Reference portion filtering automatic wafer centering process and related systems

Abstract

Each sensor in the sensor array detects when the edge of a wafer passes by the sensor on the wafer handling component of the robot arm and sends a signal. A number (N) of detected wafer edge positions are determined. Each detected wafer edge position is a set of coordinates (x, y) in the coordinate system of the wafer handling component. For each unique set (N - 1) of the number (N) of detected wafer edge positions, an estimated wafer offset that substantially minimizes a figure of merit value is determined. The estimated wafer offset is a vector extending from the center of the coordinate system of the wafer handling component to the estimated center position of the wafer. The final wafer offset is identified as the estimated wafer offset having the smallest corresponding figure of merit value. The final wafer offset is used to center the wafer at the target workstation.

Description

Technical Field

[0001] The present disclosure relates to semiconductor device manufacturing. Background Art

[0002] In the manufacture of semiconductor devices such as integrated circuits, memory cells, etc., a series of manufacturing operations are performed to define features on a semiconductor wafer (hereinafter referred to as "wafer"). The wafer includes integrated circuit devices in the form of a multi-layer structure defined on a silicon substrate. In the substrate layer, transistor devices with diffusion regions are formed. In subsequent layers, patterned interconnect metallization lines are electrically connected to the transistor devices to define the desired integrated circuit devices. The patterned conductive layers are also insulated from other conductive layers by dielectric materials.

[0003] Many and various wafer manufacturing operations require processing and placement of the wafer on a wafer support structure within a target workstation (e.g., within a processing chamber). A robotic device is used to perform such wafer placement remotely. Placing the wafer on the wafer support structure, in a known position relative to the wafer support structure, is often important. For example, it may be specified that the wafer should be centered within the wafer receiving area of the wafer support structure. However, when processing / carrying the wafer by the robotic device, it may be necessary to determine the position of the wafer relative to the robotic device in order to be able to place the wafer correctly at the target station. It is in this context that the present disclosure emerges. Summary of the Invention

[0004] In one exemplary embodiment, a method for automatically centering a wafer is disclosed. The method includes: positioning a wafer on a wafer handling component of a robotic arm. The method includes: operating the robotic arm to move the wafer handling component such that the wafer moves past a sensor array. Each sensor within the sensor array is configured to detect when an edge of the wafer passes the sensor and to emit a signal. The method includes: determining a number (N) of detected wafer edge positions. Each detected wafer edge position is defined by a set of coordinates (x, y) in a coordinate system of the wafer handling component, where any one of the sensors in the sensor array detects the wafer edge at the set of coordinates as the wafer moves past the sensor array. The method includes: for each unique set (N-1) of the number (N) of detected wafer edge positions, determining an estimated wafer offset that substantially minimizes a figure of merit value. The estimated wafer offset is defined as a vector extending from the center of the coordinate system of the wafer handling component to an estimated center position of the wafer. Each unique set (N-1) of the number (N) of detected wafer edge positions has a corresponding estimated wafer offset and a corresponding figure of merit value. The method further includes: identifying a final wafer offset as the estimated wafer offset corresponding to the unique set (N-1) of the number (N) of detected wafer edge positions having the smallest corresponding figure of merit value. And, the method includes: using the final wafer offset to center the wafer at a target station.

[0005] In one exemplary embodiment, a system for automatically centering a wafer is disclosed. The system includes a sensor array. Each sensor within the sensor array is configured to detect when an edge of the wafer passes the sensor. The system also includes a controller configured to receive data indicative of the position of the wafer handling component of the robotic arm when the wafer handling component of the robotic arm moves with a wafer held on the wafer handling component. The controller is configured to receive signals from the sensor array when the wafer moves past the sensor array, the signals indicating when the wafer edge passes a particular sensor of the sensor array. The controller is further configured to determine a number (N) of detected wafer edge positions, each detected wafer edge position defined by a set of coordinates (x, y) in the coordinate system of the wafer handling component, at which any one of the sensors in the sensor array detects the wafer edge. The controller is further configured to determine, for each unique set (N - 1) of the number (N) of detected wafer edge positions, an estimated wafer offset that substantially minimizes a figure of merit value. The estimated wafer offset is defined as a vector extending from the center of the coordinate system of the wafer handling component to an estimated center position of the wafer. Each unique set (N - 1) of the number (N) of detected wafer edge positions has a corresponding estimated wafer offset and a corresponding figure of merit value. The controller is further configured to identify a final wafer offset as the estimated wafer offset corresponding to the unique set (N - 1) of the number (N) of detected wafer edge positions having the smallest corresponding figure of merit value. The controller is configured to use the final wafer offset to guide the robotic arm to center the wafer at the target station.

[0006] In one exemplary embodiment, a method for automatic wafer centering is disclosed. The method includes: obtaining a plurality (N) of detected wafer edge positions. Each detected wafer edge position is defined by a set of coordinates (x, y) in the coordinate system of a wafer handling component. The method further includes: for each unique set (N - 1) of the plurality (N) of detected wafer edge positions, determining a minimized performance index value. The minimized performance index value has an associated estimated wafer offset, and the associated estimated wafer offset corresponds to the center of a circle that is best fit to the corresponding unique set (N - 1) of the plurality (N) of detected wafer edge positions within the coordinate system of the wafer handling component. The method further includes: determining a minimum minimized performance index value for the plurality (N) of detected wafer edge positions. The method further includes: using the minimum minimized performance index value and its associated wafer offset to center the wafer at a target station. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1A A top view of a wafer relative to a sensor is shown in accordance with some embodiments.

[0008] Figure 1B A side view of a wafer relative to a sensor is shown in accordance with some embodiments, where the sensor includes two sensor components.

[0009] Figure 1C A wafer is shown in accordance with some embodiments, where the outer leading edge of the wafer reaches and blocks the beam of the sensor.

[0010] Figure 1D A wafer is shown in accordance with some embodiments as it continues to move past the sensor, where the beam is blocked by the wafer.

[0011] Figure 1E A wafer is shown in accordance with some embodiments, where the outer trailing edge of the wafer passes the sensor such that the beam of the sensor is not blocked.

[0012] Figure 1F A vector diagram of detected wafer edge positions in a blade coordinate system is shown in accordance with some embodiments.

[0013] Figure 2 A top view of a wafer relative to three sensors is shown in accordance with some embodiments.

[0014] Figure 3 A vector diagram is shown in accordance with some embodiments that depicts a wafer offset O and a vector Ba for a given detected wafer edge position (i). i, and the estimated wafer radius E i The relationship therebetween.

[0015] Figure 4 FIG. shows a flowchart of a method for automatic wafer centering according to some embodiments.

[0016] Figure 5 FIG. shows a flowchart of a method for automatic wafer centering according to some embodiments.

[0017] Figure 6 FIG. shows a system for automatic wafer centering according to some embodiments.

[0018] Figure 7 FIG. shows an exemplary diagram of a controller 607 according to some embodiments. DETAILED DESCRIPTION

[0019] In the following description, numerous specific details are set forth in order to provide an understanding of embodiments of the present disclosure. However, it will be apparent to one of ordinary skill in the art that embodiments of the present disclosure may be practiced without some or all of these specific details. In other instances, well-known process operations have not been described in detail so as not to unnecessarily obscure the present disclosure.

