Optical constant determination device and optical constant determination method

Abstract

To provide an optical constant determination device and an optical constant determination method that can obtain the optical constant of an unknown thin film and substrate.SOLUTION: An optical constant determination device 100 according to the present disclosure comprises: an optical detector 19 that detects reflected light illuminating a sample 30 through a confocal optical system 101, and acquires images of the sample 30 using rays of light in a plurality of wavelengths included in the reflected light; and a processing unit 20 that acquires measurement data of reflectances for the wavelengths based on the images. The processing unit 20 receives input of a plurality of candidate materials selected as candidates for materials of a thin film 32 and a substrate 31 and having a known optical constant; for each of the candidate materials, refers to calculation data in which the relationship between the wavelength and reflectance is indicated for every film thickness of the thin film 32 to calculate a determination coefficient in approximating and calculating the film thickness from the measurement data; and based on the calculated determination coefficient, determines the optical constant of at least one of the thin film 32 and the substrate 31 from optical constants of the plurality of candidate materials.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an optical constant determination device and an optical constant determination method. [Background technology]

[0002] Coatings on substrates such as wafers and optical components (photoresists, anti-reflective coatings, transparent conductive resins, wear inhibitors, lubricants, etc.) are prone to widespread film unevenness and localized film thickness distribution. Thin films coated on substrates exhibit widespread film unevenness and localized film thickness distribution. Technologies for inspecting and measuring the film thickness of such thin films, including these film unevenness and distributions, are essential for improving the uniformity and other quality aspects of wafers and optical components.

[0003] One method for measuring film thickness variations and reflectivity variations caused by film thickness variations is the optical film thickness measurement method. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Patent No. 5871242 [Overview of the project] [Problems that the invention aims to solve]

[0005] Optical methods for measuring film thickness include (1) reflection spectroscopy, (2) polarization analysis, (3) white interference fringe scanning, and (4) confocal microscopy. However, each of these methods has the following problems.

[0006] (1) The method by reflection spectroscopy measures the spectral spectrum by spectroscopically analyzing the reflected light of a sample illuminated with white light. Among the methods by reflection spectroscopy, the peak valley method (PV method) is a method of analyzing the film thickness by obtaining the wavelengths corresponding to the maximum and minimum of the reflectance from the spectral spectrum. In this method, when visible light is used, for a film thickness of about the wavelength (about 0.5 μm) or less, a spectral pattern of maximum and minimum cannot be obtained. Therefore, it cannot be applied to the measurement of film thickness of about the wavelength or less.

[0007] Also, among the methods by reflection spectroscopy, the curve fitting method theoretically calculates the wavelength dependence of the absolute reflectance in the form of the spectral spectrum according to the conditions determined by the wavelength of light, the film thickness of the thin film, the refractive index (n) and extinction coefficient (k) of the thin film and the substrate. Then, the film thickness is determined by parameter fitting of the theoretically calculated absolute reflectance to the measurement data. Generally, the film thickness can be obtained by the curve fitting method from several nm to several μm. However, a sufficient number of sampling points is required for each wavelength of the illumination light used in the measurement.

[0008] A general spectrometer cannot measure the spatial distribution of the film thickness with high resolution and can only obtain the average value of the area illuminated by light. When there is a film thickness distribution (change) within the illumination area, the reflectance is averaged, so an accurate absolute reflectance cannot be obtained. Thus, the curve fitting method cannot measure the film thickness distribution with high resolution.

[0009] (2) The method by polarization analysis (ellipsometry) determines the film thickness by analyzing the reflected light. However, since it is necessary to obliquely incident the illumination light on the sample, the spatial distribution of the film thickness cannot be measured with high resolution and can only obtain the average value of the area illuminated by light (about 1 mm□ to 100 μm□). Furthermore, when the thin film itself has polarization characteristics, it becomes difficult to measure the film thickness with high resolution.

[0010] (3) The method using white interference fringe scanning calculates the maximum intensity position of the interference fringes of white light by performing a z-scan on the sample or the objective lens. Thereby, the film thickness of the thin film that generated the interference fringes can be obtained. In order to measure the film thickness of the thin film, it is necessary to separately measure the interference intensity signals from the underlying substrate and the thin film surface. However, when the film thickness is 1 μm or less, it becomes difficult to separate these intensity signals. Furthermore, when a transparent thin film with a low reflectivity exists on a substrate with a high reflectivity, it is impossible to precisely measure the interference intensity from the transparent thin film, and the measurement of the film thickness becomes even more difficult. Measuring the film thickness of multilayer films and the like becomes even more difficult.

[0011] (4) The method using a confocal microscope measures the film thickness by changing the distance between the objective lens and the sample through z-scanning and detecting the focus position. When the film thickness is several μm or less, the reflected signals from the underlying substrate and the film surface cannot be separated, so the film thickness cannot be measured. Since the signal separation limit greatly depends on the numerical aperture of the objective lens, the film thickness measurement limit is several μm at high magnification and several tens of μm at low magnification wide field of view, making it difficult to measure the film thickness.

[0012] (1) - (4) Since the spatial resolution of these methods is low, when the film thickness continuously increases or decreases from several nm to several μm within a region of several μm² to several hundred μm², it is difficult to evaluate the quality with the average value of arbitrary measurement points. Therefore, in order to increase the local spatial resolution of a sample with unevenness or warpage, for example, the (5) spectroscopic film thickness distribution measurement method as in Patent Document 1 is suitable.

[0013] (5) The spectroscopic film thickness distribution measurement method measures the film thickness by switching the light of multiple wavelengths using a confocal optical system. Specifically, the measurement data of the reflectivity for the first wavelength and the second wavelength are obtained respectively, and referring to the calculation data in which the relationship between the wavelength and the reflectivity is shown for each film thickness, the film thickness is approximately calculated from the measurement data. By applying the curve fitting method to each pixel from the images of multiple wavelengths, the film thickness can be calculated and the film thickness distribution can be obtained.

[0014] Incidentally, using a conventional spectrometer, the optical constants of materials such as dielectrics and metals can be obtained by applying an appropriate dispersion formula depending on the material. However, using an image sensor, it is difficult to analyze the optical constants of an unknown material from measurement data at several wavelengths.

[0015] Specifically, (5) the spectroscopic film thickness distribution measurement method calculates the film thickness and obtains the film thickness distribution by applying the curve fitting method to each pixel from images of multiple wavelengths. This method is effective for film thicknesses where the optical constants and film structure are known, but curve fitting becomes difficult when this information is insufficient. When calculating optical constants from reflectance using a dispersion formula, the reflectance over several wavelengths is insufficient information, and curve fitting is also difficult.

[0016] The object of the present invention is to solve these problems and to provide an optical constant determination device and optical constant determination method that can determine the optical constants of an unknown thin film. [Means for solving the problem]

[0017] The optical constant determination device according to this disclosure is an optical constant determination device for determining the optical constants of at least one of a thin film and a substrate of a sample including a substrate and a thin film provided on the substrate, comprising: a photodetector that detects reflected light reflected from the sample by illumination light that illuminates the sample via a confocal optical system, via the confocal optical system, and acquires images of the sample for each of multiple wavelengths of light contained in the reflected light; and a processing unit that acquires measurement data of each reflectance for each wavelength based on each image, wherein the processing unit receives a plurality of candidate materials having known optical constants selected as candidates for the materials of the thin film and the substrate, and calculates a coefficient of determination for approximating the film thickness from the measurement data by referring to calculation data in which the relationship between the wavelength and the reflectance for each candidate material is shown for each film thickness of the thin film, and determines the optical constants of at least one of the thin film and the substrate from among the optical constants of the plurality of candidate materials based on the calculated coefficient of determination.

[0018] In the optical constant determination device described above, the processing unit may receive input of a candidate number of layers selected as a candidate number of layers of films included in the thin film, input of a plurality of candidate materials having known optical constants selected as candidate materials for each film, and calculate the coefficient of determination used when approximating the film thickness from the measurement data by referring to the calculation data synthesized for each film, and determine the number of layers of the thin film from the plurality of candidate numbers of layers based on the calculated coefficient of determination.

[0019] In the optical constant determination device described above, the photodetector may acquire images of multiple minute regions of the sample, and the processing unit may map and display the optical constants, film thicknesses, and number of layers determined in the multiple analysis regions.

[0020] The optical constant determination device described above further includes a wavelength selection unit that selects the wavelength such that the illumination light illuminating the sample via the confocal optical system includes at least one of the wavelengths in the blue wavelength band, the green wavelength band, and the red wavelength band, and the photodetector may have a blue image sensor that acquires the image with light of the wavelength in the blue wavelength band, a green image sensor that acquires the image with light of the wavelength in the green wavelength band, and a red image sensor that acquires the image with light of the wavelength in the red wavelength band.

[0021] The optical constant determination device described above may further include an optical constant file in which the calculation data is stored, showing the relationship between the wavelength and the reflectance for each film thickness for a plurality of candidate materials having the known optical constants.

[0022] The optical constant determination method according to this disclosure is an optical constant determination method for determining the optical constants of at least one of a thin film and a substrate of a sample including a substrate and a thin film provided on the substrate, comprising: a focus composite image acquisition step of detecting reflected light reflected by the sample from illumination light that illuminates the sample via a confocal optical system and acquiring images of the sample for each of multiple wavelengths of light contained in the reflected light; a candidate material selection step of receiving input from a plurality of candidate materials having known optical constants selected as candidates for the materials of the thin film and the substrate; a film thickness analysis step of acquiring measurement data of each reflectance for each wavelength based on each image; a coefficient of determination calculation step of calculating a coefficient of determination for approximating the film thickness from the measurement data by referring to calculation data that shows the relationship between the wavelength and the reflectance for each candidate material for each film thickness of the thin film; and an optical constant determination step of determining the optical constants of at least one of the thin film and the substrate from among the optical constants of the plurality of candidate materials based on the calculated coefficient of determination.

