DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Drawings
The drawings are objected to because:
In FIG 4, P230 should read “first threshold”.
In FIG 7, P310 should read “Determine second similarity coefficient” (para 67).
In FIG 8B, reference numeral 701R should be “705R” (para 70).
Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. The figure or figure number of an amended drawing should not be labeled as “amended.” If a drawing figure is to be canceled, the appropriate figure must be removed from the replacement sheet, and where necessary, the remaining figures must be renumbered and appropriate changes made to the brief description of the several views of the drawings for consistency. Additional replacement sheets may be necessary to show the renumbering of the remaining figures. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Claim Objections
Claims 1, 3, 11, 16-17, and 19 are objected to because of the following informalities:
In claim 1, “first similarity coefficient corresponds a similarity” should read “first similarity coefficient corresponds to a similarity”.
In claims 1, 11, 17, and 19:
“at least one of a first similarity coefficient and a second similarity coefficient” should read “at least one of a first similarity coefficient or a second similarity coefficient” and/or
“at least one of the first similarity coefficient and the second similarity coefficient” should read “at least one of the first similarity coefficient or the second similarity coefficient”.
In claim 3, “cropping a dummy area in the normal overlay image in each of the plurality of target overlay images and the normal overlay image” should read “cropping a dummy area in each of the plurality of target overlay images and the normal overlay image”.
In claim 16, “measurement target; ; and” should read “measurement target; and”.
Appropriate correction is required.
Claim Interpretation
Claims 8, 14, and 20 recite “at least one of a similarity between the first horizontal key and the second horizontal key”. In each case, the claim is interpreted to include both of the listed elements (elements conjoined by “and”), based on the plain English meaning and in accordance with the Federal Circuit’s 2004 Superguide Corp. v. DirecTV Enterprises, Inc. decision.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2, 4-6, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Tarshish-Shapir (U.S. Patent No. 2014/0136137 A1), hereinafter Tarshish, in view of Lee et al. (U.S. Patent No. 11,354,798 B2), hereinafter Lee, in further view of Nakagaki et al. (U.S. Patent No. 2012/0257041 A1), hereinafter Nakagaki.
Regarding claim 1, Tarshish teaches an overlay measurement device (Tarshish, para 43: “a scatterometry overlay (SCOL) metrology system, comprising target characterization module 140, at least partly embedded in computer hardware, and arranged to identify specified metrology target abnormalities using a plurality of selected metrics 110”) comprising:
a light source configured to direct an illumination to an overlay measurement target (Tarshish, scatterometry system utilizes light measurements, see examples of illumination settings in para 44 and 52, for example);
obtain a plurality of target overlay images (Tarshish, target images, para 77: “analyze targets 80 (e.g., the target images or signals received from the targets, see, e.g., FIGS. 2I and 2J) using the analyzed metric(s) and with respect to the SCOL measurements”);
a controller (Tarshish, computer, para 79: “Deriving 210, calculating 220 and/or analyzing 240, as well as any of the following stage may be carried out by at least one computer processor”) configured to: determine at least one of a first similarity coefficient or a second similarity coefficient for each of the plurality of target overlay images (Tarshish, metrics include similarity coefficient, see second similarity coefficient described at the end of the claim below, para 37: “Metrics 110 may be selected to identifier outliers, i.e., targets which diverge extremely (e.g., several .sigma.'s from the target model), and these outliers may then be removed from the measurement process”; para 63: “Outlier targets may be removed from the metrology measurements, or targets may be weighted according to the number of metrics which indicate them as outliers” see examples of high and low values in para 56); and
determine a defective overlay image from among the plurality of target overlay images based on at least one of the first similarity coefficient or the second similarity coefficient for each of the plurality of target overlay images (Tarshish, see para 63 citation above – outlier images are determined based on the metrics values; metrics are determined for a plurality of targets); and
a memory configured to store a normal overlay image and defect data corresponding to the defective overlay image (Tarshish, computer hardware, introduced in para 43 cited above, contains a memory configured to store images and defect data, see further detail in para 43),
wherein the second similarity coefficient corresponds to a similarity between a plurality of overlay keys of each of the plurality of target overlay images (Tarshish, metrics include quantified target asymmetry, which corresponds to a similarity between a plurality of overlay keys in an image – symmetry compares overlay keys in one half of the image with that in the other half of the image; a degree of asymmetry indicates less similarity between keys across the image; para 72: “quantitative estimation of the effect of such asymmetry”; see an example using box in box keys in FIG. 4A, attached below, para 76: “The non-limiting example presents box in box targets exhibiting three noise signature clusters 83A (symmetric noise signature) and 83B, 83C (reciprocally asymmetric noise signatures)”).
