JP2009150910A - Defect inspection device, defect inspection method, optical scanning device, and production method of semiconductor device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a defect inspection device, capable of implementing defect inspection with correct, high accuracy, and less loss in the signal intensity. <P>SOLUTION: The defect inspection device of the present invention is equipped with a laser light source 1, emitting optical beam; a one-dimensional diffraction grating converting light into one-dimensional multi-beam; an objective lens 9 for converging multi-beams; a half-mirror 6a separating the optical beam reflected from a sample from the optical beam entering into the sample; and a wedge 17 bifurcating the separated optical beams. Then, a sample stand 11 is subjected to raster scan to inspect whole areas of the sample. The defect inspecting device is also equipped with a photodetector 19a for detecting the optical beam of one side bifurcated by the wedge and a photodetector 19b for detecting the optical beam of another side. Concavity and convexity-like defects are inspected by addition signal based on the signals from the photodetectors 19a and 19b concerned. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a defect inspection apparatus that detects and inspects a defect in a sample, a sample inspection method, an optical scanning apparatus, and a semiconductor manufacturing method using the defect inspection apparatus or the optical scanning apparatus. The present invention relates to a defect inspection apparatus, a sample inspection method, an optical scanning apparatus, and a semiconductor manufacturing method using the defect inspection apparatus or optical scanning apparatus, which are suitable for defect inspection of mask blanks.

  In order to improve the yield of a semiconductor device, a defect inspection apparatus that detects and inspects a surface defect of a semiconductor device, a semiconductor wafer, a mask blank, or the like is used. This defect inspection apparatus is required to perform accurate and high-precision inspection with higher spatial resolution in accordance with recent miniaturization of LSI.

  For the inspection of the surface of the semiconductor wafer or the like, a laser scattered light type defect inspection apparatus is often used. This laser scattered light type defect inspection apparatus utilizes the fact that the intensity and angle of scattered light differ depending on the surface state of the sample. That is, in this method, the defect inspection is performed by irradiating the sample surface with laser light and detecting the intensity of the scattered light. That is, when a laser beam is irradiated to a portion having a defect, the scattered light becomes stronger than a portion having no defect. Therefore, the presence / absence of a defect is determined from the intensity of the detected scattered light. Then, one or both of the sample and the laser beam are scanned to inspect the entire surface of the sample.

  On the other hand, a bright-field defect inspection apparatus is known that irradiates a sample with a laser beam and detects specular reflection light from the sample surface (for example, Patent Document 1 and Patent Document 2). . The applicant of the present application has developed and disclosed a multi-beam type defect inspection apparatus (for example, Patent Document 3). In this multi-beam type defect inspection apparatus, a diffraction grating is used to obtain a one-dimensional line-shaped beam from one laser beam, which is irradiated onto the sample, and the reflected light on the sample surface is detected. is there. In this method, a higher resolution can be obtained by using a confocal optical system (confocal optical system). The applicant of the present application has further developed and disclosed a defect detection apparatus that generates an array of multi-beams using a two-dimensional diffraction grating and irradiates a sample (for example, Patent Document 4). In this apparatus, a sample is irradiated with a multi-beam in a two-dimensional array, and the sample is scanned in a spiral shape. Therefore, this apparatus can perform high-resolution inspection in a short time.

  Regarding these defect inspection apparatuses, it is disclosed to inspect whether a defect is a convex defect or a concave defect in addition to the presence or absence of a defect. Alternatively, it is disclosed to check the height of a defect. In order to examine detailed information on these defects, conventionally, a light shielding plate (light shielding plate, knife edge) is inserted on the optical path to shield light half. And the change of the light quantity was detected using one or two or more detectors, and the defect information was examined.

  However, these defect detection apparatuses have the following problems. In the conventional defect inspection apparatus, since the detection light is half shielded, the received light signal intensity is lost. Therefore, there is a problem in that it is easily affected by shot noise and thermal noise of the received light signal, and pseudo defects are generated, and it is difficult to increase the detection sensitivity for the defects. As miniaturization of semiconductor devices progresses, a device capable of high-sensitivity and high-speed inspection is required. For example, in EUVL (extreme ultraviolet light) mask blanks, it is necessary to detect a defect having a height of several nm and a size of several tens of nm. Therefore, this loss of signal intensity has been a problem in performing highly sensitive and accurate inspection.

Japanese Patent Laid-Open No. 2002-22415 JP 2001-74423 A JP 2001-027611 A Japanese Patent No. 3210654

  The present invention has been made in view of the above problems, and has a defect inspection apparatus, a defect inspection method, an optical scanning apparatus, and a defect thereof that can perform measurement accurately and accurately with little loss of signal intensity. It is an object of the present invention to provide a semiconductor manufacturing method using an inspection device or an optical scanning device.

  A defect detection apparatus according to the present invention includes a light source that generates a light beam, a sample stage on which a sample is placed, a sample placed on the sample stage, and a light beam that is emitted from the light source to the sample. Scanning means for relatively moving, light branching means for branching the light beam reflected by the sample into a plurality of different light beams, and first light for detecting the first light beam branched by the light branching means A detection means; a second light detection means for detecting the second light beam branched by the light branching means; a signal output from the first light detection means; and a signal output from the second light detection means. A circuit for obtaining a difference signal based on the difference between the signals, a first comparison circuit for comparing the difference signal with a predetermined first value, the difference signal and a predetermined second value, A second comparison circuit for comparing the first ratio and the first ratio Based on the comparison result of the circuit and a second comparator circuit, and performs defect inspection of the sample. Thereby, defect inspection can be performed with high accuracy.

  The defect inspection apparatus may further include a diffraction grating that converts light from the light source into a one-dimensional multi-beam, and the first light detection unit includes a plurality of light receiving elements arranged in a line. Each of the plurality of light receiving elements of the second light detecting means is paired with each of the light receiving elements of the first light detecting means. Thus, a difference signal may be obtained for each of the signals detected by the paired light receiving elements. As a result, high-speed and high-precision inspection can be performed.

  The light branching means may be provided at a position other than the position of the pupil. Thereby, the freedom degree of arrangement | positioning of an optical system expands.

