Semiconductor wafer visual inspection method, visual inspection apparatus, and judgment program

Abstract

This invention provides a semiconductor wafer visual inspection method that can determine the pass / fail status of the edge shape of a semiconductor wafer at high speed and with high accuracy. [Solution] The semiconductor wafer appearance inspection method according to the present invention is a semiconductor wafer appearance inspection method using an appearance inspection apparatus 1 comprising an imaging device 11 for photographing the bevel portion, which is the edge portion of the outer circumference of the wafer, and a control device 12 having a control unit 21 that determines whether the edge shape is acceptable or unacceptable using the image data of the bevel portion photographed by the imaging device 11. In the control device 12, the control unit 21 performs a region setting step of setting an ROI that includes the wafer edge forming the bevel portion in the image data, and a determination step of determining whether the edge shape is acceptable or unacceptable using brightness information in the ROI and contour information relating to the contour of the semiconductor wafer formed within the ROI.

Description

[Technical Field]

[0001] The present invention relates to a method for visual inspection of the edge portion (bevel portion) of a semiconductor wafer, a visual inspection apparatus, and a judgment program. [Background technology]

[0002] The shape of the bevel portion of a semiconductor wafer is one of the important conditions that must be met to improve the manufacturing yield of devices, and it is standardized that it should be processed into a predetermined shape. Therefore, it is desirable to accurately measure whether the bevel portion of a semiconductor wafer is formed into the predetermined shape.

[0003] Patent Document 1 discloses a measuring device and a measuring method that can measure the edge shape of a semiconductor wafer with high precision. In this measuring device and method, the relative position of a light source irradiating the edge portion of the semiconductor wafer is changed, and the shape of the edge portion is measured based on the position where the intensity of the reflected light from the edge portion peaks. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2020-20717 [Overview of the project] [Problems that the invention aims to solve]

[0005] However, the measuring device described in Patent Document 1 requires complex structures such as optical systems and moving mechanisms, and the optical head must be moved in θ-steps relative to the edge of the semiconductor wafer. This poses a problem in that it takes a considerable amount of time to accurately determine whether the semiconductor wafer edge shape is acceptable or not. Furthermore, the complexity of the measuring device's structure results in significant costs for its introduction.

[0006] The present invention has been made in view of the above problems, and aims to provide a semiconductor wafer visual inspection method, visual inspection apparatus, and judgment program that can determine the pass / fail status of the edge shape of a semiconductor wafer at high speed, with high accuracy, and at low cost. [Means for solving the problem]

[0007] The present invention relates to a semiconductor wafer appearance inspection method, which is an appearance inspection method using an appearance inspection apparatus comprising: an imaging device for photographing the bevel portion, which is the edge portion of the outer periphery of the wafer; and a control device having a control unit that determines whether the edge shape is acceptable or unacceptable using the image data of the bevel portion photographed by the imaging device.

[0008] In the control device for this visual inspection apparatus, the control unit is characterized by performing a region setting step of setting an image processing region in the image data that includes the wafer edge forming the bevel portion, and a determination step of determining whether the edge shape is acceptable or unacceptable using brightness information in the image processing region and contour information relating to the contour of the semiconductor wafer formed within the image processing region.

[0009] More specifically, the determination step includes: a brightness information calculation step of calculating the average brightness within the image processing area as a first characteristic value that serves as an index for determining whether the edge shape is acceptable or unacceptable; a contour information calculation step of extracting the contour of a semiconductor wafer from image data within the image processing area, calculating a first function that is a function of the contour perpendicular to the wafer surface, then calculating a second function that is the derivative of the first function, and calculating the coefficient of determination when the second function is fitted to an approximate function as a second characteristic value that serves as an index; and a pass / fail determination step of determining whether the edge shape of the outer periphery of the wafer is acceptable or unacceptable based on the first characteristic value and the second characteristic value.

[0010] In the contour information calculation step, in the calculation of the coefficient of determination, a linear equation and a cubic equation are used as approximation functions, and the first coefficient of determination when approximating the second function with the linear equation and the second coefficient of determination when approximating the second function with the cubic equation are calculated, and the larger of these values is selected as the coefficient of determination for judgment.

[0011] In the method for inspecting the appearance of a semiconductor wafer according to the present invention, it is preferable to preset a boundary line for distinguishing between pass and fail on a plane formed by the range of possible values of the first characteristic value and the range of possible values of the second characteristic value. Thereby, in the pass / fail determination step, it is possible to determine pass / fail at a position on the plane specified by the first characteristic value and the second characteristic value.

