JP5201649B2 - Inspection apparatus, inspection method, and pattern substrate manufacturing method - Google Patents

Description

  The present invention relates to an inspection apparatus, an inspection method, and a pattern substrate manufacturing method.

  In a manufacturing process of a display panel used for a liquid crystal display or the like, if there is a defect in a pattern, it causes a defect such as a wiring insulation failure or a short circuit, resulting in a decrease in yield. For example, it becomes a defect in a portion where foreign matters are attached to the pattern or a portion where the pattern is missing or protruding. Therefore, an inspection apparatus for inspecting whether or not a foreign substance is attached to a substrate for a display device or a pattern substrate such as a photomask used in the manufacturing process and whether a pattern is normally formed is used. .

  There are mainly two types of photomask inspection apparatuses: a die-to-database method and a die-to-die method. In the die-to-database method, a defect is detected by comparing an actually detected image with CAD data stored in a processing device such as a computer. On the other hand, in the die-to-die method, defects having the same shape formed on the mask are compared to detect a defect. In such an inspection apparatus, a defect can be detected by taking a difference between detection data and reference data.

  In a photomask used in a manufacturing process of a liquid crystal display panel, a binary mask in which a chromium film serving as a light shielding film is formed on a transparent glass substrate is usually used. Furthermore, in recent years, a halftone mask having an intermediate transmittance pattern having a transmittance between a chromium film and glass and a gray tone mask having a fine pattern below the resolution limit of an exposure apparatus are used. It has become. As a defect inspection apparatus for this halftone mask, an apparatus for adjusting offset and gain is disclosed (Patent Document 1). Also, a graytone mask inspection apparatus is disclosed that detects a defect by setting a different threshold value for a graytone pattern (Patent Document 2).

JP-A-8-247738 JP 2003-307500 A

    However, the conventional photomask inspection apparatus has the following problems. This problem will be described with reference to FIG. 2, FIG. 3, and FIG. FIG. 2 is a diagram illustrating an example of a halftone mask pattern. FIG. 3 is a diagram showing data corresponding to the pattern shown in FIG. Here, FIG. 2A shows a comparative pattern, that is, a pattern of a halftone mask in a normal portion having no defect, and FIG. 2B shows a pattern configuration similar to that of FIG. It is a figure which shows the location where a defect exists. FIG. 3A is a diagram showing data corresponding to the pattern shown in FIG. 2A, and FIG. 3B is a diagram showing data corresponding to the pattern shown in FIG. The data shown in FIGS. 3A and 3B are compared to detect a defect. The data shown in FIG. 3A is generated from mask design data, for example.

  As shown in FIG. 2A, a halftone pattern 61 and a light shielding pattern 62 are provided on the sample 13. A region other than the halftone pattern 61 and the light shielding pattern 62 on the sample 13 becomes a transmission pattern 60. Here, as shown in FIG. 2B, the white defect 66 and the black defect 65 exist on the halftone pattern 61. Further, a black defect 67 exists on the transmission pattern 60. A white defect 68 exists on the light shielding pattern 62. Here, the transmittance of the transmissive pattern 60 and the white defects 66 and 68 is 100%, the transmittance of the light shielding pattern 62 and the black defects 65 and 67 is 0%, and the transmittance of the halftone pattern 61 is 50%.

  In the pattern configuration shown in FIG. 2A, the data corresponds to the pattern transmittance. On the other hand, at the defect location, the transmittance further varies depending on the type of defect. Therefore, as shown in FIG. 3B, the detection data fluctuates with respect to the data of FIG. 3A at the position where the defect exists. Here, in order to detect a defect, the difference between the reference data shown in FIG. 3A and the detection data shown in FIG. The difference data indicating the difference between the reference data and the detection data is as shown in FIG.

  A defect is detected by comparing the difference data with the threshold values TH1 and TH2. That is, a portion exceeding the thresholds TH1 and TH2 is a defective portion. Here, the data of the defects (corresponding to the black defect 67 and the white defect 68 in FIG. 2) on the transmission pattern 60 and the light shielding pattern 62 have a large amplitude and can be easily detected. However, the amplitude of the defect of the halftone pattern 61 (corresponding to the black defect 65 and the white defect 66 in FIG. 2) becomes small. Accordingly, when the difference in transmittance between the defective portion and the halftone pattern 61 is small, the amplitude between the defect level and the normal portion level becomes smaller. In this case, in order to detect a defect reliably, it becomes difficult to set the threshold value strictly. Therefore, there is a problem that the detection sensitivity is lowered due to noise or the like and a defect is erroneously detected. As described above, the conventional inspection apparatus has a problem that a pseudo defect occurs and the inspection cannot be performed accurately.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to provide an inspection apparatus and an inspection method capable of accurately performing an inspection, and a pattern substrate manufacturing method.

The inspection apparatus according to the first aspect of the present invention includes a transmission pattern that transmits illumination light, a light-shielding pattern that blocks illumination light, and a transmitted light amount of the illumination light is between the light-shielding pattern and the transmission pattern. An inspection apparatus that inspects a sample provided with an intermediate pattern, and detects and detects transmitted light that has passed through the sample out of illumination light emitted from the light source and illumination light emitted from the light source. A detector that outputs detection data corresponding to the amount of light, a pattern information storage unit that stores pattern information based on a pattern provided on the sample, and a detection data from the detector that is stored in the pattern information storage unit A variable amplifier that amplifies the amplified data based on the pattern information and generates amplification data, and the amplified data and the comparison data from the variable amplifier. In which and a defect detection section for detecting a defect and compare. Thereby, an inspection can be performed accurately.

  The inspection apparatus according to a second aspect of the present invention is the above-described inspection apparatus, wherein the amplification factor of the variable amplifier with respect to the intermediate pattern is a light amount detected by the detector in the light shielding pattern or the transmission pattern, The intermediate pattern is set based on the ratio to the amount of light detected by the detector. Thereby, an inspection can be performed more accurately.

