Apparatus and methods for inspecting reticles

APPARATUS AND METHODS FOR INSPECTING RETICLES

CROSS-REFERE^NCE TO.RELATED.API*LICAJIO.NS

1'0001] This application claims the benefit of priority of U.S. Patent Application No.

15/803628 filed 3 Nov 2017 by Rui-fimg Shi et al. This application also claims the benefit of priority of U.S. Provisional Application No. 62/508,369 filed. 18 May 2017. These applications and patent are herein incorporated by reference in their entirety for all purposes.

TECHNICAL HELD OF THE INVENTION

[0002] The invention generally relates to the field of reticle inspection. More particularly the present invention relates to pattern qualification.

BACKGROUND

100031 Generally, the industry of semiconductor manufacturing involves highly complex technique's for fabricating integrating circuits using semiconductor materials which are layered and patterned onto a substrate, such as silicon. Due to the large scale of circuit integration and the decreasing size of serniconduciot devices, the fabricated devices have become increasingly sensitive to defects. That is, detects which cause faults in the device are becoming increasingly smaller. The device is fault free prior to. shipment to the end users or customers.

[0004] An integrated circuit is typically fabricated from. a plurality of reticles. Initially, circuit designers provide circuit pattern data, which describes a particular integrated circuit (IC) design, to a reticle production system, or reticle writer. The circuit pattern 25 data is typically in the form of a representational layout of the physical layers of the fabricated IC device. The representational layout includes a representational layer for

each physical layer of the IC device (e,g., gate oxide, polysilicon, metallization, etc,), wherein each representational-layer is composed of a plurality of polygons- that define a layer's patterning of the particular IC device. The reticle writer uses the circuit pattern data to write (e.g., typically, an electron beam writer or laser scanner is used to expose -a 5 reticle pattern) a plurality of reticles ihat will later be used to fabricate the particular IC design.

[0005] Some reticles or photomasks are in the form of ah optical element containing at least transparent and opaque regions, semi-transparent and phase shifting regions, or absorber and reflective regions, which together define the pattern of coplanar features in 10 an electronic device such as an integrated circuit. Reticles are used during photolithography to define specified regions of a semiconductor wafer for etching, ion implantation, or other fabrication processes.

[0006] .After fabrication of each reticle or group of reticles, each, new reticle typically is: qualified for use in wafer fabrication. For example, reticle patterns need to be free of 15 printable defects. Additionally, any wafer that is fabricated with the reticle needs to be free of defects. Thus, there is a continuing need for improved reticle and wafer inspection and qualification techniques.

SUMMARY

10007} The following presents a simplified summary of the disclosure in order to provide a basic understanding of certain embodiments of the invention. This summary is not an extensive overview of the disclosure and it does not identify key/critical elements of the invention or delineate the scope of the inven tion. Its sole purpose is to present some concepts disclosed herein in a simplified form as a prelude to the more detailed description that is presented later.

[0008] In one embodiment, a method of qualifying a photolithographic· reticle is disclosed. An imaging tool is used to acquire a plurality of images at different 10 illumination configurations andfor different imaging configurations from each of a plurality of pattern areas of a test reticle. A reticle near field is recovered for each of the pattern areas of the test reticle based on the acquired Images from each pattern area of the test reticle. The recovered reticle near field is then used to determine whether the test reticle or another reticle will likely result in unstable wafer pattern or a defective wafer.

jooooi In one implementation, the reticle near field is directly analyzed to determine whether the test reticle or another reticle will likely result in an unstable wafer pattern or a defective wafer. In another aspect, the recovered reticle near field is used to detect defects in the test: reticle or in simulated wafer images that are simulated from the recovered reticle near field, wherein defect detection Includes comparing intensity and/or 20 phase for a same die at different times, for adjacent dies, for a die and its corresponding golden, die, or for a die and a corresponding die from a reticle copy with identical design to the test reticle.

jooio} In one .aspect, the images are acquired at a field plane or a pupil plane. In a specific embodiment, the reticle near field is recovered without using a design database 25 that was used to fabricate the reticle. In another aspect, the acquired images include at least three refieetivedransmissive images that are acquired at different imaging conditions that are selected to result in a same reticle near field. In this aspect, the different imaging conditions include different focus settings and different pupil shapes, and the different

illumination conditions include different source intensity distribution and/or polarization settings.

[0011] In an alternative implementation, the method includes (i) applying a lithography model to the reticle near field for the test reticle to simulate a plurality of test wafer 5 images and (ii) analyzing the simulated test water images to determine whether the test reticle will likely result in an unstable or defective wafer. In this aspect, the lithography model is configured to simulate a photolithography process. In a further aspect, thelithography model simulates an illumination. source having a different shape than an illumination shape of an. inspection tool for acquiring images of the test reticle or another 10 reticle or wafer, hi another aspect, the lithography model is calibrated with images rendered from a design database fbr a calibration reticle. In another example, the lithography model is calibrated with images acquired from a calibration reticle. In yet a further aspect, the lithography model is applied to the reticle near field, which was recovered tor the test reticle, under a plurality of different lithograph) proce» conditions. 15 and analyzing the simulated test wafer images includes determining whether the test reticle will likely result in an unstable wafer under the different lithography process conditions by comparing portions of the simulated test images associated with different process conditions and a same reticle area.

[0012] In an alternative embodiment, the invention pertains to an inspection system for 20 qualifying a photolithographic reticle. The system includes a light source for generating an incident beam and an illumination optics module fbr directing the incident beam onto a reticle. The system also includes a collection optics module for directing an output beam from each pattern area of the reticle to at least one sensor for detecting the output beam and generating an image or signal based on the output beam. The system farther 25 comprises a controller that is configured to perform operations that are similar to one or more of the above described method operations.

[0013] These and other aspects of the invention, are described further below with reference to th© figures.

BRIEF DESCRIPTION OF THE DRAWINGS

[0014| Figure 1 is a flaw chart illustrating a mask near field recovery procedure in accordance with one embodiment of the present invention.

[0015] Figure 2 is a flow chart illustrating a model calibration process in accordance 5 with a specific implementation of the presentation invention.

[0016] Figure 3 illustrates a flow chart representing a reticle qualification process in accordance with one embodiment of the present. invention.

[0017] Figure 4A. is a flow chart illustrating a process for determining reticle pattern stability in accordance with an example application of the present invention,

[0018] Figure 4B is a flow chart illustrating a defect inspection, procedure in accordance with an another embodiment of the present invention.

[0019] Figure 5 is a flow chart illustrated a reticle qualification process that is applied to a recovered .mask near field image or results in accordance with an alternative embodiment of the present invention.

[0020] Figure 6 is a diagrammatic representation of an example inspection system in which techniques of the present invention may be implemented.

[0021] Figure 7A is a simplified schematic representation of a lithographic system for transferring a mask pattern from a photomask onto a wafer in accordance with certain embodiments.

[0022] Figure 7B provides a schematic representation of a photomask inspection apparatus in accordance with certain embodiments.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

J0023| In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. The present .invention may be practiced without some or all of these specific details. In other instances, well known 5 process operations or apparatus components have not been described in detail to not unnecessarily obscure the present invention. While the invention will be described in conjunction with the specific embodiments, it will be understood that it is not intended to limit the invention to the embodiments.