PNG
media_image1.png
129
297
media_image1.png
Greyscale
While Tarshish teaches a scatterometry and imaging system (Tarshish, para 43 citation above and para 87: “Certain embodiments provide an imaging tool inspection on scatterometry targets”), Tarshish fails to explicitly teach a lens assembly configured to condense the illumination at a measurement position at any one point on the overlay measurement target; and a detector configured to obtain a plurality of target overlay images through a beam reflected from the measurement position. Further, while Tarshish teaches multiple metrics for quantifying target outliers (Tarshish, para 37: “to quantify target regularity, target asymmetry and/or ROI position in the target, and thereby be used to identify exceptional targets 80 and/or divergent ROIs 85, as exemplified below. Metrics 110 may be selected to identifier outliers, i.e., targets which diverge extremely”), Tarshish fails to teach wherein the first similarity coefficient corresponds to a similarity between the normal overlay image and one of the plurality of target overlay images.
However, Lee teaches a similar system for detecting offset defects (Lee, abstract). Lee teaches a lens assembly configured to condense the illumination at a measurement position at any one point on the overlay measurement target (Lee, col 3, ln 10-13: “A condensing lens 110 is configured to condense light having been reflected by a reflector 161 and having passed through a lower aperture 108 so that the light may be transmitted to a first surface of a photomask 150”); and a detector configured to obtain a plurality of target overlay images through a beam reflected from the measurement position (Lee, col 3, ln 28-33: “transmitted light travels along the transmitted light path 100A to pass through the photomask 150 and be incident on the optical sensor 160 and reflected light travels along the reflected light path 100B to be reflected from the surface of the photomask 150 and be incident on the optical sensor 130”). Tarshish discloses a base method for capturing images using a scatterometry system, but does not specify specific methods for utilizing a lens assembly and detector. Lee teaches a known technique of capturing images using a lens assembly and detector in the same manner as claimed. A person having ordinary skill in the art, before the effective filing date of the claimed invention, could have applied the known technique, as taught by Lee, in the same way to the scatterometry system in the device of Tarshish and achieved predictable results of capturing accurate and precise image data of the overlay keys of a wafer.
Additionally, Nakagaki teaches a system/method for comparing a semi-conductor image to a non-defective image (Nakagaki, para 72 and FIG. 6). Nakagaki discloses a first similarity coefficient that corresponds to a similarity between the normal overlay image and one of the plurality of target overlay images (Nakagaki, calculated difference; comparison between a non-defective reference image and an inspection image with defects, see FIG. 6 attached below, para 72: “This method is an inspection method in which a difference between an inspection image and a reference image (a non-defective item image) is calculated and defect information is acquired from the result”). Tarshish utilizes ideal target data to characterize metrics extracted from the target images (Tarshish, para 49: “target characterization module 140 may comprise a database of metrics 110, extracted from target images and designed to reflect on the target appearance with respect to an ideal target, regarding, for example, target symmetry, target periodicity, target uniformity, noise levels etc.”); however, Nakagaki provides a method for directly comparing a target image (inspection image) to an ideal image (image without defects). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the first similarity coefficient, taught by Nakagaki, with the metrics and device of Tarshish in order to quantify how different the target pattern is from what it should look like.
