JP6851178B2 - Pattern inspection method and pattern inspection equipment - Google Patents

Description

The present invention relates to a pattern inspection method and a pattern inspection apparatus. For example, the present invention relates to a pattern inspection technique for inspecting a pattern defect of a sample object used in semiconductor manufacturing, and relates to an inspection device and a method for inspecting a photomask used when manufacturing a semiconductor element or a liquid crystal display (LCD).

In recent years, with the increasing integration and capacity of large-scale integrated circuits (LSIs), the circuit line width required for semiconductor elements has become narrower and narrower. These semiconductor elements use an original image pattern (also referred to as a mask or reticle, hereinafter collectively referred to as a mask) on which a circuit pattern is formed, and the pattern is exposed and transferred onto a wafer by a reduction projection exposure apparatus called a stepper. Manufactured by forming a circuit. Therefore, in manufacturing a mask for transferring such a fine circuit pattern to a wafer, a pattern drawing device using an electron beam capable of drawing the fine circuit pattern is used. A pattern circuit may be drawn directly on the wafer using such a pattern drawing device. Alternatively, an attempt is being made to develop a laser beam drawing apparatus that draws using a laser beam in addition to the electron beam.

Further, improvement of the yield is indispensable for manufacturing an LSI, which requires a large manufacturing cost. However, as represented by 1 gigabit class DRAM (random access memory), the patterns constituting the LSI are on the order of submicron to nanometer. One of the major factors for reducing the yield is a pattern defect of a mask used when exposing and transferring an ultrafine pattern on a semiconductor wafer by photolithography technology. In recent years, with the miniaturization of LSI pattern dimensions formed on semiconductor wafers, the dimensions that must be detected as pattern defects have become extremely small. Therefore, it is necessary to improve the accuracy of the pattern inspection apparatus for inspecting defects in the transfer mask used in LSI manufacturing.

As an inspection method, an optical image obtained by capturing a pattern formed on a sample such as a lithography mask using a magnifying optical system at a predetermined magnification is compared with a design data or an optical image obtained by capturing the same pattern on the sample. There is a known method of inspecting by doing so. For example, as a pattern inspection method, "die to die inspection" in which optical image data obtained by capturing the same pattern in different places on the same mask are compared with each other, or a pattern is used as a mask using CAD data with a pattern design. The drawing data (design pattern data) converted into the device input format for input by the drawing device at the time of drawing is input to the inspection device, a design image (reference image) is generated based on this, and the pattern is imaged. There is a "die to database inspection" that compares the measured optical image with the measured data. In the inspection method in such an inspection apparatus, the sample is placed on the stage, and the moving of the stage causes the luminous flux to scan the sample to perform the inspection. The sample is irradiated with a luminous flux by a light source and an illumination optical system. The light transmitted or reflected from the sample is imaged on the sensor via the optical system. The image captured by the sensor is sent to the comparison circuit as measurement data. In the comparison circuit, after the images are aligned with each other, the measurement data and the reference data are compared according to an appropriate algorithm, and if they do not match, it is determined that there is a pattern defect.

For semiconductor products with short product cycles, shortening the manufacturing time is an important item. When a defective mask pattern is exposed and transferred to a wafer, the semiconductor device made from the wafer becomes defective. Therefore, it is important to perform a pattern defect inspection of the mask. Then, the defect found in the defect inspection is corrected by the defect correction device. However, fixing all the defects found will increase the manufacturing time and reduce the product value. As the development of the inspection device progresses, the inspection device determines that there is a pattern defect even if a very small deviation occurs. However, when the mask pattern is transferred onto the wafer with an actual exposure device, it is not caused by a defect such as a circuit disconnection or / or short circuit on the wafer, and if the line width of the pattern is within the permissible value, the integrated circuit Can be used as. Therefore, it is desired to use the mask for exposure transfer while leaving the state of the defect, instead of correcting the defect each time. For that purpose, it is sufficient that the pattern size of the pattern image when transferred by the exposure apparatus can be measured. However, while the exposure apparatus reduces the mask pattern and forms an image on the wafer, the inspection apparatus enlarges the mask pattern and forms an image on the sensor. Therefore, the configuration of the optical system on the secondary side with respect to the mask substrate is different in the first place. With such a configuration, the inspection device has a higher resolution for the mask than the exposure device and can detect minute defects. On the other hand, the inspection device can detect the shape of the pattern image when transferred by the exposure device. It is difficult to reproduce.

Here, a dedicated machine for inspecting an exposure image to be exposed and transferred by an exposure apparatus using an image called an aerial image is disclosed (see, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2001-235853

Here, although it seems that it is not yet known at the time of filing the application of the present application, the optics of the exposure apparatus is used when an optical image is acquired by the inspection apparatus in order to approach the pattern image when the image is transferred by the exposure apparatus described above. It is being studied to take an image with a low resolution according to the conditions. However, since the reference image created for comparison with the optical image is based on the design data, the reference image was captured under the optical conditions of the exposure apparatus including the illumination shape including the complicated aperture condition in the generation accuracy of the reference image. It is difficult to match the image to the extent that it can be compared with the optical image. Therefore, when the optical image captured under the optical conditions of the exposure apparatus and the reference image are compared, there is a problem that an error occurs in the pattern size in the image. Therefore, an error occurs in the pattern size even in the normal part where there is no defect. In such a state, there is a problem that it is difficult to obtain the pattern size of the defective portion with high accuracy.

Therefore, the present invention provides an inspection method and an apparatus capable of inspecting the pattern size of a defective portion from an optical image captured in order to bring it closer to a pattern image when transferred by an exposure apparatus.

The pattern inspection method of one aspect of the present invention is
The process of acquiring an optical image of the mask substrate on which the pattern is formed, and
The process of creating a reference image corresponding to an optical image,
A second used for a reference image in which the pattern dimension of the pattern normal part in the optical image specified using the first threshold level of the image gradation value profile and the pattern dimension of the corresponding part in the reference image are the same size. The process of measuring the threshold level and
Using the second threshold level, the step of measuring the pattern dimension of the corresponding part in the reference image corresponding to the pattern abnormality part in the optical image, and the process of measuring the pattern dimension of the corresponding part in the reference image.
The first threshold level is used to measure the tentative pattern size of the pattern abnormality portion in the optical image, and the pattern dimension of the corresponding portion in the reference image using the second threshold level is the first threshold value. A step of calculating the value obtained by dividing the value obtained by dividing the temporary pattern dimension of the pattern abnormality portion in the optical image using the level by the set dimension as the pattern dimension of the pattern abnormality portion in the optical image.
Regarding the pattern abnormality location, the process of comparing and outputting the pattern dimensions measured from the reference image and the pattern dimensions measured from the optical image, and
It is characterized by being equipped with.

Further, it is preferable to further include a step of detecting a plurality of edge pairs in the same direction of the pattern in the optical image.

Further, as a region of a pattern abnormality portion in the optical image, a defective pixel in which the gradation difference between the pixel in the optical image and the corresponding pixel in the reference image exceeds the third threshold value and a plurality of pixels around the defective pixel are used. Pixel group to be composed is used
When measuring the pattern size of the pattern abnormality part in the optical image, the size between a plurality of edge pairs in the pixel group is measured, and the size is measured.
When measuring the pattern size of the corresponding part in the reference image corresponding to the pattern abnormality part in the optical image, the dimension between the plurality of edge pairs corresponding to the plurality of edge pairs in the pixel group is measured.
When comparing the pattern dimensions measured from the optical image and the pattern dimensions measured from the reference image, the edge pair that maximizes the difference between the optical image and the reference image among the dimensions between the plurality of edge pairs in the pixel group. It is preferable to determine the pattern dimension between them as the pattern dimension of the pattern abnormality part and output it.

Further, a step of acquiring a second optical image of the mask substrate using a luminous flux having a numerical aperture higher than the numerical aperture (NA) when the luminous flux used for the optical image is incident on the objective lens from the mask substrate. When,
The process of creating a second reference image corresponding to the second optical image, and
A process of comparing the second optical image and the second reference image pixel by pixel, and
As a result of comparing the second optical image with the second reference image, the step of identifying the pattern abnormality portion in the optical image using the defective pixels determined to be defective, and
It is preferable to further provide.

The pattern inspection device of one aspect of the present invention is
An optical image acquisition mechanism that acquires an optical image of the mask substrate on which the pattern is formed,
A reference image creation unit that creates a reference image corresponding to an optical image,
A second used for a reference image in which the pattern dimension of the pattern normal part in the optical image specified by using the first threshold level of the image gradation value profile and the pattern dimension of the corresponding part in the reference image are the same size. A threshold level measuring unit that measures the threshold level and
A first pattern dimension measuring unit that measures the pattern dimensions of the corresponding points in the reference image, which corresponds to the pattern abnormality points in the optical image, using the second threshold level.
The first threshold level is used to measure the tentative pattern size of the pattern abnormality portion in the optical image, and the pattern dimension of the corresponding portion in the reference image using the second threshold level is the first threshold value. A second pattern dimension measuring unit that calculates the value obtained by dividing the value obtained by dividing the temporary pattern dimension of the pattern abnormality portion in the optical image using the level by the set dimension as the pattern dimension of the pattern abnormality portion in the optical image. When,
A comparison unit that compares the pattern dimensions measured from the optical image with the pattern dimensions measured from the reference image for the pattern abnormality points.
It is characterized by being equipped with.

