CN109725502B

CN109725502B - Double-sided exposure apparatus and double-sided exposure method

Info

Publication number: CN109725502B
Application number: CN201811284543.5A
Authority: CN
Inventors: 名古屋淳
Original assignee: Aditech Engineering Co ltd
Current assignee: Aditech Engineering Co ltd
Priority date: 2017-10-31
Filing date: 2018-10-31
Publication date: 2024-02-06
Anticipated expiration: 2038-10-31
Also published as: JP6994806B2; KR102622294B1; KR20190049563A; CN109725502A; JP2019082612A; TW201933432A; KR20240005244A; CN117806134A; TWI772549B

Abstract

The purpose of the present invention is to solve the problem of shielding of a calibration mark of a mask when the calibration mark of the mask is photographed by a camera through an opening for calibration of a substrate. The exposure unit (2) irradiates light onto the substrate (W) through a pair of first and second masks (3, 4) arranged at positions sandwiching the substrate (W) intermittently fed by the conveying system (1) to expose the substrate (W). Before exposure, the calibration means performs calibration based on imaging data from a camera (8) that images the calibration marks (31) of the first mask (3), the calibration marks (41) of the second mask (4), and the calibration openings (Wm) of the substrate (W). When the mask marks (31, 41) are not imaged through the calibration aperture (Wm) and are blocked, the positions of the mask marks (31, 41) at the completion of calibration at the previous exposure are read out from the storage unit (60), and the movement amount for blocking elimination is calculated based on the positions, and the masks (3, 4) are moved to eliminate the blocking.

Description

Double-sided exposure apparatus and double-sided exposure method

Technical Field

The present invention relates to a double-sided exposure apparatus such as a roll-to-roll (roll-to-roll) system used for manufacturing flexible printed boards and the like.

Background

An exposure apparatus that irradiates a predetermined pattern of light onto a target to expose the target is used for various purposes as a key element technique of photolithography. There are various types of exposure apparatuses, and there are two-sided exposure apparatuses that expose both sides of a long substrate in a belt shape.

For example, in the case of an apparatus for exposing a flexible substrate such as a flexible printed board, a structure is adopted in which the substrate is exposed while being transported in a roll-to-roll manner. A pair of exposure units are disposed on both sides (generally up and down) of the conveyance path of the substrate. The device comprises masks, and an exposure unit irradiates light of a predetermined pattern from both sides via each mask to perform exposure

The substrate pulled out from the roll is intermittently transported, and light of a predetermined pattern is irradiated to both surfaces of a portion between a pair of exposure units of the substrate stopped after the transport, and both surfaces are simultaneously exposed.

Such a double-sided exposure apparatus is also one type of exposure apparatus, and therefore, the accuracy of calibration (alignment) becomes a problem. In the case of an apparatus for exposing a long substrate in a band shape, such as a roll-to-roll apparatus, the substrate is cut at an appropriate position in the longitudinal direction after the end of photolithography, and a final product is obtained. Since the cutting position can be appropriately selected, calibration in the longitudinal direction in the exposure apparatus has not been a problem in the past. On the other hand, the positional relationship between the pair of masks needs to be maintained with high accuracy. That is, if the accuracy of the positional relationship between the pair of masks is poor, the pattern on one side of the substrate and the pattern on the other side in the final product are not identical, and product defects are likely to occur. Therefore, as in patent document 1 or patent document 2, the pair of masks are mutually aligned so as to avoid inconsistency of the pattern formed.

In the past, the alignment with respect to the substrate is required to be performed with a sufficiently high accuracy even though it is not sufficient to calibrate a pair of masks with each other. As one background, there are many cases where a complex structure such as a multilayer wiring is provided with higher performance of a product.

In the case of introducing a complicated structure such as a multilayer wiring into a flexible printed circuit board, a pattern is formed on a tape-shaped substrate, and a resist is applied thereon to expose the substrate. The conventional pattern is formed in plural at intervals along the longitudinal direction of the tape-shaped substrate, and the portion where each pattern is formed is finally formed into each product. In this case, in the further exposure, it is necessary to expose the already formed pattern with a desired positional accuracy, and alignment with respect to the substrate is necessary.

In addition, according to the product, another flexible square substrate may be laminated on a portion where a pattern has been formed, and the other substrate (hereinafter referred to as an upper substrate) may be exposed to light to form a pattern. In this case, too, since the upper substrate is laminated in plural at intervals along the longitudinal direction of the strip-shaped substrate, it is necessary to perform exposure in a state of alignment with respect to each upper substrate.

In this case, the alignment of the pair of masks with respect to the substrate is required, and the pair of masks must be aligned with respect to the substrate while maintaining the alignment state. Accordingly, patent document 2 adopts a configuration in which alignment marks of masks on both sides are photographed by a camera via alignment marks provided on a substrate.

Patent document 1: japanese patent laid-open No. 2000-155430

Patent document 2: japanese patent laid-open No. 2006-278648

As described above, in patent document 2, when alignment with respect to the substrate is required in addition to alignment of the pair of masks, a structure is proposed to fulfill this requirement.

However, according to the studies of the inventors, it is actually difficult to perform each calibration with a required accuracy only by the structure disclosed in patent document 1 or patent document 2. This will be explained below.

First, in patent document 1, although the alignment of the pair of masks is described, the alignment of the pair of masks and the member to be exposed is not described. In patent document 1, the formation of the alignment mark M on the member to be exposed is described, but the formation of the alignment mark by the sensitization of the resist layer is described, and it is understood that the alignment mark is not present at the time of exposure.

In patent document 2, it is described that adjustment is performed so that the calibration mark AM is made ₁₁ 、AM _12a 、AM _12B On the optical axes of the imaging devices 13A, 13B, but there is no description of how the specific configuration is adjusted.

According to the study of the inventors, in the above-described configuration in which exposure is performed while intermittently feeding a long, band-shaped substrate, there is a problem in that accuracy of a stop position of the substrate at the time of intermittent feeding affects calibration. As one of the problems, there is a problem that the calibration marks of the mask are blocked.

More specifically, in a configuration in which the alignment mark of the substrate is an opening, and the camera photographs the alignment mark of the mask through the opening, if the accuracy of the stop position of the intermittent feed is low, the alignment mark of the mask may be blocked by the substrate even if the alignment mark of the mask is photographed by the camera after the stop. In this case, the calibration marks of the mask cannot be captured by the camera, and as a result cannot be calibrated. Such shielding of the alignment mark of the mask also occurs when the accuracy of the formation position of the opening of the substrate as the alignment mark is low.

Disclosure of Invention

The present invention has been made in consideration of the above-described problems, and an object of the present invention is to effectively solve the problem of shielding of a calibration mark of a mask in a structure in which the calibration mark of a substrate is an opening and the calibration mark of the mask is photographed by a camera through the opening.

