Defect detection system and defect detection method

Abstract

The invention provides a defect detection system and a defect detection method, the defect detection system comprises an objective table, an illumination channel and a plurality of detection channels, the illumination channel comprises a light source, and the light source is used for providing detection light spots for a detection area of a to-be-detected object. The detection channel comprises an objective lens and a detector, the objective lens is used for collecting scattered light generated by the detection light spot in the detection area, and the detector is used for determining the signal intensity of the scattered light so as to determine defect information of the detection area based on the signal intensity. The incident plane is perpendicular to a second plane where the optical axes of the detection channels are located, deviation angles exist between the optical axes of the detection channels and the normal direction of the surface of the to-be-detected object, and at least two deviation angles in all the deviation angles meet the angle difference condition. According to the device, scattered light of different types of defects can be efficiently collected, concave defects and convex defects can be accurately detected through a plurality of detection channels, missing detection can be avoided, and the two types of defects existing on the surface of an object to be detected can be accurately detected.

Description

Technical Field

[0001] This application relates to the field of semiconductor testing, and in particular to a defect detection system and defect detection method. Background Technology

[0002] Defect detection involves the detection and location of physical and chemical defects such as particles and scratches on the surface of an object under test. It directly impacts chip quality and yield, and is a crucial step in the semiconductor manufacturing process. Defect detection can be performed in dark-field conditions, where the area under test is illuminated and scattered light is collected to determine the presence or absence of defects. Alternatively, a high-precision moving platform can be used to move the object under test, thus enabling defect detection in different areas of the surface. Any contaminants (such as foreign objects or particles generated during the manufacturing process) or defects (such as scratches and dents) on a semiconductor chip with physical dimensions close to or larger than the final product feature size are harmful and must be strictly monitored throughout the manufacturing process. Therefore, surface defect detection on patternless wafers has important applications in many areas of the semiconductor industry, including: product quality inspection by wafer manufacturers; incoming inspection by chip manufacturers; contamination monitoring of process equipment and processes; and equipment contamination inspection by semiconductor equipment manufacturers. The sources of different types of contamination vary. Generally, particulate protrusion defects mainly originate from the environment inside the semiconductor equipment cavity, pit defects mainly originate from epitaxy and crystal pulling processes, and scratch defects originate from errors in the polishing process. In defect detection, obtaining information on the type of defect helps users in different fields efficiently pinpoint the source of defects, effectively reducing the number of repeated inspections required. Currently, accurate detection of protrusion and pit defects is not possible. Summary of the Invention

[0003] In view of this, the purpose of this application is to provide a defect detection system and method that can accurately detect dents and protrusions, avoiding missed detections. The specific solution is as follows:

[0004] On the one hand, this application provides a defect detection system, which includes:

[0005] A stage; the stage is used to hold the object to be measured.

[0006] An illumination channel and multiple detection channels are located on the same side (upper side of the stage); the illumination channel includes a light source for providing a detection spot to the detection area of ​​the object under test; the detection channel includes an objective lens and a detector, the objective lens for collecting scattered light generated by the detection spot in the detection area, and the detector for determining the signal intensity of the scattered light so as to determine the defect information of the detection area based on the signal intensity;

[0007] The incident surface is perpendicular to the second plane, which is the plane containing the optical axis of the incident light emitted by the light source and the normal direction of the surface of the object to be tested, and the second plane is the plane containing the optical axis of the plurality of detection channels;

[0008] The optical axis of the detection channel has an offset angle from the normal direction of the surface of the object to be tested, and at least two of the offset angles satisfy the angle difference condition.

[0009] Optionally, the detection spot is a dot spot.

[0010] Optionally, the detection spot is a linear spot, and the first plane is perpendicular to the second plane; the first plane is the plane containing the optical axis of the illumination channel and the long side of the detection spot.

[0011] Optionally, the detector has a photosensitive surface, the long side of which is perpendicular to the second plane.

[0012] Optionally, the short side width of the detection spot is greater than or equal to 10 micrometers and less than or equal to 30 micrometers.

[0013] Optionally, the plurality of detection channels include a first detection channel and a second detection channel, wherein the optical axis of the first detection channel has a first deviation angle from the normal direction, and the optical axis of the second detection channel has a second deviation angle from the normal direction, wherein the first deviation angle is less than 45°, and the second deviation angle is greater than or equal to 45°.

[0014] Optionally, the angle between the optical axis of the illumination channel and the normal direction of the surface of the object under test is less than or equal to 90°.

[0015] Optionally, when the illumination channels include multiple channels, the detection light spots of all the illumination channels overlap.

[0016] Optionally, when the lighting channel includes multiple light sources, all of the light sources emit light beams with the same wavelength, or at least two of the light sources emit light beams with different wavelengths.

[0017] Optionally, when the illumination channel includes multiple light sources, and at least two of the light sources emit light beams with different wavelengths, the detection channel includes a filter module for filtering the scattered light entering the detector.

[0018] Optionally, the illumination channel includes a polarization adjustment module, which is used to adjust the polarization state of the light beam emitted by the light source.

[0019] Optionally, the illumination channel includes a shaping module for shaping the light beam emitted by the light source to obtain the detection spot.

[0020] Furthermore, embodiments of this application also provide a defect detection method applied to a defect detection system. The defect detection method includes:

[0021] During the rotation of the stage, defects are detected in different areas of the object to be tested located on the stage through the illumination channel and the multiple detection channels.

[0022] When performing defect detection on the detection area of ​​the object under test, the light source provides a detection spot to the detection area through the illumination channel. The scattered light generated by the detection spot in the detection area passes through the objective lens and is received by the detector. The detector determines the signal intensity of the scattered light and determines the defect information of the detection area based on the signal intensity.

[0023] Optionally, the plurality of detection channels include a first detection channel and a second detection channel. The optical axis of the first detection channel has a first deviation angle from the normal direction, and the optical axis of the second detection channel has a second deviation angle from the normal direction. The first deviation angle is smaller than the second deviation angle. Determining the defect information of the detection area based on the signal intensity includes:

[0024] Based on the first initial signal strength obtained by the first detector of the first detection channel and the second initial signal strength obtained by the second detector of the second detection channel, the signal strength ratio of the first detection channel and the second detection channel is determined.

[0025] When the signal strength ratio is greater than a preset threshold, first defect information of the detection area is determined; the first defect information is used to identify a dent-type defect in the detection area.

[0026] When the signal strength ratio is less than or equal to the preset threshold, second defect information of the detection area is determined; the second defect information is used to identify a protrusion-type defect in the detection area.

[0027] Optionally, determining the signal strength ratio between the first detection channel and the second detection channel based on the first initial signal strength obtained by the first detector of the first detection channel and the second initial signal strength obtained by the second detector of the second detection channel includes:

[0028] The noise signal intensity of the test object is removed from the first initial signal intensity obtained from the first detector of the first detection channel to obtain the first signal intensity of the first detection channel;

[0029] The noise signal intensity is removed from the second initial signal intensity obtained from the second detector of the second detection channel to obtain the second signal intensity of the second detection channel;

[0030] Based on the first signal strength and the second signal strength, the signal strength ratio of the first detection channel and the second detection channel is determined.

[0031] Optionally, determining the signal strength ratio between the first detection channel and the second detection channel based on the first signal strength and the second signal strength includes:

[0032] Based on the system energy difference between the first detection channel and the second detection channel, the compensation parameters are determined;

[0033] Based on the first signal strength, the second signal strength, and the compensation parameter, the signal strength ratio between the first detection channel and the second detection channel is determined.

[0034] Optionally, determining the defect information of the detection area based on the signal intensity includes:

[0035] Determine the calibration curve corresponding to the detection channel; the calibration curve is used to represent the relationship between the defect size of the test object and the signal intensity;

[0036] The signal intensity obtained by the detector is matched with the calibration curve to obtain the defect information of the detection area; the defect information includes the defect size of the defect in the detection area.

[0037] Optionally, the first signal strength for:

[0038]

[0039] Where, ∑ The initial signal strength is... The noise signal intensity of the test object in the first detection channel;

[0040] Second signal strength for:

[0041]

[0042] in, The second initial signal strength, The noise signal intensity of the test object in the second detection channel;

[0043] The signal strength ratio K is:

[0044] K =I SN / (I SW -B)

[0045] Wherein, B is the compensation parameter.

