Defect inspection device and defect inspection method

Abstract

To provide a defect inspection device with which it is possible to automatically detect an optimum optical condition for capturing an image of a region that includes a defect by a simple method.SOLUTION: A defect inspection device 1 comprises an optical microscope 10 that is provided with an image-capturing unit 14 that captures an image of a chip 31, a control unit 20 that controls an optical condition for capturing the image of a chip, and a storage unit 25 in which the preliminarily detected positions of defects in the chip are stored. The image-capturing unit includes detection means that captures a first image of a region that includes the defect stored in the storage unit while changing the optical condition, and captures a second image of a region in the same region as the region that includes the defect, that does not include a defect, and compares the feature quantity of the first image with the feature quantity of the second image, thereby detecting an optimum optical condition for capturing an image of the region that includes a defect.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a defect inspection apparatus and a defect inspection method for inspecting defects in chips formed on a wafer. [Background technology]

[0002] Semiconductor devices are manufactured by repeatedly forming multiple semiconductor device circuits (chips) in a matrix on a single semiconductor wafer, then dicing the wafer into individual chips and packaging the diced chips.

[0003] Before the individual chips are separated, the external appearance pattern of each chip formed on the wafer is sequentially imaged, and the inspection image and the reference image are compared pixel by pixel, and defects within the chip are detected based on the difference in brightness values ​​between the compared pixels (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2007-155610 Summary of the Invention [Problem to be solved by the invention]

[0005] Inspection of defects within a chip is carried out by capturing an image of the chip using an optical microscope, but it is necessary to set the optical conditions of the optical microscope so that defects can be clearly seen.

[0006] When manually setting optical conditions, even if the location of the defect is known, a person must move the optical microscope to the area containing the defect, change the optical conditions, and visually check the inspection image to find the optical conditions under which the defect is clearly visible.

[0007] However, since there are many optical conditions, it is difficult for a person to search for all of the optical conditions, and since the judgment is made by a person visually, there is a risk that defect detection will vary depending on the person.

[0008] The present invention has been made in consideration of the above points, and its main purpose is to provide a defect inspection device and a defect inspection method that use an optical microscope to inspect defects in chips formed on a wafer, and that can automatically detect, in a simple manner, the optimal optical conditions for imaging an area containing a defect. [Means for solving the problem]

[0009] The defect inspection device of the present invention is a defect inspection device that inspects defects in chips formed on a wafer, and is equipped with an optical microscope equipped with an imaging unit that captures images of the chip, a control unit that controls the optical conditions for capturing images of the chip with the optical microscope, and a memory unit that stores the positions of defects within the chip that have been detected in advance.The imaging unit is equipped with a detection means that, while changing the optical conditions with the control unit, captures a first image of an area containing a defect stored in the memory unit, and a second image of an area in the same area as the area containing the defect but not containing the defect, and detects the optimal optical conditions for capturing the area containing the defect by comparing the features of the first image with the features of the second image. [Effects of the Invention]

[0010] According to the present invention, a defect inspection device and a defect inspection method can be provided that use an optical microscope to inspect defects in chips formed on a wafer, and that can automatically detect the optimal optical conditions for imaging an area containing a defect in a simple manner. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram schematically illustrating the configuration of a defect inspection apparatus 1 according to a first embodiment of the present invention. [Figure 2]5 is a flowchart showing a method for detecting optimal optical conditions for imaging an area including a defect using the defect inspection apparatus according to the first embodiment. [Figure 3] 3(A) to 3(C) are diagrams illustrating a method for detecting optimal optical conditions for imaging an area including a defect, according to the flowchart shown in FIG. [Figure 4] 10A to 10C are diagrams illustrating a method for detecting optimal optical conditions in a modified example of the first embodiment. [Figure 5] 10(A) to 10(C) are diagrams illustrating a method for creating learning data for a plurality of non-defective chips in a second embodiment of the present invention. [Figure 6] 10A and 10B are diagrams showing the luminance value gi and sensitivity value Ai of each pixel in the defective area A. [Figure 7] 10A and 10B are diagrams showing the luminance value g′i and sensitivity value A′i of each pixel in the non-defective area B. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the following embodiment. Furthermore, appropriate modifications are possible within the scope of the effects of the present invention. (First embodiment) FIG. 1 is a diagram schematically showing the configuration of a defect inspection apparatus 1 according to a first embodiment of the present invention.

[0013] As shown in Figure 1, the defect inspection device 1 in this embodiment is a defect inspection device that inspects defects within chips formed on a wafer, and is equipped with an optical microscope 10, a control unit 20 that controls the optical conditions for capturing an image of the chip, a memory unit 25 that stores the positions of defects within the chip that have been detected in advance, and a detection means 40 that detects the optimal optical conditions for capturing an image of an area containing a defect using the optical microscope 10.

