Thickness measurement of a substrate using color measurement

Abstract

A metrology system for obtaining a measurement representative of a thickness of a layer on a substrate includes a camera positioned to acquire a color image of at least a portion of the substrate. A controller is configured to receive the color image from the camera; store a predetermined path in a coordinate space having at least two dimensions including a first color channel and a second color channel; store a function that provides a value representative of thickness as a function of position on the predetermined path; determine coordinates of a pixel in the coordinate space from color data for the pixel in the color image; determine a position of a point on the predetermined path that is closest to the coordinates of the pixel; and calculate a value representative of thickness from the function and the position of the point on the predetermined path.

Description

[0001] This application is a divisional application of the invention patent application filed on August 25, 2017, with application number "201780057683.6" and invention title "Thickness Measurement of Substrate Using Color Measurement". Technical Field

[0002] This disclosure relates to optical measurements (e.g., to detecting the thickness of layers on a substrate). Background Technology

[0003] Integrated circuits are typically formed on a substrate by sequentially depositing conductive, semiconducting, or insulating layers onto a silicon wafer. One fabrication step involves depositing a filler layer on a non-planar surface and planarizing the filler layer. For some applications, the filler layer is planarized until the upper surface of the patterned layer is exposed. Conductive filler layers, for example, can be deposited on a patterned insulating layer to fill trenches or holes within the insulating layer. After planarization, portions of a metal layer remaining between the raised patterns of the insulating layer are formed: vias, plugs, and lines providing conductive paths between thin-film circuits located on the substrate. For other applications (such as oxide polishing), the filler layer is planarized until a predetermined thickness is left on the non-planar surface. Furthermore, photolithography often requires planarization of the substrate surface.

[0004] Chemical mechanical polishing (CMP) is an acceptable planarization method. This planarization method typically requires mounting a substrate on a carrier or polishing head. The exposed surface of the substrate is usually positioned to abut against a rotating polishing pad. The carrier head provides a controlled load on the substrate to advance it against the polishing pad. An abrasive slurry with abrasive properties is typically provided to the surface of the polishing pad.

[0005] Variations in slurry distribution, pad condition, relative velocity between the pad and substrate, and load on the substrate can cause variations in the material removal rate. These variations, along with variations in the initial thickness of the substrate layer, cause variations in the time required to reach the grinding endpoint. Therefore, defining the grinding endpoint solely as a function of grinding time can lead to over- or under-grinding of the substrate.

[0006] For example, at inline or stand-alone measurement stations, various optical measurement systems (e.g., spectral or ellipsometric measurements) can be used to measure the thickness of the substrate layer before and after polishing. Furthermore, various in-situ monitoring techniques (such as monochromatic optics or eddy current monitoring) can be used to detect the polishing endpoint. Summary of the Invention

[0007] In one aspect, a measurement system for obtaining a measurement representing the thickness of a layer on a substrate includes: a support for holding a substrate for manufacturing an integrated circuit; a color camera positioned to acquire a color image of at least a portion of the substrate held by the support; and a controller. The controller is configured to receive the color image from the camera; store a predetermined path in a coordinate space having at least two dimensions including a first color channel and a second color channel; store a function that provides: a numerical value representing the thickness as a function of the position on the predetermined path; for a pixel of the color image, determining the coordinates of the pixel in the coordinate space from color data for that pixel in the color image; determining the position of a point on the predetermined path closest to the coordinates of the pixel; and calculating a value representing the thickness from the function and the position of the point on the predetermined path.

[0008] Another aspect is a computer program product comprising instructions for causing a processor to obtain a measurement representing the thickness of a layer on a substrate. Yet another aspect is a method for obtaining a measurement representing the thickness of a layer on a substrate.

[0009] Any implementation includes one or more of the following features.

[0010] The coordinate space can be two-dimensional or three-dimensional.

[0011] The first color channel and the selected color channel can be chosen from a group of color channels that include hue, saturation, brightness, X, Y, Z, red chroma, green chroma, and blue chroma. For example, the first color channel can be red chroma, and the second color channel can be green chroma.

[0012] The controller can be configured to determine the position of a point on a predetermined path by identifying a point on that path, where the normal vector of the path with respect to the thickness passes through that coordinate. The controller can be configured to resolve the degeneracy in the predetermined path by analyzing the coordinates of a cluster of pixels associated with a given physical region on the substrate surrounding that pixel. The controller can be configured to resolve the degeneracy in the predetermined path by determining the principal axis of that cluster in the coordinate space and selecting the branch of the predetermined path that is most closely parallel to the principal axis.

