JPH0752157B2 - Method and apparatus for detecting defective portion of test body - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a method and apparatus for inspecting a defective portion generated during manufacturing of an integrated circuit, and more particularly to a method and an apparatus for detecting the presence of the defective portion in a specimen by an optical pattern corresponding to the specimen. is there.

It has previously been proposed to use an optical image associated with a spatial filter to determine the presence of defects in photomasks used in the fabrication of microminiaturized integrated circuits. A device intended to do such a work is disclosed in U.S. Pat. No. 4,000,949.

Optical images composed of holograms incorporated into an optical processor are particularly advantageous in that the reverse optical tracing technique can be used during reproduction of the hologram. The backtracking light traces are advantageous in eliminating aberrations introduced by the optical components in the device and that would spread the rays that represent defects in the exposed photomask pattern to form the hologram. Spreading the light beam in this way impairs the ability to detect a defective portion. An apparatus using the reverse optical tracing technique is disclosed in US Pat. No. 4,516,833.

The defect image information in the light pattern can be detected based on the following treatment. A light sensitive device having a defined imaging field is positioned, the device is exposed, the light intensity in the defect image area of the size and shape of the imaging field is measured, and then the device is placed next to the device. Move to the area and repeat the measurement process. A problem with such detection procedures is that it takes a relatively long time to stabilize the photosensitive device after moving it from one measuring position to the next adjacent position. This is because it is necessary to stabilize the alignment position of the photosensitive device with respect to the defect image area in order to be able to detect image defects of submicron dimensions. For this reason, such detection schemes require relatively high overall defect detection speeds (ie, speeds that are capable of completely inspecting array-type photomasks, which usually have identical elements). It cannot be applied to the intended use.

An object of the present invention is to provide a method and an apparatus for detecting the presence of a defective portion in a test body by an optical pattern corresponding to the test body.

It is another object of the present invention to provide an apparatus for continuously scanning a light image pattern to detect defects of sub-micron size in a specimen at high efficiency.

A further object of the present invention is to adjust the scanning and exposure scanning,
An object of the present invention is to provide an apparatus of the type described above, which is provided with a function of shortening the time required to inspect a light pattern for a defective portion in a test body.

Yet another object of the present invention is to provide such a method that facilitates detection of relatively low light intensity defect information during successive scans of the light image.

The method and apparatus according to the invention will be described by way of example with a holographic imaging device which is effective for detecting photomasks and for detecting defects in these photomasks. The imager reveals a defect image field whose size and contour correspond to the physical contours of the photomask under examination.
The inspection area of the image field is divided into striped areas each having a width of about 0.65 mm. The width of one stripe area corresponds to the maximum size of the image of the photo-detecting means, ie the camera, ie the optical window used to scan each stripe area. The camera is a charge-coupled device attached to an XY positioning table that moves the camera along the striped area in raster scan format. The control circuitry for the XY positioning table causes the camera to move at a nominal constant velocity as it traverses each stripe area.

The presence of the defective portion in the photomask is determined by detecting the light existing portion of each stripe area divided into the second pixel array. The charge-coupled device (camera) is composed of a large number of photodetector elements arranged in a first array of 256 rows and 403 columns.
In the preferred embodiment of the invention, each photodetector element has a rectangular aperture of 16 microns by 20 microns. First
The size of the array is defined as the size of the second pixel array, and each pixel has a size proportional to the size of the photodetecting element. The number of rows of photodetectors is equal to the number of rows of pixels, and the number of rows of photodetectors is defined by the width of the stripe area. The size of the pixel and photodetector is made equal by a lens that magnifies the image in the defect image field by a factor of 10.

During the scanning period, the charge-coupled device continuously traverses the columns of the first array in a direction along the length of the fringe area, and obtains a row-by-row measured energy value corresponding to the light intensity at the pixels with which the photodetector elements are aligned. To do. In the preferred embodiment, the amount of charge generated in the photodetection element of the charge coupled device is representative of the measured energy value. This energy value is proportional to the product of the amount of charge generated by the light intensity and the exposure time of the photodetector.

The amount of charge measured at each pixel in a row is synchronized with the movement of the camera device along the striped area, but is sequentially shifted along the corresponding column in the opposite direction. In the last row of the first array, the total amount of electric charge of each photodetecting element normally corresponds to 256 times the amount of electric charge of the pixel in which the photodetecting element is aligned. Since the photo-detecting element does not have matching electrical characteristics, the amount of charge generated in response to light existing in a given pixel is different from that of another photo-detecting element. Therefore, the total charge amount of each photodetecting element in the last row of the first array represents the average total charge amount for the corresponding pixel.