PNG
media_image2.png
279
361
media_image2.png
Greyscale
Regarding claim 2 (dependent on claim 1), Tarshish in view of Lee and Nakagaki teaches wherein the controller is further configured to perform preprocessing on each of the plurality of target overlay images and the normal overlay image, and wherein the controller is configured to determine the defective overlay image based on the plurality of preprocessed target overlay images and the preprocessed normal overlay image (Nakagaki, FIG. 18a-d and para 69 – images are preprocessed to extract inspection regions 1811 to 1814; para 70: “Subsequently, an inspection method is set to the set inspection regions (S105). In this embodiment, any one of three methods, the comparison of a defect image with a reference image (in the following, image comparison)…are used for inspection methods”; see FIG. 6 where a region of the overall pattern is what is compared in both images).
Regarding claim 4 (dependent on claim 1), Tarshish in view of Lee and Nakagaki teaches wherein the controller is configured to determine the first similarity coefficient by comparing the normal overlay image with a first target overlay image among the plurality of target overlay images (Comparison taught by Nakagaki in combination with the plurality of target overlay images of Tarshish, see claim 1 rejection).
Regarding claim 5 (dependent on claim 4), Tarshish in view of Lee and Nakagaki teaches wherein the controller is configured to: compare the first similarity coefficient with a preset first threshold (Nakagaki, predetermined threshold determines defect portions – see FIG. 6 attached in claim 1 rejection, para 72: “FIG. 6 illustrates a result image 604 that binarization 603 is performed on the arithmetic operation of a differential image between an inspection image 602 and a reference image (a non-defective item image) 601 at a predetermined threshold and a portion with a large gray scale value difference is expressed in a blank and a portion with a small gray scale value difference is expressed in dark between a defect image and a non-defective item image.”); and determine the first target overlay image as the defective overlay image based on the first similarity coefficient being smaller than the preset first threshold (Nakagaki, if the difference image contains blank portions, or defect areas, para 72: “On this result image, dark defects on circuit patterns are detected as three blank portions on the result image. The number of blank portions and the areas of the blank portions, for example, are calculated from this image to obtain defect information 605”).
Regarding claim 6 (dependent on claim 4), Tarshish in view of Lee and Nakagaki teaches wherein the controller is configured to determine the first similarity coefficient by: comparing a first overlay key of the normal overlay image with a first overlay key of the first target overlay image; or comparing a second overlay key of the normal overlay image with a second overlay key of the first target overlay image (Nakagaki, first overlay key of each image is compared - see FIG. 6 wherein a first region of circuit pattern is compared in the normal/reference and target/inspection images).
Regarding claim 17, Tarshish teaches an overlay measurement method (Tarshish, see following para 84 citation) comprising: obtaining a plurality of target overlay images by repeating measurement of an overlay measurement target (Tarshish, plurality of measurements are collected, including target images, para 84: “scatterometry overlay (SCOL) measurements and further comprise analyzing the targets using the at least one analyzed metric and with respect to the SCOL measurements (stage 270), e.g. target analysis may relate to target images, target asymmetry measures, ROI parameters, target clustering or any other criterion which is derived from any of the applied metrics”) where a first overlay key on a first layer and a second overlay key on a second layer that is on the first layer are located (Tarshish, data collected for different target elements on different layers, para 38: “multiple target signals (e.g., kernels 90) may be derived from different parts of each target 80, such as target elements on different layers, inner and outer target elements etc.”). Claim 1 is the corresponding apparatus configured to perform operations equivalent to the method claimed in claim 17. Therefore, further limitations of claim 17 are met and rendered obvious by the combination of Tarshish in view of Lee and Nakagaki, demonstrated in the rejection of claim 1 above.
Claims 3, 7-9, 11-16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Tarshish in view of Lee, in further view of Nakagaki and Chen et al. (U.S. Patent No. 2013/0259358 A1), hereinafter Chen.
