Dual scan beam split defect scanner and optical inspector with independent angles of incidence

Abstract

A method includes: enabling a first radiation source configured to output a first light beam; enabling a second radiation source configured to output a second light beam; directing the first light beam to illuminate a sample at an angle of incidence that differs from Brewster's angle by ten degrees or less; directing the second light beam to illuminate the sample at an angle of incidence that is ten degrees or less (the first and second light beams pass through only a single scan lens prior to illuminating the sample); measuring scattered radiation data resulting from illuminating the sample with the second light beam; measuring reflected radiation resulting from illuminating the sample with the first light beam; and determining the presence of a defect based at least in part on one of the measurements. The radiation sources can be enabled in an alternating fashion to improve resulting measurement performance.

Description

Cross-references to related applications

[0001] This application is a PCT international application that claims the benefit of U.S. Patent Application Serial No. 18 / 520,279, filed November 27, 2023, entitled “DUAL SCAN BEAM SEPARATION WITH INDEPENDENT ANGLE OF INCIDENCE DEFECT SCANNER AND OPTICAL INSPECTOR,” which is a partial continuation of Non-Provisional U.S. Patent Application Serial No. 18 / 092,385, filed January 2, 2023, entitled “TIME DOMAIN MULTIPLEXED DEFECT SCANNER,” and claims priority to that Non-Provisional U.S. Patent Application pursuant to 35 USC §120, the subject matter of which is incorporated herein by reference. Non-provisional U.S. Patent Application Serial No. 18 / 092,385 is a partial continuation of Non-provisional U.S. Patent Application Serial No. 17 / 685,719, filed March 3, 2022, entitled “SLOPE, P-COMPONENT AND S-COMPONENT MEASUREMENT,” and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 685,719 is a partial continuation of Non-provisional U.S. Patent Application Serial No. 17 / 576,986, filed January 16, 2022, entitled “REGION PROBER OPTICAL INSPECTOR,” and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 576,986, filed 35 U.SC §120, and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 576,986, filed January 16, 2022, entitled “REGION PROBER OPTICAL INSPECTOR,” and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 576,986, filed 35 U.SC §120, and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 576,986, and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 576,986, filed 35 U.SC §120, and claims priority to that non-provisional U.S. Patent Application Serial No. 17 / 576,986. Non-provisional U.S. Patent Application Serial No. 17 / 576,986 is a partial continuation of Non-provisional U.S. Patent Application Serial No. 16 / 838,026, filed April 2, 2020, entitled “REGION PROBER OPTICAL INSPECTOR,” and claims priority to that Non-provisional U.S. Patent Application pursuant to 35 USC §120, the subject matter of which is incorporated herein by reference.Non-provisional U.S. Patent Application Serial No. 16 / 838,026 is a partial continuation of Non-provisional U.S. Patent Application Serial No. 16 / 289,632, filed on February 28, 2019, entitled “PHASE RETARDANCE OPTICALSCANNER,” and claims priority to that Non-provisional U.S. Patent Application pursuant to 35 USC §120, the subject matter of which is incorporated herein by reference. Technical Field

[0002] This invention relates generally to systems and methods for detecting properties in materials. More specifically, this invention relates to detecting properties in materials by measuring light reflected from the material. Background Technology

[0003] Many thin films are used in high-tech products. For example, thin films are used on glass in many high-tech products such as televisions, monitors, and mobile devices. Inspecting glass is challenging due to its low reflectivity and high transparency. Previous techniques required rotating the glass sample for inspection. Rotating the sample can cause problems with fragile, asymmetrical, or large glass samples. Regardless of these issues, fragile, asymmetrical, or large glass samples still need to be tested for defects before being used in expensive manufacturing processes and integrated into costly high-tech products. Summary of the Invention

[0004] In a first novel aspect, a method includes: (a) activating a first radiation source configured to output a first beam; (b) activating a second radiation source configured to output a second beam; (c) directing the first beam to irradiate a sample at an incident angle differing from Brewster's angle by ten degrees or less, wherein the first beam passes through a scanning lens before irradiating the sample; (d) directing the second beam to irradiate the sample at an incident angle of ten degrees or less, wherein the second beam passes through the scanning lens before irradiating the sample, and wherein neither the first beam nor the second beam passes through the second scanning lens before irradiating the sample; (e) measuring scattered radiation data obtained by irradiating the sample with the second beam; (f) measuring reflected radiation obtained by irradiating the sample with the first beam; and (g) determining the presence of a defect based at least in part on the measurement of (e) or (f).

[0005] In a second novel aspect, the first beam has a first wavelength, the second beam has a second wavelength, and the first wavelength is different from the second wavelength.

[0006] In a third novel aspect, the first radiation source is deactivated before the measurement in (e) is performed, and the second radiation source is deactivated before the measurement in (f) is performed.

[0007] In a fourth novel aspect, a method includes: (a) activating a first radiation source configured to output a first beam; (b) activating a second radiation source configured to output a second beam; (c) directing the first beam to irradiate a sample at an incident angle differing from Brewster's angle by ten degrees or less, wherein the first beam passes through a scanning lens before irradiating the sample; (d) directing the second beam to irradiate the sample at an incident angle of ten degrees or less, wherein the second beam passes through the scanning lens before irradiating the sample, and wherein neither the first beam nor the second beam passes through the second scanning lens before irradiating the sample; (e) measuring scattered radiation data obtained by irradiating the sample with the second beam; (f) measuring scattered radiation obtained by irradiating the sample with the first beam; and (g) determining the presence of a defect based at least in part on the measurement of (e) or (f).

[0008] In a fifth novel aspect, an apparatus includes: a first radiation source configured to output a first light beam; a second radiation source configured to output a second light beam; a dichroic mirror configured to guide the first light beam and the second light beam toward a time-varying beam reflector, wherein the time-varying beam reflector is configured to reflect the first light beam and the second light beam toward a scanning lens; a first reflector configured to reflect the first light beam and the second light beam; a second reflector configured to reflect the second light beam, wherein the second reflector is configured not to reflect the first light beam; a third reflector configured to reflect the first light beam; and a fourth reflector configured to reflect the second light beam.

[0009] In a sixth novel aspect, the third reflector is configured to reflect the first beam toward the sample at an incident angle less than or equal to ten degrees different from Brewster's angle.

[0010] In a seventh novel aspect, the fourth reflector is configured to reflect the second beam toward the sample at an incident angle of less than or equal to ten degrees.

[0011] Further details, embodiments, and techniques are described in the following detailed description. This summary is not intended to limit the invention. The invention is defined by the claims. Attached Figure Description

[0012] The accompanying drawings illustrate embodiments of the invention, wherein the same numbers indicate the same components.

[0013] Figure 1 This is a cross-sectional view of the thin film deposited on top of the glass sample.

[0014] Figure 2This is a graph illustrating the relationship between the phase retardation of a bare glass sample, a 10 Å thin film deposited on a glass sample, and a 100 Å thin film deposited on a glass sample and the incident angle of the irradiated light.

[0015] Figure 3 This is a diagram of a phase delay optical detector.

[0016] Figure 4 This is a diagram of a dual-unit detector.

[0017] Figure 5 This is a graph illustrating the phase delay caused by the input and output mirrors when both are operating out of phase. More specifically, the graph illustrates the relationship between the phase of the beams reflected from the input and output mirrors and the rotation angle of the input mirror.

[0018] Figure 6 This is a graph illustrating the combined phase delay of in-phase and out-of-phase mirror operations. More specifically, the graph illustrates the relationship between the phase of the beam after reflection by both the input and output mirrors and the rotation angle of the input mirror.

[0019] Figure 7 This is a graph illustrating the relationship between the total phase delay of a phase-delay optical inspector and the field of view of the phase-delay optical inspector at the position of the phase-delay detector.

[0020] Figure 8 This is a table listing examples of the rotation angles of the input and output mirrors when they are operating in out-of-phase mode.

[0021] Figure 9 This is a diagram illustrating a beam delay mapping method for detecting defects by detecting changes in beam delay.

[0022] Figure 10 This is a diagram of a scattered radiation optical detector.

[0023] Figure 11 This is a diagram illustrating the scattered radiation mapping of defects by detecting changes in the intensity of scattered radiation.

[0024] Figure 12 This is a flowchart illustrating the steps used to perform phase delay defect detection.

[0025] Figure 13 This is a flowchart illustrating the steps used to perform scattered radiation defect detection.

[0026] Figure 14 A novel optical inspector for area detectors is illustrated.

[0027] Figure 15 An example location to be probed on a transparent sample is shown.

[0028] Figure 16 An example is shown of the area of ​​the reflected light as viewed from the path of the reflected light blocked by the blocker 119.

[0029] Figure 17 An example of the desired region to be probed in a transparent sample is shown.

[0030] Figure 18 This is a table illustrating how to use thresholds to detect the presence of defects in a probe area.

[0031] Figure 19 This is a flowchart 300 illustrating a region detection step for detecting defects within a desired area of ​​a transparent sample.

[0032] Figure 20 Two scattered radiation events caused by inclusions are illustrated.

[0033] Figure 21 Two scattered radiation events caused by particles on the top surface are illustrated.

[0034] Figure 22 An example of a scattered radiation event caused by particles on the bottom surface is illustrated.

[0035] Figure 23 This is an example of the range of intervals between scattering radiation events of an exemplary transparent sample.

[0036] Figure 24 This is a flowchart 400 illustrating the steps for performing scattered radiation defect depth detection.

[0037] Figure 25 This is a flowchart 500 illustrating a defect detection process using multiple measured polarization values.

[0038] Figure 26 This is a diagram of another phase delay optical detector.

[0039] Figure 27 This is a diagram of another optical detector for scattered radiation.

[0040] Figure 28 This is another flowchart illustrating the steps used to perform phase delay defect detection.

[0041] Figure 29 This is another flowchart illustrating the steps used to perform scattered radiation defect detection.

[0042] Figure 30 This is a diagram of an optical tester for surface height that is independent of angle.

[0043] Figure 31This is a simplified example of an angle-independent optical surface height inspector at different sample heights.

[0044] Figure 32 This is a simplified example of an angle-independent optical surface height inspector at different sample angles.

[0045] Figure 33 This is a diagram of the electrical system used in a beam-tracking positioning sensitive detector.

[0046] Figure 34 This is a flowchart of an optical inspection device for surface height that is independent of angle.

[0047] Figure 35 This is a diagram of a time-domain multiplexing defect scanner.

[0048] Figure 36 This is a diagram of a scattered radiation optical detector.

[0049] Figure 37 This is a timing diagram illustrating the control of a radiation source.

[0050] Figure 38 This is a diagram illustrating an exemplary wafer scanning operation performed using time-division multiplexing of a radiation source.

[0051] Figure 39 This is a diagram illustrating an exemplary single wafer scan performed using time-division multiplexing of a radiation source.

[0052] Figure 40 This is an illustration of an exemplary single wafer scan used for a first radiation source.

[0053] Figure 41 This is an illustration of an exemplary single wafer scan used for a second radiation source.

[0054] Figure 42 This is a flowchart of a dual-source scanning process.

[0055] Figure 43 This is a flowchart of dual-wavelength scanning using time-division multiplexing.

[0056] Figure 44 It is a flowchart of a dual-wavelength scan using two radiation sources and time-division multiplexing.

[0057] Figure 45 It is a flowchart of a dual-wavelength scan using a single radiation source and time-division multiplexing.

[0058] Figure 46 This is a diagram of a dual-scanning beam separation defect scanner with independent incident angles.

[0059] Figure 47This is a diagram of a dual-scan beam separation scattered radiation optical detector with independent incident angles.

[0060] Figure 48 This is a diagram illustrating the capture of photoluminescence by a Huygens wavelet with an incident angle of 90 degrees.

[0061] Figure 49 This is a diagram illustrating photoluminescence captured by a Huygens wavelet with an incident angle of zero degrees.

[0062] Figure 50 This is a flowchart of a dual-scanning beam separation defect scanner and an optical inspector with independent incident angles. Detailed Implementation

[0063] Reference will now be made in detail to the background examples and some embodiments of the invention, which are illustrated in the accompanying drawings. In the following description and claims, relational terms such as “top,” “down,” “upper,” “lower,” “top,” “bottom,” “left,” and “right” are used to describe the relative orientation between different parts of the described structure, and it should be understood that the overall structure described can indeed be oriented in any way in three-dimensional space.

[0064] Many high-tech products involve depositing films on glass or other transparent substrates. An important process control metric is measuring film thickness and defects on the glass substrate. This has proven difficult due to the low reflectivity of glass and the challenge of separating signals from the bottom surface of the glass substrate. Another problem with measuring film thickness and defects on glass substrates is that current technology does not allow for scanning transparent samples of many different shapes and sizes.

[0065] The following solutions are required: (i) accurately separating signals from the top surface of a glass substrate from signals from the bottom surface of the glass substrate; (ii) detecting the presence of defects in response to small changes in signals from the surface of the glass substrate; and (iii) allowing scanning of transparent samples of many different shapes and sizes.

[0066] This invention provides a solution to this problem by providing a scanning method that illuminates the transparent sample at or near the Brewster angle. This scanning method, which illuminates the transparent sample at or near the Brewster angle, also provides scanning in the xy coordinate system, enabling the invention to scan any substantially flat object shape or size.

[0067] Transparent sample surfaces (such as glass) typically have a thin film deposited on their surface. Figure 1This is a cross-sectional view of the thin film deposited on top of a glass sample. The goal is to be able to inspect the surface cleanliness of the transparent sample before deposition and to inspect for film defects after deposition. To achieve this, the technique needs to be highly sensitive to the film on the surface of the transparent sample. The technique also needs to be able to separate the received signal from the top surface of the transparent sample from the received signal from the bottom surface of the transparent sample.