[00241 It would be beneficial to detect each mask’s defects and otherwise characterize ID various aspects of the mask (e.g. pattern stability, CD, CD uniformity) prior to shipping masks to a fabrication facility, prior to fabricating wafers using such mask for fabrication, and/or for periodic requalificafion of such mask after such mask has been used for certain periods of time in the fabrication process,

[00251 One embodiment of the present invention includes techniques for recovering 15 near field images of a reticle based on reticle images obtained from an inspection tool under a plurality of different imaging parameters. This reticle near field image may then be used in a number of reticle qualification applications. In one example, the reticle near field image may be input to a lithography model to predict a wafer image or various wafer pattern characteristics related to how the resulting, patterns will be printed on the 20 wafer. The predicted wafer image and/or various wafer characteristics may then be analyzed for detect detection, reticle qualification or requalifieation, and/or any other suitable metrology or inspection applications. The reticle near field image may also itself be analyzed for various purposes as further described herein.

[0026| The terms reticle”, “mask5', and “photomask” are used herein interchangeably 25 and generally each may include a transparent substrate, such as glass, borosilicate glass, quartz, or fused silica having a layer of opaque material formed thereon. The opaque (or substantially opaque) material may include, any suitable material that completely or partially blocks photolithographic light (e.g., deep UV or extreme UV). Example materials include chrome, molybdenum silicide (MoSi), tantalum silicide, tungsten

silicide, opaque MoSi on glass (OMOG), etc. A polysilicon film may also be added between the opaque.layer and transparent substrate.to improve adhesion. A low reflective film, such as molybdenum oxide (MoO>), tungsten oxide (WO2), titanium oxide (T1O2), or chromium oxide (CrCh) may be formed over the opaque materiaL In a specific 5 example, an EIJV reticle may include multiple layers having alternating layers of different refractive Indices with low absorption characteristics (such as molybdenum (Mo) and silicon (Si)) and an absorber materials (such as a tantalum boron nitride film topped by a thi n, anti-reflecti ve ox ide).

[0027] The term reticle refers to different types of reticles including, but not limited to, 10 a clear-field reticle, a dark-field reticle, a binary reticle, a phase-shift, mask (PSM), an alternating PSM, an attenuated or halftone PSM, a ternary attenuated PSM, a chromeless phase lithography PSM, and chromeless phase lithography (CPL). A clear-field reticle has field or background areas that are transparent, and a dark-field reticle has field or background areas that are opaque. A binary reticle is a reticle having patterned areas that 15 arc either transparent or opaque. Fur example, a photomask made from a transparent fused silica blank with a pattern defined by a chrome metal adsorbing film can be used.

Binary reticles are different, from phase-shift masks (PSM), one type of which may include films that only partially transmit light, and these reticles may be commonly referred to as halftone or embedded phase-shift masks (EPSMs), such as ArF and KrF 20 masks. If a phase-shifting material is placed on alternating clear spaces of a reticle, the reticle is referred to. as an alternating PSM, an ALT PSM, or a Levenson PSM. One type of phase-shifting material that is applied to arbitrary .layout patterns is referred to as an. attenuated or halftone PSM, which may be fabricated by replacing the opaque material with a partially transmissive or halftone film. A ternary attenuated PSM is an attenuated 25 PSM that includes completely opaque features as welt

[0028J The next generation lithography has ushered in the use of extreme ultra-violet radiation (EUV, wavelength 13,5 nm), which is absorbed in normal atmosphere as well as in glass. For this reason, lithography EUV processes take place under vacuum, and optical reflective lenses.'mlrrors are used for focusing onto the EUV photomasks, which 30 will have reflective and absorber patterns, instead of translucent and opaque patterns.

[002θ| Figure 1 is a flow chart illustrating a mask near field recovery procedure 100 in accordance with one embodiment of the present invention. The following mask recovery process 100 may be performed for a particular reticle or set of reticles at any suitable time in a reticle’s lifecycle as further described below in various use cases for the recovered 5 mask near field. By way of examples, a mask near field may be recovered prior to fabrication of any wafers with such reticle(s), prior to commencement of high volume nafcr fabrk-afiun, or daring requalideation of such rcticlepO(00301 Initially, at least three images of die mask are acquired at different imaging configurations using a mask inspection tool in operation 102. Alternatively, two images 10 may be used, but using three images has been found to work well. Acquisition with different imaging configurations may be simultaneous or sequential. The acquired images do not have to be at the field planes. By way of examples, the two or more images can be acquired at the pupil planes at which the diffraction intensity may be directly accessed.

[0031] Various suitable combinations of illumination and/or collection configurations

1.5 may be utilized for acquiring the two or more images. The different imaging configurations are generally selected to provide images from which the mask near field may be calculated. Any suitable imaging or optical configurations may be selected so that the mask near field remains the same under the different operating conditions. Examples, include different focus settings, different illumination shapes (e.g. different directions or patterns), different polarization for the entire illumination pupil or different parts of the illumination pupil, different apodization settings to obscure different portions of the collection beam, etc. In one embodiment, different focus· settings, through focus and defocus (such as 0 focus, ±800 or ±1600 defocus, etc.), may be used to acquire the different images. In another example, different quadrants of the illumination pupil may have different polarization settings. In another-example, the imaging configurations may include high resolution images, such as transmitted images (&g., for ArF masks) with different pupil shapes and/or different focal conditions. In another embodiment, three or more reflected images with different pupil shapes and/or different focal conditions may be obtained (eg., for EUV masks).

[0032] The reticle may be imaged at a “substantially low resolution” using a relatively low NA (e.g,, less than 0.5), In contest* a “substantially high resolution image” generally refers to an image of a reticle in which features printed on the reticle appear substantially as they are formed on the reticle (within the optical limitations of the reticle inspection 5 .system used to generate the image). A “substantially high resolution image” of a reticle is an image that is generated by imaging the physical reticle at the reticle plane with a substantially high resolution, reticle inspection system (e.g., a numerical aperture (NA) of greater than 0.8). The “substantially low NA” used to generate a reticle image may be substantially the same as the NA on the reticle side that is used by an 10 exposure/lithography system to project an image of the reticle onto a wafer, thereby transferring features of the reticle onto the wafer. In the substantially low NA image (or LNI), the reticle features may have a substantially different appearance than the actual reticle features. For example, reticle features may appear to have more rounded comers in an LNI of a feature than the actual feature that is formed on the reticle.

[0033] In general, any suitable imaging tool may be used for the mask near field recovery' process. In certain embodiments described herein, the results of an initial recovery' process may be later used for pattern, stability or defect detection assessments with respect to the same reticle or other reticles based on additional reticle images from a particular inspection tool. For consistency in these use cases, the images of the reticle lor 20 mask near field recovery may be acquired with the detector of the reticle inspection system that will be used for subsequent inspection of the same or other reticles or acquired with a similarly configured detector of a similarly configured reticle inspection system (e.g., a different reticle inspection system of the same make and model as the reticle inspection system that will be used for inspection). In other words, the images that 25 may be used for mask recovery may be acquired under the same optical conditions as wil l be used during subsequent mask inspection, or qualification processes. In this manner, the interaction of the reticle with the illuminating electromagnetic waves of the inspection system may be measured as directly as possible.

[0034] In alternative embodiments, the tool used for mask. near field recovery may differ from a reticle inspection system. For instance, the imaging tool may utilize the same

wavelengths (e.g. wavelength (I93.3nm for DUV or 13.5nm for EUV) as.the lithography system in which the reticle will be used for wafer manufacturing. In fact, any suitable electromagnetic wavelength, may be used for the mask near field recovery .

[0035] Referring back to the illustrated example, the three or more images may then be 5 aligned with each other or each image may be aligned to the post-OPC database in operation 104. For instance, the acquired images may be aligned via spatial-domain or frequency-domain methods. Alignment adjustments may depend oh specific geometries of the imaging system being used. If different images are obtained using different collection paths, some adjustment of the images can be made to compensate for 10 differences in optical paths.

[0036] In the imaging tool, a reticle having various patterns is illuminated by electromagnetic (EM)-waves that are incident from many directions. This incident light is diffracted from different points of the mask pattern at different electromagnetic field phases which interfere with each other differently. The near-field. of the reticle is the 15 electromagnetic field at a proximate distance of a few wavelengths from the reticle.

