Mask substrate defect detection method and device

Abstract

The invention provides a mask substrate defect detection method and device, and the method comprises the following steps: scanning a to-be-detected mask substrate by using a laser confocal microscope to obtain confocal signals of particles on the surface of the mask substrate; scanning the mask substrate by adopting a white light interferometer to obtain surface contour data of the mask substrate; morphological feature information of the particles is obtained from the confocal signals of the particles, and background morphological feature information of particle signal points is obtained from the surface contour data of the mask substrate; calculating a deviation difference value between the morphology feature information of the particles and the background morphology feature information of the particle signal points, and comparing the deviation difference value with a preset deviation threshold value; when the deviation difference value is greater than a preset deviation threshold value, marking the particles as real particles; and when the deviation difference value is not greater than a preset deviation threshold value, filtering out confocal signals of the particles. The method has the beneficial effects of high detection rate and low false alarm rate, and the reliability of defect detection of the mask substrate can be remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of integrated circuit manufacturing, and in particular to a method and apparatus for detecting defects in mask substrates. Background Technology

[0002] A photomask substrate is a substrate on a highly flat transparent quartz substrate on which functional thin films such as chromium films, chromium compound films, and molybdenum silicide films are formed through a precision deposition process. It serves as the master template for manufacturing photomasks, and its surface quality directly determines the imaging accuracy and product yield of the final photomask. Therefore, rigorous detection and control of defects on its surface and within the film layers during the photomask substrate manufacturing process are crucial steps in ensuring the yield of advanced process nodes.

[0003] Generally, during the deposition of functional thin films on a mask substrate, the sputtering parameters (temperature, pressure, gas composition, etc.) result in inherent, non-smooth surface morphologies. Metal thin film surfaces typically exhibit a polycrystalline structure, characterized by columnar or elongated polycrystalline grains, micropores (<10 nm), and surface roughness features such as xy-direction extension. Simultaneously, during sputtering, foreign particles (such as metal particles) may be introduced due to target shape, edge protection structures, and temperature cycling, leading to defects such as abnormal protrusions / depressions in the film. These inherent micromorphologies (polycrystalline structures) and actual defects (foreign particles) intertwine and resemble each other at the submicron to nanometer scale, making the mask substrate surface extremely complex and posing significant challenges to accurate film defect identification. Traditional detection methods are no longer sufficient to meet the surface defect detection requirements of mask substrates. For example, while scanning electron microscopy offers high resolution, it primarily provides surface morphology information and struggles to effectively distinguish between polycrystalline grains, pores, or particles. Traditional microscopes, due to their limited depth of field and signal confusion, cannot extract clear signals of specific defects from complex background scattered light, resulting in low image contrast and a high misjudgment rate.

[0004] To address these issues, the industry has attempted to use laser confocal microscopy to detect surface defects in mask substrates. Laser confocal microscopy utilizes spatial pinholes to filter out defocused light, possessing excellent optical tomography capabilities and higher lateral resolution, enabling the acquisition of clearer images than wide-field microscopy. However, existing confocal detection methods employ a global threshold for a single target signal. When faced with the inherent high-contrast roughness of the mask substrate, these methods suffer from the following problems: Confocal detection methods primarily rely on reflected light intensity signals to determine defect information. When the grain size of the polycrystalline structure and the size of the foreign particle are the same or similar, the detection laser scan may generate similar intensity scattered signals when passing over a real foreign particle and a rough grain protrusion. The system cannot distinguish whether the signal source is an "abnormal defect to be captured" or an "allowable substrate feature" based solely on light intensity information. This leads to numerous false alarms of "false defects" caused by the rough background when setting a high sensitivity threshold; conversely, increasing the threshold to reduce false alarms may result in the missed detection of weak, real defects. This discrepancy between the detection rate and the false alarm rate seriously affects the reliability of testing and the production cycle.

[0005] Therefore, there is an urgent need for a mask substrate defect detection method that can accurately, efficiently, and non-destructively identify real defects on the mask substrate under complex background interference. Summary of the Invention

[0006] The purpose of this invention is to provide a method and apparatus for detecting mask substrate defects that can effectively distinguish between real defects on the surface of the mask substrate and inherent background features of the substrate.

