Search

EP-3813100-B1 - ALIGNER AND CORRECTION VALUE CALCULATION METHOD FOR ALIGNER

EP3813100B1EP 3813100 B1EP3813100 B1EP 3813100B1EP-3813100-B1

Inventors

  • TAMATSUKURI, DAIGO
  • YAMAMOTO, YASUHARU
  • ANDOH, Katsumi

Dates

Publication Date
20260506
Application Date
20190620

Claims (20)

  1. A correction value calculation method for an aligner (1) including a motor (9) allowing rotation angle control, a driving pulley (10) provided with external teeth (10a) driven by the motor (9), a driven pulley (12) provided with external teeth (12a) configured at a predetermined rotation ratio with respect to the driving pulley (10), a toothed belt (13; 24) meshing with the driving pulley (10) and the driven pulley (12) and hung between the driving pulley (10) and the driven pulley (12), a spindle (8) concentrically fixed to the driven pulley (12) and having fixing means for fixing a semiconductor wafer (W), an alignment sensor (19) detecting a peripheral edge of the semiconductor wafer (W) fixed on the spindle (8), a pulley detection sensor (20) provided with a light projecting portion (19a) and a light receiving portion (19b), detecting a rotation position of the driven pulley (12) at a time when the driving pulley (10) is driven by the motor (9) and the driven pulley (12) is rotated once as the amount of light received by the light receiving portion (19b), and outputting a magnitude thereof as a detection value, and a control unit (14) provided with an input-output unit, a calculation unit (29), and a storage unit and controlling operation of each of the units, the correction value calculation method being for calculating a correction value for correcting misalignment in a direction of rotation of the driven pulley (12) attributable to a manufacturing error of the toothed belt (13; 24) by means of the control unit (14) and comprising: a reference value storage step of detecting the rotation position of the driven pulley (12) at a time when each of the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) is at a predetermined reference position by means of the pulley detection sensor (20) and storing the detected value in the control unit (14) as a reference value; a detection value storage step of executing the detection value of the pulley detection sensor (20) at a time when the motor (9) is caused to perform predetermined rotation operation for rotating the driven pulley (12) once until the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) return to the reference position without exception and storing the detection value in the control unit (14) for each rotation of the driven pulley (12); and a correction value calculation step of calculating the correction value for correcting the amount of deviation of the rotation position per rotation of the driven pulley (12) from the amount of change in the detection value stored in the control unit (14).
  2. The correction value calculation method for an aligner (1) according to claim 1, wherein the number of the external teeth (12a) of the driven pulley (12) is an integer multiple of the number of the external teeth (10a) of the driving pulley (10), the toothed belt (13; 24) is provided with internal teeth (13a), and the number of the internal teeth (13a) is a prime number.
  3. The correction value calculation method for an aligner (1) according to claim 2, wherein the correction value is a control quantity for controlling the rotation angle of the motor (9).
  4. The correction value calculation method for an aligner (1) according to any one of claims 1 to 3, comprising a reference detection data acquisition step of acquiring reference detection data by creating a graph in which a vertical axis represents the detection value or the correction value and a horizontal axis represents a number assigned to each internal tooth (13a) formed on the toothed belt (13; 24) based on the detection value detected in the reference value storage step and the detection value storage step or the correction value calculated in the correction value calculation step and calculating an approximate straight line for each predetermined range of the graph and a slope thereof.
  5. The correction value calculation method for an aligner (1) according to claim 4, further comprising: a calibration reference position storage step of performing origin search by operating the motor (9) with rotation angle information lost, detecting the rotation position of the driven pulley (12) after the origin search operation by means of the pulley detection sensor (20), and storing the detected value in the control unit (14); a calibration detection value storage step of causing the motor (9) to perform rotation operation for rotating the driven pulley (12) a predetermined number of times and then repeating the operation of detecting the rotation position of the driven pulley (12) by means of the pulley detection sensor (20) a predetermined number of times and storing in the control unit (14) a calibration detection value detected by the pulley detection sensor (20) each time the rotation operation is performed; a calibration reference detection data acquisition step of creating a graph of the calibration detection value stored in the control unit (14) and calculating a slope for each predetermined detection range of the graph; and a phase specifying step of specifying phases of the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) by comparing the graph created in the reference detection data acquisition step with the graph created in the calibration reference detection data acquisition step.
