Polarization Modification of the Sample Surface in Interference Defect Inspection

Abstract

Defects are detected using data acquired from an interference channel and a polarization modification channel in an interferometer. An interference objective lens splits a polarized illumination beam into a reference illumination and a sample beam, the reference illumination being reflected by a reference surface without modifying polarization and the sample beam being reflected by a sample surface with possible polarization modification. Light from the sample beam having no polarization change is combined with the reference illumination and directed to the interference channel, which can measure the reflectivity and / or topography of the sample. Light from the sample beam having modified polarization is directed to the polarization modification channel. The intensity of the light detected at the polarization modification channel can be used together with the reflectivity and topography data to identify defects or other characteristics of the sample.

Description

[0001] Cross - reference to related patent applications

[0002] This application claims priority under 35 USC 119 to U.S. Provisional Application No. 62 / 906,055, filed on September 25, 2019, entitled "SAMPLE SURFACE POLARIZATION MODIFICATION IN INTERFEROMETRIC DEFECT INSPECTION", and to U.S. Non - Provisional Application No. 16 / 927,571, filed on July 13, 2020, entitled "SAMPLE SURFACE POLARIZATION MODIFICATION IN INTERFEROMETRIC DEFECT INSPECTION", both of which are incorporated herein by reference in their entireties. Technical Field

[0003] The present invention relates to the optical inspection of samples (such as semiconductor wafers) using surface topography or reflectivity, and the detection of polarization modification. Background Art

[0004] Semiconductor and other similar industries commonly use optical tools for non - contact evaluation of samples during the manufacturing process. One type of evaluation is defect inspection. Defects (e.g., particles or other irregularities on the sample) can interfere with the performance of the finished device. Conventionally, optical tools for defect detection use bright - field detection and dark - field detection. Bright - field detection and dark - field detection tools detect defects based on the scattering of light caused by the defects. Defect detection requires obtaining data from a large portion of the sample and typically from the entire sample. As the size of semiconductor devices continues to shrink and the designs become more complex, it becomes increasingly difficult to detect smaller defects with conventional optical inspection tools (e.g., bright - field detection tools and dark - field detection tools). Summary of the Invention

[0005] Defects are detected using data collected from an interferometric channel and a channel that detects polarization modification in a detection interferometer. An interferometric objective lens splits a polarized illumination beam into a reference illumination and a sample beam. The reference illumination is reflected by a reference surface without modifying the polarization, and the sample beam is reflected by the sample surface and may modify the polarization. Light from the sample beam with no polarization change is combined with the reference illumination and directed to the interferometric channel, which can measure the reflectivity and / or topography of the sample. Light from the sample beam with modified polarization is directed to the polarization - modification channel. The intensity of the light detected at the polarization - modification channel can be used together with the reflectivity and topography data of the sample to identify defects on the sample.

[0006] In one embodiment, an optical inspection device may be configured to detect defects on a sample. The optical inspection device includes an interference objective lens that includes a polarization beam splitter and a reference mirror. The polarization beam splitter may be configured to receive a polarized illumination beam and direct a first portion of the illumination beam toward the reference mirror as reference illumination. The polarization beam splitter may further direct a second portion of the illumination beam toward the sample as sample illumination. The reference mirror may be configured to reflect the first portion of the illumination beam to produce reflected reference illumination. The polarization beam splitter may receive the reflected reference illumination from the reference mirror and reflected sample illumination from the sample, and combine a first portion of the reflected reference illumination and the reflected sample illumination into an interference beam. The interference beam is directed along an interference channel. A second portion of the reflected sample illumination is directed along a polarization modification channel that is different from the interference channel. The optical inspection device further includes: a first detector in the interference channel, which is configured to receive the interference beam from the sample and generate a first set of optical data from the sample; and a second detector in the polarization modification channel, which is configured to receive the second portion of the reflected sample illumination and generate a second set of optical data indicative of the polarization modification of the sample illumination. At least one processor coupled to the first detector and the second detector may be configured to use the first set of optical data and the polarization modification from the second set of optical data to detect defects on the sample.

[0007] In one embodiment, a method of detecting defects on a sample may include directing a first portion of a polarized illumination beam toward a reference mirror as reference illumination. The reference mirror reflects the first portion of the illumination beam to produce reflected reference illumination. The method may further include directing a second portion of the illumination beam toward the sample as sample illumination, and the sample reflects the second portion of the illumination beam to produce reflected sample illumination, wherein an area of the sample partially modifies the polarization of the sample illumination. The reflected reference illumination and a first portion of the reflected sample illumination from the area of the sample are combined into an interference beam that is directed along an interference channel based on polarization. A second portion of the reflected sample illumination from the area of the sample is directed along a polarization modification channel based on polarization, wherein the polarization modification channel is different from the interference channel. Detecting the interference beam in the interference channel to generate a first set of optical data, and detecting the second portion of the reflected sample illumination in the polarization modification channel to generate a second set of optical data indicative of the polarization modification of the sample illumination. Using the first set of optical data and the polarization modification from the second set of optical data to detect defects on the sample.

[0008] In one embodiment, an optical inspection device may be configured to detect defects on a sample. The optical inspection device may include means for directing a first portion of a polarized illumination beam towards a reference mirror as reference illumination and directing a second portion of the illumination beam towards the sample as sample illumination. The reference mirror reflects the first portion of the illumination beam to produce reflected reference illumination, and the sample reflects and partially modifies the polarization of the second portion of the illumination beam to produce reflected sample illumination. The optical inspection device may include means for combining the reflected reference illumination and a first portion of the reflected sample illumination into an interference beam and directing the interference beam along an interference channel. The optical inspection device may include means for directing a second portion of the reflected sample illumination along a polarization modification channel that is different from the interference channel. The optical inspection device may include means for detecting the interference beam in the interference channel to produce a first set of optical data, and means for detecting the second portion of the reflected sample illumination in the polarization modification channel to produce a second set of optical data indicative of the polarization modification of the sample illumination. The optical inspection device may include means for detecting defects on the sample using the first set of optical data and the polarization modification from the second set of optical data. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 Schematic diagram showing a phase-shifting interferometer having a polarization modification channel for detecting modification of polarized incident light by a sample.

[0010] Figure 2A and Figure 2B Side perspective view and top plan view showing a phase mask.

[0011] Figure 2C Schematic diagram showing a unit portion of a phase mask, the unit portion including a 2x2 polarizer pixel array having four discrete polarizations that repeat across the phase mask.

[0012] Figure 3 Schematic diagram showing another phase-shifting interferometer having a polarization modification channel for detecting modification of polarized incident light by a sample.

[0013] Figure 4 Schematic diagram showing a Mirau objective.

[0014] Figure 5 Schematic diagram showing a scanning interferometer having a polarization modification channel for detecting modification of polarized incident light by a sample.

[0015] Figure 6 and Figure 7 are simplified diagrams respectively showing the evolution of polarization in the sample optical path and the reference mirror optical path of an electric field system model for a polarization modification channel.

[0016] Figure 8Shows an interference image received by a detector in an interference channel and a polarization-modified image received by a detector in a polarization modification channel.

[0017] Figure 9 Is a flowchart showing a characteristic determination process using interference data and polarization mixing data.

[0018] Figure 10 Is a flowchart showing a defect determination process that can be performed by an optical inspection tool. Detailed Description

[0019] An interferometer is an optical tool commonly used in optical metrology to measure the height difference, i.e., surface topography, on the surface of a sample. An interferometer can measure the height difference on an object by determining the phase of the interference signal at each pixel. Determining the phase of the signal requires obtaining multiple measurement results from each point on the sample. A scanning interferometer can use polarized light and physically move (scan) the sample or a reference surface along the optical axis, which is functionally perpendicular to the surface of the sample, to modify the signal phase. A scanning interferometer typically scans in steps that produce a quarter-wavelength phase change. By processing multiple measurement results at different phases from each point on the sample, the vertical height (Z) of the surface at each point on the sample can be determined, and the heights of each point can be combined to determine the topography of the sample.

[0020] A phase-shifting interferometer uses polarized light and a phase mask with a pixel array that produces multiple phase shifts, such that interferometric data with multiple phase shifts can be acquired using a single exposure. Thus, unlike a scanning interferometer, the acquisition time for each measurement is only limited by the time for movement, focusing, and performing pattern recognition, as well as the image transfer rate, and is not a function of the scanning time. Additionally, since a phase-shifting interferometer obtains data with a single exposure, the effects of vibrations on all axes are reduced, especially those at low frequencies.

[0021] Semiconductor wafer patterns include features that act as polarizers. For example, the main pattern in a semiconductor memory device is a series of parallel lines that change the polarization of incident light. Other types of semiconductor devices similarly include patterns with parallel lines or other structures that change the polarization of incident light.

[0022] If a sample that modifies the polarization of reflected light is measured using an interferometer that uses polarized light (e.g., a phase-shift interferometer or a scanning interferometer that includes a polarizer), the reflectivity (e.g., the intensity of the detected light) will decrease when the incident polarization vector and any surface polarization vectors are misaligned. Although circularly polarized light can be used to reduce the effect of surface polarization of the sample, the reflected intensity of a surface that acts as a perfect linear polarizer will still be reduced by 50% compared to an unpolarized surface. The change in the polarization orientation of the incident light by the sample surface causes the reflected light to have a combination of different polarization orientations that are different from the polarization of the incident light, which is referred to herein as polarization modification.

