TW200952103A - Inspection device - Google Patents

Abstract

An inspection device (1) includes: an illumination optical system (10) which applies an illumination light onto a surface of a wafer (W); a detection optical system (20) for observing the surface of the wafer (W) illuminated by the illumination light in an enlarged scale; detection elements (36a, 36b, 36c) for detecting luminance of reflected light from the wafer (W) on a pupil surface in the inspection optical system (20); a CPU (43) which inspects the surface of the wafer (W) in accordance with the luminance detected by the detection elements (36a, 36b, 36c); and a scan unit (60) which scans the range of the wafer (W) corresponding to the pupil surface on which the luminance is detected.

Description

The present invention relates to an inspection apparatus for detecting a state of a pattern formed on a surface of a substrate to be inspected during manufacture of a semiconductor element or a liquid crystal display element or the like. [Prior Art] Conventionally, a device that uses a reflected light emitted from a pattern formed on a surface of a substrate to be inspected (such as a semiconductor wafer or a liquid crystal glass substrate) to inspect defects such as unevenness or damage on the surface of the substrate is known. There are various proposals (for example, refer to Patent Document 1). In particular, in recent years, as the semiconductor process has been miniaturized, there has been a higher accuracy requirement for defect management of the substrate to be inspected. For example, when measuring the pattern width of the substrate to be inspected by SEM (scanning electron microscope), although the measurement accuracy is high, since the observation magnification is also high, it is necessary to sample a few points for measurement, and thus it is consumed in the measurement. lot of time. In response to this, there has been proposed a method of transmitting light of a predetermined wavelength emitted from a light source through a polarizing element and an objective lens, irradiating the surface of the substrate to be inspected by epi-illumination, using a transmissive objective lens, and satisfying the polarizing element and the orthogonal polarization. The image obtained by the photodetecting element under the condition of the mirror evaluates the reflected light from the substrate to be detected in the illumination. [Patent Document 1] Japanese Laid-Open Patent Publication No. 2000-155099. SUMMARY OF THE INVENTION However, even when the above method is used, since the observation magnification is high, it is necessary to move the observation range in order to inspect the entire substrate to be inspected in 200952103. At this time, in general, the stage holding the substrate to be inspected is moved to move the observation range. However, since the stage is moved more time, the inspection time is increased. The present invention has been made in view of the above problems, and an object thereof is to provide an inspection apparatus capable of performing inspection at a high speed with a high sensitivity.

In order to achieve the above object, an inspection apparatus according to the present invention includes: an illumination unit that irradiates illumination light onto a surface of a substrate to be inspected; and an observation optical system that amplifies observation of a surface of the substrate to be inspected by the illumination light; and a detection unit detects In the observation optical system, the brightness of the reflected light from the substrate to be detected is on the pupil plane; the inspection unit checks the surface of the substrate to be detected based on the brightness detected by the detecting unit; and the scanning unit scans and detects the brightness. The pupil plane corresponds to the range of the substrate to be inspected. Further, in the above invention, preferably, the scanning unit includes a plate-shaped member including an optical opening as a field stop, and a driving unit that drives the optical opening, and the optical opening is scanned by the driving unit. range. Further, in the above invention, it is preferable that the plate-like member is formed in a disk shape, and the plurality of hole portions as the optical opening π portion are respectively formed in the diametrical direction and the circumference different from each other in the plate-like structure 2. a direction position; the drive unit is configured to rotate the target axis of the plate-shaped member to "rotate the plate-like structure", and use the phase-moving portion to rotate the plurality of holes formed in the plate-shaped member = Forming a plurality of holes formed in the plate-shaped member from the plurality of holes in the plurality of holes, and sequentially aligning the direction of the linear movement direction on the 200952103-shaped member. The portion moves to scan the range sequentially from any of the plurality of holes. Further, in the above invention, it is preferable that the detecting unit detects the brightness of the portion of the pupil surface that is largely changed in brightness due to a change in the surface (4) of the substrate to be detected. Further, in the above invention, preferably, the illumination light is applied to linearly polarized light having a surface of the substrate to be detected having a reverse pattern; and the detecting portion detects that a polarization direction of light from the substrate to be detected is substantially orthogonal to the linearly polarized light. The polarizing component. Further, in the above invention, it is preferable that the illumination unit irradiates the illumination light onto the surface of the substrate to be detected by epi-illumination. According to the present invention, inspection can be performed at high speed with high sensitivity. [Embodiment] Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is an inspection apparatus of the present invention. The main structure of the inspection apparatus 本 according to the present embodiment includes the wafer stage 5, the objective lens 6, the half mirror 7, the illumination optical system 10, the detection optical system 20, the scanning unit 60, the imaging unit 3A, and Control unit 40. The wafer stage 5 is loaded with a semiconductor wafer W (hereinafter referred to as a wafer w) which is a substrate to be inspected in a state in which the pattern (reverse pattern) is formed to face upward. The wafer stage 5 is configured to be movable in three directions of x, y, and z axes orthogonal to each other (in addition, the upper and lower directions in Fig. 1 is the z-axis direction). Thereby, the 200952103 wafer stage 5' selects the wafer W to be movable in the X, y, and z axis directions. Further, the wafer stage 5 can be rotated around the z-axis. The illumination optical system 1 〇 'from the left side to the right side of FIG. 1 has, in order of its arrangement, a light source 1 1 (for example, a white LED or a halogen lamp, etc.), a condensing lens 12, an illuminance uniformization unit 13, and an aperture. The aperture 14, the field stop 15, the collimator lens 16, and the detachable polarizing element 17 (polarizing filter). Here, the light emitted by the light source 1 1 of the illumination optical system 10 is transparent.