  An inspection apparatus according to a third aspect of the present invention is the above-described inspection apparatus, wherein the value of the comparison data corresponding to the transmission pattern and the value of the comparison data corresponding to the intermediate pattern are substantially the same. It is what. Thereby, inspection can be performed more accurately.

  An inspection apparatus according to a fourth aspect of the present invention is the inspection apparatus described above, wherein the comparison data is generated based on pattern information stored in the pattern information storage unit.

  An inspection apparatus according to a fifth aspect of the present invention is the above-described inspection apparatus, wherein the comparison data is generated by amplifying detection data from the detector with the variable amplifier.

  An inspection apparatus according to a sixth aspect of the present invention is the inspection apparatus described above, wherein illumination light incident on the sample includes at least one of g-line, h-line, and i-line. To do. Thereby, a test | inspection can be performed more correctly.

  The inspection method according to the seventh aspect of the present invention is an inspection method for inspecting a sample having a transmission pattern that transmits illumination light, a light-blocking pattern that blocks illumination light, and an intermediate pattern between the light-blocking patterns. The illumination light is incident on the sample, the illumination light transmitted through the sample is detected, the detection data from the detector is amplified at an amplification factor according to the type of pattern, and amplified data is generated, A defect is detected by comparing the amplified detection data with the comparison data. Thereby, an inspection can be performed accurately.

  The inspection method according to an eighth aspect of the present invention is the inspection method described above, wherein an amplification factor for the intermediate pattern is a ratio of a detected light amount in the light shielding pattern or the transmissive pattern and a detected light amount in the intermediate pattern. It is set based on. Thereby, an inspection can be performed accurately.

  An inspection method according to a ninth aspect of the present invention is the inspection method described above, wherein the value of the comparison data corresponding to the transmission pattern and the value of the comparison data corresponding to the intermediate pattern are substantially the same. It is what. Thereby, an inspection can be performed accurately.

  An inspection method according to a tenth aspect of the present invention is the inspection method described above, wherein the illumination light incident on the sample includes at least one of g-line, h-line, and i-line. To do. Thereby, a test | inspection can be performed more correctly.

  According to an inspection method of an eleventh aspect of the present invention, an inspection for inspecting a photomask is performed by the above-described inspection method, a defect of the inspected photomask is corrected, and the defect is corrected via the corrected photomask. The substrate is exposed, and the exposed substrate is developed. Thereby, the productivity of the pattern substrate can be improved.

  ADVANTAGE OF THE INVENTION According to this invention, the inspection apparatus and inspection method which can be correctly test | inspected based on a transmitted image and a reflected image, and the manufacturing method of a pattern substrate can be provided.

  Hereinafter, embodiments to which the present invention can be applied will be described. The following description explains the embodiment of the present invention, and the present invention is not limited to the following embodiment. For clarity of explanation, the following description is omitted and simplified as appropriate. Further, those skilled in the art will be able to easily change, add, and convert each element of the following embodiments within the scope of the present invention. In addition, what attached | subjected the same code | symbol in each figure has shown the same element, and abbreviate | omits description suitably.

  An inspection apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a diagram showing a configuration of an inspection apparatus according to the present invention. Here, an inspection apparatus for inspecting a halftone mask used in the manufacturing process of the liquid crystal display will be described as an example. The inspection apparatus according to the present embodiment is a die-to-database inspection apparatus. 11 is a light source, 12 is a stage, 13 is a sample, 14 is a lens, 15 is a detector, 16 is a stage drive unit, 17 is a control unit, and 100 is a processing device. Sample 13 is a halftone mask, and two types of patterns having different transmittances are provided on a transparent glass substrate. That is, a transmission pattern 60, a halftone pattern 61, and a light shielding pattern 62 are formed on the sample 13. That is, the portion of the sample 13 other than where the halftone pattern 61 or the light shielding pattern 62 is provided becomes the transmission pattern 60.

  The light source 11 is, for example, a lamp light source, and emits illumination light in the direction of the sample 13. The sample 13 is placed on the transparent stage 12. The stage 12 is, for example, an XYZ stage, and operates according to a drive signal from the stage drive unit 16. Illumination light from the light source 11 passes through the stage 12 and enters the sample 13. Thereby, the sample 13 is illuminated uniformly from the back side. In the transmission pattern 60 of the sample 13, most of the illumination light is transmitted. On the other hand, in the light shielding pattern 62 of the sample 13, most of the illumination light is shielded. In the halftone pattern 61, approximately half of the illumination light passes through the sample 13.

  The transmitted light that has passed through the sample 13 enters the lens 14. The transmitted light that has entered the lens 14 is refracted and enters the detector 15. Thereby, an image of the sample 13 is formed on the light receiving surface of the detector 15. The detector 15 is a CCD line sensor, for example, and the light receiving elements are arranged in a line. Here, the direction in which the light receiving elements of the detector 15 are arranged is the X direction, and the direction parallel to the sample surface and perpendicular to the X direction is the Y direction. The detector 15 outputs detection data based on the detected light amount detected by each light receiving element to the control unit 17. The detection data of the detector 15 indicates the transmittance of the sample 13. The control unit 17 is provided with an arithmetic processing unit 100. The arithmetic processing unit 100 performs arithmetic processing for detecting defects. The arithmetic processing in the arithmetic processing unit 100 will be described later.

