Computer-aided design and analysis of circuits and semiconductor – Nanotechnology related integrated circuit design
Reexamination Certificate
2002-05-29
2004-08-17
Siek, Vuthe (Department: 2825)
Computer-aided design and analysis of circuits and semiconductor
Nanotechnology related integrated circuit design
C716S030000, C716S030000
Reexamination Certificate
active
06779159
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a defect inspection method and a defect inspection apparatus for inspecting defects in an object to be inspected, such as a silicon wafer.
2. Description of the Related Art
Crystal defects of a silicon wafer or the like include oxidation induced stacking faults (OSFs) and bulk microdefects (BMDs). When a silicon wafer is etched, etch pits of a few &mgr;m depth may appear in the surface thereof. Etch pits exhibiting regularity corresponding to a face-centered tetragonal structure of a silicon wafer are referred to as the OSFs, while etch pits having irregular shapes are referred to as BMDs of oxidation deposits, microdisplacement, and stacking faults. Usually, the inspection for the OSFs or BMDs is visually performed with the aid of a microscope. The defects are visually checked and the number of such defects in an inspection visual field is counted to implement quality assurance and quality control of wafers on the basis of the counting result.
There has also been proposed methods whereby to capture microscopic images and process the images so as to detect and count defects. Examples of such methods are found in Japanese Unexamined Patent Application Publication No. 61-194737 entitled “SILICON WAFER OSF DENSITY INSPECTION METHOD” and Japanese Examined Patent Application Publication No. 6-71038 entitled “CRYSTAL DEFECT RECOGNIZING METHOD.”
Here, according to a first defect inspection method, potential defects that have been processed into binarized images on the basis of the shape characteristics of defects are checked to determine whether they are truly defects. For instance, according to the method disclosed in Japanese Unexamined Patent Application Publication No. 61-194737, potential defects are checked against a reference shape to determine whether the candidates are truly defects. The method according to Japanese Examined Patent Application Publication No. 6-71038 involves the length and aspect ratio of a defect. These inspection methods focus only on the shapes of defects to make decisions. More specifically, only a reference shape or the shape of a defect involving the length or aspect ratio of the defect are used as determining parameters to decide whether the defect is a true defect.
According to a second defect inspection method, a predetermined threshold value is set for a captured image to binarize the image thereby to detect potential defects and to determine whether they are true defects on the basis of the characteristic amounts of the potential defects. Then, the quantity of the true defects is counted.
The first inspection method, however, presents the following problem since it uses only the shape of a potential defect as the determining parameter.
In the case of the potential defects shown in
FIG. 2
, judging from the shapes thereof, they may be determined to be OSFs, while it still remains uncertain from the aspect of their distribution of occurrence. More specifically, of the seven potential defects found in the visual field, six are very closely located. Such potential defects are normally produced from accidental contact of tweezers or some other foreign matters to the surface of a wafer during a process of etching a silicon wafer, and therefore, are not true defects that occurred on the surface of the wafer.
The conventional inspection method, however, makes it impossible to accurately discriminate the potential defects caused by tweezers coming in contact with the surface of a wafer. As a result, it is determined that there is a greater number of defects than there really is, thus interfering with accurate quality control.
The second inspection method presents the following problem.
The number of defects in an area measuring a few hundreds of &mgr;m squares per visual field is counted, meaning that the inspection of the whole surface of a wafer involves a few hundreds or tens of thousands of visual fields to be inspected. In the case of visual inspection, at least about one second is required to inspect one visual field, so that inspecting all visual fields takes too much time. Furthermore, human factors, such as the degree of fatigue of an inspector, inevitably lead to inconsistency in inspection results.
Although the method in which microscopic images are processed to count the number of potential defects solves the problem in the above visual inspection, the method presents the problem described below. Since captured images are simply processed into binarized images to detect defects, if there are many defects in an inspection visual field, as shown in
FIG. 3A
, it may happen that a plurality of closely located defects candidates shown in
FIG. 3B
are overlapped and misjudged to be a single defect. Especially in the case of BMDs, numerous defects frequently take place, so that the defects may be overlapped and detected as a single defect, thus posing a problem in achieving reliable quality assurance and quality control.
In addition, if a standard microscope is used, a captured defect of a few &mgr;m depth may be overlooked due to poor contrast, presenting another problem in achieving reliable quality assurance and quality control.
SUMMARY OF THE INVENTION
Accordingly, the present invention has been made with a view toward solving the problems discussed above, and it is an object of the present invention to provide a defect inspection method and a defect inspection apparatus that improve determination accuracy by using the distribution of detected defects as a parameter for determining whether a potential defect is a true defect, and also improve defect detection accuracy by preventing human errors and by accurately identifying micro-defects and closely located defects.
To this end, according to one aspect of the present invention, there is provided a defect inspection method for inspecting potential defects observed on a surface of an object to be inspected, wherein the number and the positions of the detected potential defects are used as determining parameters for determining whether the potential defects are true defects.
With this arrangement, the number of detected potential defects and the locations of the detected potential defects are used as the parameters for determining whether the potential defects are true defects, making it easier to recognize the potential defects (not true defects) that occurred due to contact of a foreign object, including tweezers, with a surface of an object to be inspected, such as a silicon wafer. The contact by a foreign object, such as tweezers, causes potential defects to take place closely located. More specifically, such potential defects may densely occur around a point or along a line. Therefore, it is possible to determine whether detected potential defects are true defects by measuring the density of the potential defects based on the number of detected potential defects and also the positions thereof.
Preferably, the density of potential defects is determined on the basis of the aforesaid number of potential defects and the positions thereof, and the determined density is compared with a preset value so as to decide whether the potential defects are true defects.
With this arrangement, if the determined density of potential defects based on the number of potential defects and the positions thereof exceeds the preset value, then it is determined that the potential defects are not true defects.
According to another aspect of the present invention, there is provided a defect inspection method wherein, to inspect potential defects observed on a surface of an object to be inspected, if a total number of potential defects in an inspection visual field is smaller than a preset value, then an area that is smaller than the inspection visual field is set, the gravity center position of a detected potential defect is aligned with the central position of the foregoing area, the number of potential defects in the area is counted, the density of the potential defects is determined on the basis of the ratio of the counted numbe
Nakashima Yutaka
Yokoyama Hirokazu
Lin Sun James
Lorusso, Loud & Kelly
Siek Vuthe
Sumitomo Mitsubishi Silicon Corporation
LandOfFree
Defect inspection method and defect inspection apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Defect inspection method and defect inspection apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Defect inspection method and defect inspection apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3311968