The condensing lens 12 and the illuminance uniformizing unit 13 are guided to the aperture light blue 14 A and the field stop 15 . The illuminance uniformization unit 13 disperses the illumination light to uniformize the light amount distribution. Also, an interference filter can be included. The aperture stop 14 and the field stop 15' are configured such that the size and position of the opening can be changed with respect to the optical axis of the illumination optical system 1A. Therefore, the illumination optical system 1 can change the size and position of the illumination area and the adjustment of the aperture angle of the illumination by the operation of the aperture stop 14 and the field stop 15. Further, the light passing through the aperture stop 14 and the field stop 15 is collimated by the collimator lens 16 and then incident on the half mirror 7 through the polarizing element 17. The half mirror 7 reflects the light from the illumination optical system 1 to the lower side and guides the objective lens 6. Thereby, the light from the illumination optical system 1 through the objective lens 6 is incident on the wafer W by the epi-illumination. On the other hand, the light that is incident on the wafer w can be reflected back on the wafer W and returned to the objective lens 6, and transmitted through the half mirror 7 to the detection optical system 20. The detecting optical system 20 is sequentially provided from the lower side toward the upper side of FIG. 1 in the order of its arrangement: a removable photodetecting element 21 (polarizing filter), a lens 22, a half turn 23, and a Bühler lens (Bertrand Lens) 24) and a disc member 63 having a function of a field of view of 200920091. The light detecting element 21 is disposed in a state orthogonal to the polarizing element 17 of the illumination optical system 10 (a state in which the polarization directions are orthogonal). Since the polarizing element 17 of the illumination optical system 10 and the photodetecting element 21 of the detecting optical system 2 have a condition of a crossed polarizer (cr〇ssedNic〇i), as long as the pattern of the wafer w does not rotate the polarizing spindle, the detection is performed. The amount of light detected by the optical system 2 会 will be close to zero. The half prism 23 divides the incident beam into two directions. The image of the wafer w is imaged by the Bühler lens 24 to the disk member 63 having the function of the field of view by the beam of one of the half beams, and the pupil plane of the objective lens 6 = image (luminance distribution) The DMD (Digital Micromirror Element) element 31 is projected onto the imaging unit 30. Since the two-dimensional photographic element 33 of the photographic portion 30 is conjugated with the dmd element η, the luminance distribution of the objective lens 6 on the pupil plane can be reproduced on the photographic surface of the two-dimensional photographic element 33. The image of the Fourier-transformed Crystal® W (hereinafter referred to as Fourier image) is photographed. Further, a Bertrand lens 24 generally refers to a converging lens that images an image of a focal plane behind a objective lens on a focal plane of an eyepiece, but an optical system such as a microscope is generally a telecentric state, and after the objective lens Since the focal plane of the portion is a pupil plane, in the present embodiment, a lens for imaging the image of the pupil plane of the objective lens 6 on the imaging surface of the two-dimensional imaging element 33 is referred to as a Bühler lens 24. Further, the other light beam passing through the half turn 23 is guided to the second imaging unit 5 for photographing the normal wafer w image not subjected to Fourier transform. By displaying the image of the wafer w imaged by the second imaging unit 50 on the monitor 44, the operator can visually observe the surface of the wafer w. 200952103 Here, the Fourier image (i.e., the image of the pupil plane of the objective lens 6) is taken during the defect inspection of the present embodiment for the following reasons. If the defect is checked

t Using an image directly taken by the pattern of the wafer w, when the pattern I is below the resolution of the inspection device, the defect of the pattern cannot be optically detected. On the other hand, in the Fourier image, if the pattern of the wafer W has a defect, the symmetry of the reflected light is destroyed. The brightness or color of the portions orthogonal to the optical axis of the Fourier image due to the structural birefringence. Participation changes. Therefore, even if the distance between the patterns is below the resolution of the inspection apparatus, the defect (change) of the pattern can be detected by detecting the above change in the Fourier image. Further, the following is a map? [Fig. 2] illustrates the relationship between the incident angle of the illumination light incident on the wafer W and the position of the point image in the pupil plane. As shown by the dotted line in Figure 2, the imaging position of the Zhaoming this is the center of the pupil: another λ ^ ^ , on the other hand, as shown by the solid line in Figure 2, :==6: (equivalent When the imaging position on the pupil is ❹ = the outer edge, that is, the incident angle of the illumination light incident on the wafer W, the optical axis corresponding to the position mu in the radial direction in the pupil is imaged at the same radius The light entering the crystal 圊 w. (4) First h (four) _ angle shot _ board two °:::::, and:: said circular plate member 63, rotating rotary encoder 62. Motor 61, with a circular plate: The rotation angle of the member 63 is a central axis, and the disk member 63 is rotationally driven in a plane perpendicular to the optical axis of the axis of symmetry. d Optical system 20 9 200952103 The disk member 63' is as shown in FIG. It is formed into a disk shape having 17 pinholes 63ai to 7a. Each of the pinholes 63a1 to 63an is disposed in a diametrical position and a circumferential position which are different from each other on the disk member 63, and the i-th to the 17th pins are respectively arranged. Any one of the holes 63al to 63al7 is located within the optical field of view of the detecting optical system 20. Thereby, the detection range (field of view) on the surface of the wafer w is determined. Further, in the present embodiment, the pinholes 63ai to 63ai7 are arranged at equal intervals in the circumferential direction of the disk member 63, and from the first pinhole 63a1 at one end to the 17th pinhole 63ai7 at the other end. The diametrical position is offset from the center (rotational symmetry axis) of the disc member 63 at equal intervals. Further, the diameter of each of the pinholes 63a1 to 63al7 is 0 mm, and when the objective lens 6 is used at a magnification of 1 ,, the wafer is used. The detection range (field of view) on the surface of w becomes a diameter of 0 〇 / / m. Therefore, the shape of each of the pinholes 63a1 to 63al7 is preferably a circular (or rectangular) optical field D as shown in Fig. 4, indicating the surface of the wafer w The optical field of view of the detection optical system 20 is 025 〇m in the present embodiment. The scanning area c located inside the optical field of view D is a square area for obtaining inspection data, and the i side is about 17 〇"m. When the disc member 63 is rotated by the motor 61, the first pinhole 63al sweeps the data acquisition position A (1, ", A (2, ", ... ( 1) the first to the Uth pinholes 63a1 to 63al7 are respectively shifted by one scanning amount in the scanning area C In the sequential scanning mode, the disk member 63 is offset from the diametrical position. Therefore, after the scanning of the first pinhole 63ai is completed, the second pinhole 63a2 is scanned from the scanning position of the third pinhole 63al. The position of the sweeping position is obtained by the position A (1, 2), A (2, 2), ... a (i7, 200952103 2). Therefore, the plurality of pinholes do not enter the scanning area c at the same time. In the same manner, when scanning the data acquisition position at the position of the offset scan amount, the 17th pinhole 63al7 scans the most data acquisition position A(l, 17), A(2, 17), ... a (17) in the scan area c. , 17).

