Methods for visual inspection and manufacturing of semiconductor wafers

Abstract

To provide an appearance inspection method setting a plurality of imaging ranges in an element area, in which the positions of the imaging ranges can be accurately identified.SOLUTION: Provided is an appearance inspection method for a semiconductor wafer (10) with element areas (14), the method including the steps of: forming a plurality of alignment marks (30) on the semiconductor wafer; and setting each of imaging ranges (40) such that at least one of the plurality of alignment marks is included in each of the imaging ranges and each element area is divided into a plurality of sections by each of the imaging ranges, and photographing an image of each of the imaging ranges.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The technology disclosed in this specification relates to a method for inspecting the appearance of a semiconductor wafer.

[0002] Patent Document 1 discloses an inspection method for inspecting the appearance of a semiconductor wafer by taking a photograph.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] There is an appearance inspection method in which a plurality of imaging ranges are set in an element region provided on a semiconductor wafer, and images of each imaging range are taken. By dividing the element region into a plurality of imaging ranges in this way, the element region can be imaged with higher accuracy, and defects in the element region can be accurately detected. However, when there is no characteristic pattern within the imaging range, the position of the imaging range cannot be specified from the taken image. Although it is possible to specify the position of the imaging range by the mechanical accuracy of the imaging device, the mechanical accuracy of the imaging device is not so high, and the position of the imaging range cannot be accurately specified. In this specification, a technique capable of accurately specifying the position of an imaging range in an appearance inspection method for setting a plurality of imaging ranges in an element region is proposed.

Means for Solving the Problems

[0005] An appearance inspection method for a semiconductor wafer having an element region, comprising: forming a plurality of alignment marks on the semiconductor wafer; setting each imaging range such that at least one of the plurality of alignment marks is included in each imaging range and the element region is divided into a plurality by each imaging range; and taking an image of each imaging range.

[0006] In this visual inspection method, each imaging range is configured such that it contains at least one alignment mark. Therefore, the captured image contains an alignment mark. Consequently, the position of the imaging range in the image can be accurately determined based on the alignment mark in the image. [Brief explanation of the drawing]

[0007] [Figure 1] Plan view of a semiconductor wafer. [Figure 2] A plan view of a semiconductor wafer showing the element region. [Figure 3] Enlarged view of the element region. [Figure 4] This figure shows image 40bg, which was captured within the shooting range 40b. [Figure 5] A diagram showing image 40bg and comparison image 40bg-r superimposed. [Figure 6] An explanatory diagram showing how to set the image to be inspected and the comparison image in adjacent element regions. [Figure 7] A flowchart for repeatedly photographing and processing the element area. [Figure 8] A diagram showing a modified version of the alignment mark. [Modes for carrying out the invention]

[0008] An example of a visual inspection method disclosed herein may further include the step of performing a process to identify defects from the image of a target shooting range by comparing the image of the target shooting range with other images of shooting ranges corresponding to the target shooting range.

[0009] This configuration allows for the precise identification of the relative positions of the target and corresponding shooting ranges based on their respective alignment marks. Therefore, by comparing the images of these shooting ranges, defects can be accurately identified.

[0010] In one example of a visual inspection method disclosed herein, the semiconductor wafer may have a plurality of element regions. The corresponding imaging range may be an imaging range set in another element region of the same semiconductor wafer as the element region in which the imaging range of the target is set.

[0011] This configuration allows for more accurate identification of defects.

[0012] An example of a visual inspection method disclosed herein may further include a processing step of processing the element region. The step of capturing images of each imaging range may be performed both before and after the processing step.

[0013] With this configuration, if a defect exists, it is possible to check the change in the appearance of the defect before and after the processing step.

[0014] In one example of a visual inspection method disclosed herein, the semiconductor wafer may have a plurality of element regions. In the step of forming the plurality of alignment marks, the plurality of alignment marks may be formed on scribe lines provided at the boundaries of the plurality of element regions.

[0015] In one example of a visual inspection method disclosed herein, each imaging range may be set such that at least two of the plurality of alignment marks are included in each imaging range, and the element region is sandwiched between the at least two of the alignment marks within each imaging range.