[0023] As discussed herein, an interferometer that uses polarized light can detect the magnitude of polarization modification at the sample surface, thereby enabling the attribution of changes in image intensity to the correct cause, such as changes in the reflectivity of the sample surface or polarization modification caused by the sample surface. Additionally, the detection of polarization modification at the sample surface provides an additional channel of data that can be used to determine desired characteristics of the sample, including defect detection.

[0024] Figure 1 A schematic diagram of a phase-shift interferometer 100 is shown that is capable of measuring the surface height (Z) (e.g., topography) of at least a portion of a sample and can identify changes in reflectivity or polarization modification at the sample surface. For example, the sample can be a semiconductor wafer, a flat substrate, or other types of samples. As described herein, the sample surface topography and the identification of changes in reflectivity or polarization modification can be used to evaluate the surface of the sample, such as for defect inspection.

[0025] The interferometer 100 includes an interference channel 150 that detects the interference intensity between light reflected from the sample 140 and light reflected from a reference mirror 138. Data from the interference channel 150 can be used to determine the reflectivity of the sample and / or the topology of the sample 140. The interferometer 100 further includes a polarization modification channel 120 that detects the occurrence and magnitude of polarization modification caused by the sample 140.

[0026] It should be understood that although Figure 1 the polarization modification channel 120 in the phase-shift interferometer 100 is shown, the use of the polarization modification channel 120 is not limited thereto. For example, one of ordinary skill in the art can incorporate the use of the polarization modification channel 120 in other types of interferometers (e.g., a scanning interferometer as described in Figure 5 ).

[0027] As shown, interferometer 100 includes a light source 110 for generating light 112. The light source 110 can be a narrowband light source that generates light of a desired wavelength (e.g., about 450 nm). For example, the light source 110 can be an LED, a laser, or an incandescent light source such as a tungsten lamp, or a plasma source or an arc lamp, or any other suitable high-brightness light source. One or more appropriate filters can be used in combination with a broadband light source to generate light of one or more desired wavelengths. By way of example, a light source (such as an LED) having a full width at half maximum (FWHM) bandwidth of 20 nm can be used. One or more condenser lenses 114 and a field stop 115 can be used to collect light 112 from the light source 110. If desired, Kohler illumination, critical illumination, or other intermediate forms of illumination or other distributions, such as annular distributions, can be used. For example, in Kohler illumination, an aperture stop (not shown) is imaged by an illumination lens 116 in the rear focal plane of the objective in the interference objective 130, and the field stop 115 is imaged on the sample 140. The light 112 passes through a polarizer, and a beam splitter 118 is used to direct a portion of the light 112 toward the interference objective 130. The beam splitter 118 can be a non-polarizing beam splitter. The light 112 passes through a polarizer 119, which is a linear polarizer and can have an adjustable variable orientation to maximize fringe contrast. If desired, the light source 110 can generate polarized light, which can obviate the need for the polarizer 119.

[0028] For example, Figure 1 The interference objective 130 in Figure 1 is shown in Linnik geometry, but other interference objectives, such as a Michelson objective or a Mirau objective, can be used if desired. The interference objective 130 is configured to split the incident polarized light 112 into sample illumination 135 reflected from the sample 140 and reference illumination 139 reflected from a reference surface 138, and recombine the reflected sample illumination from the sample with the reflected reference illumination from the reference surface. The interference objective 130 can include a beam splitter 132, a sample objective 134 for imaging the surface of the test sample, and a reference mirror 138. In some embodiments, such as in Figure 1As shown, in the Linnik objective geometry, the interference objective 130 may further include a reference objective 136 complementary to the sample objective 134. In the Michelson objective geometry, the reference objective 136 may be removed from the reference path. The beam splitter 132 is a polarization beam splitter that transmits one polarization component and reflects the orthogonal polarization component. The beam splitter 132 separates the incident light 112 between the sample 140 and the reference mirror 138, wherein the reference illumination 139 directed towards the reference surface 138 is orthogonally polarized with respect to the sample illumination 135 directed towards the sample 140. For example, the beam splitter 132 transmits light linearly polarized in the plane of the angled surface of the beam splitter 132 and reflects light with orthogonal polarization. Any form of polarization beam splitter may be used; examples with good efficiency include those with wire grid polarization elements, or MacNeille cubes with appropriate thin film coatings on the internal angled surfaces of the beam splitter cube. The characteristics of the beam splitter 132 may be matched to the bandwidth of the light source 110, since the variation of polarization efficiency with wavelength will change the balance of light in the reflected and transmitted beams or mix the polarization states of this light. Additionally, the polarizer 119 may be set at an angle of 45° with respect to the axis of the polarization beam splitter 132, or some other angle if needed, to balance the reflectivities of the sample 140 and the reference mirror 138.

[0029] Figure 1 An actuator 137 is shown, which may optionally be attached to the reference objective 136 to move the reference objective 136 along the optical axis to change the optical path difference between the sample illumination 135 incident on the sample 140 and the reference illumination 139 incident on the reference mirror 138. The actuator 137, if included, may be used for example to optimize the contrast of the interference fringes at the measurement position and may be used to scan the reference surface to acquire multiple phase shifts at a specific position. In implementation, the entire optical assembly of the individual reference objective 136, the individual reference mirror 138, the sample 140, or the interference objective 130 may be moved along the optical axis to change the optical path difference between the sample illumination 135 and the reference illumination 139. However, it should be understood that with other interference objectives, the path difference may be changed in other ways, such as by moving the vertical position of the sample 140. From an optical perspective, there is no difference between moving the sample relative to each other or the entire imaging system; however, there is a practical significance in that the quality of the optical system may limit the choice of stage, which in turn may limit the lowest stage accuracy. It should be understood that the use of the actuator 137 does not eliminate the need for a focusing system, for example, by moving the sample relative to the optical assembly, or moving the optical assembly relative to the sample, or a combination of both.

[0030] As shown in the figure, the sample illumination 135 passes through the sample objective lens 134 and is incident on the sample 140, which is held on a chuck 142 mounted on a stage 144. The stage 144 can move horizontally in Cartesian (i.e., X and Y) coordinates or polar (i.e., R and θ) coordinates or some combination of both in order to correctly position each desired location on the sample relative to the interference objective lens 130 used for measurement. The stage can also move vertically along the z coordinate, for example, for focusing or for changing the path difference. The sample illumination 135 is reflected by the sample 140, and the reflected sample illumination passes through the sample objective lens 134. Similarly, in use, the reference illumination 139 passes through the reference objective lens 136 and is incident on the reference mirror 138. The reference illumination 139 is reflected by the reference mirror 138, and the reflected reference illumination passes through the reference objective lens 136.

[0031] Generally speaking, a surface will reflect incident light with a polarization state different from that of the incident light. The reference surface 138 is selected to minimize this effect. Thus, the reference surface 138 (e.g., a mirror) is non-polarizing and will reflect the reference illumination 139 without modifying the polarization state of the reference illumination 139. Therefore, the reflected reference illumination will be directed by the polarization beam splitter 132 in the same manner as the reference illumination 139 obtained from the polarization beam splitter 132. For example, as Figure 1 shown, the reference illumination 139 obtained from the polarization beam splitter 132 is reflected, and thus, the polarization beam splitter 132 will reflect the reflected reference illumination towards the interference channel 150 with high efficiency.

[0032] If the sample 140 is also non-polarizing, the sample illumination 135 incident on the sample 140 and reflected by the sample will behave in a manner similar to the reference illumination 139. In other words, the polarization state of the sample illumination 135 will not be modified by the sample 140, and after being reflected by the sample 140, the reflected sample illumination 135 will have the same polarization state as the incident sample illumination. In this case, the polarization beam splitter 132 will direct the reflected sample illumination in the same manner as the sample illumination 135 obtained from the polarization beam splitter 132. For example, as Figure 1 shown, the sample illumination 135 obtained from the polarization beam splitter 132 is transmitted, and thus, the polarization beam splitter 132 will transmit the reflected sample illumination towards the interference channel 150 with high efficiency, where the reflected sample illumination and the reflected reference illumination will interfere with each other.

[0033] The interference channel 150 receives the combined beam 151, which includes the reflected reference illumination 139 and a portion of the reflected sample illumination 135 that has a polarization that is not modified by the sample 140. The interference channel 150 includes, for example, one or more optical components 152, such as lenses and a detector 154 that receives the combined beam 151. A quarter-wave plate 153 in the interference channel 150 converts orthogonally polarized beams into circularly polarized light in opposite directions. For example, the p-polarized sample illumination from the sample 140 is converted into right-handed circular polarization, and the s-polarized reference illumination from the reference mirror 138 is converted into left-handed circular polarization. The optical components 152 image the combined light 151 onto the detector 154, which can be a camera. The detector 154 includes a phase mask 156 that has a pixel array that generates multiple phase shifts in front of a detector array 158, such as a CCD array. The pixel array is aligned with the pixel array in the phase mask 156 on a pixel-by-pixel basis and is located in substantially the same image plane. The combined beam 151 passes through the pixel array of the phase mask 156, thereby generating multiple (N) interleaved samples of an interference pattern on the detector array 158, where, for example, the phase difference between each sample has the same magnitude. Thus, the detector 154 receives an image of the unpolarized aspects of the sample combined with the reference signal, which produces an interferogram due to the small path difference. Different phase-shifting elements in the phase mask 156 generate interleaved images of multiple (N) samples, each sample having the same phase shift, i.e., N different samples of the interferogram exist according to the phase. Adjacent groups of samples (pixels) with different phase shifts can be processed together to obtain the local height. For example, the phase mask 156 can be a pixel array of linear polarizers arranged in groups of four, for example, with polarizer orientations of 0°, 45°, 90°, and 135°, which introduce a phase shift that is twice the reference polarizer orientation between the signal and the reference illumination. The phase mask 156 and the detector array 158 can be, for example, a Phasecam manufactured by Onto Innovation using a wire-grid polarizer array manufactured by Moxtek.

