Inspection Equipment

Abstract

An inspection apparatus for inspecting a sample having a surface formed by a transparent film and an opaque substance, the apparatus comprising: a first optical unit that condenses first reflected light obtained as a result of illumination light, emitted from a light source, being reflected on the sample; a second optical unit that condenses second reflected light obtained as a result of the illumination light being emitted toward and reflected by a reflection mirror; an optical interference unit that causes the first reflected light and the second reflected light to interfere with each other to obtain interference light; a plurality of interference light sensors that detect the reflected light intensity of the interference light; and a signal-processing device that processes the amount of light detected by the interference light sensors, wherein the signal-processing device determines whether a given coordinate point of the sample is in the transparent film or in the opaque substance, on the basis of the amount of light detected by the interference light sensors and the refractive indexes of the transparent film and the opaque substance, and measures the surface height or the thickness of the sample at the coordinate point.

Description

[Technical Field]

[0001] The present invention relates to an inspection device. [Background technology]

[0002] In manufacturing lines for semiconductor substrates, thin-film substrates, and the like, the surfaces of semiconductor substrates, thin-film substrates, and the like are inspected to improve product yield. Nanometer-order smoothness is required for the surfaces of semiconductor substrates and thin-film substrates. When inspection light passes through the substrate surface, there are known techniques for high-speed measurement of surface irregularities on substrates without patterns, such as differential interference contrast. However, when a transparent film is formed on the substrate surface, differential interference contrast cannot determine whether the phase difference occurs on or within the film. Known inspection devices suitable for substrates with such transparent films irradiate the sample surface with multiple light beams of different wavelengths while changing the working distance, and measure the reflected light before and after changing the working distance to measure the film thickness and surface height of the transparent film (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-6242 Summary of the Invention [Problem to be solved by the invention]

[0004] Typically, semiconductor substrates are wafers with a diameter of 300 mm, and it is required to inspect the entire surface of this wafer in about one minute. However, the technology of Patent Document 1, which measures while changing the working distance, requires re-measurement at the same coordinates by changing the distance between the optical system and the sample, making it difficult to inspect the sample at high speed.

[0005] Furthermore, semiconductor substrates often contain circuit patterns using copper wiring within transparent films. The refractive index of copper varies with wavelength. If n is the refractive index and k is the attenuation coefficient, then the difference is large: (n, k) = (1.15, 2.47) for blue light (470 nm) and (n, k) = (0.35, 3) for red light (600 nm). In this case, the ratio of the intensity of the reflected light from the pattern within the transparent film varies greatly between blue and red light. However, silicon dioxide, a common transparent film material, has an (n, k) ≒ (1.46, 0) for both blue and red light.

[0006] In contrast, the technology of Patent Document 1, as described in paragraph 0048, treats the AC component of the surface brightness of the reflected light from the transparent film on the surface and the AC component of the surface brightness of the reflected light from the sample being observed as fixed values ​​independent of wavelength. When copper wiring is not present in the film, the measured reflected light is generally considered to be reflected light from silicon underneath the film. Silicon has a refractive index of (n, k) = (4.4, 0.13) for blue light and (n, k) = (3.95, 0.025) for red light, which is smaller than that of copper. However, if the wavelength of the inspection light is shortened to, for example, around 405 nm to improve resolution, the refractive index of silicon changes significantly, to (n, k) = (5.42, 0.31). Therefore, even if the target is silicon, it is difficult to treat the ratio of the AC component of the reflection intensity as a constant. When inspecting semiconductor substrates, the light reflected from beneath the film is mainly light reflected by copper or silicon, and the structure beneath the film is often unknown at the inspection stage. This also makes it difficult to generally apply the technology of Patent Document 1 to the inspection of actual electronic product substrates.

[0007] Furthermore, since the silicon dioxide that makes up the transparent film has high transmittance and the amount of reflected light generated on the surface is weak, there is little change in the amount of light that accompanies changes in the film thickness or surface height of the transparent film, making it difficult to measure the surface height with high accuracy.

[0008] Additionally, when inspecting a semiconductor substrate, it is necessary to inspect not only the film thickness and surface height of the transparent film but also foreign matter, etc., but the technology of Patent Document 1 makes it difficult to inspect foreign matter, etc.

[0009] An object of the present invention is to provide an inspection device capable of measuring the film thickness or surface height of a transparent film on a substrate surface with high precision and at high speed in a manufacturing line for semiconductor substrates, thin film substrates, etc. [Means for solving the problem]

[0010] In order to achieve the above object, the present invention provides an inspection device for inspecting a sample whose surface is formed of a transparent film and an opaque material through which light passes, the inspection device comprising: a light source; a first optical unit that irradiates the sample with illumination light emitted from the light source and collects first reflected light reflected by the sample; a second optical unit that irradiates a reflecting mirror with the illumination light and collects second reflected light reflected by the reflecting mirror; an interference optical unit that causes the first reflected light and the second reflected light to interfere with each other to obtain interference light; a plurality of interference light sensors that detect the reflected light intensity of a predetermined polarization component of the interference light; and a signal processing device that processes the amount of light detected by the interference light sensors, the signal processing device being configured to process the amount of light detected by the interference light sensors for a plurality of illumination lights having different wavelengths. 、 Refractive index of the transparent film and the opaque material , and wavelength and calculates, for each wavelength of the illumination light, an estimated value of the surface height or film thickness of the sample when any coordinate of the sample is assumed to be the transparent film or the opaque material, and, based on the estimated value calculated for each wavelength, identifies whether the coordinate is the transparent film or the opaque material, and measures the surface height or film thickness of the sample at the coordinate. [Effects of the Invention]

[0011] According to the present invention, in a manufacturing line for semiconductor substrates, thin film substrates, etc., it is possible to measure the film thickness or surface height of a transparent film on the surface of a substrate with high precision and at high speed. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 is a schematic diagram illustrating an example of the configuration of an inspection device according to a first embodiment of the present invention. [Figure 2] Schematic diagram of a semiconductor wafer, a typical inspection target [Figure 3] A diagram showing a beam spot produced by illumination light from a light source provided in the inspection device of Figure 1. [Figure 4] Schematic diagram of the optical scanning unit installed in the inspection device in Figure 1 [Figure 5] Schematic diagram of the light receiving surface of the interference light sensor installed in the inspection device in Figure 1 [Figure 6] Schematic diagram of the spatial filter unit installed in the inspection device in Figure 1 [Figure 7] Schematic diagram showing an example of a pattern formed on a sample. [Figure 8] Arrow cross-sectional view of the sample taken along section line VIII in Figure 7 [Figure 9] A graph showing the relationship between the amount of light detected by each sensor and the surface height of the pattern when illumination light with a wavelength of 405 nm is incident on the surface of the pattern. [Figure 10] A graph showing the relationship between the estimated value and actual value of the surface height of the pattern calculated based on the light intensity measured in Figure 9. [Figure 11] A graph showing the relationship between the amount of light detected by each sensor and the surface height of the pattern when illumination light with a wavelength of 660 nm is incident on the surface of the pattern. [Figure 12] A graph showing the relationship between the estimated value of the surface height of the pattern calculated based on the light intensity in Figure 11 obtained by measurement and the actual value. [Figure 13] A graph showing the relationship between the amount of light detected by each sensor and the surface height of a transparent film when illumination light with a wavelength of 405 nm is incident on the surface of a transparent film with a certain film thickness. [Figure 14] A graph showing the estimated thickness of the transparent film calculated based on the light intensity measured in Figure 13. [Figure 15] A graph showing the estimated surface height of the transparent film calculated based on the light intensity measured in Figure 13. [Figure 16] A graph showing the relationship between the amount of light detected by each sensor and the surface height of a transparent film when illumination light with a wavelength of 660 nm is incident on the surface of a transparent film with a constant film thickness. [Figure 17]A graph showing the estimated thickness of the transparent film calculated based on the light intensity measured in Figure 16. [Figure 18] A graph showing the estimated surface height of the transparent film calculated based on the light intensity measured in Figure 16. [Figure 19] A graph showing the relationship between the amount of light detected by each sensor and the surface height of a transparent film when illumination light is incident on the surface of a transparent film with varying film thickness. [Figure 20] A graph showing the estimated thickness of the transparent film calculated based on the light intensity measured in Figure 19. [Figure 21] A graph showing the estimated surface height of the transparent film calculated based on the light intensity measured in Figure 19. [Figure 22] A graph showing the profile of all detected light intensity obtained by each sensor when the thickness of the transparent film is changed. [Figure 23] A graph showing the estimated thickness of the transparent film calculated based on the light intensity measured in Figure 22. [Figure 24] A graph showing the estimated surface height of the transparent film calculated based on the light intensity measured in Figure 22. [Figure 25] 1 is a flowchart showing an example of a procedure for calculating the surface height and film thickness of an arbitrary measurement portion of a sample by processing interference data in a signal processing device provided in the inspection device of FIG. [Figure 26] 1. A flowchart showing another example of the procedure for calculating the surface height and film thickness of an arbitrary measurement portion of a sample by processing interference data in a signal processing device provided in the inspection device of FIG. [Figure 27] FIG. 2 is a block diagram showing an example of a functional block related to the defect inspection of a sample by a signal processing device provided in the inspection device of FIG. 1. [Figure 28] A diagram showing an example of an output screen [Figure 29] FIG. 10 is a schematic diagram illustrating a configuration example of an inspection device according to a second embodiment of the present invention. [Figure 30] Schematic diagram of an interference light sensor provided in an inspection device according to a second embodiment of the present invention. [Figure 31] Schematic diagram of the photodetector array in the sensor shown in Figure 30 [Figure 32]FIG. 10 is a schematic diagram illustrating a configuration example of an inspection device according to a third embodiment of the present invention. [Figure 33] Diagram showing the reflectivity of silicon dioxide [Figure 34] 33 is a diagram showing a scanning trajectory of a stage provided in the inspection device shown in FIG. 32. [Figure 35] FIG. 33 is a diagram showing another example of a scanning trajectory by the stage provided in the inspection apparatus shown in FIG. 32. [Figure 36] Schematic diagram of the holey mirror of the dark-field optical unit provided in the inspection device shown in Figure 32. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0014] (overview) The inspection device described in the following embodiments as an application of the present invention is used to inspect the surface of a sample (e.g., a semiconductor silicon wafer) during the manufacturing process of, for example, semiconductors, etc. The inspection device according to each embodiment is suitable for quickly measuring the surface height of a sample and the film thickness of a transparent film (including the height of the interface of the transparent film), detecting minute defects such as foreign matter, and acquiring data on the number, position, size, and type of defects.