[0007] To achieve the above objectives, the present invention provides a method for detecting defects in a mask substrate, comprising the following steps: A laser confocal microscope was used to scan the mask substrate to be tested, and the confocal signal of the particles on the surface of the mask substrate was obtained. A white light interferometer was used to scan the mask substrate to obtain the surface contour data of the mask substrate; The morphological features of the particles are obtained from the confocal signals of the particles, and the background morphological features of the particle signal points are obtained from the surface contour data of the mask substrate. Calculate the deviation difference between the morphological feature information of the particles and the background morphological feature information of the particle signal points, and compare the deviation difference with a preset deviation threshold. When the deviation difference is greater than the preset deviation threshold, the particle is marked as a real particle; When the deviation difference is not greater than the preset deviation threshold, the confocal signal of the particles is filtered out.

[0008] The aforementioned mask substrate defect detection method constructs a "background noise elimination" mechanism by fusing the surface contour data of the mask substrate obtained by a white light interferometer with the confocal optical signals of particles obtained by a confocal microscope. Based on the high-precision surface contour data obtained by the white light interferometer, the confocal signals of particles detected by the confocal microscope are checked for consistency to determine whether the particles are real particles. Specifically, the mask substrate to be tested is scanned by a laser confocal microscope and a white light interferometer, respectively, to obtain the confocal signals of particles on the surface of the mask substrate and the surface contour data of the mask substrate. The morphological feature information of the particles is obtained from the confocal signals, and the background morphological feature information of the particle signal points is obtained from the surface contour data of the mask substrate. The deviation difference between the morphological feature information of the particles and the background morphological feature information of the particle signal points is further calculated, and the deviation difference is compared with a preset deviation threshold to determine whether the particles are real particles. When the deviation difference is greater than the preset deviation threshold, the morphological feature information of the particle is inconsistent with the background morphological feature information of the particle signal point, and the particle is determined to be a real particle and marked. When the deviation difference is not greater than the preset deviation threshold, the morphological feature information of the particle is consistent with the background morphological feature information of the particle signal point, and the confocal signal of the particle is determined to be a background noise signal. The confocal signal of the particle is filtered out, thereby eliminating false defect misjudgment and effectively reducing the false alarm rate.

[0009] The aforementioned mask substrate defect detection method verifies the consistency between the confocal signal of the particles and the background morphology features of the particle signal points. It distinguishes and filters out background interference signals caused by the surface roughness and grain characteristics of the mask substrate's thin film, thereby achieving accurate identification of real foreign particles or abnormal defects. This method can actively identify and filter out most background signals generated by the inherent roughness of the polycrystalline structure of the metal thin film, significantly reducing the false alarm rate. Furthermore, the aforementioned mask substrate defect detection method no longer relies on simply increasing the light intensity detection threshold of the laser confocal microscope to reduce the false alarm rate. The detection threshold of the laser confocal microscope can be set to remain sensitive to weaker abnormal signals. By analyzing the "inconsistency" between the confocal signal and the background morphology features, even abnormal defects with low signal intensity (such as low-reflectivity contamination particles) can be effectively captured, thus effectively improving the detection rate of real defects while reducing the false alarm rate.

[0010] In summary, the aforementioned mask substrate defect detection method effectively solves the technical challenge of simultaneously achieving a high detection rate and a low false alarm rate in mask substrate defect detection. This method maintains a high defect detection rate while significantly reducing the false alarm rate of false defects, improving the reliability and efficiency of mask substrate defect detection and contributing to increased mask substrate production efficiency. Furthermore, both white light interferometer and laser confocal microscope are non-contact detection methods, enabling non-destructive detection of true defects in the mask substrate, thus saving on mask substrate manufacturing costs. Therefore, the aforementioned mask substrate defect detection method can accurately, efficiently, and non-destructively identify true defects on the mask substrate, meeting the reliability requirements of large-scale production, improving mask substrate production efficiency, and helping to save manufacturing costs.

[0011] In one embodiment, the step of acquiring the confocal signal of particles on the surface of the mask substrate includes: The peak brightness and its coordinate position at the focal plane are identified by shallow depth-of-field scanning, and a focused image is obtained. A three-dimensional image is constructed by integrating all focused images acquired by shallow depth-of-field scanning. The distribution of particles on the surface of the mask substrate and in its thin film stacking structure is analyzed to obtain the confocal signal of the particles.

