JP2009010307A - Bonding apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a bonding apparatus for correcting displacement in X-axis, Y-axis and θ-axis directions of a component imaging means and a substrate imaging means for accurate alignment of an electronic component. <P>SOLUTION: The bonding apparatus includes the component imaging means, the substrate imaging means and a target 8 with an alignment mark. The target 8 and the both imaging means are relatively movable. A relative positional relation between the both imaging means is calculated from an image picked-up with the both imaging means coaxially positioned with the target 8 interposed therebetween to align a substrate with the electronic component based on the calculated relative positional relation. The alignment mark on the target 8 has two intersecting points apart from each other by a predetermined distance. The positions of two intersecting points are recognized from each picked-up image to calculate the relative positional relation between both imaging means from their displacement in the X-axis, Y-axis and rotating directions. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an improvement of a bonding apparatus for bonding an electronic component such as a semiconductor chip to a substrate, and has been developed mainly for alignment of the electronic component and the substrate.

  In the flip chip bonder, a change over time such as expansion and contraction occurs in each part of the apparatus due to a change in ambient temperature. Due to this change, a shift also occurs in the relative positional relationship between the component recognition imaging means and the board recognition imaging means in the electronic component / substrate alignment apparatus. When such positional relationship of the image pickup means is deviated, the bonding position is also deviated.

  In order to solve this problem, as disclosed in Japanese Patent Application Laid-Open No. 2004-228688, the amount of deviation in the X-axis direction and the Y-axis direction of the component recognition imaging unit and the board recognition imaging unit is detected and corrected to obtain an accurate position. A bonding apparatus for matching is already disclosed by the present applicant. However, in order to improve bonding accuracy, it is not sufficient to correct the shift amount only in the X-axis direction and the Y-axis direction as in the prior art, and the shift amount in the θ-axis direction (rotation direction) cannot be ignored. It is coming.

Japanese Patent No. 2780000

  The present invention corrects the deviation amount in the θ-axis direction (rotation direction) in addition to the deviation amount correction in the X-axis direction and the Y-axis direction of the image recognition means for component recognition and the imaging means for board recognition. An object of the present invention is to provide a bonding apparatus capable of aligning an electronic component and a substrate.

The first invention employs the following means in the bonding apparatus in order to solve the above problems.
First, a component recognition imaging unit that images the electronic component sucked and held by the bonding head from below, a substrate recognition imaging unit that images a substrate placed on the bonding stage from above, and a component recognition imaging And a target having an alignment mark disposed between the imaging means for substrate recognition.
Secondly, the target and the both imaging means are installed so as to be relatively movable together with the above-described configuration.
Third, the relative positional relationship between the two image pickup means is calculated from the images picked up by the two image pickup means while the component recognition image pickup means and the board recognition image pickup means are positioned coaxially with the target interposed therebetween. Based on the relationship, a bonding apparatus that performs bonding by aligning the substrate whose position is recognized by the substrate recognition imaging unit and the electronic component whose position is recognized by the component recognition imaging unit.
Fourth, the alignment mark of the target has two intersections that are separated by a predetermined distance.
Fifthly, by recognizing the positions of the two intersections from the images picked up by the respective image pickup means, the amount of deviation in the orthogonal X-axis direction and Y-axis direction, the plane formed in the X-axis direction and the Y-axis direction The relative positional relationship between the two image pickup means consisting of the amount of deviation in the rotation direction is calculated.

  In a second aspect based on the first aspect, the image pickup position is set at two locations, and the two image pickup means pick up one intersection at one image pickup position and then a predetermined distance to the other image pickup position. It is a bonding apparatus to which a shift amount is calculated by moving and imaging the other intersection.

  The present invention recognizes the position of two intersections at a predetermined distance from the target from the images by the component recognition imaging means and the board recognition imaging means, so that the orthogonal X-axis direction and Y-axis direction, It has become possible to calculate the amount of deviation between the two image pickup means in the rotational direction in the plane formed by the X-axis direction and the Y-axis direction. As a result, a highly accurate semiconductor chip can be mounted on the substrate.

  According to the second aspect of the invention, the two image pickup means, after imaging one intersection, move a predetermined distance and calculate the amount of deviation by imaging the other intersection. The distance between the intersections can be made long, and the amount of deviation can be calculated with high accuracy especially in the rotation direction.

