Wafer wobbling detection method and wafer wobbling detection system

Abstract

To provide a method and its system capable of detecting wobbling of a wafer while an operation is performed in a semiconductor processing device.SOLUTION: A wafer wobbling detection method is applied to a semiconductor processing device that performs a process while an internally mounted wafer is rotated. The method includes: a step of securing a wafer surface image such that a reflection image which is reflecting, on a surface of a wafer, a fixture mounted inside the processing device and having a fixed position is included in the wafer surface image; and a step of detecting wobbling of the wafer by analyzing the reflection image in the secured image, the reflection image being moved due to wobbling.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a method and system for sensing wobbling of a wafer, and more particularly, to a method and system for sensing wobbling of a wafer rotating at high speed in semiconductor equipment.

Background Art

[0002] Generally, many semiconductor equipments used to perform specific processes on semiconductor wafers in the process of manufacturing semiconductors perform the processes with the wafers rotating at high speed. For example, a wafer cleaning equipment operates in a manner that a cleaning liquid is discharged onto the surface of the wafer while the wafer is rotating.

[0003] In this way, a wafer rotating at high speed in semiconductor equipment wobbles due to the influence of vibrations of the equipment or the like. At this time, wobbling within a predetermined range does not affect the process, but if excessive wobbling occurs, problems will occur in the process. Excessive wobbling may be caused by problems occurring in the equipment itself, such as a rotating device, or may also occur because the wafer to be processed is not accurately mounted in the equipment.

[0004] In order to prevent such problems, in the process of performing preventive maintenance or emergency maintenance on semiconductor equipment, a technique has been developed to obtain wobbling information on the wafer and compare it with standard information to determine an operation error.

[0005] However, if excessive wobbling occurs before maintenance, it is difficult to confirm. If a problem occurs in the semiconductor equipment itself, all wafers that have been put into work must be discarded until maintenance is performed or a defect in the wafer is discovered. Also, even if the problem is not with the semiconductor equipment itself but with the wafer mounting process, it may cause a failure of the precision equipment. Therefore, a technology that can detect wafer wobbling in real time during the operation of the semiconductor equipment is required.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0007] The present invention is for solving the problems of the above-described prior art, and its object is to provide a method and a system for sensing wafer wobbling during the operation of semiconductor equipment.

Means for Solving the Problems

[0008] The wafer wobbling sensing method according to the present invention for achieving the above object is a wafer wobbling sensing method applied to semiconductor equipment that performs a process while rotating a wafer installed inside, including: a step of acquiring a surface image of the wafer so that a reflected image reflected from a fixed object installed inside the equipment and having a fixed position is included in the surface of the wafer; and a step of analyzing the reflected image that moves due to wobbling in the acquired image to sense the wobbling of the wafer.

[0009] The process of analyzing the reflected image can be performed by setting a predetermined position of the image as a normal region, setting a first abnormal region arranged to surround the normal region, and calculating the ratio of each region to which the reflected image in the image belongs.

[0010] At this time, an offset region having a predetermined width can be further set between the normal region and the first abnormal region.

[0011] In addition, a second abnormal region arranged to surround the first abnormal region can be further set, upper limits of the ratios in which the reflected image can be included in the first abnormal region and the second abnormal region are set respectively, and when the ratios of the reflected image calculated from the image are higher than the respective upper limits, it can be determined as abnormal wobbling.

[0012] The process of analyzing the reflected image can be performed by setting a closed curve in which pixels are connected at a predetermined position of the image, numbering each pixel included in the closed curve, and then detecting the occurrence direction of wobbling from the numbers of the pixels where the reflected image is mainly located.

[0013] A wafer wobbling sensing system according to another aspect of the present invention is a wafer wobbling sensing system applied to semiconductor equipment that performs a process while rotating a wafer installed inside, including a fixed object installed inside the equipment and having a fixed position, a camera that acquires a surface image of the wafer so that the reflected image of the fixed object reflected on the surface of the wafer is included, and a machine vision analysis unit that analyzes the acquired image to sense the wobbling of the wafer, wherein the machine vision analysis unit senses the wobbling of the wafer by analyzing a reflected image that moves due to wobbling in the acquired image.

[0014] The machine vision analysis unit can set a predetermined position of the image as a normal region, set a first abnormal region arranged to surround the normal region, and calculate the ratio of each region to which the reflected image in the image belongs.

[0015] At this time, the machine vision analysis unit can further set an offset region having a predetermined width between the normal region and the first abnormal region.