[0046] Optionally, the method further includes:

[0047] Determine the detection time corresponding to the detection image generated by the detector based on the scattered light;

[0048] Based on the detection time and the movement trajectory of the detection spot on the object under test during the detection time period, the location of the area to be tested in the object under test is determined.

[0049] Based on the location of the region and the location of the defect in the detection image, the defect location in the object under test is determined.

[0050] This application provides a defect detection system and method. The defect detection system includes a stage, an illumination channel, and multiple detection channels. The stage carries the object to be tested. The illumination channel and the multiple detection channels are located on the same side of the stage. The illumination channel includes a light source that provides a detection spot to the detection area of ​​the object to be tested. Each detection channel includes an objective lens and a detector. The objective lens collects scattered light generated by the detection spot in the detection area, and the detector determines the signal intensity of the scattered light to determine defect information in the detection area based on the signal intensity. The incident surface is perpendicular to a second plane, which is the plane containing the optical axis of the incident light emitted from the light source and the normal direction of the surface of the object to be tested. The second plane is the plane containing the optical axes of the multiple detection channels. The optical axis of the detection channel has an offset angle from the normal direction of the surface of the object to be tested, and at least two of the offset angles satisfy an angle difference condition. In this way, by ensuring that at least two detection channels have a significant angular difference between their deviation angles—meaning that multiple detection channels maintain specific and asymmetrical angles with the normal direction of the object's surface—the spatial distribution of these detection channels becomes more diffuse. This allows for efficient collection of scattered light from different types of defects. Multiple detection channels enable accurate detection of both concave and convex defects, avoiding missed detections and ensuring precise detection of both types of defects on the object's surface. Furthermore, it avoids the risk of contact between detection channels due to similar deviation angles, and allows for greater flexibility in the layout of each detection channel. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This paper shows a schematic diagram of the structure of a defect detection system provided in an embodiment of the present application;

[0053] Figure 2 This illustration shows a schematic diagram of the relative positions of planes in a defect detection system provided in an embodiment of this application;

[0054] Figure 3 This paper shows a schematic diagram of the structure of another defect detection system provided in an embodiment of the present application;

[0055] Figure 4 This is a side view of a defect detection system provided in an embodiment of this application;

[0056] Figure 5 A flowchart illustrating a defect detection method provided in an embodiment of this application is shown;

[0057] Figure 6 This illustration shows a schematic diagram of the signal strength in detection channels of different types and sizes in different directions, according to an embodiment of this application.

[0058] Figure 7 This illustration shows a schematic diagram of signal strength in detection channels of different directions for defects of different types and sizes, as provided in another embodiment of this application.

[0059] Figure 8 A scatter plot showing the experimental measurement results of a surface defect of a test object according to an embodiment of this application is provided.

[0060] Figure 9 A schematic diagram of a calibration curve provided in an embodiment of this application is shown;

[0061] Figure 10 A schematic diagram of a lighting channel provided in an embodiment of this application is shown;

[0062] Figure 11 A schematic diagram illustrating the signal intensity of scattered light from different types of defects, provided for an embodiment of this application;

[0063] Figure 12 A schematic diagram of a detection channel provided in an embodiment of this application;

[0064] Figure 13 A schematic diagram of the signal distribution of scattered light from a protrusion-type defect is provided in an embodiment of this application;

[0065] Figure 14 This is a schematic diagram of the signal distribution of scattered light from a concave defect, provided as an embodiment of this application.

[0066] Figure label:

[0067] Stage-101, Object to be tested-102, Illumination channel-103, First detection channel-104, Second detection channel-105, Light source-106, Objective lens-107, Detector-108, Detection spot-109, First plane-201, Second plane-202, Polarization adjustment module-110, Shaping module-111. Detailed Implementation

[0068] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the specific embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0069] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, this application may also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0070] Secondly, this application provides a detailed description in conjunction with schematic diagrams. When detailing the embodiments of this application, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of this application. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0071] As described in the background section, current related technologies cannot accurately detect both protrusion and dent defects. Based on these technical problems, embodiments of this application provide a defect detection system and method. By setting multiple detection channels at specific and asymmetrical angles to the normal direction of the surface of the object under test, the spatial distribution of these detection channels becomes more dispersed. Through multiple detection channels, accurate detection of both dent and protrusion defects can be achieved, avoiding missed detections and enabling precise detection of both types of defects present on the surface of the object under test.

[0072] For ease of understanding, the following detailed description, in conjunction with the accompanying drawings, provides an embodiment of a defect detection system and a defect detection method provided in this application.

[0073] refer to Figure 1The diagram shown is a structural schematic of a defect detection system provided in an embodiment of this application. The defect detection system includes a stage 101, an illumination channel 103, and multiple detection channels.

[0074] The stage 101 is used to support the test object 102, which is an object that needs to be defect-detected. The test object 102 is, for example, a wafer. The test object 102 can be placed on the stage 101.

[0075] The lighting channel 103 and multiple detection channels are located on the same side of the stage 101, which facilitates defect detection. For example, if the object to be tested 102 is placed above the stage 101, then the lighting channel 103 and multiple detection channels are also located above the stage 101.

[0076] The illumination channel 103 may include a light source 106, which provides a detection spot 109 to the detection area of ​​the object under test 102. The specific type of the light source 106 is not limited here; for example, it can be a laser light source 106. The shape of the detection spot 109 is also not limited; for example, it can be a dot spot, a line spot, or an elliptical spot. The diameter of the dot spot is in the range of tens to hundreds of micrometers. The width of the long side of the detection line spot 109 can range from 0.5 to 20 mm, and the width of the short side can range from 1 to 100 μm. The detection area is the region of the object under test 102 where defect detection is required. The detection area can be a part or the entire surface of the object under test 102. Typically, when performing defect detection on the detection area, the detection spot 109 emitted by the light source 106 falls within the detection area.

[0077] Furthermore, the illumination channel 103 can be vertically or obliquely positioned, and the number of illumination channels 103 can be one or more; this application does not limit this. When there are multiple illumination channels, a vertically positioned illumination channel 103 can achieve vertical light incidence, while an obliquely positioned illumination channel 103 can achieve oblique light incidence. This allows for the combined use of vertical and oblique illumination methods, thereby enriching the scattering excitation conditions of the object under test and improving the full coverage capability of defect detection.

[0078] Specifically, the vertically arranged illumination channel 103 can emit a detection beam along the normal direction of the surface of the object under test 102, forming a symmetrical vertical illumination field on the surface of the object under test 102. This is suitable for exciting targets with obvious forward scattering, such as surface microparticles and shallow concave defects. The incident direction has high incident uniformity and can form a regular and strong scattering signal, which is beneficial to the signal extraction and imaging stability of subsequent detection channels.

[0079] The inclined illumination channel 103 can emit a detection beam at an oblique angle, so that the detection beam is incident on the surface of the test object 102 at a certain angle. It forms oblique scattering with the structural edges or concave-convex interfaces of the test object 102, which helps to excite asymmetric signals of defects with large surface contour changes or more significant lateral scattering (such as pit edges, protruding particle sidewalls, etc.). This incident direction significantly expands the distribution range of signal light scattering direction and improves the response capability of the detection system to different types of defects.

[0080] In practical applications, when it is necessary to detect defects in the entire surface area or part of the surface area of ​​the object to be tested 102, the detection spot 109 can be moved so that the detection spot 109 can traverse the entire surface area or part of the surface area to achieve defect detection.

[0081] The detection channel may include an objective lens 107 and a detector 108. The detector 108 is located on the light-emitting side of the objective lens 107, and may be, for example, a camera discrete photomultiplier tube (PMT), a charge-coupled device (CCD), a time delay integration (TDI), a complementary metal oxide semiconductor (CMOS) sensor, a PMT array, an electron bombardment CCD (EB-CCD), an electron-multiplying CCD (EM-CCD), an enhancement photodiode, or an avalanche photodiode (APD) array. The objective lens 107 is used to collect the scattered light generated by the detection spot 109 in the detection area, and the detector 108 is used to determine the signal intensity of the scattered light so as to determine the defect information of the detection area based on the signal intensity. Each detection channel can be equipped with one or more detectors 108, and the detection surface (i.e., photosensitive surface) of the detector 108 is parallel to the detection spot 109. Preferably, the detector 108 can be a line detector 108, with the detection surface perpendicular to the second plane 202. In addition, the detection channel can also be equipped with a switching device to switch between objectives 107 with different magnifications, thereby adjusting the detection accuracy and speed of the system. The lower the magnification, the faster the detection speed, but the lower the accuracy.