[0014] The optical microscope 10 comprises a stage 11 on which the wafer 30 is placed, a microscope body 12 that holds the stage 11, an illumination unit 13 that irradiates illumination light L onto the wafer 30, and an imaging unit 14 that captures an image of the chip.

[0015] Illumination light L emitted from the illumination unit 13 is condensed by a condenser lens 15, and then deflected by a half mirror 16 along the optical axis of an objective lens 17, and is irradiated onto the surface of a wafer 30. Light reflected from the wafer 30 passes through the objective lens 17, the half mirror 16, and a coupling lens 18, and is incident on an image sensor 19. The image captured by the image sensor 19 is processed by an image processing unit 35.

[0016] In this embodiment, the detection means 40 detects the optimal optical conditions for imaging an area including a defect based on image information of the chip imaged by the imaging unit 14. The control unit 20 sets the optical microscope 10 to the optimal optical conditions detected by the detection means 40, images the chip, and detects defects in the wafer 30.

[0017] In this embodiment, the optical conditions of the optical microscope 10 controlled by the control unit 20 are diverse, including the magnification of the objective lens 17, the type (coaxial, oblique, transmitted) and brightness of the illumination light L, the position of the differential interference filter (not shown), and the filter (not shown) for the illumination light L.

[0018] FIG. 2 is a flowchart showing a method for detecting optimal optical conditions for imaging an area including a defect using the defect inspection apparatus 1 of this embodiment.

[0019] First, as shown in FIG. 3(A), a chip 31 formed on a wafer 30 is imaged in advance using an optical microscope 10 to detect defects within the chip 31, and the positions of the detected defects within the chip 31a are registered in the memory unit 25 (step S101).

[0020] Next, as shown in FIG. 3(B), the control unit 20 changes the optical conditions of the optical microscope 10 (step S102), and captures an image (first image) of an area A (defect area) including the defect 50 (step S103).

[0021] 3(B), under the same optical conditions, an image (second image) of a region (non-defective region) B of chip 31b that does not contain defect 50 is captured in the same region as region A containing defect 50 (step S104). Here, defect region A and non-defective region B are the same region within the chip.

[0022] Next, the luminance value (feature amount) of the image of the defective area A (first image) is compared with the luminance value (feature amount) of the image of the non-defective area B (second image) (step S105).

[0023] Next, if the search for the optimal optical conditions has not been completed (No in step S106), the optical conditions of the optical microscope 10 are changed (step S102), and steps S103 to S105 are repeated. If the search for the optimal optical conditions has been completed (Ys in step S106), the optical microscope 10 is set to the optimal optical conditions, and defects in the wafer 30 are detected.

[0024] Table 1 shows examples of the luminance values ​​(A) of defective area A, the luminance values ​​(B) of non-defective area B, and the difference (B-A) between the two luminance values ​​obtained by capturing an image of defective area A and an image of non-defective area B under each of the four optical conditions. The luminance values ​​are expressed in arbitrary units.

[0025] [Table 1]

[0026] Even when the same areas A and B are imaged, if the optical conditions change, the brightness values ​​of the images imaged in areas A and B will also change, and the higher the image clarity, the greater the difference between the two. Therefore, the difference (B-A) between the brightness value (A) of defective area A and the brightness value (B) of non-defective area B is an evaluation value for determining the optimal optical conditions.

[0027] In the example shown in Table 1, of the four optical conditions, the difference in brightness value (BA) between the two is greatest under optical condition 2. Therefore, in this case, optical condition 2 can be said to be the most suitable optical condition.

[0028] In this manner, in this embodiment, it is possible to detect the optimum optical conditions for imaging an area containing a defect by comparing the brightness values ​​of the image (first image) of the defect area A with the brightness values ​​of the image (second image) of the non-defective area B. This allows for a margin in the threshold for detecting defects, thereby improving the detection accuracy when detecting defects within a chip.

[0029] In Table 1, the optimal optical conditions were determined from the difference (B-A) between the brightness value (A) of defective area A and the brightness value (B) of good area B. However, the brightness value (A) of defective area A and the brightness value (B) of good area B may also be compared using the ratio (B / A) of the two, as shown in Table 2.

[0030] In the example shown in Table 2, of the four optical conditions, optical condition 2 has the largest ratio of the two luminance values ​​(B / A). Therefore, in this case, optical condition 2 can be said to be the most suitable optical condition.