[0023] The above optical constant determination method further includes a layer count selection step in which a candidate number of layers selected as a candidate number of layers of films to be included in the thin film is input, a candidate material selection step in which a plurality of candidate materials having the known optical constants selected as the candidate material for each film is input, a coefficient of determination calculation step in which the coefficient of determination used when approximating and calculating the film thickness from the measurement data is calculated by referring to the calculation data synthesized for each film, and in the optical constant determination step in which the number of layers of the thin film is determined from the plurality of candidate layer counts based on the calculated coefficient of determination.

[0024] The above optical constant determination method further includes a micro-region setting step of setting a micro-region after the focus-combined image acquisition step, and a mapping display step of mapping and displaying each optical constant, each film thickness, and each number of layers determined in a plurality of the micro-regions after the optical constant determination step, and in the focus-combined image acquisition step, each image including a plurality of the micro-regions of the sample may be acquired.

[0025] In the optical constant determination method described above, in the focus stacking image acquisition step, the wavelength may be selected such that the illumination light used to illuminate the sample via the confocal optical system includes at least one of the wavelengths in the blue wavelength band, the wavelength in the green wavelength band, and the wavelength in the red wavelength band, and the image obtained using the light of the wavelength in the blue wavelength band, the image obtained using the light of the wavelength in the green wavelength band, and the image obtained using the light of the wavelength in the red wavelength band.

[0026] In the above optical constant determination method, in the candidate material selection step, an optical constant file may be prepared in which the calculation data showing the relationship between the wavelength and the reflectance for each film thickness is stored for a plurality of candidate materials having the known optical constants. [Effects of the Invention]

[0027] According to the present invention, it is possible to provide an optical constant determination device and an optical constant determination method that can determine the optical constants of an unknown thin film. [Brief explanation of the drawing]

[0028] [Figure 1] This figure illustrates the configuration of the optical constant determination device according to Embodiment 1. [Figure 2] This graph illustrates the spectral characteristics of a photodetector in the optical constant determination device according to Embodiment 1, with the horizontal axis representing wavelength and the vertical axis representing transmittance. [Figure 3] This graph illustrates the transmission wavelength range of the bandpass filter according to Embodiment 1, with the horizontal axis representing wavelength and the vertical axis representing transmittance. [Figure 4] This figure illustrates the transmission wavelength of the bandpass filter according to Embodiment 1. [Figure 5] This graph illustrates the transmission wavelength range of another bandpass filter according to Embodiment 1, with the horizontal axis representing wavelength and the vertical axis representing transmittance. [Figure 6]This figure illustrates the transmission wavelengths of other bandpass filters according to Embodiment 1. [Figure 7] This is a cross-sectional view illustrating the analysis region of a sample according to Embodiment 1. [Figure 8] This is a cross-sectional view illustrating a sample containing a thin film comprising two layers according to Embodiment 1. [Figure 9] This figure illustrates the display of the calculation process for film thickness and coefficient of determination in the optical constant determination device according to Embodiment 1. [Figure 10] This figure illustrates a list of coefficients of determination arranged in descending order in the optical constant determination device according to Embodiment 1. [Figure 11] This is a cross-sectional view illustrating the material and number of layers in the thin film of a sample in the optical constant determination device according to Embodiment 1. [Figure 12] This is a plan view illustrating the material and number of layers in the thin film of the sample in the optical constant determination device according to Embodiment 1. [Figure 13] This is a plan view illustrating the integrated film thickness of a sample thin film in grayscale in the optical constant determination device according to Embodiment 1. [Figure 14] This is a grayscale plan view illustrating the number of layers of a thin film in a sample in the optical constant determination device according to Embodiment 1. [Figure 15] This is a flowchart illustrating the optical constant determination method according to Embodiment 1. [Figure 16] This flowchart illustrates another method for determining optical constants according to Embodiment 1. [Figure 17] This flowchart illustrates yet another method for determining optical constants according to Embodiment 1. [Figure 18] This is an optical model illustrating a substrate on which a thin film according to Embodiment 1 is formed. [Figure 19] This is an optical model illustrating a substrate on which a thin film according to Embodiment 1 is formed. [Figure 20] This figure illustrates the optical path difference due to the refraction of reflected light on a substrate on which a thin film according to Embodiment 1 is formed. [Figure 21]This figure illustrates interference-free multiple reflections within a thin film according to Embodiment 1. [Figure 22] This graph illustrates the relationship between wavelength and reflectance according to Embodiment 1, with the horizontal axis representing wavelength and the vertical axis representing absolute reflectance. [Figure 23] This graph illustrates the relationship between wavelength and reflectance according to Embodiment 1, with the horizontal axis representing wavelength and the vertical axis representing absolute reflectance. [Figure 24] This graph illustrates the relationship between wavelength and reflectance according to Embodiment 1, with the horizontal axis representing wavelength and the vertical axis representing absolute reflectance. [Figure 25] This graph illustrates the relationship between the film thickness and residual of a thin film according to Embodiment 1, with the horizontal axis representing the film thickness and the vertical axis representing the residual. [Figure 26] This is a bird's-eye view illustrating the film thickness of the thin film according to Embodiment 1. [Figure 27] This figure illustrates how to handle the calculation of reflectance when the film thickness is zero in the case of complete non-interference according to Embodiment 1. [Modes for carrying out the invention]

[0029] The specific configuration of this embodiment will be described below with reference to the drawings. The following description illustrates preferred embodiments of the present invention, and the scope of the present invention is not limited to these embodiments. In the following description, the same reference numerals indicate substantially the same function.

[0030] (Embodiment 1) The optical constant determination device and optical constant determination method of this embodiment utilize a reflectance spectroscopy (spectral interference) method for measuring the spectral film thickness distribution as its basic principle. In this embodiment, a large amount of data on optical constants and thin film structures is prepared in advance. By incorporating these into the fitting parameters, the optical constants of an unknown thin film are determined and the film thickness is analyzed. For this purpose, we consider a reflectance spectroscopy method for measuring the spectral film thickness distribution using the following eight points. (1) Search for the optimal optical constant file and perform film thickness analysis. For samples with the same thin film structure, the optical constants determined by the search can be used to analyze the two-dimensional film thickness distribution. (2) The optical constant file contains material information, so it is possible to map the material distribution of the thin film simultaneously with the film thickness distribution. It can handle cases where multiple materials such as insulating films and metal films are present on the sample. (3) Even if there are areas on the sample surface with different thin film structures, the film thickness distribution and the number of layers can be mapped simultaneously. For example, even if an SiO2 film covers the entire surface of a Si wafer and a photoresist film is partially patterned, the film thickness, number of layers, and film type can be mapped. (4) By using a confocal optical system, it is possible to remove the back surface reflection of the transparent substrate, and it is expected that the net surface reflection can be measured accurately. (5) Instead of spectrally analyzing reflected light as in conventional spectrophotometers, the reflectance of the sample is measured using monochromatic illumination. The reflectance is then measured for the same field of view each time the illumination wavelength is changed. Therefore, the wavelength dependence of the reflectance, i.e., the reflectance spectrum, can be obtained. By imaging at each wavelength, the reflectance can be visualized in two dimensions. (6) By using a confocal optical system and performing a z-scan, the surface shape can be measured in three dimensions, although with an error of the thickness of the film. If the substrate has irregularities or warping, that information can be measured simultaneously. (7) By using a confocal optical system and performing focus stacking by z-scan, a full-focus image can be created, enabling measurements that focus on the entire sample surface. Therefore, the influence of measurements such as film pattern edges can be reduced without degrading spatial resolution. (8) Using a confocal optical system, a large numerical aperture (NA) objective lens can be used to measure the film thickness distribution of fine patterns, allowing for the acquisition of reflected images even with pattern widths of several micrometers.

[0031] From (1) to (8), a spectral spectrum can be obtained for each pixel of the observed image. By converting this to absolute reflectance and performing a fitting with the film thickness and optical model (optical constants and film structure) as parameters, the film thickness of an unknown material can be determined. The wavelength can be, for example, several wavelengths between 400 and 700 nm.

[0032] In a confocal optical system, changing the observation wavelength can alter the focal point, leading to expected changes in brightness. This can be canceled out by pre-memorizing the focal point shift for each wavelength in a PC and automatically adjusting the z-position of the sample or objective lens by the corresponding shift for each wavelength. Alternatively, a full-focus image can be created for each wavelength using a z-scan.

[0033] By pre-measuring samples with known reflectance spectra, such as silicon or quartz glass, as reference samples for absolute reflectance, the wavelength dependence of the observation optical system itself can be corrected by calculation.

[0034] Even if the optical constants and the layered structure of the thin film are unknown, the thickness of the thin film can be measured using the curve fitting method. Even if the thin film on the sample surface is composed of multiple types of thin films, the film type distribution and the layer number distribution can be determined simultaneously by performing a thickness distribution analysis while determining the film type.

[0035] By using a confocal microscope and measuring the reflectance intensity while switching the illumination wavelength, the film thickness distribution on a substrate can be measured. Even with significant variations in film thickness distribution, measurements can be performed non-contact, non-destructively, and quickly. It can also handle film thickness measurements on surfaces with steps greater than the depth of field of an optical microscope. A confocal optical system is suitable for increasing spatial resolution.

[0036] The optical constant determination apparatus and optical constant determination method according to Embodiment 1 will be described in detail below. First, the configuration of the optical constant determination apparatus will be described. Then, the optical constant determination method will be described, followed by the film thickness analysis.

