METHOD AND IMAGING SYSTEM FOR QUALIFYING A PHOTOLITHOGRAPHIC RETICLE

Abstract

A method for qualifying a photolithographic reticulum, comprising: using an imaging tool to capture a variety of images from each of a variety of pattern regions of a test reticulum under different illumination configurations and / or different imaging configurations; reconstructing a reticulum near-field for each of the pattern regions of the test reticulum based on the captured images of each pattern region of the test reticulum; and using the reconstructed reticulum near-field to determine whether the test reticulum is likely to result in an unstable wafer pattern or a defective wafer, characterized in that the reticulum near-field is directly analyzed to determine whether the test reticulum is likely to result in an unstable wafer pattern or a defective wafer.

Description

REFERENCE TO RELATED REGISTRATIONS

[0001] This application claims the priority benefit of U.S. patent application no. 15 / 803628, filed on November 3, 2017, by Rui-fang Shi et al. This application also claims priority of U.S. preliminary application no. 62 / 508,369, filed on May 18, 2017. This application and this patent are hereby incorporated by reference in their entirety for all purposes. TECHNICAL AREA OF INVENTION

[0002] The invention relates generally to the field of reticle inspection. In particular, the present invention relates to sample qualification. BACKGROUND

[0003] In general, the semiconductor manufacturing industry involves highly complex techniques for producing integrated circuits using semiconductor materials that are layered and structured onto a substrate, such as silicon. Due to the large scale of circuit integration and the decreasing size of semiconductor devices, the manufactured components have become increasingly susceptible to defects. That is, defects that cause failures in the device are becoming ever smaller. The device must be defect-free before being shipped to end users or customers.

[0004] An integrated circuit is typically fabricated from a multitude of reticles. Initially, circuit designers provide circuit pattern data, describing a specific integrated circuit (IC) design, to a reticle production system or reticle writer. The circuit pattern data is typically in the form of a representative layout of the physical layers of the fabricated IC device. The representative layout includes a representative layer for each physical layer of the IC device (e.g., gate oxide, polysilicon, metallization, etc.), with each representative layer consisting of multiple polygons that define the structure of a layer of the specific 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 reticule pattern) of a variety of reticles that are later used to manufacture the specific IC design.

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

[0006] After the production of each reticle or group of reticles, each new reticle is typically qualified for use in wafer fabrication. For example, reticle samples must be free of printable defects. Furthermore, each wafer produced using the reticle must be defect-free. Therefore, there is a continuing need for improved techniques for the inspection and qualification of reticles and wafers. OVERVIEW

[0007] The following is a simplified summary of the disclosure to provide a basic understanding of certain embodiments of the invention. This summary is not a comprehensive overview of the disclosure and does not identify any essential / critical elements of the invention or outline the scope of the invention. Its sole purpose is to present some concepts disclosed herein in a simplified form as a prelude to the more detailed description that will be presented later.

[0008] In one embodiment, a method for qualifying a photolithographic reticle is disclosed. An imaging tool is used to acquire a multitude of images of each of a multitude of pattern regions of a test reticle under different illumination configurations and / or different imaging configurations. A reticle nearfield is reconstructed for each of the pattern regions of the test reticle based on the images acquired from each pattern region of the test reticle. The reconstructed reticle nearfield is then used to determine whether the test reticle or another reticle is likely to result in an unstable wafer pattern or a defective wafer.

[0009] In one implementation, the reticulum near-field is directly analyzed to determine whether the test reticulum or another reticulum is likely to result in an unstable wafer pattern or a defective wafer. In another aspect, the recovered reticulum near-field is used to detect defects in the test reticulum or in simulated wafer images generated from the recovered reticulum near-field. This defect detection involves comparing the intensity and / or phase for an identical die at different times, for adjacent dies, for a die and its associated golden die, or for a die and its associated die from a reticulum copy with an identical design to the test reticulum.

[0010] In one aspect, the images are acquired in a field plane or a pupil plane. In a specific embodiment, the reticle near field is reconstructed without using a design database that was used to manufacture the reticle. In another aspect, the acquired images comprise at least three reflection / transmission images acquired under different imaging conditions, selected to produce the 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 settings for the intensity distribution and / or polarization of the source.

[0011] In an alternative implementation, the procedure involves (i) applying a lithography model to the reticle near-field for the test reticle to simulate a variety of test wafer images, and (ii) analyzing the simulated test wafer images to determine whether the test reticle is likely to result in an unstable or defective wafer. In this aspect, the lithography model is configured to simulate a photolithography process. In another aspect, the lithography model simulates an illumination source with a different shape than the illumination shape of an inspection tool for capturing images of the test reticle or any other reticle or wafer. In yet another aspect, the lithography model is calibrated using images rendered from a design database for a calibration reticle. In another example, the lithography model is calibrated using images captured by a calibration reticle.In yet another aspect, the lithography model is applied to the reticule near field obtained for the test reticule under a variety of different lithography process conditions, and the analysis of the simulated test wafer images includes determining whether the test reticule is likely to lead to an unstable wafer under the different lithography process conditions by comparing parts of the simulated test images associated with different process conditions and the same reticule area.

[0012] In an alternative embodiment, the invention relates to an inspection system for qualifying a photolithographic reticle. The system comprises a light source for generating an incident beam and an illumination optics module for directing the incident beam onto a reticle. The system also comprises a collecting 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 further comprises a controller configured to perform operations similar to one or more of the process operations described above.

[0013] These and other aspects of the invention are described below with reference to the figures. List of characters Fig. Figure 1 is a flowchart illustrating a near-field mask recovery procedure according to an embodiment of the present invention. Fig. Figure 2 is a flowchart that represents a model calibration process according to a specific implementation of the invention. Fig. Figure 3 shows a flowchart illustrating a reticle qualification process according to an embodiment of the present invention. Fig. Figure 4A is a flowchart illustrating a process for determining reticle pattern stability according to an exemplary application of the present invention. Fig. Figure 4B is a flowchart showing a defect inspection procedure according to another embodiment of the present invention. Fig. Figure 5 is a flowchart illustrating a reticle qualification process applied to a recovered near-field mask image according to an alternative embodiment of the present invention. Fig. Figure 6 is a schematic representation of an exemplary inspection system in which techniques of the present invention can be implemented. Fig. Figure 7A is a simplified schematic representation of a lithography system for transferring a mask pattern from a photomask to a wafer according to certain embodiments. Fig. Figure 7B shows a schematic representation of a photomask inspection device according to certain embodiments. DETAILED DESCRIPTION OF EXAMPLES OF EXECUTION

[0014] The following description provides numerous specific details to facilitate a thorough understanding of the present invention. The present invention can be implemented without some or all of these specific details. In other cases, known processes or device components have not been described in detail to avoid unnecessarily obscuring the present invention. Even when the invention is described in conjunction with specific embodiments, it is understood that the invention is not intended to be limited to these embodiments.

[0015] It would be advantageous to record the defects of each mask and also to characterize various other aspects of the mask (e.g., pattern stability, CD, CD uniformity) before masks are transported to a manufacturing facility, before wafers are manufactured using such a mask, and / or for periodic requalification of such a mask after it has been used for certain periods in the manufacturing process.

[0016] One embodiment of the present invention comprises techniques for reconstructing near-field images of a reticle from reticle images obtained by an inspection tool under a variety of different imaging parameters. This reticle near-field image can then be used in a range of reticle qualification applications. In one example, the reticle near-field image can be inputted into a lithography model to predict a wafer image or various wafer pattern properties relating to how the resulting patterns will be printed onto the wafer. The predicted wafer image and / or various wafer properties can then be analyzed for defect detection, reticle qualification or requalification, and / or any other suitable metrology or inspection application. The reticle near-field image itself can also be analyzed for various purposes, as further described herein.

[0017] The terms "reticule," "mask," and "photomask" are used interchangeably here and can generally each comprise a transparent substrate such as glass, borosilicate glass, quartz, or fused silica with a layer of opaque material formed on top. The opaque (or essentially opaque) material can be any suitable material that completely or partially blocks photolithographic light (e.g., deep UV or extreme UV). Example materials include chromium, molybdenum silicide (MoSi), tantalum silicide, tungsten silicide, opaque MoSi on glass (OMOG), etc. A polysilicon film can also be added between the opaque layer and the transparent substrate to improve adhesion. A weakly reflective film such as molybdenum oxide (MoO2), tungsten oxide (WO2), titanium oxide (TiO2), or chromium oxide (CrO2) can be formed over the opaque material.In a specific example, an EUV reticle can comprise multiple layers with alternating layers having different refractive indices with low absorption properties (such as molybdenum (Mo) and silicon (Si)) and absorber materials (such as a tantalum boron nitride film with a thin antireflection oxide layer overlaid on top).