Since the disk member 63 and the wafer W are in a conjugate position, when the disk member 63 is rotated by the motor 61, the scanning portion C of the wafer w can be sequentially scanned from any of the pinholes, and the image capturing portion 3 is used. Each of the detecting elements 36a, 3 is complemented, and 3 6c detects the data about the scanning area c. In this manner, the scanning is performed using the pinholes 63a1 to 63al7 of the field of view, whereby the wafer stage 5 does not need to be moved over a wide range of about 17 〇 #m in the present embodiment, and can be performed at a high speed. The change in sensitivity was measured in the same manner as the SEM (sweep electron microscope). Further, with the rotation of the disk member 63, the respective pinholes 63a1 to 63ai7 rotate, but since the radius of the scanning region c with respect to the disk member 63 is small, the scanning of the respective pinholes 63a1 to 63al7 is close to a straight line. The configuration of the photographing unit 30 is as shown in FIG. 1 and has a DMD (Digital).

Micromirror Device 31, lens 32, two-dimensional imaging element 33, lens 34 disposed on the opposite side, beam splitter 35, and three detecting elements 36b, 36c. The DMD element 31 has a plurality of movable micromirrors (not shown) arranged side by side on a plane. The micromirror of the DMD element 31 is electrically driven, and is tilted in the ON state so that the light from the detecting optical system 2A is reflected toward the two-dimensional imaging element 33, and is tilted in the 〇FF state so as to be from the detecting optical system 20 The light is reflected toward the direction of the detecting elements 36a, 36b, 36c (the beam splitter 35). Therefore, the light from the detecting optical system 11 200952103 20 reflected by the micromirror in the ON state is guided to the photographing surface of the two-dimensional imaging element 33 by the lens 32 (angle adjustment (Aorizu optical system)). On the other hand, the light from the detecting optical system 20 reflected by the micromirror of the 0FF state is divided into R (red), G (green), and B (blue) by the lens 35 (angle adjusting optical system). After the light of the color, the three detecting elements 36a, 36b, 36c are respectively guided. Further, the photoelectric signals obtained by the respective detecting elements 36a, 36b, 36c are transmitted to the CPU 43 by wires not shown. The two-dimensional imaging element 33 is a CCD or CMOS having a Bayer array of color filter arrays for performing imaging on the aforementioned Fourier image. Further, the three detecting elements 36a, 36b, 36c, the high-sensitivity photodetecting element 'such as a photodiode or a sag element, respectively detect R (red), G (green), B after being split by the splitter 35 (blue) light. The control unit 40 has a recording unit 41 for recording data of a Fourier image, an input interface 42, a CPU 43 for performing various arithmetic processing, a monitor 44, and an operation unit 45 for performing overall control of the inspection device. . Further, the recording unit 41, the input interface 42, the monitor 44, and the operation unit 45 are electrically connected to the CPU 43, respectively. The cpiJ43 analyzes the Fourier image by the execution of the program, and obtains a region having high sensitivity to the pattern change in the Fourier image captured by the two-dimensional imaging element 33. Further, the input interface 42 has a connector for connecting a recording medium (not shown) or a connection terminal for connecting to an external computer (not shown) for reading data from a recording medium or a computer. A method of inspecting the wafer W using the inspection apparatus i having the above configuration will be described below with reference to a flowchart shown in Figs. 8 to 10 . First, a method of creating a pixel corresponding table of the two-dimensional imaging element 33 and the DMD 31 will be described using the flowchart shown in Fig. 8 12 200952103. In the method of creating the pixel correspondence table, first, in step S101, the polarizing element 17 of the illumination optical system 1 and the light detecting element 21 of the detecting optical system 20 are separated from the optical axis. Next, in step S102, the disk member 63 is positioned such that the checkpoint (any of the pinholes 63a1 to 63al7) becomes the center of the field of view. In the next step S103, the unpatterned wafer W is moved to the lower side (observation position) of the objective lens 6 by the wafer stage 5. In the next step S104, the light source of the illumination optical system 1 is turned on. At this time, the illumination light emitted by the light source 11 passes through the condenser lens j 2 and the illuminance uniformizing unit 13, passes through the aperture stop 14 and the field stop 15 , passes through the collimator lens 16 and becomes parallel light, and is guided by the half mirror 7 . After the reflection, the objective lens 6 is irradiated onto the wafer W. Then, the reflected light from the wafer w is incident on the detection optical system 20 through the objective lens 6 and the half mirror 7, and the light incident on the detection optical system 20 passes through the lens 22, the half turn 23, the Bühler lens 24, and the circle. The plate member 63 projects the Fourier image onto the DMD element 31 of the imaging unit 30. In the next step S105', one pixel (micromirror) of the DMD element 31 is turned on, and the other pixels (micromirrors) are turned off. Thus, the light from the detecting optical system 20 reflected by the pixels of the 〇N state is guided to the imaging surface of the two-dimensional imaging element 33 through the lens 32. In the next step S106', the two-dimensional imaging element 33 performs imaging, and the light reflected by the pixel (micromirror) in the ON state is detected, and the CPU 43 calculates the light reflected by the pixel in the ON state on the imaging surface (two-dimensional The pixel position of the photographic element 33). In the next step S107, the CPU 43 compares the pixel position of the two-dimensional camera 13 200952103 image element 33 obtained in step S106 with the pixel position of the (the position of the micromirror) # '70 pixel 31'. The pixel correspondence table of the recording unit 41.隹-人-Step S108, the CPU 43 determines whether or not the B/A has been determined by all the pixels of the DMD element 31. If you want to clip it, the right is π", the completion of the pixel correspondence table is completed, and if it is judged as "*", the process proceeds to step S109. S^Sl〇9 to display the pixel of the DMD element 31 in the N state (micro-brother) ) Change to a pixel that has not completed the measurement and return to the step. The relationship between the two-dimensional imaging element, the pixel of the member 33 and the pixel of the DMD element 31 can be registered in the pixel correspondence table by the above-described step '. Next, a method of determining a region having high sensitivity to pattern change in the Fourier image captured by the two-dimensional imaging device 33 will be described using a flowchart shown in FIG. The method for determining the Q-domain of the high-salt sound 4 first, in step S201, the first 7L member 17 of the illumination optical system 1 and the detection element core of the detection optical system 20 are orthogonal polarized mirrors (10) The mode is inserted onto the optical axis. Next, in the step coffee, the disc member W is placed in such a manner that the checkpoint (any of the pinholes 63U to 7) becomes the center of