Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
  • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
  • G01N21/84—Systems specially adapted for particular applications
  • G01N21/88—Investigating the presence of flaws or contamination
  • G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
  • G01N21/956—Inspecting patterns on the surface of objects
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
  • H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor

Definitions

• the present invention relates generally to systems and methods for optical inspection, and specifically to inspection of in-fabrication patterned substrates, such as integrated circuit wafers and flat panel displays.
• Optical micro-inspection systems are commonly used for detecting microscopic defects that can occur in production of electronic devices, such as integrated circuit wafers and flat panel displays.
• Such systems typically comprise an illumination source and an image capture device, which scan over the surface of the sample under inspection in order to form a large-scale, high-resolution image of the entire surface.
• the system typically images the surface of the sample with resolution on the order of 3 - 10 ⁇ m/pixel or less, in order to detect microscopic defects that can cause device failures .
• defects are commonly identified by automatically comparing an image of each cell in the pattern to an image of another cell in the pattern.
• inspection systems include the InVisionTM and SupervisionTM systems produced by Orbotech Ltd.
• Some types of manufacturing defects occur on a macro scale, and may not be detected using automated micro- inspection techniques known in the art.
• Such macro-defects may occur, inter alia, as the result of a deformation of the wafer or flat panel substrate during manufacturing or of a process non-uniformity.
• Defects of this sort may be manifested, for example, as a slight shift in the position of a feature, such as a line, or a variation in thickness of a thin film layer over a large group of cells .
• each cell may still appear to be within the specified tolerance of the reference image.
• the macro-defect may still affect the performance of the end-product. For example, a defect of this sort in a flat panel display may cause the display to have non-uniform brightness.
• U.S. Patent 4,589,140 describes a method and apparatus for real-time inspection based on storing digital signal mask information from optical scans of objects at different magni ications but with substantially the same field of view. Digital mask information obtained from run scans of objects to be inspected at the different magnifications is compared with the stored mask information in order to identify known or unknown portions of the objects.
• U.S. Patent 4,692,943 describes a method and system for optical inspection of a two- dimensional pattern on an object, such as a printed board. A micro-inspection is performed by applying a sequence of pixel- by-pixel picture operations for inspection of dimensions and spacings . At the same time, a macro-inspection is carried out by combining the scanned pixels to frames and by reduction thereof to a single characteristic picture information. This approach is said to make possible fully automatic real-time inspection for both minute defects and macro-defects.
• an automated optical micro-inspection system is used to identify macro- defects in a patterned substrate.
• macro- defect is used to refer to defects that affect multiple, neighboring cells of the pattern, and which might be below a threshold of detectability or significance on a cell-by-cell basis.
• aspects of the present invention provide novel image processing methods that are applied to micro-inspection results, which may be generated by an existing optical micro- inspection system, in order to extract macro-defect information.
• an optical micro-inspection system acquires an image of a patterned substrate.
• the image is processed in order to detect, or isolate, a selected feature in each cell of the pattern, and to associate a numerical value with the selected feature.
• the feature may comprise a line in the pattern that appears in each cell, such as a gate or source-drain conductor in a flat panel display, and the numerical value may correspond to the relative position of the line.
• the numerical value may indicate an average grayscale intensity of the pixels in a selected area of each cell.
• the system thus assembles a matrix of numerical values, corresponding to the matrix of cells in the pattern.
• Variations in this matrix of values are indicative of non- uniformities in the selected feature over an area of the pattern that is considerably larger than a single cell, and thus may be used to detect macro-defects that are manifested by such non-uniformities.
• the system creates a synthetic image, using the numerical feature values as pixel values, which may be viewed by an operator in order to visually identify and evaluate the non-uniformities in a convenient and reliable manner.
• the synthetic image may be automatically analyzed in order to identify and evaluate macro-defects.
• apparatus for inspection including: optical inspection functionality for providing at least one optical inspection output representative of a pattern of repeating cells on a substrate; analysis functionality operative to receive the at least one optical inspection output and to isolate pre-selected features of the repeating cells; and analysis reporting functionality operative on the isolated pre-selected features of the repeating cells for providing an output indication of variations in at least one of the isolated pre-selected features, which variations occur over plural ones of the repeating cells .
• the output indication includes a display showing one or more defects that are below a threshold of detectability or significance on a cell-by-cell basis.
• the output of the analysis reporting functionality includes a synthetic image, and at least one of the analysis functionality and the analysis reporting functionality includes enhancement functionality for enhancing a visibility the variations in the synthetic image.
• the pre-selected feature includes a line in the pattern, and the output indication is indicative of a position of the line in each of the plural ones of the repeating cells.
• the preselected feature extends over a group of pixels in the optical inspection output, and the output indication is indicative of an average gray level of the group of pixels in each of the plural ones of the repeating cells.
• a method for inspection including: capturing an electronic image of an area of a substrate on which a pattern of repeating cells is formed; detecting a predetermined feature in each of a plurality of the cells in the electronic image; and analyzing a change in the predetermined feature over the plurality of the cells in order to detect a macro-defect on the substrate.
