JP4446051B2 - Pattern substrate defect inspection apparatus, defect inspection method, and pattern substrate manufacturing method - Google Patents

Description

  The present invention relates to a pattern substrate defect inspection method, a defect inspection method, and a pattern substrate manufacturing method, and more particularly, a defect inspection apparatus, a defect inspection method, and an identical pattern for inspecting defects on a substrate on which the same pattern is repeatedly provided. The present invention relates to a method of manufacturing a patterned substrate in which is repeatedly provided.

  In a mask used for exposure of a color filter used in a liquid crystal display device or a semiconductor device, a semiconductor device, a CCD array, etc., minute elements are regularly arranged. In such a pattern substrate on which a pattern in which the same elements are regularly repeated is formed, even if a defect occurs in one element, the entire apparatus is defective. Therefore, it is desired to inspect the pattern substrate for defects.

  In the defect inspection apparatus for the patterned substrate on which the same pattern is repeatedly formed, the light emitted from the light source is divided and irradiated onto the substrate as two spot lights, and the intensity of each light is adjusted by two detectors. To detect. The entire surface of the substrate is inspected by moving the relative position between the detector and the substrate. The light intensity signals detected by the two detectors are compared, and a defect is detected based on the difference. The method using two detectors is disadvantageous from the viewpoint of cost because it is necessary to provide two detection systems.

  In order to solve this point, a defect inspection apparatus using one detector is disclosed by the applicant of the present application (for example, Patent Document 1 and Patent Document 2). This defect inspection apparatus will be described with reference to FIG. FIG. 13 is a diagram showing an example of the configuration of a conventional color filter defect inspection apparatus.

  A color filter substrate 11 to be inspected is placed on an XY stage 12. The XY stage 12 is made of a transparent material so as to transmit light from the light source 13. Light from the light source 13 is applied to the substrate 11. The light transmitted through the substrate 11 enters the detector 15 through the lens 14. The detector 15 receives the transmitted light and outputs a signal based on the intensity of the light. The output signal is delayed through the delay circuit 18 and then supplied to one input terminal of the comparison circuit 19. A signal that is not delayed is supplied to the other input terminal of the comparison circuit 19. The detector 15 or the XY stage 12 is scanned in the direction of the arrow at a constant speed, and the presence / absence of a defect is determined based on the change in light intensity.

  As shown in FIG. 14, the substrate 11 of the color filter usually includes cells (pixels) 51 and a BM (black matrix) 50 provided therebetween. The cell 51 is abbreviated as a single cell divided into R, G, and B. Part of the light from the light source 11 irradiated to the cell 51 passes through the substrate 11 and enters the detector 15. Since the BM 50 is a light shielding layer that blocks the irradiated light, the light from the light source 11 irradiated on the BM 50 does not enter the detector 15. When the XY stage 12 is moved in the X direction, the detection spot moves in the direction of the arrow, and one line of the substrate 11 is inspected.

  Here, the defect detection when one white defect 52 exists in the uppermost line of the substrate 11 will be described. FIG. 15 shows the waveform of the output signal of the detector 25 when the uppermost line is inspected. FIG. 15A shows the signal waveform A output from the detector 15, and FIG. 15B shows the signal waveform B delayed by the delay circuit 18. In FIGS. 15A and 15B, reference numeral 61 denotes a cell signal 61 corresponding to one cell 51. The cell signal 61 is a trapezoidal signal having a width corresponding to the size of the cell 51 and the scanning speed of the detector 15 or the moving speed of the XY stage. In a cell having no defect, the cell signal 61 has substantially the same waveform. Between each cell signal 61, the output is low corresponding to the BM 50 that blocks light. Further, in the cell corresponding to the white defect 52, a defect signal 62 whose output is higher than the other cell signals 61 appears. That is, in the cell having the white defect 52, the transmitted light becomes strong, and the output of the detector 15 becomes high. The signal B has a waveform delayed from the signal A by a time corresponding to one cell pitch.

  FIG. 15C shows the waveform of the difference signal C between the signals A and B, that is, AB. The difference signal C is 0 because the signal A and the signal B are the same signal at a place where neither the signal A nor the signal B has the white defect 52. However, the difference signal between the signal A and the signal B does not become zero at a place where the white defect 52 exists in either the signal A or the signal B. Therefore, a positive peak and a negative peak appear in the continuous signal C as shown in FIG. It is determined whether or not this peak exceeds slice levels H and L. When the slice level H is exceeded in the positive direction, it is determined that a positive peak has occurred. When the slice level L is exceeded in the negative direction, it is determined that a negative peak has occurred. When a positive peak and a negative peak occur in this order in a cell, it is determined that there is a white defect. If the order of occurrence of positive and negative peaks is opposite, it is determined that there is a black defect that blocks light.

  However, this defect inspection method has the following problems. For example, in the color filter substrate shown in FIG. 14, it is assumed that the white defect 51 is continuously present in two cells as in the second line from the top. The signal when this line is inspected is as shown in FIG. 16A shows the signal waveform A of the light intensity detected by the detector 15, and FIG. 16B shows the signal B delayed by one cell pitch. Two defect signals 62 appear continuously for both signals A and B. Since the defect signals 62 of the signals A and B are shifted by one cell pitch, the positions of one defect signal 62 are overlapped.

