Image based metrology of surface deformations

Abstract

A method for detecting a local tilt region on a sample using imaging reflectometer measurements includes the steps of acquiring a first image without blocking any light reflected from the sample, and acquiring a second image while blocking some light reflected from the sample at the aperture plane. The local tilt region is detected by comparing the first reflectance intensity values ​​of pixels in the first image with the second reflectance intensity values ​​of corresponding pixels in the second image.

Description

Technology Field Cross-references regarding related applications

[0001] This application claims priority to US No. 16 / 877,866, filed May 19, 2020. The disclosure of such document is incorporated herein by reference in whole and for all purposes. field

[0002] The embodiments described herein generally relate to mapping a sample shape and / or detecting local slope regions on a sample using image-based metrology measurements. Background Technology

[0003] The deposition of thin films on samples, such as semiconductor wafers, often results in the occurrence of localized stress. If left untreated, the resulting stress profile can distort the sample shape in the form of two-dimensional warping, sometimes referred to as "potato chipping." In some subsequent processing steps, the sample can be chucked to flatten it and reduce warping, but a certain amount of uncorrected residual displacement may remain. This can cause unacceptable errors in processes such as photolithography.

[0004] A map of the sample geometry can be used to adjust subsequent processes in an effort to reduce or correct deformation. It is desirable for such mapping technology to be integrated into the machining tool. As such, it is desirable for the measurement solution to be robust, reliable, and relatively inexpensive.

[0005] Current methods for providing information on sample geometry include global sample interferometers, multi-spot local gradient measurements, and shear interferometers of the entire sample using large grids. Multi-spot methods lack continuity in mapping, and global interferometer systems are complex and expensive. Furthermore, these systems are typically designed to perform only a single function—namely, to detect surface deformation without providing any accompanying information, such as images or arbitrary pattern features. In practice, many existing methods for determining sample geometry or topography often struggle to operate on patterned samples.

[0006] Accordingly, improved techniques are desired for mapping sample geometry and topographic changes that cause local inclination or distortion.

[0007] The embodiments described herein provide improved methods for mapping sample shape and detecting regions of local tilt or distortion on the sample. While various embodiments utilize image-based reflection measurements, it should be recognized that other imaging techniques may also be used to map sample shape and detect regions of local tilt in a similar manner. Only, as an example of an embodiment, multiple image-based reflection measurements are used to detect changes in reflectance intensity. One image may be acquired without blocking any light at the aperture plane of the detector, and another image may be acquired while blocking a portion of the light at the aperture plane. The blocked light may be a portion reflected from non-flat portions of the sample. These portions of the sample are distorted or have a local tilt. Light from these portions of the sample is shifted at the aperture plane to block the portion thereof, and accordingly, corresponding points in the image have reduced intensity. Accordingly, a comparison of reflectance intensity between images provides a means for detecting distortion or local tilt on the sample. Other embodiments and technologies are described herein.

[0008] According to a specific embodiment, a method for detecting local tilt regions on a sample, for example, using an imaging reflectometer, comprises the steps of illuminating a measurement area on the sample using an input beam and performing a first imaging reflectance measurement. The step of performing the first imaging reflectance measurement includes receiving a first imaging beam reflected from the sample at an imaging sensor. An iris positioned on the aperture plane of the imaging reflectometer is configured such that the first imaging beam passes through the pupil of the iris and the iris does not block any portion of the first beam reflected from the sample. A first image of the measurement area is obtained using the first imaging beam reflected from the sample and received at the imaging sensor. The method also includes performing a second imaging reflectance measurement by receiving a second imaging beam reflected from the sample at an imaging sensor. An iris positioned on the aperture plane of the imaging reflectometer is configured such that the second imaging beam passes through the pupil of the iris and the iris blocks a portion of the second beam reflected from the sample. A second image of the measurement area is obtained using the second imaging beam reflected from the sample and received at the imaging sensor. The present method also includes detecting local slope regions within a measurement area by comparing first reflectance intensity values ​​of pixels in a first image with second reflectance intensity values ​​of corresponding pixels in a second image.

[0009] According to another embodiment, a method for detecting local tilt regions on a sample using an imaging reflectometer configured for dark-field measurements comprises: a step of illuminating a measurement area on a sample using an input beam, wherein a portion of the input beam passes through an illumination pupil; a step of receiving portions of an imaging beam reflected from local tilt regions on a sample at an imaging sensor, wherein portions of the imaging beam reflected from local tilt regions pass through an imaging pupil, wherein the imaging pupil is configured to block portions of the imaging beam reflected from flat regions on the sample; a step of acquiring a first image of the measurement area using portions of the imaging beam reflected from local tilt regions on the sample and received at the imaging sensor; and a step of detecting local tilt regions within the measurement area based on the locations of bright spots in the first image.

[0010] According to another embodiment, a method for detecting local tilt regions on a sample using an imaging reflectometer comprises the steps of illuminating a measurement area on the sample using an input beam and performing a first imaging reflectance measurement. The step of performing the first imaging reflectance measurement includes receiving a first imaging beam reflected from the sample at an imaging sensor. The pupil of an iris oriented in the aperture plane of the imaging reflectometer is laterally offset with respect to the first beam reflected from the sample such that a portion of the first imaging beam passes through the pupil of the iris and is blocked at the iris. A first image of the measurement area is obtained using the portion of the first imaging beam received at the imaging sensor that passes through the pupil of the iris. The present method also includes performing a second imaging reflectance measurement by receiving a second imaging beam reflected from the sample at an imaging sensor. The pupil of an iris oriented in the aperture plane of the imaging reflectometer is laterally offset with respect to the second beam reflected from the sample such that a portion of the second imaging beam passes through the pupil of the iris and is blocked at the iris. A second image of the measurement area is obtained using a portion of the second imaging beam received from the imaging sensor after passing through the pupil of the iris. The present method also includes detecting local slope regions within the measurement area by comparing the first reflectance intensity values ​​of pixels in the first image with the second reflectance intensity values ​​of corresponding pixels in the second image.