[0068] By considering Figure 2 The information illustrated herein can help address sensitivity to films on transparent samples. Figure 2 This is a graph illustrating the relationship between the phase retardation and the incident angle of the irradiated light for a bare glass sample, a 10 Å thin film deposited on a glass sample, and a 100 Å thin film deposited on a glass sample. More specifically, Figure 2 The retardation (Φ) of typical transparent samples (such as glass (BK7) with MgF2 film thickness as a parameter) is shown. P -Φ S The relationship between (+π) and the incident angle. On a glass substrate, the retardation decreases only to the phase of the P-wave (Φ). P ).according to Figure 2 One conclusion drawn is that, in order to detect films on transparent samples (such as glass), it is desirable to operate with a polarization close to P-polarization (polarization parallel to the incident plane). According to Figure 2 Another conclusion is that it is desirable to perform scanning by irradiating the transparent sample at an incident angle equal to or close to the Brewster angle of the transparent sample. For example, by operating at 57 degrees instead of 60 degrees, a tenfold (10x) increase in sensitivity can be achieved. Sensitivity is defined as the difference between the 10 Å film curve and the bare glass curve. However, a trade-off exists because operating at the exact Brewster angle in P-polarization would result in no signal reflection from the transparent solid. Therefore, it is desirable to perform scanning by irradiating the sample at an incident angle no greater than or no less than one degree than the Brewster angle of the transparent sample. It is also desirable to perform scanning by irradiating the sample with a polarization no more than 20 degrees different from that of P-polarization.

[0069] Figure 3 This is a diagram of a phase delay optical inspector. The phase delay optical inspector includes a radiation source 10, a half-wave plate 28, a feed mirror 11, an input mirror (input time-varying beam reflector) 12, a scanning lens 13, a de-scanning lens 15, an output mirror (output time-varying beam reflector) 16, a feed mirror 17, a focusing lens 18, a stopper 19 located at the focal plane 20 of the focusing lens 18, a collimating lens 21, a half-wave plate 22, a phase delay measurement unit 29, a processor 26 (optional), and a memory 27 (optional). The phase delay measurement unit includes a polarization beam splitter 23, a first detector 24, and a second detector 25.

[0070] In one example, the input mirror 12 and the output mirror 16 are linear time-varying beam reflectors whose reflection angles change linearly as they rotate. The input mirror 12 and the output mirror 16 can also be controlled by electrical signals, such as a signal generator. The input mirror 12 and the output mirror 16 may be referred to as galvanometer mirrors.

[0071] In one example, the first detector and / or the second detector are two-cell detectors. Figure 4 An example of a dual-cell detector is illustrated. The dual-cell detector has a center line 31 separating a first light sensor and a second light sensor. The dual-cell detector can be configured to output a first signal indicating the difference between the light intensity measured on one side of the detector and the light intensity measured on the other side of the detector. The dual-cell detector can also be configured to output a second signal indicating the sum of the light intensity measured on one side of the detector and the light intensity measured on the other side of the detector.

[0072] In operation, the phase delay optical inspector measures the delay by measuring the polarization change of the signal from the transparent sample, caused by the scanning beam illuminating the transparent sample as it travels across it. To accurately measure the polarization change caused by the signal from the transparent solid, rather than by the phase delay inherent in the inspector itself, the optics that generate the moving beam and the optics that descan and guide the signal from the transparent solid need to produce minimal polarization change (delay).

[0073] The main sources of polarization change (delay) caused by the phase delay optical inspector are the input mirror 12 and the output mirror 16. Figure 5 The diagram illustrates the polarization change (delay) caused by the input mirror 12 and the output mirror 16. The input and output mirrors 12 and 16 can be operated in any desired manner based on their electrical control signals; however, the manner in which they are operated has a significant impact on the polarization change (delay) introduced by the phase-delay optical inspector. The polarization change (delay) produced by each mirror is a function of the angle of incidence of light on that mirror. The larger the angle of incidence, the greater the polarization change (delay).

[0074] In one example, input mirror 12 and output mirror 16 can be operated in phase such that each mirror is rotated to have the same angle relative to the beam. This "in-phase" operation of input mirror 12 and output mirror 16 causes the maximum polarization change (delay) because the amount of polarization change caused by the input mirror increases as input mirror 12 rotates to increase its angle relative to the beam, and the amount of polarization change caused by the output mirror increases as output mirror 16 rotates to increase its angle relative to the beam. Therefore, the polarization change (delay) caused by each mirror is always in the same direction and results in the maximum polarization change. Figure 6 The example illustrates the maximum change in phase change (delay) during "in-phase" operation. Figure 6 It is clearly shown that as the rotation angle of the mirror increases, the polarization change (delay) also increases.

[0075] In another example, the input mirror 12 and the output mirror 16 can be operated such that these mirrors are out of phase, such that each mirror is rotated such that each mirror has an opposite angle relative to the beam. Figure 8 This table lists examples of the rotation angles of the input and output mirrors when operating in out-of-phase mode. This out-of-phase operation of input mirror 12 and output mirror 16 results in a minimum polarization change (delay) because the amount of polarization change caused by the input mirror decreases as input mirror 12 rotates to increase its angle relative to the beam, and the amount of polarization change caused by output mirror 16 decreases as output mirror 16 rotates to increase its angle relative to the beam. Therefore, the polarization change (delay) caused by each mirror is always in opposite directions and results in a minimum. Figure 6 The example also illustrates the minimum change in phase (delay) during "out-of-phase" operation. Figure 6 It is clearly shown that as the mirror rotation angle increases, the polarization change (delay) does not exceed two degrees.

[0076] Regarding out-of-phase operation, it should also be noted that this technique not only reduces the polarization change (delay) across the field of view, but also reduces the reflectivity variation caused by each mirror. This is because the reflectivity of the input mirror decreases as its angle of incidence increases, while the reflectivity of the output mirror increases as its angle of incidence decreases. These two effects almost cancel each other out, resulting in a very small change in reflectivity relative to the angle of incidence.

[0077] Another major source of polarization change (delay) caused by the phase-retardation optical inspector is the feed angle of the scanning beam on the input mirror 12. As discussed above, the polarization change (delay) decreases as the incident angle approaches zero degrees. However, from a practical point of view, a feed angle of approximately five degrees is possible. In the phase-retardation optical inspector, the feed angle is controlled by the configuration of the radiation source 10 and the feed mirror 11. In one example, the radiation source 10 and the feed mirror 11 are configured such that the resulting feed angle to the input mirror 12 is twelve degrees. This fixed feed angle of twelve degrees minimizes the polarization change (delay) caused by the input mirror 12.

[0078] In one example, when the input mirror 12 is positioned at the midpoint of its rotation range, the light reflected by the feed mirror illuminates the input mirror 12 at an angle not exceeding thirty degrees from the normal angle of the input mirror 12 (the first time-varying beam reflector).

[0079] Similarly, another major source of polarization change (delay) caused by the phase-retardation optical inspector is the feed angle of the signal from the transparent sample at the output mirror 16. As discussed above, the polarization change (delay) decreases as the incident angle approaches zero degrees. However, from a practical point of view, feed angles of approximately five to fifteen degrees are possible. In the phase-retardation optical inspector, the feed angle is controlled by the configuration of the descanning lens 15, the output mirror 16, and the feed mirror 17. In one example, the descanning lens 15, the output mirror 16, and the feed mirror 17 are configured such that the resulting feed angle to the output mirror 17 is twelve degrees. This fixed feed angle of twelve degrees minimizes the polarization change (delay) caused by the feed mirror 17.

[0080] The combination of out-of-phase operation of the input mirror 12 and the output mirror 16, which have minimum feed angles and feed angles, enables the optical inspector to produce minimal polarization change (delay). Figure 7 An example illustrating the relationship between the polarization change (delay) obtained at the phase delay measurement unit 29 and the field of view of the optical inspector is shown. This clearly demonstrates the advantages of out-of-phase mirror operation over in-phase mirror operation.

[0081] The half-wave plate 28 can be used to adjust the polarization of the scanning beam output from the radiation source 10. In one example, the half-wave plate 28 is used to adjust the polarization of the scanning beam to be as close as possible to P-polarization. As discussed above, it is advantageous to use a near-P-polarized scanning beam to scan transparent samples.

[0082] The scanning lens 13 operates to focus the scanning beam onto the transparent sample. In one example, the scanning lens 13 is a telecentric scanning lens. The scanning lens 13 can be configured such that the scanning beam output from the scanning lens 13 illuminates the transparent sample at an angle not exceeding one degree from the Brewster angle of the transparent sample. Another type of lens that can replace the scanning lens is an achromatic lens.

[0083] In one example, the transparent sample is glass. In another example, the transparent sample is a thin film deposited on a transparent material. Other examples of transparent samples include, but are not limited to, sapphire, fused silica, quartz, silicon carbide, and polycarbonate.

[0084] Descanning lens 15 operates to focus the signal from the transparent sample onto output mirror 16. In one example, descanning lens 15 is an achromatic lens. Achromatic lenses can be used for descanning because critical focusing and telecentrism are not required when receiving the signal from the transparent sample. This is an economic benefit, as achromatic lenses are much cheaper than telecentric lenses. Other examples of descanning lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0085] Feed mirror 17 operates to reflect the signal from output mirror 16 to focusing lens 18. Focusing lens 18 has a focal plane 20. At the focal point of focusing lens 18, there will be two spots (assuming the sample is transparent), and these spots correspond to signals from the top and bottom surfaces of the sample. In one example, focusing lens 18 is an achromatic lens. Other examples of focusing lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0086] In one example, when the output mirror 16 is positioned at the midpoint of its rotation range, the light reflected by the output mirror 16 illuminates the feed mirror at an angle not exceeding thirty degrees from the normal angle of the output mirror 16 (the second time-varying beam reflector).

[0087] The blocker 19 is located near the focal plane 20 and operates to block a portion of the signal from a specific surface of the transparent sample. Figure 3 As illustrated, blocker 19 can be configured to block signals from the bottom surface of a transparent sample while allowing signals from the top surface of the transparent sample to pass through. Although not illustrated, blocker 19 can also be configured to block signals from the top surface of a transparent sample while allowing signals from the bottom surface of the transparent sample to pass through. In this way, the phase-retardation optical inspector is able to distinguish signals from the top of the transparent sample from signals from the bottom of the transparent sample. In one example, blocker 19 is a mirror. Other examples of blockers include, but are not limited to, absorbing materials, blackened aluminum sheets, and black-painted metal sheets.

[0088] Collimating lens 21 operates to collimate the signal from the transparent sample that is not blocked by blocker 19. In one example, collimating lens 21 is an achromatic lens. Other examples of collimating lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0089] The half-wave plate 22 operates to adjust the polarization of the signal before it illuminates the polarization beam splitter 23 of the phase delay measurement unit 29 from the transparent sample. In one example, the half-wave plate 22 adjusts the polarization of the signal from the transparent sample such that the signals incident on detectors 24 and 25 are approximately equal.

[0090] When illuminated, polarizing beamsplitter 23 allows all light polarized in one direction to pass through to detector 25 and reflects all light polarized in the other direction to detector 24. The plane of polarizing beamsplitter 23 is the same as the plane of the sample. Detector 24 outputs a signal indicating the intensity of the light illuminating detector 24. Detector 25 outputs a signal indicating the intensity of the light illuminating detector 25. The difference between the signals from the two detectors is proportional to the polarization change (delay) of the scanning beam caused by defects in the transparent sample or the film on the transparent sample. Any small changes in film thickness or properties can be detected by comparing the output signals of detectors 24 and 25. The sum of the signals from the two detectors is proportional to the reflectivity of the transparent sample or the film on the transparent sample.

[0091] When the detector is a dual-cell detector, the phase delay measurement unit can also determine the change in the surface slope of the transparent sample.

[0092] Processor 26 (optional) can be used to read the output signals from detectors 24 and 25. Processor 26 can execute code to calculate the difference between the output signals and determine whether a defect exists on the transparent sample and what type of defect it is. Processor 26 can also store the intensity value indicated by the output signal in memory 27 (optional). Processor 26 can also read instructions from memory 27. Processor 26 can also read one or more thresholds to help determine whether a defect exists and, if present, the type of defect.

[0093] Figure 9 This is an illustration of a beam delay map that detects defects by detecting changes in beam delay. This map can be created manually based on monitoring the output of the phase delay measurement unit 29. Alternatively, the map can be created automatically by a processor that samples the output of the phase delay measurement unit 29 and stores the resulting difference in memory 27. The beam delay map can then be used to determine the presence of defects on a transparent sample or a thin film deposited on a transparent sample. For example, a decrease in beam delay can indicate a change in film thickness, film defects, or particle defects on the transparent sample. Alternatively, an increase in beam delay can indicate a change in film thickness, film defects, or particle defects on the transparent sample. This beam delay map can also be output as a digital file that can be shared with the consumer of the transparent sample.

[0094] Figure 10 This is a diagram of a scattered radiation optical detector. The scattered radiation optical detector includes a radiation source (not shown) for output source beam 40, a time-varying beam reflector 41, a telecentric scanning lens 42, a focusing lens 46, a spatial filter 48, a collimating lens 49, and a detector 50.