Regarding claim 3 (dependent on claim 2), Tarshish in view of Lee and Nakagaki teaches wherein the controller is configured to perform the preprocessing by: cropping a dummy area in each of the plurality of target overlay images and the normal overlay image (Nakagaki, areas outside the inspection region is cropped so that only inspection regions remain, para 69: “Inspection regions in (b) of FIG. 18 are four rectangular regions 1811 to 1814”), the dummy area being different from a region of interest (Nakagaki, inspection region is the region of interest); and wherein the controller is further configured to store each of the plurality of preprocessed target overlay images and the preprocessed normal overlay image in the memory (Computer storage, taught by Tarshish in claim 1 rejection, can store cropped images), but fails to teach masking a second overlay key among a first overlay key and the second overlay key of the region of interest.
However, Chen teaches a similar system/method for detecting overlay errors (Chen, abstract: “A method of determining overlay error in semiconductor device fabrication includes receiving an image of an overlay mark formed on a substrate”), disclosing masking a second overlay key among a first overlay key and the second overlay key of the region of interest (Chen, FIGs. 11-12 attached below; para 34: “a region 1102 including the unwanted feature 902 is defined. The pixel values of this region 1102 are modified…The image 1202 includes the image feature 504, which is representative of a feature formed on a second layer of the substrate. It is noted that the image 1202 no longer includes an unwanted feature (e.g., an image representing a feature formed on another layer of the substrate)”; see FIGs. 8-9 where region of interest 802 is defined). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the masking of an overlay key in the image, taught by Chen, with the images and device of Tarshish in view of Lee and Nakagaki in order to isolate one type of pattern for analysis at a time (Chen, para 35-36: “the method 300 provides a first and second image, each image representative of features of an overlay mark formed on a single layer of the semiconductor substrate (e.g., images 702 and 1202)…Thus, in block 310 an asymmetry index or indexes are calculated for each of an image representing a first layer and an image representing a second layer of the overlay mark”).
PNG
media_image3.png
386
490
media_image3.png
Greyscale
Regarding claim 7 (dependent on claim 4), Tarshish in view of Lee and Nakagaki teaches wherein the controller is configured to determine the second similarity coefficient based on a second target overlay image among the plurality of target overlay images (Tarshish, metrics are calculated for a plurality of target images, para 84: “target analysis may relate to target images, target asymmetry measures, ROI parameters, target clustering or any other criterion which is derived from any of the applied metrics”) and wherein the second similarity coefficient is determined based on a first horizontal key, a second horizontal key, a first vertical key, and a second vertical key (Tarshish, see FIG. 4A above, the pattern analyzed for symmetry has top and bottom horizontal key components and left and right vertical key components), but fails to explicitly specify wherein the second similarity coefficient is determined based on a first horizontal key, a second horizontal key, a first vertical key, and a second vertical key constituting a first overlay key in the second target overlay image (emphasis added).
However, Chen teaches a similar second similarity coefficient (Chen, para 29: “an asymmetry index (AI) is calculated for the image provided in block 308. Referring to the example of FIG. 7, one or more AI indicators are calculated from the first single layer overlay mark image 702”) that is determined based on a first horizontal key, a second horizontal key, a first vertical key, and a second vertical key constituting a first overlay key in the second target overlay image (Chen, overlay key, or second feature 504 from para 25, in the second single layer overlay mark image 1202 from para 36 and FIG. 12, attached in the claim 3 rejection above – the second feature has top and bottom horizontal key components and left and right vertical key components).
Tarshish, as demonstrated by FIG. 4A above, teaches determining the second similarity coefficient based on horizontal and vertical keys, but fails to explicitly teach that these keys constitute an overlay key in a second target overlay image. Chen discloses two single layer images that each contain an overlay key and wherein the second similarity coefficient is determined for that second target overlay image, mapped above (Chen, para 35-36). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the target overlay images of Chen with the images and device of Tarshish in view of Lee and Nakagaki in order to isolate one type of pattern for analysis at a time (Chen, para 35-36: “the method 300 provides a first and second image, each image representative of features of an overlay mark formed on a single layer of the semiconductor substrate (e.g., images 702 and 1202)…Thus, in block 310 an asymmetry index or indexes are calculated for each of an image representing a first layer and an image representing a second layer of the overlay mark”).