  6. A correction value calculation method for an aligner (1) including a motor (9) allowing rotation angle control, a driving pulley (10) driven by the motor (9), a driven pulley (12) configured at a predetermined rotation ratio with respect to the driving pulley (10), a toothed belt (13; 24) meshing with the driving pulley (10) and the driven pulley (12) and hung between the driving pulley (10) and the driven pulley (12), a spindle (8) concentrically fixed to the driven pulley (12) and having fixing means for fixing a semiconductor wafer (W), an alignment sensor (19) detecting a peripheral edge of the semiconductor wafer (W) fixed on the spindle (8), and a control unit (14), the correction value calculation method being for calculating a correction value for correcting misalignment in a direction of rotation of the driven pulley (12) attributable to a manufacturing error of the toothed belt (13; 24) and comprising: a fixing step of fixing the semiconductor wafer (W) at a predetermined position on the spindle (8) such that a center position of the semiconductor wafer (W) matches a rotation center axis (C2) of the spindle (8); a second reference value storage step of rotating the semiconductor wafer (W) by operating the motor (9), detecting a position of a notch (N) of the semiconductor wafer (W) by means of the alignment sensor (19), and storing the detected value in the control unit (14) as a reference position detection value; a second detection value storage step of causing the motor (9) to perform predetermined rotation operation for rotating the driven pulley (12) once and then repeating the detection operation of detecting the notch (N) position of the semiconductor wafer (W) by means of the alignment sensor (19) until the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) return to the respective reference positions without exception, and storing detection values per rotation without exception; and a second correction value calculation step of calculating the correction value for correcting the misalignment per rotation of the semiconductor wafer (W) from the detection value per rotation and the reference position detection value stored in the control unit (14).
  7. The correction value calculation method for an aligner (1) according to claim 6, comprising a second reference detection data acquisition step of acquiring reference detection data by creating a graph of the detection value or the correction value in which a vertical axis represents the detection value or the correction value and a horizontal axis represents a number assigned to each internal tooth (13a) formed on the toothed belt (13; 24) based on the detection value detected in the second detection value storage step and calculating an approximate curve for each predetermined range of the graph and a slope thereof.
  8. The correction value calculation method for an aligner (1) according to claim 7, further comprising: a second calibration reference position storage step of performing origin search by operating the motor (9) with rotation angle information lost, detecting a notch (N) of the semiconductor wafer (W) by means of the alignment sensor (19) after the origin search operation, and storing the detected value of the notch (N) in the control unit (14); a second calibration detection value storage step of causing the motor (9) to perform rotation operation for rotating the driven pulley (12) a predetermined number of times and then repeating the operation of detecting the notch position of the semiconductor wafer (W) by means of the alignment sensor (19) a predetermined number of times and storing in the control unit (14) a detection value detected by the alignment sensor (19) each time the rotation operation is performed; a second calibration reference detection data acquisition step of creating a graph of the detection value stored in the control unit (14) and calculating a slope for each predetermined detection range of the graph; and a second phase specifying step of specifying phases of the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) by comparing the graph created in the second reference detection data acquisition step with the graph created in the second calibration reference detection data acquisition step.
  9. The correction value calculation method for an aligner (1) according to any one of claims 6 to 8, wherein there is no common divisor other than 1 in the number of the internal teeth (13a) formed on the toothed belt (13; 24), the number of the external teeth (10a) formed on the driving pulley (10), and the number of the external teeth (12a) formed on the driven pulley (12).
  10. The correction value calculation method for an aligner (1) according to any one of claims 6 to 9, wherein the number of the internal teeth (13a) formed on the toothed belt (13; 24) is a prime number.
  11. The correction value calculation method for an aligner (1) according to any one of claims 6 to 8, wherein the number of the external teeth (12a) formed on the driven pulley (12) is an integer multiple of the number of the external teeth (10a) formed on the driving pulley (10).
  12. An aligner (1) comprising: a motor (9) allowing rotation angle control; a driving pulley (10) driven by the motor (9); a driven pulley (12) configured at a predetermined rotation ratio with respect to the driving pulley (10); a toothed belt (13; 24) meshing with the driving pulley (10) and the driven pulley (12) and hung between the driving pulley (10) and the driven pulley (12); a spindle (8) concentrically fixed to the driven pulley (12) and having fixing means for fixing a semiconductor wafer (W); an alignment sensor (19) detecting a peripheral edge of the semiconductor wafer (W) fixed on the spindle (8); a pulley detection sensor (20) detecting a rotation position of the driven pulley (12); and a control unit (14), wherein in calculating a correction value for correcting misalignment in a direction of rotation of the driven pulley (12) attributable to a manufacturing error of the toothed belt (13; 24), the control unit (14) includes: a first reference