[0015] A semiconductor silicon wafer, a typical example of a sample, has a structure in which a transparent film or a pattern of a microstructure is formed on the surface of a silicon substrate. The transparent film is made of, for example, silicon dioxide, and has the property of transmitting the illumination light used in the inspection device of the present invention (transparent to the illumination light). The pattern is made of, for example, copper, and has the property of reflecting the illumination light used in the inspection device of the present invention (opaque to the illumination light). The pattern may be located inside the transparent film, or may be exposed on the surface of the transparent film. In this way, the surface of a semiconductor silicon wafer is formed of either a transparent film or an opaque material (pattern). Silicon, which is the material of the substrate, is also an opaque material and reflects the illumination light.

[0016] The inspection device of the present invention inspects a sample whose surface is formed of a transparent film through which illumination light passes and an opaque material that reflects illumination light, and measures the surface height of the sample or the film thickness of the transparent film, etc. Essential components of the inspection device are a light source, a first optical unit, a second optical unit, an interference optical unit, multiple interference light sensors, and a signal processing device.

[0017] The light source is a unit that emits illumination light, and in each embodiment described later, the light source 30 (FIGS. 1, 29, and 32) corresponds to this.

[0018] The first optical unit is a unit that irradiates the sample with illumination light emitted from the light source and collects the first reflected light reflected by the sample, and in each embodiment described below, at least the objective lens 43 (FIGS. 1, 29, and 32) corresponds to this. There are two types of illumination for the sample: epi-illumination, in which illumination light is incident perpendicularly on the surface of the sample, and oblique illumination, in which illumination light is incident obliquely on the surface of the sample. Epi-illumination is possible in all of the examples shown in FIG. 1, 29, and 32. In the example of FIG. 29, it is also possible to switch the optical path of the illumination light to obliquely illuminate the sample with S-polarized illumination light.

[0019] The second optical unit is an optical unit that illuminates the reflective mirror with illumination light and collects the second reflected light reflected by the reflective mirror. In the embodiments described later, the second optical unit and the reflective mirror correspond to at least the objective lens 44 and the reflective mirror 45 (FIGS. 1, 29, and 32), respectively.

[0020] The interference optical unit is an optical unit that causes the first reflected light and the second reflected light to interfere with each other to obtain interference light, and in each embodiment described below, at least the polarizing beam splitter 41 (Figures 1, 29, and 32) corresponds to this.

[0021] The multiple interference light sensors are sensors that detect the reflected light intensity of predetermined polarization components of the interference light, and in each embodiment described below, at least interference light sensors 55A-55D (FIGS. 1 and 32) and interference light sensors 56A-56D (FIG. 29) correspond to these. The interference light sensors detect reflected light of illumination light for each polarization component and for each light of different wavelengths. In each embodiment, four interference light sensors 55A-55D or 56A-56D detect light whose polarization directions are shifted by 45° each.

[0022] A polarizing filter or a polarizing beam splitter can be used as a means for extracting specific polarization components of reflected light. In each of the embodiments described below, a configuration will be described in which a half beam splitter 51 or a polarizing beam splitter 52 (FIGS. 1, 29, and 32) is used to separate reflected light according to its polarization components, and the separated reflected light is detected by an interference light sensor. Furthermore, the configuration for detecting reflected light for each wavelength can be a configuration in which multiple illumination lights with different wavelengths are simultaneously emitted and their reflected light is detected, or a configuration in which reflected light is separated according to wavelength. In the first embodiment (FIG. 1) and the third embodiment (FIG. 32), a light source that emits multiple monochromatic lights with different wavelengths is used as illumination light, and reflected light for each wavelength is individually detected by multiple light-receiving surfaces provided in each interference light sensor. In the second embodiment (FIG. 29), a configuration will be described in which each interference light sensor separates reflected light for each wavelength using a prism and individually detects the reflected light on the light-receiving surface.

[0023] The signal processing device is a computer that processes the amount of light detected by the interference light sensor, and corresponds to signal processing device 7 (FIGS. 1, 29, and 32) in each embodiment described below. The signal processing device can be configured as a single computer, or can be configured as multiple computers that share functions. This signal processing device identifies whether an arbitrary coordinate on the sample (for convenience, referred to as coordinate C) represents a transparent film or an opaque material based on the amount of light detected by the interference light sensor and the refractive index of the transparent film and the opaque material, and measures the surface height or film thickness of the sample at coordinate C by calculation.

[0024] The surface of the sample to be inspected is made of a limited number of materials, and the materials that make up coordinate C are limited to a few candidates. For example, if coordinate C is a position where a pattern is exposed, the illumination light reflects off the surface of the pattern and returns to the first optical unit. In this case, the material is, for example, copper. If coordinate C is a position where no pattern is present, the illumination light enters a transparent film, reflects off the substrate below the film, and returns to the first optical unit. In this case, the material is, for example, silicon dioxide and silicon. Also, if coordinate C is a position where a pattern is present in the film, the illumination light enters a transparent film, reflects off the pattern in the film, and returns to the first optical unit. In this case, the material is, for example, silicon dioxide and copper. When calculating the surface height of a sample from the detected light intensity, if the sample surface at coordinate C is an opaque material, a unique surface height can be calculated under the same conditions. However, if the sample surface at coordinate C is a transparent film, multiple surface heights can be calculated even under the same conditions (described below).

[0025] Therefore, in the inspection device of the present invention, the signal processing device calculates one or more estimated values ​​of the surface height of the sample or the thickness of the transparent film for an arbitrary coordinate C based on the detected light amount for each wavelength, and compares the one or more estimated values ​​for each wavelength. Through this comparison, the signal processing device selects and outputs one of the one or more estimated values ​​calculated for each wavelength as the measured value of the surface height or thickness of the sample at coordinate C. For example, the signal processing device stores at least one refractive index for each candidate material for the transparent film (e.g., silicon dioxide) and one refractive index for each candidate material for the opaque material (e.g., copper, silicon). The signal processing device distinguishes between cases where the illumination light incident on coordinate C is reflected directly by the opaque material without passing through a transparent film, and cases where the illumination light is reflected by the opaque material via a transparent film, and assumes one or two candidate materials for coordinate C. Then, one or two estimated values ​​of the surface height or thickness of the sample are calculated for each wavelength based on the refractive index of the assumed candidate material and the detected light amount for each wavelength obtained by each interference light sensor. The estimated values ​​calculated for each wavelength are compared, and one of the one or two estimated values ​​calculated for each wavelength is selected and output as the measured value of the surface height or film thickness of the sample. In addition, as the signal processing device determines the measured value in this way, it can identify the candidate material related to the estimated value adopted as the measured value as the material of coordinate C.

[0026] The above-mentioned comparison process by the signal processing device can employ cross-validation, utilizing the detected light intensity obtained for each wavelength. For example, since the estimated value calculated based on the detected light intensity for each wavelength is calculated equally for other wavelengths for the same candidate material, the amount of light at other wavelengths that should be detected is calculated and compared with the detected light intensity. In this case, the estimated value that least deviates from the calculated values ​​obtained for the other wavelengths can be determined as the measured value. For example, if the surface height h is calculated based on the detected light intensity I1 of illumination light with wavelength λ1 under the assumption that a pattern exists in the film at coordinate C, the amount of light λ2 that should be detected is calculated so that the surface height h can be calculated with illumination light with wavelength λ2 under the same assumption. If the difference between the calculated light intensity λ2 and the detected light intensity λ1 is minimal or zero (or less than a predetermined tolerance for identity determination), the estimated value can be considered the measured value for coordinate C. At the same time, the candidate material for the assumption related to the estimated value can be identified as the actual material for coordinate C. A specific example of this will be described later using FIG. 25.

[0027] Furthermore, the above-mentioned comparison process by the signal processing device is not limited to the above example. As another example, a method can be adopted in which estimated values ​​calculated based on different wavelengths for the same candidate material are compared, and the candidate material for which the estimated values ​​for all wavelengths match or the difference is within the above-mentioned tolerance is identified as the material of coordinate C, and the estimated value for that material is determined to be the measured value. A specific example of this method will be described later using Figure 26.

[0028] The signal processing device can also inspect the sample for defects based on the measured surface height. For example, the signal processing device measures the surface height of the sample for each predetermined area, determines whether the measurement results are within a predetermined range, and outputs any areas where the measurement results are outside the predetermined range as defects. A specific example of this will be described later using FIG. 27.

[0029] The inspection device of each embodiment includes an optical path branching unit, a spatial filter, and a dark-field optical sensor, and extracts dark-field light (light scattered from the sample) from the interference light used to measure the surface height of the sample, allowing simultaneous inspection of defects such as foreign matter on the sample surface. Specifically, the optical path branching unit separates dark-field light from the reflected light collected by the first optical unit and the second optical unit. The spatial filter removes diffracted light from the dark-field light separated by the optical path branching unit, and the dark-field light transmitted through the spatial filter is detected by the dark-field optical sensor. The signal processing device detects defects on the sample based on the output of the dark-field optical sensor. In each embodiment described below, the perforated mirror 60, spatial filter unit 61, and dark-field optical sensor 63 (FIGS. 1, 29, 32, and 36) correspond to the optical path branching unit, spatial filter, and dark-field optical sensor, respectively.

[0030] Several specific embodiments of the inspection apparatus outlined above will be described below.