[0018] The term reticulum refers to various types of reticula, including, but not limited to, a clear-field reticulum, a dark-field reticulum, a binary reticulum, a phase-shift mask (PSM), an alternating PSM, an attenuated or halftone PSM, a ternary attenuated PSM, a chromium-less phase lithography PSM, and chromium-less phase lithography (CPL). A clear-field reticulum has transparent field or background regions, and a dark-field reticulum has opaque field or background regions. A binary reticulum is a reticulum with structured regions that are either transparent or opaque. For example, a photomask can be used that is made from a transparent quartz glass blank with a pattern defined by a chromium metal adsorption film.Binary reticles differ from phase-shift masks (PSMs), one type of which may include films that transmit only partial light. These reticles may be commonly referred to as halftone or embedded phase-shift masks (EPSMs), such as ArF and KrF masks. When a phase-shifting material is placed on alternating free areas of a reticle, the reticle is called an alternating PSM, ALT-PSM, or Levenson-PSM. A type of phase-shifting material applied to any layout pattern is called a damped or halftone PSM, which can be created by replacing the opaque material with a partially transmitting or "halftone" film. A ternary damped PSM is a damped PSM that also includes completely opaque features.

[0019] The next generation of lithography has introduced the use of extreme ultraviolet (EUV, wavelength 13.5 nm) radiation, which is absorbed by both normal atmosphere and glass. For this reason, EUV lithography processes take place under vacuum, and optically reflective lenses / mirrors are used to focus the EUV light onto photomasks that have reflective and absorber patterns instead of translucent and opaque patterns.

[0020] Fig. Figure 1 is a flowchart illustrating a near-field mask recovery procedure. 100 according to one embodiment of the present invention. The following mask restoration process 100This can be performed for a specific reticle or set of reticles at any suitable point in the reticle's life cycle, as further described below in various use cases for the restored near-mask field. For example, a near-mask field can be restored before the fabrication of wafers using such reticles, before commencing large-scale wafer fabrication, or during the requalification of such reticles.

[0021] At the beginning of the process 102At least three images of the mask are acquired using a mask inspection tool with different imaging configurations. Alternatively, two images can be used, but three images have proven to be the most effective. Acquisition with different imaging configurations can be performed simultaneously or sequentially. The acquired images do not need to be in the field planes. For example, the two or more images can be acquired in the pupil planes, allowing direct access to the diffraction intensity.

[0022] Various suitable combinations of illumination and / or collection configurations can be used to capture the two or more images. The different imaging configurations are generally selected to provide images from which the near-field mask can be calculated. Any suitable imaging or optical configurations can be chosen so that the near-field mask remains consistent under different operating conditions. Examples include different focus settings, different illumination shapes (e.g., different directions or patterns), different polarization for the entire illuminated pupil or different parts of the illuminated pupil, different apodization settings to obscure different parts of the collecting beam, and so on.In one embodiment, different focus settings can be used by focusing and defocusing (such as 0-focus, ±800 or ±1600 defocus, etc.) to capture the different images. In another example, different quadrants of the illumination pupil can have different polarization settings. In yet another example, the imaging configurations can include high-resolution images, such as transmission images (e.g., for ArF masks), with different pupil shapes and / or different focus conditions. In another embodiment, three or more reflected images with different pupil shapes and / or different focus conditions can be obtained (e.g., for EUV masks).

[0023] The reticel can be imaged at a “substantially low resolution” using a relatively low numerical aperture (NA) (e.g., less than 0.5). In contrast, a “substantially high-resolution image” generally refers to an image of a reticel in which features printed on the reticel appear substantially as they are formed on the reticel (within the optical limits of the reticle inspection system used to generate the image). A “substantially high-resolution image” of a reticel is an image produced by imaging the physical reticel onto the reticle plane using a substantially high-resolution reticle inspection system (e.g., a numerical aperture (NA) greater than 0.8).The "essentially low NA" used to generate a reticule image can be essentially the same as the NA on the reticule side used by an exposure / lithography system to project an image of the reticule onto a wafer, thereby transferring features of the reticule to the wafer. In the essentially low NA (or LNI) image, the reticule features may have a substantially different appearance than the actual reticule features. For example, reticule features may have more rounded corners in an LNI of a feature than the actual feature formed on the reticule.

[0024] In general, any suitable imaging tool can be used for the mask near-field recovery process. In certain embodiments described herein, the results of an initial recovery process can later be used for pattern stability or defect detection evaluations with respect to the same reticulum or other reticulums based on additional reticulum images from a specific inspection tool. For consistency in these use cases, the reticulum images for mask near-field recovery can be acquired with the detector of the reticulum inspection system used for subsequent inspection of the same or other reticulums, or with a similarly configured detector from a similarly configured reticulum inspection system (e.g., another reticulum inspection system of the same make and model as the reticulum inspection system used for the inspection).In other words, the images that can be used for mask recovery can be captured under the same optical conditions used in subsequent mask inspection or qualification processes. This allows the interaction of the reticule with the illuminating electromagnetic waves of the inspection system to be measured as directly as possible.

[0025] In alternative embodiments, the tool used for near-field mask restoration may differ from a reticle inspection system.

[0026] For example, the imaging tool can use the same wavelengths (e.g., 193.3 nm for DUV or 13.5 nm for EUV) as the lithography system in which the reticule is used for wafer fabrication. In fact, any suitable electromagnetic wavelength can be used for mask near-field reconstruction.

[0027] Referring again to the illustrated example, the process can be described as follows: 104 The three or more images can then be aligned to each other, or each image can be aligned to the post-OPC database. For example, the acquired images can be aligned using spatial domain or frequency domain methods. Alignment adjustments may depend on specific geometries of the imaging system used. If different images are obtained using different collection paths, image adjustments can be made to compensate for differences in the optical paths.

[0028] In this imaging tool, a reticulum with various patterns is illuminated by electromagnetic (EM) waves incident from multiple directions. This incident light is diffracted at different points on the mask pattern at different electromagnetic field phases, which interfere with each other in different ways. The near field of the reticulum is the electromagnetic field at a small distance of a few wavelengths from the reticulum.

[0029] The converging optics generally direct a diffraction-limited portion of the light from the reticle to a detector (or wafer) to generate an image. The detector detects the intensity, which is the result of interference due to the near-field mask, but does not detect the phase.

[0030] Although far-field intensity is preserved in the acquired signals, it is desirable to reconstruct the near-field mask, which includes amplitude and phase. In the illustrated embodiment, the near-field mask is reconstructed and stored based on such acquired mask images, as described in the procedure. 106As shown, multiple images (or signals) are generally used to reconstruct the mask near-field, which contains both phase and amplitude components. The near-field data can be determined by a regression technique based on the images acquired by the reticle. For example, the near-field of a selected section of the reticle can be reconstructed (determined by regression) from its acquired optical images using a quasi-Newtonian or conjugate gradient technique, or from the intensity of images acquired in a detector plane. Additionally, any other suitable regression method and / or algorithm can be used to determine the near-field data from the one or more actual images.

[0031] Near-field mask recovery can generally be achieved by solving an optimization problem that attempts to minimize the difference between observed intensity images and resulting images of the assumed optical mask field. Specifically, recovering the near field of a reticle from its intensity images is an inverse or regression problem. The near field can be recovered iteratively by minimizing a cost function (e.g., energy or penalty function). The minimized quantity can be the sum of the squared differences between the acquired images and the intensity images at the detector, calculated from the near-field mask. In other words, intensity images can be computed from the final near-field mask for various sets of optical system properties, and these computed images will most closely match the acquired images once the near-field mask has been found.Various mask near-field recovery methods and system implementations are further described in U.S. Patent No. 9,478,019, issued on October 25, 2016, by Abdurrahman Sezginer et al., which patent is incorporated herein by reference in its entirety for all purposes.

