Patent Application
20050224899
Semiconductor processing systems are characterized by extremely clean environments and extremely precise semiconductor wafer movement. Industries place extensive reliance upon high-precision robotic systems to move substrates, such as semiconductor wafers, about the various processing stations within a semiconductor processing system with the requisite precision.
Reliable and efficient operation of such robotic systems depends on precise positioning, alignment, and/or parallelism of the components. Accurate wafer location minimizes the chance that a wafer may accidentally scrape against the walls of a wafer processing system. Accurate wafer location on a process pedestal in a process chamber may be required in order to optimize the yield of that process. Precise parallelism between surfaces within the semiconductor processing systems is important to ensure minimal substrate sliding or movement during transfer from a robotic end effector to wafer carrier shelves, pre-aligner vacuum chucks, load lock elevator shelves, process chamber transfer pins and/or pedestals. When a wafer slides against a support, particles may be scraped off that cause yield loss. Misplaced or misaligned components, even on the scale of fractions of a millimeter, can impact the cooperation of the various components within the semiconductor processing system, causing reduced product yield and/or quality.
This precise positioning must be achieved in initial manufacture, and must be maintained during system use. Component positioning can be altered because of normal wear, or as a result of procedures for maintenance, repair, alteration, or replacement. Accordingly, it becomes very important to automatically measure and compensate for relatively minute positional variations in the various components of a semiconductor processing system.
In the past, attempts have been made to provide substrate-like sensors in the form of a substrate, such as a wafer, which can be moved through the semiconductor processing system to wirelessly convey information such as substrate inclination and acceleration within the semiconductor system. As used herein, “substrate-like” is intended to mean a sensor in the form of substrate such as a semiconductor wafer, a Liquid Crystal Display glass panel or reticle. Attempts have been made to provide wireless substrate-like sensors that include additional types of detectors to allow the substrate-like sensor to measure a host of internal conditions within the processing environment of the semiconductor processing system. Wireless substrate-like sensors enable measurements to be made at various points throughout the processing equipment with reduced disruption of the internal environment as well as reduced disturbance of the substrate handling mechanisms and fabrication processes (e.g.: baking, etching, physical vapor deposition, chemical vapor deposition, coating, rinsing, drying etc.). For example, the wireless substrate-like sensor does not require that a vacuum chamber be vented or pumped down; nor does it pose any higher contamination risk to an ultra-clean environment than is suffered during actual processing. The wireless substrate-like sensor form factor enables measurements of process conditions with minimal observational uncertainty.
Since wireless substrate-like sensors are transported through the actual semiconductor processing environment, it is important that they not adversely affect the environment itself. Thus, such sensors should not allow particles to break off therefrom, nor outgas. Moreover, in order to ensure that such sensors can move to every location within the semiconductor processing environment that a normal substrate could move to, the dimensions of the sensor should be at least as small as a maximum substrate size, but preferably smaller. Finally, in order to ensure accuracy of measurements of the sensor, it is important that the sensor's weight does not cause any significant deflection or other form of displacement on the handling apparatus. Thus, such sensors should be relatively light-weight.
Thus, there exists a current need in the field of wireless substrate-like sensors for devices that are clean, light-weight, and low-profile.
In accordance with an aspect of the present invention, a wireless substrate-like sensor is configured to ensure it does not contaminate a semiconductor processing chamber. The sensor is sealed except for one or more apertures. In one embodiment, a vent is disposed proximate the apertures. In another embodiment, the aperture is coupled to a pressure equalization member that deforms in response to a differential in pressure between the sensor interior and exterior.
Sensor 112 is preferably constructed from light-weight, dimensionally stable materials. Sensor 112 is preferably constructed from a base material that has a high stiffness such as an aluminum alloy, aluminum, magnesium, and/or a ceramic. The sensor housing itself may be coated with any suitable coatings including aluminum oxide, nickel, or ceramics in order to improve mechanical or chemical properties.