  FIG. 16C shows a difference signal C between the signal A and the signal B. The difference signal C is 0 at the location where the defect signal 62 overlaps. Therefore, the signal between the positive peak and the negative peak becomes 0, and it is not determined that there is the white defect 51. As described above, the conventional defect inspection apparatus erroneously detects. Moreover, although the example in which the white defect 51 continues is shown in FIG. 16, it will not be judged that there is a black defect in the same manner. In addition, if two or more cells have consecutive white or black defects, if black and white defects are continuous, or if there is a defect in one cell, a defect is overlooked, There is a problem that an erroneous determination of a defect and a defect size occurs and the defect inspection cannot be performed accurately.

JP-A-1-173172 JP-A-6-357416

  As described above, there is a problem that the conventional defect inspection apparatus cannot accurately perform defect inspection.

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

A defect inspection apparatus according to the present invention is a defect inspection apparatus that inspects a pattern defect of a pattern substrate (for example, the substrate 11 in the embodiment of the present invention) on which substantially the same pattern is repeatedly formed. Based on the detected light intensity by detecting transmitted light or reflected light from the light source that irradiates light (for example, the light source 13 in the embodiment of the present invention) and the light source irradiated to the pattern substrate. A detector for outputting a signal (for example, the detector 15 in the embodiment of the present invention) and a scanning unit for scanning the detection position of the detector on the pattern substrate (for example, the XY stage 12 in the embodiment of the present invention) ) And a delayed signal for generating a 0th to Nth order delayed signal obtained by delaying a signal detected in accordance with the scanning of the scanning unit by 0 to N ( N is an integer of 4 or more ) pattern The generation unit (for example, the + n-order delay signal generation unit 71n to the -n-order delay signal generation unit 72n in the embodiment of the present invention) and one of the 0 to N-order delay signals as a reference signal, A difference signal generation unit that generates N difference signals based on the difference between the 0th to Nth order delay signals excluding the reference signal and the reference signal (for example, + nth order difference signal generation in the embodiment of the present invention) 73n to -n-th order difference signal generation unit 74n), a comparison unit that compares each of the N difference signals with a threshold value (for example, comparison unit 75 in the embodiment of the present invention), and 1 of the pattern substrate Defect determination that determines the number of difference signals that exceed the threshold among the N difference signals corresponding to one pattern and determines whether or not a defect exists based on the number of difference signals that exceed the threshold (E.g. missing in the embodiment of the present invention) Determination unit 76) and those provided with. Thereby, the defect inspection can be accurately performed.

In the above-described defect inspection apparatus, it is determined that a defect is present when the number of difference signals exceeding the threshold is greater than or equal to a set number among the N difference signals corresponding to one pattern of the pattern substrate. It is desirable. Thereby, even when a large-size defect exists or a defect exists continuously in a plurality of patterns, the defect inspection can be performed accurately.
The above number is preferably a majority of N.

  In the above-described defect inspection apparatus, the threshold is provided on the positive side and the negative side, and the defect determination unit detects a white defect or a black defect based on the number of difference signals exceeding the positive threshold or the negative threshold. It is desirable to determine whether it exists. Thereby, the presence or absence of a white defect and a black defect can be determined.

A defect inspection apparatus according to the present invention includes a step of irradiating light to a pattern substrate on which substantially the same pattern is repeatedly formed, a step of detecting transmitted light or reflected light of the light irradiated to the pattern substrate, and the pattern substrate. Scanning the detection position of the transmitted light or the reflected light at, and delaying a signal based on the intensity of the light detected by the detector from 0 to N ( N is an integer equal to or greater than 4 ) 0 to Nth order Generating a delayed signal; Generating N difference signals based on the difference between the reference signal and the 0th to Nth order delay signals using one of the 0th to Nth order delay signals as a reference signal; Comparing each of the difference signals with a threshold, determining the number of difference signals exceeding the threshold among the N difference signals corresponding to one pattern of the pattern substrate, and exceeding the threshold And determining whether a defect exists based on the number of difference signals. Thereby, the defect inspection can be accurately performed.

In the above-described defect inspection method, it is determined that a defect exists when the number of difference signals exceeding the threshold is equal to or greater than a set number among the N difference signals corresponding to one pattern on the pattern substrate. It is desirable. Thereby, even when a large-size defect exists or a defect exists continuously in a plurality of patterns, the defect inspection can be performed accurately.
The set number is preferably a majority of N.

The pattern substrate manufacturing method according to the present invention includes a step of forming substantially the same pattern on the substrate, a step of irradiating the substrate on which the pattern is formed, and a transmitted light or a reflection of the light irradiated on the substrate. detecting the light, the detector 0~N a signal based on the intensity of light detected by the (N is an integer of 4 or more) generating a number of pattern delaying 0~N next delay signal , One of the 1st to Nth order delayed signals as a reference signal, and a 1st to Nth order difference signal based on the difference between the 0th to Nth order delayed signals excluding the reference signal and the reference signal A step of comparing the 1st to Nth order difference signals with a threshold value, and of the 1st to Nth order difference signals corresponding to one pattern of the pattern substrate, the difference signal exceeding the threshold value Obtaining a number; and The method includes a step of determining whether or not a defect exists based on the number of difference signals exceeding a threshold value, and a step of correcting a pattern in which the defect exists. Thereby, an accurate pattern can be formed on the substrate, and the productivity of the patterned substrate can be improved.

  ADVANTAGE OF THE INVENTION According to this invention, the defect inspection apparatus and defect inspection method which can perform a defect inspection correctly, and the manufacturing method of a pattern board | substrate can be provided.

