JP2009158875A - Alignment method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent coordinate shift due to an error in detecting a detection point. <P>SOLUTION: A wafer stage 11 is moved to the designed coordinate 52 of a first alignment(S1), and the coordinate of a first alignment mark 81 detected by an image obtained by a camera 31 is recorded (S5). The wafer stage 11 is moved to the designed coordinate 53 of a second alignment (S6), and the coordinate of a second alignment mark 91 detected by an image obtained by the camera 31 is recorded (S10). A shift amount at each point is calculated from the difference between the respective designed coordinate of alignment 52 and the coordinates of respective alignment marks 81, 91, and the shift in the XYθ direction is corrected based on this shift amount (S11). The wafer stage 11 is moved to an edge coordinate using a corrected standard coordinate 51 as a standard (S12), and an edge 55 of a wafer 22 is detected from the image obtained by the camera 31 (S13). The standard coordinate 51 is corrected based on the detected coordinate of the edge 55 and the designed edge coordinate (S17). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an alignment method for aligning coordinates indicating the position of a wafer.

  Conventionally, a wafer stage provided in a bonding apparatus is configured to support a wafer ring, and a wafer is attached on a wafer sheet stretched on the wafer ring.

  In this bonding apparatus, alignment of coordinates indicating the position of the wafer is performed in a pre-process for starting bonding (for example, Patent Document 1).

  As this alignment method, two sufficiently separated images in the wafer, for example, images of street intersections, are registered as alignment marks in association with design coordinates indicating the positions.

  Then, the wafer is moved in accordance with the design coordinates, and pattern matching is performed with the registered alignment mark of the image of the portion, and the deviation amount between the coordinates that can be matched with the design coordinates is calculated. Next, the amount of deviation in the XYθ direction of the wafer is calculated from the amount of deviation calculated at both design coordinates.

In the correction stage, the XY coordinate deviation is corrected by parameter correction, while the θ deviation is corrected by the rotation of the θ stage. Thus, bonding is started after alignment between the reference coordinates recognized by the bonding apparatus and the position coordinates indicated by the reference position on the wafer.
JP-A-7-263294

  However, in such a conventional alignment method, when pattern matching is performed, there are cases where two patterns such as street intersection images exist in the field of view of the camera.

  At this time, if two patterns are detected in both design coordinates set at different positions, and a pattern located on the same direction side is erroneously detected as a detection point by pattern matching, between the two detection points The distance is as designed. For this reason, it cannot be processed as an error.

  In this case, the reference coordinates are corrected based on the erroneously detected detection points, and there is a fear that coordinate deviation occurs.

  The present invention has been made in view of such a conventional problem, and an object of the present invention is to provide an alignment method capable of preventing coordinate shift due to erroneous detection of a detection point.

  In order to solve the above-mentioned problem, in the alignment method of the present invention, the detection point is detected near the position of the wafer indicated by the different design coordinates calculated from the reference coordinates, the coordinates of each detection point are acquired, and both detections are performed. In the alignment method for correcting the deviation of the reference coordinates based on the coordinates acquired at the points and the corresponding design coordinates, the coordinates of the edge of the wafer near the position of the edge coordinates of the wafer calculated from the corrected reference coordinates , And the reference coordinates are corrected based on the coordinates acquired at the edge and the edge coordinates.

  That is, detection points are detected in the vicinity of the wafer position indicated by different design coordinates calculated from predetermined reference coordinates, and coordinates are acquired at each detection point. Then, based on the coordinates acquired at both detection points and the design coordinates corresponding thereto, the deviation of the reference coordinates is corrected.

  At this time, when a plurality of locations indicating detection points are detected in the vicinity of the position of the wafer indicated by the both design coordinates, and an erroneous location located on the same direction side is detected as a detection point, an error is detected. In other words, the reference coordinates may be corrected to an incorrect value.

  Therefore, the edge of the wafer is detected in the vicinity of the position of the edge coordinate of the wafer calculated using the corrected reference coordinate, the edge coordinate is acquired, and the acquired coordinate and the edge coordinate are obtained. To correct the reference coordinates. At this time, there is one edge of the wafer, and a plurality of edges are not detected in the vicinity of the portion indicated by the edge coordinates.

