JP2008010504A - Device and method of alignment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device and a method of alignment for accurately superposing a substrate and a mask. <P>SOLUTION: Virtual lines L3 and L4 for respectively connecting substrate marks 1a and 1a and substrate angle marks 1b and 1b of a glass substrate 1 are parallel to each other. Virtual points 1c and 1c located on respective virtual perpendicular lines L7 and L8 for the virtual lines L3 and L4 are respectively specified to be point-symmetrical with respect to a center point P1 of the glass substrate 1. While the glass substrate 1 and a metal mask 2 are overlappingly placed, the substrate marks 1a and 1a, the substrate angle marks 1b and 1b, and mask marks 2a and 2a are acquired as measurement images by cameras 4A and 4B. Coordinate values of the virtual points 1c and 1c and coordinate values of the mask marks 2a and 2a are calculated from the measurement images. Respective center coordinate values P11 and P21 of line segments L12 and L11 each connecting the coordinates are calculated from the coordinate values. The differential value of the coordinates is obtained. The differential value is used as a correction value to align the glass substrate 1 with the metal mask 2 in position. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an alignment apparatus and method for accurately overlapping a substrate and a mask.

  Conventionally, there is a method of forming a pattern on a substrate using a mask. For example, in the manufacturing process of an organic EL display, a technique is used in which a metal mask is stacked on a glass substrate and an organic EL layer is formed by vapor deposition. Specifically, the metal mask has a plurality of small holes corresponding to the size of the pixel (RGB compatible sub-pixel), and the pixel portion extends from the hole in a state where the metal mask is overlaid on the glass substrate. For this purpose, RGB organic dyes are vapor-deposited, and an organic EL layer is formed on a glass substrate. Then, for the three colors of RGB, the R dye, the G dye, and the B dye are deposited three times while sequentially shifting the metal mask by the pixel pitch, thereby forming an organic EL layer on the glass substrate surface.

  In order to manufacture a high-definition organic EL display, it is necessary to perform vapor deposition with high accuracy. For that purpose, it is important to perform alignment (alignment) between the glass substrate and the metal mask with high accuracy.

  For example, Patent Document 1 discloses an alignment apparatus that accurately aligns the glass substrate and the metal mask. In the following, the reference numerals in parentheses are those described in Patent Document 1.

  In this alignment apparatus (100), alignment between the glass substrate (1) and the deposition mask (2) is performed as follows. That is, on the glass substrate (1), a pair of substrate marks (1a) are written in the vicinity of both ends of the diagonal line, and a parallel reference is provided near each substrate mark (1a). Each mark (1b) is marked. The parallel reference mark (1b) is written at a position where virtual lines L1 and L2 connecting the substrate mark (1a) and the parallel reference mark (1b) are parallel to each other. On the other hand, a pair of mask marks (2a) are also opposed to the film forming mask (2) in the vicinity of both ends of the diagonal line, and the mask frame (3) is applied in a state where tension is applied in the horizontal direction. ) Is placed on top. Then, with the glass substrate (1) placed on the film formation mask (2), the two cameras 10, 10 respectively use the substrate mark (1a), the parallel reference mark (1b), and the mask mark. The mark (2a) is photographed, and the coordinate values of the substrate mark (1a) and the mask mark (2a) are calculated from the photographed image. After calculating the center coordinate values of the substrate marks (1a, 1a) and the mask marks (2a, 2a) from the calculated coordinate values, the coordinate difference values of the respective center coordinate values are obtained, Using this coordinate difference value as a correction value, alignment between the glass substrate (1) and the film formation mask (2) is performed.

  Here, in order to increase the pattern definition in recent years, the film-formation mask (2) is formed very thin, and the film-formation mask (2) can be bent even if it is simply placed on the mask frame (3). is there. Therefore, in this alignment apparatus (100), tension is applied to the film formation mask (2) in the horizontal direction, and the film formation mask (2) is fixed to the mask frame (3) in the stretched state to prevent the film formation mask (2) from bending. is doing. Further, when a tension is applied to the film formation mask (2), the position of the mask mark (2a) is shifted, so that the shift is caused by obtaining the coordinate difference value of the center coordinate value as described above. In this case, the film-formation mask (2) and the glass substrate (1) can be accurately aligned. As described above, in this alignment apparatus (100), the film forming mask (2) is prevented from being bent, and the alignment can be performed with high accuracy even when the mask mark (2a) is displaced.

  However, in this alignment apparatus (100), it is difficult for the camera 10 to recognize the mask mark (2a), and the coordinates of the mask mark (2a) may be erroneously recognized. More specifically, in this alignment apparatus (100), by aligning the mask mark (2a) with the substrate mark (1a), the glass substrate (1) and the film formation mask (2) are aligned. Although it is configured to perform alignment, what is actually being matched is the coordinate value of the center point of each mark. Therefore, in this alignment apparatus (100), after the mask mark (2a) and the substrate mark (1a) are photographed by the camera, the coordinate value of each center point is determined based on the photographed image. There is a need. Although the method for determining the coordinate value of the center point is not specifically disclosed in Patent Document 1, the normalized correlation method is frequently used. In this normalized correlation method, target points on an image are measured about 100 times, and whether or not 3σ (σ: standard deviation) falls within a predetermined accuracy range when applied to a normal distribution is determined. I do. For the determination of coordinate points, a method (model-based search) is used in which a registered model is scanned on the image plane for each measurement, and a point whose area ratio matches the model is scored and evaluated (model-based search). is there.