[0032] In the case where multiple images are taken under different optical conditions, the recovered near-field mask m, which carries the phase and amplitude information, can be determined by the following equation: m ' = arg min ∑ α c α ∑ x , y [ I α − ∑ i λ i ( α ) | m ⊕ ψ i ( α ) | 2 ] 2 <?page 7=""?>

[0033] In equation 1 above, I α the measured image for imaging condition a, ψ i ( α ) a set of eigenvectors that describe the inspection imaging system, λ i ( α ) a set of corresponding eigenvalues ​​for the mapping system and c αis a non-negative weighting factor between 0 and 1. The equation above can be solved iteratively, for example, using methods such as the quasi-Newton method or the conjugate gradient method.

[0034] Another example is the Gerchberg-Saxton algorithms, which can use a combination of field plane images and pupil plane diffraction orders to solve for both the amplitude and the phase of the object.

[0035] In one embodiment, the near-field of the mask can be determined based on the acquired images using a Hopkins approximation. In another embodiment, the regression does not include any thin-mask approximations. For example, the near-field of the reticle is the electromagnetic field calculated to be present near the surface of the reticle when illuminated by a normally incident plane wave. In lithography and inspection, a reticle is illuminated by plane waves incident from many directions. If the direction of incidence changes, the directions of the diffraction orders change according to the Hopkins approximation, but their amplitudes and phases remain approximately unchanged. The embodiments described here can use the Hopkins phase approximation but do not perform so-called thin-mask or Kirchhoff approximations.

[0036] The recovery formula can also be varied with different norms or by adding a regularization term R, which suppresses near-field vibrations, as follows: m ^ = arg min ∑ α c α ∑ x , y [ I α − ∑ i λ i ( α ) | m ⊕ Ψ i ( α ) | 2 ] l + R ( m ) where the regularization term R can include prior information about the near field or expectations based on a physical understanding of the mask substrate / material. Furthermore, the norm used for image difference can be an I-norm and adapted based on specific requirements of the optimization function.

[0037] Interestingly, the interference of the electromagnetic field vectors of the mask is greater as a result of a higher NA (than with an inspection system with a lower NA), because the range of light incidence angles and the associated interfering electric field components is larger for a larger NA.

[0038] The actual mask may differ from the intended design patterns due to the mask writing process. Obtaining the near-field mask from images of the mask means that such a near-field mask is derived from the actual physical mask rather than from the design database. This means the mask's near field can be reconstructed without using the design database.

[0039] The results for the near-mask field can then be used in various applications. In one embodiment, near-mask field results can be used to predict wafer patterns using one or more models. That is, the recovered near-mask field can be used to simulate lithographic images. Any suitable technique can be used to simulate lithographic images based on the near-mask field images. One embodiment involves calculating the lithographic image using the partial coherence model. I l i t h o ( x ; f , z ) = ∑ i λ i | a ^ ( x ) ⊕ ψ i s ( x ; f , z ) | 2 where λ i represents the eigenvalues ​​of the lithography TCC (transfer cross coefficients); ψ i s ( ) represents eigenvectors (kernels) of TCC; s is the wafer stack, including layer refractive indices; f is the focus; and z is the vertical position of the lithography plane in the photoresist material. The transmission cross coefficients (TCC) of Equation 2 can include a vectorial propagation of the field through the lithography projector, including the layer stack on the wafer.

[0040] Before using a model to predict wafer outcomes, the model can be calibrated to achieve the most accurate results possible. The model can be calibrated using any suitable technique. Certain embodiments of the present invention provide techniques for calibrating a lithography model based on near-field mask results obtained from a calibration mask. In alternative embodiments, the design database is used to calibrate the model. For example, images of the calibration reticle can be rendered from the design database.

[0041] A calibration reticulum is typically designed to have properties substantially similar to those of the reticulum to be inspected for defect detection or measured for metrology purposes. For example, the calibration reticulum and the test reticulum are preferably made of substantially the same materials with substantially the same thicknesses and compositions. Additionally, the two reticulums may have been formed using the same processes. The two reticulums do not necessarily need to have the same patterns printed on them, as long as the patterns on the reticulums can be divided into segments that are substantially the same (e.g., lines of similar widths, etc.). Furthermore, the reticulum to be inspected and the reticulum used to capture the images may be one and the same reticulum.

[0042] Fig. 2 is a flowchart that models a calibration process. 200This represents a specific implementation of the present invention. As shown, in operation 208 the photolithography process and the photoresist are modeled as they appear on near-field mask images ( 201 ) are applied, which were obtained from a calibration reticle using an initial set of model parameters. Alternatively, the calibration process can be 200 simulated calibration reticle images ( 202) which are simulated from the design database. Reticle images can be rendered from the database by simulating reticle fabrication and imaging processes within the design database. Any suitable model can be used to generate optical images for the features of the design database. For example, such a simulation might involve the use of the Sum-Of-Coherent-Systems (SOCS) or Abbe methods described here. 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 the Fraunhofer IISB in Erlangen, Germany. In the case of simulating an image from the design database 202The near field can first be simulated, which can be done using the software package mentioned above as well as several other packages, including Prolith from KLA-Tencor and HyperLith from Panoramic Technologies.

[0043] The model for generating wafer images based on reticulated near-field images can include only the effect of the photolithography scanner, or it can include the effect of photoresist, etching, CMP, or other wafer processes. An example of a process simulation model tool is Prolith, available from KLA-Tencor Corp., Milpitas, CA. Photoresist and etching processes can be modeled strictly or approximately. In a specific embodiment, the model can be a compact photoresist model that includes 3D acid diffusion within a specific photoresist material and configuration with associated boundary conditions, as well as a single threshold for generating the latent image.

[0044] It is noted that the modeled lithography tool may have a different illumination form or source than the reticle inspection tool for capturing actual images of the reticle. In certain embodiments, the modeled lithography tool may have the same or a similar source as a reticle inspection tool.

[0045] Other simulation approaches, such as SOCS or Abbe, can be used. The algorithm commonly known as Sum-Of-Coherent-Systems (SOCS) attempts to transform the imaging system into a series of linear systems whose output signals are squared, scaled, and summed. The SOCS method has been described elsewhere, including in Nicolas Cobb's doctoral dissertation, "Fast Optical and Process Proximity Correction Algorithms for Integrated Circuit Manufacturing," University of California, Berkeley, Spring 1998. The Abbe algorithm involves calculating the image of the object for each point source sequentially and then summing the intensity images, taking into account the relative intensity of each source point.

[0046] The input for the model and its modeling parameters includes a set of process conditions that are applied to a reconstructed near-field mask. That is, the model is 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 fabrication process parameters that characterize, or partially characterize, the wafer fabrication process for forming a wafer pattern from the mask. For example, a specific focus and exposure setting can be input into the model.Other adjustable model parameters may also include one or more of the following: a projection lens wavefront parameter, an apodization parameter, a focus error parameter for chromatic aberrations, a vibration parameter, a photoresist profile index, a photoresist impurity metric, a top-loss metric, etc. Using such a model with different sets of process conditions can result in a set of simulated wafer or photoresist pattern images generated by the reconstructed near-field mask under various processing conditions, and these simulated wafer images can be used to evaluate pattern stability and defect detection, as further described herein.

[0047] A calibration reticulum can also be used to produce a calibration wafer, from which the process 216Actual images will be obtained. In one example, the actual images are acquired using a critical dimension (CD) scanning electron microscope (SEM). Other imaging tools can be used, but a high-resolution tool is preferred.

[0048] In general, the calibration wafer contains any number of known structures, which can vary considerably. The structures can be in the form of lattices, which are typically periodic. Each lattice can be periodic in one direction ( X or Y ) be, for example as a line space pattern, or it can be periodic in two directions ( X and Y) be, for example, a grid space pattern. Examples of a grid space lattice might include an arrangement of lines in the Y-direction, with each line segmented in the X-direction. Another example of a grid space is an array of point structures. That is, each structure can take the form of a line space lattice, a grid space lattice, a checkerboard pattern structure, etc. The structure design features can each include the line width (width at a given height), line spacing, line length, shape, sidewall angle, height, pitch, grid orientation, top profile (degree of top rounding or T-termination), bottom profile (plan view), etc. The calibration wafer can contain structures with different combinations of these feature properties. It is understood that different structure properties (such as different widths, spacings, shapes, pitch, etc.)) show different reactions to focusing, which is why the calibration mask preferably includes different structures with different properties.