[0095] In operation, a radiation source emits a source beam 40, which illuminates a time-varying beam reflector 41. The time-varying beam reflector 41 reflects the source beam 40 to a telecentric scanning lens 42. The time variation of the time-varying beam reflector 41 causes a moving spot (scanning beam 43) to illuminate a transparent sample 44. The time-varying beam reflector 41 and the telecentric scanning lens 42 are configured to illuminate the transparent sample 44 with the scanning beam 43 at an incident angle not exceeding one degree from the Brewster angle of the transparent sample 44. A focusing lens 46 is configured to be illuminated by scattered radiation from the transparent sample 44. The scattered radiation radiates from the top surface and the bottom surface of the transparent sample 44. The focusing lens 46 may be referred to as a light collector. In one example, the focusing lens 46 is configured to be oriented along an axis incident on a plane perpendicular to the scanning beam 43. In one example, the focusing lens 46 is a low F-number camera lens. The focusing lens 46 focuses the light onto a focal plane 47. Spatial filter 48 is located at focal plane 47 and operates to filter out scattered radiation from the bottom surface of transparent sample 44, while allowing scattered radiation from the top surface of transparent sample 44 to pass through to collimating lens 49. Collimating lens 49 is configured along an axis perpendicular to scanning beam 43. In one example, spatial filter 48 is a slit-shaped spatial filter to remove scattered light from the bottom surface of transparent sample 44. In another example, collimating lens 49 is a pair of achromatic lenses that shape the scattered radiation into a circular spot illuminating detector 50. In yet another example, detector 50 is a photomultiplier tube.

[0096] In another example, the scattered radiation optical inspector also includes a processor and a memory. The processor is used to read the output signal generated by detector 50 and store the light intensity value indicated by the output signal in the memory. The processor can also be used to determine the presence and type of defects. The processor can also be used to generate a defect map across a region of the transparent sample. The processor can also be configured to communicate the defect map to another device or to a monitor.

[0097] The scattered radiation optical inspector described above collects scattered radiation from the irradiated transparent sample 44 at an angle nearly perpendicular to the incident angle of the scanning beam 43. Furthermore, the scattered radiation optical inspector can separate scattered radiation from the top surface of the transparent sample from scattered radiation from the bottom surface of the transparent sample, providing a valuable capability to detect defects on one side of the transparent sample.

[0098] The scattered radiation optical inspection instrument can be used with Figure 3 The phase delay optical inspection is integrated because both inspections require the transparent sample to be illuminated at an incident angle that differs from the Brewster angle of the transparent sample by no more than one degree.

[0099] Figure 11 This is an illustration of a scattered radiation map that demonstrates the detection of defects by detecting changes in the intensity of scattered radiation from a surface of a transparent sample. For example, the scattered radiation map may show the location of a decrease in intensity measured by detector 50. This decrease in measured intensity could be an indication of a change in film thickness, a film defect, or a particle defect. In another example, the scattered radiation map may show the location of an increase in intensity measured by detector 50. This increase in measured intensity could also be an indication of a change in film thickness, a film defect, or a particle defect.

[0100] Figure 12 This is a flowchart 100 illustrating the steps for performing phase delay defect detection. In step 101, the transparent sample is illuminated at an incident angle no greater than one degree larger than or no greater than one degree smaller than the Brewster angle of the transparent sample. In step 102, the input and output mirrors are rotated such that the input mirror is at its minimum incident angle when the output mirror is at its maximum incident angle. In step 103, light reflected from the bottom surface of the sample is blocked. In step 104, the delay (polarization change) of light reflected from the top surface is measured. In step 105, the presence of a defect at the scanned location is determined based on the measured delay and one or more thresholds.

[0101] Figure 13 This is a flowchart 200 illustrating the steps for performing scattered radiation defect detection. In step 201, the transparent sample is irradiated at an incident angle no greater than one degree larger than or no greater than one degree smaller than the Brewster angle of the transparent sample. In step 202, the scattered radiation from the transparent sample is focused using a lens oriented at a normal angle to the incident plane of the moving irradiation spot on the sample. In step 203, scattered radiation from the bottom surface of the transparent sample is blocked. In step 204, scattered radiation from the top surface of the transparent sample is collimated. In step 205, the intensity of the scattered radiation from the top surface of the transparent sample is measured. In step 206, the presence of a defect at the scanned location is determined based on the measured intensity and one or more threshold values.

[0102] Area Detector Optical Inspection Figure 14A novel optical inspector for a region detector is illustrated. In one embodiment, the optical inspector for a region detector includes a radiation source 110, a feed mirror 111, a linear time-varying beam reflector 112, a scanning lens 113, a descanning lens 115, a linear time-varying beam reflector 116, a feed mirror 117, a focusing lens 118, a stopper 119 located at the focal plane 120 of the focusing lens, and a detector 121. The optical inspector for a region detector may also include a processor 122 and a memory 123 configured to process and store intensity and / or phase output signals received from the detector 121.

[0103] In one example, the input mirror 112 and the output mirror 116 are linear time-varying beam reflectors whose reflection angles change linearly as they rotate. The input mirror 112 and the output mirror 116 can also be controlled by electrical signals, such as a signal generator. The input mirror 112 and the output mirror 116 may be referred to as galvanometer mirrors.

[0104] In one example, the detector is a two-cell detector. Figure 4 An example of a dual-cell detector is illustrated. The dual-cell detector has a center line 31 separating a first light sensor and a second light sensor. The dual-cell detector can be configured to output a first signal indicating the difference between the light intensity measured on one side of the detector and the light intensity measured on the other side of the detector. The dual-cell detector can also be configured to output a second signal indicating the sum of the light intensity measured on one side of the detector and the light intensity measured on the other side of the detector.

[0105] In operation, radiation source 110 outputs a laser beam. In an alternative embodiment, the phase of the output laser beam can be adjusted by a half-wave plate positioned along the path of the output laser beam. The output laser beam illuminates input mirror 111 and is reflected toward input mirror 112, and then reflected to scanning lens 113. Scanning lens 113 operates to focus the scanning beam onto the transparent sample. In one example, scanning lens 113 is a telecentric scanning lens. Scanning lens 113 can be configured such that the scanning beam output from scanning lens 113 illuminates transparent sample 114 at an angle not exceeding one degree different from the Brewster angle of transparent sample 114. Another type of lens that can replace the scanning lens is an achromatic lens.

[0106] In one example, the transparent sample is glass. In another example, the transparent sample is a thin film deposited on a transparent material. Other examples of transparent samples include, but are not limited to, sapphire, fused silica, quartz, silicon carbide, and polycarbonate.

[0107] The descanning lens 115 operates to focus the signal from the transparent sample onto the output mirror 116. In one example, the descanning lens 115 is a telecentric scanning lens that is substantially the same as the scanning lens 113. The descanning lens 115 can be a telecentric lens or an achromatic lens. Utilizing substantially the same lens for both the scanning lens 113 and the descanning lens 115 allows the system to focus light reflected from a very thin cross-sectional region of the transparent sample. Other examples of descanning lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0108] Feed mirror 117 operates to reflect the signal from output mirror 116 to focusing lens 118. Focusing lens 118 has a focal plane 120. At the focal point of focusing lens 118, two spots will exist (assuming the sample is transparent and defect-free), and these spots correspond to signals from the top and bottom surfaces of the sample. In one example, focusing lens 118 is an achromatic lens. Other examples of focusing lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0109] In one example, when the output mirror 116 is positioned at the midpoint of its rotation range, the light reflected by the output mirror 116 illuminates the feed mirror at an angle not exceeding thirty degrees from the normal angle of the output mirror 116 (the second time-varying beam reflector).

[0110] The blocker 119 is located near the focal plane 120 and operates to block all signals from the transparent sample except for a portion of the signal within the desired region of the transparent sample. For example, as Figure 14 As illustrated, the blocker 119 can be configured to block signals from the bottom surface and the top surface of the transparent sample, while allowing signals from a desired area of ​​the transparent sample to pass through. Figure 14 An extended illustration of one embodiment of the blocker 119 is shown. The extended illustration shows the blocker as a slit with a central opening that allows light to pass through while blocking the remainder of the blocked light. It should be noted herein that other blocker configurations can be used to achieve the desired blocking. Figure 15 An example location to be probed on a transparent sample is shown.

[0111] Figure 17 An example of the desired region to be probed in a transparent sample is shown. Figure 17 The examination showed that the desired region to be probed within the transparent sample did not include the top and bottom surfaces of the transparent solid. Instead, the desired region to be probed only included a narrow internal region of the transparent sample. It should be noted that... Figure 17The desired detection area illustrated herein is exemplary, and in operation, any area up to and including the top and bottom surfaces of the transparent sample may be part of the desired detection area.

[0112] Figure 16 The region of the reflected light is illustrated from the perspective of the path of the reflected light blocked by the blocker 119. The position of the light reflected from different regions of the transparent sample varies from the center point (reflected from the top surface) along a line that terminates at the position of the bottom surface. Figure 16 The selectable portion of the reflected light illustrated is controlled by the blocker 119 as described above. Therefore, adjusting the slit width and position of the blocker 119 will adjust the height and position of the area of ​​the transparent sample that will travel through the blocker 119 and be subsequently detected for defect detection.

[0113] In this way, the area detection optical inspector can distinguish between reflections originating from a desired area of ​​a transparent sample and reflections originating from outside the desired area of ​​the transparent sample. In one example, the blocker 119 is a mirror. Other examples of blocker materials include, but are not limited to, absorbing materials, blackened aluminum sheets, and black-painted metal sheets.

[0114] An optional collimating lens (not shown) may be used to collimate a signal from a transparent sample that is not blocked by the blocker 119. In one example, the collimating lens is an achromatic lens. Other examples of collimating lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0115] Unobstructed light reflected from the desired area of ​​the transparent sample then illuminates detector 121. In response to illumination, detector 121 outputs a signal proportional to the intensity of the light illuminating it. The output signal is then processed by processor 122 to determine whether defects exist in the desired area of ​​the transparent sample 114.

[0116] Figure 18 An example is illustrated using a threshold to determine the presence of a defect in a desired region of a transparent sample. The measured intensity of reflected light from the desired region of the transparent sample through blocker 119 is compared to the threshold. In one example, when the measured intensity is greater than the threshold, a defect is determined to exist in the desired detection region. Alternatively, when the measured intensity is less than or equal to the threshold, a defect is determined to not exist in the desired detection region.

[0117] It should also be noted in this article that Figure 3 The phase delay optical inspector illustrated above can also be used to implement an optical inspector for a region detector. (See above regarding...) Figure 3 As described, a phase-delay optical inspector can be used to measure the phase change of reflected light from a transparent sample. (The following is a continuation of the previous sentence, which is not directly related to the description of the inspector.) Figure 14 The described blocker 119 and Figure 3 The phase delay measurement unit combination allows for defect detection by analyzing the phase change of the reflected light rather than its intensity change. In this scenario, the polarization of the reflected light passing through the blocker 119 can be compared with one or more thresholds to determine whether a defect exists in the desired detection area.

[0118] Figure 19 This is a flowchart 300 illustrating the region detection steps for detecting defects within a desired area of ​​a transparent sample. In step 301, the transparent sample is illuminated at an incident angle between zero and ninety degrees relative to its surface. In step 302, the reflected radiation from the transparent sample is focused at a focal plane using a scanning lens and a descanning lens. In step 303, one or more portions of the reflected radiation are blocked at the focal plane using a blocker. In step 304, the intensity of the reflected radiation not blocked by the blocker is measured. In step 305, a defect is determined, at least in part, to exist in an area of ​​the transparent sample where the unblocked portion of the reflected light is reflected, based on the measured intensity and one or more threshold values.

[0119] Defect detection using multiple measured polarization values In addition to determining the presence of a defect in the desired detection area by comparing a single intensity or single-phase measurement with a set threshold, the presence of a defect can also be determined by comparing a single measurement with a set of other measurements obtained from the same sample. Figure 3 The optical inspector can be used at multiple locations on a transparent sample, as described above. The measurements taken at these multiple locations can then be post-processed to determine if defects exist at each measurement location on the transparent sample.

[0120] In one example, the presence of a defect at a first location is determined by comparing a single measured polarization at a first location on the sample with the average of measured polarization values ​​over a predefined distance from the first location. In this way, all measured polarization values ​​over the predefined distance from the first location are summed and divided by the count of acceptable measurements. The resulting average polarization value is then compared with a single polarization measurement performed at the first location.

[0121] In a first embodiment, if the measured phase value is greater than or less than the average polarization value of the group plus or minus a certain threshold, a defect is determined to exist at the first location. If the measured phase value is less than or equal to the average polarization value of the group plus or minus a certain threshold, a defect is determined not to exist at the first location. Depending on the nature of the defect, a measured phase value greater than or less than the average value plus or minus a threshold can be considered a defect.

[0122] In the second embodiment, if the measured phase value is greater than the average polarization value of the group by more than a threshold, a defect is determined to exist at the first location. If the measured phase value is less than the average polarization value of the group by more than a threshold, a defect is determined to not exist at the first location.

[0123] In another example, the presence of a defect at a first location is determined by comparing a single measured polarization at a first location on the sample with the median of measured polarization values ​​within a predefined distance from the first location. In this way, all measured polarization values ​​within the predefined distance from the first location are sorted and then counted. The value at the location whose total is divided by two is selected as the median. The resulting median polarization value of this group is then compared with a single polarization measurement performed at the first location.

[0124] In the first embodiment, if the measured phase value is greater than the median polarization value of the group by more than a threshold, a defect is determined to exist at the first location. If the measured phase value is less than the median polarization value of the group by more than a threshold or equal to the median polarization value of the group, a defect is determined to not exist at the first location.