  In the above defect inspection apparatus, a two-dimensional diffraction grating that converts a light beam from the light source into a two-dimensional multi-beam, a focusing unit that focuses the multi-beam reflected by the sample, and a pupil position of the focusing unit And optical branching means for branching the multibeam into two multibeams, wherein the first light detection means is a light beam of each of the multibeams branched by the light branching means. The second light detecting means is a plurality of light receiving elements for detecting the light beams of the other multi-beams branched by the light branching means. The circuit includes a plurality of light receiving elements paired with each of the plurality of light receiving elements included in the first light detection means, and the circuit for obtaining the difference signal includes the light receiving element of the first light detection means. Each difference signal is obtained based on the signal output from each pair of light receiving elements of the second photodetecting means, and the defect inspection apparatus compares the result of comparison between the first comparison circuit and the second comparison circuit. It may be determined whether the defect is a convex defect or a convex defect. Thereby, defect inspection can be performed at high speed and with high accuracy.

  The light branching means is disposed at a position occupying approximately half of the light path of the light beam reflected from the part free from defects of the sample, and the light beam is changed by changing the traveling direction of the light beam incident on the light branching means. It is preferable to be branched. For the light branching means, a wedge disposed on the optical path of the light beam may be used.

  The defect inspection apparatus includes a circuit for obtaining a sum signal based on a signal output from the first light detection means and a signal output from the second light detection means, and the sum signal is predetermined. It is preferable that a third comparison circuit for comparing with the third value is further provided, and a defect is detected based on a comparison result of the third comparison circuit. As a result, a spot-like defect can be detected with high accuracy.

  A defect detection method according to the present invention is an inspection method for inspecting a defect of a sample by irradiating the sample with a light beam and detecting the reflected light, the step of irradiating the sample with the light beam, and the light Moving the position of the beam relative to the sample, branching the light beam reflected by the sample into a first light beam and a second light beam, and the light intensity of the first light beam Detecting a light intensity of the second light beam; and a difference signal between a signal based on the light intensity of the first light beam and a signal based on the light intensity of the second light beam. A step of obtaining, and inspecting the convex defect and the concave defect based on the difference signal. By having this configuration, defect inspection can be performed with high accuracy.

  Preferably, the method further comprises comparing the difference signal with a predetermined first value and comparing the difference signal with a predetermined second value, wherein the difference signal is the first value. A convex defect and a concave defect are discriminated based on the order of timing exceeding the value and timing exceeding the second value.

  Further, in the step of branching the reflected light beam into the first light beam and the second light beam, it is desirable to branch the light beam reflected from a portion having no defect of the sample into approximately half. Thereby, an inspection can be performed with high accuracy.

  In the above inspection method, a step of obtaining a sum signal of a signal based on the light intensity of the first light beam and a signal based on the light intensity of the second light beam; And a step of comparing with a value, and a spot-like defect can be detected based on the comparison result. Thereby, a spot-like defect can be detected with high accuracy.

  An optical scanning device according to the present invention includes a light source, a means for scanning a light beam from the light source on an object, and a first light beam and a second light beam from the light beam reflected by the object. Means for branching, the means for branching so that the relative intensities of the first light beam and the second light beam change based on the state of the object, and the branched first light beam. Means for detecting, means for detecting the branched second light beam, means for comparing the intensity of the detected first light beam with the intensity of the second light beam, and means for comparing And means for detecting the state of the object based on the comparison result of By having this configuration, highly accurate measurement can be performed.

  The branching unit may include a wedge optical element provided on an optical path of light reflected by a flat portion of the object, and an end of the branching unit is in the center of the optical path. Thereby, arrangement | positioning of an optical system can be simplified.

  The comparing means generates a difference signal based on a difference between the intensity of the light detected by the means for detecting the first light beam and the intensity of the light detected by the means for detecting the second light beam. The means for outputting and detecting the state of the object can determine the state of the object based on a change in the difference signal accompanying the light beam scanning.

  It is preferable that the branching unit branches the reflected light into the first light beam and the second light.

  A method of manufacturing a semiconductor device according to the present invention includes a step of irradiating an original plate with a light beam, a step of relatively moving positions of the light beam and the original plate, and a light beam reflected by the original plate in a first manner. Branching into a light beam and a second light beam; detecting a light intensity of the first light beam; detecting a light intensity of the second light beam; and the first light beam. Obtaining a difference signal between a signal based on the light intensity of the light and a signal based on the light intensity of the second light beam, inspecting a convex defect and a concave defect based on the difference signal, and the inspection The method includes a step of setting the original plate on an exposure apparatus, a step of exposing the wafer with the exposure pattern of the original plate, and a step of developing the exposed wafer. By having this structure, it can contribute to the improvement of the yield of a semiconductor device.

  According to another aspect of the present invention, there is provided a method of manufacturing a semiconductor device, the step of scanning the light from the light source on the original plate, and the step of branching the first light beam and the second light beam from the light reflected by the original plate. And branching so that the relative intensities of the first light beam and the second light beam change based on the state of the original plate, and detecting the branched first light beam; In the comparison result of the step of detecting the branched second light beam, the step of comparing the detected intensity of the first light beam and the intensity of the second light beam, and the comparing step. A step of inspecting a state of the original plate, a step of setting the inspected original plate in an exposure apparatus, a step of exposing a wafer with an exposure pattern of the original plate, and an exposed wafer. Comprising the steps of, and has a. By having this structure, it can contribute to the improvement of the yield of a semiconductor device.

  According to another aspect of the invention, there is provided a method for manufacturing a semiconductor device, comprising: irradiating a wafer with a light beam; relatively moving a position of the light beam with respect to the wafer; and a light beam reflected by the wafer. Branching into a first light beam and a second light beam, detecting a light intensity of the first light beam, detecting a light intensity of the second light beam, and the first Obtaining a difference signal between a signal based on the light intensity of the light beam and a signal based on the light intensity of the second light beam, inspecting a convex defect and a concave defect based on the difference signal, and the inspection Forming a resist layer on the exposed wafer, exposing the wafer on which the resist layer is formed according to a mask pattern, developing the exposed wafer, Those having. By having this structure, it can contribute to the improvement of the yield of a semiconductor device.

  According to another aspect of the present invention, there is provided a method of manufacturing a semiconductor device, the step of scanning the light from the light source on the wafer, and the step of branching the first light beam and the second light beam from the light reflected by the wafer. And branching so that the relative intensities of the first light beam and the second light beam change based on the state of the wafer, and detecting the branched first light beam; In the comparison result of the step of detecting the branched second light beam, the step of comparing the detected intensity of the first light beam and the intensity of the second light beam, and the comparing step. A step of inspecting the state of the wafer, a step of forming a resist layer on the inspected wafer, and a step of exposing the wafer on which the resist layer has been formed according to a mask pattern. And up, and has the steps of: developing the exposed wafer. By having this structure, it can contribute to the improvement of the yield of a semiconductor device.