[0031] [Table 2]

[0032] (Modification of the first embodiment) In the above embodiment, the brightness values ​​of the image of the defect area A (first image) and the brightness values ​​of the image of the good area B (second image) were compared to detect the optimal optical conditions for imaging the area containing the defect, but the optimal optical conditions may also be detected by adding the difference in brightness values ​​between images of different good areas to the evaluation value.

[0033] As shown in Fig. 4, an image (second image) of area B' (non-defective area) of chip 31c, which is the same area as area A including defect 50 shown in Fig. 3(B) but is different from chip 31b shown in Fig. 3(C) and does not include defect 50, is captured. Here, defect area A and non-defective area B' are the same area within the chip.

[0034] Next, the brightness value (B1) of the image of the non-defective area (first non-defective area) B shown in FIG. 3(C) is compared with the brightness value (B2) of the image of the non-defective area (second non-defective area) B' shown in FIG. 4.

[0035] Table 3, like Table 1, shows examples of the luminance value (A) of defective area A, the luminance value (B1) of first non-defective area B, the luminance value (B2) of second non-defective area B', the difference between luminance value (A) and luminance value (B1) (first luminance value difference) (D1=B1-A), the difference between luminance value (B1) and luminance value (B2) (second luminance value difference) (D2=B1-B2), and the difference between the first luminance value difference (D1) and the second luminance value difference (D2) (D1-D2) under various optical conditions. The luminance values ​​are expressed in arbitrary units.

[0036] Here, the first brightness value difference (D1=B1−A) is the same as the difference (BA) between the brightness value (A) of the defective area A and the brightness value (B) of the non-defective area B shown in Table 1.

[0037] [Table 3]

[0038] Even when good areas B and B' are imaged under the same optical conditions, the brightness values ​​B1 and B2 of the images captured from good areas B and B' will vary due to variations between chips. However, the higher the image clarity, the smaller the difference between the two. Therefore, the difference between the brightness values ​​of the good areas (B1 - B2) is also an evaluation value for determining the optimal optical conditions. Therefore, the difference between the brightness value (A) of defective area A and the brightness value (B1) of good area B (first brightness value difference: D1) and the difference between the brightness values ​​of good areas B and B' (second brightness value difference: D2) (D1 - D2) is an evaluation value for determining the more optimal optical conditions.

[0039] In the example shown in Table 3, among the four optical conditions, optical condition 2 has the largest difference in brightness value (D1-D2). Therefore, in this case, optical condition 2 can be said to be the most suitable optical condition. In addition, by reducing the difference in brightness value between non-defective areas, it is possible to suppress the detection of false defects that may result in non-defective products being identified as defective.

[0040] (Second embodiment) In the first embodiment, the brightness values ​​of the images captured in the non-defective area and the defective area were used as the feature values ​​that serve as evaluation values ​​for determining the optimal optical conditions, but the "sensitivity value" described below may also be used as the feature value.

[0041] A known defect inspection method is a non-defective learning method called the DSI (Die-to-Statistical Image) comparison method.

[0042] FIG. 5 is a diagram illustrating a method for creating learning data for a plurality of non-defective chips.

[0043] First, as shown in Fig. 5(A), images of multiple good chips are captured to obtain an image of the same area as the good area B shown in Fig. 3(C). Here, the number of pixels P in area B is set to 12 (3 × 4).

[0044] Since images of the same area are acquired, the distribution of the brightness values ​​of each pixel forms a normal distribution that reflects the variation of non-defective products. Therefore, by performing statistical processing for each pixel, the average value G i (i=1~12) and standard deviation σ i (i=1~12) can be calculated.

[0045] Using the learning data obtained in this way, the sensitivity value A of each pixel in the defect area A and the non-defective area B shown in Figures 3(B) and 3(C) is calculated for the wafer to be inspected. i is calculated based on the following formula (1), where g iindicates the luminance value of each pixel.

[0046]

number

[0047] FIG. 6A shows the brightness value g of each pixel in the defective area A. i 6B shows the sensitivity value A of each pixel calculated using equation (1). i Shows.

[0048] Similarly, FIG. 7A shows the brightness value g' of each pixel in the non-defective area B. i 7B shows the sensitivity value A' of each pixel calculated using equation (1). i Shows.

[0049] In this way, the sensitivity value A of each pixel in the defect area A is calculated while changing the optical conditions of the optical microscope. i , and the sensitivity value A' of each pixel in the non-defective area B i Get.

[0050] Then, under each optical condition, the sensitivity value A of each pixel in the defective area A is i The maximum value Amax of the sensitivity of each pixel in the non-defective area B is i The maximum value A'max of the above is compared.

[0051] Table 4 shows the sensitivity value (Amax) of defective area A, the sensitivity value (A'max) of non-defective area B, and the difference between the two sensitivity values ​​(A'maxB-Amax) under each of the four optical conditions. The sensitivity values ​​are in arbitrary units.