In order to solve the above-described problem, the invention according to claim 1 of the present application has the following configuration: a transport system that pulls out and intermittently feeds a flexible substrate, which is wound up in a roll, and in which a plurality of target exposure areas are set at intervals along a longitudinal direction; a pair of first and second masks arranged at positions sandwiching the substrate to be fed; an exposure unit that irradiates light to target exposure areas on both sides of the substrate via each mask to perform exposure after the substrate is stopped and aligned by the transport system; a storage unit; the substrate has an alignment opening provided in a predetermined positional relationship with respect to a region to be exposed; the first mask has a first mask mark as a mark for calibration; the second mask has a second mask mark as a mark for calibration; a camera capable of photographing the first mask mark, the second mask mark, and the alignment opening of the substrate is provided; a calibration mechanism for aligning the first mask and the second mask with respect to the region to be exposed of the substrate based on imaging data from a camera that images the first mask mark, the second mask mark, and the calibration opening; the alignment device is provided with a storage means for storing the position of the first mask mark and/or the position of the second mask mark as the previous mark position in the storage unit when alignment by the alignment means is completed, and a temporary alignment means; the temporary calibration mechanism has a structure as follows: when the alignment mechanism performs alignment, if the first mask mark or the second mask mark is not captured through the alignment opening of the substrate and is blocked by the substrate, the position of the previous mark at the completion of the alignment at the time of the previous exposure is acquired from the storage unit, and the direction and distance for moving the substrate or the first and second masks for blocking removal are calculated based on the position, and the substrate or the first and second masks are moved toward the direction by the distance.

In order to solve the above-described problem, the invention according to claim 2 has the following structure: in the structure according to claim 1, a mask moving mechanism is provided for moving the first and second masks; the temporary alignment mechanism moves the first and second masks to remove the shadow.

In order to solve the above-described problem, the invention according to claim 3 has the following structure: in the configuration of claim 2, the mask moving mechanism is a mechanism capable of integrally moving the first and second masks.

In order to solve the above-described problem, the double-sided exposure method according to claim 4 of the present application is a method in which a flexible substrate, which is wound up in a roll, is pulled out by a conveyance system, in which a plurality of target exposure areas are set at intervals in a longitudinal direction, and is intermittently fed, and the substrate, which is fed and stopped, is exposed by irradiating light to the target exposure areas on both sides of the substrate by an exposure unit through a pair of first and second masks disposed with the substrate interposed therebetween; the substrate has an alignment opening provided in a predetermined positional relationship with respect to a region to be exposed; the first mask has a first mask mark as a mark for calibration; the second mask has a second mask mark as a mark for calibration; before exposure, performing alignment of the first mask mark, the second mask mark, and the alignment opening of the substrate with respect to the region to be exposed of the substrate based on the obtained imaging data while imaging the first mask mark, the second mask mark, and the alignment opening of the substrate with respect to the region to be exposed; the method comprises the following steps: a storage step of storing the positions of the first mask mark and/or the second mask mark at the time of completion of calibration before exposure to a certain target exposure area as previous mark positions in a storage unit; a movement amount calculation step of, when calibration is performed before exposure of a next target exposure area located upstream in the feed direction from the target exposure area, acquiring a previous mark position from a storage unit and calculating an orientation and a distance of movement of the substrate or the first and second mask for removing the shielding of the first mask mark or the second mask mark, based on the position, when the first mask mark or the second mask mark is not captured through the calibration opening of the substrate and is shielded by the substrate in the image data from the camera; and a movement step of moving the substrate or the first and second masks in the direction and distance calculated in the movement amount calculation step, thereby eliminating the occlusion.

The invention has the advantages that:

as described below, according to the invention described in claim 1 or 4 of the present application, when the mask mark is blocked by the substrate, the temporary alignment mechanism eliminates the blocking, so that the alignment can be performed normally. In this case, the temporary alignment mechanism calculates the movement amount based on the position of the mask mark in the alignment at the previous exposure, and moves the substrate or the mask by the movement amount to eliminate the blocking, so that the problem of productivity degradation does not occur.

Further, according to the invention described in claim 2, in addition to the above-described effects, since the temporary calibration means is means for removing the mask by moving the mask, movement for removing the mask can be easily performed, and it is possible to contribute to performing calibration with high accuracy.

In addition, according to the invention described in claim 3, in addition to the above-described effects, since the blocking can be removed by integrally moving the first and second masks, a control signal and an operation for removing the blocking can be simplified.

Drawings

Fig. 1 is a schematic front sectional view of a double-sided exposure apparatus according to an embodiment.

Fig. 2 is a perspective view schematically showing a calibration mark required for calibration.

Fig. 3 is a flow chart schematically showing extraction of a portion associated with calibration in the main sequence program.

Fig. 4 is a perspective view schematically showing a case where an alignment opening of a substrate is found by an opening search program.

Fig. 5 is a plan view schematically showing the determination of the deletion of the calibration aperture by the aperture deletion determination program and the deletion removal by the aperture deletion removal program.

Fig. 6 is a plan view schematically showing the marker occlusion determination by the marker occlusion determination program and the occlusion removal by the temporary calibration program.

Fig. 7 is a plan view schematically showing temporary calibration performed by the temporary calibration program.

Fig. 8 is a plan view schematically showing a marker deletion determination program and a marker deletion removal program.

Fig. 9 is a plan view schematically showing the main calibration performed by the main calibration program.

Description of the reference numerals

1. Conveying system

2. Exposure unit

21. Light source

22. Optical system

3. First mask

31. First mask mark

4. Second mask

41. Second mask mark

5. Mask moving mechanism

6. Main controller

61. Storage unit

7. Main sequence program

71. Program for determining whether or not opening exists

72. Opening search program

73. Program for determining missing opening

74. Program for eliminating opening defect

75. Mark occlusion determination program

76. Temporary calibration procedure

77. Marker absence determination program

78. Marker deletion removal program

79. Formal calibration procedure

8. Camera with camera body

81. Camera moving mechanism

W substrate

Wm calibration opening

V field of view

Detailed Description

Next, specific embodiments (hereinafter referred to as embodiments) of the present application will be described.

Fig. 1 is a schematic front sectional view of a double-sided exposure apparatus according to an embodiment. The apparatus of the embodiment is an apparatus for exposing a flexible and band-shaped substrate W such as polyimide. As shown in fig. 1, the double-sided exposure apparatus includes a conveyance system 1 and an exposure unit 2.

