Charged Particle Inspection for Patterning Parameter Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining patterning parameters in semiconductor manufacturing, such as Diffraction Based Focus and Astigmatism Based Focus, require special marks on the substrate, consume patternable area, and do not provide in-device information, especially being unsuitable for extreme ultra violet (EUV) applications and unable to determine multiple patterning process parameters simultaneously.
Innovation Solution
A method using a charged particle inspection system, like a scanning electron microscope, to inspect patterned substrates and determine patterning parameters like focus and dose by establishing a relationship with substrate dimensions through training substrates patterned across a process window, employing multivariable regression or machine learning models, without the need for special marks and enabling EUV patterning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Diffraction Based Focus or Astigmatism Based Focus techniques are used to determine patterning parameters, then focus and dose can be measured, but special marks are required on the substrate which consume patternable area
Solution Approach 1:
The invention extracts the measurement function from dedicated special marks and integrates it into the actual IC patterns themselves. By using the product patterns as measurement targets, the need for separate calibration marks is eliminated, thereby preserving patternable area on the substrate.
Solution Approach 2:
The invention makes the IC patterns serve dual purposes: both as the functional circuit elements and as the measurement targets for determining patterning parameters. This multi-functionality eliminates the need for separate measurement marks and maximizes the use of substrate area for productive purposes.
2Measurement precision
If DBF or ABF techniques are used for patterning parameter determination, then focus measurement is enabled, but in-device information is not provided
Solution Approach 1:
The invention extracts measurement information directly from the actual IC device patterns rather than from separate calibration marks. This allows the measurement process to simultaneously provide information about both the patterning parameters and the actual device features, eliminating information loss about the in-device characteristics.
Solution Approach 2:
The actual IC patterns serve as an intermediary that connects the measurement process with the device information. By using the device patterns themselves as the measurement target, the system simultaneously obtains patterning parameter data and device feature data, making the patterns a dual-purpose mediator.
3Measurement precision
If DBF technique is used for focus measurement, then focus parameter can be determined, but it is not suitable for extreme ultra violet (EUV) patterning applications
Solution Approach 1:
The invention changes the measurement approach from optical diffraction-based methods to charged particle beam-based methods. This parameter change in the inspection technology enables compatibility with EUV patterning applications, as charged particle microscopes can effectively inspect EUV-patterened substrates where optical methods fail.
Solution Approach 2:
The invention replaces the optical measurement system (DBF using light diffraction) with a charged particle beam system. This substitution enables the measurement technique to work with EUV-patterened substrates, as charged particles interact differently with the materials and can provide the necessary contrast and resolution for EUV features.
4Measurement precision
If traditional inspection methods are used, then single parameter measurement is possible, but multiple patterning process parameters cannot be determined simultaneously
Solution Approach 1:
The invention merges multiple measurement functions into a single inspection process. By capturing an image containing multiple pattern types with different sensitivities to various patterning parameters, the system simultaneously determines multiple parameters (focus, dose, and others) from a single measurement, greatly improving productivity.
Solution Approach 2:
The inspection system is designed to perform multiple measurement functions simultaneously. The same charged particle microscope image and analysis process extract information about multiple patterning parameters at once, making the measurement process universal and highly efficient compared to traditional single-parameter methods.
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 allows for accurate determination of patterning parameters directly from actual IC features, enhancing the accuracy and efficiency of semiconductor lithography processes by providing in-device information and supporting EUV applications with single-measurement multi-parameter analysis.
Implementation Method 1
high secondary-electron and backscattered-electron signal detection efficiencies
Implementation Method 2
Backscattered electrons have higher emission energy to escape from deeper layers of a sample
Data Source
AI summary
A method of obtaining focus and dose data that requires no special marks and that uses images of in-die features is described. A focus/dose matrix wafer is created. Dimensions such as critical dimension (CD), CD uniformity (CDU), edge placement error (EPE), etc., at in-die locations are measured using a charged particle inspection system having a large field of view. Machine learning or regression methods are used to determine a relationship between focus and dose and the measured data. The same dimensions can then be measured on a production wafer and the relationship can be utilized to determine the focus and dose for the production wafer.