[0034] Figure 2A and 2BA side perspective view and a top plan view of the phase mask 156 are shown. FIG. 2c shows a unit portion of the phase mask 156, which includes a 2x2 polarizer pixel array 202, 204, 206, and 208 having four discrete polarizer orientations (0°, 45°, 90°, 135°), where the unit portion is repeated over the entire phase mask 156 such that the phase mask 156 includes a repeating pixel array having discrete polarizer orientations. The polarizer pixels 202, 204, 206, and 208 oriented at 0°, 45°, 90°, and 135° are capable of interfering with phase shifts of 0°, 90°, 180°, and 270° respectively between the sample illumination 135 and the reference illumination 139. The size and pitch of the pixels in the array of the phase mask 156 match the size and pitch of the pixels in the detector array 158 such that each pixel in the detector array 158 matches (i.e., is aligned with) the corresponding pixel of the phase mask 156.

[0035] Other arrangements of the detector 154 are possible. For example, a phase step size other than 90° may be used if desired. Additionally, it may be advantageous to use more than four phase shifts in the unit. Additionally, the arrangement of the pixels need not be in the Figure 2C order shown. Other types of arrangements of the pixels that may be used are further described in U.S. Patent Application No. 16 / 197,929, titled "Interferometer with Pixelated Phase Shift Mask," filed on November 21, 2018, which is incorporated herein by reference in its entirety.

[0036] The interference channel 150 detects the phase difference between the sample illumination 135 and the reference illumination 139, which are orthogonally polarized by the polarization beam splitter 132. The quarter-wave plates 153 convert the linearly polarized sample illumination 135 and reference illumination 139 into left-handed circular polarization and right-handed circular polarization, which interfere after passing through the phase mask 156. The detector array 158 receives the light resulting from the interference and detects the intensity at each pixel in the detector array 158.

[0037] As described above, generally, a surface will reflect incident light with a polarization state different from that of the incident light. Thus, returning to the reference Figure 1, the surface of the sample 140 may change the polarization state of the incident light. If one or more regions within the illumination spot on the surface of the sample 140 modify the polarization of the sample illumination 135, the reflected sample illumination 135 from these regions will have a mixture of polarization components. Some light may become depolarized, but for regular structures encountered, for example, on a semiconductor wafer, the main return signal will remain polarized. The mixture of polarization components in the reflected sample illumination will be the sum of a component having the same polarization state as the incident sample illumination 135 and another component orthogonal thereto. As discussed above, the polarization beam splitter 132 will direct the reflected sample illumination having a polarization component parallel to the incident sample illumination 135 towards the interference channel 150. For example, as Figure 1 shown, the portion of the reflected sample illumination having the same polarization component as the incident sample illumination will be transmitted towards the interference channel 150, where the portion of the reflected sample illumination and the reflected reference illumination will interfere with each other. After the test path signal and the reference path signal are combined, the modulation intensity of the interference signal detected in the interference channel 150 will be reduced by the amount of light lost due to the polarization modification caused by the sample 140.

[0038] The portion of the reflected sample illumination having a polarization component orthogonal to the incident sample illumination 135 will be directed by the polarization beam splitter 132 towards the polarization modification channel 120, as shown by the arrow 121. For example, as Figure 1 shown, the sample illumination 135 is obtained by transmission from the polarization beam splitter 132, and thus, the polarization beam splitter 132 will reflect with high efficiency towards the polarization modification channel 120 the portion of the reflected sample illumination having orthogonal polarization components. Of course, if the regions of the sample 140 modify the polarization to a state not completely orthogonal to the polarization state of the sample illumination, the polarization beam splitter 132 will still direct the reflected sample illumination using the modified polarization, but with lower efficiency. For example, less than all of the reflected sample illumination having the modified polarization will be directed to the polarization modification channel 120, and the remaining portion of the reflected sample illumination having the modified polarization may be directed to the interference channel 150.

[0039] The polarization modification channel 120 includes, for example, one or more optical components 122, such as lenses and a detector 124. For example, the detector 124 can be a camera that receives the modified polarization sample illumination. Thus, the detector 124 receives the modified polarization sample illumination to detect the occurrence and magnitude of the polarization modification at the sample 140. The detector 124 may receive an image of the sample 140, and based on this image, it can detect the magnitude of the polarization modification generated by the sample. The image of the sample 140 received by the detector 124 can be aligned with the image of the sample received by the detector 154 in the interference channel 150, for example, such that corresponding pixels in the individual detectors 124 and 154 correspond to the same location on the sample 140.

[0040] The interference detector 154 and the polarization modification detector 124 are coupled to a computer 170, such as a workstation, a personal computer, a central processing unit, or other suitable computer system or systems. The computer 170 is preferably included within the interferometer 100, or connected to the interferometer 100 or otherwise associated with the interferometer. The computer 170 can also control the movement of the stage 144 and the operation of the chuck 142. The computer 170 also collects and analyzes interference data and polarization modification data obtained from the interference detector 154 and the polarization modification detector 124, as discussed herein. For example, the computer 170 can analyze the interference data and the polarization modification data to determine one or more physical characteristics of the sample 140 (e.g., the presence of defects), as discussed below. The computer 170 includes at least one processor 172 having a memory 174, and a user interface including, for example, a display 176 and an input device 178. The computer 170 can use a non-transitory computer-usable storage medium 179 embodying computer-readable program code for causing the at least one processor to control the interferometer 100 and perform functions including the analysis described herein. In accordance with the present disclosure, those of ordinary skill in the art can implement the data structures and software code described in this specific embodiment for automatically implementing one or more actions, and store the data structures and software code on, for example, the computer-usable storage medium 179, which can be any device or medium capable of storing code and / or data for use by a computer system such as the processor 172. The computer-usable storage medium 179 can be, but is not limited to, magnetic storage devices and optical storage devices, such as flash drives, disk drives, magnetic tapes, optical disks, and DVDs (Digital Versatile Discs or Digital Video Discs). A communication port 177 can also be used to receive instructions for programming the computer 170 to perform any one or more of the functions described herein, and can represent any type of communication connection, such as a communication connection to the Internet or any other computer network. The communication port 177 can further export signals (e.g., having measurement results and / or instructions) in a feedforward or feedback process to another system (such as an external processing tool) for adjusting process parameters associated with the manufacturing process steps of the sample based on the measurement results. Additionally, the functions described herein can be embodied in whole or in part in the circuitry of an application specific integrated circuit (ASIC) or a programmable logic device (PLD), and these functions can be embodied in a computer-readable descriptor language usable to create an ASIC or PLD that operates as described herein.

[0041] As Figure 1 shown, in the case of using a phase mask, the surface height z at a single position represented by a pixel unit can be determined when signals are detected simultaneously at four different phases. If all four pixels are at the same height z, the signal S received in the 2×2 pixel unitι (where i is the phase shift at each pixel expressed as a multiple of 90°, and it should be understood that a phase shift of 360° is the same as a phase shift of 0°) can be used to calculate the surface height z. For example, using

[0042]

[0043] Thus, if the wavelength λ of the light source is known, Equation 1 from the four phase signals can be used to determine the height z. Other ways of calculating these parameters based on the four measured intensities or based on different amounts of intensity are possible, as will be apparent to those of ordinary skill in the art.

[0044] For example, three known phase shifts can be used to determine the surface height z. For example, for three samples (S i , i = 1, 2, 4), the surface height z can be determined by the following:

[0045]

[0046] It should be understood that different samples S i can be selected, and other ways of calculating these parameters based on the three measured intensities are possible, as will be apparent to those of ordinary skill in the art.

[0047] Advantageously, since phase-shifting interferometers (e.g., interferometer 100) can determine the topography of a sample surface from a single captured image, these devices are capable of rapidly inspecting the entire surface of a sample (e.g., a semiconductor wafer) for defects at a rate comparable to that of conventional inspection tools (e.g., bright-field inspection tools and dark-field inspection tools), while providing additional information not available in conventional tools, such as topography and surface polarization. Additionally, in the case of obtaining all data within a single exposure, the effects of vibrations on all axes are reduced, especially those at low frequencies. The sensitivity of the phase-shifting interferometer is sufficient to detect surface height changes caused by defects, and the lateral resolution can be configured to have, for example, a desired image size (e.g., pixel size and array size) and objective magnification to capture desired defects of interest, such as those having a lateral size from 0.2 μm to at least 100 μm. For example, in one embodiment, the pixel size can match the resolution of the interferometer. In the case of a wavelength of 450 nm and an instrument numerical aperture (NA) of 0.30, the instrument resolution using the Sparrow limit (λ / 2NA) is 750 nm. Using a pixel size of 750 nm sets the effective tool resolution to 1.5 μm (since the measurement uses a 2×2 pixel kernel), while providing the largest possible acquisition area from each camera image and thus the best sample acquisition rate. Detection of even smaller defects can be performed and used herein and is further described in U.S. Patent Application No. 16 / 197,737, titled "Sub-Resolution Defect Detection," filed on November 21, 2018, which is incorporated herein by reference in its entirety.