[0049] In one specific embodiment, the calibration wafer can have the form of a Design of Experiments (DOE) wafer with different measurement sites that have been exposed 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 as a DOE wafer). In this way, the measurement sites correspond to different locations on the wafer surface to which different process parameter values ​​are assigned. In one example, the DOE pattern is a focus / exposure matrix (FEM) pattern. Typically, a DOE wafer exhibiting an FEM pattern contains a grid pattern of measurement sites. In one grid direction (e.g., the x-direction), the exposure dose is varied while the focus depth is kept constant. In the orthogonal grid direction (e.g.,(in the y-direction) the focus depth is varied while the exposure dose is kept constant. In this way, measurement data collected from the FEM wafer include data that can be correlated with known variations in the focus and dosing process parameters.

[0050] FEM measurement points are generally distributed across the focus-exposure matrix wafer. In fact, there can generally be one or more measurement points per field. Each field can be created using a different combination of focus and exposure (or can be just focus or exposure). For example, a first field can be created using a first combination, and a second field can be created using a second combination that differs from the first. The multiple combinations can be created using varying focus and varying exposure, varying focus and constant exposure, constant focus and varying exposure, and so on.

[0051] The number of measurement points can also vary. The number of points per field is generally lower for production wafers because the surface area on production wafers is so valuable. Due to time constraints in production, fewer measurements are also performed on a product wafer than on a focus exposure matrix wafer. In one embodiment, a single point per field is measured. In another embodiment, multiple points per field are measured.

[0052] In most FEM applications, the measurement point structures are formed from identically designed patterns using different processing parameters. However, it should be noted that different focus exposure matrices can have different structures. For example, a first matrix can be created using a first grid type, and a second matrix can be created using a second grid type that differs from the first.

[0053] In an alternative embodiment, simulated calibration images ( 202), which are rendered from the design database for a calibration reticle, are used as input to the model. That is, the model can be calibrated without reconstructing the near field from a physical calibration reticle. Instead, the lithography image is simulated by simulating (not reconstructing) the near field from the design database and applying the lithography imaging model to the simulated near field to arrive at the lithography result, which corresponds to the actual results from the wafer ( 216 ) is compared.

[0054] In general, optical signal data associated with known variations of any set of process parameters, structural parameters, or both are considered. Regardless of their shape, the calibration wafer structures can be printed in a variety of different wafer layers. Specifically, the printed structures are generally printed in a layer of photoresist using standard lithography techniques (e.g., projecting a circuit diagram through a reticle and onto a photoresist-coated silicon wafer). The wafer can be a calibration wafer with material layers corresponding to those typically found on product wafers at this stage of the testing process. The printed structures can be printed over other structures in underlying layers. The calibration wafer can also be a product wafer with the potential to produce functional devices.The calibration wafer can be a simple wafer used solely for calibrating the model. It can also be the same wafer used to calibrate the OPC design model. More than one calibration wafer can be used to calibrate the lithographic model. When using multiple calibration wafers, the same or different calibration reticles can be used. These different calibration reticles can have patterns with varying dimensions to generate a wider range of image data.

[0055] The process parameters used to create the calibration structures are generally configured to ensure that the pattern characteristics remain within the desired specifications. For example, the calibration structures can be printed onto a calibration wafer as part of a calibration process, or they can be printed onto a production wafer during manufacturing. In manufacturing, the calibration structures are typically printed in the write line between component areas (e.g., dies that define the IC) arranged on a production wafer. The calibration sites can be separate calibration structures arranged around the component structures, or they can be part of the component structure itself (e.g., a periodic part). It is understood that using part of the component structure can be more difficult, but tends to be more accurate because it is a part of the component structure.In another embodiment, the calibration structures can be printed across an entire calibration wafer.

[0056] With renewed reference to Fig. 2. Corresponding modeled and calibration results (e.g., images) can be incorporated into the process. 210 They can be compared. Then, in the process, 212 It will be determined whether the model parameters need to be adjusted. If model parameters need to be adjusted, they will be processed in the following steps. 214 adapted, and the procedure 200 repeats the process 208for modeling the lithography process (and the photoresist) using adjusted parameters. Model parameters can be adjusted until the quantification of the differences between the model and the calibration images reaches a minimum that is also below a predefined threshold. The minimized value can be the sum of the squared differences between the acquired calibration images and the simulated images. The output of this process 200 This is a lithography / photoresist model and its final model parameters. This set of model parameters overcomes the technical hurdle associated with modeling mask processes and calculating 3D mask diffraction due to the use of near-field masking.

[0057] The simulated wafer patterns, based on the recovered mask near-field results, can be used for a variety of mask inspection, metrology, and / or qualification purposes. In one embodiment, reticle qualification is performed by evaluating whether the recovered mask near-field is likely to result in wafer pattern defects under a range of simulated wafer fabrication conditions. For defect detection, the printability of a reticle defect on the wafer is important, and the printability of reticle defects depends directly on the reticle near-field and the lithography system.

[0058] Once a final calibrated lithography / photoresist / etching model has been obtained for a specific process—regardless of how such a model was obtained—it can be used to generate accurate wafer plane photoresist images from a mask (e.g., after development or after etching), prior to wafer fabrication using such a mask, or for requalifying such a mask. These photoresist images allow for highly reliable inspection of the wafer images for pattern recognition across various focus and exposure settings or other lithography parameters. Because this evaluation process can be performed prior to wafer fabrication, qualification and defect detection cycles can be significantly reduced.Simulated wafer images can also enable the separation of different causes of structuring problems by comparing the simulated wafer images after lithography, after applying the photoresist model, and after etching.

[0059] Fig. Figure 3 shows a flowchart illustrating a reticule qualification process. 300 according to one embodiment of the present invention. In process 302 A mask near-field image is restored, for example, for a specific reticulum based on images acquired from that specific reticulum. This process can supplement the mask near-field restoration operations of Fig. 1. Once a near-mask field has been obtained, the lithography process (and photoresist) can also be carried out using the final model parameters with respect to the restored near-mask field. 303can be modeled. For example, the final model is used to simulate wafer images using a near-field mask image.

[0060] The simulated wafer pattern can then be used in the process. 322 The reticles can be evaluated to determine pattern stability and / or to locate defects. It can be generally determined whether the corresponding reticle is likely to lead to unstable or defective wafer patterns. In one embodiment, the model is applied to the mask near-field image or mask near-field results using a variety of different process conditions, such as focus and dose, to evaluate the stability of the reticle design under varying process conditions.

[0061] Fig. 4A is a flowchart that shows a process 400 for determining wafer pattern stability according to an exemplary application of the present invention. Initially, each test image can be used in the process.402 The images are aligned to their corresponding reference images, which were also generated by the model under different sets of process conditions. The various test and reference images are calculated by the model under different processing conditions / parameters.

[0062] Each pair of aligned images can be processed 404 They are compared with each other to obtain one or more wafer pattern differences. In process 406Thresholds can then be associated with each wafer pattern difference. The wafer pattern differences and their associated thresholds can be used together to characterize pattern stability. That is, the extent of deviation of a given pattern (pattern differences) under various simulated process conditions, and whether such deviation exceeds an associated threshold, together characterize pattern stability. The process window of a manufacturing process specifies an expected or defined level of process deviation under which the resulting patterns are evaluated to ensure that they remain stable or within certain specified deviation tolerances (e.g., threshold).

[0063] Different regions of the reticle, and thus corresponding wafer patterns, can be assigned different thresholds for assessing pattern stability. These thresholds can all be the same or different, based on various factors such as the pattern design context, the pattern MEEF level (or the Mask Error Enhancement Factor, as described below), or the sensitivity of the device function to variations in the wafer pattern, etc. For example, a narrower threshold can be chosen for patterns in a dense region compared to a semi-dense region of the reticle.

[0064] Optionally, an initial set of hot spots or areas of pattern weakness can be identified in both the reference and test mask patterns. For example, a designer can provide a list of design hot spot coordinates that are critical to the component's function. Areas defined as hot spots can be assigned a detection threshold, while non-hot spot areas can be assigned a higher threshold (for defect detection). This distinction can be used to optimize inspection resources.