  A stage driving unit 16 is connected to the stage 12. The stage drive unit 16 outputs a drive signal for driving the stage 12 according to a control signal from the control unit 17. The stage 12 is driven by a drive signal from the stage drive unit 16. Specifically, the stage 12 is driven in a direction (Y direction) orthogonal to the direction in which the light receiving element of the detector 15 is provided. Thereby, the strip-shaped area | region in the sample 13 is imaged. That is, a rectangular area corresponding to the width of the detector 15 is imaged. Then, after the stage 12 is shifted in the X direction by a length corresponding to the width of the detector 15, the stage is driven in the -Y direction. Thereby, the strip-shaped area adjacent to the strip-shaped area that has already been imaged is imaged. By repeating this, the detected area moves in a zigzag manner. Therefore, the area to be inspected can be raster scanned. As described above, the stage 12 and the stage driving unit 16 that drives the stage 12 function as a scanning unit that changes the relative position between the sample 13 and the illumination light. Thereby, the substantially whole surface of the sample 13 can be irradiated with illumination light. The sample 13 is controlled by the control unit 17.

  The arithmetic processing unit 100 executes arithmetic processing for detecting defects on the sample 13. And the control part 17 pinpoints the position on the stage 12 of the detected defect. That is, it is determined to which position on the sample 13 the timing when the defect is detected corresponds. Then, the defect position on the sample 13 is stored. Thereby, the defect location on the sample 13 can be specified. Furthermore, pattern information corresponding to the pattern of the sample 13 is stored in the arithmetic processing unit 100. That is, data corresponding to the shapes of the transmission pattern 60, the halftone pattern 61, and the light shielding pattern 62 is stored as pattern information. Based on this pattern information, reference data serving as a reference for the detection data is generated. Thereby, the inspection of a die-to-database system can be performed.

  Next, an example of a pattern formed on the sample 13 will be described with reference to FIG. Here, FIG. 2A shows a comparative pattern, that is, a pattern of a halftone mask in a normal portion having no defect. FIG. 2B is a diagram showing a pattern in which a defect exists in the same pattern configuration as in FIG. FIG. 2A shows a pattern of normal locations. FIG. 2B is a diagram showing a location where a defect exists in the same pattern configuration as in FIG. The case where the pattern image shown in FIG. 2B is captured by the detector 15 will be described below. The pattern shown in FIG. 2A is a comparison pattern, and the pattern shown in FIG. 2B is a detection pattern. In the die-to-database inspection apparatus, information about the comparison pattern shown in FIG. 2A is stored in the database.

  As shown in FIG. 2A, a rectangular halftone pattern 61 and a rectangular light shielding pattern 62 are provided in the transmission pattern 60 on the sample 13. A region other than the halftone pattern 61 and the light shielding pattern 62 on the sample 13 becomes a transmission pattern 60. Therefore, three types of patterns having different transmittances are formed on the sample 13. The halftone pattern 61 is arranged side by side with the light shielding pattern 62. A transmission pattern 60 is provided between the halftone pattern 61 and the light shielding pattern 62. Accordingly, when scanning is performed in the direction of the arrow in FIG. 2A, illumination is performed in the order of the transmission pattern, the halftone pattern 61, the transmission pattern 60, the light shielding pattern 62, and the transmission pattern 60.

  Here, as shown in FIG. 2B, the white defect 66 and the black defect 65 exist on the halftone pattern 61. Further, a black defect 67 exists on the transmission pattern 60 between the halftone pattern 61 and the light shielding pattern 62. A white defect 68 exists on the light shielding pattern 62. Accordingly, when scanning is performed in the direction of the arrow in FIG. 2B, defects are detected in the order of the black defect 65, the white defect 66, the black defect 67, and the white defect 68. Here, the transmittance of the transmissive pattern 60 and the white defects 66 and 68 is 100%, the transmittance of the light shielding pattern 62 and the black defects 65 and 67 is 0%, and the transmittance of the halftone pattern 61 is 50%. A difference occurs in the value of the detected data at the defective portion. White defects 66 and 68 are white defects through which illumination light is transmitted, and black defects 65 and 67 are defects in which illumination light is shielded. For example, white defects 66 and 68 are formed by pattern omission, and black defects are formed by adhesion of foreign matter or the like.

  Data corresponding to the comparison pattern shown in FIG. 2A is shown in FIG. 3A, and data 36 corresponding to the detection pattern shown in FIG. 2B is shown in FIG. Here, the data shown in FIG. 3A is set as reference data, and the data shown in FIG. 3B is set as detection data. The reference data shown in FIG. 3A and the detection data shown in FIG. 3B respectively correspond to the pattern transmittance. In the inspection apparatus of the die-to-database method, the reference data shown in FIG. 3A is generated based on the pattern information stored in the database.

  Here, the reference data value in the transmission pattern 60 is 100, the reference data value in the halftone pattern 61 is 50, and the reference data value in the light shielding pattern 62 is 0. On the other hand, the detection data 36 has substantially the same value as the reference data except for the defective portion. However, at the defective portion, the value of the detection data 36 is different from the value of the reference data. That is, since the detected light amount detected by the detector 15 is high in the white defects 66 and 68 and the detected light amount is low in the black defects 65 and 67, the detection data value is deviated from the reference data value. That is, in the black defect 65, the detected data value is lower than the reference data value, and in the white defects 66 and 68, the detected data value is higher than the reference data value. Therefore, as shown in FIG. 3B, a maximum value and a minimum value corresponding to the defect appear in the detection data. If the difference between the reference data shown in FIG. 3 (a) and the detection data shown in FIG. 3 (b) is taken directly, the detection sensitivity will decrease as described in the prior art.

  Therefore, in the present embodiment, the amplification factor for the detection data is changed according to the type of pattern on the sample 13. That is, the amplification factor of the halftone pattern 61 is set to a value corresponding to the transmittance ratio. Specifically, since the transmittance of the halftone pattern 61 is 50% and the transmittance of the transmission pattern 60 is 100%, the amplification factor in the halftone pattern 61 is set to twice the amplification factor in the transmission pattern 60. As a result, the halftone pattern 61 and the transmission pattern 60 are at substantially the same level. The light shielding pattern 62 has the same amplification factor as that of the transmission pattern 60. Further, for the comparison data to be compared with the amplified data, the halftone pattern 61 and the transmission pattern 60 are set at substantially the same level. Thus, detection sensitivity can be increased by setting different amplification factors for different patterns. Of course, when the ratio between the transmittance of the halftone pattern 61 and the transmittance of the transmission pattern is different, the amplification factor has a different value.