[0013] The function may contain: the difference between a first value for the first vertex and a second value for the second vertex of the predetermined path, based on the distance of a point from the first vertex. The first vertex is the starting point of the predetermined path, and the second vertex is the ending point of the predetermined path. The controller may be configured to store the predetermined path by storing one or more Bezier functions or polylines.

[0014] The controller can be configured to, for each pixel in a plurality of pixels of a color image, determine the coordinates of that pixel in the coordinate space from color data for that pixel in the color image; determine the position of the point on a predetermined path closest to that pixel's coordinates; and calculate a value representing the thickness from the function and the position of the point on the predetermined path; providing multiple values ​​representing the thickness for multiple different locations on the substrate. The controller can be configured to calculate the number of pixels from the plurality of pixels where the value representing the thickness does not meet a threshold. The controller can be configured to apply an image mask to the color image.

[0015] A color camera can be configured to acquire an image smaller than the entire substrate and configured to scan across the substrate, wherein a controller is configured to generate a complete two-dimensional color image of the substrate from multiple images from the color camera. The color camera can be: a color line-scanning camera with a detector assembly arranged along a first axis parallel to the surface of the substrate during scanning; and the system can include: an elongated white light source having a longitudinal axis parallel to the first axis and configured to guide light toward the substrate at a non-zero angle of incidence during scanning; a frame supporting the light source and the camera; and a motor for causing relative movement between the frame and the support along a second axis perpendicular to the first axis, thereby enabling the light source and camera to scan across the substrate.

[0016] The support can be positioned within an inline measurement station of a semiconductor manufacturing tool. The semiconductor manufacturing tool can be a chemical mechanical polisher, and the tool can include a robot for conveying a substrate to the inline measurement station, and a controller can be configured to measure the substrate before or after its surface has been polished by the chemical mechanical polisher.

[0017] The implementation may include one or more of the following potential advantages: The thickness at a location on the substrate layer can be determined absolutely or relative to other portions of the substrate layer, and unacceptable variations can be detected. This information can be used in feedforward or feedback applications to control polishing parameters, providing improved thickness uniformity. The algorithm for determining variations can be simple and has a low computational load.

[0018] Details of one or more embodiments are set forth in the accompanying drawings and the following description. Other aspects, features, and advantages will become apparent from the description and drawings, as well as from the claims. Attached Figure Description

[0019] Figure 1 A schematic diagram of an example of an inline optical measurement system is shown.

[0020] Figure 2 This is a flowchart of a method for determining layer thickness.

[0021] Figure 3 This is a schematic top view of the substrate.

[0022] Figure 4 This is a schematic diagram of a mask.

[0023] Figure 5 An example diagram is shown, illustrating the evolution of the color of light reflected from a substrate in a coordinate space with two color channels.

[0024] Figure 6 An example diagram is shown, which illustrates a predetermined path in a coordinate space with two color channels.

[0025] Figure 7 This is a flowchart of a method for determining layer thickness from color image data.

[0026] Figure 8 An example graph is shown, which illustrates a histogram in a coordinate space with two color channels derived from a color image of the test substrate.

[0027] The same reference numerals in the various figures indicate the same components. Detailed Implementation

[0028] The thickness of layers on a substrate before or after polishing can be measured optically, for example, at an inline or stand-alone measurement station. However, some optical techniques (such as spectral measurements) require expensive spectrometers and computationally intensive operations on the spectral data. Even beyond the computational load, in some cases, the results of the algorithms cannot meet the increasing accuracy requirements of users. Another measurement technique, however, involves capturing a color image of the substrate and analyzing the image in color space to determine the layer thickness. Specifically, the position along the path in a two-dimensional color space can provide information about the current state of polishing (e.g., the amount removed or the amount of material remaining).

[0029] refer to Figure 1The grinding apparatus 100 includes an in-line (also known as sequential) optical measurement system 160 (e.g., a color imaging system).

[0030] The polishing apparatus 100 includes: one or more carrier heads 126, each carrier head configured to carry a substrate 10; one or more polishing stations 106; and a transfer station for loading and unloading substrates from the carrier heads. Each polishing station 106 includes a polishing pad 130 supported on a worktable 120. The polishing pad 130 may be a double-layered polishing pad having an outer polishing layer and a softer backing layer.

[0031] The carrier head 126 can be suspended from the support 128 and can move between grinding stations. In some implementations, the support 128 is an overhead track and the carrier head 126 is coupled to a carriage 108, which is mounted to the track. The overhead track 128 allows each carriage 108 to be selectively positioned above the grinding station 124 and the conveyor station. Alternatively, in some implementations, the support 128 is a rotatable turntable, and the rotation of the turntable simultaneously moves the carrier head 126 along a circular path.