[0011] According to another embodiment, a method for detecting local slope regions on a sample using an imaging reflectometer comprises the steps of: illuminating a measurement region on a sample using an input beam; receiving an imaging beam reflected from the sample at an imaging sensor, wherein the imaging beam reflected from the sample passes through an imaging pupil; and acquiring a first image of the measurement region using the imaging beam reflected from the sample and received at the imaging sensor. The method further comprises the steps of: illuminating a measurement region on a sample using an input beam, wherein a portion of the input beam is blocked at an illumination pupil; receiving an imaging beam reflected from the sample at an imaging sensor, wherein the imaging beam reflected from the sample passes through an imaging pupil; and acquiring a second image of the measurement region using the imaging beam reflected from the sample and received at the imaging sensor. The method further comprises the steps of: detecting local slope regions within the measurement region; and determining the slope of the local slope regions based on a change in the amount of luminance of pixels in the second image compared with the luminance of corresponding pixels in the first image.

[0012] In one embodiment, approximately half of the input beam is blocked at the illumination pupil.

[0013] In another embodiment, local slope regions within the measurement area of ​​the second image having positive and negative slopes appear as regions of greater or lower luminance, or opposite, compared to the first image.

[0014] Additional embodiments, advantages, and features are apparent from the claims, description, and the attached drawings. Brief explanation of the drawing

[0015] The various embodiments described herein, along with their features and advantages, as well as both the organization and the method of operation, can be best understood by referring to the following detailed description and the attached drawings.

[0016] Figure 1 is a simplified cross-sectional view of an imaging reflector.

[0017] Figure 2 is a simplified cross-sectional view of a multi-wavelength light source.

[0018] Figure 3 is a flowchart schematically illustrating a method for measuring the reflectance of a sample.

[0019] FIGS. 4a to 4c are simplified cross-sectional views of light passing through an aperture in the aperture plane of an imaging reflector according to some embodiments.

[0020] FIG. 5 is a simplified diagram illustrating reflectance intensities measured at different points around a local slope area on a sample according to one embodiment.

[0021] FIG. 6 is a simplified diagram illustrating reflectance intensities measured at different points around a local slope area on a sample according to another embodiment.

[0022] FIGS. 7a and 7b are simplified diagrams illustrating how knife edge arrays can be used to determine the local slope of a provided direction according to one embodiment.

[0023] FIG. 8 is a simplified diagram illustrating a lighting-based implementation according to one embodiment.

[0024] FIGS. 9a and 9b are simplified diagrams of a dark field implementation according to one embodiment.

[0025] FIGS. 10a and FIGS. 10b are simplified diagrams of a dark-field implementation according to a different embodiment.

[0026] FIGS. 11 and FIGS. 12 are flowcharts illustrating methods for detecting local tilt regions on a sample using an imaging reflectometer according to some embodiments.

[0027] For the sake of simplicity and clarity of the example, it should be recognized that the elements depicted in the drawings are not necessarily drawn in a fixed proportion. For example, the dimensions of some elements may be exaggerated relative to others for clarity. Additionally, where deemed appropriate, reference numbers may be repeated between drawings to indicate corresponding or similar elements. Specific details for implementing the invention

[0028] In the following detailed description, numerous specific details are described to provide a complete understanding of the embodiments described herein. However, it should be understood that various embodiments may be practiced without these specific details. In other cases, widely known methods, procedures, and components are not described in detail so as not to obscure the described features.

[0029] Various embodiments will be described in detail, and one or more examples thereof are illustrated in the drawings. Each example is provided for illustrative purposes only and is not intended to be limiting. Additionally, features illustrated or described as part of one embodiment may be used in or together with other embodiments to form other embodiments. The description is intended to include such variations and changes.

[0030] The terms “specimen,” “sample,” or “wafer” as used herein include, but are not limited to, semiconductor wafers, semiconductor workpieces, photolithography masks, and other workpieces, e.g., memory disks, etc. According to some embodiments which may be combined with other embodiments described herein, systems and methods are configured or applied for surface metrology applications, e.g., reflection measurements.

[0031] The embodiments described herein generally relate to improved methods for mapping wafer geometry and / or detecting local tilt regions on a wafer. In some embodiments, one image is acquired without blocking any light at the aperture plane of the detector, and another image is acquired while blocking some of the light at the aperture plane. Changes in reflectance intensity between the images can be used to identify warping and / or local tilt on the wafer.

[0032] FIG. 1 is a simplified cross-sectional view of an imaging reflectometer (100) according to one embodiment. The imaging reflectometer (100) illustrated in this example may be used to implement the methods described herein. However, the imaging reflectometer (100) is illustrated only as an example, and other imaging reflectometers and / or other imaging techniques may perform the described methods. As only an example, imaging reflectometers configured to acquire images of a portion of a sample rather than the entire sample, and / or imaging reflectometers using single or multi-wavelength light sources may be used to perform the methods.

[0033] In the example of FIG. 1, light from the source module (102) is transmitted to the homogenizer (108) through the light guide (106). Light (116) from the homogenizer (108) passes through the illumination pupil (114) and is guided toward the beam splitter (120). A portion (138) of the light (116) is reflected by the beam splitter (120) toward the reference sensor (134), and a portion (122) of the light passes through the beam splitter (120) and continues toward the sample (130) along the optical path.

[0034] A portion (122) of the light (116) that has passed through the beam splitter (120) is imaged onto the sample (130) by the lens (126). The light reflected from the sample (130) is guided through at least a portion of the lens (126) and reflected toward the imaging sensor (158) by the beam splitter (120).