25 accumulated for each pixel in the last row of photo detectors
The total energy value of 6 is converted into digital form by a video amplifier and an analog-to-digital converter and transferred to a threshold detector, which indicates whether the light quantity indicates the presence of defects in the photomask. To decide. When the charge-coupled device travels one row width along the striped area, all 403 charge quantities are processed by the threshold detector. When the threshold detector detects the presence of a defect, the position coordinates of the positioning table in the XY directions are stored in a large digital memory unit in combination with the information identifying the column of pixels and the total amount of charge in digital form. To be done.

Once the charge coupled device has completed scanning one stripe area,
The step of the positioning table moves back to another starting position, and scanning of the next adjacent fringe area is started. X due to overscan boundaries at both start and stop positions for each stripe area
The acceleration and deceleration areas of the positioning table in the -Y direction are defined, which allows continuous scanning at a nominal constant speed over the entire stripe area. This process is repeated until the entire inspection area of the defective optical image field has been scanned.

By continuously scanning the defect image field with the charge coupled device during the acquisition of energy values, the photomask can be inspected at high processing speeds. Only by combining the threshold information in the defect image field with continuous scanning capability, sub-micron sized defects represented by relatively low intensity light regions can be quickly detected.

The present invention will be described below with reference to the drawings.

FIG. 1 shows diagrammatically a transparent photomask 10, such as a glass plate coated with chrome, which is a planer consisting of generally identical photographic image elements 12 separated by a distance 14. Contains an array. Each image element 12 defines a circuit pattern that corresponds to a layer of an integrated circuit device that is partially fabricated using a photomask. Photomask 10 has 16 sides
It is a rectangle of a, 16b, 16c and 16d. The number of image elements 12 is typically 20 to 500, but may be smaller or larger than that in specific cases. A remarkable type of photomask, called a reticle, has only one image element.

2A and 2B respectively show an optical processing device for producing a hologram of the photomask 10 and for causing a defect image corresponding to the photomask to appear from the hologram. The optical process of examining such photomasks for defects using such an apparatus is described in detail in U.S. Pat. No. 4,516,833.

Referring to FIG. 2A, the main beam 18 illuminates a photomask 10, which diffracts the light rays incident on it.
The light beam 20 diffracted by the photomask passes through the imaging lens 22 and interferes with the reference beam 24 at a position beyond the Fourier transform plane 26 in the back focal plane of the imaging lens.
Ray 20 and reference beam 24 interfere at the photographic recording medium to form hologram 28. This hologram 28 is a photomask 10
Pattern and the optical information relating to the defective portion existing in this photomask are recorded.

Referring to FIG. 2B, the three-dimensional image 10 'of the photomask 10 is reconstructed from the hologram 28 in the same optical processing system, which also includes a spatial filter 30 which is an error free photomask. Fourier transform the reference pattern. The spatial filter 30 is located on the Fourier transform plane 26.
The hologram 28 is illuminated with a light beam 24 ', which is assumed to be coherent light but propagate in the conjugate direction of the reference beam 24. The spatial filter 30 removes the optical information corresponding to the photomask pattern from the reproduced holographic light pattern. The optical information or light rays 20 'that were not blocked by the spatial filter 30 are re-imaged by the imaging lens 22 which represents the defects present in the photomask pattern. The holographic image 10 'becomes a light region having a relatively low intensity. This holographic image 10 'allows photomask defect features of sub-micron dimensions such as about 0.5 microns to be characterized. The light areas in the holographic defect image 10 'appear at the same relative positions where the corresponding defect appears in the photomask 10.

FIG. 3 diagrammatically shows a scanning mechanism for detecting the presence of defective portions in the photomask 10 in the holographic defect image 10 'corresponding to these defective portions. The defect image field 32 represents a region where the static spatial light image 10 'in which the light intensity in the optical processing system is relatively low appears. The defect image field 32 is formed on each side 10a, 16b, 16c of the photomask 10.
, And 16d corresponding to sides 34a, 34b, 34c and 34d, but not necessarily exactly the same size as photomask 10. In the preferred example, the lens is shown in phantom
The size of the defect image 10 'present in the defect image field 32 is 10 times the size of the defect in the photomask 10 for reasons to be described in more detail below with reference to 35.