[0065] This assessment of pattern stability can be used to facilitate reticle qualification and thereby overcome many challenges in this field. With the increasing density and complexity of integrated circuits (ICs), the inspection of photolithographic mask patterns becomes increasingly difficult. Each new generation of ICs features denser and more complex patterns that currently reach and exceed the optical limits of lithography systems. To overcome these optical limits, various resolution enhancement techniques (RETs) have been introduced, such as optical proximity correction (OPC). For example, OPC helps to overcome some diffraction limitations by modifying photomask patterns so that the resulting printed patterns match the originally desired patterns.Such modifications can include changes to the sizes and edges of the main IC features, i.e., the printable features. Other modifications include adding serifs to pattern corners and / or providing sub-resolution assist features (SRAFs) nearby, which are not expected to result in printable features and are therefore referred to as non-printable features. These non-printable features are expected to cancel out pattern defects that would otherwise have occurred during the printing process. However, OPC makes mask patterns even more complex and typically very dissimilar to the resulting wafer images. Furthermore, OPC defects often do not result in printable defects.The increased complexity of the photomask pattern, and the fact that not all pattern elements are intended to directly influence the printed pattern, significantly complicates the inspection of the photomask for significant pattern defects. As the semiconductor industry moves towards ever smaller features, leading manufacturers are beginning to use even more exotic OPCs, such as inverse lithography (ILT), which result in highly complex patterns on the mask. Therefore, it is highly desirable to know the mask writing fidelity and its wafer printing quality before the physical fabrication of the wafer.

[0066] A measure of the significance of a defect is its MEEF, or Mask Error Enhancement Factor. This factor relates the size of the defect in the mask plane to the magnitude of its impact on the printed image. Defects with a high MEEF have a significant impact on the printed pattern; defects with a low MEEF have little or no impact. A feature of the main pattern that is too small in a pattern section with dense, fine lines is an example of a defect with a high MEEF, where a small size error in the mask plane could cause the printed pattern to collapse completely. An isolated small pinhole is an example of a defect with a low MEEF, where the defect itself is too small to be printed and is far enough from the nearest main pattern edge 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 pattern context in which the defect is located.

[0067] In addition to mask defects with higher MEEF, which cause more significant wafer defects, certain design patterns and corresponding mask patterns may be more robust to process changes than other design and mask patterns. When the manufacturing process begins to deviate from optimal process conditions, certain mask patterns can lead to more significant wafer pattern disturbances and defects.

[0068] Fig. 4B is a flowchart that describes a defect inspection procedure. 450 according to another embodiment of the present invention. In process 452Each modeled test wafer image can be aligned with its corresponding reference image. In one embodiment, die-to-die or cell-to-cell alignment can be achieved. In another embodiment, the modeled test wafer image is aligned with a rendered reference image from the corresponding post-OPC design. For example, the post-OPC design is processed to simulate the reticle fabrication process for such a design, such as rounding corners. Generally, a reference image can be from the same die as the test image at an earlier time point, from an adjacent identical die, or rendered from the design database. In a specific example, the reference image is obtained from a "golden" die that has been proven to be free of defects (e.g., immediately after the reticle has been fabricated and qualified).The golden reticle images obtained from the reticulum when it was known to be non-defective can be saved and later used to calculate the golden reticle near-field images and wafer images if needed. Alternatively, the golden reticle near-field images can be saved for easy access without requiring recalculation of the near field during future inspections.

[0069] Each pair of aligned test and reference images is compared based on an assigned threshold to determine the process. 454 To locate reticular defects, any suitable mechanism can be used to assign threshold values ​​to specific reticular areas, as described above. Any suitable metrics of the test and reference images can be compared. For example, the contour of test and reference wafer images can be compared as a metric for edge placement error (EPE).

[0070] For each reticular defect, the process can then be carried out 456 The corresponding simulated wafer defect area is compared with its corresponding reference pre-OPC area. That is, the simulated wafer patterns are evaluated to determine whether the reticulated defect leads to a wafer defect that deviates from the intended design.

[0071] With renewed reference to Fig. 3 can then be performed based on the simulated reticle images in process 324It must be determined whether the design is defective. In one embodiment, it is determined whether the design pattern leads to an unacceptable variation in the wafer pattern under a specific 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 can be considered defective. These systematic defects are referred to as hot spots. It can also be determined whether any difference between a simulated wafer pattern from the reticle and its corresponding pre-OPC pattern exceeds a predefined threshold. If the design is found to be defective, the design can be modified in process. 332 be modified.

[0072] Once the design of a reticle has been reviewed, the reticle may still contain hot spots that should be monitored. The following operations are described as being implemented on a mask for which at least some identified hot spots exist. If the mask does not contain any identified hot spots, the following operations can, of course, be performed by Fig. Steps 3 can be skipped, and the mask can be used without hot spot monitoring during manufacturing and inspection.

[0073] If the design in the illustrated example is not considered faulty, then the process can proceed. 326 It will be determined whether any hot spots can be monitored. If it is determined that the hot spots can be monitored, they can then be monitored during the wafer process. 334These parameters must be monitored. For example, hot spot patterns during wafer fabrication can be monitored to determine if the process has deviated from specifications and caused the corresponding wafer pattern to exhibit critical parameters that have changed to unacceptable values. One implementation might involve setting a relatively high MEEF level to inspect the corresponding reticle and / or wafer pattern of the hot spot. If conditions deviate further from nominal process conditions, CD or EPE can become large and compromise the integrity of the wafer fabrication process.

[0074] Hot-spot patterns can only be identified when a test mask pattern changes by a predefined amount, regardless of how such a change relates to the originally intended design (e.g., pre-OPC data). In other words, a significant change in the physical mask pattern under different process conditions can indicate a problem with the intended design pattern. Differences between the corresponding modeled image portions represent differences in the impact of 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 simply "hot spots" and represent weaknesses in the design with respect to the specific process conditions under investigation, and potentially also with respect to the manufactured mask.Examples of the types of differences that can occur between modeled images for different process conditions are CD (critical dimension) or EPE (edge ​​placement error).

[0075] In another embodiment, when the model is applied to the post-OPC design database, the resulting wafer pattern can correspond to the pattern intended by the designer to be printed on the wafer. Optionally, the results from applying the model to the post-OPC database can be used with the modeled images to improve hot spot detection. For example, a post-OPC database model that considers only design effects can be used to separate the impact of the wafer process on the design from the impact of the wafer process on the manufactured mask. Modeled patterns from the near-mask area can be compared with modeled wafer images from corresponding post-OPC patterns.For example, if a set of modeled wafer patterns for various process changes matches corresponding modeled post-OPC wafer patterns for the same process changes, it can be determined that the changes in the wafer pattern (or photoresist pattern) due to process changes 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 the wafer due to process variations from the post-OPC database differ from those on the wafer due to the same process variations from the reconstructed mask (or near-mask area), these hot spots are considered to originate from a hot spot in the actual mask, which can be repaired or monitored.

[0076] The simulated wafer image differences can also be analyzed to determine metrics for wafer CD uniformity (CDU, CD uniformity) across the die or over time if reticulation changes occur during exposure during the manufacturing process. For example, the CD for each target of each image can be measured 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 can be calibrated and converted into CD variations, as described in U.S. Patent Application No. 14 / 664,565, filed on March 20, 2015, by Carl E. Hess et al., and in U.S. Patent Application No. 14 / 390,834, filed on October 6, 2014, by Rui-fang Shi et al., which applications are incorporated herein by reference for all purposes.

[0077] It can also be in process 328It must be determined whether the reticulum should be repaired. It may be determined that the expected variations in the wafer pattern are outside the specification for the process window to be used during the lithography process. In certain cases, the reticulum may contain a defect that can be addressed in the process. 336 The reticle will be repaired. It can then be requalified. Otherwise, the reticle can be moved to the next step. 330 It will be discarded if it is not repairable. A new reticle can then be manufactured and requalified.