[0031] (First embodiment) 1. Inspection equipment FIG. 1 is a schematic diagram showing an example of the configuration of an inspection device 100 according to a first embodiment of the present invention. The inspection device 100 shown in FIG. 1 is an inspection device that inspects a sample 1, measures the height of the surface of this sample 1, and simultaneously inspects the surface of the sample 1 for defects such as minute foreign objects and dents. If there is a transparent film on the surface of the sample 1, it is also possible to measure the film thickness and interface height of the transparent film. A typical example of the sample 1 is a disk-shaped semiconductor silicon wafer with a flat surface on which a pattern is formed.

[0032] The inspection apparatus 100 includes a stage 2, an illumination optical unit 3, an illumination / detection optical unit 4, an interference optical unit 5, a dark-field optical unit 6, a signal processing device 7, a control device 81, a user interface 82, a monitor 83, and a storage device 84. The storage device 84 stores processing parameters applied when the signal processing device processes the detection signal, and the results of processing by the signal processing device.

[0033] 2. Stage 2 The stage 2 includes a sample stage 2a and a sample drive stage 2b. The sample stage 2a supports the sample 1. The sample drive stage 2b drives the sample stage 2a to change the relative position between the sample 1 and the illumination and detection optical unit 4. Although not shown in detail, the stage 2b includes an XY stage and a Z stage. The sample stage 2a is supported on the XY stage via the Z stage. The Z stage adjusts the height of the surface of the sample 1. The XY stage is driven by a control signal from the control device 81 so that the desired inspection area of ​​the sample 1 is within the field of view of the illumination and detection optical unit 4. After detection of one inspection area by the interference optical unit 5 and the darkfield optical unit 6 is completed, the control device 81 controls the sample drive stage 2b so that the next inspection area is within the field of view of the illumination and detection optical unit 4. The sample drive stage 2b performs a step-and-repeat operation. In other words, the sample driving stage 2b moves the desired inspection point on the sample 1 to the illumination position where illumination light is irradiated by the illumination / detection optical unit 4, stops it temporarily, and after completing the image data of the inspection point, moves the next inspection point to the illumination position, repeating this operation.

[0034] FIG. 2 is a schematic diagram of a semiconductor wafer, a typical object to be inspected. Multiple chips are formed in a matrix on the surface of the semiconductor wafer sample 1. The figure shows a situation in which the field of view 1ijk is positioned on chip 1ij among multiple chips formed on the surface of the sample 1 as the sample driving stage 2b moves the sample 1 relative to the field of view 1ijk of the illumination and detection optical unit 4 during inspection of the sample 1. In the inspection apparatus 100 of this embodiment, an arbitrary inspection area on the surface of the sample 1 is placed in the field of view 1ijk of the illumination and detection optical unit 4 to acquire image data. After acquiring image data of that inspection area, the sample 1 is moved to acquire image data of the next inspection area. In this way, the sample 1 is scanned using a step-and-repeat method.

[0035] 3. Illumination optical unit 3 1 includes a group of optical elements and irradiates a sample 1 placed on a sample stage 2a with desired illumination light. The illumination optical unit 3 includes a light source 30, an illumination shaping unit 31, a half beam splitter 32, an optical scanning unit 33, relay lenses 34a and 34b, and relay lenses 35a and 35b.

[0036] 3-1.Light source 30 The light source 30 is a unit that emits a laser beam as illumination light, and in this embodiment, a multi-line laser light source that emits multiple monochromatic beams is used. The light source 30 in this embodiment simultaneously emits blue (wavelength 405 nm), green (532 nm), and red (660 nm) light beams with long coherence lengths, arranged in ascending order of wavelength. FIG. 3 is a diagram showing beam spots formed by the illumination light emitted by the light source 30. The beam spots 40r, 40g, and 40b formed by the illumination light from the light source 30 each have a Gaussian profile with a minor axis of several tens to several hundreds of micrometers and a major axis of several millimeters to several tens of millimeters. The beam spot 40r is red light (wavelength 660 nm), the beam spot 40g is green light (wavelength 532 nm), and the beam spot 40b is blue light (wavelength 405 nm). They are formed close to each other and aligned in the minor axis direction in the field of view 1ijk of the illumination and detection optical unit 4.

[0037] 3-2. Lighting shaping unit 31 The illumination shaping unit 31 includes anamorphic prisms 31a and 31b. The three-color illumination light from the light source 30 is expanded in specific directions by the anamorphic prisms 31a and 31b. The light beam shaped by the illumination shaping unit 31 has an elliptical shape (Fig. 3) with a large aspect ratio between the minor axis and the major axis in a cross section perpendicular to the optical axis.

[0038] 3-3.Half Beam Splitter 32 The half beam splitter 32 guides the light shaped by the illumination shaping unit 31 to the optical scanning unit 33, and also guides the light that is collected by the illumination / detection optical unit and guided via the optical scanning unit 33 to relay lenses 35a and 35b.

[0039] 3-4. Optical scanning unit 33 FIG. 4 is a schematic diagram of the optical scanning unit 33. The optical scanning unit 33 is typically a MEMS optical scanner, and an electrostatic type is shown here as an example. The optical scanning unit 33 includes a reflecting surface 33a and driving electrodes 33b and 33c. The reflecting surface 33a is supported by a rotation axis 33d and tilts around the rotation axis 33d. By applying a voltage to the driving electrodes 33b and 33c, the angle of the reflecting surface 33a changes around the rotation axis 33d, thereby changing the reflection direction of the light guided from the half beam splitter 32 and performing scanning. Note that in this embodiment, an electrostatic MEMS mirror is used for the reflecting surface 33a. However, an electromagnetic type may be used instead of the electrostatic type, or a galvanometer mirror may be used instead of the MEMS mirror.

[0040] 3-5. Relay lenses 34a and 34b The relay lenses 34a and 34b relay the light guided from the reflecting surface 33a of the optical scanning unit 33 to the illumination and detection optical unit 4, and form an image of that light on the pupil plane of the objective lenses 43 and 44 (described later) of the illumination and detection optical unit 4. In other words, the relay lenses 34a and 34b are adjusted so that the reflecting surface 33a is positioned conjugate with the pupil planes of the objective lenses 43 and 44. This allows the angle of the reflecting surface 33a to be changed so that the field of view 1ijk can be scanned with beam spots 40r, 40g, and 40b. In addition, the light collected by the illumination and detection optical unit 4 is relayed by the relay lenses 34a and 34b and returns to the reflecting surface 33a of the optical scanning unit 33.

[0041] 3-6. Relay lenses 35a and 35b The relay lenses 35 a and 35 b relay the light collected by the illumination and detection optical unit 4 and guided via the optical scanning unit 33 and half beam splitter 32 to the interference optical unit 5 and dark-field optical unit 6 .

[0042] 4. Illumination and detection optical unit 4 The illumination and detection optical unit 4 includes a quarter-wave plate 42 , a polarizing beam splitter 41 , objective lenses 43 and 44 , and a reflecting mirror 45 .

[0043] Light guided from the illumination optical unit 3 via relay lenses 34a and 34b is converted from linearly polarized light to circularly polarized light by a quarter-wave plate 42, the fast axis or slow axis of which is rotated by 45°. This circularly polarized light is then split into two beams by a polarizing beam splitter 41 according to the difference in polarization direction, and the split beams enter objective lenses 43 and 44, respectively.

[0044] The light incident on the objective lens 43 forms beam spots (beam spots 40r, 40g, 40b) on the surface of the sample 1. The reflected light generated at the beam spots is collected by the objective lens 43 and returns to the polarizing beam splitter 41.

[0045] On the other hand, the light incident on the objective lens 44 similarly forms a beam spot on the surface of the reflecting mirror 45. The reflected light generated at this beam spot is collected by the objective lens 44 and returns to the polarizing beam splitter 41.

[0046] The reflected light reflected by the sample 1 and the reflecting mirror 45 and returned to the polarizing beam splitter 41 has its polarization direction converted by the quarter-wave plate 42 and returns to the optical scanning unit 33 via the relay lenses 34a and 34b. The light returned to the optical scanning unit 33 is reflected by the reflecting surface 33a and relayed to the interference optical unit 5 and the dark-field optical unit 6 via the relay lenses 34a and 34b.

[0047] 5. Interference Optical Unit 5 The interference optical unit 5 is a unit that performs interference measurement of the light collected by the illumination and detection optical unit 4. The interference optical unit 5 includes an imaging lens 50, a half beam splitter 51, a polarizing beam splitter 52, a quarter-wave plate 53, a polarizing beam splitter 54, and interference light sensors 55A, 55B, 55C, and 55D.

[0048] The light reflected by the sample 1 and the reflecting mirror 45 and transmitted via the relay lenses 35a and 35b passes through an imaging lens 50 and enters a half beam splitter 51, where it is split into two beams.

[0049] One of the beams split by the half beam splitter 51 is guided to the polarizing beam splitter 52 and further split into two beams according to the polarization direction. The two beams split by the polarizing beam splitter 52 form interference images of the beam spots formed on the sample 1 and the reflecting mirror 45 on the light-receiving surfaces of the interference light sensors 55A and 55B, respectively.

[0050] The other light beam split by half beam splitter 51 has its polarization direction changed by quarter-wave plate 53, whose fast axis or slow axis is rotated by 45°, and is then guided to polarizing beam splitter 54, where it is further split into two beams according to its polarization direction. The two beams split by polarizing beam splitter 54 form interference images of the beam spots formed on sample 1 and reflecting mirror 45 on the light-receiving surfaces of interference light sensors 55C and 55D, respectively.

[0051] FIG. 5 is a schematic diagram of the light-receiving surfaces of the interference light sensors 55A-55D. Each of the interference light sensors 55A-55D has three light-receiving surfaces 55r, 55g, and 55b. Reflected light from beam spots 40r, 40g, and 40b is incident on the light-receiving surfaces 55r, 55g, and 55b, respectively, and images of the beam spots 40r, 40g, and 40b are formed. The beam spots 40r, 40g, and 40b scan the surface of the sample 1 or the reflecting mirror 45 by changing the angle of the reflecting surface 33a. However, as described above, the reflected light is reflected by the same reflecting surface 33a and directed to the interference optical unit 5. Therefore, even when the beam spots 40r, 40g, and 40b are scanned, the images of the beam spots 40r, 40g, and 40b are always formed on the corresponding light-receiving surfaces 55r, 55g, and 55b.