  According to the present invention, it is possible to provide an optical scanning device that can perform accurate and highly accurate measurement with little loss of signal intensity.

It is a block diagram which shows the structure of the defect inspection apparatus concerning Embodiment 1 of this invention. It is a figure which shows the laser beam in the wedge in this invention. 2A is a cross-sectional view showing laser light at the wedge. FIG. 2B is a top view showing laser light around the wedge. It is a figure which shows the laser beam with a wedge at the time of the defect detection in this invention. FIG. 3A is a cross-sectional view showing laser light at the wedge. FIG. 3B is a top view showing laser light around the wedge. It is a figure which shows the laser beam with a wedge at the time of the defect detection in this invention. 4A is a cross-sectional view showing laser light at the wedge. FIG. 4B is a top view showing laser light around the wedge. It is a figure which shows the signal of the photodetector at the time of detecting a concave defect. It is a figure which shows the signal of the photodetector at the time of a spot-like defect detection. It is a circuit diagram which shows the signal processing circuit concerning this invention. It is a block diagram which shows the structure of the defect inspection apparatus concerning Embodiment 2 of this invention. It is a block diagram which shows the structure of the defect inspection apparatus concerning Embodiment 3 of this invention. It is a block diagram which shows the structure of the defect inspection apparatus concerning Embodiment 3 of this invention. It is a block diagram which shows the structure of the defect inspection apparatus concerning Embodiment 4 of this invention. It is sectional drawing which shows the spot of the two-dimensional multi-beam on a photodetector. It is a figure which shows the laser beam of a one-dimensional multibeam. FIG. 13A is a cross-sectional view showing a one-dimensional multi-beam spot on the wedge. FIG. 13A is a top view showing a one-dimensional multi-beam around the wedge.

Embodiment 1 of the Invention
A defect inspection apparatus according to a first embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a configuration diagram showing the overall configuration of the defect inspection apparatus. 2 to 4 are diagrams showing detection of reflected light. 5 and 6 are diagrams showing detection signals when a defect is detected. FIG. 7 is a circuit diagram showing signal processing of the detection signal. 1 is a laser, 6a is a half mirror, 10 is a sample, 11 is a sample stage, 16 is a relay lens, 17 is a prism wedge, 19a is a photodetector, and 19b is a photodetector.

  The substantially parallel light beam emitted from the laser 1 enters the half mirror 6a. Half of the light beam passes through the half mirror 6a and enters the objective lens. The light beam incident on the objective lens 9 is condensed to an appropriate spot diameter and incident on the sample 10. The light incident on the sample 10 is reflected on the surface. The reflected light passes through the objective lens 9 and becomes a substantially parallel light beam. The substantially parallel light beam is separated from the light beam incident on the sample by the half mirror 6a. Therefore, half of the light beam is reflected in the direction of the photodetectors 19a and 19b. The light reflected by the half mirror 6 a enters the photodetector 19 through the relay lens 16. In the present embodiment, a prism wedge 17 is inserted in the optical path between the half mirror 6 a and the relay lens 16. The laser beam is split into two beams by the prism wedge 17 and enters the photodetectors 19a and 19b, respectively. Each of the surface states of the sample is inspected based on the light intensity detected by the photodetector 19. Furthermore, in this embodiment, the sample stage 11 is driven two-dimensionally and the entire surface of the sample is inspected. In this embodiment, the entire sample surface can be inspected by raster scanning.

  Detection of reflected light using this prism wedge 17 will be described with reference to FIGS. In the following, detection of reflected light will be described as a method of explanation from the viewpoint of geometric optics. FIG. 2 shows a state in which a normal part having no defect is inspected. FIG. 2 is a cross-sectional view of the laser beam with the prism wedge 17 inserted at this time. FIG. 2B is a top view showing the optical path of the laser light at this time. Here, 30 is a laser beam reflected by the sample 10. For the sake of explanation, the relay lens 16, the half mirror 6a, the objective lens 9, and the like are omitted in FIG.

  When a normal part having no defect is inspected in the sample 10, the wedge 17 is inserted at a substantially half position of the laser beam 30 as shown in FIG. The wedge 17 has a wedge shape as shown in FIG. 2B, and the angles of the entrance surface and the exit surface are different. Therefore, the light incident on the wedge 17 is emitted with its traveling direction changed. Accordingly, half of the original laser beam 30 enters the wedge 17 and is emitted with the traveling direction changed. A photodetector 19b is provided to detect this light. The remaining half is not incident on the wedge, but is incident on the photodetector 19a with the same traveling direction. Therefore, when inspecting a normal part without a defect, the light intensity of the photodetector 19a and the photodetector 19b is the same. The light detector 19a and the light detector 19b are installed so as to have a distance and an interval that allow the light that has passed through the wedge 17 to be separated.

  Next, a state when inspecting a convex defect will be described with reference to FIGS. 3 and 4 show the state of the sample surface and the corresponding optical path of the laser beam. 31 is a convex defect. The sample stage on which the sample is placed moves and the convex defect 31 is irradiated with laser light. As shown in FIG. 3, first, the laser light reaches an uphill slope. Then, after the sample stage 11 moves and exceeds the apex, the downhill slope is irradiated with laser light as shown in FIG.

  When the uphill slope is irradiated with laser light, its reflection angle changes. Accordingly, the laser beam 30 is incident with a deviation from a substantially half position of the wedge 17. That is, the laser beam 30 shown in FIG. 2A is shifted to the left. Accordingly, the intensity of the light incident on the photodetector 19a is larger than the intensity of the light incident on the photodetector 19b. On the other hand, when the light is irradiated on the slope of the downhill, the laser beam 30 shown in FIG. 2A is shifted to the right. Therefore, the light intensity of the photodetector 19b becomes larger than the light intensity of the photodetector 19a. A convex defect and a concave defect are detected based on changes in signal intensity of the photodetector 19a and the photodetector 19b. The wedge 17 is desirably inserted in a direction corresponding to the scanning direction of the sample stage 11. In scanning with laser light, the wedge 17 is inserted in a direction in which the optical path is inclined geometrically and optically due to the concavo-convex defect, whereby a more accurate inspection can be performed.

