Search

EP-4435840-B1 - WIRELESS SUBSTRATE-LIKE TEACHING SENSOR FOR SEMICONDUCTOR PROCESSING

EP4435840B1EP 4435840 B1EP4435840 B1EP 4435840B1EP-4435840-B1

Inventors

  • CHEN, Ferris J.
  • MARK, ROBERT M.
  • DUQUETTE, DAVID W.

Dates

Publication Date
20260513
Application Date
20190422

Claims (12)

  1. A wireless substrate-like sensor (100) for teaching transfer coordinates to a robotic substrate handling system, the sensor (100) comprising: a base portion (104) sized and shaped like a substrate handled by the robotic substrate handling system; an electronics housing (102) coupled to the base portion (104); a power module (200) disposed within the electronics housing (102) and configured to power components of the sensor (100); at least one edge camera (108, 110, 112, 114, 222, 224) disposed near an edge (116) of the base portion (104), the at least one edge camera (108, 110, 112, 114, 222, 224) having a field of view that images an edge of a process chuck configured to receive a semiconductor wafer within the field of view of the at least one edge camera of the wireless substrate-like sensor; at least one structured illuminator mounted relative to the base portion and operably coupled to the at least one edge camera, the at least one structured illuminator being configured to generate structured illumination in the field of view of the at least one edge camera; a controller (206) disposed within the electronics housing (102) coupled to the at least one edge camera (108, 110, 112, 114, 222, 224) and the at least one structured illuminator, the controller (206) being configured to obtain an image from the at least one edge camera (108, 110, 112, 114, 222, 224) having the structured illumination and perform a calibration operation to calibrate apparent range based on the structured illumination, and compensate for lateral shift caused by slight changes in range using the apparent range between the wireless substrate-like sensor and the process chuck , and transmit the compensated location of the process chuck to the robotic substrate handling system to teach the robotic substrate handling system the determined position.
  2. The wireless substrate-like sensor (100) of claim 1, wherein the at least one edge camera (108, 110, 112, 114, 222, 224) includes four edge cameras (108, 110, 112, 114), each disposed proximate an edge (116) of the base portion (104).
  3. The wireless substrate-like sensor (100) of claim 1, wherein the controller (206) is configured to employ a camera (106, 220) proximate to the center of the sensor base to attempt to image a central aperture of the object and selectively engage the at least one edge camera (108, 110, 112, 114, 222, 224) if the attempt to image the central aperture of the object fails.
  4. The wireless substrate-like sensor of claim 1, wherein the wireless substrate-like sensor comprises at least four edge cameras and a structured illuminator disposed proximate each respective edge camera.
  5. The wireless substrate-like sensor of claim 1, wherein the structured illumination is in the form selected from the group consisting of a dot, a circle, a plus symbol, a plurality of parallel lines, a sinusoidally varying intensity pattern in at least one of X and Y direction, diamond shapes, and a checkerboard.
  6. The wireless substrate-like sensor of claim 1, wherein the structured illumination is configured to allow the calibration operation in order to calibrate the apparent range based on the structured illumination.
  7. The wireless substrate-like sensor of claim 4, wherein the controller is configured to perform the calibration operation for each pair of edge camera and structured illuminator.
  8. The wireless substrate-like sensor of claim 1, wherein the controller is configured to utilize stored calibration information in order to accurately calculate range based on the appearance of the structured illumination in the image.
  9. A method of measuring the position of a process chuck (138) of a semiconductor processing tool relative to a wireless substrate-like sensor (100) held by a robotic manipulator of the semiconductor processing tool, the method comprising: providing a wireless substrate-like sensor (100) having at least one edge detecting camera (108, 110, 112, 114, 222, 224) and a structured illuminator for each of the at least one edge camera; causing the structured illuminator to generate structured illumination in a field of view of the at least one edge detecting camera; causing the at least one edge detecting camera (108, 110, 112, 114, 222, 224) to obtain at least one image of a round outer surface of the process chuck (138), the at least one image having the structured illumination; performing a calibration operation to calibrate apparent range based on analyzing the at least one image of the round outer surface of the process chuck (138) having the structured illumination, and compensating for lateral shift caused by slight changes in range using the apparent range between the wireless substrate-like sensor and the process chuck; and communicating the compensated position of the process chuck (138) to a controller (206) of the robotic manipulator.
  10. The method of claim 9, and further comprising using the calibrated sensor (100) to obtain at least one subsequent image of the round outer surface of the process chuck (138).
  11. The method of claim 9, wherein communicating the compensated position includes transmitting wireless data indicative of the compensated position to the controller (206) of the robotic manipulator.
  12. The method of claim 9, wherein the at least one edge detecting camera (108, 110, 112, 114, 222, 224) includes four edge detecting cameras (108, 110, 112, 114).