The defective portion where the photomask 10 exists is determined by detecting a region where the light intensity is equal to or higher than a predetermined limit value in the inspection region 36 of the defect image field 32. The inspection area 36 includes the space contained within the fracture outline 38 defined by the sides of the image element 12 adjacent each side of the photomask 10 of FIG.

The defect image 10 'appearing in the image field 32 is formed by the rays 20' which converge in the horizontal plane 40. Lens 35 is horizontal 40
A defect image 10 times larger than the defect image appearing on this surface is formed on the lower side. The detection of the defective portion is carried out by scanning the inspection area 36 in the form of a raster type with the light detecting means, that is, the camera 42,
It does this by sensing the presence of light. The scanning procedure in this case is carried out in a plane 44 parallel to the horizontal plane 40 and the lens 35, and below them. In the preferred example, the camera 42 is a solid-state imager comprising a charge coupled device.
Such a device is, for example, the model SID 504 charge coupled device manufactured by RCA Corporation of the United States.

The camera 42 and the lens 35 are supported on a plate 46, which plate 46 forms part of a table 48 for positioning in the XY direction, by means of which the camera in the plane 44 lies.
42 can be positioned to align with any position in defect image field 32. The top of the positioning table 48, the Y-step 50, supports the plate 46 and moves the camera 42 in the plane 44 along the Y-direction. The bottom of the positioning table 48, the X-step 52, directs the camera 42 to the plane 44.
Move along the X-direction inside. A suitable such XY positioning table is, for example, the Model 8500 available from Richmond of California, USA.

4A and 4B are enlarged diagrams showing the outline of the lower left corner of the inspection area 36 of FIG. 3 and the raster scanning path of the camera 42 that scans the outline. FIG. 5 is an enlarged view of the striped area in FIG. 4B, which shows the dimensions of the photodetector elements of the first array and the dimensions of the pixels of the second array obtained by enlarging the lens 35 by 10 times. In a one-to-one correspondence.

Referring to Figures 3, 4A, 4B and 5, the camera 42 has an optical window 54 through which the light to be detected is passed. The optical window 54 is rectangular with sides 56 and 58 defining its length and sides 60 and 62 defining the width of the optical window.

In normal scanning operation, the positioning table 48 causes the camera 42 to move from the starting position 66 in the X-direction to the scan path segment 64a.
Move along to align the side 62 of the optical window 54 with the segment 68 of the examination region 36. The sides 56 and 58 of the optical window 54 are parallel to the Y-direction, and these sides are the striped areas 72 (3 in FIG. 3).
The width 70 of the two striped regions is shown), each striped region representing the portion of the camera that traverses the inspection region 36 as the camera 42 moves in the X-direction. After the side 56 of the optical window 54 has moved past the segment 74 of the examination area 36, the positioning table 48 causes the camera 42 to move the return path along the segment 76a to the second starting position 78 to move the second path. Scan adjacent stripe regions 72. During such a reverse operation (hereinafter also referred to as "retrace operation") of the camera, Y-
Step 50 moves camera 42 in the Y-direction by a distance equal to the width of stripe region 72, while X-step 52 moves camera 42 in the X-direction in the opposite direction by a distance equal to the length of scan path segment 64a. After the retrace operation, the positioning table 48 moves the camera 42 along the scan path segment 64b from the starting position 78 in the X-direction to a second width 70.
Cross the camera in the striped area 72.

The above-described scanning and retrace processing is repeated until the camera 42 crosses the entire inspection area 36. However, there are differences in the lengths of the scan path and retrace path segments that correspond to the size differences of the inspection region 36 in the X-direction.

In particular, referring to FIG. 5, the camera 42 includes an RCA model SID50 that includes an array 82 of photo detectors 84 arranged in rows 86 and columns 88.
Preferably it is a four-charge coupled device. The array 82 has photo detectors 84 in 256 rows and 403 columns. The row 86 is a group of photodetector elements 84 arranged in a line perpendicular to the scanning direction (that is, the Y-direction), and the column 88 is a line parallel to the scanning direction (the X-direction). It is a group of photodetector elements 84 arranged. The length of each row 86 and each column 88 is respectively
6.5 mm and 4.9 mm. Each sensing element 84 has a length of 20 microns and a width of 16 microns. Therefore, the width of each stripe area 72 is equal to the total length of the 403 photodetection elements. Each photo-detecting element 84 receives a light beam emitted from a portion of the inspection region 36 aligned with that element through the optical window 54 and outputs an amount of electric charge corresponding to the intensity of the light beam incident on this element, that is, a measured energy value. It accumulates in the potential well of the device.