  Hereinafter, embodiments of the present invention will be described together with examples according to the drawings. FIG. 1 shows an embodiment of a flip chip bonding apparatus used in the present invention, and is a schematic perspective view showing a state in which the bonding head 1 is above the bonding position in the flip chip bonding apparatus, and FIG. FIG. 3 is an explanatory plan view, and FIG. 3 is an explanatory front view thereof.

  In the bonding apparatus, a semiconductor chip as an electronic component is sucked and held, and the bonding head 1 for bonding the semiconductor chip to the substrate 2 (shown in FIGS. 2 and 3) and the position of the substrate 2 are recognized from above. A substrate recognition camera 3 (which serves as an imaging means for substrate recognition in the present invention) and a Y-axis drive that enables the bonding head 1 and the substrate recognition camera 3 to reciprocate in the Y-axis direction (the left-right direction in FIGS. 2 and 3). A mechanism 4, a substrate stage 5 that supports the substrate 2 at the bonding position, and an X-axis drive mechanism 6 that enables the substrate stage 5 to move in the X-axis direction (direction orthogonal to the movement direction of the Y-axis drive mechanism 4). , A chip recognition camera 7 for recognizing the semiconductor chip attracted and held by the bonding head 1 from below (becomes an imaging means for component recognition of the present invention), a target 8, It is provided.

The bonding head 1 includes a head Z-axis drive mechanism 9 that controls movement in the vertical direction and a head θ-axis drive mechanism 10 that controls movement (rotation) in the rotation direction, and is mounted on the movable plate 11 of the Y-axis drive mechanism 4. Has been. Therefore, the bonding head 1 can move in the Y-axis direction, the Z-axis direction, and the θ-axis direction. The bonding head 1 in the embodiment does not have a drive mechanism in the X-axis direction. Therefore, when the semiconductor chip is bonded to the substrate 2, the relative movement in the X-axis direction during alignment is performed by the X-axis drive mechanism 6.
Although not shown, a chip supply unit for supplying semiconductor chips to the bonding head 1 is disposed on the right front side of the substrate transport guide 30 in FIG. 2, and this chip supply unit is formed by dicing the wafer. There are provided a chip support part for supporting a plurality of semiconductor chips, a pickup head for picking up one semiconductor chip from the chip support part and transferring it to the bonding head 1 by turning it upside down.

  The substrate recognition camera 3 is a camera that can recognize the substrate 2 and has a camera field of view 20 in which two intersections 18 and 19 of the target 8 can be recognized simultaneously from above as shown in FIG. The substrate recognition camera 3 includes a Z-axis drive mechanism 16 and is attached to the movable plate 12 of the Y-axis drive mechanism 4. Therefore, the substrate recognition camera 3 can only move in the Z-axis direction and the Y-axis direction, and the movement in the X-axis direction at the time of substrate recognition is performed by the X-axis drive mechanism 6, and the target 8 and the chip The relative movement of the recognition camera 7 to the reference position in the X-axis direction is performed by a drive mechanism in the X-axis direction of the target 8 and the chip recognition camera 7. The substrate recognition camera 3 and the bonding head 1 are installed at the same position in the X-axis direction.

  For the Y-axis drive mechanism 4, a linear motor including a fixed guide 13 linearly arranged in the Y-axis direction and movable plates 11 and 12 that run in the Y-axis direction along the fixed guide 13 is used. The substrate stage 6 can be reciprocated between the bonding position and the substrate supply position by the X-axis drive mechanism 6, and the substrate transport guide 30 is provided to be able to move up and down along the movement path of the substrate stage 6. When the substrate 2 is conveyed on the substrate conveyance guide 30 to the substrate supply position by a conveyance claw (not shown), the substrate conveyance guide 30 is lowered and placed on the substrate stage 6 and is moved to the bonding position by the X-axis drive mechanism 6. It has become so. After the completion of bonding, the substrate transport guide 30 is raised, so that the substrate 2 is transferred from the substrate stage 6 to the substrate transport guide 30 and unloaded from the bonding position by a transport claw (not shown).