Description

BACKGROUND Virtually all modern electronics use integrated circuits. These integrated circuits are produced using a semiconductor material, such as silicon, in a semiconductor processing facility. The semiconductor is generally provided in the form of a round, thin wafer that is processed into potentially hundreds or thousands of integrated circuits. The semiconductor processing facility uses automated robots to handle material inside the processing chamber. Wafers must be positioned for processing with high accuracy and repeatability. Wafers must be returned to their storage locations carefully to avoid collisions that can produce debris; the processing environment is required to be extremely clean, since debris or dust can have a significant negative impact on processing yield. Wafer handling systems for semiconductor processing can take many forms. These systems are employed to remove a wafer from a storage pod, move a wafer to a processing station, move a wafer between processing stations and/or return the processed wafer to the pod. Typically, the wafer handling system includes a robotic arm having an end effector that is shaped and sized to be inserted into the pod for wafer removal/replacement. The end effector may include a vacuum system for gripping or otherwise adhering to the wafer to ensure that it does not slide or shift during transport. The end effector must be positionable with precision. Positioning errors in the wafer handling system can cause inadvertent contact between wafers, the end effector and/or other components of the semiconductor processing system. Even minor impacts can dislodge particles that may contaminate other wafers resulting in very costly scrap. In some processing steps, it is necessary to center the wafer very accurately on the processing tool to assure uniform process results across the wafer surface. Accordingly, handling system positioning errors must be minimized. During semiconductor processing, a wafer will generally need to be transferred to and properly positioned on a platform generally called the 'process chuck'. A process chuck is a device that holds a wafer during one or more process steps. It is often close to the size and circular shape of the wafer. The chuck may also have vacuum apertures such that when the wafer is placed upon the process chuck, it can be held securely thereto using vacuum. If a wafer is not placed correctly on the process chuck, the edge dies of the wafer will not be processed correctly. These dies will have defects, resulting in loss of yield. For some processing steps, such as ion etching and others, the chuck may include special sidewalls that help guide the process. The wafer must be placed very accurately within the sidewalls for the process to be successful. In order to reduce positional errors, it is useful periodically to re-align the wafer handling system to known markers in the process tool, the wafer storage and transfer pod, or both. This is commonly called 'teaching' of wafer transfer coordinates to the wafer handling system in order to reduce or eliminate any disparities between the wafer handling system's coordinate system and the coordinate system of the semiconductor processing system. One commercially available system is sold under the trade designation WaferSense ATS by CyberOptics Corporation of Golden Valley, MN. The WaferSense ATS (Auto-Teaching System), uses machine vision to provide a calibrated "eye" that helps teach wafer transfer coordinates to a wafer handling system. This system employs a sensor that is shaped like a wafer so that it can be handled exactly like a process wafer. In this way, the sensor can be moved about the semiconductor processing system to measure robot handoffs without opening the process chamber. Opening and closing the process chamber can be an expensive hours- to days-long interruption in production. A wafer-like sensor also can access difficult-to-reach stations like pre-aligners and load locks. The WaferSense ATS sensor is taught to recognize circular target features, such as a central hole in the chuck, and uses machine vision and its on-board signal processing to measure X-Y-Z offset from the sensor's calibrated geometric center to the center of the target feature. The ATS sensor allows equipment users to save the time during equipment maintenance without opening chamber and the accurate robot teaching ensures the wafer is transferred properly to the center of the process chuck. The current ATS sensor requires a round shape target in the center of the wafer chuck so the ATS can use it for calculating the deviation between the wafer and the chuck. Also, the robot blade or end effector needs to have an unobstructed opening or port so the camera of ATS sensor can see through it to find the round shape of the process chuck. However, a significant percentage of semiconductor processing equipment (~>70%) does not have the round shape target on the process chuck or the end effector does