Film cutting method and apparatus

Abstract

The embodiment of the present disclosure provides a film cutting method and device, wherein the film covers a crystal face of a wafer, the wafer has a notch mark, and the film cutting method comprises the following steps: determining position information of the notch mark on the crystal face, the position information being obtained when the crystal face is not covered by the film; obtaining light intensity values at different positions on the film surface when the crystal face is covered by the film; determining a region where the notch mark is covered by the film according to the light intensity values at the different positions on the film surface; and generating a first cutting instruction according to the region where the notch mark is covered by the film and the position information, so as to remove the film on the notch mark. By using the above technical solution, the film on the notch mark can be removed.

Description

Technical Field

[0001] This disclosure relates to the field of thin film cutting technology, and more particularly to a thin film cutting method and apparatus. Background Technology

[0002] In semiconductor manufacturing, there are several processes that treat the back of the wafer, such as wafer back-end polishing and metallization. These processes require a film-coating process to apply a protective film (BG tape) to the wafer's crystal face (the side opposite the back of the wafer) to protect the devices on the wafer's crystal face from damage.

[0003] In the actual film application process, a portion of the protective film needs to be removed to expose the notch.

[0004] Therefore, how to provide a technical solution to remove the protective film on the notch mark has become an urgent technical problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present disclosure provide a film cutting method and apparatus capable of removing the film from a notch mark.

[0006] This disclosure provides a thin-film dicing method, wherein the thin film covers the crystal surface of a wafer, the wafer having notch markings, and the thin-film dicing method includes: The position information of the notch mark on the crystal plane is determined, and the position information is obtained when the crystal plane is not covered by the thin film; When the crystal plane is covered by the thin film, the light intensity values ​​at different positions on the surface of the thin film are obtained; The area covered by the film for the notch mark is determined based on the light intensity value at different locations on the surface of the film. Based on the area covered by the film and the location information of the notch mark, a first cutting command is generated to remove the film located on the notch mark.

[0007] Optionally, determining the position information of the notch mark on the crystal plane includes: When the crystal plane is not covered by the thin film, obtain the coordinate information of different points on the crystal plane, as well as the sampled image containing the notch mark; Based on the coordinate information of different locations, the sampled image is divided into multiple sub-images, such that the overlapping area of ​​at least two sub-images surrounds the notch mark; The feature maps corresponding to each sub-image are determined respectively, and the confidence scores of different points in the overlapping region are determined based on the feature maps. The position information of the notch mark on the crystal plane is determined based on the confidence level of different locations within the overlapping region.

[0008] Optionally, determining the feature maps corresponding to each sub-image and determining the confidence levels of different locations within the overlapping region based on the feature maps includes: By using extraction parameters at different scales, image features of each sub-image at different scales are obtained; The image features of each sub-image at different scales are weighted according to weights that are appropriate for each scale, so as to obtain the feature map corresponding to each sub-image. Based on a pre-trained neural network model, forward propagation processing is performed on the feature maps corresponding to each sub-image, and the confidence scores of each sub-image at each location point in the overlapping region are output. The confidence levels of different locations within the overlapping region are determined based on the confidence levels of the same location point in different sub-images.

[0009] Optionally, determining the position information of the notch mark on the crystal plane based on the confidence level of different locations within the overlapping region includes: Select a first location point, a second location point, and a third location point that satisfy a preset confidence level from the confidence levels of different location points, wherein the third location point is located between the first location point and the second location point; The first contour line passing through the first position point and the third position point is used as the first boundary of the notch mark, the second contour line passing through the second position point and the third position point is used as the second boundary of the notch mark, and the third position point is used as the vertex of the notch mark.

[0010] Optionally, determining the area covered by the film for the notch mark based on the light intensity values ​​at different locations on the film surface includes: Calculate the light intensity difference between any two locations; If the light intensity difference is determined to be greater than the preset light intensity, the combination of all regions corresponding to any two positions will be used as the area covered by the film as the notch mark.

[0011] Optionally, generating a first cutting instruction based on the area covered by the film and the location information of the notch mark includes: The cutting area is determined based on the area covered by the film and the location information of the notch mark; Based on the cutting area, a first cutting instruction is determined for the film in the cutting area.

[0012] Optionally, based on the first cutting instruction, after removing the film covering the notch mark and exposing the edge position of the notch mark, a second cutting instruction is generated to cut the film covering the edge of the wafer along the edge of the wafer, and the cutting path is: along the edge of the wafer, from the first edge position of the notch mark to the second edge position of the notch mark.

[0013] Accordingly, this disclosure also provides a thin-film dicing apparatus, wherein a thin film covers a wafer, the wafer having a notch mark, and the thin-film dicing apparatus includes: A control module, the control module being adapted to execute the thin film cutting method as described in any of the foregoing embodiments, and to generate a first cutting command; A first cutting tool, electrically connected to the control module, is adapted to remove the film located on the notch mark in response to the first cutting command.