100371 The collection opties generally directs a diffracfioii-limited portion of the light from the reticle towards a detector (or wafer) to form an image. The detector detects intensity which is the result of interference due to the mask near field, but does not detect the phase.

[0038] Although far-field intensity is obtained in the detected signals, it is desirable to recover the mask near field, which includes amplitude and phase. In the illustrated embodiment, the mask near field is recovered and stored based on such acquired mask images, as illustrated in operation 106. Multiple images (or signals) are generally used to recover the mask near field, which includes both phase and amplitude components. The 25 near-field data may be determined by a regression technique based on the images acquired from the reticle. For example, the near-field of a selected portion of the reticle can be recovered (regressed), using a quasi-Newton or conjugate gradient technique, from its acquired optical images or intensity of images recorded at a detector plane. In

addition, any other suitable regression method and/or algorithm may be used to determine the near*field data from the one or more actual images.

[0039] Mask near field recovery can generally be achieved by solving an optimization problem that seeks to minimize -the difference between observed .intensity images and resulting images of the assumed mask optical field. In particular, recovering the near-field of a reticle from its intensity images is an inverse problem or a regression, problem. The near-field can be recovered iteratively by minimizing a cost function (e.g., energy or penalty .function). The quantity that is minimized can. be the sum of squared differences, between the acquired images and intensity images at the detector that are calculated from

1.0 the mask near-field. In other words, intensity images can be calculated from the final mask near field for various sets of optical system properties, and these calculated images will most closely match the acquired images when the mask near field is found. Various mask near field recovery methodology and system embodiments are described further in U.S. Patent No. 9,478,019 issued 25 October 2016 by Abdurrahman Sezginer et al., which patent is incorporated herein by reference in its entirety for ali purposes.

10040] In the case where multiple images are acquired under various optical conditions, the recovered near field mask m, which carries the phase and amplitude information, can be determined by the following equation:

[0041[ rn'^argmin Im P'f Equation I

[0042] In. the above Equation I, la is the measured image for imaging condition a, ^/e) is a set of eigenvectors describing the inspection imaging system, A*if) is a set of corresponding eigenvalues for the imaging system, and c., is a non-negative weighting factor between 0 and 1. The above equation can be solved iteratively through, for example, methods such as quasi-Newton or conjugate gradient.

[00431 Another example is the Gerchberg-Saxton algorithms in which a combination of field plane images and pupil plane diffraction orders can be utilized to solve both the amplitude and phase of the object.

[0044] In one embodiment, the mask near field may be determined based on the acquired images via a Hopkins approximation. In another embodiment, the regression does not include thin-mask approximations. For example, the near-field of the -.reticle is the electromagnetic field that is calculated to be present near the surface of the reticle 5 when it is illuminated by a nonnally-mcident plane wave. In. lithography and inspection.

a reticle: is illuminated by plane-waves that are incident from many directions. When the direction of incidence changes, according to the Hopkins approximation, the directions of the dlftfaction orders change but their amplitudes and phases remain approximately unchanged. Hie embodiments described herein can use the Hopkins’ phase approximation but do not make the so-called thin-mask or Kirchhoff approximations.

[0045] The recovery formulation can also be varied with different norms or the addition of a mgiihiri/aiion term λ', w Inch penalizes oscillations m the near Ikdd. as foilon s:.

[0046] m ~ arg min Z^R^Z^h^^f « L where regularization term R, can incorporate prior information about the near field or expectations based on the physical understanding of the mask substrate/material. In addition, the norm used for image difference can be a knorm and adjusted based on specific needs of the optimization function.

[0047] As an interesting note, the interference of the mask electromagnetic field vectors as a result of a higher NA will be greater (than a lower NA inspeetion system) due 20 to the wider range of incident angles of light and associated interfering electric field components for a higher NA.

10048] The actual mask may vary from the intended design patterns due to the mask writing process. Obtaining the near field mask from images of the mask means that such near field mask is obtained from the actual physical mask, rather than the design database. That is, mask near field may be recovered without using the design database.

[0049] The mask near field results can then, be used in various applications. In one embodiment, mask near field results can be used to predict wafer patterns using one or

more models. That is, the recovered mask near field may be used to simulate lithography images. Any suitable technique may be: utilized to simulate lithography images based on the mask, near field images. One enfoodmient includes computation of the lithography image through the Partial Coherence Model:

[0050] ~^(x;/,z)|

Equation 2

[0051] where λί represents the Eigen values of the lithography TCC (transfer cross coefficients); y/* () represents Eigen vectors (kernels) of TCC; s is the wafer stack, including film refractive indices', f is focus; and ,r is vertical position of the lithography plane in the resist material. The transfer cross coefficients (TCC) of Equation 2 may 10 include vectorial propagation of the field through the lithography projector including the film stack on the wafer.

[0052] Prior to using a model for predicting wafer results, the model may be calibrated to produce as accurate results as possible. The model may be calibrated using any suitable technique. Certain embodiments of the present invention provide techniques for 15 calibrating a lithography model based on mask, near field results that are recovered from a calibration mask. In alternative embodiments, the design database is used to calibrate the model. For instance, calibration reticle images may be rendered from the design database.

|0053|A calibration reticle will typically be designed to have characteristic(s) that are substantially similar to the reticle to be inspected for defect detection or to be measured 20 for metrology purposes. For example, the calibration reticle and the test reticle, are preferably formed from substantially the same materials having substantially the same thicknesses and compositions. In addition, the two reticles may have been formed using the same processes. The two reticles may not necessarily have the same patterns printed thereon as long as the patterns on the reticles can be broken up into segments that are 25 substantially the same (e.g., lines having similar widths, etc.). In addition, the reticle that will be inspected and the reticle that is used to acquire the images may be one and the same reticle.

[0054] Figure 2 is a flow chart illustrating a model calibration process 200 in accordance with a specific implementation of the presentation invention. As shown,, the photolithography process and photoresist can be modeled as applied to mask near field images (201) recovered from a calibration reticle using an initial set of model parameters 5 in operation 208. Alternatively^ the calibration process 200 may use simulated calibration reticle images ¢202) that are simulated from the design database. Reticle, images may be rendered from the database by simulating the reticle fabrication and imaging processes on the design database. Any suitable model may be used to generate optical images for the features of the design database. By way of example, such simulation may include using 10 Sum Of Coherent Systems (SOCS) or Abbe methodologies described herein. Several software packages exist that can simulate intensity images of an optical system from a known design database. One example is Dr.LiTHO developed at Fraunhofer IISB in Frlangen Germany. In the case of simulating an image from design database 202, the near field may be simulated first, which may be done by the software package referenced 15 above, as well as several other packages, including Proiith by KLA-Tencor, HyperLith by

Panoramic Technologies, among others.

[0055| The model tor producing wafer images based on reticle near field images may include just the effect of the photolithography scanner, and it may also include the effect of resist, etch, CMP or any other wafer processes. One example process simulation 20 model tool is Proiith available from KLA-Tencor Corp, of Milpitas, CA. Resist and etch processes can be modeled rigorously or approximately. In a specific embodiment, the model may be in the form of a compact resist model that includes 3D acid diffusion inside a particular resist material and configuration, with boundary conditions imposed therewith, as well as a single threshold being applied to form the latent image.