The conveyance system 1 is a mechanism that pulls out and intermittently feeds out a flexible substrate W wound into a roll. The term "flexible" means flexibility to such an extent that the flexible printed circuit board can be wound into a roll, and examples thereof include a substrate for a flexible printed circuit board.

In this embodiment, the transport system 1 is a mechanism for horizontally pulling out the substrate W and transporting the substrate W in a horizontal posture. Specifically, the conveying system 1 includes: a delivery-side core roller 11 around which the unexposed substrate W is wound; a delivery-side pressing roller 12 that pulls out the substrate W from the delivery-side core roller 11; a winding-side core roller 13 for winding the exposed substrate W; and a winding-side pressing roller 14 that pulls out the exposed substrate W and winds the substrate W around the winding-side core roller 13. The feeding direction of the substrate W by the transport system 1 is defined as the X direction, and the horizontal direction perpendicular thereto is defined as the Y direction. The Y direction is the width direction of the substrate W. Let the direction perpendicular to the XY plane be the Z direction.

An exposure operation position is set between the delivery-side platen roller 12 and the take-up-side platen roller 14. The exposure operation position is a position where both surfaces of the substrate W are simultaneously exposed by the exposure unit 2.

As shown in fig. 1, a pair of masks 3 and 4 are disposed at the exposure operation position with the substrate W interposed therebetween. Hereinafter, the upper mask 3 is referred to as a first mask, and the lower mask 4 is referred to as a second mask. Each mask 3, 4 is in a horizontal posture.

The exposure unit 2 is also provided with two corresponding masks 1, 2. The exposure unit 2 for exposure through the first mask 3 is provided above the first mask 3, and irradiates light downward to expose the mask. The exposure unit 2 for exposing light through the second mask 4 is provided below the second mask 4, and irradiates light upward to expose the light.

The two exposure units 2 are arranged symmetrically up and down, and are identical in structure. That is, each exposure unit 2 includes a light source 21, an optical system 22 for irradiating light from the light source 21 to the masks 3 and 4, and the like. As will be described later, the apparatus of this embodiment is an apparatus for performing contact exposure, and each exposure unit 2 is a unit for irradiating parallel light to each of the masks 3 and 4. Thus, the optical system 22 includes a collimator lens.

The conveying system 1 includes buffers 101, 102 on the upstream side and the downstream side of the exposure operation position. The conveying system 1 includes a first driving roller 15 disposed on the upstream side of the exposure operation position and a second driving roller 16 disposed on the downstream side of the exposure operation position. Each of the driving rollers 15, 16 is a pressing roller.

As shown in fig. 1, a delivery-side buffer 101 is provided between the delivery-side pressure roller 12 and the first drive roller 15. Further, between the second driving roller 16 and the winding-side pressing roller 14 is a winding-side buffer 102.

The first driving roller 15 and the second driving roller 16 are members for intermittently feeding the substrate W passing through the exposure operation position. That is, the first driving roller 15 and the second driving roller 16 are rollers that operate in synchronization, and are configured to feed the substrate W at a predetermined stroke. This stroke is a distance for feeding the substrate W at the time of one intermittent feeding, and is hereinafter referred to as a feeding stroke.

On the other hand, the delivery-side core roller 11 and the delivery-side pressing roller 12 are synchronously driven in accordance with the amount of slack of the substrate W in the delivery-side buffer 101. A sensor, not shown, is disposed in the delivery-side buffer 101, and if the amount of slack becomes smaller, the delivery-side core roller 11 and the delivery-side pressure roller 12 operate synchronously to deliver the substrate W to a maximum value set.

The winding-side buffer 102 is also provided with a sensor, not shown. In response to a signal from the sensor, if the amount of slack becomes to a limit, the winding-side pressing roller 14 and the winding-side core roller 13 operate synchronously to wind up the substrate W so that the amount of slack is reduced to a set minimum value.

In the intermittent feeding of the above-described conveying system 1, after the feeding in the feeding stroke, both sides of the substrate W are exposed by each exposure unit 2 in the stop of the substrate W, but before that, calibration is performed by the calibration mechanism. The alignment mechanism aligns the first and second masks 3 and 4 with respect to the region of the substrate W to be exposed.

In this embodiment, the alignment is finally performed by aligning the pair of masks 3, 4 with respect to the areas to be exposed on the substrate W. Thus, as shown in fig. 1, the pair of masks 3, 4 includes a mask moving mechanism 5, and the mask moving mechanism 5 is included in the alignment mechanism. The mask moving mechanism 5 moves the masks 3 and 4 in the XY directions to change positions. The mask moving mechanism 5 is a mechanism capable of moving the first mask 3 and the second mask 4 independently and integrally moving the two masks 3 and 4. Such a mechanism can be easily manufactured, for example, by fixing a mechanism for moving the first mask 3 in the XY direction to the first base plate, fixing a mechanism for moving the second mask 4 in the XY direction to the second base plate, and providing a mechanism for integrally moving the first and second base plates in the XY direction.

Further, a Z-direction moving mechanism, not shown, is provided on each of the masks 3 and 4. The Z-direction moving mechanism moves the masks 3 and 4 toward the substrate W for contact exposure, and brings the masks into close contact with the substrate W.

As shown in fig. 1, the apparatus includes a main controller 6 that controls each section including the conveyance system 1 and the mask moving mechanism 5. The main controller 6 is provided with a main sequence program 7 for controlling the respective units of the device to operate in a predetermined order. That is, the main sequence program 7 is stored in the storage unit 60 of the main controller 6, and can be executed by a processor (not shown) of the main controller 6. In addition, the main controller 6 includes a display 61 for performing error display or the like.

For calibration, a mark as a sign is required. Fig. 2 is a perspective view schematically showing a calibration mark required for calibration. As shown in fig. 2, alignment marks 31 and 41 are formed on the masks 3 and 4. Hereinafter, the alignment mark 31 provided on the first mask 3 is referred to as a first mask mark, and the alignment mark 41 provided on the second mask 4 is referred to as a second mask mark. As shown in fig. 2, in this embodiment, the first mask mark 31 is circular, and the second mask mark 41 is a point of a circle smaller than the first mask mark 31.

As shown in fig. 2, a calibration mark Wm is also formed on the substrate W for calibration. The alignment mark Wm of the substrate W is an opening. Hereinafter referred to as an opening for calibration. In this embodiment, the calibration opening Wm is circular.

As described above, the alignment is an operation of aligning the pair of masks with each other and aligning the pair of masks with respect to the substrate. For this reason, it is convenient to calibrate the substrate by taking a state in which a pair of mask marks overlap with a calibration mark of the substrate as a reference and setting this state to be an ideal state (reference of accuracy). As shown in fig. 2, the "overlapped state" is typically a case where the centers of the marks 31, 41, wm are positioned on a straight line (on a straight line perpendicular to the substrate W), but other states may be used as a reference for accuracy.