Each stripe area 72 of the inspection area 36 is divided into an array 90 composed of pixels 92, and the size of each of these pixels is made the same as that of the photo-detecting element 84 of the array 82 by the action of the lens 35. The pixels 92 of the array 90 are arranged in rows 94 and columns 96, the number of pixels in each row is 403, and the number of pixels in each column is determined by the length of the stripe area 72. The presence of light in the striped area is determined by the optical window of camera 42.
The pixel 54 is moved along each stripe region 72, and the energy value corresponding to the light intensity is obtained at each pixel 92 according to the following processing to detect.

In the X-stage 52, the side 58 of the optical window 54 is a segment 98 of the examination area 36.
Optical window 54 from the starting position 66 until X-
The scanning process is started by accelerating to the right of the direction. The X-stage 52 then moves the optical window 54 along the fringe area 72 at nominal speed.

The photodetector elements 84 in the first row 86a of the first array 82 are second
In aligning with pixel 92 in the first row 94a of array 90, the following events occur. Each photodetector 84 in row 86a
A charge is generated in the potential well of. This charge amount corresponds to the intensity of light existing in the pixel. (The potential wells of photodetector element 84 do not have any charge stored on them prior to scanning stripe region 72.) The row transfer clock signal provided to each row 86 of array 82 is Although the charges from the photodetector 84 are in the same row 88, they are next to each other,
That is, the light is transferred to the photodetector in the second row 86b. This transfer occurs when the photodetector and the pixel are aligned with each other.
(Since the X-stage 52 continuously moves the optical window 54 along the striped region 72, elephant degradation due to aliasing between adjacent rows of pixels is negligible.) Charge from row 86a to row 86b After the transfer, the accumulated charge does not exist in the potential well of the photodetector element 84 in the row 86a.

The photo detectors 84 in the second row 86b are the second row 94b of the array 90.
When aligned with pixel 92 at, the following events occur. Electric charges are generated in the potential wells of the photo-detecting elements 84 of the rows 86a and 86b. The amount of charge generated in each photodetector element 84 in row 86b is added to the charge previously transferred to that element. Therefore, the charge amount of the photodetector element 84 in row 86b represents two energy values corresponding to the light intensity present in the pixel 92 in each column of the row 94a of the Murray 90. The row transfer clock signal transfers the charges from the photodetecting elements 84 in the rows 86b and 86a to the photodetecting elements in the third row 86c and the second row 86b that are laterally adjacent to each other, though in the same column 88.

The photodetector element 84 in row 86 obtains the energy value corresponding to the light intensity of the pixel 92 aligned with this detector element (1), and that energy value is in the same column 88,
The pixel 92 is transferred to the photodetector element 84 of an adjacent row 86 that has not been previously aligned (2) The above procedure is repeated for 255 cycles of the row transfer clock signal.

After 255 such row-to-row transfers are completed,
The 256th or final row 86d of the array 82, or photodetector element in the last row, is aligned with the pixel 92 in the first row 94a of the array 90. First line 94a
The previously accumulated energy value of 255 for each pixel 92 in the 256 rows of 256 pixels acquired by each photodetector element 84 in the last row 86d.
Is added to the th energy value. Prior to the generation of the 256th row transfer clock signal, 403 corresponding to the pixel 92 of row 94a.
The energy value accumulated in each photodetector element 84 is serially read by the high speed data transfer clock signal. Pixel
Whether the accumulated energy value for 92 is converted to digital form and processed by a threshold detector to determine whether the amount of light present at each pixel 92 indicates the presence of a defect at the corresponding location on the photomask 10. To decide.

When the 256th cycle of the row transfer clock signal occurs, the energy value previously accumulated for each pixel 92 in the second row 94b becomes the 256th energy value acquired by each photodetector element 84 in the last row 86d. Is added. Before the 257th cycle of the row transfer clock signal occurs, the contents of the 403 photodetector elements corresponding to each pixel 92 in row 94b are read and processed as described above.