[0016] In addition, the machine vision analysis unit can further set a second abnormal region arranged to surround the first abnormal region, and set upper limits of the ratios at which reflected images can be included in the first abnormal region and the second abnormal region respectively. When the ratios of the reflected images calculated from the image are higher than the respective upper limits, it can be determined as abnormal wobbling.

[0017] The machine vision analysis unit can set a closed curve in which pixels are connected at a predetermined position of the image, number each pixel included in the closed curve, and then detect the occurrence direction of wobbling from the numbers of the pixels where the reflected image is mainly located.

Advantages of the Invention

[0018] As configured above, the present invention has the effect of being able to sense the wobbling of a wafer in a state where a process is being performed on the wafer by analyzing an image of the rotating wafer with machine vision.

[0019] In addition, since it is possible to sense the wobbling of the wafer in a state where a process is being performed on the wafer, there is an excellent effect that measures such as stopping the equipment in real time can be taken when abnormal wobbling occurs.

Brief Description of the Drawings

[0020]

Figure 1

Figure 2

Figure 3

Figure 4

Embodiments for Carrying Out the Invention

[0021] Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

[0022] However, the embodiments of the present invention can be modified into various other forms, and the scope of the present invention is not limited to the embodiments described below. The shapes and sizes of elements in the drawings may be exaggerated for clearer explanation, and elements denoted by the same reference numerals in the drawings are the same elements.

[0023] Also, throughout the specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected", but also the case where it is "electrically connected" with another element interposed therebetween. Further, when a part "includes" or "comprises" a certain component, this means that, unless otherwise stated to the contrary, it does not exclude other components, but can further include or comprise other components.

[0024] Also, terms such as "first" and "second" are for distinguishing one component from another and do not limit the scope of rights. For example, the first component can be named the second component, and similarly, the second component can also be named the first component.

[0025] FIG. 1 is a diagram for explaining the structure of a wafer wobbling sensing system according to an embodiment of the present invention.

[0026] The wafer wobbling detection system according to an embodiment of the present invention includes a fixture 100, a camera 200, and a machine vision analysis unit (not shown).

[0027] The object on which the wobbling detection system of this embodiment is installed is all semiconductor equipment in which the wafer rotates during the process, and it is not particularly limited as long as the features of the present invention are not impaired.

[0028] The fixture 100 is a component installed inside the semiconductor equipment, and a reflected image 110 of the fixture 100 reflected on the surface of the wafer 10 photographed by the camera 200 is photographed. If the reflected image 110 of the fixture 100 is confirmed within the photographed range 210 photographed by the camera 200, the position and the like are not particularly limited. Therefore, the fixture 100 may be a structure originally installed inside the equipment, but it is preferably a separately added component to enhance the discrimination ability. In this embodiment, a square sticker is attached to the ceiling of the equipment.

[0029] When using a separately added component as the fixture, the material of the fixture can be selected so that the reflected image is photographed more clearly. For example, a fixture made of a material that absorbs light can be attached. Specifically, by applying a dark material or a material that absorbs incident light by a predetermined amount or more, the reflected image can be configured to be photographed clearly.

[0030] On the other hand, when using a separately added component as the fixture, the color of the fixture and the component to which the fixture is attached may be similar, and it may be difficult to distinguish the reflected image from the component to which the fixture is attached in the photographed image. To prevent such a problem, a structure in which a material that is easy to contrast with the fixture is arranged around the fixture can be applied. By arranging a bright material or a material that can reflect incident light by a predetermined amount or more around the fixture, the reflected image of the fixture can be configured to be photographed clearly.

[0031] The camera 200 is a component that captures an image for sensing wobbling, and captures the image such that at least a part of the rotating wafer 10 is included in the imaging range 210. The camera used in the present invention is not particularly limited as long as it has a resolution capable of analyzing the image by machine vision. In the present invention, the imaging range 210 captured by the camera 200 is important. Specifically, the camera 200 is installed such that the reflected image 110 of the fixed object 100 reflected on the surface of the wafer 10 is included in the imaging range 210.

[0032] At this time, since the movement of the wafer surface due to wobbling increases as the distance from the center of the wafer 10 increases, it is more advantageous for sensing wobbling that the reflected image 110 of the fixed object 100 is closer to the edge of the wafer. Therefore, it is preferable to adjust the relative position and angle between the fixed object 100 and the camera 200 so that the reflected image 110 of the fixed object 100 is located on the edge side of the wafer 10. In particular, the distance of the reflected image 110 from the center of the wafer in the image captured by the camera 200 is preferably 60% or more of the wafer radius, preferably 70% or more of the wafer radius, and more preferably 80% or more of the wafer radius. On the other hand, when the reflected image 110 of the fixed object 100 is excessively on the edge side of the wafer 10, there is a possibility that the problem of the reflected image being out of the wafer region and not being detected due to wobbling may occur. Therefore, the distance of the reflected image 110 from the center of the wafer is preferably 95% or less of the wafer radius.