[0052] The light-receiving surfaces 55r, 55g, and 55b operate as individual TDI sensors, each synchronizing its own output with the beam spot scanning by the reflecting surface 33a. Each of the light-receiving surfaces 55r, 55g, and 55b has a configuration in which n lines of linearly arranged line sensors (a group of light-receiving elements) are arranged in the S1 direction in FIG. 5. Each time the beam spots 40r, 40g, and 40b formed on the sample 1 move a distance equivalent to one pixel size of the interference light sensors 55A-55D on the sample surface, the interference light sensors 55A-55D sequentially output one line's worth of signals from each of the light-receiving surfaces 55r, 55g, and 55b. For example, the amount of light received by the line sensor in the ith column of the light-receiving surface 55r at a specific time t is defined as SR(i,t), and the interval between outputting one line is defined as ΔT. When the beam spot 40r moves in the S1 direction in FIG. 5, the signals SRO(t) output from the interference light sensors 55A-55D at a specific time t are calculated by the following equation.

[0053]

number

[0054] 6. Darkfield Optical Unit 6 The dark field optical unit 6 is a unit that performs dark field detection, and includes a perforated mirror 60 , a spatial filter unit 61 , an imaging lens 62 , and a dark field optical sensor 63 .

[0055] The perforated mirror 60 is conjugate with the pupils of the objective lenses 43 and 44 with respect to the relay lenses 34a, 34b, 35a, and 35b. The perforated mirror 60 does not interfere with the optical axes of the relay lenses 34a and 34b, but reflects light that deviates from the optical axis by more than a predetermined distance. Since the reflected light from the smooth reflecting mirror 45 travels uniformly along the optical axis, all of it passes through the perforated mirror 60 and is guided to the interference optical unit 5. On the other hand, the light collected by the objective lens 43 includes not only direct reflected light from the sample 1 but also scattered light generated by foreign matter. The directly reflected light traveling along the optical axis passes through the perforated mirror 60 together with the reflected light from the reflecting mirror 45 and is guided to the interference optical unit 5. On the other hand, the scattered light traveling off the optical axis is reflected by the perforated mirror 60 and is guided to the spatial filter unit 61. All of the light guided to the spatial filter unit 61 is scattered light generated by the sample 1. The spatial filter unit 61 is located close to the perforated mirror 60 and, like the perforated mirror 60, is arranged at a position conjugate with the pupils of the objective lenses 43 and 44. Light that has passed through the spatial filter unit 61 is imaged by an imaging lens 62 onto the light receiving surface of a dark-field optical sensor 63. The dark-field optical sensor 63 detects an image of a foreign particle on the surface of the sample 1 using light with the shortest wavelength (405 nm), for example.

[0056] 6-1. Spatial Filter Unit 61 FIG. 6 is a schematic diagram of the spatial filter unit 61. The spatial filter unit 61 includes rods 61a-61j that are driven in translation by a motor. Rods 61a-61e extend vertically in the figure and are aligned parallel to the horizontal direction, while rods 61f-61j extend horizontally in the figure and are aligned parallel to the vertical direction. Rods 61a-61e and rods 61f-61j overlap, forming a mesh-like configuration. The spatial filter unit 61 uses the mesh-like rods 61a-61j to remove diffracted light from a periodic pattern formed on the surface of the sample 1 and transmit scattered light from foreign matter through the mesh. Rods 61a-61e can be translated horizontally, and rods 61f-61j can be translated vertically. The mesh size can be adjusted by changing the spacing between rods 61a-61e and rods 61f-61j. The spatial filter unit 61 is combined with a band-pass filter that transmits only light of a wavelength (405 nm) to be detected by the dark-field optical sensor 63, if necessary.

[0057] Figure 7 is a schematic diagram showing an example of a pattern formed on the sample 1. This figure shows the field of view 1ijk of the illumination and detection optical unit 4. Regular patterns, such as contact patterns 10aa-10ad, 10ba-10bd, 10ca... shown in the figure, are formed on the surface of the sample 1. The diffracted light from these patterns reduces the sensitivity of dark-field detection of foreign particles. The contact patterns are copper patterns for electrical conductivity and are often arranged in a houndstooth pattern. In the example shown in the figure, the contact patterns 10aa-10ad, 10ba-10bd, 10ca... are aligned in a line in the X direction, with a constant pitch in the Y direction. The alignment of the contact patterns between adjacent rows in the Y direction is offset by 1 / 2 pitch in the X direction. In the sample 1 with such patterns, the non-patterned portions are covered with a transparent silicon dioxide film 11, which is optically transparent to the illumination light typically used for foreign particle inspection. The target of detection by the dark-field optical unit 6 is foreign particles 12 present in the transparent film 11 in the non-patterned portions.

[0058] The rods 61a-61f of the spatial filter unit 61 shown in Figure 6 block diffracted light from the pattern, thereby suppressing deterioration of foreign particle inspection sensitivity due to diffracted light from the pattern. The optical path of the diffracted light changes depending on the wavelength of the illumination light. By combining a bandpass filter, the light blocking targets of the rods 61a-61f are narrowed down to only diffracted light of wavelengths that pass through the spatial filter unit 61, making it easier to adjust the position of the rods 61a-61f. Although not shown, the adjustment of the rods 61a-61f can be further facilitated by installing a camera that monitors the light passing through the spatial filter unit 61 and a movable mirror that switches the optical path of the light passing through the spatial filter unit 61 toward the camera.

[0059] 7. Signal Processing Device 7 The signal processing device 7 is, for example, a computer, and performs the functions of dark-field data processing 71, interference data processing 72, and processing result integration 73.

[0060] 7-1. Dark-field data processing 71 The signal processing device 7 receives image data from the dark-field optical sensor 63 and stores it sequentially in memory. In the dark-field data processing 71, images of identically designed portions of the sample 1 are compared, and coordinate regions with large data differences (e.g., exceeding a threshold) are identified as defects. Examples of the dark-field data processing 71 include die comparison, cell comparison, and a combined die-cell comparison, as described in U.S. Patent No. 7,889,911. Cell comparison is a known method for detecting defects in memory cells where the same pattern is repeatedly formed by comparing images of positions shifted by the cell pitch (i.e., positions with the same design). The contact pattern illustrated in FIG. 7 is similar to a memory cell in that the same pattern is repeatedly formed, and defect detection algorithms such as cell comparison can be easily applied to defect inspection of a sample 1 with contact patterns.

[0061] 7-2. Interference data processing 72 Interference data processing 72 is processing for calculating the surface height of sample 1. The surface of sample 1 includes a copper contact pattern portion and a silicon dioxide transparent film portion, and interference data processing 72 calculates the surface height for both the contact pattern portion and the silicon dioxide transparent film portion.

[0062] Figure 8 is a cross-sectional view of sample 1 taken along line VIII in Figure 7. As shown in the figure, sample 1 has a configuration in which a transparent film 11 made of silicon dioxide is formed on a silicon substrate 14 made of silicon, and contact patterns 10aa-10ad formed on the surface are embedded in transparent film 11. In the example of Figure 8, an internal pattern 13 made of copper exists deeper than contact patterns 10aa-10ad (near silicon substrate 14).

[0063] 7 and 8, illumination light incident on the sample 1 from the objective lens 43 is incident on the contact patterns 10aa-10ad or the transparent film 11. The illumination light incident on the contact patterns 10aa-10ad is reflected on the surfaces of the contact patterns 10aa-10ad and collected by the objective lens 43. The illumination light incident on the transparent film 11 passes through the transparent film 11 and reaches the internal pattern 13 in the transparent film 11 or the silicon substrate 14 below the film, and the reflected light is collected by the objective lens 43.

[0064] (1) Pattern surface height First, consider the light incident on contact patterns 10aa-10ad. These contact patterns 10aa-10ad are made of copper and are exposed on the surface of transparent film 11. The phase difference generated on the surface of contact patterns 10aa-10ad is Δh, the amplitude reflectance on the sample surface is R, and the phase difference generated on reflecting mirror 45 is Δh2. For simplicity, we assume that the reflectance of reflecting mirror 45 is 1, and the illumination light incident on quarter-wave plate 42 is P-polarized. In this case, the following two relational expressions hold for the light intensities I1(λ), ​​I2(λ), I3(λ), and I4(λ) of light with wavelength λ detected by interference light sensors 55A, 55B, 55C, and 55D:

[0065]

number

[0066]

number

[0067] The following relational expression can be obtained from the above two equations.

[0068]

number

[0069] 9 is a graph showing the relationship between the amount of light detected by interference light sensors 55A, 55B, 55C, and 55D and the surface height of contact patterns 10aa-10ad when illumination light with a wavelength of 405 nm is incident on the surface of contact patterns 10aa-10ad. The X-axis in the graph represents the surface height of contact patterns 10aa-10ad, and the Y-axis represents the amount of light. Light amount 801 is the amount of light I1(λ) detected by interference light sensor 55A, light amount 802 is the amount of light I2(λ) detected by interference light sensor 55B, light amount 803 is the amount of light I3(λ) detected by interference light sensor 55C, and light amount 804 is the amount of light I4(λ) detected by interference light sensor 55D.