According to the present invention, it is possible to inspect the pattern size of a defective portion from an optical image captured so as to be close to a pattern image when transferred by an exposure apparatus.

It is a block diagram which shows the structure of the pattern inspection apparatus in Embodiment 1. FIG. FIG. 5 is a conceptual diagram for comparing the numerical aperture of the inspection device and the numerical aperture of the exposure device according to the first embodiment. It is a flowchart which shows a part of the main part process of the inspection method in Embodiment 1. FIG. It is a flowchart which shows the rest of the main part process of the inspection method in Embodiment 1. FIG. It is a figure which shows an example of the structure of the optical system in the high NA inspection mode in Embodiment 1. FIG. It is a conceptual diagram for demonstrating the inspection area in Embodiment 1. FIG. It is a figure for demonstrating the filter processing in Embodiment 1. FIG. It is a block diagram which shows the internal structure of the low NA inspection circuit in Embodiment 1. FIG. It is a figure which shows an example of the structure of the optical system in the low NA inspection mode in Embodiment 1. FIG. It is a figure which shows an example of the image gradation value profile of the pattern in an optical image and the pattern in a reference image in Embodiment 1. It is a figure which shows the defect area image and an example of an edge pair in Embodiment 1. FIG. It is a figure for demonstrating the method of edge recognition of a pattern in Embodiment 1. FIG. It is a figure which shows an example of the mechanism which changes the illumination shape in Embodiment 1. FIG. It is a figure which shows another example of the mechanism which changes the illumination shape in Embodiment 1. FIG. It is a figure which shows another example of the mechanism which changes the illumination shape in Embodiment 1. FIG.

Embodiment 1.
FIG. 1 is a configuration diagram showing a configuration of a pattern inspection device according to the first embodiment. In FIG. 1, the inspection device 100 for inspecting defects in the pattern formed on the mask substrate 101 includes an optical image acquisition mechanism 150 and a control system circuit 160 (control unit).

The optical image acquisition mechanism 150 includes a light source 103, a transmission illumination optical system 170, an illumination shape switching mechanism 171, a movablely arranged XYθ table 102, an objective lens 104, a reflection illumination optical system 172, an aperture diaphragm 180, and an imaging optical system. It has 178, a photodiode array 105 (an example of a sensor), a sensor circuit 106, a stripe pattern memory 123, and a laser length measuring system 122. The mask substrate 101 is placed on the XYθ table 102. The mask substrate 101 includes, for example, a photomask for exposure that transfers a pattern onto a wafer. Further, in this photomask, a pattern composed of a plurality of graphic patterns to be inspected is formed. The mask substrate 101 is arranged on the XYθ table 102, for example, with the pattern forming surface facing downward.

In the control system circuit 160, the control computer 110, which is a computer, transmits the position circuit 107, the comparison circuit 108, the expansion circuit 111, the reference circuit 112, the autoloader control circuit 113, the table control circuit 114, and the autofocus (AF) via the bus 120. ) Control circuit 140, inspection mode switching control circuit 144, low NA inspection circuit 146, magnetic disk device 109, magnetic tape device 115, flexible disk device (FD) 116, CRT 117, pattern monitor 118, and printer 119. .. Further, the sensor circuit 106 is connected to the stripe pattern memory 123, and the stripe pattern memory 123 is connected to the comparison circuit 108. Further, the XYθ table 102 is driven by an X-axis motor, a Y-axis motor, and a θ-axis motor. Further, the reflection illumination optical system 172 includes an objective lens 104 and a polarization beam splitter 174 (polarizing element). The polarization beam splitter 174 is moved in and out of the optical path by the drive mechanism 176.

In the inspection device 100, a high-magnification inspection optical system is configured by a light source 103, an XYθ table 102, a transmission illumination optical system 170, an objective lens 104, a photodiode array 105, and a sensor circuit 106. For example, an inspection optical system having a magnification of 200 to 300 times is configured. Further, the XYθ table 102 is driven by the table control circuit 114 under the control of the control computer 110. It can be moved by a drive system such as a three-axis (XY-θ) motor that drives in the X, Y, and θ directions. As these X motors, Y motors, and θ motors, for example, linear motors can be used. The XYθ table 102 can be moved in the horizontal direction and the rotational direction by the motor of each axis of the XYθ. Then, the focal position (optical axis direction: Z-axis direction) of the XYθ table 102 is dynamically adjusted to the pattern forming surface of the mask substrate 101 by the AF control circuit 140 under the control of the control computer 110. For example, the focal position of the XYθ table 102 is adjusted by being moved in the optical axis direction (Z-axis direction) by the piezo element 142. The moving position of the mask substrate 101 arranged on the XYθ table 102 is measured by the laser length measuring system 122 and supplied to the position circuit 107.

Design pattern data (drawing data) that is the basis for pattern formation of the mask substrate 101 is input from the outside of the inspection device 100 and stored in the magnetic disk device 109.

Here, FIG. 1 describes the components necessary for explaining the first embodiment. It goes without saying that the inspection apparatus 100 may usually include other necessary configurations.

FIG. 2 is a conceptual diagram for comparing the numerical aperture of the inspection device and the numerical aperture of the exposure device according to the first embodiment. FIG. 2A shows a part of the optical system of the exposure apparatus. In the exposure apparatus, illumination light (not shown) is illuminated on the mask substrate 300, the transmitted light 301 from the mask substrate 300 is incident on the objective lens 302, and the light 305 passing through the objective lens 302 is connected to the semiconductor substrate 304 (wafer). To image. Although one objective lens 302 is shown in FIG. 2A, it goes without saying that a combination of a plurality of lenses may be used. Here, in the current exposure apparatus, the pattern formed on the mask substrate 300 is reduced to, for example, 1/4 and exposed and transferred to the semiconductor substrate 304. The numerical aperture NAi (numerical aperture on the image i side) of the exposure apparatus with respect to the semiconductor substrate 304 at that time is set to, for example, NAi = 1.2. In other words, the numerical aperture NAi of the objective lens 302 that can pass through the objective lens 302 is set to, for example, NAi = 1.2. In the exposure apparatus, the transmitted light image from the mask substrate 300 is reduced to 1/4, so that the sensitivity of the objective lens 302 to the mask substrate 300 is 1/4. In other words, the numerical aperture NAo (numerical aperture on the object o side) of the objective lens 302 that can be incident when the transmitted light is incident from the mask substrate 300 to the objective lens 302 is 1/4 of NAi, and NAo = 0.3. It becomes. Therefore, in the exposure apparatus, the transmitted light image from the mask substrate 300 having a luminous flux with a numerical aperture of NAo = 0.3 is exposed and transferred to the semiconductor substrate 304.

On the other hand, in the inspection device 100 according to the first embodiment, as shown in FIG. 2B, illumination light (not shown) is illuminated on the mask substrate 101, and the transmitted light 190 from the mask substrate 101 is the objective lens 104. The light 192 incident on the photodiode array 105 (sensor) is imaged with the light 192 that has passed through the objective lens 104. At that time, the numerical aperture NAo (numerical aperture on the object o side) of the objective lens 104 that can be incident when the transmitted light 190 is incident from the mask substrate 101 to the objective lens 104 is set to, for example, NAo = 0.9. Therefore, in the inspection device 100, the photodiode array 105 receives a transmitted light image from the mask substrate 101 having a luminous flux with a numerical aperture of NAo = 0.9. Therefore, in the inspection device 100, defect inspection is performed using light having a higher resolution than the light that is exposed and transferred to the semiconductor substrate 304 by the exposure device. By using such high-resolution light, it is possible to perform high-precision inspection capable of detecting minute defects on the order of nanometers, for example. Although only the objective lens 104 is shown in FIG. 2B, a plurality of lenses (not shown) are arranged between the objective lens 104 and the photodiode array 105.

However, as described above, in the inspection device 100, since the photodiode array 105 receives light having a higher resolution than the light that is exposed and transferred to the semiconductor substrate 304 by the exposure device, the pattern is exposed and transferred to the semiconductor substrate 304 by the exposure device. In this case, it is difficult for the inspection device 100 to reproduce the same image as the pattern image formed on the semiconductor substrate 304. Therefore, in the first embodiment, the numerical aperture 180 is arranged to narrow the light beam of the transmitted light from the mask substrate 101 to adjust the light beam diameter of the transmitted light, and the same as the light to be exposed and transferred to the semiconductor substrate 304 by the exposure apparatus. The inspection device 100 can generate light having a numerical numerical aperture NAo (numerical aperture on the object o side) (similar resolution).

