# US 10,872,403 B2 · System for predicting properties of structures, imager system, and related methods

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 10,872,403 B2
- Document ID: US10872403B2
- Status: Granted
- Kind: B2
- Application number: 16/100,729
- Filed: 2018-08-10
- Published: 2020-02-13
- Granted: 2020-12-22
- Earliest claimed: 2018-08-10
- Assignees: Micron Technology, Inc.
- Applicants: Micron Technology, Inc.
- Inventors: Amitava Majumdar; Qianlan Liu; Pradeep Ramachandran; Shawn D. Lyonsmith; Steve K. McCandless; Ted L. Taylor; Ahmed N. Noemaun; Gordon A. Haller
- Web page: https://patentgrep.com/patent/US10872403B2

## CPC

- G06T 7/0004: {Industrial image inspection}
- G01B 1/00: Measuring instruments characterised by the selection of material therefor
- G01B 11/00: Measuring arrangements characterised by the use of optical techniques
- G01N 21/211: {Ellipsometry (optical thickness measurement G01B11/06)}
- G01N 21/4738: {Diffuse reflection (precedence is given to G01N21/55 - G01N21/57 if specular component is taken into consideration), e.g. also for testing fluids, fibrous materials}
- G03F 7/705: {Modelling or simulating from physical phenomena up to complete wafer processes or whole workflow in wafer productions}
- G03F 7/70616: {Monitoring the printed patterns}
- G03F 7/70625: {Dimensions, e.g. line width, critical dimension [CD], profile, sidewall angle or edge roughness}
- G06F 30/20: Design optimisation, verification or simulation (optimisation, verification or simulation of circuit designs G06F30/30)
- G06F 30/367: Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
- G06N 20/00: Machine learning
- H10P 74/203: {Structural properties, e.g. testing or measuring thicknesses, line widths, warpage, bond strengths or physical defects}
- H10P 74/23: {characterised by multiple measurements, corrections, marking or sorting processes}
- G01B 2210/56: Measuring geometric parameters of semiconductor structures, e.g. profile, critical dimensions or trench depth
- G06F 2111/10: Numerical modelling
- G06T 2207/10024: Color image
- G06T 2207/10152: Varying illumination
- G06T 2207/20081: Training; Learning
- G06T 2207/30148: Semiconductor; IC; Wafer
- H10P 74/207: {Electrical properties, e.g. testing or measuring of resistance, deep levels or capacitance-voltage characteristics}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G01B: MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q)
- G03F: PHOTOMECHANICAL PRODUCTION OF TEXTURED OR PATTERNED SURFACES, e.g. FOR PRINTING, FOR PROCESSING OF SEMICONDUCTOR DEVICES; MATERIALS THEREFOR; ORIGINALS THEREFOR; APPARATUS SPECIALLY ADAPTED THEREFOR; (phototypographic composing devices B41B; photosensitive materials or processes for photographic purposes G03C; electrophotography, sensitive layers or processes therefor G03G)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H10P: GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10

## Abstract

A method of predicting virtual metrology data for a wafer lot that includes receiving first image data from an imager system, the first image data relating to at least one first wafer lot, receiving measured metrology data from metrology equipment relating to the at least one first wafer lot, applying one or more machine learning techniques to the first image data and the measured metrology data to generate at least one predictive model for predicting at least one of virtual metrology data or virtual cell metrics data of wafer lots, and utilizing the at least one generated predictive model to generate at least one of first virtual metrology data or first virtual cell metrics data for the first wafer lot.

## Claims

### Claim 1

A method, comprising:
receiving first image data from an imager system, the first image data relating to at least one first wafer lot;
receiving measured metrology data from metrology equipment relating to the at least one first wafer lot;
applying one or more machine learning techniques to the first image data and the measured metrology data to generate at least one predictive model for predicting at least one of virtual metrology data or virtual cell metrics data of wafer lots;
utilizing the at least one generated predictive model to generate at least one of first virtual metrology data or first virtual cell metrics data for the at least one first wafer lot;
receiving second image data from the imager system, the second image data relating to at least one second wafer lot; and
generating, via the at least one generated predictive model, at least one of second virtual metrology data or second virtual cell metrics data for the second wafer lot based on the second image data.

### Claim 2 (depends on 1)

The method of claim 1, further comprising not receiving measured metrology data for the second wafer lot.

### Claim 3 (depends on 1)

The method of claim 1, wherein generating, via the predictive model, the at least one of second virtual metrology data or the second virtual cell metrics data for the second wafer lot comprises generating at least one of the second virtual metrology data or the second virtual cell metrics data without receiving measured metrology data related to the at least one second wafer lot.