[0070] FIG. 10 is a graph showing the relationship between the estimated and actual surface heights of the contact patterns 10aa-10ad calculated using the relationship in Equation 4 based on the measured light intensity shown in FIG. 9. The X-axis in FIG. 10 is the same as the X-axis in FIG. 9 and represents the height of the contact patterns 10aa-10ad used in the measurement. The Y-axis in FIG. 10 represents the surface height of the contact patterns 10aa-10ad calculated using Equation 4. The estimated surface height 901 is obtained as shown in FIG. In the inspection system 100, because the phase shift occurs between the illumination and detection sides, the detection range is half the wavelength of 405 nm of the illumination light. The estimated value 901 contains points where the value changes significantly and discretely at a period of half the wavelength of 405 nm. However, if the hypothesis that the surface of the contact patterns 10aa-10ad is smooth is established, the periodic estimated value 901 can be considered to be linearly connected by offsetting the discrete changes in the value by the step difference. As shown in FIG. 10, the surface height of the contact patterns 10aa-10ad can be accurately determined from the measured light intensity.

[0071] FIG. 11 is a graph showing the relationship between the light intensity detected by interference light sensors 55A, 55B, 55C, and 55D and the surface height of contact patterns 10aa-10ad when illumination light with a wavelength of 660 nm is incident on the surface of contact patterns 10aa-10ad. The measurement conditions in FIG. 11 are the same as those in FIG. 9, except that the wavelength of the illumination light is 660 nm. Light intensities 1001-1004 are the light intensities obtained by interference light sensors 55A-55D, respectively. FIG. 12 is a graph showing the relationship between the estimated and actual surface heights of contact patterns 10aa-10ad calculated using the relationship in Equation 4 based on the measured light intensities in FIG. 11. Comparing estimated value 1006 in FIG. 12 with estimated value 901 in FIG. 10 reveals that the positions at which the values ​​change discretely differ due to differences in the wavelength of the illumination light, and that the surface heights of contact patterns 10aa-10ad are accurately determined from the measured light intensity. By using the measured light amounts obtained by illuminating a plurality of lights with different wavelengths in this way, the dynamic range can be greatly improved.

[0072] (2) Surface height of a transparent film with a constant thickness Next, let us consider the light incident on the transparent film 11. In order to calculate the surface height of the transparent film 11, it is necessary to know the refractive index of the transparent film 11 in advance. If the refractive index of the transparent film 11 for illumination light of wavelength λ is n1(λ), ​​the distance (film thickness) D(λ) to a reflective object inside or below the transparent film 11 can be calculated using the following equation:

[0073]

number

[0074] Let ρ01(λ) be the reflectance at the boundary between air and the transparent film, and ρ12(λ) be the reflectance at the boundary between the transparent film and a reflective object within the transparent film. Based on the measured light intensities I1(λ)-I4(λ), the reflected light intensity α(λ) can be calculated from Equation 2-Equation 4 as follows:

[0075]

number

[0076]

number

[0077] Also, σ is 1 or -1, m D is an integer value and mh is an integer, the surface height of the transparent film 11 can be calculated as follows:

[0078]

number

[0079]

number

[0080]

number

[0081] 13 is a graph showing the relationship between the amount of light detected by interference light sensors 55A, 55B, 55C, and 55D and the surface height of transparent film 11 when illumination light with a wavelength of 405 nm is incident on the surface of transparent film 11 with a constant film thickness. The X axis in the graph represents the surface height of transparent film 11, and the Y axis represents the amount of light. Light amount 1101 is the amount of light I1(λ) detected by interference light sensor 55A, light amount 1102 is the amount of light I2(λ) detected by interference light sensor 55B, light amount 1103 is the amount of light I3(λ) detected by interference light sensor 55C, and light amount 1104 is the amount of light I4(λ) detected by interference light sensor 55D.

[0082] FIG. 14 is a graph showing estimated values ​​of the film thickness of the transparent film 11 calculated using the relational expression [Equation 5] based on the light intensity shown in FIG. 13 obtained by measurement. As shown in the figure, it can be seen that two estimated values ​​1201 and 1202 can be obtained based on the detected light intensity. FIG. 15 is a graph showing estimated values ​​of the surface height of the transparent film 11 calculated using the relational expression [Equation 8] based on the light intensity shown in FIG. 13 obtained by measurement. As shown in the figure, it can be seen that two estimated values ​​1301 and 1302 can be obtained.

[0083] 16 is a graph showing the relationship between the light intensity detected by the interference light sensors 55A, 55B, 55C, and 55D and the surface height of the transparent film 11 when illumination light with a wavelength of 660 nm is incident on the surface of the transparent film 11 with a constant film thickness. Light intensities 1401-1404 are I1(λ)-I4(λ), respectively.

[0084] FIG. 17 is a graph showing the estimated film thickness of the transparent film 11 calculated using the relational expression [Equation 5] based on the measured light intensity shown in FIG. 16. Two estimated values ​​1501 and 1502 are obtained based on the detected light intensity. FIG. 18 is a graph showing the estimated surface height of the transparent film 11 calculated using the relational expression [Equation 8] based on the measured light intensity shown in FIG. 16. Comparing the estimated values ​​1601 and 1602 in FIG. 18 with the estimated values ​​1301 and 1302 in FIG. 15, it is found that the two solutions obtained for different wavelengths share a common solution; specifically, the estimated values ​​1301 and 1601 are consistent. Therefore, it can be seen that an accurate value for the surface height of the transparent film 11 can be obtained by selecting the common solution from the two solutions obtained for multiple wavelengths.

[0085] (3) Surface height of transparent film with varying thickness FIG. 19 is a graph showing the relationship between the light intensity detected by interference light sensors 55A, 55B, 55C, and 55D and the surface height of transparent film 11 when illumination light is incident on the surface of transparent film 11 with varying film thickness. The X-axis of the graph represents the film thickness of transparent film 11, and the Y-axis represents the light intensity. In this example, the surface height of transparent film 11 is assumed to be constant. Light intensity 1701 is the light intensity I1(λ) detected by interference light sensor 55A, light intensity 1702 is the light intensity I2(λ) detected by interference light sensor 55B, light intensity 1703 is the light intensity I3(λ) detected by interference light sensor 55C, and light intensity 1704 is the light intensity I4(λ) detected by interference light sensor 55D. While the detected light intensity varies trigonometrically with height in FIGS. 9, 11, 13, and 16, the brightness in FIG. 19 varies in a complex manner with film thickness.

[0086] FIG. 20 is a graph showing estimated values ​​of the film thickness of the transparent film 11 calculated using the relational expression [Equation 5] based on the light intensity shown in FIG. 19 obtained by measurement. Two estimated values ​​1801 and 1802 are obtained based on the detected light intensity. FIG. 21 is a graph showing estimated values ​​of the surface height of the transparent film 11 calculated using the relational expression [Equation 8] based on the light intensity shown in FIG. 19 obtained by measurement. Two estimated values ​​1901 and 1902 are obtained based on the detected light intensity. In the inspection device 100, the interference light sensors 55A-55D simultaneously obtain detected light intensities for three illumination lights with different wavelengths. That is, the interference light sensors 55A-55D each obtain data for a wavelength of 405 nm, and also obtain data for wavelengths of 532 nm and 660 nm.

[0087] FIG. 22 is a graph showing the profiles of all detected light intensities obtained by the interference light sensors 55A-55D when the film thickness of the transparent film 11 is changed. FIG. 23 is a graph showing estimated values ​​of the film thickness of the transparent film 11 calculated using the relational expression [5] based on the measured light intensities shown in FIG. 22. FIG. 24 is a graph showing estimated values ​​of the surface height of the transparent film 11 calculated using the relational expression [8] based on the measured light intensities shown in FIG. 22. Two estimates are calculated for each wavelength for both film thickness and surface height, but for any given value on the X-axis, there are four estimated values ​​of surface height, as shown in FIG. 24. This indicates that some estimates at different wavelengths are consistent. In this way, by comparing and voting surface heights calculated using multiple wavelengths, it is possible to calculate multiple calculated surface heights or transparent film thicknesses.

[0088] To obtain the surface height and film thickness of the transparent film 11, the refractive index, i.e., the material of the transparent film 11, must be known in advance. In the above explanation, boundary reflectances such as ρ01 and ρ12 are used, but these can be calculated from the refractive index. In this case, the material can also be identified by voting. Although the material of the portion of the measurement sample being measured cannot be clearly determined, in evaluations of semiconductor wafers and other manufacturing processes, the sample 1 is assumed to be made of several materials, such as silicon dioxide, copper, and silicon. For example, the material of the transparent film 11 is limited to silicon dioxide, and the surface material of the sample 1 can be selected from two options: silicon dioxide and copper. The material from which light incident on the transparent film 11 is reflected can be selected from two options: copper and silicon. Given these options, the surface height of the sample 1 and / or the film thickness of the transparent film 11 based on the detected light intensity are calculated using the refractive index for each option, and the estimated values ​​calculated for each material selection setting are compared. In the test area where the options match the actual material, the estimated values ​​are partially common across multiple wavelengths as shown in Figure 24, and the material can be identified based on the voting results for that identity.

[0089] (4) Calculation procedure FIG. 25 is a flowchart showing the procedure for calculating the surface height and film thickness of an arbitrary measurement portion of the sample 1 by the interference data processing 72.

[0090] First, Step 1 is a loop for the wavelength of light used in measurement, and the signal processing device 7 repeats the procedure of Steps 2 to 5 by sequentially switching the wavelength of light for an arbitrary measurement coordinate P1. In this embodiment, there are three wavelengths of light: 405 nm, 532 nm, and 660 nm, so Step 1 is repeated three times.

[0091] In Step 2, the signal processing device 7 determines whether the surface of the measurement coordinate P1 is not silicon dioxide (transparent film 11) under the wavelength conditions set in Step 1, that is, copper The surface height is calculated assuming that the pattern is a contact pattern. The calculation method is as described above.

[0092] Step 3 is a loop for selecting options for the transparent film 11, assuming that the measurement coordinate P1 is the transparent film 11. In the case of a semiconductor substrate or the like, the material of the transparent film 11 assumed in the measurement stage by the inspection device 100 is often limited to silicon dioxide, and in this embodiment, the options for the material of the transparent film 11 set in Step 3 are limited to silicon dioxide. Therefore, the number of repetitions of Step 2 is 1.