[0082] In other words, the light beam in the illumination channel 103 is incident on the detection area to form a detection spot 109. The scattered light generated by the detection spot 109 is received by the detector 108 through the objective lens 107. The detector 108 can determine the defect information of the detection area based on the signal intensity of the scattered light. Specifically, the signal intensity of the scattered light is significantly different when there is a defect in the detection area and when there is no defect. Therefore, the defect information can be determined by collecting the scattered light. The defect information may include information such as the size, type, and location of the defect.

[0083] Combination Figure 1 The defect detection principle is explained as follows: When the detection spot 109 illuminates a defect on the surface of the test object 102, it generates scattered light in the surrounding space. The distribution and intensity of the scattered light are related to the type and size of the defect. At this time, the objective lens 107 and detector 108 of the detection channel are used to capture and collect these scattered light signals and analyze them to finally determine the location, size, and type (pit or bump) of the defect on the wafer. Among them, the location information is given by scanning combined with imaging detection, and the size and defect type are obtained by joint judgment of the dual-channel signal intensity.

[0084] The incident surface is perpendicular to the second plane 202. The incident surface is the plane containing the optical axis of the incident light emitted from the light source and the normal direction of the surface of the object under test, i.e., the plane containing the incident light. The second plane 202 is the plane containing the optical axes of multiple detection channels. (Reference) Figure 2 As shown, the second plane 202 is illustrated, and the incident plane is the plane where the incident light emitted from the illumination channel 103 is located. Figure 2 The location of the first plane 201 is the location of the incident surface. By setting the incident surface to be perpendicular to the second plane 202, the subsequent defect detection process can be realized.

[0085] There is a deviation angle between the optical axis of the detection channel and the normal direction of the surface of the object to be tested 102, and at least two of the deviation angles satisfy the angle difference condition.

[0086] The deviation angle is the angle between the optical axis of the detection channel and the normal direction of the surface of the object under test 102. Satisfying the angle difference condition can be understood as the angle difference between two deviation angles being greater than a preset threshold, i.e., the difference between the two deviation angles is significant. When at least two deviation angles satisfy the angle difference condition, it indicates that at least two detection channels have a significant difference in tilt. For example, some detection channels have a deviation angle of 30°, and some have a deviation angle of 80°; these two detection channels satisfy the angle difference condition.

[0087] In this way, by ensuring that at least two detection channels have a significant angular difference between their deviation angles—meaning that multiple detection channels maintain specific and asymmetrical angles with the normal direction of the object's surface—the spatial distribution of these detection channels becomes more diffuse. This allows for efficient collection of scattered light from different types of defects. Multiple detection channels enable accurate detection of both concave and convex defects, avoiding missed detections and ensuring precise detection of both types of defects on the object's surface. Furthermore, it avoids the risk of contact between detection channels due to similar deviation angles, and allows for greater flexibility in the layout of each detection channel.

[0088] Since multiple objectives are used to inspect the same area at the same time, at least one objective's optical axis is not perpendicular to the surface of the object 102. For non-point spot scanning methods, such as line spot or elliptical spot scanning, some positions in the area to be inspected are prone to defocusing, which affects the accuracy of defect detection.

[0089] To address the defocusing issue during linear illumination, in one possible implementation, the detection spot can be a linear spot. The first plane 201 is the plane containing the optical axis of the illumination channel 103 and the long side of the detection spot 109, and the second plane 202 is the plane containing the optical axes of multiple detection channels. The first plane 201 is perpendicular to the second plane 202.

[0090] refer to Figure 2 The diagram shown is a schematic representation of the relative positions of planes in a defect detection system according to an embodiment of this application. To clearly illustrate the first plane 201 and the second plane 202, [the diagram is shown in the original text]. Figure 2 The diagram provides a simplified representation of the lighting channel 103 and the detection channel, with the lighting channel 103 indicated by a dashed line and the detection channel by a solid line. Figure 2 The diagram shows two detection channels and two illumination channels 103.

[0091] The first plane 201 is the plane formed by the optical axis of the light beam and the long side of the detection spot 109, and the second plane 202 is the plane formed by the optical axes of the cameras of multiple detection channels. For example... Figure 2 As shown, both detection channels are located within the second plane 202. The first plane 201 is coplanar with the long side of the detection spot 109, and the two planes are perpendicular to each other.

[0092] In addition, by setting all detection channels to be coplanar, that is, their optical axes are all located within the second plane 202, it can be ensured that the detection spot 109 is always clearly focused, thereby improving detection accuracy.

[0093] Next, the relationship between the first plane 201, the second plane 202, and the incident plane will be explained. The incident plane and the first plane 201 are different planes; they can be located at the same position or at different positions. The incident plane is the plane containing the incident light, and the first plane 201 is the plane containing the long side of the line spot. That is to say, when the detected light spot is a line spot, the incident plane and the first plane 201 are not necessarily at the same position. Figure 2 As an example, when the incident surface is at the location of the first plane 201, by adjusting the component parameters in the illumination channel 103, it can be achieved that... Figure 2 The line spot rotates, which in turn causes the position of the first plane 201, where the long side of the line spot is located, to change.

[0094] In this application, when the detected light spot is a linear light spot, to solve the defocusing problem, the first plane 201 is set to be perpendicular to the second plane 202. Since the incident surface is also perpendicular to the second plane 202, the position of the first plane 201 is the same as the position of the incident surface; for example, both are perpendicular. Figure 2 The location of the first plane 201. In this application, by setting two planes perpendicular to each other, the long side direction of the detection spot 109 is perpendicular to the second plane 202. Furthermore, since the optical axes of the detection channels in the second plane 202 all intersect at a single point, and this point lies on the intersection line of the two planes, the long side direction of the detection spot 109 is perpendicular to the optical axis of the detection channel (i.e.,...). Figure 2 The long side of the detection spot 109 is perpendicular to the two solid lines within the second plane 202. Since the optical axis of the detection channel is perpendicular to the light-gathering surface of the objective lens 107, the long side of the detection spot 109 is parallel to the light-gathering surface of the objective lens 107. Consequently, the two endpoints of the detection spot 109 along its long side are equidistant from the objective lens 107. Therefore, the entire detection spot 109 is located on the object-side focal plane of the objective lens 107, preventing defocusing. The entire detection spot 109 can be focused within the detector, ensuring that all positions within the line area illuminated by the spot are in focus. This guarantees the accuracy of imaging reception at each position within the line area, thus ensuring the accuracy of defect detection and improving the utilization rate of the detection spot 109. In short, the detection spot 109 is parallel to the first plane 201, and the detector 108 is perpendicular to its photosensitive surface and parallel to the long side of the detection spot 109, thereby achieving focusing. Furthermore, the dual-channel detection system overcomes the limitation of single-channel systems in detecting pits and bumps. Compared to three or four channels, dual-channel systems occupy less space and allow for the use of a large NA objective lens (107) to enhance detection capabilities. In addition, the defect detection system features an online learning mechanism that continuously updates model parameters based on newly detected defect data, improving the model's adaptability and generalization ability.

[0095] Next, combined Figure 3The relative positions of the first plane 201 and the second plane 202 will be explained in detail again.

[0096] refer to Figure 3 Figure (b) is a top view of the defect detection system, illustrating the example of a beam incident directly onto the surface of the object under test 102. Only the detection channel on the right side is shown. The detection channel includes an objective lens 107 and a detector 108. The light source 106 is incident from above. To more clearly illustrate the relative positions of the beam and its components, the directly incident beam emitted by the light source 106 is shown in (b). It can be seen that the light source 106 forms a detection spot 109 on the surface of the object under test 102. The detection spot 109 is a linear spot. The first plane 201 and the second plane 202 are perpendicular to each other, so the distance from the two endpoints of the long side of the detection spot 109 to the objective lens 107 is the same, thus preventing defocusing. All of the detection spots 109 can be used for detection and imaging.