[0048] Figure 3 Another phase-shifting interferometer 300 is shown using the polarization modification channel 120 to detect polarization modification of the sample 140. The phase-shifting interferometer 300 is similar to Figure 1 the phase-shifting interferometer 100 shown, and like-designated elements are the same. The phase-shifting interferometer 300 differs in layout from the phase-shifting interferometer 100 and uses circular polarization instead of the linear polarization used by the phase-shifting interferometer 100.

[0049] As Figure 3 shown, the phase-shifting interferometer 300 includes a beam splitter 118 that directs (e.g., reflects) light toward the interference objective 130 via a polarizer 119 and directs (e.g., transmits) the reflected sample illumination having a polarization component modified by the sample 140 toward the polarization modification channel 120. The polarization beam splitter 132 is shown reflecting the sample illumination toward the sample 140 and transmitting the reference illumination 139 toward the reference surface 138 and transmitting light toward the interference channel 150.

[0050] The polarizer 119 can be, for example, a linear polarizer, but in some embodiments can be a circular polarizer and has an adjustable variable orientation to maximize fringe contrast. In use, the quarter-wave plates 131 and 133 are positioned in front of the sample 140 and the reference mirror 138, for example, between the beam splitter 132 and the objective lenses 134 and 136. As shown, the sample illumination 135 passes through the quarter-wave plate 131 and the sample objective lens 134 and is incident on the sample 140. The sample illumination 135 is reflected from the sample 140, and the reflected sample illumination returns through the sample objective lens 134 and the quarter-wave plate 131. Similarly, in use, the reference illumination 139 passes through the quarter-wave plate 133 and the reference objective lens 136 and is incident on the reference mirror 138. The reference illumination 139 is reflected from the reference mirror 138, and the reflected reference illumination returns through the reference objective lens 136 and the quarter-wave plate 133.

[0051] The quarter-wave plates 131 and 133 convert the orthogonally polarized beams from the polarization beam splitter 132 into circularly polarized light with opposite directions. For example, the p-polarized beam is converted into right-handed circular polarization, and the s-polarized beam is converted into left-handed circular polarization. Thus, if the fast axes of the quarter-wave plates 131 and 133 are parallel, the sample illumination 135 and the reference illumination 139 are circularly polarized with opposite senses after passing through the quarter-wave plates 131 and 133. In the case of using circularly polarized light in each sample path and reference path, each path must undergo one reflection and one transmission in the polarization beam splitter 132, thereby balancing the effect of any imbalance in the polarization efficiency of the polarization beam splitter 132.

[0052] The reference surface 138 is non-polarizing and will reflect the reference illumination 139 without modifying the polarization state of the reference illumination 139. After returning through the quarter-wave plate 133, the reflected reference illumination will become linearly polarized but with a polarization state orthogonal to the polarization state of the incident reference illumination. Thus, the polarization beam splitter 132 will direct (e.g., reflect) the reflected reference illumination towards the interference channel 150 with high efficiency, as Figure 3 shown.

[0053] If the sample 140 is non-polarizing, the sample illumination 135 incident on and reflected by the sample 140 will behave in a manner similar to the reference illumination 139. In other words, the polarization state of the sample illumination 135 will not be modified by the sample 140, and after being reflected by the sample 140 and returning through the quarter-wave plate 131, the reflected sample illumination 135 will become linearly polarized with a polarization state orthogonal to the polarization state of the incident sample illumination. Thus, the polarization beam splitter 132 will direct (e.g., transmit) the unmodified reflected sample illumination towards the interference channel 150 with high efficiency, as Figure 3As shown, where the reflected sample illumination and the reflected reference illumination will interfere with each other.

[0054] If the surface of sample 140 modifies the polarization of sample illumination 135, then after returning through quarter-wave plate 131, the reflected sample illumination 135 will have a mixture of linearly polarized components. After returning through quarter-wave plate 131, the portion of the reflected sample illumination with polarization components unmodified by sample 140 will have a polarization state orthogonal to the polarization state of the incident sample illumination and will be directed by polarization beam splitter 132 towards interference channel 150 (e.g., as Figure 3 shown, transmitted), where the sample illumination and the reflected reference illumination will interfere with each other, as discussed above. After the test path signal and the reference path signal are combined, the modulation intensity of the interference signal detected in interference channel 150 will be reduced by the amount of light lost due to the polarization modification caused by sample 140.

[0055] The portion of the reflected sample illumination 135 with polarization modified by sample 140 will pass through quarter-wave plate 131 and will be directed at least in part by polarization beam splitter 132 (e.g., as Figure 3 shown, reflected) back in the direction of incident beam 112, i.e., reflected towards polarization modification channel 120. For example, if the circular polarization of the sample illumination is completely converted by sample 140 to the opposite circular polarization state, then after returning through quarter-wave plate 131, the reflected sample illumination will become linearly polarized, where the polarization components are orthogonal to the polarization components of the incident sample illumination. Polarization beam splitter 132 will direct a portion of the reflected sample illumination with modified polarization towards polarization modification channel 120 with high efficiency. Of course, if the region of sample 140 modifies the polarization to a state not completely orthogonal to the polarization state of the sample illumination, polarization beam splitter 132 will still direct the reflected sample illumination using the modified polarization, but with lower efficiency, e.g., less than all of the reflected sample illumination with modified polarization will be directed to polarization modification channel 120, and the remainder of the reflected sample illumination with modified polarization may be directed to interference channel 150.

[0056] The portion of the sample illumination with polarization modified by sample 140 is directed by beam splitter 118 towards polarization modification channel 120, which detects the magnitude of the polarization modification produced by the sample, as discussed above.

[0057] Although Figure 3 the interference objective 130 shown in has a Linnik geometry (or a Michelson geometry in the case of removing reference objective 136), other interference objective geometries may be used, such as a Mirau objective, as discussed above. Figure 4Shows the Mirau objective lens 130M that can be used as an interference objective lens in the interferometer 100. In the case of the Mirau objective lens 130M, the polarization beam splitter 132 receives the polarized illumination beam from the polarizer 119 and guides all the illumination beams towards the Mirau objective lens 130M. For example, the polarizer 119 can be set at an angle of 0° with respect to the axis of the polarization beam splitter 132. The Mirau objective lens 130M includes a beam splitter 132M and a pupil 134M (which can be a lens) and a reference mirror 138M positioned on the pupil 134M (e.g., at the center). As shown in the figure, a quarter-wave plate 131 is provided between the beam splitter 132 and the Mirau objective lens 130M, and in the absence of polarization modification of the sample, it makes the polarization of the reflected reference illumination and the reflected sample illumination at the beam splitter 132 orthogonal to their incident linear polarization direction. Similar to Figure 3 the operation described in, the polarization beam splitter 132 guides a portion of the reflected reference illumination and the reflected sample illumination with unmodified polarization to the interference channel 150, and guides the reflected sample illumination with modified polarization to the polarization modification channel 120. The choice of the objective lens can be restricted depending on the polarization state of the light in the instrument. If desired, the function of the quarter-wave plate 131 can be incorporated into the beam splitter 132M and the pupil 134M. For example, the reference mirror 138M on the pupil 134M can include a circular polarizer. The orientations of the polarization of the reflected sample illumination and the reflected reference illumination are parallel, while Figure 3 the phase mask interferometer shown in requires orthogonal polarization. Therefore, the Mirau objective lens can preferably be used together with the Figure 5 scanning interferometer described below. If linear polarization is used, as Figure 1 discussed in, rather than Figure 3 the circular polarization discussed in, the Mirau objective lens 130M can be used without the quarter-wave plate 131, and the interference channel 150 and the polarization modification channel 120 are respectively parallel and perpendicular to the light from the sample 140.

[0058] Figure 5Another scanning interferometer 500 is shown that can use a polarization modification channel 120 to detect the polarization modification of a sample 140. The scanning interferometer 500 can be similar to the phase-shifting interferometer 100, and like-designated elements are the same. For example, the scanning interferometer 500 can include an interference channel 150 and a polarization modification channel 120, as discussed. The scanning interferometer 500 can include an actuator 502 coupled to an objective lens 134 controlled by a computer system 170 to adjust the vertical position (Z-height) of the objective lens 134 along the optical axis such that interference signals at multiple phase shifts can be acquired. Additionally or alternatively, the vertical position (Z-height) of the sample 140 can be adjusted using a stage 144. As another option, scanning can be achieved by moving the reference mirror in a direction parallel to the optical axis or by separately moving the reference objective lens 136 using an actuator 137. The scanning interferometer 500 can use a detector 554 (e.g., a camera) without a phase array to detect interference signals at multiple phase shifts, whereby the surface height at each position can be generated and thus the topography of the sample can be generated, or reflectivity measurements at each position of the sample can be generated, as is well known in the art.