the field of view. In the next step S203, all the pixels (micromirrors) of the DMD element 31 are brought into the 〇n state, and the light from the wafer w is completely reflected in the direction of the two-dimensional photographic element 33. In the next step S 2 0 4, the light source 11 of the illumination optical system 10 is turned on by gg gp jueg. In the step S205, the wafer w on which the reverse pattern is formed is placed on the wafer stage 5, and a portion of the irradiation pattern on the wafer w is moved by the wafer stage 5 to the lower side of the objective lens 6. At this time, a wafer W in which a plurality of patterns of the same shape having different exposure conditions (dose and focus) are formed is used. 200952103 In this way, the illumination light emitted by the light source 11 passes through the condensing lens i 2 and the illuminance homogenization unit 13 , passes through the aperture stop 14 and the field of view light 5 , passes through the collimator lens 16 and becomes parallel light, and passes through the polarizing element I. It is reflected by the half mirror 7, and then irradiated onto the wafer w through the objective lens 6. Further, the reflected light from the wafer W enters the detection optical system 20 through the objective lens 6 and the half mirror 7, and enters the light of the detection optical system 20, passes through the light detecting element 21, the lens 22, the half mirror 23, and the ray. The lens 24 and the disk member 63 project the Fourier image on the DMD element 31 of the imaging unit 30. At this time, since all the pixels (micromirrors) of the DMD element 31 are in the on state, the light reflected by the dmD element 31 projects the Fourier image on the photographic surface of the two-dimensional photographic element 33 through the lens 32'. The Fourier image is imaged by the two-dimensional imaging element 33 in the next step S206', and the captured Fourier image is recorded on the recording unit 41. At the next step S207, it is determined by the CPU 43 whether or not the measurement has been performed on all necessary patterns on the wafer w. If the determination is YES, the process proceeds to step S208. If the determination is "NO", the process returns to step S205, and the pattern (other irradiation region) that has not been measured is moved to the lower side of the objective lens 6, and the process proceeds to step S206. Photography. Thereby, in the recording unit 41, color information of a plurality of Fourier images having different exposure conditions is recorded for the same-shaped pattern. ^ In step S208, CpU43 generates luminances (average values) of R (red), G (green), and B (blue) at each position of the Fourier image for each Fourier image. For the method of obtaining the luminance data, first, as shown in FIG. n, the Fourier image (for example, the Fourier image Fli of one frame) is divided into square lattices at equal intervals in the longitudinal direction and the horizontal direction of 15 200952103. The plurality of divided regions P' are averaged for each of the divided regions p of the Fourier image, and each of the RGB values are obtained according to the color. This step is then performed for each Fourier image. Thereby, for each of the divided regions P of the Fourier images from the frame to the nth-long Fourier image FI1 to FIn, the brightness of each of the R, G, and B color components is generated. In the next step S209, focusing on the same divided area as shown in (10), the CPU 43 generates grayscale difference data for each color component of R, G, and B, and the grayscale difference data is expressed in the same divided region in Fourier. Image! The gray level difference between the two. Specifically, if any of the divided regions on the Fourier image is Pm, first, for each Fourier image, FIn is extracted, and luminance data of each color component of each of the divided regions p is extracted (in step S208). The obtained one is connected to the English, and the eight ^)# extracts the maximum and minimum values of the R, G, and B color components from the grayscale values of the luminance data corresponding to the sub-long field ^ The difference between the value and the minimum value. In addition, these steps are performed for all the divided regions. ^ Thus, for the Fourier image, the divided regions are formed according to the R, G, and B. The value of the difference between the maximum value and the minimum value of the gray-scale difference between the gray-scale differences between the Fourier images of the Fourier image. In step S21, the CPU 43 obtains the maximum value of the gray scale in the divided region of the Fu image based on the gray-scale difference beaker (the maximum value of the gray scale and the final difference value) obtained in the step s2〇9. The color of the big color and the segmentation area are determined as the most favorable area, and the divided area is determined as the high-sensitivity area. 16 200952103 This is determined as the detection condition. Fig. 13 to Fig. 5 show the distribution state of the gray scale difference of each divided region of the Fourier image in accordance with each color component. In the example of Figs. 13 to 15, the upper left area of the gray level difference of B shown in Fig. 15 is the area of the maximum sensitivity. Thereby, it can be determined that it is preferable to use which of the R, G, and b colors and which of the Fourier images are used in order to detect the change in the line width or the distribution state of the pattern with high sensitivity. As described above, the change of the unknown pattern can be detected by the image captured by the two-dimensional imaging element 33. However, the reflected light from the wafer W is weak, and the exposure time of the two-dimensional photographic element 33 becomes long, and there is a case where the yield cannot be improved. Here, a method of detecting a pattern change at a high speed with high sensitivity will be described using a flowchart shown in Fig. 1A. In the method of detecting the pattern, first, in step S301, the polarizing element 17 of the illumination optical system 1 and the light detecting element 21 of the detecting optical system 20 are inserted on the optical axis. Then, in step S302, the CPU 43 determines the ΟΝ/OFF state in the pixel (micromirror) of the DMD element 31 to direct the reflected light from the wafer w to the direction of each of the detecting elements 36a, 36b, 36c. Specifically, referring to the pixel correspondence table of the two-dimensional imaging element 33 and the DMD element 31 obtained in steps S101 to S109, the pixel of the DMD element 31 is obtained and found in steps S2〇1 to S210. The high-sensitivity pixel area (divided area) on the two-dimensional imaging element 33 corresponds to each other. In the next step S303', the CPU 43 sets the pixel of the DMD element 31 corresponding to the high-sensitivity pixel area (divided area) determined in step S302 to the OFF state ' to guide the directions of the respective detecting elements 36a, 36b, 36c, and The other pixels are set to the ON state to guide the direction of the two-dimensional imaging element 33. 17 200952103 In the next step S304, the light source 丨丨 of the illumination optical system 1 is turned on β. Next, in step S305, the disk member 63 is rotated by the motor 61 at a constant speed. In the next step S306, the wafer w to be inspected is loaded on the wafer stage 5, and the pattern to be inspected (the amount of one irradiation area) on the wafer W is moved to the lower side of the objective lens 6 by the wafer stage 5.