• detecting the predetermined feature includes assigning respective numerical values to the plurality of the cells responsive to the feature detected therein, and analyzing the change includes analyzing a variation in the numerical values over the plurality of the cells.
• analyzing the variation includes displaying a synthetic image including pixels corresponding to the plurality of the cells and having pixel values determined by the respective numerical values assigned to the cells.
• detecting the predetermined feature includes detecting a line in the pattern, and assigning the respective numerical values includes determining a position of the line in each of the plurality of the cells, and assigning the respective numerical values responsively to the position.
• determining the position includes finding the position of the line in each of the plurality of the cells to an accuracy of less than one pixel in the electronic image.
• the predetermined feature extends over a group of pixels in the electronic image, and assigning the respective numerical values includes determining an average gray level of the group of pixels in each of the plurality of the cells, and assigning the respective numerical values responsively to the average gray level.
• Analyzing the variation may include detecting one or more macro-defects that are below a threshold of detectability or significance on a cell-by-cell basis.
• a method for inspection including: capturing an electronic image of an area of a substrate on which a pattern of repeating cells is formed; detecting a predetermined feature in each of a plurality of the cells in the electronic image; assigning respective numerical values to the plurality of the cells responsive to the feature detected therein; generating a synthetic image including pixels corresponding to the plurality of the cells and having pixel values determined by the respective numerical values assigned to the cells; and analyzing the synthetic image in order to detect a defect on the substrate.
• generating the synthetic image includes processing the synthetic image so as to enhance a visibility of the defect in the synthetic image.
• processing the synthetic image includes enhancing the visibility of a variation in the pixel values that extends over multiple, neighboring pixels in the synthetic image. Additionally or alternatively, processing the synthetic image includes suppressing high-frequency variations in the synthetic image.
• apparatus for inspection including: an image capture device, which is adapted to capture an electronic image of an area of a substrate on which a pattern of repeating cells is formed; and an image processor, which is adapted to detect a predetermined feature in each of a plurality of the cells in the electronic image, and to analyze a change in the predetermined feature over the plurality of the cells in order to detect a macro-defect on the substrate.
• the image processor is further adapted to assign respective numerical values to the plurality of the cells responsive to the feature detected therein, and to analyze a variation in the numerical values over the plurality of the cells.
• the apparatus includes an output device, wherein the image processor is coupled to drive the output device to display a synthetic image including pixels corresponding to the plurality of the cells and having pixel values determined by the respective numerical values assigned to the cells.
• apparatus for inspection including: an image capture device, which is adapted to capture an electronic image of an area of a substrate on which a pattern of repeating cells is formed; an output device; and an image processor, which is adapted to detect a predetermined feature in each of a plurality of the cells in the electronic image, to assign respective numerical values to the plurality of the cells responsive to the feature detected therein, and to drive the output device to display a synthetic image including pixels corresponding to the plurality of the cells and having pixel values determined by the respective numerical values assigned to the cells.
• a computer software product including a computer-readable medium in which program instructions are stored, which instructions, when read by a computer, cause the computer to receive an electronic image of an area of a substrate on which a pattern of repeating cells is formed, and to process the electronic image so as to detect a predetermined feature in each of a plurality of the cells in the electronic image, to assign respective numerical values to the plurality of the cells responsive to the feature detected therein, and to analyze a variation in the numerical values over the plurality of the cells in order to detect a macro-. defect on the substrate.
• a computer software product including a computer-readable medium in which program instructions are stored, which instructions, when read by a computer, cause the computer to receive an electronic image of an area of a substrate on which a pattern of repeating cells is formed, and to process the electronic image so as to detect a predetermined feature in each of a plurality of the cells in the electronic image, to assign respective numerical values to the plurality of the cells responsive to the feature detected therein, and to display a synthetic image including pixels corresponding to the plurality of the cells and having pixel values determined by the respective numerical values assigned to the cells.
• an inspection device for inspecting an in-fabrication patterned substrate having a pattern of repeating cells, including: a camera capturing an image of cells defining the patterned substrate, the camera including pixels; feature extraction functionality extracting selected features from cells in the image; location measurement functionality measuring a location of the selected features to a sub-pixel degree of precision; and image analysis and reporting functionality analyzing measured locations for a plurality of cells, and reporting deviations from a predicted position of the selected features.
• FIG. 1 is a schematic, pictorial illustration of an optical inspection system, in accordance with an embodiment of the present invention
• FIGs. 2A and 2B are schematic side and top views, respectively, of a substrate under inspection, showing a type of macro-defect that is detectable using an embodiment of the present invention
• Fig. 3 is a flow chart that schematically illustrates a method for macro-defect detection, in accordance with an embodiment of the present invention.
• Fig. 1 is a schematic, pictorial illustration of an optical inspection system 20, in accordance with an embodiment of the present invention.
• system 20 is used to inspect a substrate 22 on which a pattern of repeating cells 24 is formed.
• substrate 22 may be a glass substrate on which a flat panel display is to be fabricated, and cells 24 may correspond to individual picture elements of the display.