[0037] [Configuration of the optical constant determination device] Figure 1 is a diagram illustrating the configuration of an optical constant determination device according to Embodiment 1. As shown in Figure 1, the optical constant determination device 100 according to this embodiment includes a light source 11, a wavelength selection unit 12, lenses 13a, 13b, 13c, a slit 14, a beam splitter 15, a vibrating mirror 16, an objective lens 17, a stage 18, a photodetector 19, and a processing unit 20. The light source 11 and the wavelength selection unit 12 constitute the light source unit 10. The lenses 13a, 13b, 13c, the slit 14, the beam splitter 15, the vibrating mirror 16, and the photodetector 19 constitute the confocal optical system 101. A sample 30 is placed on the stage 18. The optical constant determination device 100 determines the optical constants of the thin film 32 of the sample 30, which includes a substrate 31 and a thin film 32 provided on the substrate 31. The optical constants include the refractive index and extinction coefficient of the material of the thin film 32. The optical constant determination device 100 also measures the film thickness of the thin film 32. Furthermore, if the thin film 32 is absent, the optical constant determination device 100 can determine the optical constants of the substrate 31.

[0038] The optical constant determination device 100 may include an optical constant file 21. The optical constant file 21 is, for example, a file listing optical constants for each 1 nm increment for wavelengths from 400 nm to 700 nm. The optical constants stored in the optical constant file 21 may be measured values, theoretically calculated values, or values ​​described in literature. The optical constant file 21 may also store information such as the material name, material type (metal, glass, etc.), and the relationship between wavelength and reflectance for each film thickness.

[0039] Furthermore, the optical constant file 21 is not limited to one provided in the optical constant determination device 100; it may also be stored in an external storage device or obtained from the cloud via the internet. Additionally, the wavelength is not limited to 400nm to 700nm; it may be less than 400nm or greater than 700nm.

[0040] As the light source 11, a white light source containing multiple emission lines in a continuous spectrum, such as a mercury lamp or a xenon lamp, is used. For example, a xenon lamp with a broad continuous spectrum from ultraviolet to infrared (185 nm to 2000 nm) may be used. Of course, the light source 11 is not limited to a xenon lamp; a white diode, white laser, etc., may also be used. As will be described later, any light source can be used as long as the wavelength can be selected.

[0041] The optical system for observing the sample 30 using light from the light source 11 will now be described. The light emitted from the light source 11 passes through the wavelength selection unit 12 and is converted into illumination light containing light of a specific wavelength.

[0042] The wavelength selection unit 12 selects the wavelength of illumination light that illuminates the sample 30 via the confocal optical system 101. The wavelength selection unit 12 is, for example, a bandpass filter BPF. The bandpass filter BPF may be a multibandpass filter that transmits multiple wavelength bands, or a bandpass filter BPF that transmits a single wavelength band. In this specification, the term "bandpass filter BPF" includes both multibandpass filters and bandpass filters. The relationship between the bandpass filter BPF and the photodetector 19 will be described later.

[0043] The wavelength selection unit 12 may use multiple bandpass filters (BPFs) that selectively transmit light of specific wavelengths. This allows for the selective transmission of multiple single-wavelength illumination lights. For example, the wavelengths of the illumination light can be selected from 405nm, 436nm, 488nm, 546nm, 578nm (wavelengths corresponding to the emission lines of a mercury xenon lamp), and 514nm and 633nm (wavelengths not corresponding to the emission lines of a mercury xenon lamp). When using a mercury xenon lamp, it is also possible to select light of wavelengths other than those corresponding to the emission lines using the bandpass filters (BPFs). Since light of wavelengths other than those corresponding to the emission lines has low intensity, this can be balanced by widening the full width at half maximum (FWHM) of the bandpass filters (BPFs). The wavelength selection unit 12 may switch between multiple wavelengths. The wavelengths may be switched continuously or intermittently. For example, 5 to 7 wavelengths between 400nm and 650nm may be selected.

[0044] Furthermore, a laser light source that emits single-wavelength laser light may be used as the light source 11, and a wavelength conversion element may be provided. For example, the wavelength of the single-wavelength light incident on the wavelength conversion element can be converted by generating a second harmonic. It is also possible to use a tunable wavelength laser as the light source 11. In addition, multiple laser light sources that emit laser light of different wavelengths may be provided, and light of a desired wavelength may be selected from among the multiple laser light sources.

[0045] The confocal optical system 101 guides illumination light from the light source 10 to the sample 30 and the reflected light from the sample 30 to the photodetector 19. The illumination light that has passed through the wavelength selection unit 12 passes through the lens 13a and enters the slit 14. The illumination light is shaped into a line in the X direction through the slit 14. The line-shaped illumination light then enters the beam splitter 15. The beam splitter 15 splits the light so that the amount of reflected light and transmitted light is approximately 1:1, regardless of the polarization state. Therefore, approximately half of the illumination light passes through the beam splitter 15.

[0046] Subsequently, the light traveling to the right in Figure 1 is incident on the vibrating mirror 16. The vibrating mirror 16 scans the sample 30 in the Y direction with a line of illumination light in the X direction. This allows scanning of the sample 30 surface in both X and Y directions. For example, a galvanometer mirror, a polygon mirror, etc., can be used as the vibrating mirror 16.

[0047] The illumination light reflected downward by the vibrating mirror 16 is focused by the objective lens 17 and irradiated onto the sample 30. The sample 30 is placed on the stage 18. The reflected light from the sample 30 then passes through the objective lens 17 again, is reflected again by the vibrating mirror 16, and enters the beam splitter 15. Approximately half of the incident light is then reflected by the beam splitter 15 and enters the lens 13c. The lens 13c images the combined light onto the light-receiving surface of the photodetector 19. The light that has passed through the lens 13c is received by the photodetector 19.

[0048] The photodetector 19 detects the reflected light reflected from the sample 30 and acquires an image of the measurement area of ​​the sample 30. The photodetector 19 may also be a 3CCD (Charged Coupled Device) line sensor that captures a confocal image of the sample 30. For example, the photodetector 19 may have three image sensors: a blue image sensor, a green image sensor, and a red image sensor. The blue image sensor acquires an image using light with wavelengths in the blue wavelength band. The green image sensor acquires an image using light with wavelengths in the green wavelength band. The red image sensor acquires an image using light with wavelengths in the red wavelength band.

[0049] In this way, the photodetector 19 detects the reflected light from the sample 30, which is illuminated by the confocal optical system 101, via the confocal optical system 101, and acquires images of the sample 30 using multiple wavelengths of light contained in the reflected light. Note that the scanning method may differ as long as the confocal optical system 101 is used, and the slit 14 and photodetector 19 can be appropriately used in accordance with the method. For example, an acousto-optic element (AOD) can be used instead of the vibrating mirror 16. Also, instead of the combination of the slit 14 and vibrating mirror 16 when a line sensor is used for the photodetector 19, the combination of XY 2D scanning when a point sensor is used for the photodetector 19 may be applied.

[0050] Stage 18 has a Z-axis drive motor (not shown) that can move the sample 30 in the Z direction. By moving Stage 18 in the Z direction, the sample surface is controlled to be at the focal point. Alternatively, instead of moving Stage 18 in the Z direction, the focal position can be adjusted by moving the objective lens 17. Focus stacking using a variable focus lens or the like may also be used.

[0051] In the confocal optical system 101, the focal point position may change depending on the wavelength being observed, and a change in brightness is expected as a result. In this case, the shift in the focal point position for each wavelength can be measured in advance and stored in the processing unit 20, and the shift can be automatically corrected by fine-tuning the Z position of the sample 30 or objective lens 17 by the amount of the shift for each wavelength. Alternatively, a full-focus image may be created for each wavelength by Z-scanning. The Z-scan range may be widened to cover the focal points for all wavelengths. The wavelength dependence of the observation optical system itself can be corrected by calculation by pre-measuring samples with known reflection spectra, such as silicon or quartz glass.

[0052] The optical constant determination device 100 of this embodiment can acquire full-focus reflection images of the warping of the substrate 31 and the uneven surface of the substrate 31 by using a confocal optical system 101. The optical constant determination device 100 may be configured to acquire multiple images in a single focus stacking by combining a color confocal optical system 101 composed of a wavelength selection unit 12 such as a bandpass filter BPF that transmits multiple wavelengths from white light generated from a light source 11 and a photodetector 19 such as a 3CCD camera.

[0053] Next, we will describe the combination of the photodetector 19 and the wavelength selection unit 12 when the wavelength selection unit 12 includes a multibandpass filter. First, we will describe the spectral characteristics of the photodetector 19. Figure 2 is a graph illustrating the spectral characteristics of the photodetector 19 in the optical constant determination device 100 according to Embodiment 1, where the horizontal axis represents wavelength and the vertical axis represents transmittance. As shown in Figure 2, the photodetector 19 is designed to have spectral characteristics that detect reflected light transmitted through a visible light region divided into three parts. For example, the photodetector 19 has a blue image sensor, a green image sensor, and a red image sensor corresponding to channel B-ch for detecting light in the blue wavelength band, channel G-ch for detecting light in the green wavelength band, and channel R-ch for detecting light in the red wavelength band, respectively. Note that one of the blue wavelength band, green wavelength band, and red wavelength band may be referred to as the first wavelength band and the second wavelength band. By using the signals received by these three image sensors, a color image of the entire visible light can be obtained. Note that the image obtained by detecting illumination light including visible light is used as a color review image.