[0012] In one embodiment, the spot area of ​​the detection beam of the white light interferometer is 1 mm. 2 ~100mm 2 .

[0013] In one embodiment, the spot area of ​​the laser beam in the laser confocal microscope is no greater than 1 mm. 2 .

[0014] On the other hand, the present invention also provides a mask substrate defect detection device, which is used to perform the mask substrate defect detection method of any of the above claims. The mask substrate defect detection device includes: a white light interferometer, a laser confocal microscope, and a controller. The white light interferometer and the laser confocal microscope are arranged adjacent to each other. The controller is electrically connected to the white light interferometer and the laser confocal microscope respectively. The controller is used to determine whether the particles are real particles based on the confocal signal of the particles on the surface of the mask substrate obtained by the confocal microscope and the surface contour data of the mask substrate obtained by the white light interferometer.

[0015] The aforementioned mask substrate defect detection device is used to perform any of the mask substrate defect detection methods described above. This mask substrate defect detection device can accurately, efficiently, and non-destructively identify real defects on the mask substrate, meet the requirements of large-scale production for detection reliability, improve the production efficiency of mask substrates, and help save manufacturing costs.

[0016] In one embodiment, the controller includes: Signal receiving unit: used to receive the confocal signal of the particles sent by the laser confocal microscope and the surface contour data of the mask substrate sent by the white light interferometer; Information extraction unit: used to obtain the morphological feature information of particles from the confocal signal of particles, and to obtain the background morphological feature information of particle signal points from the surface contour data of the mask substrate; Information comparison unit: used to calculate the deviation difference between the morphological feature information of the particle and the background morphological feature information of the particle signal point, and compare the deviation difference with a preset deviation threshold; Signal filtering unit: Used to receive the comparison result from the information comparison unit, where the deviation difference is not greater than a preset deviation threshold, and filter out the confocal signal of the particles based on the comparison result; Particle marking unit: Used to receive the comparison result from the information comparison unit when the deviation difference is greater than a preset deviation threshold, mark the particle as a real particle according to the comparison result, and send the confocal signal of the particle to the storage unit. Storage unit: Used to store preset deviation thresholds, surface contour data of the mask substrate, and confocal signals of the actual particles.

[0017] In one embodiment, the mask substrate defect detection device further includes a display device connected to the controller, the display device being used to display the confocal signal of the real particles and the surface contour data of the mask substrate.

[0018] In one embodiment, the mask substrate defect detection device further includes a robot arm connected to a controller, the robot arm being used to transfer the mask substrate.

[0019] In one embodiment, the mask substrate defect detection device further includes: a device body and an FFU. The device body has a mounting chamber, and a mounting stage is provided in the mounting chamber. A white light interferometer and a laser confocal microscope are respectively mounted on the mounting stage, and the FFU is located on the top of the device body.

[0020] In one embodiment, the mask substrate defect detection device further includes an ion generator connected to the main body of the device. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the principle of a mask substrate defect detection method according to one embodiment; Figure 2 This is a structural cross-sectional view of a mask substrate defect detection device according to an embodiment; Figure 3 This is a schematic diagram of the controller in one embodiment.

[0022] Explanation of reference numerals in the attached figures: 10-Main body of the device, 20-Controller, 11-White light interferometer, 12-Laser confocal microscope, 12-Robot arm, 14-FFU, 15-Mounting stage, 16-Mounting chamber, 17-Ion generator. Detailed Implementation

[0023] The present invention will be described more clearly and completely by way of embodiments and in conjunction with the accompanying drawings, but the present invention is not limited to the scope of the following embodiments.

[0024] Please see Figure 1 One embodiment of the mask substrate defect detection method includes the following steps: Step S11: Use a laser confocal microscope to scan the mask substrate to be tested and obtain the confocal signal of the particles on the surface of the mask substrate.

[0025] Specifically, a laser confocal microscope is used to scan the mask substrate to be tested. The laser confocal microscope acquires the reflected light signal of each detection point and compares the reflected light signal of the detected detection point with the set light intensity detection threshold to determine whether there are particles on the surface of the mask substrate. When there are particles on the surface of the mask substrate, the confocal signal of the particles is acquired.