[0097] refer to Figure 3 Figure (a) is a schematic diagram of the defect detection system, illustrating the example of a beam obliquely incident on the surface of the object under test 102. Only the detection channel on the right side is shown. The detection channel includes an objective lens 107 and a detector 108. A light source 106 obliquely incidents from above, forming a detection spot 109 on the surface of the object under test 102. The detection spot 109 is a line spot. The first plane 201 and the second plane 202 are not perpendicular. At this time, the two endpoints A and B of the detection spot 109 are at different distances from the objective lens 107. Point A is closer to the objective lens 107, and point B is farther away. Therefore, the images of both ends of the detection spot 109 on the camera will be out of focus; only the center point of the detection spot 109 is in focus, resulting in extremely low spot utilization. Therefore, ensuring that the first plane 201 and the second plane 202 are perpendicular prevents the detection spot 109 from easily defocusing, increases its utilization, and improves system stability and yield.

[0098] It is worth noting that whether the beam is incident normally or obliquely does not affect the degree of focusing of the detection spot 109 in the detector. The main reason that affects the degree of focusing is whether the first plane 201 and the second plane 202 are perpendicular.

[0099] In other possible implementations, when the detection spot is a point spot, the first plane 201 can be set to be perpendicular to the second plane 202, thereby enabling more precise focusing when using the point spot for defect detection. In this case, the first plane 201 can be considered as the plane containing the optical axis of the illumination channel and the normal direction of the surface of the object under test 102.

[0100] In one possible implementation, detector 108 has a photosensitive surface whose long side is perpendicular to the second plane 202.

[0101] Specifically, the photosensitive surface of detector 108 is used to collect the scattered light from detection spot 109. When the length and width of the photosensitive surface are different, in order to achieve clear imaging, the long side direction of the photosensitive surface can be made perpendicular to the second plane 202, so that the long side direction of the photosensitive surface is parallel to the first plane 201 and parallel to the long side direction of the detection spot 109. This allows the scattered light from detection spot 109 to be fully collected by the photosensitive surface, increasing the amount of scattered light entering detector 108, achieving a complete and clear image of detection spot 109, and thus improving detection efficiency. In one possible implementation, the angle between the optical axis of illumination channel 103 and the normal direction of the surface of object 102 is less than or equal to 90°.

[0102] When the object under test 102 is placed horizontally on the stage 101, the normal direction of the surface of the object under test 102 is vertical. When the angle between the optical axis of the illumination channel 103 and the normal direction of the surface of the object under test 102 is 90°, the light source 106 of the illumination channel 103 is parallel to the surface of the object under test 102, and the detection light spot 109 is perpendicularly incident on the surface of the object under test 102, achieving vertical illumination. (Reference) Figure 1 As shown, an illumination channel 103 is provided directly above the object to be tested 102, and the light emitted by the light source 106 (shown by dashed lines) is incident perpendicularly onto the surface of the object to be tested 102.

[0103] When the angle between the optical axis of the illumination channel 103 and the normal direction of the surface of the object under test 102 is less than 90°, the light source 106 of the illumination channel 103 is not parallel to the surface of the object under test 102, and the detection spot 109 is obliquely incident on the surface of the object under test 102, achieving oblique illumination. (Reference) Figure 1 As shown, an illumination channel 103 is provided obliquely above the object to be tested 102, and the light emitted by the light source 106 (shown by dashed lines) is obliquely incident on the surface of the object to be tested 102. In addition, when irradiated at an oblique angle, the angle between the optical axis of the illumination channel 103 and the normal direction of the surface of the object to be tested 102 can be in the range of 60° to 80° in order to improve the detection accuracy.

[0104] In practical applications, the detection spot 109 can be selected to irradiate the surface of the object to be tested 102 vertically or obliquely, depending on the actual needs. Of course, multiple illumination channels 103 can also be set, some of which are vertically incident and some of which are obliquely incident.

[0105] Furthermore, when the first plane 201 and the second plane 202 are not perpendicular, there is a risk of contact between the illumination channel 103 and the objective lens 107 in the detection channel. To avoid this contact, the placement and angle of the objective lens 107 must be restricted. However, when the first plane 201 and the second plane 202 are perpendicular, it is less likely for the illumination channel 103 and the objective lens 107 in the detection channel to contact each other. The layout of the detection channel is unrestricted, and a larger numerical aperture (NA) objective lens 107 can be used in the detection channel (the larger the NA, the larger the size of the objective lens 107). The larger the NA, the wider the range of collected scattered light angles, the higher the object-side resolution, the stronger the detection, and the further improved the detection accuracy.

[0106] In one possible implementation, when there are multiple illumination channels 103, the detection spots 109 of all illumination channels 103 overlap.

[0107] In other words, when the defect detection system includes multiple illumination channels 103, the detection spots 109 of all illumination channels 103 overlap, resulting in only one illumination spot on the surface of the object under test 102. This illumination spot is formed by the convergence of multiple illumination spots, resulting in higher spot quality, such as higher brightness, thereby improving the defect detection accuracy of the detection area. Furthermore, the overlap of detection spots 109 in this application does not limit the detection spots 109 to complete overlap; they only need to be within the overlapping area, or substantially completely overlap, thus taking into account the overlap error caused by the system differences of different illumination systems.

[0108] In one possible implementation, when the lighting channel 103 includes multiple channels, all light sources 106 emit light beams with the same wavelength, or at least two of the light sources 106 emit light beams with different wavelengths.

[0109] Specifically, when the defect detection system has multiple illumination channels 103, each illumination channel 103 has a light source 106. The wavelengths of the light beams emitted by all the light sources 106 can be equal. In this case, when performing defect detection, these light sources 106 can be controlled to irradiate at different time periods, thereby avoiding mutual interference between the light sources 106 when irradiating at the same time period and ensuring the accuracy of detection.

[0110] Of course, at least two of the light sources 106 can emit beams with different wavelengths. Furthermore, all the light sources 106 can emit beams with different wavelengths, thereby increasing the richness of the light sources 106 and enabling defect detection using light of different wavelengths.

[0111] In one possible implementation, when the illumination channel 103 includes multiple light sources 106, and at least two of the light sources 106 emit light beams with different wavelengths, the detection channel includes a filter module for filtering the scattered light entering the detector.

[0112] Specifically, the detection channel may include a filter module, which detects the scattered light generated by the beam in the detection area when using multiple wavelengths of light for defect detection. The filter module can retain only one wavelength of scattered light, thereby achieving separation between multiple wavelengths of scattered light and ensuring that a detection channel only detects one wavelength of scattered light, thus guaranteeing the detection accuracy of each detection channel.

[0113] In one possible implementation, the multiple detection channels include a first detection channel 104 and a second detection channel 105. The optical axis of the first detection channel 104 has a first deviation angle between itself and the normal direction, and the optical axis of the second detection channel 105 has a second deviation angle between itself and the normal direction. The first deviation angle is less than 45°, and the second deviation angle is greater than or equal to 45°.

[0114] The first detection channel 104 and the second detection channel 105 are any two of the total detection channels. The first detection channel 104 has a smaller first deviation angle, i.e., less than 45°, for example, 30°. The second detection channel 105 has a larger second deviation angle, i.e., greater than or equal to 45°, for example, 60°. For ease of description, the first detection channel 104 can also be referred to as the narrow channel, or simply the N channel, and the second detection channel 105 can be referred to as the wide channel, or simply the W channel.

[0115] In this way, by setting the first deviation angle of the narrow channel to be less than 45° and the second deviation angle of the wide channel to be not less than 45°, it is possible to detect the same detection area at different deviation angles, thereby improving detection accuracy and avoiding the situation where all detection channels are excessively close to or far from the normal direction of the surface of the object under test 102, resulting in detection results limited to a certain deviation angle range and poor detection accuracy. Furthermore, during multi-channel signal fusion, fusing the W channel, N channel signals, and polarization information can greatly improve the detection capability and distinguish between pits and protrusions.

[0116] In one possible implementation, the short side width of the detection spot 109 can be greater than or equal to 10 micrometers and less than or equal to 30 micrometers.