[0032] Each grinding station 106 of the grinding apparatus 100 may include, for example, a port at the end of an arm 134 to apply a grinding fluid 136 (such as a slurry having a grinding effect) to the grinding pad 130. Each grinding station 106 of the grinding apparatus 100 may also include a pad adjustment device to grind the grinding pad 130 and maintain the grinding pad 130 in a consistent grinding state.

[0033] Each bearing head 126 is operable to hold the substrate 10 against the polishing pad 130. Each bearing head 126 can independently control polishing parameters (e.g., pressure) associated with each individual substrate. Specifically, each bearing head 126 may include a retaining ring 142 to hold the substrate 10 below the elastic membrane 144. Each bearing head 126 also includes a plurality of independently controllable pressurized chambers (e.g., three chambers 146a-146c) defined by the membrane, which can apply independently controllable pressure to an associated area on the elastic membrane 144 (and therefore on the substrate 10). Figure 1 Only three chambers are shown for simplification; there may be one or two chambers, or four or more chambers (e.g., five chambers).

[0034] Each bearing head 126 is suspended from the support 128 and connected to a bearing head rotary motor 156 via a drive shaft 154, allowing the bearing head to rotate about axis 127. Alternatively, each bearing head 140 can vibrate laterally, for example, by driving a carrier 108 on the track 128 or by the rotational vibration of the turntable itself. In operation, the table rotates about its central axis 121, and each bearing head rotates about its central axis 127 and translates laterally across the upper surface of the grinding pad. The lateral scan is along a direction parallel to the grinding surface 212. The lateral scan can be a linear or arcuate motion.

[0035] A controller 190 (such as a programmable computer) is connected to each motor to independently control the rotational rate of the table 120 and the carrier head 126. For example, each motor may include an encoder that measures the angular position or rotational rate of the associated drive shaft. Similarly, the controller 190 is connected to actuators in each carrier 108 and / or the rotary motors for the turntable to independently control the lateral movement of each carrier head 126. For example, each actuator may include a linear encoder that measures the position of the carrier 108 along the track 128.

[0036] Controller 190 may include a central processing unit (CPU), memory and supporting circuitry (e.g., input / output circuitry), a power supply, clock circuitry, cache, etc. Memory is connected to the CPU. Memory is a non-transitory, computable, and readable medium and may be one or more readily available memories (such as random access memory (RAM), read-only memory (ROM), floppy disk, hard disk, or other forms of digital storage). Furthermore, although shown as a single computer, controller 190 may be a distributed system (e.g., containing multiple independently operating processors and memories).

[0037] An inline optical measurement system 160 is installed within the polishing apparatus 100, but does not perform measurements during the polishing operation; instead, measurement results are collected between polishing operations (e.g., when a substrate is moved from one polishing station to another (or from a transfer station or to a transfer station)).

[0038] The inline optical measurement system 160 includes a sensor assembly 161 supported at a location between two grinding stations in the grinding station 106 (e.g., between two worktables 120). Specifically, the sensor assembly 161 is positioned such that a carrier head 126 supported by a support member 128 can position the substrate 10 above the sensor assembly 161.

[0039] In an implementation where the polishing apparatus 100 includes three polishing stations and sequentially transports a substrate from a first polishing station to a second polishing station and then to a third polishing station, one or more sensor assemblies 161 may be positioned between a transfer station and the first polishing station, between the first polishing station and the second polishing station, between the second polishing station and the third polishing station, and / or between the third polishing station and the transfer station.

[0040] The sensor assembly 161 may include a light source 162, a photodetector 164, and a circuit system 166 for transmitting and receiving signals between the controller 190 and the light source 162 and the photodetector 164.

[0041] Light source 162 can be operated to emit white light. In one embodiment, the emitted white light comprises light with a wavelength of 200-800 nanometers. Suitable light sources are white light-emitting diodes (LEDs), or arrays of xenon lamps or mercury-xenon lamps. Light source 162 is oriented to guide light 168 onto the exposed surface of substrate 10 at a non-zero incident angle α. The incident angle α can be, for example, from approximately 30° to 75°, such as 50°.

[0042] The light source can illuminate a substantially linear, elongated region spanning the width of substrate 10. Light source 162 may include optics (e.g., a beam extender) to extend light from the light source into the elongated region. Alternatively or additionally, light source 162 may comprise a linear array of light sources. Both the light source 162 itself and the region illuminated on the substrate can be elongated and have a longitudinal axis parallel to the surface of the substrate.

[0043] A diffuser 170 may be placed in the path of light 168, or a light source 162 may include a diffuser to diffuse the light before it reaches the substrate 10.

