Sample surface quality control device

Abstract

Provided is a sample surface quality management device that measures the microroughness of a sample, the sample surface quality management device comprising a scattered light measurement device that measures scattered light produced at the sample, an interference light measurement device that measures interference light that includes reflected light produced at the sample, and a signal processing device that processes signals from the scattered light measurement device and the interference light measurement device. The signal processing device computes a first evaluation value for the microroughness of the sample on the basis of a signal from the interference light measurement device, computes a scattering characteristics signal on the basis of a signal from the scattered light measurement device, and computes, for a spatial frequency band for which the first evaluation value is not computed, a second evaluation value for the microroughness on the basis of the first evaluation value and the scattering characteristics signal.

Description

[Technical Field]

[0001] The present invention relates to a sample surface quality control apparatus for measuring the microroughness of a sample such as a wafer in a semiconductor device manufacturing process or the like and controlling the quality of the sample surface. [Background technology]

[0002] The increasing integration density of semiconductor devices has led to demands for ever higher quality silicon wafers. One element of wafer quality is surface roughness, which affects the electrical characteristics of devices. Therefore, it is important to inspect the entire surface of all samples in in-line inspection during the wafer manufacturing process. Sample microroughness, or in other words, the microscopic flatness of the sample surface, is generally measured using an AFM (atomic force microscope). However, because AFM measurements require a long time, it is difficult to introduce it into in-line inspection.

[0003] A known technique for measuring microroughness in inline inspection is to measure microroughness based on the haze value, which is measured using a scattered light measurement device used to inspect sample surfaces for foreign matter, and which takes advantage of the fact that the haze value correlates with the microroughness (Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6043813 Summary of the Invention [Problem to be solved by the invention]

[0005] However, because the haze value varies depending on various conditions, such as the material of the sample, the state of the optical system of the inspection device, and machine differences, quantitative evaluation of microroughness presents a challenge. In response to this, the technology disclosed in Patent Document 1 calibrates the haze value by comparing it with previously obtained AFM measurement results. In this case, whenever conditions that affect the haze value, such as the material of the sample, change, a calibration work must be performed by performing a prior measurement with the AFM, increasing the number of steps.

[0006] An object of the present invention is to provide a sample surface quality control device capable of measuring the micro-roughness of a sample surface over the entire surface at high speed. [Means for solving the problem]

[0007] In order to achieve the above object, the present invention provides a sample surface quality control device for measuring the microroughness of a sample, comprising: a stage device for holding the sample and moving the sample in the direction of the sample surface; a scattered light measurement device for measuring scattered light generated by the sample; an interference light measurement device for measuring interference light including reflected light generated by the sample; and a signal processing device for processing signals from the scattered light measurement device and the interference light measurement device, wherein the signal processing device calculates a first evaluation value of the microroughness of the sample based on the signal from the interference light measurement device, calculates a scattering characteristic signal based on the signal from the scattered light measurement device, and calculates a second evaluation value of the microroughness based on the first evaluation value and the scattering characteristic signal for a spatial frequency band in which the first evaluation value is not calculated. [Effects of the Invention]

[0008] According to the present invention, the micro-roughness of the surface of a sample can be measured over the entire surface at high speed. [Brief explanation of the drawings]

[0009] [Figure 1] Schematic diagram of a sample surface quality control device according to a first embodiment of the present invention. [Figure 2] FIG. 1 is a diagram showing an example of a scanning trajectory of a sample in a sample surface quality control device according to a first embodiment of the present invention; [Figure 3] FIG. 1 is a schematic diagram showing an example of the configuration of a scattered light intensity measurement system provided in a sample surface quality control device according to a first embodiment of the present invention; [Figure 4] FIG. 1 is a schematic diagram showing the positional relationship between a scattered light intensity measurement system and a beam spot provided in a sample surface quality control device according to a first embodiment of the present invention. [Figure 5] FIG. 1 is a diagram illustrating an example of an output signal (scattered light signal) of a scattered light intensity measurement system provided in a sample surface quality control device according to a first embodiment of the present invention. [Figure 6] FIG. 1 is a schematic diagram showing an example of the configuration of an interference light measurement device provided in a sample surface quality control device according to a first embodiment of the present invention; [Figure 7] 1 is a flowchart showing an example of a procedure for evaluating the micro-roughness of a sample by a signal processing device provided in a sample surface quality control device according to a first embodiment of the present invention. [Figure 8] FIG. 1 is an explanatory diagram of a spatial frequency band that cannot be measured by DIC measurement using the sample surface quality control device according to the first embodiment of the present invention. [Figure 9A] An explanatory diagram of the function used as the PSD model (first model) [Figure 9B] An explanatory diagram of the function used as the PSD model (second model) [Figure 9C] An explanatory diagram of the function used as a PSD model (third model) [Figure 10A] FIG. 10 is an explanatory diagram of a procedure for calculating a second evaluation value based on a first evaluation value and a scattering property signal in the first embodiment of the present invention. [Figure 10B] FIG. 10 is an explanatory diagram of a procedure for calculating a second evaluation value based on a first evaluation value and a scattering property signal in the first embodiment of the present invention. [Figure 11A] FIG. 10 is an explanatory diagram of a first example of handling a haze value having a different spatial direction from the first evaluation value. [Figure 11B] FIG. 10 is an explanatory diagram of a second example of handling a haze value in a different spatial direction from the first evaluation value. [Figure 11C] FIG. 10 is an explanatory diagram of a third example of handling a haze value having a different spatial direction from the first evaluation value. [Figure 12] Explaining PSD data calculation in the direction perpendicular to the shear direction [Figure 13] FIG. 10 is a schematic diagram showing an example of the configuration of an interference light measurement device provided in a sample surface quality control device according to a second embodiment of the present invention; [Figure 14A] FIG. 10 is an explanatory diagram of a procedure for calculating a second evaluation value based on a first evaluation value and a scattering property signal in a second embodiment of the present invention. [Figure 14B] FIG. 10 is an explanatory diagram of a procedure for calculating a second evaluation value based on a first evaluation value and a scattering property signal in a second embodiment of the present invention. [Figure 15] FIG. 10 is a schematic diagram showing an example of the configuration of an interference light measurement device provided in a sample surface quality control device according to a third embodiment of the present invention. [Figure 16] FIG. 10 is an explanatory diagram of a spatial frequency band that cannot be measured by DIC measurement using the sample surface quality control device according to the third embodiment of the present invention. [Figure 17] FIG. 10 is a schematic diagram showing an example of the configuration of a scattered light intensity measurement system provided in a sample surface quality control device according to a fourth embodiment of the present invention. [Figure 18] Conceptual diagram showing partial overlap of beam spots between adjacent scanning tracks [Figure 19A] FIG. 10 is a schematic diagram showing an example of the configuration of a detection optical system of a scattered light measurement device provided in a sample surface quality control device according to a fifth embodiment of the present invention. [Figure 19B] FIG. 10 is an explanatory diagram illustrating the relationship between the emission direction of scattered light in the detection optical system of the scattered light measurement device provided in the sample surface quality control device according to the fifth embodiment of the present invention and pixel coordinates on the pupil plane. [Figure 20A] FIG. 10 is a diagram showing an example of a display in which the processing results for each display unit area are displayed as a surface map of the sample 1. [Figure 20B] A diagram showing an example of a histogram display of the processing results. [Figure 20C] A diagram showing an example of displaying the processing results (PSD data) as a scatter plot. [Figure 20D] A diagram showing an example of displaying the processing results (model parameters) in a table format. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0011] (First embodiment) [Sample surface quality control device] 1 is a schematic diagram of a sample surface quality control device according to a first embodiment of the present invention. The sample surface quality control device shown in the figure is configured to include an inspection device 100 and a signal processing device 200. A typical example of a sample 1, whose microroughness, in other words, the microscopic flatness of the surface, is inspected and controlled by this sample surface quality control device, is a disk-shaped semiconductor silicon wafer with no patterns formed on it and a flat surface.

[0012] [Inspection device 100] The inspection apparatus 100 includes an optical system for measuring scattered light generated on the surface of the sample 1, and an optical system for measuring the surface shape of the sample 1. The method for measuring the surface shape of the sample 1 is interference light measurement, which can measure the entire surface of the sample at high speed. Examples of interference light measurement include optical interferometry, phase-shift interferometry, and wavelength-shift interferometry, and any of these methods may be used. Specifically, the inspection apparatus 100 includes a stage device 110, a scattered light measurement device 120, an interference light measurement device 130, and a signal processing unit 190.

[0013] [Stage device 110] The stage device 110 is a device that holds the sample 1 relative to the beam spot and moves the sample 1 in the direction of the sample surface, and is equipped with a sample stage 111, a rotary stage 112, and a linear stage 113. The sample stage 111 is, for example, a chuck table that holds the sample 1 by suction or the like. The linear stage 113 moves the rotary stage 112 linearly in the radial direction of the sample 1, and the rotary stage 112 rotates the sample stage 111 (spins on its axis).