Regarding claim 8 (dependent on claim 7), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the controller is configured to determine the second similarity coefficient based on at least one of a similarity between the first horizontal key and the second horizontal key, and a similarity between the first vertical key and the second vertical key (Tarshish, see FIG. 4A where 83B and 83C indicate asymmetric targets, stated in para 76; degree of symmetry is based on the similarity between the left and right vertical key components and the top and bottom horizontal components, as reflected in the difference between 83B/83C and the symmetric example in the middle, 83A).
Regarding claim 9 (dependent on claim 7), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the controller is configured to determine the second target overlay image as the defective overlay image based on the second similarity coefficient being smaller than a preset second threshold (Chen, para 51: “For example, if the determined overlay error (e.g., as output from the quality indicator system 1808 and the metrology tool 1804) exceeds a given allowable threshold or control limit (e.g., for the allowable overlay error of the process), material (e.g., lot or wafer) may be held for engineering review, reworked, scrapped, and/or other suitable corrective actions”).
As demonstrated in claim 7, Chen teaches a similar second similarity coefficient to that of Tarshish; however, Tarshish fails to teach a preset threshold for the second similarity coefficient. Chen teaches the known image analysis technique of comparing values to a threshold to determine if the metric indicates the presence or absence of an abnormality. A person having ordinary skill in the art, before the effective filing date of the claimed invention, could have applied the known technique, as taught by Chen, in the same way to the images and device of Tarshish in view of Lee and Nakagaki and achieved predictable results of detecting defective overlay image areas based on the calculated similarity coefficient.
Regarding claim 11, all claim limitations are met and rendered obvious by Tarshish in view of Lee, Nakagaki, and Chen, as demonstrated by the rejection of claims 1-3 above. Claim 11 solely differs from the overlay measurement device of claim 3 (dependent on claims 1-2) in that the determining of similarity coefficients is performed on the plurality of preprocessed target overlay images (emphasis added). As demonstrated in the rejection of claim 2 above, Nakagaki teaches performing preprocessing on each of the target overlay images before further similarity calculations are performed, thus teaching determine at least one of a first similarity coefficient or a second similarity coefficient for each of the plurality of preprocessed target overlay images (emphasis added).
Regarding claim 12 (dependent on claim 11), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the controller is configured to determine the first similarity coefficient based on the plurality of preprocessed target overlay images and the preprocessed normal overlay image (Nakagaki, FIG. 18a-d and para 69 – images are preprocessed to extract inspection regions 1811 to 1814 before difference, mapped to first similarity coefficient, is calculated; para 70: “Subsequently, an inspection method is set to the set inspection regions (S105). In this embodiment, any one of three methods, the comparison of a defect image with a reference image (in the following, image comparison) …are used for inspection methods”; see FIG. 6 where a region of the overall pattern is what is compared in both images).
Regarding claim 13 (dependent on claim 11), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the controller is configured to:
determine the first similarity coefficient by comparing the preprocessed normal overlay image with a first target overlay image among the plurality of preprocessed target overlay images (Nakagaki, FIG. 18a-d and para 69 – images are preprocessed to extract inspection regions 1811 to 1814 before difference, mapped to first similarity coefficient, is calculated; para 70: “Subsequently, an inspection method is set to the set inspection regions (S105). In this embodiment, any one of three methods, the comparison of a defect image with a reference image (in the following, image comparison) …are used for inspection methods”; see FIG. 6 where a region of the overall pattern is what is compared in both images; first target overlay image taught by Tarshish, see note below);
compare the first similarity coefficient with a preset first threshold (Nakagaki, predetermined threshold determines defect portions – see FIG. 6 attached in claim 1 rejection, para 72: “FIG. 6 illustrates a result image 604 that binarization 603 is performed on the arithmetic operation of a differential image between an inspection image 602 and a reference image (a non-defective item image) 601 at a predetermined threshold and a portion with a large gray scale value difference is expressed in a blank and a portion with a small gray scale value difference is expressed in dark between a defect image and a non-defective item image.”),
determine the second similarity coefficient based on the first target overlay image among the plurality of target overlay images (Tarshish, metrics may be determined for each target, see citations in claim 1 rejection regarding the second similarity coefficient); and
compare the second similarity coefficient with a preset second threshold (Chen, para 51: “For example, if the determined overlay error (e.g., as output from the quality indicator system 1808 and the metrology tool 1804) exceeds a given allowable threshold or control limit (e.g., for the allowable overlay error of the process), material (e.g., lot or wafer) may be held for engineering review, reworked, scrapped, and/or other suitable corrective actions”; see note on the combination below).