[0056] It is noted that the modeled lithography tool may have a different illumination shape or source as the reticl e inspection tool for acquiring actual images of the reticle. In certain embodiments, the modelled lithography tool may have a same or similar source as a reticle inspector tool.

|0057| Other simulation approaches such as SOCS or Abbe may be used. The algorithm 30 generally known as the Sum Of Coherent Systems (SOCS) attempts ίο convert the

imaging system into a bank of linear systems whose outputs are squared, scaled and summed. The SOCS method has been described elsewhere, including in the Ph.D. thesis of Nicolas Cobb, “Fast Optical, and Process Proximity Correction Algorithms for Integrated Circuit Manufacturing”, University of California, Berkeley, Spring, 1998. The 5 Abbe algorithm includes calculating the image of the object for each point source, one at a time, and then summing up the intensity images together and taking the relative intensity of each source point into consideration.

[00581 The input for the model and its modeling parameters includes a set of process conditions that are applied to a recovered near field mask. That is, the model is 10 configured to simulate different sets of process conditions on the reconstructed near field mask (or-simulated mask image). Each set of process conditions generally corresponds to a set of wafer manufacturing process parameters that characterize, or partially characterize the wafer process for forming a wafer pattern from the mask. For example, a particular setting of focus and exposure can be input to the model. Other adjustable 15 model parameters may also include one or more of the following parameters: a projection lens wavefront parameter, an apodization parameter, a chromatic aberrations focus error parameter, a vibration parameter, a resist profile index, a resist scum metric, top loss metric, etc. Use of such .model with -different sets of process conditions can result in a set of simulated wafer or resist pattern, images formed, by the reconstructed near field mask 20 under .-different processing conditions, and these simulated -wafer images can be used for pattern stability and defect detection assessment as further described herein.

[0059] A calibration reticle may also be used to fabricate a calibration wafer from which actual images arc obtained in operation 216, In one example, the actual images are acquired using a critical dimension (CD) scanning electron microscope (SEM). Other 25 imaging tools may be utilized, but a high resolution tool is preferred.

|0060| In general, the calibration wafer will contain any number of known structures, which, may widely vary. The structures may be in the form of gratings which are •typically periodic. Each grating may be periodic in one direction (X or Y) as, for example, a line space grating, or it may be periodic in two directions (X and Y) as, for 30 example, a grid space grating. Examples of a grid space grating may include an array of

lines in the '¥ direction with each line being segmented in the X direction. Another grid space example is an array of dot structures. That is, each structure may take the form of a line space grating, grid space grating, checkerboard pattern structure, etc. The structure design characteristics may each include line width (width at a specific height), line space 5 width, line length, shape, side wall angle, height, pitch, grating orientation, top-profile (degree of top rounding or T topping), bottom profile (footing), etc. The calibration wafer may contain structures with different combinations of these feature characteristics. As should be appreciated, different structure characteristics (such as different widths, spacing, shapes, pitch, etc.) exhibit different response to focus and, therefore, the 10 calibration mask preferably includes different structures with different characteristics.

[0061J In. a specific embodiment, the calibration wafer may take the form of a “Design of Experiments (DOE) wafer having different measurement sites that were subject to different processing conditions. In more general embodiments, process parameter variations are organized in a pattern on the surface of a semiconductor wafer (referred to 15 as u DOS water), in tins manner, the measurement sires correspond to different locations on the wafer surface having different associated process parameter values. In one example, the DOE pattern is a Foeus/Exposure Matrix (FEM) pattern. Typically, a DOE wafer exhibiting a FEM pattern includes a grid pattern of measurement sites. In one grid direction (e.g,, the x-direction), the exposure dosage is varied while the depth of focus is 20 held constant In the orthogonal grid direction (e.g., the y-direction), the depth of focus is varied while the exposure dosage is held constant. In this manner, measurement data collected from the FEM wafer includes data associated with known variations in the focus and dosage process parameters.

J00621 FEM measurement sites are generally located across the focus exposure matrix 25 wafer. In fact, there may generally be one or more measurement sites per field. Each field may be formed using a different combination of focus and exposure (or may be focus or exposure only). For example, a first field may be produced using a first combination, and a second field may be produced using a second combination that is different than the first combination. The multiple .combinations can be produced using

varying focus and varying exposure, varying focus - constant exposure, constant focus — varying exposure, and. the like.

[0063] The number of measuremen t sites may also differ. The number of sites per field is generally smaller on production wafers since the real estate on production wafers is so 5 valuable. Also, fewer measurements are made on a product wafer than mi a focus exposure matrix wafer due to time constraints in production. In one embodiment, a single site is measured per field. In another embodiment, multiple sites are measured per field.

[0064] In most FEM cases, the'measurement site structures are formed .from identically designed patterns using different processing parameters. It should be noted, however, 10 that different focus exposure matrices may have different structures. For example, a first matrix may be performed using a first grating type and a second matrix may be performed using a second grating type that is different than the first grating type,

[0065] In an alternative embodiment, simulated, cafibration images (202) that are rendered from the design database for a calibration reticle may be used as input to the 15 model. That is, the model can be calibrated without recovering the near-field from a physical calibration reticle. Instead, tire lithography image is simulated by simulating (not recovering) the near field from the design database and applying the lithography imaging model to the simulated near field to arrive at the lithography result that is compared to the actual results from the wafer (216).

[0066] In general, optical signal data associated with known variations in any set of process parameters, structural parameters, or both, are contemplated. Regardless of form, the calibration wafer structures may be printed in a variety of different wafer layers. In particular, the printed structures are generally printed in a layer of photoresist using standard lithography processes (e.g., projecting a circuit image through a reticle and onto 25 a silicon waler coated with photoresist). The wafer may be a calibration wafer with layers of materials that correspond to the materials typically present on product wafers at that step in the test process. The printed structures may be printed over other structures in underlying layers. The calibration wafer may be a product wafer that has the potential to produce working devices. The calibration wafer may be a simple wafer that is only used

for calibrating the model. The calibration wafer may be the same wafer that is used to calibrate the OPC tiesign model. More than one calibration wafer may be used to calibrate the lithography model. When using .multiple calibration wafers, the same or different calibration reticles may be used. The different calibration reticles may have patterns with different dimensions- so as to produce a wider range of image data.

[0067] The process parameters used to form the calibration structures are generally configured to keep the pattern’s characteristics within desired specifications. For example, the calibration structures may be printed on a calibration wafer as a part of a calibration procedure or they may be printed on a production wafer during production. In 1.0 production, the calibration structures are typically printed in the scribe line between device areas (e.g., dies that define the IC) disposed on a production wafer. The measurement sites may be dedicated calibration structures disposed around, the device structures or they may be a portion of the device structure (e.g., a periodic portion). As should be appreciated, using a portion of the device structure may be more difficult, but it 15 tends to be more accurate since it is a portion of the device structure. In another embodiment, the calibration structures may be printed across an entire calibration wafer.

[0068] Referring back to Figure 2, corresponding modeled and calibration results (e.g., linages) may be compared in operation 210. It may then be determined whether the model parameters are to be adjusted in operation 212, If model parameters are to be 20 adjusted, they are adjusted in operation 214 and the procedure 200 repeats the operation

208 for modeling the lithography process (and resist) using the adjusted parameters. Model parameters may be adjusted until a quantification of the differences between the model and calibration images have reached a minimum that is also below a predefined threshold. The quantity that is minimized can be the sum of squared differences between 25 the acquired calibration images and the simulated images. The output of this process 200 is a lithography/resist model and its fina l model parameters. This set of model parameters, by the nature of using mask near field, overcomes the technical hurdle associated with mask process modeling and mask 3D diffraction calculation.