  The chip recognition camera 7 can recognize the semiconductor chip attracted and held by the bonding head 1 from below and has a camera field of view 20 that can simultaneously recognize the two intersections 18 and 19 of the target 8 from below as shown in FIG. It has a camera. The chip recognition camera 7 is fixed to the XY drive table 21 so as to be movable in the X-axis direction and the Y-axis direction.

  The target 8 has a disk shape, and an alignment mark 15 that can be recognized from the front and back is drawn. The first embodiment of the alignment mark 15 shown in FIG. 5 is composed of three orthogonal vertical and horizontal straight lines, a central intersection 17 for setting a reference position, a first intersection 18 and a first intersection 18 which are located away from the central intersection 17. Two intersection points 19 are set. Here, the three linear alignment marks 15A, 15B, 15C in the X-axis direction are parallel at equal intervals, and the three linear alignment marks 15D, 15E, 15F in the Y-axis direction are also parallel at equal intervals. The middle point of the line segment connecting the first intersection 18 and the second intersection 19 is the central intersection 17.

  The target 8 has an X-axis cylinder 14 that can be moved back and forth to a predetermined position in the X-axis direction, and can be moved below the substrate recognition camera 3 on the same axis during calibration. Yes. Thereby, the board | substrate recognition camera 3 and the chip | tip recognition camera 7 can be located on the same axis on both sides of the target 8.

Next, the correction operation according to the embodiment will be described.
1 to 3, the rod of the X-axis cylinder 14 of the target 8 is in a contracted state, and the target 8 is in the retracted position. Further, as shown in FIG. 2, the substrate recognition camera 3 and the chip recognition camera 7 are separated at the same position at the X-axis position and different positions at the Y-axis position. From this state, the reference position is set as the first stage.

  From the above state, the X-axis cylinder 14 operates, the rod extends, and the target 8 enters a predetermined position and stops. The stop position is the same position as the X-axis position of the substrate recognition camera 3. Thereafter, the Y-axis drive mechanism 4 operates, the substrate recognition camera 3 moves above the target 8, and the movement of the substrate recognition camera 3 stops at a position where the center of the substrate recognition camera 3 and the center intersection 17 of the target 8 coincide. . Next, the XY drive table 21 of the chip recognition camera 7 operates, moves below the target 8, and stops at a position where the center of the chip recognition camera 7 coincides with the center intersection 17 of the target 8. In this state, as shown in FIG. 4, the center of the substrate recognition camera 3, the center of the chip recognition camera 7, and the center intersection 17 of the target 8 are located on the same axis.

  When the center of the substrate recognition camera 3 and the chip recognition camera 7 and the center intersection 17 of the target 8 are located coaxially, the substrate recognition camera 3 and the chip recognition camera 7 set the first intersection 18 and the second intersection 19 of the target 8. Take an image. After imaging, the coordinates of the first intersection 18 and the second intersection 19 are obtained by image processing, and the inclination angle of the straight line connecting the two points with the center point coordinate (reference position) between the two points is calculated from the coordinates of the two points. The calculated center point coordinates are the reference position, and the tilt angle is the reference angle, which are recorded in the storage device as the reference positions of the substrate recognition camera 3, the chip recognition camera 7, and the target 8, and both imaging means (substrates) at the reference position The relative positional relationship between the recognition camera 3 and the chip recognition camera 7) is also recognized. The above is the reference position setting operation as the first stage.

  Subsequently, as a second stage, a semiconductor chip is bonded to the substrate 2 by a normal method for a predetermined number of times or a predetermined time. That is, the position of the semiconductor chip attracted to the bonding head 1 is recognized by the chip recognition camera 7, the position of the substrate 2 placed on the substrate stage 6 is recognized by the substrate recognition camera 3, and based on the calculation. The semiconductor chip is bonded onto the substrate 2 at the predetermined position.

  As a third stage, a change in relative positional relationship is detected after a predetermined number of times or a predetermined time of bonding operation. That is, the imaging position is set near or at the reference position stored in the reference position setting operation. In principle, it is preferable to set the imaging position as the reference position. In this embodiment, the imaging position is set as the reference position. The substrate recognition camera 3, the chip recognition camera 7, and the target 8 are moved to the imaging position, and the first intersection 18 and the second intersection 19 are captured by both recognition cameras 3, 7.