In this way, the illumination light emitted from the light source 11 passes through the condensing lens 丨 2 and the illuminance uniformizing unit 13 , passes through the aperture stop 14 and the field stop 15 , passes through the collimator lens 16 , becomes parallel light, and passes through the polarizing element 17 . After the reflection of the half mirror 7, the wafer w is irradiated through the objective lens 6. Then, the reflected light from the wafer W enters the detection optical system 20 through the objective lens 6 and the half mirror 7, and enters the light of the detection optical system 2, passes through the light detecting element 21, the lens 22, the half prism 23, and the ray. The lens 24 and the disk member 63 reach the DMD element 31 of the imaging unit 30. At this time, the reflected light of the region having high sensitivity to the pattern change of the wafer W is subjected to the inverse of the pixel (micromirror) of the UFF state in the DMD element 31.

The light passing through the lens 34' is guided to the Wth measuring element by the beam splitter 35, the red light is directed to the second detecting element 36b, and the blue light is directed to the third detecting element #H. The rotating member 63 of the scanning unit 60 is rotated, and the scanning area C of the crystal scanning layer W is sequentially scanned from any of the first to the πth pinholes 63a1 to 63al7, and the detecting elements are scanned at high speed. 36a, 36b, 36c detect data about the -%-dimensional scan area C. In the next step S307, the CPU 43 creates a data transfer table and records it in the recording unit 41. The data acquisition table is a table showing the correspondence between the above-mentioned data acquisition position A(l, 1)... and the count value of the rotary encoder 62 (i.e., the rotation angle of the disk member 63) 18 200952103. In the first method, first, the disk member 63 is rotated, and M. al (see Fig. 4) reaches the data acquisition position A (l, and the count value is registered in the data acquisition table. Next, the third = 彳·^ When the 1) rotation code (4) is reached, the data acquisition position l (ti is used to register the first pinhole 63al and the count value is registered to the data acquisition H07, 1). In this way, the registration of the data acquisition table of one scan amount is completed. "Tian Yu confuses the village. Rotation=:6 The second pinhole 63a2 reaches the data acquisition position Α(1,2) and the value of D 2 is registered in the data acquisition form. Next, when the 2nd f 2 arrives at the data acquisition position A (2, 2), the rotation encoder 62 arrives at the data acquisition, and when Α (17, 2), the juice value of the rotary encoder 62 is registered to the data acquisition table. . • Sample mode 'For each of the pinholes 63ai to 63an, the count value of the rotary encoder 62 when the position is obtained by the Becker is registered to the data 1° to complete the data acquisition table. The _ 5 shows an example of a data acquisition table when the encoder of the 编码 〇〇〇〇 〇〇〇〇 丄 丄 62 is used as the rotary code hacker 62. In addition, the interval at which the data is acquired can also be changed by the resolution. : Sub-step measurement, _ using the data created in step (10) to determine whether or not to take all the data acquisition locations by the respective detection elements =, 36c to obtain information about the scanning area. When the data acquisition is not completed, the process proceeds to step S3G9, and when the data acquisition is completed, the process proceeds to step s(1). 19 200952103 In addition, 'the order of obtaining the data, according to the data acquisition location ι(ι,1)~A(i 7, 0, A(l' 2)~A(17, 2),...A(l,17)~A The order of (17, 17) may be performed randomly, or may be performed at random. In step S309, the CPU 43 determines whether the pinhole has reached the unfinished data by the rotation of the disk member 63 by the data acquisition table created in step S307. If the data has not been reached, the process returns to step S3 to 9. If it has arrived, the process proceeds to step S3 10. In step S310, the CPU 43 obtains the position by any of the detecting elements 36a, 36b, 36c. The reflected light is detected from the high sensitivity guided by the DMD element 31, and the brightness (gray scale) of the reflected light is measured from the detection signal. At this time, as described above, the light detecting diodes are used for the respective detecting elements 36a, 36b, 36c or The smashing element, etc., can convert the weak signal corresponding to the reflected light from the wafer W into an electrical signal (detection signal) at the same speed (for example, the CCD is about 100 ms, and the sag component is sampled accordingly) The frequency is several kHz, i sampling time is a few ms)) 'High-speed detection of the wafer w The state of the case (change). In addition, in the example of Fig. 13 to Fig. 15, the blue light detected by the third detecting element 36c is used. Further, the pixel (micromirror) of the DMD element 31 has a higher positional accuracy even in the 〇N state. High, but still set to guide each of the detecting elements 36a, 36b, 3 6c in the 〇FF state, and by making the lens 34 a reduced lens, it can be controlled within the allowable range even if the direction of reflection in the FF state is deviated. Further, the brightness data of the detected reflected light is recorded in the recording unit 41, and the data acquisition table of the 5 recorded portion 4 1 is registered in the data acquisition position at which the detection is performed. Then, after the processing of step S310 is completed, Returning to step S3 08. As described above, when all the data acquisitions have been completed in step S3, the process proceeds to 20 200952103 to step S311. Next, in step S3U, CpU43 creates the brightness data and data from the measurement table in step S31. The two-dimensional luminance distribution (gray scale distribution) shown in Fig. 16. Fig. 16 is an example of the result of the measurement, and a luminance material is described in one grid (in the case of the present embodiment, the wafer W table) The positional relationship of each of the squares of the figure of Fig. 16 and the position A of the data of Fig. 4 (l, ^~8(1), ~A(i7,2), ···(7)~a(1) 17) Correspondence. The 'and'-dimensionality distribution is displayed on the monitor 44, and the operator can visually confirm the measurement result (inspection result), thereby detecting the change of the pattern on the wafer W (that is, the drawing ^^ *, _ 缺陷 tea defects). And 'by changing the two-dimensional brightness distribution to a two-dimensional chart not shown in Figure 17, the pattern change can be easily identified. This can also automatically determine the pros and cons of the pattern from the measured brightness data. Regarding the automatic 敎 method of the pattern, cpU43 first calculates the average value of the brightness (gray scale) from the measurement