[0020] Automatic wafer centering (AWC) is a process in which a robotic arm moves a semiconductor wafer (hereinafter referred to as a "wafer") past a set of sensors to attempt to precisely determine where the wafer is positioned on the blades of the robotic arm (i.e., the end effector or wafer handling component). The purpose of the set of sensors is to generate a signal when the edge of the wafer moves through a defined point in the wafer motion plane. Typically, through-beam optical sensors are positioned such that the leading edge of the wafer will interrupt the beam, while the trailing edge of the wafer will expose the beam. At each beam transition, a signal is sent to the robotic arm, and when the sensor transitions, the robotic arm immediately stores its position. In this way, the position of the robotic arm's blade relative to the sensor is recorded when the sensor transitions.

[0021] The wafer offset is defined as the spatial relationship between the center of the wafer and the center of the blade coordinate system. Thus, the wafer offset is defined as the vector extending from the center of the blade coordinate system to the center of the wafer. In the case of zero wafer offset, the center of the wafer is located at the center of the blade coordinate system. When the manipulator moves the blade, the center of the blade coordinate system is tracked and known. Thus, by knowing the position of the center of the blade coordinate system at any given position of the blade, when the wafer moves past the set of sensors, the AWC process can determine the wafer offset based on the interaction between the wafer and the set of sensors. Then, when the wafer is placed at the target station (e.g., on the chuck in the processing chamber), any error in the position of the wafer on the blade, i.e., the wafer offset, can be corrected.

[0022] Figure 1A A top view of a wafer 101 relative to sensors 103A and 103B is shown in accordance with some embodiments. Figure 1B A side view of a wafer 101 relative to sensor 103A is shown in accordance with some embodiments, where the sensor includes two sensor components 103A-1 and 103A-2. In some embodiments, sensors 103A and 103B are configured as light-emitting diode (LED) beam sensors that operate by transmitting signals when the LED beam is interrupted and when the LED beam is reformed. For example, as Figure 1B shown, the first sensor component 103A-1 operates as a beam emitter to transmit a beam 109 to the second sensor component 103A-2, and the second sensor component 103A-2 operates as a beam detector to transmit a signal when the beam is interrupted and to transmit a signal when the beam is reformed. It should be understood that the operations of sensor components 103A-1 and 103A-2 can be reversed. It should also be understood that in other embodiments, a variety of other types of sensors can be used in place of LED beam sensors 103A and 103B, provided that the sensors are capable of detecting when the edge of the wafer 101 passes through the detection position of the sensors.

[0023] In Figure 1A the example, the wafer 101 is positioned on the blade of the manipulator and is moved past sensors 103A and 103B by the manipulator in the direction 105. Thus, when the wafer 101 moves past sensors 103A and 103B, the beams of sensors 103A and 103B travel along paths 107A and 107B, respectively. Figure 1AThe example shows that as the wafer 101 moves past sensors 103A and 103B, the light beams of sensors 103A and 103B travel across the wafer 101 along straight paths 107A and 107B. However, it should be understood that in some embodiments, the wafer 101 may be moved past sensors 103A and 103B by a robotic arm in a non-linear manner, which will cause the light beams of sensors 103A and 103B to travel across the wafer 101 along non-linear paths. When the wafer 101 moves in direction 105, the peripheral leading edge of the wafer 101 reaches position Ba 1 and blocks the light beam of sensor 103A, and reaches position Ba 2 and blocks the light beam of sensor 103B. When the wafer 101 continues to move in direction 105, the peripheral trailing edge of the wafer 101 passes by sensors 103A and 103B, such that the light beams of sensors 103A and 103B are respectively at position Ba 3 and Ba 4 not blocked. In this way, when the wafer 101 moves past sensors 103A, 103B, the leading edge of the wafer 101 interrupts the sensor light beam 109, and the trailing edge of the wafer 101 reforms the sensor light beam 109. Each time the sensor light beam 109 changes, i.e., is interrupted or reformed, an interrupt signal is transmitted from sensors 103A, 103B to trigger the recording of the current blade position. Each detected wafer edge position Ba B and Y B is specified within the coordinate system of the blade represented by the coordinate axes X 1 、Ba 2 、Ba 3 and Ba 4 .

[0024] Figure 1B shows the wafer 101 approaching sensor 103A in direction 105. In Figure 1B , the light beam 109 is not interrupted between the first sensor component 103A-1 and the second sensor component 103A-2. Figure 1C shows the wafer 101, at this time, the peripheral leading edge of the wafer 101 reaches position Ba 1 and blocks the light beam 109 of sensor 103A. Figure 1D shows the wafer 101 continuing to move past sensor 103A, and the light beam 109 is blocked by the wafer 101. Figure 1E shows the wafer 101, at this time, the peripheral trailing edge of the wafer 101 passes by sensor 103A, such that the light beam 109 of sensor 103A reforms (is not blocked) at position Ba 4 .

[0025] In some embodiments, the robotic arm has a digital input / output (I / O) card. Sensors 103A and 103B are connected to the digital I / O card of the robotic arm. When sensors 103A, 103B are triggered (by detecting the blockage of beam 109 or by detecting the non-blockage of beam 109), sensors 103A, 103B send an interrupt signal to the digital I / O card of the robotic arm, causing the robotic arm to store its current blade position data in a digital memory. In some embodiments, after the wafer 101 has passed sensors 103A and 103B, the blade position data corresponding to the time when sensors 103A and 103B are triggered is downloaded from the robotic arm to a control system for performing AWC processing. In some embodiments, the control system for performing AWC processing is implemented on the robotic arm. In these embodiments, the control system for performing AWC processing on the robotic arm can utilize the blade position data corresponding to the time when sensors 103A and 103B are triggered without downloading the blade position data to another control system separate from the robotic arm.

[0026] A calibration procedure can be performed to determine the position of the beams 109 of sensors 103A and 103B relative to the center of the blade coordinate system, where a calibration wafer (having an accurate and uniform diameter) is positioned on the blade in a centered manner (such that the wafer is precisely centered at the center of the blade coordinate system), and the calibration wafer is moved through sensors 103A and 103B. By knowing the calibration position of the beams 109 of sensors 103A and 103B relative to the center of the blade coordinate system of the truly centered wafer, the detected positions Ba 1 、Ba 2 、Ba 3 、and Ba 4 can be used to determine the center coordinates of the non-centered wafer within the blade coordinate system and, in turn, determine the wafer offset of the non-centered wafer. In Figure 1A , the wafer offset is depicted by a vector O extending from the center of the blade coordinate system to the center 104 of the wafer 101.

[0027] Most wafers have a fiducial feature, namely a notch, which is used to determine and set the angular position of the wafer. This fiducial feature is typically in the form of a notch at a position on the wafer edge. For example, Figure 1A shows a wafer 101 having a fiducial feature 102. The basis of the AWC algorithm for calculating the wafer offset O is the assumption that sensors 103A, 103B accurately determine points on the circumference of the wafer 101. If the fiducial feature 102 moves under one of sensors 103A, 103B, as shown in Figure 1A for example, the wafer edge position reported by the sensor will not be on the wafer circumference, and the result of the AWC calculation of the wafer offset O will be incorrect.