[0035] The imaging reflector (100) may include a plurality of other lenses (e.g., 110, 112, 118, 136) that shape and / or guide light along optical paths to illuminate a sample (130), illuminate a reference sensor (134), guide light to other lenses (e.g., 120, 140, 144), and guide light reflected from the sample to an imaging sensor (158). In some embodiments, for example, light may pass through one or more polarizers (e.g., polarizers (110, 154)). One or more polarizers may be inserted into the illumination and / or imaging paths to provide enhanced sensitivity to different features of the sample (130). Wave plates may also be inserted to change the phase of the polarization. Wave plates and / or polarizers may be intended to provide polarized reflection measurements at fixed angles, or may be rotated to provide elliptical polarization analysis measurements. It should be recognized that imaging reflectometers according to the embodiments described herein may not include all of the optical elements illustrated in the example of FIG. 1 and / or may include other optical elements not included in such example.

[0036] In this example, the source module (102) provides a multi-wavelength light source capable of sequentially generating different light beams, each having a narrow wavelength range. In some embodiments, the multi-wavelength light source is provided by a plurality of light sources that can be individually activated. Each of the light sources generates a light beam, and at least some of the light beams have different nominal wavelengths.

[0037] In other embodiments, a multi-wavelength light source is provided by adjusting the source power to the source module (102) to generate light beams having different nominal wavelengths. The power for each wavelength can be controlled independently to optimize the dynamic range of the reflectance measured at each wavelength.

[0038] In other embodiments, a multi-wavelength light source is provided by a broadband light source and a set of bandpass filters. The broadband light source can be used with bandpass filters to generate light beams at selected nominal wavelengths.

[0039] In other embodiments, the source module (102) may include both a plurality of light sources, a broadband light source, and a set of bandpass filters.

[0040] In one embodiment, the lens (126) has a measurement field size (or illumination area) slightly larger than the size of the sample (130) so that whole-sample images can be acquired by the imaging sensor (158) without scanning the light or moving the stage (132). For example, the lens (126) may have a measurement field size of 300 mm or more to measure a semiconductor wafer having a diameter of 300 mm. The lens (126) may be a telecentric lens that causes the light rays traveling from the lens (126) to the sample (130) to be approximately parallel to an optical axis that is substantially perpendicular to the surface of the sample (130). This provides substantially normal illumination over the entire sample (130) or across the entire measurement area. This can reduce measurement error because the illumination angles are approximately the same. Telecentric imaging allows light reflected at substantially the same angle across the entire field to reach the imaging sensor. In one embodiment, for example, the light illuminating the sample (130) may have a telecentricity error of less than 0.3 degrees over a wavelength range of about 350 nm to about 1200 nm, and in some embodiments, a telecentricity error of less than 1% over a wavelength range of about 350 nm to about 1100 nm. As used herein, the telecentricity error is a measure of the angular deviation of a light ray that is incident and reflected from the wafer surface normal (or optical axis).

[0041] In some embodiments, the lens (126) may be a lens having a field size smaller than that of the sample (130). In this case, the area (or measurement area) is imaged, and the optics and / or stage (132) may be moved and / or the optical module may be scanned to image adjacent fields. Depending on the application, the measurement area may be approximately the same size as the die or stepper field. Adjacent images may be stitched together using known techniques to provide multi-field or full-sample images.

[0042] The imaging sensor (158) may be an area imaging sensor comprising one or more digital cameras for capturing light (142) reflected from the sample (130) and passing through the pupil in the aperture plane (150). The imaging sensor (158) provides an image of the sample (130) based on the received light (142). In some embodiments, the imaging sensor (158) may include a single camera configured to image the entire surface of the sample (130). In other embodiments, the imaging sensor (158) may include a single camera configured to image a portion of the sample (130). In yet another embodiment, the imaging sensor (158) may include multiple cameras, each imaging adjacent or slightly overlapping fields (or measurement areas) on the sample (130). Adjacent images may be stitched together using known techniques. Image resolution may be increased by using a higher resolution imaging sensor or by using multiple imaging sensors, each imaging a smaller field.

[0043] The imaging reflector (100) includes an illumination path that provides light to the sample (130) and an imaging path that provides reflected light to the imaging sensor (158). This allows for independent control of the illumination aperture number (NA) and the imaging NA. As just one example, if the imaging sensor (158) has an array size of 5120 pixels × 5120 pixels and the imaging NA is about 0.004, the pixel size on the sample (130) is about 60 µm for a 300 mm wafer, which has a Rayleigh resolution of about 55 µm at a wavelength of 365 nm and a Rayleigh resolution of about 153 µm at a wavelength of 1 µm. Generally, the illumination NA is larger than the imaging NA to correct for residual chromatic telecentric errors and to provide tolerance for the tilt and bow of the sample (130). In some embodiments, the illumination NA may be in the range of about 0.005 to about 0.5, and the imaging NA may be in the range of about 0.003 to about 0.2.

[0044] The reference sensor (134) may include one or more digital cameras for capturing light (138) reflected from the beam splitter (120). The reference sensor (134) may have a lower resolution than the imaging sensor (158). The reference sensor (134) may be used to monitor the uniformity and stability of the light (138) and to provide real-time corrections to the reflectance measurements taken by the imaging sensor (158). Measurements from the reference sensor (134) may be used to adjust the characteristics of the light sources (e.g., output power) to provide spatial and temporal corrections.

[0045] FIG. 2 is a simplified cross-sectional view of a multi-wavelength light source according to one embodiment. The multi-wavelength light source may be used, for example, as part of the source module (102) of the imaging reflector (100) of FIG. 1. The multi-wavelength light source comprises a plurality of light sources (202) and a plurality of optical fibers (206). Each of the light sources (202) may include one or more light-emitting diodes (LEDs) and / or laser diodes (LDs). Each of the light sources (202) is optically coupled to a homogenizer (208) by one of the optical fibers (206). Each of the light sources (202) generates a beam, and at least some of the beams may have different nominal wavelengths. Light from the homogenizer (208) may be guided to a lens and used to image a sample, as described in relation to FIG. 1.