Embodiments of the present invention will be described below with reference to the drawings. The following description shows preferred examples of the present invention, and the scope of the present invention is not limited to the following examples. In the following description, the same reference numerals denote the same contents.
Example 1.

  A defect inspection apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a diagram showing the configuration of a color filter defect inspection apparatus. Reference numeral 11 denotes a substrate, 12 denotes an XY stage, 13 denotes a light source, 14 denotes an objective lens, 15 denotes a detector, 16 denotes a driving mechanism, and 17 denotes a processing apparatus. In this embodiment, the substrate 11 to be inspected is a color filter substrate for a liquid crystal display device. Examples of the defect of the color filter include a white defect in which a part of the color element passes through and a black defect that does not transmit light due to adhesion of foreign matter, etc., and the defect inspection apparatus according to the present invention has a position of a substrate-like defect. And the type of defect can be inspected. The present invention is not limited to a color filter, and can be used for a substrate on which the same pattern is provided at a constant repetition pitch.

  First, the configuration of the defect inspection apparatus will be described. The substrate 11 is placed on the XY stage 12. The XY stage 12 is made of a transparent material so as to transmit light from the light source 13. For example, an illumination light source such as a xenon lamp or a halogen lamp is suitable for the light source 13, and the entire surface of the substrate 11 is irradiated with uniform light. Of course, other illumination light sources may be used, and a laser light source or the like may be used. The light emitted from the light source 13 enters the substrate 11. The light transmitted through the substrate 11 enters the detector 15 through the lens 14. An image sensor such as a CCD camera is used for the detector 15, and the light transmitted through the substrate 11 is received and a transmission image of the substrate 11 is captured. The detector 15 A / D converts a signal corresponding to the intensity of light received at each pixel of the CCD camera and outputs the signal to the processing device 17. A signal corresponding to the light intensity is stored in the processing device 17 in association with each pixel of the CCD camera. When the detection of the imaging area of the CCD camera is completed, the detector 15 is moved by the drive mechanism 16 to detect the entire substrate. The drive mechanism 16 outputs a position signal corresponding to the position of the detector 15 to the processing device 17. At this time, the objective lens 14 is also moved in the same manner as the detector 15. The processing device 17 includes hardware for performing image processing and a computer for performing arithmetic processing. Based on this position signal, the processing device 17 specifies the position on the substrate detected by the detector 15. The computer of the processing device 17 includes a CPU (central processing unit), a memory, and the like, and delays the output signal from the detector for a time corresponding to the cell pitch and compares the output signal with the delayed signal. Further, the processing device 17 specifies the position, size, and type of the defect based on the detection pixel of the detector 15, the position signal of the detector 15, and the comparison result.

  Next, the color filter substrate to be inspected will be described. FIG. 2 is a plan view schematically showing a partial configuration of the color filter substrate. As shown in FIG. 2, the color filter substrate 11 to be inspected is usually formed with a plurality of cells 51 having the same pattern at the same repetition pitch. Between adjacent cells (pixels) 51, a BM (black matrix) 50 is provided. Note that the cell 51 is abbreviated as a cell divided into R, G, and B. Part of the light from the light source 11 irradiated to the cell 51 passes through the substrate 11 and enters the detector 15. Since the BM 50 is a light shielding layer that blocks the irradiated light, the light from the light source 11 irradiated on the BM 50 does not enter the detector 15. A defect is detected by scanning a detection spot in the imaging region along the cell pattern. When the scanning of the imaging area is completed, the detector 15 is moved in the X direction, and the adjacent imaging area is similarly scanned. Thereby, the detection position on the substrate is scanned. When the inspection of the substrate in the X direction is completed, the detector 15 or the XY stage 12 is shifted in the Y direction and the inspection is performed in the same manner. By repeating this scan, the entire substrate is inspected for defects. When the XY stage 12 is moved, since the position of the XY stage 12 is also input to the processing device 17 in the same manner as the detector 15, the detection position on the substrate can be specified.

  In the first embodiment, as shown in the uppermost line in FIG. 2, a line in which a white defect 52 exists in one cell 51 is inspected. FIG. 3 shows a detection signal when the uppermost line is inspected. FIG. 3 is a diagram illustrating a signal output from the detector 15 and a signal processed by the processing device 17. FIG. 3A shows an output signal of the detector and a signal obtained by delaying the output signal by 1 cell pitch from 4 cell pitches.

  In FIG. 3A, the uppermost signal A is an output signal output from the detector. First, the signal A output from the detector will be described. In FIG. 3, reference numeral 61 denotes a cell signal 61 corresponding to one cell 51. The cell signal 61 is a trapezoidal signal having a width corresponding to the size of the cell 51 and the scanning speed of the detector 15, and the cell signal 61 has the same waveform in a cell having no defect. Between each cell signal 61, the output is low corresponding to the BM 50 that blocks light. It is assumed that the second cell signal from the left in the signal A is a signal corresponding to the cell in which the white defect 52 exists. In the cell in which the white defect 52 exists, the transmitted light is stronger than in the other cells, so that the output of the detector 15 is increased. Therefore, in the cell corresponding to the white defect 52, the white defect signal 62 whose output is higher than that of the other cell signals 61 appears.