[0044] Detector 164 is a light-sensitive color camera from light source 162. The camera includes an array of detector components. For example, the camera may include a CCD array. In some implementations, the array is a single column of detector components. For example, the camera may be a line-scan camera. The column of detector components may extend parallel to the longitudinal axis of the elongated area illuminated by light source 162. If light source 162 includes a column of light-emitting components, the column of detector components may extend along a first axis parallel to the longitudinal axis of light source 162. A column of detector components may contain 1024 or more components.

[0045] Camera 164 is configured with appropriately focusing optics 172 to project the field of view of the substrate onto the detector assembly 178 of the array. The field of view can be long enough to observe the entire width of the substrate 10 (e.g., 150 to 300 mm long). Camera 164 (including associated optics 172) can be configured such that each pixel corresponds to an area having a length equal to or less than approximately 0.5 mm. For example, assuming a field of view of approximately 200 mm long and detector 164 containing 1024 components, the image produced by a line-scanning camera can have pixels with a length of approximately 0.5 mm. To determine the length resolution of the image, the length of the field of view (FOV) can be divided by the number of pixels to which the FOV is imaged.

[0046] Camera 164 can also be configured such that the pixel width is comparable to the pixel length. For example, the advantage of a line scan camera is its very fast frame rate. The frame rate can be at least 5 kHz. The frame rate can be set at a frequency such that when the imaging area is scanned across the substrate 10, the pixel width is comparable to the pixel length (e.g., equal to or less than approximately 0.3 mm).

[0047] The light source 162 and the photodetector 164 can be supported on the stage 180. In the case where the photodetector 164 is a line-scanning camera, the light source 162 and the camera 164 can be moved relative to the substrate 10 so that the imaging area can be scanned across the length of the substrate. Specifically, the relative movement can be in a direction parallel to the surface of the substrate 10 and perpendicular to the detector assembly of that column of the line-scanning camera 164.

[0048] In some implementations, the stage 182 is stationary, and the carrier head 126 is moved, for example, by the movement of the carrier 108 or by the rotational vibration of the turntable. In some implementations, the stage 180 is movable while the carrier head 126 remains stationary for image acquisition. For example, the stage 180 can be moved along the track 184 by a linear actuator 182. In either case, this allows the light source 162 and the camera 164 to remain in a fixed position relative to each other as the scanned area moves across the substrate 10.

[0049] The potential advantages of a line-scanning camera and light source that move together across the substrate are, for example, that the relative angle between the light source and the camera remains constant for different positions across the wafer, compared to a conventional 2D camera. Therefore, artifacts caused by changes in viewing angle can be reduced or eliminated. Furthermore, line-scanning cameras can eliminate perspective distortion, whereas conventional 2D cameras inherently exhibit perspective distortion, which then requires image transformation to correct.

[0050] Sensor assembly 161 may include components for adjusting the vertical distance between substrate 10 and light source 162 and detector 164. For example, sensor assembly 161 may include an actuator for adjusting the vertical position of stage 180.

[0051] Optionally, polarizing filter 174 can be positioned in the path of light, for example, between substrate 10 and detector 164. Polarizing filter 184 can be a circular polarizer (CPL). A typical CPL is a combination of a linear polarizer and a quarter-wave plate. Proper orientation of the polarization axis of polarizing filter 184 can reduce haze in the image and sharpen or enhance desired visual features.

[0052] Assuming the outermost layer on the substrate is a semi-transparent layer (e.g., a dielectric layer), the color of the light detected at detector 164 depends on, for example, the composition of the substrate surface, the smoothness of the substrate surface, and / or the amount of interference between light reflected from different interfaces of one or more layers (e.g., dielectric layers) located on the substrate.

[0053] As described above, the light source 162 and the photodetector 164 can be connected to a computing device (e.g., a controller 190) that can be operated to control the operation of the light source 162 and the photodetector 164 and to receive signals from the light source 162 and the photodetector 164.

[0054] refer to Figure 2 The controller combines the individual image lines from the photodetector 164 into a two-dimensional color image (step 200). The camera 164 may include separate detector components for each of the red, blue, and green channels. The two-dimensional color image may include monochrome images 204, 206, and 208 for each of the red, blue, and green channels.

[0055] The controller can apply offset and / or gain adjustments to the intensity values ​​of the image in each color channel (step 210). Each color channel can have a different offset and / or gain.

[0056] Optionally, the image can be normalized (step 220). For example, the difference between the measured image and a standard predefined image can be calculated. For example, the controller can store background images for each of the red, green, and blue channels, and can subtract the background images from the measured images for each color channel. Alternatively, the measured image can be divided by a standard predefined image.