In addition, in the exposure apparatus, the transmitted light from the mask substrate 300 is transferred to the semiconductor substrate 304, but the reflected light is not used. Therefore, the polarizing element is not arranged on the optical path from the mask substrate 300 to the semiconductor substrate 304. On the other hand, in the inspection device 100, in addition to the inspection using the transmitted light from the mask substrate 101, the inspection using the reflected light from the mask substrate 101 can also be performed. Therefore, in addition to the inspection light for the transmitted light inspection (first inspection light or the first illumination light), the inspection light for the reflected light inspection (second inspection light or the second illumination). Light) is needed. Therefore, in the inspection device 100, a polarization beam splitter 174 (polarizing element) is arranged on the optical path for illuminating the mask substrate 300 with the inspection light for the reflected light inspection. Therefore, the photodiode array 105 receives the light that has passed through the polarizing beam splitter 174 (polarizing element). As a result, the light received by the photodiode array 105 in the inspection device 100 has a different polarization component from the light formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 in the exposure apparatus. In this respect as well, it is difficult for the inspection device 100 to reproduce the same image as the pattern image formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 by the exposure apparatus. Therefore, in the first embodiment, the polarization beam splitter 174 (polarizing element) is configured to be able to be excluded from the optical path.

In addition, in the inspection device 100 and the exposure device, dipole, 4-pole lighting, ring-shaped lighting, and other arbitrary aperture conditions optimized by computer simulation, that is, various so as to improve the resolution characteristics of the pattern of the shape of the illumination light. The shape is shaped and exposed. Further, the polarization condition of the illumination light is set to direct polarization, azimuth polarization, radial polarization, no polarization, or the like for exposure. As a result, the light received by the photodiode array 105 in the inspection device 100 has a different illumination shape and polarization conditions from the light formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 by the exposure apparatus. Become. In this respect as well, it is difficult for the inspection device 100 to reproduce the same image as the pattern image formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 by the exposure apparatus. Therefore, in the first embodiment, the illumination shape and the polarization condition can be changed.

A point of generating light having the same numerical aperture NAo (numerical aperture on the object o side) (similar resolution) as the light to be exposed and transferred to the semiconductor substrate 304 by the above exposure apparatus, the polarizing beam splitter 174 (polarizing element) is placed on the optical path. By implementing the point of excluding from the light and the point of changing the illumination shape and the polarization condition to the same shape as the exposure device by the inspection device 100, the inspection device 100 causes the pattern of the mask substrate to be formed on the semiconductor substrate or the like by the exposure device. It is possible to reproduce an image that is substantially the same as or close to the image that is exposed and transferred. Therefore, in the first embodiment, the image is substantially the same as the pattern image formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 by the high NA inspection mode in which the normal defect inspection is performed and the exposure apparatus. It is configured so that it can be switched between the low NA inspection mode that reproduces and inspects a close image.

FIG. 3 is a flowchart showing a part of the main steps of the inspection method according to the first embodiment.
FIG. 4 is a flowchart showing the rest of the main steps of the inspection method according to the first embodiment. In FIGS. 3 and 4, the inspection methods according to the first embodiment are a high NA inspection mode setting step (S102), a high NA scanning (scan) step (S104), and an autofocus (AF) data creation step (S105). The reference image (1) creation step (S106), the comparison step (S108), the defect region identification step (S110), the low NA inspection mode switching step (S112), and the low NA scanning (scanning) step (S120). ), The AF control step (S122), the reference image (2) creation step (S124), the measurement image edge pair detection step (S130) of the normal portion, and the reference image edge pair detection step (S132) of the normal portion. The image threshold level measurement step (S134), the measurement image edge pair detection step (S136) of the defect portion, the reference image edge pair detection step (S138) of the defect portion, the reference image pattern dimension measurement step (S140) of the defect portion, A series of steps of a defect measurement image pattern dimension measurement step (S142) and a comparison step (S144) are carried out.

Further, in the low NA inspection mode switching step (S112), as internal steps, a polarizing element evacuation step (S114), an NA squeezing step (S116), and an illumination shape changing step (S118) are carried out.

As the high NA inspection mode setting step (S102), the inspection mode switching control circuit 144 sets each configuration to the high NA inspection mode for performing a normal high NA defect inspection. Specifically, it operates as follows. Upon receiving the control signal from the inspection mode switching control circuit 144, the illumination shape switching mechanism 171 sets the shape of the illumination light (inspection light) to the illumination shape for normal inspection. Further, in the inspection mode switching control circuit 144, the numerical aperture NAo (numerical aperture on the object o side) of the objective lens 104 that can be incident when the transmitted light 190 is incident from the mask substrate 101 to the objective lens 104 is, for example, NAo = 0. The squeeze value of the numerical aperture 180 is controlled so as to be 9. Alternatively, the aperture stop 180 may be fully opened. Further, the inspection mode switching control circuit 144 controls the drive mechanism 176 to arrange the polarization beam splitter 174 on the optical path. If the polarizing beam splitter 174 is located outside the optical path, it is moved onto the optical path. If the polarizing beam splitter 174 is on the optical path, its position may be maintained.

As a high NA scanning step (S104), the numerical aperture (NA) that can be incident from the mask substrate 101 to the objective lens 104 that magnifies the transmitted light or the reflected light from the mask substrate 101 on which the pattern is formed is high. A pattern formed on the mask substrate 101 by using the photodiode array 105 by scanning the mask substrate 101 in a state where the photodiode array 105 (sensor) can receive the transmitted light or the reflected light in the NA state. The transmitted light image or the reflected light image is imaged. In the first embodiment, a low NA scan is performed as described later. Here, before performing the low NA scan, the optical image acquisition mechanism 150 has a numerical aperture (NA) larger than the numerical aperture (NA) when the light beam used for the low NA optical image described later is incident on the objective lens 104 from the mask substrate 101. An optical image (high NA optical image: second optical image) of the mask substrate 101 is acquired using the light beam in a state of a high numerical aperture. Specifically, it operates as follows.

FIG. 5 is a diagram showing an example of the configuration of the optical system in the high NA inspection mode in the first embodiment. In FIG. 5, laser light (for example, DUV light) having a wavelength below the ultraviolet region, which is the inspection light, is generated from the light source 103. The generated light is polarized through the polarizing plate 60, and when the transmitted light inspection is performed, the light for the transmission inspection illumination (one of the p wave and the s wave) is passed. Then, the light for transmission inspection illumination (one of the p wave and the s wave) is reflected by the polarization beam splitter 61. On the other hand, when the reflected light inspection is performed, the light for the reflection inspection illumination (the other of the p wave and the s wave) is passed. Then, the light for reflection inspection illumination (the other of the p wave and the s wave) passes through the polarizing beam splitter 61. For example, by rotating the polarizing plate 60, the light that passes through can be restricted. This allows you to choose between transmitted light inspection and reflected light inspection.

When performing the transmitted light inspection, the light for the transmitted inspection illumination (first inspection light) is illuminated on the mask substrate 101 by the transmitted illumination optical system 170. Hereinafter, a specific example will be described based on the example of FIG. In the transmission illumination optical system 170, the light for transmission inspection illumination branched by the polarization beam splitter 61 is reflected by the mirror 63, passes through the lenses 64, 65, 66, and is reflected by the mirror 67. Then, the light reflected by the mirror 67 is imaged on the pattern forming surface of the mask substrate 101 from the back surface side opposite to the pattern forming surface of the mask substrate 101 by the condenser lens 68. The transmitted light transmitted through the mask substrate 101 enters the imaging optical system 178 via the objective lens 104 and the polarizing beam splitter 174. Then, the image is formed on the photodiode array 105 (an example of the sensor) by the imaging optical system 178, and is incident on the photodiode array 105 as an optical image. The light that has passed through the polarizing beam splitter 174 is imaged on the photodiode array 105 by the lens 72 through the lenses 69 and 70 and the aperture diaphragm 180 in the imaging optical system 178. As the photodiode array 105, for example, it is preferable to use a TDI (time delay integration) sensor or the like. The photodiode array 105 (sensor) captures an optical image of a pattern formed on the mask substrate 101 while the XYθ table 102 on which the mask substrate 101 is placed is moving.