  Next, a signal obtained by delaying the signal A will be described. In the present invention, the signal A is delayed by 1 to 4 cells to generate a delayed signal. In FIG. 3A, the second signal B from the top is a signal obtained by delaying the signal A by one cell pitch. In FIG. 3A, the third signal C from the top is a signal obtained by delaying the signal A by two cell pitches. In FIG. 3A, the fourth signal D from the top is a signal obtained by delaying the signal A by 3 cell pitches. In FIG. 3A, the fifth signal E from the top is a signal obtained by delaying the signal A by 4 cell pitches. The time corresponding to the cell pitch is determined based on the scanning speed of the detector 15 and the cell size.

  Since the signals A to E are delayed by one cell pitch, the white defect signal 62 is shifted by one cell pitch. Here, the signal C is set as a reference signal Sig (0) as a reference. Since the signal A is a signal obtained by advancing the signal C, that is, Sig (0) by two cell pitches, it is assumed to be Sig (-2). Since the signal B is a signal obtained by advancing Sig (0) by one cell pitch, it is set to Sig (-1). On the other hand, since the signal D is a signal obtained by delaying Sig (0) by one cell pitch, it is set to Sig (+1). Since the signal E is a signal obtained by delaying Sig (0) by two cell pitches, it is set to Sig (+2). The reference Sig (0) is a zero-order delay signal, and Sig (-2) is a -second delay signal. Similarly, Sig (−1), Sig (+1), and Sig (+2) are defined as a −1st order delay signal, a + 1st order delay signal, and a + 2nd order delay signal, respectively.

  Next, a difference signal is generated by subtracting Sig (0) from each delayed signal of Sig (+2) to Sig (-2). In FIG. 3B, the signal F is a difference signal obtained by subtracting Sig (−2) from Sig (−2), and the signal F is referred to as a −second-order difference signal Dif (−2). That is, −2nd order difference signal Dif (−2) = Sig (−2) −Sig (0). The signal G is a difference signal obtained by subtracting Sig (0) from Sig (−1), and the signal G is defined as a −1st order difference signal Dif (−1). That is, −1st order difference signal Dif (−1) = Sig (−1) −Sig (0). The signal H is a difference signal obtained by subtracting Sig (+1) from Sig (+1), and the signal H is defined as a + first order difference signal Dif (+1). That is, the first order difference signal Dif (+1) = Sig (+1) −Sig (0). The signal I is a difference signal obtained by subtracting Sig (0) from Sig (+2), and the signal I is a + second-order difference signal Dif (+2). That is, the second-order difference signal Dif (+2) = Sig (+2) −Sig (0).

  In the difference signal between Dif (−2) and Dif (+2), the portion that does not correspond to the defect signal 62 is 0, but a positive peak or a negative peak appears at the portion corresponding to the white defect signal 62. It is determined whether this peak exceeds the slice level H or the slice level L. When the slice level H is exceeded in the positive direction, it is determined that the cell has a white side signal. On the other hand, when the slice level L is exceeded in the negative direction, it is determined that the cell has a black side signal. The slice levels H and L are set at the same level in all the difference signals.

  In Dif (−2) to Dif (+2), the white side signal is formed at the same cell position. This is because the white side signal is generated based on the defect signal 62 of the reference signal Sig (0). On the other hand, the black side signals are generated in different cells in Dif (−2) to Dif (+2). This is because the black side signal is generated based on the defect signals 62 of Sig (−2) to Sig (+2) excluding Sig (0).

  Dif (−2) to Dif (+2) are arranged on the time axis with Sig (0) as a reference, and it is determined that a cell having three or more white side signals simultaneously has a white defect. Further, it is determined that a cell in which three or more black side signals appear has a black defect. That is, in the fourth cell from the left indicated by an arrow in FIG. 3B, a white side signal appears in any difference signal of Dif (−2) to Dif (+2), and the white side signal is 4 Counted times. Therefore, it is determined that this cell has a white defect. On the other hand, the black side signal appears in different cells in Dif (−2) to Dif (+2). That is, as shown in FIG. 3B, one black side signal appears in each of the second, third, fifth and sixth cells from the left. Accordingly, there is only one black side signal that appears at the same time, and three or more black side signals are not counted. Therefore, it is determined that there is no black defect. In this way, a plurality of difference signals are arranged on the time axis, and it is determined whether or not there is a white defect (black defect) based on the number of white side signals (black side signals) that appear at the same time. Then, the position of the white defect or black defect on the substrate is specified based on the time in the difference signal, that is, the time of the reference signal Sig (0). Thereby, a defect inspection can be performed accurately.

In the above description, the output signal of the detector 15 is delayed up to 4 cells. However, the present invention is not limited to this, and it may be delayed up to an arbitrary M ( M is an integer of 4 or more ) cells. In this case, a 0th to Mth order delay signal delayed from 0 to M cells from the output signal is generated. Using one of the 0th to Mth delay signals as a reference signal, a difference signal is generated based on the difference between the 0th to Mth delay signals excluding the reference signal and the reference signal. That is, M difference signals are generated using one signal out of (M + 1) delayed signals as a reference Sig (0). Then, M difference signals are arranged on the time axis, and the presence or absence of a white defect is determined based on the number of white side signals that appear at the same time. Similarly, the presence or absence of a black defect is determined based on the number of black side signals. At this time, it is determined that a white defect (black defect) exists when the white side signal (black side signal) is counted in a difference signal of a set number or more among the M difference signals. Here, as a preferred embodiment, the above-described number setting assumes that a white defect (black defect) exists when a white signal (black signal) is counted among M majority difference signals. Thereby, a defect inspection can be performed accurately. Furthermore, by determining the presence or absence of a defect based on the number of white side signals (black side signals) in the M difference signals, the size of the defect that can be accurately detected can be increased. For example, when a delay signal of 1 to 2n (where n is a natural number of 2 or more ) is generated and 2n difference signals are generated, a huge defect corresponding to n cells or consecutively present in n cells It is possible to detect defects that occur.