]0069| The simulated wafer patterns based on. the recovered mask near field results may be used for a number of mask inspection, metrology, and/orx|uaiification purposes. In one

embodi ment, a reticle qualification is performed by assessing whether the recovered mask near field will likely result in wafer pattern defects under a -range of simulated wafer fabrication conditions. For defect detection, the prift.tabil.ily of a reticle defect on the wafer is important, and the printability of reticle defects depends directly on the reticle near field and lithography system,

[0070] After a final calibrated lithographyZresist/etch model for a particular process is obtained-regardless of how such model is obtained, such model may be used to generate: accurate wafer plane resist images (e.g., after development or after etch) from a mask prior to wafer fabrication with such mask or for requalification of such mask. These resist 10 images will allow one to assess the wafer images for any inspection patterns with high fidelity and through different focus and exposure settings or other lithography parameters.

Since this assessment process can occur prior io wafer fabrication, qualification and defect detection cycles can be significantly shortened. Simulated wafer images may also enable separation of different patterning problem roof causes by comparing the simulated 15 wafer images after lithography, after resist model application, and after etch.

[00711 Figure 3 illustrates a flow chart representing a reticle qualification process 300 in accordance with one embodiment. of the present, invention. In operation 302, a mask near field image is recovered, e.g., for a particular reticle based on images acquired from such particular reticle. This operation may include the mask, near field, recovery' operations of 20 Figure 1. After a mask near field is obtained, the lithography process (and resist) may also be modeled using the final model parameters with respect to the recovered, near mask field in operation 303. For instance, the final model is used to simulate wafer images using a mask near field, image.

|0072| The simulated wafer pattern may then be assessed to determine' pattern stability 25 and/or locate defects in operation 322. It may generally be determined whether the corresponding reticle will likely result in unstable or defective wafer patterns. In one embodiment, the model is applied to the mask near field image or results using a plurality of different, process conditions, such as focus and dose, to assess the reticle design stability under varying process conditions.

[0073] Figure 4A is a flow chart Illustrating a process 400 for determining wafer pattern stability in accordance with an example application of the present invention. Initially»· each test image may be aligned with its corresponding reference image, which images were also produced by the model under different sets of process conditions, in operation 5 402. fhe different test and reference images are -calculated by the model under different processing conditions/parameters.

[0074] Each pair of aligned images may be compared to each other to obtain one or more: wafer pattern differences in operation 404. Thresholds may then be associated with each wafer pattern difference in operation 406. The wafer pattern differences and their 10 associated thresholds can be used together to characterize the pattern stability. That is, a particular pattern's amount of deviation tpattern differences) under different simulated process conditions and whether such deviation crosses an associated threshold together characterize pattern stability. The process window of a fabrication process specifies an expected or defined amount of process deviation under which the resulting patterns are 15 assessed to ensure that they will remain stable or within certain specified tolerances of deviation (e.g., threshold).

[0075] Different thresholds for assessing pattern stability may be assigned to different areas of the reticle and, thereby, corresponding wafer patterns. The thresholds may all be the same or be different based on various factors, such as pattern design context, pattern 20 .MEBF (or Mask Error Enhancement Factor as described, further below) level or sensitivity of the device performance to wafer pattern variation, etc. For instance, one may choose a tighter threshold for patterns in a dense region, compared to a semi-dense region of the reticle.

[0076] An initial set of hot spots or areas of pattern weakness may be optionally identified in both the reference and test mask pattern. For instance, -a designer may provide a list of design hot. spot coordinates that are critical to device function. For example, areas defined as hot spots may be assigned one detection threshold, while nonhot spot areas may be assigned a higher threshold (for defect detection), This differentiation can be used to optimize inspection resources.

|0077.| This pattern stability assessment may be used to facilitate reticle qualification, 'thereby,· overcoming many challenges in this field. As densities and complexities of integrated circuits (ICs) continue to increase, inspecting photolithographic mask patterns become progressively more challenging. Every new generation of ICs has denser and 5 more complex patterns that currently reach and exceed optical limitations of lithographic systems. To overcome these optical limitations, various Resolution Enhancement Techniques (RET), such as Optical. Proximity Correction (OPC), have been introduced. For example, OPC helps to overcome some diffraction limitations by modifying photomask patterns such that the resulting printed patterns con'cspond to the original 10 desired patterns. Such modifications can include perturbations to sizes and edges of main

IC features, i.e., printable features. Other modifications involve additions of serifs to pattern, corners and/or providing nearby sub-resolution assist features (SRAFs), which, are not expected to result in printed features and, therefore, are referred to as non-printable features. These nun-printable features are expected to cancel pattern perturbations that 15 would otherwise have occurred during the printing process. However, OPC makes mask patterns even more complex and usually very dissimilar to resulting wafer images. Furthermore, OPC defects often dp not translate into printable defects. The increased complexity of the photomask pattern and fact that not all pattern elements are expected to directly affect the printed pattern makes the task of inspecting the photomask for 20 meaningful pattern defects much more difficult. As the semiconductor industry moves to ever smaller features, leading-edge manufacturers are starting to use even more exotic OPC, such as inverse lithography technology (l.LT), which result in highly complex patterns on the mask, Thus, it is highly desirable to know the mask writing fidelity and its wafer printing quality prior to physically making the wafer.

[0078] One measure of a defect's importance is its MEEF or Mask Error Enhancement

Factor. This factor relates the size of the defect in the mask plane to the magnitude of the impact, it will have on the printed image. High MEEF defects have high impact on the printed pattern; low MEEF defects have little or no impact, on the printed pattern. An undersized main pattern fea ture in a dense fine-l ine portion of a. pattern is an example of a 30 defect w ith high MEEF where a small mask plane sizing error could cause a complete

collapse of the printed pattern. An isolated small pinhole is an example of a defect with low MEEF where the defect itself is too small to print and is distant enough from the nearest main pattern edge so as not to affect how that edge is printed. As these examples show the MEEF of a defect is a somewhat complicated function of the defect type and the 5 pattern context in which the defect is located.

[007^1 In addition to higher MEEF mask defects causing more significant wafer defects, certain design patterns and corresponding mask patterns may be more robust than other design and mask patterns to process changes. When the fabrication process begins to drift from optimal process conditions, certain mask patterns may result in more significant 10 -wafer pattern perturbations and defects.

[()0801 Figure 4B is a flow chart illustrating a defect inspection procedure 450 In accordance with another embodiment of the present invention, hi operation 452, each modelled test wafer image may be aligned with its corresponding reference image. In one embodiment, a die-to-die or cell-to-cell alignment may be accomplished. In another 15 embodiment, the modelled test wafer image is aligned with a rendered reference image from the corresponding post-OPC design. For instance, the post-ΟΡί' design is processed so as to simulate the reticle fabrication process for such design. For example, corners are rounded. In general, a reference image may originate from the same die as the test image at an earlier time, from an 'adjacent identical die, or rendered from the design database, In 20 a specific example, the reference image is obtained from a ‘‘golden” die, which was proven to be without defects (e.g., immediately after the reticle was manufactured and qualified). The golden reticle images, which were obtained from the reticle when it was known to not be defective, can be stored and later used to compute the golden reticle near field images and wafer images on demand vvhen needed. Alternatively, the golden reticle 25 near field images can be stored for ready access without the need to re-compute the near field in future inspections.

[0081] Each pair of aligned test and reference images is compared based on an associated threshold to locate reticle defects in operation 454, Any suitable mechanism may be used, to associate thresholds to particular reticle areas as further described above, 30 Any suitable metrics of the test and reference images may be compared. For instance, the

contour of test and reference wafer images can be compared as a metric for edge placement error (EPE).

[0082J For each reticle defect, the corresponding .simulated- wafer defect area may then be compared to its corresponding reference pre-OPC area in operation 456. That is, the simulated wafer patterns arc assessed to determine whether the reticle defect results in a wafer defect that varies from the intended design.