  On the other hand, in this alignment apparatus (100), since the purpose is to finally match the mask mark (2a) with the substrate mark (1a), the substrate mark ( There may occur a case where 1a) and the mask mark (2a) overlap. FIG. 9 shows the state. FIG. 9 shows the camera (10) from the substrate mark (1a), the parallel reference mark (1b), and the mask mark (2a) in a state where the glass substrate (1) and the film formation mask (2) are overlapped. It is a figure which shows the image for a measurement acquired by (1). Since the mask mark (2a) is located on the back side of the glass substrate (1) when viewed from the camera (10), it is difficult for the camera (10) to recognize the portion overlapping the substrate mark (1a). In such a state, if the coordinate point is determined by the model base search and the coordinate value of the center point is determined using the normalized correlation method, the coordinate value of the center point may not be obtained accurately, and as a result There is a problem that alignment takes time and alignment accuracy also decreases.

JP 2004-303559 A

  The present invention has been made to solve the above-described problems, and the object of the present invention is to prevent marks from overlapping each other and reduce the time required for alignment when aligning a substrate and a mask. Another object is to provide an alignment apparatus and method capable of improving the alignment accuracy.

  In order to achieve the above object, the present invention provides an alignment apparatus that performs alignment when a substrate and a mask are overlapped, and the substrate includes a substrate mark formed near one corner of the substrate mark, A substrate angle mark formed in the vicinity, and the substrate mark and the substrate angle mark are point-symmetric with respect to a center point of the substrate, and the substrate mark and the substrate angle mark are Are formed in the vicinity of the diagonal of the one corner so that virtual lines connecting the two are parallel to each other, the mask has a mask mark formed in the vicinity of each of the pair of diagonals, and the alignment apparatus is In a state where virtual point setting means for setting a virtual point that is located on a virtual perpendicular to the virtual line and is point-symmetric with respect to the center point of the substrate, and the substrate and the mask are overlaid, Image acquisition means for acquiring the substrate mark, the substrate angle mark, and the mask mark as a measurement image, the coordinate value of the substrate mark and the substrate angle from the measurement image acquired from the substrate mark and the substrate angle mark And calculating the coordinate value of the virtual mark from the coordinate value, calculating the coordinate value of the mask mark from the measurement image acquired from the mask mark, and further calculating the virtual point A coordinate value calculating means for calculating a first center point coordinate value on a first line segment connecting each other and a second center point coordinate value on a second line segment connecting the mask marks; and the first center A correction value calculation means for calculating a coordinate difference value between the point coordinate value and the second center point coordinate value, and a position correction for aligning the substrate and the mask using the coordinate difference value as a correction value. And having a means.

  Further, the present invention is an alignment method for performing alignment when a substrate and a mask are overlapped, and the substrate includes a substrate mark formed in the vicinity of the corner and a substrate formed in the vicinity of the substrate mark. An angle mark, and the substrate mark and the substrate angle mark are point-symmetric with respect to the center point of the substrate, and virtual lines connecting the substrate mark and the substrate angle mark are The mask is formed in the vicinity of the diagonal of the one corner so as to be parallel, the mask has mask marks formed in the vicinity of the pair of diagonals, and the alignment method is a virtual for the virtual line. A virtual point setting step for setting a virtual point that is located on a vertical line and is point-symmetric with respect to the center point of the substrate, and the substrate mark and the mask in a state in which the substrate and the mask are arranged in an overlapping manner. An image acquisition step of acquiring the plate angle mark and the mask mark as a measurement image, and the coordinate value of the substrate mark and the coordinate value of the substrate angle mark from the measurement image acquired from the substrate mark and the substrate angle mark Are calculated, the coordinate value of the virtual point is calculated from these coordinate values, the coordinate value of the mask mark is calculated from the measurement image acquired from the mask mark, and a first point connecting the virtual points is further calculated. A coordinate value calculating step of calculating a first center point coordinate value on a line segment and a second center point coordinate value on a second line segment connecting the mask marks; the first center point coordinate value; and A correction value calculation step for calculating a coordinate difference value with respect to the second center point coordinate value, and a position correction step for aligning the substrate and the mask using the coordinate difference value as a correction value. The features.

  As the substrate, a transparent glass substrate, an acrylic substrate, or the like is applicable. In addition, as the mask, a so-called metal mask made of metal or a mask using other materials is applicable, and this mask is provided with a slit for forming a pattern on the substrate. Pattern formation is performed by overlaying this mask on the substrate. Since the position of the pattern formed on the substrate is predetermined, alignment (alignment) is required so that the pattern is formed at a predetermined position on the substrate when the mask and the substrate are overlapped. In order to perform this alignment, a substrate mark and a substrate angle mark are formed on the substrate, while a mask mark is formed on the mask.