[0125] Figure 25 This is a flowchart 500 illustrating defect detection steps using multiple measured polarization values. In step 501, a zone-detection optical inspector is used to scan the sample to measure polarization values ​​across the sample. In step 502, a processing filter is applied to the raw measured data. This step is optional. In step 503, a group of polarization values ​​(mean or median) for a set of polarization measurements is determined. In step 504, the group of polarization values ​​is compared with individual measured polarization values. In step 505, based on the comparison result between the group of polarization values ​​and the individual measured polarization values, it is determined that a defect exists at the location where the individual measured polarization was measured.

[0126] Scattered radiation defect depth detection As described above, using Figure 10 The scattered radiation optical detector can detect defects by analyzing a single scattered radiation measurement. However, Figure 10 The scattered radiation optical detector can be used in a new and novel way to detect defects by analyzing the distance between multiple scattered radiation events. Figure 10 This new and novel use of the scattered radiation optical detector also removes the requirement for the scanning beam to be oriented at an incident angle no greater than or less than the Brewster angle of the transparent sample by no more than one degree. Instead, Figure 10 The novel and innovative use of the scattered radiation optical detector allows the scanning beam to be oriented at any incident angle.

[0127] Figure 10This novel application of the scattered radiation optical detector allows for the determination of the x, y, and z positions of defects by analyzing the distances between multiple scattered radiation events. Scattered radiation events are those where the measured intensity of scattered radiation exceeds a threshold intensity value.

[0128] Figure 20 The first example of defect depth detection is illustrated below. An inclusion (defect) 251 is located at a depth within a transparent sample 250. d At time t1, the incident beam is perpendicular to the normal. The incident angle of radiation on the transparent sample is 250 degrees. Due to the refractive index of the transparent sample... n When light travels from the top surface of a transparent sample to its bottom surface, it is redirected to different angles. Upon reaching the bottom surface, the light is reflected upwards at an equal but opposite angle. This same opposite angle continues until the reflected light reaches the top surface, where it is redirected to an angle equal to but opposite to the angle of incidence. Assuming the path of light through the transparent sample is known, the presence and depth of defects can be detected by analyzing multiple scattered radiation events measured along a single axis.

[0129] The following equation represents the thickness of the transparent sample ( t ), the interval between scattered radiation events ( x ), refractive index of transparent samples ( n ), the incident angle of the scanning beam ( ) and the depth of the defect ( d The relationship between ).

[0130]

[0131] The equation is correct as long as the intervals between scattered radiation events are within a certain range.

[0132] When the interval between scattered radiation events is greater than a first threshold, the defect or particle is determined to be located on the bottom surface of the transparent sample.

[0133] Figure 21Two scattered radiation events caused by top surface particles are illustrated. At time t1, the incident light beam enters the transparent sample, is reflected from the bottom surface of the transparent sample, and illuminates the top surface particles as the beam leaves the top surface of the transparent sample. This causes the first scattered radiation event at time t1. At time t2, before the beam has a chance to enter the transparent sample, the incident light beam illuminates the top surface particles. This causes the second scattered radiation event. For a given transparent wafer with constant thickness and refractive index, these two scattered radiation events caused by the top surface particles will always be separated by the same distance. Therefore, any interval between scattered radiation events greater than or less than this fixed distance is not a top surface particle. Another criterion for the two scattered events is that they must be directly above each other, i.e., they must have the same x-coordinate (within the specified tolerance).

[0134] Figure 20 Two scattered radiation events caused by an inclusion (a defect within a transparent sample) are illustrated. At time t1, the incident light beam enters the transparent sample, is reflected from the bottom surface, and travels towards the top surface, illuminating the inclusion. This causes the first scattered radiation event at time t1. At time t2, the incident light beam enters the transparent sample and then illuminates the inclusion. This causes the second scattered radiation event. Compared to the interval between the two scattered radiation events caused by the top surface particles, the two scattered radiation events caused by the inclusion will always be closer together. Therefore, any interval between scattered radiation events caused by a single inclusion (a defect within a transparent sample) will never be greater than the interval between the scattered radiation events caused by the top surface particles. Another criterion for the two scattered events is that they must be directly above each other, i.e., they must have the same x-coordinate (within the specified tolerance).

[0135] Therefore, two scattered radiation events with an interval larger than the fixed interval between scattered radiation events caused by top surface particles (and having the same x-coordinate) are not caused by top surface particles, nor by inclusion particles. Thus, through the elimination process, the two scattered radiation events with an interval larger than the fixed interval between scattered radiation events caused by top surface particles must be a result of bottom surface particles.

[0136] When the interval between scattered radiation events is less than a first threshold and greater than a second threshold, the defect or particle is determined to be located on the top surface of the transparent sample. Figure 21 The reasoning for this second threshold is illustrated below. As discussed above, for a given transparent wafer with constant thickness and refractive index, the two scattered radiation events caused by the top surface grains will always be separated by the same distance. Therefore, as Figure 23As shown, the spacing between the top surface particles is relatively narrow because all top surface particles should cause two scattering events at substantially similar distances. In practice, a narrow range is used to identify top surface particles due to measurement accuracy and potential variations in particle size and shape. Another criterion for the two scattering events is that they must be directly above each other, i.e., they must have the same x-coordinate (within the specified tolerance).

[0137] When the interval between scattered radiation events is less than a second threshold and greater than a third threshold, the defect or particle is determined to be an inclusion defect (not a surface defect or particle) located within the transparent sample, and the above equation applies to determine the depth of the defect within the transparent sample. (See above regarding...) Figure 20 As discussed, any interval between scattered radiation events caused by a single inclusion (a defect within a transparent sample) will never be greater than the interval between scattered radiation events caused by top surface particles. Therefore, any substantial interval between scattered radiation events smaller than the fixed interval between scattered radiation events caused by top surface particles must be an inclusion.

[0138] When the interval between scattered radiation events is less than the third threshold, the defect or particle is determined to be located on the bottom surface of the transparent sample. Figure 22 The inference for this first threshold is illustrated in the example. Figure 22 The example illustrates that the bottom surface particles are only irradiated by the incident beam at time t1. Therefore, theoretically, there should only be a single scattered radiation event for the bottom surface particles. However, due to measurement accuracy and variations in the size and shape of the bottom surface particles, the practical range of very closely spaced scattered radiation events has been determined to be caused by the bottom surface particles.

[0139] Figure 23 This is an example of the range of intervals between scattered radiation events in a 500-micrometer-thick transparent sample with a scanning beam incident angle of 56.3° and a refractive index of 1.5. When the interval between scattered radiation events is greater than a first threshold of 686 micrometers, the defect or particle is determined to be located at the bottom surface of the transparent sample, and the above equation does not apply. When the interval between scattered radiation events is less than the first threshold of 686 micrometers and greater than a second threshold of 646 micrometers, the defect or particle is determined to be located at the top surface of the transparent sample. When the interval between scattered radiation events is less than the second threshold of 646 micrometers and greater than a third threshold of 20 micrometers, the defect or particle is determined to be an inclusion defect (not a surface defect or particle) located within the transparent sample, and the above equation applies to determine the depth of the defect in the transparent sample. Another criterion proposed for two scattered radiation events is that they must be directly above each other, i.e., they must have the same x-coordinate (within the specified tolerance). When the interval between scattered radiation events is less than the third threshold of 20 micrometers, the defect or particle is determined to be located at the bottom surface of the transparent sample.

[0140] Figure 24 This is a flowchart 400 illustrating the steps for performing scattered radiation defect depth detection. In step 401, a transparent sample is illuminated. In step 402, the scattered radiation from the transparent sample is focused using a lens. In step 403, the intensity of the scattered radiation from the transparent sample is measured along a single axis. In step 404, the locations where the scattered radiation increases along the single axis are determined. In step 405, the distance between the locations where the scattered radiation increases along the single axis is used to determine the presence and depth of the defect.

[0141] Figure 26 This is a diagram of another phase-retardation optical inspector. Additional tests and studies have indicated that the phase-retardation optical inspector described above functions well even at differences of up to plus or minus ten (10) degrees from the Brewster angle (not just plus or minus 1 degree as disclosed in the parent application). Performance varies with distance from the Brewster angle. For example, if the device operates at an incident angle of 56.8 degrees (0.1 degrees from the Brewster angle), the difference between 100 angstroms of MgF2 and the bare substrate is a retardation of about 70 degrees. However, if the device operates at an incident angle of 60 degrees, the difference between 100 angstroms of MgF2 and the bare substrate is a retardation of about 10 degrees. Thus, a sevenfold increase in sensitivity is sacrificed, but the inspector still functions even when operated at an angle 3.3 degrees different from the Brewster angle of the substrate.

[0142] Figure 27 This is a diagram of another optical detector for scattered radiation, depicting an incident angle that differs from Brewster's angle by positive or negative ten (10) degrees.

[0143] Figure 28 This is another flowchart illustrating the steps for performing phase delay defect detection using an incident angle that differs from Brewster's angle by positive or negative ten (10) degrees.

[0144] Figure 29 This is another flowchart illustrating the steps for performing scattered radiation defect detection using an incident angle that differs from Brewster's angle by positive or negative ten (10) degrees.

[0145] Figure 30This is a diagram of an angle-independent surface height optical inspector. In one embodiment, the angle-independent surface height optical inspector includes a radiation source 510, an entrance lens 511, a linear time-varying beam reflector 512, a scanning lens 513, a descanning lens 515, a linear time-varying beam reflector 516, an exit lens 517, a focusing lens 518, an unpolarized beam splitter 519, a collimating lens 523, a half-wave plate 524, a polarized beam splitter 525, detectors 526 and 527, an unpolarized beam splitter 520, detectors 521 and 522, a processor 528, and a memory 529. Optionally, the angle-independent surface height optical inspector may also include a half-wave plate 526.

[0146] The descanning lens 515 is located at a distance of approximately one focal length from the sample 514. Similarly, the linear time-varying beam reflector 516 is located at a distance of approximately one focal length from the descanning lens 515.

[0147] Processor 528 and memory 529 are configured to process and store intensity, irradiation position and / or phase output signals received from detectors 521, 522, 526 and 527.

[0148] In one example, the input mirror 512 and the output mirror 516 are linear time-varying beam reflectors whose reflection angles change linearly as they rotate. The input mirror 512 and the output mirror 516 can also be controlled by electrical signals, such as a signal generator. The input mirror 512 and the output mirror 516 may be referred to as galvanometer mirrors.

[0149] In one example, detectors 521, 522, 526, and 527 are any combination of dual-cell detectors, quad-cell detectors, and / or localization-sensitive detectors. Figure 4An example of a dual-unit detector is illustrated. The dual-unit detector has a center line 31 separating a first light sensor and a second light sensor. The dual-unit detector can be configured to output a first signal indicating the difference between the light intensity measured on one side of the detector and the light intensity measured on the other side. The dual-unit detector can also be configured to output a second signal indicating the sum of the light intensity measured on one side of the detector and the light intensity measured on the other side. A quad-unit detector includes four light sensors. Typically, the four sensors are of uniform size and separated by a central vertical line and a central horizontal line. The quad-unit detector can be configured to output a signal indicating the difference between the light intensities measured by each light sensor. The quad-unit detector can also be configured to output a second signal indicating the sum of the light intensities measured by each light sensor. A positioning-sensitive detector is a light detector that can typically measure the positioning of a light spot in one or two dimensions at relatively high speeds. The positioning-sensitive detector can be configured to output a first signal indicating the positioning of the illuminated area on the detector. The positioning-sensitive detector can also be configured to output a second signal indicating the intensity of the light illuminating the detector.

[0150] In operation, radiation source 510 outputs a laser beam. In an alternative embodiment, the phase of the output laser beam can be adjusted by a half-wave plate positioned along the path of the output laser beam. The output laser beam illuminates input mirror 511 and is reflected toward input mirror 512, and then reflected to scanning lens 513. Scanning lens 513 operates to focus the scanning beam onto the transparent sample. In one example, scanning lens 513 is a telecentric scanning lens. Scanning lens 513 can be configured such that the scanning beam output from scanning lens 513 illuminates transparent sample 514 at an angle not exceeding ten (10) degrees from the Brewster angle of sample 514. Another type of lens that can replace the scanning lens is an achromatic lens.

[0151] In one example, sample 514 is opaque. In another example, sample 514 is translucent. In yet another example, sample 514 is transparent. For example, transparent sample 514 can be glass, a thin film deposited on a transparent material, sapphire, fused silica, quartz, silicon carbide, and polycarbonate.

[0152] Descanning lens 515 operates to focus the signal from sample 514 onto output mirror 516. In one example, descanning lens 515 is a telecentric scanning lens that is substantially the same as scanning lens 513. Descanning lens 515 can be a telecentric lens or an achromatic lens. Utilizing substantially the same lens for scanning lens 513 and descanning lens 515 allows the system to focus light reflected from a very thin cross-sectional area of ​​a transparent sample. For example, two telecentric scanning lenses can have a field of view greater than one hundred (100) mm. Other examples of descanning lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0153] Feed mirror 517 operates to reflect the signal from output mirror 516 to focusing lens 518. Focusing lens 518 has a focal length based on the characteristics of a focusing lens. At the focal point of focusing lens 518, there will be two spots (assuming the sample is transparent and free of defects), and these spots correspond to signals from the top and bottom surfaces of the sample. In the case of an opaque sample, there will be a single spot from the top surface of the sample. In one example, focusing lens 518 is an achromatic lens with a focal length of 400 (400) mm. Other examples of focusing lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0154] In one example, when the output mirror 516 is positioned at the midpoint of its rotation range, the light reflected by the output mirror 516 illuminates the feed mirror at an angle not exceeding thirty degrees from the normal angle of the output mirror 516 (the second time-varying beam reflector).

