System and method for defect detection threshold determination in an workpiece surface inspection system

• US 20080075353A1
• Filed: 09/22/2006
• Published: 03/27/2008
• Est. Priority Date: 09/22/2006
• Status: Active Grant

• Current Assignee: Andrey Vertikov, Mehmet Tek, Timothy Tiemeyer
• Original Assignee: Andrey Vertikov, Mehmet Tek, Timothy Tiemeyer
• Inventors: Tek, Mehmet, Vertikov, Andrey, Tiemeyer, Timothy
• Granted Patent: US 8,260,035 B2
• Time in Patent Office: Days
• US Class Current: 382/145
• CPC Class Codes: G01N 2021/4711 Multiangle measurement
• CPC Class Codes: G01N 2021/4735 Solid samples, e.g. paper, glass
• CPC Class Codes: G01N 2021/8861 Determining coordinates of flaws
• CPC Class Codes: 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
• CPC Class Codes: G01N 21/9501 Semiconductor wafers manufacturing processes per se of semiconductor devices implementing a measuring step H01L22/10
• CPC Class Codes: G06F 18/2415 based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
• CPC Class Codes: G06V 10/145 Illumination specially adapted for pattern recognition, e.g. using gratings
• CPC Class Codes: G06V 10/30 Noise filtering
• CPC Class Codes: G06V 10/98 Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns

Abstract

A method and system for inspecting a surface of a semiconductor workpiece comprises providing a surface inspection system and using the surface inspection apparatus to cause laser light to impinge upon a test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light; collecting the returned light and generating a signal from the returned and collected light, the signal comprising a signal value representative of a characteristic of the workpiece surface at the test location; providing a plurality of threshold candidates and causing the surface inspection system to select a threshold from among the plurality of threshold candidates; comparing the threshold to the signal value to obtain a difference value; using the difference value to assess the characteristic of the workpiece surface at the test location; and using the surface inspection system to automatically cause the method to be repeated for a plurality of test locations on the workpiece surface.

Claims

1. A method for inspecting a surface of a semiconductor workpiece, the method comprising:

providing a surface inspection system and using the surface inspection apparatus to cause laser light to impinge upon a test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light;

collecting the returned light and generating a signal from the returned and collected light, the signal comprising a signal value representative of a characteristic of the workpiece surface at the test location;

providing a plurality of threshold candidates and causing the surface inspection system to select a threshold from among the plurality of threshold candidates;

comparing the threshold to the signal value to obtain a difference value;

using the difference value to assess the characteristic of the workpiece surface at the test location; and

using the surface inspection system to automatically cause the method to be repeated for a plurality of test locations on the workpiece surface.

2. A method as recited in claim 1, wherein the collected and returned light consists of scatter light.

3. A method as recited in claim 1, wherein the collected and returned light consists of reflected light.

4. A method as recited in claim 1, wherein the providing of the plurality of threshold candidates comprises providing at least one of the threshold candidates using an estimate of the signal value.

5. A method as recited in claim 1, wherein the providing of the threshold candidates comprises using statistics based upon a plurality of signal values for a corresponding plurality of test locations on the workpiece surface to provide at least one of the threshold candidates.

6. A method as recited in claim 1, wherein the providing of the threshold candidates comprises using a size that is representative of a minimum size expected for the characteristic that is detectable given the sensitivity of the surface inspection system to provide at least one of the threshold candidates.

7. A method as recited in claim 6, further comprising using a constant sensitivity (CSENS) value to comprise the at least one threshold candidate.

8. A method as recited in claim 1, wherein the providing of the threshold candidates comprises using a false alarm-based threshold that is representative of a minimum size expected for the characteristic that is detectable given the sensitivity of the surface inspection system and a desired maximum probability of false identifications of the characteristic in the signal value for the surface inspection system to provide at least one of the threshold candidates.

9. A method as recited in claim 1, wherein the providing of the threshold candidates comprises using a false alarm-based threshold candidate that is representative of a maximum acceptable probability of false defect detection for the surface inspection system as at least one of the threshold candidates.

10. A method as recited in claim 1, wherein the selection of the threshold comprises using an estimate of the signal value to select the threshold from among the threshold candidates.

11. A method as recited in claim 1, wherein the selection of the threshold comprises using statistics based upon a plurality of signal values for a corresponding plurality of test locations on the workpiece surface.

12. A method as recited in claim 1, wherein the selection of the threshold comprises assigning a value to each of the threshold candidates and selecting the threshold by selecting one of the threshold candidates based on the threshold candidate values.

13. A method as recited in claim 12, wherein the selection of the one of the threshold candidate values comprises selecting a maximum of the threshold candidate values.

14. A system for inspecting a surface of a semiconductor workpiece, the system comprising:

a laser source that causes a laser light to impinge upon a test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light;

a collection subsystem that collects the returned light and generates a signal from the returned and collected light, the signal comprising a signal value representative of a characteristic of the workpiece surface at the test location;

a processing device that compares the signal value to a plurality of threshold candidates and selects a threshold from among the plurality of threshold candidates, and that uses the threshold to assess the characteristic of the workpiece surface at the test location;

wherein the surface inspection system to automatically analyzes the workpiece surface at a plurality of test locations by making said comparisons at each of the test locations.