On the other hand, when the reflected light inspection is performed, the light for the reflection inspection illumination (second inspection light) is illuminated on the mask substrate 101 by the reflection illumination optical system 172. The reflection illumination optical system 172 has an objective lens 104 and a polarization beam splitter 174 (polarization element), and the mask substrate 101 is illuminated with light for reflection inspection illumination by using the objective lens 104 and the polarization beam splitter 174, and the mask substrate is used. The reflected light from 101 is passed through. Hereinafter, a specific example will be described based on the example of FIG. In the reflection illumination optical system 172, the light for reflection inspection illumination that has passed through the polarization beam splitter 61 passes through the lenses 80, 81, 82, and 83 and is reflected by the polarization beam splitter 174. The light reflected by the polarizing beam splitter 174 is incident on the objective lens 104, and is imaged on the pattern forming surface of the mask substrate 101 from the pattern forming surface side of the mask substrate 101 by the objective lens 104. The reflected light reflected from the mask substrate 101 passes through the objective lens 104 and the polarizing beam splitter 174 and is incident on the imaging optical system 178. Then, the image is formed on the photodiode array 105 (an example of the sensor) by the imaging optical system 178, and is incident on the photodiode array 105 as an optical image. The photodiode array 105 (sensor) captures an optical image of a pattern formed on the mask substrate 101 while the XYθ table 102 on which the mask substrate 101 is placed is moving.

At that time, a part of the reflected light is branched by the mirror 73 arranged between the lenses 69 and 70, and the branched light is imaged on the AF sensor 76 by the AF optical system 173. The light receiving surface of the AF sensor 76 is arranged so as to have a confocal relationship with the light receiving surface of the photodiode array 105 with respect to the beam splitter 73. The output signal from the AF sensor 76 is transmitted to the AF control circuit 140. When the photodiode array 105 receives the reflected light in a state where the numerical aperture is high, the AF control circuit 140 measures the focal position of each position on the mask substrate 101 using the reflected light. Then, the AF control circuit 140 dynamically controls the piezo element 142 to move the XYθ table 102 in the optical axis direction (Z-axis direction) in real time based on the signal to dynamically form the pattern of the mask substrate 101. The focal position (optical axis direction: Z-axis direction) of the objective lens 104 is adjusted on the surface. Even when the transmitted light inspection is performed, a part of the reflected illumination is illuminated on the mask substrate 101 for AF by the reflected illumination optical system 172. Then, the reflected light corresponding to a part of the reflected illumination is reflected by the mirror 73 and imaged on the AF sensor 76. As a result, the focal position (optical axis direction: Z-axis direction) of the objective lens 104 can be dynamically adjusted on the pattern forming surface of the mask substrate 101 even in the inspection of only transmission.

The inspection device 100 can perform one or both of such a scan of transmitted light and a scan of reflected light. The reflected light reflected from the mask substrate 101 includes pattern height information that is difficult to obtain with the transmitted light from the mask substrate 101. Therefore, the AF control circuit 140 performs AF control using the reflected light from the mask substrate 101 instead of the transmitted light from the mask substrate 101. When both the transmitted light scan and the reflected light scan are performed, it does not matter which one is performed first. Alternatively, another photodiode array 105 is arranged, the position of illuminating the mask substrate 101 is shifted between the scan of transmitted light and the scan of reflected light, and the transmitted light is received by one of the photodiode arrays 105, and the reflected light is received. Is reflected by the mirror 73 and then received by the other photodiode array 105, so that both the transmitted light scan and the reflected light scan may be performed at the same time.

FIG. 6 is a conceptual diagram for explaining an inspection area according to the first embodiment. As shown in FIG. 6, the inspection area 10 (the entire inspection area) of the mask substrate 101 is virtually divided into a plurality of strip-shaped inspection stripes 20 having a scan width W, for example, in the y direction. Then, the inspection device 100 acquires an image (stripe region image) for each inspection stripe 20. For each of the inspection stripes 20, a laser beam (DUV light) is used to capture an image of a graphic pattern arranged in the stripe region in the longitudinal direction (x direction) of the stripe region. By moving the XYθ table 102, the mask substrate 101 is moved, and as a result, an optical image is acquired while the photodiode array 105 is relatively continuously moving in the x direction. The photodiode array 105 continuously captures an optical image having a scan width W as shown in FIG. In other words, the photodiode array 105, which is an example of the sensor, captures an optical image (measurement image) of the pattern formed on the mask substrate 101 by using the inspection light while moving relative to the XYθ table 102. In the first embodiment, after the optical image of one inspection stripe 20 is captured, the optical image of the scan width W is similarly moved to the position of the next inspection stripe 20 in the y direction and this time in the opposite direction. Images are taken continuously. That is, the imaging is repeated in the forward (FWD) -back forward (BWD) directions in the opposite directions on the outward and return paths.

Here, the direction of imaging is not limited to the repetition of forward (FWD) -back forward (BWD). You may take an image from one direction. For example, FWD-FWD may be repeated. Alternatively, BWD-BWD may be repeated.

The image of the pattern formed on the photodiode array 105 is photoelectrically converted by each light receiving element of the photodiode array 105, and further A / D (analog-digital) converted by the sensor circuit 106. Then, the pixel data of the inspection stripe 20 to be measured is stored in the stripe pattern memory 123. When capturing such pixel data (striped region image), the dynamic range of the photodiode array 105 uses, for example, a dynamic range in which the maximum gradation is when the amount of illumination light is 60% incident. Further, when acquiring the optical image of the inspection stripe 20, the laser length measuring system 122 measures the position of the XYθ table 102. The measured position information is output to the position circuit 107. The position circuit 107 (calculation unit) calculates the position of the mask substrate 101 using the measured position information.

After that, the stripe region image is sent to the comparison circuit 108 together with the data indicating the position of the photomask 101 on the XYθ table 102 output from the position circuit 107. The measurement data (pixel data) is, for example, 8-bit unsigned data, and represents the gradation (light amount) of the brightness of each pixel. The stripe area image output in the comparison circuit 108 is stored in a storage device (not shown).

In the comparison circuit 108, stripes of a predetermined size (for example, the same width as the scan width W) in the x direction so as to cut out the frame image of the target frame area 30 from the stripe area image (optical image) of the inspection stripe 20. The area image is divided. For example, it is divided into a frame image of 512 × 512 pixels. In other words, the stripe area image for each inspection stripe 20 is divided into a plurality of frame images (optical images) with a width similar to the width of the inspection stripe 20, for example, a scan width W. By such processing, a plurality of frame images (optical images) corresponding to the plurality of frame areas 30 are acquired. As described above, for the plurality of frame images, one image (measurement image: optical image) data to be compared for inspection is generated.

As an autofocus (AF) data creation step (S105), the AF control circuit 140 focuses on the height position of the mask substrate 101 surface used for AF control at the same time as scanning with the reflected light or transmitted light described above. The position information is stored in the magnetic disk device 109. Therefore, in the magnetic disk device 109 (storage device), the reflected light from the mask substrate 101 illuminated with the light through the reflected illumination optical system 172 in a state where the polarization beam splitter 174 (polarizing element) is moved on the optical path. The information on the focal position of each position on the mask substrate 101 measured using the above is stored.

When only scanning the transmitted light is performed in the high NA scanning step (S104) described above, the focal position information depending on the height position of the mask substrate 101 surface may be acquired in advance using the reflected light. .. When scanning the transmitted light in the high NA scanning step (S104) described above, AF control may be performed using the focal position information obtained at that time.

Reference image (1) As a creation step (S106), the reference image creation unit creates a reference image (second reference image) corresponding to the high NA optical image (second optical image). First, the development circuit 111 (a part of the reference image creation unit) develops an image based on the design pattern data that is the basis of pattern formation of the mask substrate 101 to create a design image. Specifically, the design data is read from the magnetic disk device 109 through the control computer 110, and each graphic pattern in the area of the target frame 30 defined in the read design data is converted into binary or multi-valued image data (). Image development) to create a design image.

Here, the figure defined in the design pattern data is, for example, a basic figure of a rectangle or a triangle, for example, the coordinates (x, y) at the reference position of the figure, the length of the side, the rectangle, the triangle, or the like. Graphical data (vector data) that defines the shape, size, position, etc. of each pattern graphic is stored with information such as a graphic code that serves as an identifier for distinguishing graphic types.

When the information of the design pattern that becomes the graphic data is input to the expansion circuit 111, the data is expanded to the data for each graphic, and the graphic code, the graphic dimension, etc. indicating the graphic shape of the graphic data are interpreted. Then, binary or multi-valued design image data is developed and output as a pattern arranged in a grid having a grid of predetermined quantization dimensions as a unit. In other words, the design data is read, the inspection area is virtually divided into squares with a predetermined dimension as a unit, the occupancy rate of the figure in the design pattern is calculated for each square, and the n-bit occupancy rate data is obtained. Output. For example, it is preferable to set one square as one pixel. Then, when to have a resolution of 1/2 8 ^{(= 1/256)} to 1 pixel, the occupancy rate of the pixel allocated the small area region amount corresponding 1/256 of figures are arranged in a pixel Calculate. Then, a design image of 8-bit occupancy rate data is created for each pixel. The design image data is output to the reference circuit 112.

The reference circuit 112 (another part of the reference image creating unit) filters the design image to create a reference image (for high NA).