[0028] Many tools within a semiconductor manufacturing facility have an aligner (wafer aligner) somewhere in the wafer flow of the tool. One use of the aligner is to determine the position of fiducial 102 and rotate the wafer so that fiducial 102 is at a specified azimuthal position. Typically, fiducial 102 is positioned so that it will not pass through any of the sensors 103A, 103B used in the AWC process, regardless of the target workstation to which the wafer can be transported. This eliminates the requirement that the AWC process and associated algorithms account for the possible interference of fiducial 102 with sensors 103A, 103B. However, some drawbacks to using an aligner in the wafer flow of a tool are the cost of the aligner and the reduction in wafer throughput due to the time required to move the wafer into the aligner, operate the aligner to determine and position the wafer fiducial 102, and retrieve the wafer from the aligner. Thus, in some cases, adding an aligner to the wafer flow of a tool is not reasonable, and the AWC system must attempt to determine if there is interference of fiducial 102 with either of the sensors 103A, 103B, and if so, affect the sensor data.

[0029] As Figure 1A shown, an exemplary AWC process uses a sensor array including two sensors 103A and 103B to generate a total of four sensor signals corresponding to detected wafer edge positions Ba 1 、Ba 2 、Ba 3 、and Ba 4 as the wafer moves past the sensor array. Each sensor signal and associated robot position are used to calculate the corresponding detected wafer edge positions (Ba 1 、Ba 2 、Ba 3 、and Ba 4 ) in the blade coordinate system. Figure 1F Shows a vector diagram of the detected wafer edge positions Ba 1 、Ba 2 、Ba 3 、and Ba 4 in the blade coordinate system. The purpose of the AWC process is to determine the wafer offset O, i.e., to determine the position of the wafer center relative to the center of the blade coordinate system. Traditional AWC processes include selecting any three of the detected wafer edge positions (Ba 1 、Ba 2 、Ba 3 、and Ba 4Any three of them) are used for analysis to determine the wafer offset O. It is assumed that the three detected wafer edge positions selected will uniquely define a circle corresponding to the estimated wafer radius R and the estimated wafer center 117. The estimated wafer center 117 is given by the intersection of the perpendicular bisectors 113 and 115 of lines 112 and 114 respectively, and lines 112 and 114 extend between adjacent detected wafer edge positions (Ba 1 Ba 2 Ba 3 ). The estimated wafer radius R is the distance between the estimated wafer center 117 used in the analysis and any adjacent detected wafer edge position (Ba 1 Ba 2 Ba 3 ).

[0030] For the AWC process that uses two sensors 103A and 103B to generate four sensor signals corresponding to the detected wafer edge positions Ba 1 Ba 2 Ba 3 Ba 4 , there are four sets of three detected wafer edge positions: Group 1 = (Ba 1 Ba 2 Ba 3 ); Group 2 = (Ba 2 Ba 3 Ba 4 ); Group 3 = (Ba 1 Ba 3 Ba 4 ); and Group 4 = (Ba 1 Ba 2 Ba 4 ). Figure 1F Shows the use of Group 1 to determine the estimated wafer center 117 and the estimated wafer radius R. In the AWC process, each of Group 1, Group 2, Group 3, and Group 4 is used to determine the estimated wafer center 117 and the estimated wafer radius R. If the reference portion 102 of the wafer interferes with any one of the four sensor signals, only one of Group 1, Group 2, Group 3, and Group 4 will be unaffected by the reference portion 102. For example, if the sensor signal corresponding to the detected wafer edge position Ba 3 is affected by the reference portion 102, then only Group 4 = (Ba 1 Ba 2 Ba 4)There will be no error introduced by the interference of the reference portion 102. Assuming that the interference of the reference portion 102 has occurred, by comparing the estimated wafer radius R associated with each of the first group, the second group, the third group, and the fourth group and the known wafer radius, to determine the correct set of three detected wafer edge positions (the first group, the second group, the third group, or the fourth group) for determining the estimated wafer center 117 and the corresponding wafer offset O. Assuming that the interference of the reference portion 102 has occurred, for the four sets of three detected wafer edge positions (the first group, the second group, the third group, the fourth group), only one R value should be correct. And, the particular set of three detected wafer edge positions (the first group, the second group, the third group, the fourth group) that provides the correct R value is the set of three detected wafer edge positions for determining the estimated wafer center 117 and the corresponding wafer offset O. Again, the wafer offset O is the vector that extends from the center of the blade coordinate system to the wafer center when the wafer is held by the blade. Therefore, the coordinates of the estimated wafer center 117 in the blade coordinate system define the wafer offset O.

[0031] It should be noted that the reference portion 102 will only affect one of the detected wafer edge positions Ba 1 , Ba 2 , Ba 3 , and Ba 4 . Moreover, it should be noted that the reference portion 102 will rarely pass directly under the sensors 103A, 103B. More often, some parts of the reference portion 102 closer to the wafer edge affect the sensor transition, and when determining the estimated wafer radius R based on a given set (the first group, the second group, the third group, the fourth group) of three detected wafer edge positions, the reference portion 102 only causes a small error. In these cases, it becomes difficult to separate the error caused by the interference of the reference portion 102 from other errors in the AWC system (e.g., errors in the calibration of the sensor position and / or errors in tracking the exact position of the center of the blade coordinate system when the blade moves). In most cases, the above AWC process algorithm uses an error threshold in the estimated wafer radius R to determine whether the interference of the reference portion 102 has occurred. For example, the difference between the known wafer radius and the estimated wafer radius R determined based on a given set (the first group, the second group, the third group, the fourth group) of three detected wafer edge positions will have to exceed a preset threshold of the error in the estimated wafer radius R to conclude that the estimated wafer radius R is affected by the interference of the reference portion 102. However, setting the preset threshold of the error in the estimated wafer radius R is an arbitrary action and is itself error-prone.

[0032] The present disclosure relates to methods and systems for benchmark filtering of AWC processes and algorithms that use data returned from AWC sensors to: 1) determine whether a fiducial portion 102 has caused one of the sensors to report invalid wafer edge data, and 2) eliminate erroneous wafer edge data so that the determined wafer offset O is accurate. Figure 2 Shown is a top view of a wafer 101 relative to three sensors 201A, 201B, and 201C according to some embodiments. In some embodiments, sensors 201A, 201B, and 201C are configured as LED beam sensors, similar to sensors 103A and 103B described above, and operate by transmitting a signal when an LED beam is interrupted and when the LED beam is reformed. However, it should be understood that in other embodiments, a variety of other types of sensors may be used in place of LED beam sensors 201A, 201B, and 201C as long as the sensors are capable of detecting when the edge of the wafer 101 passes through the detection location of the sensor.

[0033] In Figure 2 the example of, the wafer 101 is positioned on a blade of a robot and is moved by the robot in direction 105 past sensors 201A, 201B, and 201C. Thus, as the wafer 101 moves past sensors 201A, 201B, and 201C, the beams of sensors 201A, 201B, and 201C travel along paths 203A, 203B, and 203C, respectively. In Figure 2 the example of, the beam of sensor 201A encounters the fiducial portion 102 of the wafer 101. Figure 2 The example of shows that as the wafer 101 moves past sensors 201A, 201B, and 201C, the beams of sensors 201A, 201B, and 201C travel along straight paths 203A, 203B, and 203C across the wafer 101. However, it should be understood that in some embodiments, the wafer 101 may be moved past sensors 201A, 201B, and 201C by the robot in a non-linear manner, which would cause the beams of sensors 201A, 201B, and 201C to travel along non-linear paths across the wafer 101.