[0046] In one embodiment, a multi-wavelength light source sequentially generates different beams among the input beams and sequentially generates combinations of multiple input beams. The beams may be generated sequentially at a switching rate generally equal to the frame rate of an imaging sensor (e.g., the imaging sensor (158) shown in FIG. 1) to achieve one image per wavelength of the same field on the sample. In some embodiments, the frame rate of the sensor may be faster than the wavelength switching rate. A faster switching rate may average multiple images at each wavelength to achieve a higher signal-to-noise ratio. The output power and / or integration time for each of the light sources (202) may be independently controlled and adjusted so that the sensor signal at each wavelength is close to saturation to maximize the signal-to-noise ratio. Each of the light sources (202) may have sufficient output power to enable high-speed measurements (or measurements at or near the reading speed of the imaging sensor).

[0047] In some embodiments, optical throughput may be increased by inserting diffusers between the optical fibers (206) and the homogenizer (208). Multiple light sources (202) may be combined by other means such as dichromatic beam splitters, and the light sources (202) may be combined with the homogenizer (208) by other means such as free space optics relays.

[0048] In some embodiments, bandpass filters may be inserted between each of the light sources (202) and their individual optical fibers (206) to narrow the bandwidth of each wavelength. Narrower bandwidths may provide better sensitivity for measuring thick film stacks or dense patterns on the surface of a sample. Bandpass filters can accurately define measurement wavelengths by eliminating wavelength shifts of the LEDs to improve measurement accuracy.

[0049] An imaging sensor (e.g., the imaging sensor (158) illustrated in FIG. 1) can have a high reading rate (e.g., 50 to 1,000 frames per second (FPS) or more, and up to 100 million pixels per frame or more). As an example, at a reading rate of 100 FPS, the imaging sensor can perform 6,000 reflectance measurements per minute. The measurements can be taken at the same or different wavelengths. Acquiring multiple measurements at the same wavelength can improve the signal-to-noise ratio and improve measurement sensitivity.

[0050] FIG. 3 is a flowchart schematically describing a method for measuring the reflectance of a sample using an imaging reflectometer. This method includes sequentially generating a plurality of input beams at a first switching rate (302). In some embodiments, each of the plurality of input beams is generated by a different light source, and at least some of the plurality of input beams may have nominal wavelengths and / or integration times different from the others of the plurality of input beams. In other embodiments, at least some of the plurality of input beams are generated by a broadband light source, and the wavelength of each of the plurality of input beams is defined using a set of bandpass filters.

[0051] Each of the plurality of input beams is guided through an illumination pupil having a first NA (304). The illumination pupil may be arranged along a first optical path. In some embodiments, each of the plurality of input beams may be divided, and a first portion of each of the plurality of input beams may be guided toward a reference sensor along the first optical path, and a second portion of each of the plurality of input beams may continue along the first optical path.

[0052] At least a portion of each of the plurality of input beams is provided as substantially telecentric illumination over the sample being imaged using a lens (306). A portion of each of the plurality of input beams may also be provided to a reference sensor to monitor the uniformity and stability of the input beams. In some embodiments, the measurement field of the lens may be larger than the imaged sample to provide whole-sample measurements.

[0053] The reflected portions of substantially telecentric illumination reflected from the sample are received at the lens and guided through an imaging pupil having a second NA lower than the first NA of the illumination pupil (308). The reflected portions can be guided through the imaging pupil using a beam splitter.

[0054] Reflected parts are received, and corresponding image information is generated in an imaging sensor module, wherein the image information is generated at a frame rate equal to or faster than the first switching rate (310). The image information can be corrected or normalized based on information from a reference sensor.

[0055] In some embodiments, images may be processed to identify process excursions. Images may be processed according to known anomaly identification techniques. For example, reflectances measured at multiple wavelengths may be compared with modeled reflectances or known good samples. Measured patterns may also be compared at different locations across the sample to identify variation and / or outliers. Measured variation may be quantified by calculating the root-mean-squared (RMS) difference at multiple wavelengths. Measurement sensitivity may be enhanced by selecting the wavelength or wavelengths with the highest sensitivity based on the measurement data. Multiple wavelength reflectances may be processed by non-linear regression to a theoretical model to derive the CD of the film thickness and / or pattern.

[0056] It should be recognized that the imaging reflectometers described herein may be configured as standalone metrology tools or integrated with other metrology or process tools. As an example, an imaging reflectometer as described herein may be integrated with a process tool and positioned outside a window separating the process chamber from the imaging reflectometer. In some embodiments, a lens positioned outside the window provides illumination for a sample positioned inside the process chamber. The lens may be configured to provide illumination for all or part of the sample (e.g., the measurement area may be approximately the same size as the die or stepper field, or may be of a different size, such as several centimeters on the side). This enables reflection measurements to be performed during and / or immediately after processing while the samples are inside the vacuum chamber. This can shorten control loops, improve process control, and prevent material damage caused by air in the environment.

[0057] Imaging reflectometers may utilize a spectral imaging configuration, wherein fields of various sizes are imaged sequentially at different illumination wavelengths. The integration time per field may vary depending on the sample reflectance at such wavelengths. Images acquired at each wavelength may be stored in the imaging reflectometer or in a remote storage location.

[0058] The imaging path includes a pupil or aperture in the aperture plane. The aperture plane is located along the imaging path in the area through which substantially all light reflected from the sample passes. This is illustrated in FIG. 4a, in which reflected light (442) at points on the sample (432) within the measurement area is focused by a lens (440) and passes through the aperture (452) of the aperture plane (450). After passing through the aperture (452), the light (442) is guided by a lens (444) to an imaging sensor (458).