[0117] refer to Figure 4The figure shows a side view of a defect detection system provided in an embodiment of this application. It illustrates an inclined detection channel and the object to be tested 102. When the narrow side length (or linewidth) of the detection spot 109 is the length of segment AB, the distance L from the endpoint to the object-side focal plane is length A1. When the linewidth is the length of segment CD, the distance L from the endpoint to the object-side focal plane is length C4. Since the linewidth CD is longer than the linewidth AB, and the distance C4 from the endpoint of linewidth CD to the object-side focal plane is longer than the distance A1 from the endpoint of linewidth AB to the object-side focal plane, the detection spot 109 with a linewidth of CD exhibits more severe defocusing.

[0118] In other words, by setting the short side width of the detection spot 109 to be smaller, even if the objective lens 107 or detector 108 of the detection channel has a small angle of deflection in the narrow side direction, the endpoint of the narrow side of the detection spot 109 will be smaller than the object-side focal plane of the objective lens 107 due to the smaller short side width. Therefore, even if there is a small angle of shift in the narrow side direction, the resulting distance L deviation will not be too large. That is, after the shift, the distance between the endpoint of the narrow side of the detection spot 109 and the object-side focal plane of the objective lens 107 will still be small, preventing defocusing. In short, a wider linewidth leads to more defocusing, and conversely, a shorter linewidth makes defocusing less likely, thereby improving focusing accuracy, image quality, and stability.

[0119] In one possible implementation, the illumination channel 103 may include a polarization adjustment module 110, which is used to adjust the polarization state of the light beam emitted by the light source 106.

[0120] Specifically, the beam emitted by the light source 106 passes through a polarization conversion module to adjust its polarization state. For example, it can be adjusted to linearly polarized light, circularly polarized light, ellipsoidally polarized light, left-handedly polarized light, right-handedly polarized light, etc., and can be set according to actual needs. As an example, the polarization conversion module can be an optical element such as a polarizer or a half-wave plate.

[0121] This allows for defect detection using beams with various polarization states, improving detection accuracy. Furthermore, the polarization adjustment module 110 can enhance particle signals. Some particles are more sensitive to light with specific polarization directions; rotating the polarization direction enhances the particle's scattered signal, increasing detection sensitivity. This means increasing signal strength and improving the signal-to-noise ratio (SNR). The polarization adjustment module 110 can also suppress background noise. Surface roughness and thin-film interference on the test object 102 generate background noise; rotating the polarization direction suppresses background noise in specific directions, improving the SNR. This means reducing noise intensity and increasing the SNR.

[0122] In one possible implementation, the illumination channel 103 may include a shaping module 111, which is used to shape the light beam emitted by the light source 106 to obtain the detection spot 109. Figure 1 The polarization adjustment module 110 and the shaping module 111 are not shown in the inclined lighting channel 103.

[0123] Specifically, the light emitted by the light source 106 can sequentially pass through a polarization conversion module and a beam shaping module 111. The shaping module 111 can be a collimating flat-top shaping element, or it can be a freeform surface shaping element. The freeform surface shaping element can provide single-degree-of-freedom linewidth compression along the narrow side direction. The shaping module 111 can be implemented as a diffractive optical element (DOE), a microlens array, a prism, etc. The shaping module 111 can also include a one-dimensional beam expander, a collimating flat-top shaping element, a freeform surface shaper, or any other suitable shaping module 111 known in the art. As an example, the shaping module 111 can shape a circular detection spot 109 into a linear detection spot 109.

[0124] refer to Figure 5 The diagram shown is a flowchart of a defect detection method provided in an embodiment of this application, which is applied to a defect detection system and may include the following steps.

[0125] S101, during the rotation of the stage 101, defects are detected in different areas of the object to be tested 102 located on the stage 101 through the illumination channel 103 and multiple detection channels.

[0126] When performing defect detection on different areas of the surface of the object under test 102, the stage 101 can rotate the object under test 102, so that the detection spot 109 illuminates each area of ​​the surface of the object under test 102, providing incident light to the area to be tested. At this time, the detection spot 109 emitted by the illumination channel 103 is incident on the surface of the object under test 102, and the detection channel collects the scattered light for detection.

[0127] S102, when performing defect detection in the detection area of ​​the test object 102, the light source 106 provides a detection spot 109 to the detection area through the illumination channel 103. The scattered light generated by the detection spot 109 in the detection area passes through the objective lens 107 and is received by the detector 108. The detector 108 determines the signal intensity of the scattered light and determines the defect information of the detection area based on the signal intensity.

[0128] Specifically, during the defect detection process in the detection area, the light source 106 of the illumination channel 103 emits a light beam, which is incident on the detection area to form a detection spot 109. The detection spot 109 generates scattered light in the detection area. The scattered light is received by the detector 108 through the objective lens 107. The detector 108 can determine the signal intensity of the scattered light. Since the signal intensity and the distribution of the scattered light are related to information such as the type and size of the defect, the defect information of the detection area can be determined based on the signal intensity.

[0129] In summary, when using a defect detection system with a significant angular difference between the deviation angles of at least two detection channels, the spatial distribution of these channels is more dispersed because they maintain specific and asymmetrical angles with the normal direction of the surface of the object under test. This allows for accurate detection of both dent and protrusion defects, avoiding missed detections and ensuring precise detection of both types of defects on the surface of the object. Furthermore, it avoids the risk of contact between detection channels due to similar deviation angles, and allows for greater flexibility in the layout of each detection channel.

[0130] In one possible implementation, the multiple detection channels include a first detection channel 104 and a second detection channel 105. The optical axis of the first detection channel 104 has a first deviation angle between itself and the normal direction, and the optical axis of the second detection channel 105 has a second deviation angle between itself and the normal direction. The first deviation angle is smaller than the second deviation angle. The defect information of the detection area is determined based on the signal intensity, which may include S1021-S1023.

[0131] S1021, based on the first initial signal strength obtained by the first detector 108 of the first detection channel 104 and the second initial signal strength obtained by the second detector 108 of the second detection channel 105, determine the signal strength ratio between the first detection channel 104 and the second detection channel 105.

[0132] The first initial signal intensity is the sum of the scattered light intensities output by the first detector 108 in the first detection channel 104, and the second initial signal intensity is the sum of the scattered light intensities output by the second detector 108 in the second detection channel 105. Since the first detection channel 104 is a narrow channel and the second detection channel 105 is a wide channel, the two detection channels collect scattered light from different directions. Based on the first and second initial signal intensities, the signal intensity ratio between the two detection channels can be determined. The magnitude of the signal intensity ratio can be used to determine whether the defect type is a depression or a protrusion. The angular distribution patterns of scattered light from depression and protrusion defects are different; the defect type can be distinguished by combining the W and N channels.

[0133] The following section will explain in detail the principle of determining the defect type based on signal strength.

[0134] The defect detection method provided in this application is based on defect scattering. When a defect exists in the detection area, the detection spot 109 illuminating that area will scatter, and the scattered light will exit into the three-dimensional space above the object under test 102. By constructing a near-field physical scattering model based on the finite-difference time-domain method and solving the three-dimensional Maxwell's equations, the near-field distribution of defect scattering can be obtained. Furthermore, by performing ray tracing based on Debye integrals to achieve joint simulation of near-field optics and imaging optics, the distribution of different types of defect scattered light can be obtained.

[0135] refer to Figure 6 The diagram illustrates the signal intensity in detection channels of different types and sizes in different directions, according to an embodiment of this application. Specifically, two different sizes of protruding and concave defects were analyzed, and the signal intensity of scattered light collected by the detection channels in different directions was compared. The horizontal axis is sin(θ), where θ (the angle between the optical axis of the detection channel and the normal direction of the surface of the test object 102) represents the detection channel in different directions. The left vertical axis represents the signal intensity of scattered light corresponding to protruding defects, and the right vertical axis represents the signal intensity of scattered light corresponding to concave defects. The dashed line corresponds to a defect size of 41 nm, and the solid line corresponds to a defect size of 31 nm. The red line indicates the signal light intensity of concave defects at different angles, and the blue line indicates the signal light intensity of protruding defects at different angles. Furthermore, refer to... Figure 12 The diagram shown is a schematic of a detection channel provided in an embodiment of this application. In the diagram, θ is the angle between the optical axis of the second detection channel 105 and the normal direction of the surface of the object to be tested 102.