[0034] As the wafer 101 moves in direction 105, the leading edge of the wafer 101 reaches position Ba 1 and blocks the beam of sensor 201A, and reaches position Ba 2 and blocks the beam of sensor 201B, and reaches at position Ba 3 and blocks the beam of sensor 201C. As the wafer 101 continues to move in direction 105, the trailing edge of the wafer 101 reaches position Ba 4and does not block the beam of sensor 201C and reaches position Ba 5 and does not block the beam of sensor 201B and reaches position Ba 6 and does not block the beam of sensor 201A. In this way, as the wafer 101 moves past sensors 201A, 201B, and 201C, the leading edge of the wafer 101 interrupts the sensor beam 109, and the trailing edge of the wafer 101 reforms the sensor beam 109. Each time the sensor beam 109 transitions, i.e., is interrupted or reformed, an interrupt signal is transmitted from the sensor to trigger the recording of the current blade position. In the blade coordinate system represented by the coordinate axes X B and Y B each detected wafer edge position Ba is specified 1 、Ba 2 、Ba 3 、Ba 4 、Ba 5 、and Ba 6 . In some embodiments, three sensors 201A, 201B, and 201C are positioned such that at least one sensor tracks each half of the wafer 101. For example, in Figure 2 sensor 201A tracks on the left half of the wafer 101, and sensors 201B and 201C track on the right half of the wafer 101. Moreover, in some embodiments, sensors 201A, 201B, 201C are positioned such that only one sensor transition will occur at a given time. For example, in Figure 2 sensors 201A, 201B, 201C are positioned such that as the wafer moves in direction 105, sensor 201A is triggered first, followed by sensor 201B, followed by sensor 201C, followed by sensor 201C again, followed by sensor 201B, followed by sensor 201A. In this way, each signal received from sensors 201A, 201B, 201C can be associated with a particular one of sensors 201A, 201B, 201C based on the order in which the signal is received.

[0035] A calibration procedure can be performed, in which a calibration wafer (having an exact and uniform diameter) is positioned on the blade in a centered manner (centering the calibration wafer precisely at the center of the blade coordinate system), and the calibration wafer is moved past sensors 201A, 201B, and 201C. This calibration procedure gives the positions of sensors 201A, 201B, 201C relative to the center of the blade coordinate system for the centered wafer, where the centered wafer causes transitions of sensors 201A, 201B, 201C. The results of the calibration procedure, along with the geometry of the manipulator and some coordinate transformations, enable the definition of a set of vectors Ba i(i = 1 to N), the vector extends from the center of the blade coordinate system to the detected wafer edge position (e.g., Ba 1 、Ba 2 、Ba 3 、Ba 4 、Ba 5 、and Ba 6 ), at which position the sensor transition occurs. More specifically, as Figure 2 shown, each detected wafer edge position Ba 1 、Ba 2 、Ba 3 、Ba 4 、Ba 5 、and Ba 6 respectively defines a corresponding vector Ba 1 、Ba 2 、Ba 3 、Ba 4 、Ba 5 、Ba 6 , the vector emanates from the center of the blade coordinate system as represented by the coordinate axes X B and Y B . When the robotic arm moves the blade in the direction 105 to pass the wafer 101 through the sensors 201A, 201B, and 201C, the vectors Ba i (i = 1 to 6) are determined. The set of vectors Ba i (i = 1 to 6) can be used to determine the coordinates of the center 104 of the non-centered wafer within the blade coordinate system and thereby determine the wafer offset O of the non-centered wafer.

[0036] The number (N) of the vectors Ba i (i = 1 to N) depends on the number of sensors used in the AWC process. In some embodiments, the fiducial filtering AWC process uses at least three sensors 201A, 201B, and 201C. Each sensor 201A, 201B, and 201C typically transitions twice (once when the beam is interrupted and once when the beam reforms). Thus, the number (N) of the vectors Ba i is twice the number of the sensors 201A, 201B, and 201C. However, sometimes, the beam of a given sensor may be interrupted by the leading edge of the wafer 101 but not reformed due to the geometry of the robotic arm blade. In this case, there will be only one vector for the given sensor. The fiducial filtering AWC process requires specifying the number (N) of the vectors Ba i and the rated wafer diameter, such as 200 mm, 300 mm, etc.

[0037] The fiducial filtering AWC process uses an optimal fit algorithm to determine the wafer offset O and to determine if fiducial interference has occurred. The optimal fit algorithm determines the position of the wafer center 104 within the blade coordinate system, the optimal fit being defined by the detected wafer edge positions defined by the set of vectors Ba i (i = 1 to N). The output of the optimal fit algorithm includes the x coordinate (O x ) and the y coordinate (O y ) of the wafer center 104 in the blade coordinate system, with any effects from fiducial 102 interference removed. The x coordinate (O x ) and the y coordinate (O y ) of the wafer center 104 define the wafer offset O. The optimal fit algorithm determines the best fit of a circle to the detected wafer edge positions defined by the set of N vectors Ba i (i = 1 to N). The optimal fit algorithm includes the definition of a performance index (IP), which is the sum of the squares of the differences between the estimated wafer radius E and the rated wafer radius R for all N detected wafer edge positions for a given wafer offset O.

[0038] Figure 3 FIG. shows a vector diagram according to some embodiments, the vector diagram depicting the relationship between the wafer offset O, the vector Ba i , and the estimated wafer radius E i for a given detected wafer edge position (i). As Figure 3 shown, for each of the N vectors Ba i (i = 1 to N), there is the relationship of Equation 1. In Equation 1, O is the current estimate of the wafer offset O, and E i is the vector extending from the current estimate of the wafer offset O to the ith detected wafer edge position Ba i corresponding to the ith estimated wafer radius.

[0039] Equation 1: Ba i = O + E i

[0040] The relationship of Equation 1 can be rewritten as shown in Equation 2.

[0041] Equation 2: E i = Ba i - O

[0042] The magnitude of E i is given by Equation 3, where Ba ix is the x coordinate of the ith detected wafer edge position Ba i in the blade coordinate system, Ba iyis the edge position Ba of the i-th detected wafer i The y coordinate in the blade coordinate system, O x is the estimated x coordinate of the wafer center in the blade coordinate system, O y is the estimated y coordinate of the wafer center in the blade coordinate system.

[0043] Equation 3:

[0044]

[0045] If O x and O y 's current values are close to the wafer center, then the magnitude of E i will be close to the rated wafer radius R. As shown in Equation 4, the performance index (IP) for any estimated wafer center (O x , O y ) is the sum of the squares of the differences between the estimated wafer radius (|E i |) at the i-th detected wafer edge position and the rated wafer radius R for N detected wafer edge positions.