[0012] In the method for inspecting the appearance of a semiconductor wafer according to the present invention, the length in the vertical direction of the image processing area preferably includes both axial ends of the wafer end, and the length in the horizontal direction of the image processing area is desirably shorter than the length in the vertical direction and includes a sufficient wafer area. Also, the horizontal range of the image processing area is desirably a region extending from the most protruding portion in the radial direction of the wafer end toward the wafer center, or a region extending from the outside of the wafer outer circumference toward the wafer center so as to include the most protruding portion of the wafer end.

[0013] In the method for inspecting the appearance of a semiconductor wafer according to the present invention, it is desirable that the imaging device captures the shadow of the semiconductor wafer projected by light irradiated from above the extension in the radial direction of the semiconductor wafer.

[0014] The appearance inspection apparatus according to the present invention includes an imaging device that captures an image of a bevel portion, which is an edge portion of the outer circumference of a wafer, and a control device that determines whether the edge shape is acceptable or not using the image data of the bevel portion captured by the imaging device. The control device sets an image processing region including the wafer end portion that forms the bevel portion in the image data, calculates the average luminance in the image processing region as a first characteristic value serving as an index for determining whether the edge shape is acceptable or not, extracts the contour of the semiconductor wafer from the image data in the image processing region, calculates a first function that is a function in the wafer surface normal direction of the contour, then calculates a second function that is the derivative of the first function, calculates the coefficient of determination when the second function is fitted to an approximation function as a second characteristic value serving as the index, and determines whether the edge shape of the outer circumference of the wafer is acceptable or not based on the first characteristic value and the second characteristic value.

[0015] In the appearance inspection apparatus according to the present invention, it is desirable that the imaging device captures the shadow of the semiconductor wafer projected by light irradiated from the extension in the radial direction of the semiconductor wafer.

[0016] The determination program according to the present invention is a determination program for a control unit of a control device that receives image data of a bevel portion, which is an edge portion of the outer circumference of a wafer, to determine whether the edge shape is acceptable or not, and causes a computer operating as the control unit of the control device to execute a series of processes according to the above steps.

[0017] According to the appearance inspection method, appearance inspection apparatus, and determination program for a semiconductor wafer according to the present invention, since the edge shape of the semiconductor wafer can be quantitatively evaluated and its acceptability can be automatically determined, it is possible to remove the subjectivity of the inspector from the acceptability determination and achieve stabilization of inspection quality. That is, it is possible to determine whether the edge shape of the semiconductor wafer is acceptable or not at high speed and with high accuracy. In addition, since the device structure does not become complicated and an expensive device is not required, it is possible to determine whether the edge shape of the semiconductor wafer is acceptable or not at low cost.

Effects of the Invention

[0018] According to the semiconductor wafer visual inspection method, visual inspection apparatus, and judgment program of the present invention, the pass / fail status of the edge shape of a semiconductor wafer can be determined quickly, with high accuracy, and at low cost. [Brief explanation of the drawing]

[0019] [Figure 1] Figure 1 shows an embodiment of the visual inspection apparatus according to the present invention. [Figure 2] Figure 2 shows an example of the shape of the bevel portion of a semiconductor wafer. [Figure 3] Figure 3 shows an example of the shape of the bevel portion of a semiconductor wafer. [Figure 4] Figure 4 is a schematic diagram showing an example of an image processing area. [Figure 5] Figure 5 is a flowchart showing an example of shape determination processing by a control device. [Figure 6] Figure 6 schematically shows the difference in average brightness within the image processing area due to the shape of the wafer edge. [Figure 7] Figure 7 schematically shows an image of the function of the contour perpendicular to the wafer surface extracted from the image data. [Figure 8] Figure 8 shows the image data and its function approximations using linear and cubic equations. [Figure 9] Figure 9 shows the results of the judgment program plotted on a plane. [Modes for carrying out the invention]

[0020] Hereinafter, embodiments of the semiconductor wafer visual inspection method, visual inspection apparatus, and judgment program according to the present invention will be described in detail with reference to the drawings. However, the present invention is not limited to these embodiments. Furthermore, in the specification and drawings of this application, elements that can be similarly described are denoted by the same reference numerals, thereby omitting redundant explanations.

[0021] <Visual Inspection Equipment> Figure 1 shows an embodiment of the visual inspection apparatus according to the present invention. Figures 2 and 3 show examples of the shape of the bevel portion of a semiconductor wafer, and Figure 4 is a schematic diagram showing an example of an image processing region (ROI: Region of Interest).