  Since the substrate mark, substrate angle mark, and mask mark are all marks necessary for alignment, any shape can be applied. However, in order to easily identify each mark, each mark It is better to have different shapes. For example, if the substrate mark is formed in a cross shape, the substrate angle mark is square, and the mask mark is formed in a circular shape, it is preferable that each is easily identified.

  As the shape of the substrate and the mask, for example, if the alignment between the substrate and the mask, such as a square shape or a rectangular shape, is possible and a pattern can be formed at a predetermined position on the substrate, what is the shape? Any shape is acceptable.

  In addition, since the mark is acquired as a measurement image by an image acquisition unit such as a CCD camera in a state where the substrate and the mask are overlapped, the substrate mark, the substrate angle mark, and the mask mark include the substrate and the mask. It must be formed at a position corresponding to the overlap. For example, when both the substrate and the mask are square, if the substrate mark and the substrate angle mark are formed in the vicinity of the upper left corner of the substrate, the mask mark also has the upper left corner when the mask and the substrate are overlapped. It must be formed in the vicinity. Furthermore, these marks are also provided point-symmetrically near the diagonal. That is, in the above case, these marks are also formed in point symmetry with respect to the center point of the substrate or mask in the vicinity of the lower right corner which is the diagonal of the upper left corner.

  The image acquisition means can acquire various images such as a CCD camera and a CMOS camera if the substrate mark, the substrate angle mark, and the mask mark can all be acquired as one measurement image when the substrate and the mask are overlapped. Means are applicable. Further, the installation position may be either on the substrate side or on the mask side as long as all the marks can be acquired as measurement images when the substrate and the mask are overlapped.

  Here, in the past, alignment was performed by matching the substrate mark and the mask mark based on the measurement image, but in the present invention, the virtual point and the mask mark are matched. Alignment is performed. This virtual point is located on a virtual perpendicular to a virtual line connecting the substrate mark and the substrate angle mark, and is set point-symmetrically with respect to the center point of the substrate, with this virtual line as a boundary. It can be set to either of the two areas to be divided. Further, the virtual line, the virtual perpendicular, and the virtual point are virtual ones obtained by calculation, and are not directly formed on the substrate or the mask.

  This virtual point is formed at a position where a conventional substrate mark is formed. Conventionally, since the substrate mark and the mask mark are aligned, that is, the mark formed on the substrate and the mark formed on the mask are aligned with each other, the first time when the substrate and the mask are overlapped, The marks sometimes overlapped. On the other hand, in the present invention, since the point calculated in calculation called a virtual point and the mask mark are matched, the mark does not overlap from the beginning when the substrate and the mask are overlapped.

  Here, each coordinate value from the measurement image acquired from the substrate mark, the substrate angle mark, and the mask mark, that is, the coordinate value at the center point of each mark is obtained by the above-described normalized correlation method, for example. In addition, you may obtain | require by the edge recognition method. This edge recognition method is a method for learning all the information on the edge of the shape or information on a part of the edge based on the brightness of the target shape as information on the image, and determining the center point of each point from the information It is.

  The coordinate value of the virtual point is calculated from the calculated coordinate value of the substrate mark and the substrate angle mark. For example, when a virtual point is set in advance on a virtual perpendicular that passes through the substrate mark, if the distance from the virtual line to the virtual point is set in advance, the coordinate value of the virtual point is determined from the coordinate value of the substrate mark and the substrate angle mark. Can be calculated.

  In the present invention, the alignment between the substrate and the mask includes the first center point coordinate value on the first line segment connecting the virtual points and the second center point coordinate value on the second line segment connecting the mask marks. , And a coordinate difference value of each center point coordinate value is calculated, and this coordinate difference value is used as a correction value. If the substrate and the mask are overlapped at a predetermined position from the beginning, the coordinate values of these center points coincide and the coordinate difference value becomes zero. However, if there is a positional deviation, the coordinate difference value does not become zero. Alignment is performed by calculating this coordinate difference value and setting it to zero. At this time, if the horizontal plane of the substrate is the XY plane, there are positional displacement in the XY direction and positional displacement in the direction around the axis centered on the central point of the substrate. This is done so as to correct the misalignment of both.

  In addition, the position correction unit may align the substrate and the mask so that the first line segment and the second line segment coincide with each other.

  Moreover, you may make it have a mask frame which can arrange | position the said mask by applying tension | tensile_strength in a horizontal direction.

  Further, the coordinate value calculation means calculates the coordinate values after correcting the inclination angle when the image acquisition means acquires the measurement image with an inclination angle with respect to the coordinate system. It may be.