[0059] Substantially all reflected light (442) passes through the aperture (452) during normal operation of the imaging reflector. Normal operation is when the surface of the sample (432) is substantially orthogonal to the optical axis. However, when the sample includes a region of local inclination, the position of the light returning from the inclination point is shifted. This is better understood by referring to FIG. 4b. As illustrated in this example, the sample (432) includes an inclination region (456) that causes the shift of the reflected light (442a) in the inclination region (456). The reflected light (442) in other regions of the sample (432) is not shifted. As illustrated in this example, the reflected light (442a) in the inclination region (456) is shifted at the position where the aperture plane is located. The aperture plane illustrated in FIG. 4a is not illustrated in FIG. 4b to more clearly illustrate the shift of the reflected light (442a) in the inclination region (456).

[0060] Under normal conditions, the imaging system is designed to withstand this type of movement. As can be seen in FIG. 4a, as long as the optical field is not blocked at the aperture plane (450), the reflected light (442) from the sample (432) passes through the unobstructed aperture (452) to the intended image point on the imaging sensor (458). In the scenario of FIG. 4b, the only effect of the inclined region (456) is that the reflected light (442a) reaches the corresponding image point at an angle that is usually imperceptible because the angle of the inclined region (456) is generally very small.

[0061] In some embodiments, the aperture (452) is a variable aperture or iris. Some imaging reflection measurements may be performed with the aperture (452) open so that the reflected light (442) is not blocked at the aperture plane (450). Other imaging reflection measurements may be performed with a reduced aperture (452) to contain only the physical magnitude of the reflected light under normal conditions. In this configuration, any movement of the reflected light (442) will result in a suitable change in the amount of energy passing through the aperture (452) toward the imaging sensor (458). This can be illustrated with reference to FIG. 4c, in which a portion of the moved reflected light (442a) is blocked at the aperture plane (450) (or aperture plate). In this simplified example, the shifted reflected light (442a) is depicted as passing through the aperture plane (450), but in actual operation, the aperture plane (450) blocks a portion of the reflected light (442a) so that less light reaches the imaging sensor (458). This will cause a change in the signal level detected in the inclined area (456) on the sample (432).

[0062] The magnitude of the signal change can be estimated. When the sample (432) is in the nominal focal plane of the objective lens, the tilt θ of the sample results in a lateral shift of Fθ, where F is the focal length of the objective lens.

[0063] The position of the reflected light (442a) in the inclined area (456) is moved in the aperture plane (450) regardless of the position of the inclined area (456) on the sample (432). The reflected light in all inclined areas within the field of view is moved. An aperture (452) of the same size in the aperture plane (450) that causes an intensity change due to the inclined area at one point on the sample (432) causes an intensity change for all inclined areas within the measurement area (or field of view). That is, in addition to providing reflectance intensity values, the image can provide information about the local terrain across the measurement area.

[0064] For any provided point on the sample (432) and for any provided wavelength of light, the image of the measurement area on the sample (432) can be represented as K(λ) * M(λ) * I(λ) * R(λ), where I is the illumination intensity at the provided wavelength (λ) and R is the reflectance at the wavelength (λ). Here, M represents the percentage of light passing through the aperture (452), and K is a proportionality constant. This representation is reasonable for a properly designed telecentric lens because the reflected light from all points on the sample will fill the aperture (452) substantially uniformly, and accordingly, the positional shift of the light distribution in the aperture plane (450) is represented as a multiplier of the amount of energy.

[0065] According to some embodiments, both reflectance and gradient information may be obtained from imaging reflection measurements. Reflectance information is obtained by performing measurements with the aperture or pupil of the iris wide open so that the iris does not block light reflected from the sample. Gradient information is obtained by (i) performing measurements with the aperture or pupil of the iris wide open, and (ii) performing measurements with the aperture or pupil size reduced. The aperture or pupil size is reduced so that light shifted by the gradient regions is blocked at the iris (or aperture plane). All measurements may be performed at a single wavelength or multiple wavelengths to obtain wavelength sensitivity information.

[0066] The image obtained in step (i) provides reflectance intensity values. By normalizing the image obtained in step (ii) to the image obtained in step (i) at the point level, the parameter (M) corresponding to the slope can be extracted at the point level.

[0067] The dependence of the detected signal on the slope is not a direct measurement of the local terrain, but rather a derivative of the terrain (i.e., local slope) of the sample (432) along a specific direction. For example, the local slope on the sample (432) along the x-direction causes the reflected light (442a) in the aperture plane (452) to move in the x-direction. Similarly, the local slope on the sample (432) along the y-direction causes the reflected light (442a) in the aperture plane (452) to move in the y-direction. The slope along other directions similarly causes the reflected light (442a) in the aperture plane (452) to move in individual directions.

[0068] FIG. 5 is a simplified diagram illustrating reflectance intensities measured at different points around a local slope region on a sample according to one embodiment. The description of the effect of the slope on the image has so far assumed an array of circular light distributions interrupted by a circular iris in the aperture plane. For the case of the sample's nominal orthogonality to the illumination light, and for the case of symmetric orientation of the light distribution with respect to the iris's pupil, the provided slope will cause the light passing through the iris's pupil to have reduced intensity compared to the light in neighboring flat regions on the sample. This reduction in the reflectance signal applies whether the slope has a positive or negative slope. Accordingly, the image from the symmetric protrusion or depression on the sample is a dipole having negative-going peaks as illustrated. In other words, an array having symmetric localization of the light distribution with respect to the iris pupil enables the detection of features, but it cannot determine whether the feature is a protrusion or a depression.

[0069] This is illustrated in FIG. 5, in which the reflected light from the flat regions on the sample, including the peak of the protrusion, substantially fills the iris's pupil in the aperture plane without being blocked at the aperture plane. The reflectance intensity measured at these points is a nominal value. In the inclined regions on the sample, specifically the inclined regions on each side of the protrusion, the reflected light is moved in the aperture plane. The portions of the beam outside the iris's pupil are blocked at the aperture plane, thereby reducing the reflectance intensity measured at these points. The right axis of the plot represents the surface topography, and the left axis represents the reflectance intensity. The circles on the plot represent the aperture in the aperture plane for corresponding points on the sample, and the parts of the circular shades represent the light in the aperture plane. In reality, all light passes through a single aperture in the aperture plane, but the aperture and the light at each point are shown separately in this figure to illustrate the movement of the reflected light in the inclined regions. Similar reflectance intensity values ​​will be obtained from depressions on a sample having inwardly inclined sidewalls of similar size.