[0014] FIG. 2 is a diagram showing an example of a sample scanning trajectory. Light is irradiated onto the surface of the sample 1 from the scattered light measurement device 120 and the interference light measurement device 130, forming a beam spot 2. The beam spot 2 shown in the figure is formed by both the beam spots of the scattered light measurement device 120 and the interference light measurement device 130 at substantially the same coordinates (or coordinates whose distance from each other is within a predetermined distance). During inspection, the sample 1 is rotated in the circumferential direction of the sample 1 by the rotation stage 112, and the beam spot 2 rotates on the surface of the sample 1 in the direction of arrow θ. At the same time, the sample 1 is driven linearly by the linear stage 113, and the beam spot 2 moves linearly on the surface of the sample 1 in the direction of arrow R. By combining this rotational movement and linear movement, the stage device 110 moves the sample 1 relative to the beam spot 2, and the beam spot 2 traces a spiral trajectory to scan the entire surface of the sample 1.

[0015] 1 and 2, the stage device 110 may be provided with a linear stage having a movement axis (Y-axis here) that horizontally intersects with the movement axis (X-axis here) of the linear stage 113. In this case, the sample 1 can be scanned along an orbit in which the beam spot 2 repeatedly moves in the X direction and the -X direction while being sequentially shifted in the Y-axis direction.

[0016] [Scattered light measurement device 120] The scattered light measuring device 120 is an optical system that illuminates the sample 1 and measures the scattered light generated at the beam spot 2 on the sample surface, and is configured to include an illumination optical system 121 and a detection optical system 122 .

[0017] The illumination optical system 121 is an optical system that guides a light source for scattered light and light emitted from the light source to the beam spot 2, and includes a plurality of optical elements such as lenses. In this embodiment, the illumination optical system 121 is capable of oblique illumination, in which light is irradiated obliquely onto the surface of the sample 1. The illumination optical system 121 may also be provided with a mechanism for switching the optical path to epi-illumination, in which light is irradiated perpendicularly onto the surface of the sample 1.

[0018] The detection optical system 122 is an optical system that detects scattered light generated at the beam spot 2 on the sample surface by dividing it in spatial directions, and in this embodiment, it is equipped with multiple scattered light intensity measurement systems 123-126 (four are shown in Figure 1) that have different azimuth angles and elevation angles relative to the beam spot 2.

[0019] 3 is a schematic diagram showing an example of the configuration of the scattered light intensity measurement system 123-126. The scattered light intensity measurement system 123-126 has a detection optical system 127 and a scattered light sensor 128.

[0020] The detection optical system 127 includes multiple lenses (lens group) and constitutes a so-called focusing optical system or imaging optical system. The detection optical system 127 may be equipped with a spatial filter or a polarizing filter and have the function of blocking undesired light that becomes noise. The beam spot 2 is located on an extension of the optical axis 129 of the detection optical system 127. The optical axes 129 of the scattered light intensity measurement systems 123-126 extend in different spatial directions from the beam spot 2. The optical axes 129 of the scattered light intensity measurement systems 123-126 are inclined with respect to the normal N of the sample surface that passes through the beam spot 2 (intersect with the normal N at the beam spot 2). However, the inspection device 100 may be equipped with a scattered light intensity measurement system whose optical axis 129 coincides with or is parallel to the normal N.

[0021] The scattered light sensor 128 is a photoelectric conversion element, and preferably has a high gain in order to measure weak scattered light, and a photomultiplier tube or an avalanche photodiode array can be used. Other possible sensor types include a photomultiplier tube, SiPM, CMOS sensor, and CCD.

[0022] Figure 4 is a schematic diagram showing the positional relationship between the scattered light intensity measurement systems 123-126 and the beam spot 2. When expressing the position on the surface of the sample 1 using two-dimensional coordinates (X, Y), the X-axis is the projection direction onto the XY plane of the incident light axis (the center line of the light incident on the beam spot 2 from the illumination optical system 121) incident at an elevation angle θi. The detection optical system 127, which has a numerical aperture corresponding to the solid angle ω, is positioned at an elevation angle θs and an azimuth angle φs relative to the beam spot 2. The combination of the elevation angle θs and the azimuth angle φs differs for each scattered light intensity measurement system 123-126. Of the scattered light generated by the beam spot 2, the scattered light flux that is emitted from the beam spot 2 at an elevation angle θs within the range of the solid angle ω of the azimuth angle φs is measured by the scattered light intensity measurement system positioned in that direction. In each of the scattered light intensity measurement systems 123-126, the scattered light collected by the detection optical system 127 is photoelectrically converted into a current or voltage signal by the scattered light sensor 128, and then further AD converted and processed by the signal processing unit 190 (Figure 1).

[0023] -Example of scattered light signal- FIG. 5 shows an example of the output signal (scattered light signal) of each scattered light intensity measurement system 123-126. The horizontal axis of FIG. 5 represents the θ coordinate on the sample surface along the spiral trajectory of the beam spot 2 shown in FIG. 2, corresponding to time. The vertical axis of FIG. 5 represents the magnitude of the scattered light signal output from the scattered light sensor 128. Scattered light caused by microroughness of the sample 1 is incident on each scattered light intensity measurement system 123-126, and each scattered light intensity measurement system 123-126 produces a scattered light signal S1 with a waveform as shown in FIG. 5. A data set of the value of this scattered light signal S1 and the θ coordinate is stored for each scattered light intensity measurement system 123-126, for example, in the signal processing unit 190. Furthermore, when the beam spot 2 crosses a defect (foreign matter, etc.), a defect signal S2, which is a particularly large scattered light signal S1, is detected. This defect signal S2 is separated from the scattered light signal S1 by, for example, a high-pass filter (HPF) in the signal processing unit 190, and the value and coordinates are stored in, for example, the signal processing unit 190 as a defect detection signal.

[0024] Note that methods for separating the defect signal S2 from the scattered light signal S1 can be not only separation in the frequency domain, such as the above-mentioned high-pass filter (HPF), but also separation based on signal magnitude. In other words, this method determines that signals below a preset threshold are scattered light signals caused by microroughness, and signals above the threshold are defect signals S2. The threshold can be a predetermined fixed value, or can be set in real time based on a signal that can be clearly determined to be a defect signal S2.

[0025] Furthermore, as long as the scattered light signal S1 due to microroughness is extracted, it is not necessarily necessary to separate the defect signal S2. For example, a method can be applied in which the scattered light signals S1 of the scattered light intensity measurement systems 123-126 are averaged (merged) at predetermined time intervals or for each predetermined region on the sample surface. One averaging method is to group the scattered light intensity measurement systems 123-126 and average the scattered light signals S1 for each group. As a specific example, the scattered light intensity measurement systems 123 and 126 are grouped as a first group, and the scattered light intensity measurement systems 124 and 125 are grouped as a second group, and the scattered light signals S1 are averaged for each group. The combination pattern of the scattered light intensity measurement systems 123-126 can be changed arbitrarily, but a combination that accurately reflects changes in the scattered light signal S1 due to microroughness is desirable.

[0026] Furthermore, if the sampling interval of the scattered light signal S1 is sufficiently short, the proportion of the defect signal S2 in the overall scattered light signal S1 is extremely small. In this case, even if the defect signal S2 is larger than the scattered light signal S1, averaging the scattered light signal S1 including the defect signal S2 will result in almost no change in the average value, and can be considered essentially the same as the average value excluding the defect signal S2. Averaging the scattered light signal S1 also has the advantage of reducing the processing load on the signal processing unit 190.

[0027] [Interference light measuring device 130] 6 is a schematic diagram showing an example of the configuration of an interference light measurement device 130. The interference light measurement device 130 is an optical system that measures interference light including reflected light generated at a beam spot 2 on the sample 1 and measures the surface shape of the sample 1. In this embodiment, an interference light measurement device 130 that detects differential interference contrast (DIC) and calculates the height of the sample surface will be described as an example.

[0028] The interference light measurement device 130 forms a beam spot 2 consisting of two polarized illumination spots of light with different polarizations on the surface of a sample, and generates an image of the interference light by collecting reflected light from the two polarized illumination spots. Specifically, the interference light measurement device 130 includes a light source 131, a differential interference illumination system 132, a beam splitter 133, a quarter-wave plate 134, a Nomarski prism 135, an objective lens 136, an imaging lens 137, and an interference light sensor 138.

[0029] Light emitted from the light source 131 passes through a differential interference illumination system 132 that includes a beam spot shaping unit and an illumination lens. The light that passes through the differential interference illumination system 132 is linearly polarized light. The light that has passed through the differential interference illumination system 132 is incident on a quarter-wave plate 134 via a beam splitter 133. The quarter-wave plate 134 is installed so that its fast axis is at an angle of 45° with respect to the direction of incident polarization. The light that passes through the quarter-wave plate 134 becomes circularly polarized light. The light that has passed through the quarter-wave plate 134 is incident on a Nomarski prism 135.