  Next, a method for examining detailed information on defects using signals detected by the photodetector will be described with reference to FIG. FIG. 5 is a diagram showing the signal intensity corresponding to the surface state of the sample. As shown in FIG. 5A, it is assumed that a portion having a convex defect is scanned on the surface of the sample. The horizontal axis indicates the position or time of the sample. The vertical direction indicates the height of the defect. FIG. 5B to FIG. 5D show signal intensities corresponding to FIG. FIG. 5B shows the signal intensity of the photodetector 19a, and FIG. 5C shows the signal intensity of the photodetector 19b. FIG. 5E shows the sum of the signal intensity of the photodetector 19a and the signal intensity of the photodetector 19b, and FIG. 5D shows the difference between the signal intensity of the photodetector 19a and the signal intensity of the photodetector 19b. ing. For the sake of explanation, it is assumed that the signal intensity of the photodetector 19a is A and the signal intensity of the photodetector 19b is B. Accordingly, FIG. 5B shows A and FIG. FIG. 5D shows AB and FIG. 5E shows A + B.

  A and B show the same signal intensity when a laser beam is irradiated to a portion having no irregular defect. In addition, each signal strength in a location without a defect is set to 0. When light is applied to the position corresponding to the slope of the uphill, the laser light is detached from the wedge 17 and A gradually increases. Thereafter, the laser light returns in the direction of the wedge 17, and A gradually decreases. At the apex of the defect, A and B show the same signal intensity. When light is applied to the downhill slope beyond the apex of the defect, the light goes toward the wedge 17 and A gradually decreases. Thereafter, the laser beam is removed from the wedge 17, and A gradually increases. When the light is removed from the defect, the original state is restored. Thus, A has a positive peak first, followed by a negative peak. B shows the opposite behavior to A as shown in FIG. Thus, B has a negative peak first, followed by a positive peak. When the convex defect is irradiated with the laser beam and the sample on the sample stage is scanned in this way, A shows a negative value after showing a positive value. On the contrary, B shows a positive value after swinging negatively. Using this order, it is possible to determine whether the defect is a convex defect or a concave defect.

  AB is a signal waveform in which the peak of A is enhanced as shown in FIG. Further, as shown in FIG. 5E, the signal intensity of A + B does not change with time (sample position) and remains 0.

  The inspection apparatus of the present embodiment compares the signals from the two detectors and detects the state of the sample by utilizing the difference in relative intensity. Specifically, it is determined whether the defect is convex or concave by using the signal AB. A slice level H and a slice level L shown in FIG. When the order of exceeding the slice level is H → L, it is determined as a convex defect. In A-B, since the A signal is enhanced, the loss of signal strength is small, and an accurate and highly accurate inspection can be performed. Since no shielding plate is provided on the optical path, detection can be performed without reducing the light intensity. Therefore, loss of signal intensity is reduced, and defects can be detected with a high S / N ratio. Thus, since the defect inspection apparatus according to the present invention can perform inspection with high signal quality, it can perform more accurate and accurate inspection. Alternatively, it is effective in that common mode noise such as noise of the laser light source 1 can be effectively removed. In other words, even when the intensity of light incident on the defect fluctuates due to noise or the like and becomes a vertically asymmetric signal, it is possible to accurately perform classification at the slice level.

  In addition, when detecting a concave defect, the signal of A and B will interchange and it will show the behavior which AB reversed. Therefore, when the order of exceeding the slice level is L → H, a concave defect is indicated. Therefore, it is possible to determine whether the defect is a convex defect or a concave defect based on the order exceeding the slice level. This makes it possible to accurately determine whether the sample has a convex defect or a concave defect.

  Next, defect detection when there is a spot-like defect on the sample surface will be described with reference to FIG. Here, a spot-like defect can be detected as a defect whose reflectance is changed only partially. FIG. 6 is a diagram showing the signal intensity when the sample has a spot-like defect and the laser beam is irradiated to a portion where the reflectance of the surface is weakened only partially. 6B to 6E show the same signals as in FIG. 6B shows the signal A, FIG. 6C shows the signal B, FIG. 6D shows the signal AB, and FIG. 6E shows the signal A + B.

  In this case, since the reflectance of the sample surface is weakened only partially, the spot position of the laser light is not changed, and only the light intensity is weakened. Accordingly, as shown in FIGS. 6B and 6C, the signal strengths of both A and B become weak. Therefore, AB is substantially constant at 0 as shown in FIG. 6 (d). Therefore, the slice levels H and L shown in FIG. 5 are not exceeded. On the other hand, when the reflectance of the sample surface is weakened only partially, the A + B signal is an enhancement of the A (B) signal. Therefore, if A + B exceeds the slice level L, it can be determined that the reflectance of the sample surface is weakened only partially. Further, by setting a slice level H (not shown) and exceeding this A + B, it is possible to determine that the reflectance of the sample surface is only partially increased. Thereby, a spot-like defect can be detected. In A + B, the signal of A (or B) is enhanced, and since no shielding plate is provided on the optical path, there is little loss of signal intensity, and an accurate and highly accurate inspection can be performed.

  With the above signal processing, loss of signal intensity is reduced, and convex defects, concave defects, and spot-like defects can be detected with a high S / N ratio. Then, the defect detection signal and the position of the sample stage are made to correspond to inspect which position of the sample has which type of defect. Thus, since the defect inspection apparatus according to the present invention can perform inspection with high signal quality, it can perform more accurate and accurate inspection. Note that different numbers can be added / subtracted / divided from / to the signals A and B if necessary.

  Next, a signal processing circuit for detecting defects will be described with reference to FIG. FIG. 7 is a circuit diagram for performing signal processing. 50 is a resistor, 51 is a subtracting circuit, 52 is an adding circuit, and 53 to 56 are comparators. Reference numerals 57 to 60 denote defect detection signal generation circuits. These circuits can be incorporated into the photodetector 19 or can be separate circuits.

  The photodetectors 19a and 19b output currents corresponding to the intensity of the received light, and these currents are converted into voltages by the resistors 50, respectively. The signals converted into this voltage are A and B, respectively. A and B are input to the subtraction circuit 51 and the addition circuit 52, respectively. A subtracting circuit 51, which is a comparison means, generates an AB signal that is a difference signal. This AB signal is input to the comparator 53 and the comparator 54. The comparator 53 sets a negative slice level L, and outputs a pulse signal when the AB signal exceeds this slice level L (less than the slice level L). The comparator 54 sets a positive slice level H, and outputs a pulse signal when the AB signal exceeds the slice level H (when it is greater than the slice level H).