  The processing in the above-described defect inspection is executed in the processing device 17. The configuration of the processing device 17 for performing the above-described processing will be described with reference to FIG. FIG. 4 is a block diagram showing the configuration of the processing device 17. In FIG. 4, 70 is a zero-order delay signal generator, 71a is a primary delay signal generator, 71n is an n-order delay signal generator, 72a is a −1st-order delay signal generator, and 72n is a −n-order delay signal generator. is there. 73a is a + 1st order difference signal generation unit, 73n is a + nth order difference signal generation unit, 74a is a −1st order difference signal generation unit, and 74n is a −nth order difference signal generation unit. 75 is a comparison unit, and 76 is a defect determination unit. Here, n is an arbitrary natural number, and in FIG. 3, it is assumed that n = 2, but here, it will be described as general ± n. In FIG. 4, the secondary to (n−1) th order configurations are not shown.

  The + n-order delay signal generator 71n to + 1-order delay signal generator 71a, the 0-order delay signal generator 70, and the −1st-order delay signal generator 72a to −n-order delay signal generator 72n each output the detector output signal. A delay signal delayed corresponding to the cell pitch is generated. That is, a delay signal delayed from 0 to 2n cell pitch is generated from the output signal of the detector. Therefore, a total of (2n + 1) delay signals are generated. In this case, a signal that is not delayed is also included in the delayed signal as a signal delayed by 0 pitch. The time corresponding to this cell pitch is determined based on the scanning speed of the detector 15 and the size of the cell 51, and specifically, is (cell size) / (scanning speed of the detector 15). Here, among the signals delayed by 0 to 2n cells, a signal delayed by (n + 1) cells is set as a 0th-order delayed signal which is a reference signal.

  The + n-order difference signal generator 73n generates a + n-order difference signal obtained by subtracting the zero-order delay signal generated by the zero-order delay signal generator 70 from the n-order delay signal generated by the + n-order delay signal generator 71n. . Similarly, the −nth order difference signal generation 74n generates a −nth order difference signal obtained by subtracting the 0th order delay signal from the −nth order delay signal. Further, other order difference signals are also generated by the difference of the 0th order delay signal when the order is a delayed signal. Thereby, a difference signal of −n order to + n order excluding the 0th order is generated. That is, 2n difference signals are generated.

  The comparison unit 75 compares the -n-order to + n-order difference signals with the slice levels H and L, and determines whether a black signal or a white signal appears in each difference signal. The defect determination unit 76 determines whether a black defect or a white defect exists based on the number of black-side signals or the number of white-side signals that appear among the −n-order to + n-order difference signals. That is, the presence / absence of a white defect (black defect) is determined depending on how many white side signals (black side signals) appear in the same cell among 2n difference signals. At this time, if a white side signal (black side signal) appears in a predetermined number or more of the 2n difference signals, it is determined that a white defect (black defect) exists. It should be noted that if a white side signal (black side signal) appears in a majority of 2n difference signals, it is desirable to determine that a white defect (black defect) exists. That is, it is desirable to determine that a white defect exists if the white side signal is n + 1 or more of 2n difference signals, and to determine that a black defect exists if the black side signal is n + 1 or more.

  Since the position signal of the detector 15 by the drive mechanism 16 is input to the processing device 17, the defect position on the substrate 11 is specified. That is, the position on the substrate where the defect exists is specified based on the time of the reference signal corresponding to the white side signal and the position signal of the drive mechanism 16. In this way, the presence / absence of a defect, the specification of the defect position, and the determination of a white defect or a black defect are performed. Then, the detector 15 is scanned by the drive mechanism 16 so as to inspect the entire surface of the substrate.

In the above description, a ± n-order difference signal is generated. However, a delayed signal of any order can be used as a reference signal. That is, 2n difference signals can be generated on the basis of one of arbitrary signals of 0th to 2nth order as a reference signal. It is possible to determine whether or not a defect exists based on the number of difference signals exceeding the slice level (threshold value) among the 2n difference signals.
Example 2

  In this embodiment, the defect determination when the white defect 52 exists in two consecutive cells 51 will be described. Since the apparatus configuration and signal processing of this embodiment are the same as those of the first embodiment, description thereof is omitted. Here, as shown in FIG. 5, there are white defects 52 in two adjacent cells, and the detector 15 is scanned to detect these two white defects 52 in succession. FIG. 6 shows a detection signal of a portion where the white defect 52 continues. FIG. 6 is a diagram illustrating a signal output from the detector 15 and a signal processed by the processing device 17. FIG. 6A shows the output signal of the detector and a signal obtained by delaying the output signal by 4 cell pitches from 1 cell pitch.

