TW200535453A - Exposing apparatus - Google Patents

Abstract

In this exposure device 1, a laser beam radiated from an exposure optical system 3 relatively scans a glass substrate 8 in the direction orthogonal to the movement direction (arrow A) thereof to draw the function patterns with a predetermined pitch directly on the glass substrate 8. The exposure device 1 comprises an imaging means 5 for imaging the pixels of a black matrix previously formed on the glass substrate 8; and an optical system control means 7 for carrying out a predetermined image process to the image data of the pixels acquired by the imaging means 5, for forming a defect-free pixel row image by removing defects in a pixel row image of one scan region according to the laser beam, for detecting the reference position of the exposure start or end in the defect-free pixel row image, and for controlling the irradiation start or stop of the laser beam with reference to the reference position.

Description

200535453 V. Description of the invention (1) [Technical field to which the invention belongs] The present invention relates to an exposure device that exposes a functional pattern on an object to be exposed, and particularly relates to an exposure device obtained by using a photographing mechanism on the object to be exposed. A pre-formed reference position set on a functional pattern that is used as a reference is detected and detected. The reference position is used as a reference to control the start or stop of irradiation of the light beam in order to improve the accuracy of the functional pattern registration. At the same time, the cost of the exposure device is suppressed. : [Prior art] A conventional exposure device uses a mask pattern previously formed with a functional pattern corresponding to a glass substrate _ as a mask, and the mask pattern is transferred to the exposed body for exposure. For example, steppers (S t e p p e r), micro symmetric projection (Micro Mirror Projection) or proximity function (Proximity) are all available. However, when these conventional exposure devices are laminated to form a plurality of layers of functional patterns, the accuracy of the functional patterns of the layers is very problematic. Especially for the large masks used in the formation of large LCD TFTs and color filters, the pattern arrangement of the masks must be required to have an absolutely accurate size, which increases the cost of the masks. In order to obtain the above-mentioned coincidence accuracy, it is necessary to align the functional pattern of the bottom layer and the pattern of the photomask, especially for large photomasks, this kind of accurate operation is difficult. On the other hand, there is an exposure device that does not use a photomask and directly draws a CAD data pattern on an object to be exposed using an electric beam or a laser beam. This exposure device has a laser light source, an exposure optical system for reciprocating scanning of the laser light source _, and a state of carrying an exposed object.

Page 5 200535453 V. Description of the invention (2) According to the CAD data, the conveying mechanism controls the emission status of the laser light source, scans the laser beam back and forth, and simultaneously conveys the exposed object in the scanning direction of the laser beam. In the orthogonal direction, a CAD data pattern corresponding to a functional figure is formed in a two-dimensional manner on the object to be exposed (for example, refer to Japanese Patent Application Laid-Open No. 2000-1-1 4 4 4 1 5). However, in such a direct drawing type conventional exposure device, the arrangement of CAD data patterns must have absolute dimensional accuracy, which is the same as the exposure / light device using a photomask. In the process of forming a functional pattern using a plurality of exposure devices, if the accuracy of the exposure devices varies, the accuracy of the coincidence of the functional patterns may decrease. Therefore, in order to cope with this problem, a high-precision exposure device is required, and the cost of the exposure device is increased. In addition, it is necessary to align the functional pattern and CAD data pattern of the bottom layer in advance, which is the same as that of other exposure devices using a photomask. There are the same problems as above. SUMMARY OF THE INVENTION An object of the present invention is to provide an exposure device which can cope with the above problems to improve the registration accuracy of a functional pattern, and at the same time, can suppress the cost of the exposure device. φ In order to achieve the above object, the exposure apparatus of the first invention is an exposure apparatus that relatively scans the light beam irradiated by the exposure optical system in a direction orthogonal to the direction of exposure and the movement of the body, and on the subject to be exposed Those who expose the functional pattern at a predetermined pitch include a photographing mechanism that can take a reference functional pattern formed in advance on the exposed body as a reference exposure position; and

Page 6 200535453 V. Description of the invention 3) Optical control mechanism for correcting ^^

The type of portrait data is processed according to the standard functional pattern of the reference machine obtained by the shadow mechanism in the field. The second processing is to exclude the light that corresponds to the light. Non-defective: = reference drinking shift, trap reference function diagram image at the beginning or end of exposure. The light beam is started or irradiated on the portrait image, and the reference position is used as a reference to control. . 'I Bismuth LV 氺 氺 氺 氺 氺 枇 枇 丨 丨 All the data of the reference functional pattern obtained by Xie 7 structure are processed by the "" system for predetermined image processing and excluded ;: the reference functional pattern in the field of f image beam scanning Defects of daylight image, based on :; light reference functional pattern daylight image, the reference position where the start or end of exposure is detected on the non-defective reference functional pattern, ==, and the reference position ^ is entered in two The irradiation of the light 'I is controlled to start or stop. With this, even if the reference functional pattern pre-made on the object to be exposed is defective, it is possible to form a high-precision / predetermined functional pattern at a predetermined position. Therefore, in the case of forming a functional pattern of several layers, the combination of the functional patterns of each layer can also be improved. Furthermore, in the case of forming a multi-layer pattern using a plurality of exposure devices, I ^ is used to eliminate the cause of the problem; the problem of the low accuracy of the coincidence of the functional patterns caused by the accuracy difference between the exposure devices, and the cost of the exposure device is suppressed. Kou Yueyou, the 4-light local control mechanism, uses the two portrait data in the basic functional pattern obtained by the photographic mechanism to select the theoretical data of each portrait data to make the non-defective reference functional map. Type portrait. Therefore, the two daytime image data in the reference functional map obtained by the photography agency were used to obtain the theoretical sum of each daytime image data by the optical system 1 control mechanism to create a defect-free reference functional pattern portrait.