[0016] With this configuration, even if one alignment mark is unrecognizable due to dirt or other reasons, the position of the shooting range can be determined by the other alignment mark. [Examples]

[0017] FIG. 1 shows a semiconductor wafer 10 used in manufacturing the semiconductor device of Example 1. The semiconductor wafer 10 is composed of Si, SiC, GaN, Ga2O3, etc. Hereinafter, a direction parallel to the surface 10a of the semiconductor wafer 10 is referred to as the x-direction, and a direction parallel to the surface 10a of the semiconductor wafer 10 and orthogonal to the x-direction is referred to as the y-direction. A positioning mark 12 is provided on the surface 10a of the semiconductor wafer 10. In the manufacturing method of Example 1, first, as shown in FIG. 2, a plurality of element regions 14 are formed on the surface 10a of the semiconductor wafer 10. Each element region 14 is a region where the structure of a semiconductor element is formed. Here, positioning is performed based on the positioning mark 12, and a plurality of element regions 14 are formed in a matrix. The element region 14 is formed by performing ion implantation, epitaxial growth, etching, electrode formation, etc. on the semiconductor wafer 10. FIG. 3 is an enlarged view of the element region 14. As shown in FIG. 3, the element region 14 has two main electrodes 20. Also, as shown in FIGS. 2 and 3, scribe lines 16 are provided so as to partition the plurality of element regions 14. The scribe line 16 is a region that is scraped off by dicing or the like after the completion of the semiconductor element in each element region 14. The scribe line 16 extends in a grid pattern along the x-direction and the y-direction.

[0018] Next, as shown in FIG. 3, a plurality of alignment marks 30 are formed on the surface 10a of the semiconductor wafer 10. The alignment mark 30 is an optically detectable mark and is composed of a recess or a thin film or the like. Here, positioning is performed based on the positioning mark 12 shown in FIG. 1, and a plurality of alignment marks 30 are formed. Also, the plurality of alignment marks 30 are formed in the scribe line 16 extending along the y-direction. Also, a pair of alignment marks 30a, 30b are formed at intervals in the x-direction within the scribe line 16 extending along the y-direction. Also, a plurality of pairs of alignment marks 30a, 30b are formed at regular intervals in the y-direction within the scribe line 16 extending along the y-direction.

[0019] Next, each element region 14 is photographed by the imaging device. Here, images are taken of the imaging ranges 40a to 40e shown in Figure 3. That is, for one element region 14, images are taken in multiple imaging ranges 40a to 40e that are shifted in position in the y direction. The imaging device has a line sensor camera and can photograph imaging ranges 40a to 40e that are long in the x direction. Each imaging range 40a to 40e is set up so that the alignment mark 30a is included at the left end of each imaging range 40a to 40e and the alignment mark 30b is included at the right end of each imaging range 40a to 40e. That is, in each imaging range 40a to 40e, the element region 14 is sandwiched between the alignment mark 30a and the alignment mark 30b. Therefore, as illustrated in Figure 4, an image is taken in which the element region 14 is positioned between the alignment mark 30a and the alignment mark 30b. The imaging device captures each imaging range 40a to 40e of each element region 14 by moving the line sensor camera and the semiconductor wafer 10 relative to each other in the x and y directions. Therefore, due to mechanical errors in the imaging device, an error occurs between the actual position and the design position of the imaging range 40a to 40e. However, even if an error occurs in the position of the imaging range 40a to 40e, the alignment marks 30a and 30b will not move outside the imaging range 40a to 40e.

[0020] Next, by analyzing the images of each imaging range 40a to 40e of each element region 14, defects on the surface of the element region 14 are detected. Next, in the image in which a defect is detected, the position of the defect is specified. Here, alignment marks 30a and 30b are used as references to specify the position of the defect (that is, the x-coordinate and the y-coordinate). As shown in FIG. 4, there is no pattern in the element region 14 within the imaging range 40b that can specify the position in the y direction. However, since the alignment marks 30a and 30b are provided on the scribe line 16, the position of the defect can be specified based on the alignment marks 30a and 30b. Thereafter, the semiconductor wafer 10 is diced along the scribe line 16 to divide the semiconductor wafer 10 into a plurality of semiconductor devices. Next, the semiconductor devices in which defects are detected are recovered, and the semiconductor devices without defects are shipped. For the semiconductor devices in which defects are detected, the defects can be observed. For example, cross-sectional analysis of the defects can be performed. Since the position of the defect is specified, the defect can be appropriately observed.

[0021] As described above, in this appearance inspection method, the position of the defect is specified based on the alignment marks 30a and 30b. Therefore, even if an error occurs in the positions of the imaging ranges 40a to 40e, the position of the defect can be accurately specified.

Example

[0022] In the manufacturing method of Example 2 as well, in the same manner as the manufacturing method of Example 1, each imaging range 40a to 40e of each element region 14 is imaged. Next, each of the captured images is subjected to image analysis by a defect detection device to detect defects from each image. The image analysis by the defect detection device will be described below.