During each successive cycle of the row transfer clock signal, array 90
Of 256 energy values for each pixel 92 in row 94 and column 96 of the array of photodetectors 84 in corresponding columns 88 and 86 of array 82.
Accumulated in.

There are several general aspects to the accumulation of energy values that characterize the scanning method of the present invention. First, each photodetector element 84 in row 86a never stores more than the energy value for that pixel in any of the pixels 92 aligned with that element. Second, the pixel 92 in a particular 94
The photodetector elements 84 in one row 86 of the array 82 that are currently aligned with are always higher than the energy values previously accumulated in those elements, and the adjacent row previously aligned with that particular row 94 of pixels 92. It becomes higher than the energy value in the photodetector element 84 of 86. Third, each photodetector element 84 in row 86d accumulates 256 energy values corresponding to the light present in the pixel 92 with which it is aligned.

After the side 56 of the optical window 54 has passed the segment 74 of the examination area 36, the scanning of the stripe area 72 is complete and the last row 94 of the array 90 is reached.
The cumulative energy value for pixel 92 at d is array 8
This means that the data is read from the photodetector element 84 in the second last row 86d. The X-stage 52 slows down the optical window 54 and stops at the stop position 100. In FIG. 5, the optical window 54 is indicated by a broken line at this stop position. The X-stage 52 and the Y-stage 50 reverse the camera 42 back along the passage segment 76a to position the optical window 54 at the next starting position 78. Next horizontal stripe 72
During such time in preparation for scanning the
Clear the potential well of. The second and subsequent stripe region 72 scans and retrace scans are performed as previously described.

FIG. 6 is a block diagram of an apparatus according to the present invention for scanning a holographic defect image for defects. FIG. 7 is a flow chart showing the interphase relationship of various functions executed by the general purpose computer included in the apparatus of FIG. The following description of the operation of the scanning device of FIG.
This is performed by referring to the appropriate processing in the flowchart of FIG. 7 or the inset phrase in the decision block.

Referring to FIGS. 6 and 7, a general-purpose computer 202 is used as the scanning device 200, and a positioning table 4 in the XY directions is used.
In addition to controlling the operation of the array 90, the measured energy value corresponding to the intensity of the light present in the pixel 92 of the array 90 is displayed in the array 90 row.
Acquire and identify by 94 and column 96. Prior to starting the scanning process, certain initial information is placed in the memory of computer 202 (processing block 300 of FIG. 7). The initial information includes the contour and size of the inspection area 36 corresponding to a particular photomask 10, the scan and acceleration and deceleration rates of the X-stage 52, and the width 70 of the stripe area 72. The XY positioning table 48 has precision optical encoders for each X and Y stage. Each optical encoder produces one pulse for each stage of 4 micron displacement. The exact instantaneous position of the camera 42 is set by first visually aligning its optical window side 58 with the segment 98 of the examination area 36 and then recording the X-stage 52 and Y-stage 50 position coordinates. (FIG. 7, processing block 302). This reference position is input to the memory of the computer 202, and the instantaneous position of the camera 42 with respect to this reference position is monitored.

The computer 202 calculates the starting position 66 and the stopping position 100 for the camera 42 in preparation for scanning the first striped area 72 from the contour and reference position information of the photomask stored in the memory of the computer (FIG. 7, processing block). 304).
Computer 202 also calculates the position of camera 42 with respect to array 90 and generates a readable signal during this time. The readable signal allows the accumulated threshold energies to be stored in memory along with the matrix and column locations of the fringe areas from which their energy values were dictated (FIG. 7, processing block 304). ).

The commands generated by computer 202 are X-stage 52 and Y-stage 5
It is supplied to the optical encoder pulse counter 204 which forms part of the servo mechanism for the drive motor for 0. The drive motor is integrated with the positioning table 48. The above command is bidirectional in X and Y to move the camera 42 to its starting position.
Includes 0.4 micron increments. The optical encoder pulse counter 204 supplies a control signal to the X-stage and Y-stage motor drive circuit 206, which in turn drives the drive circuit 206.
Applies drive signals to X-stage 52 and Y-stage 50 to move these stages to a starting position 56 (FIG. 7, processing block 30).
6). The optical encoder pulse counter 204 counts the pulses generated at the X and Y direction optical encoder starting terminals included in the positioning table 48, and the number of counted pulses corresponds to the increment number of 0.4 micron increments supplied from the computer 202. Continue to supply drive signal. The optical encoder pulse counter 204 has the camera 42 at the starting position 66.
The supply of the drive signal to the motor drive circuit 206 is stopped immediately after reaching.