[0022] The appearance inspection apparatus 1 of this embodiment includes an imaging device 11 that photographs the bevel portion 2a, which is the edge portion of the outer circumference of the wafer, and a control device 12 that sets a predetermined ROI 2b in the image data of the bevel portion 2a photographed by the imaging device 11, and determines whether the edge shape is acceptable or not using brightness information in the ROI 2b and information regarding the contour of the semiconductor wafer 2 (hereinafter sometimes referred to as wafer 2) formed within the ROI 2b (contour information).

[0023] As an example of the shape of the bevel portion 2a of the object being photographed, Figure 2 shows the shape of the bevel portion 2a in a normal wafer 2, and Figure 3 shows the shape of the bevel portion 2a in a defective wafer 2. In these figures, the bevel portion 2a on the outer circumference of the wafer is the part where beveling has been performed, and has a tapered portion 2c that narrows at a certain angle of inclination, and a wafer edge 2d that is connected to the tip of the tapered portion 2c. In a normal wafer 2, the wafer edge 2d has a smooth shape from one tapered portion 2c to the other, but in a defective wafer 2, this shape is distorted and an uneven surface is formed (defective shape).

[0024] The wafer edge 2d is defined in the shape (image data) shown in Figures 2 and 3 as the area between two points where two straight lines m, which are tangent to the tapered portion 2c at the above-mentioned inclination angle, intersect with a straight line n in the wafer axial direction that is tangent to the wafer edge 2d.

[0025] As described above, the shape defects of the bevel portion 2a can cause chipping (such as chipping or cracking) during the semiconductor wafer manufacturing and semiconductor device manufacturing processes. The fragments generated by chipping can damage other wafers, leading to a deterioration in the quality of semiconductor wafers and semiconductor devices. Therefore, defective wafers 2 as shown in Figure 3 must be removed as soon as possible.

[0026] Therefore, in the visual inspection apparatus 1 of this embodiment, the control device 12 uses image data of the bevel portion 2a (edge ​​shape) captured by the imaging device 11 to analyze the edge shape of the wafer 2 with high precision and quantitatively determine whether the edge shape is acceptable or not. This makes it possible to remove defective wafers 2 at an early stage after bevel processing.

[0027] In Figure 1, the imaging device 11 captures the shape of the bevel portion 2a, which is the edge portion of the outer circumference of the wafer, and transmits the image data to the control device 12. During imaging, light is shone from above the radial extension of the wafer 2, and the imaging device 11 captures the projected shadow of the wafer 2. The image data is output as binarized data of white (8-bit Gray Scale: 255) and black (8-bit Gray Scale: 0). That is, the wafer portion is black, and the parts other than the wafer are white.

[0028] The control device 12 sets an ROI 2b in the image data of the bevel portion 2a captured by the imaging device 11, and determines whether the edge shape is acceptable or not using the average brightness (brightness information) in the ROI 2b and a function of the contour of the wafer 2 formed within the ROI 2b in the direction perpendicular to the surface (contour information).

[0029] Figure 4 schematically shows an example of ROI2b, where ROI2b is indicated as the shaded area in the figure. The vertical length (first length) of ROI2b preferably includes both ends of the wafer edge 2d in the axial direction, for example, it is preferable to specify the range of both ends of the wafer edge 2d in the axial direction. The horizontal length (second length) of ROI2b preferably is shorter than the first length and sufficiently includes the wafer region, but any length that sufficiently includes the wafer region may be set as an analysis constant.

[0030] Furthermore, the lateral range of ROI2b is defined as the region (second length) extending from the radially protruding part of the wafer edge 2d towards the wafer center, but ROI2b is not limited to this. For example, as shown in Figure 4, the left and right edges may be shifted so as to include the most protruding part of the wafer edge 2d, thereby including a region outside the outer edge of the wafer. In other words, the lateral range of ROI2b may be the region extending from outside the outer edge of the wafer towards the wafer center, so as to include the most protruding part of the wafer edge 2d. Alternatively, it may be the region extending from inside the outer edge of the wafer towards the wafer center. The length and position of ROI2b when setting it are based on the captured image, and the unit is pixels (px).

[0031] The semiconductor wafer used in this embodiment may be a wafer on which a device has been formed. Furthermore, the material (raw material) of the semiconductor wafer used in this embodiment is not particularly limited to silicon (Si), but may be, for example, germanium (Ge), selenium (Se), gallium arsenide (GaAs), silicon carbide (SiC), indium phosphide (InP), gallium nitride (GaN), synthetic diamond (C), etc.