[0057] The image can be filtered to remove low-frequency spatial variations (step 230). In some implementations, the image is converted from a red-green-blue (RGB) color space to a hue, saturation, and lightness (HSL) color space, a filter is applied to the HSL color space, and then the image is converted back to the RGB color space. For example, in the HSL color space, the lightness channel can be filtered to remove low-frequency spatial variations; that is, the hue and saturation channels are not filtered. In some implementations, the lightness channel is used to generate a filter, which is then applied to the red, green, and blue images.

[0058] In some implementations, smoothing is performed only along the first axis. For example, the brightness values ​​of pixels 186 along the direction of travel can be averaged together to provide an average brightness value, which is a function of position only along the first axis. Then, each column of image pixels can be divided by the corresponding portion of the average brightness value, which is a function of position along the first axis.

[0059] The controller can use image processing techniques to analyze images and locate wafer orientation features 16 (e.g., wafer notches or wafer flat edges) on the substrate 10 (see image processing). Figure 4 (Step 240). Image processing techniques can also be used to locate the center 18 of the substrate 10 (see...). Figure 4 ).

[0060] Based on this data, the image is transformed (e.g., scaled and / or rotated and / or translated) to a standard image coordinate system (step 250). For example, the image may be translated so that the wafer center is located at the center point of the image and / or the image may be scaled so that the edge of the substrate is located at the edge of the image and / or the image may be rotated so that there is a 0° angle between the x-axis of the image and the radial segment connecting the wafer center and the chip orientation feature.

[0061] Optionally, an image mask may be applied to mask portions of the image data (step 260). For example, refer to... Figure 3 A typical substrate 10 contains multiple dies 12. Scribe lines 14 separate the dies 12. For some applications, it may be useful to process only the image data corresponding to the die. In this case, a reference... Figure 4The image mask can be stored by the controller, wherein the unmasked region 22 corresponds in spatial location to the die 12 and the masked region 24 corresponds to the scribing line 14. Image data corresponding to the masked region 24 is not processed or used during the thresholding process step. Alternatively, the masked region 24 may correspond to the die such that the unmasked region corresponds to the scribing line, or the unmasked region may be only a portion of each die, wherein the remainder of each die is masked, or the unmasked region may be a specific die(s) where the remainder of the die and the scribing line are masked, and the unmasked region may be only a portion of a specific die(s) where the remainder of each die of the substrate is masked. In some implementations, the user can define the mask using a graphical user interface on the controller 190.

[0062] The color data at this stage can be used to calculate a value representing the thickness (step 270). This value can be the thickness, the amount of material removed, or a value indicating the amount of progress achieved through the grinding process (e.g., compared to a reference grinding process). The calculation can be performed for each unmasked pixel in the image. This value can then be used in feedforward or feedback algorithms to control grinding parameters, providing improved thickness uniformity. For example, the value for each pixel can be compared to a target value to generate an error signal image, which can then be used for feedforward or feedback control.

[0063] Some background will be discussed to help understand the calculation of representative numerical values. For any given pixel from a color image, a pair of values ​​corresponding to the two color channels can be extracted from the color data for that pixel. Therefore, each pair of values ​​can be confined to coordinates in a coordinate space having a first color channel and a distinct second color channel. Possible color channels include: hue, saturation, lightness, X, Y, Z (e.g., from the CIE 1931 XYZ color space), red chromaticity, green chromaticity, and blue chromaticity.

[0064] refer to Figure 5 For example, when grinding begins, the logarithmic values ​​(e.g., V10, V20) are defined at initial coordinates 502 in a coordinate space 500 with two color channels. However, because the spectrum of the reflected light changes as grinding progresses, the color composition of the light changes, and the values ​​(V1, V2) in the two color channels will change. Therefore, the position of the coordinates in the coordinate space with two color channels will change as grinding progresses, depicting a path 504 in coordinate space 500.

[0065] refer to Figure 6 and Figure 7To calculate a value representing the thickness, a predetermined path 604 (step 710) is stored in the memory of controller 190 in a coordinate space 500 with two color channels. This predetermined path is generated before measuring the substrate. Path 404 can travel from start coordinate 402 to end coordinate 406. Path 404 can represent the entire polishing process, where start coordinate 402 corresponds to the initial thickness of the layer on the substrate, and end coordinate corresponds to the final thickness of the layer. Alternatively, the path can represent only a portion of the polishing process (e.g., the expected distribution of layer thickness across the substrate at the polishing endpoint).

[0066] In some implementations, to generate a predetermined path 404, the substrate is ground to approximately the target thickness to be used as the substrate in the device. An optical measurement system 160 is used to obtain a color image of the substrate. Because the grinding rate across the substrate is typically non-uniform, different locations on the substrate will have different thicknesses, thus reflecting different colors, and therefore having different coordinates in a coordinate space having a first color channel and a second color channel.