[0136] When the defect is fixed, setting the detection channel within the angle range of strong scattered light can effectively obtain signal intensity, thereby improving detection accuracy. Considering that different types of defects have different scattered light intensity distributions, it is necessary to analyze and summarize the distribution patterns of scattered light from various defects and rationally set the detection channels so that the detection channels have good scattered light collection effects for different types of defects, and can integrate signals collected from multiple channels to identify the type of defect. Since it is unknown whether a defect is a concave or convex defect before detection, the defect detection system provided in this application can identify concave and convex defects, thereby avoiding the omission of certain defects and ultimately improving the sensitivity and accuracy of detection.

[0137] from Figure 6It can be observed that the scattered light is non-uniformly distributed above the test object 102, and its distribution characteristics are related to the defect type. When the incident surface is perpendicular to the second plane 202, the scattered light from protruding defects is mainly distributed in the large-angle channel, while the scattered light from concave defects is mainly distributed in the small-angle channel. That is, the smaller the angle between the optical axis of the detection channel and the normal direction of the surface of the test object 102, the greater the signal intensity of the scattered light from the concave defects; the larger the angle, the greater the signal intensity of the scattered light from the protruding defects. In addition, for the same type of defect, the signal intensity of the scattered light is closely related to the defect size; the larger the size, the greater the signal intensity of the scattered light. However, for different types of defects, the signal intensity of the scattered light produced by protruding and concave defects of the same size also differs significantly (visible on the corresponding dual y-axis scale).

[0138] refer to Figure 11 The diagram shown illustrates the signal intensity of scattered light from different types of defects according to an embodiment of this application. The horizontal axis represents the signal intensity ratio I. N / I W The vertical axis represents the number of particles (particle cnt). It can be observed that the signal intensity ratio corresponding to protrusion-type defects is smaller than that corresponding to pit-type defects, indicating that the signal intensity distribution of these two types of defects is different.

[0139] Although the angular spectrum distribution of the scattered light (i.e. Figure 6 As shown, the defect type can be clearly displayed. However, in imaging collection based on objective lens 107, the scattered light transmitted at different angles will converge to the same position of the photodetector and be detected together, making it impossible to obtain its angular spectrum distribution. Considering the influence of defect type on the scattered light distribution, this application proposes to set up a detection channel, namely the N channel and the W channel, in the small angle and large angle ranges respectively, as shown. Figure 7 As shown, changes in defect size cause a corresponding change in the signal intensity collected by each detection channel. However, since the signals from both detection channels originate from the scattered light of the same defect, the signal intensity of the two channels is greater than that of I. N / I W Different patterns are observed for different defect types: Protrusion-type defects I N / I W Smaller, and concave defects I N / I W Larger. Figure 7In the diagram, the central angle of the W channel represents the angle by which the optical axis of the W channel deviates from the normal direction of the surface of the test object 102, and the central angle of the N channel represents the angle by which the optical axis of the N channel deviates from the normal direction of the surface of the test object 102. The dashed line corresponds to a defect size of 41 nm, and the solid line corresponds to a defect size of 31 nm. The red line indicates the signal light intensity of pit-type defects at different angles, and the blue line indicates the signal light intensity of protrusion-type defects at different angles.

[0140] exist Figure 7 middle, This indicates the signal strength detected by the wide channel for large-sized protrusion-like defects. This indicates the signal strength detected by the narrow channel for large-sized protrusion-like defects. This indicates the signal strength detected by the wide channel for small-sized protrusion-like defects. This indicates the signal strength detected by the narrow channel for small-sized protrusion-like defects. This indicates the signal intensity detected by the wide channel for large-sized indentation-type defects. This indicates the signal intensity detected by the narrow channel for large-sized indentation-like defects. This indicates the signal intensity detected by the wide channel for small-sized dent-like defects. This indicates the signal strength detected by the narrow channel for small-sized dent-like defects.

[0141] Taking the large-size defect represented by the dashed line as an example, the signal strength ratio I is calculated. N / I W Calculations show that for large-sized protruding defects, the signal strength ratio is... = = For large-sized dent-like defects, the signal strength is higher than... = = It can be observed that the signal intensity ratio of concave defects is greater, while the signal intensity ratio of convex defects is smaller.

[0142] Therefore, the larger the signal intensity ratio, the more likely the defect type is a depression, and the smaller the signal intensity ratio, the more likely the defect type is a protrusion.

[0143] refer to Figure 13 and Figure 14 As shown, Figure 13 This is a schematic diagram of the signal distribution of scattered light from a protrusion-like defect, provided in an embodiment of this application. Figure 14 This is a schematic diagram illustrating the signal distribution of scattered light from a depression-like defect, provided as an embodiment of this application. Figure 13 (a) shows a yellow raised defect located on the surface of the test object. Figure 13In the diagram, (b) represents the light distribution. By setting a detection channel, the light distribution in the area corresponding to that channel can be collected. Figure 14 (a) shows the test object with a depression-like defect. Figure 14 (b) in the diagram represents the corresponding light distribution.

[0144] S1022, when the signal strength ratio is greater than a preset threshold, determine the first defect information of the detection area.

[0145] The first defect information is used to identify dent-type defects in the detection area. The preset threshold is a pre-set threshold used as the basis for distinguishing between dent-type and convex-type defects. If the signal intensity ratio is greater than the preset threshold, it indicates that the signal intensity is relatively large, and the detected defect type can be determined to be a dent-type defect.

[0146] S1023, when the signal strength ratio is less than or equal to a preset threshold, determine the second defect information of the detection area.

[0147] The second defect information is used to identify convex defects in the detection area. If the signal strength ratio is not greater than the preset threshold, it means that the signal strength is relatively small, and the detected defect type is a convex defect.

[0148] In short, if the signal strength ratio K is greater than the preset threshold A, it is a dent-type defect; if the signal strength ratio K is not greater than the preset threshold A, it is a bulge-type defect. The preset threshold A can vary depending on the system layout, for example, it can vary between 1 and 5.

[0149] In this way, by determining the ratio of the signal intensity based on the first initial signal intensity acquired through the narrow channel to the second initial signal intensity acquired through the wide channel, the type of defect can be determined, thereby accurately distinguishing between protruding and concave defects and improving the accuracy of defect type detection.

[0150] In one possible implementation, S1021 determines the signal strength ratio of the first detection channel 104 and the second detection channel 105 based on the first initial signal strength obtained by the first detector 108 of the first detection channel 104 and the second initial signal strength obtained by the second detector 108 of the second detection channel 105, which may include S10211-S10213.

[0151] S10211, the noise signal intensity of the test object 102 is removed from the first initial signal intensity obtained from the first detector 108 of the first detection channel 104 to obtain the first signal intensity of the first detection channel 104.

[0152] S10212, remove the noise signal intensity from the second initial signal intensity obtained from the second detector 108 of the second detection channel 105 to obtain the second signal intensity of the second detection channel 105.

[0153] Specifically, when collecting the signal intensity of scattered light from the test object 102 using the detection channel, not only the signal intensity generated by the defect scattering light, but also the signal intensity generated by the rough area of ​​the substrate of the test object 102 scattering light, is collected. The scattered light from the rough area of ​​the substrate has virtually no effect on defect detection, so this part of the useless signal intensity can be discarded.

[0154] The noise signal intensity of the test object 102 is the signal intensity generated by the rough portion of the substrate within the detection area of ​​the test object 102. This signal intensity does not help in classifying the defect type. Therefore, the noise signal intensity can be removed from the first initial signal intensity acquired by the first detector 108 to obtain the first signal intensity, and the noise signal intensity can be removed from the second initial signal intensity acquired by the second detector 108 to obtain the second signal intensity.

[0155] In one possible implementation, the first signal strength can be represented by the following formula. :

[0156] ;

[0157] Where ∑ represents the superposition of signal intensities across all pixel regions detected by the first detector 108. This represents the signal strength detected by the first detection channel 104 within a single pixel area. This represents the noise signal intensity detected by the first detection channel 104 within a single pixel area.

[0158] In other words, the initial signal strength is ∑ The noise signal intensity of the test object 102 corresponding to the first detection channel 104 is ∑ .

[0159] In one possible implementation, the second signal strength It can be represented as:

[0160] ;

[0161] Where ∑ represents the superposition of signal intensities across all pixel regions detected by the second detector 108. This indicates the signal strength detected by the second detection channel 105 within a single pixel area. This indicates the intensity of the noise signal detected by the second detection channel 105 within a single pixel area.