  In FIG. 6A, signals A to E are signals obtained by shifting the detection signals by 0 to 4 cell pitches as in FIG. The signals A to E are set to Sig (−2) to Sig (+2) as in FIG. In Sig (−2) to Sig (+2), the two defect signals 62 are shifted one cell at a time. Difference signals Dif (−2) to Dif (+2) from Sig (−2), Sig (−1), and Sig (+1) Sig (+2) when Sig (0) is used as a reference zero-order signal. ) Is as shown in FIG.

Since the conventional defect inspection apparatus generates only one difference signal, the white side signal and the black side signal do not continuously occur as shown by Dif (−1). Therefore, a determination is made erroneously without detecting a white defect. However, in this embodiment, among the four difference signals in the two cells indicated by the arrows shown in FIG. 6B, the white side signal appears by three difference signals. Further, the white side signal appears only in two difference signals among the four difference signals in the cells before and after the two cells. That is, the black side signal appears only in Dif (−2) and Dif (−1) in the cell immediately before the two cells, and only Dif (+2) and Dif (+1) are black in the next cell. Side signal does not appear. Therefore, if a white defect is determined when three or more white side signals are generated, it is determined that a white defect exists in two cells indicated by arrows. Thereby, the defect inspection can be accurately performed. In the above description, the white defect has been described. Similarly, for the black defect, if three black-side signals are generated among the four difference signals, it is determined that the black defect exists, thereby performing an accurate inspection. It can be carried out.
Example 3 FIG.

  In this embodiment, the defect determination when a huge white defect 52 extending over two cells 51 exists will be described. Since the apparatus configuration and signal processing of this embodiment are the same as those of the first embodiment, description thereof is omitted. Here, as shown in FIG. 7, there is a huge white defect 52 extending over two cells, and the detector 15 is scanned to detect the huge white defect 52. FIG. 8 shows a detection signal at a location where the huge white defect 52 exists. FIG. 8 is a diagram illustrating a signal output from the detector 15 and a signal processed by the processing device 17. FIG. 8A shows the output signal of the detector and a signal obtained by delaying the output signal by 4 cell pitches from 1 cell pitch.

  In FIG. 8A, signals A to E are signals obtained by shifting the detection signals by 0 to 4 cell pitches as in FIG. The signals A to E are set to Sig (−2) to Sig (+2) as in FIG. In Sig (−2) to Sig (+2), the defect signal 62 extending over two cells is shifted by one cell. Difference signals Dif (−2) to Dif (+2) from Sig (−2), Sig (−1), and Sig (+1) Sig (+2) when Sig (0) is used as a reference zero-order signal. ) Is as shown in FIG.

Since the conventional defect inspection apparatus generates only one difference signal, the white side signal and the black side signal do not continuously occur as shown by Dif (−1). Therefore, a determination is made erroneously without detecting a white defect. However, in this embodiment, among the four difference signals in the two cells indicated by the arrows shown in FIG. 8B, the white side signal appears by three difference signals. Further, the white side signal appears only in two difference signals among the four difference signals in the cells before and after the two cells. That is, the black side signal appears only in Dif (−2) and Dif (−1) in the cell immediately before the two cells, and only Dif (+2) and Dif (+1) are black in the next cell. Side signal does not appear. Therefore, if a white defect is determined when three or more white side signals are generated, it is determined that a white defect exists in the two cells indicated by the arrows. Thereby, the defect inspection can be accurately performed. Although the white defect has been described in the above description, the black defect is also detected in the same manner by determining that a black defect exists when three black side signals are generated among the four difference signals. It can be carried out.
Example 4

  In this embodiment, the defect determination when a normal cell 51 exists between two cells having the white defect 52 will be described. Since the apparatus configuration and signal processing of this embodiment are the same as those of the first embodiment, description thereof is omitted. Here, as shown in FIG. 9, a normal cell 51 exists between two cells having a white defect 52, and the detector 15 is scanned to detect this white defect. FIG. 10 shows a detection signal at a location where the huge white defect 52 exists. FIG. 10 is a diagram illustrating a signal output from the detector 15 and a signal processed by the processing device 17. FIG. 10A shows the output signal of the detector and a signal obtained by delaying the output signal by 1 cell pitch from 4 cell pitches.

  In FIG. 10A, signals A to E are signals obtained by shifting the detection signals by 0 to 4 cell pitches as in FIG. The signals A to E are set to Sig (−2) to Sig (+2) as in FIG. In Sig (−2) to Sig (+2), the defect signals 62 extending over two cells are shifted one cell at a time. Difference signals Dif (−2) to Dif (+2) from Sig (−2), Sig (−1), and Sig (+1) Sig (+2) when Sig (0) is used as a reference zero-order signal. ) Is as shown in FIG.

Since the conventional defect inspection apparatus generates only one difference signal, the white side signal and the black side signal are repeatedly generated as shown in Dif (−1). For this reason, if there is a black defect between the white defects, it is erroneously determined. However, in this embodiment, among the four difference signals in the two cells indicated by the arrows shown in FIG. 8B, the white side signal appears by three difference signals. Further, the cell between the two cells in which the white side signal appears, the black side signal appears only by two difference signals among the four difference signals. Therefore, if a white defect is determined when three or more white side signals are generated, it is determined that a white defect exists in two cells indicated by arrows. Similarly, if a black defect is determined when three or more black side signals are generated, it is determined that the cell between the two cells indicated by the arrows is a normal cell. As described above, according to the present invention, it is possible to accurately perform defect inspection even for scattered defects.
Embodiment 5 FIG.