  The data amplified at the above amplification factor will be described with reference to FIG. FIG. 4 shows data in the case where the amplification factor of the halftone pattern 61 is twice that of the transmission pattern 60 with respect to the data of FIG. Here, FIG. 4A shows data when the level of the halftone pattern 61 is matched with the level of the transmission pattern 60 in the reference data shown in FIG. FIG. 4B shows data when the amplification factor of the halftone pattern 61 is doubled in the detection data shown in FIG. 3B. FIG. 4C shows the amplification factor. Here, the data shown in FIG. 4A is referred to as comparison data 29, and the data shown in FIG. That is, the amplified data 37 means data obtained by amplifying the detection data 36 with an amplification factor set according to the pattern. For example, in the amplification data 37 shown in FIG. 4B, the amplification factor in the transmission pattern 60 and the light shielding pattern 62 is set to 1 with respect to the detection data shown in FIG. 2. Here, the amplification factor changes as shown in FIG.

  In the comparison data 29, similarly to the reference data, the data value is 100 in the region corresponding to the transmission pattern 60, and the data value is 0 in the region corresponding to the light shielding pattern 62. However, in the area corresponding to the halftone pattern 61, the value of the comparison data is 100, unlike the reference data. That is, the value of the halftone pattern 61 and the value of the transmission pattern 60 are made substantially equal. Similarly, the amplified data 37 other than the defective portion has a data value of 100 in the transmission pattern 60, a data value of 0 in the light shielding pattern 62, and a data value of 100 in the halftone pattern 61. Further, in the amplified data 37, the data value is amplified even in the defect on the halftone pattern 61. Specifically, in the amplified data 37, the value of the white defect 66 on the halftone pattern 61 whose data value is 100 is amplified to 200. On the other hand, since the value of the white defect 66 on the halftone pattern 61 is 0, it remains 0 even when amplified.

  In the white defect 68 on the light shielding pattern 62, the black defect 65 on the halftone pattern 61, and the black defect 67 on the transmission pattern 60, the data value of the defect portion does not change, but in the white defect 66 on the halftone pattern 61, The data value changes. That is, in the amplified data 37, the difference in luminance level between the defective portion and the normal portion in the halftone pattern 61 becomes large. In other words, in the amplified data 37, the amplitude caused by the defect increases in the halftone pattern 61.

In order to detect a defect, the difference between the amplified data 37 and the comparison data 29 is taken. Here, the comparison data 29 is subtracted from the amplified data 37. Here, the difference between the amplified data 37 and the comparison data 29 is shown as difference data 41 in FIG. As shown in FIG. 5, since there is no difference between the comparison data 29 and the amplified data 37 at a normal part without a defect, the difference data 41 is zero. On the other hand, the data value of the white defect 68 in the light shielding pattern 62 is 100, and the data value of the black defect 67 in the transmission pattern 60 is -100. Further, the data value of the white defect 66 in the halftone pattern 61 is 100, and the data value of the black defect 65 in the halftone pattern 61 is -100. Thus, in the difference data 41, the amplitude with respect to the normal part of the defect in the halftone pattern 61 becomes large.

  Here, a threshold value TH1 and a threshold value TH2 are set in order to specify the defect position. A portion that exceeds the threshold TH1 and a portion that exceeds the threshold TH2 are defective portions. The threshold value TH1 is set to a positive level, and the threshold value TH2 is set to a negative level. Therefore, the detected light amount is higher than the normal location at the location where the amplified data 37 exceeds the threshold value TH1, and the detected light amount is lower than the normal location at the location where the amplified data 37 is below the threshold value TH2. Therefore, it is determined whether or not a white defect exists based on the threshold value TH1, and it is determined whether or not a black defect exists based on the threshold value TH2. Here, also in the halftone pattern 61, the amplitude with respect to the normal part is large in the defective part. Thus, since the difference between the level of the defective portion and the level of the normal portion is large in the halftone pattern 61, the defect can be detected more accurately. Thereby, the threshold values TH1 and TH2 can be strictly set even for a mask in which the halftone pattern 61 exists. Therefore, even when defect detection is performed with high sensitivity, generation of pseudo defects can be prevented.

  Next, a configuration for executing the above processing will be described with reference to FIG. FIG. 6 is a diagram illustrating a configuration of the arithmetic processing unit 100. The arithmetic processing unit 100 includes a halftone pattern information storage unit 21, a binary pattern information storage unit 22, a convolution processing unit 23, a superimposing circuit 24, an amplifier 31, an A / D converter 32, a variable amplifier 33, a difference circuit 40, and a comparator. 45 and a comparator 46. First, in order to execute the above processing, the detection data from the detector 15 is amplified by the amplifier 31 and then A / D converted by the A / D converter 32. As a result, the detection data is converted from analog data to digital data. The detection data 36 is as shown in FIG. The detection data 36 converted into digital data is amplified by the variable amplifier 33. Amplified data 37 is generated by the variable amplifier 33. Here, the amplification factor of the variable amplifier 33 is set based on an LUT (look-up table) 25. In the LUT 25, different amplification factors are set according to patterns. With this LUT 25, the amplification factor of the variable amplifier 33 is switched according to the pattern corresponding to the detected data. That is, the variable amplifier 33 is set with a different amplification factor depending on the position. For example, in the halftone pattern 61, the amplification factor is 2, and otherwise it is 1. Amplified data 37 from the variable amplifier 33 is input to the difference circuit 40. The amplified data 37 is as shown in FIG. As described above, the LUT 25 sets the amplification factor of the variable amplifier 33 based on the ratio between the light amount detected by the detector in the light shielding pattern or the transmission pattern and the light amount detected by the detector in the halftone pattern 61.

