Patent Application
20210096062
Ellipsometers are optical metrology devices that detect changes in the polarization state of light reflected from a surface of a sample in order to measure characteristics of the sample. By way of example,
The ellipsometer 10 must be properly focused on the sample. Some systems use independent focusing systems, i.e., systems that attached to the ellipsometer, but that use an independent light path to determine the position of the focusing system, and thus, the ellipsometer, with respect to the sample. Such focusing systems, however, require very precise alignment, which is expensive and difficult.
Accordingly, an improved focusing system for ellipsometers is desired.
An ellipsometer includes a focusing system that uses an image of the measurement spot to determine a best focal position for the ellipsometer. The focus signal is produced by splitting off the reflected measurement signal before the measurement signal is analyzed by a polarizer thereby avoiding a modulated intensity in the imaged measurement spot. Additionally, splitting off the reflected measurement signal before the analyzer avoids relative movement of the imaged measurement spot from different samples that is caused by differences in reflected elliptical polarizations across the spread of incident angles. The focus signal is imaged on a sensor array and based on the position of the spot on the sensor array, the focal position of the ellipsometer may be determined. A single image may be used to determine the focal position of the ellipsometer permitting a real time focus position measurement.
In one implementations, an ellipsometer includes a source that emits light along a path; a polarizer that polarizes the light to produce a sample beam that interacts with a sample and is reflected to produce a reflected beam; a compensator disposed in the path of the sample beam or the reflected beam, the compensator induces phase retardations of a polarization state of the light, wherein at least one of the polarizer and the compensator rotates about an axis parallel to a propagation direction of the light; a beamsplitter positioned in the path of the reflected beam after the compensator, the beam splitter positioned to receive the reflected beam and to direct a first portion of the reflected beam to a focusing system and to direct a second portion of the reflected beam to an analyzer, wherein the first portion of the reflected beam and the second portion of the reflected beam both include an entire cross-section of the reflected beam; the focusing system positioned to receive the first portion of the reflected beam from the beamsplitter, the focusing system comprising: a lens system that receives the first portion of the reflected beam, the lens system magnifies any deviation from a best focus position of the ellipsometer; and a first detector positioned to receive the first portion of the reflected beam from the lens system, wherein the first detector comprises a two-dimensional sensor and the lens system in the focusing system produce a spot on the two-dimensional sensor and the two-dimensional sensor produces an image of the spot; the analyzer positioned to receive the second portion of the reflected beam from the beamsplitter; a second detector positioned to receive the reflected beam from the analyzer; and a processor that receives the image and is configured to find a location of the spot on the two-dimensional sensor in the image to determine the deviation from the best focus position of the ellipsometer.
In one implementations, a method of focusing an ellipsometer includes generating light along a path; polarizing the light to produce a sample beam that interacts with a sample and is reflected to produce a reflected beam; inducing phase retardations of a polarization state of the light in the sample beam or the reflected beam; rotating at least one of the polarization or the phase retardations of the polarization state; splitting the reflected beam and directing a first portion of the reflected beam along a first path and directing a second portion of the reflected beam along a second path, wherein the first portion of the reflected beam and the second portion of the reflected beam both include an entire cross-section of the reflected beam; focusing the first portion of the reflected beam into a spot on a two-dimensional sensor, wherein any deviation from a best focus position of the ellipsometer is magnified; producing an image of the spot; determining a location of the spot on the two-dimensional sensor in the image; determining the deviation from the best focus position using the location of the spot on the two-dimensional sensor; adjusting a focal position of the ellipsometer based on the deviation from the best focus position; analyzing the polarization state of the second portion of the reflected beam; and detecting the analyzed second portion of the reflected beam.
In one implementations, an ellipsometer includes emits light along a path; a polarizer that polarizes the light to produce a sample beam that interacts with a sample and is reflected to produce a reflected beam; a compensator disposed in the path of the sample beam or the reflected beam, the compensator induces phase retardations of a polarization state of the light; an analyzer positioned in the path of the reflected beam after the compensator, wherein at least one of the polarizer, the compensator, and the analyzerrotates about an axis parallel to a propagation direction of the light; a focusing system positioned in the path of the reflected beam after the compensator and before the analyzer, the focusing system comprising: a beamsplitter positioned to receive the reflected beam and to direct a first portion of the reflected beam to a lens system and to direct a second portion of the reflected beam to the analyzer, wherein the first portion of the reflected beam and the second portion of the reflected beam both include an entire cross-section of the reflected beam; the lens system positioned to receive the first portion of the reflected beam and to direct the first porition of the reflected beam to a first detector, the lens system magnifies any deviation from a best focus position of the ellipsometer; and the first detector positioned to receive the first portion of the reflected beam from the lens system, wherein the first detector comprises a two-dimensional sensor and the lens system in the focusing system produce a spot on the two-dimensional sensor and the two-dimensional sensor produces an image of the spot; a second detector positioned to receive the reflected beam from the analyzer; and a processor that receives the image and is configured to find a location of the spot on the two-dimensional sensor in the image to determine the deviation from the best focus position of the ellipsometer.
