# US 10,504,805 B2 · Method of examining defects in a semiconductor specimen and system thereof

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

- Number: US 10,504,805 B2
- Document ID: US10504805B2
- Status: Granted
- Kind: B2
- Application number: 15/685,781
- Filed: 2017-08-24
- Published: 2019-02-28
- Granted: 2019-12-10
- Earliest claimed: 2017-08-24
- Assignees: Applied Materials Israel Ltd.
- Applicants: Applied Materials Israel Ltd.
- Inventors: Yotam Sofer; Ariel Hirszhorn
- Web page: https://patentgrep.com/patent/US10504805B2

## CPC

- H10P 74/235: {comprising optical enhancement of defects or not-directly-visible states}
- G01N 15/1456: {without spatial resolution of the texture or inner structure of the particle, e.g. processing of pulse signals}
- G05B 19/41875: {characterised by quality surveillance of production}
- 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}
- Y02P 90/02: Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
- H10P: GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10
- 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)
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- Y02P: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS

## Abstract

There is provided a method of examining defects in a semiconductor specimen and a system thereof. The method comprises: processing a first defect map obtained for a first semiconductor specimen to assign cluster-seed (CS) scores to at least some of the classified defects presented therein, wherein a CS score of a given defect is indicative of a number of neighboring defects classified to the same class as the given defect; upon obtaining a second defect map for a second semiconductor specimen, using the CS scores assigned to the at least some of the classified defects presented in the first defect map to select, among defects presented in the second defect map, defects for review, wherein the first defect map and the second defect map present respective defects in an attribute hyperspace; and reviewing defects selected in the second defect map, thereby obtaining classified defects presented in the second semiconductor specimen.

## Claims

### Claim 1

An examination system comprising:
a memory; and
a processor, operatively coupled with the memory, to:
upon obtaining a first defect map representing defects in a first semiconductor specimen in an attribute hyperspace, process the first defect map to assign cluster-seed (CS) scores to at least part of classified defects presented in the first defect map, wherein a CS score of a given defect of the classified defects in the first defect map is indicative of a number of neighboring defects in the first defect map that are classified to the same class as the given defect;
upon obtaining a second defect map representing defects in a second semiconductor specimen in the attribute hyperspace, use the CS scores assigned to the at least part of classified defects presented in the first defect map to select, among defects presented in the second defect map, defects for review; and
review the selected defects in the second defect map for classification of the selected defects.

### Claim 2 (depends on 1)

The examination system of claim 1, wherein to use the CS scores, the processor is further to:
a. apply to the second defect map a set of one or more sampling utilities, each sampling utility providing for a given defect a respective score, thereby obtaining for the given defect a set of one or more sampling scores, wherein the set of sampling scores comprises a score obtained by using CS scores of defects in the first semiconductor specimen and provided by a CS score-based utility;
b. calculate a resulting “review” score of the given defect as a function of weighed sampling scores; and
c. select the defects for review in accordance with the resulting “review” score.

### Claim 3 (depends on 2)

The examination system of claim 2, wherein the processor is further to use the CS scores assigned to the at least part of classified defects presented in the first defect map to generate a sampling recipe usable for selecting defects in the second semiconductor specimen for reviewing by an examination tool.

### Claim 4 (depends on 3)

The examination system of claim 3, wherein the processor is further to generate the CS score-based utility and to use the generated CS score-based utility to generate the sampling recipe.

### Claim 5 (depends on 1)

The examination system of claim 1, wherein the processor is further to:
a. upon obtaining data informative of classification of defects presented in the second defect map, assign respective CS scores to classified defects presented in the second defect map;
b. accumulate CS scores assigned to defects in the first and the second defect maps;
c. use the accumulated CS scores to select, among defects presented in a next defect map, defects for review;
d. upon obtaining data informative of classification of defects presented in the next defect map, process the next defect map to assign CS scores to the classified defects respectively presented therein; and
e. accumulate CS scores assigned to defects in the first defect map, the second defect map and the next defect map, the accumulated CS scores usable for selecting defects for review in a further next defect map.