FIG. 7 is a diagram for explaining the filtering process in the first embodiment. The measurement data as an optical image obtained from the sensor circuit 106 is in a state in which the filter acts due to the resolution characteristics of the objective lens 104, the aperture effect of the photodiode array 105, and the like, in other words, in an analog state in which the image changes continuously. By filtering the reference design image data whose intensity (shade value) is the image data on the design side of the digital value, it can be adjusted to the measurement data. In this way, a reference image to be compared with the frame image (optical image) is created. The created reference image (for high NA) is output to the comparison circuit 108, and the reference image output in the comparison circuit 108 is stored in a storage device (not shown). As a result, the other image (reference image) data to be compared for inspection is generated.

As a comparison step (S108), the comparison circuit 108 inspects the defect of the pattern by using the optical image of the transmitted light image or the reflected light image of the imaged pattern of the mask substrate 101. The comparison circuit 101 (comparison unit) compares a frame image (optical image: second optical image) and a reference image (for high NA: second reference image) for each pixel. Specifically, first, the frame image (optical image) to be compared and the reference image to be compared are aligned by a predetermined algorithm. For example, alignment is performed using the method of least squares. Then, the comparison circuit 108 compares the two for each pixel according to a predetermined determination condition, and determines the presence or absence of a defect such as a shape defect. As a determination condition, for example, the presence or absence of a defect is determined by comparing the two for each pixel according to a predetermined algorithm. For example, it is determined whether or not the difference between the pixel values of both images is larger than the determination threshold value (third threshold value). Then, a pixel whose gradation difference between the pixel in the frame image (optical image) and the corresponding pixel in the reference image (for high NA) exceeds the determination threshold value (third threshold value) is determined as a defective pixel. Then, the comparison result is output. The comparison result may be output from the magnetic disk device 109, the magnetic tape device 115, the flexible disk device (FD) 116, the CRT 117, the pattern monitor 118, or the printer 119.

By the above high NA scan inspection (normal inspection), it is possible to identify the pixel at the defect position where the shape defect occurs. Next, the pattern size of the defect portion is measured in order to confirm whether the dimensional error caused by such a defect can withstand the actual use.

FIG. 8 is a block diagram showing an internal configuration of the low NA inspection circuit according to the first embodiment. In FIG. 8, in the low NA inspection circuit 146, storage devices 50, 52, 59 such as a magnetic disk device, an area identification unit 51, an alignment unit 56, a measurement image edge pair detection unit 56, a reference image edge pair detection unit 58, The threshold level measurement unit 40, the measurement image pattern dimension measurement unit 42, the reference image pattern dimension measurement unit 44, and the comparison unit 46 are arranged. Here, what is described as "~ part" has a processing circuit, and as the processing circuit, an electric circuit, a computer, a processor, a circuit board, a quantum circuit, a semiconductor device, or the like can be used. Further, a common processing circuit (same processing circuit) may be used for each “~ part”. Alternatively, different processing circuits (separate processing circuits) may be used.

As a defect region identification step (S110), the region identification unit 51 determines that the frame image (optical image: second optical image) and the reference image (for high NA: second reference image) are defective as a result of comparison. Using the defective pixels, the region of the pattern abnormality portion in the frame image (optical image: second optical image) is specified. Specifically, inspection result information (for example, coordinate information of defect positions) is input from the comparison circuit 108, and a region in which a defect is detected (hereinafter, defect region) is specified based on the inspection result information. As the defect region, for example, the frame region 30 including the defect position can be mentioned. However, the region size is not limited to this, and various area sizes can be selected. The information of the identified defective region is stored in the storage device 59.

As the low NA inspection mode switching step (S112), the inspection mode switching control circuit 144 switches the setting of each configuration from the high NA inspection mode described above to the low NA inspection mode for performing low NA defect inspection. As a result, the photodiode array 105 can receive an image substantially similar to or close to the pattern image formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 by the exposure apparatus. Therefore, the following internal steps are carried out.

As the polarizing element retracting step (S114), the inspection mode switching control circuit 144 controls the drive mechanism 176 to move the polarizing beam splitter 174 from the optical path to the outside of the optical path. In this way, the drive mechanism 176 can move the polarizing element from the outside of the optical path to the outside of the optical path, and can move the polarizing element from the top of the optical path to the outside of the optical path. Further, since the reflected illumination light is not used due to the retracting of the polarization beam splitter 174, the inspection mode switching control circuit 144 passes the polarizing plate 60 so that only the light for transmission inspection illumination (one of the p wave and the s wave) passes through. Rotate.

As the NA squeezing step (S116), the inspection mode switching control circuit 144 switches the numerical aperture NA from the state of the high numerical aperture to the state of the low numerical aperture. Specifically, the numerical aperture NAo (numerical aperture on the object o side) of the objective lens 104 that can be incident when the transmitted light 190 is incident from the mask substrate 101 to the objective lens 104 has the same characteristics as the exposure device. Control. For example, the aperture value of the aperture stop 180 is controlled so that NAo = 0.3. In the first embodiment, since the numerical aperture is changed from the high numerical aperture to the low numerical aperture, the aperture of the aperture diaphragm 180 is narrowed down. In this way, the aperture diaphragm 180 allows the numerical aperture (NA) at which transmitted light can be incident from the mask substrate 101 to the objective lens 104 to be switched between a high numerical aperture state and a low numerical aperture state. Adjust the light beam diameter of the transmitted light and / and the reflected light. In the example of FIG. 5, the aperture diaphragm 180 is arranged between the lenses 70 and 72 in the imaging optical system 178, but is not limited to this. It suffices that the luminous flux of the light incident on the objective lens 104 or the light incident on the photodiode array 105 through the objective lens 104 can be narrowed down and the luminous flux diameter can be adjusted. Therefore, the aperture diaphragm 180 may be arranged between the mask substrate 101 and the photodiode array 105. More preferably, the aperture diaphragm 180 is arranged at the pupil position.

As the illumination shape changing step (S118), the control signal from the inspection mode switching control circuit 144 is received, and the illumination shape switching mechanism 171 uses the shape of the illumination light (inspection light) in the same manner as the illumination shape used in the exposure apparatus. Change to the lighting shape. The illumination shape switching mechanism 171 includes a mechanism capable of setting a plurality of aperture diaphragms having an illumination shape manufactured in advance by depositing Cr on a glass plate, and a mechanism for mechanically switching them. Further, the polarization condition of the illumination light is set so that the polarization condition of the illumination light becomes the polarization condition of the exposure apparatus.

In the low NA scanning step (S120), the photodiode array 105 can receive light corresponding to the state where the numerical aperture is low, and the light is between the mask substrate 101 and the photodiode array 105. Transmitted light of a pattern formed on the mask substrate 101 using the photodiode array 105 by scanning the specified defect region of the mask substrate 101 with a predetermined illumination light in the absence of a polarizing element on the optical path. Take an image. Specifically, it operates as follows.

FIG. 9 is a diagram showing an example of the configuration of the optical system in the low NA inspection mode according to the first embodiment. In FIG. 9, the light generated from the light source 103 is polarized through the polarizing plate 60, and only the light for transmission inspection illumination (one of the p wave and the s wave) is passed. Then, the light for transmission inspection illumination (one of the p wave and the s wave) is reflected by the polarization beam splitter 61. The light for transmission inspection illumination (first inspection light) is illuminated on the mask substrate 101 by the transmission illumination optical system 170. At that time, the illumination shape switching mechanism 171 changes the configuration of the transmission illumination optical system 170 to change the illumination shape of the light for transmission inspection illumination (first inspection light). As described above, the transmission illumination optical system 170 is configured so that the illumination shape of the light for transmission inspection illumination (first inspection light) can be changed. In the example of FIG. 9, the light for transmission inspection illumination reflected by the polarizing beam splitter 61 reaches the lens 65, and then the illumination shape is changed between the lenses 65 and 66. The light for transmission inspection illumination whose illumination shape has been changed is reflected by the mirror 67. Then, the light reflected by the mirror 67 is imaged on the pattern forming surface of the mask substrate 101 from the back surface side opposite to the pattern forming surface of the mask substrate 101 by the condenser lens 68. As a result, the mask substrate 101 can be irradiated with light having an illumination shape similar to that used in the exposure apparatus. The transmitted light transmitted through the mask substrate 101 enters the imaging optical system 178 via the objective lens 104. Then, the image is formed on the photodiode array 105 (an example of the sensor) by the imaging optical system 178, and is incident on the photodiode array 105 as an optical image. At that time, since the numerical aperture is limited by the aperture diaphragm 180, an optical image of low NA light is incident on the photodiode array 105. In the example of FIG. 9, the light that has passed through the objective lens 104 passes through the lenses 69 and 70 and the aperture diaphragm 180 in the imaging optical system 178, and is imaged on the photodiode array 105 by the lens 72. As described above, the photodiode array 105 receives the transmitted light from the mask substrate 101 illuminated with the light for transmission inspection illumination (first inspection light) while the XYθ table 102 is moving.