[0059] Figure 6 and Figure 7 are simplified diagrams respectively showing the evolution of polarization in the sample optical path and the reference mirror optical path of an electric field system model for the polarization modification channel 120 shown in Figure 3 and Figure 5 The polarization can be targeted for Figure 1Generate a similar electric field system model. For example, remove the quarter-wave plates 131 and 133. An electric field system model for the polarization modification channel 120 can be generated based on the following assumptions: ignore depolarization caused by optical elements, assume that the polarizers are ideal, the input light is completely unpolarized, does not include polarization bidirectional attenuation, and for the purpose of discussion only, includes only axial rays (both the incident angle and the azimuth angle are zero). Each of these assumptions can be removed to improve the system model as needed. The input polarizer 119 is set at an angle of 45° (or some other angle if needed) with respect to the axis of the polarization beam splitter 132 to balance the reflectivities of the sample 140 and the reference mirror 138. The polarization beam splitter 132 acts as a linear polarizer, sending light in one polarization state to the sample 140 and light in an orthogonal polarization state to the reference mirror 138. For convenience, the equivalent linear polarizer is denoted as P0 for the light guided to the sample and P90 for the light guided to the reference mirror, but the angles can vary according to the operation of the polarization beam splitter 132. The quarter-wave plates 131 and 133 are set at an angle of 45° with respect to the polarization direction. If needed, the fast axes of the quarter-wave plates 131 and 133 in both the signal and reference channels can be parallel to each other or can be at an angle of 90° to each other. The light is reflected by the sample 140 or reflected by the reference mirror 138, and when returning, the polarization components operate as if rotated by 90°. Under these assumptions, for the configuration of the optical metrology device 100 with the polarizers 119, 132, and 131, the outgoing electric field vector of the rays irradiating the sample 140 and entering the polarization modification channel 120 can be written in the following matrix form.

[0060] E 样品 =P -45 P0Q -45 F3Q 45 P0P 45 E in

[0061] (Equation 1)

[0062] where each term is described in Table 1 below. Figure 6 Shows the evolution of polarization in the sample optical path from the light source 110 to the polarization modification detector 124, where Figure 6 the top part shows the optical path emitted by the light source 110 and incident on the sample 140, and Figure 6 the lower part shows the optical path of the light reflected by the sample 140 and received by the polarization modification detector 124.

[0063] Similarly, for the configuration of the optical metrology device 300 with the polarizers 119, 132, and 133, the outgoing electric field vector of the rays irradiating the reference mirror 138 and entering the polarization modification channel 120 can be written in the following matrix form.

[0064] E镜 = P -45 P 90 Q -45 FRQ 45 P 90 P 45 E in

[0065] (Equation 2)

[0066] Each term is described in Table 1 below. Figure 7 Shows the evolution of polarization in the reference mirror optical path from the light source 110 to the polarization modification detector 124, where Figure 7 The top part of shows the optical path emitted by the light source 110 and incident on the reference mirror 138, and Figure 7 The lower part of shows the optical path of the light reflected by the reference mirror 138 and received by the polarization modification detector 124.

[0067]

[0068] Table 1

[0069] If desired, the effect of reflection on the electric field can be included in the calculation of the Jones matrix for the sample S or the reference mirror R, in which case the reflection term F will not be used in Equation 1 or 2.

[0070] In the general case, the incident electric field E In will vary with the spatial frequency k and the position in the pupil. Additionally, the model includes contributions from the sample S and the polarizer and beam splitter, including the effect of polarization bidirectional attenuation in the beam splitter and other components.

[0071] In the approximation r sp = r ps = 0, the signal intensity in the polarization modification channel 120 is

[0072] I 偏振混合 = |E 样品 + E 镜 | 2

[0073] (Equation 3)

[0074]

[0075] If the reference mirror 138 and the sample 140 are perfect reflectors such that r pp = r ss , s pp = s ss and s ps = s s= 0, then, as expected, there is no signal in the polarization modification channel 120. Unless the reference mirror 138 changes, the term |r pp - r ss | is constant. Thus, the signal in the polarization modification channel 120 indicates different sample reflection coefficients s pp and s ss or significant depolarization terms s sp and s ps . For a perfect reference mirror (r pp = r ss , r ps = r p = 0), Equation 4 becomes

[0076]

[0077] By comparison, in the absence of a path difference between the sample and reference illuminations and with a vertical (0°) phase mask element, the simple form of the magnitude of the signal in the interference channel 150 is

[0078]

[0079] A lower reflectivity region on the sample 140 that does not mix the polarization of the sample illumination 135 (i.e., the region under the conditions of s pp = s ss < 1 and s sp = s ps = 0) will produce a lower intensity image at the interference detector 154 than a region under the condition of perfect reflectivity (s pp = s ss = 1), but will not produce a signal in the polarization modification detector 124.

[0080] Regions on the sample 140 under the conditions of s pp ≠ s ss and / or s sp + s ps ≠ 0 will produce a lower intensity image at the interference detector 154, but will also produce a signal in the polarization modification detector 124. Examples of such regions are a set of parallel lines, as commonly found in semiconductor devices. These regions act as linear polarizers, and one of s pp and s ss is close to zero. Other regions with more complex two-dimensional or three-dimensional patterns and regions with anisotropic materials (such as diamond-like carbon) can also significantly mix the polarization states such that one or both of s sp and s ps are non-zero.

[0081] If the reference mirror 138 and the sample 140 are perfect elements, such as r pp= r ss , s pp = s ss and s ps = s sp = 0, there is no signal in the polarization modification channel 120. Therefore, this channel can operate with a higher gain than the interference channel 150, and thus can detect smaller imbalances in the sample reflection coefficient. Generally, the off - diagonal terms s ps and s sp in the sample reflectivity Jones matrix are much smaller than the diagonal terms s pp and s ss . Since the "normal" signal in the polarization mixing channel 120 is zero, operating at a high gain allows detection of regions with weak depolarization (s sp + s ps ≠ 0), even with a small impact on the main interference signal, because s sp + s ps << s pp + s ss . Thus, the polarization modification channel 120 can identify certain differences between regions with presumably the same pattern that cause changes in the off - diagonal sample reflectivity terms. Examples of such effects include changes in pattern linewidth, relative height, and material properties, which are well - known as they are the basis for measuring reflectivity changes of sub - resolution pattern properties using optical critical dimension (OCD) tools (also known as scatterometers). An example of such a device is the Atlas III OCD system manufactured by Nanometrics, Inc. In the current embodiment, the variation in the signal in the polarization modification channel 120 between regions with the same expected pattern is additional information that can be used to identify defects in the pattern through the modification of the local reflectivity coefficient by the defects.

[0082] Therefore, a comparison of the intensities of the signals received at the interference detector 154 and the polarization modification detector 124 will enable determination of the magnitudes of the polarization modification and reflectivity changes. Additionally, the corresponding pixels of the interference detector 154 and the polarization modification detector 124 can be aligned and mapped to physical locations on the sample 140. Thus, by imaging the sample 140 at both the interference detector 154 and the polarization modification detector 124, the polarization modification and reflectivity changes can be mapped to specific locations on the sample 140.

[0083] Figure 8An example showing the intensity of the interference image 802 of the sample 140 received, for example, by the detector 154 in the interference channel 150 and the intensity of the polarization-modified image 812 of the sample 140 received, for example, by the detector 124 in the polarization-modification channel 120. The first region 804 of the interference image 802 corresponds to the first region 814 of the polarization-modified image 812, and both are mapped to the same position on the sample 140. Similarly, the region 806 in the interference image 802 and the region 816 in the polarization-modified image 812 correspond to each other and correspond to the same position on the sample 140, and likewise, the region 808 in the interference image 802 and the region 818 in the polarization-modified image 812 correspond to each other and correspond to the same position on the sample 140. The regions 804, 806, and 808 in the interference image 802 represent regions with reduced intensity relative to the surrounding region 803, and the regions 814, 816, and 818 in the polarization-modified image 812 represent regions with the same or increased intensity relative to the surrounding region 813. As Figure 8 shown by the shading in, white indicates high intensity, i.e., the interference image 802 and the polarization-modified image 812 receive large signals, and black indicates low intensity or no intensity, i.e., the interference image 802 and the polarization-modified image 812 receive very little signal or no signal. Figure 8 The use of, for example, the dark shading in the regions 804 and 808 in the interference image 802 relative to the surrounding region 803 indicates that the intensity in the regions 804 and 808 is reduced by the same amount, while the use of the black shading in the region 806 indicates that the intensity in the region 806 is reduced by a greater amount, i.e., the interference image 802 receives very little signal or no signal in the region 806. The reason for the reduction in intensity in the regions 804, 806, and 808 may be due to a change in the reflectivity of the sample 140 (s pp = s ss < 1) and / or polarization modification of light caused by the sample 140 (s pp ≠ s ss and / or s ps + s sp ≠ 0). In the case where the polarization-modification channel 120 is not used, the corresponding polarization-modified image 812 is generated, and the reason for the reduction in intensity in the regions 804, 806, and 808 in the interference image 802 is unknown. The regions 814, 816, and 818 in the polarization-modified image 812 correspond to the regions 804, 806, and 808 in the interference image 802, and the differences in the amount of polarization modification caused by the sample 140 are shown using different levels of shading.

[0084] As Figure 8As shown, the first region 804 in the interference image 802 and the first region 814 in the polarization-modified image 812 represent regions of lower reflectivity on the sample 140 that do not modify the polarization of the sample illumination 135: s pp = s s s <1, s ps + s sp = 0, where for convenience, the reflectivity of the bright surrounding region is considered s pp = s ss = 1, s ps = s sp = 0. Since the reflectivity from the corresponding region on the sample 140 is lower than that of other regions on the sample 140, the region 804 in the interference image 802 has a reduced intensity relative to the surrounding region 803 in the image 802 (represented by the Figure 8 shading). However, the region 814 in the polarization-modified image 812 does not receive a signal (as indicated by the black shading) because the corresponding region on the sample 140 does not mix the polarization of the sample illumination 135. It should be understood that the black and white boundaries of the different regions in the interference image 802 and the polarization-modified image 812 only indicate the positions of the regions in the figure and do not represent signal intensity.