  As described above, according to the present invention, the center point coordinate value between the substrate marks and the center point coordinate value between the mask marks are not matched with each other, but the center point coordinate value between the virtual points and The center point coordinate values of the mask marks are made to coincide with each other. For this reason, when the substrate is attached to the mask, it is possible to prevent a state in which the substrate mark and the mask mark are overlapped from the beginning as in the conventional case. As a result, in the present invention, erroneous recognition of the coordinate value of the mask mark by the camera when aligning the substrate and the mask mark can be eliminated, and the alignment time can be shortened and the alignment accuracy can be improved.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing a schematic configuration of an alignment apparatus of the present invention, FIG. 2 is a plan view of a glass substrate, and FIG. 3 is a plan view of a metal mask.

  The alignment apparatus in the present embodiment is configured to assume alignment between a glass substrate and a metal mask in a vacuum vapor deposition step of an organic EL layer at the time of manufacturing an organic EL display.

  In FIG. 1, the alignment apparatus 100 includes two cameras 4A and 4B, a camera I / F 5, a transport driving unit 6, a control unit 7, an image processing unit 8, and a holding unit 9, and includes a glass substrate 1 and a metal. Alignment with the mask 2, that is, alignment is performed.

  The glass substrate 1 is a transparent square substrate as shown in FIG. 2, and a substrate mark 1a is written at one end of a diagonal line L1 indicated by a broken line in the drawing, that is, near the upper left corner in FIG. Further, a substrate angle mark 1b is written in the vicinity of the substrate mark 1a. On the other hand, the substrate mark 1a and the substrate angle mark 1b are also written on the other end of the diagonal line L1, that is, in the vicinity of the lower right corner in FIG. In FIG. 2, L3 and L4 are virtual lines connecting the center point of the substrate mark 1a and the center point of the substrate angle mark 1b, and the substrate mark 1a and the substrate angle mark 1b are arranged so that the virtual lines L3 and L4 are parallel to each other. Oppositely written on the glass substrate 1. Furthermore, the substrate mark 1a and the substrate angle mark 1b are written on the glass substrate 1 so as to be point-symmetrically opposed to the center point P1 of the glass substrate 1, that is, the intersection of the diagonal lines L1 and L2. The diagonal lines L1 and L2 and the virtual lines L3 and L4 are not solid lines directly written on the glass substrate 1, but are virtual lines obtained by calculation.

  The metal mask 2 is a mask formed of a material such as stainless steel, titanium, or nickel, for example, and has a mask pattern 2b formed of a plurality of slits as shown in FIG. The metal mask 2 is a square mask that is slightly larger than the glass substrate 1, and mask marks 2a are opposed to both ends of a diagonal line L5 indicated by a broken line in the drawing, that is, in the vicinity of each pair of diagonal lines. Is provided. The mask mark 2a is formed by a circular hole, and is provided on the metal mask 2 so as to be point-symmetrically opposed to the center point P2 of the metal mask 2 that is the intersection of the diagonal lines L5 and L6. As will be described later, the mask mark 2a is a position to be photographed together with the substrate mark 1a and the substrate angle mark 1b when the state in which the glass substrate 1 is superimposed on the metal mask 2 is photographed with the cameras 4A and 4B. Is provided.

  Further, the metal mask 2 is arranged in a state in which tension is applied in the horizontal direction by a mask frame 3 formed in a square cylindrical structure. Since the tension is applied as described above, the metal mask 2 can be prevented from being bent.

  As shown in FIG. 1, the metal mask 2 is placed on the mask frame 3 in a state where tension is applied, and the glass substrate 1 is placed on the metal mask 2 so as to overlap. The glass substrate 1 may be disposed at a predetermined position of the metal mask 2 when stacked, but if the glass substrate 1 is not disposed at the predetermined position, the glass substrate 1 or the metal mask 2 is moved to correct the position. Therefore, it is preferable that the glass substrate 1 and the metal mask 2 are arranged so as to overlap each other through a minute gap so that the movement is easy.

  The cameras 4A and 4B are composed of a CCD camera, a CMOS camera, or the like. When the metal mask 2 and the glass substrate 1 are arranged so as to overlap each other as shown in FIG. 1, the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a Two units are arranged facing each other, that is, above the vicinity of both ends of the diagonal line L1 shown in FIG. The two cameras 4A and 4B are connected to the control unit 7 via the camera I / F 5. The cameras 4A and 4B function as image acquisition means for acquiring the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a as measurement images.

  The conveyance driving unit 6 is a power source for causing the holding unit 9 that holds the glass substrate 1 above the metal mask 2 to move in the two-dimensional X and Y axes and to rotate in the horizontal direction. It is configured. Here, the holding unit 9 is assumed to hold the glass substrate 1, but if the glass substrate 1 and the metal mask 2 can be aligned by the action of the holding unit 9, the holding unit 9 is the glass substrate 1. It is not necessary to hold. For example, a mechanism that holds the mask frame 3 and moves the metal mask 2 by moving the mask frame 3 to perform alignment may be used. In addition, under the control of the control unit 7, the conveyance driving unit 6 moves the holding unit 9 to a predetermined position, thereby serving as a position correction unit that aligns the glass substrate 1 and the metal mask 2.

  The control unit 7 is connected to the conveyance drive unit 6, the image processing unit 8, and the cameras 4 </ b> A and 4 </ b> B via the camera I / F 5, and controls the operation of the entire alignment apparatus 100.