[0054] Next, the transmission wavelength band of the bandpass filter BPF will be described as an example of the wavelength selection unit 12. Figure 3 is a graph illustrating the transmission wavelength band of the bandpass filter BPF according to Embodiment 1, where the horizontal axis represents wavelength and the vertical axis represents transmittance. Figure 4 is a diagram illustrating the transmission wavelength of the bandpass filter BPF according to Embodiment 1. As shown in Figures 3 and 4, each bandpass filter BPF-1 to BPF-5 transmits a wavelength band that includes a specific wavelength. The specific wavelength may be called the first wavelength, second wavelength, and third wavelength. Specifically, the wavelength selection unit 12 may select the first wavelength, second wavelength, and third wavelength from one of the wavelengths in the blue wavelength band, the green wavelength band, and the red wavelength band, respectively. Each bandpass filter BPF-1 to BPF-5 may transmit a wavelength band with a specific wavelength as the center wavelength, or it may transmit a wavelength band that has a peak at a specific wavelength.

[0055] For example, bandpass filter BPF-1 transmits illumination light containing three wavelengths: b1, g1, and r1. Bandpass filter BPF-2 transmits illumination light containing three wavelengths: b2, g2, and r2. Bandpass filter BPF-3 transmits illumination light containing three wavelengths: b3, g3, and r3. Bandpass filter BPF-4 transmits illumination light containing two wavelengths: c4 and r4. Bandpass filter BPF-5 transmits illumination light containing wavelength y5.

[0056] Wavelengths b1, b2, and b3 are included in the wavelength band of channel B-ch. Wavelengths g1, g2, and g3 are included in the wavelength band of channel G-ch. Wavelengths r1, r2, r3, and r4 are included in the wavelength band of channel R-ch. Wavelength c4 is included in the wavelength band between channel B-ch and channel G-ch (e.g., the cyan wavelength band). Wavelength y5 is included in the wavelength band between channel G-ch and channel R-ch (e.g., the yellow wavelength band). For wavelength bands between different channels, the signal obtained by summing the signals from both channels is used.

[0057] Thus, the wavelength selection unit 12, such as the bandpass filter BPF, selects wavelengths such that the illumination light illuminating the sample 30 via the confocal optical system 101 includes at least a first wavelength in the first wavelength band and a second wavelength in a second wavelength band different from the first wavelength band. The photodetector 19 then detects the reflected light reflected from the sample 30 via the confocal optical system 101 and acquires images of the measurement region of the sample 30 for each wavelength of light contained in the reflected light. Therefore, by performing five focus stacking operations using bandpass filters BPF-1 to BPF-5, images of 12 wavelengths can be obtained.

[0058] Figure 5 is a graph illustrating the transmission wavelength bands of other bandpass filters BPF according to Embodiment 1, where the horizontal axis represents wavelength and the vertical axis represents transmittance. Figure 6 is a diagram illustrating the transmission wavelengths of other bandpass filters BPF according to Embodiment 1. As shown in Figures 5 and 6, bandpass filters BPF-1 to BPF-3 and BPF-6 to BPF-8 each transmit illumination light of specific wavelengths. Bandpass filters BPF-1 to BPF-3, as described above, transmit illumination light including three wavelengths b1, g1, and r1, three wavelengths b2, g2, and r2, and three wavelengths b3, g3, and r3, respectively.

[0059] Bandpass filter BPF-6 transmits illumination light containing two wavelengths, v1 and y1. Bandpass filter BPF-7 transmits illumination light containing two wavelengths, c1 and r4. Bandpass filter BPF-8 transmits illumination light containing two wavelengths, c2 and r5.

[0060] Wavelengths v1, b1, b2, and b3 are included in the wavelength band of channel B-ch. Wavelengths g1, g2, and g3 are included in the wavelength band of channel G-ch. Wavelengths r1, r2, r3, r4, and r5 are included in the wavelength band of channel R-ch. Wavelengths c1 and c2 are included in the wavelength band between channel B-ch and channel G-ch. Wavelength y1 is included in the wavelength band between channel G-ch and channel R-ch.

[0061] By performing six focus stacking operations using bandpass filters BPF-1 to BPF-3 and BPF-6 to BPF-8, images at 15 different wavelengths can be obtained.

[0062] The processing unit 20 receives each image of each wavelength acquired by the photodetector 19 as input. Based on each input image, measurement data for reflectance at each wavelength is acquired. The processing unit 20 converts the brightness (gradation level) of the acquired images at each wavelength into absolute reflectance (R) using a previously acquired reflectance reference (sample with known reflectance) and dark luminance (unlit state). The processing unit 20 calculates the absolute reflectance (R) value for each wavelength (λ) for all pixels in the field of view.

[0063] The optical constant determination device 100 of this embodiment determines the optical constants of an unknown thin film 32 on the substrate 31. Furthermore, as described above, if there is no thin film 32, the optical constant determination device 100 can determine the optical constants of the unknown substrate 31. Figure 7 is a cross-sectional view illustrating the analysis area of ​​the sample 30 according to Embodiment 1. As shown in Figure 7, the material and number of layers of the thin film 32 of the sample 30 are unknown. The processing unit 20 first determines the optical constants using the average value of the 100 × 100 pixel analysis area 40 in the center of the acquired field of view V image.

[0064] In the optical constant determination device 100, in order to determine the coefficient of determination of an unknown thin film 32, the processing unit 20 first receives input from a plurality of candidate materials having known optical constants that have been selected as candidates for the material of the thin film 32. The known optical constants of the candidate materials are stored, for example, in the optical constant file 21. Thus, the optical constant file 21 stores calculation data showing the relationship between wavelength and reflectance for each film thickness for a plurality of candidate materials having known optical constants. The processing unit 20 may select the optical constants of a plurality of candidate materials by selecting from the optical constant file 21. The processing unit 20 may also receive input from a candidate number of layers selected as a candidate for the number of layers of films included in the thin film 32. If there is no thin film 32, in order to determine the coefficient of determination of an unknown substrate 31, the processing unit 20 receives input from a plurality of candidate materials having known optical constants that have been selected as candidates for the material of the substrate 31. In this embodiment, a plurality of candidate materials and candidate number of layers are selected for a sample 30 whose material and number of layers are unknown.

[0065] First, we will explain the case where there is one thin film 32 on the substrate 31 as an optical model. We will assume that the optical constants of the substrate 31 are known. Using the optical constants of candidate materials for the single thin film 32, the processing unit 20 can calculate and obtain the theoretical absolute reflectance of the optical model from the film thickness analysis described later, that is, from calculations using the Fresnel coefficient.

[0066] The processing unit 20 calculates and obtains the absolute reflectance by changing the film thickness from 0 nm to 1000 nm in 1 nm increments while keeping the optical constants fixed. The processing unit 20 then calculates and obtains the absolute reflectance for multiple candidate materials of the thin film 32, successively swapping them out. The processing unit 20 then approximates and calculates the film thickness from the measurement data by referring to the calculation data that shows the relationship between wavelength and reflectance for each film thickness for each candidate material obtained in this way. For example, the processing unit 20 performs fitting using the least squares method to obtain the most probable value of the film thickness of the thin film 32. The processing unit 20 also calculates the coefficient of determination when approximating and calculating the film thickness from the measurement data by referring to the calculation data.

[0067] The coefficient of determination (R) is a measure of how well a model fits when using the least squares method to fit film thickness as a parameter based on reflectance measurements at each wavelength. The maximum value of the R is 1. The closer the R is to 1, the better the fit is considered to be. Generally, the R is used to express the R value. 2 It is expressed as follows. In this embodiment, the coefficient of determination is D in order to distinguish it from the reflectance R. 2 It can sometimes be expressed as follows. The coefficient of determination can be, for example, the value obtained by subtracting from 1 the sum of the squares of the residuals divided by the sum of the squares of the deviations from the mean when fitting, but it is not limited to this as long as it shows a measure of how well the model fits.

[0068] The processing unit 20 generates a list of coefficients of determination sorted in descending order. In the list, candidate materials and film thicknesses are associated with the coefficients of determination. Therefore, a list is obtained in which candidate materials and film thicknesses are sorted in descending order of their coefficients of determination. Based on the calculated coefficients of determination, the processing unit 20 determines the optical constants of the thin film 32 from among the optical constants of multiple candidate materials. For example, the processing unit 20 determines the optical constants of the thin film 32 based on the list sorted in descending order of coefficients of determination. In this way, by referring to this list, the processing unit 20 can estimate the material of the thin film 32 and obtain reasonable optical constants. The processing unit 20 can determine the distribution of the final determined optical constants and film thicknesses by performing film thickness analysis on all pixels in the field of view. If there is no thin film 32 in the sample 30, in order to determine the optical constants of the substrate 31, the processing unit 20 calculates the coefficient of determination by performing the above least squares method fitting with the film thickness of the thin film 32 set to 0. Then, based on the calculated coefficients of determination, the processing unit 20 determines the optical constants of the substrate 31 from among the optical constants of multiple candidate materials.

[0069] Next, we will describe the case of a thin film 32 containing two or more layers of film as an optical model. Figure 8 is a cross-sectional view illustrating a sample 30 containing a thin film 32 with two layers of film according to Embodiment 1. Figure 9 is a diagram illustrating the display of the calculation process for film thickness and coefficient of determination in the optical constant determination device 100 according to Embodiment 1.

[0070] As shown in Figures 8 and 9, in the case of a thin film 32 containing two or more layers of films L1 and L2, the processing unit 20 may independently set optical constant files 21 for each of the films L1 and L2 and perform analysis of the multilayer film. Specifically, the processing unit 20 is input with a candidate number of layers selected as a candidate number of layers of films contained in the thin film 32. The processing unit 20 is then input with a plurality of candidate materials having known optical constants selected as candidate materials for each film. The processing unit 20 approximates and calculates the film thickness from the measurement data by referring to the synthesized calculation data for each film. Even in the case of a multilayer film, the reflectance calculation data can be obtained by synthesizing the Fresnel coefficient from the substrate 31 side. The processing unit 20 calculates the coefficient of determination for each film when approximating and calculating the film thickness.