[0085] The second region 806 in the interference image 802 and the second region 816 in the polarization-modified image 812 represent regions on the sample where there is no reduction in reflectivity on the sample 140, but there is a complete polarization modification of the sample illumination 135: s pp = -s ss ≈ 1, s ps + s sp ≠ 0. Thus, as can be seen, the region 806 in the interference image 802 has a greatly reduced intensity or zero intensity, i.e., the interference image 802 does not receive a signal (as indicated by the black shading), and the region 816 in the polarization-modified image 812 receives an increased intensity due to the polarization modification of the sample illumination 135 (represented by the white shading).

[0086] The third region 808 in the interference image 802 and the third region 818 in the polarization-modified image 812 represent regions on the sample where there is a reduction in reflectivity on the sample 140 and there is a polarization modification of the sample illumination 135: s pp ≠ s ss , s pp + s ss <1, s ps + s sp≠ 0. It can be seen that the region 808 in the interference image 802 has a reduced intensity (represented by the dark shading), and the region 818 in the polarization-modified image 812 receives a small increase in intensity, i.e., the polarization-modified image 812 receives an increased signal due to the polarization modification of the sample illumination 135 (shown as a lighter shading relative to the surrounding region 813). Thus, it can be determined that the reduction in intensity in the region 808 is due to both a change in reflectivity on the sample 140 and a polarization modification of the sample illumination 135 caused by the sample 140.

[0087] Thus, a comparison of the intensities in corresponding regions of the interference image 802 and the polarization-modified image 812 enables determination of whether the intensity change in the interference image 802 is due to a reflectivity change, a polarization modification, or both a reflectivity change and a polarization modification of the sample 140. Additionally, the polarization modification and the reflectivity change can be mapped to specific locations on the sample 140.

[0088] Furthermore, as discussed above, data from the interference channel 150 can be used to determine the topography of the sample 140. If desired, the determined topography can be compared to one or more reference surfaces (e.g., other samples) or other locations in the same sample 140 to determine whether the topography is nominal or abnormal. A comparison of the topography characteristics (e.g., subtracting the topography characteristics of the region of interest of the sample from the topography characteristics of a reference location on a pixel-by-pixel basis) removes common pattern structures or characteristics and leaves only the variations. The resulting variations can be determined as defects, e.g., using thresholding to identify candidate defects and filtering the candidate defects based on one or more defect characteristics (e.g., height, size, shape, texture, etc.). The topography of the sample 140 determined from the interference channel 150 can be used together with the polarization modification data from the polarization modification channel 120, e.g., to detect defects.

[0089] Thus, even in cases where a change in the polarization pattern does not manifest as a reflectivity change or a topography in the interference channel 150, a change in the polarization pattern on the sample 140 can be detected as a change in the polarization modification signal using the polarization modification channel 120. Additionally, defects in a blurred pattern layout or a damaged pattern structure on the sample 140 can be detected by the effect of the defect on the polarization modification signal, signal intensity, or topography. Thus, the polarization modification channel 120 is not only capable of analyzing the polarization modification effect as part of the surface data (e.g., reducing the sensitivity of the polarization effect at the sample surface), but also provides information that can be used to detect and classify defects that may not be identifiable based solely on the image intensity in the interference channel or the measured sample topography.

[0090] Thus, as discussed below, one or more determined physical characteristics of sample 140 (e.g., the presence of defects, including size, location, type, etc.) can be determined by computer system 170 and can be transmitted and stored, for example, in a memory or database. Defect data can be transmitted to adjust one or more process tools associated with a particular manufacturing process step in the manufacturing sequence (e.g., the process tool responsible for the detected defect) or to adjust the manufacturing sequence of the sample itself, such as by rejecting or discarding the sample or a portion of the sample.

[0091] For example, a semiconductor process line can use defect data by correlating the expected yield loss or kill rate with each detected defect type. The kill rate of a defect is typically used in conjunction with survival statistics to determine which die are likely to produce defects. To estimate the yield impact from the defects reported on a given wafer, classification can be performed. In one implementation, features calculated for each defect can be used to perform the classification in an automated manner. In another implementation, the wafer can be loaded onto another tool with higher spatial resolution and images of each defect can be captured such that the defects can be manually classified, for example, by an operator. Defect data, which can include location and classification, can be used in a variety of ways. For example, defect data can be used to adjust the manufacturing sequence of the sample itself based on the number, type, and location of the detected defects, such as by rejecting or discarding the sample or a portion of the sample (e.g., the entire wafer or a die from the wafer). For example, if the wafer exceeds a threshold for yield loss, the entire wafer can be scrapped, thus avoiding the consumption of downstream process tool resources.

[0092] In addition, defect data can be used to adjust one or more process tools associated with a particular manufacturing process step in a manufacturing sequence (e.g., the process tool responsible for the detected defect), including adjusting process parameters or removing the process tool from the production line. For example, defect data can be compared with defect data from wafers that were processed slightly differently upstream, so that the process parameters of a particular process tool can be adjusted to reduce defects. In one example, the slurry composition of a chemical mechanical polishing (CMP) tool can be changed based on defect data. By comparing the number of defects before and after the slurry change, the slurry composition that produces fewer defects can be identified and used. In this way, the in-line yield learning provided by the defect inspection tool provides rapid tuning of the process flow, e.g., during process development. Additionally, defect data can be used to monitor the health of upstream process tools and identify and take corrective action with respect to any tool responsible for a defect. For example, by understanding the most likely root cause of each defect type at a given processing step, defect data can be used to monitor the health of upstream process tools. For example, a statistical process control (SPC) system can be used to perform this type of offset monitoring to flag samples as out of specification due to one or more defect types. Based on the type of defect, the upstream process tool responsible for the offset can be identified and appropriately addressed, e.g., by adjusting process parameters or by taking the processing tool out of service for maintenance before further samples are affected. Defect data can also be used to optimize the process tool preventive maintenance schedule.

[0093] If desired, other types of interferometers can be used to obtain the surface topography of the sample. For example, instead of using a phase mask, the interferometer can split the combined sample and reference illumination for detection by multiple detectors, each detector having a single linear polarization element. By splitting the combined test and reference illumination into multiple cameras (each camera having a different linear polarizer angle), an alternative arrangement can be achieved that allows simultaneous detection of interference signals having multiple phase shifts. Such an arrangement allows for maximum pixel density, where pixels in different paths sample the same point on the sample, but this comes at an additional cost and it is difficult to maintain alignment between the split channels and avoid modifying the polarization state during beam splitting. If the system is configured such that the light incident on the sample and reference mirror is linearly polarized, a circular polarizer can be used in this arrangement, which can be a single circular polarizer placed before the combined beam split, or each path after the combined beam split can have its own circular polarizer.

[0094] Figure 9 is a flowchart showing an example of a defect determination process 900 that can be performed by an optical inspection tool, such as interferometer 100. As Figure 9 shown, a polarized illumination beam (902) is generated, e.g., as Figure 1 、Figure 3 , Figure 4 or Figure 5 as shown by the light source 110 and polarizer 119 in. A first portion of the illumination beam is directed towards a reference mirror as reference illumination, and the reference mirror reflects the first portion of the illumination beam to produce reflected reference illumination (904), e.g., as Figure 1 , Figure 3 or Figure 5 as shown by the polarization beam splitter 132 in or Figure 4 as shown by the beam splitter 132M in. A second portion of the illumination beam is directed towards the sample as sample illumination, and the sample reflects the second portion of the illumination beam to produce reflected sample illumination, wherein an area of the sample partially modifies the polarization of the sample illumination (906), e.g., as Figure 1 , Figure 3 or Figure 5 as shown by the polarization beam splitter 132 in or Figure 4 as shown by the beam splitter 132M in.

[0095] The polarization beam splitter receives the reflected reference illumination from the reference mirror and the reflected sample illumination from the sample (908), e.g., as Figure 1 , Figure 3 , Figure 4 and Figure 5 as shown by the polarization beam splitter 132 in. The first portions of the reflected reference illumination and the reflected sample illumination are combined into an interference beam, and the interference beam is directed by the polarization beam splitter along an interference channel (910), e.g., as Figure 1 , Figure 3 , Figure 4 and Figure 5 as shown by the polarization beam splitter 132 in. The second portion of the reflected sample illumination from the area of the sample is directed by the polarization beam splitter along a polarization modification channel different from the interference channel (912), e.g., as Figure 1 , Figure 3 , Figure 4 and Figure 5 as shown by the polarization beam splitter 132 in.

[0096] The interference beam is detected in the interference channel to produce a first set of optical data (914), e.g., as Figure 1 , Figure 3 or Figure 5 as shown by the detectors 154, 554 in. The second portion of the reflected sample illumination is detected in the polarization modification channel to produce a second set of optical data indicative of the polarization modification of the sample illumination (916), e.g., as Figure 1 , Figure 3 or Figure 5 as shown by the detector 124 in. At least one of the topography or intensity determined from the first set of optical data and the polarization modification from the second set of optical data is used to detect defects on the sample (918).