On the other hand, as shown in FIG. 9, since the light for reflection inspection illumination is blocked by the polarizing plate 60, the light for reflection inspection illumination does not enter the reflection illumination optical system 172. Even if the light for reflection inspection illumination is incident on the reflection illumination optical system 172, the polarization beam splitter 174 (polarizing element) is retracted from the optical path in the reflection illumination optical system 172, so that the light is reflected by the mask substrate 101. It does not illuminate the light for inspection lighting.

As described above, the photodiode array 105 receives an image substantially similar to or close to the pattern image formed on the semiconductor substrate 304 when the pattern is exposed and transferred to the semiconductor substrate 304 by the photodiode array 105. it can.

The image of the pattern formed on the photodiode array 105 is photoelectrically converted by each light receiving element of the photodiode array 105, and further A / D (analog-digital) converted by the sensor circuit 106. Then, the pixel data of the defect area to be measured is stored in the stripe pattern memory 123. When capturing such pixel data (defect region image), the dynamic range of the photodiode array 105 uses, for example, a dynamic range whose maximum gradation is when the amount of illumination light is 60% incident. Further, when acquiring the optical image of the defect region, the laser length measuring system 122 measures the position of the XYθ table 102. The measured position information is output to the position circuit 107. The position circuit 107 (calculation unit) calculates the position of the mask substrate 101 using the measured position information.

After that, the defect region image is sent to the low NA inspection circuit 146 together with the data indicating the position of the photomask 101 on the XYθ table 102 output from the position circuit 107. The measurement data (defect area image: pixel data) is, for example, 8-bit unsigned data, and represents the gradation (light amount) of the brightness of each pixel. The defect region image output in the low NA inspection circuit 146 is stored in the storage device 52. As a result, one image (measured image) data to be compared for the low NA test is generated.

Here, in the low NA inspection mode, as described above, since the polarization beam splitter 174 is moved from the optical path to the outside of the optical path, the light passing through the reflection illumination optical system 172 is not illuminated on the mask substrate 101. Therefore, the reflected light from the mask substrate 101 cannot be obtained. Therefore, the reflected light from the mask substrate 101 cannot be used for AF control. Therefore, it is configured as follows.

As an AF control step (S122), in the AF control circuit 140 (a part of the focusing mechanism), the photodiode array 105 is low while the XYθ table 102 is moving while the polarization beam splitter 174 is moved out of the optical path. While receiving the transmitted light corresponding to the numerical aperture, the transmitted light is dynamically focused by using the focal position information stored in the magnetic disk device 109. Specifically, the AF control circuit 140 controls the piezo element 142 (a part of the focusing mechanism) based on the already measured focal position information stored in the magnetic disk device 109, and the XYθ table 102. Is moved in the optical axis direction (Z-axis direction) in real time to follow the progress of the XYθ table 102 and dynamically move the focal position of the objective lens 104 on the pattern forming surface of the mask substrate 101 (optical axis direction: Z-axis). Direction) is adjusted.

In the reference image (2) creation step (S124), the reference circuit 112 (an example of the reference image creation unit) filters the design image to match the defect region image (low NA optical image) and refers to the reference image (low NA). For). Here, it is necessary to use a filter calculation that reproduces the low NA illumination shape, polarization conditions, and imaging optical conditions, which are different from the filter calculation that matches the optical image obtained by the high NA scan (scan in the normal inspection mode). This is done using an optical filter calculation based on partial coherent imaging theory. The created reference image (for low NA) is sent to the low NA inspection circuit 146. The reference image (2) output in the low NA inspection circuit 146 is stored in the storage device 50. Of the reference images (for high NA) of each frame region 30 created in the reference image (1) creation step (S106), the reference image (for high NA) corresponding to the defect region image (low NA optical image). May be diverted as a simple reference image (for low NA) by smoothing with a filter process.

FIG. 10 is a diagram showing an example of an image gradation value profile of a pattern in an optical image and a pattern in a reference image according to the first embodiment. The pattern dimension (CD) of the pattern formed on the mask substrate 101 is created from the gradation value profile of the pattern in the optical image obtained by capturing the pattern forming region of the mask substrate 101 and the design data corresponding to the pattern. It is obtained by comparing with the gradation value profile of the pattern in the reference image. In FIG. 10, in the gradation value profile of the pattern in the reference image, the floor of the pattern in the optical image _{has a detection threshold Th 0 (first threshold level) at which the dimension between EFs in FIG. 10 becomes the desired design width dimension.} By measuring the inter-AB dimension (for example, the width of the space pattern) from the threshold value profile, it should have been possible to obtain the width dimension CD of the pattern originally formed on the mask substrate 101.

Here, in the optical image obtained by the high NA scan (scan in the normal inspection mode), if it is a normal part that is not a defect position, the gradation value profile of the pattern in the optical image and the pattern in the reference image The coefficients of the filter calculation are set so that the gradation value profiles substantially match. Therefore, the EF dimension of the pattern in the reference image and the AB dimension of the pattern in the optical image match. However, in the optical image obtained by the low NA scan, since the reference image is generated by the filter calculation based on the optical theory, the shape of the gradation value profile of the pattern in the optical image and the reference image (for low NA) The shape of the gradation value profile of the pattern may not match as shown in FIG. Therefore, the EF dimension of the pattern in the reference image (for low NA) and the AB dimension of the pattern in the optical image do not match. In addition to the reference image (for high NA) that has been smoothed and filtered to match the optical image obtained by high NA scan (scanning in normal inspection mode), the optical image obtained by low NA scan is separately used. It is difficult to match the gradation value profile even with the reference image (for low NA) created so as to match. The reason for this is that even if you try to create a reference image to match the optical image obtained by low NA scan, the accuracy of the filtering process is not yet sufficient, and as a result, the gradation value profiles can be matched. It has become difficult. Specifically, due to the complexity of the illumination conditions and other conditions to get closer to the exposure image, a certain difference, that is, an error, remains in the reference image generated by the filtering process on the optical image obtained by the low NA scan. become. Therefore, as it is, it is not possible to accurately obtain the pattern dimensions in the optical image obtained by the low NA scan. Therefore, in the first embodiment, the detection threshold Th ₀ (threshold level) at which the desired design width dimension is obtained is changed in the gradation value profile of the pattern in the reference image. Specifically, using the pattern of the normal part that is not the defect position, from the gradation value profile of the pattern in the optical image obtained by the low NA scan at _{the original detection threshold Th 0 (first threshold level).} The detection threshold Th'(second threshold level) of the gradation value profile of the pattern in the reference image (for low NA) corresponding to the obtained inter-AB dimension is measured. Then, the dimension between E'F'is measured from the gradation value profile of the pattern in the reference image at the detection threshold Th'. On the other hand, at the original detection threshold Th ₀ , the dimension between AB is measured from the gradation value profile of the pattern in the optical image obtained by the low NA scan. Since the dimension between E'F'should correspond to the desired design width dimension, the width dimension L of the pattern formed on the mask substrate 101 is L = (dimension between E'F') / (dimension between AB) × ( It can be approximated by the design width dimension).

As a measurement image edge pair detection step (S130) of a normal portion, first, the alignment portion 56 relates to a defect region image (low NA optical image) of the size of the frame region 30 to be compared and a reference image to be compared. , Align with a predetermined algorithm. For example, alignment is performed using the method of least squares.

Next, the measurement image edge pair detection unit 56 detects the edge pair for the pattern in the normal portion of the defect region image (low NA optical image) that is not the defect position stored in the storage device 59.

FIG. 11 is a diagram showing an example of a defect region image and an edge pair according to the first embodiment. In the example of FIG. 11, in the defect region image (low NA optical image), line patterns 12a, 12b, 12c, 12d extending in the y direction and line patterns 13 extending in the x direction, which are arranged in the x direction, are shown. Has been done. A space pattern 14a is formed between the line patterns 12a and 12b. A space pattern 14b is formed between the line patterns 12b and 12c. A space pattern 15 is formed between the line patterns 12c, 12c, and 13. The defect region image (low NA optical image) is composed of an image of the size of the frame region 30 including the pixel 36 determined to be the defect portion (part A) as shown in FIG. Therefore, in the defect region image (low NA optical image), there is also a pattern of a normal portion that is not a defect portion (part A). The measurement image edge pair detection unit 56 recognizes the position of the edge portion (outer periphery) of the pattern for the pattern of the normal portion in the defect region image (low NA optical image), and constitutes the width dimension (CD) of the pattern. Detects an edge pair consisting of edges at both ends (around both outer edges). In the example of FIG. 11, the case where the edges a and b at both ends (around both outer edges) constituting the CD in the space portion (white part) between the adjacent patterns 12a and 12b are detected is shown. The number of edge pairs to be detected in the pattern of the normal portion may be one, but it is preferable to detect a plurality of pairs of edge pairs. In such a case, for example, it is detected in pixel units. For example, for the frame region of the x-direction line-and-space pattern, a pair is detected for each pixel on the outer periphery of the line pattern whose outer periphery extends in the y-direction. For the frame area of the y-direction line-and-space pattern, pairs are detected for each pixel on the outer periphery of the line pattern whose outer periphery extends in the x-direction. For the frame areas of the plurality of rectangular patterns, pairs are detected for each pixel on the outer periphery extending in the y direction of each rectangular pattern. For the frame area of the plurality of rectangular patterns, a pair is detected for each pixel on the outer periphery extending in the x direction of each rectangular pattern.