In this embodiment, in order to enable the calibration to be performed with high accuracy and ease, the calibration opening Wm is larger than the first mask mark 31 and larger than the second mask mark 41. That is, in the aligned state, the two mask marks 31 and 41 are distinguishable from each other in the alignment opening Wm when viewed from the direction perpendicular to the substrate W.

As shown in fig. 1, the apparatus includes a camera 8 for photographing the calibration marks 31, 41, wm. The camera 8 is connected to the main controller 6, and the photographing data of the camera 8 is transmitted to the main controller 6.

As shown in fig. 2, in this embodiment, four first mask marks 31 and four second mask marks 41 are provided, respectively. Four cameras 8 are also provided in match with them. The first mask mark 31 and the second mask mark 41 are provided at positions corresponding to the corners of the square, and the camera 8 is also provided at positions corresponding to the corners of the square.

Each camera 8 is arranged so that an optical axis (optical axis of the built-in lens) a is perpendicular, and is mounted in a posture of photographing below. A camera moving mechanism 81 for changing the position of the camera 8 in the XY direction is provided on the pedestal on which each camera 8 is mounted.

The first mask mark 31 and the second mask mark 41 are provided at positions corresponding to corners of squares of the same size and shape. This position is known as design information, and the four cameras 8 are provided in a state adjusted to have the same positional relationship in the horizontal direction. However, it is not necessary that the optical axes a of the four cameras 8 be coaxial with the centers of the mask marks 31, 41, and it is sufficient that the mask marks 31, 41 come within the field of view of the cameras 8.

The alignment opening Wm of the substrate W is a mark indicating the position of an area to be exposed (hereinafter referred to as a "target exposure area") and is provided in a predetermined positional relationship with respect to the target exposure area. The target exposure region is a region to which the pattern of each mask 3, 4 is transferred, and is indicated by a broken line in fig. 2. The alignment opening Wm is formed outside the target exposure region R at a position corresponding to a corner of a square having the same size and shape as the first and second mask marks 41.

The target exposure region R corresponds to a portion of the substrate W used when 1 product is produced. Thus, as shown in fig. 2, a plurality of target exposure regions R are set at intervals along the longitudinal direction of the strip-shaped substrate W. The alignment openings Wm are also provided in the same positional relationship in design with respect to the respective target exposure regions R. The pitch of each target exposure field R corresponds to the feeding stroke (denoted by Lf in fig. 2) by the conveying system 1.

The calibration means is constituted by the hardware provided in the above-described device and software including the main sequence program 7 installed in the main controller 6. In this embodiment, a temporary calibration mechanism is provided for temporarily performing calibration before final calibration, which is a large feature point. These mechanisms are specifically described below.

Fig. 3 is a flowchart schematically showing the extraction of the calibration-related portion in the main sequence program 7. The calibration is performed after the intermittent feeding of the substrate W by the transport system 1 is completed. For calibration, the main sequence program 7 is provided with, as shown in fig. 3: an opening presence/absence determination step S1 of determining whether or not all of the calibration openings Wm are imaged; an opening missing determination step S2 of determining whether or not all the calibration openings Wm are recognized in a state of no missing; a mark shielding determination step S3 of determining whether or not each of the mask marks 31, 41 is not shielded by the substrate W, when all of the alignment openings Wm are recognized in a state of no missing; a mark missing determination step S4 of determining whether or not the mask marks 31, 41 are imaged without missing, when it is determined that the mask marks 31, 41 are not blocked; and a main calibration step S5 of performing main calibration when it is determined that none of the mask marks 31, 41 is missing.

The main controller 6 is provided with a presence/absence of opening determination program 71, an opening search program 72, an opening loss determination program 73, an opening loss removal program 74, a mark shielding determination program 75, a temporary calibration program 76, a mark loss determination program 77, a mark loss removal program 78, and a main calibration program 79 as subroutines to be called and executed from the main sequence program 7. In addition, the provisional calibration program 76 constitutes a provisional calibration mechanism.

The opening presence/absence determination step S1 is a step of executing the opening presence/absence determination program 71 to obtain a return value. The opening search program 72 is a program executed when at least 1 calibration opening Wm is determined not to be in the field of view of the camera 8.

The loss-of-opening determination step S2 is a step of executing the loss-of-opening determination program 73 to obtain a return value. The opening defect removal program 74 is a program executed when it is determined that there is a defect with respect to at least 1 calibration opening Wm.

The marker occlusion determination step S3 is a step of executing the marker occlusion determination program 75 to acquire a return value thereof. The provisional calibration program 76 is a program executed when it is determined that the mask mark is blocked by the substrate W in the image data from at least 1 camera 8.

The marker determination step S4 is a step of executing the marker deletion determination program 77 to acquire a return value thereof.

The main calibration routine 79 is a routine executed when all the mask marks 31 and 41 are determined to be calibratable without being blocked by the substrate W.

Next, the configuration of each step and each subroutine will be described in order. First, the opening presence/absence determination step S1 and the opening presence/absence determination program 71 will be described.

As shown in fig. 3, the main sequence program 7 executes the opening presence/absence determination program 71 after the intermittent feed is completed. The return value of the opening presence/absence determination program 71 returns to the normal value when all the calibration openings Wm are imaged, and returns to the abnormal value when not.

The opening presence/absence determination program 71 is programmed to process image data from each camera 8 and determine whether or not an image of the opening Wm for calibration is included by pattern matching. In this embodiment, the calibration opening Wm is circular, and its diameter is known as design information. Therefore, the opening presence/absence determination program 71 searches for an opening Wm for calibration, which can be determined from among those whose boundary line is regarded as a circle. With respect to at least one image data, an abnormal value is returned if not regarded as the opening Wm for calibration, and a normal value is returned if not.

As shown in fig. 3, the main sequence program 7 is programmed to execute the opening search program 72 when the return value of the opening presence determination program 71 is an abnormal value. Fig. 4 is a perspective schematic view showing an example of the case where the opening for calibration Wm is searched by the opening search program 72.

The opening search program 72 is a program for outputting a control signal to the transfer system 1 to move the substrate W and putting the calibration opening Wm into the field of view of the camera 8 when the calibration opening Wm is not imaged. Fig. 4 shows a field of view V of a camera 8 and a calibration opening Wm to be found. The calibration openings Wm are provided in a predetermined positional relationship with respect to the target exposure region R.

The movement of the substrate W for placing the alignment opening Wm into the field of view of the camera 8 is the movement in the X direction. The travel of this movement is slightly shorter than the length of the field of view in the X direction. Hereinafter, this stroke will be referred to as a search stroke.