result, and compares the calculated average value with the = setting: good product range. Then, if the calculated average value is a good product, it is judged to be a good product, and if it is outside the good product range, it is judged to be bad. Thereby, = the merits of the pattern. In addition, the range of the data to be averaged by the change is determined by the effect of the size of the field of view on the K w: 2, ☆ outside the average value of the brightness, by calculating the maximum or minimum Local defects can also be found by comparing the range of good products. Furthermore, the standard deviation can be used to check the deviation of the pattern. Further, although the measured luminance data is directly used in SE=, the line width of the pattern is calculated from the luminance data measured in 21 200952103 by correlating the result with the pattern shape of the (scanning electron microscope), and the calculated line width is used. The line width of the pattern may be determined in the same manner (refer to the formula (1) described later). As described above, the inspection apparatus according to the present embodiment includes the scanning unit 60 that scans the range of the wafer W corresponding to the pupil plane of the luminance detected by each of the detecting elements 36a, 36b, and 36c (that is, In the scanning area C)', the wafer stage 5 is not moved in the two-dimensional inspection area, and the pattern formed on the surface of the wafer W can be inspected at high speed with the same high sensitivity as that of the SEM (scanning electron microscope). Further, the scanner unit 60 moves the pinhole (hole portion) of the disk member 63 formed in the field of view by the motor 61 to scan the scanning region c. Thus, the inspection can be performed at a high speed with a simple structure composed of the motor 61 and the disk member 63. Further, at this time, the plurality of pinholes 63a1 to 63ai7 formed in the disc member 63 are respectively rotated by the motor 61, and the scanning region is sequentially scanned from any of the plurality of pinholes 63 & 1 to 63al7. c. Thus, the scanning of the scanning region c is performed by the rotational movement of the disk member 63, so that the size of the scanning portion 60 can be reduced. Further, as described above, each of the detecting elements 36a, 36b, and 36c detects the change in the surface state of the wafer W (the change in the pattern) and the brightness of the portion in which the brightness greatly changes in the pupil plane (Fourier image). Perform a high-sensitivity inspection. Further, the one-dimensional imaging element 33 and each of the detection elements 36a, 36b, and 36c detect a polarization component that is substantially orthogonal to the polarization direction of the illumination light belonging to the linearly polarized light from the wafer W, thereby forming a so-called orthogonality. Polarized mirror 22 200952103 state, and (4) structural birefringence for high sensitivity inspection. Further, the polarization direction of the polarizing element 17 and the light detecting element 21 is not limited to 9 inches. The state of the (orthogonal polarizer) can also be adjusted in a slight adjustment by the rotation of the circularly polarized light caused by the structural birefringence generated by the inspection target pattern. Further, at this time, the surface of the wafer w is illuminated by epi-illumination, whereby the size of the device can be reduced. Further, in the above embodiment, the size of each of the pinholes 63a1 to 63ai7 is 0 mm, but the size is not limited thereto, and the size may be 0 to 5 mm in diameter. At this time, as shown in Fig. 6, 34 pin holes 63A to 63b34 are formed in the disk member 63. Further, in the same manner as in the above-described embodiment, each of the pinholes 63b1 to 63b34 is disposed in the disk member 63' in the diametrical position and the circumferential direction, and is in the first to 34th pinholes 63b1 to 63b34. Any one of the pinholes is located within the optical field of view of the detection optical system 2〇. Thus, when the objective lens 6 having a magnification of 100 times is used, the detection range (field of view) of the surface of the wafer w becomes a diameter of 05 // m, and the resolution can be changed. Further, at this time, after the disk member 63 is rotated by the motor 61, as shown in Fig. 7, the first to 34th pinholes 63b1 to 63b34' respectively acquire the position B (l, 1) in the scanning area c. Scanning is performed in the order of ~B(34, 1), B(1,2)~B(34, 2), ...B(l, 34)~B(34, 34). Further, in the above embodiment, the configuration of the scanning unit 60 includes the motor 61 and the disk member 63, but the invention is not limited thereto. Next, a modification of the scanning unit will be described with reference to Figs. 18 and 19 . In the configuration of the scanning unit 1 60 of the modified example, as shown in FIG. 18, a view plate 163 having a rectangular plate shape, a linear drive unit 16 that linearly moves the view stop plate 163, and a detection field stop plate 163 are provided. 23 200952103 Position linear encoder 1 62. The linear drive unit 161 is composed of a fixed portion 161a and a movable portion 161b. The fixing portion 161a is fixed to the device (detection optical system 20), and drives the movable portion 161b on a straight line. A field stop plate 1 63 is attached to the movable portion 161b. The field stop plate 163 is driven by the linear drive unit 161 in the vertical direction of the optical axis of the detection optical system 20. The field stop plate 163, as shown in Fig. 19, is formed in a rectangular plate shape having 17 pinholes 163 & 1 to 163 & Each of the pinholes 163 & 1 to 163 & 17 is arranged obliquely in the direction of linear movement of the field-of-view aperture plate 163 on the field-of-view aperture plate 163, and any one of the first to seventeenth pinholes 163a1 to 163al7 The holes are located within the optical field of view of the inspection optical system 20. Thereby, the detection range (field of view) on the surface of the wafer W is determined. Further, each of the pinholes 163a1 to I63al7 has a diameter of 0 lmm, and when the objective lens 6 having a magnification of 100 times is used, the detection range (field of view) of the surface of the wafer W becomes a diameter of 0 〇//m. After the linear driving unit 161 linearly moves the field stop plate 163, as in the case of the above-described embodiment, the pinhole hole 63a scans the uppermost data acquisition position 扫描 (1, υ, Α (2) in the scanning area c. , υ, ... A (i7, 1) (refer to Fig. 4). The first to seventeenth pinholes 163a1 to 163al7 are sequentially scanned in the scanning area C offset i reading (four) position, respectively, offset When the scanning of the first pinhole 163a1 is completed, the second pinhole 163a2 scans the position A (1, 2) from the scanning position of the first pinhole 163a1, A(2, 2), scan offset 1 scan position