  The image processing unit 8 includes a coordinate value calculation unit 81, a correction value calculation unit 82, and a storage unit 83. The coordinate value calculation means 81 calculates the substrate mark 1a, the substrate angle mark 1b, the coordinate value of the virtual point 1c described later, and the coordinate value of the mask mark 2a from the images taken by the cameras 4A and 4B, and the virtual value described later. The center point coordinate value of the line segment connecting the points 1c and the center point coordinate value of the line segment connecting the mask marks 2a are calculated. The correction value calculation means 82 is configured to calculate a correction value for position correction from each of the calculated center point coordinate values. The storage unit 83 is configured to temporarily store the images taken by the cameras 4A and 4B. As will be described later, the distance value between the substrate mark 1a and the substrate angle mark 1b, the virtual line L3, The distance value between each of L4 and the virtual point 1c and the direction of the virtual point 1c with respect to each of the virtual lines L3 and L4 are stored in advance. The virtual point 1c is set by the distance value and the direction stored in advance in the storage unit 83, and functions as a virtual point setting unit.

  FIG. 4A shows a measurement image acquired from the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a before alignment by the cameras 4A and 4B in a state in which the glass substrate 1 is placed on the metal mask 2. FIG. FIG. 4B is an enlarged view of a portion A1 in FIG. Since the virtual point 1c is a virtual point obtained by calculation, it is not photographed by the cameras 4A and 4B, but the positional relationship with each of the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a is shown. For the sake of clarity, FIG. 4 shows a virtual point 1c.

  In FIG. 4A, a positional deviation occurs between the glass substrate 1 and the metal mask 2, and the virtual point 1c and the mask mark 2a do not coincide. Here, the virtual point 1c is set as shown in FIG. 4B in the present embodiment. That is, it is set on a virtual perpendicular L7 that is perpendicular to the virtual line L3 and passes through the substrate mark 1a. This virtual point 1c is a distance of 1000 μm from the substrate mark 1a, and the virtual line in FIG. It is set on the upper side of the region divided vertically by L3. Here, the distance between the substrate mark 1a and the substrate angle mark 1b is preset to 1000 μm. Therefore, if the position coordinates of the substrate mark 1a and the substrate angle mark 1b can be acquired by the cameras 4A and 4B, the position coordinates of the virtual point 1c can also be calculated from these position coordinates.

  Further, the alignment apparatus 100 is configured not to match the substrate mark 1a and the mask mark 2a but to match the virtual point 1c and the mask mark 2a. Therefore, the substrate mark 1a and the mask mark 2a are not formed at positions where the glass substrate 1 and the metal mask 2 overlap each other at the beginning, as shown in FIGS. 4 (a) and 4 (b). The substrate mark 1a and the mask mark 2a are in a positional relationship that does not overlap.

  The distances described above are not limited to 1000 μm, and can be set by the user as appropriate depending on the size of the glass substrate 1, the size of the metal mask 2, the pattern of the slits 2b of the metal mask 2, and the like. The position of the virtual point 1c is not limited to the above, and may be set to the lower side, for example.

  Further, the position coordinates at the center points of the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a can be calculated by the above-described normalized correlation method or edge recognition method.

  The operation of the alignment apparatus 100 configured as described above will be described with reference to FIGS.

  5 is an explanatory diagram for correcting the inclination of the measurement image in FIG. 4, FIG. 6 is an explanatory diagram for obtaining the coordinate values of the mask mark 2a and the virtual point 1c, and FIG. 7 is the coordinate value of the mask mark 2a and the virtual value. FIG. 8 is an explanatory flowchart showing a method of performing alignment from the coordinate value of the point 1c, and FIG. 8 is an operation flowchart showing the alignment operation.

  As shown in FIG. 1, when the glass substrate 1 is transferred by a transfer means (not shown) and placed above the metal mask 2 set on the mask frame 3, the alignment operation shown in FIG. 8 starts.

  When this operation starts, first, under the operation control of the control unit 7, the imaging of the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a is started by each of the camera 4A and the camera 4B (S100, S200). . As shown in FIG. 4A, this image is taken as a measurement image (A1 and A2 portions in the figure), converted into a form that can be processed by the image processing unit 8 by the camera I / F 5, and controlled. The image is input to the image processing unit 8 via the unit 7.

  The image processing unit 8 checks whether all the marks in the camera field of each of the cameras 4A and 4B, that is, all of the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a are detected (S101, S201). The camera 4A, 4B continues shooting until the mark is detected (N in S101, N in S201), but if all the marks 1a, 1b, 2a are detected (Y in S101, Y in S201) The coordinate value calculation means 81 calculates each coordinate value from each mark 1a, 1b, 2a (S102, S202).

  Each coordinate value is a coordinate value at the center point of each mark, that is, the substrate mark 1a, the substrate angle mark 1b, and the mask mark 2a. This center point is obtained by the above-described normalized correlation method or edge recognition method. Since the virtual point 1c is a virtual point set on the glass substrate 1, it is not captured by the cameras 4A and 4B, but the positional relationship with the marks 1a, 1b, and 2a can be understood. 4A and 4B show a virtual point 1c.