[0026] Further, in one embodiment, the step of acquiring the confocal signal of particles on the surface of the mask substrate includes: identifying the peak brightness and coordinate position at the focal plane through shallow depth-of-field scanning to acquire a focused image; integrating all focused images acquired by shallow depth-of-field scanning to construct a three-dimensional image, analyzing the distribution of particles on the surface of the mask substrate and in its thin film stack structure, and acquiring the confocal signal of the particles. Specifically, the laser confocal microscope performs "focus scanning" with a tiny laser spot to detect particle information, identifies the peak brightness at the focal plane in the Z-axis direction and locates the particle position using shallow depth-of-field scanning, then integrates the "full-focus image" to observe the distribution of particles in the thin film stack, and generates a three-dimensional image to complete the acquisition of the confocal signal of the particles. In this embodiment, the confocal signal of the particles includes at least the reflected light intensity of the particles, the position information of the particles, and the morphological feature information of the particles. Furthermore, to effectively ensure the resolution of the detection results, a tiny laser spot is used for scanning. Preferably, the spot area of ​​the laser beam of the laser confocal microscope is no greater than 1 mm. 2 .

[0027] Step S12: Use a white light interferometer to scan the mask substrate and obtain the surface contour data of the mask substrate.

[0028] Specifically, a white-light interferometer is used to scan and measure the surface morphology of the mask substrate. Then, a Fourier transform (along the x and y directions) is used to present the spatial frequency distribution curve of the polycrystalline surface structure, obtaining the surface contour data of the mask substrate (e.g., repeatability distance data in the x and y directions). This embodiment uses a white-light interferometer to acquire the surface contour data of the mask substrate, which can accurately characterize each grain protrusion, micropore, and overall roughness contour of the mask substrate surface, establishing a high-precision reference surface and forming a "reference morphology map." Furthermore, to ensure detection efficiency, the preferred spot area of ​​the detection beam of the white-light interferometer is 1 mm². 2 ~100mm 2 .

[0029] Step S13: Obtain the morphological feature information of the particles from the confocal signal of the particles, and obtain the background morphological feature information of the particle signal points from the surface contour data of the mask substrate.

[0030] Specifically, the confocal signal of the particles is analyzed to obtain the morphological feature information of the particles and the position information of the particles. The position of the particle signal point is determined, and then the "baseline morphological map" is queried based on the position information of the particles. Regular background morphological feature information (such as repeating distance information in the x and y directions) of the background morphology of the particle signal point is obtained from the surface contour data of the mask base plate.

[0031] Step S14: Calculate the deviation difference between the morphological feature information of the particle and the background morphological feature information of the particle signal point, and compare the deviation difference with the preset deviation threshold.

[0032] Specifically, the deviation difference between the morphological feature information of the particle and the background morphological feature information of the particle signal point is calculated, and the deviation difference is compared with a preset deviation threshold to determine whether the morphological feature of the particle matches the background morphological feature of the inherent roughness of the mask substrate surface, and further determine whether the particle is a real particle. Step S15 or step S16 is executed according to the comparison result.

[0033] Step S15: When the deviation difference is greater than the preset deviation threshold, mark the particle as a real particle.

[0034] Specifically, when the deviation difference exceeds a preset deviation threshold, the particle's morphological feature information is either greater or less than the background morphological feature information of the particle signal point, and the difference between the particle's morphological feature information and the background morphological feature information of the particle signal point is significant. The particle's morphological feature information deviates significantly from the background morphological feature information of the particle signal point, and the particle's morphological feature does not match the background morphological feature of the inherent roughness of the mask substrate surface. Therefore, the particle is determined to be a real particle. Thus, when the deviation difference exceeds the preset deviation threshold, the particle's morphological feature information is determined to be inconsistent with the background morphological feature information of the particle signal point, and the particle is identified as a real particle and marked.

[0035] Step S16: When the deviation difference is not greater than the preset deviation threshold, filter out the confocal signal of the particles.