Tarshish teaches calculating multiple metric values for each target image (Tarshish, para 37: “Metrics 110 may be selected to quantify target regularity, target asymmetry and/or ROI position in the target, and thereby be used to identify exceptional targets 80 and/or divergent ROIs 85”). In the combination with Nakagaki and Chen, determining both the first and second similarity coefficient for each target image is taught, thus also teaching the determining of the first and second similarity coefficient for the first target overlay image.
Chen teaches a similar second similarity coefficient to that of Tarshish; however, Tarshish fails to teach a preset threshold for the second similarity coefficient. Chen teaches the known image analysis technique of comparing values to a threshold to determine if the metric indicates the presence or absence of an abnormality. A person having ordinary skill in the art, before the effective filing date of the claimed invention, could have applied the known technique, as taught by Chen, in the same way to the images and device of Tarshish in view of Lee and Nakagaki and achieved predictable results of detecting defective overlay image areas based on the calculated similarity coefficient.
Regarding claim 14 (dependent on claim 13), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the controller is configured to determine the first similarity coefficient by comparing a first overlay key of the preprocessed normal overlay image with a first overlay key of the first target overlay image or by comparing a second overlay key of the preprocessed normal overlay image with a second overlay key of the first target overlay image (Nakagaki, a first overlay key of each image is compared - see FIG. 6 wherein a first region of circuit pattern is compared in the normal/reference and target/inspection images), wherein the controller is configured to determine the second similarity coefficient based on a first horizontal key, a second horizontal key, a first vertical key, and a second vertical key constituting the first overlay key in the first target overlay image (Tarshish, see FIG. 4A above, the pattern analyzed for symmetry has top and bottom horizontal key components and left and right vertical key components), and wherein the controller is further configured to determine the second similarity coefficient based on at least one of a similarity between the first horizontal key and the second horizontal key, and a similarity between the first vertical key and the second vertical key (Tarshish, see FIG. 4A where 83B and 83C indicate asymmetric targets, stated in para 76; degree of symmetry is based on the similarity between the left and right vertical key components and the top and bottom horizontal components, as reflected in the difference between 83B/83C and the symmetric example in the middle, 83A).
Regarding claim 15 (dependent on claim 13), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the controller is further configured to: determine the first target overlay image as the defective overlay image based on the first similarity coefficient being less than the preset first threshold (Nakagaki, if the difference image contains blank portions, or defect areas, para 72: “On this result image, dark defects on circuit patterns are detected as three blank portions on the result image. The number of blank portions and the areas of the blank portions, for example, are calculated from this image to obtain defect information 605”), or based on the second similarity coefficient being less than the preset second threshold.
Regarding claim 16 (dependent on claim 11), wherein the controller is further configured to: measure an overlay for the overlay measurement target (Tarshish, scatterometry overlay measurements in para 84); and generate a final target overlay image excluding the defective overlay image among the plurality of preprocessed target overlay images (Metrics with removed outliers, see note below regarding images), and wherein the controller is further configured to measure an overlay based on the final target overlay image (Tarshish, metrics may include image comparisons, see target images or partial images in para 51, and overlay measurement is based on weighted/removed metrics, para 63: “Outlier targets may be removed from the metrology measurements, or targets may be weighted according to the number of metrics which indicate them as outliers”; para 84: “method 200 may comprise weighting a plurality of metrics according to the extent the respective target abnormalities (which the respective metrics is selected to identify, stage 222) influence the corresponding metrology measurements (stage 275)”).