[0097] In one embodiment, a first portion of the illumination beam is directed towards a reference mirror by a polarization beam splitter and is linearly polarized in a first polarization orientation, and a second portion of the illumination beam is directed towards a sample by the polarization beam splitter and is linearly polarized in a second polarization orientation. For example, the reference illumination incident on the reference mirror can be linearly polarized and have a first polarization orientation, and the sample illumination incident on the sample can be linearly polarized and have a second polarization orientation orthogonal to the first polarization orientation. A first portion of the reflected sample illumination can have a first polarization orientation component aligned with the second polarization orientation, and a second portion of the reflected sample illumination can have a second polarization orientation aligned with the first polarization orientation and orthogonal to the first polarization orientation component of the first portion of the reflected sample illumination. In one embodiment, the first portion of the illumination beam incident on the reference mirror can be circularly polarized with a first circular polarization handedness, and the second portion of the illumination beam incident on the sample can be circularly polarized with a second circular polarization handedness different from the first circular polarization handedness. In one embodiment, the reflected reference illumination received by the polarization beam splitter is linearly polarized in the second polarization orientation, and wherein a first portion of the reflected sample illumination is linearly polarized in a direction aligned with the first polarization orientation, and a second portion of the reflected sample illumination is linearly polarized in a direction aligned with the second polarization orientation. For example, the sample illumination and the reflected sample illumination can pass through a first quarter-wave plate, and the reference illumination and the reflected reference illumination can pass through a second quarter-wave plate.

[0098] In one embodiment, the interference beam can pass through a quarter-wave plate.

[0099] In one embodiment, the intensity variation from a first set of optical data of the interference beam and the polarization modification from a second set of optical data can be used to detect a polarization modification at a sample region. For example, in one embodiment, the region of the sample is a first region, and wherein a second region of the sample partially modifies the polarization of the sample illumination by a different amount than the first region of the sample, and the process can further include using the intensity variation from the first set of optical data of the interference beam, in the second region, and the polarization modification from the second set of optical data, in the second region, to detect the polarization modification at the second region of the sample.

[0100] In one embodiment, a first set of optical data and a second set of optical data can be used to detect a lower reflectivity at a sample region. For example, the intensity decrease from a first set of optical data of the interference beam and the polarization modification from a second set of optical data can be used to detect a lower reflectivity at a sample region.

[0101] In one embodiment, the first data set can be a first image of the sample generated from the interference beam, and the second data set can be a second image of the sample generated from a second portion of the sample reflected illumination.

[0102] The process may further include directing the illumination beam through a second beam splitter towards a polarization beam splitter, e.g., as shown by beam splitter 118 as shown in Figure 3 and Figure 5 The second portion of the reflected sample illumination may be received from the polarization beam splitter by the second beam splitter and directed along a polarization modification channel, e.g., as shown by beam splitter 118 as shown in Figure 3 and Figure 5 shown.

[0103] In one embodiment, the first portion of the illumination beam may be directed towards a reference mirror as reference illumination, and the second portion of the illumination beam may be directed towards the sample through a polarization beam splitter that is part of an interference objective as sample illumination, e.g., as shown by polarization beam splitter 132 as shown in Figure 1 , Figure 3 and Figure 5 shown. By way of example, the interference objective may be one of a Linnik objective or a Michelson objective.

[0104] In one embodiment, the first portion of the illumination beam may be directed towards a reference mirror as reference illumination, and the second portion of the illumination beam may be directed towards the sample through a second beam splitter that is part of an interference objective as sample illumination, e.g., as shown by beam splitter 132M as shown in Figure 4 shown. By way of example, the interference objective may be a Mirau objective, where the reference mirror is positioned at the pupil of the interference objective through which the illumination beam, the reflected reference illumination, and the reflected sample illumination pass, e.g., as shown by pupil 134M and reference mirror 138M as shown in Figure 4 . The illumination beam and the reflected reference illumination and the reflected sample illumination may pass through a quarter-wave plate positioned between the polarization beam splitter and the pupil, e.g., as shown by circular polarizer 131 in Figure 4 shown.

[0105] Figure 10 is a flowchart showing an example of a defect determination process 1000 that may be performed by an optical inspection tool such as interferometer 100. As shown in Figure 10As shown, at block 1002, a first portion of the polarized illumination beam is directed toward a reference mirror as reference illumination, and the reference mirror reflects the first portion of the illumination beam to produce reflected reference illumination. Additionally, at block 1004, a second portion of the illumination beam is directed toward a sample as sample illumination, and the sample reflects the second portion of the illumination beam to produce reflected sample illumination, where a region of the sample partially modifies the polarization of the sample illumination. For example, the reference illumination incident on the reference mirror can be linearly polarized and have a first polarization orientation, and the sample illumination incident on the sample can be linearly polarized and have a second polarization orientation orthogonal to the first polarization orientation. In another example, the first portion of the illumination beam incident on the reference mirror can be circularly polarized with a first circular polarization handedness, and the second portion of the illumination beam incident on the sample can be circularly polarized with a second circular polarization handedness different from the first circular polarization handedness. An apparatus for directing a first portion of a polarized illumination beam toward a reference mirror as reference illumination, where the reference mirror reflects the first portion of the illumination beam to produce reflected reference illumination, for example, can include a polarization beam splitter 132 or a beam splitter 132M and a quarter-wave plate 131, as Figure 1 , Figure 3 , Figure 4 or Figure 5 shown. An apparatus for directing a second portion of an illumination beam toward a sample as sample illumination, where the sample reflects the second portion of the illumination beam to produce reflected sample illumination, and where a region of the sample partially modifies the polarization of the sample illumination, for example, can include a polarization beam splitter 132 or a beam splitter 132M, and a quarter-wave plate 131 or 133, as Figure 1 , Figure 3 , Figure 4 or Figure 5 shown.

[0106] At block 1006, the reflected reference illumination and a first portion of the reflected sample illumination from the region of the sample are combined into an interference beam, and the interference beam is directed along an interference channel based on polarization. Additionally, at block 1008, a second portion of the reflected sample illumination from the region of the sample is directed along a polarization modification channel based on polarization, where the polarization modification channel is different from the interference channel. An apparatus for combining the reflected reference illumination and a first portion of the reflected sample illumination from the region of the sample into an interference beam and directing the interference beam along an interference channel based on polarization; for example, can include Figure 1 , Figure 3 , Figure 4 or Figure 5 the polarization beam splitter 132, beam splitter 132M, quarter-wave plates 131, 133 shown. An apparatus for directing a second portion of the reflected sample illumination from the region of the sample along a polarization modification channel based on polarization, where the polarization modification channel is different from the interference channel, for example, can include Figure 1 , Figure 3 ,Figure 4 or Figure 5 the polarization beam splitter 132, beam splitter 132M, and quarter-wave plate 131 shown in

[0107] At block 1010, an interference beam in an interference channel is detected to generate a first set of optical data. For example, the first set of optical data may be an image of a sample generated by the interference beam. Additionally, at block 1012, a second portion of the reflected sample illumination in a polarization modification channel is detected to generate a second set of optical data indicative of the polarization modification of the sample illumination. For example, the second set of optical data may be a second image of the sample generated by the second portion of the reflected sample illumination. The apparatus for detecting the interference beam in the interference channel to generate the first set of optical data may, for example, include Figure 1 , Figure 3 or Figure 5 the detectors 154, 554, and quarter-wave plate 153 shown in Figure 1 , Figure 3 or Figure 5 the detector 124 and linear polarizer 119 in

[0108] At block 1014, defects on a sample are detected using a first set of optical data and polarization modifications from a second set of optical data. An apparatus for detecting defects on a sample using a first set of optical data and polarization modifications from a second set of optical data can include, for example, one or more processors 172 having dedicated hardware or executable code or software instructions implemented in a memory 174 or a non-transitory computer-usable storage medium 179. In some embodiments, intensity variations from the first set of optical data of an interference beam and polarization modifications from the second set of optical data can additionally be used to detect polarization modifications at a region of the sample. For example, the region of the sample can be a first region, and a second region of the sample can partially modify the polarization of sample illumination by a different amount than the first region of the sample. Polarization modifications can be detected at the second region of the sample using intensity variations from the first set of optical data of the interference beam, at the second region, and polarization modifications from the second set of optical data, at the second region. An apparatus for detecting polarization modifications at a region of a sample using intensity variations from the first set of optical data of an interference beam and polarization modifications from the second set of optical data can include, for example, one or more processors 172 having dedicated hardware or executable code or software instructions implemented in a memory 174 or a non-transitory computer-usable storage medium 179. In some embodiments, the first set of optical data and the second set of optical data can additionally be used to detect a lower reflectivity at a region of the sample. An apparatus for detecting a lower reflectivity at a region of a sample using the first set of optical data and the second set of optical data can include, for example, one or more processors 172 having dedicated hardware or executable code or software instructions implemented in a memory 174 or a non-transitory computer-usable storage medium 179.

[0109] References throughout this specification to "one example", "an example", "certain examples", or "exemplary embodiments" mean that a particular feature, structure, or characteristic described in connection with the feature and / or example is included in at least one feature and / or example of the claimed subject matter. Thus, the phrases "in one example", "an example", "in certain examples", or "in certain embodiments", or other similar phrases that appear throughout this specification are not necessarily all referring to the same feature, example, and / or limitation. Additionally, the particular features, structures, or characteristics may be combined in one or more examples and / or features.