[0071] For example, as shown in Figure 9, the measurement range (analysis area 40) is set to a width of 100 nm and a height of 100 nm from the 461 coordinate system, with both the starting point X and starting point Y within the field of view V. The X and Y directions refer to the width and height directions. The processing unit 20 analyzes the average reflectance of the analysis area 40, for example, if the analysis area 40 is uniform. The processing unit 20 then approximates and calculates film thicknesses of 92 nm and 102 nm from the measurement data, using calculation data with candidate materials such as silicon nitride (Si3N4) and silicon oxide (SiO2), respectively, as film L1 and film L2 of the thin film 32. When calculating the film thickness, the processing unit 20 calculates a coefficient of determination of 0.997537. The processing unit 20 successively changes multiple candidate materials and candidate number of layers for the thin film 32, calculates the film thickness, and calculates the coefficient of determination when calculating the film thickness.

[0072] Figure 10 is an example of a list arranged in descending order of the coefficient of determination in the optical constant determination device 100 according to Embodiment 1. The processing unit 20 arranges the candidate materials and candidate number of layers in descending order of the calculated coefficient of determination. In this way, the processing unit 20 can obtain a list in which the combinations of optical constants of each film are arranged in order of the coefficient of determination. Then, based on the list arranged in descending order of the coefficient of determination, the processing unit 20 determines the optical constant of the thin film from among the optical constants of multiple candidate materials. Thus, the material name of the thin film 32 of sample 30 can be obtained. At the same time, the processing unit 20 can obtain the number of layers of the thin film 32 of sample 30 and the thickness of each film contained in the thin film 32. If the material obtained as a result of the processing can be rejected from other information, a second candidate may be selected, or the rejected optical constant may be excluded from the list and re-analyzed.

[0073] The processing unit 20 may determine the final optical constants and film thickness distribution by performing film thickness analysis on all pixels within the field of view V. Furthermore, the photodetector 19 may acquire images of each wavelength of light contained in the reflected light for multiple other fields of view V of the sample 30, and the processing unit 20 may determine the optical constants and film thickness distribution for each field of view V.

[0074] Figure 11 is a cross-sectional view illustrating the material and number of layers of the thin film 32 of the sample 30 in the optical constant determination device 100 according to Embodiment 1. Figure 12 is a plan view illustrating the material and number of layers of the thin film 32 of the sample 30 in the optical constant determination device 100 according to Embodiment 1. Figure 13 is a plan view illustrating the accumulated film thickness of the thin film 32 of the sample 30 in grayscale in the optical constant determination device 100 according to Embodiment 1. Figure 14 is a plan view illustrating the number of layers of the thin film 32 of the sample 30 in grayscale in the optical constant determination device 100 according to Embodiment 1.

[0075] As shown in Figures 11 and 12, the processing unit 20 may map and display each optical constant (material having optical constants), each film thickness, and each number of layers determined for each minute region, such as a pixel, within a predetermined range such as the entire surface of the sample 30, within the field of view V, or multiple analysis regions 40. In this way, the processing unit 20 can not only read optical constants from an optical constant file, but also obtain information about the film structure, such as which layer a particular optical constant material is in. Furthermore, the processing unit 20 determines the film thickness and number of layers from the combination of optical constants that has the closest coefficient of determination to 1 among all combinations of optical constants, including the case of one layer and the case of two layers. This allows the processing unit 20 to determine the film thickness and number of layers even when it is unknown whether the thin film on the substrate 31 is one layer or two layers thick. Also, as shown in Figures 13 and 14, the processing unit 20 can display the film thickness and number of layers of the thin film 32 of the sample 30 in grayscale.

[0076] The processing unit 20 may be, for example, an information processing device such as a server or a personal computer. The information processing device has a control unit, a communication unit, a storage unit, and an interface unit (not shown). The control unit, communication unit, storage unit, and interface unit each have the functions of a control means, a communication means, a storage means, and an interface means, respectively.

[0077] The control unit includes, for example, a processor such as a CPU (Central Processing Unit), MPU (Micro Processing Unit), ECU (Electronic Control Unit), FPGA (Field-Programmable Gate Array), or ASIC (Application Specific Integrated Circuit). The control unit functions as a computing device that performs control processing and arithmetic processing. The control unit also controls the operation of the communication unit, storage unit, interface unit, and each component necessary for executing the functions of each device.

[0078] Each component of an information processing device can be realized, for example, by executing a program controlled by a control unit. More specifically, each component can be realized by the control unit executing a program stored in a memory unit. Alternatively, each component may be realized by recording the necessary programs on any non-volatile recording medium and installing them as needed. Furthermore, each component is not limited to being realized by software programs, but may also be realized by any combination of hardware, firmware, and software.

[0079] The communication unit performs the necessary communications for the information processing device to perform information processing. The storage unit is, for example, ROM (Read Only Memory) or RAM (Random Access Memory). The storage unit has the function of storing control programs and arithmetic programs executed by the control unit. The storage unit also has the function of temporarily storing processing data.

[0080] The interface unit is, for example, a user interface. The interface unit is connected to input means such as a keyboard, touch panel, or mouse, and output means such as a display or speaker. The interface unit accepts data input operations from the user (operator, etc.) and outputs information to the user.

[0081] [Optical constant determination method] Next, the optical constant determination method will be described. The optical constant determination method determines the optical constants of the thin film 32 and the substrate 31 of a sample 30 which includes a substrate 31 and a thin film 32 provided on the substrate 31. The optical constant determination method also calculates the film thickness of the thin film 32. Figure 15 is a flowchart illustrating the optical constant determination method according to Embodiment 1. As shown in Figure 15, the optical constant determination method comprises a focus stacking image acquisition step (S11), a candidate material selection step (S12), a film thickness analysis step (S13), a coefficient of determination calculation step (S14), and an optical constant determination step (S15).

[0082] First, in the focus-combined image acquisition step (S11), the photodetector 19 detects the reflected light from the sample 30, which was illuminated by the confocal optical system 101, via the confocal optical system 101. The photodetector 19 then acquires images of the sample 30 using multiple wavelengths of light contained in the detected reflected light.

[0083] Next, in the candidate material selection step (S12), a plurality of candidate materials having known optical constants are selected as candidates for the materials of the thin film 32 and the substrate 31. For example, a plurality of optical constant files containing candidate materials may be selected. In this way, the processing unit 20 receives input of a plurality of candidate materials having known optical constants that have been selected as candidates for the materials of the thin film 32 and the substrate 31.

[0084] Next, in the film thickness analysis step (S13), the processing unit 20 acquires measurement data of the reflectance for each wavelength based on each image of the sample 30 taken with light of multiple wavelengths.

[0085] Next, in the coefficient of determination calculation step (S14), the processing unit 20 approximates the thickness of the thin film from the measurement data by referring to the calculation data which shows the relationship between wavelength and reflectance for each thin film thickness for each candidate material. Then, the processing unit 20 calculates the coefficient of determination used when approximating the thickness of the film. If there is no thin film 32, the thickness of the thin film 32 is set to 0 for the calculation.

[0086] Next, in the optical constant determination step (S15), the processing unit 20 determines the optical constants of the thin film 32 and the substrate 31 from among the optical constants of multiple candidate materials based on the calculated coefficient of determination. For example, the processing unit 20 determines the optical constants of the thin film 32 and the substrate 31 from among the optical constants of multiple candidate materials based on a list of optical constants arranged in descending order of coefficient of determination. In this way, the optical constants of the thin film 32 and the substrate 31 of the sample 30 can be determined.

[0087] Next, another method for determining optical constants will be described. Figure 16 is a flowchart illustrating another method for determining optical constants according to Embodiment 1. As shown in Figure 16, the other method for determining optical constants further comprises a step of selecting the number of layers (S11a), a step of deciding whether to select the next candidate material (S14a), and a step of deciding whether to select the next number of layers (S14b).

[0088] The layer count selection step (S11a) is performed after the focus stacking image acquisition step (S11) and before the candidate material selection step (S12). In the layer count selection step (S11a), candidate numbers of layers of films to be included in the thin film 32 are selected. Therefore, the processing unit 20 is input the candidate number of layers selected as the candidate number of layers of films to be included in the thin film 32. In this case, in the candidate material selection step (S12), the processing unit 20 is input a plurality of candidate materials having known optical constants that have been selected as candidate materials for each film. In the coefficient of determination calculation step (S14), the processing unit 20 calculates the coefficient of determination when approximating and calculating the film thickness from the measurement data by referring to the synthesized calculation data for each film.

[0089] After step S14, determine whether to select the next candidate material, as shown in step S14a. If the answer in step S14a is YES (selecting the next candidate material), return to step S12 and repeat steps S12 to S14a. If the answer in step S14a is NO (not selecting the next candidate material), determine whether to select the next number of layers, as shown in step S14b.

[0090] In step S14b, if the next number of layers is selected (YES), the process returns to step S11a and steps S11a to S14b are repeated. In step S14b, if the next number of layers is not selected (NO), in step S15, the processing unit 20 determines the number of layers of the thin film 32 from among several candidate numbers of layers, in addition to the optical constants, based on the calculated coefficient of determination. In this way, the number of layers of the thin film 32 of the sample 30 can be determined.

[0091] Figure 17 is a flowchart illustrating yet another optical constant determination method according to Embodiment 1. As shown in Figure 17, the optical constant determination method further comprises a minute region setting step (S11x), a step of determining whether to set the next minute region (S15a), and a mapping display step (S16).