[0014] Optionally, the control module is further adapted to generate a second cutting command when the edge of the notch mark is exposed after the first cutting tool removes the film covering the notch mark; The thin film cutting device further includes: a second cutting tool electrically connected to the control module, adapted to cut a thin film covering the edge of the wafer along the edge of the wafer in response to the second cutting command, and the cutting path is: along the edge of the wafer, from the first edge position of the notch mark to the second edge position of the notch mark.

[0015] Optionally, the film cutting device satisfies at least one of the following: The notch mark is in the shape of at least one of U-shape or V-shape; The first cutting tool includes: a laser or a blade; The second cutting tool includes: a blade or a laser; A stage for holding the wafer, and the first dicing tool and the first dicing tool are also connected to the stage.

[0016] Compared with the prior art, the technical solution of the present disclosure has the following advantages: The thin-film dicing method provided in this disclosure improves the accuracy of the notch mark position information by acquiring the position information of the notch mark on the crystal surface when the crystal surface of the wafer is not covered by the thin film. Furthermore, after determining the area of ​​the notch mark covered by the thin film, a first dicing command can be generated simultaneously based on the area and position information of the notch mark covered by the thin film to remove the thin film from the notch mark. Therefore, by employing the above technical solution, the thin film on the notch mark can be removed. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the description of the embodiments of this disclosure or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A schematic diagram illustrating the principle of membrane removal is shown.

[0019] Figure 2 A flowchart of a thin film cutting method according to an embodiment of the present disclosure is shown.

[0020] Figure 3 A flowchart illustrating the determination of the position information of the notch mark on the crystal plane is shown in one embodiment of the present disclosure.

[0021] Figure 4 A schematic diagram of the sampling image segmentation in one embodiment of this disclosure is shown.

[0022] Figure 5 A schematic diagram illustrating the principle of a membrane removal method according to an embodiment of the present disclosure is shown.

[0023] Figure 6 A schematic diagram of a thin film cutting structure according to an embodiment of the present disclosure is shown. Detailed Implementation

[0024] As described in the background section, in the process of processing the back side of a wafer, the protective film covers notch marks and the edges of the wafer in addition to the wafer face. The notch marks serve to determine the wafer position, thus requiring a step of removing the protective film to expose the notch marks.

[0025] See Figure 1 The diagram shown illustrates the principle of membrane removal. Figure 1 As shown in sub-figure (a), the thin film 20 is attached to the wafer 10 and covers the notch mark N. During the film removal operation, the cutting path typically starts from the first cutting point Ct1 and follows the direction indicated by arrow D, first removing the thin film covering the edge of the wafer 10. Then, the thin film 20 covering the notch mark N is removed following the cutting paths indicated by the second cutting point Ct2 and the third cutting point Ct3.

[0026] At the location of the notch mark N, due to the blade's width, when the blade is completely flush with the edge of the notch mark N during the cutting operation, the blade cannot rotate within the space of the notch mark N. As shown in sub-figure (b), the film 20 covering the notch mark N cannot be completely removed. Film 20 remains on the notch mark N.

[0027] The thin film 20 remaining on the notch mark N will reduce the accuracy of subsequent alignment operations on wafer 10.

[0028] It should be noted that, in order to illustrate the residual problem of film 20 on the notch mark N, Figure 1 The protective film covering the crystal surface is not shown.

[0029] To address the aforementioned technical problems, this disclosure provides a thin-film dicing method. By acquiring the positional information of the notch mark on the crystal surface when it is not covered by a thin film, the accuracy of the acquired notch mark positional information can be improved. Furthermore, after determining the area of ​​the notch mark covered by the thin film, a first dicing command can be generated simultaneously based on the area and positional information of the notch mark covered by the thin film to remove the thin film from the notch mark. Therefore, by employing the above technical solution, the thin film on the notch mark can be removed, improving the accuracy of subsequent wafer alignment operations.

[0030] To make the above-mentioned objects, features and advantages of the embodiments of this disclosure more apparent and understandable, the specific embodiments of this disclosure will be described in detail below with reference to the accompanying drawings.

[0031] See Figure 2 The flowchart shown is a thin film cutting method in one embodiment of the present disclosure, wherein the cutting scheme in this application is used to remove a portion of the thin film covering the wafer.

[0032] More specifically, the thin film covers the crystal face of the wafer, and in the case of notch markings on the wafer, the thin film also covers the notch markings. The dicing scheme in this application can be used to remove at least the thin film located on the notch markings, as well as the thin film covering the wafer edges.

[0033] Correspondingly, such as Figure 2 As shown, the following steps can be performed: S10, determine the position information of the notch mark on the crystal plane, the position information being obtained when the crystal plane is not covered by the thin film.

[0034] Specifically, when executing the thin film dicing scheme, the position information of the notch mark on the crystal surface is first obtained when the wafer is in its initial state, that is, when it is not covered by the thin film.