[0155] Unpolarized beam splitter 519 is illuminated by light focused by focusing lens 518. When illuminated, unpolarized beam splitter 519 allows a designated portion of the light intensity to pass through to collimating lens 523, and allows the remainder to be reflected back to unpolarized beam splitter 520. The polarization of the transmitted and reflected beams from unpolarized beam splitter 519 remains unchanged. The planes of unpolarized beam splitters 519 and 520 may be parallel to or perpendicular to the plane of the sample. When illuminated, unpolarized beam splitter 520 allows a designated portion of the light intensity to pass through to detector 521, and reflects the remaining intensity back to detector 522 in the opposite direction. The polarization of the transmitted and reflected beams from unpolarized beam splitter 520 remains unchanged. Detectors 521 and 522 are located approximately one focal length from focusing lens 518.

[0156] Detector 521 is configured to track the focus of the beam. Detector 521 has a low bandwidth of approximately 20 kHz. Detector 521 provides feedback to a sample height control system that precisely maintains the sample at the correct focus. An example of the sample height control system is shown in... Figure 33Examples are shown in the text and described in the accompanying description below.

[0157] Detector 522 is configured to measure the micro-surface profile of the sample. Detector 522 has a large bandwidth of approximately 2 MHz. The output signal from detector 522 is used to measure submicron height changes on the sample.

[0158] Collimating lens 523 is configured to redirect the focused beam from focusing lens 518 into a collimated (i.e., parallel) beam. The collimated beam illuminates half-wave plate 524. Half-wave plate 524 rotates the polarization plane of the collimated beam. The rotated collimated beam then illuminates polarizing beam splitter 525. When illuminated, polarizing beam splitter 525 allows all light polarized in one direction to pass through to detector 527 and reflects all light polarized in the other direction to detector 526. The plane of polarizing beam splitter 525 is the same as the plane of the sample.

[0159] Detector 526 is configured to measure the s-component of a polarized beam. Detector 526 may be a dual-cell, quad-cell, position-sensitive detector, PIN diode (single element), or other type of Si-based detector.

[0160] Detector 527 is configured to measure the slope of the sample surface and the p-component of the polarized beam. Detector 527 can be a dual-cell, quad-cell, positioning-sensitive detector, PIN diode (single element), or other type of Si-based detector.

[0161] Figure 31This is a simplified illustration of an angle-independent surface height optical inspector at different sample heights. In the first scenario, sample 600 is located at height "A" and is illuminated by an illumination beam. The illumination beam reflects off the top surface of sample 600, generating a height "A" reflection, which centers on the descanning lens 601. The descanning lens 601 is located at a descanning lens focal length from the illumination point on sample 600. The descanning lens 601 then centers the height "A" reflection onto the focusing lens 602. The focusing lens 602 then centers the height "A" reflection onto a point on detector 603. In the second scenario, the sample is located at height "B" and is illuminated by an illumination beam. Assuming height "B" is lower than height "A", different points on the sample surface are illuminated by a stationary illumination beam. The illumination beam reflects off the top surface of sample 600, generating a height "B" reflection, which is then centered onto an off-center location on the descanning lens 601. The descanning lens 601 is located at a focal length of the descanning lens at a distance from the sample 600. The descanning lens 601 then guides the height "B" reflection to the focusing lens 602. Because the height "B" reflection is off-center from illuminating the descanning lens 601, the height "B" reflection emitted by the descanning lens 601 is not parallel to the height "A" reflection and is also off-center from illuminating the focusing lens 602. The focusing lens 602 then guides the height "B" reflection to a non-centered point on the detector 603. The distance between the illumination point of the height "A" reflection on the detector and the illumination point of the height "B" reflection on the detector is proportional to the change in sample height between the two measurements.

[0162] Figure 32This is a simplified illustration of an angle-independent surface height optical inspector at different sample angles. In the first scenario, sample 610 is positioned at angle "A" and is illuminated by an illumination beam. The illumination beam is reflected from the top surface of sample 610, generating an angle "A" reflection that centers on the descanning lens 611. The descanning lens 611 is located at a descanning lens focal length from the illumination point on sample 610. The descanning lens 611 then centers the angle "A" reflection to the focusing lens 612. The focusing lens 612 then centers the angle "A" reflection on the detector 613. In the second scenario, the sample is positioned at angle "B" and is illuminated by an illumination beam. Assuming angle "B" is different from angle "A", the same point on the surface of the sample is illuminated by a stationary illumination beam, but the angle of the reflected beam is different. The illumination beam is reflected from the top surface of sample 610, generating an angle "B" reflection, which is guided to an off-center position on descanning lens 611. Descanning lens 611 is located at a descanning lens focal length from the illumination point on sample 610. Descanning lens 611 then guides the angle "B" reflection to focusing lens 612. When the angle "B" reflection irradiates descanning lens 611 off-center, due to the focusing function of descanning lens 611, the angle "B" reflection emitted by descanning lens 611 is parallel to the angle "A" reflection. The angle "B" reflection also irradiates focusing lens 612 off-center. By providing additional focusing, focusing lens 612 then guides the angle "B" reflection to a centered point on detector 613. The distance between the illumination point of angle "A" reflection on the detector and the illumination point of angle "B" reflection on the detector is almost zero. Therefore, using this optical system, consistent top surface optical measurements can be performed regardless of the angle of the sample being measured. In this sense, the optical system is independent of the angle of the sample being measured.

[0163] Figure 33 This is a diagram of the electrical system used for a beam-tracking positioning sensitive detector. During normal operation, the sample (or wafer 714) is supported by a chuck 713, which rests on a Z-motor 711 (providing upward and downward movement). The Z-motor 711 is part of a feedback loop that originates from the beam-tracking detector (…). Figure 30Detector 521 receives height information and uses this information to maintain the sample height via lead screw 712 (by moving the Z-motor up or down). This action means that the focus of the beam from focusing lens 518 is always maintained at the position of detector 521 (and detector 522). The system is implemented such that there is no focusing error at detectors 521 and 522. If the sample height is not maintained at a constant position, focusing positioning errors will occur at detectors 521 and 522. Detector 521 is not used to directly measure the sample height, but rather to determine the sample height based on the distance the Z-motor has moved to maintain focus (i.e., to keep the feedback error signal to a minimum). The bending or warping of the wafer is inversely proportional to the amount of movement of the Z-stage. This has the added benefit of keeping the focus of the beam on the sample at a constant size regardless of sample bending or warping. The micro-analyzer channel (detector 522) provides a high-frequency channel (bandwidth of several MHz). Given the required high physical movement speed, no mechanical Z-motion can follow this bandwidth. Therefore, detector 521 and the feedback loop maintain focus on detector 522, and the high-frequency signal received from detector 522 is a direct measurement of the height of defects or features on the sample (wafer) surface. The height is then normalized by the reflectivity of the sample. In one example, normalization is performed by dividing the difference between intensity measurements by the sum of the intensity measurements. Thus, detector 521 is a low-frequency positioning sensitive detection channel (operating in the kHz range) that measures wafer bending or warping by calculating the bending or warping of the wafer based on the movement of the Z-motor, and simultaneously maintains the focus of focusing lens 518 on detectors 521 and 522 (thus minimizing focusing error), while detector 522 is a high-frequency positioning sensitive detector channel (in the MHz range) that directly measures the height of defects or features on the wafer surface after normalization is applied.

[0164] Figure 34 This is a flowchart 800 of an angle-independent surface height optical inspector. In step 801, a light beam is generated. In step 802, the light beam is guided to the sample. In step 803, the reflected light beam from the sample is descanned to generate a first descanned beam. Descanning is performed at approximately one focal length from the point where the beam illuminates the sample, which is the location of the descanning lens. In step 804, the first descanned beam is focused to generate a focused beam. In step 805, the position of the focused beam is measured. This measurement is performed at approximately one focal length from the focusing lens.

[0165] Temporal multiplexing defect scanner Figure 35This is a diagram of a time-domain multiplexed defect scanner. The time-domain multiplexed defect scanner is also capable of performing the functionality described above regarding the angle-independent surface height optical inspector. In one embodiment, the time-domain multiplexed defect scanner includes a first radiation source 810, a second radiation source 830, an acousto-optic modulator 831, an optional half-wave field 833, a dichroic mirror 832, a feed mirror 811, a linear time-varying beam reflector 812, a scanning lens 813, a descanning lens 815, a linear time-varying beam reflector 816, a feed mirror 817, a focusing lens 818, an unpolarized beam splitter 819, an optional collimating lens 823, a half-wave plate 824, a polarized beam splitter 825, detectors 826 and 827, unpolarized beam splitters 820, 821, and 822, a processor 828, and a memory 829.

[0166] The descanning lens 815 is located at a distance of approximately one focal length from the sample 814. Similarly, the linear time-varying beam reflector 816 is located at a distance of approximately one focal length from the descanning lens 815.

[0167] The processor 828 and memory 829 are configured to process and store intensity, irradiation position and / or phase output signals received from detectors 821, 822, 826 and 827.

[0168] In one example, the input mirror 812 and the output mirror 816 are linear time-varying beam reflectors whose reflection angles change linearly as they rotate. The input mirror 812 and the output mirror 816 can also be controlled by electrical signals, such as a signal generator. The input mirror 812 and the output mirror 816 may be referred to as galvanometer mirrors.

[0169] In another example, detectors 821, 822, 826, and 827 are any combination of dual-cell detectors, quad-cell detectors, and / or localization-sensitive detectors. Figure 4An example of a dual-unit detector is illustrated. The dual-unit detector has a center line 31 separating a first light sensor and a second light sensor. The dual-unit detector can be configured to output a first signal indicating the difference between the light intensity measured on one side of the detector and the light intensity measured on the other side. The dual-unit detector can also be configured to output a second signal indicating the sum of the light intensity measured on one side of the detector and the light intensity measured on the other side. A quad-unit detector includes four light sensors. Typically, the four sensors are of uniform size and separated by a central vertical line and a central horizontal line. The quad-unit detector can be configured to output a signal indicating the difference between the light intensities measured by each light sensor. The quad-unit detector can also be configured to output a second signal indicating the sum of the light intensities measured by each light sensor. A positioning-sensitive detector is a light detector that can typically measure the positioning of a light spot in one or two dimensions at relatively high speeds. The positioning-sensitive detector can be configured to output a first signal indicating the positioning of the illuminated area on the detector. The positioning-sensitive detector can also be configured to output a second signal indicating the intensity of the light illuminating the detector.

[0170] In operation, both radiation sources 810 and 830 output laser beams. In an alternative embodiment, the phase of the output laser beam can be adjusted by a half-wave plate 833 positioned along the path of the laser beam output from the first radiation source 810. The laser beam output from the first radiation source 810 then passes through a dichroic mirror 832 and illuminates a feed mirror 811, which then reflects the laser beam toward an input mirror 812, which in turn reflects it toward a scanning lens 813, which then guides the laser beam toward the sample 814.

[0171] In a second alternative embodiment, the amplitude of the output laser beam can be adjusted by an acousto-optic modulator 831 positioned along the path of the laser beam output from the second radiation source 830. The laser beam output from the second radiation source 830 is then reflected by a dichroic mirror 832 toward a feed mirror 811, which in turn reflects the laser beam toward an input mirror 812. The input mirror 812 then reflects the laser beam toward a scanning lens 813, which then guides the laser beam toward the sample 814. The scanning lens 813 operates to focus the scanning beam onto the sample.

[0172] In another embodiment, the first wavelength is approximately 532 nanometers and the second wavelength is approximately 349 nanometers.

[0173] In one example, scanning lens 813 is a telecentric scanning lens designed to operate at two wavelengths. Scanning lens 813 can be configured such that the scanning beam output from scanning lens 813 illuminates sample 814 at an angle not exceeding ten (10) degrees from Brewster's angle of sample 814. Another type of lens that can be used as an alternative to scanning lens is an achromatic lens.

[0174] In this way, the laser beam output from the first radiation source 810 or the laser beam output from the second radiation source 830 can be adjusted at any time and used as a single scanning laser beam for scanning the sample 814.

[0175] In one example, sample 814 is opaque. In another example, sample 814 is translucent. In yet another example, sample 814 is transparent. For example, sample 814 can be glass, a thin film deposited on a transparent material, sapphire, fused silica, quartz, silicon carbide, and polycarbonate.

[0176] The descanning lens 815 operates to focus the signal from sample 814 onto output mirror 816. In one example, descanning lens 815 is a telecentric scanning lens that is substantially the same as scanning lens 813. Descanning lens 815 can be a telecentric lens or an achromatic lens. Utilizing substantially the same lens for scanning lens 813 and descanning lens 815 allows the system to focus light from the top of a transparent sample and the bottom of the same sample into two distinct spots, such as... Figure 26 The focal plane is shown at 620°. Using two identical scanning lenses 813 and 815 means that any optical aberrations are minimized, making the top spot distinct from the bottom spot and therefore easily separable. As a result, specular signals from the top and bottom surfaces of a transparent surface with a thickness of less than 0.5 mm can be easily separated. Telecentric scanning lenses can be designed to have a large field of view, making substrate scanning more efficient. For example, two telecentric scanning lenses can have a field of view greater than one hundred (100) mm. Other examples of descanning lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0177] Feed mirror 817 operates to reflect the signal from output mirror 816 to focusing lens 818. Focusing lens 818 has a focal length based on the characteristics of a focusing lens. At the focal point of focusing lens 818, there will be two spots (assuming the sample is transparent and free of defects), and these spots correspond to signals from the top and bottom surfaces of the sample. In the case of an opaque sample, there will be a single spot from the top surface of the sample. In one example, focusing lens 818 is an achromatic lens with a focal length of 400 (400) mm. Other examples of focusing lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0178] In one example, when the output mirror 816 is positioned at the midpoint of the rotation range of the output mirror 816, the light reflected by the output mirror 816 illuminates the feed mirror at an angle not greater than thirty (30) degrees from the normal angle of the output mirror 816 (the second time-varying beam reflector).