[0110] Some portions of the detailed description contained herein are presented in terms of algorithms or symbolic representations of operations on binary digital signals stored within the memory of a specific device or a dedicated computing device or platform. In the context of this particular specification, the term specific device, etc. includes a general-purpose computer once programmed to perform specific operations in accordance with instructions from program software. The algorithmic descriptions or symbolic representations are examples of techniques used by those of ordinary skill in the signal processing or related arts to convey the substance of their work to others skilled in the art. An algorithm is here generally considered to be a self-consistent sequence of operations or a similar sequence of signal processing, resulting in a desired outcome. In this context, an operation or process involves the physical manipulation of physical quantities. Usually, though not necessarily, such quantities may take the form of electrical or magnetic signals capable of being stored, transmitted, combined, compared, or otherwise manipulated. It has proven convenient at times (mainly for reasons of common usage) to refer to such signals as bits, data, values, elements, symbols, characters, terms, numbers, numerals, etc. However, it should be understood that all such or similar terms will be associated with appropriate physical quantities and are merely convenient labels. Unless otherwise specifically stated, as will be apparent from the discussion herein, it should be understood that throughout the specification the terms such as "processing", "computing / calculating", "determining", etc. refer to the actions or processes of a specific device, such as a dedicated computer, a dedicated computing device, or a similar dedicated electronic computing device. Thus, in the context of this specification, a dedicated computer or a similar dedicated electronic computing device is capable of manipulating or transforming signals, typically represented as physical electronic or magnetic quantities within a memory, a register, or other information storage devices, transmission devices, or display devices of the dedicated computer or a similar dedicated electronic computing device.

[0111] In the foregoing detailed description, numerous specific details have been set forth in order to provide a thorough understanding of the claimed subject matter. However, those skilled in the art will understand that the claimed subject matter may be practiced without these specific details. In other instances, methods and devices known to those of ordinary skill in the art have not been described in detail so as not to obscure the claimed subject matter.

[0112] As used herein, the terms "and", "or", and "and / or" may have multiple meanings, which depend at least in part on the context in which such terms are used. Generally, if used to associate a list, such as A, B, or C, "or" is intended to mean A, B, and C, used here in an inclusive sense, as well as A, B, or C, used here in an exclusive sense. Additionally, as used herein, the term "one or more" may be used to describe any feature, structure, or characteristic in the singular, or may be used to describe multiple features, structures, or characteristics or some other combination thereof. It should be noted, however, that this is merely an illustrative example and the claimed subject matter is not limited to this example.

[0113] Although the presently considered exemplary features have been described and illustrated, those skilled in the art will understand that various other modifications can be made and equivalents can be substituted without departing from the claimed subject matter. Additionally, many modifications can be made to adapt a particular situation to the teachings of the claimed subject matter without departing from the central concept described herein.

[0114] Although the present invention has been shown in connection with specific embodiments for purposes of illustration, the present invention is not limited thereto. Various adaptations and modifications can be made without departing from the scope of the present invention. Accordingly, the spirit and scope of the appended claims should not be limited to the foregoing description.

Claims

1. An optical inspection device configured to detect defects on a sample, the optical inspection device comprising: An interference objective lens, the interference objective lens including a polarization beam splitter configured to receive a polarized illumination beam and direct a first portion of the illumination beam toward a reference mirror as reference illumination and configured to direct a second portion of the illumination beam toward the sample as sample illumination, wherein the reference mirror is configured to reflect the first portion of the illumination beam to produce reflected reference illumination, the polarization beam splitter further being configured to receive the reflected reference illumination from the reference mirror and receive reflected sample illumination from the sample, the polarization beam splitter being configured to combine a first portion of the reflected sample illumination and the reflected reference illumination into an interference beam, the interference beam being directed along an interference channel, and a second portion of the reflected sample illumination being directed along a polarization modification channel different from the interference channel; A first detector in the interference channel and configured to receive the interference beam and produce a first set of optical data from the sample; A second detector in the polarization modification channel and configured to receive the second portion of the reflected sample illumination and produce a second set of optical data indicative of polarization modification of the sample illumination; and At least one processor coupled to the first detector and the second detector and configured to detect a defect on the sample using the first set of optical data and polarization modification from the second set of optical data.

2. The optical inspection device according to claim 1, wherein, The interference objective lens is one of a Linnik objective lens or a Michelson objective lens.

3. The optical inspection device according to claim 1, further comprising: A first quarter-wave plate between the polarization beam splitter and the reference mirror; And A second quarter-wave plate between the polarization beam splitter and the sample.

4. The optical inspection device according to claim 1, further comprising a quarter-wave plate between the polarization beam splitter and the first detector.

5. The optical inspection device according to claim 1, wherein, The at least one processor is further configured to detect polarization modification at a region of the sample using intensity variation from the first set of optical data of the interference beam and polarization modification from the second set of optical data.

6. The optical inspection device according to claim 5, wherein, The region of the sample is a first region, and wherein the sample further includes a second region that partially modifies the polarization of the sample illumination by an amount different from that of the first region of the sample, wherein the at least one processor is further configured to detect polarization modification at the second region of the sample using intensity variation from the first set of optical data of the interference beam, in the second region, and polarization modification from the second set of optical data, in the second region.

7. The optical inspection device according to claim 1, wherein, The at least one processor is further configured to detect a lower reflectivity at a region of the sample relative to a surrounding region of the region using the first set of optical data and the second set of optical data.

8. The optical inspection device according to claim 7, wherein, The at least one processor is configured to detect a lower reflectivity at a region of the sample relative to a surrounding region of the region using intensity reduction from the first set of optical data of the interference beam and polarization modification from the second set of optical data.

9. The optical inspection device according to claim 1, wherein, The first detector receives a first image of the sample from the interference beam, and the second detector receives a second image of the sample from a second portion of the sample illumination.

10. The optical inspection device according to claim 1, further comprising: A second beam splitter that directs the illumination beam towards the interference objective and directs a second portion of the reflected sample illumination along a polarization modification channel that includes the second detector.

11. A method for detecting defects on a sample, the method comprising: Direct a first portion of the polarized illumination beam towards a reference mirror as reference illumination, and the reference mirror reflects the first portion of the illumination beam to produce reflected reference illumination; Direct a second portion of the illumination beam towards the sample as sample illumination, and the sample reflects the second portion of the illumination beam to produce reflected sample illumination, wherein an area of the sample partially modifies the polarization of the sample illumination; Combine a first portion of the reflected sample illumination from the area of the sample and the reflected reference illumination into an interference beam, and direct the interference beam along an interference channel based on polarization; Direct a second portion of the reflected sample illumination from the area of the sample along a polarization modification channel based on polarization, wherein the polarization modification channel is different from the interference channel; Detect the interference beam in the interference channel to generate a first set of optical data; Detect a second portion of the reflected sample illumination in the polarization modification channel to generate a second set of optical data indicative of the polarization modification of the sample illumination; and Detect a defect on the sample using the polarization modification from the second set of optical data and the first set of optical data.

12. The method according to claim 11, wherein, The reference illumination incident on the reference mirror is linearly polarized and has a first polarization orientation, and the sample illumination incident on the sample is linearly polarized and has a second polarization orientation orthogonal to the first polarization orientation.

13. The method according to claim 11, wherein, The first portion of the illumination beam incident on the reference mirror is circularly polarized with a first chirality, and the second portion of the illumination beam incident on the sample is circularly polarized with a second chirality different from the first chirality.

14. The method according to claim 11, further comprising detecting a polarization modification at a region of the sample using an intensity change of the interference beam from the first set of optical data and a polarization modification of the second set of optical data.

15. The method according to claim 14, wherein The area of the sample is a first area, and wherein the sample further includes a second area that partially modifies the polarization of the sample illumination by an amount different from that of the first area of the sample, and the method further includes using the intensity change in the second area from the first set of optical data of the interference beam and the polarization modification in the second area from the second set of optical data to detect the polarization modification at the second area of the sample.

16. The method according to claim 11, further comprising detecting a lower reflectivity at a region of the sample relative to a surrounding region of the region using the first set of optical data and the second set of optical data.

17. The method according to claim 16, wherein Detecting a lower reflectivity of the area of the sample relative to the surrounding area of the area includes using a decrease in intensity from the first set of optical data of the interference beam and the polarization modification from the second set of optical data.

18. The method according to claim 11, wherein The first set of optical data includes a first image of the sample generated from the interference beam, and the second set of optical data includes a second image of the sample generated from the second portion of the reflected sample illumination.

19. An optical inspection device configured to detect a defect on a sample, the optical inspection device comprising: A device for guiding a first part of a polarized illumination beam towards a reference mirror as reference illumination and guiding a second part of the illumination beam towards the sample as sample illumination, the reference mirror reflecting the first part of the illumination beam to produce reflected reference illumination, wherein the sample reflects and partially modifies the polarization of the second part of the illumination beam to produce reflected sample illumination; A device for combining a first part of the reflected sample illumination and the reflected reference illumination into an interference beam and guiding the interference beam along an interference channel; A device for guiding a second part of the reflected sample illumination along a polarization modification channel, wherein the polarization modification channel is different from the interference channel; A device for detecting the interference beam in the interference channel to produce a first set of optical data; A device for detecting the second part of the reflected sample illumination in the polarization modification channel to produce a second set of optical data indicating the polarization modification of the sample illumination; and A device for detecting a defect on the sample using the polarization modification from the second set of optical data and the first set of optical data.

20. The optical inspection device according to claim 19, further comprising means for generating circular polarization of a first handedness for a first portion of the illumination beam incident on the reference mirror and circular polarization of a second handedness for a second portion of the illumination beam incident on the sample, the second handedness being different from the first handedness.

Images