  These pulse signals are input to the defect detection signal generation circuit 57. When pulse signals are input in the order of the comparator 53 and the comparator 54, the defect detection signal generation circuit 57 outputs a concave defect detection signal. When pulse signals are input in the order of the comparator 54 and the comparator 53, the defect detection signal generation circuit 57 outputs a convex defect detection signal. By performing such differential detection, it is possible to accurately determine whether the defect is a convex defect or a concave defect.

  The adder circuit 52 generates an A + B signal that is a sum signal. This A + B signal is input to 55 and 56. In 55, a negative slice level L is set, and when this slice level L is exceeded, a pulse signal is output. The comparator 56 sets a positive slice level H, and a pulse signal is output when the slice level H is exceeded.

  The pulse signal output from the comparator 55 is input to the defect detection signal generation circuit 59. Therefore, when a defect detection signal is output from the defect detection signal generation circuit 59, this indicates that there is a defect having a reduced reflectance on the sample surface. The pulse signal output from the comparator 56 is input to the defect detection signal generation circuit 60. Therefore, when a defect detection signal is output from the defect detection signal generation circuit 60, this indicates that there is a defect having a high reflectance on the sample surface. By comparing these defect detection signals with the scanning timing of the sample stage, the location where the sample defect exists is specified. Thereby, the position and kind of the defect which exists in a sample can be specified. In addition, an amplifier circuit or the like may be provided in the photodetector to amplify the signal. In this case, the signals A and B are amplified signals. The amplification factors of A and B can be different. The difference signal can be generated, for example, as a difference between the outputs after performing different operations on the outputs from the photodetectors. The sum signal can also be generated as a sum of outputs by subjecting each output from the detector to signal processing as appropriate.

  The branching of the reflected light can be performed using, for example, a knife edge prism. However, by performing defect detection with the above-described optical system using the wedge 17, the optical components for the photodetector (for example, the relay lens 16 and the wedge 17) can perform differential detection with one set. Therefore, the number of parts can be reduced. Therefore, cost reduction can be achieved. Further, the optical system is not limited to the illustrated optical system, and other optical components may be added or changed. For example, the positions of the laser beam 1 and the photodetector 19 in FIG. 1 may be changed. Even in this case, since the light beam reflected by the sample 10 is separated by the half mirror 6a, the same effect can be obtained by inserting the wedge 17 or the like on the optical path. Furthermore, detection by a confocal optical system can be easily performed. That is, it is possible to detect with the confocal optical system by forming an image in front of the light detector 19 and providing a pinhole at the image formation point. Thereby, a highly accurate defect inspection can be performed.

Embodiment 2 of the Invention
The defect inspection apparatus according to this embodiment will be described with reference to FIG. FIG. 8 is a configuration diagram showing the configuration of the optical system of the defect inspection apparatus according to this embodiment. In this embodiment, the sample stage is not scanned, but the light beam is scanned using a vibrating mirror. Since the same reference numerals as those in FIG. 1 indicate the same configuration, the description thereof is omitted. 4a is a vibrating mirror, 6 is a polarization beam splitter, 5a and 5b are relay lenses, and 8 is a quarter-wave plate.

  The light beam emitted from the laser 1 is transmitted through the polarization beam splitter 6 only by the P-polarized light. This light beam is reflected in the direction of the sample 10 by the vibrating mirror 4a. The light passes through the relay lens 5a, the relay lens 5b, and the quarter wavelength plate and enters the objective lens 9. The light incident on the objective lens 9 is condensed to an appropriate spot diameter and enters the sample 10. Then, the light beam again passes through the objective lens 9, the quarter wavelength plate 8, the relay lens 5b, and the relay lens 5a, and enters the vibrating mirror 4a. The light incident on the oscillating mirror 4 a is reflected in the direction of the polarization beam splitter 6. Up to this point, the light passes through the quarter-wave plate 8 twice in a reciprocating manner, so that P-polarized light is polarized into S-polarized light. Accordingly, the light beam is reflected by the polarization beam splitter 6. Thereby, the fall of light intensity can be suppressed. Of course, a polarizing beam splitter and a quarter wavelength plate may be used in the optical system shown in the first embodiment.

  The light reflected by the polarization beam splitter 6 is condensed on the photodetector 19 by the relay lens 16. As in the first embodiment, a wedge 17 is inserted on this optical path. Accordingly, half of the light beam is incident on the photodetector 19a and the other half is incident on the photodetector 19b, and the light intensity is detected. Since the detection here is the same as the detection described in the first embodiment, the description thereof is omitted. In the present embodiment, a light beam is scanned with a vibration beam to detect defects on the entire surface of the sample. As the vibrating mirror 4a, for example, a galvanometer mirror or a polygon mirror can be used. The same effect as in the first embodiment can be obtained by scanning using such a vibrating mirror.

Embodiment 3 of the Invention
The defect inspection apparatus according to this embodiment will be described with reference to FIGS. 9 and 10 are configuration diagrams showing the configuration of the defect inspection apparatus. 1 and 8 indicate the same configuration and the description thereof is omitted. FIG. 9 includes the optical system corresponding to the first embodiment, that is, FIG. FIG. 10 includes an optical system corresponding to the second embodiment, that is, FIG.

  This embodiment is different from the first and second embodiments in that the wedge 17 is provided between the relay lens 16 and the photodetector 19. Since other optical system configurations, detection methods, and signal processing are the same as those in the first and second embodiments, description thereof is omitted.

  In this embodiment, the wedge 17 can be arranged at a position other than the position of the pupil. That is, in this embodiment, the wedges 17 may be inserted in half of the spots so that the laser light can be divided into two. Accordingly, the half mirror 6 and the photodetector 19 shown in FIG.

If the light reflected by the sample is on an optical path separated from the light incident on the sample, the wedge 17 can be disposed. Therefore, it can also be arranged at a position other than the pupil. The same effects as those of the first and second embodiments can be obtained by the optical system having such an arrangement.

  In the defect inspection apparatus according to the present invention, since the wedge 17 does not need to be arranged at the position of the pupil, it is not subject to restrictions on the arrangement of optical components, and the degree of freedom of design can be expanded.