[0032] <Control device detailed configuration> Here, the configuration of the control device 12 will be described in more detail.

[0033] In Figure 1, the area enclosed by the thick line shows an example of the hardware configuration of a computer operating as the control device 12 according to the present invention. The control device 12 in this embodiment operates as a host computer that determines whether the edge shape of the outer circumference of the wafer is acceptable or not when it receives image data of the bevel portion 2a captured by the imaging device 11.

[0034] In Figure 1, the control device 12 comprises a control unit 21 composed of an integrated circuit such as a CPU (Central Processing Unit), a storage unit 22 including various memories, an input unit 23 including a user interface such as a keyboard and mouse, an interface unit (I / F unit) 24 that processes input / output data such as image data from the imaging device 11 and print and scan data, a display unit 25 which is a display, and a communication unit 26 that communicates with the outside via a predetermined network. In Figure 1, the control device 12 is shown to include an input unit 23 including a user interface such as a keyboard and mouse, but the control device 12 of this embodiment is not limited to this, and the input unit 23 may be omitted by giving the display unit 25 a touch panel function, or the input unit 23 and the display unit 25 may be used in combination.

[0035] In Figure 1, the control unit 21 executes a "judgment program" to implement the process (shape judgment process) performed by the control device 12 of this embodiment to determine whether the edge shape of the outer circumference of the wafer is pass or fail. The storage unit 22 stores the judgment program of this embodiment, various information (image data, ROI setting information, approximation function, selection conditions, boundary conditions, etc.), and various data obtained during the processing (ROI, brightness information, contour information, pass / fail judgment result, etc.). The control unit 21 executes the shape judgment process of this embodiment by reading the judgment program stored in the storage unit 22.

[0036] Furthermore, the storage unit 22 is not limited to internal memory; for example, it may be an external storage medium such as a DVD (Digital Versatile Disc) or SD memory, or it may consist of both internal memory and an external storage medium (DVD, SD memory, etc.). Also, for the sake of explanation, the hardware configuration of the control device 12 in this embodiment is a list of the configurations related to the shape determination process and does not represent all the functions of the computer that constitutes the control device 12.

[0037] Furthermore, while the control device 12 in this embodiment is assumed to be a general-purpose PC such as a desktop computer or a notebook computer, it is not limited to these. For example, it may be a mobile device such as a smartphone or tablet, or it may be a dedicated PC manufactured for use in a visual inspection device.

[0038] <Visual Inspection Method (Shape Determination Process)> Next, the shape determination process in the visual inspection device 1 (control device 12) of this embodiment will be described in detail. Figure 5 is a flowchart showing an example of the shape determination process (determination program) by the control device 12.

[0039] When the control unit 21 receives image data of the image captured by the imaging device 11 via the I / F unit 24, it stores the acquired image data in the storage unit 22 and extracts from this image data the length in the wafer axis direction, position (contour), and position of the most protruding part in the wafer radial direction as information about the wafer edge 2d (step S1). The extracted information is then stored in the storage unit 22.

[0040] Next, the control unit 21 sets ROI2b in the acquired image data based on the information extracted in step S1, the width (second length) of the image processing region (ROI), and the offset value of the lateral position of the ROI (step S2). Then, the set ROI2b is stored in the storage unit 22. It is assumed that the width and offset value are set in advance before starting the shape determination process. In this embodiment, as shown in Figure 4, ROI2b is set as a rectangle including the wafer edge 2d.

[0041] Next, the control unit 21 calculates the average brightness L within the ROI 2b as one of the indicators (first characteristic value) for determining whether the edge shape of the outer circumference of the wafer is acceptable (step S3). Then, the calculated average brightness L is stored in the storage unit 22.

[0042] Figure 6 schematically illustrates the difference in average brightness L within ROI2b depending on the shape of the wafer edge 2d, where (a) schematically shows an image of the bevel portion 2a of a normal wafer, and (b) schematically shows an image of the bevel portion 2a of a defective wafer. For example, if the left edge of ROI2b is the radially protruding part of the wafer edge 2d, the image within ROI2b of a normal wafer 2 will be mostly wafer portion (black: brightness 0), as shown in Figure 6(a). Therefore, a normal wafer 2 will have an average brightness L within ROI2b close to 0. On the other hand, the image within ROI2b of a defective wafer 2 will be such that most of the wafer edge 2d is located away from the left edge of ROI2b, as shown in Figure 6(b), and there will be a larger area of ​​non-wafer portion (white: brightness 255) compared to a normal wafer 2. Therefore, a defective wafer 2 tends to have a higher average brightness L within ROI2b. In other words, the higher the average brightness L within ROI2b, the higher the probability that the wafer is defective. Therefore, in this embodiment, the average brightness L within ROI2b is used as one of the indicators (first characteristic value) for determining whether the edge shape of the outer circumference of the wafer is acceptable or not.