[0078] In addition to or as an alternative to using a recovered near-field mask image to simulate wafer images in a qualification process, a near-field mask image or near-field mask results can also be evaluated directly in a reticule qualification process. Fig. 5 is a flowchart that illustrates a reticle qualification process. 500illustrated, which is applied to a recovered near-field mask image or recovered results for a near-field mask, according to an alternative embodiment of the present invention. Initially, in the process 502 Near-field mask results are recovered from a reticulum. This near-field mask image can be recovered for a specific reticulum based on images acquired from that particular reticulum. This process can be performed similarly to the near-field mask recovery operations of Fig. 1 can be executed. Additionally, several of the operations can be performed. Fig. 5 in a similar way to processes of Fig. 3 will be implemented, however with the recovered reticle near-field image, including intensity and / or phase components of such an image.

[0079] As shown, the mask near-field results can then be used in the process 522The images can be evaluated to characterize and / or locate defects. Essentially, it can be determined whether the corresponding reticle is defective or exhibits hot spots requiring monitoring. In particular, some of the techniques described herein for evaluating the simulated wafer images can be implemented on the mask near-field images. In a defect detection process, any suitable metrics from test and reference mask near-field images can be compared. For example, intensity and / or phase can be compared. Different defect types have different effects on intensity and / or phase values. It can be determined that these differences are true defects (as opposed to interfering flaws without effect) that are likely to lead to a defective wafer, or the differences can identify hot-spot patterns or areas that can be repaired or monitored.

[0080] For example, in the process 524 It will be determined whether the design is faulty. If the design is found to be faulty, the design can be modified in the process. 532 can be modified. For example, it can be determined whether a difference between a reticle near-field image and its corresponding post-OPC-based near-field exceeds a predefined threshold for detecting defects. The procedure 500 The process can continue to determine whether wafer hot spots should be monitored, the reticulum repaired, or the reticulum redesigned as described above. If the design is not considered faulty, the process can then proceed. 526 It can be determined whether any hot spots can be monitored. For example, it can be determined whether any intensity and / or phase difference between a test and a reference reticle near-field image is close to an assigned threshold.

[0081] If it is determined that the hot spots can be monitored, the hot spots can, for example, be detected during the wafer process. 534 These parameters must be monitored. For example, hot spot patterns during wafer fabrication can be monitored to determine if the process has deviated from specifications, causing the corresponding wafer pattern to exhibit critical parameters that have changed to unacceptable values. One implementation might involve setting a relatively high sensitivity level for inspecting the reticle and / or wafer pattern of the relevant hot spot. As conditions deviate further from nominal process conditions, CD defects or EPE can become large and compromise the integrity of the wafer fabrication process.

[0082] It can be in process 528It will also be determined whether the reticulum should be repaired. In certain cases, the reticulum may contain a defect that must be addressed in the process. 536 The reticle will be repaired. It can then be requalified. Otherwise, the reticle can be moved to the next step. 530 It will be discarded if it is not repairable. A new reticle can then be manufactured and requalified.

[0083] Certain techniques of the present invention provide for mask pattern qualification and early detection of faint patterns or hot spots on the physical mask before wafer fabrication begins. In addition to restoring the reticle near-field based on reticle images, a full range of wafer process effects (including many focus and exposure settings, as well as the effect of wafer photoresist, etching, CMP, and other wafer processes) can be considered with regard to how they affect the wafer patterns. No prior knowledge of the mask is required, as the mask near-field is restored using only reticle images without the use of reticle design data. Since mask patterns are generally 4xSince the patterns are larger than wafer patterns, more precise pattern positions relative to the design database can be determined. The techniques described above can also be extended to any suitable mask type, such as the pattern qualification of EUV masks.

[0084] Techniques of the present invention can be implemented in any suitable combination of hardware and / or software. Fig. Figure 6 is a schematic representation of an exemplary inspection system. 600 , in which techniques of the present invention can be implemented. The inspection system 600 can an input 602 from an inspection tool with high NA or an inspector with low rank NA received, which mimics a scanner (not shown). The inspection system can also be a data distribution system (e.g. 604a and 604b ) to distribute the received input 602, an intensity signal (or patch) processing system (e.g., patch processors) and a reticle qualification system (e.g., 612 ) for near-field mask and wafer restoration, process modeling, etc., a network (e.g., the switched network) 608 ) to enable communication between the inspection system components, an optional mass storage device 616 and one or more inspection control and / or verification stations (e.g. 610 ) to check the intensity and phase of the mask near field (values, images, or differences), the reticle / wafer images, the identified hot spots, the CD, the CDU maps, the process parameters, etc. Each processor of the inspection system 600It can typically contain one or more integrated microprocessor circuits and may also contain integrated interface and / or memory circuits and may additionally be coupled with one or more shared and / or global storage devices.

[0085] The inspector or data acquisition system (not shown) for generating input data 602 can take the form of any suitable instrument (e.g., as further described herein) for obtaining intensity signals or images of a reticulum. For example, the inspector with low NA Create an optical image or generate intensity values ​​of a portion of the reticle based on a portion of the detected light that is reflected, transmitted, or otherwise directed onto one or more light sensors. The low NA inspector can then output the intensity values ​​or the image.

[0086] The inspection tool with low NA It can be operated to detect and collect reflected and / or transmitted light when an incident optical beam scans across each patch of a reticle. As mentioned above, the incident optical beam can scan across strips of reticles, each containing a multitude of patches. Light is collected from a multitude of points or sub-areas of each patch in response to this incident beam.

[0087] The inspection tool with low NA It can generally be operated in such a way that it converts detected light into detected signals corresponding to intensity values. The detected signals can take the form of an electromagnetic waveform with amplitude values ​​corresponding to different intensity values ​​at different locations on the reticulum. The detected signals can also take the form of a simple list of intensity values ​​and associated reticulum point coordinates. Alternatively, the detected signals can take the form of an image with different intensity values ​​corresponding to different positions or scan points on the reticulum.Two or more images of the reticle can be generated after all positions of the reticle have been scanned and converted into detected signals, or parts of two or more images can be generated when each reticle section is scanned, with the last two or more images for the reticle being complete after the entire reticle has been scanned.

[0088] The detected signals can also be in the form of aerial images. This means that an aerial imaging technique can be used to simulate the optical effects of the photolithography system, generating an aerial image of the photoresist pattern exposed onto the wafer. Generally, the optics of the photolithography tool are emulated to generate an aerial image based on the detected signals from the reticle. The aerial image corresponds to the pattern created from the light directed through the photolithography optics and the reticle onto the photoresist layer of a wafer. Additionally, the photoresist exposure process can also be emulated for a specific type of photoresist material.

[0089] The incident or detected light can be guided through any suitable spatial aperture to generate any desired incident or detected light profile at any suitable angle of incidence. For example, programmable illumination or detection apertures can be used to generate a specific beam profile such as a dipole, quadrupole, quasar, ring, etc. In a particular example, source mask optimization (SMO) or any pixelated illumination technique can be implemented. The incident light can also be guided through a linear polarizer to linearly polarize the entire illumination pupil or a portion thereof in one or more polarizations. The detected light can be guided through apodization components to block specific regions of the collecting beam.

[0090] Intensity or image data 602can be accessed from the data distribution system via the network 608 to be received. The data distribution system can be assigned to one or more storage devices, such as RAM buffers, to store at least part of the received data. 602 to maintain. Preferably, the total memory is large enough to accommodate an entire data field. For example, one gigabyte of memory is suitable for a field corresponding to 1 million by 1000 pixels or points.

[0091] The data distribution system (e.g. 604a and 604b ) can also be the distribution of parts of the received input data 602 to the processors (e.g. 606a and 606b ) control. For example, the data distribution system can send data for an initial patch to an initial patch processor. 606a route and send data for a second patch to the patch processor 606broute. Multiple data sets for multiple patches can also be forwarded to each patch processor.

[0092] The patch processors can receive intensity values ​​or an image corresponding to at least one part or patch of the reticle. Each patch processor can also be coupled to or integrated with one or more memory devices (not shown), such as DRAM elements, that provide local memory functions, such as holding the received data segment. Preferably, the memory is large enough to store data corresponding to one patch of the reticle. For example, eight megabytes of memory is suitable for intensity values ​​or an image corresponding to a 512 x 1024 pixel patch. Alternatively, the patch processors can share memory.