In this embodiment, the opening search program 72 first performs movement (return) toward the opposite side to the intermittent feeding of the substrate W, and performs movement (feeding) toward the same side as the intermittent feeding even if the alignment opening Wm is not found in this way. For example, the search stroke is set to be moved up to 2 times. In this example, the range of the search calibration opening Wm is a range of 5 search strokes. The example shown in fig. 4 is an example in which searching is performed in the order of (1) → (2) → (3) → (4) → (5), and is an example in which the searching is performed in the first 2 searching strokes, and then the feeding is performed in the 4 searching strokes, and the result calibration opening Wm is brought into the visual field V.

As shown in fig. 3, the main sequence program 7 is programmed to acquire the return value from the opening search program 72, and if the return value is an abnormal value, the error processing is performed to stop the program since the calibration opening Wm is not found. The error processing includes an operation of displaying, on the display 61 of the main controller 6, the content that cannot be photographed to the opening Wm for calibration.

As shown in fig. 3, when the return value of the opening search program 72 is a normal value or when the normal value is returned during the execution of the first opening presence/absence determination program 71, the main sequence program 7 executes the opening absence determination program 73. Fig. 5 is a schematic plan view showing the deletion determination of the calibration aperture by the aperture deletion determination program 73 and the deletion removal by the aperture deletion removal program.

When the intermittent feeding of the substrate W by the conveyor system 1 is completed or the opening search process 72 is normally completed, the alignment opening Wm may be completely in the field of view of the camera 8, but some of the openings may be missing without being in the field of view. An example of deletion elimination is shown in fig. 5 (1). The opening missing determination program 73 processes the image data from each camera 8, and determines whether or not the calibration opening Wm is imaged in a state of no missing in all the image data. The opening deletion determination program 73 is programmed to return a normal value to the main sequence program 7 if the image is captured without deletion, and to return an abnormal value if it is determined that there is a deletion with respect to the image data from 1 or more cameras 8.

As shown in fig. 3, when an abnormal value is returned from the open deletion determination program 73 (when it is determined that there is a deletion), the main sequence program 7 calls and executes the open deletion removal program 74. The aperture loss removal program 74 processes image data from each camera 8, and calculates the amount of movement (orientation and distance) of the substrate W or the camera 8 required for the loss removal. The opening defect removal program 74 is programmed to transmit the calculated movement amount to the conveyor system 1 and/or the camera movement mechanism 81, and move the substrate W and/or the camera 8. At this time, the movement in the X direction may be performed by moving the substrate W or by moving the camera 8. Further, the camera 8 is moved with respect to the Y direction. That is, the opening defect removal program 74 is programmed to transmit the movement amount (direction and distance) in the X direction for the removal of the defect to the conveying system 1 and transmit the movement distance in the Y direction to the camera movement mechanism 81.

In any case, if the opening deletion elimination program 74 is executed, as shown in fig. 5 (2), four calibration openings Wm are imaged in a state where the deletion is eliminated. Since the amount of the loss is generally different for each image data, the image data having the largest loss of the calibration openings Wm is determined for the image data from the four cameras 8, and the amount of movement for eliminating the loss in the image data is transmitted to the conveying system 1 and/or the camera moving mechanism 81.

Next, the temporary calibration mechanism will be described.

As described above, the provisional calibration means is means for performing provisional calibration before final calibration. The temporary calibration is an operation of eliminating the masking of the mask marks 31 and 41 by moving the final mask 3 and 4 with reference to the position thereof during the previous exposure calibration. Hereinafter, description will be made specifically.

As shown in fig. 3, the main sequence program 7 is programmed to execute the marker occlusion determination program 75 after executing the opening deletion removal program 74. Fig. 6 is a plan view schematically showing the marker occlusion determination by the marker occlusion determination program 75 and the occlusion removal by the temporary calibration program 76.

When the normal value is returned by the aperture loss determination program 73 or when the aperture loss removal program 74 is completed, the alignment aperture Wm is imaged in the state where no aperture is present in each camera 8, but the pair of mask marks 31, 41 are not located in each alignment aperture Wm and may be blocked by the substrate W. Fig. 6 (1) shows an example of a situation in which such a pair of mask marks 31, 41 is blocked.

The mark shielding determination program 75 is a program for processing image data from each camera 8 and determining whether or not an image of a pair of mask marks 31 and 41 exists in each calibration opening Wm. In this embodiment, since the first mask marks 31 are circles smaller than the openings Wm for alignment and the second mask marks 41 are dots of circles smaller than the second mask marks 41, it is determined whether or not they are present in the openings Wm for alignment by pattern matching. The marker occlusion decision program 75 is programmed to send a normal value back to the main sequence program 7 if present and to return an abnormal value if not present.

As shown in fig. 3, the main sequence program 7 executes a temporary calibration program 76 in the case where an abnormal value is returned from the marker occlusion determination program 75. Fig. 7 is a plan view schematically showing temporary calibration performed by the temporary calibration program 76.

In this embodiment, the temporary calibration program 76 is a program for eliminating the masking of the mask marks 31, 41 with reference to the final positions of the masks 3, 4 in the calibration at the time of the previous exposure (exposure to the previous target exposure region R).

Specifically, as will be described later, the main sequence program 7 constitutes a storage means and includes a storage step of storing the center positions (positions in XY coordinates) of the mask marks 31 and 41 in the storage unit 60 when the main calibration is completed. The storing step is a step of storing each position of the four mask marks 31 of the first mask 3 and each position of the four mask marks 41 of the second mask 4, and 8 pieces of position information are stored in total. The storage is performed in a form of updating information of the positions of the mask marks 31 and 41 in the calibration at the previous exposure. That is, the positional information stored in the storage unit 60 is information at the time of the immediately preceding exposure. Hereinafter, these pieces of position information will be referred to as previous position information.

The provisional calibration program 79 first processes the image data from each camera 8 to obtain the position of the center of the calibration opening Wm. This position is acquired as a position in a coordinate system with the optical axis a of the camera 8 as the origin in each image data.

Next, the previous position information is read from the storage unit 60. In each image data, although the occlusion is detected, the movement amount (direction and distance) for eliminating the occlusion is calculated assuming that each mask mark 31, 41 is present at the position of the previous position information. Fig. 7 shows calculation of the amount of movement in the image data from a certain camera 8. In FIG. 7, the center of the obtained calibration opening Wm is denoted by C ₀ . Further, the center position of the first mask mark 31 in the previous position information read from the storage unit 60 is set to be C ₁ Let the center of the second mask mark 41 be C ₂ 。

As shown in fig. 7, the provisional calibration program 79 calculates the center C for marking the first mask in the previous position information ₁ Center C with the opening Wm for calibration ₀ The uniform movement amount, the center C for marking the second mask is also calculated ₂ Center C with the opening Wm for calibration ₀ A uniform amount of movement. As shown in fig. 8, the movement amount of the first mask mark 31 is set to Q ₁ Let the movement amount of the second mask mark 41 be Q ₂ 。Q ₁ 、Q ₂ Is a vector.