Embodiment 4 of the Invention
A defect inspection apparatus according to Embodiment 4 of the present invention will be described with reference to FIG. FIG. 11 is a configuration diagram showing the overall configuration of the defect inspection apparatus. 1 is a laser, 2 is a diffraction grating, 3 is a Fourier transform lens, 4 is a reflection mirror, 5 is a relay lens, 6 is a polarization beam splitter, 7 is a dichroic mirror, 8 is a quarter-wave plate, 9 is an objective lens, 10 Is a sample, 11 is a sample stage, 12 is an autofocus optical system, 13 is a relay lens, 14 is a reflection mirror, 15 is a reflection mirror, 16 is a relay lens, 17 is a wedge, 18 is a relay lens, and 19 is a photodetector. is there. In the present embodiment, inspection is performed by irradiating a sample with a single light beam by the diffraction grating 2 as a plurality of light beams.

  Laser light is irradiated on the sample surface by these optical systems, and the scattered light can be detected. 20 is a personal computer (hereinafter referred to as PC), 21 is a defect data memory, 22 is a stage controller, and 23 is a stage drive unit. These perform control for scanning the sample on the sample stage 11 and processing of detection data. Thereby, the defect position of a sample can be specified.

  First, the optical system of the defect inspection apparatus according to the present invention will be described with reference to FIG. Light emitted from the laser 1 enters the diffraction grating 2 and is converted into a multi-beam. Here, since a two-dimensional diffraction grating is used for the diffraction grating 2, a beam array of a matrix-like two-dimensional light beam is obtained. Hereinafter, this beam array is referred to as a multi-beam. The multi-beam is incident on the Fourier transform lens 3 and becomes a substantially parallel light beam. The multi-beam is reflected by the reflecting mirror 4 in the direction of the sample. Further, the multi-beam passes through the relay lens 5 and enters the polarization beam splitter 6. In the polarization beam splitter, only P-polarized light is transmitted. Further, the multi-beam passes through the dichroic mirror 7 and the quarter wavelength plate 8 and enters the objective lens 9. The objective lens 9 focuses the multi-beam on a minute spot and makes it incident on the sample 10. When the sample 10 has defects such as foreign matter, steps, and scratches, the incident light is scattered by these defects. Due to this scattering, a difference occurs in the amount of detected light depending on the presence or absence of defects. Therefore, the defect information of the sample can be detected by detecting the reflected light from the sample surface.

  Next, an optical path where the light reflected by the sample reaches the detector will be described. The light reflected by the sample 10 passes through the objective lens 9 and the quarter-wave plate 8 and enters the dichroic mirror 7. A part of the light is reflected by the dichroic mirror 7 and enters the autofocus optical system 12. The autofocus optical system 12 includes a light source that emits light having a wavelength different from that of the laser 1 and reflected by a dichroic mirror. The light from this light source is reflected by the dichroic mirror 7 and applied to the sample 10. The reflected light is received by the autofocus optical system 12. The position of the objective lens 9 or the sample 10 is adjusted using the output signal of the received light, and the distance between the sample 10 and the objective lens 9 is optimized. A half mirror or the like may be used instead of the dichroic mirror 7.

  The light transmitted through the dichroic mirror 7 is incident on the polarization beam splitter 6. Since it has passed through the quarter-wave plate 8 twice in the optical path so far, the light that was P-polarized light is polarized to S-polarized light. Therefore, it is reflected by the polarization beam splitter 6. The light reflected by the polarization beam splitter 6 enters the relay lens 13, the reflection mirror 14, the reflection mirror 15, and the relay lens 16 in this order. The light transmitted through the relay lens 16 is once focused and then enters the zoom lens 18.

  The light transmitted through the zoom lens 18 enters the photodetector 19. The light detector 19 includes light receiving elements configured in a matrix, and can detect the amount of light corresponding to each light beam of the multi-beams. For example, a photodiode array is used as the light receiving element. The zoom lens 18 is composed of a zoom lens system, and the magnification can be adjusted. Therefore, it can adjust so that each light beam may inject into each light receiving element.

  In the present embodiment, a wedge 17 is arranged at the pupil between the relay lens 16 and the zoom lens 18. The wedge 17 is inserted at a position substantially half the cross section of the focused light spot. In the two-dimensional multi-beam, a wedge is arranged at the position of the pupil, and each light beam is split in half. Each of the branched multi-beams passes through the relay lens 18 to become a substantially parallel light beam and enters the photodetector 19.

  Then, the sample stage 11 is scanned in a spiral manner to scan the entire surface of the sample. This spiral scanning is performed by rotating and translating the sample stage by a stage controller 22 connected to the PC. The rotational movement is performed by the θ stage 23d and the translational movement is performed by the r stage 23b. The PC 21 stores stage control data such as the moving speed in the r direction and the rotational speed in the θ direction. Based on this data, the stage controller 22 drives the r stage 23b and the θ stage 23d. The r stage 23b is moved by, for example, a linear guide, a ball screw, and a servo motor. The rotation of the θ stage 23d is performed by a drive motor. The sample is chucked by the chuck mechanism of the sample stage. Thereby, spiral scanning can be performed.

  Further, the θ stage angle detection system 23 and the r stage position detection system are provided with an encoder 8 (not shown) so that the respective angles and positions can be grasped. The output from this encoder is stored in the defect data memory 21. Since these position signals are synchronized with the defect detection signal, the position of the defect can be checked. The data stored in the defect data memory is transmitted to the PC, and detailed information on the defect of the sample is displayed.

  Next, the multi-beams incident on the photodetector 19 will be described with reference to FIG. FIG. 12 is a cross-sectional view of the multi-beam on the photodetector 19. A white spot (line A) indicates a beam whose traveling direction is not changed by the wedge 17. A shaded spot indicates a beam whose traveling direction has been changed by the wedge 17. The positions of A and B are shifted by the amount that the traveling direction of the beam has changed to the wedge 17. That is, one beam in the multi-beam is branched into A and B by the wedge 17 respectively. The A and B signals are detected, respectively, and the defect is inspected. The photodetector 19 includes two sets of light receiving elements arranged in a 6 × 3 matrix in order to detect this multi-beam. Each of the 6 × 3 light receiving elements is paired, and signal processing is performed on the signals of the paired light receiving elements. That is, the above-described signal processing is performed on each pair of 6 × 3 light beams. Since each signal processing is the same as the processing described in the first embodiment, the description is omitted. If a defect is detected, the defect type is detected by specifying the type and location. Therefore, loss of signal intensity can be reduced, and accurate and highly accurate defect detection can be performed. Thus, the same effect can be obtained even if a two-dimensional multi-beam is used. In the description of FIG. 12, the light detectors 19a and the light detectors 19b are alternately arranged. However, the light receiving elements of the light detectors are placed at arbitrary positions according to the positions of the light spots. Can be arranged. For example, all the light receiving elements of the photodetector 19a and all the light receiving elements of the photodetector 19b may be arranged separately. Of course, it may be in the middle. Since the arrangement of the photodetectors 19a and 19b is determined by the angle of the wedge and the optical path length, it can be adjusted to an arbitrary position. Thereby, a large number of photodetectors can be easily combined.