[0043] In parallel with the calculation of the average brightness L, the control unit 21 extracts the contour of wafer 2 (wafer edge 2d) from the image data within ROI 2b and calculates the function y perpendicular to the wafer surface of this contour (step S4). Subsequently, it calculates the derivative y' of the function y (step S5), and the coefficient of determination R when this derivative y' is fitted to the approximation function. 2 Calculate (Step S6). Then, calculate the function y, derivative y', and coefficient of determination R. 2 The data is stored in the memory unit 22. In this embodiment, the function y, the derivative y', and the coefficient of determination R are stored. 2 This is sometimes referred to as contour information.

[0044] Figure 7 schematically shows the image of the function y perpendicular to the wafer surface of the contour extracted from the image data. As shown in Figure 7, this function y is represented as a one-dimensional array where the index is the number of pixels from the top to the bottom edge of ROI2b, and the value is the number of pixels from the left edge of ROI2b to the contour of wafer 2 (contour of wafer edge 2d). If there are no abnormalities in the shape of wafer edge 2d, the function y becomes a relatively simple function, and its derivative y' is similar. Therefore, the coefficient of determination R when the derivative y' when there are no abnormalities in the shape of wafer edge 2d is fitted with an approximation function is 2 The coefficient of determination R tends to be close to 1. Conversely, if there is an abnormality in the shape of the wafer edge 2d, the function y becomes a complex curve and cannot be approximated by a simple function, and the coefficient of determination R 2 The coefficient of determination R tends to be small. Therefore, in this embodiment, taking these tendencies into consideration, 2 This was used as the second indicator (second characteristic value) for determining whether the edge shape of the wafer's outer circumference was acceptable or not.

[0045] Furthermore, in this embodiment, a linear function (linear equation) and a cubic function (cubic equation) are used as approximation functions. Therefore, in step S6 above, the derivative y', the linear equation and the cubic equation are input to the control unit 21, and the first coefficient of determination R when the derivative y' is approximated by the linear equation is 2 And the second coefficient of determination R when the derivative y' is approximated by a cubic equation. 2 And, we seek.

[0046] Next, the control unit 21 reads the first determination coefficient R 2 and the second determination coefficient R 2 from the storage unit 22, and selects the one with the larger value as the determination coefficient R 2 for determination (step S7). Then, the selected determination coefficient R 2 is stored in the storage unit 22.

[0047] As described above, in the present embodiment, the larger value (selection condition) between the first determination coefficient R 2 and the second determination coefficient R 2 when a good approximation can be made is selected as the determination coefficient R 2 for determination. At this time, the approximation by the linear equation is regarded as being able to make a good approximation without particular limitation. On the other hand, the cubic equation is limited to those without extreme values, and the approximation by the cubic equation is made under the condition that the inflection point does not deviate from a certain range centered on the assumed initial value. This is a measure to prevent the determination coefficient R 2 from becoming high as the degree of the polynomial increases and misjudging a defective wafer as qualified.

[0048] Finally, the control unit 21 reads the average luminance L (the first characteristic value) and the determination coefficient R 2 (the second characteristic value) from the storage unit 22, determines the pass / fail of the edge shape of the wafer outer periphery based on the predetermined boundary conditions, the first characteristic value, and the second characteristic value, and displays the pass / fail determination result on the display unit 25 (step S8). Then, the pass / fail determination result is stored in the storage unit 22.

[0049] In this embodiment, the pass / fail determination is made based on the position on a plane formed by the range of possible values ​​of the first characteristic (e.g., the horizontal axis) and the range of possible values ​​of the second characteristic (e.g., the vertical axis). In this case, the plane formed by the two characteristics is divided into a pass region and a fail region by a boundary line. The conditions of this boundary line (boundary conditions) are predetermined as described above and can be set arbitrarily according to the operation. For example, points (positions) specified by the two characteristic values ​​are plotted on the plane, and the determination is made based on whether the plotted position belongs to the pass or fail region.