position detection unit (27) detecting the rotation position of the driven pulley (12) at a time when each of the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) is at a reference position by means of the pulley detection sensor (20) and storing the reference position detection value; a rotation position detection unit executing the operation of detecting the rotation position of the driven pulley (12) by means of the pulley detection sensor (20) until the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) return to the respective reference positions without exception after the motor (9) is caused to perform predetermined rotation operation for rotating the driven pulley (12) once and storing a rotation position detection value per rotation of the driven pulley (12); and a correction value calculation unit (29) calculating the correction value for correcting the misalignment per rotation of the driven pulley (12) from the reference position detection value and the rotation position detection value.
  13. The aligner (1) according to claim 12, wherein the control unit (14) further includes a graph creation unit creating a graph in which a vertical axis represents the detection value or the correction value and a horizontal axis represents a number assigned to each internal tooth (13a) formed on the toothed belt (13; 24) based on the calculated detection value and calculating an approximate straight line for each predetermined detection range of the graph and a slope thereof.
  14. The aligner (1) according to claim 13, wherein the control unit (14) includes: a second reference position detection unit (31) performing origin search by operating the motor (9) with rotation angle information lost, detecting the rotation position of the driven pulley (12) at a time when the driving pulley (10) has returned to the reference position by means of the pulley detection sensor (20), and storing the detected value; a calibration rotation position detection unit (32) causing the motor (9) to repeat a predetermined number of times rotation operation for rotating the driven pulley (12) a predetermined number of times and then detecting the rotation position of the driven pulley (12) by means of the pulley detection sensor (20) and storing the detected value; a calibration graph creation unit (33) creating a graph of the detection value detected by the calibration rotation position detection unit (32); and a phase specifying unit (34) calculating a slope for each predetermined detection range of the graph and specifying phases of the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) by comparing the graph created by the graph creation unit with the graph created by the calibration graph creation unit (33).
  15. The aligner (1) according to any one of claims 12 to 14, wherein there is no common divisor other than 1 in the number of the internal teeth (13a) formed on the toothed belt (13; 24), the number of the external teeth (10a) formed on the driving pulley (10), and the number of the external teeth (12a) formed on the driven pulley (12).
  16. The aligner (1) according to any one of claims 12 to 14, wherein the number of the internal teeth (13a) formed on the toothed belt (13; 24) is a prime number.
  17. The aligner (1) according to any one of claims 12 to 14, wherein the number of the external teeth (12a) formed on the driven pulley (12) is an integer multiple of the number of the external teeth (10a) formed on the driving pulley (10).
  18. A correction value calculation method for an aligner (1) according to claim 1 wherein the control unit (14) rotates the driving pulley (10) n times by means of the motor (9), detects the rotation position of the driven pulley (12) by means of the pulley detection sensor (20) for each of the rotation operations, executes the operation of detecting the rotation position of the driven pulley (12) by means of the pulley detection sensor (20) for each of the rotation operations of the driven pulley (12) by the number of internal teeth (13a) formed on the toothed belt (13; 24), and calculates, from a detection value detected by the pulley detection sensor (20), a correction value corresponding to the detection value of each rotation of the driven pulley (12), as a correction value corresponding to each rotation position of the toothed belt (13; 24).
  19. The correction value calculation method for an aligner (1) according to claim 18, wherein reference detection data is acquired by a graph in which a vertical axis represents the detection value or the correction value and a horizontal axis represents a number assigned to each internal tooth (13a) formed on the toothed belt (13; 24) being created based on each of the detection values and an approximate curve for each predetermined range of the graph and a slope thereof being calculated.
  20. The correction value calculation method for an aligner (1) according to claim 19, wherein in a case where rotation angle information is lost, origin search is performed by the motor (9) being operated, the rotation position of the driven pulley (12) after the origin search operation is detected by means of the pulley detection sensor (20), subsequently the motor (9) is caused to perform rotation operation for rotating the driven pulley (12) a predetermined number of times, the rotation position of the driven pulley (12) is detected by means of the pulley detection sensor (20), a graph of the detection value after the origin search is created, and then a slope is calculated for each predetermined detection range of the graph, and calibration detection data for specifying phases of the driving pulley (10), the driven pulley (12), and the toothed belt (13; 24) is acquired by the slope of the graph after the origin search being compared with a slope of a graph of the reference detection data.