A real time focus system for an ellipsometer may use an image of the ellipsometer measurement spot to determine a best focal position for the ellipsometer. The focus signal is produced by splitting off the reflected measurement signal before the measurement signal is analyzed by a polarizer thereby avoiding a modulated intensity in the imaged measurement spot. Additionally, splitting off the reflected measurement signal before the analyzer avoids relative movement of the imaged measurement spot from different samples that is caused by differences in reflected elliptical polarizations across the spread of incident angles. Accordingly, a single image may be used to determine the focal position of the ellipsometer, permitting a real time focus position adjustment.
In order to properly measure the sample 101, the ellipsometer 100 is positioned at a best focal position with respect to the sample 101. The ellipsometer 100, thus, includes an integrated auto focusing system 150 that images the same light rays that are used by the ellipsometer 100 and additionally magnifies the deviation from a best focus position. Focusing system 150 includes a beam splitter 152 that directs a portion of the reflected light 113 to a lens system 154 that focuses the light onto a detector 156.
As illustrated in
As illustrated in
The beam splitter 152 may be a pellicle beam splitter, which may have a thinness, e.g., 0.002 mm, that does not significantly impact the optical path length or aberrations in a converging beam. The effect on optical path length and aberrations would be even less in a collimated beam, as illustrated in
As illustrated in
If the processor 132 is, e.g., a microprocessor, that carries out instructions of a computer program, the data structures and software code for automatically implementing one or more acts described in this detailed description can be implemented by one of ordinary skill in the art in light of the present disclosure and stored, e.g., on a computer readable storage medium, such as memory 134 or disk 142, which may be any device or medium that can store code and/or data for use by a computer system. The computer readable storage medium 134/142 may be, but is not limited to, magnetic and optical storage devices such as disk drives, magnetic tape, compact discs, and DVDs (digital versatile discs or digital video discs). A communication port 144 may also be used to receive instructions that are used to program the processor 132 to perform any one or more of the functions described herein and may represent any type of communication connection, such as to the internet or any other computer network. Additionally, the functions described herein may be embodied in whole or in part within the circuitry of an application specific integrated circuit (ASIC) or a programmable logic device (PLD), and the functions may be embodied in a computer understandable descriptor language which may be used to create an ASIC or PLD that operates as herein described. For example, as discussed above, a field programmable gate array (FPGA) may be used. The FPGA may be either in the detector 156 or on a frame grabber board 157 internal or external to the computer 130. Where processor 132 is an FPGA, computer readable storage medium 134/142 may provide the programming file to embed the desired configuration in the processor 132, which may be performed one time for a non-volatile FPGA, or otherwise at power-up. By avoiding the use of the main system CPU to perform the necessary calculations for auto focusing, the CPU is not slowed down. Further, a dedicated processor increases the image processing speed. Thus, the stage servo controller 108cont may be directly coupled to the frame grabber board 157, which may provide a signal directly to the stage servo controller 108cont through a Serial Peripheral communication Interface (SPI) channel.
As is known in the art, an ellipsometer, such as ellipsometer 100, used to measure the properties and/or structures of a sample vary the polarization state of the light using the polarization state generator 103 or polarization state analyzer 115. For example, the polarization state of the light may be varied by rotating at least one of the polarizer 104 and the compensator 105 about an axis parallel to the propagation direction of the light. In an ideal case, the rotating optic (i.e., the polarizer 104 or compensator 105) rotates around an axis that is perfectly parallel to the optical axis, in which case the location of the illumination spot on the sample 101 will not move. In practice, however, the motor and bearings that rotate the rotating optic are not perfect, resulting in a wobble of the rotating optic. In addition, the input beam to the rotating optic and the output beam after the rotating optic will not be perfectly parallel. As a result of these two effects, the illumination spot on the sample 101 moves, which will cause the spot imaged on the two-dimensional sensor of the detector 156 of the focus system 150 to move as the rotating optic rotates. The movement of the illumination spot in ellipsometer 100 may be small, e.g., less than 2 μm, but, when present, may cause problems for the auto focus system 150. Accordingly, the auto focus system 150 may be configured to compensate for the movement of the illumination spot.