  In the present embodiment, a description will be given of defect determination when white and black semi-defects are intricately present. Since the apparatus configuration and signal processing of this embodiment are the same as those of the first embodiment, description thereof is omitted. Here, it is assumed that the half-defect is one in which the intensity of transmitted light is changed to a level that does not cause a practical problem of the color filter as compared with a normal cell. That is, the white semi-defect has a higher intensity of transmitted light than a normal cell, but the signal intensity does not exceed the slice level H. On the other hand, the black semi-defect has a lower transmitted light intensity than that of a normal cell, but the signal intensity does not exceed the slice level L.

  Here, as shown in FIG. 11, a case where a black half defect 54 exists between two white half defects 53 will be described. An area having the white half defect 53 and the black half defect 54 is scanned by the detector 15. The detection signal at this time is as shown in FIG. FIG. 12 is a diagram illustrating a signal output from the detector 15 and a signal processed by the processing device 17. FIG. 12A shows the output signal of the detector and a signal obtained by delaying the output signal by 1 cell pitch from 4 cell pitches.

  In FIG. 12A, signals A to E are signals obtained by shifting the detection signals by 0 to 4 cell pitches as in FIG. The signals A to E are set to Sig (−2) to Sig (+2) as in FIG. In Sig (−2) to Sig (+2), the white half-defect signal 63 and the black half-defect signal 64 are shifted one cell at a time. In this embodiment, the black semi-defect 54 has a larger variation in transmittance with respect to a normal cell than the white semi-defect 53. Accordingly, as shown in FIG. 12A, the height of the convex portion of the white half defect signal 63 is lower than the depth of the concave portion of the black half defect signal 64. Difference signals Dif (−2) to Dif (+2) from Sig (−2), Sig (−1), and Sig (+1) Sig (+2) when Sig (0) is used as a reference zero-order signal. ) Is as shown in FIG.

  In this embodiment, since a line having only a half defect is scanned, it must be determined that no defect exists. Accordingly, the slice level H is set above the convex portion of the white half defect signal 63, and the slice level L is set below the concave portion of the black half defect signal 64. In the conventional defect inspection apparatus, as shown by Dif (−1), the height of the black half-defect signal is emphasized by the front and rear white half-defect signals, so that the slice levels H and L are exceeded. Therefore, it is erroneously determined that a black defect exists in a cell in which the black half defect 53 exists.

  However, in the defect inspection apparatus according to the present invention, the presence or absence of a black defect is determined based on the number of black side signals. Of the four difference signals in the cell corresponding to the black half defect, the black side signal appears in Dif (-1) and Dif (+1), but the black side signal appears in Dif (-2) and Dif (+2). do not do. Since the black side signal appears only with two difference signals out of the four difference signals, it is determined that there is no black defect. As described above, in the defect inspection apparatus according to the present invention, since the presence or absence of the black defect is determined based on the number of black side signals, the defect inspection can be accurately performed even for the half defect. In the above-described embodiment, the black defect has been described. However, it is possible to accurately inspect the white defect by determining the presence or absence of the white defect based on the number of white side signals.

  In addition, although the defect inspection apparatus shown in FIG. 1 detected the transmitted light from the board | substrate 11, you may make it detect reflected light. In this case, a half mirror or the like may be placed between the substrate 11 and the detector 15 and the substrate 11 may be illuminated via the half mirror. By using the reflected light, it is possible to accurately inspect even a substrate 11 for a semiconductor device such as a Si wafer that is not transparent. In this case, for example, it is suitable for inspection of a pattern such as a memory. Furthermore, it is possible to inspect the photomask for defects using transmitted light. In the above-described embodiment, both the white defect and the black defect are detected, but only the white defect or the black defect may be detected. In the above-described embodiment, the detector 15 is used to detect transmitted light in a plurality of cells. However, other scanning means, for example, a scanning means that moves the XY stage 12 or the like may be used. After repeatedly forming the same pattern on the substrate, the defect is inspected using the above-described defect inspection method. And the pattern board | substrate with which the pattern was correctly formed can be manufactured with sufficient productivity by correcting the location where the defect of the board | substrate exists.

In the above-described embodiment, the 0th to 2nth order delay signal is generated, but the 0th to (2n + 1) th order delay signal may be generated. In this case, (2n + 1) difference signals are generated, and if a white side signal (black side signal) appears in a difference signal of a majority of the difference signals, it is determined that there is a white defect (black defect). Is desirable. That is, four or more delay signals are generated, and the presence / absence of a defect is determined based on the number of white side signals (black side signals) among the four or more difference signals. Further, any signal of 0 to 2n or 0 to (2n + 1) may be used as a reference signal. Note that the delayed signal in the above description is not limited to time delay due to analog processing, but includes a signal generated by processing for displacing a signal for n cells by digital signal processing. In the first to fifth embodiments, 0 to 4 delay signals are generated. However, any 0 to n ( n is an integer of 4 or more ) delay signals can be generated to perform defect determination. That is, by generating a difference signal from Dif (−n) to Dif (+ n), it is possible to accurately detect a large defect up to n cells, a defect continuously present in n cells, or a defect scattered. Can do. Alternatively, the user may set the number of white side signals (black side signals) determined to be present by inputting an appropriate number to the processing device 17. Thereby, since an arbitrary number can be selected, the convenience can be improved.