[0033] The machine vision analysis unit is a component that senses the wobbling of the rotating wafer 10 by analyzing the images captured by the camera. Since the position of the camera 200 is fixed, as the rotating wafer 10 wobbles, the angle of its surface relative to the camera 200 continuously changes. As a result, in the captured image, the reflected image 110 of the fixed object 100 reflected on the surface of the wafer 10 is not fixed and can only move. Therefore, by detecting the position and movement of the reflected image 110 in the image of the wafer's surface, the wobbling of the wafer can be sensed. The method by which the machine vision analysis unit detects the position of the reflected image 110 can be applied in various ways, and a method based on pixel detection may be applied.

[0034] On the other hand, in order to improve the wobbling sensing performance, a plurality of fixed objects 100 can be attached and configured such that a plurality of reflected images 110 are captured in one or more captured images. By detecting the position and movement of the reflected images 110 reflected at different positions on the surface of the wafer 10, the wobbling sensing performance can be improved.

[0035] Furthermore, pixel detection can be performed on each of the plurality of reflected images 110, and one or more relatively clearer captured reflected images can be selected and applied to the sensing of wobbling. As will be described below, it is preferable that the outer edge frame of the reflected image is clearly confirmed during the process of sensing wobbling, and the wobbling sensing performance can be improved by selecting a clear reflected image.

[0036] At this time, the process of detecting the position of the reflected image 110 and selecting one or more relatively clearer captured reflected images from among the plurality of reflected images can apply a detection algorithm and a recommendation algorithm based on deep learning.

[0037] In addition, in order to improve the detection performance of the reflected image 110, image processing can be performed on the captured image. Specifically, the reflected image 110 can be detected by converting it to grayscale. At this time, the detection performance of the reflected image 110 can be further improved by reflecting the characteristics of the wafer 10 on the grayscale-converted image. Due to factors such as the thickness of the wafer 10 and the current process step and process method, there are color differences in the color reflected from the surface. Such color differences of the wafer can be more clearly distinguished in the grayscale-converted image, and the detection performance of the reflected image 110 can be further improved by reflecting the characteristics of the wafer in the process of detecting the reflected image 110 reflected on the wafer.

[0038] FIG. 2 is a diagram for explaining a sensing method based on the position of a reflected image in a wafer wobbling sensing system according to an embodiment of the present invention.

[0039] FIG. 2 is a diagram showing an enlarged view of a portion where the reflected image 110 is located in the image captured by the camera 200 in FIG. 1.

[0040] In this embodiment, regions are set so that the machine vision analysis unit can sense and evaluate the wobbling state according to the position of the reflected image. First, the normal region 300 is a region where the reflected image is located when wobbling occurs within a normal range. The first abnormal region 310 is arranged so as to surround the outside of the normal region 300, and the second abnormal region 320 is arranged so as to surround the outside of the first abnormal region 310.

[0041] The machine vision analysis unit calculates the ratio of each region to which the reflected image belongs in real time. At this time, an upper limit is set based on the yield data for the ratio of each region to which the reflected image belongs. When the ratio of the region to which the reflected image belongs exceeds this limit, it can be determined as abnormal wobbling, that is, a device failure or a wafer installation error, and the operation of the semiconductor equipment can be configured to stop.

[0042] FIG. 3 is an illustration of the position of the reflected image in the wafer wobbling sensing system according to an embodiment of the present invention and the resulting yield data.

[0043] As shown in the figure, in one embodiment, when the ratio of the first abnormal region 310 is 15%, the yield is at the 90% level, and when the ratio of the first abnormal region 310 exceeds 15%, it is set to determine that abnormal wobbling has occurred. On the other hand, for the second abnormal region 320, when the ratio exceeds 2%, the yield drops below 89%, and thus, when the ratio of the second abnormal region 320 exceeds 2%, it is set to determine that abnormal wobbling has occurred. In this embodiment, only the yield data is applied to determine abnormal wobbling, but it may also be determined by including time and material costs together with the yield.

[0044] On the other hand, since the surface of the wafer 10 is not excellent in reflection characteristics like a mirror, the reflected image 110 is not only affected by diffuse reflection, but also may be distorted in the reflected image 110 due to the flow of the liquid applied to the surface of the wafer 10 during the process.

[0045] Therefore, in this embodiment, an offset region 301 with a 1 mm interval is set between the normal region 300 and the first abnormal region 310, and the ratio of the first abnormal region 310 is calculated in a range outside the offset region 301. The width of the offset region 301 is not limited to the numerical value of this embodiment and can be changed variously.