You 1

200535453 V. Description of the invention (4) Furthermore, the optical system control mechanism uses the previous reference functional pattern image data obtained by the photographing mechanism and the newly obtained reference functional pattern image data to obtain the object located in the exposed body. Theoretical sum of each image data in the same position in front and back of the moving direction to make a defect-free reference functional image portrait. Therefore, the previous reference functional pattern day image data obtained by the photographing agency and the newly obtained reference functional pattern portrait data are used to obtain each of the same positions in front and back of the exposed body by the optical system control mechanism. Painting_Theoretical sum of image data to make a defect-free reference functional portrait. By doing so, defects such as foreign matter adhering to the platform carrying the object to be exposed can also be removed. This prevents exposure due to the mistaken recognition of the defect as a reference position, so that the exposure accuracy of a predetermined functional pattern can be improved. In addition, the optical system control mechanism uses adjacent reference functional pattern image data among the reference functional pattern image data obtained by the photographing mechanism to obtain the theoretical sum of each day image data, and thus creates a defect-free reference function. Sexual pattern portrait. In this way, the day-to-day reference functional pattern day image data of the reference functional graphic image obtained by the photographing mechanism is used, and the theory and sum of each image data are obtained by the optical system control mechanism to create a defect-free reference functional pattern day image. image. Therefore, even if there is a defect in the beam scanning area of the last column, the two adjacent functional patterns can be compared to remove the defect. φ In accordance with the second embodiment of the exposure apparatus of the present invention, a beam scanned by the exposure optical system is relatively scanned in a direction orthogonal to the moving direction of the object to be exposed, and the object is exposed at a predetermined pitch on the object. From the functional pattern of exposure. Including a photographing mechanism for taking a reference functional pattern formed in advance on an object to be exposed as a reference light position; and an optical control device

Page 8 200535453 V. Description of the invention (5) '-;--: Guiyu copied the reference functional map of the established field obtained by the photography > organization ^ ^ Γ image data and copied it to the established field Follow-up field, and complement the reference functional pattern image that was not obtained by the ί Ϊ mechanism, and detect the reference position at which the exposure starts or ends on the second image of the complement 2> 2 I, . 1 is used as a reference to control the start or stop of irradiation of the beam 2 Φ Α, using the optical control mechanism to copy the day-to-day image data obtained by the photography mechanism from the previous functional standard pattern to the predetermined field The daylight image is supplemented with a reference functional map that the photography agency has failed to obtain. Or it is supplemented with a reference functional map. The daylight image is detected and the exposure starts and the exposure starts. The beam scanning side 1 that controls the light beam as a reference until the emission stops. Therefore, the imaging mechanism can also be used to scan the collar when the photographic mechanism cannot obtain all the high-accuracy reference patterns formed on the exposed object. ', The nature of the machine is in a predetermined position. If this is the case, for example, the photographic field of a photo camera's I! Domain photography agency is relatively small, and the full scan device cost image data can be made in a complete form based on the image data of the image in the photographic field. It can be reduced by reducing the number of photographing agencies. • Implementation method] 1 p ^ i, device 1 is a conceptual diagram of the embodiment of the exposure device of the present invention. Exposure light source 2, for exposing the functional pattern on the object to be exposed, and equipped with a thunder and light mechanism, north exposure optical system 3, transport mechanism 4, photographing mechanism 5, as a lunar surface light irradiation mechanism 6, And optical system control mechanism 7. Again

200535453 V. Description of the invention (6) The functional pattern refers to the pattern of the necessary components of the product for its original purpose, such as the black matrix pixel pattern of color filters, and red, The patterns of the blue and green color filters are wiring patterns and various electrode patterns in the case of semiconductor components. In the following description, a glass substrate for a color filter as an exposed object will be described as an example. The above-mentioned laser light source 2 is used to emit a light beam, for example, it belongs to a fully fixed mode-locked laser light source that generates 3 5 5nm ultraviolet rays and outputs a high output of more than 4W. -An exposure optics 3 is provided in front of the beam emitting direction of the laser light source 2 described above. This exposure optical system 3 is used to scan the laser beam as a light beam on the glass substrate 8 reciprocatingly. It has a light switch 9 and a light deflection mechanism 10 in order from the front of the laser beam emission direction. A mirror 11, a polygon mirror 12, an f0 lens 13, and a second mirror 14. The optical switch 9 is used to switch the irradiation and stop states of the laser beam. For example, as shown in FIG. 2, the first and second polarizing elements 15A and 1 5B are separately arranged so that the polarization axes P of the respective polarizing elements 15 A and 1 5 B are orthogonal to each other (the polarizing element 1 in the figure). The polarizing axis P of 5 A is set in the vertical direction, and the polarizing axis P of polarizing element 1 5 B is set in the horizontal direction). Between the two polarizing elements 15 A and 1 5 B, an electro-optic modulator 16 is arranged. When the electro-optical scheduler 16 φ is applied with a voltage, it operates to rotate the polarization surface of polarized light (linear polarized light) at a high speed of a number n e s c. For example, when the applied voltage is zero, in the same figure (a), the linearly polarized light having a polarizing surface, such as a vertical direction, which is selectively transmitted by the first polarizing element 1 5 A, passes through the electro-optic modulator 16 as it is, and Up to the > second polarizing element 15B. Since this second polarizing element 15B is formulated to enable