[0023] The defect detection device stores comparison images for each of the shooting ranges 40a to 40e. In Example 2, the comparison image is an image of the shooting range 40a to 40e of the element region 14 that does not have defects (a so-called good product image). In other words, in Example 2, the comparison image is a pre-prepared image. Each comparison image includes alignment marks 30a and 30b. The defect detection device detects defects by overlaying the image of the shooting range 40a to 40e of the element region 14 to be inspected (hereinafter referred to as the inspection target image) onto the corresponding comparison image. At this time, the defect detection device positions the inspection target image on the comparison image using the alignment marks 30a and 30b as a reference. For example, Figure 5 shows the state in which the inspection target image 40bg of Figure 4 (i.e., the image of the shooting range 40b) is overlaid on the comparison image 40bg-r of the shooting range 40b. Here, the inspection image 40bg and the comparison image 40bg-r are superimposed so that the alignment marks 30a and 30b match between them. Note that there is an error in the imaging range of the imaging device, so there is a misalignment between the contour of the inspection image 40bg and the contour of the comparison image 40bg-r. By superimposing the inspection image onto the comparison image in this way, using the alignment marks 30a and 30b as a reference, the inspection image can be accurately positioned relative to the comparison image. By positioning in this way, defects can be detected. For example, by subtracting the pixel values ​​of each image in the comparison image from the pixel values ​​of each image in the inspection image, differences between the inspection image and the comparison image can be detected. These differences can then be detected as defects.

[0024] Furthermore, as shown in Figure 5, for example, a dot pattern 20d extends intermittently along the y-direction on the edge of the main electrode 20. When the inspection image and the comparison image are superimposed, if the dot pattern 20d of the inspection image is shifted in the y-direction relative to the dot pattern 20d of the comparison image, differences will occur at each dot. Therefore, if the inspection image and the comparison image cannot be accurately superimposed, defects near the dot pattern 20d cannot be accurately detected. In contrast, by positioning the inspection image and the comparison image based on the alignment marks 30a and 30b and superimposing them, the dot pattern 20d can be made to almost perfectly match between the inspection image and the comparison image, as shown in Figure 5. Therefore, defects near the dot pattern 20d can be accurately detected.

[0025] As described above, according to the visual inspection method of Example 2, the inspection target image is positioned based on the alignment marks 30a and 30b and compared with the comparison image, so defects can be accurately detected. [Examples]

[0026] In the manufacturing method of Example 3, similar to the manufacturing method of Example 2, each imaging range 40a to 40e of each element region 14 is imaged, and defects are detected by comparing the image to be inspected with a comparison image. However, in Example 3, the comparison image is an image taken on the same semiconductor wafer 10 as the image to be inspected.

[0027] Figure 6 shows the imaging ranges 40a to 40e of two element regions 14a and 14b. In Example 3, the visual inspection device detects defects in the imaging ranges 40a to 40e of element region 14a by comparing the image of the imaging range 40a to 40e of element region 14a with the image of the imaging range 40a to 40e of the adjacent element region 14b. For example, if the image to be inspected is image 40bg of the imaging range 40b of element region 14a, the visual inspection device selects the image of the imaging range 40b of the adjacent element region 14b as comparison image 40bg-r. Then, as shown in Figure 5, the image to be inspected 40bg is superimposed on comparison image 40bg-r so that the alignment marks 30a and 30b match, and these differences are detected as defects.

[0028] The visual inspection method of Example 3 can accurately detect defects, similar to the visual inspection method of Example 2. Furthermore, when forming alignment marks 30a and 30b, errors may cause the positions of the alignment marks 30a and 30b relative to the element region 14 to deviate from their design positions. However, within the same semiconductor wafer 10, the direction and amount of the deviation of the alignment marks 30a and 30b relative to the element region 14 are approximately the same for all element regions 14. Therefore, by using an image from the same semiconductor wafer 10 as the image to be inspected as a comparison image, the image to be inspected can be accurately superimposed onto the comparison image without being affected by positional errors of the alignment marks 30a and 30b. Thus, the visual inspection method of Example 3 allows for even more accurate superimposition of the image to be inspected onto the comparison image than Example 2, enabling more accurate detection of defects.

[0029] In Example 3, an image of element region 14b adjacent to element region 14a to be inspected was used as a comparison image. However, an image of another element region 14 that is included in the same semiconductor wafer 10 as element region 14a but is not adjacent to element region 14a may also be used as a comparison image. [Examples]

[0030] Figure 7 shows the manufacturing method of the semiconductor device in Example 4. First, in step S2, alignment marks 30a and 30b are formed. That is, in Example 4, alignment marks 30a and 30b are formed before processing the element region 14. Here, similar to Example 1, alignment marks 30a and 30b are formed on the scribe line 16. Next, in step S4, each imaging range 40a to 40e of each element region 14 is photographed in the same manner as in Example 1. Here, similar to Example 1, defects within each element region 14 are detected and the location of each defect is identified. Next, in step S6, processing is performed on each element region 14. Next, in step S8, each imaging range 40a to 40e of each element region 14 is photographed in the same manner as in step S4. Here, similar to step S4, defects within each element region 14 are detected and the location of each defect is identified. Next, in step S10, processing is performed on each element region 14. Next, in step S12, each imaging range 40a to 40e of each element region 14 is imaged in the same manner as in step S4. Here, defects within each element region 14 are detected and their locations are identified, in the same manner as in step S4.