[0046] FIG. 4 is a diagram for explaining a sensing method based on the movement of the reflected image in the wafer wobbling sensing system according to an embodiment of the present invention.

[0047] As shown in the figure, a closed curve composed of connected pixels is set, and numbering is given to each pixel included in the closed curve for classification.

[0048] When performing pixel analysis on the position of the reflected image, it is possible to confirm the position where the reflected image passes through the pixels included in the closed curve. From the numbers of the pixels where the reflected image is mainly located, it is possible to know the direction in which the reflected image mainly moves, and by using this, it is possible to detect the occurrence direction of wobbling.

[0049] In this way, by analyzing the position and movement of the reflected image reflected on the surface of the wafer to sense wobbling, it is possible to accurately sense the wobbling of the wafer even while the semiconductor equipment is performing a process.

[0050] As described above, the present invention has been described by way of preferred embodiments. However, the above-described embodiments are merely illustrative explanations of the technical idea of the present invention, and various changes can be made without departing from the technical idea of the present invention, which should be understood by those having ordinary knowledge in the art. Therefore, the protection scope of the present invention should be interpreted by the matters described in the claims rather than specific embodiments, and all technical ideas within the equivalent scope should also be construed as being included in the scope of rights of the present invention.

Explanation of Reference Numerals

[0051] 10 Wafer 100 Fixture 110 Reflected Image 200 Camera 210 Shooting Range 300 Normal Region 301 Offset Region 310 First Abnormal Region 320 Second Abnormal Region

Claims

1. A method for detecting wafer wobbling applied to a semiconductor device that performs a process while rotating a wafer installed therein, acquiring a surface image of the wafer such that the surface image includes a reflection image of a fixed object that is installed in the equipment and fixed in position on the surface of the wafer; and detecting wobbling of the wafer by analyzing a reflected image that moves due to wobbling in the acquired image.

2. analyzing the reflected image, A predetermined position of the image is set as a normal area, and a first abnormal area is set so as to surround the normal area; 2. The method of claim 1, wherein the method is performed by calculating the proportion of each area in the image to which the reflected image belongs.

3. 3. The method of claim 2, further comprising: setting an offset area having a predetermined width between the normal area and the first abnormal area.

4. 3. The method of claim 2, further comprising: setting a second abnormal area surrounding the first abnormal area.

5. 5. The method of claim 4, further comprising: setting upper limits for a ratio of a reflected image that may be included in the first abnormal area and the second abnormal area; and determining that the reflected image is abnormal when the ratio is higher than the respective upper limits of the ratio of the reflected image calculated from the image.

6. analyzing the reflected image, A closed curve is set by connecting pixels at a predetermined position of the image, 2. The method of claim 1, wherein the direction of wobbling is detected from the number of the pixel where the reflected image is mainly located after numbering each pixel included in the closed curve.

7. A wafer wobble detection system applied to semiconductor equipment that performs a process while rotating a wafer installed inside, A fixed object that is installed inside the equipment and has a fixed position; a camera for acquiring a surface image of the wafer so as to include a reflection image of the fixed object reflected on the surface of the wafer; a machine vision analysis unit that analyzes the acquired image to detect wobbling of the wafer; The machine vision analysis unit is A wafer wobbling detection system, characterized in that the wobbling of the wafer is detected by analyzing a reflected image that moves due to the wobbling in an acquired image.

8. The machine vision analysis unit is 8. The wafer wobble sensing system of claim 7, further comprising: setting a predetermined position of the image as a normal area; setting a first abnormal area arranged to surround the normal area; and calculating the proportion of each area to which the reflected image in the image belongs.

9. The machine vision analysis unit is 9. The wafer wobble sensing system of claim 8, further comprising an offset region having a predetermined width between the normal region and the first abnormal region.

10. 9. The wafer wobble detection system of claim 8, wherein the machine vision analysis unit further defines a second abnormal area disposed to surround the first abnormal area.

11. The machine vision analysis unit is 11. The wafer wobble detection system of claim 10, further comprising: setting upper limits for the proportion of the reflected image that may be included in the first abnormal area and the second abnormal area, and determining that abnormal wobbling occurs when the proportion of the reflected image calculated from the image is higher than the respective upper limits.

12. The machine vision analysis unit is A closed curve is set by connecting pixels at a predetermined position of the image, 8. The wafer wobble sensing system according to claim 7, wherein the direction of wobble can be detected from the number of the pixel where the reflected image is mainly located after numbering each pixel included in the closed curve.

Images