Description

TECHNICAL FIELD The present invention relates to a technique for improving the alignment accuracy of a wafer aligner on which a semiconductor wafer is placed and the notch or orientation flat formed in the outer peripheral edge portion of the semiconductor wafer is detected for positioning at a predetermined rotation position. BACKGROUND ART In a semiconductor device manufacturing process, a plurality of semiconductor wafers as semiconductor device substrates are transported after being stored, in a clean room, in a closed multi-shelf container called a front opening unified pod (FOUP). The semiconductor wafer stored in the FOUP and transported is taken out of the FOUP in a highly clean atmosphere called a mini-environment space and undergoes various processes such as inspection and processing. In addition, in a process in which semiconductor wafer position information is essential, examples of which include various inspections and processing such as electronic circuit patterning, vapor deposition, and chemical vapor deposition, it is an important pre-stage work to always accurately position the notch portion formed in the outer peripheral edge portion of a semiconductor wafer, such as a notch and orientation flat, and the center point of the semiconductor wafer at predetermined positions. Accordingly, it is necessary to perform delivery to various processing and inspection devices after placing the semiconductor wafer on a wafer positioning device called an aligner, detecting the position of the center point of the semiconductor wafer and the position of the notch portion, and accurately moving the semiconductor wafer to a correct position before the processing, manufacturing, and inspection processes. Further, in recent years, positioning at an accuracy level higher than in the related art has been required from semiconductor wafer circuit pattern miniaturization. In general, the aligner is formed in a columnar shape and includes a base, a spindle as a wafer mount rotatably disposed on the base, a line sensor disposed at an end of the base and detecting the peripheral edge portion of a semiconductor wafer, and a spindle rotation mechanism rotating the spindle. Further, some aligners are provided with a spindle moving mechanism moving a spindle and a rotary driving unit in X-, Y-, and Z-axis directions. The spindle rotation mechanism is provided with a driving pulley fixed to the output shaft of a motor, a driven pulley attached to a support shaft fixed coaxially with the spindle, and a toothed belt hung around the driving pulley and the driven pulley. In addition, a servomotor or a stepping motor allowing easy driving shaft rotation angle control is used as the motor rotationally driving the spindle. The spindle is a wafer mount on which the semiconductor wafer is placed horizontally. An adsorption hole for adsorbing and holding a wafer W horizontally placed on the spindle is formed in the spindle. The adsorption hole is connected to a vacuum source via a piping member. In the aligner configured as described above, the semiconductor wafer is rotated by the motor with the semiconductor wafer placed on the spindle held and the line sensor measures the peripheral edge portion of the semiconductor wafer. As a result, the amount of deviation of the semiconductor wafer with respect to the spindle rotation center axis is accurately detected. However, in recent years, semiconductor design rules have become fine and aligners have been required to position semiconductor wafers more accurately than in the related art. Here, a pitch accuracy of the teeth on the toothed belt that does not meet the positioning accuracy required for the aligner is an obstacle to the positioning accuracy improvement. Accordingly, Patent Document 1 discloses a technique for improving positioning accuracy by attaching an encoder coaxially with the support shaft of a spindle and directly detecting the rotation position of the spindle. In addition, a technique for cutting a toothed belt into a plurality of belts and hanging the belts around a pulley with a phase shift so as to cancel periodic fluctuations in the pitch width of the belts as described in Patent Document 2 is disclosed as a method for mitigating misalignment attributable to the pitch width fluctuations of the internal teeth formed on the toothed belt. JP H06 249306 A discloses a method for calculating a correction value from a rotation amount difference attributable to the presence of tolerances in the diameter of two pulleys of a rotation transmission mechanism. JP 2008 002664 A discloses a detection method for identifying tooth skipping of a torque transmission device. CITATION LIST PATENT DOCUMENT Patent Document 1: JP 2002-164419 APatent Document 2: JP 2013-157462 A SUMMARY OF THE INVENTION PROBLEM TO BE SOLVED BY THE INVENTION However, although the encoder attachment has led to positional accuracy improvement in the technique described in Patent Document 1 described above,