  The coordinates of the first intersection 18 and the second intersection 19 are obtained by image processing, and the inclination angle of the straight line connecting the two points with the center point coordinate between the two points is calculated from the coordinates of the two points. At this time, if the center position between the two points is changed from the reference position (the center position between the two points detected by the reference position setting operation), it is influenced by the reasons such as thermal expansion, that is, the X-axis direction and the Y-axis. Axial misalignment. Furthermore, if the inclination angle of the straight line connecting the two points changes from the reference angle, it means that a deviation in the θ-axis direction (rotation direction) has occurred.

  This deviation amount is detected as a relative deviation amount between the target 8 and the substrate recognition camera 3 and as a relative deviation amount between the target 8 and the chip recognition camera 7. Then, the relative displacement between the target 8 and the substrate recognition camera 3 and the relative displacement between the target 8 and the chip recognition camera 7 are calculated, and the relative displacement between the substrate recognition camera 3 and the chip recognition camera 7 is calculated. Recognized as a relative positional relationship. This deviation amount is used as a correction amount to remove an error during bonding.

  The above has described an example using the alignment mark 15 of the first embodiment. That is, the first intersection 18 and the second intersection 19 enter the camera field of view 20 of the substrate recognition camera 3 and the chip recognition camera 7, and the alignment marks 15 that can recognize the positions of both the intersections 18 and 19 by one imaging are provided. Using. On the other hand, as shown in FIG. 6, the alignment mark 25 of the second embodiment in which the first intersection 28 and the second intersection 29 are set at a relatively long distance in the Y-axis direction that does not enter the same camera field of view 20 is shown. It is also possible to calculate the shift amount by moving the substrate recognition camera 3 and the chip recognition camera 7 in the Y-axis direction using the target 8 that is provided.

  When the target 8 of the alignment mark 25 of the second embodiment is used, the operation is as follows. First, the centers of the substrate recognition camera 3 and chip recognition camera 7 and the center intersection 27 of the target 8 are positioned on the same axis. From there, the center of the substrate recognition camera 3 and the chip recognition camera 7 and the first intersection 28 of the target 8 are moved coaxially, and the first intersection of the target 8 is detected by the substrate recognition camera 3 and the chip recognition camera 7. 28 is imaged. After imaging, the coordinates of the first intersection point 28 are obtained by image processing.

  Subsequently, both the cameras 3 and 7 are moved by a predetermined distance in the Y-axis direction, the center of both the cameras 3 and 7 and the second intersection 29 of the target 8 are positioned on the same axis, and the second intersection is detected by both the cameras 3 and 7. 29 is imaged, and the coordinates of the second intersection point 29 are obtained after imaging. From the coordinates of these two points, the inclination angle of the straight line connecting the two points with the center point coordinate (reference position) between the two points is calculated. The calculated center point coordinates and tilt angle are recorded in the storage device as reference positions and reference angles of the substrate recognition camera 3, the chip recognition camera 7 and the target 8, and both image pickup means (substrate recognition camera 3 and The relative positional relationship of the chip recognition camera 7) is also recognized.

  Subsequently, as in the case of the first embodiment of the alignment mark 15, as a second step, a semiconductor chip is bonded to the substrate 2 by a normal method for a predetermined number of times or for a predetermined time. That is, the position of the semiconductor chip attracted to the bonding head 1 is recognized by the chip recognition camera 7, the position of the substrate 2 placed on the substrate stage 6 is recognized by the substrate recognition camera 3, and based on the calculation. The semiconductor chip is bonded onto the substrate 2 at the predetermined position.

  As a third stage, a change in relative positional relationship is detected after a predetermined number of times or a predetermined time of bonding operation. At this time, the first intersection 28 and the second intersection 29 are imaged by both the cameras 3 and 7, but two imaging positions are provided, and the imaging position of the first intersection 28 is the first when the reference point is set. It is set to be the same as or near the coordinate position of the intersection 28. Similarly, the imaging position of the second intersection point 29 is set to be the same as or near the coordinate position of the second intersection point 29 when the reference position is set.

  Here, the center of the substrate recognition camera 3 and the chip recognition camera 7 and the first intersection 28 of the target 8 are moved to the imaging position, and when they are located on the same axis, the substrate recognition camera 3 and the chip recognition camera 7 use the first of the targets 8. One intersection 28 is imaged. After imaging, the coordinates of the first intersection point 28 are obtained by image processing.

