Method for evaluating pattern defects on a wafer surface

• US 20010036306A1
• Filed: 03/05/2001
• Published: 11/01/2001
• Est. Priority Date: 03/08/2000
• Status: Active Grant

• Current Assignee: Leica Microsystems Gmbh (Danaher Corporation)
• Original Assignee: Leica Microsystems Gmbh (Danaher Corporation)
• Inventors: Wienecke, Joachim
• Granted Patent: US 6,941,009 B2
• Time in Patent Office: Days
• US Class Current: 382/149
• CPC Class Codes: G01N 2021/8867 using sequentially two or more inspection runs, e.g. coarse and fine, or detecting then analysing
• CPC Class Codes: G01N 21/9501 Semiconductor wafers manufacturing processes per se of semiconductor devices implementing a measuring step H01L22/10
• CPC Class Codes: G01N 21/95607 using a comparative method
• CPC Class Codes: H01L 2223/54466 Located in a dummy or reference die

Abstract

The invention concerns a method for evaluating pattern defects on a wafer surface, comprising the following steps: acquiring the surface data of a plurality of individual image fields (4) of a series-produced wafer (1); storing the data in a reference data set and making it available as reference data for the inspection of further wafers of the same series; inspecting, successively in time, the individual image fields (4) on the surface of a wafer (1) presently being examined; retrieving from the reference data set a reference datum corresponding to the respective individual image field (4) presently being inspected; comparing the surface of each individual image field (4) currently being inspected to the corresponding reference datum; if one or more deviations are identified, subsequently classifying the deviations into critical and noncritical defects in terms of the functionality of the chip; and simultaneously updating or adding to the reference data set.

Claims

1. A method for evaluating pattern defects on a wafer surface, in which surface data of an individual image field (4) of the wafer surface that have just been acquired are compared to electronically stored reference data that have been generated from previously acquired surface data of identically located individual image fields (4) from a plurality of wafer surfaces of the same production series and pattern.

2. The method as defined in

claim 1

, comprising the following steps:

acquiring the surface data of a plurality of differently located individual image fields (4) of the wafer surface;

comparing (S8) the surface data of each individual image field (4) to a reference datum from a reference data set in which all the previously acquired surface data of the identically located individual image fields (4) of wafers having the same pattern are stored;

ascertaining deviations as a result of the comparison;

classifying the deviations into critical and noncritical defects; and

outputting advisories as to the individual image fields (4) of the wafer surface having critical deviations.

3. The method as defined in

claim 2

, characterized in that the comparison of the surface data of a first individual image field (4) to the reference data set, the ascertaining of deviations, and the classification thereof are accomplished in temporally parallel fashion with the acquisition of the surface data of a second individual image field (4);

then the comparison of the surface data of the second individual image field (4) to the reference data set, the ascertaining of deviations, and the classification thereof are performed in temporally parallel fashion with the acquisition of the surface data of a third individual image field (4); and

so forth.

4. The method as defined in

claim 2

, characterized in that in order to obtain a reference data set, the surface data of identically located individual image fields (4) of a plurality of wafers having the same pattern are acquired, and are analyzed according to various parameters.

5. The method as defined in

claim 1

, characterized in that when a deviation is identified, the surface data of the corresponding individual image field (4) are stored in an intermediate buffer (19), and from there are conveyed to an analysis apparatus (17) in which classification based on defect types is performed.

6. The method as defined in

claim 5

, characterized in that the associated reference data sets are also stored in the intermediate buffer (19), and from there are conveyed to the analysis apparatus (17) along with the buffered surface data of individual image fields (4).

7. The method as defined in

claim 1

, characterized in that the deviations are classified into different defect types, and when output are displayed visually with the allocation as to the respective defect type.

8. The method as defined in

claim 1

, characterized in that a segmentation of an individual image field (4) into regions each having similar surface patterns is performed;

and the classification of the defect types is performed according to location with respect to the various surface pattern regions.

9. The method as defined in

claim 1

, characterized in that the definition of the defect types is performed on the basis of a combination of analysis parameters of the surface data of known defects.

10. The method as defined in

claim 1

, characterized in that for all individual image fields (4), the same type of reference data set is selected for a first comparison (S8) of the acquired surface data.

11. The method as defined in

claim 1

, characterized in that a specific type of reference data set is selected for a first comparison (S8) as a function of the individual image field (4) to be examined, the individual image fields (4) being categorized into classes and one type of reference data set being allocated to each individual image field class.

12. The method as defined in

claim 11

, characterized in that the information regarding an allocation of an individual image field (4) to a class is stored together with a location datum in the information base (16);

and that upon inspection of an individual image field (4), the reference data set to be used for the first comparison (S8) is selected on the basis of the class that is stored in the information base (16) and allocated to the relevant individual image field (4).

13. The method as defined in

claim 11

, characterized in that upon inspection of an individual image field (4), firstly its class is ascertained from data of the individual image field (4), and with them the reference data set is then selected.

14. The method as defined in

claim 1

, characterized in that a predetermined method for inspecting the wafer (1) and/or specific inspection parameters are established as a function of the class of the individual image field (4).

15. The method as defined in

claim 14

, characterized in that the inspection parameters comprise the resolution and a focus state of a focusable inspection apparatus.

16. The method as defined in

claim 1

, characterized in that after classification of the deviations in terms of defect type, the defect-affected individual image field (4) are examined as to whether they are directly adjacent to one another.

17. The method as defined in

claim 1

, characterized in that after classification of the deviations in terms of defect type, all defects are examined in terms of their position on the wafer (1).