[0083] Referring back to Figure 3, it may then be determined whether the design is defective based on the simulated reticle images- in operation 324. hi one embodiment, it is determined whether the design pattern results in. unacceptable water pattern variation 10 under a specified range of process conditions (or process window). It is determined whether there is a significant difference due to process variability. If the difference between differently processed wafer patterns is higher than a corresponding threshold, such wafer patterns may be deemed defective. These systematic defects are known as hot spots. It may also be determined whether any difference between a simulated wafer 15 pattern from reticle and its corresponding pre-OPC pattern is above a predefined threshold. If the design is determined to be defective, the design may be modified in operation 332.

[0084] Once a reticle’s design has been verified, the reticle may likely still contain hot. spot that should be monitored. The following operations are described as being 20 implemented on a mask for which there are at least some identified hot. spots. Of course, if the mask does not contain any identified hot spots, the following operations of Figure 3 may be skipped and the mask used without hot spot monitoring being performed during fabrication and inspection.

[0085] In the illustrated example, if the design is not deemed defective, it may then be 25 determined whether any hot spots can be monitored in operation 326, If it is determined that, the hot spots can be monitored, the hot spots may then be monitored during the-wafer process in operation 334. For instance, the hoi spot patterns may be monitored during wafer fabrication to determine whether the process has drifted out of specification and has caused the corresponding wafer pattern to have critical parameters that change to

unacceptable values. One implementation may involve setting a relatively high MEEF level for inspection of the corresponding hot spot’s reticle and/or wafer pattern. As conditions get further away from nominal process conditions, CI) or EPE can become large and endanger the integrity of the wafer manufacturing process.

[0086| Hot spot patients may he identified merely when a test mask pattern changes by a predefined amount regardless of how such change compares to the original intended design (e.g., pre-OPC data). In other words, a. significant change in the physical mask pattern under different process conditions may indicate a problem with the intended design pattern. Differences between the corresponding modeled image portions represent 10 differences, in the effect of the process conditions’ on the designed pattern and the manufactured mask. Differences associated with a particular design pattern are commonly referred to as “design hot spots”, or just “hot spots”, and represent weak points in the design with .respect to the particular process conditions that have been examined, maybe also, with respect to the manufactured mask. Examples of the kinds of differences that 15 might be found between modeled images for different process conditions are CD (critical dimension) or EPE (edge placement eirof).

[0087] In another embodiment, if the model is applied to the post-OPC design database, the resulting wafer pattern can correspond to the pattern that is intended by the designer to be printed on the waler. Optionally, the results from applying the model to the post20 OPC database can be used with the modeled images to improve hot spot detection. For example, a model of the post-OPC database takes into account design effects only, and so can be used to separate the effect of the wafer process on the design and the effect of the wafer process on the manufactured mask. Modelled patterns from the mask near field may be compared to modelled wafer images from corresponding post-OPC patterns. For 25 instance, when a set of modelled wafer patterns for different process changes match corresponding modelled post-OPC wafer patterns for the same process changes, the changes in wafer pattern (or resist pattern) due to process change can be determined to originate from the design pattern, which can be redesigned or monitored, rather than from a defect in the mask pattern. However, if the changes on wafer due to process variations 30 from the post-OPC database are different from those on wafer due to the same process

variations from the recovered mask (or mask near field), then these hot spots are considered originating from a hot spot from the actual mask, which, can be repaired or monitored.

[0088] The simulated wafer image differences may also be analyzed to determine wafer

CD uniformity (CDU) metrics across the die or over time as reticle changes occur during exposure in the fabrication process. For example, the CD can be measured tor each target Of each image by analyzing and measuring the distance between the target edges if the resolution is high enough. Alternatively, the intensity differences between reference and test images may be calibrated and transformed into CD variations as described further in 10 U.S. Patent Application No. 14/664.565 filed 20 March 2015 by Carl E. Hess et al, and

U.S. Patent Application No. 14/390,834 filed 6 October 2014 by Rui-faiig Shi et al, which applications are incorporated herein by reference in their entirety for all purposes.

[0089] It may also be determined whether the reticle is to be repaired in operation 328. The anticipated wafer pattern variations may be determined to be out of specification for 15 the process window that is expected to be used during the lithography process. In certain cases, the reticle may contain a defect that is repaired in operation 336. The reticle may then be reqnalified, Otherwise, the reticle may be discarded if it is not repairable in operation 330. A new reticle may then be manufactured and requalified.

[0090] In addition or alternatively to using a recovered mask near field image for 20 simulating wafer images in a qualification process, a mask near field image or results may also be assessed directly in a reticle qualification process. Figure 5 is a flow chart illustrated a reticle qualification process 500 that is applied to a recovered mask near field Image or results in accordance with an alternative embodiment of the present invention.

Initially. mask near field results are recovered from a reticle in operation 502. This mask 25 near field image may be recovered for a particular reticle based on images acquired from such particular reticle. This operation may be practiced similar to the mask near field recover operations of Figure 1. Additional!}. several of the operations of 1 igure 5 may he implemented in a similar manner as operations of Figure 3, but with to the recovered reticle near field, image, including intensity and/or phase components of such image.

[00911 As shown, the mask near field results may then be assessed to characterize and/or locate defects in operation 522,. It may generally be determined. whether the corresponding reticle is defective or has hot spots that need monitoring. More specifically, some of the techniques described herein for assessing the simulated wafer 5 images may be implemented on the mask near field images. In a defect detection process.

any suitable metrics of test and reference mask near field images may be compared. For instance,, intensity and/or phase may be compared. Different defect types will, have different impacts on intensity and/or phase values. 'These differences can be determined to be real defects (as opposed to non-impacting nuisance defects) that will likely result, in a 10 defective wafer or identify hoi spot patterns or areas that arc repairable or monitorable.

[0092.1 For instance, it may then be determined whether the design is defective in operation 524. If the design is determined to be defective, the design may be modified in operation 532. For instance, it may be determined whether any difference between a reticle near field image and its corresponding post-OPC-based near field is above a 15 predefined threshold for detecting defects. The procedure 500 may continue so as to determine whether to monitor wafer hot spots, repair the reticle, or redesign the reticle as described above. If the design is not deemed defective, it may then be determined whether any hot spots can be monitored in operation 526. For example, it may be determined that any intensity and/or phase difference between a test and reference reticle 20 near field image is close to an associated threshold.

[0093.1 If it is determined that the hot spots can be monitored, the hot spots may then be monitored during the wafer process in operation 534, for instance. For instance, the hot spot patterns may be monitored during wafer fabrication to determine whether the process has drifted out of specification and has caused the corresponding wafer pattern to have 25 critical parameters that change to unacceptable values. One implementation may involve setting a relatively high sensitivity level for inspection of the corresponding hot spot’s reticle and/or wafer pattern. As conditions get further away from nominal process conditions, CD error or EPE can. become large and endanger the integrity of the wafer manufacturing process.

[0094] It may also be determined whether foe reticle is to be repaired in operation 528. In certain, .cases, the reticle may contain, a defeci that is repaired, in operation 536. The reticle may then be requalified. Otherwise, the reticle may be discarded if it is not repairable in operation 530. A new reticle may then be manufactured and requalified.

[0095] Certain techniques of the present invention, provide mask pattern qualification and early detection of weak patterns or hot spots on. the physical mask before beginning wafer --manufacturing. In addition to providing recovery of the reticle near field based on reticle images, a full range of wafer process effects (including many settings of focus and exposure, and the effect, of wafer resist, etch, CMP, and any other wafer processes) can be 10 considered for how they affect the wafer patterns. No prior knowledge of the mask is required since the mask near field is recovered using only reticle images without using reticle design data. Since mask patterns are generally 4x larger than wafer patterns, more exact locations of patterns with respect to the design database can be determined. The above techniques can also be extended to any suitable type of masks, such as pattern 15 qualification of EUV masks.