Set the data of the position of the offset by one time...A(17, 2). In the following section, when the position is obtained in the same manner, the 17th pinhole 24 200952103 163al7 scans the scanning area c for the most recent data acquisition position, 17), A(2, 17), ..... a (17, 17 ). Further, the 'viewing aperture plate 163 is disposed at a position shared with the wafer W in the same manner as the disk member 63. Therefore, when the field stop plate 163 is linearly moved by the linear drive unit ι 61, it can be taken from any pinhole. The two-dimensional data on the scanning area C is detected by the respective detecting elements 36a, 36b, 36c of the photographing unit 30 while sequentially scanning the scanning area c' of the wafer w. In this way, as in the case of the above-described embodiment, the wafer mount 5 can be moved at a high speed and the same sensitivity as that of the SEM (scanning electron microscope) without moving the wafer stage 5 to a wide range of about 17 Å. Pattern changes. Further, the position of the field stop plate 163 (movable portion l61b) is detected by the linear encoder 162. If the count value of the linear encoder 162 is used, the above-described data acquisition table can be similarly created. Further, in the above-described embodiment, the pixels of the DMD element 31 corresponding to the sensitivity pixel region (divided region) obtained in the step S302 are set to the OFF state, and the directions of the respective detecting elements 36a, 36b, and 36c are guided. Further, the other pixels are turned to the ON state to be directed to the direction of the two-dimensional imaging element 33, but the invention is not limited thereto. For example, as shown in FIG. 20, the half turn 38 may be disposed between the DMD element 31 and the lens 32 such that a portion of the light directed from the DMD element 31 toward the one-dimensional imaging element 33 passes through the lens 34 by the half prism 38. And the dichroic prism 35 is guided to the respective detecting elements 36a, 36b, 36c. In this case, the CPU 43 in step S303 sets the pixel of the DMD element 31 corresponding to the high-sensitivity pixel area (divided area) obtained in step S302 to the ON state, and guides the two-dimensional imaging element 33 and each detection element. The directions of 36a, 36b, 36c, and the other pixels are set to the 0FF state without being directed to the directions of the respective detecting elements 25 200952103 pieces 36a, 36b, 36c. Thereby, the pixels of the DMD element 31 can be placed in an ON state with a high positional accuracy, and the light from the wafer w can be directed to the direction of each of the detecting elements 36a, 36b, 36c. Further, in the above-described embodiment, the inspection apparatus 1 for performing defect inspection of the wafer W is exemplified, but the substrate to be inspected is not limited to the wafer W' such as a liquid crystal glass substrate. Further, in the above-described embodiment, the region having high sensitivity to the pattern change is determined based on the gray scale difference data (difference value between the maximum value and the minimum value of the gray scale), but the present invention is not limited thereto. Here, a modification of the method for determining the sensitivity region will be described using the flowchart of Fig. 21. In this method, as in the case of the above-described embodiment, a wafer w in which a plurality of patterns of the same shape having different exposure conditions (dose and focus) are formed is used, and the line width information of the Fourier image and each pattern of each pattern is used. To determine the area with high sensitivity to pattern changes. Further, the line width data corresponding to the above-described pattern is measured by, for example, a line width measuring device such as a scatterometer or a scanning electron microscope (SEM), and the (4) group of line widths is input in advance through an input interface. The input of 42 is recorded in the recording unit 4, and as the scanning unit, the scanning unit 6A having the above-described disc member 63 is used. First, in the same manner as in the above embodiment, in step S25i, the value of the illumination optical system 1 is erected; >, and the detection of the eccentricity 17 and the detection optical system 2 The jaw 21 is inserted into the optical axis. Then, in step s252, the disc member 63 is positioned such that the checkpoint (the pinholes 63a1 to 63al7) becomes the center of the field of view. In the next step S253, all the pixels (micromirrors) of the DMD element 31 are brought into the ON state, and the light from the wafer W is completely reflected toward the direction of the 26200952103 of the two-dimensional imaging element 33. At the next step S254, the light of the illumination optical system 10 is turned on. The K member, in the next step S255, will form a pattern of 7::=11 and focus from a plurality of different patterns of the same shape, and the pattern to be determined on the top (:2::-stage 5) One part of the ',, and shot area is moved to the lower side of the objective lens 6 by the round stage 5. The next step S256, the Fourier image is photographed by the Φ 二维 two-dimensional imaging element 33, and the measurement is performed. The obtained Fourier image is recorded in the recording unit 41. In the next step S257, the CPU 43 determines whether or not the pattern on the wafer w has been measured. If the determination is "Yes", then the image is made. In step S258, if the determination is "NO", the process returns to step S255, and the pattern (other irradiation area) that has not been measured is moved to the image of the objective lens 6 to perform the shooting in step S256. In step S258, the CPU 43 and the above In the same manner as in the embodiment, for each Fourier image, luminance data of R (red) 'G (green) and B (blue) are generated in the divided regions of the respective Fourier images (average, then, sub-step S259' Focus on the same split area CPU" according to R For each color component of G and B, an approximate expression that can represent the rate of change of the grayscale value and the pattern line width in the same divided region of each Fourier image FL to FIn is obtained. Specifically, if the Fourier image is arbitrarily selected First, the divided area is set to Pm'. First, the data of the pattern line width corresponding to each of the Fourier images FI and FIn is read from the recording unit 41. Also, at this time, for each of the Fourier images FI1 to FIn, The luminance data of each color component in the divided region pm is extracted (the one obtained in step S258). Next, for each of the Fourier images FI! to FIn of 27 200952103, the pattern line width and the grayness of the luminance data in the divided region pm are obtained. The correspondence between the order values.

Then, based on the correspondence relationship between the pattern line width and the gray scale value in the divided region Pm, an approximate expression indicating the rate of change of the gray scale value and the pattern line width in the divided region pm is obtained by the least square method. Here, 'the line width corresponding to the pattern of each Fourier image FI! to FIn is set to y, and the gray scale value of B (4 R or G) of the divided area Pm is set to x 'the slope is set to a, y When the distance is set to b, the approximate expression can be expressed by the following formula (1).

y = ax + b (1) Furthermore, the absolute value of the coefficient a corresponds to the inverse of the gray-scale change of the change in the line width of the corresponding pattern (that is, the inverse of the sensitivity of the detection of the pattern change) ). In other words, the smaller the absolute value of the above coefficient a, the more the difference in line width is, the more the gray scale change of the Fourier image is increased, and therefore the detection sensitivity for the pattern change will be higher. Further, for all the divided regions, these steps are performed for each of the color components of R, 'G, and B.