The provisional calibration program 79 performs the above-described arithmetic processing on the image data from each camera 8, and calculates Q ₁ 、Q ₂ . And calculate Q ₁ 、Q ₂ Average value of (2). Namely, about Q in four image data ₁ An average of the orientations (vector synthesis) is calculated, and an average of the distances is calculated. In addition, regarding Q ₂ The average of the orientation and the average of the distance are also calculated.

Next, the provisional calibration program 79 calculates the average Q ₁ 、Q ₂ As control signals, the control signals are output to the mask moving mechanism 5, and are moved. That is, Q is averaged by the first mask 3 ₁ To average Q of the second mask 4 ₂ Is moved by the motion of the moving object. Thereby the processing time of the product is reduced,as shown in fig. 6 (2), the masking marks 31 and 41 are blocked.

The temporary calibration is a calibration in which the movement for eliminating the mark occlusion is performed assuming that the masks 3 and 4 are continuously positioned at the positions of the previous position information. The masks 3 and 4 are moved in the Z direction so as to be closely adhered to the substrate W during exposure, and are also moved in the Z direction so as to return to a height away from the substrate W after exposure. In this case, the XY-direction position may be shifted, but even if the shift is small, there is no problem as a reference in calculating the amount of movement for eliminating mark shielding.

In the above configuration, the positions of the masks 3 and 4 are aligned in the calibration at the previous exposure, so that the centers are aligned within the range of the required accuracy. Thus, there is no need to add Q ₁ 、Q ₂ The average value of (a) is output to the mask moving mechanism 5 to move the masks 3 and 4 independently, or the two masks 3 and 4 may be integrally moved. In this case, it is possible to output an average Q ₁ And integrally move, or output an average Q ₂ And integrally moves. Furthermore, the average Q may be ₁ Q and Q ₂ The average value is further calculated and outputted as an integrally moving amount.

It is not necessary to store the positional information of both of the pair of mask marks 31 and 41 in the storage unit 60 for use. As described above, since the pair of masks 3 and 4 are aligned with each other in the previous alignment, the position information of one mask mark is stored, and the occlusion can be removed by calculating the movement distance based on the stored position information and outputting the calculated movement distance.

In any case, the temporary alignment mechanism is in a state where the masking marks 31 and 41 are blocked off as shown in fig. 6 (2). As shown in fig. 3, when the provisional calibration program 76 is executed, the main sequence program 7 executes the marker occlusion determination program 75 again to determine whether or not there is no marker occlusion. When the return to the normal value is confirmed, the main sequence program 7 performs a marker deletion determination program 77. Fig. 8 is a plan view schematically showing the marker deletion determination program 77 and the marker deletion removal program 78.

The mark absence determination program 77 is a step of determining whether or not each of the mask marks 31, 41 completely enters the alignment opening Wm. Similarly, the step of determining whether or not the image of each mask mark 31, 41 is taken into the opening Wm for calibration is performed by pattern matching. As shown in fig. 8 (1), when it is determined that there is a missing of the pair of mask marks 31, 41 in the image data from the at least one camera 8, the mark missing determination program 77 returns an abnormal value, and if not, returns a normal value.

The mark deletion removal program 78 calculates the amount of movement (direction and distance) required to remove the deletion of the mask mark with respect to the image data in which the mark is missing in the mark deletion determination program 77. In this embodiment, since the first mask mark 31 is large, the mark deletion eliminating program 78 determines an arc determined to be a part of the first mask mark 31, and obtains the center of the arc. Next, the shortest movement amount (distance and direction) required to separate the center from the outline of the calibration aperture Wm by a distance equal to or greater than the radius (radius of the circular arc of the first mask mark 31) is obtained. The mark absence removing program 78 is programmed to send a signal to the mask moving mechanism 5 to move the pair of masks 3 and 4 by the movement amount. When there is a marker deficiency in the image data from two or more cameras 8, the marker deficiency removing program 78 calculates the amount of movement for each image data, and calculates the average value of these amounts. Since the movement amount is the distance and the direction, the average distance and the average direction are obtained. The calculated average movement amount is transmitted to the mask movement mechanism 5. As a result, the pair of masks 3 and 4 is moved, and the missing of the mask marks 31 and 41 is eliminated as shown in fig. 8 (2).

When the mark deletion eliminating program 78 is executed, the main sequence program 7 executes the mark deletion determining program 77 again to determine whether there is no mask mark deletion, and executes the main calibration program 79 when it is confirmed that the normal value is returned. Fig. 9 is a plan view schematically showing the main calibration performed by the main calibration program 79.

The main calibration program 79 processes the imaging data from each camera 8 in a state where the main calibration is possible. The main calibration program 79 first obtains the center of the first mask mark 31 and the center of the second mask mark 41 in a coordinate system with the point on the optical axis a as the origin. Then, it is determined whether or not the centers of the first mask mark 31 and the second mask mark 41 match with a desired accuracy, and if they do not match, a signal is sent to the mask moving mechanism 5 to move one or both of the masks to match. The two are usually brought into agreement with each other with a required accuracy at the time of the previous exposure, and this state is maintained.

After confirming that the centers of the first mask mark 31 and the second mask mark 41 agree with each other with a desired accuracy, the intermediate point of the centers is found by the main calibration routine 79. The main calibration program 79 calculates the center of the calibration opening Wm of the substrate W, calculates the offset from the center point of the center of the pair of mask marks 31, 41, and calculates the direction and distance of movement of each mask 3, 4 to cancel the offset.

The main calibration program 79 performs the above-described data processing on the imaging data from each camera 8, and calculates the direction and distance of movement of each mask 3, 4 for eliminating the deviation. Then, an average value is obtained for the direction and distance of movement obtained from each of the imaging data, and the average value is returned to the main sequence program 7 as a movement command for each of the masks 3 and 4 for final main calibration. Since the direction and distance of movement are grasped as respective vectors (indicated by arrows in fig. 9), the directions of the respective vectors are combined and the length is averaged.

The main sequence program 7 transmits a movement command as a return value to the mask movement mechanism 5, and integrally moves the pair of masks 3 and 4 so that the centers are aligned on a straight line with a desired accuracy. Thus, the actual calibration is ended. Although not shown in fig. 3, the main sequence program 7 performs a storage step of storing the coordinates of the centers of the mask marks 31 and 41 at the time of completion of the main calibration in the storage unit 60 for calibration at the time of exposure of the next target exposure region R.