[0030] The Nomarski prism 135 is made of a birefringent optical material and separates the circularly polarized incident light into two linearly polarized lights l1 and l2 with mutually orthogonal vibration planes, for example, in the X direction. The linearly polarized lights l1 and l2 are, for example, S-polarized and P-polarized lights, respectively. The linearly polarized lights l1 and l2 separated by the Nomarski prism 135 enter an objective lens 136 for DIC. This objective lens 136 is mounted on a stage (not shown), and its pupil position coincides with the separation position of the Nomarski prism 135. The two linearly polarized lights l1 and l2 that pass through the objective lens 136 travel parallel to each other and are irradiated perpendicularly onto the sample surface, forming a beam spot 2 consisting of two polarized illumination spots 2a and 2b.

[0031] The Nomarski prism 135 can be moved in the X direction by a drive mechanism (not shown), and the phase difference between the separated linearly polarized beams 11 and 12 can be adjusted by adjusting its position in the X direction. The separation width of the linearly polarized beams 11 and 12 is called the shear amount δ.

[0032] As shown in an exaggerated manner in Figure 6, if there is a step in the polarized illumination spots 2a, 2b, i.e., a difference in height in the direction of propagation of the linearly polarized light l1, l2, the phase difference between the linearly polarized light l1, l2 will change. In DIC measurement, a larger shear amount δ increases the contrast, but reduces the measurable height difference per unit distance in the horizontal direction. In this embodiment, emphasis is placed on contrast, and the shear amount δ is set larger than the optical resolution and sampling interval of the interference light measurement device 130.

[0033] In DIC measurement, the height difference (differential height Δh) between polarized illumination spots 2a and 2b is measured based on the phase difference between linearly polarized light l1 and l2 reflected from the surface of sample 1. The linearly polarized light l1 and l2 reflected from the surface of sample 1 are collimated by objective lens 136 and recombined into the same optical path by Nomarski prism 135 to become interference light, which enters interference light sensor 138 via imaging lens 137. In this embodiment, a polarizing beam splitter 139 is placed after imaging lens 137 to split the interference light into two orthogonal polarization directions, and the interference intensities of these two interference lights are measured by different interference light sensors 138.

[0034] The interference light sensor 138 is a photoelectric conversion element like the scattered light sensor 128, but because it detects light directly reflected from the sample surface, it can have a lower gain than the scattered light sensor 128. The interference light sensor 138 can be a point sensor, area sensor, or multi-line sensor, and the sensor type can be a photomultiplier tube, SiPM, CMOS sensor, CCD, or the like. Because the interference intensity of the interference light changes depending on the differential height Δh, the interference intensity of the interference light measured by the interference light sensor 138 can be AD converted and processed by, for example, a signal processing unit 190 (FIG. 1), thereby making it possible to measure the differential height Δh.

[0035] DIC measurement by the interference light measurement device 130 can be performed simultaneously (during the same scan) with scattered light measurement by the scattered light measurement device 120 by wavelength separation or spatial separation. This enables faster measurement of the surface of the sample 1. This also applies when a surface shape measurement method other than DIC measurement is applied to the measurement method of the interference light measurement device 130.

[0036] [Signal processing device 200] The signal processing device 200 is one or more computers that process signals from the scattered light measurement device 120 and the interference light measurement device 130. In this embodiment, the signal processing device 200 includes a data input unit 210, a data processing unit 220, and the signal processing unit 190 of the inspection device 100. For example, measurement data of the sample surface acquired by the detection optical system 122 and the detection optical system 130b and processed by the signal processing unit 190 is input to the data input unit 210. The data processing unit 220 calculates an evaluation value of the microroughness of the sample 1 based on the data input to the data input unit 210. The example of FIG. 1 illustrates a case in which the signal processing device 200 includes the signal processing unit 190 of the inspection device 100 and is configured by multiple computers. However, the functions of the data input unit 210, the signal processing unit 190, and the data processing unit 220 may be provided in a single computer, and the signal processing device 200 may be configured by a single computer.

[0037] [Micro-roughness evaluation procedure] The evaluation of the microroughness of the sample 1 performed by the data processing unit 220 will now be described. In this embodiment, the signal processing device 200 (for example, the data processing unit 220) calculates a first evaluation value of the microroughness of the sample 1 based on the signal from the interference light measurement device 130. At the same time, the signal processing device 200 calculates a scattering characteristic signal based on the signal from the scattered light measurement device 120, and calculates a second evaluation value of the microroughness based on the first evaluation value and the scattering characteristic signal for a spatial frequency band in which the first evaluation value is not calculated by the interference light measurement device 130. In this embodiment, the signal processing device 200 calculates the first and second evaluation values ​​of the microroughness based on detection signals of the interference light and scattered light that are generated simultaneously.

[0038] The first evaluation value and the second evaluation value are values ​​that correlate with the microroughness of the surface of the sample 1. The microroughness can be calculated from the first evaluation value and the second evaluation value. In the example described below, a case will be described in which PSD (Power Spectral Density) data of the surface of the sample 1 is calculated as the first evaluation value and the second evaluation value, and a haze value is calculated as the scattering property signal.

[0039] Furthermore, the upper limit of the spatial frequency band associated with the second evaluation value calculated from the scattering characteristic signal is higher than the upper limit of the spatial frequency band associated with the first evaluation value calculated from the signal of the interference light measurement device 130. In particular, this embodiment describes an example in which the lower limit of the spatial frequency band associated with the second evaluation value is higher than the upper limit of the spatial frequency band associated with the first evaluation value, and the spatial frequency band in which the first PSD data can be acquired does not overlap with the spatial frequency band in which the second PSD data can be acquired.

[0040] The PSD data or haze value can be obtained by dividing the entire surface of the sample 1 into a plurality of processing unit areas and processing each processing unit area with the signal processing unit 190. In order to evaluate the microroughness with high precision, it is desirable for the signal processing device 200 to calculate the second evaluation value based on the first evaluation value (PSD data) and scattering property signal (haze value) for the same area of ​​the same sample.

[0041] When the surface shape of sample 1 is expressed in three-dimensional coordinates (X, Y, Z), the height Z is subjected to a two-dimensional Fourier transform with respect to (X, Y), and the value obtained by squaring the amplitude can be calculated as the spatial frequency spectrum. The spatial frequency spectrum is expressed as a function P(fx, fy) with the inverse (fx, fy) of (X, Y) as a variable. The PSD function P(fr) is obtained by expressing this spatial frequency spectrum P(fx, fy) in terms of fr. fr is a value calculated by fr = √(fx × fx + fy × fy).

[0042] The PSD function P(fr) contains information about the magnitude and period of surface roughness. In other words, the PSD function is one of the functions that expresses the spatial frequency spectrum. The value of this PSD function (PSD data) is essentially equivalent to data on the surface shape of sample 1 related to micro-roughness. By integrating the PSD function P(fr) over any spatial frequency band (f1 to f2), the surface RMS roughness (root mean square roughness) of sample 1 can be determined.

[0043] The haze value is expressed as the ratio of the scattered light signal S1 to the amount of light incident on the beam spot 2 in the scattered light measurement device 120, and can be calculated by dividing each scattered light signal S1 measured by the scattered light intensity measurement systems 123-126 by the amount of incident light. The scattered light signal S1 can be the value of the signal output in real time from the scattered light intensity measurement systems 123-126, or it can be a value stored in the signal processing unit 190 and read out later for use.

[0044] 7 is a flowchart showing an example of the procedure for evaluating the microroughness of the sample 1 by the signal processing device 200. The flow in the figure is broadly divided into process 710 for calculating a first evaluation value of the microroughness, process 720 for calculating a scattering characteristic signal, and process 730 for calculating a second evaluation value of the microroughness. As will be described in detail later, in process 710, the first evaluation value is calculated from measurement data of the surface shape of the sample 1 based on the interference light. In process 720, a scattering characteristic signal of a predetermined spatial frequency in a predetermined spatial direction is calculated from signals from multiple scattered light intensity measurement systems 123-126. In process 730, a second evaluation value is calculated based on the first evaluation value and the scattering characteristic signal.

[0045] -Process 710- The process 710 for calculating the first evaluation value includes step 711 for calculating the differential height Δh, step 712 for calculating the surface shape of the sample 1, and step 713 for calculating the first evaluation value.