### Claim 6 (depends on 1)

The examination system of claim 1, wherein to select, among defects presented in the second defect map, defects for review, the processor is further to:
a. select, among defects presented in the first defect map, defects having an assigned CS score and meeting one or more predefined rules, thus giving rise to seed defects;
b. for each given seed defect:
i. identify a virtual location of the given seed defect in the second defect map;
ii. identify in the second defect map a predefined number of neighboring defects being the nearest neighbors to the virtual location of the given seed defect;
iii. assign to each identified neighboring defect an intermediate “seed” score being in direct relationship with the CS score of the given seed defect and in inverse relationship with a distance between the respective virtual location and the neighboring defect;
c. for each defect identified in the second defect map as a neighboring defect of at least one seed defect, calculate a resulting “seed” score as a predefined function of one or more intermediate “seed” scores assigned to the given defect and using the resulting “seed” scores to obtain resulting “review” scores of respective defects; and
d. assign the obtained resulting “review” scores to respective defects, the resulting “review” scores being usable for selecting defects in the second defect map for review.

### Claim 7 (depends on 1)

The examination system of claim 1 further comprising a recipe node operatively connected to an examination tool, wherein the processor is implemented as a part of the recipe node.

### Claim 8

A method comprising:
processing a first defect map obtained for a first semiconductor specimen to assign cluster-seed (CS) scores to at least part of the classified defects presented in the first defect map therein, wherein a CS score of a given defect of the classified defects in the first defect map is indicative of a number of neighboring defects in the first defect map that are classified to the same class as the given defect;
upon obtaining a second defect map for a second semiconductor specimen, using, by a processor, the CS scores assigned to the at least part of the classified defects presented in the first defect map to select, among defects presented in the second defect map, defects for review, wherein the first defect map and the second defect map present respective defects in an attribute hyperspace; and
reviewing the selected defects in the second defect map to obtain classified defects in the second semiconductor specimen.

### Claim 9 (depends on 8)

The method of claim 8, wherein selecting, in the second defect map, defects for review comprises assigning scores to defects presented therein and selecting defects matching a predefined score-related criterion, wherein the scores are calculated using the CS scores assigned to the at least some of classified defects presented in the first defect map.

### Claim 10 (depends on 9)

The method of claim 9, wherein the score-related criterion specifies a number of top-ranked defects to be selected in accordance with a predefined number of defects to be selected for review.

### Claim 11 (depends on 8)

The method of claim 8, wherein using the CS scores comprises:
a. applying to the second defect map a set of one or more sampling utilities each providing for a given defect a respective score, thereby obtaining for the given defect a set of one or more sampling scores, wherein the set of sampling scores comprises a score obtained by using CS scores of defects in the first semiconductor specimen;
b. calculating a resulting “review” score of the given defect as a function of weighed sampling scores; and
c. selecting the defects for review in accordance with the resulting “review” score.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
a. applying to the second defect map the set of sampling utilities to select a first subgroup of defects for review, wherein the set of sampling utilities comprises two or more sampling utilities;
b. upon classifying defects in the first subgroup, repeating until a stop condition is fulfilled:
i. modifying the set of sampling utilities to reduce the weight of the score obtained by using CS scores of defects in the first semiconductor specimen;
ii. applying to the second defect map a modified set of sampling utilities to select a next subgroup of defects for review; and
iii. classifying defects in the next subgroup.

### Claim 13 (depends on 12)

The method of claim 12, wherein the weight of the score obtained by using CS scores of defects in the first semiconductor specimen is equal to 100% when selecting the first subgroup of defects for review, and wherein the weight is exponentially reduced for selection of each next subgroup.

### Claim 14 (depends on 8)

The method of claim 8, wherein selecting, among defects presented in the second defect map, defects for review comprises:
a. selecting, among defects presented in the first defect map, defects having an assigned CS score and meeting one or more predefined rules, thus giving rise to seed defects;
b. for each given seed defect:
i. identifying a virtual location of the given seed defect in the second defect map;
ii. identifying in the second defect map a predefined number of neighboring defects being the nearest neighbors to the virtual location of the given seed defect;
iii. assigning to each identified neighboring defect an intermediate “seed” score being in direct relationship with the CS score of the given seed defect and in inverse relationship with the distance between the respective virtual location and the neighboring defect;
c. for each defect identified in the second defect map as a neighboring defect of at least one seed defect, calculating resulting “seed” score as a predefined function of one or more intermediate “seed” scores assigned to the given defect and using the resulting “seed” scores to obtain resulting “review” scores of respective defects; and
d. assigning the obtained resulting “review” scores to respective defects, the resulting “review” scores being usable for selecting defects in the second defect map for review.