  Therefore, the reference coordinates are not corrected based on the coordinates of the edge detected in error, and even if the reference coordinates are corrected to an incorrect value in the previous process, The value is corrected to an appropriate value for detecting the position on the wafer.

  As described above, in the alignment method of the present invention, the reference coordinates corrected based on the coordinates detected at different detection points on the wafer and the corresponding design coordinates are corrected to incorrect values. Even in such a case, the reference coordinates can be corrected to an appropriate value by correcting the reference coordinates based on the coordinates acquired at the edge of the wafer and the edge coordinates.

  For this reason, in the vicinity of the position of the wafer indicated by the different design coordinates, the reference coordinates are erroneously detected when an error cannot be detected by detecting an erroneous location located in the same direction from a plurality of detection points. Thus, it is possible to save the trouble of having to perform re-measurement after confirming the measurement result as in the conventional case where the value is corrected.

  As a result, it is possible to reduce unnecessary man-hours in the production line and improve productivity.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing a bonding apparatus 1 that executes an alignment method according to the present embodiment. The bonding apparatus 1 is an apparatus that picks up a chip with a bonding head outside the figure, transfers it to a workpiece, and bonds it. .

  The bonding apparatus 1 is provided with a wafer stage 11, and the wafer stage 11 is configured to hold a wafer ring 12 in an exchangeable manner.

  A wafer sheet 21 is stretched on the wafer ring 12, and a thin disk-shaped wafer 22 is stuck on the wafer sheet 21 as shown in FIG. The wafer 22 is diced in the vertical direction and the horizontal direction, and a plurality of chips 23 are formed on the wafer sheet 21. Between the chips 23,..., Streets 24,... Extending vertically and horizontally are formed, and the adjacent chips 23,.

  In the present embodiment, rectangular chips 23,... That are long in the vertical direction are shown as an example in FIG. 2, and the wafer 22 on which the vertically long chips 23,. A case where alignment is performed will be described as an example.

  The wafer stage 11 is configured by a combination of an XY stage and a rotary stage, and each stage is configured to be operated by a motor constituting a drive unit. Thereby, the wafer stage 11 is configured to move the held wafer ring 12 in the X direction and the Y direction and to rotate in the θ direction.

  As shown in FIG. 1, a camera 31 is fixed to the upper portion of the wafer stage 11 so that an image of the wafer sheet 21 held on the wafer stage 11 can be acquired. The camera 31 is set at a high magnification, and is configured so that a predetermined portion of the wafer 22 can be magnified and imaged.

  The camera 31 and the drive unit of the drive mechanism of the wafer stage 11 are connected to a control unit 41 mainly composed of a microcomputer incorporating a ROM and a RAM, and the microcomputer operates according to a program stored in the ROM. Thus, the alignment process is executed.

  In the RAM of the control unit 41, as shown in FIG. 2, reference coordinates 51 serving as a reference for moving the wafer stage 11 and design positions of detection points used during alignment are shown. The first alignment design coordinates 52 and the second alignment design coordinates 53, alignment image data 54 for detecting detection points by pattern matching at these design coordinate positions, and the position of the design edge 55 of the wafer 22 are shown. As the alignment image data 54, an intersection 61 of streets 24, 24 extending vertically and horizontally is given as an example. In the present embodiment, the intersection 61 is a detection point. And registered in the alignment image data 54.

  The operation of the bonding apparatus 1 according to the present embodiment having the above configuration will be described with reference to the flowchart shown in FIG.

  That is, when the control unit 41 operates according to a program stored in the ROM and an alignment process is called from the main routine, the wafer stage 11 is moved to the first alignment with reference to the reference coordinates 51 recorded in the RAM. The wafer 22 moves relative to the design coordinates 52, and the part of the wafer 22 located at the first alignment design coordinates 52 is arranged directly below the camera 31 (S1).

  Then, the camera 31 images the part of the wafer 22 located at the first alignment design coordinates 52, and pattern-matches the acquired image with the alignment image data 54 stored in the RAM, whereby the vertical and horizontal streets 24, 24 are obtained. Is detected as a first alignment mark 81 that is a detection point (S2).

  At this time, if the intersection 61 is not detected in the acquired image (S3), error processing such as setting an error flag is performed (S4), and then the process returns to the main routine while the acquired image is returned. When the intersection 61 is detected (S3), the intersection 61 is used as the first alignment mark 81, and its coordinates are associated with the first alignment mark 81 and stored in the RAM (S5).