[0093] Each set of input data 602can correspond to a strip of the reticle. One or more data records can be stored in the memory of the data distribution system. This memory can be controlled by one or more processors within the data distribution system, and the memory can be divided into multiple partitions. For example, the data distribution system can receive data corresponding to part of a strip in a first memory partition (not shown), and the data distribution system can receive further data corresponding to another strip in a second memory partition (not shown). Preferably, each of the memory partitions of the data distribution system contains only the portions of the data that are to be forwarded to a processor assigned to that memory partition. For example, the first memory partition of the data distribution system can hold initial data and forward it to the patch processor. 606adirect, and the second storage partition can hold second data and send it to the patch processor. 606b lead.

[0094] The data distribution system can define and distribute each data record based on any suitable data parameters. For example, the data can be defined and distributed based on the corresponding position of the patch on the reticulum. In one embodiment, each strip is assigned a range of column positions that correspond to horizontal pixel positions within the strip. For example, the columns 0 until 256 of the strip correspond to a first patch, and the pixels in these columns comprise the first image or the first set of intensity values ​​that is passed on to one or more patch processors. Likewise, the columns 257 until 512of the band correspond to a second patch, and the pixels in these columns comprise the second image or the second set of intensity values ​​that is directed to (another) patch processor(s).

[0095] The inspection device can be suitable for inspecting semiconductor devices or wafers and optical reticles, as well as EUV reticles or masks. Examples of suitable inspection tools include the Teron™, which operates at 193 nm, or the TeraScan™ DUV reticle inspection tools, available from KLA-Tencor of Milpitas, CA. Other types of samples that can be inspected or imaged using the inspection device of the present invention include any surface, such as a flat-screen display.

[0096] An inspection tool may include at least a light source for generating an incident beam of light, illumination optics for directing the incident beam onto a sample, collecting optics for directing an output beam emitted by 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 enabling the near-field mask generation and analysis techniques as further described herein.

[0097] In the following exemplary inspection systems, the incident beam can be in any suitable form of coherent light. Additionally, any suitable lens arrangement can be used to direct the incident beam onto the sample and to direct the output beam from the sample onto a detector. The output beam can be reflected or scattered by the sample, or transmitted through the sample. In EUV reticle inspection, the output beam is typically reflected by the sample. Likewise, any suitable detector type or number of detection elements can be used to receive the output beam and provide an image or signal based on the properties (e.g., intensity) of the received output beam.

[0098] First, a generalized photolithography tool is described, although an EUV photolithography tool typically only has a reflective optic. Fig. 7A is a simplified schematic representation of a typical lithography system 700 , which can be used to transfer a mask pattern from a photomask M to a wafer W according to certain embodiments. Examples of such systems include scanners and steppers, in particular the TWINSCAN NXT:1970Ci step-and-scan system available from ASML in Veldhoven, Netherlands. In general, a light source directs 703 a beam of light through a lighting optic 707 (e.g. lens) 705 ) onto a photomask M, which is located in a mask layer 702 The lighting lens is located. 705 has a numerical aperture 701 at this level 702 The value of the numerical aperture 701This influences which defects on the photomask are lithographically significant and which are not. A portion of the beam passing through the photomask M forms a structured optical signal, which is then processed by the imaging optics. 713 The light beam is directed onto a wafer W to initiate structure transfer. In a reflection system (not shown), the illumination beam is reflected by certain sections of the mask M (and absorbed by other sections of this mask M) and forms a structured signal that is guided through the reflective imaging optics onto a wafer W.

[0099] The inspection tool can use similar components or be configured similarly to the photolithography tool described above, e.g., LNI capability. However, the inspection tool can also be configured to generate high-resolution images, either alternatively or additionally. Fig. Figure 7B presents a schematic representation of an exemplary inspection system. 750 ready, that's a lighting optic 751a features an imaging lens with a relatively large numerical aperture 751b in a reticulum level 752 according to certain embodiments. For example, the numerical aperture 751b in the reticulum level 752 the inspection system's aperture must be considerably larger than the numerical aperture. 701 in the reticulum level 702 of the lithography system 700 , which would lead to differences between test inspection images and actual printed images.

[0100] The inspection techniques described herein can be implemented on various specially configured inspection systems, such as the one schematically depicted in Figure 7B. The system shown 750 includes a light source 760, which produces a beam of light that passes through the lighting optics 751a on a photomask M in the reticulation plane 752 is directed. Examples of light sources include a coherent laser light source (e.g., a deep-UV or gas laser generator), a filtered lamp, an LED light source, etc. In certain embodiments, a light source can generally provide a high pulse repetition rate, low noise, high power, stability, reliability, and scalability. While an EUV scanner operates at a wavelength of 13.5 nm, an inspection tool for an EUV reticle does not need to operate at the same wavelength, although this is possible. In one example, the source is a 193 nm laser.

[0101] The lighting optics 751aIt may include a beam steering device for precise beam positioning and a beam conditioning device that 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 devices, for example, from a laser. The illumination optics 751a It may also include optics for controlling polarization, focus, magnification, illumination intensity distributions, etc.

[0102] As explained above, the inspection system 750 a numerical aperture 751b in the reticulum level 752 exhibiting elements equal to or greater than a numerical aperture in the reticular plane (e.g., element 701 in 7A) of the corresponding lithography system. The photomask to be inspected M is placed on a mask table in the reticulum plane752 placed and exposed to the source.

[0103] The inspection system shown 750 can detection optics 753a and 753b including, which may also contain microscopic magnification optics designed, for example, to provide a magnification of 60 until 200 to provide specialist or more for improved inspection. The collecting optics 753a and 753b They can include any suitable optics for conditioning the output light / beam. For example, the converging optics can 753a and 753b Included are optics for controlling focus, pupil shapes, polarization analyzer settings, etc.

[0104] In a transmission mode, the structured image can be transferred from the mask. M through a number of optical elements 753a are directed to project the structured image onto a sensor 754aproject. In a reflection mode, collecting elements (e.g., beam splitters) direct the beam. 776 and detection lens 778 ) the reflected light from the mask M onto the sensor 754b and capture it. Although two sensors are shown, a single sensor can be used to capture reflected and transmitted light during different scans of the same reticular area. Suitable sensors include charge-coupled devices (CCDs), CCD arrays, time-delay integration (TDI) sensors, TDI sensor arrays, photomultiplier tubes (PMTs), and other sensors.

[0105] The illumination optics column can be moved relative to the mask table and / or the table relative to a detector or camera by any suitable mechanism to scan patches of the reticulum. For example, a motor mechanism can be used to move the table. The motor mechanism could consist of, for example, a worm drive and stepper motor, a linear actuator with feedback position, or a belt actuator and a stepper motor. The system 700 can use one or more motor mechanisms to move one of the system components with respect to the optical illumination or collecting paths.

[0106] The data from each sensor (e.g. 754a and / or 754b The captured signals can be processed by a computer system 773or more generally, processed by one or more signal processing devices, each of which may contain an analog-to-digital converter configured to convert analog signals from each sensor into digital signals for processing. The computer system 773 typically features one or more processors connected to input / output ports and one or more memories via suitable buses or other communication mechanisms.

[0107] The computer system 773 It may also include one or more input devices (e.g., a keyboard, a mouse, a joystick) for providing user input, such as changing focus and other inspection recipe parameters. The computer system 773It can also be connected to the table to control, for example, a sample position (e.g., focusing and scanning) and to other inspection system components to control other inspection parameters and configurations of such inspection system components.

[0108] The computer system 773 It can 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), reticle / wafer images, identified hot spots, CD, CDU maps, process parameters, etc. The computer system 773 It can be configured to analyze the intensity, phase, and / or other properties of reflected and / or transmitted, detected and / or simulated signals or images, recovered reticle near-field results, etc. The computer system 773It can be configured (e.g., with programming instructions) to provide a user interface (e.g., on a computer screen) for displaying resulting intensity and / or phase values, images, and other inspection properties. In certain embodiments, the computer system 773 configured to perform inspection techniques as described above.

[0109] Since such information and program instructions can be implemented on a specially configured computer system, such a system contains program instructions / computer code for performing various operations described herein, which can be stored on a computer-readable medium. Examples of machine-readable media include, but are not limited to, magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROM discs; magneto-optical media such as optical disks; and hardware devices specifically configured to store and execute program instructions, such as read-only memory (ROM) and random-access memory (RAM). Examples of program instructions include both machine code, as generated by a compiler, and higher-level code files that can be executed by the computer using an interpreter.