[0046] Step S711 The signal processing device 200 calculates the differential height Δh based on the change in the phase of the two linearly polarized light beams l1 and l2 of the light emitted from the light source 131. In this embodiment, the light recombined by the Nomarski prism 135 is split by polarization by the polarizing beam splitter 139 and detected by two interference light sensors 138. The time average of the interference intensity of the interference light separated and detected by the two interference light sensors 138 is calculated, and the phase shift of the linearly polarized light beams l1 and l2 is corrected based on the calculated time average interference intensity, thereby improving the calculation accuracy of the differential height Δh. The phase shift here refers to the amount of phase shift between the two linearly polarized light beams l1 and l2 that occurs due to factors other than the differential height Δh, such as the tilt of the sample 1 or fluctuations in the output of the light source 131.

[0047] Step S712 Next, the signal processing device 200 calculates the surface shape of the sample 1 based on the differential height Δh calculated in step S711. The differential height Δh calculated in step 711 is the difference in height between the polarized illumination spots 2a and 2b of the two linearly polarized light beams l1 and l2 that are separated by a shear amount δ. Therefore, by accumulating the data of the differential height Δh obtained by scanning for each shear amount δ, it is possible to calculate data on the surface shape of the sample 1.

[0048] Step S713 In the following step 713, the signal processing device 200 calculates a first evaluation value for microroughness based on the surface shape data of the sample 1 calculated in step 712. In this embodiment, the first evaluation value is calculated by Fourier transforming the surface shape data of the sample 1 calculated in step 712 to calculate PSD data. Hereinafter, this PSD data calculated as the first evaluation value based on the signal from the interference light measurement device 130 will be referred to as "first PSD data." In this embodiment, the first PSD data is calculated only in the shear direction, that is, only in the same spatial direction as the straight line passing through the centers of the two polarized illumination spots 2a and 2b. The spatial direction and spatial frequency band of the first PSD data are determined by the measurement technique for the surface shape of the sample 1 and the configuration of the interference light measurement device 130. For example, in DIC measurement, the upper limit of the spatial frequency band is determined by the Nyquist criterion for the larger of the two parameters: the spatial sampling interval and the optical system resolution.

[0049] DIC measurement is characterized by its low sensitivity to the first PSD data in the spatial frequency band corresponding to the shear amount δ. In this embodiment, because the shear amount δ is greater than the optical resolution, there exists a spatial frequency band that cannot be measured even if it is below the upper limit of the spatial frequency of the first PSD data. This is shown in FIG. 8. The horizontal axis of the figure represents fr, and the vertical axis represents the magnitude of the spatial frequency spectrum. As mentioned above, the upper limit 804 of the spatial frequency band 803 measurable by the interference light measurement device 130 is determined by the resolution and sampling interval of the optical system. Even for bands lower than this upper limit 804, the sensitivity is low for a specific band 806 including a spatial frequency 805 corresponding to the shear amount δ, due to the principle of DIC, which measures the difference in height (differential height Δh) between two beam spots separated by the shear amount δ. While it is possible to not calculate the first PSD data in band 806, in this embodiment, interpolation can be performed based on the first PSD data surrounding band 806.

[0050] -Process 720- Next, a description will be given of process 720 for calculating a haze value as a scattering characteristic signal from the scattered light signal S1. Process 720 includes step 721 for acquiring the scattered light signal S1, and step 722 for calculating the scattering characteristic signal from the scattered light signal S1.

[0051] Step 721 In step 721, the signal processing device 200 acquires the scattered light signal S1 measured by the scattered light intensity measurement systems 123-126. As described above, the scattered light signal S1 can use the value of the signal output in real time from the scattered light intensity measurement systems 123-126, or it can be a value stored in the signal processing unit 190 and read out later for use.

[0052] Step 722 In the next step 722, the signal processing device 200, for example in the signal processing unit 190, divides the scattered light signal S1 of each scattered light intensity measurement system 123-126 acquired in step 721 by the amount of incident light. This means calculating the ratio of the scattered light signal S1 of each scattered light intensity measurement system 123-126 to the amount of incident light. This signal ratio is the haze value.

[0053] -Process 730- Next, a process 730 for calculating a second evaluation value of microroughness based on the first evaluation value of microroughness and the scattering property signal will be described.

[0054] In this embodiment, the haze value is converted into a bidirectional reflectance distribution function (BRDF) that represents the reflection and scattering characteristics of the material surface, and is stored together with this BRDF in the signal processing device 200. This BRDF has information about spatial frequency and spatial direction. With reference to FIG. 4 described above, the BRDF is defined by the following (Equation 1) using the incident light amount Ii on the beam spot 2, the elevation angle θs of the scattered light beam, the solid angle ω, and the detected light amount Iω.

[0055]

number

[0056] Moreover, the BRDF can be modeled as shown in (Equation 2).

[0057]

number

[0058] θi is the elevation angle of the incident light beam relative to the beam spot 2, and λ is the wavelength of the incident light and scattered light. Q is a parameter determined by the refractive index of the sample 1, the elevation angle θi of the incident light beam, the elevation angle θs of the scattered light beam, and the azimuth angle φs. The spatial frequencies fx and fy of the PSD function on the right side of (Equation 2) are expressed as (Equation 3) using the parameters of the incident light beam and scattered light beam.

[0059]

number

[0060] As shown in (Equation 1), the BRDF can be calculated from the haze value (Iω / Ii) obtained by scattered light measurement and the parameters of the incident and outgoing light fluxes. This BRDF is related to the PSD function as shown in (Equation 2). All terms on the right side of (Equation 2) other than the PSD function are parameters determined by the configurations of the inspection device 100 and the sample 1. Therefore, in principle, PSD data of a predetermined spatial frequency in a predetermined spatial direction can be calculated from the haze value and parameters of the inspection device 100 and the sample 1. In this embodiment, the PSD data calculated based on this haze value is referred to as "second PSD data."

[0061] However, the haze value actually measured varies depending on the reflectance of the sample 1 and fluctuations in the output of the light source 131, and the second PSD data is affected by these factors. As a result, a discrepancy may occur between the first PSD data and the second PSD data. Therefore, in order to quantitatively evaluate the microroughness of the sample 1 using the first PSD data and the second PSD data, a calibration algorithm for the second PSD data is required.

[0062] 9A to 9C are explanatory diagrams of functions used as PSD models. Each model shown in Fig. 9A to 9C is represented on a double logarithmic graph, with the horizontal axis representing fr and the vertical axis representing the magnitude of the spatial frequency spectrum.

[0063] The first PSD model shown in FIG. 9A is called the ABC model. The ABC model uses parameters A, B, and C for the spatial frequency fr of the surface roughness, and is expressed as PSD(fr)=A / (1+Bfr 2 ) C / 2 In this ABC model of microroughness, the PSD has a constant value in a specific low-frequency spatial frequency band (a band below 1 / B), and in a high-frequency band (a band above 1 / B), the PSD decreases monotonically according to fr. The constant value of the PSD in the low-frequency band is A, the slope of the PSD in the high-frequency band is -C / 2, and the spatial frequency at which the PSD changes from a constant value to a monotonically decreasing value is 1 / B.

[0064] The ABC model is not limited to the example shown in FIG. 9A, and examples include the second PSD model shown in FIG. 9B and the third PSD model shown in FIG. 9C. The ABC model shown in FIG. 9B is called a Fractal ABC model, and uses parameters A, B, C, K, and M, and is expressed as PSD(fr)=A / (1+Bfr 2 ) C / 2 +K / fr M The Fractal ABC model is characterized by the fact that in the low frequency band of the ABC model shown in Figure 9A, the PSD increases with a decrease in fr, with an intercept K and a slope -M. The ABC model shown in Figure 9C is called the Double ABC model, and uses parameters A1, B1, C1, A2, B2, and C2, and is expressed as PSD(fr) = A1 / (1 + B1fr 2 ) C1 / 2 +A2 / (1+B2fr 2 ) C2 / 2 The Double ABC model is the sum of two different ABC models.

[0065] In consideration of the above, in process 730, the signal processing device 200 calculates second PSD data based on the first PSD data and the Haze value using model functions such as those illustrated in Figures 9A to 9C. In this embodiment, a case will be described in which the ABC model illustrated in Figure 9A is used.

[0066] Process 730 includes steps 731 to 734. Specifically, in step 731, the signal processing device 200 calculates a portion (A in this example) of model function parameters (hereinafter referred to as model parameters) representing the PSD of the surface of the sample 1 for the ABC model based on the first PSD data. In the following step 732, the signal processing device 200 calculates a calibration coefficient used to convert the haze value to PSD for a predetermined spatial direction and spatial frequency band, based on the correspondence between the model parameters and the haze value calculated in step 731. Then, in step 733, the signal processing device 200 calibrates the haze value using the calibration coefficient and calculates second PSD data as a second evaluation value of the microroughness of the sample 1. Finally, in step 734, the signal processing device 200 calculates the remaining model parameters based on the first PSD data and the second PSD data. Regarding process 730, the procedure for determining the model function, spatial direction, spatial frequency band, and parameters used in the calculation varies depending on the characteristics of the sample 1 and the configuration of the inspection device 100, and is therefore not necessarily limited to the procedure illustrated in FIG. 7.