FIG. 12 is a diagram for explaining a method of edge recognition of a pattern in the first embodiment. In the example of FIG. 12, a part of the defect region image (low NA optical image) shown in FIG. 11 is shown. For example, at the edge portion of the line pattern 12c extending in the y direction in the line width direction, the gradation value profile in the x direction changes sharply, whereas the gradation value profile in the y direction maintains a constant value. Become. Similarly, at the edge portion of the line pattern 12d extending in the y direction in the extending direction, the gradation value profile in the y direction changes sharply, whereas the gradation value profile in the x direction maintains a constant value. Become. As described above, there is an edge portion in the direction of maximum inclination at the position where the inclination of the gradation value profile is maximum. Such a position may be recognized as an edge portion.

In the example of FIG. 11, the CD of the space portion (white portion) between the adjacent patterns 12a and 12b is shown, but the present invention is not limited to this. It is also preferable that an edge pair is detected for a black CD in which a pattern is present.

As the reference image edge pair detection step (S132) of the normal portion, the reference image edge pair detection unit 58 uses the reference image (for low NA) corresponding to the defect region image (low NA optical image) and uses the defect region image (low NA). The edge pair in the reference image corresponding to the edge pair detected from the optical image) is detected.

As a reference image threshold level measurement step (S134), the threshold level measurement unit 40 uses the detection threshold Th ₀ (first threshold level) of the image gradation value profile to identify a defect region image (low NA optical image). The pattern dimension between the edge pairs at the normal part (for example, the dimension between AB) and the pattern dimension between the edge pairs at the corresponding part in the reference image (for low NA) are the same dimension (dimension between E'F'). The detection threshold Th'(second threshold level) used for the reference image (for low NA) is measured. When a plurality of edge pair pairs are detected, the statistical value (for example, mean value, median value, maximum value, or minimum value) of the detection threshold value Th'(second threshold value level) obtained in each pair is used. It is preferable to calculate. For example, if the average value or the median value is used, the measurement error of the detection threshold value Th'(second threshold value level) can be averaged. Since it is assumed that the optimum value of the detection threshold value Th'(second threshold value level) differs depending on the direction, it is preferable to obtain the detection threshold value for each dimensional direction. For example, it is preferable to obtain the detection threshold Th'for the x-direction dimension and the detection threshold Th'for the y-direction dimension.

As the measurement image edge pair detection step (S136) of the defect portion, the measurement image edge pair detection unit 56 detects the edge pair of the pattern in the region of the pattern abnormality portion in the defect region image (low NA optical image). Even if defective pixels can be identified by high NA inspection (normal inspection), the maximum value of the pattern dimension at the pattern abnormality location cannot be immediately determined. Therefore, in the first embodiment, not only the pattern size in which the edge is located in the defective pixel but also the defective pixel 36 and the plurality of areas around the defective pixel 36 are used as the area of the pattern abnormality portion in the defective region image (low NA optical image). The pattern size of the pixel group 38 composed of the pixels of the above is measured. For example, the pattern size in the 3 × 3 pixel group 38 centered on the defective pixel 36 is measured. Therefore, the measurement image edge pair detection unit 56 detects a plurality of edge pairs in the same direction of the pattern in the defect region image (low NA optical image). Specifically, a plurality of edge pairs including at least one edge in the pixel group 38 are detected. In the example of FIG. 11, three sets of edge pairs in the same direction in the pixel group 38 in the defect region image (low NA optical image) are shown. If the dimensional directions are different, the dimensions in the directions different from the dimensions of the defective portion to be obtained will be included. Therefore, a plurality of edge pairs in the same direction in the pixel group 38 are detected.

As a reference image edge pair detection step (S138) of the defect portion, the reference image edge pair detection unit 58 uses the reference image (reference image) corresponding to the edge pair of the pattern detected in the region of the pattern abnormality portion in the defect region image (low NA optical image). Detects edge pairs of patterns in (for low NA). Specifically, the reference image edge pair detection unit 58 detects a plurality of edge pairs corresponding to a plurality of edge pairs in the same direction including at least one edge in the pixel group 38 in the defect region image (low NA optical image). To do. In the example of FIG. 11, three sets of edge pairs in the same direction in the pixel group 38 in the reference image (for low NA) are shown.

As the reference image pattern dimension measurement step (S140) of the defect portion, the reference image pattern dimension measurement unit 44 (first pattern dimension measurement unit) uses the detection threshold Th'(second threshold level) to image the defect region. The pattern dimension of the corresponding part in the reference image (for low NA) corresponding to the pattern abnormality part in the (low NA optical image) is measured. Specifically, the edge-pair dimension (E'F'-dimension) is measured for each of a plurality of detected edge pairs in the same direction including at least one edge in the pixel group 38. The edge pair dimension (E'F' dimension) corresponds to the design dimension of the pattern. The design dimensions can be obtained from the design data (drawing data) before image development. In this way, when measuring the pattern dimensions of the corresponding parts in the reference image (for low NA) corresponding to the pattern abnormal parts in the defect region image (low NA optical image), a plurality of edge pairs in the pixel group 38 are used. The dimensions between the corresponding edge pairs are measured.

As a measurement image pattern dimension measurement step (S142) of a defect portion, the measurement image pattern dimension measurement unit 42 (second pattern dimension measurement unit) uses the original detection threshold Th ₀ (first threshold level) to obtain a defect. The pattern size of the pattern abnormality part in the region image (low NA optical image) is measured. Specifically, the edge pair dimension (AB dimension) is measured for each of a plurality of detected edge pairs in the same direction including at least one edge in the pixel group 38. Then, the pattern dimension is measured (approximate) by calculating the pattern dimension L (= (dimension between E'F') / (dimension between AB) × design dimension) for each edge pair. In this way, when measuring the pattern size of the pattern abnormality portion in the defect region image (low NA optical image), the size between the plurality of edge pairs in the pixel group 38 is measured.

As a comparison step (S144), the comparison unit 46 compares the pattern dimensions measured from the reference image (for low NA) with the pattern dimensions measured from the defect region image (low NA optical image) for the pattern abnormality portion. .. Specifically, the comparison unit 46 is in the pixel group 38 when comparing the pattern size measured from the defect region image (low NA optical image) with the pattern size measured from the reference image (for low NA). Of the pattern dimensions between the plurality of edge pairs, the pattern dimension between the edge pairs that maximizes the difference between the defect region image (low NA optical image) and the reference image (for low NA) is determined as the pattern dimension of the pattern abnormality portion. Then, the comparison result (judgment result) is output. The comparison result may be output from the magnetic disk device 109, the magnetic tape device 115, the flexible disk device (FD) 116, the CRT 117, the pattern monitor 118, or the printer 119. If the pattern dimension measured in the defect region image (low NA optical image) deviates from the pattern dimension of the reference image by more than several percent, it can be determined as an unusable defect pattern. .. For example, if a deviation of more than 5 to 10% occurs, it can be determined that the defect pattern cannot be used.

As described above, in the first embodiment, the pattern size of the defective portion is evaluated by comparing with the reference image created based on the design data. For example, in a circuit such as a logic circuit in which the same pattern is not repeatedly formed in the mask substrate 101, die-die inspection is difficult. Therefore, the method of the first embodiment comparing with the reference image uses such a logic circuit. It is especially effective when inspecting.

FIG. 13 is a diagram showing an example of a mechanism for changing the illumination shape in the first embodiment. In FIG. 13, the illumination shape switching mechanism 171 moves the polarizing element 90 between the lenses 65 and 66. Thereby, the configuration of the transmission illumination optical system 170 can be changed, and the transmission illumination optical system 170 can change the illumination shape. When the light for transmission inspection illumination passes through the polarizing element 90, the illumination shape of the light for transmission inspection illumination can be changed to polarized illumination in which the polarization direction is manipulated. By replacing the polarizing element 90 with a diffraction element, the illumination shape may be changed to N-pole illumination such as 2-pole illumination or 4-pole illumination, or ring-shaped illumination. In the example of FIG. 13, the polarizing element 90 is arranged between the lenses 65 and 66, but the present invention is not limited to this. It suffices if the light can be polarized before reaching the mask substrate 101. Therefore, the polarizing element 90 may be arranged on any of the optical paths from the polarizing plate 60 to the condenser lens 68.

