, based on the measurement distribution, the characteristic of a plurality of the semiconductor devices that is virtual, and calculating a defect rate in the virtual distribution.