[0067] The diagram shown in Figure 10A is a double logarithmic graph, with the horizontal axis representing fr, the vertical axis (left) representing the magnitude of the spatial frequency spectrum, and the vertical axis (right) representing the signal ratio between the amount of incident light and the scattered light signal. In Figure 10A, the common horizontal axis represents the first PSD data represented by the frequency spectrum and the haze value represented by the signal ratio. Steps 731-734 of process 730 illustrated in Figure 7 will be explained using Figures 10A and 10B.

[0068] Step 731 In step 731 of process 730, the signal processing device 200 first calculates some model parameters for the model function of Fig. 9A, in this example, parameter A, which is a constant value in the low-frequency band, from the first PSD data. Parameter A can be calculated as a statistical value such as the mean value or median value of the first PSD data.

[0069] Step 732 In the next step 732, the signal processing device 200 calculates a calibration coefficient for converting a haze value included in a spatial frequency band 1001 expressed by the parameter A, for example, a haze value 1002 equal to or lower than a predetermined spatial frequency set in advance, into a PSD expressed by the parameter A. In this case, the calibration coefficient is calculated so that a statistical value such as the average value or median value of the haze value 1002 matches the PSD expressed by the parameter A.

[0070] Step 733 Next, in step 733, signal processing device 200 converts all haze values, including haze value 1002, into PSD data using the calculated calibration coefficient, and calculates second PSD data. As a result, PSD data for both the frequency band measurable by interference light measurement device 130 and the frequency band measurable by scattered light measurement device 120 can be obtained, as shown in Fig. 10B. Specifically, PSD data for the frequency band measurable by interference light measurement device 130 (first PSD data) and PSD data for the measurable frequency band expanded by cooperation between interference light measurement device 130 and scattered light measurement device 120 (second PSD data) are acquired.

[0071] Step 734 Next, in step 734, the signal processing device 200 calculates the remaining model parameters B and C based on the second PSD data calculated in step 733. If necessary, the RMS roughness can also be calculated by integrating the model function 1003 constructed based on the parameters A, B, and C in an arbitrary spatial frequency range f1 to f2.

[0072] The signal processing device 200 outputs the data calculated in the process shown in Fig. 7 to the display device (monitor) 230 in a timely or sequential manner, and displays the calculation process or results numerically or graphically, allowing the operator to confirm the validity of the measurement and calibration.

[0073] [Handling of haze values ​​with different shear and haze directions] Signal processing device 200 calculates a second evaluation value based on a scattering property signal whose spatial direction corresponds to the first evaluation value (a haze value corresponding to microroughness in the shear direction). Here, several examples of how to handle haze values ​​corresponding to microroughness in a direction different from the shear direction are shown using FIGS. 11A to 11C. The example shown in FIG. 11A is referred to as the first example, the example shown in FIG. 11B as the second example, and the example shown in FIG. 11C as the third example. That is, the first, second, and third examples are examples of how to handle haze values ​​whose spatial direction differs from the first evaluation value when calculating a second evaluation value. The left and right diagrams shown in FIGS. 11A to 11C are double logarithmic graphs, with the horizontal axis representing fr and the vertical axis representing the magnitude of the spatial frequency spectrum.

[0074] As mentioned above, the scattered light intensity measurement systems 123-126 each measure scattered light in a different spatial direction, and therefore the signal from a specific scattered light intensity measurement system matches the spatial direction of the first PSD data. The left diagrams in each of Figures 11A to 11C show a mixture of haze values ​​1101 whose spatial direction matches or has a certain degree of approximation to the first PSD data, and haze values ​​1102 whose spatial direction differs from that of the first PSD data. For convenience of illustration, the number of scattered light sensors 128 shown in FIGS. 1 and 3 differs from the number of haze values ​​shown in the left diagrams of FIGS. 11A-11C, but in reality they are the same number (for example, a dozen).

[0075] 11A is an example in which a correction coefficient set in advance according to the characteristics of the sample 1 and the configuration of the inspection apparatus 100 is used to correct the haze value 1102 by the difference between the haze values ​​1101 and 1102 due to differences in spatial directions, as indicated by the white arrow in the right diagram of the figure. In this case, the correction coefficient can be obtained, for example, by performing AFM measurement in advance and comparing the AFM measurement with scattered light measurement. The haze value 1101 shown in the right diagram of FIG. 11A, including the value obtained by correcting the haze value 1102, is used to calculate the second evaluation value.

[0076] 11B is an example in which, under the assumption that the microroughness of sample 1 is isotropic, a haze value 1102 having a different spatial direction from the first PSD data is treated in the same way as a haze value 1101 having the same spatial direction as the first PSD data. The haze value 1102 is included in the basis for calculating the second evaluation value, just like the haze value 1101 (right diagram in FIG. 11B), and the second evaluation value is calculated from the haze values ​​1101 and 1102.

[0077] 11C is an example in which the haze value 1102, which has a different spatial direction from the first PSD data, is excluded. The haze value 1102 is excluded from the basis for calculating the second evaluation value (right diagram in FIG. 11C), and the second evaluation value is calculated based only on the haze value 1101.

[0078] [others] In this embodiment, the shear direction is unidirectional, and in process 710 (FIG. 7), the first PSD data is calculated only for the spatial direction that is the same as the single shear direction. In other words, the case where the first evaluation value in process 710 is a one-dimensional value has been described. However, DIC measurement does not necessarily provide only one-dimensional roughness evaluation values. For example, the Nomarski prism 135 in FIG. 6 can be replaced with two Nomarski prisms with orthogonal polarization directions to separate the light into four linearly polarized beams with shear amounts δ in the X and Y directions, i.e., the linearly polarized beams l1 and l2 can be further separated in the Y direction. With this configuration, two-dimensional first PSD data can be calculated.

[0079] Furthermore, when the sample 1 is highly flat, such as when the sample 1 is a wafer, even if the shear direction is a single direction, PSD data can be obtained in the spatial direction (orthogonal direction 1202) perpendicular to the shear direction 1201, as shown in FIG. 12 . FIG. 12 shows a portion of the sample surface as a collection of rectangular sampling points 1203. In this embodiment, the surface shape, i.e., height, is calculated by accumulating differential heights in the shear direction 1201. In this case, the surface shape in the orthogonal direction 1202 can also be calculated by repeating the scan of the sample 1. However, if there is an error in setting the initial value for the height accumulation calculation, accuracy will be lost by the amount of the error. On the other hand, when the surface of the sample 1 is highly flat, the signal processing device 200 can suppress the setting error of the initial value, thereby improving the calculation accuracy of the surface shape of the sample 1. For example, height measurements are performed in a region of a set area where a certain level of flatness is expected, and statistical values ​​(such as the average value) are calculated. These values ​​can then be used as initial values ​​that are expected to have a certain level of accuracy. It is also possible to perform processes such as smoothing and fitting at sampling points 1203 adjacent to the orthogonal direction 1202. By increasing the accuracy of setting the initial values ​​in this way, highly reliable PSD data can be obtained also in the orthogonal direction 1202.

[0080] -effect- (1) The interference light measurement by the interference light measurement device 130 can be performed simultaneously (at the same scanning opportunity) with the scattered light measurement of the entire surface of the sample 1 by the scattered light measurement device 120, and the interference light and scattered light can be measured at high speed over the entire surface of the sample 1. In particular, when the sample 1 is rotated and scanned in a spiral orbit as shown in Figure 2, the scanning does not involve a reciprocating motion, and therefore the measurement can be performed at higher speed.

[0081] Here, interferometric light measurement can measure microroughness faster than AFM, which is generally used for measuring microroughness, but on the other hand, it has lower resolution than AFM, so from the perspective of resolution, AFM measurements cannot simply be replaced by interferometric light measurement.

[0082] In contrast, the maximum spatial frequency of microroughness measurable using scattered light is generally higher than the maximum spatial frequency of microroughness measurable using interference light. As described above, both interference light and scattered light can be measured simultaneously and at high speed when scanning the sample 1. Therefore, by calculating a first evaluation value of microroughness using interference light measurement and calculating a second evaluation value of microroughness using scattered light measurement for the spatial frequency band where the first evaluation value is not calculated, the lack of resolution of the interference light measurement can be compensated for. As described above, a discrepancy may occur between the value based on scattered light and the value based on interference light, so treating them as such is not desirable. However, in this embodiment, the discrepancy between the two is corrected by calculating the calibration coefficient as described above. This allows the first evaluation value and the second evaluation value to be treated in the same way. Furthermore, since the calibration coefficient can be calculated sequentially from the scanning data of the sample 1 using the scattered light measurement device 120 and the interference light measurement device 130, there is no need to separately collect basic data for calculating the calibration coefficient.