Page 10 200535453 V. Description of the invention (7) ~-The linearly polarized light with the horizontally polarized surface is selectively transmitted. Therefore, the linearly polarized light with the vertically polarized surface cannot be transmitted. In this case, the belt is in the irradiation stop state. Tian, on the other hand, as shown in Figure (b), a voltage is applied to the electro-optic modulator 16 and when the linearly polarized polarization plane incident on the electro-optical modulator 6 is rotated 90 degrees, it has a vertical The linearly polarized light of the directional polarization plane becomes a horizontally polarized light plane when the air optical scheduler 16 is issued, and the linear-polarized light can pass through the second polarizing element 15β. As a result, the laser beam is irradiated. ", Shifted from the top and the 1st, for example, the polarizing mechanism 10, is used to orthogonalize the scanning direction of the laser beam scanning position I (the direction of the arrow A in the movement pattern of the glass substrate 8 is the same) , And adjusted to scan the acoustic optical element (A0 element) at the correct position. It is also used to direct the laser beam passing through the polarizing mechanism 10. Bend to the installation direction of the polygon mirror 12 described later, = plane mirror In addition, the polygon mirror 12 can scan the laser back and forth. For example, the regular octagonal columnar body has 8 mirrors. A reflected laser beam is scanned along with the direction of multiple preparation, seeing, and going. , 俟 Laser beam ': rotation' returns to the complex direction between -dimensional, and again with more :; shot: move to the lower-mirror surface, and scan in the direction of "quot 角 开 7 镜 12 之" Rotate and start-the dimensional input plate 8 becomes constant velocity, and is configured so that its beam scanning speed is slightly the same at the base of the glass. The second mirror 4f and the polygon mirror 12 are known Reflection through f 0 lens 1 3

$ 11 pages 200535453 V. Description of the invention (8) ~~ 1 * The laser beam makes it slightly perpendicular to the 8 sides of the glass substrate and belongs to a flat mirror. In addition, on the part near the emission side of the lens 13 toward the scanning beam scanning start side, a line sensor 17 is provided to scan the% ^ to detect the laser beam and set the scanning position and travel. The deviation of the two scanning positions, and the laser beam scanning start time is detected at the same time: In addition to the sensor side, the sensor 17 can be set at any position as long as it can detect the laser beam knowing the starting point of the laser beam, such as: Field and glass substrate transfer platforms can be set at 8 sides. Behind this second mirror 丨 4, there is a transport mechanism 4. This moving plate 8 is placed on the platform 18, and the laser beam is drawn at a predetermined speed in the direction of being a positive father. It has useful functions. The mobile platform worker, for example, transports the transport roller i 9 and drives the transport roller i 9; the transport drive unit 20. 1 J Goma is moved above the conveying mechanism 4, and a photographing mechanism 5 is provided on the front side of the beam position in the conveying direction as shown by the arrow 10. This photography is used to capture the black matrix pixels of the reference machine 'and the functional pattern previously formed on the broken glass substrate 8 as the exposure reference, such as CCDs with light receiving elements arranged in a row *. Here, as shown in FIG. 3, the distance D between the photographing position e of the photographing mechanism 5 and the beam scanning position F is set to an integer multiple (n times) of the pitch P arranged in the direction of the black matrix 21 蚩 #. By this, when the broken glass = plate 8 is carried to the center of the pixel 22 and the scanning position of the laser beam coincides, the laser beam starts scanning. The smaller the distance D, the better. The position of the laser beam scanning element 22 can be accurately determined by reducing the movement error of the glass substrate 8 '. In addition, in Figure i, three cameras are provided as an example:

200535453 V. Description of the invention (9) '-~ If the laser beam scanning range is narrower than the image processing field of the photography mechanism 5, using one photography mechanism 5 is fine. If the scanning range is wider than the imaging mechanism 5-Taiwan's In the field of day image processing, multiple cameras should be set up5. ≫ A back light irradiation mechanism 6 is provided below the conveying mechanism 4. The back light illuminating mechanism 6 is used to illuminate the daylight element 22 to enable the photographing mechanism 5 to take a picture, such as a surface light source. The optical system control mechanism 7 is connected to the laser light source 2, the optical switch 9, the light deflection mechanism 10, the polygon mirror 丨 2, the line sensor 丨 7, the transport mechanism 4 and the shadow mechanism 5 ° This optical system control mechanism The 7 series is used to detect the reference position of the pixel 2 2 taken in advance in f ~ mechanism 5 and to control the laser light source 2 laser beam irradiation start based on the ° reference position. Or “irradiation stop” at the same time according to the output of the line sensor 17 to control the voltage applied to the light deflection mechanism 10 to shift the laser beam emission direction, and to control the speed of the laser beam to maintain the laser beam scanning speed If the glass substrate 8 is transported at a predetermined speed by the transport mechanism 4 at a predetermined speed, one of them is provided with a light source driving section 23 for igniting the laser light source 2 to control the start and stop of the irradiation of the laser beam. The light switch controller 24 is used to control the light deflection mechanism driving portion of the laser beam deflection vector of the 2 5 f direction mechanism 10, and the polygon driving portion 2 5 B is used to control the driving of the polygon mirror 12. f ^ Handling speed of the handling means 4 The controller 26 is used for controlling the back light controller 27 for turning on and off the light-transmitting mechanism 6 and is used for the A / D conversion section 28 ′ of the image taken by the camera 5. The image processing unit 2 9 is used to determine the laser beam irradiation start position and the irradiation stop position according to the image data based on the warp and stop, to memorize the image processing unit 29