[0046] Equation 4:

[0047]

[0048] If there are no errors during the calibration of the sensor position or the tracking of the blade position or the data collection, and if there is no fiducial interference on the sensor, then the best-fit algorithm will iterate on O x and O y until the performance index (IP) is zero. In practice, during the calibration of the sensor position and / or the tracking of the blade position and / or the data collection, there may be a certain amount of AWC system error, although very small. Therefore, in the presence of a certain amount of AWC system error, the best-fit algorithm will iterate on O x and O y until the performance index (IP) is minimized to a non-zero value, so that the wafer edge is as close as possible to the N detected wafer edge positions Bai (i = 1 to N).

[0049] As shown in Equation 4, each estimated value of the wafer offset O (i.e., O x , O y ) generates a performance index (IP) value for the N vectors Ba i (i = 1 to N). A poor estimate of the wafer offset O will generate a higher performance index (IP) value. The best estimate of the wafer offset O minimizes the performance index (IP) value. When minimizing the performance index (IP) value, the wafer edge will be as close as possible to the detected wafer edge position Ba i(i = 1 to N).

[0050] The best fit algorithm progressively completes different estimated values of the wafer offset O to identify the wafer offset O that minimizes the value of the performance index (IP). In some embodiments, the estimated wafer offset O is adjusted iteratively until the corresponding change in the minimum value of the performance index (IP) between iterations is equal to or less than a specified iteration stop value. In some embodiments, the iteration stop value can be 0.00001. However, in other embodiments, the iteration stop value can be greater than or less than 0.00001. In some embodiments, Newton's Method is used to iterate on the estimated wafer offset O to determine the minimum value of the performance index (IP). In these embodiments, Newton's Method includes determining the derivative of the performance index (IP) with respect to O x and O y and setting these derivatives equal to zero, which corresponds to the local minimum of the performance index (IP). Moreover, in some embodiments, the second derivative of the performance index (IP) with respect to O x and O y can be evaluated to determine the adjustment amount of the estimated wafer offset O from one iteration to the next when the performance index (IP) drops to its minimum value.

[0051] In the fiducial filtering AWC process, if there is a fiducial interference to the sensor, one of the N vectors Ba i (1 to N) will not be located on the peripheral edge of the wafer, i.e., on the circumference of the wafer. In this case, if there is a fiducial interference to the sensor, there will be no value of the wafer offset O that positions the wafer center such that the edge of the wafer passes through all N detected wafer edge positions Ba i (i = 1 to N).

[0052] Fiducial interference to the sensor will only affect one of the N vectors Ba i (i = 1 to N). To identify which of the N vectors Ba i (i = 1 to N) has been affected by fiducial interference, the fiducial filtering AWC process includes eliminating another one of the N vectors Ba i (i = 1 to N) each time to obtain a unique set of (N - 1) vectors Ba i and the performance of the best fit algorithm to identify the wafer offset O that minimizes the value of the performance index (IP) for each unique set of (N - 1) vectors Ba i . Equation 5 shows the value of the performance index (IP z ) for the unique set of (N - 1) vectors Ba i excluding the z-th vector Ba Z .

[0053] Equation 5:

[0054]

[0055] For each value of (z) in the range 1 to N, determine the wafer offset O that minimizes the performance index (IP Z ). After the best fit algorithm has converged to the resulting minimum of the performance index (IP i ) for the given (z) excluding the vector Ba z , for the given (z) excluding the vector Ba Z , record the resulting minimum of the performance index (IP z ) and the corresponding estimated wafer offset O. The above process will give N estimated values of the wafer offset O, and each estimated value of the wafer offset O will have a corresponding performance index (IP Z ). In other words, for each excluded vector Ba Z (z = 1 to N), there are separate estimated values of the wafer offset O and the corresponding performance index (IP z ) determined using the (N - 1) non-excluded vectors Ba i . Z

[0056] As an example, consider the AWC configuration of Figure 2 . In this configuration, there are N = 6 vectors B a1 , Ba 2 , Ba 3 , Ba 4 , Ba 5 , and Ba 6 . For this configuration, the reference section filtering AWC process involves determining six different estimated values of the wafer offset O and their corresponding performance indices (IP Z ). Specifically, for the first estimated value (z = 1), the vector Ba 1 is excluded, and the vector group (Ba 2 , Ba 3 , Ba 4 , Ba 5 , Ba 6 ) determines the estimated value of the wafer offset O. For the second estimated value (z = 2), the vector Ba 2 is excluded, and the vector group (B a1 , Ba 3 , Ba 4 , Ba 5 , Ba 6 ) determines the estimated value of the wafer offset O. For the third estimated value (z = 3), the vector Ba 3 ​, and the vector group (B a1 , Ba 2 , Ba 4 , Ba 5 , Ba 6 ) determines the estimated value of the wafer offset O. For the fourth estimate (z = 4), the vector Ba 4 is excluded, and the vector group (B a1 , Ba 2 , Ba 3 , Ba 5 , Ba 6 ) determines the estimated value of the wafer offset O. For the fifth estimate (z = 5), the vector Ba 5 is excluded, and the vector group (Ba 1 , Ba 2 , Ba 3 , Ba 4 , Ba 6 ) determines the estimated value of the wafer offset O. For the sixth estimate (z = 6), the vector Ba 6 is excluded, and the vector group (B a1 , Ba 2 , Ba 3 , Ba 4 , Ba 5 ) determines the estimated value of the wafer offset O.

[0057] Since the fiducial interference only affects one of the vectors Ba i (i = 1 to 6), only one of the six determined estimated values of the wafer offset O will not be affected by the fiducial interference. The estimated value of the wafer offset O with the lowest corresponding performance index (IP Z ) is based on the (N - 1) vectors B z that do not include the vector Ba ai affected by the fiducial interference on the sensor. Therefore, the estimated value of the wafer offset O with the lowest corresponding performance index (IP Z ) is the estimated value of the wafer offset O that has not been affected by the fiducial interference on the sensor. The basic assumption is that although each AWC sensor is expected to have a small amount of noise (i.e., errors in its position calibration and / or errors in its signal transition and / or errors in its signal transmission), the sensor that encounters fiducial interference will have a corresponding error amount when detecting the wafer edge, and this error amount is significantly greater than any error caused by noise when other sensors detect the wafer edge.

[0058] For the fiducial filtering AWC process of three AWC sensors (e.g., as Figure 2 shown), determine the unique set of five vectors Ba i(from six vectors Ba i (i = 1 to 6)). The unique set of five vectors Ba i does not include the vectors Ba affected by the interference of the reference part of the sensor i , thereby identifying the vectors Ba affected by the interference of the reference part of the sensor i . And, for the unique set of five vectors Ba i the determined estimated wafer offset O is the wafer offset O that is not affected by the interference of the reference part of the sensor.

[0059] As described above, the conventional method for reference part interference detection requires the operator to select a somewhat arbitrary error threshold in the estimated wafer radius R to determine which vectors Ba i (i = 1 to N) are affected by the interference of the reference part. Usually, the following cases are encountered: more than one combination of three vectors Ba i satisfies the error threshold selected in the estimated wafer radius R, and the selection of the wrong data becomes arbitrary. There are also cases where, even though a vector Ba i has some errors, the combinations of three vectors Ba i do not satisfy the error threshold in the estimated wafer radius R. In contrast, in the reference part filtering AWC process disclosed herein, the unique set of (N - 1) vectors Ba Z with the lowest corresponding performance index (IP i ) is determined to identify which vector Ba i has been affected by the interference of the reference part of the AWC sensor. And, the unique set of (N - 1) vectors Ba Z with the lowest corresponding performance index (IP i ) is used to determine the wafer offset O to guide the manipulator to center the wafer at the target workstation.