In order for the substrate-like sensor to accurately measure a three-dimensional offset, it is important for the sensor to deform in a manner similar to that of an actual substrate. Common wafer dimensions and characteristics may be found in the following specification: SEMI M1-0302, “Specification for Polished Monocrystaline Silicon Wafers”, Semiconductor Equipment and Materials International, www.semi.org. The center of a 300 mm silicon wafer supported at its edges will sag approximately 0.5 mm under its own weight. The difference in the deformation of the sensor and the deformation of an actual wafer should be much less than the accuracy of sensor measurement. In a preferred embodiment, the stiffness of the substrate-like sensor results in a deflection that is nearly identical to that of an actual silicon wafer. Therefore, no compensation is required to correct for any differential deflection. Alternatively, a compensation factor may be added to the measurement. Similarly, the weight of the substrate-like sensor will also deflect its support. Substrate supports include, but are not limited to: end effectors, pedestals, transfer pins, shelves, etc. The differential support deflection will be a function both of the difference in weights of the sensor and a substrate as well as the mechanical stiffness of the substrate support. The difference between deflection of the support by the sensor and that by a substrate should also be much less than the accuracy of sensor measurement, or the deflection difference should be compensated by a suitable calculation.
In the prior art, technicians have iteratively adjusted the alignment of a vacuum transfer robot end effector with a process chamber pedestal by viewing them after removing the lid of the process chamber or through a transparent window in the lid. Sometimes a snuggly fitting fixture or jig must first be placed on the process pedestal to provide a suitable reference mark. The substrate-like sensor enables an improved, technician assisted, alignment method. The substrate-like sensor provides an image of the objects being aligned without the step of removing the cover and with greater clarity than viewing through a window. The wireless substrate-like sensor saves significant time and improves the repeatability of alignment.
A wireless substrate-like sensor can transmit an analog camera image by radio.
A preferred embodiment uses a machine vision sub-system of a substrate-like wireless sensor to transmit all or a portion of the digital image stored in its memory to an external system for display or analysis. The display may be located near the receiver or the image data may be relayed through a data network for remote display. In a preferred embodiment, the camera image is transmitted encoded as a digital data stream to minimize degradation of image quality caused by communication channel noise. The digital image may be compressed using any of the well known data reduction methods in order to minimize the required data rate. The data rate may also be significantly reduced by transmitting only those portions of the image that have changed from the previous image. The substrate-like sensor or the display may overlay an electronic cross hair or other suitable mark to assist the technician with evaluating the alignment quality.
While vision-assisted teaching is more convenient than manual methods, technician judgment still affects the repeatability and reproducibility of alignment. The image acquired by a substrate-like wireless sensor camera may be analyzed using many well-known methods, including two-dimensional normalized correlation, to measure the offset of a pattern from its expected location. The pattern may be an arbitrary portion of an image that the vision system is trained to recognize. The pattern may be recorded by the system. The pattern may be mathematically described to the system. The mathematically described pattern may be fixed at time of manufacture or programmed at the point of use. Conventional two-dimensional normalized correlation is sensitive to changes in the pattern image size. When a simple lens system is used, magnification varies in proportion to object distance. Enhanced pattern offset measurement performance may be obtained by iteratively scaling either the image or the reference. The scale that results in the best correlation indicates the magnification, provided the size of the pattern is known, or the magnification, as used when the reference pattern was recorded, is known.
When the correspondence between pixels in the image plane to the size of pixels in the object plane is known, offsets may be reported in standard units of measure that are easier for technicians or machine controllers to interpret than arbitrary units such as pixels. For example, the offset may be provided in terms of millimeters such that the operator can simply adjust the systems by the reported amount. The computations required to obtain the offset in standard units may be performed manually, by an external computer, or preferentially within the sensor itself. When the sensor extracts the required information from an image, the minimum amount of information is transmitted and the minimum computational burden is placed on the technician or external controller. In this way objective criteria may be used to improve the repeatability and reproducibility of the alignment. Automated offset measurement improves the reproducibility of alignment by removing variation due to technician judgment.