[0052] [Table 4]

[0053] Even when the same areas A and B are imaged, if the optical conditions change, the sensitivity values ​​of the images imaged in areas A and B will also change; the higher the clarity of the defective area in area A, the greater the difference from the learning data, and the greater the difference with area B. Therefore, the difference (Amax - A'max) between the sensitivity value of defective area A (Amax) and the sensitivity value of non-defective area B (A'max) is the evaluation value used to determine the optimal optical conditions. The sensitivity value is the threshold value used to detect the target defect during inspection. Therefore, because it is the same evaluation value as in an actual inspection, it allows for a more accurate evaluation than using "brightness values."

[0054] In the example shown in Table 4, of the four optical conditions, optical condition 2 has the largest difference in luminance value (A'max - Amax) between the two. Therefore, in this case, optical condition 2 can be said to be the most suitable optical condition.

[0055] In Table 4, the optimal optical conditions were determined from the difference (A'max-Amax) between the sensitivity value (Amax) of defective area A and the sensitivity value (A'max) of non-defective area B. However, the sensitivity value (Amax) of defective area A and the sensitivity value (A'max) of non-defective area B may also be compared using the ratio (A'max / Amax) of the two.

[0056] While the present invention has been described above using preferred embodiments, these descriptions are not limiting and various modifications are possible. For example, in the above embodiments, the brightness values ​​and sensitivity values ​​of images captured in the non-defective area and the defective area are used as feature quantities that serve as evaluation values ​​for determining the optimal optical conditions for the optical microscope. However, the present invention is not limited to these, and other feature quantities may be used as long as they vary depending on the image sharpness, such as variance values ​​and brightness values ​​after edge enhancement filter processing. [Explanation of symbols]

[0057] 1. Defect inspection equipment 10. Optical microscope 11 Stages 12 Microscope body 13 Lighting Department 14 Imaging unit 15 Condenser lens 16 Half Mirror 17 Objective Lens 18 Coupling Lens 19 Image sensor 20 Control Unit 25 Memory section 30 wafers 31 chips 35 Image processing section 40 Detection means 50 Defects

Claims

1. A defect inspection apparatus for inspecting defects in chips formed on a wafer, comprising: an optical microscope equipped with an imaging unit that captures an image of the chip; a control unit that controls optical conditions for capturing an image of the chip with the optical microscope; a storage unit in which the positions of defects detected in advance within the chip are stored; Equipped with the imaging unit captures a first image of an area including the defect stored in the storage unit and a second image of an area that is the same area as the area including the defect but does not include the defect while changing the optical conditions using the control unit; a detection unit that detects optimal optical conditions for capturing an image of an area including the defect by comparing a feature amount of the first image with a feature amount of the second image; the imaging unit captures a plurality of the second images; The detection means detects optimal optical conditions for imaging the area including the defect by comparing the feature amounts of the first image with the feature amounts of the second image, and also by comparing the feature amounts of the plurality of second images with each other.

2. 2. The defect inspection device according to claim 1, wherein optical conditions that maximize the difference between the feature amount of the first image and the feature amount of the second image are detected as optimal optical conditions for imaging the area including the defect.

3. 2. The defect inspection device according to claim 1, wherein a difference between a feature amount of the first image and a feature amount of the second image is defined as a first difference, a difference between the feature amounts of the plurality of second images is defined as a second difference, and optical conditions under which the difference or ratio between the first difference and the second difference is maximized are detected as optimal optical conditions for imaging the area including the defect.

4. 4. The defect inspection device according to claim 1, wherein the feature amount of the first image and the feature amount of the second image are luminance values ​​or sensitivity values.

5. A defect inspection device described in any one of claims 1 to 4, wherein the optical conditions include at least one of the magnification of an objective lens that captures an image of the chip, the type of illumination light that is irradiated onto the chip, the brightness of the illumination light, and the filter of the illumination light.

6. 1. A defect inspection method for inspecting defects in a chip by capturing an image of the chip formed on a wafer with an optical microscope, comprising: a step of registering the position of the defect in the chip in advance; capturing a first image of a region including the defect and a second image of a region that does not include the defect in the same region as the region including the defect while changing optical conditions of the optical microscope; detecting optimal optical conditions for imaging the area including the defect by comparing the feature amount of the first image with the feature amount of the second image; Including, the capturing step includes capturing a plurality of the second images; A defect inspection method in which, in the detecting step, in addition to comparing the feature amounts of the first image with the feature amounts of the second image, the feature amounts of the plurality of second images are compared to each other to detect optimal optical conditions for imaging the area including the defect.

Images