[0060] It should be understood that by being able to eliminate the time required for wafer reference part detection and alignment in the dedicated aligner module, the reference part filtering AWC process and related systems disclosed herein support high wafer throughput operations in a manufacturing facility. Moreover, the reference part filtering AWC process and related systems disclosed herein provide an accurate and reliable determination of the center of a wafer with an unknown reference part position, and the unknown reference part position may have reference part interference on the AWC sensor.

[0061] Figure 4A flowchart showing a method of automatic wafer centering according to some embodiments is shown. The method includes operation 401: positioning a wafer on a wafer handling component of a robotic arm. The method further includes operation 403: operating the robotic arm to move the wafer handling component such that the wafer moves past a sensor array. Each sensor within the sensor array is configured to detect and emit a signal when an edge of the wafer passes by the sensor. In some embodiments, the sensor array includes at least three sensors. In some embodiments, the sensor array includes at least one sensor that is positioned to pass through a first half of the wafer relative to the direction of travel of the wafer center when the wafer moves past the sensor array. And, the sensor array includes at least one sensor that is positioned to pass through a second half of the wafer relative to the direction of travel of the wafer center when the wafer moves past the sensor array. In some embodiments, each sensor in the sensor array is a beam sensor, such as an LED beam sensor. In some embodiments, the beam sensor is configured to transmit a signal when the beam of the beam sensor is interrupted by the wafer, the signal indicating that the edge of the wafer has been detected. And, the beam sensor is configured to transmit a signal when the interrupted beam of the beam sensor reforms, the signal indicating that the edge of the wafer has been detected.

[0062] The method further includes operation 405: determining a number (N) of detected wafer edge positions. Each detected wafer edge position is defined by a set of coordinates (x,y) in the coordinate system of the wafer handling component, where any one of the sensors in the sensor array detects the edge of the wafer at the set of coordinates as the wafer moves past the sensor array. The method further includes operation 407, where for each unique set of (N - 1) detected wafer edge positions among the number (N) of detected wafer edge positions, an estimated wafer offset that substantially minimizes a performance index value is determined. The estimated wafer offset is defined as a vector extending from the center of the coordinate system of the wafer handling component to the estimated center position of the wafer. Each unique set of (N - 1) detected wafer edge positions among the number (N) of detected wafer edge positions has a corresponding estimated wafer offset and a corresponding performance index value. The estimated wafer offset for a given unique set of (N - 1) detected wafer edge positions among the number (N) of detected wafer edge positions corresponds to the center of the circle that best fits the given unique set of (N - 1) detected wafer edge positions within the coordinate system of the wafer handling component. The performance index (IP) of any estimated wafer center (O x , O y ) is the estimated wafer radius (|E iThe sum of the squares of the differences between the detected wafer edge positions and the rated wafer radius R. When the performance index (IP) is minimized, the performance index (IP) and the associated estimated wafer center (O x ,O y ) correspond to a circle that best fits a given unique set (N - 1) of the detected wafer edge positions among a number (N) of detected wafer edge positions within the coordinate system of the wafer handling component.

[0063] The method further includes operation 409: identifying the final wafer offset as the estimated wafer offset corresponding to the unique set (N - 1) of the detected wafer edge positions among the number (N) of detected wafer edge positions having the minimum corresponding performance index value. The unique set (N - 1) of the detected wafer edge positions having the minimum corresponding performance index value does not include: one of the detected wafer edge positions at the wafer fiducial among the number (N) of detected wafer edge positions. The method further includes operation 411: using the final wafer offset to center the wafer at the target station.

[0064] Figure 5 A flowchart of a method for automatic wafer centering according to some embodiments is shown. The method includes operation 501: obtaining a number (N) of detected wafer edge positions. Each detected wafer edge position is defined by a set of coordinates (x, y) in the coordinate system of the wafer handling component. In some embodiments, the number (N) of detected wafer edge positions is obtained by passing the wafer through a beam sensor array including at least three sensors. The method further includes operation 503: determining a minimized performance index value for a unique set (N - 1) of the detected wafer edge positions among the number (N) of detected wafer edge positions. The minimized performance index value has an associated estimated wafer offset corresponding to the center of a circle that best fits the corresponding unique set (N - 1) of the detected wafer edge positions among the number (N) of detected wafer edge positions within the coordinate system of the wafer handling component.

[0065] The method further includes operation 505 of determining the minimum minimized performance index value for the number (N) of detected wafer edge positions. The minimum minimized performance index value has an associated wafer offset. The performance index (IP) of any estimated wafer center (O x ,O y ) is the estimated wafer radius (|E iThe sum of the squares of the differences between the detected wafer edge positions and the rated wafer radius R. When the performance index (IP) is minimized, the performance index (IP) and the associated estimated wafer center (O x , O y ) correspond to a circle of a given unique set of (N - 1) detected wafer edge positions out of a number (N) of detected wafer edge positions within the coordinate system that best fits the wafer handling component. For the corresponding estimated wafer offset, the unique set of (N - 1) detected wafer edge positions out of the number (N) of detected wafer edge positions having the minimum corresponding performance index value does not include: one detected wafer edge position at the wafer reference portion among the number (N) of detected wafer edge positions. The method further includes operation 507: using the minimum minimized performance index value and its associated wafer offset to center the wafer at the target station.

[0066] Figure 6 FIG. shows a system for automatic wafer centering according to some embodiments. The system includes a robot 601 having a wafer handling component 603 (e.g., a blade). The system further includes a sensor array 605. The sensor array 605 includes a plurality of sensors 201A, 201B, 201C. Each sensor 201A, 201B, 201C within the sensor array 605 is configured to detect when the edge of the wafer 101 passes by the sensors 201A, 201B, 201C. In some embodiments, the sensor array 605 includes at least three sensors 201A, 201B, 201C. In some embodiments, the sensor array 605 includes at least one sensor 201A that is positioned to pass through the first half of the wafer 101 with respect to the direction of travel 105 of the center 104 of the wafer 101 when the wafer 101 moves past the sensor array 605 (see Figure 2 ). And, the sensor array 605 includes at least one sensor 201B, 201C that is positioned to pass through the second half of the wafer 101 with respect to the direction of travel 105 of the center 104 of the wafer 101 when the wafer 101 moves past the sensor array 605. In some embodiments, each sensor 201A, 201B, 201C in the sensor array 605 is a beam sensor. In some embodiments, the beam sensor is configured to transmit a signal when the beam of the beam sensor is interrupted by the wafer 101, the signal indicating that the edge of the wafer 101 is detected. And, the beam sensor is configured to transmit a signal when the interrupted beam of the beam sensor is reformed, the signal indicating that the edge of the wafer 101 is detected.