There are a few different ways of varying the polarization state of the light, all of which are contemplated with the present disclosure. One way to vary the polarization state is to continuously rotate the polarizer 104 about the optical axis, while the analyzer 112, which transmits only one polarization state, is fixed. In this method, there is no need for rotating compensator 105. In the simplest case, where the sample does not change the polarization of the incident light, the result is a variation in the intensity of the light after the analyzer 112. For example, the analyzer 112 could be set to transmit only horizontally polarized light and block vertically polarized light. With the polarizer 104 starting in a position in which the illuminating light 111 is horizontally polarized, the analyzer 112 would transmit 100% of the light. As the polarizer 104 rotates 90 degrees about the optical axis, the light is now vertically polarized. As a result, the analyzer 112 will block all of the light, and there will be no signal at the detector 116. As the polarizer moves another 90 degrees (180 degrees total) it is now transmitting horizontally polarized light again and the analyzer would again transmit 100% of the light. A plot of the signal intensity vs. time, thus, produces a sine wave, where the intensity of the signal varies between 0% and 100% and the frequency of the sine wave is double the frequency of the rotation of the polarizer 104. Additionally, the analyzer 112 may rotate, while the polarizer 104 is held fixed. Alternatively, a rotating compensator 105, as illustrated in
In use, a sample under test will change the polarization state of the incident light, which will change the intensity and phase of the resulting signal received by the detector 116. Using the change in intensity and phase, the material properties of the sample 101 may be determined, which is the essence of ellipsometry and is well known in the art. However, the continuous variation of intensity and phase due to the rotating optic may produce problems in an auto focus system for ellipsometers. For example, a variation in the intensity of the signal between 0% and 100% after the analyzer 112 due to a rotating optic results in no or little light being available to determine the focal position of the ellipsometer during portions of the signal cycle.
Accordingly, as illustrated in
Most digital cameras, such as those used in conventional metrology systems include an internal auto exposure control. The auto exposure control of such cameras, however, is inadequate for the present high precision focusing system 150, which uses a single spot on the CCD. Auto exposure control in conventional cameras attempts to adjust the exposure for the whole CCD and thus will lose control when there is single spot on the CCD. Moreover, if the image is downloaded to the computer 130 to perform the exposure control, there would be too much delay, which would lead to unstable control. Thus, the focusing system 150 may use the camera exposure control I/O as a slave to the image processing dedicated processor, e.g., on the detector 156 or frame grabber board 157, so the exposure control is in good correlation to what is needed in the image processing. The exposure is adjusted with respect to only the spot on the CCD and not the entire CCD, which is, e.g., 99% empty. Thus, once the spot is located on the sensor 304 of the detector 156, the exposure error based on the local intensity of the spot is calculated, as opposed to calculating the exposure error over the entire sensor.
Additionally, some samples will scatter unwanted light into the path of the auto focus system 150. The amount and pattern of unwanted light will vary depending on the sample. While the optical system of the ellipsometer 100 may be designed to minimize receiving scattered light, the problem cannot be completely eliminated. If the unwanted or scattered light is not excluded from the spot location calculations of the auto focus system, errors will be produced. Accordingly, the auto focus system 150 is configured to be insensitive to scattered light.
To extract the ROI (406), the image may be summed horizontally (X) and vertically (Y) into vectors and the maximum in both vectors is found in X and Y. Using the maximum in X and Y, the ROI can be located and extracted in the image data. The spot can then be determined from the 2D data of the image in the ROI. Masking and thresholding may be performed (410) to filter noise from the signal. In this step, for example, a histogram showing the number of pixels at a given intensity may be produced. Most of the pixels receive little or no signal, so there will be a large peak near zero intensity. The pixels that are illuminated in the auto focus spot will produce a second peak. However, some pixels outside the auto focus spot may be illuminated due to background or scattered light, which are to be removed. To determine if any given pixel is part of the auto focus spot or part of the background, a technique, such as inter class variation auto threshold algorithm, may be employed. The pixels that are part of the auto focus spot are retained, while pixels that are determined to be outside the auto focus spot, e.g., due to reflections or isolated pixels are eliminated or masked.
The spot location calculation for the pixels inside the mask may then be performed (412). In one embodiment, the spot location may be determined based on the centroid of the spot. Other techniques, however, may be used to calculate the spot location. For example, the average x, y location of the pixels in the spot may be used to determine a location of a center of the spot, or a smoothing function may be used to smooth the points in the spot and the maximum may be used as the location of the center of the spot. Alternatively, the center of the spot may be found using a large scale optimization problem, e.g., by treating the perimeter of the spot as an ellipse and finding the center of the ellipse. Of course, other techniques or variation of the above may be used if desired. By way of example, the location of the spot may be calculated as a centroid based on the gray level values (or, alternatively, the binary values) of pixels that have an intensity that is greater than the threshold. A simple centroid calculation would be to assume that all of the pixels inside of the mask are weighted equally, i.e., binary centroid calculation. However, it has been found that when the spot has a Gaussian distribution, the brighter pixels near the center of the “blob” of remaining pixels have less noise than the dimmer pixels at the edge of the blob. Accordingly, a grey level centroid may be produced using a centroid calculation that is weighted by the pixel intensity (aka grey scale), where brighter pixels have a stronger weighting in the calculation according to their intensity, which improves the focus precision. The use of a binary centroid calculation, however, may be advantageous when the internal structure of the spot is not homogenous and is changing in correlation to the changes on the wafer patterns. The auto focus system should be insensitive to the wafer pattern so in that case a binary centroid calculation may be used.