15. A method of differentiating noise from particles on a surface of a semiconductor workpiece using a threshold, wherein the threshold is used to assess a characteristic of the workpiece surface at a current test location, the method comprising:

providing a surface inspection system and using the surface inspection apparatus to cause laser light to impinge upon the current test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light;

collecting the returned light and generating a signal from the returned and collected light, the signal comprising a signal value representative of the characteristic of the workpiece surface at the current test location and a plurality of prior signal values representative of the characteristic of the workpiece surface at a corresponding plurality of secondary test locations, wherein the secondary test locations comprise at least one of the current test location and test locations other than the current test location;

causing the surface inspection system to select a threshold from among at least one threshold candidate, wherein the at least one threshold candidate is selected based on statistical data for the secondary test locations;

comparing the threshold to the signal value to obtain a difference value;

using the difference value to assess the characteristic of the workpiece surface at the current test location; and

using the surface inspection system to automatically cause the method to be repeated for a plurality of test locations on the workpiece surface.

16. A method as recited in claim 15, wherein the selection of the at least one threshold candidate comprises comparing the signal value with a statistically-determined false alarm-based threshold candidate.

17. A method as recited in claim 16, wherein the statistically-determined false alarm-based threshold candidate is based on a desired maximum probability and statistical characteristics of the signal values for the secondary test locations.

18. A method as recited in claim 15, wherein the signal value comprises a component attributable to haze and a component noise, and the selection of the at least one threshold candidate comprises using at least one of the haze component and the noise component.

19. A method as recited in claim 18, wherein the haze component and the noise component are based on the signal value and accumulated run-time statistics.

20. A system for inspecting a surface of a semiconductor workpiece, wherein the workpiece surface has a characteristic at a current test location, the system comprising:

a laser source that causes a laser light to impinge upon the current test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light;

a collection subsystem that collects the returned light and generates a signal from the returned and collected light, the signal comprising a signal value representative of the characteristic of the workpiece surface at the current test location and a plurality of prior signal values representative of the characteristic of the workpiece surface at a corresponding plurality of secondary test locations, wherein the secondary test locations comprise at least one of the current test location and test locations other than the current test location; and

a processing device that selects a threshold from among at least one threshold candidate, compares the threshold to the signal value to obtain a difference value, and uses the difference value to assess the characteristic of the workpiece surface at the current test location, wherein the at least one threshold candidate is selected based on statistical data for the secondary test locations;

wherein the surface inspection system automatically analyzes the workpiece surface at a plurality of test locations on the workpiece surface by making said comparisons at each of the test locations.

21. A method for inspecting a surface of a semiconductor workpiece, the method comprising:

providing a surface inspection system and using the surface inspection apparatus to cause laser light to impinge upon a test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light;

collecting the returned light and generating a signal from the returned and collected light, the signal comprising a signal value representative of a characteristic of the workpiece surface at the test location, wherein the signal value comprises raw data comprising a data line comprising a vector of voltage values based on intensity measurements of the returned light at locations on the workpiece surface within the test location and a selected scan line;

filtering the raw data;

extracting components of the signal value representative of the matched filtered

data and attributable to a haze component and a noise component;

performing a thresholding calculation; and

using the surface inspection system to automatically cause the method to be repeated for a plurality of test locations on the workpiece surface.

22. A method for inspecting a surface of a semiconductor workpiece, the method comprising:

providing a surface inspection system and using the surface inspection apparatus to cause laser light to impinge upon a test location on the workpiece surface and thereby cause the laser light to emerge from the surface as returned light comprising at least one of reflected light and scatter light;

collecting the returned light and generating a signal from the returned and collected light, the signal comprising a signal value representative of a characteristic of the workpiece surface at the test location;

providing a plurality of threshold candidates in respective tiers and causing the surface inspection system to select a threshold from among the plurality of threshold candidates using the tiers;

comparing the threshold to the signal value to obtain a difference value;

using the difference value to assess the characteristic of the workpiece surface at the test location; and

using the surface inspection system to automatically cause the method to be repeated for a plurality of test locations on the workpiece surface.

23. A method as recited in claim 22, wherein a first threshold tier is used to identify a first type of the characteristics, and wherein a second threshold tier is used to identify a second type of the characteristics.

24. A method as recited in claim 22, wherein the characteristic being identified comprises defects, and wherein the first type comprises a scratch defect;

and the second type comprises a point defect.

25. A method for assessing the capability of a multi-channel surface inspection system to analyze a workpiece to selected defect size identification specifications, the method comprising:

comparing channel-specific statistically determined false alarm-based threshold values for a portion of the workpiece to a specified minimum defect size, and

finding the surface inspection system suitable to analyze the workpiece if the specified minimum defect size exceeds a selected number of the channel-specific false alarm-based threshold values.

28. A method as recited in claim 25, wherein the finding of the surface inspection system to be suitable comprises finding the surface inspection system to be suitable if the specified minimum defect size exceeds at least one of the channel-specific statistically determined false alarm-based threshold values.

29. A method as recited in claim 25, wherein the comparing of the channel-specific statistically determined false alarm-based threshold values for a portion of the workpiece to associated channel-specific specified minimum defect sizes comprises providing a minimum defect size for each channel.

30. A method as recited in claim 25, wherein the minimum defect size comprises a size-determined threshold value comprising a voltage that is representative of the smallest defect size that is detectable given the sensitivity of the surface inspection system.

26. A method for assessing the capability of a multi-channel surface inspection system to analyze a semiconductor workpiece to selected defect size identification specifications, the method comprising:

measuring local area noise levels for each channel and establishing detection limits for desired measurement confidence for each channel during acquisition of data on the workpiece based on the local area noise levels.

27. A method as recited in claim 26, wherein the establishment of detection limits comprises developing mean haze and local area averages for variance for each channel as the data is acquired.