Method, computer system and apparatus for recipe generation for automated inspection of semiconductor devices

Patent metadata

Field	Value
Publication number	US-9739720-B2
Application number	US-201313885136-A
Country	US
Kind code	B2
Filing date	Mar 13, 2013
Priority date	Mar 19, 2012
Publication date	Aug 22, 2017
Grant date	Aug 22, 2017

Abstract

A method, a computer system and an apparatus are disclosed for inspection recipe generation for the automated inspection of semiconductor devices. In order to generate the inspection recipe a reference data set is used. Automatic inspection is carried out with an initial recipe on images of dies of the reference data set (reference wafermap). The detected inspection results from the automatic inspection are classified and the classified inspection results are compared with an expert classification of defects in dies. Overkill and underkill numbers are automatically generated. According to the overkill and underkill numbers the inspection recipe parameters are modified. Automatic inspection is repeated if the detection and/or the classification are below a predefined threshold.

First claim

What is claimed is:

1. A method for inspection recipe generation for automated inspection of semiconductor devices, comprising at least one of the following groups of steps based on an initial condition:
when creating an initial recipe:
using a reference data set for inspection recipe generation;
running automatic inspection with the initial recipe on captured images of dies of the reference data set with an inspection system;
classifying the detected inspection results from the automatic inspection and comparing the classified inspection results with an expert classification of defects in dies;
generating overkill and underkill numbers; and
modifying inspection recipe parameters and repeating the automatic inspection if the detection and/or the classification of a test data set is below a predefined threshold using saved images of dies to iteratively improve the inspection recipe parameters for use in future semiconductor device production; or,

when starting with the initial inspection recipe and a knowledge database of dies is not large enough:
using a reference data set for inspection recipe generation;
running automatic inspection with the initial recipe to capture images of dies of a new semiconductor device loaded into the inspection system;
classifying the detected inspection results from the automatic inspection and comparing the classified inspection results with the expert classification of defects in dies;
generating overkill and underkill numbers; and
modifying inspection recipe parameters and repeating the automatic inspection if the detection and/or the classification of the test data set is below the predefined threshold using saved images of dies to iteratively improve the inspection recipe parameters for use in future semiconductor device production.

2. The method of claim 1 , wherein the reference data set is a stored reference map or a tune map which is uploaded to a computer.

3. The method of claim 2 , wherein a refining, tuning and modifying of the existing reference map or the existing tune map is carried out prior to the upload to the computer.

4. The method of claim 1 , wherein the existing inspection recipe is stored and is uploaded to a computer.

5. The method of claim 4 , wherein a refining, tuning and modifying of the existing initial inspection recipe is carried out prior to the upload to the computer.

6. The method of claim 1 , wherein the tune map contains dies of interest and a die classification is selectable by the user.

7. The method of claim 6 , wherein dies from different semiconductor devices are assed to the tune map.

8. The method of claim 1 wherein a classification table shows comparison results between the reference data set and the test data set, wherein the test data set is a test map.

9. The method of claim 1 , wherein prior to using the reference data set for inspection recipe generation, the generation of the data set and of the initial recipe comprise the steps of:
loading a semiconductor device into an inspection system;
performing a setup and a tune alignment;
inspecting the semiconductor device with the inspection system;
classifying dies on the semiconductor devices; and
adding die images and defect information to a tune map.

10. The method of claim 9 , wherein a new semiconductor device is loaded into the inspection system if a knowledge database of dies of the tune map is not large enough and automatic inspection of the newly loaded semiconductor device is carried out.

11. A computer system for inspection recipe generation for automated inspection of semiconductor devices comprising:
an inspection system arranged to capture images of dies on one or more wafers;
a computer arranged to use a reference data set for inspection recipe generation and run an automatic inspection with an initial recipe on the images of dies of the reference data set;
a dialog, with a first window showing at least the reference data, a second window showing at least test data; a third window showing a tune map and a fourth window showing a classification table which enables a comparison between the classified inspection results and an expert classification of defects in dies, with regard to the overkill and underkill numbers, wherein the computer is arranged to perform a generation and display of overkill and underkill numbers, and to modify inspection recipe parameters and repeat the automatic inspection if the detection and/or the classification of a test data set is below a predefined threshold using saved images of dies to iteratively improve the inspection recipe parameters for use in future semiconductor device production.

12. The computer system of claim 11 , wherein the reference data set is a stored reference map or a tune map which is uploaded to the computer.

13. The computer system of claim 12 , wherein a refining, tuning and modifying of the existing reference map or the existing tune map is carried out prior to the upload to the computer.

14. The computer system of claim 11 , wherein the existing inspection recipe is stored and is uploaded to the computer system.

15. The computer system of claim 14 , wherein a refining, tuning and modifying of the existing initial inspection recipe is carried out prior to the upload to the computer.

16. The computer system of claim 11 , wherein tune map contains dies of interest and a die classification is selectable by the user via the dialog.

17. The computer system of claim 16 wherein dies from different semiconductor devices are assed to the tune map and displayed in the third window of the dialog.

18. The computer system of claim 11 , wherein a classification table of the fourth shows comparison results between the reference data set and a test data set, wherein the test data set is the test map.

19. The computer system of claim 11 , wherein the inspection system is connected to the computer system and the inspection system is used to run automatic inspection of a loaded semiconductor device.

20. The computer system for inspection recipe generation for automated inspection of semiconductor devices of claim 11 , wherein a new semiconductor device is loaded into the inspection system if a knowledge database of dies is not large enough and automatic inspection of the newly loaded semiconductor device is carried out and the new images are added to the knowledge database.

21. An apparatus for inspection recipe generation for automated inspection of semiconductor devices comprising:
an inspection system, comprising:
an incident light illumination system;
a camera arranged to receive light from a surface of the semiconductor devices, wherein the light is converted to electric image data for further analysis;

a computer arranged to use a reference data set for inspection recipe generation and run an automatic inspection with an initial recipe on images of dies of the reference data set, in order to modify inspection recipe parameters by iteratively improving the inspection recipe parameters of a test data set using saved images of dies for use in future semiconductor device production; and
at least one display, which is subdivided into a first window, a second window, a third window and a fourth window; wherein the first window showing at least the reference data, the second window showing at least test data; the third window showing a tune map and the fourth window showing a classification table which enables a comparison between images of the classified inspection results and images of an expert

Assignees

• Kla Tencor Corp

Inventors

• De Wel Koen
• Carette Cedric

Classifications

• H10P74/203

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

• H10P74/23

  characterised by multiple measurements, corrections, marking or sorting processes · CPC title

• G05B23/0216

  Human interface functionality, e.g. monitoring system providing help to the user in the selection of tests or in its configuration · CPC title

• G01N21/9501 Primary

  Semiconductor wafers (manufacturing processes per se of semiconductor devices implementing a measuring step H10P74/20) · CPC title

• G05B2219/37224

  Inspect wafer · CPC title