During alignment and calibration of semiconductor processing equipment, it is not only important to correctly position an end effector relative to a second substrate supporting structure, it is also important to ensure that both substrate supporting structures are parallel to one another. In a preferred embodiment, a machine vision subsystem of a wireless substrate-like sensor is used to measure the three dimensional relationship between two substrate supports. For example: a robotic end effector may hold a wireless substrate-like sensor in close proximity to the transfer position and a measurement of the three dimensional offset with six degrees of freedom may be made from the sensor camera to a pattern located on an opposing substrate support. One set of six degrees of freedom includes yaw, pitch, and roll as well as displacement along the x, y, and z axes of the Cartesian coordinate system. However, those skilled in the art will appreciate that other coordinate systems may be used without departing from the spirit and scope of the invention. Simultaneous measurement of both parallelism and Cartesian offset allows a technician or a controller to objectively determine satisfactory alignment. When a controller is used, alignments that do not require technician intervention may be fully automated. Automated alignments may be incorporated into scheduled preventive maintenance routines that optimize system performance and availability.
In a very general sense, operation and automatic calibration of robotic system 102 is performed by instructing robot 102 to select and convey sensor 112 to reference target 114. Once instructed, robot 102 suitably actuates the various links to slide end effector 116 under sensor 112 to thereby remove sensor 112 from container 100. Once removed, robot 102 moves sensor 112 directly over reference target 114 to allow an optical image acquisition system (not shown in
This information allows sensor 112 to be used to acquire images of additional targets, such as target 116 on system component 104 to calculate a precise position and orientation of system component 104. Repeating this process allows the controller of robot 102 to precisely map exact positions of all components within a semiconductor processing system. This mapping preferably generates location and orientation information in at least three and preferably six degrees of freedom (x, y, z, yaw, pitch and roll). The mapping information can be used by a technician to mechanically adjust the six degree of freedom location and orientation of any component with respect to that of any other component. Accurate measurements provided by the substrate-like wireless sensor are preferably used to minimize or reduce variability due to technician judgment. Preferably, this location information is reported to a robot or system controller which automates the calibration process. After all mechanical adjustments are complete; the substrate-like sensor may be used to measure the remaining alignment error. The six degrees of freedom offset measurement may be used to adjust the coordinates of points stored in the memories of the robot and/or system controllers. Such points include, but are not limited to: the position of an atmospheric substrate handling robot when an end effector is located at a FOUP slot #1 substrate transfer point; the position of an atmospheric substrate handling robot when an end effector is located at a FOUP slot #25 substrate transfer point; the position of an atmospheric substrate handling robot when an end effector is located at a substrate pre-aligner substrate transfer point; the position of an atmospheric substrate handling robot when an end effector is located at a load lock substrate transfer point; the position of an atmospheric substrate handling robot when an end effector is located at a reference target attached to the frame of an atmospheric substrate handling system; the position of a vacuum transfer robot when its end effector is located at a load lock substrate transfer point; the position of a vacuum transfer robot when an end effector is located at a process chamber substrate transfer point; and the position of a vacuum transfer robot when an end effector is located at a target attached to the frame of a vacuum transfer system.
An alternative embodiment of the present invention stores and reports the measurements. Real-time wireless communication may be impractical in some semiconductor processing systems. The structure of the system may interfere with wireless communication. Wireless communication energy may interfere with correct operation of a substrate processing system. In these cases, sensor 112 can preferably record values as it is conveyed to various targets, for later transmission to a host. When sensor 112, using its image acquisition system, or other suitable detectors, recognizes that it is no longer moving, sensor 112 preferably records the time and the value of the offset. At a later time, when sensor 112 is returned to its holster (not shown) sensor 112 can recall the stored times and values and transmit such information to the host. Such transmission may be accomplished by electrical conduction, optical signaling, inductive coupling or any other suitable means. Store and report operation of the wireless substrate-like sensor potentially: increases the reliability, lowers the cost and shortens a regulatory approval cycle for the system. Moreover, it avoids any possibility that the RF energy could interact with sensitive equipment in the neighborhood of the sensor and its holster. Store and report operation can also be used to overcome temporary interruptions of a real-time wireless communication channel.