[0096] Techniques of the present invention may be implemented in any suitable combination of hardware and/or software. Figure 6 is a diagrammatic representation of an example inspection system 600 in which techniques of the present invention may be implemented. The inspection system 600 may receive input 602 from a high NA inspection tool ora low NA inspector mimicking a scanner (not shown}. The inspection system may also include a data, distribution system (e.g., 604a and 604b) for distributing the received input 602, an Intensity signal (or patch} processing system (e.g., patch processors and reticle qualification system (e.g., oI2) for mask: near field and wafer recovery, process modelling, etc., a network (e.g., switched, network: 608) for allowing 25 communication between the inspection system components, an optional mass storage device 61.6, and one or more inspection control and/or review stations (e>g., 61.0) for reviewing mask near field intensity and phase (values, images, or differences), reticle/wafer images, identified hot spots, CD, CDU maps, process parameters, etc. Each processor of the inspection system 600 typically may include one or more microprocessor

integrated circuits and may also contain interface and/or memory integrated circuits and may additionally be coupled to one or more shared and/or global memory devices.

[0097] The inspector or data acquisition system (not shown) for generating input data 602 may take the form of any suitable instrument (e.g., as described further herein) for obtaining intensity signals or images of a reticle. For example, the low NA inspector may construct an optical image or generate intensity values of a portion of the .reticle based on a portion of detected light that is reflected, transmitted, or otherwise directed to one or more light sensors. The low NA inspector may then output the intensity values or image.

[0098] The low NA inspection tool may be operable to detect and collect reflected 10 and/or transmitted light as an incident optical beam scans across each patch of a reticle,

As .noted above, the incident, optical beam may scan across .reticle swaths that each comprises a plurality of patches. Light is collected in response to this incident beam from a plurality of points or subareas o f each patch.

[0099] The low NA inspection tool may be generally operable to convert such detected light into detected signals corresponding to intensity values. The detected .signals may take the form of an electromagnetic waveform having amplitude values that correspond to different intensity values at different locations of the reticle. The detected signals may also take the form of a simple list of intensity values and associated reticle point coordinates. The detected signals may also take the form of an image having different intensity values corresponding to different positions or scan points on the reticle. Two or more images of the reticle may be generated after all the positions of the reticle are scanned and converted into detected signals, or portions of a two or more images may be generated as each reticle portion is scanned with the final two or more images for the reticle being complete after the entire reticle is scanned.

[001.00] The detected signals may also take the form of aerial images. That is, an aerial imaging: technique may be used to simulate: the optical effects of the. photolithography system so as to produce an aerial image of the photoresist pattern that is exposed on the wafer. In general, the optics of the photolithography tool am emulated so 'as to produce an aerial image based on the detected signals from the reticle. The aerial image

corresponds to the pattern produced fen the light passed through the photolithography optics and reticle onto the photoresist, layer of a wafer. Additionally, the photoresist exposure process for the particular type of photoresist material may also be emulated.

[001011 The incident light or detected light may be passed through any suitable spatial aperture io produce any incident or detected light profile at any suitable incident angles.

By way of examples, programmable illumination or detection apertures may be utilized to produce a particular beam profile, such as dipole, quadrapole, quasar, annulus, etc. In a specific example, Source Mask Optimization (SMO) or any pixelated illumination technique may be implemented. The incident light may also be passed through a linear 10 polarizer for linearly polarizing all or a portion of the illumination pupil in one or more polarizations. The detected light may be passed through apodization components for blocking particular areas of the collection beam.

[001021 Intensity or image data 602 can be received by data distribution system via network 608. The data distribution system may be associated with one or more memory 15 devices, such as RAM buffers, lor holding at least a portion of the received data 6().2.

Preferably, the total memory is large enough to hold an entire swatch of data. For example, one gigabyte of memory works well for a swatch that is I million by 1000 pixels or points.

[001031 The data distribution system (e.g,, 604a and 604b) may also control distribution 20 of portions of the received input data 602 to the processors (e.g. 606a and 606b). For example, data distribution system may route data for a first patch to a first, patch processor 606a, and may route data for a second patch to patch processor 606b. Multiple sets of data for multiple patches may also be routed to each patch processor.

[001041 The patch processors may receive intensity values or an image that corresponds 25 to at. least a portion or patch of the reticle. The patch processors may each also be coupled to or integrated with one or more memory devices (not shown), such as DRAM devices (hat provide local memoty functions, such as holding the received- data portion.

Preferably, the memory is large enough to hold data that corresponds to a patch of the .reticle. For example, eight megabytes of memory works well for intensity values or an

image corresponding to a patch that is 512 by 1024 pixels. Alternatively, the patch processors may share memory.

[00105] Each set of input data 602 may correspond to a swath of the reticle. One or more sets of data may be stored in memory of the data distribution system. This memory .may be controlled by one or more processors within the data distribution system, and the memory may be divided into a plurality of partitions. For example, the data distribution system may receive data corresponding to a portion of a swath into a first memory partition (not shown), and the data distribution system may receive another data corresponding to another swath into a second memory partition (not shown). Preferably, 1.0 each of the memory partitions of the data distribution system only holds the portions of the data that are to be routed to a processor associated with such -memory partition. For example, the first memory' partition of the data distribution system may hold and. route first data to patch processor 606a, and the second .memory partition may hold and route second data to patch processor 606b.

[00106] The data distribution system may define and distribute each set of data of the data based on any suitable parameters of the data. For example, the data may be defined and distributed based on the corresponding posifioti of the patch on the reticle. In one embodiment, each swath is associated with a rangv of column positions that correspond to horizontal positions of pixels within the swath, for example, columns 0 through 256 of 20 the swath may correspond to a first patch, and the pixels within these columns will comprise the first image or set of intensity values, which is routed to one or more patch processors. likewise, columns 257 through 512 of the swath may correspond to a second patch, and the pixels in these columns will comprise the second, image or set of intensity values, which is routed to different patch processors).

[(101(17] The inspection apparatus may be suitable for inspecting semiconductor devices or wafers and optical reticles, as well as EUV reticles or masks. Examples of suitable inspeetion tools are the Teron™ operating at 193 nm or the TeraScan5 M DUV reticle inspection tools available from KLA-Tencor of Milpitas, CA. Other types of samples which may be inspected or imaged using the inspection apparatus of the present invention 30 include any surface, such as a flat panel display.

jOOIOS] An inspection tool may include at least one light source for generating an incident light beam, illumination optics for directing the incident beam onto a sample, collection optics for directing an output beam that is emitted from the sample in response to the incident beam, a sensor for detecting the output beam and generating an image or signal for the output beam, and a controller/processor for controlling the components of the inspection tool and facilitating the mask near field generation and analysis techniques as described- further herein.

[001091 In the following exemplary inspection systems, the incident beam may be in any suitable form of coherent light. Additionally, any suitable lens arrangement may be 10 used to direct the incident beam towards the sample and direct the output beam emanating from the sample towards a detector. The output beam may be .i eflcctvd or scattered from the sample or transmitted through the sample. For EUV reticle inspection, the output beam is typically reflected from the sample. Likewise, any suitable detector type or number of detection elements may be used to receive the output beam and provide an 15 image or a signal based on the characteristics (e.g.< intensity) of the received output beam.