  In S102 and S202, the coordinate value of the virtual point 1c is also calculated from the obtained coordinate values of the substrate mark 1a and the substrate angle mark 1b. A method of calculating the coordinate value of the virtual point 1c will be described with reference to FIG. In other words, as described above, in the storage unit 83, the distance between the substrate mark 1a and the substrate angle mark 1b, the distance between the virtual point 1c and the virtual line L3, and the virtual point 1c are positioned above the virtual line L3. Since the information on whether or not it is located in advance is stored in advance, the coordinate value of the virtual point 1c is obtained from these information and the coordinate values of the obtained substrate mark 1a and substrate angle mark 1b.

  When the coordinate values of the substrate mark 1a, the substrate angle mark 1b, the virtual point 1c, and the mask mark 2a are calculated as described above, whether or not the camera position correction is necessary in the image processing unit 8 next. (S103, S203). The marks 1a, 1b and 2a are usually affected by the mounting direction of the cameras 4A and 4B, and the marks 1a, 1b and 2a may be inclined as shown in FIG. If so, the inclination angle θ is obtained and the inclination is corrected (S104, S204).

  A method of obtaining the inclination angle θ will be described with reference to FIG. In FIG. 5, the marks 1a, 1b, 2a indicated by solid lines are measurement images captured by the cameras 4A, 4B, and the virtual point 1c indicated by the solid line is the virtual point 1c calculated from these images. Position. In FIG. 5, the virtual line L3 has an inclination of the inclination angle θ with respect to the camera coordinate system, and the marks 1a, 1b, 2a and the virtual point 1c are inclined by the inclination angle θ with respect to the camera coordinate system. Yes.

Thus, when each mark 1a, 1b, 2a is imaged with inclination with respect to the camera coordinate system, this inclination is corrected. Here, the inclination angle θ is calculated as follows. When the coordinate value of the substrate mark 1a is (X1, Y1), the coordinate value of the substrate angle mark 1b is (X2, Y2), and the coordinate value (X1, Y1) of the substrate mark 1a is the reference coordinate for correction, the inclination angle θ is obtained by the following Equation 1.
<Formula 1>
θ = arctan (Y1-Y2) / (X1-X2)

  In the camera coordinate system, when the tilt angle θ component is corrected, that is, when the coordinate values of the marks 1b and 2a and the virtual point 1c are rotated by θ in the direction of the arrow in FIG. 1b, 2a and the virtual point 1c have the coordinates and positional relationship indicated by the broken lines in FIG. Then, the coordinate values of the marks 1a, 1b, 2a and the virtual point 1c in the corrected camera coordinate system are calculated (S105, S205). As a result of obtaining the above θ, if θ = 0, the marks 1a, 1b, 2a are not inclined with respect to the camera coordinate system, so the correction is not performed, and the processes of S105 and S205 are not performed. The process proceeds to the next process S300 (N in S103, N in S203).

  Next, the camera coordinate system is converted to the substrate coordinate system (S300). Since the coordinate values of the marks 1a, 1b, 2a and the virtual point 1c obtained as described above are coordinate values based on the coordinate system that each of the cameras 4A, 4B has, the line segment L12 that connects the mask marks 2a described later. In order to obtain each coordinate value of the center point P11 of the line segment L11 connecting the virtual point 1c and the virtual point 1c, it is necessary to convert the camera coordinate system to the substrate coordinate system. A method of converting the coordinate system will be described with reference to FIG.

  Since the relative positional relationship between the substrate marks 1a and 1a on the glass substrate 1 does not change, the glass substrate 1 is used as the XY coordinate system, and the distance in the X direction and the Y direction between them is stored in the storage unit 83 in advance. The coordinate values of the substrate marks 1a and 1a in the substrate coordinate system are also stored in advance. For example, the distance between the substrate marks 1a and 1a on the glass substrate 1 is Xs in the X direction and Ys in the Y direction, and a straight line passing through one substrate mark 1a and a straight line passing through another substrate mark 1a are orthogonal to each other. Is the origin Os [0, 0] of the substrate coordinate system, the coordinates of the substrate marks 1a, 1a in the substrate coordinate system are [0, Ys], [Xs, 0], respectively. On the other hand, the coordinate values of the marks 1a, 1b and 2a and the virtual point 1c calculated by the camera coordinate system are set such that the coordinate values of the substrate marks 1a and 1a are [0, Ys] and [Xs, 0], respectively. Is converted into each coordinate value in the substrate coordinate system.

  An image of these processes is shown in FIG. 6, and on the measurement images C <b> 1 and C <b> 2 acquired in the camera coordinate system at positions of coordinate values [0, Ys] and [Xs, 0] in the substrate coordinate system. If the substrate marks 1a, 1a are respectively arranged, the coordinate values of the marks 1a, 1b, 2a and the virtual point 1c are converted from the camera coordinate system to the substrate coordinate system.