[0093] Step 4 is a loop for selecting options for the material in or under the film that reflects the illumination light, assuming that the measurement coordinate P1 is the transparent film 11. Since the material in the film at the measurement coordinate P1 is not known in advance, multiple options are assumed, such as copper and silicon.

[0094] In Step 5, the signal processing device 7 calculates the film thickness and surface height using the formulas [Equation 5] to [Equation 10] based on the wavelength set in Step 1, the refractive index of the transparent film 11 set in Step 3, and the refractive index of the material set in Step 4. Here, for example, for the surface height of the transparent film, multiple estimated values ​​are calculated as candidates due to the existence of an indefinite mh and σ that is 1 or −1 as described above.

[0095] After calculating the estimated film thickness and surface height values ​​for all combinations of options in Steps 1, 3, and 4, the signal processing device 7 proceeds to Step 6. Step 6 is a loop for the estimated film thickness and surface height values ​​(candidates) calculated in Steps 1 to 5, and Step 7 is a loop for the wavelength of light used for measurement. In Step 8, the signal processing device 7 calculates the reflected light intensity based on the settings in Steps 6 and 7. The reflected light intensity is determined by the light reflected from the reflecting mirror 45, the light reflected from the surface of the transparent film 11, the light reflected from the opaque material in the film, and the light reflected multiple times between the surface of the transparent film 11 and the surface of the opaque material. In this embodiment, particularly, the signal processing device 7 calculates the reflected light intensity for a wavelength λb, which is different from the wavelength λa related to the estimated value, based on the estimated value set in Step 6 in Step 8. Then, in the subsequent Step 9, the signal processing device 7 compares the reflected light intensity calculated in Step 8 with the detected light amount obtained at wavelength λb. To give a specific example, cross-validation is performed to calculate the reflected light intensity that should be obtained with light of wavelength 660 nm, using estimated values ​​of film thickness and surface height calculated based on the detected light amount using light of wavelength 405 nm. If the estimated values ​​of film thickness and film surface height are equal to the actual values, the estimated value of reflected light intensity and the detected light amount will match in Step 9.

[0096] After executing the process of Step 9 for all wavelengths regarding the estimated values ​​of film thickness and surface height set in Step 6, the signal processing device 7 calculates an integrated evaluation value for all wavelengths in Step 10. As an algorithm for Step 10, for example, it is possible to compare the difference between the reflected light intensity (calculated value) calculated in Step 9 and the detected light amount, and calculate the maximum difference for each wavelength as the evaluation value. As the evaluation value, it is also possible to appropriately use values ​​known by statistical methods, such as the average value or the median value. As an algorithm for Step 11, it is possible to select the minimum evaluation value calculated in Step 10, and determine the estimated values ​​of film thickness and surface height related to this minimum evaluation value as the measured values ​​of film thickness and surface height at coordinate P1.

[0097] After completing the loop of Step 6 and calculating the evaluation value in Step 10 for each estimated value, the signal processing device 7 proceeds to Step 11, extracts the estimated value of film thickness and surface height that gives the best evaluation value, and determines this as the measured value of film thickness and surface height at the coordinate P1.

[0098] In the example of Fig. 25, a method of calculating reflected light intensity and performing cross-validation was described, but it is also possible to employ an algorithm that compares the estimated values ​​of surface height and film thickness calculated in Step 5 for different wavelengths to evaluate consistency and determines the most frequently calculated estimated value as the measured value. Fig. 26 is a flowchart showing another example of the procedure for calculating surface height and film thickness using this algorithm.

[0099] In Figure 26, Steps 1 to 5 are the same processes as Steps 1 to 5 in Figure 25. After calculating the estimated values ​​of film thickness and surface height for all combinations of options in Steps 1, 3, and 4, the signal processing device 7 proceeds to Step 20. Step 20 is a loop for the estimated values ​​(candidates) of surface height and film thickness calculated using a specific wavelength λ1. For example, if the specific wavelength is 405 nm, the estimated values ​​are calculated based on the amount of light detected using light with a wavelength of 405 nm, assuming that a transparent film 11 is formed on the sample surface. The materials of the transparent film 11 and the reflective material within the film are set as conditions. This condition setting is repeated the number of times equal to the number of pairs of surface height and film thickness calculated based on the amount of light detected using light with a wavelength of 405 nm.

[0100] Step 21 is a loop for the wavelength λ2 of light used other than the specific wavelength λ1. For example, if the specific wavelength λ1 is set to 404 nm in Step 20, in this embodiment, 532 nm and 660 nm are sequentially set as the wavelength λ2 in Step 21.

[0101] In Step 22, the signal processing device 7 selects the surface height and film thickness estimates calculated for wavelength λ2 for the material set in Step 20 that are closest to the surface height and film thickness estimates calculated for light of the specific wavelength λ1. As an example, consider the examples shown in FIGS. 14, 15, 17, and 18. In this case, of the film thickness and surface height estimates 1501, 1502, 1601, and 1602 for wavelength λ2 (660 nm), estimates 1501 and 1601 are closest to the film thickness and surface height estimates 1201 and 1301 for the specific wavelength λ1 (405 nm) under the same material assumption. Under the material assumptions shown in FIGS. 14, 15, 17, and 18, estimates 1501 and 1601 are selected in Step 22 as candidate film thickness and surface height measurements for wavelength λ2. This selection of measurement value candidates is performed for all wavelengths (532 nm, 660 nm in this example) (Step 23, loop (g)).

[0102] In Step 23, the signal processing device 7 selects the value of the candidate measurement values ​​obtained in Step 22 for each wavelength λ2 for the material set in Step 20 that has the greatest deviation from the estimated value for the specific wavelength λ1 compared with the candidate measurement values ​​in Step 22 as the evaluation value for that material. If the assumption about the material is correct, the candidate measurement values ​​will match the estimated value for the specific wavelength λ1 regardless of the wavelength λ2. If the assumption about the material is incorrect, at least one of the candidate measurement values ​​selected for each wavelength λ2 will deviate from the estimated value for the specific wavelength λ1.

[0103] After determining the evaluation value in Step 23 for each material set in Step 20, the signal processing device 7 proceeds to Step 24. In Step 24, the signal processing device 7 selects the best value from the evaluation values ​​determined in Step 23 for each material, determines the material assumption related to this best value to be the appropriate assumption, and determines the measurement value candidates related to the appropriate assumption to be the measurement values ​​of the film thickness and surface height at the coordinate P1.

[0104] (5) Defect determination FIG. 27 is a block diagram showing an example of functional blocks related to defect inspection of the sample 1 by the signal processing device 7.

[0105] In the inspection system 100, 12 image data for the same region are input to the signal processing device 7, which are the product of four different polarized lights and three different wavelengths. The signal processing device 7 calculates the measurement values ​​of the height and film thickness of the sample surface from the interference data in process 7201. Specifically, the algorithm described in FIG. 25 or 26 can be used for process 7201.

[0106] In the following process 7202, the signal processing device 7 corrects the measurement values ​​of the height and film thickness of the sample surface in accordance with the height fluctuation of the sample 1. Because the measurement values ​​are on the order of nanometers, the measurement value of the height of the sample surface changes depending on the holding state of the sample 1, etc. Therefore, the difference between a representative height value near the measurement coordinate of the sample 1 and the measurement value of the surface height is calculated, and the calculated difference is used as a correction value to correct the measurement value. As the representative height value, for example, the average value or median value of the height of a predetermined region of the sample 1 including the measurement coordinate, or a height value filtered with a low-pass filter can be used. In process 7203, the signal processing device 7 stores the corrected surface height data of the sample 1 in memory. This data also includes data on the surface material (silicon dioxide, copper, etc.) for each coordinate of the sample 1.

[0107] Next, in process 7204, the signal processing device 7 judges whether there is an abnormality on the surface of the sample 1. As a method of judging whether there is an abnormality, it is possible to judge whether the measured value of the surface height after correction stored in memory in process 7203 is within an appropriate height range (set value). Furthermore, instead of comparing with a set value, it is also possible to apply a method in which the surface heights after correction are compared between portions of the same design on the same chip and the abnormality is judged whether the difference is within an appropriate range. It is also possible to apply a method in which the surface heights after correction are compared between the same positions on different samples 1, or the surface heights after correction are compared between corresponding positions on different chips on the same sample 1 and the abnormality is judged whether the difference is within an appropriate range. A combination of these methods is also applicable.

[0108] Furthermore, the signal processing device 7 extracts feature quantities of the region determined in process 7205 based on the obtained data. Feature quantities extracted here include, for example, the surface roughness and height variations (average height, maximum height, minimum height, etc.) of the region determined to be abnormal in process 7204. Roughness can be obtained from scattered light signals other than the foreign matter signal measured by the dark-field optical sensor 63. It is also possible to calculate, for example, the average value or maximum value of the step between the contact pattern (a portion whose surface material is estimated to be copper) and the transparent film 11 (a portion whose surface material is estimated to be silicon dioxide). Data on whether the step is positive or negative, i.e., whether the contact pattern or the transparent film is higher, can also be obtained.

[0109] 7-3. Integration of processing results 73 In the process of processing result integration 73, the signal processing device 7 outputs, for example, coordinates where the surface height is outside the appropriate range and coordinates where a foreign substance 12 is detected, based on the data obtained in the interference data processing 72 and the dark-field data processing 71. Furthermore, the signal processing device 7 stores the measured values ​​of the surface height and film thickness of the sample in the storage device 84, and can display, for example, a map of the surface height of the sample in an area specified by the user's operation on the monitor 83 in response to the operation of the UI 82.

[0110] FIG. 28 is a diagram showing an example of an output screen. FIG. 28 illustrates a case where the sample 1 is a semiconductor wafer. A detected defect 8302 is displayed on a map 8301 of the sample 1. The screen of FIG. 28 also displays an enlarged surface height map 8304 of a specified region 8303 specified by the user on the map 8301 in a display format in which the display color and shading vary depending on the height. By manipulating cross-sectional lines 8306 and 8307 extending vertically and horizontally on the surface height map 8304 and moving them to desired positions, cross-sectional waveforms 8308 and 8309 of the sample 1 at desired locations on the surface height map 8304 are displayed. Furthermore, by manipulating a frame 8305 on the surface height map 8304 and moving it to a desired position, feature data 8310 of the region specified by the frame 8305, such as the maximum height, minimum height, and height variance within the region of the frame 8305, is displayed.