[0070] In some applications, it may be advantageous to determine the characteristics of the features. For example, it may be advantageous to determine whether the slope is positive or negative across a local slope region. This allows the features to be identified as protrusions or depressions. According to one embodiment, the position of the pupil of the iris is shifted (or displaced) relative to the light distribution in the aperture plane depending on the direction being considered (e.g., displaced along the x-direction or y-direction). This allows the characteristics of the features to be determined. This arrangement is illustrated in FIG. 6, in which the pupil of the iris is shifted horizontally so that some light reflected from flat regions on the sample (including light reflected from the peaks of the protrusions) is blocked in the aperture plane. This is illustrated by the circular and circular shaded areas on the plot. Because some of the light is blocked, the measured reflectance intensity from these flat regions is below the maximum. Additionally, in the case of a patterned sample, this arrangement slightly reduces the magnitude of the detected spatial frequencies emanating from the sample due to scattering. In some embodiments, the movement of the iris pupil may be small so as to minimize the effect on image brightness.

[0071] As illustrated in FIG. 6, light reflected from sloped regions (or sides of the protrusion) is shifted relative to light reflected from flat regions. For sloped regions with a positive slope (on the left side of the protrusion), the reflected light is shifted so that more light is blocked at the aperture plane compared to flat regions. For sloped regions with a negative slope (on the right side of the protrusion), the reflected light is shifted so that less light is blocked at the aperture plane compared to flat regions. Accordingly, in this example, reflectance intensity values ​​increase and decrease differently for positive and negative slopes. Using this arrangement, the image of the feature is a dipole, and its polarity (i.e., negative versus positive, or positive versus negative) determines the characteristics of the feature (i.e., protrusion or depression). In some embodiments, the amount of the iris pupil shifted is determined by the expected maximum slope on the sample such that the polarity is preserved for all expected features on the sample.

[0072] The example in Fig. 6 illustrates that for an array in which there is a linear offset between the center of the iris pupil and the light distribution along a specific direction (e.g., along the x-direction), the maximum sensitivity to the gradient also follows that direction. Such an array will be less sensitive to the gradient in the orthogonal direction (e.g., along the y-direction), but some intensity variation from the gradient in the orthogonal direction may still exist.

[0073] In some applications, it may be advantageous to determine the local slope in a provided direction (e.g., x-direction) while further minimizing the influence of the local slope along the orthogonal direction (e.g., y-direction). According to some embodiments, this is provided using a knife edge array as illustrated in FIGS. 7a and 7b. In the first example illustrated in FIG. 7a, a one-dimensional barrier or knife edge is arranged to block a portion of the reflected light in the aperture plane. Two circles and a double arrow indicate the movement of the reflected light in the aperture plane from features having a slope along the x-axis. The knife edge blocks different amounts of reflected light depending on the movement of the reflected light. Accordingly, the array in FIG. 7a is sensitive to features having a slope along the x-axis.

[0074] In Fig. 7b, the knife edge is aligned with respect to the aperture plane in the same manner as in Fig. 7a. In Fig. 7b, the two circles and the double arrow represent the movement of reflected light in the aperture plane from features that have a slope along the y-axis. In this case, the knife edge blocks the same amount of reflected light regardless of the movement of the reflected light. Accordingly, the arrangement in Fig. 7b is not sensitive to features that have a slope along the y-axis. The measured intensity will be the same regardless of the movement of the reflected light.

[0075] Knife edges can be arranged in a similar manner along any direction to detect slanted features along that direction. As an example, knife edges can be arranged horizontally (rather than vertically as shown in FIGS. 7a and 7b) to detect local slanted regions along the y-direction.

[0076] FIG. 8 is a simplified diagram illustrating an illumination-based implementation according to one embodiment. In this arrangement, the illumination pupil may be partially blocked so that light from only a portion of the pupil illuminates the sample. In a suitably designed imaging system, the illumination pupil is imaged onto the imaging pupil. If the sample has no local slope regions, it is concluded that the image of the illumination pupil is a partially blocked image. However, if the sample has local slope regions, the image is shifted laterally, thereby causing an increase or decrease in the amount of light passing through the imaging pupil. Accordingly, the system can distinguish between positive and negative slopes.

[0077] FIGS. 9a and 9b illustrate a dark field implementation according to one embodiment. In this arrangement, a portion of the illumination NA is blocked at the imaging pupil (the black portions are blocked). For flat areas on the sample, all or nearly all of the reflected light is blocked at the imaging pupil (or at the aperture plane). For slanted areas on the sample, the reflected light will be shifted, so that at least a portion will not be blocked at the imaging pupil, and accordingly, in this arrangement, the slanted areas will have an increase in the measured reflectance intensity.

[0078] FIGS. 10a and 10b illustrate the measurement of a two-dimensional surface profile having a one-dimensional pupil pattern according to one embodiment. The pattern used in these figures is merely an example, and other patterns may be used in other embodiments. As illustrated in FIGS. 10a and 10b, the pattern in the illuminated pupil may be opposite to the pattern in the imaging pupil for a dark-field implementation. The pattern may provide sensitivity to local tilts extending along a specific direction. For example, the pattern used in FIGS. 10a and 10b is sensitive to local tilts extending along the x-axis but not sensitive to local tilts extending along the y-axis (light shifted along the y-axis is not detected in the dark-field implementation). According to one embodiment, the patterns are rotatable so that local tilt regions extending in different directions can be detected separately.