[0083] As described above, according to this embodiment, by quickly calculating the first evaluation value and the second evaluation value, and by calculating the second evaluation value in addition to the first evaluation value, it is possible to measure the microroughness of the entire surface of the sample quickly and with high resolution.

[0084] (2) In addition, in an inspection system for inspecting sample defects, a sphere centered on the beam spot may be divided into multiple regions, and a scattered light intensity measurement system (such as a scattered light sensor) may be placed in each region to detect light scattered in various directions from the sample surface. In other words, these multiple scattered light intensity measurement systems detect scattered light in each spatial direction, allowing data on scattered light with different spatial directions and spatial frequencies to be measured. The spatial direction and spatial frequency of microroughness, which correlates with the haze value, are determined by the incident angle, exit angle, and wavelength of light relative to the sample surface. Therefore, a sample surface quality control system can be constructed by adding an interference light measurement device 130 to a defect inspection system. Cooperation between the scattered light measurement device 120 and the interference light measurement device 130 enables high-speed measurement of microroughness over a wide spatial frequency band.

[0085] (Second embodiment) The second embodiment illustrates a case where a Michelson interferometer is used as a method for measuring the surface shape of the sample 1. In this embodiment, the spatial frequency band in which the first PSD data can be acquired and the spatial frequency band in which the second PSD data can be acquired partially overlap. In other words, the upper limit of the spatial frequency band associated with the first evaluation value calculated from the signal of the interference light measurement device 130 is lower than the upper limit of the spatial frequency band associated with the second evaluation value calculated from the scattering property signal, as in the first embodiment, but is higher than the lower limit of the spatial frequency band associated with the second evaluation value.

[0086] 13 is a schematic diagram showing an example of the configuration of an interference light measurement device 130 in this embodiment. In Fig. 13, elements that are the same as or correspond to those in the first embodiment are given the same reference numerals as in the previously mentioned drawings, and descriptions thereof will be omitted as appropriate.

[0087] The interference light measurement device 130 of this embodiment is configured to include a light source 131, a bandpass filter 232, an interference objective lens optical system 233, and an interference light sensor 138. The interference objective lens optical system 233 includes a beam splitter 133 and a reference surface (reflection mirror) 235. The interference objective lens optical system 233 is driven by a driving device 236, and is displaced, for example, in a direction toward or away from the sample 1. The interference light measurement device 130 of this embodiment is not provided with a Nomarski prism 135.

[0088] In this embodiment, the light source 131 is a white light source. Light emitted from the light source 131 enters the interference objective lens optical system 233 and is split into two by the beam splitter 133. One of the split light beams is incident on the sample 1, and the other light beam is incident on the reference surface 235. The light beams reflected by the sample 1 and the reference surface 235 are combined by the beam splitter 133 and guided to the interference light sensor 138. In this embodiment, the interference light sensor 138 is a two-dimensional sensor. In this case, the interference light sensor 138 can measure the interference light intensity of the reflected light from the sample 1 and the reflected light from the reference surface 235 while the interference objective lens optical system 233 is displaced by the drive device 236. The signal processing device 200 calculates the surface shape of the sample 1 based on the interference light intensity distribution measured by the interference light sensor 138.

[0089] In this embodiment, two-dimensional height information of the surface of the sample 1 within the field of view can be obtained at once. As in the first embodiment, the entire surface of the sample 1 can be scanned by a combination of rotational and linear scanning (FIG. 2), or by a combination of linear scanning in two directions.

[0090] Process 730 (FIG. 7) in this embodiment will be described using FIGS. 14A and 14B. The graph shown in FIG. 14A is a double logarithmic graph, with the horizontal axis representing fr, the vertical axis (left) representing the magnitude of the spatial frequency spectrum, and the vertical axis (right) representing the signal ratio between the amount of incident light and the scattered light signal. In FIG. 14A, the common horizontal axis represents the first PSD data represented by the frequency spectrum and the haze value represented by the signal ratio. The graph shown in FIG. 14B is also a double logarithmic graph, with the horizontal axis representing fr and the vertical axis representing the magnitude of the spatial frequency spectrum.

[0091] 14A, in this embodiment, the spatial frequency band in which the first PSD data is measured and the spatial frequency band in which the haze value is measured overlap in a predetermined band 1401. In other words, the upper limit value of the spatial frequency band in which the first PSD data is measured is lower than the upper limit value of the spatial frequency band in which the haze value is measured, but higher than the lower limit value of the spatial frequency band in which the haze value is measured.

[0092] In calculating the second evaluation value in this embodiment, the signal processing device 200 first calculates a calibration coefficient used to convert the haze value to a PSD so that the first PSD data 1402 and the haze value 1403 included in the predetermined band 1401 match. For example, statistical values ​​such as the average value and median value are obtained for each of the first PSD data 1402 and the haze value 1403, and a calibration coefficient is calculated so that the PSD based on the statistical value of the haze value 1403 matches the statistical value of the first PSD data 1402. Thereafter, the signal processing device 200 converts the haze values ​​inside and outside the predetermined band 1401 into PSDs using this calibration coefficient, and calculates the second PSD data.

[0093] 14B, PSD data can be obtained for the frequency band measurable with scattered light in addition to the frequency band measurable with interference light. If necessary, model parameters for a model function (such as an ABC model) can be obtained from the first PSD data and the second PSD data, as in the first embodiment. Furthermore, microroughness can be calculated by integrating the model function over a predetermined spatial frequency band.

[0094] According to this embodiment, in addition to the same effects as the first embodiment, there is an advantage that by measuring the surface shape of the sample 1 using a Michelson interferometer, two-dimensional height information of the surface of the sample 1 can be obtained at once.

[0095] Furthermore, by making the spatial frequency band of the first PSD data measured using interference light partially overlap with the spatial frequency band of the second PSD data calculated based on scattered light, it is not necessary to use a model function to calculate the calibration coefficient. In other words, in the above-mentioned predetermined band 1401, the matching target of the second PSD data based on the haze value 1403 is the PSD data itself (second PSD data 1402) measured by the same inspection device 100. This can be expected to improve the calculation accuracy of the second PSD data, and ultimately the calculation accuracy of the microroughness.

[0096] (Third embodiment) In the first embodiment, the interference light measurement device 130 employs DIC measurement, splits light emitted from a light source 131 into two linearly polarized beams with a predetermined shear amount, irradiates the sample 1 with the split beams, and measures the interference intensity of the interference light generated in the sample 1 with an interference light sensor 138. This configuration is similar to that of the present embodiment. This embodiment differs from the first embodiment in that, while the shear amount δ of the DIC measurement in the first embodiment is greater than the optical system resolution of the interference light measurement device 130, this embodiment uses interference light with a shear amount smaller than the optical system resolution of the interference light measurement device 130. Specifically, in this embodiment, the signal processing device 200 calculates a first evaluation value based on a signal related to interference light with a shear amount larger than the optical resolution of the interference light measurement device 130 and a signal related to interference light with a shear amount smaller than the optical resolution of the interference light measurement device 130. A specific example will be described with reference to the drawings.

[0097] Fig. 15 is a schematic diagram showing an example of the configuration of an interference light measurement device provided in a sample surface quality control device according to this embodiment. In Fig. 15, elements that are the same as or correspond to those in the first embodiment are given the same reference numerals as those in the previously mentioned drawings, and descriptions thereof will be omitted as appropriate.

[0098] Similar to the first embodiment, the interference light measurement device 130 of this embodiment includes a light source 131, a differential interference illumination system 132, a beam splitter 133, a quarter-wave plate 134, a Nomarski prism 135, an objective lens 136, an imaging lens 137, and an interference light sensor 138. The interference light measurement device 130 of this embodiment differs from the first embodiment in that the light source 131 is a multi-wavelength light source that emits two monochromatic lights with different wavelengths, and is configured to obtain interference light for each monochromatic light.

[0099] The interference light measurement device 130 of this embodiment is equipped with a dichroic mirror 1501 as an optical element that separates the light emitted from the light source 131 into two monochromatic lights, and is equipped with differential interference optical systems 1511 and 1512 corresponding to the two monochromatic lights. The differential interference optical system 1511 is an optical system including a quarter-wave plate 134, a Nomarski prism 135, and an objective lens 136. The differential interference optical system 1512 is an optical system similar to the differential interference optical system 1511, and is equipped with a quarter-wave plate 134', a Nomarski prism 135', and an objective lens 136'. However, the shear amount δ1 of the Nomarski prism 135 is designed to be larger than the optical resolution of the interference light measurement device 130, and the shear amount δ2 of the Nomarski prism 135' is designed to be smaller than the optical resolution of the interference light measurement device 130.