Then, in step S260, the CPU 43 obtains the correlation error between the approximate expression obtained in the step (10) and the line width of the pattern, based on the respective color components of R, G, and B on the Fourier image. Specifically, the data of each of the color components R of R, G, and B is calculated by calculating the line width of each of the Fourier images ρ to FIn and the deviation of the line width calculated by using the approximate expression. For the data of the deviation, for each of the divided regions: the color component is calculated as the standard deviation', and the value is used as the correlation error. Then, in step S261, the CPU 43 obtains, based on the correlation error obtained in the step a s259 coefficient a and the step S260, that the divided region in which the absolute value of the coefficient a is low and the correlation error is extremely small in the divided region of Fourier shadow; 28 200952103 This cutting area is determined as a region of high sensitivity, which is determined as a detection condition. Specifically, for example, each divided region is scored according to the low absolute value of the coefficient a and the correlation error is small, and the divided region having high sensitivity is determined based on the result of the scoring. In this way, it is also possible to determine which one of r, G, and B is to be used, and which of the Fourier images is used, in order to detect the change in the line width or the distribution state with high sensitivity. Further, in the above-described embodiment, the DMD element 3 is used to switch the reflection angle between the ON state and the 〇FF state. However, the present invention is not limited thereto, and an optical spatial adjustment such as an SLM (SPaCe Light Modulator) is used. For the transformer, the sampled light for R (red), G (green), and B (blue) can be taken out separately. Further, by scanning the fixed micromirror to obtain an optimum detection position, sampling with the micromirror disposed at the optimum detection position can also obtain information equivalent to the case of using the DMD element 31 and the like. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic view of an inspection apparatus of the present invention. Fig. 2 is an explanatory view showing the relationship between the incident angle of the illumination light entering the wafer and the imaging position in the pupil. Figure 3 is a top plan view of the disc member. Fig. 4 is a view showing a scanning area and a data acquisition position. Fig. 5 is a view showing a data acquisition table. Fig. 6 is a plan view showing a modification of the disk member. 29 200952103 Fig. 7 is a view showing a modification of the scanning area and the data acquisition position. Figure § is a flow chart showing a method of creating a pixel correspondence table between a two-dimensional imaging element and a DMD element. Fig. 9 is a flow chart showing a method of determining an area having high sensitivity to pattern change. Fig. 1 is a flow chart showing a method of detecting a change in a pattern at a high speed with high sensitivity. Fig. 11 is a view showing an example of a state in which a Fourier image is divided into regions. ❹ The figure shows a schematic diagram of the removal status of the brightness data. Fig. 13 is a diagram showing the distribution state of the gray level difference of R in the Fourier image. Fig. 14 is a diagram showing the distribution state of the gray level difference of G in the Fourier image. 3 Fig. 15 shows the gray scale of B in the Fourier image. Diagram of the distribution state of the difference '3 Fig. 16 is a diagram showing a display example of the inspection result. Fig. 17 is a view showing another display example of the inspection result. Fig. 1 is a schematic view showing a modification of the inspection apparatus. Figure 19 is a plan view of a field stop member. Fig. 20 is a schematic view showing a second modification of the inspection apparatus. Fig. 21 is a flowchart showing a modification of the method for determining the high sensitivity region. 0 30 200952103 Main component symbol description W Wafer (substrate to be inspected) 1 Inspection device 10 Illumination optical system (illumination portion) 17 Polarization element 20 Detection optical system ( Observation optical system 21 Photodetection element 30 Photodetection unit 36a First detection element (detection unit) 36b Second detection element (detection unit) 36c Third detection element (detection unit) 40 Control unit 43 CPU (inspection unit) 60 Scanning unit 61 Motor (drive unit) 63 Disc member (plate member) 63 Disc member (Modification) 63al to 63al7 Pinhole (hole) 63bl to 63b34 Pinhole (Modification) 160 Scanning section (Modification) 161 Linear Drive unit (161a fixing portion, 161b 163 field diaphragm member (plate member) 31

Claims (1)

200952103 VII. Patent application scope: i An inspection device comprising: an illumination unit that illuminates illumination light onto a surface of the substrate to be inspected; an observation optical system for magnifying observation of a surface of the substrate to be inspected illuminated by the illumination light; Detecting brightness of the reflected light from the detected substrate in the viewing optical system on the pupil plane; The inspection unit checks the surface of the substrate to be inspected based on the brightness detected by the detecting unit, and the scanning unit scans a range of the substrate to be inspected corresponding to the pupil surface on which the brightness is detected. 2. The inspection apparatus according to claim i, wherein the scanning unit includes: a plate-shaped member having an optical opening as a field stop, and a driving unit that drives the optical opening; The Ministry of the Ministry of the People’s Republic of China, in the form of an inspection device, Further, the plurality of hole portions which are formed in a plate shape and serve as the optical opening portion are driven in a straight line direction and a circumferential direction s which are different from each other on the plate member. The lanthanoid system is configured to rotationally drive the plate-shaped member by the axis of the plate-shaped member; (4) the driving portion is used to rotate the plate-shaped member by using the driving portion as a center, and the plurality of holes are partially U-shaped from The range is known in the middle of the hole. a hole portion is sequentially scanned 32 200952103, wherein the driving portion 4. The inspection device of claim 2 of the patent scope moves the plate member linearly; the hole portions are respectively formed in the plate-like configuration; The plurality of optical opening portions are obliquely aligned with respect to the linear movement direction. The driving portion is used to linearly move the plurality of holes formed in the plate-like configuration and the plurality of holes from the plurality of holes. Any of the holes of τ sequentially scans the range. ❹ 5. The inspection apparatus according to the scope of application of the patent item ii, wherein the detection unit detects the brightness of the portion where the brightness of the substrate is largely changed in response to a change in the surface state of the substrate to be detected. 6. The inspection apparatus according to claim 1, wherein the illumination light is irradiated onto a linearly polarized light having a surface of the detected substrate having a reverse pattern; and the detecting portion detects a polarization direction of light from the detected substrate A polarized component that is substantially orthogonal to the linearly polarized light. 7. The inspection apparatus of claim 6, wherein the illumination unit illuminates the illumination light onto the surface of the substrate to be inspected by epi-illumination. Eight, the pattern: (such as the next page) 33