[0031] As described above, in Example 4, imaging and processing are performed alternately on each element region 14. With the configuration of Example 4, defects and their locations can be identified at each processing stage of the element region 14. Therefore, it is possible to observe how the defects change during the manufacturing process. In other words, it is possible to observe the changes in defects before and after processing of the element region 14.

[0032] In Example 4, the imaging and processing of the element region were repeated multiple times. However, by taking images at least once both before and after processing the element region, it is possible to observe the changes in defects before and after the processing.

[0033] As explained above, according to the manufacturing methods of Examples 1 to 4, even if there is an error in the position of the imaging range 40a to 40e, the position of the imaging range 40a to 40e can be accurately determined based on the alignment marks 30a and 30b. Therefore, in Examples 1 and 4, the position of the defect can be accurately determined. Also, in Examples 2 and 3, the defect can be accurately detected. In the manufacturing process of Examples 1 to 4, foreign matter may adhere to one of the alignment marks 30a and 30b. Since the element region 14 is positioned between the alignment marks 30a and 30b in each imaging range 40a to 40e, the distance between the alignment marks 30a and 30b is wide. Therefore, the possibility of foreign matter adhering to both alignment marks 30a and 30b is low. Therefore, even if foreign matter adheres to one alignment mark 30, the position of the imaging range 40a to 40e can be determined using the other alignment mark 30 as a reference. In other examples, there may be only one alignment mark in each imaging range.

[0034] Furthermore, in Examples 1 to 4, each alignment mark 30 had a shape that was elongated in the y-direction. However, as shown in Figure 8, each alignment mark 30 may be composed of two marks spaced apart in the y-direction. Also, each alignment mark 30 may have other shapes.

[0035] Although embodiments have been described in detail above, these are merely illustrative and do not limit the scope of the claims. The technologies described in the claims include various modifications and changes to the specific examples illustrated above. The technical elements described in this specification or drawings exhibit technical usefulness individually or in various combinations, and are not limited to the combinations described in the claims at the time of filing. Furthermore, the technologies illustrated in this specification or drawings achieve multiple objectives simultaneously, and achieving even one of these objectives constitutes technical usefulness. [Explanation of Symbols]

[0036] 10: Semiconductor wafer, 14: Device area, 16: Scribe line, 30: Alignment mark, 40a~40e: Image range

Claims

1. A method for visual inspection of a semiconductor wafer (10) having a plurality of element regions (14), The steps include forming a plurality of alignment marks (30) on the semiconductor wafer, The steps include setting each shooting range (40) such that each shooting range includes at least one of the plurality of alignment marks and each shooting range divides each element area into a plurality, and taking an image of each shooting range, The process involves performing a step of identifying defects from the image of the target shooting range by comparing the image of the target shooting range (40bg) with another image (40bg-r) of a shooting range corresponding to the target shooting range, for each of the aforementioned shooting ranges. It has, The corresponding imaging range is an imaging range set in another element region (14b) of the same semiconductor wafer as the element region (14a) in which the target imaging range is set. Visual inspection method.

2. A method for visual inspection of a semiconductor wafer (10) having an element region (14), The steps include forming a plurality of alignment marks (30) on the semiconductor wafer, The steps include setting each shooting range (40) such that each shooting range includes at least one of the plurality of alignment marks and each shooting range divides the element area into a plurality of sections, and taking an image of each shooting range, A machining step for processing the element region, It has, The step of capturing the images of each of the aforementioned shooting ranges is performed both before and after the processing step. Visual inspection method.

3. A method for visual inspection of a semiconductor wafer (10) having multiple element regions (14), The steps include forming a plurality of alignment marks (30) on the semiconductor wafer, The steps include setting each shooting range (40) such that each shooting range includes at least one of the plurality of alignment marks and each shooting range divides each element area into a plurality, and taking an image of each shooting range, It has, In the step of forming the plurality of alignment marks, the plurality of alignment marks are formed on the scribe lines (16) provided at the boundaries of the plurality of element regions. Visual inspection method.

4. A method for visual inspection of a semiconductor wafer (10) having an element region (14), The steps include forming a plurality of alignment marks (30) on the semiconductor wafer, The steps include setting each shooting range (40) such that each shooting range includes at least one of the plurality of alignment marks and each shooting range divides the element area into a plurality of sections, and taking an image of each shooting range, It has, In the step of capturing the image of each shooting range, each shooting range is set such that at least two of the plurality of alignment marks are included in each shooting range and the element region is sandwiched between the at least two of the alignment marks within each shooting range. Visual inspection method.

5. A method for manufacturing a semiconductor device, comprising the step of performing the visual inspection method described in any one of Claims 1 to 4.

Images