[0067] refer to Figure 8 A two-dimensional (2D) histogram is calculated using pixels contained within the unmasked area. Specifically, a scatter plot 800 is generated in a coordinate space with a first color channel and a second color channel using a color image and some or all of the coordinate values ​​of pixels from the unmasked portion of the set substrate. Each point 802 in the scatter plot is a pairwise value (V1, V2) for the two color channels for a specific pixel. The scatter plot 800 can be displayed on the monitor of the controller 190 or another computer.

[0068] As mentioned above, possible color channels include: hue, saturation, lightness, X, Y, Z (e.g., from the CIE 1931 XYZ color space), red chroma, green chroma, and blue chroma. In some implementations, the first color channel is red chroma (r) and the second color channel is green chroma (g), where the red chroma and green chroma can be derived from... and The values ​​are defined by R, G, and B, which are the intensity values ​​for the red, green, and blue channels of a color image.

[0069] The thickness path 604 can be manually generated by a user (e.g., an operator of a semiconductor manufacturing facility) using a graphical user interface integrated with a computer (e.g., controller 190). For example, when a scatter plot is displayed, the user can manually create a path that follows and covers the scatter plot by, for example, using mouse operations to click on selected points displayed in the scatter plot.

[0070] Alternatively, the thickness path 604 can be generated automatically using software designed to analyze the set of coordinates in a scatter plot and, for example, to generate a path that fits the points in the scatter plot 800 by using topology structuring.

[0071] The thickness path 604 can be provided by various functions (e.g., using a single line, a polyline, one or more arcs, one or more Bezier curves, etc.). In some implementations, the thickness path 604 is provided by a polyline, which is a set of line segments plotted between discrete points in coordinate space.

[0072] refer to Figure 6 The function provides a relationship between a position and a thickness value on a predetermined thickness path 604. For example, the controller 190 may store a first thickness value for a starting point 602 of the predetermined thickness path 604 and a second thickness value for an ending point 606 of the predetermined thickness path 604. The first and second thickness values ​​may be obtained, respectively, by using a conventional thickness measurement system to measure the thickness of the substrate layer at a position corresponding to a pixel providing the point 802 closest to the starting point 602 and the ending point 606.

[0073] In operation, the controller 190 can calculate a value representing the thickness at a given point 610 on path 604 by interpolating between a first value and a second value based on the distance along path 604 from the starting point 602 to a given point 610. For example, the controller can calculate the thickness T of the given point 610 according to the following formula:

[0074]

[0075] , where T1 is the value for the starting point 602, T2 is the thickness for the ending point 606, L is the total distance along the path between the starting point 602 and the ending point 606, and D is the distance along the path between the starting point 602 and the given point 610.

[0076] As another example, controller 190 can store thickness values ​​for each vertex on a predetermined thickness path 604 and calculate a numerical value representing the thickness for a given point on the path based on interpolation between the two nearest vertices. For this configuration, various values ​​for vertices can be obtained by using a conventional thickness measurement system to measure the thickness of the substrate layer at the location corresponding to the pixel providing the closest vertex point 802.

[0077] There may be other functions that associate the position on the path with the thickness.

[0078] Furthermore, instead of using a measurement system to measure the thickness of the substrate, the thickness value can be calculated based on an optical model.

[0079] If theoretical simulations or empirical learning based on a known “setup” of the wafer are used, the thickness value can be an actual thickness value. Alternatively, the thickness value at a given point on a predetermined thickness path can be a relative value (e.g., relative to the degree of grinding of the substrate). This latter value can be scaled down downstream to obtain an empirical value, or it can simply be used to represent an increase or decrease in thickness without specifying an absolute thickness value.

[0080] refer to Figure 6 and Figure 7 For a pixel analyzed from the image of the substrate, values ​​for both color channels are extracted from the color data for that pixel (step 720). This provides coordinates 620 in a coordinate system 600 with two color channels.

[0081] Next, the point (e.g., point 610) at coordinates 620 closest to the pixel on the predetermined thickness path 604 is calculated (step 730). In this context, "closest" does not necessarily indicate geometric perfection. The "closest" point can be defined in various ways, and limitations in processing power, the choice of search function for computational convenience, the existence of multiple local maxima in the search function, etc., can prevent a geometrically ideal determination, but still provide a sufficiently good usable result. In some implementations, the closest point is defined as a point on the thickness path 604 that is defined with respect to the normal vector of the thickness path passing through the pixel coordinates 620. In some implementations, the closest point is calculated by minimizing the Euclidean distance.