[0110] Fig.Figure 7B shows an example where an illumination beam is directed at the sample surface at a substantially normal angle to the inspected surface. In other embodiments, an illumination beam can be directed at an oblique angle, allowing separation of the illumination beam and the reflected beam. In these embodiments, an attenuator can be positioned on the reflected beam path to attenuate a zero-order component of the reflected light beam before it reaches a detector. Furthermore, an imaging aperture can be positioned on the reflected beam path to shift the phase of the zero-order component of the reflected light beam.

[0111] It should be noted that the above description and drawings do not represent a limitation of the specific components of the system and that the system can be implemented in many other forms. For example, it is considered that the inspection or measurement tool could incorporate any suitable features from any number of known imaging or measurement tools arranged for detecting defects and / or resolving critical aspects of features on a reticle or wafer. For example, an inspection or measurement tool could 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 considered that single- and multiple-image techniques could be used to acquire images of the target.These methods include, for example, single-scan, dual-scan, single-scan coherence probe microscopy (CPM), and dual-scan CPM methods. Optical methods without imaging, such as scatterometry, can also be considered as part of the inspection or measurement setup.

[0112] Although the foregoing invention has been described in detail for the sake of clarity of understanding, it is obvious that certain changes and modifications can be made within the scope of protection of the appended claims. It should be noted that there are many alternative ways of implementing the processes, systems, and devices of the present invention. Accordingly, the present embodiments should be considered illustrative and not limiting, and the invention should not be restricted to the details given herein. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] US 15803628

[0001] US 62 / 508369

[0001] US 9478019

[0031] US 14 / 664565

[0076] Cited non-patent literature

[0000] Abdurrahman Sezginer et al., described which patent is incorporated herein by reference in its entirety for all purposes

[0031]

Claims

[1] Method for qualifying a photolithographic reticule, comprising: Using an imaging tool to capture a variety of images under different lighting configurations and / or different imaging configurations from each of a variety of sample areas of a test reticulum; Reconstructing a reticule near field for each of the pattern regions of the test reticule based on the captured images of each pattern region of the test reticule; and Analyzing the restored reticle near field to characterize the test reticle or to determine if such a test reticle is defective. [2] Method according to claim 1, wherein the plurality of images are captured in one pupil plane. [3] Method according to claim 1, wherein the restored reticulum near field is analyzed to detect defects in the test reticulum, the defect detection comprising comparing the intensity and / or phase for an identical 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 reticulum copy with a design identical to the test reticulum. [4] Method according to claim 1, wherein the reticle near field is restored without using a design database that was used to produce the test reticle. [5] Method according to claim 1, wherein the acquired images include at least three reflection images acquired under different imaging conditions selected to produce the same reticle near field, and wherein the different imaging conditions include different focus settings, different pupil shapes and / or polarization analyzer settings, and wherein the different illumination conditions include different source intensity distribution and / or polarization settings. [6] Method according to claim 1, wherein the acquired images include at least three transmission images acquired under different imaging conditions selected to produce the same reticle near field, and wherein the different imaging conditions include different focus settings, different pupil shapes, or polarization analyzer settings, and wherein the different illumination conditions include different source intensity distribution and / or polarization settings. [7] Method according to claim 1, further comprising: Applying a lithography model to the reticule near field for the test reticule to simulate a variety of test wafer images, and Analyzing the simulated test wafer images to determine whether the test reticulum is likely to result in an unstable or defective wafer, with the lithography model configured to simulate a photolithography process. [8] Method according to claim 7, wherein the lithography model simulates an illumination source having a shape other than an illumination shape of an inspection tool to capture images of the test reticule or of any other reticule or wafer. [9] Method according to claim 7, wherein the lithography model is calibrated with images rendered from a design database for a calibration reticle. [10] Method according to claim 7, wherein the lithography model is calibrated with images taken from a calibration reticle. [11] Method according to claim 7, wherein the lithography model includes a compact photoresist model. [12] Method according to claim 7, wherein the lithography model is applied to the reticule near field obtained for the test reticule under a plurality of different lithography process conditions, and wherein the analysis of the simulated test wafer images comprises determining whether the test reticule is likely to result in an unstable wafer under the different lithography process conditions by comparing parts of the simulated test images associated with different process conditions and the same reticule area. [13] The method of claim 7, further comprising repeating the processes of capturing images, restoring, applying the lithography model and analyzing the simulated test wafer images obtained after each application of photolithography modeling, photoresist modeling and etching modeling to separate the root causes of reticular defects. [14] Method according to claim 7, wherein the imaging tool uses the same wavelength range as a photolithography system in which the test reticulum is used for wafer production. [15] Method according to claim 7, wherein the imaging tool uses a different wavelength range than a photolithography system in which the test reticulum is used for wafer fabrication, and wherein the simulated test wafer images are analyzed to determine whether the test reticulum is likely to result in a defective wafer by performing defect detection on the test wafer images. [16] Imaging system for qualifying a photolithographic reticulum, the system comprising: a light source for generating an incident ray; a lighting optics module for directing the incident beam onto a reticle; a collecting optics module for directing an output beam from each pattern area of ​​the reticle to at least one sensor; at least one sensor for detecting the output beam and for generating an image or signal based on the output beam; and a controller configured to perform the following operations: Enabling the acquisition of a large number of images under different lighting configurations and / or different imaging configurations of each of a large number of sample areas of a test reticulum; Reconstructing a reticule near field for each of the pattern regions of the test reticule based on the captured images of each pattern region of the test reticule; and Analyzing the recovered reticule near field to determine whether the test reticule or another reticule is likely to result in an unstable wafer pattern or a defective wafer. [17] System according to claim 16, wherein the plurality of images are captured in one pupil plane. [18] System according to claim 16, wherein the restored reticulum near field is analyzed to detect defects in the test reticulum, the defect detection comprising comparing the intensity and / or phase for an identical 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 reticulum copy with a design identical to the test reticulum. [19] System according to claim 16, wherein the reticule near field is restored without using a design database that was used to produce the test reticule. [20] System according to claim 16, wherein the acquired images include at least three reflection images acquired under different imaging conditions selected to produce the same reticle near field, and wherein the different imaging conditions include different focus settings, different pupil shapes and / or polarization analyzer settings, and wherein the different illumination conditions include different source intensity distribution and / or polarization settings. [21] System according to claim 16, wherein the acquired images include at least three transmission images acquired under different imaging conditions selected to produce the same reticle near field, and wherein the different imaging conditions include different focus settings, different pupil shapes and / or polarization analyzer settings, and wherein the different illumination conditions include different source intensity distribution and / or polarization settings. [22] System according to claim 16, wherein the controller is further configured to: Applying a lithography model to the reticule near field for the test reticule to simulate a variety of test wafer images, and Analyzing the simulated test wafer images to determine whether the test reticulum is likely to result in an unstable or defective wafer, with the lithography model configured to simulate a photolithography process. [23] System according to claim 22, wherein the lithography model simulates a light source having a shape other than a light source shape of an inspection system to capture images of the test reticule or any other reticule or wafer. [24] System according to claim 22, wherein the lithography model is calibrated with images rendered from a design database for a calibration reticle. [25] Method according to claim 22, wherein the lithography model is calibrated with images acquired by a calibration reticle. [26] System according to claim 22, wherein the lithography model includes a compact photoresist model. [27] System according to claim 22, wherein the lithography model is applied to the reticule near field obtained for the test reticule under a variety of different lithography process conditions, and wherein the analysis of the simulated test wafer images comprises determining whether the test reticule is likely to result in an unstable wafer under the different lithography process conditions by comparing parts of the simulated test images associated with different process conditions and the same reticule area. [28] System according to claim 22, wherein the control is further configured to repeat the processes of capturing images, restoring, applying the lithography model and analyzing the simulated test wafer images obtained after each application of photolithography modeling, photoresist modeling and etching modeling to separate the root causes of reticular defects. [29] System according to claim 22, wherein the imaging system uses the same wavelength range as a photolithography system in which the test reticulum is used for wafer production. [30] System according to claim 22, wherein the imaging system uses a different wavelength range than a photolithography system in which the test reticulum is used for wafer fabrication, and wherein the simulated test wafer images are analyzed to determine whether the test reticulum is likely to result in a defective wafer by performing defect detection on the test wafer images.