To start scanning the first striped area 72, the computer 20
2 supplies a command to the optical encoder pulse counter 204 and X
-Accelerate the movement of the step 52 to the side 58 of the optical window in the camera 42
When is aligned with the segment 98 of the inspection area 36, the speed is brought to the nominal constant speed. X-stage 52 leaves camera 42 at position 66
When moved from, the optical encoder pulse counter 204 counts the pulses generated by the X and Y (for direction) optical encoders and outputs the X optical encoder pulses to the 1 / N counter 2.
Supply to 08. This counter produces one pulse at the starting terminal of the counter every time five pulses appear at the starting terminal of the X optical encoder (ie N = 5). The row counter 210, which is part of the camera controller 212, is a 1 / N counter 2
Receive and count the pulses that appear at the 08 departure terminal. The general purpose computer 202 provides count information corresponding to the beginning and end of the optical window 54 in the striped area 72 to the row counter 210, which can send the energy values obtained in the striped area to memory. As soon as the counter reaches the first count value, the row counter 210 issues a readable command, which allows the cumulative energy value to be stored in memory.

Each pulse appearing at the starting terminal of the 1 / N counter 208 corresponds to a 2 micron increment of the optical window 54 during scanning of the fringe area 72. In plane 40, the image that appears below lens 35 represents ten times the size of the photographic image of photomask 10, so that a 2 micron displacement of optical window 54 in the X-direction will cause X in plane 40. -Corresponds to 20 micron increments in the direction. This 20 micron distance corresponds to the 20 micron length, or row width, of pixel 92. Therefore, the pulse appearing at the starting terminal of the 1 / N counter 208 indicates that the optical window 54 is displaced by the length of one pixel 92, that is, the width of one row in the X-direction.

Therefore, the row counter 210 counts the number of rows 94 of the array 90 that the camera 42 has traversed. The camera controller 212 also includes a row transfer clock generator 214 and a data transfer clock generator and pixel counter 216. As the optical window 54 of the camera 52 traverses the defect image field 32 in the plane 44 at a rate of 29 mm / sec, the row counter 210 generates a row transfer clock in response to each pulse appearing at the starting terminal of the 1 / N counter 20. To each row of the photodetector element 84, which clocks a row transfer clock signal of 1 MHz.
Provides one three-phase clock cycle of 86. After the completion of each row transfer clock cycle, the data transfer clock generator and pixel counter 216 generates an 8 MHz data transfer clock signal, which provides row 86d of photodetector element 84 with 403 three-phase clock cycles.

The row transfer and data transfer clock signals are generated by the stripe area 72.
It is clear that the scanning is synchronized with the displacement amount of the X-stage 52 during the scanning period. For this reason, the row transfer and data transfer clock signals are only generated after the X-stage 52 has traveled a distance of 2 microns along the plane 44 (regardless of the time required for this travel). By such synchronization, even when the scanning speed of the X-stage 52 changes from its nominal value due to, for example, an externally generated vibration, the photo-detecting element 84 and the corresponding pixel 92 are aligned. Will always be maintained. Such features are important in determining the exact location of sub-micron sized defect information in the light pattern.

During the first 255 cycles of the row transfer clock signal, 255 pixels 92 in each of the 403 columns 96 of array 90 are scanned. The energy values of the 255 scan pixels 92 in each column 96 of the array 90 accumulate in the photo-detecting elements 84 of the array 82 as described above. After the 255th cycle of the row transfer clock signal, the accumulated 256 energy values for the first row 94a of pixel 92 appear in row 86d of array 82. The row counter 210 then produces a readable signal,
This allows the memory unit 218 to receive, in digital form, the accumulated energy value stored in column 88 of photodetector element 84 for row 94a of pixel 92. (FIG. 7, decision block 310). As mentioned above, the row counter 210 receives the beginning and ending row pulse counts from the computer 202 and the readable signal is generated at the beginning and ending row pulse count times. The storage capacity of the memory unit 218 is 512 kilobytes.