• Each such picture element typically include a liquid crystal display element, control connections and transistors .
• System 20 comprises a high-intensity light source 26, which illuminates a small area of substrate 22, and an image capture device 28, which provides an optical inspection functionality so as to form an image of the illuminated area.
• light source 26 illuminates a generally elongated area, which is imaged by image capture device 28.
• Substrate 22 is mounted on a translation stage 30, which scans the substrate relative to the light source and the image capture device, so that a composite image can be acquired of an extended area of the substrate, typically covering substantially all of the substrate.
• the light source and camera may be scanned relative to the substrate.
• device 28 comprises a high-resolution solid-state imaging camera comprising a sensor array, such as a linear CCD array, a multi-line CCD array operating in a TDI mode, a CMOS matrix array operating in a memory integration mode (as described, for example, in copending U.S. patent application 10/176,003, which is assigned to the assignee of the present patent application and the disclosure of which incorporated by reference in its entirety), or other suitable imaging camera.
• the camera typically comprises suitable optics for producing a magnified image of the substrate on the sensor array. Illumination and imaging arrangements of this sort are used in many micro-inspection systems known in the art, such as the above-mentioned InVision and Supervision systems.
• system 20 may comprise illumination and imaging means of other types known in the art, such as a flying-spot laser scanner and detector, or a scanned staring array sensor capturing a plurality of images under suitable flash illumination.
• the electronic images captured by imaging device 28 are digitized and are then processed by an image processor 32.
• the image processor provides an analysis functionality, so as to identify and analyze one or more characteristics of a selected feature in each of a plurality of cells 24 (wherein the plurality may include all of the cells) .
• image processor 32 provides at least a micro-defect inspection functionality operative to detect defects in the structure of individual cells, or parts thereof.
• the image processor comprises a general-purpose computer, which is programmed in software to perform the functions described hereinbelow.
• the software may be downloaded to the computer in electronic form, over a network, for example, or it may alternatively be provided on tangible media, such as CD-ROM, DVD, magnetic media or non-volatile memory.
• image processor 32 may be carried out by dedicated hardware logic circuits, or by a programmable digital signal processor (DSP) or logic array device.
• DSP programmable digital signal processor
• Image processor 32 may be configured in this manner to detect both micro-defects and macro-defects of substrate 22.
• image processor 32 detects selected features within each of cells 24 and also provides an analysis reporting functionality so as to generate a matrix of numerical values associated with the selected feature that is detected. Exemplary methods for determining and reporting these numerical values, which may be indicative of macro-defects in the pattern on substrate 22, are described hereinbelow.
• the numerical values are typically used to generate pixel values for a synthetic image 34 that is shown on an output device, typically a display 36.
• each pixel corresponds to one or more cells on substrate 22, and the pixel value of each pixel is determined by the numerical values associated with the detected features in the one or more cells.
• each pixel value corresponds to a physical attribute associated with a cell or group of cells, for example a distance between conductors in adjacent cells.
• Image 34 may be viewed by an operator of system 20 in order to identify possible macro-defects.
• the pixel values may be further processed by processor 32 in order to automatically warn of the possible existence of macro-defects, detect the macro-defects or classify the macro-defects.
• Figs . 2A and 2B are schematic side and top views of substrate 22, illustrating creation of a macro-defect that may be detected by system 20.
• the cause of the defect is a physical deformation of substrate 22, causing a section 40 of the substrate to bow, as shown in Fig.
• the bowing in this case occurs during a photolithography step that is used to print lines 44 on substrate 22.
• a mask pattern is projected onto the substrate, as indicated by arrows 42 (wherein the light rays may be angled, as shown in the figure) .
• Lines 44 are designed to be printed at a precisely fixed spacing (generally equal to the cell width) . In section 40, however, -the bowing of the substrate causes lines 46 to be displaced.
• the displacement is typically by only a small amount, often by less than the thickness of lines 44 themselves. Moreover, the difference in displacement of lines between neighboring cells typically is negligible. It is therefore difficult to detect the displacement of the lines on a cell-by-cell basis using conventional micro-defect inspection systems. Yet even such small shifts may have a critical impact on the performance of a microelectronic device such as a flat panel display. For example, if the shift of lines 46 affects one layer of the transistor elements in the cells of the flat panel display, such as the gate layer, without affecting other layers, such as the emitter layer, the small deviation in lines 46 may cause the transistors to malfunction. In other cases, the deviation of lines 46 may cause the display to have non- uniform brightness.
• Fig. 3 is a flow chart that schematically illustrates a method for detection of macro-defects in system 20, in accordance with an embodiment of the present invention.
• the method begins with acquisition of an electronic image of substrate 22, at an image acquisition operation 50.
• the image is a gray-scale image acquired optically by camera 28.
• stage 30 translates substrate 22 in a raster pattern, so that camera 28 can acquire a scanned image covering the entire substrate, or at least covering a certain area of the substrate that is to be inspected for defects.
• Processor 32 may stitch these images together in order to provide a combined electronic image of the entire area of interest (or of the entire substrate) .
• the electronic image may be acquired at operation 50 using other types of optical imaging devices and arrangements, as noted above, or using non-optical imaging modalities, - such as electron beam imaging.
• the steps that follow may be applied to pre-acquired images, which are stored and recalled from a memory for this purpose.
• Image processor 32 processes the digitized electronic image in order to detect a selected feature in each of a plurality of cells 24 on substrate 22, at a feature detection operation 52.
