Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12422378-B2
Publication Date
2025-09-23
Expiration Date
Abstract
A method includes identifying property data of a substrate processed by a substrate processing system. The method further includes identifying, based on a first subset of the property data, a plurality of regions of the substrate corresponding to a first defect category. The method further includes sub-categorizing, based on a second subset of the property data, the plurality of regions of the substrate corresponding to the first defect category into a plurality of defect sub-categories. The method further includes causing, based on one or more of the plurality of regions corresponding to at least one of the plurality of defect sub-categories, performance of a corrective action associated with the substrate processing system.
Core Innovation
A processing device performs an end-to-end substrate defect analysis workflow by identifying property data of a substrate processed by a substrate processing system. Based on a first subset of the property data, the processing device identifies a plurality of regions of the substrate corresponding to a first defect category and, based on a second subset of the property data, sub-categorizes the plurality of regions into a plurality of defect sub-categories.
The processing device determines a defect source associated with one or more of the plurality of regions corresponding to the plurality of defect sub-categories. Responsive to determining the defect source, the processing device causes performance of a corrective action associated with the substrate processing system to reduce substrate defects.
The described approach can use a trained machine learning model to support identifying the plurality of regions corresponding to the first defect category and to support sub-categorizing the plurality of regions into defect sub-categories. Predictive data output from the trained machine learning model is used as a basis for the region identification and the sub-categorization, and defect sub-categories may be filtered by a threshold level for corresponding substrate property data and selected based on user input and historical defect sub-categories and target historical property data.
Claims Coverage
The independent claims cover a method, a non-transitory computer-readable storage medium, and a system configured to perform the same end-to-end workflow. The inventive features focus on property-data-based identification of defect-category regions, sub-categorization into defect sub-categories, defect-source determination, and triggering corrective action to reduce substrate defects, with optional trained machine learning predictive data support.
Property-data-based defect-category region identification
identifying property data of a substrate processed by a substrate processing system; identifying, based on a first subset of the property data, a plurality of regions of the substrate corresponding to a first defect category.
Property-data-based defect sub-categorization
sub-categorizing, based on a second subset of the property data, the plurality of regions of the substrate corresponding to the first defect category into a plurality of defect sub-categories.
Defect-source determination tied to defect sub-categories
determining a defect source associated with one or more of the plurality of regions corresponding to the plurality of defect sub-categories.
Corrective action to reduce substrate defects
responsive to the determining of the defect source, causing performance of a corrective action associated with the substrate processing system to reduce substrate defects.
Computer-readable storage medium for end-to-end defect workflow
instructions which, when executed, cause a processing device to perform identifying property data, identifying regions corresponding to a first defect category, sub-categorizing into defect sub-categories, determining a defect source, and causing performance of a corrective action.
System for property-data defect analysis and corrective action
a processing device coupled to memory to identify property data, identify regions corresponding to a first defect category, sub-categorize into defect sub-categories, determine a defect source, and cause performance of a corrective action.
The independent claims require identifying substrate property data, using a first subset to identify regions corresponding to a first defect category, using a second subset to sub-categorize those regions into defect sub-categories, determining a defect source associated with one or more regions, and causing corrective action to reduce substrate defects. Dependent refinements add use of predictive data from a trained machine learning model and selection or filtering of defect sub-categories using threshold level, user input, and historical-data-based predictive modeling.
Stated Advantages
Reduce substrate defects by performing a corrective action associated with the substrate processing system responsive to determining a defect source.
Documented Applications
Substrate defect analysis and defect source tracing/root-cause identification to trigger corrective action in a substrate processing system, including computational process control (CPC), statistical process control (SPC), and advanced process control (APC) contexts.
Use of metrology/sensor property data, including SEM images and EDX images and associated morphology/elemental/spatial metadata, to identify substrate regions corresponding to defect categories and to sub-categorize them for defect source determination and corrective action.
Interested in licensing this patent?