Page 13 200535453 V. Description of the invention (10) The data of the laser beam irradiation start position (hereinafter referred to as the exposure start position) and the irradiation stop position (hereinafter referred to as the end of exposure position) are obtained. A memory unit 30 such as an observation table is used to prepare modulation data for the ON / OFF light switch 9 based on the exposure start position and end exposure position data read out by the memory unit 30. The modulation data creation processing unit 3 1, and a control section 32 ° for appropriately controlling the entire device to operate for a predetermined purpose. Figures 4 and 5 are block diagrams showing an example of the configuration of the image processing section 29. As shown in FIG. 4, the image processing unit 29 is provided with, for example, three ring buffer memories 3 3 A, 3 3 B, and 3 3 C in parallel; , 3 3C, for example, three line buffer memories 34A, 34B, and 3 4C; connected to the line buffer memories 34A, 34B, and 34C, and compared with a predetermined angle value to output binary grayscale data Comparison circuit 3 5; It is used to compare the output data of the above nine line buffer memories 34 A, 34B, and 34C with the first reference position portrait equivalent to the exposure start position determined from the memory section 30 shown in FIG. 1 Observation table for data (LUT for exposure start position), when the two data are consistent, the exposure start position determination circuit 36 that outputs the determination result of the exposure start position; and used to compare the above nine line buffers & 34A, 34B, 3 4C The output data of 与 is obtained from the memory section 30 shown in the figure. 二 • When determining the end position of the second reference position, the image data of the second reference position (LUT for the end of exposure position) is used to output exposure when the two data match. τ, α • Position determination circuit 37 determines the exposure end of the determination result. As shown in FIG. 5, the image processing unit 29 is provided with a number of images corresponding to the first reference position and a full exposure image, which is used to determine the starting position of the light.

Page 14 200535453 V. Description of the invention (11) Counting circuit 3 8A according to the number of coincidences; used to compare the output of the counting circuit 38 A with the exposure start day number from the memory section 30 shown in Figure 1, two data When they are consistent, the exposure start number is output. The modulation data shown in Figure 1 is used to create the comparison circuit 3 1 of the processing unit 31. The input is used to input the final exposure position determination result and calculate the image data corresponding to the second reference position. The counting circuit 38B for the number of times; used to compare the output of the counting circuit 38B with the end-of-exposure pixel number 'obtained from the memory portion 30 shown in FIG. 1 when the two values match, and the end-of-exposure / signal is output as shown in FIG. 1 The modulation data is made into a comparison circuit of the processing unit 31 / 39B; a previous Λ pixel counting circuit 4 0 for calculating the previous pixel number based on the output of the counting circuit 38A; and a comparison circuit for comparing the previous pixel counting circuit 40 0 The exposure pixel row number obtained from the memory unit 30 shown in FIG. 1 is output. When the two values are consistent, the exposure pixel row designation signal is output to the comparison circuit 4 1 of the processing unit 31 as the modulation data shown in FIG. 1. . In addition, the counting circuits 38A and 38B are reset according to the reading start signal when the camera 5 starts reading operation. In the previous pixel calculation circuit 40, when the predetermined exposure pattern has been formed, it is reset according to the exposure pattern end signal. Next, the operation and pattern formation method of the exposure apparatus 1 configured as described above will be described below. First, once the power is turned on in the exposure device 1, the optical system control mechanism 7 is driven. The laser light source 2 is activated to emit laser light φ. At the same time, the polygon mirror 12 starts to rotate and the laser beam is scanned. However, at this time, the optical switch 9 is still in the OFF state, and the laser beam is not irradiated yet. • Next, the glass substrate 8 is mounted on the platform 18 of the conveyance mechanism 4. In addition, since the transport mechanism 4 is used to transport the glass substrate 8 at a constant speed, the movement direction of the station 18 is shown by the scanning trajectory (arrow B) of the laser beam as shown in FIG. 6.