### Claim 15 (depends on 14)

The method of claim 14, wherein the one or more predefined rules specify that the seed defects are selected among defects classified as defects of interest (DOI) having CS scores meeting a threshold criteria.

### Claim 16 (depends on 8)

The method of claim 8, further comprising:
a. upon classifying defects presented in the second defect map, assigning to said classified defects respective CS scores;
b. accumulating CS scores assigned to defects in the first and the second defect maps;
c. using the accumulated CS scores to select, among defects presented in a next defect map, defects for review;
d. upon classifying defects in the next defect map, processing the next defect map to assign CS scores to the classified defects respectively presented therein; and
e. accumulating CS scores assigned to defects in the first defect map, the second defect map and the next defect map, the accumulated CS scores usable for selecting defects for review in a further next defect map.

### Claim 17

A non-transitory computer readable medium comprising instructions that, when executed by a processor, cause the processor to perform operations comprising:
processing a first defect map obtained for a first semiconductor specimen to assign cluster-seed (CS) scores to at least part of the classified defects presented in the first defect map, wherein a CS score of a given defect of the classified defects in the first defect map is indicative of a number of neighboring defects in the first defect map that are classified to the same class as the given defect;
upon obtaining a second defect map for a second semiconductor specimen, using the CS scores assigned to the at least part of classified defects presented in the first defect map to select, among defects presented in the second defect map, defects for review, wherein the first defect map and the second defect map present respective defects in an attribute hyperspace; and
reviewing the selected defects in the second defect map to obtain classified defects in the second semiconductor specimen.

### Claim 18 (depends on 17)

The non-transitory computer readable medium of claim 17, wherein selecting, among defects presented in the second defect map, defects for review comprises:
a. selecting, among defects presented in the first defect map, defects having an assigned CS score and meeting one or more predefined rules, thus giving rise to seed defects;
b. for each given seed defect:
i. identifying a virtual location of the given seed defect in the second defect map;
ii. identifying, in the second defect map, a predefined number of neighboring defects being the nearest neighbors to the virtual location of the given seed defect;
iii. assigning to each identified neighboring defect an intermediate “seed” score being in direct relationship with CS score of the given seed defect and in inverse relationship with the distance between the respective virtual location and the neighboring defect;
c. for each defect identified in the second defect map as a neighboring defect of at least one seed defect, calculating a resulting “seed” score as a predefined function of one or more intermediate “seed” scores assigned to the given defect, and using the resulting “seed” scores to obtain resulting “review” scores of respective defects; and
d. assigning the obtained resulting “review” scores to respective defects, the resulting “review” scores being usable for selecting defects in the second defect map for review.

### Claim 19 (depends on 17)

The non-transitory computer readable medium of claim 17, wherein using the CS scores comprises:
a. applying to the second defect map a set of one or more sampling utilities each providing, for a given defect, a respective score, thereby obtaining for the given defect a set of one or more sampling scores, wherein the set of sampling scores comprises a score obtained by using CS scores of defects in the first semiconductor specimen;
b. calculating a resulting “review” score of the given defect as a function of weighed sampling scores; and
c. selecting the defects for review in accordance with the resulting “review” scores.

### Claim 20 (depends on 19)

The non-transitory computer readable medium of claim 19, the operations further comprising:
a. applying to the second defect map the set of sampling utilities to select a first subgroup of defects for review, wherein the set of sampling utilities comprises two or more sampling utilities;
b. upon classifying defects in the first subgroup, repeating until a stop condition is fulfilled:
i. modifying the set of sampling utilities to reduce the weight of the score obtained by using CS scores of defects in the first semiconductor specimen;
ii. applying to the second defect map a modified set of sampling utilities to select a next subgroup of defects for review; and
iii. classifying defects in the next subgroup.