[0092] The micro-region setting step (S11x) is performed after the focus-combined image acquisition step (S11). In the micro-region setting step (S11x), a micro-region is set. A micro-region is, for example, a pixel. By making the micro-region a pixel, the resolution of the mapping can be improved. Note that the micro-region is not limited to a pixel, but may also be a collection of several pixels. After step S15, it is determined whether to set the next micro-region as shown in step S15a. If the answer in step S15a is YES, which means setting the next micro-region, the process returns to step S11x, and steps S11x to S15a are repeated. As a result, the photodetector 19 acquires optical constants and stack counts for multiple micro-regions of the sample 30. If the answer in step S15a is NO, which means not setting the next micro-region, the mapping is displayed as shown in step S16. In the mapping display step (S16), the processing unit 20 maps and displays each optical constant and each stack count determined for the multiple micro-regions.

[0093] In the focus stacking image acquisition step (S11), the wavelength of the illumination light used to illuminate the sample 30 via the confocal optical system 101 may be selected to include at least one of the wavelengths in the blue wavelength band, the green wavelength band, and the red wavelength band. The photodetector 19 may then acquire images using light with wavelengths in the blue wavelength band, images using light with wavelengths in the green wavelength band, and images using light with wavelengths in the red wavelength band. Furthermore, in the candidate material selection step (S12), an optical constant file may be prepared in which calculated data showing the relationship between wavelength and reflectance for each film thickness is stored for a plurality of candidate materials having known optical constants.

[0094] [Film thickness analysis] Next, we will explain film thickness analysis. First, we will explain the calculation data used in film thickness analysis. Then, we will explain how to measure the reflectance and calculate the film thickness by referring to the calculation data from the measured reflectance data.

[0095] <Introduction of interfering factors> To enable the calculation of reflectance even when the coherence of reflected light decreases, such as when using a high numerical aperture objective lens or when the film thickness is 1 μm or more, a coherence factor is introduced into the reflectance calculation. The reflectance R due to interference of the thin film in sample 30 can be divided into an interference component term and a non-interfering component term. Therefore, the reflectance R due to interference of the thin film in sample 30 is divided into the non-interfering component term R ic and the term R of the interference component if Using the coherence factor Γ, it can be written as in equation (1).

[0096]

number

[0097] If we define a state where reflected light within a thin film completely interferes as perfectly coherent (Γ=1) and a state where there is no interference as perfectly non-coherent (Γ=0), then the general state can be considered to be somewhere in between. When theoretically calculating reflectance, in the case of Γ=1 (R=R ch ) and when Γ=0 (R=R ic The interference component R can be calculated from the Fresnel coefficients for each of these. Therefore, by taking the difference between them, the interference component R can be calculated. if This can be calculated using equation (2).

[0098]

number

[0099] Therefore, by including the coherence factor Γ and using equation (3), the reflectance R can be theoretically calculated.

[0100]

number

[0101] Interferability includes temporal interferability and spatial interferability. Regarding the former, the interferometric distance L c can be estimated in the form of. It can be calculated as shown in Equation (4) using the wavelength band Δλ and the central wavelength λ of the illumination light. For example, when the central wavelength λ = 546 nm and the wavelength band Δλ = 16 nm, the interferometric distance L c is about 18 μm.

[0102]

Number

[0103] When the film thickness t is within a range sufficiently smaller than the interferometric distance L c the interferability factor Γ ~ 1, and in a large range, the interferability factor Γ << 1. Therefore, it is generally considered that the interferability factor Γ changes in a Gaussian function type. Although it may be distorted from the Gaussian type due to the influence of multiple reflections, it is approximated. Let β be the correction coefficient. When the refractive index of the thin film is n1, the optical path difference is 2tn1. (Later, by correcting the effect of the numerical aperture NA, tn1 is replaced with ΔL.)

[0104]

Number

Number

[0105] Factors other than the interferometric distance include the influence of the depth of focus. The deviation of the focal position between the film surface and the substrate surface acts in the direction of reducing interferability. If the half-value width when approximating the relationship between the deviation from the focal position and the received light intensity (I-Z curve) by a Gaussian function or a Lorentz function is not sufficiently small with respect to the film thickness t, the influence on interferability cannot be ignored. It is considered that the correction of this effect can be absorbed by the correction coefficient β.

[0106] By including a correction coefficient β as a parameter in the theoretical value compared to the measured value, the film thickness can be analyzed using the curve fitting method. The value of the correction coefficient β obtained from the analysis of representative points can also be used to analyze the film thickness distribution across the entire image.

[0107] <Theoretical calculation of absolute reflectance> Next, the theoretical calculation of absolute reflectance will be explained. Figures 18 and 19 are optical models illustrating a substrate 31 on which a thin film 32 according to Embodiment 1 is formed. As shown in Figures 18 and 19, the optical models show a cross-section of the structure of the sample 30 to be measured. The sample 30 includes a substrate 31 and a thin film 32 provided on the substrate 31. The substrate 31 is, for example, a silicon substrate Si, and the thin film 32 is, for example, a photoresist film (Photo Resist) PR.

[0108] The structure of sample 30 is, from top to bottom, air / photoresist film PR / silicon substrate Si, with the complex refractive indices of N0, N1, and N2 respectively with respect to wavelength λ, the refractive indices n0=1, n1, and n2, and the extinction coefficients k0=0, k1, and k2. The complex refractive index is defined as shown in equations (7) to (9). The desired thickness of the photoresist film PR is denoted as film thickness t.

[0109]

number

number

number

[0110] <1. In the case of complete interference> Next, let's consider the thin-film interference intensity due to the photoresist film PR on the silicon substrate Si. Assume that light of wavelength λ is incident perpendicularly to the thin film (in Figure 18, oblique incidence is assumed considering the numerical aperture NA of the objective lens). Part of the incident light is reflected at the air / photoresist film PR interface 0. Part of it is transmitted and reflected at the photoresist film PR / silicon substrate Si interface 1. Part of the reflected light from interface 1 is transmitted through interface 0, but some of it is reflected at interface 0 and then reflected again at interface 1. Generally, such multiple reflections occur in thin films. All of these reflected lights interfere and are summed up to obtain the reflectance R. If the complex refractive index is known, the reflectance R for perpendicular incidence is determined only by the wavelength λ and the film thickness t.

[0111] Let the amplitude reflectance at interface 0 (air / photoresist film PR interface) and interface 1 (photoresist film PR / silicon substrate Si interface) be given by equations (10) and (11), respectively. At this time, let δ and γ be the phase change and amplitude change of light that passes through the thin film once, as shown in equations (12) and (13), respectively.

[0112]

number

number

number

number

[0113] The amplitude reflectance of the entire film structure, considering multiple reflections, is given by equation (14).

[0114]

number

[0115] Since the reflectance R is the square of the absolute value of the amplitude reflectance, we get equation (15).

[0116]

number

[0117] Therefore, if the complex refractive index is known, the reflectance for normally incident light can be calculated using only the wavelength λ and film thickness t.

[0118] Here, we consider the correction of the oblique incidence effect by the numerical aperture NA of the objective lens. Figure 20 is a diagram illustrating the optical path difference due to the refraction of reflected light on a substrate on which a thin film according to Embodiment 1 is formed. As shown in Figure 20, light incident at an angle θ0 with respect to the optical axis is refracted by θ1 at interface 0 according to Snell's law (16), reflected at interface 1, and refracted again at interface 0 to become reflected light E1. Interference due to multiple reflections occurs between the reflected light E1 and the reflected light E0 reflected at interface 0. The difference from the case of normal incidence is that the optical path difference between reflected light E0 and reflected light E1 is as shown in equation (17). As the angle of incidence increases, the apparent film thickness t decreases.

[0119]

number

[0120]

number

[0121] The incident light is affected by an angle θ determined by the numerical aperture NA of the objective lens. NA Since the light includes various angles from 0 to 1, strictly speaking, the interference of reflected light needs to be calculated for all angles. However, since handling the angular dependence of amplitude reflectance and optical path difference for all angles is computationally intensive, we approximate it using the amplitude reflectance and average optical path difference for normal incidence. Since the incident light is unpolarized, we assume that the angular dependence of the optical path difference is dominant over the angular dependence of the reflectance.

[0122] The relationship between the numerical aperture NA and the angle of refraction is given by equation (18), so the average optical path difference is given by equation (17) multiplied by -θ. max ~+θmax Within this range, the average of the angular distribution can be calculated as shown in equation (19). Therefore, by substituting equation (17) with equation (19) and calculating equation (15), the reflectance for any numerical aperture NA can be easily calculated.

[0123]

number

number

[0124] The calculations for the single-layer thin film 32 described above can also be performed for multilayer films by using the composite rule of the composite Fresnel coefficients from the substrate 31 side.

[0125] <2. In the case of complete non-interference> Next, the reflectance R in the absence of thin-film interference due to the photoresist film PR on the silicon substrate Si. ic This will be examined. Figure 21 is a diagram illustrating interference-free multiple reflections within the thin film 32 according to Embodiment 1. Similar to the case of perfect interference, multiple reflections occur within the thin film 32, but since the reflected light from interface 0 and interface 1 does not interfere with each other, the reflectance R is obtained by simply summing these reflected lights. ic You can obtain this.

[0126] The reflectance at interface 0 is R0, the reflectance at interface 1 is R1, and the internal transmittance of thin film 32 is T. i Let's assume that the reflectance at each interface is given by equations (20) and (21) from the Fresnel coefficient. Internal transmittance T i Using Lambert's law and the relationship between the absorption coefficient α and the extinction coefficient k1, we obtain equation (22). The reflectance of m reflections by the substrate 31 is R. m Therefore, the sum of m from 0 to infinity is the reflectance R. ic Therefore, it can be written as in equation (23).

[0127]

number

number

number

number

[0128] Internal transmittance T i The effect of numerical aperture NA on the value is corrected. Internal transmittance T at angle θ1. i In equation (22), the film thickness t is angularly averaged. <t>You can replace it with this. The angle of incidence is 0 to θ. max Then, we get equation (24), and the angle average <t>This can be written as in equation (25). Here again, the internal transmittance T i Make it as shown in equation (26).