Page 15 200535453 i. Description of the invention (12) (arrow A) is relatively oblique. When the glass substrate 8 is set in parallel with the moving direction (arrow A), as shown in FIG. 6 (a), the scanning start pixel 2 2 a and the end pixel of the black matrix 21 are exposed at the exposure position. 2 2b is biased. In this case, as shown in Fig. 6 (b), the glass substrate 8 is tilted to the conveying direction (arrow A direction), so that the arrangement direction of the pixels 22 and the laser beam trace (arrow B) — Just get it. However, in reality, the laser beam • scanning speed is much faster than the conveying speed of the glass substrate 8, and the deviation is not large. The glass substrate 8 is set in parallel with the moving direction, and the deviation is measured based on the data captured by the camera / mechanism 5. The deviation of the light can be corrected by controlling the light deflection mechanism 10 of the exposure optical system 3. In the following description, it is assumed that the deviation of the upper number is negligible. Next, the conveyance drive unit 20 is driven to move the mobile station 18 in the direction of arrow A in FIG. 1. At this time, the conveyance driving unit 20 is controlled by the conveyance controller 26 of the optical system control mechanism 7 at a constant speed. Next, when the black matrix 21 formed on the glass substrate 8 reaches the photographing position of the photographing mechanism 5, the photographing mechanism 5 starts photographing, and the exposure start position and the end of exposure are detected based on the captured image data of the black matrix 21 . The pattern forming method will be described below with reference to the flowchart shown in FIG. 7. First, in step S1, a pixel #day image of the black matrix 21 is obtained by the photographing mechanism 5. The acquired image data is sent to the image processing unit 2g shown in FIG. 4, and three ring buffer memories 3 3A, 3 3B, and 3 3C are processed. Then, the three new data are fetched and output from each of the ring buffer memories 33A, 33B, 33 (.). At this time, for example, the first two data are output from the ring buffer memory 3 3a, and output from the ring buffer memory 33B. Previous data, self-circulating buffer memory

200535453 V. Description of the invention (13) The body 33C outputs multiplication new data. Furthermore, each of the buffer memory 3 4 A ′ 3 4B, 3 4C is configured as, for example, 3 x on the same time axis. As a result, for example, the day image as shown in FIG. 8 (a) is digitized, that is, the day image corresponding to the 3 x value of the same figure (b) is arranged on the same time axis, and can be binarized by comparison of boundary values. . For example, the critical values are, 4, " are binarized into those as in the same figure (c). Next, in step S2, the exposure start and -position are detected. Specifically, the detection of the reference position is performed at the start of exposure, and the binary data is compared with the data obtained from the LUT for the start position shown in the figure. For example, to designate the first reference position of the exposure start position, the left light of the black matrix 2 1 in the black matrix 21 shown in Fig. 9 (a) starts to be LUT, as shown in Fig. (B), as "_〇no" . Therefore, the above two values: According to the comparison of the first LUT data and '0 0 00 1 1 0 1 1 〃, the image data taken by the two data and 5 is determined as the first reference position, and the position determination circuit 36 outputs the start position determination 1 result. It shows that there are six celestial elements 2 2 juxtaposed, then each celestial element 2 2 is left-reference position. In addition, as shown in FIG. 11, when the platform 8 or the glass object or the scar is taken by the photographing agency 5 in the picture 22 of the pixel 4 2, the defect 4 2 may be mistaken for the standard invention. In one embodiment, the images obtained by the photographing mechanism 5 are shown by the CCD pixel images in three rows 3, respectively. If 3 value. These meridian comparison circuits 3 5 and Pro are the same as the reference position determination circuit 3 6 at the end of the daylight image exposure in the same figure (a). The exposure setting of the memory 30 is set at the upper corner, and the exposure starts with the LUT data. When it is used for the start of exposure, the photographing mechanism from the beginning of the exposure, as shown in FIG. 10, should belong to a defect position such as a foreign object or a foreign object on the substrate 8. So in this portrait of Prime L1

Page 17

200535453 V. Description of the invention (14) The data is stored in the memory section 20. Therefore, when taking the p * data, the day element row is read from the memory unit 20: the number of day images of the pixel row L2 (a). The image of the day image processing unit 29 shown in the figure, to the image data, is like ^ ;? 98 at the same position in front and back of each other in the same direction (arrow A squares sum up. At this time, the portrait data of the two portraits 2 2 are the same, which is extremely rare, so take a long time. ^ 9 Simultaneous existence of defect 42 ^ α & Prime 2 2 The theory of silly numbers can remove defect 42 from portrait data. An image like sigh and sorrow, like image data, detect the benchmarks as described above Position, use non-defective pixels to arrange the day of the day? Is it possible to obtain a new day of the day? • 5: The above method removes the defect 42. Adjacent pixels may be taken at this time The 1% image data of 22 is correct, and the defect 42 day image is removed. Specifically, as shown in Fig. 1 (b), the last column of pixel images is obtained, and the pixel image "day image is listed" The day image with a direction shift of one pitch, the image data of day element 22 located at the same position in front and back of the glass substrate 8 moving direction (direction of arrow A), take its theory and At this time, it is very rare that the defect 42 exists at the same position of the adjacent two-day pixel 22. At the same time, the theory of borrowing the image data of each pixel 22 and 'even the last pixel row Ln can be removed Defects in the image 4 2. So you can make defect-free pixel rows.

Secondly, according to the result of the above judgment, the counting circuit 38A shown in FIG. And the number of counts is compared in the comparison circuit 3 9 A with the exposure start pixel number obtained from the memory section 3 0 shown in FIG. 1. When the two values match, the exposure start signal is output to the adjustment shown in FIG. 1 Variable data into processing unit 31. At this time, as shown in FIG. 10, for example, in the laser beam scanning direction, the upper left corner of the first pixel 221 and the fourth pixel 224 is set in the drawing direction.