  The halftone pattern information storage unit 21 stores pattern information of the halftone pattern on the sample 13 based on the mask design data. The light shielding pattern information storage unit 22 stores pattern information of the light shielding pattern on the sample 13 based on the mask design data. Here, the pattern information indicates the configuration of the pattern formed on the sample 13 and is generated based on the mask design data. In other words, the position of the sample and the data based on the pattern existing at the position are stored in association with each other. For example, if the pattern information is binarized data, it is 1 at a location where the pattern exists, and 0 at a location where no pattern exists. Furthermore, a value of “1” or “0” corresponding to the position is stored for the entire sample. Specifically, in the pattern information stored in the halftone pattern information storage unit 21, it is 1 at a location where the halftone pattern 61 exists, and 0 at a location where it does not exist. In the pattern information stored in the light shielding pattern information storage unit 22, the value is 1 when the light shielding pattern 62 is present, and is 0 when the light shielding pattern 62 is not present. As described above, the halftone pattern information storage unit 21 and the light shielding pattern information storage unit 22 each store pattern information corresponding to the pattern on the sample. Furthermore, the halftone pattern information storage unit 21 and the light shielding pattern information storage unit 22 store pattern information as a database for each mask to be inspected. Thereby, comparison data can be generated for various masks.

  The convolution processing unit 23 performs convolution processing on the pattern information stored in the halftone pattern information storage unit 21 and the light shielding pattern information storage unit 22. In the convolution process, data corresponding to an optical image is generated from the pattern information. For example, convolution processing is performed on halftone pattern information. As a result, convolution data corresponding to the amount of light detected by the detector when no light shielding pattern is present is generated. That is, reference data for an ideal state without defects is obtained for a mask in which the light shielding pattern 62 is changed to a transmission pattern. Further, a convolution process is performed on the light shielding pattern information. Thereby, convolution data corresponding to the amount of light detected by the detector when the halftone pattern 61 does not exist is generated. In the actual convolution process, the rounding of the signal waveform based on the diffraction due to the pattern is considered.

  The superimposing circuit 24 superimposes the two convolution data generated by the convolution processing unit. That is, the convolution data based on the halftone pattern information and the convolution data based on the light shielding pattern information are added together. Thereby, the comparison data 29 about the sample 13 in which the halftone pattern 61 and the light shielding pattern 62 are formed is generated. The comparison data 29 is, for example, as shown in FIG. That is, the data value of the halftone pattern 61 and the data value of the transmission pattern are 100, and the data value of the light shielding pattern is 0. The comparison data 29 has no difference in the comparison data 29 between the halftone pattern 61 and the transmission pattern 60. Therefore, in an ideal state, the comparison data 29 can be generated based only on the light shielding pattern information.

  The comparison data 29 from the superimposing circuit 24 is input to the difference circuit 40. The amplified data 37 is also input to the difference circuit 40. The difference circuit 40 takes the difference between the amplified data 37 and the comparison data. The comparison data 29 and the amplified data 37 are input in synchronization. Therefore, the difference circuit 40 takes the difference of the position corresponding to each. That is, the difference circuit 40 takes the difference at the same position of the same pattern and outputs it as difference data 41.

  The LUT 25 is generated based on the pattern information of the halftone pattern 61 stored in the halftone pattern information storage unit 21. That is, the LUT 25 is generated based on the coordinates of the location where the halftone pattern 61 exists and the transmittance thereof. Here, the amplification factor in the halftone pattern 61 is an amplification factor according to the ratio between the detected light amount in the transmission pattern 60 and the detected light amount in the halftone pattern 61. This amplification factor is applied to a place where the halftone pattern 61 exists. A reference amplification factor is applied to the light shielding pattern 62 and the transmission pattern 60. That is, in the above case, in the LUT 25, the gain of the light shielding pattern 62 and the transmission pattern 60 is set to 1, and the gain of the halftone pattern 61 is set to 2. Specifically, the LUT 25 outputs 2 at the timing of the halftone pattern 61 and outputs 1 at other timings. This value is set as the amplification factor of the variable amplifier 33. The difference data 41 from the difference circuit 40 is input to the comparators 45 and 46.

  The comparators 45 and 46 compare the threshold value with the difference data 41. Specifically, the comparator 45 compares the threshold value TH1 with the difference data 41, and the comparator 46 compares the threshold value TH2 with the difference data 41. Then, the comparator 45 outputs a defect signal to the control unit 17 at a timing when the difference data 41 exceeds the threshold value TH1. Thereby, the position of the defect is specified. The control unit 17 stores the coordinates of the defect based on the defect signal. Specifically, the comparator 46 outputs a defect signal at the timing when the difference data 41 exceeds the threshold value TH2. Therefore, a defect exists at a location corresponding to the timing at which the defect signal is output. The difference circuit 40 and the comparators 45 and 46 function as a defect detection unit that detects defects. The comparator 45 outputs a defect signal at a timing when a white defect is detected, and the comparator 46 outputs a defect signal at a timing when a black defect is detected. That is, the defect type can be determined by comparing the two threshold values TH1 and TH2 with the difference data. Furthermore, it is possible to determine on which pattern the defect exists by referring to the pattern information.

  As described above, the defect can be accurately detected by performing the calculation process in the calculation processing unit 100 shown in FIG. Furthermore, since defects can be detected in parallel by the two comparators 45 and 46, the inspection time can be shortened. It is also possible to determine the type of defect by using the two comparators 45 and 46. In the above configuration, it is possible to detect defects on the halftone pattern as well as the transmission pattern 60 and the light shielding pattern 62 with only two comparators. Therefore, the defect detection sensitivity can be improved with a simple configuration.