[0082] Then, as described above, a value representing the thickness is calculated from the function based on the position of point 610 on path 604 (step 740). The nearest point is not necessarily one of the vertices of the polyline. As described above, in this case, interpolation can be used to obtain the thickness value (e.g., based on simple linear interpolation between the nearest vertices of the polyline).

[0083] By repeating steps 720-740 for some or all of the pixels in the color image, a mapping of the substrate layer thickness can be generated.

[0084] For some layer stacks on a substrate, a predetermined thickness path will traverse itself, leading to a condition known as degeneracy. A degeneracy point on the predetermined thickness path (e.g., point 650) has two or more thickness values ​​associated with that point. Therefore, without some additional information, it may be impossible to know which thickness value is the correct one. However, it is possible to analyze the properties of the coordinates of a cluster of pixels associated with a given physical region on the substrate (e.g., within a given die) and use this additional information to resolve the degeneracy. For example, measurements within a given small region of the substrate can be assumed not to vary significantly and thus occupy a small segment along the scatter plot (i.e., not extending along both branches).

[0085] Therefore, the controller can analyze the coordinates of a cluster that are associated with pixels from a given physical region on the substrate and surrounding the pixel whose degeneracy needs to be addressed. Specifically, the controller can determine the principal axis of the cluster in coordinate space. A predetermined branch of the thickness path that is most closely parallel to the principal axis of the cluster can be selected and used to calculate a value representing the thickness.

[0086] return Figure 2 Optionally, uniformity analysis can be performed on a per-region basis of the substrate (e.g., each die) or on the entire image (step 280). For example, the value for each pixel can be compared to a target value, and the total number of “failed” pixels (i.e., pixels that do not meet the target value) within the die can be calculated. This total number can be compared to a threshold to determine whether the die is acceptable; for example, if the total number is less than the threshold, the die is marked as acceptable. This provides a pass / fail indication for each die.

[0087] As another example, the total number of "failed" pixels in the unmasked areas of the substrate can be calculated. This total number can be compared to a threshold to determine whether the substrate is acceptable; for example, if the total number is less than the threshold, the substrate is marked as acceptable. This threshold can be set by the user. This provides a pass / fail indication for the substrate.

[0088] If a die or wafer is determined to be "failed," the controller 190 may generate an alert or cause the polishing system 100 to take corrective action. For example, an audible or visual alert indicating that a specific die is unusable may be generated, or a data file indicating that a specific die is unusable may be generated. As another example, the substrate may be sent back for rework.

[0089] Compared to spectral processing, which typically uses pixels represented by 1024 or more intensity values, pixels in color images can be represented by only 3 intensity values ​​(red, green, and blue), and only two color channels are needed for computation. Therefore, the computational load for processing color images is significantly lower.

[0090] Generally, data can be used to control one or more operating parameters of a CMP device. Operating parameters include, for example, stage rotation speed, substrate rotation speed, substrate polishing path, substrate speed across the board, pressure applied to the substrate, slurry composition, slurry flow rate, and temperature at the substrate surface. Operating parameters can be controlled in real time and adjusted automatically without further human intervention.

[0091] As used herein, the term "substrate" may include, for example, a product substrate (e.g., a product substrate comprising multiple memory or processor dies), a test substrate, a bare substrate, and a gated substrate. The substrate may be at various stages of integrated circuit manufacturing; for example, the substrate may be a bare wafer, or the substrate may comprise one or more deposited and / or patterned layers. The term "substrate" may include disks and rectangular wafers.

[0092] However, the color image processing techniques described above can be particularly useful in scenarios utilizing 3D vertical NAND (VNAND) flash memory. Specifically, the layer stacking in the fabrication of VNAND is so complex that current measurement methods (e.g., Nova spectral analysis) may not be sufficiently reliable in detecting areas with inappropriate thickness. In contrast, color image processing techniques can offer superior processing power.

[0093] The embodiments of the present invention and all the functional operations described herein can be implemented in digital electronic circuits, or in computer software, firmware, or hardware (including the structural means and their equivalents disclosed herein, or combinations thereof). Embodiments of the present invention can be implemented as one or more computer program products, i.e., one or more computer programs tangibly implemented on a non-transitory machine-readable storage medium for execution by or for controlling the operation of a data processing device (e.g., a programmable processor, a computer, or multiple processors or computers).

[0094] The term "relative positioning" is used to indicate the positioning of components of a system relative to each other (which is not necessarily relative to gravity); it should be understood that the polished surfaces and substrates may be held in a vertical direction or some other orientation.

[0095] Some implementations have already been described. Nevertheless, it will be understood that various modifications can be made. For example...

[0096] • A camera that images the entire substrate can be used, instead of a line-scanning camera. In this case, no movement of the camera relative to the substrate is required.