[0129]

number

number

number

[0130] The above calculations can be performed even in the case of a multilayer film by performing a composite calculation of the interfacial reflectance from the substrate 31 side.

[0131] <3. In the case of partial interference> Perfectly coherent reflectance R ch , the reflectance R of a completely non-coherent system ic By using the coherence factor Γ, the reflectance R of these intermediate states (partial coherence) can be calculated using equation (27).

[0132]

number

number

number

number

number

[0133] The coherence factor Γ in equation (30) was obtained by replacing the film thickness tn1 in equation (5) with ΔL using the numerical aperture NA correction in equation (19). In the case of multilayer films, the sum of the ΔL of each layer is used as the final ΔL.

[0134] Figures 22 and 23 are graphs illustrating the relationship between wavelength and reflectance according to Embodiment 1, where the horizontal axis represents wavelength and the vertical axis represents absolute reflectance. As shown in Figure 22, the dependence of the reflectance R on the correction coefficient β at wavelengths of 400 to 700 nm was calculated for the case where the numerical aperture NA = 0.3 and the film thickness t = 1000 nm. When the correction coefficient β = 0, Γ = 1; when the correction coefficient β = 1, Γ < 0.8; when the correction coefficient β = 2, Γ < 0.65; and when the correction coefficient β = 10, Γ < 0.1. As Γ decreases, the interference component is attenuated. Furthermore, since Γ is wavelength-dependent, the interference component is attenuated from the shorter wavelength side. As shown in Figure 23, when the numerical aperture NA = 0.3 and the film thickness t = 2000 nm, the interference component is attenuated as the film thickness increases.

[0135] Here, six wavelengths are selected as the illumination wavelengths: 436nm, 486nm, 514nm, 546nm, 578nm, and 633nm. The reflectance is the brightness value I of the captured image. sample The film thickness t can be determined by fitting the measured reflectance R at each wavelength to equation (27) using the least squares method, with the film thickness t and the correction coefficient β as parameters.

[0136] When measuring reflectance R, the reflectance of a reference sample with a known reflectance R is measured to correct for the characteristics of the optical system and light source used in the measurement. For example, when using silicon (Si) as a reference (quartz glass, etc., can also be used), the reflection image of silicon (Si) is captured at each wavelength (color balance, gain control, etc. are kept constant). The luminance value I of this silicon (Si) is then measured. Si The reflectance R of the sample is calculated as a relative value to [a certain value], and further, the wavelength dependence of the known reflectance of silicon (Si) is [a certain value]. Si The correction is then performed. Therefore, the reflectance of the sample can be determined as shown in equation (32). 0 This is the dark luminance value received when the shutter for illumination is closed.

[0137]

number

[0138] With the gain adjustment of the confocal microscope fixed, images of the same field of view were acquired for the reference silicon (Si) and sample 30, but at different wavelengths. The average value of the brightness value for a specified area on the image, or the brightness value of each pixel, was measured using I sample , I Si We use this and calculate the absolute reflectance from equation (32).

[0139] Figure 24 is a graph illustrating the relationship between wavelength and reflectance according to Embodiment 1, where the horizontal axis represents wavelength and the vertical axis represents absolute reflectance. Figure 21 shows a plot of the measured reflectance obtained from equation (32) and the calculated reflectance obtained from equation (27). The calculations were performed with correction coefficients β of 0 and 0.5. The residual Σ between the measured reflectance Rmes and the calculated reflectance Rcal is obtained using equation (33).

[0140] Figure 25 is a graph illustrating the relationship between the film thickness and residuals of a thin film according to the embodiment, where the horizontal axis represents the film thickness and the vertical axis represents the residuals.

[0141]

number

[0142] The film thickness t at which the residual Σ is minimized can be considered the film thickness at the measurement point of this sample. Therefore, in this case, when the correction coefficient β=0 (perfectly coherent), the film thickness t is determined to be 44 nm, and when the correction coefficient β=0.5 (partially coherent), the film thickness t is determined to be 1580 nm. The analysis results with the coherence factor Γ introduced are consistent with the measurement results of white light interference. Even with measurements at six wavelengths, by introducing the coherence factor Γ into the analysis, film thickness t of 1 μm or more can be measured with high accuracy.

[0143] Similarly, the film thickness t can be determined for each pixel of the image, allowing the film thickness distribution to be displayed. Therefore, the film thickness distribution can be easily analyzed using a histogram. Even if abnormal points occur in the measurement values ​​due to the presence of dust or foreign matter on the thin film, interpolation (noise reduction) with the surrounding normal values ​​is easily performed. As a display method, a bird's-eye view can also be used, as shown in Figure 26.

[0144] [Supplementary information on reflectance calculations] This section provides supplementary information regarding the handling of reflectance calculations when the thin film 32 in sample 30 is a multilayer film and several zero-thickness layers exist within the multilayer film. In the case of perfect coherence, the composite formula for the Fresnel coefficients of each interface can be applied directly even when the thickness is zero, so no problems arise. However, in the case of perfect non-coherence, when the thickness of the layer is zero, equations (20), (21), and (23) cannot be applied directly, so it is necessary to introduce the special calculation rules described below.

[0145] Figure 27 illustrates the treatment of reflectance calculations for a film thickness of zero in the case of complete non-interference according to Embodiment 1. As shown in Figure 27, consider the overlap of three consecutive layers in an N-layer film: a (m-1) layer, an m layer, and a (m+1) layer. Focusing on the m layer, the optical constant is n m , k m and film thickness t m The interface between the (m-1) layer and the m layer is defined as the (m-1, m) interface, and the interface between the m layer and the (m+1) layer is defined as the (m, m+1) interface. The Fresnel coefficients of the interfaces are given by r. m-1、m and r m、m+1 This can be expressed similarly to equations (10) and (11). This relationship can also be considered for the (m-1) layer and the (m+1) layer.

[0146] Combined reflectance R' up to the (m+2) layer m+2、m+3 Given that we have been able to calculate up to this point, the film thickness t of the (m+1) layer is calculated. m+1 If >0, the combined reflectance R' of the (m+1, m+2) interface. m+1、m+2 This can be calculated in the same manner as in equation (23). Here, the film thickness t of the m layer is m If = 0, the Fresnel coefficient of the upper and lower interface, r m-1、m and r m、m+1 This is replaced with special rules such as those in equations (34) and (35). Using these revised Fresnel coefficients, the reflectance of each interface is redefined as equations (36) and (37). The internal transmittance T at this point is then calculated. i These can be expressed as equations (38) and (39), respectively.

[0147]

number

number

number

number

number

number

[0148] The combined reflectance R' at the (m, m+1) interface. m。m+1 According to the composition rules, this is given by equation (40).

[0149]

number

[0150] Furthermore, the combined reflectance R' of the upper m layer. m-1、m However, according to the composition rules, we get equation (41).

[0151]

number

[0152] Upon closer examination of equation (41), it becomes clear that it is precisely the formula for reflectance when no m-layer exists. By applying this Fresnel coefficient transformation rule, it is possible to calculate the perfectly incoherent reflectance even when layers with a thickness of 0 exist in a multilayer film. The same calculation can be performed when J layers with a thickness of 0 are distributed within an N-layer film.

[0153] Next, the effects of this embodiment will be explained. The optical constant determination device 100 of this embodiment can determine the optical constants of an unknown thin film 32. Therefore, the material of the unknown thin film 32 can be estimated. Furthermore, when determining the optical constants, the optical constant determination device 100 calculates the coefficient of determination and determines the optical constants based on the coefficient of determination. Therefore, the accuracy of the optical constants and material can be improved.

[0154] Furthermore, the optical constant determination device 100 can calculate the film thickness of the thin film 32 and determine the number of layers of the thin film 32. Since the film thickness and the number of layers are determined based on the coefficient of determination, the accuracy can be improved.

[0155] When determining optical constants, materials, film thickness, and number of layers, the determination is made based on a list sorted in descending order of coefficient of determination. For example, the top-ranked item on the list is chosen. Therefore, the determination can be made easily. Furthermore, if the top-ranked material on the list obtained from the analysis results can be rejected based on other information, the second-best candidate can be easily selected for the optical constants, materials, film thickness, and number of layers.

[0156] The optical constants, film thicknesses, and number of layers determined in multiple analysis regions can be mapped and displayed on the sample 30. Therefore, information about the thin film 32 can be easily grasped visually.

[0157] A bandpass filter (BPF) selects the wavelength of the illumination light so that it includes wavelengths in multiple wavelength bands. The photodetector (19) also includes an image sensor that detects wavelengths in multiple wavelength bands. Therefore, multiple confocal composite images using light of multiple wavelengths can be obtained in a single imaging pass. This reduces the time required to measure film thickness.

[0158] Furthermore, the optical constant determination device 100 of this embodiment is equipped with a confocal optical system 101. Therefore, higher resolution images can be obtained than with a conventional optical microscope. This improves the spatial resolution of the film thickness distribution. Also, even when the substrate 31 is thin, such as a transparent film, back surface reflection can be eliminated. For this reason, analysis can be performed using a simple optical model, and experimental results and theoretical results can be easily reconciled. Even if there are undulations on the surface of the sample 30 that exceed the depth of focus, a full-focus image can be created, and the reflectance of the entire surface of the sample 30 can be measured. When using a bandpass filter BPF, even if a shift in the focal position occurs due to wavelength, the shift can be suppressed by synthesizing a full-focus image obtained by a focus scan.