FIG. 14 is a diagram showing another example of the mechanism for changing the illumination shape in the first embodiment. In FIG. 14, the illumination shape switching mechanism 171 may be moved so as to arrange the diaphragm 91 between the light source 103 and the polarizing plate 60. Thereby, the configuration of the transmission illumination optical system 170 can be changed, and the transmission illumination optical system 170 can change the illumination shape. By narrowing down the aperture 91, the illumination shape can be changed to small sigma illumination. In the example of FIG. 14, the diaphragm 91 is arranged between the light source 103 and the polarizing plate 60, but the present invention is not limited to this. It suffices to squeeze the light before it reaches the mask substrate 101. Therefore, the diaphragm 91 may be arranged on any of the optical paths from the polarizing plate 60 to the mask substrate 101.

FIG. 15 is a diagram showing another example of the mechanism for changing the illumination shape in the first embodiment. In FIG. 15, the illumination shape switching mechanism 171 may be moved so as to arrange the reduction optical system by the lenses 92 and 93 between the light source 103 and the polarizing plate 60. Thereby, the configuration of the transmission illumination optical system 170 can be changed, and the transmission illumination optical system 170 can change the illumination shape. By reducing the light with such a reduction optical system, the illumination shape can be changed to illumination light having a different illumination magnification. In the example of FIG. 11, the reduction optical system by the lenses 92 and 93 is arranged between the light source 103 and the polarizing plate 60, but the present invention is not limited to this. It suffices if the light can be reduced before the light reaches the mask substrate 101. Therefore, the reduction optical system by the lenses 92 and 93 may be arranged on any of the optical paths from the polarizing plate 60 to the condenser lens 68.

The illumination shape switching mechanism 171 may be configured so that one, any two combinations, or all combinations of the mechanisms shown in FIGS. 13, 140, and 15 can be arranged. Alternatively, it may be configured by another mechanism. The necessary optical elements may be arranged according to the specifications of the exposure apparatus.

As described above, according to the first embodiment, it is possible to improve the line width measurement accuracy of the defective portion from the optical image (low NA optical image) captured so as to be close to the pattern image when transferred by the exposure apparatus. it can. Therefore, the pattern size of the defective part can be inspected. As a result, the degree of influence of transferability when the mask pattern is transferred to a semiconductor substrate or the like can be evaluated more accurately.

In the above description, the one described as "~ circuit" has a processing circuit, and as the processing circuit, an electric circuit, a computer, a processor, a circuit board, a quantum circuit, a semiconductor device, or the like can be used. Further, a common processing circuit (same processing circuit) may be used for each “~ circuit”. Alternatively, different processing circuits (separate processing circuits) may be used. When configured by a processor or the like, the program is recorded on a recording medium such as a magnetic disk device, a magnetic tape device, an FD, or a ROM (read-only memory).

The embodiment has been described above with reference to a specific example. However, the present invention is not limited to these specific examples. For example, when the defective pixel 36 is located near the end of the frame area 30, and the gradation value of the pixel group 38 including the peripheral pixels of the defective pixel 36 cannot be obtained from one image, the adjacent frames A low NA scan of the region 30 may also be performed to create an adjacent low NA optical image.

Further, although the description of parts not directly required for the description of the present invention such as the device configuration and the control method is omitted, the required device configuration and the control method can be appropriately selected and used. For example, the description of the control unit configuration for controlling the inspection device 100 has been omitted, but it goes without saying that the required control unit configuration is appropriately selected and used.

In addition, all pattern inspection devices and pattern inspection methods that include the elements of the present invention and can be appropriately redesigned by those skilled in the art are included in the scope of the present invention.

10 Inspection area 12, 13 Line pattern 14, 15 Space pattern 20 Inspection stripe 30 Frame area 36 Pixels 38 Pixel group 40 Threshold level measurement unit 42 Measurement image pattern Dimension measurement unit 44 Reference image pattern Dimension measurement unit 46 Comparison unit 50, 52, 59 Storage device 51 Area identification unit 54 Alignment unit 56 Measurement image Edge pair detection unit 58 Reference image Edge pair detection unit 60 Plate plate 61 Polarized beam splitter 63 Mirror 64, 65, 66 Lens 67, 73 Mirror 68 Condenser lens 69, 70, 72 , 80, 81, 82, 83 Lens 90 Polarizing element 91 Aperture 92, 93 Lens 100 Inspection device 101 Mask substrate 102 XYθ Table 103 Light source 104 Objective lens 105 Photo diode array 106 Sensor circuit 107 Position circuit 108 Comparison circuit 109 Magnetic disk device 110 Control computer 111 Deployment circuit 112 Reference circuit 113 Autoloader control circuit 114 Table control circuit 115 Magnetic tape device 116 FD
117 CRT
118 Pattern monitor 119 Printer 120 Bus 122 Laser length measurement system 123 Stripe pattern memory 140 AF control circuit 142 Piezo element 144 Inspection mode switching control circuit 146 Low NA inspection circuit 150 Optical image acquisition mechanism 160 Control system circuit 170 Transmission illumination Optical system 171 Illumination Shape switching mechanism 172 Reflective illumination optical system 174 Polarized beam splitter 176 Drive mechanism 178 Imaging optical system 180 Aperture aperture 190 Transmitted light 192 Light 300 Mask substrate 301 Transmitted light 302 Objective lens 304 Semiconductor substrate 305 Light

Claims (4)

The process of acquiring an optical image of the mask substrate on which the pattern is formed, and
The process of creating a reference image corresponding to the optical image and
Used for the reference image in which the pattern dimension of the pattern normal portion in the optical image and the pattern dimension of the corresponding portion in the reference image are the same dimensions specified by using the first threshold level of the image gradation value profile. The process of measuring the second threshold level and
Using the second threshold level, the step of measuring the pattern dimension of the corresponding portion in the reference image corresponding to the pattern abnormality portion in the optical image, and the step of measuring the pattern dimension of the corresponding portion in the reference image.
The tentative pattern size of the pattern abnormality portion in the optical image is measured using the first threshold level, and the pattern dimension of the corresponding portion in the reference image using the second threshold level is determined by the first. The step of calculating the value obtained by dividing the value obtained by dividing the temporary pattern dimension of the pattern abnormality portion in the optical image by the set dimension using the threshold level of is calculated as the pattern dimension of the pattern abnormality portion in the optical image.
A step of comparing and outputting the pattern dimension measured from the reference image and the pattern dimension measured from the optical image for the pattern abnormality portion.
A pattern inspection method characterized by being equipped with. The pattern inspection method according to claim 1, further comprising a step of detecting a plurality of edge pairs in the same direction of the pattern in the optical image. The optical image is a first optical image, and the reference image is a first reference image.
As the area of the pattern abnormal part in the first optical image, the light beam used for the first optical image is a high numerical aperture state than the numerical aperture (NA) in the case of entering from the mask substrate to the objective lens The gradation difference between the pixels in the second optical image of the mask substrate imaged using the light beam and the corresponding pixels in the second reference image corresponding to the second optical image exceeds the third threshold value. A pixel group composed of a defective image and a plurality of pixels around the defective pixel is used.
When measuring the pattern dimension of the pattern abnormality portion in the first optical image, the dimension between a plurality of edge pairs in the pixel group is measured.
When measuring the pattern size of the corresponding portion in the first reference image corresponding to the pattern abnormality portion in the first optical image, between the plurality of edge pairs corresponding to the plurality of edge pairs in the pixel group. The dimensions are measured,
When comparing the first said and measured pattern dimension from the optical image of the first measurement pattern dimension from the reference image, among the plurality of dimensions between edge pair in the pixel group, the first The pattern inspection method according to claim 1 or 2, wherein the pattern dimension between the edge pairs that maximizes the difference between the optical image and the first reference image is determined to be the pattern dimension of the pattern abnormality portion and output. .. An optical image acquisition mechanism that acquires an optical image of the mask substrate on which the pattern is formed,
A reference image generating unit configured to generate a reference image corresponding to the optical image,
Used for the reference image in which the pattern dimension of the pattern normal portion in the optical image and the pattern dimension of the corresponding portion in the reference image are the same dimensions specified by using the first threshold level of the image gradation value profile. A threshold level measuring unit that measures the second threshold level,
Using the second threshold level, a first pattern dimension measuring unit for measuring the pattern dimension of the corresponding portion in the reference image corresponding to the pattern abnormality portion in the optical image, and a first pattern dimension measuring unit.
The tentative pattern size of the pattern abnormality portion in the optical image is measured using the first threshold level, and the pattern dimension of the corresponding portion in the reference image using the second threshold level is determined by the first. The second pattern dimension is calculated as the pattern dimension of the pattern abnormality portion in the optical image by multiplying the value obtained by dividing by the temporary pattern dimension of the pattern abnormality portion in the optical image using the threshold level of. With the measuring unit
With respect to the pattern abnormality portion, a comparison unit for comparing the pattern size measured from the reference image and the pattern size measured from the optical image, and a comparison unit.
A pattern inspection device characterized by being equipped with.

Images