[0035] Since the crystal plane of the wafer is not covered by the thin film, the position information of the notch mark on the crystal plane is less affected by the thin film, thus the boundary outline of the notch mark can be defined more clearly.

[0036] For example, when the notch is V-shaped, the two boundaries of the notch and the intersection of these two boundaries can be determined. This improves the accuracy of the position information of the notch on the crystal plane, which is beneficial for the precise cutting in subsequent steps.

[0037] In this embodiment, see Figure 3 and Figure 4 ,in, Figure 3 This is a flowchart illustrating the determination of the position information of the notch mark on the crystal plane in one embodiment of this disclosure. Figure 4 This is a schematic diagram of image segmentation in one embodiment of the present disclosure, as shown below. Figure 3 and Figure 4 As shown, the following steps can be performed: S11, acquire the coordinate information of different positions of the crystal surface when the crystal surface is not covered by the thin film, and the sampled image containing the notch mark.

[0038] Specifically, when a wafer is placed on a processing machine, a coordinate system can be established based on the center of the wafer, thereby determining the coordinate system of different points (as a non-limiting example, different points include at least points located on the notch markings, such as points on the boundary contour, and points within the boundary contour). Thus, when acquiring a sampled image, the sampled image also contains the coordinate information of each point.

[0039] Specifically, the wafer uses a first coordinate system, while the image acquisition device uses a second coordinate system. By using the transformation coefficient between the wafer and the image acquisition device, the transformation between the first and second coordinate systems can be achieved, thus determining the coordinate information of each location point in the sampled image and realizing the correspondence between physical coordinate information and image coordinate information.

[0040] S12, based on the coordinate information of different locations, the sampled image is divided into multiple sub-images, such that the overlapping area of ​​at least two sub-images surrounds the notch mark.

[0041] Specifically, the coordinate information of the location point can characterize different regions on the wafer surface. When dividing the sampled image, there is a division scheme such that at least two of the multiple sub-images obtained by division have overlapping regions, and the overlapping regions can surround the notch mark.

[0042] As an example, see Figure 4 By dividing the sampled image, three sub-images M1, M2, and M3 can be obtained. Sub-image M1 consists of line segment A1-B1 and arc A1-B1, sub-image M2 consists of line segment A2-B2 and arc A2-B2, and sub-image M3 consists of line segment A1-C, C-A2, and arc A1-B2. Here, C is the intersection of line segment A1-B1 and line segment A2-B2.

[0043] See next Figure 4Sub-images M1 and M2 have an overlapping region consisting of line segments B2-C, C-B1 and arcs B1-B2, and the overlapping region contains a notch marker N.

[0044] It should be noted that, Figure 4 The illustrated method of dividing the sampled image is merely an example to show that the sampled image can be divided into multiple sub-images, and that overlapping areas of the sub-images contain notch markers N. It should not be construed as a limitation of this disclosure. Other division methods may be used in some other embodiments.

[0045] S13, determine the feature map corresponding to each sub-image, and determine the confidence level of different positions in the overlapping area based on the feature map.

[0046] Specifically, a feature map corresponds to one or more specific features or attributes of a sub-image, which can reflect the features of the sub-image such as edges, textures or colors. By generating a feature map corresponding to the sub-image, the true spatial information of the sub-image can be preserved.

[0047] In this way, based on the feature maps of the sub-images, the confidence level of the same location point in each sub-image can be determined. Thus, based on the confidence level of the same location point in each sub-image, the confidence level of each location point in the overlapping area can be determined. The confidence level reflects the degree to which the location point is truly located on the gap marker N.

[0048] In this embodiment, step S13 may specifically include: S131 uses extraction parameters of different scales to obtain image features of each sub-image at different scales.

[0049] Specifically, by using extraction parameters of different scales, each sub-image can be sampled layer by layer, thereby generating images at different scales. This facilitates the extraction of key feature points in images at different scales and helps determine the confidence level of each location point.

[0050] S132, according to the weights adapted to each scale, the image features of each sub-image at different scales are weighted to obtain the feature map corresponding to each sub-image.

[0051] Specifically, the extraction parameters at different scales have different weights, resulting in varying fusion ratios for each sub-image during image fusion. This allows for weighted fusion of image features at different scales based on the weight information, thereby generating feature maps for each sub-image.

[0052] As an example, image pyramids, feature extraction (e.g., scale-invariant feature transformation or accelerated robust features), and fusion algorithms can be used to determine the feature maps corresponding to each sub-image.

[0053] S133, based on a pre-trained neural network model, performs forward propagation processing on the feature maps corresponding to each sub-image, and outputs the confidence scores of each sub-image at each location point within the overlapping region.

[0054] The neural network model can be a convolutional neural network (CNN). Based on the feature maps corresponding to each input sub-image, the neural network model can output the confidence level of each location point, which represents the probability of the location of the gap marker in the overlapping region.