[0100] In this embodiment, light passing through the beam splitter 133 is separated into two monochromatic lights by a dichroic mirror 1501, and each monochromatic light enters a differential interference optical system 1511, 1512, which then obtains an interference light in the same manner as in the first embodiment. These interference lights are combined via the dichroic mirror 1501 and enter the interference light sensor 138. Although not particularly shown, the interference light can also be separated according to wavelength as needed and measured by multiple interference light sensors 138.

[0101] The other configurations are the same as those in the first embodiment.

[0102] In this embodiment, in addition to the same effects as in the first embodiment, the following effects can be obtained: The new effects obtained in this embodiment will be described with reference to FIG.

[0103] FIG. 16 is an explanatory diagram of a spatial frequency band that cannot be measured by DIC measurement using a sample surface quality control system according to a third embodiment of the present invention, and corresponds to FIG. 8 for the first embodiment. The horizontal axis of the diagram represents fr, and the vertical axis represents the magnitude of the spatial frequency spectrum. As mentioned above, an upper limit 1604 of a spatial frequency band 1603 that can be measured by the interference light measurement system 130 of this embodiment is determined by the resolution and sampling interval of the optical system. As mentioned above, even in a band lower than the upper limit 804, sensitivity decreases for a specific band 1606 that includes a spatial frequency 1605 corresponding to the shear amount δ1.

[0104] However, this embodiment is equipped with a differential interference optical system 1512 in which the shear amount δ2 is smaller than the optical resolution. The spatial frequency 1607 corresponding to the shear amount δ2 is greater than the upper limit 1604. Therefore, the differential interference optical system 1512 can perform measurements with reasonable sensitivity over the entire spatial frequency band 1603. On the other hand, a differential interference optical system 1512 with a small shear amount δ2 will, in principle, result in reduced contrast in differential height measurement.

[0105] In this embodiment, the differential interference optical system 1511 can perform high-contrast measurement of the spatial frequency band 1603 excluding the band 1606, while the differential interference optical system 1512 can perform high-sensitivity measurement of the band 1606 where the sensitivity is reduced in the differential interference optical system 1511. In this way, this embodiment can perform high-sensitivity measurement over a wider spatial frequency band than the first embodiment.

[0106] (Fourth embodiment) Fig. 17 is a schematic diagram showing an example of the configuration of a scattered light intensity measurement system provided in a sample surface quality control device according to a fourth embodiment of the present invention. In Fig. 17, elements that are the same as or correspond to those in the first embodiment are given the same reference numerals, and descriptions thereof will be omitted where appropriate. In this embodiment, an example of the configuration and processing when an imaging optical system is included in the scattered light intensity measurement systems 123-126 will be described.

[0107] 17, the scattered light intensity measurement systems 123-126 in this embodiment include a focusing optical system 1701 that focuses the light scattered from the surface of the sample 1, and an imaging optical system 1702 that forms an image of the beam spot 2 on the light receiving surface of the scattered light sensor 128. Although not shown in FIG. 17, a focus adjustment drive device that drives at least one of the sample 1, the focusing optical system 1701, the imaging optical system 1702, and the scattered light sensor 128 may also be provided.

[0108] As in the first embodiment, the scattered light sensor 128 in this embodiment may be a photomultiplier tube, an avalanche photodiode array, or a photon counting array, and a point sensor, area sensor, or multi-line sensor may be used. In this embodiment, a line sensor is used as the scattered light sensor 128, for example. The scattered light sensor 128 obtains a scattered light signal for each pixel.

[0109] In process 720 (FIG. 7), while in the first embodiment the signal processing device 200 calculated the haze value by dividing the scattered light signal by a uniform amount of incident light, in this embodiment the haze value is calculated by dividing the scattered light signal by an amount of incident light that varies for each pixel of the scattered light sensor 128. The amount of incident light for each sensor pixel is calculated based on the intensity profile of the beam spot 2 and the parameters of the focusing optical system 1701 and the imaging optical system 1702.

[0110] Figure 18 is a conceptual diagram showing the partial overlap of beam spots between adjacent scanning tracks. Here, we will explain how to handle areas that are measured more than once during scanning of the entire sample surface (i.e., areas where beam spots 2 overlap between adjacent scanning tracks).

[0111] 18 is a helical trajectory scan that combines a rotational movement in the direction of the arrow θ and a linear movement in the direction of the arrow R, as shown in FIG. 2. In this embodiment, the width of the beam spot 2 is large in the direction R and small in the direction of the arrow θ. Such an illumination profile can be generated by introducing a beam shaping unit using an anamorphic prism or a cylindrical lens into the illumination optical system 121 of the scattered light measurement device 120.

[0112] FIG. 18 shows measurement areas 1803 and 1804 with the same θ coordinate on the nth and (n+1)th rotations. The R coordinates of the measurement areas 1803 and 1804 are shifted by the distance between adjacent scanning tracks, that is, the distance the linear stage 113 moves during one rotation of the sample 1. FIG. 18 also illustrates an example in which the measurement areas 1803 and 1804 are divided and detected by eight pixels. In this embodiment, the length of the measurement area in the R direction is longer than the distance between adjacent scanning tracks, and an overlapping area 1805 between the measurement areas 1803 and 1804 is measured twice. Possible methods for measuring the overlapping area 1805 include using only the measurement values ​​of either the measurement areas 1803 or 1804, or integrating the haze values ​​related to the measurement values ​​of the measurement areas 1803 and 1804.

[0113] When integrating the haze values ​​associated with the measurement values ​​of measurement regions 1803 and 1804 for overlap region 1805, first, for each pixel in overlap region 1805, the sum of the scattered light signal from the nth measurement and the scattered light signal from the (n+1)th measurement is calculated. The pixels in overlap region 1805 are different between the nth and (n+1)th measurements. For example, in overlap region 1805, the coordinate measured at pixel Px2, the second pixel from the top in the figure, during the nth measurement is measured at pixel Px8, the eighth pixel from the top, during the (n+1)th measurement. Therefore, for this coordinate, the scattered light signal from pixel Px2 during the nth measurement and the scattered light signal from pixel Px8 during the (n+1)th measurement are added together. Regarding the amount of incident light, for each pixel in overlap region 1805, the amount of incident light during the nth measurement and the amount of incident light during the (n+1)th measurement are calculated based on the beam profile and then added together. Then, for each pixel in the overlap region 1805, the sum of the scattered light signals during measurements on the nth and (n+1)th rotations is divided by the sum of the incident light amounts during measurements on the nth and (n+1)th rotations, thereby calculating the haze value.

[0114] In other respects, this embodiment is similar to the first embodiment.

[0115] According to this embodiment, in addition to the same effects as those of the first embodiment, there is an advantage that the calculation accuracy of the haze value at each coordinate on the surface of the sample 1, and therefore the calculation accuracy of the second evaluation value, is improved.

[0116] (Fifth embodiment) In the first to fourth embodiments, the detection optical system 122 of the scattered light measurement device 120 has been described as including a plurality of scattered light intensity measurement systems 123-126. In the first to fourth embodiments, the scattered light intensity measurement systems 123-126 are arranged in different spatial directions, so that scattered light generated in the sample 1 is detected by different scattered light sensors 128 depending on the emission direction. In contrast, the present embodiment is an example in which the detection optical system 122 is provided with only one scattered light sensor.

[0117] 19A is a schematic diagram showing an example of the configuration of the detection optical system of the scattered light measurement device provided in the sample surface quality control device according to the fifth embodiment of the present invention. An example of the configuration of the detection optical system 122 having only one scattered light sensor will be described using FIG. 19A.

[0118] The detection optical system 122 of the scattered light measurement device 120 of this embodiment has only one scattered light intensity measurement system 175. The scattered light intensity measurement system 175 is configured to include a focusing optical system 1901, a detection optical system 1903, and a scattered light sensor 1904.

[0119] Scattered light generated at beam spot 2 on the sample surface is collected by a collecting optical system 1901. A detection optical system 1903 forms an image of a pupil plane 1902 located at the back focal position of the collecting optical system 1901 onto a scattered light sensor 1904. As with the first embodiment, the scattered light sensor 1904 in this embodiment can be a photomultiplier tube, an avalanche photodiode array, or a photon counting array, and an area sensor, multi-line sensor, etc. may also be used. In this embodiment, a two-dimensional array sensor is used for the scattered light sensor 1904.

[0120] 19B is an explanatory diagram of the relationship between the emission direction of scattered light and pixel coordinates (X, Y) on pupil plane 1902. In Fig. 19B, the emission angle of scattered light corresponding to pixel coordinates (X, Y) on pupil plane 1902, that is, coordinates (X, Y), can be calculated as follows.

[0121]

number

[0122]

number

[0123] Here, f is the focal length of the focusing optical system 1901. The definition of the scattered light emission angle is as explained above in FIG. 4. The coordinates on the light receiving surface of the scattered light sensor 1904 can be found by converting X and Y in equations 4 and 5 according to the optical parameters of the detection optical system 1903. Therefore, in combination with equation 3, it is possible to calculate the spatial direction and spatial frequency of roughness corresponding to the haze value for scattered light incident on each pixel on the light receiving surface of the scattered light sensor 1904, which is a two-dimensional array sensor.