  Further, the defect inspection is performed using the two-dimensional diffraction grating in the optical system shown in FIG. 11, but a one-dimensional diffraction grating may be used. That is, one light beam emitted from the laser light source is converted into a one-dimensional line-shaped multi-beam by a one-dimensional diffraction grating. Then, this multi-beam is converted into a substantially parallel light beam by the Fourier transform lens 3. A light receiving element for receiving each light beam branched by the wedge 17 is provided in the photodetector 19. In this case, two sets of four line-shaped light receiving elements are provided and are paired. The signal processing described above is performed on the signals of the paired light receiving elements. Therefore, the same inspection can be performed with a one-dimensional multi-beam. In this case, it is desirable to insert a wedge in each half of the one-dimensional multi-beam as shown in FIG. That is, the wedge 17 can be arranged at a position other than the pupil as in the case of one light beam shown in the first to third embodiments. This eliminates the need to place a wedge at the position of the pupil, thus increasing the degree of design freedom. Note that raster scanning other than spiral scanning can be used in a one-dimensional or two-dimensional multi-beam.

Other embodiments.
In the above-described embodiment, light is branched in two directions using a wedge, but other light branching means may be used. For example, the same effect can be obtained by providing a knife edge prism that reflects incident light into two branched lights at a half position on the optical path. If the wedge is used, the detector can be placed close to the optical beam because the angle deviation of the light beam is small. Thereby, wiring for signal processing and the like can be simplified, and space saving can be achieved. Furthermore, the interval between the two photodetectors can be easily adjusted by changing the wedge angle and the optical path length to the photodetector. Therefore, the photodetector can be placed in an arrangement suitable for inspection, and is particularly suitable when using a multi-beam. Note that the number of lights branched by the wedge is not limited to two, and may be four. Further, the angle of the entrance surface and the exit surface of the wedge may be inclined with respect to the respective sides to emit the exit light obliquely. Further, the wedge 17 may be arranged at a position other than substantially half the position on the optical path. Two wedges having different emission angles or one prism having an equivalent function may be arranged on the optical path. Even in this case, substantially the same effect can be obtained. Moreover, the edge part shape of a wedge does not need to be linear in an optical path cross section.

  As the laser light scanning method, for example, a scanning method other than raster scanning such as step scanning or spiral scanning may be used. Further, the same effect can be obtained even if the number of beams used for the multi-beam is not the number shown. In the drawings shown in the above-described embodiments, the light detector is disposed at a position where the light beam is imaged, but a light detector may be disposed at a position other than the image formation position. Furthermore, a confocal optical system can be constructed by arranging a pinhole at the imaging position and detecting light that has passed through the pinhole. By using this confocal optical system, inspection with higher accuracy can be performed. Further, the light may be received by the photodetector via the optical fiber. The PC, stage controller, stage drive system, autofocus optical system, detection system, etc. used in the fourth embodiment can also be used in the first to third embodiments.

  The present invention is not limited to the above-described embodiments, and various modifications can be made. For example, the same effect can be obtained by using other various optical components and optical elements without being limited to the illustrated optical system. Further, the present invention can be used not only for defect inspection of mask blanks, masks, semiconductor devices, and semiconductor wafers but also for other samples. In particular, the manufacturing yield of semiconductor devices can be improved by using the inspection apparatus of the present invention for inspection processing of masks or mask blanks or inspection processing of semiconductor wafers. In manufacturing a typical semiconductor device, a mask original plate is set in an exposure apparatus, and a resist-formed wafer is exposed using light, an ion beam, an electron beam, or the like. The semiconductor wafer subjected to the exposure process is subjected to a development process, and a resist pattern is formed on the wafer. According to this pattern, a widely known thin film deposition process, etching process, oxidation process, ion implantation process, and the like are performed to form a semiconductor device. With the mask inspected using the inspection apparatus or inspection method of the present invention, or the mask using mask blanks, the exposure process in the manufacture of the semiconductor device can be carried out. Also, a semiconductor device manufacturing process can be performed on a wafer inspected using the inspection apparatus or inspection method of the present invention to manufacture a semiconductor device. The optical system of the present invention can be applied not only to a defect inspection apparatus that is one of sample states, but also to an optical scanning device that detects a general sample state, such as a microscope.

1 laser, 2 diffraction grating, 3 Fourier transform lens,
4 reflective mirror, 4a vibrating mirror, 5, 5a, 5b relay lens,
6 polarizing beam splitter, 7 dichroic mirror 8 1/4 wavelength plate, 9 objective lens, 10 sample, 11 sample stage,
12 autofocus optical system, 13 relay lens, 14 reflecting mirror,
15 reflection mirror, 16 relay lens, 17 wedge,
18 Relay lens 19, 19a, 19b Photodetector
20 PC, 21 defect data memory, 22 stage controller,
23 stage drive unit, 23 a r stage position detection system 23 b r stage,
23c θ stage position detection system, 23d θ stage, 30 laser beam 50 resistance, 51 subtraction circuit, 52 addition circuit
53, 54, 55, 56 Comparator,
57, 59, 60 Defect detection signal generation circuit

Claims (9)