[0079] In some embodiments, patterns may also be implemented using an LED array for illumination and / or an LCD array for detection. Patterns for illumination and imaging may be generated by electronic control and / or mechanical patterns.

[0080] FIG. 11 is a flowchart illustrating a method for detecting local tilt regions on a sample using an imaging reflectometer according to one embodiment. The method includes illuminating a measurement area on a sample using an input beam (1002).

[0081] The present method also includes performing a first imaging reflection measurement (1004). Performing the first imaging reflection measurement may include receiving a first imaging beam reflected from a sample at an imaging sensor. An iris positioned at the aperture plane of the imaging reflection system may be configured such that the first imaging beam passes through the pupil of the iris and the iris does not block any portion of the first beam reflected from the sample. A first image of the measurement area may be acquired using the first imaging beam reflected from the sample and received at the imaging sensor. In some embodiments, the size of the pupil of the iris may be variable.

[0082] The present method also includes performing a second imaging reflection measurement (1006). Performing the second imaging reflection measurement may include receiving a second imaging beam reflected from a sample at an imaging sensor. An iris positioned at the aperture plane of the imaging reflection system may be configured so that the second imaging beam passes through the pupil of the iris and the iris blocks a portion of the second beam reflected from the sample. A second image of the measurement area may be acquired using the second imaging beam reflected from the sample and received at the imaging sensor. In some embodiments, the diameter of the pupil of the iris may be smaller during the second imaging reflection measurement than during the first imaging reflection measurement.

[0083] The method also includes detecting local slope regions within a measurement area (1008). Detecting local slope regions may include comparing first reflectance intensity values ​​of pixels in a first image with second reflectance intensity values ​​of corresponding pixels in a second image. In some embodiments, detecting local slope regions may include identifying pixels in the first image, wherein the first reflectance intensity values ​​differ from the second reflectance intensity values ​​of corresponding pixels in the second image by a predetermined amount. In other embodiments, detecting local slope regions may include comparing, for each of the first plurality of pixels and each of the second plurality of pixels, the first reflectance intensity value of a pixel in the first image with the second reflectance intensity value of a corresponding pixel in the second image.

[0084] Comparing the first reflectance intensity values ​​of pixels in the first image with the second reflectance intensity values ​​of corresponding pixels in the second image may include dividing the second reflectance intensity values ​​of pixels in the second image by the first reflectance intensity values ​​of corresponding pixels in the first image.

[0085] The height difference between local sloped regions and surrounding flat regions on the sample can be determined based on the difference in reflectance intensity values ​​between pixels in the first image and corresponding pixels in the second image.

[0086] The method of FIG. 11 may also include determining the tilt direction associated with local tilt regions by arranging knife edges to block a first portion of a first imaging beam reflected from a sample and a second portion of a second imaging beam reflected from a sample. Regions within the local tilt regions can be identified when the tilt direction is orthogonal to the knife edge based on the change in reflectance intensity values ​​between the first image and the second image.

[0087] According to some embodiments, methods for detecting local slope regions on a sample using an imaging reflectometer include performing first and second imaging reflection measurements using an array in which the pupil of the iris is laterally offset with respect to a first beam reflected from the sample. Because the pupil is offset, a portion of the imaging beam passes through the pupil of the iris, and a portion of the imaging beam is blocked at the iris. Using the offset pupil, regions of positive slope and regions of negative slope can be identified within the local slope regions.

[0088] FIG. 12 is a flowchart illustrating a method for detecting local tilt regions on a sample using an imaging reflectometer configured for dark-field measurements according to one embodiment. The method includes illuminating a measurement area on the sample using an input beam (1102).

[0089] The present method also includes receiving (1104) portions of an imaging beam reflected from local sloped regions on a sample in an imaging sensor. The portions of the imaging beam reflected from local sloped regions may pass through an imaging cavity configured to block portions of the imaging beam reflected from flat regions on the sample. In some embodiments, a portion of the input beam may pass through an illumination cavity having a first NA. The imaging cavity may have a second NA greater than the first NA.

[0090] The present method also includes acquiring a first image of a measurement area (1106) using portions of an imaging beam that are reflected from local tilted regions on the sample and received by an imaging sensor.

[0091] The present method also includes detecting local slope regions within a measurement area based on the locations of bright points within a first image (1108).

[0092] In some embodiments, a one-dimensional pattern may be provided in the illumination pupil, and a one-dimensional pattern may be provided in the imaging pupil. The one-dimensional pattern in the illumination pupil may be opposite to the one-dimensional pattern in the imaging pupil. The orientation of the one-dimensional pattern in the illumination pupil may be aligned with the orientation of the one-dimensional pattern in the imaging pupil. The tilt direction associated with local tilt regions may be determined by rotating the one-dimensional pattern in the illumination pupil and the one-dimensional pattern in the imaging pupil by about 90°, acquiring a second image of the measurement region using portions of the imaging beam reflected from the local tilt regions on the sample and received by the imaging sensor, and determining the tilt direction associated with the local tilt regions based on the change in reflectance intensity values ​​between the first image and the second image. In some embodiments, a one-dimensional pattern in the illumination pupil and / or a one-dimensional pattern in the imaging pupil may be generated by electronic control, and rotation of the one-dimensional pattern in the illumination pupil and / or a one-dimensional pattern in the imaging pupil may be provided by electronically generating a new one-dimensional pattern rather than rotating a mechanical pattern generator.

[0093] It should be recognized that the specific steps illustrated in FIGS. 11 and 12 provide specific methods for measuring reflectance according to some embodiments. Different orders of the steps may also be performed according to alternative embodiments. For example, alternative embodiments may perform the steps outlined above in a different order. Additionally, the individual steps illustrated in FIGS. 10 and 11 may include a number of sub-steps that can be performed in various orders. Additionally, additional steps may be added or removed depending on the specific application.

[0094] Although the foregoing relates to specific embodiments, other and additional embodiments may be devised without departing from the basic scope thereof, the scope thereof being determined by the following claims.