[0162] In other words, the second initial signal strength is ∑ The noise signal intensity of the test object 102 corresponding to the second detection channel 105 is ∑ Furthermore, the noise at different locations of the same test object 102 is mostly similar. or The signal intensity can be obtained by taking the signal intensity collected by the detector 108 at the defect-free location of the test object 102, or by averaging the signal intensity of multiple defect-free location detectors 108.

[0163] S10213, Based on the first signal strength and the second signal strength, determine the signal strength ratio of the first detection channel 104 and the second detection channel 105.

[0164] Specifically, the signal strength ratio can be determined based on the first signal strength and the second signal strength. As an example, the signal strength ratio K can be expressed as... / .

[0165] In summary, when calculating the signal intensity ratio, the noise signal intensity generated by the rough surface of the test object 102 itself can be eliminated, so that the first signal intensity and the second signal intensity can better reflect the signal intensity caused by defect scattering, that is, the defect signal intensity, thus improving the accuracy of the signal intensity ratio.

[0166] In one possible implementation, S10213 determines the signal strength ratio of the first detection channel 104 and the second detection channel 105 based on the first signal strength and the second signal strength, which may include S102131-S102132.

[0167] S102131, Based on the system energy difference between the first detection channel 104 and the second detection channel 105, determine the compensation parameters.

[0168] S102132, based on the first signal strength, the second signal strength, and the compensation parameters, determine the signal strength ratio between the first detection channel 104 and the second detection channel 105.

[0169] Specifically, when there are multiple detection channels, there are generally system energy differences between different detection channels. In order to compensate for the error between the first signal intensity and the second signal intensity caused by the system energy difference, a compensation parameter can be determined based on the system energy difference. The compensation parameter can be represented by B, and can be determined based on the differences in objective lens 107 and detector 108 between detection channels.

[0170] The signal strength ratio K, determined based on the first signal strength, the second signal strength, and the compensation parameters, can be expressed as: K = I SN / (I SW -B).

[0171] For dark field defect detection systems, since the objective lens 107NA used in the detection channel is different, the channel loss, magnification and other parameters are also different. Therefore, the values ​​of the preset threshold A and compensation parameter B of the system are also different. The values ​​of A and B can be determined by experimental calibration.

[0172] The specific procedure is as follows: Prepare one or more sample pieces with pit-like defects of different sizes, and one or more sample pieces with protrusion-like defects of different sizes. These calibration sample pieces are measured separately using the dark field defect detection system, and two intensity distribution maps of the defect W channel and N channel are obtained respectively (similar to...). Figure 8 As shown, different sets of A and B values ​​are selected. For each set of A and B values, the defect type is determined using the method described above to obtain a set defect type value. This set of set defect type values ​​is then compared with known results (the known results are derived from the type of specially made defect sample on which the defect is located) to determine whether the set defect type value is correct. The set of A and B values ​​with the highest accuracy is taken as the judgment value of the dark field defect detection system.

[0173] In this way, by taking into account the channel differences between different detection channels when determining the signal intensity ratio, the accuracy of the signal intensity ratio can be further improved, and protrusion-type defects and depression-type defects can be accurately distinguished.

[0174] In addition to the theoretical analysis above, experimental tests have also verified the feasibility of multi-channel defect detection. (Reference) Figure 8 The image shows a scatter plot of experimental measurement results for surface defects of a test object 102 provided in an embodiment of this application. Each point represents a defect, and its horizontal axis represents the first signal intensity corresponding to channel N. The vertical axis represents the second signal strength corresponding to channel W. It can be observed that the scatter plot is densely distributed in the two central regions, the upper left and the lower right. The entire coordinate region can be divided into two regions by a straight line with a signal intensity ratio of N channel to W channel equal to k, where k is a fixed constant. Protrusion and depression defects are dispersed in these two regions respectively. The upper left region is for protrusion defects, and the lower right region is for depression defects. In addition, the correctness of these defect type classifications was verified by methods such as electron microscopy re-examination.

[0175] The process of determining the defect size will be described next. In one possible implementation, S102 determines the defect information of the detection area based on the signal intensity, which may include S201-S202.

[0176] S201, Determine the calibration curve corresponding to the detection channel.

[0177] The calibration curve is used to represent the relationship between the defect size of the test object 102 and the signal intensity. The calibration curve can be plotted in advance based on the test results of the standard piece, for reference. Figure 9 As shown, taking a protruding defect in the N-channel as an example, this is a schematic diagram of the calibration curve provided in the embodiment of this application. The horizontal axis represents the defect size (DSize), and the vertical axis represents the signal intensity (Intensity). The corresponding defect size can be found in the figure based on the signal intensity of the defect. The calibration curves are also different for different types of test objects 102, different channels, and different optical configurations, and can be calibrated separately.

[0178] S202, the signal intensity obtained by detector 108 is matched with the calibration curve to obtain the defect information of the detection area.

[0179] The defect information includes the defect size in the detection area. The output signal intensity of the detector 108 can be compared with the calibration curve to find the defect size corresponding to that signal intensity. Furthermore, different calibration curves can be used for wide channels and narrow channels, or they can use the same calibration curve.

[0180] In this way, by comparing the signal strength with the calibration curve, the size information of the defect can be determined based on the signal strength, thereby further improving the accuracy of defect detection and making the defect detection results more comprehensive.

[0181] In practical applications, one approach is to set corresponding calibration curves to determine the defect size for different types of defects (protrusion defects and depression defects). Another approach is to set a uniform calibration curve for both protrusion and depression defects to determine the defect size when it is inconvenient to prepare depression defects of different standard sizes.

[0182] For the first method, the signal intensity of protrusion-type defects of different sizes can be measured in advance to obtain the corresponding calibration curves for protrusion-type defects. Similarly, the signal intensity of depression-type defects at different sizes can also be measured to obtain calibration curves for depression-type defects. Therefore, after detecting the defect type, the corresponding calibration curve can be found based on the defect type, and the defect size can be determined based on the calibration curve, thereby improving the accuracy of defect size determination.

[0183] In addition, the signal intensity corresponding to convex defects is usually detected using the second detection channel (i.e., the W channel), while the signal intensity corresponding to concave defects is usually detected using the first detection channel (i.e., the N channel) to improve the accuracy of the calibration curve.

[0184] The second method allows for measurement of defects of different sizes without distinguishing between defect types, thus obtaining the correspondence between defect size and signal intensity, which is the calibration curve. In other words, regardless of whether the detected defect is a protrusion or a depression, the same calibration curve is used for size determination, thereby improving the efficiency of size determination.

[0185] In one possible implementation, the method also includes S301-S303.

[0186] S301, determine the detection time corresponding to the detection image generated by the detector 108 based on the scattered light.

[0187] Specifically, when the detector 108 collects the scattered light, the defect detection system can record the corresponding detection time, that is, at the detection time, there is a corresponding detection image.

[0188] S302, based on the detection time and the motion trajectory of the detection spot 109 on the object to be tested 102 during the detection time period, determine the location of the area to be tested in the object to be tested 102.

[0189] Specifically, since the stage 101 moves the object under test 102, the detection spot 109 is incident on different areas of the object under test 102. The defect detection system also records the motion trajectory of the detection spot 109 on the surface of the object under test 102. The motion trajectory can be understood as the position record of the detection spot 109 on the surface of the object under test 102 at various times. Therefore, based on the motion trajectory and the detection time, the location of the detection spot 109 on the surface of the object under test 102 at that detection time can be determined, i.e., the area location.

[0190] S303, based on the region location and the location of the defect in the detection image, determine the defect location in the test object 102.

[0191] The location of the area can indicate the approximate location of the defect on the surface of the object under test 102. Combined with the location of the defect in the detection image, the exact location of the defect on the surface of the object under test 102 can be accurately determined, thereby achieving precise detection of the defect location.

[0192] In layman's terms, the test object 102 rotates at a constant speed. When a particle at a certain position on the surface of the test object 102 moves to the area illuminated by the detection spot 109, the particle is illuminated and emits scattered light signals in all directions, which are captured by the detection channel and finally imaged by the camera (i.e., detector 108). The position of the particle on the test object 102 can be calculated based on the detection image stored by the camera and the motion trajectory of the stage 101.