Embodiment 2 of the Invention
The inspection apparatus according to this embodiment will be described with reference to FIGS. FIG. 7 is a diagram illustrating signals processed by the inspection apparatus according to the present embodiment. FIG. 8 is a diagram illustrating a configuration of an arithmetic processing unit used in the inspection apparatus according to the present embodiment. The basic configuration of the inspection apparatus according to the present embodiment is the same as that of the first embodiment, but a part of the configuration of the arithmetic processing unit 100 is different. Therefore, the description overlapping with the first embodiment is omitted. Also in the present embodiment, processing when the patterns shown in FIG. 2 are compared will be described below.

  The data for the pattern shown in FIG. 2 is as shown in FIG. 3 as described above. Also in this embodiment, the amplification factor in the halftone pattern 61 is set to be twice the amplification factor in the transmission pattern 60 and the light shielding pattern 62. Further, comparison data in which the halftone pattern 61 and the transmission pattern 60 are at the same level is generated. Thereby, the amplified data 37 and the comparison data 29 shown in FIG. 4 are generated as in the first embodiment. The comparison data 29 and the amplified data 37 are input to the difference circuit 40. Thereby, difference data 41 is generated. The processing so far is the same as in the first embodiment.

  In the present embodiment, the level shift unit 42 shifts the level of the data value of the difference data 41. Specifically, as shown in FIG. 5, the difference data 41 that is a value of −100 to +100 is used as the shift data 47 from 0 to +100. That is, 100 is added to the difference data 41, and the difference data 41 is multiplied by 1/2. That is, when the difference data is x, the shift data 47 is (x + 100) / 2. Thereby, the shift data 47 shown in FIG. 7A is generated. In the shift data 47, the value is 0 at the black defect portion and the value is 100 at the white defect portion. In the shift data 47, the value is 50 at a normal location.

  Next, the shift data 47 is input to the extraction units 43 and 44. Here, the extraction units 43 and 44 extract a part of the shift data. Specifically, the extraction unit 43 extracts and expands data values from 50 to 100 out of the shift data 47 from 0 to 100. On the other hand, the extraction unit 44 extracts and expands data values from 0 to 50 out of the shift data from 0 to 100. Here, for example, when the value of the shift data 47 is y, the extraction unit 43 sets the value to 0 when 0 ≦ y <50, and (y−50) × 2 when 50 ≦ y ≦ 100. To do. As a result, the range from 50 to 100 of the shift data 47 is expanded to 0 to 100 as shown in FIG. On the other hand, the extraction unit 44 sets the value to 0 when 50 <y ≦ 100, and 2 × y when 0 ≦ y ≦ 50. Thereby, the range from 0 to 50 of the shift data 47 is expanded to 0 to 100 as shown in FIG. Here, data generated by extracting data from 0 to 50 out of the shift data is referred to as extraction data 48, and data generated by extracting data from 50 to 100 is referred to as extraction data 49. The extraction units 48 and 49 can be configured by, for example, a lookup table in which an output value is determined according to an input value. Thereby, data conversion can be easily performed.

  The extracted data 48 is input to the comparator 45, and the extracted data 49 is input to the comparator 46. The comparator 45 compares the extracted data 48 and the threshold value TH1 as shown in FIG. Thereby, a defect signal is output at a timing exceeding the threshold value TH1. On the other hand, the comparator 46 compares the extracted data 49 with the threshold value TH2 as shown in FIG. Thereby, a defect signal is output at a timing exceeding the threshold value TH2.

  Thus, in the present embodiment, the same effect as in the first embodiment can be obtained. Furthermore, in the present embodiment, as described above, a part of the amplified data 37 is extracted and compared after being expanded. Therefore, the sensitivity can be improved only by adding a part of the configuration to the arithmetic processing unit 100 of the existing inspection apparatus. Specifically, detection with high sensitivity is possible by simply adding the LUT 25, the variable amplifier 33, the level shift unit 42, and the extraction units 43 and 43 to the arithmetic processing unit 100 of the conventional inspection apparatus. Therefore, the detection sensitivity can be improved by simply making a simple improvement over the conventional inspection apparatus.

  In the description of the first and second embodiments, the die-to-database type inspection apparatus has been described. However, the inspection apparatus according to the present invention can also be applied to a die-to-die type inspection apparatus. In a die-to-die inspection apparatus, reference data corresponding to the pattern shown in FIG. 2A is generated based on a signal from a detector. Therefore, comparison data is generated from the detection data from the detector. In this case, the amplified data 37 amplified by the variable amplifier 33 may be compared synchronously. For example, in an inspection apparatus provided with a plurality of detectors 15, two sets of amplified data may be compared in synchronization. Further, in an inspection apparatus having one detector 15, the amplified data 37 before and after being delayed may be compared.

  In the above description, the sample 13 is described as a halftone mask, but the present invention is not limited to this. For example, a gray tone mask can be used for the sample 13. The gray tone mask is a photomask that adjusts the exposure amount by forming a fine pattern that is less than the resolution limit of the exposure apparatus. The pattern formed on the sample 13 may be an intermediate pattern in which the amount of transmitted illumination light is between the light shielding pattern 62 and the transmission pattern 60. That is, when illuminated with the same amount of illumination light, the amount of light transmitted through the intermediate pattern is between the transmission pattern 60 and the light shielding pattern 62. Therefore, even when a light shielding pattern having a transmittance of approximately 0% is formed with a pattern size equal to or less than the resolution limit, it is an intermediate pattern. As described above, the inspection apparatus according to the present invention can be used for a mask on which an intermediate pattern in which the amount of transmitted illumination light is between the light shielding pattern 62 and the transmission pattern 60 is formed. Of course, the sample is not limited to a photomask. If the sample has a transmissive pattern, a light-shielding pattern, and an intermediate pattern, inspection can be performed.