A light source that generates a light beam;
A sample stage on which the sample is placed;
A one-dimensional diffraction grating for converting light from the light source into a one-dimensional multi-beam;
Scanning means for relatively moving the sample placed on the sample stage and the multi-beam irradiated to the sample from the light source;
The multi-beams are arranged at positions where the multi-beams are arranged in one row, and the direction of incident light is changed to branch each of the multi-beams reflected by the sample into two multi-beams. Means,
A plurality of light receiving elements arranged in a line, and a first light detecting means for detecting each of the multi-beams branched by the light branching means;
A second light detection comprising a plurality of light receiving elements arranged in a line in pairs with the light receiving elements of the first light detecting means, and detecting each of the other multi-beams branched by the light branching means Means,
A circuit for obtaining a difference signal for each of the multi-beams based on a difference between signals output from the pair of light receiving elements;
A first comparison circuit for comparing the difference signal with a predetermined first value;
A second comparison circuit for comparing the difference signal with a predetermined second value;
A circuit for obtaining a sum signal for each of the multi-beams based on signals output from the pair of light receiving elements;
A third comparison circuit for comparing the sum signal with a predetermined third value;
Based on the order of the timing when the difference signal exceeds the first value and the timing when the difference value exceeds the second value, it is determined whether it is a convex defect or a concave defect,
A defect inspection apparatus that detects a spot-like defect in which the reflectance changes only in part based on the comparison result of the third comparison circuit.   The light branching means is disposed at a position that occupies approximately half of the light beam reflected from a portion of the sample where there is no defect, and the multibeam is changed by changing the traveling direction of the light beam incident on the light branching means. The defect inspection apparatus according to claim 1, wherein the defect inspection apparatus is branched.   3. The defect inspection apparatus according to claim 1, wherein the light branching means is a wedge arranged on an optical path of a light beam. An inspection method for inspecting a defect of a sample by irradiating the sample with a light beam and detecting the reflected light,
Generating a light beam;
Converting light from the light source into a one-dimensional multi-beam;
Relatively moving the position of the multi-beam and the sample;
By changing the traveling direction of the incident light by the light branching means arranged at the position where the multi-beams are arranged in one row, each of the one row of multi-beams reflected by the sample is branched, and two multi-beams A step to beam,
Detecting each of one multi-beam from the light branching means by a first light detecting means having a plurality of light receiving elements arranged in a line;
Each of the other multi-beams from the light branching means is detected by a second light detecting means having a plurality of light receiving elements arranged in a line in pairs with the light receiving elements of the first light detecting means. And steps to
Obtaining a difference signal for each of the multi-beams based on a difference between signals output from the pair of light receiving elements;
Comparing the difference signal with a predetermined first value;
Comparing the difference signal with a predetermined second value;
Determining whether the difference signal is a convex defect or a concave defect based on the order of the timing when the difference signal exceeds the first value and the timing when the difference signal exceeds the second value;
Obtaining a sum signal for each of the multi-beams based on signals output from the pair of light receiving elements;
Comparing the sum signal with a predetermined third value;
A defect inspection method comprising: detecting a spot-like defect whose reflectance has changed only in part based on a comparison result between the sum signal and the third value.   5. The defect inspection according to claim 4, wherein the multi-beam is split into two multi-beams by an optical branching unit that splits a light beam reflected from a portion free of defects of the sample into approximately half. Method. A light source and means for converting a light beam from the light source into a one-dimensional multi-beam;
Means for scanning the multi-beam on an object;
The multi-beams are arranged at positions arranged in a row, and each row of multi-beams reflected by the object is branched by changing a traveling direction of incident light, so that the first multi-beams are branched. And a second multi-beam means for branching such that the relative intensities of the first multi-beam and the second multi-beam change based on the state of the object,
A plurality of light receiving elements arranged in a line, and a pair of a first light detecting means for detecting each of the first multi-beams and a light receiving element of the first light detecting means; A second light detecting means having a plurality of light receiving elements arranged and detecting each of the second multi-beams;
Means for obtaining a difference signal for each of the multi-beams based on a difference between signals output from the pair of light receiving elements;
Means for comparing the difference signal with a predetermined first value;
Means for comparing the difference signal with a predetermined second value;
Means for determining whether the difference signal is a convex defect or a concave defect based on the order of the timing when the difference signal exceeds the first value and the timing when the difference signal exceeds the second value;
Means for obtaining a sum signal for each of the multi-beams based on signals output from the pair of light receiving elements;
A third comparison circuit for comparing the sum signal with a predetermined third value;
An optical scanning device that detects a spot-like defect whose reflectance is changed only in part based on a comparison result of the third comparison circuit.   The said branching means is provided on the optical path of the light reflected by the flat part of the said target object, The edge of the said branching means has a wedge optical element which exists in the center of the said optical path. Optical scanning device. Converting the light from the light source into a one-dimensional multi-beam;
Scanning the multi-beam on an original plate;
By changing the traveling direction of the incident light by the light branching means arranged at the position where the multibeams are arranged in one row, each of the one row of multibeams reflected by the original plate is branched, Branching to change the relative intensities of the first and second multi-beams based on the state of the original plate, and
A plurality of light receiving elements arranged in a line; and detecting each of the branched first multi-beams; and being arranged in a line with the light receiving elements of the first light detecting means. Detecting each of the second multi-beams by a second light detecting means having a plurality of light receiving elements;
Obtaining a difference signal for each of the multi-beams based on a difference between signals output from the pair of light receiving elements;
Comparing the difference signal with a first value;
Comparing the difference signal with a second value;
Inspecting the convex defect and the concave defect based on the order of the timing when the difference signal exceeds the first value and the timing when the difference signal exceeds the second value;
Obtaining a sum signal for each of the multi-beams based on signals output from the pair of light receiving elements;
Comparing the sum signal with a predetermined third value;
Based on a comparison result between the sum signal and the third value, a step of inspecting a spot-like defect whose reflectance is changed only in part, and a step of setting the inspected original plate in an exposure apparatus;
Exposing the wafer with an exposure pattern of the original plate;
Developing the exposed wafer;
A method for manufacturing a semiconductor device, comprising: Converting light from the light source into a one-dimensional multi-beam;
Scanning the multi-beam over a mask;
By changing the traveling direction of the incident light by the light branching means arranged at the position where the multibeams are arranged in one row, each of the one row of multibeams reflected by the mask is branched, Branching to change the relative intensities of the first and second multi-beams based on the state of the original plate, and
A step of detecting each of the first multi-beams by a first light detecting means having a plurality of light receiving elements arranged in a line, and a pair of light receiving elements of the first light detecting means. Detecting each of the second multi-beams by a second photodetecting means having a plurality of light receiving elements arranged in
Obtaining a difference signal for each of the multi-beams based on a difference between signals output from the pair of light receiving elements;
Comparing the difference signal with a first value;
Comparing the difference signal with a second value;
Inspecting the convex defect and the concave defect based on the order of the timing when the difference signal exceeds the first value and the timing when the difference signal exceeds the second value;
Obtaining a sum signal for each of the multi-beams based on signals output from the pair of light receiving elements;
Comparing the sum signal with a predetermined third value;
Based on a comparison result between the sum signal and the third value, a step of inspecting a spot-like defect in which the reflectance changes only in part, and a step of forming a resist layer on the inspected wafer;
Exposing the wafer on which the resist layer is formed according to a mask pattern;
Developing the exposed wafer;
A method for manufacturing a semiconductor device, comprising:

Images