Page 18 200535453 V. Description of the invention (15) ---- If it is set as the reference position, then the camera corresponding to the reference position of the i% den / ^% one reference position of the class 5 will be assigned to the component address of the CCD For example, '] ηπΛ // π π π-^ i 0 〇〇 ”, 4 0 0 0 is stored in # 开关 开关 2 4. The data phase in the other road 37 is at the 12th position, and the exposure is performed when it is on ^ For example, on the one hand, the two data are judged at the end position. The above-mentioned binary data is determined at the end position of the exposure, and compared with the end position of the exposure obtained from the memory unit 30 shown in FIG. 1 by UT. For example, specifying the end of the exposure The second reference position setting of the position \ a) The black starting line 2 丨 pictured in the picture shows the end position of the pixel in the upper right corner of the pixel 22 in the right corner. The LUT is the same as that shown in Fig. 12 (b). The LUT data is '11 〇〇〇〇〇〇 (r.) Then the 2 value, compared with the end of the exposure with LUT data M 1 〇 1 1 〇 00, compared with, when the agreement, the photo agency 5 obtained The image data is determined as the exposure reference position, and the end-of-exposure position determination circuit 37 outputs the final determination result. In addition, "As in the previous case, if there are six pixels 2 2 shown in Fig. 10 juxtaposed, the upper right corner of each pixel 22 is the second position. Based on the result of this judgment," the count is shown in Fig. 5. In the loop 3 8 B, the number of coincidences is calculated. Therefore, the counted number is compared with the pixel address of the exposure end obtained from the memory portion 30 shown in FIG. 丨. In the comparison loop 3 9 β, two When the values are the same, the end-of-exposure signal is output to the reading p-variation data shown in Fig. 1 to form the processing unit 31. At this time, as shown in Fig. 10, for example, the first day element 221 is in the laser beam scanning direction. And the upper right corner of the fourth pixel 224. If the first reference position is set, the element address of the CCD in the row of the camera 5 corresponding to the second reference position is, for example, 1 g⑽ ", 4 g⑽ // It is stored in the optical switch controller 22. Therefore, the exposure opening is detected in the above manner.

Page 19 200535453 V. Description of the invention (16) After the reference position of the start position and the end position of exposure, proceed to step U. In the vlinS 3, the detection is performed on the glass substrate 8 at the position where the light is moved. Here, as shown in FIG. 3, the distance D between the scanning position F of the laser beam and the photographing position E of the exposure structure 5 is set to a pixel 2 2 moving side to an integer multiple of the machine arrangement pitch P ( η), so the exposure position can be obtained by calculating the laser beam scanning ^. For example, as shown in FIG. 13, the scanning position of the laser beam ^ and the distance D between the photographing positions of the photography mechanism 5 are set to, for example, 3 times the picture + two-column pitch P. Then, the day element 22 is detected in the step. When the end has the second and second reference positions (see the same figure (a)), when the glass substrate 8 moves and the _ ^ line in the day column reaches the photographing position of the photography mechanism 5 (see the same figure (b)), that is, Laser light. Beam scanning starts at the same time. Therefore, when the laser beam is scanned at the period T, the transport speed of the glass substrate 8 is synchronized with the period τ of the laser beam, and it is controlled to move only by a pitch corresponding to the day prime 2 2. Therefore, in the next τ, pixel 22 moves to the position shown in the same figure (c), and after 2 again, pixel 22 moves to the position shown in the same figure (d), so after 3, it is as shown in the figure ( As shown in e), the center line of the column with a diurnal speed of 2 2 reaches the laser beam scanning position. This will detect the exposure position. Next, in step S4, the exposure position is adjusted while scanning with a laser beam. Specifically, as shown in FIG. 14, the exposure position is adjusted by comparing the current laser beam scanning position (element address) detected by # 于 ί lens 1 3 line sensor 1 7 with the preset reference element address. It is only necessary to detect the amount of deflection of the cover and control the light deflection mechanism 10 so that the laser beam scanning position is consistent with the reference element address (reference scanning position). Next, in step S5, the exposure operation is started.

200535453 V. Description of the invention (17) The optical switch controller 24 is used to control the input timing of the optical switch 9. At this time, firstly, the optical switch 9 should be scanned with a laser beam in the ON state, and the line sensor 17 should be turned off immediately after detecting the laser beam scanning start time. At this time, from the modulation data preparation processing section 31, the component address `` 1 0 0 0 〃 '' of the photographing mechanism 5 corresponding to the exposure start position of, for example, FIG. 10 is read out, and the self-laser is calculated by the control section 32. Time from beam scanning start time to exposure start position 11. At this time, the scan time t 0 from the start time of the laser beam scanning to the component address v '1 摄影 of the photography mechanism-5 is measured in advance, and the scanning speed of the laser beam is synchronized with the CCD time axis CLK of the line of the photography mechanism 5, that is, Calculate the number of time axis up to the component address '' 1 0 0 0 〃, you can easily find the scan start time — 1 = t ϋ + 10 0 0CLK 〇 Then put the light switch 9 after 11 from the laser beam scan start time Start exposure. Next, in step S6, the operation of detecting the end position of exposure is performed. The exposure end position is the same as above, for example, the exposure end time of the device address 1 0 0 0 12 12 = t 0 + 1 9 0 0 C LK. The exposure is terminated by turning off the optical switch 9 after t 2 from the start of the laser beam scanning. Next, in step S7, it is determined whether the scanning of the laser beam is finished. When it is determined that it is N 0 返回, the process returns to S 2 and the above operation is repeated. Therefore, as shown in FIG. 10 in S 2, for example, when the second exposure start position, 4 0 0 " and 0 2 exposure end position are detected, and '49 0 0 〃, S5 is entered via S4, and In the same situation as above, the exposure starts at the component address '4 0 0 0 〃, and the exposure ends at the component address -v' 4 9 0 0 〃. In step S7, if it is determined as Y ES ", it returns to S1 and moves P to the operation of detecting a new exposure position. So the above actions are repeated,