[001101 A generalized photolithography tool will first be described although an EUV photolithography tool would .typically only have reflecting type optics. Figure 7A is a simplified schematic representation o f a typical lithographic system 700 that can be used to transfer a mask pattern from a photomask M onto a. wafer W in accordance with certain 20 embodiments. Examples of such systems include scanners and steppers, more specifically the TWINSCAN NXT:197OC1 Step-and~Sca.ii system available from ASML in Veldhoven., Netherlands. In -general, an illumination, source 703 directs a light beam through an illumination optics 707 (Ag., lens 705) onto a photomask M located in a mask plane 702. Illumination lens 705 has a numeric aperture 701 at that plane 702, The value 25 of the numerical aperture 701 impacts which defects on the photomask are lithographic significant defects and which ones are not. A portion of the beam that passes through the photomask M forms a patterned optical signal that is directed through imaging optics 71,3 and onto a wafer W to initiate the pattern transfer. In a reflecting system (not shown), the illumination beam is reflected from certain portions of the mask M (and absorbed by

other portions of such mask M) and forms a patterned signal that is directed through reflecting imaging optics on a wafer W.

[0011'1] The inspection tool may utilize similar components or be similarly configured as the above-described photolithography tool. e.g. LNI capability. However, the 5 inspection tool may be alternatively or additionally configurable to generate high resolution images. Figure 7B provides a schematic representation of an example inspection system 750 that has illumination optics 751a includes an imaging lens with a relative large numerical aperture 751b at a reticle plane 752 in accordance with certain embodiments. For example, numerical aperture 751b at the reticle plane 752 of the 10 inspection system may be considerable greater than the numerical aperture 701 at the reticle plane 702 of the lithography system 700, which would result in differences between test inspection images and actual printed images.

[001.12] fhe inspection techniques described herein may be implemented on various specially configured inspection systems, such as the one schematically illustrated in 15 Figure 7B. The illustrated system. 750 includes an illumination source 760 producing a light beam that is directed through illumination optics 751a onto a photomask M in the reticle plane 752. Examples of light sources include a coherent laser light source (e.g. deep UV or gas laser generator), a filtered lamp, LED light source, etc. In certain embodiments, a light source may generally provide high pulse repetition rate, low-noise, 20 high power, stability, reliability, and extendibility. It is noted that while an EUV scanner operates at 13.5 nm wavelength, an inspection tool for an EUV reticle does not have to operate at the same wavelength al though it could. In one example, the source is a 193 urn laser.

100113] The illumination optics 75 la may include a beam steering device for precise beam positioning and a beam, conditioning device, which can be used to provide light level control, speckle noise reduction, and high beam uniformity. Beam steering and/or beam conditioning devices may be separate physical dm ices from, for example, a laser. The illumination optics 751a may also include optics for controlling polarization, focus, magnification, illumination intensity distributions, etc.

[00114] As explained above, the inspection system 750 may have a numerical aperture 751b at the. reticle plane 752 that may be equal to or greater than a reticle plane numerical aperture (e.g., element 701 in Figure 7A) of the corresponding lithography system. The photomask M to be inspected is placed on a mask stage at the reticle plane 752 and 5 exposed to the source.

[00115] The depicted inspection system 750 may include detection optics 753a and 753b, which may also include microscopic magnification optics designed to provide, for example, 60-200X magnification or mure for enhanced inspection. The collection optics 753a and 753b may include any suitable optics for conditioning the output, light/beam.

For instance, the collection opties 753a and 753b may include optics for controlling focus, pupil shapes, polarization analyzer settings, etc.

[00116] In. a transmitting mode, the patterned image from the mask M may be directed through a collection of optical elements 753a, which project the patterned image onto a sensor 754a. In a reflecting mode, collection elements (e.g,t beam splitter 776 and 1.5 detection lens 778) direct and capture the reflected light from the mask. M onto sensor

754b. Although two sensors are shown, a single sensor can be used to detect reflected and transmitted light during different scans of the same reticle area. Suitable sensors include charged coupled devices (CCD), CCD arrays, time delay integration (TDI) sensors, TDI sensor arrays, photomultiplier tubes (PMT), and other sensors.

J 00117] The illumination optics column may be moved with respect to the mask stage and/or the stage moved relative to a detector or camera by any suitable mechanism so as io scan patches of the reticle. For example, a motor mechanism may be utilized to move the stage. The motor mechanism may be formed from a screw drive and stepper motor, linear drive with feedback position, or baud actuator and stepper motor, by way of examples. The system 700 may utilize one or more motor mechanisms for moving any of the system components with respect to the illumination or collection optical paths.

[00118] The signals captured by each sensor (e g,. 754a and/or 754b) can be processed by a computer system 773 or, more generally, by one or more signal processing devices, which may each include an analog-to-digital converter configured to convert analog

signals from each sensor into digital signals for processing. The computer system 773 typically has -one or more processors coupled to input/output ports, and one or more memories via appropriate buses or other communication mechanisms.

[00H9J The computer system 773 may also include one or more input devices (e.g., a:

keyboard, mouse, joystick) for providing user input, such as changing focus and other inspection recipe parameters. The computer system 773 may also be connected to the stage for controlling, for example, a sample position (e.g,, focusing and scanning) and connected to other inspection system components for controlling other inspection parameters and configurations of such inspection system components.

1001201 The computer system 773 may be configured, (e.g., with programming instructions) to provide a user interface (e.g., a computer screen) for displaying mask near field intensity and phase (values, images, or differences), reticleAvafer images, identified hot spots, CD, CDU maps, process parameters, etc. The computer system 773 may be configured to analyze intensity, phase, and/or other characteristics of reflected and/or transmitted detected and/or simulated signals or images, recovered reticle near field results, etc. The computer system 773 may be configured (e.g., with programming instruct ions) to provide a user interlace (e.g., on a computer screen) for displaying resultant intensity and/or phase values, images, and other inspection characteristics, hr certain embodiments, the computer system 773 is configured to carry out inspection techniques detailed above

[00121] Because such information and program instructions may be implemented on a specially configured computer system, such a system includes program instructions / computer code for performing various operations described herein that can be stored on a computer readable media. Examples of machine-readable media include, but are not 25 Limited to, magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROM disks; magneto-optical media such as optical disks; and hardware devices that are specially configured to store and perform program instructions, such as read-only memory devices (ROM) and random access memory (RAM). Examples of program instructions include both machine code, such as produced by a

compiler, and files containing higher level code that may be executed by the computer using an interpreter.

[00122] Figure 7B shows an example where an illuminating light beam is directed towards the sample surface at a substantially normal angle with respect to the inspected 5 surface. In other embodiments, an illuminating light beam can be directed at an oblique angle, which allows separation of the illuminating and reflected beams. In these, embodiments, an attenuator may be positioned on the reflected beam path in order to attenuate a zero order component, of the reflected light beam prior to reaching a detector. Furthermore, an imaging aperture may be positioned on the reflected beam path to shift 1.0 the phase of the zero order componen t of th e re fl ec ted light beam.

[()0123] It should be noted that the above description and drawings are not to be construed as a limitation on the specific components of the system and that the system may be embodied in many other forms. For example, it is contemplated that the inspection or measurement tool may have any suitable features from any number of 15 known imaging or metrology tools arranged for detecting defects and/or resolving the critical aspects of features of a reticle or wafer. By way of example, an inspection or niemirement tool may be adapted for bright field imaging microscopy, dark field Imaging microscopy, full sky imaging microscopy, phase contrast microscopy, polarization contrast microscopy, and coherence probe microscopy. It is also 20 contemplated that single and multiple image methods may be used in order to capture images of the target. These methods include, for example, single grab, double grab, single grab coherence probe microscopy (COM.) and double grab CPM methods, Nonimaging optical methods, such as scatterometry, may also be contemplated as forming part of the inspection or metrology apparatus.

[()0124] .Although the foregoing invention has been described in some detail for purposes of clarity of understanding, it will be apparent that certain changes and modillcmions may be practiced within the scope of the appended claims. It should be noted that there arc many alternative ways of implementing the processes, systems, and apparatus of the present invention. Accordingly, the present embodiments are to be

considered as illustrative and not restrictive, and the invention is not to be limited to the details given herein.