  Subsequently, both the cameras 3 and 7 are moved by a predetermined distance in the Y-axis direction, the center of both the cameras 3 and 7 and the second intersection 29 of the target 8 are coaxially positioned, 29 is imaged, and the coordinates of the second intersection point 29 are obtained after imaging. From the coordinates of the two points, the center point coordinate between the two points and the inclination angle of the straight line connecting the two points are calculated. At this time, if the center position between the two points has changed from the reference position (the center position between the two points detected by the reference position setting operation), a deviation in the X-axis direction and the Y-axis direction has occurred. Furthermore, if the inclination angle of the straight line connecting the two points changes from the reference angle, it means that a deviation in the θ-axis direction (rotation direction) has occurred.

  This deviation amount is detected as a relative deviation amount between the target 8 and the substrate recognition camera 3 and as a relative deviation amount between the target 8 and the chip recognition camera 7. Then, the relative displacement between the target 8 and the substrate recognition camera 3 and the relative displacement between the target 8 and the chip recognition camera 7 are calculated, and the relative displacement between the substrate recognition camera 3 and the chip recognition camera 7 is calculated. Recognized as a relative positional relationship. This deviation amount is used as a correction amount to remove an error during bonding.

  Even when the target 8 having the alignment mark 25 of the second embodiment is used, it is possible to cope with one imaging position by widening the camera field 20 of the substrate recognition camera 3 and the chip recognition camera 3. The wider the spread, the greater the distortion of the peripheral portion, causing a problem in accurate position recognition. Therefore, the substrate recognition camera 3 and the chip recognition camera 7 having a small camera field of view 20 are used, and the distance between the two points is further increased. As a result, it is possible to calculate the shift amount more accurately.

An explanatory perspective view showing an outline of a flip chip bonding apparatus. Same plane explanatory drawing Front explanatory drawing Front explanatory drawing during target observation Plane explanatory drawing of 1st Example of an alignment mark Plane explanatory drawing of 2nd Example of an alignment mark

Explanation of symbols

1. . . . . . Bonding head . . . . . Substrate 3. . . . . . Substrate recognition camera . . . . . 4. Y-axis drive mechanism . . . . . Substrate stage 6. . . . . . 6. X-axis drive mechanism . . . . . Chip recognition camera 8. . . . . . Target 9. . . . . . Head Z axis drive mechanism 10. . . . . Head θ-axis drive mechanism 11,12. . . Movable plate 13. . . . . Fixed guide 14. . . . . X-axis cylinder 15, 25. Alignment mark 16. . . . . Z-axis drive mechanism of substrate recognition camera 17, 27. Central intersection 18,28. First intersection 19,29. Second intersection 20. . . . . Camera view 21. . . . . XY drive table 30. . . . . Board transfer guide

Claims (2)

Image recognition means for recognizing components from below to image electronic components sucked and held by the bonding head;
A substrate recognition image pickup means for picking up an image of a substrate placed on the bonding stage from above, and a target having an alignment mark arranged between the component recognition image pickup means and the substrate recognition image pickup means; Both imaging means are installed so as to be relatively movable,
The relative positional relationship between the two imaging means is calculated from the images captured by the two imaging means in a state where the imaging means for component recognition and the imaging means for board recognition are positioned coaxially with the target interposed therebetween, and based on the relative positional relationship In a bonding apparatus for aligning and bonding a substrate whose position is recognized by the image recognition means for substrate recognition and an electronic component whose position is recognized by the image recognition means for component recognition,
The target alignment mark has two intersections separated by a predetermined distance,
By recognizing the position of the two intersections from the images imaged by each imaging means,
The relative positional relationship between the two image pickup means is calculated, which includes a deviation amount in the X-axis direction and the Y-axis direction orthogonal to each other and a deviation amount in the rotation direction in a plane formed by the X-axis direction and the Y-axis direction. Bonding equipment.
  The imaging position is set at two locations, and both the imaging means capture an image of one intersection at one imaging position, and then move to a different distance to image another intersection to image the other intersection. The bonding apparatus according to claim 1, wherein:

Images