[0124] According to this embodiment, by measuring the light intensity distribution on the pupil plane 1902, the emission angle distribution of scattered light can be measured with a single scattered light sensor 1904. Each pixel of the scattered light sensor 1904 corresponds to a different spatial direction and spatial frequency band. Therefore, in process 720 (FIG. 7) in the first to fourth embodiments, a specific pixel signal of the scattered light sensor 1904 can be used instead of the scattered light signal of a specific scattered light sensor corresponding to the target spatial direction, and the scattering property signal and second evaluation value can be calculated in the same way as in the first to fourth embodiments.

[0125] (Sixth embodiment) In the first to fifth embodiments, various data such as the first evaluation value and the second evaluation value acquired for the entire surface of the sample 1 can be stored in a memory and displayed on the display device 230 ( FIG. 1 ) by the signal processing device 200. In this embodiment, the display of data acquired for the entire surface of the sample 1 in the first to fifth embodiments will be exemplified. As described in the first embodiment, processing results such as PSD data ( FIG. 10B , etc.) and microroughness (data obtained through the processing of FIG. 7 ) are calculated for each processing unit area by the signal processing device 200 by dividing the entire surface of the sample 1 into multiple processing unit areas. All of these processing results for each processing unit area can be used for display on the display device 230 ( FIG. 1 ), or data obtained by statistical processing for any multiple processing unit areas can be used for display.

[0126] FIG. 20A is a diagram showing an example of a display in which the processing results are displayed as a surface map of the sample 1 for each display unit area. While FIG. 20A illustrates an example of a contour display of the processing results (sample surface height), it is also possible to use a method in which color-coded surface heights (numerical values) are displayed together with a color bar. The display example of FIG. 20A also includes an input field 2001 for specifying the spatial frequency band of the processing results to be displayed. While FIG. 20A illustrates an example in which the lower and upper limit values ​​of the spatial frequency band of the processing results to be displayed are specified numerically in the input field 2001, other methods may also be used, such as specifying the spatial frequency band using a control bar.

[0127] Fig. 20B is a diagram showing an example of a display showing the processing results in a histogram. In the example of Fig. 20B, the frequency (vertical axis) of the processing results (roughness) of a specified region of the sample surface specified in an input field (not shown) is displayed. In addition to frequency, a ratio or the like may also be used on the vertical axis.

[0128] FIG. 20C shows an example of a display showing the processing results (PSD data) in a scatter plot. In the example of FIG. 20C, PSD data (e.g., FIG. 10B) calculated based on the interference light and scattered light is displayed in a double logarithmic format, with spatial frequency on the horizontal axis and PSD magnitude on the vertical axis. In the example of FIG. 20C, PSD data for a spatial frequency band specified in, for example, input field 2001 is displayed. FIG. 20C illustrates an example of a method in which data for multiple samples 1 is plotted with different marks for each sample 1, making it easy to compare visually. Alternatively, a method of displaying a model function (e.g., model function 1003 in FIG. 10B) based on the calculated parameters may be used. This display example allows the operator to easily select an appropriate model function type that represents the PSD of the surface of sample 1.

[0129] Fig. 20D is a diagram showing an example of a display showing the processing results (model parameters) in a table format. In the example of Fig. 20D, only the model parameters calculated for the model function (such as model function 1003 in Fig. 10B) are displayed in a table format for each sample. In this display example, the data required for display, and therefore the data volume required for microroughness management for each sample, can be reduced.

[0130] (Variation) The present invention is not limited to the above-described embodiments and may include various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. It is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0131] The above-described configurations, functions, processes, processing means, etc. may be realized in part or in whole by hardware such as an integrated circuit. The above-described configurations, functions, etc. may also be realized by software, with a processor interpreting and executing a program that realizes each function. Information such as programs, tables, and files that realize each function can be stored in various storage media. Examples of various storage media include recording devices such as memory, hard disks, and solid-state drives (SSDs), as well as flash memory cards and digital versatile disks (DVDs).

[0132] In each embodiment, the signal input / output lines shown are those considered necessary for explanation, and do not necessarily show all of the signal input / output lines on the product. In reality, it can be considered that almost all components are interconnected. [Explanation of symbols]

[0133] 1... sample, 110... stage device, 120... scattered light measuring device, 130... interference light measuring device, 131... light source, 200... signal processing device, 230... display device, 1401... predetermined band, l1, l2... linearly polarized light, δ, δ2, δ2... shear amount

Claims

1. In a sample surface quality control device for measuring the micro-roughness of a sample, a stage device that holds the sample and moves the sample toward the sample surface; a scattered light measurement device for measuring scattered light generated by the sample; an interference light measurement device for measuring interference light including reflected light generated by the sample; a signal processing device that processes signals from the scattered light measurement device and the interference light measurement device, The signal processing device includes: calculating a first evaluation value of the microroughness of the sample based on the signal from the interference light measurement device; calculating a scattering characteristic signal based on the signal from the scattered light measuring device; For a spatial frequency band in which the first evaluation value is not calculated, a second evaluation value of the microroughness is calculated based on the first evaluation value and the scattering characteristic signal. Sample surface quality control device.

2. 2. The sample surface quality control device according to claim 1, the scattering characteristic signal is a haze value, The first evaluation value is PSD data. Sample surface quality control device.

3. 2. The sample surface quality control device according to claim 1, The stage device is a sample surface quality control device that moves the sample so that the entire surface of the sample is scanned.

4. 2. The sample surface quality control device according to claim 1, A sample surface quality control device in which the upper limit value of the spatial frequency band related to the second evaluation value calculated from the scattering characteristic signal is higher than the upper limit value of the spatial frequency band related to the first evaluation value calculated from the signal of the interference light measurement device.

5. 2. The sample surface quality control device according to claim 1, The sample surface quality control device, wherein the first evaluation value is calculated by a differential interference contrast method.

6. 2. The sample surface quality control device according to claim 1, the interference light measurement device splits light emitted from a light source into two linearly polarized lights with a predetermined shear amount, irradiates the light onto the sample, and measures the interference intensity of the interference light generated by the sample; The signal processing device calculates the first evaluation value based on a signal related to interference light having a shear amount larger than the optical resolution of the interference light measurement device and a signal related to interference light having a shear amount smaller than the optical resolution. Sample surface quality control device.

7. 2. The sample surface quality control device according to claim 1, The signal processing device is a sample surface quality control device that calculates the second evaluation value based on the first evaluation value and the scattering property signal related to the same region of the same sample.

8. 2. The sample surface quality control device according to claim 1, The signal processing device is a sample surface quality control device that calculates the first evaluation value and the second evaluation value based on detection signals of the interference light and the scattered light that are simultaneously generated.

9. 2. The sample surface quality control device of claim 1, The signal processing device includes: calculating a constant value based on the first evaluation value for a model function relating to the microroughness that takes a constant value in a predetermined spatial frequency band; calculating a calibration coefficient for converting the scattering characteristic signal into an evaluation value of the microroughness based on the constant value and the scattering characteristic signal in the predetermined spatial frequency band; Converting the scattering characteristic signal into the second evaluation value using the calibration coefficient. Sample surface quality control device.

10. 2. The sample surface quality control device of claim 1, a spatial frequency band associated with the first evaluation value and a spatial frequency band associated with the scattering characteristic signal overlap in a predetermined band, The signal processing device includes: calculating a calibration coefficient for converting the scattering characteristic signal into the evaluation value of the microroughness based on the scattering characteristic signal of the predetermined band and the first evaluation value; Converting the scattering characteristic signal into the second evaluation value using the calibration coefficient. Sample surface quality control device.

11. 2. The sample surface quality control device according to claim 1, The signal processing device is a sample surface quality control device that calculates the second evaluation value based on a signal of the scattering property signals that corresponds in spatial direction to the first evaluation value.

12. 12. The sample surface quality control device according to claim 11, The signal processing device corrects the scattering property signals having a spatial direction different from that of the first evaluation value using a preset correction coefficient, and includes the corrected signals as the basis for calculating the second evaluation value.

13. 12. The sample surface quality control device according to claim 11, The signal processing device includes, among the scattering property signals, signals having a spatial direction different from that of the first evaluation value as the basis for calculating the second evaluation value, in the same manner as signals corresponding to the spatial direction of the first evaluation value.

14. 12. The sample surface quality control device according to claim 11, The signal processing device excludes, from the basis for calculating the second evaluation value, signals of the scattering property signals that have a spatial direction different from that of the first evaluation value.

15. 2. The sample surface quality control device according to claim 1, A sample surface quality control device comprising a display device that displays the first evaluation value and the second evaluation value.

Images