[0055] S134, determine the confidence level of different locations within the overlapping region based on the confidence level of the same location point in different sub-images.

[0056] Specifically, the confidence level of the same location point differs in different sub-images. Based on the confidence level of the same location point in different sub-images, methods such as summation and weighted average can be used to determine the confidence level of different location points within the overlapping area. This reflects the probability value of the location point being located on the gap mark.

[0057] It should be noted that when using a weighted average, the weights used can be the weights of the sub-images themselves.

[0058] By adopting the scheme in the above example, the selection range of position points on the notch mark is narrowed by pre-selecting the overlapping area, and by performing the selection operation based on the confidence of each position point in the overlapping area, the position point that truly belongs to the notch mark can be selected, which can improve the subsequent film cutting accuracy.

[0059] S14, determine the position information of the notch mark on the crystal plane based on the confidence level of different locations within the overlapping area.

[0060] Specifically, by using steps S11 to S13 above, the confidence level of different locations within the overlapping region can be determined, allowing the selection of locations with the highest or relatively high confidence levels. Based on these selected locations, the positional information of the notch marker on the crystal plane is determined. This positional information defines parameters such as the notch marker's position and boundary information.

[0061] In this embodiment, a first location point, a second location point, and a third location point that satisfy a preset confidence level can be selected from the confidence levels of different location points, and the third location point is located between the first location point and the second location point.

[0062] In an optional embodiment, the first location point may be located on one side of the overlapping region near one of the sub-images having the overlapping region, and the second location point may be located on one side of the overlapping region near another sub-image having the overlapping region. Then, a first contour line passing through the first location point and the third location point is used as the first boundary of the notch mark, a second contour line passing through the second location point and the third location point is used as the second boundary of the notch mark, and the third location point is used as the vertex of the notch mark.

[0063] In other words, since the difference between the notch mark N and the boundary of the crystal plane is large, the confidence of the position points on the boundary in the overlapping region determined by steps S11 to S13 must be relatively high. Since the notch mark N is a symmetrical structure (e.g., V-shaped notch mark N or U-shaped notch mark N), the positions corresponding to different positions must satisfy a normal distribution. The positions with higher confidence are closer to the position of the axis of symmetry. Thus, it is possible to obtain the first position point, the second position point, and the third position point, which include three position distributions.

[0064] By obtaining the first, second, and third position points, and based on the shape of the notch mark N, curve fitting can be used to determine the contour structure formed by the first, second, and third position points, thereby determining the position and boundary of the notch mark.

[0065] S20, obtain the light intensity values ​​at different positions on the surface of the thin film when the crystal plane is covered by the thin film.

[0066] In one embodiment, the light intensity value at different locations on the thin film can be determined based on a detector and a receiver that have a cooperative relationship.

[0067] Specifically, the detection light emitted by the detector can illuminate the surface of the thin film. The reflected light reflected by the surface of the thin film can be detected by the receiver. Based on the detection result of the receiver, the light intensity value at the current incident position can be determined, that is, the light intensity value at one position on the thin film can be obtained.

[0068] By changing the emission angle of the detector, the position of the probe light on the thin film surface can be changed, thereby enabling the measurement of the entire surface of the thin film.

[0069] S30, determine the area covered by the film for the notch mark based on the light intensity value at different positions on the surface of the film.

[0070] Specifically, the light intensity value reflects the difference in the material covered at different locations of the thin film, and thus, based on the light intensity value, it can be determined whether the thin film covers a crystal plane or a thin film.

[0071] In this embodiment, for the same probe beam and the same thin film, the difference in light intensity at different locations on the surface of the thin film is reflected in the material covering the thin film.

[0072] Specifically, if the film covers the crystal surface of a wafer, the penetration of the probe beam is low, resulting in more reflection of the probe beam and thus a higher light intensity at that location. Furthermore, with other parameters remaining constant, there are multiple locations (which could refer to the locations of the covered crystal surfaces) where the light intensity is high and tends to be the same.

[0073] If the area beneath the film is a notch mark, the probe beam has high penetration, resulting in less reflection and thus a lower light intensity at that location. Furthermore, with other parameters remaining constant, multiple locations (which could refer to the locations of the covered notch mark) exhibit low and similar light intensity values.

[0074] This allows us to determine the area covered by the film for the notch mark based on changes in light intensity.

[0075] In one specific embodiment, the light intensity difference between any two locations can be calculated; if the light intensity difference is determined to be greater than a preset light intensity, the combination of all regions corresponding to any two locations will be used as the area covered by the film for the notch mark.

[0076] Specifically, when the light intensity difference is greater than the preset light intensity, it indicates that the characteristics at these two locations are significantly different. If the parameters of the probe beam (type, intensity, and even the angle between the probe beam and the location point) are consistent, the boundary between these two locations may be one of the areas of the notch mark covered by the thin film. Thus, by performing multiple light intensity difference calculations and comparisons, the area of ​​the notch mark covered by the thin film can be determined.