### Claim 4 (depends on 3)

The method of claim 3, further comprising:
receiving measured metrology data related to the at least one second wafer lot; and
validating the at least one generated predictive model with the measured metrology data related to the at least one second wafer lot.

### Claim 5 (depends on 1)

The method of claim 1, further comprising filtering the measured metrology data for the at least one first wafer lot into training metrology data and validation metrology data.

### Claim 6 (depends on 1)

The method of claim 1, further comprising filtering the first image data related to the at least one first wafer lot into training image data, validation image data, and metro-less image data.

### Claim 7

A method, comprising:
receiving first image data from an imager system, the first image data relating to at least one first wafer lot;
receiving measured metrology data from metrology equipment of the at least one first wafer lot;
training a machine learning model with the first image data and the measured metrology data;
applying one or more machine learning techniques to the first image data and the measured metrology data to generate at least one predictive model for predicting at least one of virtual metrology data or virtual cell metrics data of wafer lots;
receiving second image data from the imager system, the second image data relating to at least one second wafer lot; and
generating, via the at least one generated predictive model, at least one of second virtual metrology data or second virtual cell metrics data for the second wafer lot based on the second image data.

### Claim 8 (depends on 7)

The method of claim 7, wherein the one or more machine learning techniques are selected from the list consisting of a quadratic regression analysis, a logistic regression analysis, a support vector machine, a Gaussian process regression, an ensemble model, any other regression analysis, decision tree learning, regression trees, boosted trees, gradient boosted trees, multilayer perceptron, one-vs.-rest, Naïve Bayes, k-nearest neighbor, association rule learning, a neural network, deep learning, and pattern recognition.

### Claim 9 (depends on 7)

The method of claim 7, wherein the second virtual metrology data comprises data related to thicknesses and critical dimensions of features of wafers within the at least one second wafer lot.

### Claim 10 (depends on 7)

The method of claim 7, wherein the second virtual cell metrics data comprises data related to median threshold voltage of a given die, variation in threshold voltage within a given die, operating voltage window of a given die, endurance to read and/or write cycles of a given die, lifetime of a given die, persistence of memory within a given die, or binning of the given die to various product grades.

### Claim 11 (depends on 7)

The method of claim 7, further comprising:
receiving third image data from the imager system, the third image data relating to at least one third wafer lot;
receiving measured metrology data from the metrology equipment of the at least one third wafer lot; and
validating the at least one generated predictive model with the third image data and the measured metrology data related to the at least one third wafer lot.

### Claim 12 (depends on 7)

The method of claim 7, wherein the first image data comprises red, green, and blue intensities of wafers within the at least one first wafer lot.

### Claim 13 (depends on 7)

The method of claim 7, wherein the metrology equipment includes at least one of a scatterometer or an ellipsometer.

### Claim 14 (depends on 7)

The method of claim 7, wherein the first image data includes images of every wafer within the at least one first wafer lot and the measured metrology data of the at least one first wafer lot includes metrology measurements for only a portion of wafers of the at least one first wafer lot.

### Claim 15

A method, comprising:
receiving first image data from an imager system, the first image data relating to at least one first wafer lot; and
without receiving any measured metrology data related to the at least one first wafer lot,
utilizing at least one generated predictive model for predicting at least one of virtual metrology data or virtual cell metrics data of wafer lots to generate at least one of first virtual metrology data or first virtual cell metrics data for the at least one first wafer lot based on the first image data;
receiving second image data from the imager system, the second image data relating to at least one second wafer lot; and
utilizing the at least one generated predictive model to generate second virtual metrology data for the at least one second wafer lot based on the second image data.

### Claim 16 (depends on 15)

The method of claim 15, further comprising:
receiving measured metrology data related to the at least one second wafer lot; and
validating the at least one generated predictive model with the measured metrology data related to the at least one second wafer lot.

### Claim 17 (depends on 16)

The method of claim 16, wherein validating the at least one generated predictive model comprises determining a goodness of fit of the generated second virtual metrology data and the measured metrology data related to the at least one second wafer lot.

### Claim 18 (depends on 17)

The method of claim 17, further comprising triggering a retraining of the at least one generated predictive model if a coefficient of determination of the goodness of fit is below a threshold.

### Claim 19 (depends on 17)

The method of claim 17, further comprising triggering a retraining of the at least one generated predictive model if a coefficient of determination of the goodness of fit is below 0.60.

### Claim 20 (depends on 15)

The method of claim 15, further comprising applying one or more machine learning techniques to image data and measured metrology data related to another wafer lot to generate the at least one generated predictive model.