[0067] The system further includes a controller 607, which is configured and connected to receive signals from each of the sensors 201A, 201B, 201C within the sensor array 605. The controller 607 is also configured and connected to receive signals from the robotic arm 601. The controller 607 is further configured and connected to transmit signals to the robotic arm 601. The controller 607 is configured to receive data indicating the position of the wafer handling component 603 of the robotic arm 601 when the wafer handling component 603 carrying the wafer 101 moves. The controller 607 is configured to receive signals from the sensor array 605 when the wafer 101 moves past the sensor array 605. The signals indicate when the edge of the wafer 101 passes the specific sensors 201A, 201B, 201C of the sensor array 605. The controller 607 is configured to determine a number (N) of detected wafer edge positions. Each detected wafer edge position is defined by a set of coordinates (x, y) in the coordinate system of the wafer handling component 603, at which any one of the sensors 201A, 201B, 201C in the sensor array 605 detects the edge of the wafer 101.

[0068] The controller 607 is configured to determine, for each unique set of (N - 1) detected wafer edge positions out of the number (N) of detected wafer edge positions, an estimated wafer offset that substantially minimizes a performance index value. The estimated wafer offset is defined as a vector extending from the center of the coordinate system of the wafer handling component 603 to the estimated center position of the wafer 101. Each unique set of (N - 1) detected wafer edge positions out of the number (N) of detected wafer edge positions has a corresponding estimated wafer offset and a corresponding performance index value. The estimated wafer offset for a given unique set of (N - 1) detected wafer edge positions out of the number (N) of detected wafer edge positions corresponds to the center of a circle that is best fit to the given unique set of (N - 1) detected wafer edge positions out of the number (N) of detected wafer edge positions within the coordinate system of the wafer handling component 603. The performance index (IP) of any estimated wafer center (O x , O y ) is the sum of the squares of the differences between the estimated wafer radius (|E i |) at the i-th detected wafer edge position of the wafer 101 and the rated wafer radius R for the unique set of (N - 1) detected wafer edge positions out of the number (N) of detected wafer edge positions. When the performance index (IP) is minimized, the performance index (IP) and the associated estimated wafer center (O x , O y)corresponds to a circle that best fits a given unique set (N - 1) of the detected wafer edge positions out of several (N) detected wafer edge positions within the coordinate system of the wafer handling member 603.

[0069] The controller 607 is further configured to identify the final wafer offset as the estimated wafer offset corresponding to a given unique set (N - 1) of the detected wafer edge positions out of several (N) detected wafer edge positions that has the smallest corresponding performance index value. The given unique set (N - 1) of the detected wafer edge positions out of several (N) detected wafer edge positions that has the smallest corresponding performance index value does not include: one of the several (N) detected wafer edge positions at the fiducial portion 102 of the wafer 101. The controller 607 is configured to use the final wafer offset to guide the robotic arm 601 to center the wafer 101 at the target station 609.

[0070] Figure 7Exemplary diagram of controller 607 in accordance with some embodiments is shown. In various embodiments, controller 607 includes a processor 701, a storage hardware unit (HU) 703 (e.g., memory), an input HU 705, an output HU 707, an input / output (I / O) interface 709, an I / O interface 711, a network interface controller (NIC) 713, and a data communication bus 715. The processor 701, the storage HU 703, the input HU 705, the output HU 707, the I / O interface 709, the I / O interface 711, and the NIC 713 can communicate data with each other via the data communication bus 715. The input HU 705 is configured to receive data communication from a plurality of external devices (e.g., robotic arm 601 and sensors 201A, 201B, 201C). Examples of the input HU 705 include a data acquisition system, a data acquisition card, etc. The output HU 707 is configured to transmit data to a number of external devices, such as the robotic arm 601. An example of the output HU 707 is a device controller. Examples of the NIC 713 include a network interface card, a network adapter, etc. Each of the I / O interfaces 709 and 711 is defined to provide compatibility between different hardware units coupled to the I / O interface. For example, the I / O interface 709 can be defined to convert a signal received from the input HU 705 into a form, amplitude, and / or rate compatible with the data communication bus 715. Moreover, the I / O interface 707 can be defined to convert a signal received from the data communication bus 715 into a form, amplitude, and / or rate compatible with the output HU 707. Although various operations are described herein as being performed by the processor 701 of the controller 607, it should be understood that in some embodiments, various operations can be performed by multiple processors of the controller 607 and / or by multiple processors of multiple computing systems that communicate data with the controller 607. In some embodiments, there is also a user interface associated with the controller 607. The user interface can include a display (e.g., a display screen of the device and / or process conditions and / or a graphical software display), and user input devices such as a pointing device, a keyboard, a touch screen, a microphone, etc.

[0071] The controller 607 can be configured to execute a computer program that includes an instruction set for controlling the operation of the robotic arm 601 and performing the fiducial filtering AWC process disclosed herein. Moreover, in some embodiments, a computer program stored on a memory device associated with the controller 607 can be employed. The software for guiding the operation of the controller 607 can be designed or configured in many different ways. The computer program for guiding the operation of the controller 607 to perform the fiducial filtering AWC process and correspondingly control the robotic arm 601 can be written in any conventional computer-readable programming language, such as, for example, assembly language, C, C++, Pascal, Fortran, or others. The compiled object program code or script is executed by the processor 701 to perform the tasks identified in the program.

[0072] Broadly speaking, the controller 607 can be defined as an electronic device having a variety of integrated circuits, logic, memory, and / or software that receives instructions, sends instructions, and controls operations. The integrated circuits can include a chip storing program instructions in hardware form, a digital signal processor (DSP), a chip defined as an application-specific integrated circuit (ASIC), and / or one or more microprocessors, or a microcontroller that executes program instructions (such as software). The program instructions can be instructions passed to the controller 607 in the form of a variety of individual settings (or program files) that define the operation parameters for performing the fiducial filtering AWC process and controlling the robotic arm 601.

[0073] The embodiments described herein can also be practiced using a variety of computer system configurations, including handheld hardware units, microprocessor systems, microprocessor-based or programmable consumer electronics, minicomputers, mainframe computers, and the like. The embodiments described herein can also be practiced in a distributed computing environment where tasks are performed by remote processing hardware units linked through a network. It should be understood that the embodiments described herein can employ a variety of computer-implemented operations involving data stored in a computer system. These operations are those that require the physical manipulation of physical quantities. Any operation described herein that forms part of an embodiment is a useful machine operation. The embodiments also relate to hardware units or devices for performing these operations. The devices can be specifically constructed for a special-purpose computer. When defined as a special-purpose computer, the computer can also perform other processing, program execution, or subroutines that are not part of the special purpose, while still being able to operate for the special purpose. In some embodiments, the operations can be processed by a general-purpose computer selectively activated or configured by one or more computer programs stored in a computer memory, cache, or obtained through a network. When data is obtained through a network, the data can be processed by other computers on the network (such as the cloud of computing resources).