[0193] The following is an example illustration of the optical path diagram for lighting channel 103. (Reference) Figure 10 The diagram shows a schematic of an illumination channel 103 provided in an embodiment of this application. A light source 106 emits signal light towards the object under test 102. The beam is first split by a beam splitter 112, and then part of the light enters the first incident component. The polarization conversion module 110 converts the laser light into different polarization forms such as linearly polarized light and circularly polarized light. Linearly polarized light can be divided into p-polarized light, s-polarized light, and linearly polarized light at a specific angle. Circularly polarized light can be divided into left-handed and right-handed circularly polarized light. The polarization conversion module 110 can be an optical system consisting of waveplates or combinations of waveplates, or it can be an optical element formed from a metasurface. After the polarization-modulated beam passes through a shaping module 111 and a first reflecting mirror 116 to change its propagation path, it forms a vertically incident beam L1. The beam is first split by a beam splitter 112, and then another part of the light enters the second incident component. The second incident component includes a second reflecting mirror 114, a polarization conversion module 110, a shaping module 111, and a third reflecting mirror 117, forming an obliquely incident beam L2.

[0194] The defect detection method provided in this application can be applied to a defect detection system. Specifically, a computer subsystem can be used to control the defect detection system to achieve adjustments such as optical path adjustment and platform movement. The computer subsystem can include any suitable computer system known in the art. For example, a computer subsystem can take various forms, including personal computer systems, mainframe computer systems, workstations, image processing computers, parallel processors, or any other device known in the art. Generally, the term "computer subsystem" can be broadly defined as any device containing one or more processors that execute instructions from a storage medium.

[0195] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, the method embodiments are basically similar to the apparatus embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the apparatus embodiments.

[0196] The above description is merely a preferred embodiment of this application. Although this application has disclosed preferred embodiments above, it is not intended to limit this application. Any person skilled in the art can make many possible variations and modifications to the technical solutions of this application using the methods and techniques disclosed above, or modify them into equivalent embodiments with equivalent changes, without departing from the scope of the technical solutions of this application. Therefore, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of this application without departing from the content of the technical solutions of this application shall still fall within the protection scope of the technical solutions of this application.

Claims

1. A defect detection system, characterized in that, The defect detection system includes: A stage; the stage is used to hold the object to be measured. An illumination channel and multiple detection channels are located on the same side of the stage; the illumination channel includes a light source for providing a detection spot to the detection area of ​​the object under test; the detection channel includes an objective lens and a detector, the objective lens for collecting scattered light generated by the detection spot in the detection area, and the detector for determining the signal intensity of the scattered light so as to determine the defect information of the detection area based on the signal intensity; The incident surface is perpendicular to the second plane, which is the plane containing the optical axis of the incident light emitted by the light source and the normal direction of the surface of the object to be tested, and the second plane is the plane containing the optical axis of the plurality of detection channels; The optical axis of the detection channel has an offset angle from the normal direction of the surface of the object to be tested, and at least two of the offset angles satisfy the angle difference condition.

2. The defect detection system according to claim 1, characterized in that, The detection spot is a linear spot, and the first plane is perpendicular to the second plane. The first plane is the plane containing the optical axis of the illumination channel and the long side of the detection spot.

3. The defect detection system according to claim 2, characterized in that, The detector has a photosensitive surface, the long side of which is perpendicular to the second plane.

4. The defect detection system according to claim 2, characterized in that, The short side width of the detection spot is greater than or equal to 10 micrometers and less than or equal to 30 micrometers.

5. The defect detection system according to claim 1, characterized in that, The plurality of detection channels include a first detection channel and a second detection channel. The optical axis of the first detection channel has a first deviation angle from the normal direction, and the optical axis of the second detection channel has a second deviation angle from the normal direction. The first deviation angle is less than 45°, and the second deviation angle is greater than or equal to 45°.

6. The defect detection system according to claim 1, characterized in that, The angle between the optical axis of the illumination channel and the normal direction of the surface of the object under test is less than or equal to 90°.

7. The defect detection system according to claim 1, characterized in that, When there are multiple lighting channels, the detection light spots of all the lighting channels overlap.

8. The defect detection system according to claim 1, characterized in that, When the lighting channel comprises multiple light sources, all of the light sources emit light beams with the same wavelength, or at least two of the light sources emit light beams with different wavelengths.

9. The defect detection system according to claim 8, characterized in that, When the illumination channel includes multiple light sources, and at least two of the light sources emit light beams with different wavelengths, the detection channel includes a filter module for filtering the scattered light entering the detector.

10. The defect detection system according to any one of claims 1-9, characterized in that, The lighting channel includes a polarization adjustment module, which is used to adjust the polarization state of the light beam emitted by the light source.

11. The defect detection system according to any one of claims 1-9, characterized in that, The lighting channel includes a shaping module, which is used to shape the light beam emitted by the light source to obtain the detection spot.

12. A defect detection method, applied to the defect detection system according to any one of claims 1-11, the defect detection method comprising: During the rotation of the stage, defects are detected in different areas of the object to be tested located on the stage through the illumination channel and the multiple detection channels. When performing defect detection on the detection area of ​​the object under test, the light source provides a detection spot to the detection area through the illumination channel. The scattered light generated by the detection spot in the detection area passes through the objective lens and is received by the detector. The detector determines the signal intensity of the scattered light and determines the defect information of the detection area based on the signal intensity.

13. The defect detection method according to claim 12, characterized in that, The plurality of detection channels include a first detection channel and a second detection channel. The optical axis of the first detection channel has a first deviation angle from the normal direction, and the optical axis of the second detection channel has a second deviation angle from the normal direction. The first deviation angle is smaller than the second deviation angle. Determining the defect information of the detection area based on the signal intensity includes: Based on the first initial signal strength obtained by the first detector of the first detection channel and the second initial signal strength obtained by the second detector of the second detection channel, the signal strength ratio of the first detection channel and the second detection channel is determined. When the signal strength ratio is greater than a preset threshold, first defect information of the detection area is determined; the first defect information is used to identify a dent-type defect in the detection area. When the signal strength ratio is less than or equal to the preset threshold, second defect information of the detection area is determined; the second defect information is used to identify a protrusion-type defect in the detection area.

14. The defect detection method according to claim 13, characterized in that, Determining the signal strength ratio between the first detection channel and the second detection channel based on the first initial signal strength obtained by the first detector of the first detection channel and the second initial signal strength obtained by the second detector of the second detection channel includes: The noise signal intensity of the test object is removed from the first initial signal intensity obtained from the first detector of the first detection channel to obtain the first signal intensity of the first detection channel; The noise signal intensity is removed from the second initial signal intensity obtained from the second detector of the second detection channel to obtain the second signal intensity of the second detection channel; Based on the first signal strength and the second signal strength, the signal strength ratio of the first detection channel and the second detection channel is determined.

15. The defect detection method according to claim 14, characterized in that, Determining the signal strength ratio between the first detection channel and the second detection channel based on the first signal strength and the second signal strength includes: Based on the system energy difference between the first detection channel and the second detection channel, the compensation parameters are determined; Based on the first signal strength, the second signal strength, and the compensation parameter, the signal strength ratio between the first detection channel and the second detection channel is determined.

16. The defect detection method according to claim 15, characterized in that, The first signal strength for: ； Where, ∑ The initial signal strength is... The noise signal intensity of the test object in the first detection channel; Second signal strength for: ； in, The second initial signal strength, The noise signal intensity of the test object in the second detection channel; The signal strength ratio K is: K =I SN / (I SW -B); Wherein, B is the compensation parameter.

17. The defect detection method according to claim 12, characterized in that, Determining the defect information of the detection area based on the signal intensity includes: Determine the calibration curve corresponding to the detection channel; the calibration curve is used to represent the relationship between the defect size of the test object and the signal intensity; The signal intensity obtained by the detector is matched with the calibration curve to obtain the defect information of the detection area; the defect information includes the defect size of the defect in the detection area.

18. The defect detection method according to claim 12, characterized in that, The method further includes: Determine the detection time corresponding to the detection image generated by the detector based on the scattered light; Based on the detection time and the movement trajectory of the detection spot on the object under test during the detection time period, the location of the area to be tested in the object under test is determined. Based on the location of the region and the location of the defect in the detection image, the defect location in the object under test is determined.

Images