• the feature is assumed to be a certain line, such as lines 44, and processor 32 detects a deviation in the position of the line in each of the cells.
• a feature may be identified by a selected characteristic, such as a gray level of pixels imaging the feature. This ensures that only a selected feature of a cell is detected and other features in a cell, such as orthogonal lines, transistors, ITO electrodes and the like, are not detected.
• processor 32 may detect other characteristics, such as a critical dimension of a feature, for example a line width.
• the selected feature to be a line
• the characteristic indicative of a macro-defect to be a deviation in the line position with respect to an expected location various methods may be used to accurately detect the line and determine line position with sub-pixel precision.
• the shape of the line in the electronic image of each cell may be correlated with a template representing the expected shape of the line.
• the displacement of the template that gives the maximal correlation value represents the actual position of the line (which may be found in this manner to high precision, limited, for example, only by the signal/noise ratio of the electronic image) .
• an edge detection filter may be applied to the line
• a fitting algorithm such as a mean square fit
• the line position may be interpolated based on the gray-scale pixel values measured along a cross-section of the line. For instance, assuming that the line in question appears as a relatively thin bright band against a darker background, with the highest gray-scale value G (n) measured at pixel n, the interpolated line position n' may be calculated as:
• n- ⁇ and n+1 respectively indicate the pixels preceding and succeeding pixel n along the cross-section of the line.
• n is an integer
• n' is typically a non- integer rational value, which may be represented as a floating point number.
• the interpolated line position may be determined in this manner at a number of points along the length of the line within a cell 24, and the results averaged in order to arrive at a final representation of the line position.
• Other methods for sub-pixel measurement of the location of the line of interest will be apparent to those skilled in the art .
• the feature characteristics detected at operation 54 are typically corrected to remove non-uniformities that are due to artifacts (rather than to actual non-uniformities in the pattern on substrate 22), at an artifact removal operation 54.
• the optics of camera 28 may introduce slight magnification non-uniformities, which cause deformation of the electronic image.
• certain detector elements in the camera may deviate from normal sensitivity. These optical effects can generally be mapped in advance, using special calibration targets.
• the speed of stage 30 may not be uniform as it translates substrate 22 over the raster pattern, resulting in small deviations in the positions of the pixels in the electronic image.
• processor 32 Based on the features detected at operation 52 (and corrected at operation 54) , processor 32 assigns at least one numerical value to each cell, at a numerical assignment operation 56.
• the numerical value is indicative of a measured characteristic of the feature in question. In the case of feature position measurements, as described above, the numerical value represents the precise, sub-pixel position of the feature (for example, the line) .
• the numerical value is based on the difference between the actual position and the expected position of the feature, or between the actual position of the feature and the edge of the cell or the actual positions of the feature in neighboring cells.
• the numerical value could' be n' -N* cell_size .
• the variations of interest in the line position are generally much smaller than the cell size - typically as much as seven orders of magnitude smaller. It is therefore important at operation 56 to use a sufficient number of bits in the computation of the numerical values in order to assure that small but significant variations are preserved accurately.
• Processor 32 combines the individual numerical values calculated for each cell at operation 56 in order to make up an array of values, at a synthetic image generation operation 58.
• the synthetic image can be formed such that each pixel corresponds to each cell, and a pixel gray level value for each pixel is a function of the numerical value determined with respect to the feature in each cell .
• the pixel value for each cell is given by the line position determined " in that cell.
• the pixel value for each cell may be equal to an average of the gray-level values in the electronic image for a selected feature in each cell.
• the selected feature be an ITO electrode in each cell. Macro-defects may result from variations in the layer thickness of the ITO electrode, which may be manifested by variations in the average gray-level value .
• Processor 32 may analyze the array of values in the synthetic image in order to warn of, detect or classify macro- defects autonomously.
• the synthetic image may be presented in the form of image 34 on display 36.
• the processor typically applies image enhancement techniques to the synthetic image in order to improve the visibility of pixel-to-pixel variations in the synthetic image that extend over multiple, neighboring pixels. This sort of variations is most often indicative of macro- defects. If substrate 22 has been fabricated perfectly, so that all cells are precisely uniform, the synthetic image will be uniformly gray. In practice, however, at least small pixel-to-pixel variations are to be expected.
• a low-pass filter or spatial down-sampling may be applied to the synthetic image in order to remove noise due to insignificant local variations in the feature of interest.
• techniques such as histogram processing and pseudo-coloring may be applied in order to spread apart and make visible small but consistent variations in the numerical values of the pixels in the synthetic image.
• a human operator typically views the enhanced synthetic image 34 on display 36, at a visual inspection operation 60.
• the operator notes variations in the gray-scale or color pixel values in image 34 in order to identify the locations of macro-defects on substrate 22. Additionally or alternatively, processor 32 may extract this information automatically, as mentioned above .
• processor 32 may extract this information automatically, as mentioned above .
• the description above makes reference to detection of certain specific features and characteristics of the cells in production of microelectronic devices, and in particular to cells of a flat panel display, the principles of the present invention may similarly be applied to find macro- defects of other types, not only on microelectronic device substrates, but also on other sorts of patterned samples defined by periodic patterns.

