# US 2019/0378260 A1 · Cross layer common-unique analysis for nuisance filtering

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

- Number: US 2019/0378260 A1
- Document ID: US20190378260A1
- Status: Published application
- Kind: A1
- Application number: 16/196,090
- Filed: 2018-11-20
- Published: 2019-12-12
- Earliest claimed: 2018-06-06
- Assignees: KLA-Tencor Corporation
- Applicants: Kla-Tencor Corporation
- Inventors: Bjorn Brauer
- Web page: https://patentgrep.com/patent/US20190378260A1

## CPC

- G06T 7/001: {using an image reference approach}
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T 2207/30148: Semiconductor; IC; Wafer
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

Common events between layers on a semiconductor wafer are filtered. Common events should contain the majority of defects of interest. Only nuisance events that are common between layers on the semiconductor wafer remain, which reduces the nuisance rate. Defects that are common across layers can be filtered based on, for example, defect coordinates, a difference image, or defect attributes.

## Claims

### Claim 1

A system comprising:
a light source that generates a beam of light;
a stage configured to hold a semiconductor wafer in a path of the beam of light;
a detector that receives the beam of light reflected from the semiconductor wafer; and
a processor in electronic communication with the detector, wherein the processor is configured to report a location of a defect of interest that is common across a first layer and a second layer of the semiconductor wafer based on defect locations, difference images, or defect attributes.

### Claim 2 (depends on 1)

The system of claim 1, wherein the processor is configured to:
receive defect locations of the first layer;
receive defect locations of the second layer;
perform common-unique analysis; and
filter all common events thereby excluding a unique location, wherein the location of the defect of interest is based on one of the common events.

### Claim 3 (depends on 2)

The system of claim 2, wherein the processor is further configured to receive defect attributes of the first layer and defect attributes of the second layer, wherein filtering all the common events uses the defect attributes, and wherein the defect attributes of the first layer and the second layer include signal strength, shape, size, location relative to a point, brightness, difference in brightness relative to background, or location on a design.

### Claim 4 (depends on 3)

The system of claim 3, wherein the system further comprises a machine learning module operated by the processor, and wherein the machine learning module is configured to filter all the common events.

### Claim 5 (depends on 2)

The system of claim 2, wherein the processor is configured to provide instructions to the system to perform a hot scan of the semiconductor wafer using the light source and the detector that provides the defect locations of the first layer and the defect locations of the second layer.

### Claim 6 (depends on 1)

The system of claim 1, wherein the processor is configured to:
receive a difference image of the first layer for each defect location on the first layer, wherein the difference image of the first layer is between neighboring dies on the first layer;
receive a difference image of the second layer for each defect location on the second layer, wherein the difference image of the second layer is between neighboring dies on the second layer;
compare the difference image of the first layer and the difference image of the second layer; and
filter defects for which the difference image of the first layer and the difference image of the second layer are similar thereby determining the defect of interest.

### Claim 7 (depends on 6)

The system of claim 6, wherein the processor is configured to provide instructions to the system to perform a hot scan of the semiconductor wafer using the light source and the detector that is used to generate the difference image of the first layer and the difference image of the second layer.

### Claim 8 (depends on 1)

The system of claim 1, wherein the processor is configured to:
receive a test image and a reference image for each defect location on the first layer;
receive a test image and a reference image for each defect location on the second layer;
calculate one or more attributes using a machine learning module operated by the processor, wherein the one or more attributes filter nuisance events and are based on the test image and the reference image for the first layer and the test image and the reference image for the second layer, and wherein the test image and the reference image for the first layer and the test image and the reference image for the second layer are used as channels; and
filter nuisance events with the one or more attributes thereby determining the defect of interest.

### Claim 9 (depends on 8)

The system of claim 8, wherein the machine learning module includes a convolutional neural network.

### Claim 10 (depends on 9)

The system of claim 9, wherein the processor is configured to identify a new attribute to filter the nuisance events based on the test image and the reference image for the first layer and the test image and the reference image for the second layer.

### Claim 11 (depends on 9)

The system of claim 9, wherein the processor is configured to provide instructions to the system to perform a hot scan of the semiconductor wafer using the light source and the detector that is used to generate the test image and the reference image of the first layer and the test image and the reference image of the second layer.

### Claim 12

A method comprising:
performing a hot scan of a semiconductor wafer using a wafer inspection system; and
reporting, using a processor, a location of a defect of interest on the semiconductor wafer that is common across a first layer and a second layer of the semiconductor wafer based on defect locations, difference images, or defect attributes.

### Claim 13 (depends on 12)

The method of claim 12, wherein the method further comprises:
receiving, at the processor, defect locations of the first layer;
receiving, at the processor, defect locations of the second layer;
performing, using the processor, common-unique analysis; and
filtering, using the processor, all common events thereby excluding a unique location, wherein the reporting the location of the defect of interest is based on one of the common events.

### Claim 14 (depends on 13)

The method of claim 13, further comprising receiving, at the processor, defect attributes of the first layer and defect attributes of the second layer, wherein filtering all the common events uses the defect attributes, and wherein the defect attributes of the first layer and the second layer include signal strength, shape, size, location relative to a point, brightness, difference in brightness relative to background, or location on a design.

### Claim 15 (depends on 12)

The method of claim 12, further comprising:
receiving, at the processor, a difference image of a first layer for each defect location on the first layer, wherein the difference image of the first layer is between neighboring dies on the first layer;
receiving, at the processor, a difference image of a second layer for each defect location on the second layer, wherein the difference image of the second layer is between neighboring dies on the second layer;
comparing, using the processor, the difference image of the first layer and the difference image of the second layer; and
filtering, using the processor, defects for which the difference image of the first layer and the difference image of the second layer are similar thereby determining the defects of interest.

### Claim 16 (depends on 12)

The method of claim 12, further comprising:
receiving, at the processor, a test image and a reference image for each defect location on a first layer;
receiving, at the processor, a test image and a reference image for each defect location on a second layer;
calculating, using the processor, one or more attributes using a machine learning module operated by the processor, wherein the one or more attributes filter nuisance events and are based on the test image and the reference image for the first layer and the test image and the reference image for the second layer, and wherein the test image and the reference image for the first layer and the test image and the reference image for the second layer are used as channels; and
filtering, using the processor, nuisance events with the one or more attributes thereby determining the defect of interest.

### Claim 17 (depends on 16)

The method of claim 16, wherein a difference image is formed from the test image and the reference image corresponding to the test image, and wherein the difference image is used as one of the channels.

### Claim 18 (depends on 16)

The method of claim 16, wherein a difference image is formed between neighboring dies on the test image on the first layer or neighboring dies on the test image on the second layer.

### Claim 19 (depends on 16)

The method of claim 16, further comprising identifying a new attribute, using the processor, to filter the nuisance events based on the test image and the reference image for the first layer and the second layer.

### Claim 20 (depends on 12)

A non-transitory computer readable medium storing a program configured to instruct a processor to execute the method of claim 12.
