Optical-mode selection for multi-mode semiconductor inspection

Patent metadata

Field	Value
Publication number	US-11010885-B2
Application number	US-201916406374-A
Country	US
Kind code	B2
Filing date	May 8, 2019
Priority date	Dec 18, 2018
Publication date	May 18, 2021
Grant date	May 18, 2021

Abstract

One or more semiconductor wafers or portions thereof are scanned using a primary optical mode, to identify defects. A plurality of the identified defects, including defects of a first class and defects of a second class, are selected and reviewed using an electron microscope. Based on this review, respective defects of the plurality are classified as defects of either the first class or the second class. The plurality of the identified defects is imaged using a plurality of secondary optical modes. One or more of the secondary optical modes are selected for use in conjunction with the primary optical mode, based on results of the scanning using the primary optical mode and the imaging using the plurality of secondary optical modes. Production semiconductor wafers are scanned for defects using the primary optical mode and the one or more selected secondary optical modes.

First claim

What is claimed is:

1. A semiconductor-inspection method, comprising:
scanning one or more semiconductor wafers or portions thereof using a primary optical mode, to identify defects;
selecting a plurality of the identified defects;
reviewing the plurality of the identified defects using an electron microscope;
based on the reviewing, classifying respective defects of the plurality as defects of either a first class or a second class;
imaging the plurality of the identified defects using a plurality of secondary optical modes;
selecting one or more of the secondary optical modes for use in conjunction with the primary optical mode, based on results of the scanning and the imaging for the classified defects, comprising:
defining multiple combinations of the primary optical mode with one or more respective secondary optical modes of the plurality of secondary optical modes,
for each combination of the multiple combinations, training a respective convolutional neural network (CNN) to predict classes of the plurality of identified defects, thereby producing a plurality of CNNs, and
evaluating the plurality of CNNs for separation of defects in the first class from defects in the second class; and

scanning production semiconductor wafers using the primary optical mode and the one or more selected secondary optical modes, to identify defects.

2. The method of claim 1 , wherein:
the first class is defects of interest that impede semiconductor-die functionality; and
the second class is nuisance defects that do not impede semiconductor-die functionality.

3. The method of claim 1 , wherein:
selecting the one or more of the secondary optical modes comprises selecting a single secondary optical mode for use in conjunction with the primary optical mode; and
scanning the production semiconductor wafers is performed using the primary optical mode and the single secondary optical mode.

4. The method of claim 1 , wherein the electron microscope is a scanning electron microscope (SEM).

5. The method of claim 1 , wherein imaging the plurality of the identified defects using the plurality of secondary optical modes is performed before the reviewing and the classifying.

6. The method of claim 1 , wherein imaging the plurality of the identified defects using the plurality of secondary optical modes is performed after the reviewing and the classifying.

7. The method of claim 1 , wherein the one or more selected secondary optical modes correspond to a respective CNN that produces maximum separation between defects in the first class and defects in the second class out of the plurality of CNNs.

8. The method of claim 1 , wherein the plurality of CNNs is trained using image data annotated with defect locations.

9. The method of claim 1 , wherein the plurality of CNNs is trained using image data augmented with modified image data.

10. A non-transitory computer-readable storage medium storing one or more programs for execution by one or more processors of a semiconductor-inspection system that includes one or more semiconductor-inspection tools, the one or more programs including instructions for:
selecting a plurality of defects identified by scanning one or more semiconductor wafers or portions thereof using a primary optical mode;
based on review of the plurality of the identified defects using an electron microscope, classifying respective defects of the plurality as defects of either a first class or a second class; and
based on results of scanning the classified defects and of imaging the classified defects using a plurality of secondary optical modes, selecting one or more of the secondary optical modes for use in conjunction with the primary optical mode to scan production semiconductor wafers, comprising:
defining multiple combinations of the primary optical mode with one or more respective secondary optical modes of the plurality of secondary optical modes,
for each combination of the multiple combinations, training a respective convolutional neural network (CNN) to predict classes of the plurality of identified defects, thereby producing a plurality of CNNs, and
evaluating the plurality of CNNs for separation of defects in the first class from defects in the second class.

11. A semiconductor-inspection system, comprising:
one or more semiconductor-inspection tools;
one or more processors; and
memory storing one or more programs for execution by the one or more processors, the one or more programs comprising instructions for:
scanning one or more semiconductor wafers or portions thereof using a primary optical mode, to identify defects;
selecting a plurality of the identified defects;
based on review of the plurality of the identified defects using an electron microscope, classifying respective defects of the plurality as defects of either a first class or a second class;
imaging the plurality of the identified defects using a plurality of secondary optical modes;
selecting one or more of the secondary optical modes for use in conjunction with the primary optical mode, based on results of the scanning and the imaging for the classified defects, comprising:
defining multiple combinations of the primary optical mode with one or more respective secondary optical modes of the plurality of secondary optical modes,
for each combination of the multiple combinations, training a respective convolutional neural network (CNN) to predict classes of the plurality of identified defects, thereby producing a plurality of CNNs, and
evaluating the plurality of CNNs for separation of defects in the first class from defects in the second class; and

scanning production semiconductor wafers using the primary optical mode and the one or more selected secondary optical modes, to identify defects.

12. The system of claim 11 , wherein:
the first class is defects of interest that impede semiconductor-die functionality; and
the second class is nuisance defects that do not impede semiconductor-die functionality.

13. The system of claim 11 , wherein the instructions for imaging the plurality of the identified defects using the plurality of secondary optical modes comprise instructions for performing the imaging before the classifying.

14. The system of claim 11 , wherein the instructions for evaluating the plurality of CNNs comprise instructions for identifying a respective CNN that produces maximum separation between defects in the first class and defects in the second class out of the plurality of CNNs.

Assignees

• Kla Tencor Corp
• Kla Corp

Inventors

• Brauer Bjorn
• Wallingford Richard
• Grama Kedar
• Lee Hucheng
• Park Sangbong

Classifications

• G06V10/82

  using neural networks · CPC title

• G06V10/7715

  Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• H10P74/203 Primary

  Structural properties, e.g. testing or measuring thicknesses, line widths, warpage, bond strengths or physical defects · CPC title

• G06F18/2132

  based on discrimination criteria, e.g. discriminant analysis · CPC title