[0077] In one embodiment, the area of ​​the notch mark covered by the film can be larger than the actual area of ​​the notch mark. This ensures that the determined area can completely cover the notch mark and avoids ignoring parts of the notch mark.

[0078] In some embodiments, when the light intensity difference is determined to be greater than a preset light intensity, and the region is taken as the boundary between the notch mark and the crystal plane, the range can be appropriately expanded so that the area of ​​the notch mark covered by the thin film can be larger than the actual area of ​​the notch mark.

[0079] In other words, by making the area covered by the film of the notch mark larger than the actual area of ​​the notch mark, it is ensured that the film covering the notch mark can be detected. In addition, by using the position information determined in the above steps, the actual area covered by the film of the notch mark can be determined.

[0080] S40, based on the area covered by the film and the location information of the notch mark, a first cutting command is generated to remove the film located on the notch mark.

[0081] Specifically, the area covered by the film defines the spatial location of the notch mark, while the location information defines the boundary and contour information of the notch mark. The two are related, so the first cutting command can be generated based on the area covered by the film and the location information of the notch mark.

[0082] More specifically, a cutting area is determined based on the area covered by the film and the location information of the notch mark; and a first cutting command for the film in the cutting area is determined based on the cutting area.

[0083] In other words, the cutting area defines the thin film located only on the boundary contour of the notch mark, so that the generated first cutting command can accurately cut the thin film at the position corresponding to the boundary contour of the notch mark, so that the thin film corresponding to the notch mark and the thin film covering the crystal surface are separated at the position corresponding to the boundary contour of the notch mark.

[0084] In this embodiment, the first cutting instruction includes the travel path of the cutting tool, and in response to the first cutting instruction, the cutting tool can remove the film located on the notch mark.

[0085] In some embodiments, the cutting tool may refer to a laser, which can remove the film on the notch mark by laser ablation by causing the laser emitted from the laser to travel along the path indicated by the first cutting command.

[0086] In some embodiments, the cutting tool may refer to a blade that, by causing the cutting path of the blade to follow the travel path indicated by the first cutting instruction, can directly contact the film to remove the film on the notch mark.

[0087] In actual processing, there are also cases where thin films cover the edges of the wafer, and these films also need to be removed.

[0088] Therefore, based on the first cutting instruction, after removing the film covering the notch mark and exposing the edge position of the notch mark, a second cutting instruction is generated to cut the film covering the edge of the wafer along the edge of the wafer, wherein the cutting path is: along the edge of the wafer, from the first edge position of the notch mark to the second edge position of the notch mark.

[0089] Specifically, such as Figure 5The schematic diagram shown in one embodiment of this disclosure illustrates the principle of film removal. After performing film removal according to a first cutting command, a second cutting command can be generated after confirming that film removal is complete. In response to the second cutting command, the cutting tool can cut the remaining portion of the film along the edge of the exposed notch mark.

[0090] And during the first cutting process, such as Figure 5 As shown, the projection of the outer edge of the cut onto the film 20 corresponds to the cutting positions CA1 and CA2, where only the notch mark N is exposed, and the remaining part of the film is still in a connected state.

[0091] During the second dicing process, the direction indicated by arrow D can be followed from the first dicing point Ct4 toward the second dicing point Ct5, so that the outer edge of the cut is projected onto the edge of the wafer 10 to remove the film on the edge of the wafer 10, leaving only the film on the crystal surface. Figure 5 (Not shown).

[0092] Compared to continuing to cut the film along the extension directions of CA1 and CA2 to the edge of the film so that the film corresponding to the notch mark N is completely separated from the film in other areas, the remaining part of the film is still connected, which can enhance its bending resistance, reduce deformation during the subsequent second cutting instruction (e.g., cutting with a tool), and improve the accuracy of cutting the film at the edge of the wafer.

[0093] In some embodiments, the cutting tool may refer to a laser, which can remove the thin film covering the edge of the wafer by laser ablation by causing the laser emitted from the laser to travel along the path indicated by the second cutting command.

[0094] In some embodiments, the cutting tool may refer to a blade that, by causing the cutting path of the blade to follow the travel path indicated by the second cutting instruction, can directly contact the film and remove the film covering the edge of the wafer.

[0095] In one specific embodiment, a laser may be used to perform the first cutting process, and a blade may be used to perform the second cutting process.

[0096] It should be noted that during the first cutting process, the cutting status can be obtained in real time or at preset intervals (e.g., by acquiring an image of the wafer in its current state). After it is determined that the thin film covering the notch mark has been completely removed, a second cutting instruction is generated. The travel path indicated by the second cutting instruction can be determined in advance according to the shape of the wafer.

[0097] Accordingly, this solution also provides a thin film cutting device that can remove a portion of the thin film covering the wafer.