  Further, in the above description, for clarity of explanation, the transmittance of the black defects 65 and 67 is 0% and the transmittance of the white defects 66 and 68 is 100%. Of course, in the above inspection apparatus, , Defects having various transmittances can be detected. The transmissive pattern 60 may be a pattern having a high light transmittance, and the light shielding pattern 62 may be a pattern having a low light transmittance. A defect having a small difference from the transmittance of the halftone pattern 61 can be detected with high sensitivity. Therefore, it can test | inspect more correctly.

  In addition, the inspection method and inspection apparatus described in Embodiment 1 are suitable for inspection of a pattern substrate such as a photomask, for example. Then, by performing exposure using the inspected mask, the yield of a pattern substrate such as a color filter substrate used in a semiconductor device or a display device can be improved. That is, first, the mask is inspected by the above-described inspection method and inspection apparatus. And exposure for pattern formation is performed with respect to the board | substrate with which the photosensitive resin was apply | coated by the mask suitable for inspection, or the mask which corrected the detected defect. Then, development or the like is performed to form a pattern on the substrate. Thereby, productivity of pattern substrates, such as a color filter substrate used for a semiconductor device and a display device, can be improved.

Embodiment 3 of the Invention
In the inspection apparatus according to the embodiment of the present invention, at least one of g-line with a wavelength of 435 nm, h-line with 405 nm, and i-line with 365 nm is used for the illumination light for inspection. That is, light having a spectrum including g-line, h-line, or i-line is used as inspection illumination light. Of course, light having a spectrum including two or more of g-line, h-line, and i-line may be used as inspection illumination light. Contrast can be increased by using light having such a spectrum as inspection illumination light.

  The inspection apparatus according to this embodiment will be described with reference to FIG. In the inspection apparatus according to the present embodiment, a wavelength filter 18 is used in addition to the inspection apparatus shown in the first embodiment. The basic configuration of the inspection apparatus according to this embodiment is the same as that of the inspection apparatus shown in FIG. Also, the inspection method is the same as that in the above embodiment, and the description thereof is omitted. In the present embodiment, for example, a mercury lamp that emits g-line, h-line, and i-line is used as the light source 11. Further, the wavelength filter 18 transmits light including g-line, h-line, or i-line, and illuminates the sample 13. Of course, two or more light sources may be used in combination. For example, the wavelength filter 18 is a filter that cuts light having a wavelength longer than that of the g line (for example, e line). The wavelength filter 18 may be incorporated in the light source 11.

  Here, the reason why the contrast can be increased by using the light of the above-mentioned wavelength for the inspection illumination light will be described with reference to FIGS. FIG. 10 is a plan view showing a pattern imaged by the inspection apparatus. FIG. 11 is a diagram illustrating data of a comparative example when none of the g-line, h-line, and i-line is used. Specifically, FIG. 11 shows data when the e-line (546 nm) is used. FIG. 12 is a diagram showing data when all of the g-line, h-line, and i-line are used.

  In the present embodiment, a case where a region where the sample 13 is provided with the halftone pattern 61 and the light shielding pattern 62 is inspected as in FIG. 2 will be described. That is, as shown in FIG. 9, a portion where the halftone pattern 61 and the light shielding pattern 62 are arranged side by side in the transparent pattern 60 is imaged. Here, FIG. 10A shows a comparative pattern, that is, a pattern of a halftone mask in a normal portion having no defect. FIG. 2B is a diagram showing a pattern in which a defect exists in the same pattern configuration as in FIG. Then, as shown in FIG. 10B, black defects 65 and white defects 66 are present on the halftone pattern 61.

  Here, the transmittance of the halftone pattern 61 at a wavelength longer than the exposure wavelength may be about 90%. In this case, detection data when illuminated with e-line is as shown in FIGS. 11 (a) and 11 (b). FIG. 11A shows detection data corresponding to FIG. 10A, and FIG. 11B shows detection data corresponding to FIG. 10B. Here, the e-line transmittance in the halftone pattern 61 is about 90%. Therefore, contrast cannot be improved for the white defect 66 existing on the halftone pattern 61. That is, the white defect 66 has a transmittance of about 100%, and the halftone pattern 61 has a transmittance of about 90%. Accordingly, since the difference in transmittance is only 10%, as shown in FIG. 11B, the difference between the white defect 66 on the halftone pattern 61 and the detected data at the normal location becomes small. When such detection data is amplified as described above and the difference is taken, the result is as shown in FIG.

  Here, when the transmittance is 90%, the amplification factor is about 1.1. In this case, also in the difference data shown in FIG. 11C, the level difference between the white defect 61 and the normal portion on the halftone pattern 61 becomes small. Thus, with e-line illumination, only about 10% contrast can be obtained with or without defects. For this reason, it is difficult to improve S / N.

  Therefore, in this embodiment, S / N is improved by using illumination light in a wavelength region including g-line, h-line, and i-line. In these wavelength ranges, the transmittance of the halftone pattern 61 is about half that of the transmission pattern 60. Here, the description will be made assuming that the transmittance of the halftone pattern 61 in the above wavelength range is 60%. The detection data when illuminated using this wavelength range is as shown in FIGS. 12 (a) and 12 (b). 12A shows detection data corresponding to FIG. 10A, and FIG. 12B shows detection data corresponding to FIG. 10B.

  As shown in FIG. 12B, the contrast between the color defect 66 and the normal portion in the halftone pattern 61 is increased. When such detection data is amplified as described above and the difference is taken, the result is as shown in FIG. Here, when the transmittance is 60%, the amplification factor is about 1.67. Therefore, in the difference data, the level difference between the white defect 61 and the normal portion on the halftone pattern 61 becomes large. That is, a contrast of about 67% can be obtained with or without defects. Thereby, S / N can be made high and detection sensitivity can be improved. Further, the detection sensitivity of the black defect 65 does not decrease. Therefore, the inspection can be performed accurately.