[0159] In this embodiment, the reflectance in the calculation data, which shows the relationship between wavelength and reflectance for each thickness of the thin film 32, includes a non-coherent component term. Therefore, the film thickness can be analyzed with high accuracy even for thin films with a thickness of 1 μm or more, where coherence decreases. This allows for high-resolution analysis of the film thickness distribution in thin films 32 with a thickness of 1 μm or more.

[0160] Furthermore, the non-coherent component R of the reflectance in the calculated data. ic This term includes the sum of the reflectances at each reflection up to a predetermined number of times at the interface on the substrate 31 side. This allows for the reflection of multiple reflections within the thin film 32, and thus enables high-precision analysis of the film thickness even for thin films 32 with a thickness of 1 μm or more.

[0161] The reflectance in the calculated data is the interference component R if It includes the product of the term and the coherence factor Γ, which represents the proportion of the interfering component term. As a result, the reflectance is the non-interfering component R ic and interference component R if Even in cases of partial coherence including [specific conditions], the film thickness of thin films can be analyzed with high accuracy.

[0162] Furthermore, the reflectance in the calculated data includes correction based on the numerical aperture (NA) of the objective lens. This allows for highly accurate analysis of the film thickness even when using high numerical aperture objective lenses. It also enables high-resolution analysis of the film thickness distribution.

[0163] The optical constant determination device 100 synthesizes the reflectances shown for each layer of stacked thin films 32, even when the thin film 32 is a multilayer film, to obtain the reflectance in the calculation data. Therefore, the film thickness of a multilayer film 32 can be analyzed with high accuracy. Furthermore, the film thickness distribution in a multilayer film can be analyzed with high resolution. Even when some film thicknesses in the multilayer film are zero, the film thickness can be analyzed with high accuracy by applying the Fresnel coefficient transformation rule without changing the configuration of the optical model.

[0164] In this way, the optical constant determination device 100 can obtain a spectral spectrum for each pixel of the observed image. Then, the spectral spectrum is converted to absolute reflectance, and fitting is performed using the film thickness, which takes the coherence factor into consideration, as a parameter. This makes it possible to accurately determine not only film thicknesses of 0 to 1 μm, but also film thicknesses of 1 μm or more.

[0165] Although embodiments of the present invention have been described above, the present invention includes appropriate modifications that do not impair its purpose and advantages, and is not limited by the above embodiments. Furthermore, the components in Embodiment 1 may be combined in appropriate ways.

[0166] Furthermore, the following optical constant determination program, which causes a computer to execute the optical constant determination method of this embodiment, is also included in the technical concept of this embodiment.

[0167] An optical constant determination program for determining the optical constant of a thin film in a sample comprising a substrate and a thin film provided on the substrate, A focus-stacked image acquisition step involves detecting the reflected light from the sample, which is emitted by illuminating the sample via a confocal optical system, and acquiring separate images of the sample using multiple wavelengths of light contained in the reflected light. A candidate material selection step in which a plurality of candidate materials having known optical constants selected as candidates for the material of the thin film are input, A film thickness analysis step is performed to obtain measurement data of the reflectance for each wavelength based on each image, A step to calculate the coefficient of determination when approximating the film thickness from the measurement data, by referring to calculation data that shows the relationship between the wavelength and the reflectance for each candidate material for each film thickness, An optical constant determination step in which the optical constant of the thin film is determined from among the optical constants of a plurality of candidate materials based on the calculated coefficient of determination, A program that causes a computer to perform the task of determining optical constants.

[0168] Furthermore, the optical constant determination program described above, when loaded into a computer, includes a set of instructions (or software code) for causing the computer to perform one or more of the functions described in the embodiments. The program may be stored in a non-temporary computer-readable medium or a physical storage medium. Examples, but not limited to, include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technologies, CD-ROM, digital versatile disc (DVD), Blu-ray® disc or other optical disc storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The program may be transmitted over a temporary computer-readable medium or a communication medium. Examples, but not limited to, include temporary computer-readable medium or a communication medium that includes electrical, optical, acoustic or other forms of propagating signals. [Explanation of Symbols]

[0169] 10 Light source section 11 Light source 12 Wavelength Selection Section 13a, 13b, 13c lenses 14 slits 15 Beam Splitter 16 Vibrating Mirror 17 Objective lens 18 stages 19. Photodetector 20 Processing Units 21 Optical Constants File 30 samples 31 circuit boards 32 Thin film 40 Analysis area 100 Optical constant determination device 101 Confocal optical system B-Phoenix bandpass filter BPF-1, BPF-2, BPF-3, BPF-4, BPF-5 Bandpass Filters BPF-6, BPF-7, BPF-8 bandpass filters V field of view< / t> < / t>

Claims

1. An optical constant determination device for determining the optical constant of at least one of the thin film and the substrate of a sample including a substrate and a thin film provided on the substrate, A photodetector that detects reflected light from the sample after illumination light has been used to illuminate the sample via a confocal optical system, and acquires images of the sample using multiple wavelengths of light contained in the reflected light. A processing unit that acquires measurement data of the reflectance for each wavelength based on each image, Equipped with, The aforementioned processing unit, A plurality of candidate materials having known optical constants, selected as candidates for the materials of the thin film and the substrate, are input. For each candidate material, the relationship between the wavelength and the reflectance is shown in the calculation data for each film thickness, and the coefficient of determination is calculated when approximating the film thickness from the measurement data. The coefficient of determination is a measure of how well the thin film thickness is fitted as a parameter using the least squares method based on the measured reflectance values ​​for each wavelength. Based on the calculated coefficient of determination, the optical constant of at least one of the thin film and the substrate is determined from among the optical constants of a plurality of candidate materials. Optical constant determination device.

2. The aforementioned processing unit, The candidate number of layers selected as a candidate for the number of layers of film included in the thin film is input. A plurality of candidate materials having known optical constants, selected as candidate materials for each film, are input. The coefficient of determination is calculated when approximating the film thickness from the measurement data by referring to the synthesized calculation data for each film, Based on the calculated coefficient of determination, the number of layers of the thin film is determined from among the multiple candidate number of layers. The optical constant determination apparatus according to claim 1.

3. The photodetector acquires images of each of the multiple minute regions of the sample, The processing unit maps and displays the optical constants, film thicknesses, and number of layers determined in the plurality of minute regions. The optical constant determination apparatus according to claim 2.

4. The system further includes a wavelength selection unit that selects the wavelengths such that the illumination light illuminating the sample via the confocal optical system includes at least one of the wavelengths in the blue wavelength band, the wavelength in the green wavelength band, and the wavelength in the red wavelength band. The photodetector includes a blue image sensor that acquires the image using light of the wavelength in the blue wavelength band, a green image sensor that acquires the image using light of the wavelength in the green wavelength band, and a red image sensor that acquires the image using light of the wavelength in the red wavelength band. An optical constant determination apparatus according to any one of claims 1 to 3.

5. The system further comprises an optical constant file in which the calculation data is stored, showing the relationship between the wavelength and the reflectance for each film thickness for each of the multiple candidate materials having the known optical constants, An optical constant determination apparatus according to any one of claims 1 to 3.

6. A method for determining optical constants of a sample comprising a substrate and a thin film provided on the substrate, wherein the optical constants of at least one of the thin film and the substrate are determined. A focus-stacked image acquisition step involves detecting the reflected light from the sample, which is emitted by illuminating the sample via a confocal optical system, and acquiring separate images of the sample using multiple wavelengths of light contained in the reflected light. A candidate material selection step in which a plurality of candidate materials having known optical constants, selected as candidates for the material of the thin film and the substrate, are input; A film thickness analysis step is performed to obtain measurement data of the reflectance for each wavelength based on each image, A step to calculate the coefficient of determination when approximating the film thickness from the measurement data, by referring to calculation data that shows the relationship between the wavelength and the reflectance for each candidate material for each film thickness, An optical constant determination step in which, based on the calculated coefficient of determination, the optical constant of at least one of the thin film and the substrate is determined from among the optical constants of a plurality of candidate materials, Equipped with, The coefficient of determination is a measure of how well the model fits the thin film thickness as a parameter when fitting the measured reflectance values ​​at each wavelength using the least squares method. Method for determining optical constants.

7. The system further comprises a layer count selection step in which a candidate number of layer counts selected as a candidate number of layers for the number of layers of films included in the thin film is input, In the candidate material selection step, A plurality of candidate materials having the known optical constants, selected as candidate materials for each film, are input. In the step of calculating the coefficient of determination, The coefficient of determination is calculated when approximating the film thickness from the measurement data by referring to the synthesized calculation data for each film, In the optical constant determination step, Based on the calculated coefficient of determination, the number of layers of the thin film is determined from among a plurality of candidate layer counts. The method for determining optical constants according to claim 6.

8. A micro-region setting step is performed after the focus-combined image acquisition step, to set a micro-region, Following the optical constant determination step, a mapping display step is performed to map and display the optical constants, film thicknesses, and number of layers determined in a plurality of minute regions. Furthermore, In the focus-stacked image acquisition step, each image including a plurality of microregions of the sample is acquired. The method for determining optical constants according to claim 7.

9. In the aforementioned focus-stacked image acquisition step, The wavelengths are selected such that the illumination light used to illuminate the sample via the confocal optical system includes at least one of the wavelengths in the blue wavelength band, the wavelength in the green wavelength band, and the wavelength in the red wavelength band. The image obtained by the light of the wavelength in the blue wavelength band, the image obtained by the light of the wavelength in the green wavelength band, and the image obtained by the light of the wavelength in the red wavelength band. A method for determining optical constants according to any one of claims 6 to 8.

10. In the candidate material selection step, For a plurality of candidate materials having the known optical constants, prepare an optical constant file in which the calculated data showing the relationship between wavelength and reflectance for each film thickness is stored. A method for determining optical constants according to any one of claims 6 to 8.

Images