[0098] More specifically, a thin film covers the crystal face of the wafer, and the wafer has notch markings, which are used to enable wafer alignment operations.

[0099] Accordingly, see Figure 6 The schematic diagram of the film cutting apparatus in one embodiment of this disclosure is shown below. Figure 6 As shown, the thin film cutting device 100 may include: Control module 110, the control module 110 is adapted to execute the thin film cutting method as described in any of the foregoing embodiments, and generate a first cutting command; The first cutting tool 120 is electrically connected to the control module 110 and is adapted to remove the film located on the notch mark N in response to the first cutting command.

[0100] The method by which the control module 110 generates the first cutting command based on the thin film cutting method described in any of the foregoing embodiments can be referred to in the foregoing examples and will not be described here.

[0101] In this embodiment, the control module 110 may be composed of multiple devices or equipment with different functions. For example, in order to generate the first cutting instruction, the control module 110 may include a processor with a built-in convolutional neural network (CNN), a sampling device, and a detector and a receiver with a cooperative relationship.

[0102] More specifically, the sampling device can acquire a sampled image containing the notch mark as input to the processor.

[0103] In some embodiments, the sampling device may include a camera, CCD, or other imaging device.

[0104] A processor with a built-in convolutional neural network (CNN) can perform region segmentation, feature map construction, and confidence calculation on the sampled image to determine the location information of the notch mark on the crystal plane.

[0105] In some embodiments, the processor may include, but is not limited to, hardware circuits implemented with application-specific integrated circuits (ASICs), programmable logic devices (PLDs), microcontroller units (MCUs), microprocessor units (MPUs), digital signal processors (DSPs), or central processing units (CPUs). For example, hardware circuits implemented with PLDs may include field-programmable gate arrays (FPGAs).

[0106] The detector emits a probe light that can illuminate the surface of the thin film. The reflected light that is reflected by the surface of the thin film can be detected by the receiver. Based on the detection result of the receiver, the light intensity value at the current incident position can be determined, that is, the light intensity value at one of the positions on the thin film can be obtained.

[0107] By changing the emission angle of the detector, the position of the probe light on the thin film surface can be changed, thereby enabling the measurement of the entire surface of the thin film.

[0108] In some embodiments, the detector may include multiple lasers. The lasers may include one or more types of lasers. For example, the lasers may include semiconductor lasers, fiber lasers, or other types of lasers. For instance, semiconductor lasers may include vertical cavity surface emitting lasers (VCSELs), edge emitting lasers (EELs), distributed feedback lasers (DFBs), or similar devices. The above are merely examples, and this disclosure does not limit the type of laser used.

[0109] In some embodiments, the receiver may include one or more detection devices. For example, the receiver may include: a photodetector circuit, a pin photodiode (PINPD), an avalanche photodiode (APD), a single photon avalanche diode (SPAD), a silicon photomultiplier (SiPM), or similar devices. The above are merely examples, and the embodiments disclosed herein do not limit the type of receiver.

[0110] In some other embodiments, the light intensity values ​​at different locations on the thin film surface can also be obtained by changing the positions of the detector and receiver.

[0111] For example, a drive mechanism connected to the detector and receiver is provided. The drive mechanism is electrically connected to the processor and, in response to a motion control signal from the processor, drives the detector and receiver to move, so that the detection light can be incident on different positions on the thin film surface.

[0112] Specifically, by driving the detector and receiver, the relative positional relationship between the detector and receiver can be maintained, so that at any time, different positions on the thin film surface can receive the detection light, thus obtaining the light intensity values ​​at multiple positions.

[0113] More specifically, the driving mechanism may include: a first slide rail extending along a first direction to drive the detector and receiver to move along the first direction; a second slide rail extending along a second direction and slidingly cooperating with the first slide rail to drive the detector and receiver to move along the second direction; and a driver connected to the first slide rail and the second slide rail for generating a first driving force along the first direction and a second driving force along the second direction in response to the motion control signal.

[0114] In short, the detector and receiver can move synchronously, thus enabling the acquisition of light intensity values ​​at different locations on the thin film surface.

[0115] After the light intensity values ​​at different locations are transmitted to the processor, the processor can execute the aforementioned method to determine the area where the notch mark is covered by the thin film.

[0116] In this embodiment, the first cutting tool 120 may include a laser or a blade.

[0117] More specifically, when the first cutting tool 120 is a laser, by causing the laser emitted from the laser to follow the travel path indicated by the first cutting command, the thin film on the notch mark can be removed by laser ablation.

[0118] When the first cutting tool 120 is a blade, the blade can directly contact the film by making the cutting path of the blade follow the travel path indicated by the first cutting command, so as to remove the film on the notch mark.

[0119] In this embodiment, based on the first cutting command, the thin film covering the notch mark can be removed to expose the edge position of the notch mark, which is beneficial for performing wafer alignment operations.