Page 21 200535453 V. Description of the invention (18) Formation of exposure patterns in the desired area. According to the above-mentioned embodiment, after the image data of the first pixel row obtained by the photographing mechanism 5 and stored in the memory unit 20 is read, the theoretical sum of the two with the newly acquired image data of the next pixel row is obtained by It is the image which is different from the original pixel image caused by the defects 42 and the foreign matter and flaws attached to the platform 18 and the glass substrate 8 and can be created without defects. Thereby, it is possible to prevent the above-mentioned defects from being mistakenly regarded as a reference position and to be exposed ', so that the exposure accuracy of a predetermined functional pattern can be improved. Ben's use of the defect-free pixel row image data to detect a reference position on the picture 22 = 疋 in advance, and form an exposure pattern based on the reference position, so Yk Nan's functional pattern overlap of the day element 2 2 Precision. The reason is that if the number is plural, the optical device 1 is suitable for the occasion of forming a laminated pattern, and the accuracy of the degree of coincidence is ^ ^ w. In this way, it is not necessary to obtain the exposure units ^, the precision of the machinery, and the cost of the exposure device 1 is suppressed.

The Wt photographing mechanism 5 reads the preset reference position at the pixel 2 2 and must take 5 position as the reference to perform the exposure and stop the exposure.仃 | The alignment of the element 22 and the exposure pattern makes it easy to perform the exposure operation. This is a conceptual diagram of the embodiment of the exposure device of the second invention. Also, in this second invention, a different part is described for explanation. In ♦ # # if > 1 ^ " σ is smaller than the laser beam scanning range of the imaging agency in the field of image processing5. ο Recognition is set here in the black matrix to detect, as shown in the first m π-the reference position of the pixel 22 in / Zao Z1, as shown in Figure 16, which is obtained by the photography agency 5 Motoyuki

200535453 V. Description of the invention (19) The image data of column L 1 is compared with that shown in Fig. 5 using LUT for the exposure start position and LUT for the end position of exposure. Therefore, for example, the exposure start position is set at the top corner of the first pixel 2 2 1, and the end position of the exposure is set at the upper right corner of the fourth day pixel 2 2 4. CCD element address, such as M00CT and M00 (T is stored in the memory section 20) On the other hand, if the pixels 22 are arranged in a sufficient direction, that is, in this column direction, the image of the pixel column obtained by the photography agency 5 is copied in the subsequent In the field of the daytime image processing area 4 3 A of this photography 1. institution 5, the daytime sequence day image that the photography institution 5 cannot obtain is complemented, and exposure is performed based on the complemented daytime image of the pixel sequence. That is, if the pixel pitch in the column direction 22 is W (such as 10 ooclk), the day-to-day image of the pixel row subsequent to the image processing area 4 3 A of the photographing mechanism 5 is shown in FIG. 16 (a). Start position, 4? (= 4〇) from 1000. The exposure start position `` ΙΟΟΟ " read from the memory section 20 and the end of exposure, 49 0 (T, , 39 0 (after exposure at T CLK, 'the same exposure control (for example, exposure time, 3 9 〇 〇) Applicable to the field from the exposure start position 〇〇 (T from the field after the insertion (being printed in the pixel image field 43B), as shown in Figure 16 (b), for Germany 'in the photography agency 5 The pixel sequence of 43A in the image processing field is also full mr4. Furthermore, the same operation is repeatedly performed. According to the second invention, the pixel image obtained by the photography agency 5 can be printed on the subsequent image. The field is made of pixels, and the image sequence of the pixels is exposed, so the photography agency 5 could not obtain it;

Page 23 ZUU ^ 35453 V. Description of the invention (20) The pixel array of the domain can also be used with ^ precision and 'can reduce the need to take 彳 ^ ^ to reveal the established functional pattern. The number of installations' therefore suppresses the area 43A of the photographing mechanism 5 of the exposure full-length portrait full-length portrait, which can be narrow, so that it can be installed at a cost. In this case, the image processing reference position detection accuracy. It is possible to increase the resolution with a photographing mechanism 5 capable of providing a high resolution. More effort ° Improve the exposure precision of the exposure pattern

Λ outside; the situation above 1 I in FIG. 15 ′, but it can also be arranged at = the photographing mechanism 5 is arranged at the transporting mechanism 4 and is formed below the platform 18 to suck the transporting mechanism 4. At this time, due to the existence of f-1, etc., it is possible that the photon 2 is started. The suction mechanism of the plate 8 and the book element opening of the black matrix 21 equipped with the screw 8 are first formed on the glass substrate. In inches. At this time, as in the above case, to complement the daytime row portraits that cannot be obtained by the photographing mechanism 5, the pixel rows hidden in the suction grooves or next to the screw flash can also & accurately expose the established functionality Graphics. In the above-mentioned first and first embodiments, the illumination mechanism adopts back lighting, but the projection illumination is also used. The exposure device of the present invention is not limited to large substrates such as color filters suitable for liquid crystal displays. Suitable for exposure devices such as semiconductors.