  At this time, the chip 23 of the wafer 22 is formed in a vertically long rectangular shape, and two intersections 61 of streets 24 and 24 indicating the first alignment mark 81 are adjacent to each other in the X direction in the acquired image. May appear. In this case, in the pattern matching described above, one of the intersections 61 is detected as the first alignment mark 81, and the coordinates of the first alignment mark 81 are associated with the first alignment mark 81 and stored in the RAM. The

  Here, in the present embodiment, although the intersection 61 appearing on the X direction − side is the designed first alignment mark, the intersection 61 appearing on the X direction + side is the above-described pattern matching in the pattern matching described above. The following will be described by taking the case where the first alignment mark 81 is detected as an example.

  Next, the wafer stage 11 is moved relative to the second alignment design coordinate 53 with reference to the reference coordinate 51 recorded in the RAM, and the part of the wafer 22 located at the second alignment design coordinate 53 Is placed directly below the camera 31 (S6).

  Then, the camera 31 images the part of the wafer 22 located at the second alignment design coordinate 53, and pattern-matches this acquired image with the alignment image data 54 stored in the RAM, thereby causing the vertical and horizontal streets 24, 24 to be matched. Is detected as a second alignment mark 91 (S7).

  At this time, if the intersection 61 is not detected in the acquired image (S8), error processing such as setting an error flag is performed (S9), and then the process returns to the main routine, while the acquired image When the intersection 61 is detected (S8), the intersection 61 is used as the second alignment mark 91, and its coordinates are stored in the RAM in association with the second alignment mark 91 (S10).

  At this time, the chip 23 of the wafer 22 is formed in a vertically long rectangular shape, and two intersections 61 of streets 24 and 24 indicating the second alignment mark 91 are adjacent in the X direction in the acquired image. May appear. In this case, in the pattern matching described above, one of the intersections 61 is detected as the second alignment mark 91, and the coordinates of the second alignment mark 91 are associated with the second alignment mark 91 and stored in the RAM. The

  Here, in the present embodiment, although the intersection 61 appearing on the X direction − side is the designed second alignment mark, the intersection 61 appearing on the X direction + side is the above-described pattern matching. The following will be described by taking the case where the second alignment mark 91 is detected as an example.

  At this time, as described above, the first alignment mark 81 and the second alignment mark 91 both detect the intersection 61 appearing on the X direction + side as the first alignment mark 81 and the second alignment mark 91. Thus, the distance between the alignment marks 81 and 91 in the X direction is the same as the designed distance.

  For this reason, although the intersections 61 and 61 which are not in the design alignment mark position are detected as the first and second alignment marks 81 and 91, no error occurs, and the alignment marks 81 and 91 are not generated. Calculations using the coordinates of are performed.

  That is, the difference between the first alignment design coordinates 52 and the coordinates of the first alignment mark 81 stored in the RAM, and the second alignment design coordinates 53 and the second alignment mark 91 stored in the RAM. The amount of deviation at each location is calculated from the difference from the coordinates, and the amounts of deviation in the X direction, Y direction, and θ direction are corrected based on the amount of deviation (S11).

  More specifically, the rotation direction of the wafer 22 is determined from the inclination between the design straight line connecting the alignment design coordinates 52 and 53 and the actual straight line connecting the coordinates detected by the alignment marks 81 and 91. Detect misalignment. Further, the amount of deviation in the X direction and the amount of deviation in the Y direction of the wafer 22 in a state where the deviation in the rotational direction is eliminated are calculated.

  Thereby, the deviation in the rotation direction is corrected by operating the rotation stage constituting the wafer stage 11 and rotating the wafer 22, and the deviation in the X direction and the Y direction is corrected in the RAM. Each of the stored reference coordinates 51 is corrected by subtracting it based on the amount of deviation in each direction.

  Next, on the basis of the corrected reference coordinates 51, the wafer stage 11 starts to move in the X direction and moves relatively to the edge coordinates stored in the RAM. The position of the edge 55 of the wafer 22 is arranged immediately below the camera 31 (S12), the part of the wafer 22 located at the edge coordinates is imaged by the camera 31, and the edge 55 of the wafer 22 is detected from the acquired image. (S13).