[0111] 8.Effects (1) According to this embodiment, by using interference light as described above, the surface height and thickness of the transparent film 11, which has low reflectivity, can be measured with high throughput by a single scan without being affected by reflection from copper wiring or the like present in the optically transparent film. In this way, in a manufacturing line for semiconductor substrates, thin film substrates, etc., the film thickness and surface height of a transparent film on the substrate surface can be measured with high accuracy and high speed.

[0112] (2) Furthermore, as explained in Figures 17 and 18, the surface height and film thickness of the transparent film 11 cannot be uniquely calculated based on the detected light intensity of the interference light. In contrast, in this embodiment, interference light sensors 55A-55D simultaneously detect interference light of different wavelengths, and by comparing estimated values ​​such as the surface height calculated for each wavelength, a highly reliable measurement value can be determined from multiple estimated values. Furthermore, since the estimated values ​​to be compared are calculated by separating the candidate materials of the beam spot, the candidate materials of the beam spot are also identified when the measurement value is determined.

[0113] (3) Furthermore, defects such as abnormal film thickness can be determined from the measured data on the surface height and film thickness for each portion of the sample 1.

[0114] (4) By separating and extracting dark field light from the interference light, it is possible to simultaneously and accurately inspect the surface of the sample 1 for defects such as foreign matter.

[0115] (Second embodiment) Fig. 29 is a schematic diagram showing an example of the configuration of an inspection device according to a second embodiment of the present invention. In Fig. 29, elements that are the same as or correspond to those in the inspection device 100 of the first embodiment are given the same reference numerals as in Fig. 1, and descriptions thereof will be omitted as appropriate.

[0116] 9. Illumination Optical Unit 3 In this embodiment, a laser light source that emits white light is used as the light source 30, rather than a multi-line laser light source that individually emits three monochromatic light beams. The inspection device 100 of the first embodiment employs a configuration in which a flat Gaussian beam is manipulated by the optical scanning unit 33, but the inspection device 100 of the second embodiment employs a configuration in which an isotropic Gaussian beam is illuminated. For this reason, the illumination shaping unit 31 (FIG. 1) that uses an anamorphic prism is replaced with a beam expander 37. Accordingly, the optical scanning unit 33 (FIG. 1) is no longer necessary and is omitted in this embodiment.

[0117] 10. Beam Expander 37 The beam expander 37 is a unit that expands the beam diameter of the incident illumination light and includes multiple lenses 37a and 37b. FIG. 29 illustrates a Galilean beam expander 37 that uses a concave lens as the lens 37a and a convex lens as the lens 37b. The beam expander 37 is equipped with a mechanism (zoom mechanism) for adjusting the spacing between the lenses 37a and 37b, and the expansion ratio of the beam diameter changes by adjusting the spacing between the lenses 37a and 37b. The beam diameter expansion ratio of the beam expander 37 is, for example, approximately 5 to 10 times. In this case, if the beam diameter of the illumination light emitted from the laser light source 30 is 1 mm, the beam diameter of the illumination light is expanded to approximately 5 to 10 mm. If the illumination light incident on the beam expander 37 is not a parallel beam, collimation (quasi-parallelization of the beam) as well as the beam diameter can be achieved by adjusting the spacing between the lenses 37a and 37b. However, for collimating the light beam, a collimating lens may be installed upstream of the beam expander 37 separately from the beam expander 37. The illumination light that has passed through the beam expander 37 is guided to the illumination / detection optical unit 4 via an illumination lens 38.

[0118] 11. Illumination and detection optical unit 4 The illumination light incident on the polarizing beam splitter 41 via the illumination lens 38 and the quarter-wave plate 42 is split by the polarizing beam splitter 41 into two orthogonally polarized beams. One of the split beams is converted into circularly polarized light by a quarter-wave plate 46, whose fast or slow axis is rotated by 45°, and is then irradiated onto the sample 1 via the objective lens 43. The reflected light from the sample 1 is again incident on the quarter-wave plate 46, where its polarization is shifted by 90° from that of the illumination light incident from the polarizing beam splitter 41. The reflected light then passes through the polarizing beam splitter 41 and is guided via relay lenses 48a and 48b to the interference optical unit 5 and the dark-field optical unit 6. The other beam split by the polarizing beam splitter 41 is converted into circularly polarized light by a quarter-wave plate 47, whose fast or slow axis is rotated by 45°, and is then irradiated onto the reflecting mirror 45 via the objective lens 44. The light reflected from the reflecting mirror 45 is again incident on the quarter-wave plate 47, where its polarization is shifted by 90° from the illumination light incident from the polarizing beam splitter 41. The reflected light, whose polarization has been shifted by 90°, is then reflected by the polarizing beam splitter 41 and directed to the interference optical unit 5 and the dark-field optical unit 6 via relay lenses 48a and 48b. In this way, the interference light of the reflected lights from the sample 1 and the reflecting mirror 45 is directed to the interference optical unit 5 and the dark-field optical unit 6 via relay lenses 48a and 48b.

[0119] 12. Interference Optical Unit 5 The interference optical unit 5 differs from the inspection apparatus 100 of FIG. 1 in that the interference light sensors 55A-55D, which are TDI sensors, have been replaced with interference light sensors 56A-56D, which are 2D sensors. FIG. 30 is a schematic diagram of the interference light sensors 56A-56D, and FIG. 31 is a schematic diagram of the light-receiving element arrays provided in the interference light sensors 56A-56D. As shown in FIG. 22A, the interference light sensors 56A-56D are three-chip cameras and each include three light-receiving element arrays 56r, 56g, and 56b that detect light with wavelengths of 660 nm, 532 nm, and 405 nm. The interference light incident on the interference light sensors 56A-56D is split by three prisms into light with wavelengths of 660 nm, 532 nm, and 405 nm, respectively, and then enters the light-receiving element arrays 56r, 56g, and 56b. The light receiving element arrays 56r, 56g, and 56b each include a large number of light receiving elements arranged two-dimensionally as shown in FIG.

[0120] Other configurations of this embodiment are the same as those of Embodiment 2. In Embodiment 1, a multi-line laser light source was used to provide a plurality of illumination lights with different wavelengths, but similar effects can be obtained by using normal illumination light as in this embodiment, and dispersing the interference light according to wavelength in the process of guiding it to the light-receiving element arrays 56r, 56g, and 56b.

[0121] (Third embodiment) The inspection apparatus 100 of the first and second embodiments employs a scanning method in which the stage 2 repeats a step-and-repeat operation. In contrast, in this embodiment, the stage 2 operates continuously at a constant speed, scanning the entire surface of the sample 1 without stopping. The inspection apparatus 100 of this embodiment also includes a unit for adjusting the angle of incidence of the illumination light on the sample 1.

[0122] 13. Lighting optical unit 3 Fig. 32 is a schematic diagram showing an example of the configuration of an inspection device according to the second embodiment of the present invention. In Fig. 32, elements that are the same as or correspond to those in the inspection device 100 of the first or second embodiment are given the same reference numerals as in Fig. 1 or 29, and descriptions thereof will be omitted as appropriate.

[0123] The inspection apparatus 100 of this embodiment is equipped with an illumination incident angle adjustment unit 39 having mirrors 39a and 39b. The illumination incident angle adjustment unit 39 changes the illumination angle of the illumination light shaped into a flattened shape by the illumination shaping unit 31. The illumination angle is adjusted by driving and moving the mirror 39b using a drive device (not shown) that operates in response to commands from the control device 81. Driving the mirror 39b makes it possible to switch between oblique illumination, which illuminates the sample 1 obliquely, and epi-illumination, which illuminates the sample 1 perpendicularly, as in the first and second embodiments. The mirror 39b is positioned conjugate to the pupil planes of the objective lenses 43 and 44. The illumination light reflected by the mirror 39b enters the quarter-wave plate 42 via relay lenses 34a and 34b. The light transmitted through the quarter-wave plate 42 is split by a polarizing beam splitter 41 according to its polarization direction.

[0124] The light split by the polarizing beam splitter 41 and directed toward the objective lens 43 is irradiated obliquely onto the sample 1 using S-polarized illumination to form a beam spot. The light reflected from the sample 1 is collected by the objective lens 43 and enters a half-wave plate 46-2 whose fast axis or slow axis has been rotated by 45°. The reflected light, whose polarization direction has been rotated by 90° by the half-wave plate 46-2, passes through the polarizing beam splitter 41 and is guided to the interference optical unit 5 and the dark-field optical unit 6 via relay lenses 48a and 48b.

[0125] On the other hand, the light separated by the polarizing beam splitter 41 and directed toward the objective lens 44 is incident on a half-wave plate 47-2 whose fast axis or slow axis is rotated by 45°, becomes S-polarized light whose polarization direction is rotated by 90°, and forms a beam spot on the reflecting mirror 45. The reflected light from the reflecting mirror 45 is collected by the objective lens 44 and reflected by the polarizing beam splitter 41, and is guided as interference light with the reflected light from the sample 1 to the interference optical unit 5 and the dark-field optical unit 6 via relay lenses 48a and 48b.

[0126] As described above, both of the half-wave plates 46-2 and 47-2 cover only about half the effective pupil diameter of the objective lenses 43 and 44, and are configured to pass only light incident on or emitted from the objective lenses 43 and 44.

[0127] Figure 33 is a diagram showing the reflection characteristics of silicon dioxide, illustrating the difference in the corresponding reflectance depending on the polarization of the incident angle on the surface of the transparent film. The X-axis represents the incident angle, and the Y-axis represents the reflectance. In order to stably measure the surface height of the sample 1, it is effective to increase the reflectance on the surface of the transparent film as much as possible. Characteristic 2701 represents the reflectance characteristic of S-polarized light, and characteristic 2702 represents the reflectance characteristic of P-polarized light. It can be seen that as the incident angle increases, the reflectance of S-polarized light increases relative to P-polarized light. Therefore, when forming a beam spot on the sample 1, by irradiating illumination light obliquely with S-polarized light as in this embodiment, the surface height of the transparent film can be stably calculated.