[0074] The various embodiments described herein can be implemented via AWC process control instructions, such as computer-readable program code on a non-transitory computer-readable medium. A non-transitory computer-readable medium is any data storage hardware unit that can store data, which can then be read by a computer system. Examples of non-transitory computer-readable media include hard disks, network-attached storage devices (NAS), ROM, RAM, compact disc read-only memory (CD-ROM), CD recordable discs (CD-R), CD rewritable discs (CD-RW), magnetic tapes, and other optical and non-optical data storage hardware units. The non-transitory computer-readable medium can include a tangible computer-readable medium distributed on network-coupled computer systems such that the computer-readable program code is stored and executed in a distributed manner.

[0075] Although the foregoing disclosure includes some details for purposes of clarity of understanding, it will be apparent that certain changes and modifications can be practiced within the scope of the appended claims. For example, it will be understood that one or more features from any embodiment disclosed herein can be combined with one or more features from any other embodiment disclosed herein. Accordingly, the provided embodiments should be considered illustrative rather than restrictive, and the claimed subject matter is not limited to the details given herein but can be modified within the scope of the embodiments and equivalents thereof.

Claims

1. An automatic wafer centering method, comprising: Operating a manipulator to move a wafer handling component such that the wafer moves past a sensor array, each sensor within the sensor array being configured to detect when an edge of the wafer passes the sensor and emit a signal; Determining a number (N) of detected wafer edge positions after the wafer has passed the sensor array, each detected wafer edge position being defined by a set of coordinates (x, y) in a coordinate system of the wafer handling component; For each unique set (N - 1) of the number (N) of detected wafer edge positions, determining an estimated wafer offset that substantially minimizes a performance index value, the estimated wafer offset being defined as a vector extending from the center of the coordinate system of the wafer handling component to an estimated center position of the wafer, the estimated wafer offset corresponding to the center of a circle that is best fit to a given unique set (N - 1) of the number (N) of detected wafer edge positions within the coordinate system of the wafer handling component, wherein the performance index value is defined as the sum of the squares of the differences between the estimated wafer radius at the i-th detected wafer edge position of the wafer and a known rated wafer radius for a given unique set (N - 1) of the number (N) of detected wafer edge positions, wherein the minimized performance index value and the associated estimated wafer center correspond to the circle that is best fit to a given unique set (N - 1) of the number (N) of detected wafer edge positions within the coordinate system of the wafer handling component, each unique set (N - 1) of the number (N) of detected wafer edge positions having a corresponding estimated wafer offset and a corresponding performance index value; Identifying a final wafer offset as the estimated wafer offset corresponding to the unique set (N - 1) of the number (N) of detected wafer edge positions having the smallest corresponding performance index value; and Using the final wafer offset to center the wafer at a target station.

2. The method according to claim 1, wherein the sensor array comprises at least three sensors.

3. The method according to claim 1, wherein the sensor array comprises at least one sensor positioned to pass through a first half of the wafer relative to the direction of travel of the center of the wafer when the wafer moves past the sensor array, and wherein the sensor array comprises at least one sensor positioned to pass through a second half of the wafer relative to the direction of travel of the center of the wafer when the wafer moves past the sensor array.

4. The method according to claim 1, wherein each sensor in the sensor array is a beam sensor.

5. The method according to claim 4, wherein the beam sensor is configured to transmit a signal indicating detection of the wafer edge when the beam of the beam sensor is interrupted by the wafer, and wherein the beam sensor is configured to transmit a signal indicating detection of the wafer edge when the interrupted beam of the beam sensor re - forms.

6. The method according to claim 1, wherein the unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions having the smallest corresponding performance index value does not include: One detected wafer edge position at the reference portion of the wafer among the several (N) detected wafer edge positions.

7. A system for automatically centering a wafer, comprising: A sensor array, each sensor within the sensor array being configured to detect when the edge of a wafer passes by the sensor; and A controller configured to receive data indicating the position of the wafer handling component of the robot when the wafer handling component of the robot moves with a wafer held on the wafer handling component, The controller is further configured to determine several (N) detected wafer edge positions, each detected wafer edge position being defined by a set of coordinates (x, y) in the coordinate system of the wafer handling component, The controller is further configured to, for each unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions, determine an estimated wafer offset that substantially minimizes a performance index value, the estimated wafer offset being defined as a vector extending from the center of the coordinate system of the wafer handling component to the estimated center position of the wafer, the estimated wafer offset corresponding to the center of a circle that is best - fit to a given unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions within the coordinate system of the wafer handling component, wherein the performance index value is defined as the sum of the squares of the differences between the estimated wafer radius at the i - th detected wafer edge position of the wafer and the known rated wafer radius for a given unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions, wherein the minimized performance index value and the associated estimated wafer center correspond to the circle that is best - fit to a given unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions within the coordinate system of the wafer handling component, each unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions having a corresponding estimated wafer offset and a corresponding performance index value, The controller is further configured to identify the final wafer offset as the estimated wafer offset corresponding to the unique set of (N - 1) detected wafer edge positions among the several (N) detected wafer edge positions having the smallest corresponding performance index value, and Wherein the system is configured to use the final wafer offset to center the wafer at the target station.

8. The system according to claim 7, wherein the sensor array comprises at least three sensors.

9. The system according to claim 7, wherein the sensor array comprises at least one sensor positioned to pass through a first half of the wafer relative to the direction of travel of the center of the wafer as the wafer moves past the sensor array, and wherein the sensor array comprises at least one sensor positioned to pass through a second half of the wafer relative to the direction of travel of the center of the wafer as the wafer moves past the sensor array.

10. The system according to claim 7, wherein each sensor in the sensor array is a beam sensor.

11. The system according to claim 10, wherein the beam sensor is configured to transmit a signal indicating detection of the wafer edge when the beam of the beam sensor is interrupted by the wafer, and wherein the beam sensor is configured to transmit a signal indicating detection of the wafer edge when the interrupted beam of the beam sensor is reformed.

12. The system according to claim 7, wherein the unique set of (N-1) detected wafer edge positions among the several (N) detected wafer edge positions having the smallest corresponding performance index value does not include: One of the several (N) detected wafer edge positions at the reference portion of the wafer.

13. A method for automatically centering a wafer, comprising: Obtaining several (N) detected wafer edge positions, each detected wafer edge position being defined by a set of coordinates (x, y) in the coordinate system of the wafer handling component; For each unique set of (N-1) detected wafer edge positions among the several (N) detected wafer edge positions, determining a minimized performance index value, wherein the performance index value is defined as the sum of the squares of the differences between the estimated wafer radius of the wafer at the i-th detected wafer edge position and the known rated wafer radius for a given unique set of (N-1) detected wafer edge positions among the several (N) detected wafer edge positions, wherein the minimized performance index value has an associated estimated wafer offset corresponding to the center of the circle that best fits the corresponding unique set of (N-1) detected wafer edge positions among the several (N) detected wafer edge positions within the coordinate system of the wafer handling component; Determining the smallest minimized performance index value for the several (N) detected wafer edge positions; and Using the smallest minimized performance index value and its associated wafer offset to center the wafer at the target station.

14. The method according to claim 13, wherein said plurality (N) of detected wafer edge positions are obtained by passing said wafer through a beam sensor array comprising at least three sensors.

15. The method according to claim 13, wherein a particular unique set (N-1) of said plurality (N) of detected wafer edge positions having the smallest minimization performance index value does not include: One of said plurality (N) of detected wafer edge positions at a reference portion of said wafer.

Images