  At this time, when the edge 55 is not detected in the acquired image (S14), the wafer 22 is stepped on the XY stage from the edge coordinate in the negative X direction, that is, in the Y direction. The motor is moved by 1 pitch (S15), and the camera 31 at this position is imaged and the edge 55 is detected from the acquired image (S13).

  Also at this time, if the edge 55 is not detected in the acquired image (S14), the stepping motor of the XY stage is moved by one pitch in the + direction of the X direction of the edge coordinate as the second time as the second time. Then (S15), the imaging by the camera 31 at this position and the detection of the edge 55 from the acquired image are performed (S13).

  Also in this case, when the edge 55 is not detected in the acquired image (S14), the stepping motor of the XY stage moves the wafer 22 in the negative X direction of the edge coordinate, that is, the Y direction as the third time. Is moved by 2 pitches (S15), and the camera 31 at this position is imaged and the edge 55 is detected from the acquired image (S13).

  Also in this case, if the edge 55 is not detected in the acquired image (S14), the stepping motor of the XY stage is moved by 2 pitches in the + direction of the X direction of the edge coordinate as the fourth time as the fourth time. Then (S15), the imaging by the camera 31 at this position and the detection of the edge 55 from the acquired image are performed (S13).

  Also in this case, when the edge 55 is not detected in the acquired image (S14), error processing such as setting an error flag is performed (S16), and the process returns to the main routine.

  On the other hand, when the edge 55 is detected in the acquired image between these steps S13 to S15 (S13), the coordinates of the detected edge 55 are acquired, and the acquired coordinates and the design point are designed. The reference coordinates 51 are corrected based on the edge coordinates.

  More specifically, the reference coordinate stored in the RAM is calculated by calculating the amount of deviation in the X direction and the amount of deviation in the Y direction of the edge 55 from the detected coordinates of the edge 55 and design edge coordinates. The reference coordinate 51 is corrected by subtracting the amount of deviation in the X direction and the amount of deviation in the Y direction from 51 (S17), and the process returns to the main routine.

  At this time, in the step S5 for detecting the coordinates at the first alignment mark 81 and the step S10 for detecting the coordinates at the second alignment mark 91, intersections 61 and 61 appear at two adjacent positions, and The intersections 61 and 61 on the same direction side, that is, the X direction + side, which is an erroneous place in the present embodiment, are detected as the first and second alignment marks 81 and 91.

  For this reason, these are not detected as errors, and the reference coordinates 51 are corrected to incorrect values in the step S11. However, in steps S12 to S17, the position on the wafer 22 is detected by correcting the reference coordinates 51 based on the coordinates acquired at the edge 55 of the wafer 22 and the edge coordinates. Therefore, the reference coordinate 51 can be corrected to an appropriate value.

  Therefore, in the vicinity of the position of the wafer 22 indicated by the different design coordinates 52 and 53, a plurality of intersections 61 and 61 as detection points are detected, and the intersection 61 at the wrong location located on the same direction side is the alignment marks 81 and 91. If the error cannot be detected by the detection, the reference coordinate 51 is corrected to an incorrect value. As in the prior art, it is necessary to perform remeasurement after confirming the measurement result. Save.

  As a result, it is possible to reduce unnecessary man-hours in the production line and improve productivity.

It is a block diagram which shows one embodiment of this invention. It is explanatory drawing which shows the embodiment. It is a flowchart which shows the operation | movement of the embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Bonding apparatus 22 Wafer 23 Chip 41 Control part 51 Reference coordinate 52 First alignment design coordinate 53 Second alignment design coordinate 54 Alignment image data 55 Edge 81 First alignment mark 91 Second alignment mark

Claims (1)

The detection point is detected near the position of the wafer indicated by the different design coordinates calculated from the reference coordinates, and the coordinates of each detection point are acquired, and the reference coordinates based on the coordinates acquired at both detection points and the corresponding design coordinates. In the alignment method for correcting the deviation,
An alignment method for acquiring the coordinates of the edge of the wafer near the position of the edge coordinates of the wafer calculated from the corrected reference coordinates, and correcting the reference coordinates based on the coordinates acquired at the edge and the edge coordinates .

Images