[0128] 14. Stage 2 FIG. 34 shows the scanning trajectory of the stage 2 provided in the inspection apparatus according to this embodiment. The stage 2 provided in the inspection apparatus 100 according to this embodiment is equipped with a θ rotation stage (not shown) in addition to an XY stage. The rotation speed is set to synchronize with the data transmission rate of the light-receiving elements of the TDI sensors employed in the interference light sensors 55A-55D. The sample 1 moves and rotates relative to the beam spot through a combination of translational movement by the XY stage and rotational movement by the θ rotation stage. As shown in FIG. 34, the beam spot moves along a spiral trajectory from the center to the outer edge of the sample 1, scanning the entire surface of the sample 1. The beam spot moves in the s2 direction by a distance equal to or less than the length of the beam spot in the s2 direction while the sample 1 rotates once in the s1 direction.

[0129] FIG. 35 shows another example of the scanning trajectory of the sample 1 in this embodiment. The example in FIG. 35 shows a scanning trajectory in which only the XY stage is driven. In this example, the beam spot scans the surface of the sample 1 by overlapping linear trajectories rather than a spiral trajectory. Specifically, the X stage is driven in translation in the s1 direction at a constant speed, and the Y stage is driven in the s2 direction by a predetermined distance (for example, a distance equal to or less than the length of the beam spot BS in the s2 direction), after which the X stage is again driven in translation in the s1 direction. As a result, the beam spot repeats linear scanning in the s1 direction and movement in the s2 direction to scan the entire surface of the sample 1. Compared to this scanning method, the spiral scanning method shown in FIG. 34 does not involve reciprocating motion, and therefore inspection of the sample 1 can be completed in a short time.

[0130] 15. Dark field optical unit Figure 36 shows a schematic diagram of the perforated mirror 60 of the dark-field optical unit 6 provided in the inspection apparatus according to this embodiment. Figure 36 shows an example of the perforated mirror 60 having a configuration including two bar mirrors spaced apart in the S2 direction.

[0131] The beam spots 40r, 40g, and 40b shown in FIG. 36 are hypothetical representations of those formed on the sample 1 for ease of explanation; they are not actually formed on the perforated mirror 60 as shown in the figure. Because the beam spots 40r, 40g, and 40b are long in the S2 direction and short in the direction perpendicular to S2, the perforated mirror 60, which is arranged on the pupil-conjugate plane of the objective lenses 43 and 44, becomes a light beam that is short in the S2 direction and long in the direction perpendicular to S2. Therefore, the interference light from the sample 1 and the reflecting mirror 45 passes through the gap between two bar mirrors extending in the direction perpendicular to S2, and dark-field light (scattered light) from the sample 1 traveling along an optical path deviated from the light beam is reflected by the bar mirrors. Although not specifically described in the first and second embodiments, the inspection apparatus 100 of the first and second embodiments can also employ a perforated mirror 60 having a configuration similar to that shown in FIG. 36. In this embodiment, by configuring the illumination light to be incident on the sample 1 at an angle in the S2 direction during oblique illumination, the distance between the two bar mirrors of the perforated mirror 60 can be set large.

[0132] As described above, in the inspection apparatus 100 of this embodiment, the mirror 39b is driven to switch between oblique illumination and epi-illumination, offsetting the optical path of the illumination light. Since the optical path of the interference light is also offset accordingly, the perforated mirror 60 is configured to move in synchronization with the mirror 39b as indicated by the arrow in FIG.

[0133] Other configurations of this embodiment are the same as those of the first or second embodiment. This embodiment also provides the same effects as those of the first and second embodiments. Furthermore, by enabling oblique illumination, it is expected that the measurement accuracy of the surface height of the transparent film 11 will be improved as described above, and the accuracy of defect inspection using dark-field light will also be improved.

[0134] (Variation) The present invention is not limited to the above-described embodiments and may include various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. It is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0135] The above-described configurations, functions, processes, processing means, etc. may be realized in part or in whole by hardware such as an integrated circuit. The above-described configurations, functions, etc. may also be realized by software, with a processor interpreting and executing a program that realizes each function. Information such as programs, tables, and files that realize each function can be stored in various storage media. Examples of various storage media include recording devices such as memory, hard disks, and solid-state drives (SSDs), as well as flash memory cards and digital versatile disks (DVDs).

[0136] In each embodiment, the signal input / output lines shown are those considered necessary for explanation, and do not necessarily show all of the signal input / output lines on the product. In reality, it can be considered that almost all components are interconnected. [Explanation of symbols]

[0137] REFERENCE SIGNS LIST 1... sample, 7... signal processing device, 30... light source, 43... objective lens (first optical unit), 44... objective lens (second optical unit), 41... polarizing beam splitter (interference optical unit), 45... reflecting mirror, 55A-55D... interference light sensor, 55b, 55g, 55r... light receiving surface, 56A-56D... interference light sensor, 56b, 56g, 56r... light receiving surface, 60... perforated mirror (optical path branching unit), 61... spatial filter, 63... dark field light sensor, 100... inspection device

Claims

1. An inspection device for inspecting a sample whose surface is formed of a transparent film through which light passes and an opaque material, A light source and a first optical unit that irradiates a sample with illumination light emitted from the light source and collects first reflected light reflected by the sample; a second optical unit that illuminates a reflecting mirror with the illumination light and collects second reflected light reflected by the reflecting mirror; an interference optical unit that causes the first reflected light and the second reflected light to interfere with each other to obtain interference light; a plurality of interference light sensors for detecting the reflected light intensity of a predetermined polarization component of the interference light; a signal processing device for processing the amount of light detected by the interference light sensor; The signal processing device includes: calculating an estimated value of the surface height or film thickness of the sample for each wavelength of the illumination light, based on the detected light amounts of the interference light sensor for a plurality of illumination light beams having different wavelengths, the refractive indexes of the transparent film and the opaque material, and the wavelength, when any coordinate of the sample is assumed to be the transparent film and when any coordinate of the sample is assumed to be the opaque material; An inspection device that identifies whether the coordinates correspond to the transparent film or the opaque material based on the estimated values ​​calculated for each wavelength, and measures the surface height or film thickness of the sample at the coordinates.

2. 2. The inspection device of claim 1, the interference light sensor detects reflected light of the illumination light for each light having a different wavelength; The signal processing device includes: calculating one or more estimated values ​​of the surface height or the film thickness for each of the wavelengths based on the detected light amount; an inspection apparatus that compares the estimated values ​​calculated for each wavelength with each other, and selects and outputs one of the one or more estimated values ​​calculated for each wavelength as the measurement value of the surface height or the film thickness.

3. 2. The inspection device of claim 1, the interference light sensor detects reflected light of the illumination light for each light having a different wavelength; The signal processing device includes: At least one refractive index of each of the candidate materials for the transparent film and the opaque material is stored; The illumination light incident on the coordinate is directly reflected by the opaque material, and the illumination light is reflected by the opaque material via the transparent film. One or two candidate materials are assumed for the coordinate; calculating one or two estimated values ​​of the surface height or the film thickness for each wavelength based on the refractive index of the assumed candidate material and the detected light amount for each wavelength individually obtained by the interference light sensor; An inspection device that compares the estimated values ​​calculated for each wavelength with each other, selects one of the one or two estimated values ​​calculated for each wavelength, and determines and outputs the measured value of the surface height or the film thickness.

4. In the inspection device of claim 3, The signal processing device is an inspection device that identifies the candidate material associated with the estimated value determined as the measurement value as the material of the coordinate.

5. In the inspection device of claim 3, The signal processing device includes: The process of comparing the estimated values ​​calculated for each wavelength with each other to determine the measured value includes: calculating the amount of light of another wavelength that should be detected so that the estimated value calculated based on the amount of light detected for each wavelength can be calculated equally for the same candidate material at other wavelengths; comparing the calculated value of the amount of light at the other wavelength with the detected amount of light; An inspection device that determines the estimated value that has the smallest deviation from the calculated value as the measured value.

6. In the inspection device of claim 3, The signal processing device includes: The process of comparing the estimated values ​​calculated for each wavelength with each other to determine the measured value includes: comparing the estimated values ​​calculated based on different wavelengths for the same candidate material; An inspection device that identifies a candidate material for which the estimated values ​​for all wavelengths match or the difference is within a tolerance value as the material of the coordinates, and determines the estimated value for that material as the measured value.

7. In the inspection device of claim 3, the light source emits a plurality of monochromatic lights having different wavelengths as the illumination light, The interference light sensor is an inspection device having a plurality of light receiving surfaces that individually detect reflected light for each wavelength of the illumination light.

8. In the inspection device of claim 3, The interference light sensor is an inspection device that has a plurality of light receiving surfaces that individually detect reflected light separated by wavelength.

9. 2. The inspection device of claim 1, The signal processing device includes: measuring the surface height of the sample for each predetermined region; Determine whether the measurement result is within a predetermined range, An inspection device that outputs areas where the measurement results fall outside a predetermined range as defects.

10. 2. The inspection device of claim 1, The inspection device is provided with four of the interference light sensors, each of which detects light whose polarization direction is shifted by 45°.

11. 2. The inspection device of claim 1, The first optical unit is an inspection device that illuminates the surface of the sample with the illumination light in S-polarized light from an oblique angle.

12. 2. The inspection device of claim 1, an optical path branching unit that separates dark field light from the reflected light collected by the first optical unit and the second optical unit; a spatial filter for removing diffracted light from the dark field light separated by the optical path branching unit; a dark-field optical sensor that detects the dark-field light that has passed through the spatial filter; The signal processing device is an inspection device that detects defects in the sample based on the output of the dark field optical sensor.

Images