  In FIG. 6, (Xc3, Yc3), (Xc4, Yc4), (Xc1, Yc1), and (Xc2, Yc2) are described as the coordinate value of the virtual point 1c and the coordinate value of the mask mark 2a, respectively. However, these are all coordinate values in the camera coordinate system, and when these are converted into coordinate values in the substrate coordinate system, they are as shown in FIG. That is, (Xc3, Yc3) is (Xs8, -Ys8), (Xc4, Yc4) is (Xs9, -Ys9), (Xc1, Yc1) is (Xs5, -Ys5), (Xc2, Yc2) is Each is converted to (Xs6, -Ys6).

  Next, from the coordinate values of the virtual points 1c, 1c and the mask marks 2a, 2a converted into the substrate coordinate system, the center point P11 of the first line segment L11 connecting the virtual points 1c, 1c, the mask mark 2a, Each center point coordinate value of the center point P21 of the second line segment L12 connecting 2a is calculated (S301).

  Here, the coordinate value (Xs10, -Ys10) of the center point P11 of the first line segment L11 and the coordinate value (Xs7, -Ys7) of the center point P21 of the second line segment are obtained by the following equations 2 and 3.

<Formula 2>
(Xs10, -Ys10) = {(Xs9-Xs8) / 2, (-Ys9-(-Ys8)) / 2}

<Formula 3>
(Xs7, -Ys7) = {(Xs6-Xs5) / 2, (-Ys6-(-Ys5)) / 2}

  Next, it is checked whether or not the coordinate values of the center points P11 and P21 calculated by the above equations 2 and 3 match (S302). If they do not match (N in S302), a coordinate difference value is calculated. (S303) On the other hand, if they match (Y in S302), the coordinate difference value = 0, so the coordinate difference value is not calculated. This coordinate difference value is calculated as a positional deviation in the X coordinate direction and a positional deviation in the Y coordinate direction.

  Furthermore, whether or not the coordinate difference value is calculated and the coordinate difference value is not calculated, whether the inclination angle α1 of the first line segment L11 and the inclination angle α2 of the second line segment L12 coincide with each other. (S304). As shown in FIG. 7, even if the coordinate values of the center points P11 and P21 coincide with each other (intersection of the first line segment L11 and the second line segment L12 indicated by a broken line), the first line segment L11 and the second line segment L11 The line segment L12 may not match. That is, there is a positional shift in the direction around the axis centered on the coincident center point. If there is this positional deviation, the positions of the virtual point 1c and the mask mark 2a do not match as shown in FIG. 7, and the first line segment L11 and the second line segment L12 also do not match. It is also necessary to obtain a positional deviation (hereinafter referred to as “a positional deviation in the rotational direction”).

  Therefore, in the present embodiment, the inclination angle α1 of the first line segment L11 with respect to the X axis and the inclination angle α2 of the second line segment L12 with respect to the X axis are obtained, and the difference between α1 and α2 of each inclination angle is obtained. Thus, the positional deviation in the rotational direction is calculated. The inclination angles α1 and α2 are obtained by the following equations 4 and 5.

<Formula 4>
α1 = arctan {(− Ys9 − (− Ys8)) / (Xs9−Xs8)}

<Formula 5>
α2 = arctan {(− Ys6-(− Ys5)) / (Xs6-Xs5)}

  As a result of calculating the inclination angles α1 and α2 by the above equations 4 and 5, if the inclination angles α1 and α2 do not match (N in S304), an angle difference value which is the difference between the inclination angles α1 and α2 is obtained (S305). . On the other hand, when the inclination angles α1 and α2 match (Y in S304), α1 = α2, and it is not necessary to calculate the angle difference value.

  Next, it is finally checked whether or not there is a difference value (S306). That is, if either the coordinate difference value or the angle difference value is present (Y in S306), alignment is necessary, so the glass substrate 1 and the metal mask 2 are aligned (S307). On the other hand, if none of the difference values exists (N in S306), it is not necessary to perform alignment, and the process ends.

  For example, when this alignment procedure is imaged, s1 and s2 shown in FIG. 7 are obtained. That is, when there is a positional shift in the coordinate values of the center points P11 and P21, the second line segment L12 is translated in the direction of the first line segment L11 so that the center point P21 and the center point P11 coincide (s1). ). Next, the difference between the inclination angles α1 and α2 is obtained, and the second line segment L12 is rotated around the axis centered on the center point that is the coincident center point by the obtained angle difference (s2). Then, the virtual points 1c and 1c coincide with the mask marks 2a and 2a, and the alignment process ends.

  When the metal mask 2 is stretched by the tension of the mask frame 3, the virtual points 1c and 1c may not match the mask marks 2a and 2a even if the alignment process is performed. However, even in such a case, if the first line segment L11 and the second line segment L12 are made to coincide with each other by the alignment process, the pattern formed by the slit 2b of the metal mask 2 is a predetermined one of the glass substrate 1. Therefore, if the first line segment L11 and the second line segment L12 are matched, high-definition alignment can be realized.