Regarding claim 18 (dependent on claim 17), all claim limitations are met and rendered obvious by Tarshish in view of Lee, Nakagaki, and Chen because the method steps of claim 18 are the same as claim 3, demonstrated in the rejection of claims 1-3 above.
Regarding claim 19 (dependent on claim 18), Tarshish in view of Lee, Nakagaki, and Chen teaches wherein the determining of at least one of the first similarity coefficient or the second similarity coefficient comprises: determining the first similarity coefficient based on the plurality of preprocessed target overlay images and the preprocessed normal overlay image (Nakagaki, FIG. 18a-d and para 69 – images are preprocessed to extract inspection regions 1811 to 1814 before difference, mapped to first similarity coefficient, is calculated; para 70: “Subsequently, an inspection method is set to the set inspection regions (S105). In this embodiment, any one of three methods, the comparison of a defect image with a reference image (in the following, image comparison) …are used for inspection methods”; see FIG. 6 where a region of the overall pattern is what is compared in both images; first target overlay image taught by Tarshish).
Regarding claim 20 (dependent on claim 18), Tarshish in view of Lee, Nakagaki, and Chen teaches further comprising:
comparing the preprocessed normal overlay image with a first target overlay image among the plurality of preprocessed target overlay images (Nakagaki, FIG. 18a-d and para 69 – images are preprocessed to extract inspection regions 1811 to 1814 before difference, mapped to first similarity coefficient, is calculated; para 70: “Subsequently, an inspection method is set to the set inspection regions (S105). In this embodiment, any one of three methods, the comparison of a defect image with a reference image (in the following, image comparison) …are used for inspection methods”; see FIG. 6 where a region of the overall pattern is what is compared in both images; first target overlay image taught by Tarshish, see note below); and
comparing the first similarity coefficient with a preset first threshold (Nakagaki, predetermined threshold determines defect portions – see FIG. 6 attached in claim 1 rejection, para 72: “FIG. 6 illustrates a result image 604 that binarization 603 is performed on the arithmetic operation of a differential image between an inspection image 602 and a reference image (a non-defective item image) 601 at a predetermined threshold and a portion with a large gray scale value difference is expressed in a blank and a portion with a small gray scale value difference is expressed in dark between a defect image and a non-defective item image.”); and
comparing a first overlay key of the preprocessed normal overlay image with a first overlay key of the first target overlay image or comparing a second overlay key of the preprocessed normal overlay image with a second overlay key of the first target overlay image (Nakagaki, a first overlay key of each image is compared - see FIG. 6 wherein a first region of circuit pattern is compared in the normal/reference and target/inspection images),
determining the second similarity coefficient based on a first horizontal key, a second horizontal key, a first vertical key, and a second vertical key constituting the first overlay key in the first target overlay image among the plurality of preprocessed target overlay images (Tarshish, see FIG. 4A above, the pattern analyzed for symmetry has top and bottom horizontal key components and left and right vertical key components); and
comparing the second similarity coefficient with a preset second threshold (Chen, para 51: “For example, if the determined overlay error (e.g., as output from the quality indicator system 1808 and the metrology tool 1804) exceeds a given allowable threshold or control limit (e.g., for the allowable overlay error of the process), material (e.g., lot or wafer) may be held for engineering review, reworked, scrapped, and/or other suitable corrective actions”; see note on the combination below);
wherein the second similarity coefficient is determined based on at least one of a similarity between the first horizontal key and the second horizontal key, and a similarity between the first vertical key and the second vertical key (Tarshish, see FIG. 4A where 83B and 83C indicate asymmetric targets, stated in para 76; degree of symmetry is based on the similarity between the left and right vertical key components and the top and bottom horizontal components, as reflected in the difference between 83B/83C and the symmetric example in the middle, 83A).