[0036] Specifically, when the deviation difference is not greater than a preset deviation threshold, the particle's morphological feature information is the same as or close to the background morphological feature information of the particle signal point; the particle's morphological feature information is the same as or deviates little from the background morphological feature information of the particle signal point; and the particle's morphological feature matches the background morphological feature of the inherent roughness of the mask substrate surface. Therefore, the confocal signal of the particle is determined to be a background interference signal. Conversely, when the deviation difference is greater than the preset deviation threshold, the particle's morphological feature information is determined to be consistent with the background morphological feature information of the particle signal point. The confocal signal of that particle is then identified as a background noise signal, and the confocal signal of that particle is filtered out, eliminating false defect misjudgments.

[0037] The mask substrate defect detection method of this embodiment first uses a laser confocal microscope to scan the mask substrate to be inspected to obtain the confocal signal of the particles on the surface of the mask substrate. Then, a white light interferometer is used to scan the mask substrate to obtain the surface contour data of the mask substrate. It should be noted that in other embodiments, a white light interferometer can be used to scan the mask substrate first to obtain the surface contour data of the mask substrate, and then a laser confocal microscope can be used to scan the mask substrate to be inspected to obtain the confocal signal of the particles on the surface of the mask substrate. In practical applications, the execution order of the above steps S11 and S12 can be specifically set as needed. This invention does not specifically limit the execution order of the above steps S11 and S12.

[0038] The aforementioned mask substrate defect detection method constructs a "background noise elimination" mechanism by fusing the surface contour data of the mask substrate obtained by a white light interferometer with the confocal optical signal of the particles obtained by a confocal microscope. Based on the high-precision surface contour data obtained by the white light interferometer, the consistency of the confocal signal of the particles detected by the confocal microscope is checked to determine whether the particles are real particles. Specifically, the deviation difference between the confocal signal of the particle and the background morphological feature information of the particle signal point is calculated, and the deviation difference is compared with a preset deviation threshold. When the deviation difference is greater than the preset deviation threshold, the confocal signal of the particle is inconsistent with the background morphological feature information of the particle signal point, and the particle is determined to be a real particle and marked. When the deviation difference is not greater than the preset deviation threshold, the confocal signal of the particle is consistent with the background morphological feature information of the particle signal point, and the confocal signal of the particle is determined to be a background noise signal, and the confocal signal of the particle is filtered out, thereby eliminating false defect misjudgments and effectively reducing the false alarm rate.

[0039] The aforementioned mask substrate defect detection method verifies the consistency between the confocal signal of the particles and the background morphology features of the particle signal points. It distinguishes and filters out background interference signals caused by the surface roughness and grain characteristics of the mask substrate's thin film, thereby achieving accurate identification of real foreign particles or abnormal defects. This method can actively identify and filter out most background signals generated by the inherent roughness of the polycrystalline structure of the metal thin film, significantly reducing the false alarm rate. Furthermore, the aforementioned mask substrate defect detection method no longer relies on simply increasing the light intensity detection threshold of the laser confocal microscope to reduce the false alarm rate. The detection threshold of the laser confocal microscope can be set to remain sensitive to weaker abnormal signals. By analyzing the "inconsistency" between the confocal signal and the background morphology features, even abnormal defects with low signal intensity (such as low-reflectivity contamination particles) can be effectively captured, thus effectively improving the detection rate of real defects while reducing the false alarm rate.

[0040] In summary, the aforementioned mask substrate defect detection method effectively solves the technical challenge of simultaneously achieving a high detection rate and a low false alarm rate in mask substrate defect detection. This method maintains a high defect detection rate while significantly reducing the false alarm rate of false defects, improving the reliability and efficiency of mask substrate defect detection and contributing to increased mask substrate production efficiency. Furthermore, both white light interferometer and laser confocal microscope are non-contact detection methods, enabling non-destructive detection of true defects in the mask substrate, thus saving on mask substrate manufacturing costs. Therefore, the aforementioned mask substrate defect detection method can accurately, efficiently, and non-destructively identify true defects on the mask substrate, meeting the reliability requirements of large-scale production, improving mask substrate production efficiency, and helping to save manufacturing costs.