A phase locked loop (PLL) operation may be used (414) to synchronize with the angular position of the rotating optics, so that wobble in the rotating optics may be compensated. Additionally, the PLL locked on the synchronization signal from the rotating optics may be used to compensate for alignment errors. For example, the center of the sensor 304 might not be perfectly aligned to the motion of the spot on the sensor. Moreover, motion of the spot on the sensor 304 may not be linear, for example, because the motion of the spot on the sensor 304 is the result of angular motion, while the sensor 304 is linear. As discussed above, the controller/driver 105D for the rotating optic (i.e., compensator 105 or polarizer 104) outputs a signal indicating the angular position of the rotating optic, which is acquired during the acquisition synchronization (400) step. Using the angular position signal as a “trigger” signal, a PLL may be used to lock on the trigger signal and compensation for any wobble in the rotating optic or alignment error may be performed using a look up table (LUT) in the FPGA in the detector 156 or frame grabber board 157 (or in computer 130), where the values of the look up table are obtained through a calibration procedure. The system may be calibrated, e.g., using a silicon wafer, where the rotating optic is rotated one or more rotations. At each angular position of the rotating optic, the spot location may be measured for a number of different focal heights, and the average spot location for each focal height at each angular position of the rotating optic, i.e., for every trigger angle, is determined and loaded into the LUT. The LUT may be obtained by stepping through the Z axis and recording the Z height, or the offset from a desired focal position with respect to the measured spot location on the CCD. The spot location may be recorded in terms of pixels in the CCD and the focus errors in microns or stage encoder counts may be output from the LUT. Because the wobble and alignment error are hardware dependent, i.e., a property of the optical components of the ellipsometer, this calibration is not expected to change frequently.
The spot location may be calculated for every image frame, e.g., PLL trigger signal, and thus, in practice, there are more than eight exposures for 360° rotation of the rotating optic. As illustrated in
The ellipse 554 is a result of the systematic mechanical error, e.g., the wobble of the rotating optic, and may be calibrated away, as discussed above, so that the focus measurements may be performed with single image focal plane data acquisition. For example, the image acquisition from detector 156 may be synchronized with the angular position of the rotating optic. The rotating optic hardware may be designed to send, e.g., 13 trigger signals per 360° rotation of the rotating optic. The use of a PLL, as discussed above, permits division the ellipse 554 into any desired number of angles. During calibration, while the ellipsometer 100 is scanned through different focal positions (Z height), an XY positions of the spot 552 on the sensor 304 may be obtained for each focal position at every desired angular position of the rotating optic and stored in the compensation LUT.
For example, during calibration, the ellipsometer 100 is placed above a blank silicon wafer in that is in focus. The PLL will lock on the triggering signal from the rotating optic signal and produce camera triggers at the predefined acquisition angles. The detector 156 will capture images at every designated angle (triggering from the PLL logic). By way of example, in
Referring to
The focal position of the ellipsometer 100 may then be adjusted accordingly (418), e.g., by sending the results to a servo controller for the actuator 109 to move the stage 108 to eliminate the determined focal position offset.
Due to the configuration of the focus system 150, and in particular the location of the beam splitter 152 before the analyzer 112, the intensity of the spot imaged by the detector 156 does not change as the rotating optic rotates. For example, as illustrated in
In contrast, if the analyzer 112 is positioned before the beam splitter 152, the intensity of the spot received by the focus system will be modulated between 0% and 100% due to the rotation of the rotating optic.
In one implementation, rotating at least one of the polarization or the phase retardations of the polarization state causes the location of the spot to move on the two-dimensional sensor and the process may further include compensating for movement of the location of the spot on the two-dimensional sensor. For example, the process may include providing an indication of an angular position of the at least one of the polarization or the phase retardations of the polarization state, wherein compensating for the movement of the location of the spot on the two-dimensional sensor may use the indication of the angular position of the at least one of the polarization or the phase retardations of the polarization state.
Although the present invention is illustrated in connection with specific embodiments for instructional purposes, the present invention is not limited thereto. Various adaptations and modifications may be made without departing from the scope of the invention. Therefore, the spirit and scope of the appended claims should not be limited to the foregoing description.