[0120] In actual processing, there are also cases where thin films cover the edges of the wafer. These films can reduce the wafer thinning effect and even cause contamination of the processing equipment, so these films also need to be removed.

[0121] In view of this, the control module 110 is also adapted to generate a second cutting command when the edge position of the notch mark is exposed after the first cutting tool 120 removes the film covering the notch mark N.

[0122] Specifically, after performing the film removal process according to the first cutting instruction, a second cutting instruction can be generated after confirming that the film removal is complete. In response to the second cutting instruction, the remaining portion of the film can be cut along the edge of the exposed notch mark.

[0123] Accordingly, the thin film cutting device 100 may further include: a second cutting tool 130, electrically connected to the control module 110, adapted to cut a thin film covering the edge of the wafer along the edge of the wafer in response to the second cutting command, and the cutting path being: along the edge of the wafer, from the first edge position of the notch mark to the second edge position of the notch mark.

[0124] And during the first cutting process, such as Figure 5 As shown, the projection of the outer edge of the cut onto the film 20 corresponds to the cutting positions CA1 and CA2, where only the notch mark N is exposed, while the remaining part of the film is still in a connected state, the film still has good tension, and the film has high coverage stability.

[0125] During the second dicing process, the film on the edge of wafer 10 can be removed from the first dicing point Ct4 along the edge of wafer 10 toward the second dicing point Ct5, as indicated by arrow D, leaving only the film on the crystal surface. Figure 5 (Not shown).

[0126] That is, the outer edge of the cut is projected onto the edge of the wafer 10 in the thin film 20.

[0127] In other words, this application first removes the thin film on the notch mark N, and then removes the thin film covering the edge of the wafer. This reduces the risk of the thin film cracking or falling off the wafer during the film removal process.

[0128] In some embodiments, the second cutting tool 130 may be a blade or a finger laser.

[0129] More specifically, when the second cutting tool 130 is a blade, by making the cutting path of the blade follow the travel path indicated by the second cutting command, it can directly contact the film and remove the film covering the edge of the wafer.

[0130] When the second cutting tool 130 is a laser, by making the laser emitted from the laser follow the travel path indicated by the second cutting command, the thin film covering the edge of the wafer can be removed by laser ablation.

[0131] In some embodiments, the first cutting tool may be a laser, and the second cutting tool 130 may be a blade.

[0132] In this embodiment, the shape of the notch mark can be at least one of U-shape or V-shape. Based on the shape of the notch mark, a blade that is adapted to it can be used, or the laser emitted by the laser can completely fit the boundary contour of the notch mark.

[0133] In this embodiment, the thin film dicing apparatus may further include a stage for supporting the wafer. By placing the wafer on the stage, the stability during the film stripping process can be improved, allowing the remaining portion of the film to still better cover the crystal surface and enhancing the protective effect of the film.

[0134] Correspondingly, the first cutting tool and the first cutting tool are also connected to the stage.

[0135] It is understood that the above embodiments provide multiple implementation schemes, and these implementation schemes can be combined and cross-referenced with each other without conflict, thereby extending to multiple possible implementation schemes. These can all be considered as the implementation schemes disclosed and made public in this application.

[0136] In some embodiments, the thin film cutting device described in the above examples can be applied to a wafer thinning machine.

[0137] The specific structure and working principle of the thin film cutting device can be found in the aforementioned example.

[0138] By incorporating a thin film cutting device into the wafer thinning machine, the thin film on the notch mark can be completely removed, thus improving the thinning quality.

[0139] This disclosure provides a data processing device, which may include a memory and a processor. The memory and the processor can communicate with each other via a communication bus. The memory stores one or more computer instructions that can be executed on the processor. When the processor executes the computer instructions, it can perform the steps of the thin film cutting method described in any of the above embodiments. For details, please refer to the above-mentioned related content, which will not be repeated here.

[0140] In some examples, computer instructions may include any suitable type of code implemented using any appropriate high-level, low-level, object-oriented, visual, compiled, and / or interpreted programming language, such as source code, compiled code, interpreted code, executable code, static code, dynamic code, encrypted code, etc.

[0141] In some examples, the data processing device may further include a display interface and a display connected via the display interface. The display interface can communicate with a memory and a processor via a communication bus. The display can display various data generated during the thin film cutting process provided in the embodiments of this disclosure by the processor. In some embodiments, the data processing device may further include a data output interface.

[0142] The data output interface can communicate with the memory and processor via a communication bus to output various data during the thin film cutting process.

[0143] It should be noted that the terms "example" or "implementation" used in this specification refer to a specific feature, structure, or characteristic that may be included in at least one implementation of the embodiments of this disclosure. Furthermore, in the description of this specification, terms such as "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with terms such as "first" and "second" may explicitly or implicitly include one or more of that feature. Moreover, terms such as "first" and "second" are used to distinguish similar objects and are not necessarily used to describe a specific order or indicate importance. It is understood that such terms can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein.