[0050] In this embodiment, the derivative y' is fitted with an approximation function as an example, but this is not the only option; the function y may also be fitted with an approximation function. Furthermore, the approximation function used when fitting the derivative y' or the function y may consist of three or more functions, and may not be a polynomial.

[0051] <Effects, etc.>

[0052] As described above, the semiconductor wafer appearance inspection method of this embodiment is a semiconductor wafer appearance inspection method using an appearance inspection apparatus 1 that includes an imaging device 11 for photographing the bevel portion 2a, which is the edge portion of the outer circumference of the wafer, and a control device 12 having a control unit 21 that determines whether the edge shape is acceptable or not using the image data of the bevel portion 2a photographed by the imaging device 11.

[0053] In the control device 12 of this visual inspection apparatus 1, the control unit 21 performs a region setting step (steps S1 and S2) in which it sets an image processing region (ROI 2b) that includes the wafer edge 2d forming the bevel portion 2a in the image data, and a determination step (steps S3 to S8) in which it determines whether the edge shape is acceptable or not using brightness information in the ROI 2b and contour information relating to the contour of the wafer 2 (wafer edge 2d) formed within the ROI 2b.

[0054] In detail, the determination step includes a luminance information calculation step (step S3) which calculates the average luminance L within ROI2b as a first characteristic value that serves as an index for determining whether the edge shape is acceptable or not, a step S4 which extracts the contour of wafer 2 from image data within ROI2b and calculates a first function which is a function y of the contour perpendicular to the wafer surface, and then calculates a second function which is the derivative y' of the first function (step S5), and the coefficient of determination R when the second function is fitted to the approximation function as a second characteristic value that serves as the index. 2 The process includes a contour information calculation step (steps S6 and S7) which calculates the first characteristic value and the second characteristic value, and a pass / fail determination step (step S8) which determines whether the edge shape of the outer circumference of the wafer is pass / fail based on the first characteristic value and the second characteristic value.

[0055] Then, in the contour information calculation step, the coefficient of determination R 2 In calculating the first coefficient of determination R, a linear and a cubic equation are used as approximation functions, and the first coefficient of determination R is obtained when the second function is approximated by a linear equation. 2 And the second coefficient of determination R when the second function is approximated by a cubic equation. 2 The coefficient of determination R is calculated, and the larger of these values ​​is used as the determination coefficient R. 2 Select as (Step S7).

[0056] In the semiconductor wafer visual inspection method of this embodiment, it is preferable to pre-set a boundary line for determining pass or fail on the plane formed by the possible ranges of the first characteristic value and the possible ranges of the second characteristic value. This allows the pass or fail determination step to be made at a position on the plane specified by the first characteristic value and the second characteristic value.

[0057] Furthermore, in the semiconductor wafer visual inspection method of this embodiment, it is desirable that the length of ROI2b in the vertical direction includes both ends of the wafer edge 2d in the axial direction, and that the length of ROI2b in the horizontal direction is shorter than the length in the vertical direction and sufficiently includes the wafer area. Moreover, it is desirable that the horizontal range of ROI2b be a region that extends toward the wafer center with the radially most protruding part of the wafer edge 2d as the base, or a region that extends toward the wafer center from outside the outer edge of the wafer so as to include the radially most protruding part of the wafer edge 2d.

[0058] Furthermore, in the semiconductor wafer visual inspection method of this embodiment, it is desirable that the imaging device 11 captures the shadow of the wafer 2 projected by light irradiated from the radial extension of the wafer 2.

[0059] According to the semiconductor wafer visual inspection method and visual inspection apparatus of this embodiment, the edge shape of the semiconductor wafer can be quantitatively evaluated and its pass / fail status can be automatically determined. This eliminates the subjectivity of the inspector from the pass / fail judgment, thereby stabilizing inspection quality. In other words, the pass / fail status of the semiconductor wafer edge shape can be determined quickly and with high accuracy. Furthermore, since the apparatus structure is not complex and expensive equipment is not required, the pass / fail status of the semiconductor wafer edge shape can be determined at low cost.

[0060] It should be noted that the present invention is not limited to the embodiments described above. The embodiments described above are illustrative, and any configuration that has substantially the same technical idea as described in the claims and produces similar effects is included within the technical scope of the present invention. [Examples]

[0061] Next, the method for visual inspection of semiconductor wafers according to the present invention will be further described based on examples. However, the present invention is not limited to the following examples.