  By the way, in embodiment mentioned above, although the alignment of the glass substrate and a metal mask in the vacuum evaporation process of the organic EL layer in the case of manufacture of an organic EL display was demonstrated, of course, as an object of vacuum evaporation, The present invention is not limited to this organic EL material. For example, when manufacturing a semiconductor device, it can be applied to various devices for forming a conductor pattern on a substrate using a mask in a vacuum vessel.

1 is a block diagram showing a schematic configuration of an alignment apparatus according to an embodiment. The top view of the glass substrate of this embodiment. The top view of the metal mask of this embodiment. (A) is a top view which shows the image for a measurement acquired with a camera, (b) is an enlarged view of the image for a measurement. Explanatory drawing for correct | amending the inclination of the image for a measurement in FIG. Explanatory drawing for calculating | requiring the coordinate value of a mask mark and a virtual point. Explanatory drawing which shows the method of aligning from the coordinate value of a mask mark, and the coordinate value of a virtual point. The flowchart figure which shows the operation | movement of alignment. The top view which shows when a camera acquires the state which the mask mark and the board | substrate mark overlapped as a measurement image when using the conventional alignment method.

Explanation of symbols

1 Glass substrate 1a Substrate mark 1b Substrate angle mark 1c Virtual point 2 Metal mask 2a Mask mark 2b Slit 3 Mask frame 4A, 4B Camera 5 Camera I / F
6 Transport Drive Unit 7 Control Unit 8 Image Processing Unit 81 Coordinate Value Calculation Unit 82 Correction Value Calculation Unit 83 Storage Unit 9 Holding Unit L1, L2, L5, L6 Diagonal L3, L4 Virtual Line L7, L8 Virtual Perpendicular L11 First Line Segment L12 Second line segment P1 Center point of glass substrate P2 Center point of metal mask P11, P21 Center point α1, α2 Inclination angle C1, C2 Measurement image

Claims (3)

An alignment apparatus that performs alignment when stacking a substrate and a mask,
The substrate is
A substrate mark formed in the vicinity of the corner;
A substrate angle mark formed in the vicinity of the substrate mark,
Further, the substrate mark and the substrate angle mark are point-symmetric with respect to the center point of the substrate, and the imaginary lines connecting the substrate mark and the substrate angle mark are parallel to each other. It is also formed near the corner,
The mask is
Having a mask mark formed in the vicinity of each of the pair of diagonals;
The alignment device is
Virtual point setting means for setting a virtual point that is located on a virtual perpendicular to the virtual line and that is point-symmetric with respect to the center point of the substrate;
Image acquisition means for acquiring the substrate mark, the substrate angle mark, and the mask mark as a measurement image in a state in which the substrate and the mask are arranged in an overlapping manner;
While calculating the coordinate value of the substrate mark and the coordinate value of the substrate angle mark from the measurement image acquired from the substrate mark and the substrate angle mark, respectively, and calculating the coordinate value of the virtual point from these coordinate values, A coordinate value of the mask mark is calculated from the measurement image acquired from the mask mark, and further, a second center point that connects the first center point coordinate value on the first line segment that connects the virtual points to each other and the mask mark. Coordinate value calculating means for calculating a second center point coordinate value on the line segment;
Correction value calculating means for calculating a coordinate difference value between the first center point coordinate value and the second center point coordinate value;
An alignment apparatus comprising: position correction means for aligning the substrate and the mask using the coordinate difference value as a correction value.   The alignment apparatus according to claim 1, wherein the position correcting unit aligns the substrate and the mask so that the first line segment and the second line segment coincide with each other. An alignment method for performing alignment when stacking a substrate and a mask,
The substrate is
A substrate mark formed in the vicinity of the corner;
A substrate angle mark formed in the vicinity of the substrate mark,
Further, the substrate mark and the substrate angle mark are point-symmetric with respect to the center point of the substrate, and the imaginary lines connecting the substrate mark and the substrate angle mark are parallel to each other. It is also formed near the corner,
The mask is
Having a mask mark formed in the vicinity of each of the pair of diagonals;
The alignment method is
A virtual point setting step of setting a virtual point that is located on a virtual perpendicular to the virtual line and that is point-symmetric with respect to the center point of the substrate;
An image acquisition step of acquiring the substrate mark, the substrate angle mark, and the mask mark as a measurement image in a state in which the substrate and the mask are arranged in an overlapping manner,
While calculating the coordinate value of the substrate mark and the coordinate value of the substrate angle mark from the measurement image acquired from the substrate mark and the substrate angle mark, respectively, and calculating the coordinate value of the virtual point from these coordinate values, The coordinate value of the mask mark is calculated from the measurement image acquired from the mask mark, and further, the second central point coordinate value on the first line segment connecting the virtual points and the second mask point connecting the mask marks. A coordinate value calculating step of calculating a second center point coordinate value on the line segment;
A correction value calculating step of calculating a coordinate difference value between the first center point coordinate value and the second center point coordinate value;
And a position correction step of aligning the substrate and the mask using the coordinate difference value as a correction value.

Images