Claims

Claim 1 A method for detecting local tilt regions on a sample using an imaging reflectometer configured for dark field measurements, comprising: a step of illuminating a measurement region on the sample using an input beam, wherein a portion of the input beam passes through an illumination pupil; a step of receiving portions of the imaging beam reflected from local tilt regions on the sample at an imaging sensor, wherein portions of the imaging beam reflected from the local tilt regions pass through an imaging pupil, and the imaging pupil is configured to block portions of the imaging beam reflected from flat regions on the sample; a step of acquiring a first image of the measurement region using portions of the imaging beam reflected from the local tilt regions on the sample and received at the imaging sensor; and a step of detecting local tilt regions within the measurement region based on the locations of bright points in the first image. Claim 2 A method according to claim 1, wherein the illumination pupil has a first numerical aperture (NA) and the imaging pupil has a second NA that is larger than the first NA. Claim 3 The method of claim 1, wherein a one-dimensional pattern is provided to the illumination pupil, a one-dimensional pattern is provided to the imaging pupil, the one-dimensional pattern in the illumination pupil is inverse of the one-dimensional pattern in the imaging pupil, and the orientation of the one-dimensional pattern in the illumination pupil is aligned with the orientation of the one-dimensional pattern in the imaging pupil, and the method further comprises determining a tilt direction associated with the local tilt regions, wherein determining the tilt direction comprises rotating the one-dimensional pattern in the illumination pupil and the one-dimensional pattern in the imaging pupil by 90°; acquiring a second image of the measurement region using portions of the imaging beam reflected from the local tilt regions on the sample and received by the imaging sensor; and determining the tilt direction associated with the local tilt regions based on the change in reflectance intensity values ​​between the first image and the second image. Claim 4 A method according to claim 3, wherein the one-dimensional pattern in the illumination pupil and / or the one-dimensional pattern in the imaging pupil is generated by electronic control, and the rotation of the one-dimensional pattern in the illumination pupil and / or the one-dimensional pattern in the imaging pupil is provided by electronically generating a new one-dimensional pattern rather than by rotating a mechanical pattern generator. Claim 5 A method for detecting local tilt regions on a sample using an imaging reflectometer, comprising: a step of illuminating a measurement area on the sample using an input beam; a step of performing a first imaging reflectance measurement, wherein the step of performing the first imaging reflectance measurement comprises receiving a first imaging beam reflected from the sample at an imaging sensor, wherein the pupil of an iris oriented at the aperture plane of the imaging reflectometer is laterally offset with respect to the first beam reflected from the sample such that a portion of the first imaging beam passes through the pupil of the iris and a portion of the first imaging beam is blocked at the iris; A method comprising: a step of acquiring a first image of the measurement area using a portion of the first imaging beam received by an imaging sensor passing through the pupil of the iris; a step of performing a second imaging reflection measurement, wherein the step of performing the second imaging reflection measurement includes receiving a second imaging beam reflected from the sample at an imaging sensor, wherein the pupil of the iris, which is oriented on the aperture plane of the imaging reflection system, is laterally offset with respect to the second beam reflected from the sample such that a portion of the second imaging beam passes through the pupil of the iris and a portion of the second imaging beam is blocked at the iris; and a step of acquiring a second image of the measurement area using a portion of the second imaging beam received by an imaging sensor passing through the pupil of the iris; and a step of detecting local tilt areas within the measurement area, wherein the step of detecting local tilt areas includes comparing first reflectance intensity values ​​of pixels in the first image with second reflectance intensity values ​​of corresponding pixels in the second image. Claim 6 A method according to claim 5, further comprising identifying positive slope regions and negative slope regions within the local slope regions based on a comparison of first reflectance intensity values ​​of pixels in the first image and second reflectance intensity values ​​of corresponding pixels in the second image. Claim 7 A method according to claim 5, wherein the step of detecting the local slope regions further comprises identifying pixels in the first image in which the first reflectance intensity values ​​differ from the second reflectance intensity values ​​of corresponding pixels in the second image by more than a predetermined amount. Claim 8 A method according to claim 5, wherein comparing the first reflectance intensity values ​​of pixels in the first image with the second reflectance intensity values ​​of corresponding pixels in the second image comprises dividing the first reflectance intensity values ​​of pixels in the first image by the second reflectance intensity values ​​of corresponding pixels in the second image. Claim 9 A method according to claim 5, wherein the step of detecting the local gradient regions comprises comparing a first reflectance intensity value of a pixel in the first image with a second reflectance intensity value of a corresponding pixel in the second image for each of the first plurality of pixels in the first image and each of the second plurality of pixels in the second image. Claim 10 A method for detecting local slope regions on a sample using an imaging reflectometer, comprising: illuminating a measurement region on the sample using an input beam; receiving an imaging beam reflected from the sample passing through an imaging pupil at an imaging sensor; acquiring a first image of the measurement region using the imaging beam reflected from the sample and received at the imaging sensor; illuminating a measurement region on the sample using the input beam, wherein a portion of the input beam is blocked at an illumination pupil; receiving an imaging beam reflected from the sample passing through an imaging pupil at an imaging sensor; acquiring a second image of the measurement region using the imaging beam reflected from the sample and received at the imaging sensor; detecting the local slope regions within the measurement region; and determining the slope of the local slope regions based on a change in the amount of luminance of pixels in the second image compared with the luminance of corresponding pixels in the first image. Claim 11 In paragraph 10, a method in which half of the input beam is blocked at the illumination pupil. Claim 12 A method according to claim 10, wherein local slope regions within the measurement area of ​​the second image, having positive and negative slopes, exist as regions of greater or lesser luminance than the first image, or vice versa. Claim 13 delete Claim 14 delete Claim 15 delete Claim 16 delete Claim 17 delete Claim 18 delete Claim 19 delete Claim 20 delete

Images