[0062] <Example 1> In Example 1, image data of a wafer known to have a normal bevel shape (normal wafer) was input to a computer acting as a control device, and the above-described determination program (shape determination process: steps S1 to S8) was executed using this computer. At this time, the settings for the computer were the width of the ROI (=100 μm), the lateral offset value of the ROI (=0: no offset (the most protruding part is the base)), and the determination coefficient R for determination. 2 The selection criteria and boundary conditions for determining pass / fail were pre-set to suit the operational requirements. In addition to selecting the larger value, the selection criteria included that the cubic equation must not have extrema and that there must be an acceptable range for the deviation of the inflection point when approximating with a cubic equation. The boundary conditions were "L<32" or "R 2 A score of >0.9 was considered a "pass".

[0063] <Example 2> In Example 2, image data of a wafer known to have an abnormally shaped bevel (a defective wafer) was input to a computer acting as a control device, and the judgment program described above (shape determination process: steps S1 to S8) was executed using this computer. The settings for the computer were the same as in Example 1.

[0064] <<Pass / Fail Judgment>> Figure 8 shows the image data and the linear and cubic function approximations for each embodiment. (a) shows the image data for Embodiment 1, (b) shows the linear approximation for Embodiment 1, (c) shows the cubic approximation for Embodiment 1, (d) shows the image data for Embodiment 2, (e) shows the linear approximation for Embodiment 2, and (f) shows the cubic approximation for Embodiment 2. In Figures 8(b), (c), (e), and (f), the vertical axis Y represents the derivative y'[px / px] and the horizontal axis X represents the index[px].

[0065] In Example 1 (normal wafer), the average brightness L (first characteristic value) within the ROI was 39.305 (8-bit Grayscale). Furthermore, in Example 1, the first coefficient of determination R when the derivative y' was approximated by a linear equation was...2 R is 0.9871, and is the second coefficient of determination when the derivative y' is approximated by a cubic polynomial. 2 The value was 0.9480. As a result, the first coefficient of determination R, which has a large value, was 2 The coefficient of determination R is used for the determination. 2 This was selected as the (second characteristic value). Subsequently, based on the predetermined boundary conditions and the obtained first and second characteristic values, the pass / fail status of the wafer's outer edge shape was determined, and the pass / fail result was displayed on the screen. The pass / fail result was "Pass," indicating that the correct result was obtained.

[0066] In Example 2 (defective wafer), the average brightness L (first characteristic value) within the ROI was 66.902 (8-bit Grayscale). Also, in Example 2, the first coefficient of determination R when the derivative y' was approximated by a linear equation was... 2 R is 0.5860, and is the second coefficient of determination when the derivative y' is approximated by a cubic polynomial. 2 The value was 0.7491. Subsequently, based on predetermined boundary conditions and the obtained first and second characteristic values, the pass / fail status of the wafer's outer edge shape was determined, and the pass / fail result was displayed on the screen. The pass / fail result was "fail," and the correct result was obtained.

[0067] In Example 2, the coefficient of determination R 2 The coefficient of determination R differs significantly between the linear approximation (0.5860) and the cubic approximation (0.7491). Numerically, the cubic approximation shows better agreement, but the inflection point of the cubic (approximation function) is far from the center of the Profile, so it does not satisfy the selection criteria. 2 The (second characteristic value) is the first coefficient of determination R 2 (=0.5860) is selected. This prevents defective wafers from being mistakenly classified as acceptable.

[0068] <Example 3> In Example 3, a judgment program (shape judgment process: steps S1 to S8) was executed on 100 wafers whose edge shape had been previously judged as pass or fail by an inspector, using a computer acting as a control device.

[0069] Figure 9 shows the results of the judgment program plotted on a plane, with the horizontal axis representing the average brightness L and the vertical axis representing the coefficient of determination R. 2 This represents the following. In the figure, the plots marked with ○ represent wafers that the inspector judged as "pass," and the plots marked with × represent wafers that the judge judged as "fail." The boundary conditions are "L>32" and "R 2 A score of <0.9 was considered a "failure".