[0097] • The camera can cover a width smaller than the entire width of the substrate. In this case, the camera will need to move in two vertical directions (e.g., supported on an XY coordinate stage) to scan the entire substrate.

[0098] • The light source can illuminate the entire substrate. In this case, the light source does not need to move relative to the substrate.

[0099] • The light detector can be a spectrometer, rather than a color camera; the spectral data can then be reduced to the appropriate color channels.

[0100] • Although the coordinates represented by numerical pairs in a two-dimensional coordinate space were discussed above, this technique can be applied to coordinate spaces with three or more dimensions defined by three or more color channels.

[0101] • The sensing components do not need to be located in an inline system between grinding stations or between grinding stations and transfer stations. For example, the sensor components can be located within a transfer station, within a cassette interface unit, or can be a stand-alone system.

[0102] • The uniformity analysis step is optional. For example, an image generated by applying threshold transformation can be fed into a feedforward process to adjust subsequent processing steps for the substrate, or fed into a feedback process to adjust subsequent processing steps for the substrate.

[0103] While an inline system has been described, this technique can be applied to in-situ measurements (i.e., measurements taken when the substrate is being polished). In this case, various components of the monitoring system can be mounted in recesses within a stage, and the rotation of the stage allows the components to scan across the substrate. Instead of constructing an image for in-situ measurements, the monitoring system can simply detect the color of a white beam of light reflected from points on the substrate and use this color data to determine the thickness using the techniques described above.

[0104] While this description has focused on polishing, the technique can be applied to other types of semiconductor manufacturing processes (such as etching and deposition) that can be monitored using optical methods, such as adding or removing layers.

[0105] Therefore, other implementations are within the scope of the claims.

Claims

1. A computer program product for obtaining values ​​of multiple thicknesses representing multiple locations on a substrate, the computer program product being tangibly implemented in a non-transitory computer-readable medium, the computer program product comprising instructions for causing a processor to perform the following steps: Store an image mask representing the portion of the image that will be masked; Receive color images of the substrate from a color camera; Transform the color camera to a standard coordinate system; The image mask is applied to the color image in the standard coordinate system to produce a masked color image; For each unmasked pixel in the masked color image, The coordinates of the pixel in a coordinate space containing at least two dimensions, including a first color channel and a second color channel, are determined from the color data in the color image for the pixel. The thickness of the layer on the substrate is determined based on the coordinates of the pixel in the coordinate space.

2. The computer program product of claim 1, wherein the image mask comprises a plurality of rectangular unmasked regions, the plurality of rectangular unmasked regions being separated by gaps providing a portion of the image to be masked.

3. The computer program product of claim 1, wherein the portion of the image to be masked represents a scribing line on a substrate.

4. The computer program product as claimed in claim 1, comprising the following instructions: locating wafer orientation features and wafer center.

5. The computer program product of claim 4, comprising the following instructions: rotating the color image such that there exists a [missing information] between the x-axis of the image and the radial segment connecting the wafer center and the wafer orientation feature. horn.

6. The computer program product of claim 1, wherein the first color channel and the second color channel are selected from the group consisting of color channels including hue, saturation, brightness, X, Y, Z, red chromaticity, green chromaticity, and blue chromaticity.

7. The computer program product of claim 6, wherein the first color channel is red chromaticity and the second color channel is green chromaticity.

8. The computer program product of claim 1, further comprising the following instructions: A storage function that provides a value representing the thickness as a function of the position on a predetermined path in the coordinate space; For each pixel among multiple pixels at different locations in the color image Determine the position of the point on the predetermined path that is closest to the coordinates of the pixel; and The value representing the thickness of the layer is calculated from the function and the position of the point on the predetermined path.

9. The computer program product of claim 1, further comprising the following instructions: Each image is received to acquire a series of color images smaller than the entirety of the substrate, and a complete two-dimensional color image of the substrate is generated from the series of images.

10. A measurement system for obtaining a measurement representing the thickness of a layer on a substrate, comprising: A support member for holding a substrate used in integrated circuit manufacturing; A color camera is positioned to acquire a color image of at least a portion of the substrate held by the support. as well as Controller, the controller is configured to: Store an image mask representing the portion of the image that will be masked; Receive color images of the substrate from a color camera; Transform the color camera to a standard coordinate system; The image mask is applied to the color image in the standard coordinate system to produce a masked color image; For each unmasked pixel in the masked color image, The coordinates of the pixel in a coordinate space containing at least two dimensions, including a first color channel and a second color channel, are determined from the color data of the pixel in the color image. The thickness of the layer on the substrate is determined based on the coordinates of the pixel in the coordinate space.

Images