The energy value read from row 86d of photodetector elements 84 is appropriately adjusted and amplified by preamplifier 220, and
Converted to digital form by analog-to-digital converter 222. The energy value represented by each digital word is provided to the threshold detector 224 for identification. Only the digital world of pixels 92 whose accumulated energy value is above the expected threshold level of threshold detector 224 are stored in memory unit 218 and the remaining digital words are discarded. Pixel 92 is identified in memory unit 218 by its position in row 94 and column 96 of array 90, the row is identified by the output of row counter 210, and the column is when continuous energy values are read from row 86d of array 82. It is identified by the pixel number recorded in the data transfer clock generator / pixel counter 216. The computer 202 inputs into its internal memory a data word identifying the striped area and retrieves the stored pixels from the data.

The data transfer clock generator and pixel counter 216 provides 403 cycles for each cycle of the row transfer clock signal of 255 or more until the cumulative energy value for all pixels 92 in the striped area 72 is read from row 86d of array 82. Continue to provide the data transfer clock signal. Side 56 of optical window 54
After passing through the segment 74 of the examination area 36, the row counter 210 ceases to generate a readable signal, which means that the energy value of the pixel 92 in the row 94d of the array 90 is in the array 8.
It is indicated that the data is read from the photodetector element 84 in the second row 86d (FIG. 7, decision block 312). At the end of the readable signal, the optical encoder pulse counter 204 causes the camera 4
The movement of 2 is decelerated and the camera is stopped at the stop position 100 (FIG. 7, decision block 314). The data captured by the memory unit 218 for the first striped area 72 is read and processed by the general purpose computer 202 (FIG. 7, processing block 316).

After the camera 42 reaches the stop position 100, the computer 2
02 calculates the start position 78 and the stop position 102 of the next adjacent side, that is, the second striped area 72 (FIG. 7, processing block 30).
Four). The computer 202 returns the camera 42 back to the second starting position 78 along the return path segment 76a and then begins scanning the adjacent striped area 72 in a manner similar to that described above (FIG. 7, processing block 306). And 308). The scanning of the inspection area 36 is continued until the energy values have been captured for all pixels in each stripe area 72.

The present invention is not limited to the examples described above, and it goes without saying that many modifications can be made.

[Brief description of drawings]

FIG. 1 is a diagram showing an array-type photomask applied to the present invention; FIGS. 2A and 2B are respectively for producing a hologram of a photomask pattern and reproducing a defective optical image corresponding to the photomask from the hologram. 3 is a diagram showing an optical processing device; FIG. 3 is a perspective view showing a scanning mechanism for detecting the presence of a defective portion in the photomask of FIG. 1 in the holographic optical image field; FIG. 4A is the light of FIG. 3 in the lower left corner of the statue field
FIG. 4B is an enlarged diagram of two stripe areas; FIG. 4B is an enlarged raster area of the stripe areas of FIGS. 3 and 4A, and a raster scan path is followed by the photodetection stage to detect a defective image in the optical image field according to the present invention. FIG. 5 shows a second array of pixels in the optical image field,
FIG. 6 is a diagram showing a first array of photodetection elements of a charge-coupled device used in the present invention; FIG. 6 is a block diagram showing an example of an apparatus for scanning an optical image field for a defective portion in the photomask of FIG. 1 according to the present invention. Diagram; FIG. 7 is a flow chart showing the function of defect image processing executed by a general-purpose computer incorporated in the apparatus of FIG. 10 ... Photomask, 10 '... Defect image 12 ... Photo image element, 22 ... Imaging lens 28 ... Hologram, 30 ... Spatial filter 32 ... Defect image field, 35 ... Lens 36 ... Detection area 42 ... Photodetector (camera, charge-coupled device) 48 ... Positioning table, 54 ... Optical window 72 ... Stripe area, 82 ... Photodetector array 84 ... Photodetector, 90 ... Pixel array 92 ...... Pixel, 200 ...... Scanning device 202 …… General-purpose computer 204 …… Optical encoder pulse counter 206 …… Motor drive circuit, 208 …… 1 / N counter 210 …… Line counter, 212 …… Camera controller 214 …… Row transfer clock generator 216 …… Data transfer clock generator and pixel counter 218 …… Memory unit, 220 …… Preamplifier 222 …… Analog-digital converter 224 …… Threshold detector

Claims (6)