[0070] As a result, it was found that the results were almost identical to those of the inspectors' pass / fail judgments. In addition, there were some wafers that the inspectors judged as "pass" but this judgment program judged as "fail," but it was found that the program was able to detect abnormal shapes that could not be identified by visual inspection by the inspectors, indicating that the wafer edge shape can be judged with higher accuracy. [Explanation of symbols]

[0071] 1. Visual inspection device 2. Semiconductor wafer (wafer) 2a Bevel section 2b Image Processing Area (ROI) 2c Tapered section 2D wafer edge 11 Imaging device 12 Control device 21 Control Unit 22 Memory section 23 Input section 24 Interface section (I / F section) 25 Display section 26 Communications Department

Claims

1. A method for inspecting the appearance of a semiconductor wafer using an appearance inspection apparatus comprising: an imaging device for photographing the bevel portion, which is the edge portion of the outer circumference of the wafer; and a control device having a control unit that determines whether the edge shape is acceptable or unacceptable using the image data of the bevel portion photographed by the imaging device; The control unit, A region setting step in which an image processing region including the wafer edge forming the bevel portion is set in the image data, A determination step of determining whether the edge shape is acceptable or unacceptable using brightness information in the image processing region and contour information relating to the contour of the semiconductor wafer formed within the image processing region. Execute A method for visual inspection of a semiconductor wafer, characterized by the following features.

2. The aforementioned determination step is, A luminance information calculation step, which calculates the average luminance within the image processing area as a first characteristic value that serves as an indicator for determining whether the edge shape is acceptable or unacceptable, A contour information calculation step comprising: extracting the contour of a semiconductor wafer from image data within the image processing area; calculating a first function which is a function of the contour perpendicular to the wafer surface; then calculating a second function which is the derivative of the first function; and calculating the coefficient of determination when the second function is fitted to an approximation function as a second characteristic value that serves as an index; A pass / fail determination step in which the edge shape of the outer circumference of the wafer is determined to be acceptable or unacceptable based on the first characteristic value and the second characteristic value, including, The method for visual inspection of a semiconductor wafer according to feature 1.

3. In the contour information calculation step, when calculating the coefficient of determination, a linear equation and a cubic equation are used as approximation functions, and a first coefficient of determination is calculated when the second function is approximated by the linear equation, and a second coefficient of determination is calculated when the second function is approximated by the cubic equation. The larger of these two values ​​is selected as the coefficient of determination for determination. The method for visual inspection of a semiconductor wafer according to feature 2.

4. A boundary line for determining pass or fail is predetermined on the plane formed by the range of possible values ​​for the first characteristic and the range of possible values ​​for the second characteristic. In the pass / fail determination step, pass or fail is determined at the position on the plane specified by the first characteristic value and the second characteristic value. The method for visual inspection of a semiconductor wafer according to feature 2.

5. The vertical length of the image processing area includes both ends of the wafer edge in the axial direction, and the horizontal length of the image processing area is shorter than the vertical length and sufficiently includes the wafer area. The method for visual inspection of a semiconductor wafer according to feature 2.

6. The lateral range of the image processing area is defined as the region extending towards the wafer center, with the radially protruding portion of the wafer edge as the base. The method for visual inspection of a semiconductor wafer according to feature 5.

7. The lateral range of the image processing area is defined as a region extending from the outer edge of the wafer toward the wafer center, such that the most protruding part of the wafer edge is included. The method for visual inspection of a semiconductor wafer according to feature 5.

8. The imaging device captures the shadow of the semiconductor wafer projected by light irradiated from the radial extension of the semiconductor wafer. The method for visual inspection of a semiconductor wafer according to feature 1.

9. An imaging device for photographing the bevel portion, which is the edge portion of the outer circumference of the wafer, A control device that determines whether the edge shape is acceptable or unacceptable using image data of the bevel portion captured by the aforementioned imaging device, Equipped with, The control device is In the aforementioned image data, an image processing region including the wafer edge forming the bevel portion is set. As a first characteristic value that serves as an indicator for determining whether the edge shape is acceptable or not, the average brightness within the image processing area is calculated. The contour of the semiconductor wafer is extracted from the image data within the image processing area, a first function is calculated which is a function of the contour perpendicular to the wafer surface, a second function is calculated which is the derivative of the first function, and the coefficient of determination is calculated as the second characteristic value that serves as the index when the second function is fitted to an approximate function. Based on the first characteristic value and the second characteristic value, the pass / fail status of the edge shape of the outer circumference of the wafer is determined. A visual inspection device characterized by the following features.

10. The imaging device captures the shadow of the semiconductor wafer projected by light irradiated from the radial extension of the semiconductor wafer. The visual inspection apparatus according to feature 9.

11. The control unit of the control device, which receives image data of the bevel portion, which is the edge portion of the outer circumference of the wafer, has a determination program for determining whether the edge shape is acceptable or not. The computer that operates as the control unit of the control device, The process is performed according to each step of the semiconductor wafer visual inspection method described in any one of claims 1 to 8. A determination program characterized by the following:

Images