[0144] While the embodiments disclosed herein are as described above, this disclosure is not limited thereto. Any person skilled in the art can make various alterations and modifications without departing from the spirit and scope of this disclosure; therefore, the scope of protection of this disclosure should be determined by the scope defined in the claims.

Claims

1. A method for cutting a thin film, characterized in that, A thin film covers the crystal face of a wafer, the wafer having notch markings, and the thin film dicing method includes: The position information of the notch mark on the crystal plane is determined, and the position information is obtained when the crystal plane is not covered by the thin film; When the crystal plane is covered by the thin film, the light intensity values ​​at different positions on the surface of the thin film are obtained; The area covered by the film for the notch mark is determined based on the light intensity values ​​at different locations on the surface of the film. Based on the area covered by the film and the location information of the notch mark, a first cutting command is generated to remove the film located on the notch mark.

2. The thin film cutting method according to claim 1, characterized in that, Determining the position information of the notch mark on the crystal plane includes: When the crystal plane is not covered by the thin film, obtain the coordinate information of different points on the crystal plane, as well as the sampled image containing the notch mark; Based on the coordinate information of different locations, the sampled image is divided into multiple sub-images, such that the overlapping area of ​​at least two sub-images surrounds the notch mark; The feature maps corresponding to each sub-image are determined respectively, and the confidence scores of different points in the overlapping region are determined based on the feature maps. The position information of the notch mark on the crystal plane is determined based on the confidence level of different locations within the overlapping region.

3. The thin film cutting method according to claim 2, characterized in that, The step of determining the feature maps corresponding to each sub-image and determining the confidence levels of different locations within the overlapping region based on the feature maps includes: By using extraction parameters at different scales, image features of each sub-image at different scales are obtained; The image features of each sub-image at different scales are weighted according to weights that are appropriate for each scale, so as to obtain the feature map corresponding to each sub-image. Based on a pre-trained neural network model, forward propagation processing is performed on the feature maps corresponding to each sub-image, and the confidence scores of each sub-image at each location point in the overlapping region are output. The confidence levels of different locations within the overlapping region are determined based on the confidence levels of the same location point in different sub-images.

4. The thin film cutting method according to claim 2, characterized in that, Determining the position information of the notch mark on the crystal plane based on the confidence levels of different locations within the overlapping region includes: Select a first location point, a second location point, and a third location point that satisfy a preset confidence level from the confidence levels of different location points, wherein the third location point is located between the first location point and the second location point; The first contour line passing through the first position point and the third position point is used as the first boundary of the notch mark, the second contour line passing through the second position point and the third position point is used as the second boundary of the notch mark, and the third position point is used as the vertex of the notch mark.

5. The thin film cutting method according to claim 1, characterized in that, Determining the area covered by the film for the notch mark based on the light intensity values ​​at different locations on the film surface includes: Calculate the light intensity difference between any two locations; If the light intensity difference is determined to be greater than the preset light intensity, the combination of all regions corresponding to any two positions will be used as the area covered by the film as the notch mark.

6. The thin film cutting method according to claim 1, characterized in that, The step of generating a first cutting instruction based on the area covered by the film and the location information of the notch mark includes: The cutting area is determined based on the area covered by the film and the location information of the notch mark; Based on the cutting area, a first cutting instruction is determined for the film in the cutting area.

7. The thin film cutting method according to any one of claims 1 to 6, characterized in that, Based on the first cutting instruction, after removing the film covering the notch mark and exposing the edge position of the notch mark, a second cutting instruction is generated to cut the film covering the edge of the wafer along the edge of the wafer, and the cutting path is: along the edge of the wafer, from the first edge position of the notch mark to the second edge position of the notch mark.

8. A thin film cutting device, characterized in that, A thin film covers a wafer, the wafer having notch markings, and the thin film dicing apparatus includes: The control module is adapted to execute the thin film cutting method as described in any one of claims 1 to 7 and generate a first cutting command; A first cutting tool, electrically connected to the control module, is adapted to remove the film located on the notch mark in response to the first cutting command.

9. The thin film cutting apparatus according to claim 8, characterized in that, The control module is also adapted to generate a second cutting command when the edge of the notch mark is exposed after the first cutting tool removes the film covering the notch mark; The thin film cutting device further includes: a second cutting tool electrically connected to the control module, adapted to cut a thin film covering the edge of the wafer along the edge of the wafer in response to the second cutting command, and the cutting path is: along the edge of the wafer, from the first edge position of the notch mark to the second edge position of the notch mark.

10. The film cutting apparatus according to claim 9, characterized in that, Meet at least one of the following: The notch mark is in the shape of at least one of U-shape or V-shape; The first cutting tool includes: a laser or a blade; The second cutting tool includes: a blade or a laser; A stage for holding the wafer, and the first dicing tool and the first dicing tool are also connected to the stage.

Images