By finally performing the main calibration routine 79 in this way, the pair of masks 3 and 4 are calibrated with each other and the pair of masks 3 and 4 are calibrated with respect to the substrate W. The main sequence program 7 is programmed to perform the respective determination steps as described above, and perform calibration while executing the respective subroutines as needed.

Next, the overall operation of the double-sided exposure apparatus according to the embodiment of the above-described configuration will be described in general. The following description is also a description of an embodiment of the invention of the double-sided exposure method. The invention of the double-sided exposure method can be referred to as a method for manufacturing an object such as a substrate having both sides exposed.

The pair of masks 3 and 4 are located at standby positions apart from the substrate W in the Z direction. This position is a position where the XY plane where the calibration of each mask 3, 4 is performed exists.

A control signal is sent from the main controller 6 that executes the main sequence program 7 to the conveyance system 1 to feed the substrate W by the amount of the feed stroke Lf. Thus, the first driving roller 15 and the second driving roller 16 operate synchronously, and the substrate W is fed forward in the X direction (winding side) by the feeding stroke Lf.

If a feeding completion signal is returned from the conveyance system 1 to the main controller 6, the main sequence program 7 performs the series of calibration operations described above. That is, whether or not the calibration opening Wm is present in the field of view of each camera 8 is determined, and if not, the opening search program 72 is executed, and then it is determined that the opening is missing. If any of the calibration openings Wm is missing, an opening missing removal program 74 is executed, and then whether or not the mark is blocked is determined. When there is a mark occlusion in certain image data, a temporary calibration program 76 is executed. Further, when the mask marks 31 and 41 are missing and photographed, the mark missing removal program 78 is executed. Then, the main sequence program 7 executes the formal calibration program 79. Thereby, the calibration is completed.

Then, the main sequence program 7 transmits a control signal to a not-shown Z-direction moving mechanism, and moves the pair of masks 3 and 4 in the Z-direction, thereby bringing the masks 3 and 4 into close contact with the substrate W. In this state, the main sequence program 7 acquires imaging data from each camera 8, and determines whether or not the calibrated state is maintained (whether or not the centers of the marks 31, 41, wm coincide with a required accuracy). If so, the main sequence program 7 sends a control signal to each exposure unit 2 to perform exposure.

After exposure for a prescribed time for the required exposure amount, each exposure unit 2 stops light irradiation. Then, the main sequence program 7 transmits a control signal to a not-shown Z-direction moving mechanism to separate the pair of masks 3 and 4 from the substrate W and return the masks to the first standby position.

If it is confirmed that each of the masks 3 and 4 has returned to the standby position, the main sequence program 7 transmits a control signal to the transport system 1 to feed the substrate W forward in the X direction by the feed stroke Lf. Then, the same operation as described above is repeated, and the exposure operation is performed after calibration during the intermittent feeding of the substrate W in the feeding stroke Lf.

When the substrate W is relaxed in the delivery-side buffer 101 during the repetitive operation, the delivery-side core roller 11 and the delivery-side pressure roller 12 operate synchronously, and the substrate W is delivered to the delivery-side buffer 101. Further, if the amount of slack of the substrate W in the winding-side buffer 102 increases, the winding-side core roller 13 and the winding-side pressure roller 14 operate synchronously, and the substrate W is wound around the winding-side core roller 13.

According to the two-sided exposure apparatus of the embodiment of the structure and operation described above, when the mask marks 31 and 41 are blocked by the substrate W, the temporary alignment mechanism calculates the movement amount based on the positions of the mask marks 31 and 41 in the previous alignment at the time of exposure, and performs the temporary alignment in which the masks 3 and 4 are moved by the movement amount, so that the blocking of the mask marks 31 and 41 is eliminated, and the main alignment can be performed normally.

In the case where the temporary alignment mechanism is not provided, as a structure for eliminating the shielding of the mask marks 31 and 41, a structure in which a pair of masks 3 and 4 are moved little by little in the X direction or the Y direction to search for the mask marks 3 and 4 may be considered. The pair of masks 3 and 4 is moved stepwise by a stroke slightly shorter than the width of the alignment opening Wm, and the mask marks 3 and 4 are searched. In such a configuration, it is necessary to move stepwise in the X direction and the Y direction, respectively. Since the time until the start of the main calibration becomes long, the total time required for one exposure becomes long, and there is a problem that productivity is lowered. On the other hand, if the temporary calibration mechanism is provided as in the embodiment, the shielding can be eliminated with one movement, so that the problem of the reduction in productivity does not occur.

In the above configuration, the pair of masks 3 and 4 are moved to remove the blocking, but the substrate W may be moved to remove the blocking. When the substrate W is moved to remove the blocking, the first driving roller 15 and the second driving roller 16 are operated in synchronization with each other in the X direction, and the substrate W is moved to the front side or the rear side in the X direction. In the Y direction, a Y-direction moving mechanism is provided to the first driving roller 15 and the second driving roller 16, and moves in the Y direction with the substrate W sandwiched between the first driving roller 15 and the second driving roller 16.

However, the conveyance system 1 is a mechanism for intermittent feeding of the substrate W, and the stroke of the intermittent feeding is considerably longer than the moving distance for eliminating the shielding. In the movement of a short subtle distance for eliminating the occlusion, the conveying system 1 is not suitable in terms of accuracy. Further, if the first and second driving rollers 15 and 16 are integrally moved for movement in the Y direction, meandering of the substrate W tends to occur, and the accuracy of the stop position of the substrate W after intermittent feeding also deteriorates. The mask moving mechanism 5 is more likely to move with high accuracy, and is preferable for performing calibration with high accuracy.

In the apparatus according to the embodiment, it is determined whether or not the alignment opening Wm of the substrate W is within the field of view of the camera 8 at the time of alignment after completion of intermittent feeding, and if the alignment opening Wm is not within the field of view of the camera 8, the substrate W is moved to bring the alignment opening Wm into the field of view of the camera 8, so that an error in alignment (incapability of alignment) due to incapability of photographing the alignment opening Wm is prevented. Therefore, even when the accuracy of the forming position of the alignment opening Wm is low or the accuracy of the intermittent feeding of the substrate W is low, the alignment cannot be performed, and the problem of the productivity degradation due to the abnormal stop of the apparatus is prevented. At this time, the camera 8 may be moved so that the alignment opening Wm is brought into the field of view, but the pair of masks 31 and 41 need to be moved together in many cases, and the movement distance is long, so that it is preferable to move the substrate W.

In addition, when there is a defect in the alignment opening Wm of the substrate W, the defect is eliminated and then the alignment is performed in a state where the entire alignment opening Wm is taken into the image data, so that there is an effect that the alignment accuracy is improved.