[0041] On the other hand, the present invention also provides a mask substrate defect detection device for performing the mask substrate defect detection method described in any of the above embodiments. For example... Figure 2 , Figure 3 As shown, a mask substrate defect detection device of one embodiment includes a white light interferometer 11, a laser confocal microscope 12, and a controller 20. The white light interferometer 11 and the laser confocal microscope 12 are arranged adjacent to each other. The controller 20 is electrically connected to both the white light interferometer 11 and the laser confocal microscope 12. The controller 20 is used to determine whether the particles are real particles based on the confocal signal of the particles on the surface of the mask substrate obtained by the confocal microscope 12 and the surface contour data of the mask substrate obtained by the white light interferometer 11.

[0042] Specifically, one embodiment of the controller includes a signal receiving unit, an information extraction unit, an information comparison unit, a signal filtering unit, a particle marking unit, and a storage unit. The signal receiving unit receives the confocal signal of the particles from a laser confocal microscope and the surface contour data of the mask substrate from a white light interferometer. The information extraction unit, connected to the signal receiving unit, extracts the morphological feature information of the particles from the confocal signal and the background morphological feature information of the particle signal points from the surface contour data of the mask substrate. The information comparison unit, connected to the information extraction unit, calculates the deviation difference between the morphological feature information of the particles and the background morphological feature information of the particle signal points, compares the deviation difference with a preset deviation threshold, and outputs the comparison result. The signal filtering unit, connected to the information comparison unit, receives the comparison result from the information comparison unit where the deviation difference is not greater than the preset deviation threshold, and filters out the confocal signal of the particles based on the comparison result. The particle marking unit, connected to the information comparison unit, receives the comparison result from the information comparison unit where the deviation difference is greater than the preset deviation threshold, marks the particles as real particles based on the comparison result, and sends the confocal signal of the particles to the storage unit. The storage unit is connected to the signal receiving unit and the particle marking unit respectively, and is used to store the preset deviation threshold, the surface contour data of the mask substrate and the confocal signal of the real particles.

[0043] The aforementioned mask substrate defect detection device is used to execute any of the mask substrate defect detection methods described above. This device can accurately, efficiently, and non-destructively identify actual defects on the mask substrate, meeting the reliability requirements of large-scale production, improving mask substrate production efficiency, and helping to save manufacturing costs. Furthermore, the device stores surface contour data of the mask substrate and confocal signals of actual particles, facilitating historical data tracking, production control, and further improving production efficiency.

[0044] In one embodiment, the mask substrate defect detection device further includes a display device connected to the controller. The display device is used to display the confocal signal of the actual particles and the surface contour data of the mask substrate. Specifically, the display device can intuitively display the signal information of the actual particles and the surface contour information of the mask substrate, facilitating manual review or auditing of the detection data. Further, the controller and display device can be separate independent controller host and display, or they can be integrated all-in-one machines or tablet computers; this embodiment does not impose specific limitations.

[0045] In one embodiment, the mask substrate defect detection device further includes a robot arm 13, which is electrically connected to the controller 20. The robot arm 13 is used to transfer the mask substrate. Specifically, the robot arm 13 is used to transfer the mask substrate between the laser confocal microscope 12 and the white light interferometer 11, and is used for loading, unloading, and positioning the mask substrate. The robot arm 13 can quickly transfer the mask substrate, which helps to improve detection efficiency.

[0046] In one embodiment, the mask substrate defect detection device further includes: a device body 10 and an FFU 14. The device body 10 has a mounting chamber 16, and a mounting stage 15 is disposed within the mounting chamber 16. A white light interferometer 11 and a laser confocal microscope 12 are respectively disposed on the mounting stage 16, and the FFU 14 is disposed on the top of the device body 10. In this embodiment, the device body 10 and FFU 14 effectively maintain the working cleanliness of the mask substrate defect detection device and prevent the introduction of external particles during the detection process from causing particulate contamination to the surface of the mask substrate. Specifically, to avoid the vibration generated by the FFU during the detection process affecting the detection results, it is preferable to turn off the FFU during the detection process and turn on the FFU during non-detection periods to purify the mask substrate defect detection device, thereby ensuring the working cleanliness of the mask substrate defect detection device.

[0047] Furthermore, to further ensure the cleanliness of the work area, the mounting table 15 is preferably made of stainless steel perforated plate. Even further, in one embodiment, the mask substrate defect detection device further includes an ion generator 17, which is connected to the device body 10. The ion generator 17 is used to eliminate static electricity and prevent electrostatic contamination of the mask substrate.