[0179] Unpolarized beam splitter 819 is illuminated by light focused by focusing lens 818. When illuminated, unpolarized beam splitter 819 allows a designated portion of the light intensity to pass through to collimating lens 823, and allows the remainder to be reflected back to unpolarized beam splitter 820. The polarization of the transmitted and reflected beams from unpolarized beam splitter 819 remains unchanged. The planes of unpolarized beam splitters 819 and 820 may be the same as or perpendicular to the plane of the sample. When illuminated, unpolarized beam splitter 820 allows a designated portion of the light intensity to pass through to detector 821, and reflects the remaining intensity back to detector 822 in the other direction. The polarization of the transmitted and reflected beams from unpolarized beam splitter 820 remains unchanged. Detectors 821 and 822 are located approximately one focal length from focusing lens 818.

[0180] Detector 821 is configured to track the focus of the beam. Detector 821 has a low bandwidth of approximately 20 kHz. Detector 821 provides feedback to a sample height control system that precisely maintains the sample at the correct focus. An example of the sample height control system is shown in... Figure 33 Examples are provided and described in the accompanying description above.

[0181] Detector 822 is configured to measure the micro-surface profile of the sample. Detector 822 has a large bandwidth of approximately 2 MHz. The output signal from detector 822 is used to measure submicron height changes on the sample.

[0182] Collimating lens 823 is optional and configured to redirect the focused beam from focusing lens 818 into a collimated (i.e., parallel) beam. The collimated beam illuminates half-wave plate 824. Half-wave plate 824 rotates the polarization plane of the collimated beam. The rotated collimated beam then illuminates polarizing beam splitter 825. When illuminated, polarizing beam splitter 825 allows all light polarized in one direction to pass through to detector 827 and reflects all light polarized in the other direction to detector 826. The plane of polarizing beam splitter 825 is the same as the plane of the sample.

[0183] Detector 826 is configured to measure the S-component of a polarized beam. Detector 826 can be a dual-cell, quad-cell, position-sensitive detector, PIN diode (single element), or other type of Si-based detector.

[0184] Detector 827 is configured to measure the slope of the sample surface and the p-component of the polarized beam. Detector 827 can be a dual-cell, quad-cell, positioning-sensitive detector, PIN diode (single element), or other type of Si-based detector.

[0185] The first radiation source 810 and the second radiation source 830 can be enabled or disabled by various means. For example, the system can utilize a beam blocker, a radiation source power switch, a shutter, a chopper, an acousto-optic modulator, or an electro-optic modulator. Those skilled in the art will readily understand after reading this application that the desired operation can be achieved using any means for turning each radiation source on or off. Furthermore, those skilled in the art will readily understand after reading this application that the desired operation can be achieved using any means for blocking the output of the radiation sources.

[0186] Figure 36 This is a diagram of an improved optical detector for scattered radiation. The optical detector for scattered radiation includes one or more radiation sources (not shown) for the output source beam 840, a time-varying beam reflector 841, a telecentric scanning lens 842, a focusing lens 846, a spatial filter 848, a collimating lens 849, a detector 850, a compound ellipsoid collector 853, an optional spatial filter 854, and a detector 855.

[0187] In operation, one of one or more radiation sources emits a source beam 840, which illuminates a time-varying beam reflector 841. The time-varying beam reflector 841 reflects the source beam 840 to a telecentric scanning lens 842. The time variation of the time-varying beam reflector 841 causes a moving spot (scanning beam 843) to illuminate a sample 844. The time-varying beam reflector 841 and the telecentric scanning lens 842 are configured to illuminate a transparent sample 844 with the scanning beam 843 at an incident angle not exceeding ten degrees from the Brewster angle of the sample 844. A focusing lens 846 is configured to be illuminated by scattered radiation (also referred to as “tilted scattered radiation”) from the sample 844. The scattered radiation radiates from the top surface and the bottom surface of the sample 844. The focusing lens 846 may be referred to as a light collector. In one example, the focusing lens 846 is configured to be oriented along an axis incident on a plane perpendicular to the scanning beam 843. In one example, the focusing lens 846 is a low F-number camera lens. A focusing lens 846 focuses light onto a focal plane 847. A spatial filter 848 is located at the focal plane 847 and operates to filter out scattered radiation from the bottom surface of the transparent sample 844, while allowing scattered radiation from the top surface of the sample 844 to pass through to the collimating lens 849. The collimating lens 849 is configured along an axis perpendicular to the scanning beam 843. In one example, the spatial filter 848 is a slit-shaped spatial filter to remove scattered light from the bottom surface of the sample 844 and other light not from the top surface of the sample 844. In another example, the collimating lens 849 is a pair of achromatic lenses that shape the scattered radiation into a circular spot illuminating the detector 850. In yet another example, the detector 850 is a photomultiplier tube.

[0188] Irradiating sample 844 also induces normal scattered radiation 856 that does not irradiate focusing lens 846. A compound ellipsoidal collector 853 is used to collect at least a portion of the normal scattered radiation 856 and redirect it to the focusing region. Normal scattered radiation 856 is radiation scattered within an angle range between +45 degrees and -45 degrees with respect to the surface normal. The radiation collected by the compound ellipsoidal collector 853 may optionally be filtered by a spatial filter 854. The spatial filter 854 blocks light that is not from the focal point of the compound ellipsoidal collector. The collected radiation then irradiates normal scattered radiation detector 855. In one example, normal scattered radiation detector 855 is a photomultiplier tube detector. Normal scattered radiation detector 855 measures the intensity of the collected normal scattered radiation, which is used to determine normal dark-field defects. Normally scattered dark-field defects can be photoluminescent defects in SiC, such as stacking faults, triangular defects, step clustering, basal defects, or other types of photoluminescent defects. Other examples of dark-field defects caused by normal scattering are pits, scratches, and grains.

[0189] In another example, the scattered radiation optical inspector also includes a processor and a memory. The processor reads the output signals generated by detectors 850 and 855 and stores the light intensity values ​​indicated by the output signals in the memory. The processor can also be used to determine the presence and type of defects. The processor can also be used to generate a defect map across regions of a transparent sample. The processor can also be configured to communicate the defect map to another device or to a monitor.

[0190] The tilt-scattering radiation optical inspector 845 described above collects scattered radiation from the irradiated sample 844 at an angle nearly perpendicular to the incident angle of the scanning beam 843. If the wavelength of the laser 810 or 830 is short enough (less than about 360 nm), the substrate, such as SiC, will become opaque, and the tilt-scattering radiation optical detector 850 and the normal-scattering radiation optical detector 855 will detect only the scattered radiation from the top surface of the sample, such as a transparent sample like SiC. This provides the valuable ability to detect top-side defects in a single scan.

[0191] Figure 37 This is a timing diagram illustrating the control of a radiation source. The diagram illustrates three values ​​along a single time axis. The first value is the angle of the linearly time-varying beam reflector. The second value is the state of the first radiation source (on / off). The third value is the state of the second radiation source (on / off). Figure 37 In one example illustrated, as the angle of the linearly time-varying beam reflector increases, the first radiation source is activated and the second radiation source is deactivated. This configuration causes the first source beam (output from the first radiation source) to illuminate the sample 844 during a first duration. In the next duration, as the angle of the linearly time-varying beam reflector decreases, the first radiation source is deactivated and the second radiation source is activated. This configuration causes the second source beam (output from the second radiation source) to illuminate the sample 844 during a second duration. Figure 37 In the example illustrated, the pattern is repeated with a 50% duty cycle. However, it should be noted that those skilled in the art will readily recognize that the duty cycle can be varied while still achieving the desired results. The results of this scanning method are illustrated in the accompanying figures below.

[0192] Figure 38 This is a diagram illustrating an exemplary wafer scanning operation using time-division multiplexing of radiation sources. In this example, the system scan width is set to one-third of the sample (wafer) width. The first patterned line scan from left to right is performed using a first light source beam (output from a first radiation source). The second patterned line scan from right to left is performed using a second light source beam (output from a second radiation source). As mentioned above... Figure 37 The alternating scans discussed are repeated across the entire sample (wafer).

[0193] Figure 39 This is a diagram illustrating an exemplary single wafer scan performed using time-division multiplexing of radiation sources. The single scan further illustrates the alternating nature of scans performed using a first radiation source and scans performed using a second radiation source. How these scans are organized and processed is illustrated in the following figures.

[0194] Data measured by the various detectors in the system of the present invention can be organized according to the time of measurement. The system can also store which radiation source, either the first or the second, was activated during each measurement. In this way, the system can perform time-division multiplexing to separate the measured data obtained from either the first light source beam (output from the first radiation source) or the second light source beam (output from the second radiation source). Figure 40 An example is illustrated of sample (wafer) scanning using a first light source beam (output from a first radiation source). Alternatively, Figure 41 An example of sample (wafer) scanning using a second light source beam (output from a second radiation source) is illustrated. Therefore, the system is able to separate data obtained using different radiation sources and aggregate data obtained using the same radiation source. Thus, the system can create two separate aggregated datasets (one from illumination from the first radiation source and the other from illumination from the second radiation source) and process each aggregated dataset separately to detect different sets of potential defects.

[0195] This novel system has many benefits.

[0196] First, time-domain multiplexing of radiation sources allows for the separation of scans using different radiation sources without requiring any source separation mechanism. If the system utilizes two radiation sources simultaneously, it will need to be able to separate the resulting radiation from each of the simultaneous sources. The means for separating the resulting radiation from multiple sources require additional cost and design complexity. Furthermore, systems using two or more simultaneous sources are inherently more susceptible to measurement noise caused by simultaneous secondary radiation sources.

[0197] Secondly, the scanning time of this invention is inversely proportional to the rotational speed of the time-varying beam reflector. Other systems require the sample (wafer) itself to rotate while the beam is stationary. This severely limits the scanning rate of other systems because there are practical limitations on how fast the sample (wafer) can rotate, while the rotational speed of the time-varying beam reflector is several orders of magnitude greater than the fastest possible rotational speed of the sample (wafer). For example, time-varying beam reflectors (such as galvanometers or polygons) can rotate at speeds up to 2 kHz, while the sample (wafer) cannot rotate at speeds higher than 100 Hz without risking damage to the sample (wafer).

[0198] Figure 42This is flowchart 870 for dual-radiation-source scanning. In step 871, the sample is scanned using a first radiation source. In step 872, the obtained radiation is measured based on the scan performed using the first radiation source. In step 873, the sample is scanned using a second radiation source. In step 874, the obtained radiation is measured based on the scan performed using the second radiation source. In step 875, the measured radiation obtained from the scan performed using the first radiation source is processed. In step 876, the presence of defects on the sample is determined at least in part based on the processed measured radiation obtained from the scan performed using the first radiation source. In step 877, the measured radiation obtained from the scan performed using the second radiation source is processed. In step 878, the presence of defects on the sample is determined at least in part based on the processed measured radiation obtained from the scan performed using the second radiation source.

[0199] Figure 43 This is flowchart 880, which describes a dual-wavelength scan using time-division multiplexing. In step 881, a first radiation source operating at a first wavelength is enabled. In step 882, a second radiation source operating at a second wavelength is disabled. In step 883, radiation is measured. In step 884, the first radiation source operating at the first wavelength is disabled. In step 885, the second radiation source operating at the second wavelength is enabled. In step 886, the resulting radiation is measured.

[0200] Figure 44 This is a flowchart 890 of a dual-wavelength scan using two radiation sources and time-division multiplexing. In step 891, a first radiation source operating at a first wavelength is enabled. In step 892, a second radiation source operating at a second wavelength is disabled. In step 893, the radiation received from the sample due to irradiation by the first radiation source is measured. In step 894, the first radiation source operating at the first wavelength is disabled. In step 895, the second radiation source operating at the second wavelength is enabled. In step 896, the radiation received from the sample due to irradiation by the second radiation source is measured. In step 897, a first set of data, including all radiation measurements taken when the first radiation source is enabled, is aggregated. In step 898, a second set of data, including all radiation measurements taken when the second radiation source is enabled, is aggregated. In step 899, the first set of data is processed to determine if defects exist on the sample. In step 900, the second set of data is processed to determine if defects exist on the sample.

[0201] Figure 45This is a flowchart 910 of a dual-wavelength scan using a single radiation source and time-division multiplexing. In step 911, the first radiation source is configured to operate at a first wavelength. In step 912, the resulting radiation from the sample caused by irradiation of the first radiation source operating at the first wavelength is measured. In step 913, the first radiation source is configured to operate at a second wavelength. In step 914, the resulting radiation from the sample caused by irradiation of the first radiation source operating at the second wavelength is measured. In step 915, a first set of data including all resulting radiation measurements taken when the first radiation source operates at the first wavelength is aggregated. In step 916, a first set of data including all resulting radiation measurements taken when the first radiation source operates at the second wavelength is aggregated. In step 917, the first set of data is processed to determine if defects exist on the sample. In step 918, a second set of data is processed to determine if defects exist on the sample.