Both sensor 112 and sensor 118 employ central region 120. A portion of the underside of central portion 120 is disposed directly over an access hole 126 as illustrated in
Illumination module 150, which preferably comprises a number of Light Emitting Diodes (LEDs), and image acquisition system 152 are coupled to digital signal processor 144 through camera controller 154. Camera controller 154 facilitates image acquisition and illumination thus providing relevant signaling to the LEDs and image acquisition system 152 as instructed by digital signal processor 144. Image acquisition system 152 preferably comprises an area array device such as a Charge Coupled Device (CCD) or Complementary Metal Oxide Semiconductor (CMOS) image device coupled preferably to an optical system 156, which focuses images upon the array. Preferably, the image acquisition device is available from Kodak under the trade designation KAC-0310. Digital signal processor 144 also preferably includes a number of I/O ports 158, 160. These ports are preferably serial ports that facilitate communication between digital signal processor 144 and additional devices. Specifically, serial port 158 is coupled to radio-frequency module 162 such that data sent through port 158 is coupled with external devices via radio frequency module 162. In one preferred embodiment, radio frequency module 162 operates in accordance with the well-known Bluetooth standard, Bluetooth Core Specification Version 1.1 (Feb. 22, 2001), available from the Bluetooth SIG (www.bluetooth.com). One example of module 162 is available from Mitsumi under the trade designation WML-C11.
Detectors 164 may take any suitable form and provide relevant information regarding any additional conditions within a semiconductor processing system. Such detectors can include one or more thermometers, accelerometers, inclinometers, compasses (Magnetic field direction detectors), light detectors, pressure detectors, electric field strength detectors, magnetic field strength detectors, acidity detectors, acoustic detectors, humidity detectors, chemical moiety activity detectors, or any other types of detector as may be appropriate.
In accordance with another embodiment of the present invention, a wireless substrate-like sensor is provided with improved safeguards against contaminating a semiconductor wafer processing chamber. It is extremely important that such sensors measure the physical properties while not contaminating the processing chamber. Moreover, such sensors must be dimensionally stable. Well known sensor materials and components may shed particles that could contaminate the wafer processing chamber. If a wireless substrate-like sensor is sealed to isolate potentially contaminating materials inside the sensor, a pressure differential may arise between the interior and exterior. If sufficiently extreme, the pressure differential could potentially deform the housing, or even cause a rupture. This is particularly so for a light-weight substrate-like sensor housing which may be mechanically weak due to the desire to minimize the total weight of the housing.
Wireless substrate-like sensors generally have an internal space and an external surface. Some of the sensor apparatus is contained within the internal space. The sensor housing includes a seal that prevents gas, particles or molecules from entering or leaving the internal space except through a vent that is specifically provided for that purpose. A filter is provided across the vent that allows the passage of gas, but prevents the passage of particles or molecules too large to fit through the filter. Preferably, the external surface of the sensor is constructed from or coated or deposited with chemically unreactive materials such as: nickel, polyethylene or polycarbonate. The shape and finish of the sensor housing is also preferably selected such that the sensor itself is easy to clean. External crevices and corners where particles may become trapped are also preferably minimized.
Although the present invention has been described with reference to preferred embodiments, workers skilled in the art will recognize that changes may be made in form and detail without departing from the spirit and scope of the invention.
The present application claims priority to previously filed co-pending provisional application Ser. No. 60/551,460, filed Mar. 9, 2004, entitled WIRELESS SUBSTRATE-LIKE SENSOR, which application is incorporated herein by reference in its entirety; and the present application is a Continuation-In-Part of U.S. patent application Ser. No. 10/356,684, filed Jan. 31, 2003, entitled WIRELESS SUBSTRATE-LIKE SENSOR.
| Number | Date | Country | |
|---|---|---|---|
| 60551460 | Mar 2004 | US | |
| 60354551 | Feb 2002 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10356684 | Jan 2003 | US |
| Child | 11075357 | Mar 2005 | US |