Further, since the mark missing determination is performed and the main calibration is performed in a state where the mark missing is eliminated in the case of the mark missing, the calibration is performed after the complete images of the pair of mask marks 31 and 41 are taken in, and therefore, there is an effect that the calibration accuracy is improved in this point.

In the above embodiment, the transport system 1 transports the substrate W in a roll-to-roll manner, but a structure in which only the delivery side is a roll may be adopted. That is, the double-sided exposure apparatus of the present invention may be used in a process of cutting the exposed substrate W at a predetermined position and performing a subsequent process.

In addition, as the transport system 1, there is a case where the feeding direction of the substrate W is the up-down direction. In this case, the exposure units 2 are disposed in the left and right sides so as to expose both sides of the substrate W in a vertical posture through the mask.

In the above embodiment, the calibration opening Wm is circular, but this is merely an example, and other shapes such as square and triangle may be used. In addition, the edge may not be entirely circumferential like a shape cut from the side edge of the substrate W.

The term "opening" is a term "opening" in the sense of passing light. It is assumed that the substrate W is light-shielding, and a typical example thereof is a case where a resist is applied. Since the opening is open in the sense of passing light, the light-transmitting member may be used instead of the through-hole. That is, it means the degree to which the light-shielding layer is opened.

The first mask mark 31 and the second mask mark 41 may have shapes other than a circular shape or a circular shape. For example, one may be circular and the other may be cross-shaped. In addition, calibration may be performed in a state where the first mask mark 31 is located inside the second mask mark 41.

Further, the mask mark on the side closer to the camera 8 than the substrate W is not blocked by the substrate W, and thus may be larger than the alignment opening Wm. However, when the contrast between the substrate W and the mask is small, there is a problem in that processing of image data becomes difficult. In the structure in which the alignment is performed in a state where the pair of mask marks are located in the alignment opening, the contrast between the substrate W and the mask marks is not a problem, which is preferable.

In the configuration of the temporary alignment mechanism described above, the positional information of the mask marks 31 and 41 used is positional information at the time of completion of alignment at the time of the previous exposure, but may be positional information during the previous exposure. When the accuracy of the intermittent feeding of the substrate W by the conveyor system 1 is not poor and the accuracy of the formation position of the alignment opening Wm is not poor, the mask marks 31 and 41 are not blocked when the intermittent feeding of the substrate W is completed, and the masks 3 and 4 are often not moved significantly during the alignment. When calibration and occlusion after exposure are repeated in such a state, occlusion may be eliminated by referring to position information during exposure several times before. In this case, the storage means may store all the positional information of the mask marks 31 and 41 at the completion of the calibration at each exposure (not the aforementioned configuration).

The device of the above embodiment performs exposure in a contact manner, but the above calibration structure also has an effect even in the proximity type or projection type exposure, and therefore, these methods may be employed.

In addition, in the case of the proximity method or the projection exposure method, since it is not necessary to bring a pair of masks into close contact with the substrate, there are cases where a mechanism for moving the masks in the Z direction is not provided.

The main controller 6 is an example of the control unit, but other configurations are possible. For example, a control unit may be provided separately from the main controller 6, or a part of the main controller 6 may correspond to the control unit.

Claims

1. A double-sided exposure apparatus is characterized in that,

the device is provided with:

a transport system that pulls out and intermittently feeds a flexible substrate, which is wound up in a roll, and in which a plurality of target exposure areas are set at intervals along a longitudinal direction;

a pair of first and second masks arranged at positions sandwiching a target exposure region of the substrate to be fed;

an exposure unit that irradiates light to target exposure areas on both sides of the substrate via each mask to perform exposure after the substrate is stopped and aligned by the transport system; and

A storage unit;

the substrate has an alignment opening provided in a predetermined positional relationship with respect to a region to be exposed;

the first mask has a first mask mark as a mark for calibration;

the second mask has a second mask mark as a mark for calibration;

a camera capable of photographing the first mask mark, the second mask mark, and the alignment opening of the substrate is provided;

a calibration mechanism for aligning the first mask and the second mask with respect to the region to be exposed of the substrate based on imaging data from a camera that images the first mask mark, the second mask mark, and the calibration opening;

the mask alignment device is provided with a storage means for storing the positions of the first mask marks and the positions of the second mask marks as previous mark positions in a storage unit, a temporary alignment means, and a mask movement means;

the temporary calibration mechanism is such that: when the alignment mechanism performs alignment, if the first mask mark or the second mask mark is not captured through the alignment opening of the substrate and is blocked by the substrate, the position of the previous mark at the completion of the alignment at the time of the previous exposure is acquired from the storage unit, and the direction and distance of the movement of the substrate for blocking elimination, the directions and distances of the movement of the first mask and the second mask are calculated based on the position, and the substrate is moved to the direction by the distance, or the first mask and the second mask are moved independently by the mask moving mechanism.

2. The two-sided exposure apparatus according to claim 1, wherein,

the mask moving mechanism is a mechanism capable of integrally moving the first and second masks.

3. A double-sided exposure method in which a flexible substrate, which is wound up in a roll and has a plurality of target exposure regions set at intervals in the longitudinal direction, is pulled out by a transport system and intermittently fed, and the substrate, which is fed and stopped, is exposed by irradiating light to the target exposure regions on both sides of the substrate by an exposure unit through a pair of first and second masks disposed with the substrate interposed therebetween;

the substrate has an opening for calibration provided in a predetermined positional relationship with respect to the target exposure region;

the first mask has a first mask mark as a mark for calibration;

the second mask has a second mask mark as a mark for calibration;

before exposure, performing alignment of the first mask mark, the second mask mark, and the alignment opening of the substrate with respect to the region to be exposed of the substrate based on the obtained imaging data while imaging the first mask mark, the second mask mark, and the alignment opening of the substrate with respect to the region to be exposed;

the method is characterized by comprising the following steps of:

a storage step of storing positions of the first mask mark and the second mask mark at the time of completion of calibration before exposure to a certain target exposure area as previous mark positions in a storage section;

A movement amount calculation step of, when calibration is performed before exposure of a next target exposure area located upstream in the feed direction from the target exposure area, acquiring a previous mark position from a storage unit and calculating, based on the position, an orientation and a distance of movement of the substrate or an orientation and a distance of movement of the first and second masks for eliminating the shielding of the first and second mask marks when the first and second mask marks are not captured through the calibration opening of the substrate and are shielded by the substrate in image data from the camera; and

and a movement step of removing the mask by moving the substrate in the direction and distance calculated in the movement amount calculation step or by moving the first and second masks in the direction and distance calculated in the movement amount calculation step independently.