[0202] Dual-scanning beam separation defect scanner and optical inspector with independent incident angles Figure 46 This is a diagram of a dual-scan beam separation defect scanner with independent incident angles. The dual-scan beam separation defect scanner with independent incident angles is also capable of performing the functionality described above regarding the angle-independent surface height optical inspector. In one embodiment, a dual-scanning beam splitting defect scanner with independent incident angles includes a first radiation source 920, a beam expander 921, an optional half-wave plate 922, a second radiation source 924, a beam expander 925, a dichroic mirror 923, a feed mirror 926, a linear time-varying beam reflector 927, a dual-wavelength scanning lens 928, reflectors 929 to 932, a descanning lens 935, a linear time-varying beam reflector 936, a feed mirror 937, a focusing lens 938, an unpolarized beam splitter 939, an optional collimating lens 943, a half-wave plate 944, a polarized beam splitter 945, a detector 947, a detector 946, an unpolarized beam splitter 940, a detector 942, a detector 941, a processor 948, and a memory 949.

[0203] The descanning lens 935 is located at a distance of approximately one focal length from the sample 934. Similarly, the linear time-varying beam reflector 936 is located at a distance of approximately one focal length from the descanning lens 935.

[0204] The processor 948 and memory 949 are configured to process and store intensity, irradiation position and / or phase output signals received from detectors 941, 942, 946 and 947.

[0205] In one example, the input mirror 927 and the output mirror 936 are linear time-varying beam reflectors whose reflection angles change linearly as they rotate. The input mirror 927 and the output mirror 936 can also be controlled by electrical signals, such as a signal generator. The input mirror 927 and the output mirror 936 may be referred to as galvanometer mirrors.

[0206] In another example, detectors 941, 942, 946, and 947 are any combination of dual-cell detectors, quad-cell detectors, and / or localization-sensitive detectors. Figure 4 An example of a dual-unit detector is illustrated. The dual-unit detector has a center line 31 separating a first light sensor and a second light sensor. The dual-unit detector can be configured to output a first signal indicating the difference between the light intensity measured on one side of the detector and the light intensity measured on the other side. The dual-unit detector can also be configured to output a second signal indicating the sum of the light intensity measured on one side of the detector and the light intensity measured on the other side. A quad-unit detector includes four light sensors. Typically, the four sensors are of uniform size and separated by a central vertical line and a central horizontal line. The quad-unit detector can be configured to output a signal indicating the difference between the light intensities measured by each light sensor. The quad-unit detector can also be configured to output a second signal indicating the sum of the light intensities measured by each light sensor. A positioning-sensitive detector is a light detector that can typically measure the positioning of a light spot in one or two dimensions at relatively high speeds. The positioning-sensitive detector can be configured to output a first signal indicating the positioning of the illuminated area on the detector. The positioning-sensitive detector can also be configured to output a second signal indicating the intensity of the light illuminating the detector.

[0207] In operation, both radiation sources 920 and 924 output laser beams. A beam expander 921 can optionally be used to expand the output beam from the first radiation source. The beam expander allows adjustment of the input beam size, which is inversely proportional to the focused spot size. Common beam expanders are Galilean or Keplerian beam expanders. A commercially available beam expander is the ZBE3UVB manufactured by Thor Labs. In an alternative implementation, the phase of the output laser beam can be adjusted by a half-wave plate 922 positioned along the path of the laser beam output from the first radiation source 920. The laser beam output from the first radiation source 920 then passes through a dichroic mirror 923 and illuminates a feed mirror 926, which then reflects the laser beam toward an input mirror 927, which then reflects the laser beam toward a dual-wavelength scanning lens 928, which then guides the laser beam toward a first reflector 929 (see [link to relevant documentation]). Figure 47 (Alternative view).

[0208] Optionally, a beam expander 925 can be used to expand the output beam from the second radiation source 924. The laser beam output from the second radiation source 924 is then reflected by a dichroic mirror 923 and illuminates a feed mirror 926. The feed mirror 926 then reflects the laser beam toward an input mirror 927. The input mirror 927 then reflects the laser beam toward a dual-wavelength scanning lens 928. The dual-wavelength scanning lens 928 then guides the laser beam toward a first reflector 929 (see...). Figure 47 (Alternative view).

[0209] In one embodiment, the output beam from the first radiation source has a wavelength of approximately 515 nanometers (“long wavelength”), and the output beam from the second radiation source has a wavelength of approximately 349 nanometers (“short wavelength”).

[0210] Reflector 929 is configured to reflect a short-wavelength output beam from a first radiation source and a long-wavelength output beam from a second radiation source toward a long-pass reflector 930. Long-pass reflector 930 is configured to reflect the short-wavelength output beam from the first radiation source toward reflector 932, but not the long-wavelength output beam from the second radiation source, thus allowing the long-wavelength output beam from the second radiation source to illuminate reflector 931. Reflectors such as 931 are commonly referred to as long-pass dichroic mirrors. Some examples of long-pass dichroic mirrors are manufactured by Semrock, part number Di03-R405-t1, or by Thorlabs, part number DMLP245. Reflector 932 is configured to reflect the short-wavelength output beam from the first radiation source toward sample 934 at an angle of incidence less than or equal to ten degrees. Reflector 931 is configured to guide the long-wavelength output beam from the second radiation source toward sample 934 at an angle of incidence within ten degrees of Brewster's angle.

[0211] In one example, the dual-wavelength scanning lens 928 is a telecentric scanning lens designed to operate at two wavelengths. Another type of lens that can be used as an alternative to a scanning lens is an achromatic lens.

[0212] In this way, Figure 46 Defect scanner and Figure 47 The optical inspector can illuminate the sample with a first beam of light having a first wavelength at a first incident angle, and with a second beam of light having a second wavelength at a second incident angle. This capability offers numerous advantages. One advantage is... Figure 48 and Figure 49 Example in.

[0213] Figure 48 An example is given of Huygens wavelet-captured photoluminescence with an incident angle of almost 90 degrees. Figure 48An example is shown of the resulting spherical wavelet 971 caused by incident light. The spherical wavelet 971 has an area 970 of emitted radiation that is less than fifty percent of the surface area of ​​the spherical wavelet. The remaining radiation from the spherical wavelet 971 is trapped in the sample by total internal reflection or forward scattering and therefore cannot escape for measurement.

[0214] In comparison, Figure 49 An example is given of Huygens wavelet-captured photoluminescence with an incident angle of zero degrees. Figure 49 An example is shown of the resulting spherical wavelet 975 induced by incident light. The spherical wavelet 975 has an area 974 of emitted radiation that is approximately fifty percent of the surface area of ​​the spherical wavelet. The remaining radiation from the spherical wavelet 975 is trapped in the sample by total internal reflection or forward scattering and therefore cannot escape for measurement.

[0215] Figure 48 and Figure 49 The illustration clearly demonstrates the benefits of irradiating a sample at an incident angle of zero or near zero degrees. Irradiating the sample at an incident angle of zero or near zero degrees maximizes the amount of emissive radiation available for measurement by an optical inspector. Assuming the increase is an increase in area, and not just a linear increase in length, the increase in emissive radiation is exponential. This increase in emissive radiation allows the optical inspector to become many times more sensitive to changes in the sample. This increased sensitivity to changes in the sample improves the performance of the inspector.

[0216] In one example, sample 934 is opaque. In another example, sample 934 is translucent. In yet another example, sample 934 is transparent. For example, sample 934 can be glass, a thin film deposited on a transparent material, sapphire, fused silica, quartz, silicon carbide, and polycarbonate.

[0217] The descanning lens 935 operates to focus the signal from sample 934 onto output mirror 936. In one example, the descanning lens 935 is a telecentric scanning lens that is substantially the same as the scanning lens 928. The descanning lens 935 can be a telecentric lens or an achromatic lens. Utilizing substantially the same lens for scanning lens 928 and descanning lens 935 allows the system to focus light from the top of a transparent sample and the bottom of the same sample into two distinct spots, such as... Figure 26The focal plane is shown. Using two identical scanning lenses 928 and 935 means that any optical aberrations are minimized, making the top spot distinct from the bottom spot and therefore easily separable. As a result, specular signals from the top and bottom surfaces of a transparent surface with a thickness of less than 0.5 mm can be easily separated. Telecentric scanning lenses can be designed to have a large field of view, which makes substrate scanning more efficient. For example, two telecentric scanning lenses can have a field of view greater than one hundred (100) mm. Other examples of descanning lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0218] Feed mirror 937 operates to reflect the signal from output mirror 936 to focusing lens 938. Focusing lens 938 has a focal length based on the characteristics of a focusing lens. At the focal point of focusing lens 938, there will be two spots (assuming the sample is transparent and free of defects), and these spots correspond to signals from the top and bottom surfaces of the sample. In the case of an opaque sample, there will be a single spot from the top surface of the sample. In one example, focusing lens 938 is an achromatic lens with a focal length of 400 (400) mm. Other examples of focusing lenses include, but are not limited to, spherical single lenses, spherical double lenses, triple lenses, or aspherical lenses.

[0219] In one example, when the output mirror 936 is positioned at the midpoint of the rotation range of the output mirror 936, the light reflected by the output mirror 936 illuminates the feed mirror at an angle not exceeding thirty (30) degrees from the normal angle of the output mirror 936 (the second time-varying beam reflector).

[0220] Unpolarized beam splitter 939 is illuminated by light focused by focusing lens 938. When illuminated, unpolarized beam splitter 939 allows a designated portion of the light intensity to pass through to collimating lens 943, and allows the remainder to be reflected back to unpolarized beam splitter 940. The polarization of the transmitted and reflected beams from unpolarized beam splitter 939 remains unchanged. The planes of unpolarized beam splitters 939 and 940 may be the same as or perpendicular to the plane of the sample. When illuminated, unpolarized beam splitter 940 allows a portion of the light intensity to pass through to detector 941, and reflects the remainder in another direction to detector 942. The polarization of the transmitted and reflected beams from unpolarized beam splitter 940 remains unchanged. Detectors 941 and 942 are located approximately one focal length from focusing lens 938.

[0221] Detector 941 is configured to track the focus of the beam. Detector 941 has a low bandwidth of approximately 20 kHz. Detector 941 provides feedback to a sample height control system that precisely maintains the sample at the correct focus. An example of the sample height control system is shown in... Figure 33Examples are provided and described in the accompanying description above.

[0222] Detector 942 is configured to measure the micro-surface profile of the sample. Detector 942 has a large bandwidth of approximately 2 MHz. The output signal from detector 942 is used to measure submicron height changes on the sample.

[0223] Collimating lens 943 is optional and configured to redirect the focused beam from focusing lens 938 into a collimated (i.e., parallel) beam. The collimated beam illuminates half-wave plate 944. Half-wave plate 944 rotates the polarization plane of the collimated beam. The rotated collimated beam then illuminates polarizing beam splitter 945. When illuminated, polarizing beam splitter 945 allows all light polarized in one direction to pass through to detector 947 and reflects all light polarized in the other direction to detector 946. The plane of polarizing beam splitter 945 is the same as the plane of the sample.

[0224] Detector 946 is configured to measure the S-component of a polarized beam. Detector 946 can be a two-cell, four-cell, position-sensitive detector, PIN diode (single element), or other type of Si-based detector.

[0225] Detector 947 is configured to measure the slope of the sample surface and the p-component of the polarized beam. Detector 947 can be a dual-cell, quad-cell, position-sensitive detector, PIN diode (single element), or other type of Si-based detector.

[0226] The first radiation source 920 and the second radiation source 924 can be enabled or disabled by various means. For example, the system can utilize a beam blocker, a radiation source power switch, a shutter, a chopper, an acousto-optic modulator, or an electro-optic modulator. Those skilled in the art will readily understand, upon reading this application, that the desired operation can be achieved using any means for turning each radiation source on or off. Furthermore, those skilled in the art will readily understand, upon reading this application, that the desired operation can be achieved using any means for blocking the output of the radiation sources.

[0227] Figure 47 This is a diagram of an improved optical detector for scattered radiation. The optical detector for scattered radiation includes one or more radiation sources (not shown) that output a short-wavelength source beam 961 and a long-wavelength source beam 962, a time-varying beam reflector 927, a dual-wavelength telecentric scanning lens 928, a reflector 929, a long-pass reflector 930, reflectors 931 and 932, a focusing lens 956, a spatial filter 958, a collimating lens 959, a detector 960, a compound ellipsoidal collector 950, an optional spatial filter 951, and a detector 952.

[0228] In operation, one of one or more radiation sources emits a short-wavelength source beam 961 and a long-wavelength source beam 962 that illuminate the time-varying beam reflector 927. The time-varying beam reflector 927 reflects the short-wavelength source beam 961 and the long-wavelength source beam 962 onto the telecentric scanning lens 928.

[0229] A long-wavelength output beam 962 then illuminates reflector 929 towards long-pass reflector 930. The long-wavelength output beam 962 passes through long-pass reflector 930 and illuminates reflector 931. The long-wavelength output beam 962 is then reflected from reflector 931 towards sample 934. The time-varying beam reflector 927 causes a moving spot (scanning beam) to illuminate sample 934. The time-varying beam reflector 927, the telecentric scanning lens 928, and reflectors 929 to 931 are configured to illuminate sample 934 using a scanning long-wavelength output beam at an incident angle less than or equal to ten degrees from the Brewster angle of sample 934. A focusing lens 956 is configured to be illuminated by scattered radiation 955 (also referred to as "tilted scattered radiation") from sample 934. Scattered radiation 955 radiates from the top surface and bottom surface of sample 934. The focusing lens 956 may be referred to as a light collector. In one example, focusing lens 956 is configured to be oriented along an axis incident on a plane perpendicular to the scanning beam. In another example, focusing lens 956 is a low F-number camera lens. Focusing lens 956 focuses light onto focal plane 957. Spatial filter 958 is located at focal plane 957 and operates to filter out scattered radiation from the bottom surface of transparent sample 934, while allowing scattered radiation from the top surface of sample 934 to pass through to collimating lens 959. Collimating lens 959 is configured along an axis perpendicular to the scanning beam. In one example, spatial filter 958 is a slit-shaped spatial filter to remove scattered light from the bottom surface of sample 934 and other light not from the top surface of sample 934. In another example, collimating lens 959 is a pair of achromatic lenses that shape the scattered radiation into a circular spot illuminating detector 960. In yet another example, detector 960 is a photomultiplier tube.