Tarshish teaches calculating multiple metric values for each target image (Tarshish, para 37: “Metrics 110 may be selected to quantify target regularity, target asymmetry and/or ROI position in the target, and thereby be used to identify exceptional targets 80 and/or divergent ROIs 85”). In the combination with Nakagaki and Chen, determining both the first and second similarity coefficient for each target image is taught, thus also teaching the determining of the first and second similarity coefficient for the first target overlay image.
Chen teaches a similar second similarity coefficient to that of Tarshish; however, Tarshish fails to teach a preset threshold for the second similarity coefficient. Chen teaches the known image analysis technique of comparing values to a threshold to determine if the metric indicates the presence or absence of an abnormality. A person having ordinary skill in the art, before the effective filing date of the claimed invention, could have applied the known technique, as taught by Chen, in the same way to the images and method of Tarshish in view of Lee and Nakagaki and achieved predictable results of detecting defective overlay image areas based on the calculated similarity coefficient.
Claims 10 are rejected under 35 U.S.C. 103 as being unpatentable over Tarshish in view of Lee, in further view of Nakagaki and Harada et al. (U.S. Patent No. 2014/0169657 A1), hereinafter Harada.
Regarding claim 10 (dependent on claim 1), Tarshish in view of Lee and Nakagaki teaches wherein the controller is further configured to: generate a defect map based on the defect data stored in the memory (Nakagaki, result image 604 in FIG. 6, attached in claim 1 rejection, displays defect regions where they were detected; memory taught in combination with Tarshish in claim 1 rejection); but fails to teach display coordinates of the defective overlay image and a defect type within a wafer.
However, Harada teaches a system for product defect detection (Harada, abstract: “A defect inspection method for inspecting a defect on a semiconductor wafer”). Harada teaches to display coordinates of the defective image (Harada, para 51: “And there are further provided an interface 1207 where a result of reading detection defect coordinates, as set, and hot-spot coordinates are plotted on a wafer map to be displayed, and an interface 1208 for displaying a location list of captured-image coordinates, and attributes added thereto”; see FIG. 12 attached below) and a defect type within a wafer (Harada, para 55: “output resulting from execution of the defect review, and the hot-spot inspection. It is possible to display occurrence frequency by the defect type (FIG. 17A), an occurrence tendency in a wafer plane by the defect type (FIG. 17B)”; see FIG. 17a-b attached below). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the display of defect location and type, taught by Harada, with the images and device of Tarshish in view of Lee and Nakagaki in order to aid an operator in understanding where/what type of defects occur in order to remedy them (Harada, para 56: “the use of the present invention enables a user to more quickly get hold of the occurrence frequency by the defect type, and the occurrence tendency in the wafer plane with a higher accuracy, so that the user can quickly obtain guidelines for deciding process-improvement guidelines”).
PNG
media_image4.png
409
457
media_image4.png
Greyscale
PNG
media_image5.png
504
442
media_image5.png
Greyscale
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Do (U.S. Patent No. 7,626,691 B2) teaches a similar method for inspecting overlay patterns (see abstract, col 4, ln 37-end, FIGs. 2, 4a-c).
Tsai et al. (Tsai, D. M., & Lai, S. C. (2008). Defect detection in periodically patterned surfaces using independent component analysis. Pattern Recognition, 41(9), 2812-2832.) teaches a defect detection method including comparing sub-images using a correlation value (see abstract and FIG. 9 on pg. 2828).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to EMMA E DRYDEN whose telephone number is (571)272-1179. The examiner can normally be reached M-F 9-5 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, ANDREW BEE can be reached at (571) 270-5183. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/EMMA E DRYDEN/Examiner, Art Unit 2677
/ANDREW W BEE/Supervisory Patent Examiner, Art Unit 2677