Landscapes

• Engineering & Computer Science (AREA)
• Biochemistry (AREA)
• Pathology (AREA)
• Analytical Chemistry (AREA)
• Power Engineering (AREA)
• Physics & Mathematics (AREA)
• Health & Medical Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Chemical & Material Sciences (AREA)
• Microelectronics & Electronic Packaging (AREA)
• Manufacturing & Machinery (AREA)
• General Physics & Mathematics (AREA)
• General Health & Medical Sciences (AREA)
• Immunology (AREA)
• Computer Hardware Design (AREA)
• Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
• Image Analysis (AREA)
• Length Measuring Devices By Optical Means (AREA)
• Testing Or Measuring Of Semiconductors Or The Like (AREA)
• Image Processing (AREA)

Priority Applications (2)

Applications Claiming Priority (2)

Publications (2)

Family

ID=33027706

Family Applications (1)

Country Status (5)

Cited By (3)

Families Citing this family (14)

Citations (3)

Family Cites Families (7)

• 2003
  • 2003-03-17 JP JP2003071791A patent/JP3948728B2/ja not_active Expired - Lifetime
• 2004
  • 2004-03-17 CN CN2004800123171A patent/CN1839306B/zh not_active Expired - Lifetime
  • 2004-03-17 KR KR1020057017277A patent/KR101146081B1/ko active IP Right Grant
  • 2004-03-17 WO PCT/IL2004/000253 patent/WO2004083901A2/en not_active Application Discontinuation
  • 2004-03-17 TW TW093107098A patent/TWI318680B/zh not_active IP Right Cessation

Patent Citations (3)

Also Published As

Similar Documents

Legal Events