  In this way, illumination is performed using light in a wavelength region where the contrast between the transmission pattern 60 and the halftone pattern 61 is high. That is, the wavelength filter 18 that cuts light having a wavelength longer than the exposure wavelength is used. As a result, it is possible to block light having a wavelength at which the contrast deteriorates. That is, it can be illuminated with light including light having an exposure wavelength with high contrast. Therefore, the S / N can be increased and the detection sensitivity of the white defect 66 on the halftone pattern 61 can be improved.

It is a figure which shows typically the structure of the test | inspection apparatus concerning this invention. It is a figure which shows an example of the pattern imaged by the test | inspection apparatus concerning this invention. It is a figure which shows the detection data corresponding to the pattern in FIG. FIG. 4 is a diagram showing amplified data obtained by amplifying the detection data shown in FIG. 3 and comparison data. It is a figure which shows the difference data which is a difference of the comparison data shown by Fig.4 (a), and the amplification data shown by FIG.4 (b). It is a figure which shows the structure of the arithmetic processing part of the test | inspection apparatus concerning Embodiment 1 of this invention. It is a figure which shows the data processed by the arithmetic processing part of the test | inspection apparatus concerning Embodiment 2 of this invention. It is a figure which shows the structure of the arithmetic processing part of the test | inspection apparatus concerning Embodiment 2 of this invention. It is a figure which shows the structure of the test | inspection apparatus concerning Embodiment 3 of this invention. It is a figure which shows an example of the pattern imaged with the inspection apparatus of Embodiment 3 of this invention. It is a figure which shows the data of the comparative example in Embodiment 3 of this invention. It is a figure which shows the data in Embodiment 3 of this invention. It is a figure for demonstrating the process currently performed with the conventional test | inspection apparatus.

Explanation of symbols

11 light source, 12 stage, 13 sample, 14 lens, 15 detector,
16 stage drive unit, 17 control unit,
21 halftone pattern information storage unit, 22 light-shielding pattern information storage unit,
23 convolution processing unit, 24 superimposing circuit, 25 LUT, 31 amplifier,
32 A / D converter, 33 variable amplifier, 36 detection data, 37 amplification data,
40 difference circuit, 41 difference data, 42 level shift unit, 43 extraction unit,
44 extraction unit, 45 comparator, 46 comparator, 47 shift data,
48 extracted data, 49 extracted data,
60 transmission pattern, 61 halftone pattern, 62 shading pattern,
65 black defect, 66 white defect, 67 black defect, 68 white defect, 100 arithmetic processing unit

Claims (9)

Inspection for inspecting a sample provided with a transmission pattern that transmits illumination light, a light shielding pattern that blocks illumination light, and an intermediate pattern in which the amount of transmitted light of the illumination light is between the light shielding pattern and the transmission pattern A device,
A light source that emits illumination light;
A detector that detects transmitted light that has passed through the sample out of the illumination light emitted from the light source, and outputs detection data corresponding to the detected light amount;
A pattern information storage unit for storing pattern information based on design data of a pattern provided in the sample;
A variable amplifier that amplifies detection data from the detector based on pattern information stored in the pattern information storage unit at an amplification factor according to the type of pattern, and generates amplified data;
A defect detection unit that detects a defect by comparing the amplified data from the variable amplifier and the comparison data generated based on the pattern information stored in the pattern information storage unit ,
The amplification factor of the variable amplifier with respect to the intermediate pattern is larger than the amplification factor of the variable amplifier with respect to the light shielding pattern and the transmission pattern,
In the region corresponding to the intermediate pattern, the value of the comparison data is different from the value of the reference data corresponding to the transmittance of the pattern generated based on the design data .   The amplification factor of the variable amplifier with respect to the intermediate pattern is set based on a ratio between a light amount detected by the detector in the light shielding pattern or the transmission pattern and a light amount detected by the detector in the intermediate pattern. The inspection apparatus according to claim 1.   The inspection apparatus according to claim 1, wherein a value of the comparison data corresponding to the transmission pattern and a value of the comparison data corresponding to the intermediate pattern are substantially the same. 4. The inspection apparatus according to claim 1 , wherein the illumination light incident on the sample includes at least one of g-line, h-line, and i-line. An inspection method for inspecting a sample having a transmission pattern that transmits illumination light, a light shielding pattern that shields illumination light, and an intermediate pattern between the light shielding patterns,
Make the illumination light incident on the sample,
Detecting illumination light transmitted through the sample;
Amplification data is generated by amplifying detection data from the detector at an amplification factor corresponding to the type of pattern based on pattern information stored based on pattern design data provided on the sample. And
A defect is detected by comparing the amplified detection data with comparison data generated based on the pattern information stored in the pattern information storage unit ,
The amplification factor for the intermediate pattern is greater than the amplification factor for the light shielding pattern and the transmission pattern;
In the region corresponding to the intermediate pattern, the value of the comparison data is different from the value of the reference data corresponding to the transmittance of the pattern generated based on the design data . The inspection method according to claim 5 , wherein an amplification factor for the intermediate pattern is set based on a ratio between a detected light amount in the light shielding pattern or the transmissive pattern and a detected light amount in the intermediate pattern. The inspection method according to claim 5, wherein a value of the comparison data corresponding to the transmission pattern and a value of the comparison data corresponding to the intermediate pattern are substantially the same. Inspection method according to any one of claims 5-7 in which the illumination light is characterized in that it comprises at least one g-line, h-line, and i-rays incident on the sample. Inspecting a photomask by the inspection method according to any one of claims 5 to 8 ,
Correcting defects in the inspected photomask;
Exposing the substrate through a photomask in which the defect is corrected;
A method for producing a patterned substrate, wherein the exposed substrate is developed.

Images