Page 24 200535453 Brief description of drawings [Simplified description of drawings] Fig. 1 is a conceptual diagram of an embodiment of an exposure device of the present invention. Fig. 2 is a perspective view illustrating the structure and operation of the optical switch. Fig. 3 is an explanatory diagram showing the relationship between the laser beam scanning position and the photographing position of the camera. Fig. 4 is a block diagram showing the first half of the processing system in the internal configuration of the day image processing section. • Figure 5 is a block diagram showing the second half of the processing system in the internal configuration of the image processing section. Figure 6 is an explanatory diagram of the relationship between the black ® matrix moving in the direction orthogonal to the laser beam scanning direction and the laser beam scanning performance. Fig. 7 is a flowchart illustrating a pattern forming method and procedure using the exposure apparatus. Fig. 8 is an explanatory diagram showing a state of an output of a binary ring buffer memory. Fig. 9 is an explanatory diagram showing a day image and an observation table of an exposure start position set in advance in a black matrix pixel. Fig. 10 is an explanatory diagram showing a relationship between a reference position set in advance in the black matrix and a photographic mechanism element. φ Fig. 11 is an explanatory diagram showing a state in which a defect image existing in a black matrix pixel is excluded. Fig. 12 is an explanatory diagram showing a day image and an observation table of an exposure start position set in advance on a black matrix day element. Figure 13 shows the pixel exposure detected in the glass substrate transport direction

P.25 200535453 Brief description of the drawing Fig. 14 is an explanatory diagram showing the state of the laser beam scanning position correction. FIG. 15 is an explanatory view of an embodiment of an exposure apparatus according to the second invention. Fig. 16 is an explanatory view showing a state in which a missing black matrix day image column image is made and exposed. [Description of main component symbols] 1: exposure device 2: laser light source 3: exposure optical system 4: conveying mechanism 5: photographing mechanism 6: back-illuminating mechanism * 7: optical control mechanism 8: glass substrate 9 · light switch 1 0 : Polarizing mechanism 1 1: First mirror 1 2: Polygon mirror 1 3: f (9 lens 1 4: Second mirror 1 5 A ·. First polarizing element 1 5 C: Second polarizing element 16: Electrical optical adjustment Inverter 17: Row sensor 18: Platform 19: Carousel 2 0: Carrying drive unit 2 1: Black matrix 2 2: Daylight unit 2 3: Light source drive unit φ 2 4: Optical switch controller 2 5 A: Light deflection mechanism drive unit 2 5 B: Polygon drive unit 2 6: Carrying controller 2 7: Back light controller 2 8: A / D conversion unit 2 9: Image processing unit 3 0: Memory unit 3 1 Variable data creation processing unit 3 2: control unit

Page 26 200535453 Brief description of the diagram 3 5: Comparison circuit 3 3A, 33B, 3 3C: Ring buffer memory 34A, 34B, 34C: Line buffer memory 3 6: Exposure start position determination circuit 3 7: Exposure end position Decision circuit 3 8 A: Counting circuit 3 9 A: Comparison circuit 3 8 B: Counting circuit 3 9 B: Comparison circuit 4 0: Previous pixel counting circuit 2 2a: Scanning start pixel 2 2 b: Scanning end pixel 4 2: Defect ❿

Page 27

Claims (1)

200535453 VI. Application for Patent Scope 1. An exposure device that scans the light beam irradiated by the exposure optics in a direction orthogonal to the moving direction of the object to be exposed, and exposes the function of the object at a predetermined pitch on the object to be exposed. The patterner includes: a photographing mechanism for capturing a reference functional pattern previously formed on the subject as a reference exposure position; and a day image of the reference functional pattern obtained by the photographing mechanism. The data is subjected to a predetermined image processing to eliminate defects in the reference functional pattern portrait corresponding to the beam scanning field, to create a non-defective reference functional pattern day image, and inspect the non-defective reference functional pattern day image. An optical system control mechanism that outputs a reference position at which exposure starts or ends, and uses the reference position as a reference to control the start or stop of P irradiation of the light beam. 2. For the exposure device according to item 1 of the scope of patent application, wherein the optical system control mechanism uses two portrait data in the reference functional pattern obtained by the photography agency to select the theoretical sum of each portrait data, The person making the defect-free reference functional pattern portrait. 3. For the exposure device according to item 2 of the scope of patent application, wherein the optical system control mechanism uses the previous reference functional pattern image data obtained by the photographing mechanism and the newly obtained reference functional pattern image data to obtain The pixel data located at the same position before and after the moving direction of the object to be exposed are combined to form a defect-free reference functional pattern portrait. 4. For the exposure device according to item 2 of the patent application scope, wherein the optical system control mechanism uses adjacent reference functional pattern image data among the reference functional pattern image data obtained by the photographing mechanism to obtain each day Like the theory of data, and make a defect-free reference functional figure portrait. Page 28 200535453 VI. Application for Patent Scope 5. — An exposure device that relatively scans the light beam irradiated by the exposure optics in a direction orthogonal to the moving direction of the object to be exposed, and uses the predetermined section on the object to be exposed. The distance recording function pattern includes: a photographing mechanism for taking a reference functional pattern formed in advance on the object to be used as a reference exposure position; and a photographing mechanism for taking the photographic mechanism in a given field. The reference functional pattern image data is copied in the subsequent fields of the established field, and the reference functional pattern day image that is not obtained by the photography agency is supplemented. In the complement, the reference functional pattern image A reference position that ends after exposure starts is detected, and the irradiation start or stop of the beam is controlled by the reference position. Page 29