[0048] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the present invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure shall fall within the scope of the present invention.

Claims

1. A method for detecting defects in a mask substrate, characterized in that, Includes the following steps: A laser confocal microscope was used to scan the mask substrate to be tested, and the confocal signal of the particles on the surface of the mask substrate was obtained. The mask substrate is scanned using a white light interferometer to obtain the surface contour data of the mask substrate; The morphological feature information of the particles is obtained from the confocal signal, and the background morphological feature information of the particle signal points is obtained from the surface contour data of the mask substrate. Calculate the deviation difference between the morphological feature information of the particle and the background morphological feature information of the particle signal point, and compare the deviation difference with a preset deviation threshold; When the deviation difference is greater than a preset deviation threshold, the particle is marked as a real particle; When the deviation difference is not greater than a preset deviation threshold, the confocal signal of the particles is filtered out.

2. The mask substrate defect detection method according to claim 1, characterized in that, The step of acquiring the confocal signal of the particles on the surface of the mask substrate includes: The peak brightness and its coordinate position at the focal plane are identified by shallow depth-of-field scanning, and a focused image is obtained. A three-dimensional image is constructed by integrating all focused images acquired by shallow depth-of-field scanning. The distribution of particles on the surface of the mask substrate and in its thin film stacking structure is analyzed to obtain the confocal signal of the particles.

3. The mask substrate defect detection method according to claim 1, characterized in that, The spot area of ​​the detection beam of the white light interferometer is 1 mm. 2 ~100mm 2 .

4. The mask substrate defect detection method according to claim 1, characterized in that, The laser beam spot area of ​​the laser confocal microscope is no greater than 1 mm. 2 .

5. A mask substrate defect detection device, wherein the mask substrate defect detection device is used to perform the mask substrate defect detection method as described in any one of claims 1 to 4, characterized in that, The mask substrate defect detection device includes a white light interferometer, a laser confocal microscope, and a controller. The white light interferometer and the laser confocal microscope are respectively arranged adjacent to each other. The controller is electrically connected to both the white light interferometer and the laser confocal microscope. The controller is used to determine whether the particles are real particles based on the confocal signal of the particles on the surface of the mask substrate acquired by the confocal microscope and the surface contour data of the mask substrate acquired by the white light interferometer.

6. The mask substrate defect detection device according to claim 5, characterized in that, The controller includes: Signal receiving unit: used to receive the confocal signal of the particles sent by the laser confocal microscope and the surface contour data of the mask substrate sent by the white light interferometer; Information extraction unit: Obtains the morphological feature information of the particles from the confocal signal of the particles, and obtains the background morphological feature information of the particle signal points from the surface contour data of the mask substrate; Information comparison unit: used to calculate the deviation difference between the morphological feature information of the particle and the background morphological feature information of the particle signal point, and compare the deviation difference with a preset deviation threshold; Signal filtering unit: used to receive the comparison result from the information comparison unit, where the deviation difference is not greater than a preset deviation threshold, and filter out the confocal signal of the particles according to the comparison result; Particle marking unit: Used to receive the comparison result from the information comparison unit where the deviation difference is greater than a preset deviation threshold, mark the particle as a real particle according to the comparison result, and send the confocal signal of the particle to the storage unit; Storage unit: Used to store preset deviation thresholds, surface contour data of the mask substrate, and confocal signals of the actual particles.

7. The mask substrate defect detection device according to claim 5, characterized in that, The mask substrate defect detection device further includes a display device, which is connected to the controller and is used to display the confocal signal of the real particles and the surface contour data of the mask substrate.

8. The mask substrate defect detection device according to claim 5, characterized in that, The mask substrate defect detection device further includes a robotic arm, which is connected to the controller and is used to transfer the mask substrate.

9. The mask substrate defect detection device according to claim 5, characterized in that, The mask substrate defect detection device further includes: a device body and an FFU. The device body has a mounting chamber, and a mounting platform is provided in the mounting chamber. The white light interferometer and the laser confocal microscope are respectively mounted on the mounting platform, and the FFU is located on the top of the device body.

10. The mask substrate defect detection device according to claim 9, characterized in that, The mask substrate defect detection device further includes an ion generator, which is connected to the main body of the device.

Images