[0230] In operation, one of one or more radiation sources emits a short-wavelength source beam 961 and a long-wavelength source beam 962 that illuminate the time-varying beam reflector 927. The time-varying beam reflector 927 reflects the short-wavelength source beam 961 and the long-wavelength source beam 962 onto the telecentric scanning lens 928.

[0231] The short-wavelength output beam 961 then illuminates reflector 929 towards long-wavelength reflector 930. Long-wavelength reflector 930 reflects the short-wavelength output beam 961 towards reflector 932. The short-wavelength output beam 961 is then reflected from reflector 932 towards sample 934. The time-varying beam reflector 927 causes a moving spot (scanning beam) to illuminate sample 934. The time-varying beam reflector 927, telecentric scanning lens 928, reflectors 929 to 930, and reflector 932 are configured to illuminate sample 934 with a scanning short-wavelength output beam at an incident angle of less than or equal to zero degrees.

[0232] As mentioned above, Figure 48 and Figure 49 The illustration clearly demonstrates the benefits of irradiating a sample at an incident angle of zero or near zero degrees. Irradiating the sample at an incident angle of zero or near zero degrees maximizes the amount of emissive radiation available for measurement by an optical inspector. Assuming the increase is an increase in area, and not just a linear increase in length, the increase in emissive radiation is exponential. This increase in emissive radiation allows the optical inspector to become many times more sensitive to changes in the sample. This increased sensitivity to changes in the sample improves the performance of the inspector.

[0233] Irradiating sample 934 with a short-wavelength source beam 961 at an incident angle of less than or equal to ten degrees will induce normal scattered radiation 953 that does not illuminate the focusing lens 956. A compound ellipsoidal collector 950 is used to collect at least a portion of the normal scattered radiation 953 and redirect it to the focusing region. The normal scattered radiation 953 is radiation scattered within an angle range between +45 degrees and -45 degrees with respect to the surface normal. The radiation collected by the compound ellipsoidal collector 950 may optionally be filtered by a spatial filter 951. The spatial filter 951 blocks light that does not originate from the focal point of the compound ellipsoidal collector. The collected radiation then illuminates a normal scattered radiation detector 952. In one example, the normal scattered radiation detector 952 is a photomultiplier tube detector. The normal scattered radiation detector 952 measures the intensity of the collected normal scattered radiation, which is used to determine normal dark-field defects. Dark-field defects that scatter normally can be photoluminescent defects in SiC, such as stacking faults, triangular defects, step clusters, basal defects, or other types of photoluminescent defects. Other examples of dark-field defects that scatter normally are pits, scratches, and grains.

[0234] In another example, the scattered radiation optical inspector also includes a processor and a memory. The processor reads the output signals generated by detectors 952 and 960 and stores the light intensity values ​​indicated by the output signals in the memory. The processor can also be used to determine the presence and type of defects. The processor can also be used to generate a defect map across a region of the transparent sample. The processor can also be configured to communicate the defect map to another device or to a monitor.

[0235] The tilt-scattering radiation optical inspector 960 described above collects scattered radiation from the irradiated sample 934 at an angle nearly perpendicular to the incident angle of the scanning beam. If the wavelength of the scanning beam is short enough (less than about 360 nm), the substrate, such as SiC, will become opaque, and the tilt-scattering radiation optical detector 960 and the normal-scattering radiation optical detector 952 will detect only the scattered radiation from the top surface of the sample, such as a transparent sample like SiC. This provides the valuable ability to detect top-side defects in a single scan.

[0236] Figure 37 This is a timing diagram illustrating optional control of the radiation source. The timing diagram illustrates three values ​​along a single time axis. The first value is the angle of the linearly time-varying beam reflector. The second value is the state of the first radiation source (on / off). The third value is the state of the second radiation source (on / off). Figure 37 In one example illustrated, as the angle of the linearly time-varying beam reflector increases, the first radiation source is activated and the second radiation source is deactivated. This configuration causes the first source beam (output from the first radiation source) to illuminate the sample 934 during a first duration. In the next duration, as the angle of the linearly time-varying beam reflector decreases, the first radiation source is deactivated and the second radiation source is activated. This configuration causes the second source beam (output from the second radiation source) to illuminate the sample 934 during a second duration. Figure 37 In the example illustrated, the pattern is repeated with a 50% duty cycle. However, it should be noted that those skilled in the art will readily recognize that the duty cycle can be varied while still achieving the desired results. The results of this scanning method are illustrated in the accompanying figures below.

[0237] Figure 38 This is a diagram illustrating an exemplary wafer scanning operation using time-division multiplexing of radiation sources. In this example, the system scan width is set to one-third of the sample (wafer) width. The first patterned line scan from left to right is performed using a first light source beam (output from a first radiation source). The second patterned line scan from right to left is performed using a second light source beam (output from a second radiation source). As mentioned above... Figure 37 The alternating scans discussed are repeated across the entire sample (wafer).

[0238] Figure 39 This is a diagram illustrating an exemplary single wafer scan performed using time-division multiplexing of radiation sources. The single scan further illustrates the alternating nature of scans performed using a first radiation source and scans performed using a second radiation source. How these scans are organized and processed is illustrated in the following figures.

[0239] Data measured by the various detectors in the system of the present invention can be organized according to the time of measurement. The system can also store which radiation source, either the first or the second, was activated during each measurement. In this way, the system can perform time-division multiplexing to separate the measured data obtained from either the first light source beam (output from the first radiation source) or the second light source beam (output from the second radiation source). Figure 40 An example is illustrated of sample (wafer) scanning using a first light source beam (output from a first radiation source). Alternatively, Figure 41 An example of sample (wafer) scanning using a second light source beam (output from a second radiation source) is illustrated. Therefore, the system is able to separate data obtained using different radiation sources and aggregate data obtained using the same radiation source. Thus, the system can create two separate aggregated datasets (one from illumination from the first radiation source and the other from illumination from the second radiation source) and process each aggregated dataset separately to detect different sets of potential defects.

[0240] This novel system has many benefits.

[0241] First, time-domain multiplexing of radiation sources allows for the separation of scans using different radiation sources without requiring any source separation mechanism. If the system utilizes two radiation sources simultaneously, it will need to be able to separate the resulting radiation from each of the simultaneous sources. The means for separating the resulting radiation from multiple sources require additional cost and design complexity. Furthermore, systems using two or more simultaneous sources are inherently more susceptible to measurement noise caused by simultaneous secondary radiation sources.

[0242] Secondly, the scanning time of this invention is inversely proportional to the rotational speed of the time-varying beam reflector. Other systems require the sample (wafer) itself to rotate while the beam is stationary. This severely limits the scanning rate of other systems because there are practical limitations on how fast the sample (wafer) can rotate, while the rotational speed of the time-varying beam reflector is several orders of magnitude greater than the fastest possible rotational speed of the sample (wafer). For example, time-varying beam reflectors (such as galvanometers or polygons) can rotate at speeds up to 2 kHz, while the sample (wafer) cannot rotate at speeds higher than 100 Hz without risking damage to the sample (wafer).

[0243] Figure 50This is a flowchart 980 describing the operation of a dual-beam separation defect scanner and optical inspector with independent incident angles. In step 981, a first radiation source is activated to output a first beam with a first wavelength. In step 982, a second radiation source is activated to output a second beam with a second wavelength. In step 983, the first beam is reflected along a path such that the first beam illuminates the sample at an incident angle less than or equal to ten degrees from the Brewster angle. In step 984, the second beam is reflected along a second path such that the second beam illuminates the sample at an incident angle less than or equal to ten degrees. In step 985, the oblique scattered radiation obtained by the first beam illuminating the sample is measured. In step 986, the normal scattered radiation obtained by the second beam illuminating the sample is measured. In step 987, a first set of data including all measured reflected radiation measurements is aggregated. In step 988, a second set of data including all measured scattered radiation measurements is aggregated. In step 989, the first set of data is processed to determine whether a defect exists on the sample. In step 990, the second set of data is processed to determine whether there are defects on the sample.

[0244] Although certain specific embodiments have been described above for guidance purposes, the teachings of this patent document are of general applicability and are not limited to the specific embodiments described above. Therefore, various modifications, adaptations, and combinations of the features of the described embodiments can be practiced without departing from the scope of the invention as set forth in the claims.

Claims

1. A method, the method comprising: (a) Enable the first radiation source configured to output the first beam; (b) Enable the second radiation source configured to output the second beam; (c) The first beam is directed to illuminate the sample at an incident angle that differs from Brewster's angle by ten degrees or less, wherein the first beam passes through a scanning lens before illuminating the sample; (d) The second beam is directed to illuminate the sample at an incident angle of ten degrees or less, wherein the second beam passes through the scanning lens before illuminating the sample, and wherein the first beam and the second beam do not pass through the second scanning lens before illuminating the sample; (e) Measure the scattered radiation data obtained by irradiating the sample with the second beam; (f) Measuring the reflected radiation obtained by irradiating the sample with the first beam; and (g) The presence of a defect is determined at least in part based on the measurement in (e) or (f).

2. The method of claim 1, wherein the first beam has a first wavelength, wherein the second beam has a second wavelength, wherein the first wavelength is different from the second wavelength, wherein the first radiation source is deactivated before the measurement in (e), and wherein the second radiation source is deactivated before the measurement in (f).

3. The method of claim 1, wherein the guiding in (c) further comprises: (c1) Use a first time-varying beam reflector to reflect the first source beam.

4. The method of claim 1, wherein the guidance in (c) further comprises: (c1) Use a dichroic mirror to reflect the second beam.

5. The method of claim 1, wherein the guidance in (c) further comprises: (c1) Focus the first beam onto the surface of the sample.

6. The method of claim 1, wherein the scattered radiation data obtained by irradiating the sample with the second beam is guided at least in part by a composite ellipsoidal collector.

7. The method of claim 1, wherein the measurement in (e) further comprises: (e1) Read the output of the detector that is irradiated by scattered radiation in response to the second beam irradiating the sample.

8. The method according to claim 7, wherein the detector is a photomultiplier tube.

9. The method of claim 1, wherein the determination in (g) further comprises: (g1) Store in memory a first value representing the measured scattered radiation or a second value representing the reflected radiation; as well as (g2) Process the first value or the second value, wherein the processing is performed by one or more processing circuits.

10. A method, the method comprising: (a) Enable the first radiation source configured to output the first beam; (b) Enable the second radiation source configured to output the second beam; (c) The first beam is directed to illuminate the sample at an incident angle that differs from Brewster's angle by ten degrees or less, wherein the first beam passes through a scanning lens before illuminating the sample; (d) The second beam is directed to illuminate the sample at an incident angle of ten degrees or less, wherein the second beam passes through the scanning lens before illuminating the sample, and wherein the first beam and the second beam do not pass through the second scanning lens before illuminating the sample; (e) Measure the scattered radiation data obtained by irradiating the sample with the second beam; (f) Measuring the scattered radiation obtained by irradiating the sample with the first beam; and (g) The presence of a defect is determined at least in part based on the measurement in (e) or (f).

11. An apparatus comprising: A first radiation source, configured to output a first beam; A second radiation source, configured to output a second beam; A dichroic mirror, configured to guide the first beam and the second beam toward a time-varying beam reflector, wherein the time-varying beam reflector is configured to reflect the first beam and the second beam toward a scanning lens; A first reflector, configured to reflect the first beam and the second beam; A second reflector, configured to reflect the second light beam, wherein the second reflector is configured not to reflect the first light beam; A third reflector, the third reflector being configured to reflect the first beam; and A fourth reflector, which is configured to reflect the second beam.

12. The apparatus of claim 11, wherein the third reflector is configured to reflect the first beam toward the sample at an incident angle less than or equal to ten degrees different from Brewster's angle.

13. The apparatus of claim 11, wherein the fourth reflector is configured to reflect the second beam toward the sample at an incident angle of less than or equal to ten degrees.

14. The apparatus of claim 11, wherein the first reflector is a broadband reflector.

15. The apparatus of claim 11, wherein the second reflector is a long-pass reflector.

16. The apparatus of claim 11, further comprising: A composite ellipsoidal collector configured to collect scattered radiation obtained by irradiating a sample with the second beam.

17. The apparatus of claim 16, further comprising: A spatial filter configured to filter at least a portion of the scattered radiation collected by the composite ellipsoidal collector.

18. The apparatus of claim 16, further comprising: A detector configured to measure the scattered radiation collected by the composite ellipsoid collector.

19. An apparatus comprising: A first radiation source, configured to output a first beam; A second radiation source, configured to output a second beam; A dichroic mirror, configured to guide the first beam and the second beam toward a time-varying beam reflector, wherein the time-varying beam reflector is configured to reflect the first beam and the second beam toward a scanning lens; A first component is used to guide the first beam toward the sample at an incident angle within ten degrees of Brewster's angle. and The second component is used to guide the second beam toward the sample at an incident angle of less than or equal to ten degrees.

20. The apparatus of claim 19, wherein the first component comprises: A first reflector, configured to reflect the first beam and the second beam; and A second reflector, configured to reflect the first beam toward the sample; and The second component includes: A third reflector, the third reflector being configured to reflect the second beam, wherein the third reflector is configured not to reflect the first beam; and A fourth reflector, configured to reflect the first beam toward the sample.

Images