It is a figure which shows the structure of the defect inspection apparatus concerning this invention. It is a top view which shows the structure of the defective part in the color filter board | substrate used as a test object. It is a figure which shows the signal waveform detected in the defective part shown by FIG. It is a block diagram which shows the structure of the processing apparatus of the defect inspection apparatus concerning this invention. It is a top view which shows the structure of the defective part in the color filter board | substrate used as a test object. It is a figure which shows the signal waveform detected in the defective part shown by FIG. It is a top view which shows the structure of the defective part in the color filter board | substrate used as a test object. It is a figure which shows the signal waveform detected in the defective part shown by FIG. It is a top view which shows the structure of the defective part in the color filter board | substrate used as a test object. It is a figure which shows the signal waveform detected in the defective part shown by FIG. It is a top view which shows the structure of the defective part in the color filter board | substrate used as a test object. It is a figure which shows the signal waveform detected in the defective part shown by FIG. It is a figure which shows the structure of the conventional defect inspection apparatus. It is a top view which shows the structure of the defective part in the color filter board | substrate used as a test object. It is a figure which shows the signal waveform detected in the defective part shown by FIG. It is a figure which shows the signal waveform detected in the defective part shown by FIG.

Explanation of symbols

11 substrate, 12 XY stage, 13 light source, 14 objective lens, 15 detector,
16 drive mechanism, 17 processing device, 18 delay circuit, 19 comparison circuit 50 BM (black matrix), 51 cell, 52 white defect, 53 white half defect,
54 black defect, 61 cell signal, 62 defect signal,
70 0th-order delay signal generation unit, 71a + 1st-order delay signal generation unit,
71n n-order delay signal generator, 72a -first-order delay signal generator,
72n-nth order delay signal generator, 73a + 1 order difference signal generator,
73n + n-order difference signal generation unit, 74a-primary difference signal generation unit,
74n-nth order difference signal generation unit, 75 comparison unit, 76 defect determination unit

Claims (8)

A defect inspection apparatus for inspecting a pattern defect of a pattern substrate on which substantially the same pattern is repeatedly formed,
A light source for irradiating the pattern substrate with light;
A detector that detects transmitted light or reflected light from the light source irradiated on the pattern substrate and outputs a signal based on the intensity of the detected light;
Scanning means for scanning the detection position of the detector on the pattern substrate;
A delay signal generator for delaying a signal detected according to the scanning of the scanning unit from 0 to N ( N is an integer of 4 or more ) pattern to generate a 0th to Nth order delayed signal;
N difference signals are generated based on a difference between the 0th to Nth order delay signals excluding the reference signal and the reference signal, using one of the 0th to Nth order delay signals as a reference signal. A difference signal generator;
A comparison unit that compares each of the N difference signals with a threshold;
A defect inspection unit including a defect determination unit that determines whether or not a defect exists based on the number of difference signals exceeding the threshold value among the N difference signals corresponding to one pattern of the pattern substrate. apparatus.   The defect inspection according to claim 1, wherein a defect is determined to be present when the number of difference signals exceeding the threshold is greater than or equal to a set number among the N difference signals corresponding to one pattern of the pattern substrate. apparatus.   The defect inspection apparatus according to claim 2, wherein a defect is determined to exist when the number of the difference signals is a majority of N.   The threshold is provided on the positive side and the negative side, and the defect determination unit determines whether a white defect or a black defect exists based on the number of difference signals exceeding the positive threshold or the negative threshold. The defect inspection apparatus according to claim 1 or 2. Irradiating light onto a patterned substrate on which substantially the same pattern is repeatedly formed; and
Detecting transmitted light or reflected light of the light irradiated on the pattern substrate;
Scanning a detection position of the transmitted light or the reflected light on the pattern substrate;
A signal based on the intensity of light detected by the detector is delayed by 0 to N ( N is an integer of 4 or more ) pattern to generate a 0th to Nth order delayed signal;
N difference signals are generated based on a difference between the 0th to Nth order delay signals excluding the reference signal and the reference signal, using one of the 0th to Nth order delay signals as a reference signal. Steps,
Comparing each of the N difference signals to a threshold;
Obtaining the number of difference signals exceeding the threshold among the N difference signals corresponding to one pattern of the pattern substrate;
A defect inspection method comprising a step of determining whether or not a defect exists based on the number of difference signals exceeding the threshold.   The defect inspection according to claim 5, wherein a defect is determined to be present when the number of difference signals exceeding the threshold value is greater than or equal to a set number among the N difference signals corresponding to one pattern of the pattern substrate. Method. The defect inspection method according to claim 6, wherein a defect is determined to exist when the number of the difference signals is a majority of N. Irradiating light onto a patterned substrate on which substantially the same pattern is repeatedly formed; and
Detecting transmitted light or reflected light of the light irradiated on the pattern substrate;
Scanning a detection position of the transmitted light or the reflected light on the pattern substrate;
And generating the 0~N a detection signal based on the intensity of light detected by the detector (N is an integer of 4 or more) pattern delaying by 0~N following delay signal,
N difference signals are generated based on a difference between the 0th to Nth order delay signals excluding the reference signal and the reference signal, using one of the 0th to Nth order delay signals as a reference signal. Steps,
Comparing the first to Nth order difference signals with a threshold;
Obtaining the number of difference signals exceeding the threshold among the 1st to Nth order difference signals corresponding to one pattern of the pattern substrate;
Determining whether a defect exists based on the number of difference signals exceeding the threshold;
And a step of correcting a pattern in which the defect exists.

Images