[Claims] 1. A method for detecting the presence of a defective portion in a test body by a light pattern having a first width corresponding to the test body, wherein the light pattern has a different light intensity in a defective region having a predetermined light intensity. A method for detecting a defect in a test object having in the background is: preparing a photodetector of a charge coupled device for detecting light from a window region of the light pattern, the window region being parallel to the first width. Has a second width narrower than the first width, the photodetectors comprising a plurality of photodetector elements arranged in a first array of rows and columns, each photodetector element having a respective second photodetector element. A photodetector preparation step operative to provide a measured charge value corresponding to the intensity of light received by the photodetector; and the photopattern defining a plurality of adjacent striped regions of a second width. Optical pattern the photodetector in the scanning direction at right angles to the first and second widths. Repeatedly scanned with respect to each other, each striped region including a plurality of pixels arranged in a second array of rows and columns, the rows of the first and second arrays extending in a direction perpendicular to the scanning direction. And the columns extend along the scanning direction, and the pixels are defined as respective positions in the light pattern which are sequentially imaged on the respective photodetecting elements in each column of the first array, the columns of the first array A scanning step in which each photodetector element in step S1 sequentially acquires a charge value corresponding to a pixel; and the charge value measured by the photodetector element to which the pixel is directed is synchronized with the scanning of the photodetector for the light pattern. Therefore, the adjacent photodetecting elements in the columns of the first array are shifted in the direction opposite to the scanning direction along the columns of the first array according to the row transition clock signal, and the photodetecting elements in the columns of the first array are shifted. Many equal to the number of A charge shift step of giving a total charge value representing the sum of the charge values of the photodetector to the last row of the photodetector elements; A comparison step of determining whether or not the light intensity received from the pixel represents a defective portion in the test object by comparing the charge value of the threshold value; Storing in a memory together with information for identifying the position of the container and the column of pixels, and a method of detecting a defective portion of a test object. 2. The pixels are sequentially connected onto the photodetector elements in the columns of the first array by stationary the light pattern and continuously moving the first array of photodetector elements along the striped areas of the light pattern. A method according to claim 1, characterized in that it is imaged. 3. A method according to claim 1, characterized in that the defect in the test specimen is represented by a total charge value which is greater than a threshold charge value. 4. An apparatus for detecting the presence of a defective portion in a test body by a light pattern of a first width corresponding to the test body, wherein the light pattern has a different light intensity in a defect area having a predetermined light intensity. An apparatus for detecting a defective portion of a test body, which is configured to have in a background, is: Photodetection for detecting light of a light pattern from a window region having a second width narrower than the first width in parallel with the first width. A photodetector of a charge-coupled device comprising a plurality of photodetector elements arranged in a first array of rows and columns; an optical pattern being imaged on the photodetector elements, each photodetector element being An optical element operable to supply a measured charge value corresponding to the intensity of the light received by each of these photo-detecting elements; defining a plurality of adjacent striped areas of second width in the light pattern. To scan the light in the scanning direction perpendicular to the first and second widths. Out device is repeatedly scanned to light pattern includes a plurality of pixels each banded region is arranged in a second array of rows and columns,
Rows of the first and second arrays extend in a direction perpendicular to the scanning direction and columns extend along the scanning direction;
Further, the pixels are defined as respective positions in the light pattern sequentially imaged on the respective photodetection elements in each column of the first array, and each photodetection element in the column of the first array provides a charge value corresponding to the pixel. Scanning means adapted to sequentially acquire; a charge value measured by a photodetector to which the pixel is coupled,
First in synchronization with the scanning of the photodetector for the light pattern
The adjacent photodetecting elements in the array are shifted in the direction opposite to the scanning direction along the columns of the first array to generate the first
Means for forming a column transfer clock signal to provide a total charge value to the last row of photodetectors representing a sum of a number of charge values equal to the number of photodetectors in a column of the array; Data transfer signal forming means for transferring the total charge value; comparing the total charge value with a charge value of a preselected threshold value;
Comparing means for determining whether the light intensity of the pixel represents a defect in the specimen; the total charge value of the pixel representing the defect in the specimen together with information identifying the position of the photodetector and the row of pixels. A means for storing in a memory; and a defect detecting device for a test body, comprising: 5. The light pattern remains stationary and the scanning means also includes positioning means such that the light detection means traverses each of the striped areas and continuously scans the striped areas. The device according to claim 4, characterized in that 6. The apparatus according to claim 4, wherein the defect in the test piece is represented by a total charge value larger than a threshold charge value.